Digitalization and Energy - International Energy Agency

Digitalization and Energy - International Energy Agency

Secure Sustainable Together Digitalization & Energy INTERNATIONAL ENERGY AGENCY The International Energy Agency (IEA), an autonomous agency, was es...

3MB Sizes 1 Downloads 108 Views

Secure Sustainable Together

Digitalization & Energy

INTERNATIONAL ENERGY AGENCY The International Energy Agency (IEA), an autonomous agency, was established in November 1974. Its primary mandate was – and is – two-fold: to promote energy security amongst its member countries through collective response to physical disruptions in oil supply, and provide authoritative research and analysis on ways to ensure reliable, affordable and clean energy for its 29 member countries and beyond. The IEA carries out a comprehensive programme of energy co-operation among its member countries, each of which is obliged to hold oil stocks equivalent to 90 days of its net imports. The Agency’s aims include the following objectives: n Secure member countries’ access to reliable and ample supplies of all forms of energy; in particular, through maintaining effective emergency response capabilities in case of oil supply disruptions. n Promote sustainable energy policies that spur economic growth and environmental protection in a global context – particularly in terms of reducing greenhouse-gas emissions that contribute to climate change. n Improve transparency of international markets through collection and analysis of energy data. n Support global collaboration on energy technology to secure future energy supplies and mitigate their environmental impact, including through improved energy efficiency and development and deployment of low-carbon technologies. n Find solutions to global energy challenges through engagement and dialogue with non-member countries, industry, international organisations and other stakeholders.

© OECD/IEA, 2017 International Energy Agency Website: www.iea.org

IEA member countries: Australia Austria Belgium Canada Czech Republic Denmark Estonia Finland France Germany Secure Greece Sustainable Hungary Together Ireland Italy Japan Korea Luxembourg Netherlands New Zealand Norway Poland Portugal Slovak Republic Spain Sweden Switzerland Turkey United Kingdom United States

Please note that this publication is subject to specific restrictions that limit its use and distribution. The terms and conditions are available online at www.iea.org/t&c/

The European Commission also participates in the work of the IEA.

Foreword Digital technologies permeate modern life, affecting everything from the way we work

and travel, to the way we live and play. Digitalization holds great promise to help improve the safety, productivity, efficiency and sustainability of energy systems worldwide. But it also raises questions of security, privacy and economic disruption.

The International Energy Agency (IEA) has been focused on the interplay between

digitalization and energy for many years. We formulated smart grid roadmaps in 2011 and 2017. We have tracked the progress of electric vehicles and smart charging, and analysed renewables integration and energy use by connected devices.

We are finding that digitalization is becoming increasingly important to a wide variety of energy sectors. Furthermore, with digital technologies changing so rapidly, there are

many unknowns about how technology, behaviour and policy will evolve over time and how these dynamics will impact energy systems into the future.

To ensure a more systematic and co-ordinated approach to our efforts, I established a cross-agency Digitalization and Energy Working Group in 2016. Drawing on expertise from across the agency, we are now presenting this first comprehensive examination of digitalization and energy.

Our aim is to provide an accurate and balanced view of what is happening today. Using

our analytical tools, we are also beginning to paint a picture of what could happen

next. In addition to this report, a new interactive website complements our efforts and provides an alternative way to access our findings.

This effort has benefited from the guidance and feedback of IEA member, association and partner countries as well as energy and technology companies (including those of

the IEA Energy Business Council). All of these governments and companies are seeking insights into the opportunities and challenges of digitalization across the whole energy system.

Our hope is that this report will be a clear and accessible step toward helping all global actors better navigate the rapidly changing digitalization and energy landscape.

Dr Fatih Birol

International Energy Agency Foreword

3

© OECD/IEA, 2017

Executive Director

© OECD/IEA, 2017

Acknowledgements This study was prepared by a cross-agency Digitalization and Energy Working Group drawn from all relevant directorates and offices of the IEA. The study was designed

and directed by Dave Turk, Director (Acting) of Sustainability, Technology and Outlooks and Head of the Energy Environment Division and Laura Cozzi, Head of the Energy Demand Outlook Division.

The chapters on impacts on energy demand (Chapter 2) and energy supply (Chapter 3)

were co-ordinated by Ellina Levina with Stéphanie Bouckaert (buildings), John Dulac (buildings), Carlos Fernandez Alvarez (coal), Araceli Fernandez Pales (industry), George Kamiya (transport), Markus Klingbeil (oil and gas), and Brent Wanner (power).

The chapter on system-wide impacts (Chapter 4) was led by Vincenzo Franza and Luis Munuera. The chapter on energy use by ICT (Chapter 5) was led by George Kamiya and Eric Masanet. The chapters on risks (Chapter 6) and on policy (Chapter 7) were

led by Jesse Scott. Eric Masanet also led on demand sectors during the earlier stages

of the project.

Other principal contributors to the report were: Thibaut Abergel (buildings), Jan Bartos

(cybersecurity), Elie Bellevrat (industry), Simon Bennett (investment and jobs),

Thomas Berly (carbon capture and storage), Vlad Kubecek (statistics), Gee Yong Law

(Singapore case study), Christophe McGlade (oil and gas), David Morgado (energy

efficiency), Duncan Millard (statistics), Joe Ritchie (industry), Melanie Slade (energy

efficient appliances), Jacob Teter (transport), Michael Waldron (investment), Molly Walton (energy access), and Kira West (industry). Rob Stone led on video and Jon Custer on the website. Jad Mouawad helped to shape the text, and with the design of the website and video. Teresa Coon and Lisa-Marie Grenier provided essential support.

Valuable input, comments and feedback were provided by the senior management and many other colleagues within the IEA – in particular, Paul Simons, Keisuke Sadamori,

Rebecca Gaghen, Laszlo Varro, Aad van Bohemen, Alessandro Blasi, Tyler Bryant,

Pierpaolo Cazzola, Timon Dubbeling, Paolo Frankl, Peter Fraser, Jean-François Gagné, Christina Hood, Zoe Hungerford, Simone Landolina, Louis Mark, Tomi Motoi, Simon

Mueller, Yoko Nobuoka, Peter Sopher, Glenn Sondak, Samuel Thomas, Aya Yoshida, and Martin Young. Experts from the OECD Going Digital project team also contributed to the report, in particular Andrew Wyckoff, Molly Lesher and Laurent Bernat.

Therese Walsh of the IEA Communication and Information Office for their help in

Acknowledgements

5

© OECD/IEA, 2017

Thanks also to Muriel Custodio, Astrid Dumond, Katie Russell, Bertrand Sadin and

producing the final report. Trevor Morgan (Menecon) carried editorial responsibility. Justin French-Brooks (Word to Dialogue) was the copy-editor.

A number of workshops and meetings were organised to receive essential input to this study, including: Digitalization & Energy Workshop, Paris, 5 April 2017 and Energy

Sector Cybersecurity Workshop, Paris, 6 April 2017. The participants offered valuable

new insights, feedback and data for this analysis. Further details are available at: www.iea.org/workshops/digitalization-and-energy.html.

Additional valuable inputs and insights were provided by Brian Neal, David Scuderi and

colleagues (Amazon), Jeffrey Dahmus and Craig Arnold (Apple), Caroline Baylon (Axa),

Julian Gray (BP), Rob Jenks (C3IoT), Jean Lamming and colleagues (California Public Utilities Commission), Taishi Sugiyama (Canon Institute for Global Studies), Karthik

Ganesan (CEEW), Sudeep Maitra and colleagues (Centrica), Mushfiq Sarker (Drift), Daniel Wragge and colleagues (EEX and EpexSpot), Michael Ding and Tim Naylor (Envision Energy), Jens Malmodin (Ericsson), Sonal Pandya Dalal, Dimitry Gershenson

and colleagues (Facebook), Luis Neves and Chiara Venturini (Global e-Sustainability

Initiative), Gary Cook (Greenpeace), Umeshwar Dayal and colleagues (Hitachi Insight

Group), Erol Gelenbe (Imperial College London), Peter Evans (KPMG), Lynn Price, Bruce Nordman, Alan Meier and colleagues (Lawrence Berkeley National Laboratory),

Michelle Patron, Brian Janous, Brian Marrs and colleagues (Microsoft), Doug Arent and colleagues (National Renewable Energy Laboratory), Jorge Vasconcelos (Newes),

Tom Martin and Jane Kruse McClellan (PG&E), Jean-Jacques Marchais (Schneider Electric), Alexander Vasylchenko (Sofitto), Sonia Dunlop (Solar Europe), Jonathan

Koomey (Stanford University), Sarah Van Cleve (Tesla), Igor Linkov (United States

Army Engineer Research and Development Center), and Aloys Nghiem (Wind Europe). Peer reviewers

Many senior government officials and international experts provided input and reviewed preliminary drafts of the report. Their comments and suggestions were of

great value. They include:

Florian Ausfelder

DECHEMA

Joerg Boettcher

Honda

Steven Beletich Carlo Bozzoli

Stefano Bracco Vytautas Butrimas Geoffrey Cann

Yin Cao

Russell Conklin 6

Sony

Connected Devices Alliance

Enel

Agency for the Cooperation of Energy Regulators, European

Union

NATO Energy Security Center of Excellence Deloitte

Energy Blockchain Labs

United States Department of Energy Acknowledgements

© OECD/IEA, 2017

Joshua Aslan

Giorgio Corbetta Mike Corcoran

Vlad C. Coroama

EEX

University of Warwick Centre for Cyber Security, United

Kingdom

ETH Zurich, Switzerland

Philippine De T'Serclaes

Schneider Electric

Mark Ellis

IEA 4E TCP

Giles Dickson

Alexander Folz

Wind Europe

Federal Ministry for Economic Affairs and Energy, Germany

Christoph Frei

World Energy Council

Livio Gallo

Enel

William Garcia

Cefic

Torben Funder-Kristensen Danfoss

Matt Golden

Avi Gopstein

Energy Market Authority, Singapore OpenEE

National Institute of Standards and Technology, United States

Peter Gornischeff

Permanent Representation to the OECD, Republic of Estonia

Bil Hallaq

University of Warwick Centre for Cyber Security, United

Angelo Guardo

RES4MED

Kingdom

Carol Hawk

United States Department of Energy

Erfan Ibrahim

National Renewable Energy Laboratory, United States

Steve Heinen Agnes Koh

Jonathan Koomey Lorenzo Kristov Anna Lerner

Kaili Lévesque Igor Linkov

Claude Lorea

Don MacKenzie Jens Malmodin

Kiyoshi Matsuda

University College Dublin Electricity Research Centre, Ireland Energy Market Authority, Singapore Stanford University, United States California ISO World Bank

Natural Resources Canada

United States Army Engineer Research and Development

Center

CEMBUREAU

University of Washington, United States Ericsson

Mitsubishi Chemicals Holdings Corporation

Pat Meehan

National Oil Reserves Agency, Ireland

Simone Mori

Enel

Satoshi Mizunoe Luis Neves

Nicola Rega

Ministry of Economy, Trade and Industry, Japan Global e-Sustainability Initiative / Deutsche Telekom

Confederation of European Paper Industries

Brian Ricketts

Euracoal

Miguel Sánchez-Fornié

University of Comillas, Spain / Iberdrola

Laurent Schmitt

ENTSO-E

Arman Shehabi

Lawrence Berkeley National Laboratory, United States

Seth Roberts Avik Sarkar

Susan Shaheen

Acknowledgements

Dow / International Council of Chemical Associations NITI Aayog, India

University of California Berkeley, United States

7

© OECD/IEA, 2017

Faith Gan

Hans-Paul Siderius

IEA 4E TCP

Jorge Vasconcelos

Newes

Kazushige Tanaka

Francesco Venturini Zia Wadud

Ministry of Economy, Trade and Industry, Japan Enel

University of Leeds, United Kingdom

Thomas Weisshaupt

Wirepas

Roberto Zangrandi

European DSO for Smart Grids

Taiki Yamada

Christian Zinglersen

Ministry of Economy, Trade and Industry, Japan

Clean Energy Ministerial

The individuals and organisations that contributed to this study are not responsible for

any opinions or judgments it contains. All errors and omissions are solely the responsibility of the IEA.

Comments and questions are welcome and should be addressed to: Dave Turk and Laura Cozzi Directorate of Sustainability, Technology and Outlooks International Energy Agency 31-35, rue de la Fédération 75739 Paris Cedex 15 France Email: [email protected]

We look forward to your continued feedback and partnership, including suggestions and guidance on further needed analysis.

8

Acknowledgements

© OECD/IEA, 2017

More information about the report is available at: www.iea.org/digital.

Table of contents Executive summary ............................................................................................................................ 15 Digitalization: A new era in energy? ................................................................................................ 21 Highlights...................................................................................................................................................................... 21 What is digitalization? ................................................................................................................................................ 22 How pervasive is digitalization? ............................................................................................................................... 22 A new era of digitalization in energy? ..................................................................................................................... 25 Purpose and structure of report .............................................................................................................................. 26 References ................................................................................................................................................................... 28

Impact of digitalization on energy demand in transport, buildings and industry ...................... 29 Highlights...................................................................................................................................................................... 29 Introduction .................................................................................................................................................................. 30 Transport ...................................................................................................................................................................... 31 Deployment of digital technologies in transportation ................................................................................... 31 Implications for energy use and emissions: Focus on road transport ....................................................... 35

Barriers to digitalization and policy considerations ....................................................................................... 39 Buildings ....................................................................................................................................................................... 41 Digitalization of the provision of energy services ........................................................................................... 41 Energy implications of digitalization in buildings ........................................................................................... 42 Barriers to further digitalization and policy considerations ......................................................................... 48 Industry ......................................................................................................................................................................... 49 Digitalization within industrial plants ............................................................................................................... 51 Digitalization beyond the industrial plant fence ............................................................................................. 55

Barriers to further digitalization and policy considerations ......................................................................... 56 References ................................................................................................................................................................... 58

Impact of digitalization on oil and gas, coal, and power supply .................................................. 65 Highlights...................................................................................................................................................................... 65 Oil and gas.................................................................................................................................................................... 67

Table of contents

9

© OECD/IEA, 2017

Introduction .................................................................................................................................................................. 66

Potential for further digitalization in the upstream sector ........................................................................... 67 Monitoring of methane emissions ..................................................................................................................... 70 Improved downstream operations ..................................................................................................................... 72 Barriers to further digitalization ......................................................................................................................... 72 Coal ................................................................................................................................................................................ 73 Deployment of digital technologies ................................................................................................................... 73 Potential impacts of digitalization ..................................................................................................................... 75

Barriers to further digitalization ......................................................................................................................... 75 Power sector ................................................................................................................................................................ 76 Data and analytics in power plants and electricity networks ...................................................................... 77 Potential benefits of digital data and analytics .............................................................................................. 78

Barriers to further digitalization ......................................................................................................................... 81 References ................................................................................................................................................................... 82

System-wide impacts: From energy silos to digitally interconnected systems......................... 83 Highlights...................................................................................................................................................................... 83 Digitalization is transforming the way the electricity system functions .......................................................... 84 Smart demand response ........................................................................................................................................... 89 Integration of variable renewables .......................................................................................................................... 93 The growing role of EVs............................................................................................................................................. 94 Distributed generation, mini-grids and “prosumers” ........................................................................................... 97 Towards ever-smarter energy systems .................................................................................................................. 99 References ................................................................................................................................................................ 101

Energy use by information and communications technologies ................................................. 103 Highlights................................................................................................................................................................... 103 Introduction ............................................................................................................................................................... 104 Current energy use and near-term outlook ........................................................................................................ 105 Data centres........................................................................................................................................................ 105 Data transmission networks ............................................................................................................................ 108

Connected devices............................................................................................................................................. 112 Long-term energy demand outlook ...................................................................................................................... 113 Policy considerations .............................................................................................................................................. 116

10

Table of contents

© OECD/IEA, 2017

References ................................................................................................................................................................ 118

Cross-cutting risks: Cybersecurity, privacy and economic disruption ...................................... 123 Highlights................................................................................................................................................................... 123 Digital security.......................................................................................................................................................... 124 Cyber threats to energy security ..................................................................................................................... 124 Building resilience.............................................................................................................................................. 127 Best practices and policy ................................................................................................................................. 130 Data privacy and ownership .................................................................................................................................. 134 Balancing privacy and other policy objectives ............................................................................................. 136 Economic disruption: Jobs and skills in an increasingly digitalized energy world ...................................... 139 References ................................................................................................................................................................ 142

Policy .................................................................................................................................................. 145 Highlights................................................................................................................................................................... 145 Promoting energy access, sustainability and security ..................................................................................... 146 Energy access ..................................................................................................................................................... 146 Environmental sustainability ........................................................................................................................... 149 Energy security ................................................................................................................................................... 152 Digital tools to improve the policy-making process .......................................................................................... 152 Better energy statistics..................................................................................................................................... 152

Data on energy efficiency standards for appliances .................................................................................. 154 Policy frameworks and market design................................................................................................................. 157 No-regrets policy recommendations .................................................................................................................... 159 References ................................................................................................................................................................ 163

Glossary.............................................................................................................................................. 165 Abbreviations and acronyms ........................................................................................................... 183 Units of measure ...................................................................................................................................................... 185

List of figures Figure 1.1

Global internet traffic ...................................................................................................................... 23

Figure 1.3

Largest companies by market capitalisation ............................................................................. 24

Figure 1.4 Figure 2.1 Figure 2.2 Figure 2.3

Global trends in connectivity ......................................................................................................... 23 Investments in digital electricity infrastructure and software ................................................ 26 Digitalization’s potential impact on transport, buildings, and industry ................................ 30 Range of possible energy impacts from vehicle automation in the United States............. 37 Digitalization’s impact on energy use and emissions reductions in road freight ............... 38

Table of contents

11

© OECD/IEA, 2017

Figure 1.2

Figure 2.4

Cumulative energy savings in buildings from widespread digitalization .............................. 42

Figure 2.6

Global energy use and average energy intensity of active controls in buildings ................ 44

Figure 2.5 Figure 2.7 Figure 2.8 Figure 2.9

Figure 2.10 Figure 2.11 Figure 2.12 Figure 3.1 Figure 3.2

Figure 3.3 Figure 3.4 Figure 3.5 Figure 3.6 Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4 Figure 4.5 Figure 4.6 Figure 4.7 Figure 4.9 Figure 5.1 Figure 5.3 Figure 5.4 Figure 5.5 Figure 6.1 Figure 6.2 Figure 7.1 Figure 7.2

Household electricity consumption of appliances and other small plug loads ................... 44 Potential electricity savings from smart lighting in buildings to 2040 .................................. 47 Application of digital technologies and strategies in industry ................................................ 50 Stages of digitalization in industrial process control................................................................ 51 Energy savings, payback and investment in digitally enabled optimisation of process

controls in the United States......................................................................................................... 52 Potential energy savings from improvements in process control enabled by digitalization

by subsector in Australia, 2010-11 ............................................................................................... 52 Number of industrial robots worldwide by sector ..................................................................... 53 Digitalization’s potential impact in oil and gas, coal, and power ........................................... 66 Example of digital visualisation of an oil well and reservoir ................................................... 68 Impact of digitalization on global technically recoverable oil and gas resources............... 69 Digitally-connected remote operations in oil and gas fields .................................................. 71 Impact of digitalization on electricity sector assets ................................................................. 77 Illustrative potential worldwide cost savings from enhanced digitalization in power

plants and electricity networks over 2016-40 ............................................................................ 78 Traditional structure of electricity sector ................................................................................... 85 The role of digitalization in reshaping the electricity sector ................................................... 85 Possible steps in the digital transformation of the electricity system .................................. 86 Unit costs of key emerging electricity technologies ................................................................. 88 Four key elements in the creation of a digitally interconnected electricity system ........... 89 Impact of demand response on the daily load curve................................................................ 90 Global potential of demand response and its implications ..................................................... 91 Impact of smart versus standard charging of EVs .................................................................... 96 Global data centre energy demand by end use and data centre type ............................... 106 Electricity use by internet data transmission networks ........................................................ 109 Global IP traffic by access mode ............................................................................................... 110 Electricity intensity of network transmission by access type in 2015................................ 111 Capabilities of companies to handle cyber-attacks .............................................................. 130 Behind-the-meter energy use data ........................................................................................... 135 Mobile phone ownership and electricity access in selected sub-Saharan African

countries ......................................................................................................................................... 147 Example of a QR code on an appliance energy efficiency label in China ......................... 156

List of figures in boxes Figure 2.13

Potential reductions in metals demand and fuel use in the US commercial aircraft fleet

Figure 4.8

Flexibility measures and their implications in the European Union in 2040........................ 94

Figure 5.6 12

Data centre energy demand in India by end use and data centre type ............................. 108 Corporate renewable power purchase agreements by type of buyer ................................ 114 Table of contents

© OECD/IEA, 2017

Figure 5.2

from the use of 3D printing............................................................................................................ 54

List of tables Table 6.1 Table 7.1

Open source information regarding cyber-attacks affecting energy infrastructure........ 125 Illustrative electricity offers from PAYG market leaders in Kenya ..................................... 147

List of boxes Box 2.1

Drones ................................................................................................................................................ 34

Box 2.3

3D printing in US aircraft manufacturing .................................................................................... 54

Box 4.1 Box 4.2 Box 4.3 Box 4.4 Box 4.5 Box 4.6 Box 5.1 Box 5.2 Box 6.1 Box 6.2 Box 6.3 Box 6.4 Box 6.5 Box 7.1 Box 7.2 Box 7.3

LEDs as a value-added tool for energy services ....................................................................... 47 From one-way electricity grids to smart energy systems ........................................................ 87 Interruption of large electricity consumers as the first form of demand response ............ 90 Demand response market and aggregators ............................................................................... 92 Digitalization can help to decarbonise the EU electricity sector ........................................... 93 Vehicle-to-grid: EVs as mobile batteries .................................................................................... 96 Blockchain, transactive energy and peer-to-peer trading ....................................................... 97 Two futures of data centre energy use in India...................................................................... 107 Data centres, renewable energy and demand response potential ..................................... 114 Malware, ransomware, phishing and bots ............................................................................... 124 IEA digital resilience efforts........................................................................................................ 133 Interoperability and standardisation ......................................................................................... 136 Examples of specific approaches to privacy regulation ........................................................ 138 Automation and jobs: Focus on autonomous cars and trucks ............................................ 141 Digitalization and CCS ................................................................................................................. 150 Types of digital data..................................................................................................................... 153 Singapore’s Smart Nation Vision ............................................................................................... 158

Table of contents

13

© OECD/IEA, 2017

Box 2.2

© OECD/IEA, 2017

Executive summary Over the coming decades, digital technologies are set to make energy systems around

the world more connected, intelligent, efficient, reliable and sustainable. Stunning advances in data, analytics and connectivity are enabling a range of new digital applications such as smart appliances, shared mobility, and 3D printing. Digitalized

energy systems in the future may be able to identify who needs energy and deliver it at the right time, in the right place and at the lowest cost. But getting everything right will not be easy.

Digitalization is already improving the safety, productivity, accessibility and

sustainability of energy systems. But digitalization is also raising new security and privacy risks. It is also changing markets, businesses and employment. New business

models are emerging, while some century-old models may be on their way out.

Policy makers, business executives and other stakeholders increasingly face new and

complex decisions, often with incomplete or imperfect information. Adding to this challenge is the extremely dynamic nature of energy systems, which are often built on

large, long-lived physical infrastructure and assets.

This report seeks to provide more clarity for decision makers on what digitalization means for energy – shining a light on both its enormous potential and most pressing

challenges. It is the first effort from the International Energy Agency (IEA) to comprehensively depict how digitalization is reshaping the energy sector. It is also intended to serve as a springboard for future work. Digitalization: A new era in energy? The energy sector has been an early adopter of digital technologies. In the 1970s, power utilities were digital pioneers, using emerging technologies to facilitate grid

management and operation. Oil and gas companies have long used digital

technologies to model exploration and production assets.

Recent technological advances and trends are truly astounding. Data are growing at an

exponential rate – internet traffic has tripled in only the past five years. There are now

more mobile phone subscriptions than people in the world.

Advancing technology, falling costs, and ubiquitous connectivity are opening the door build new architectures of interconnected energy systems, including breaking down traditional boundaries between demand and supply. Executive summary

15

© OECD/IEA, 2017

to new models of producing and consuming energy. Digitalization holds the potential to

The impact of these tremendous digital advances and their rapid deployment across

the energy landscape raise the fundamental question of whether we are on the cusp of

a new digital era in energy and, if so, what are the emerging trends. This report attempts to answer these questions.

All energy demand sectors are feeling the effects Digital technologies are already widely used in energy end-use sectors, with the

widespread deployment of potentially transformative technologies on the horizon, such as autonomous cars, intelligent home systems and machine learning. While these

technologies could improve efficiency, some could also induce rebound effects that increase overall energy use.

In the transport sector, cars, trucks, planes, ships, trains and their supporting infrastructure are all becoming smarter and more connected, improving safety and efficiency. Digitalization could have its biggest impact on road transport, where

connectivity and automation (alongside further electrification) could dramatically

reshape mobility. Meanwhile, the overall net impacts on energy use are highly uncertain. Over the long term, under a best-case scenario of improved efficiency through automation and ride-sharing, energy use could halve compared with current

levels. Conversely, if efficiency improvements do not materialise and rebound effects

from automation result in substantially more travel, energy use could more than double.

In buildings, our analysis shows that digitalization could cut energy use by about 10% by using real-time data to improve operational efficiency. Smart thermostats can

anticipate the behaviour of occupants (based on past experience) and use real-time weather forecasts to better predict heating and cooling needs. Smart lighting can

provide more than just light when and where it is needed; light-emitting diodes (LEDs)

can also include sensors linked to other systems, for example, helping to tailor heating and cooling services.

In industry, many companies have a long history of using digital technologies to improve safety and increase production. Further cost-effective energy savings can be achieved through advanced process controls, and by coupling smart sensors and data

analytics to predict equipment failure. 3D printing, machine learning and connectivity could have even greater impacts. For example, 3D printing can be used to make

16

Executive summary

© OECD/IEA, 2017

aircraft lighter, reducing both the materials to build the plane and the fuel to fly it.

Energy suppliers will reap greater productivity and improve safety The oil and gas industry has long used digital technologies, notably in upstream, and

significant potential remains for digitalization to further enhance operations. Widespread use of digital technologies could decrease production costs between 10% and 20%, including through advanced processing of seismic data, the use of sensors,

and enhanced reservoir modelling. Technically recoverable oil and gas resources could be boosted by around 5% globally, with the greatest gains expected in shale gas.

In the coal industry, digital technologies are increasingly being used in geological

modelling, process optimisation, automation, predictive maintenance, and to improve

worker health and safety. Specific examples include driverless trucks and tele-remote equipment operated from the control room. Digitalization’s overall impact, however, may be more modest than in other sectors.

In the power sector, our analysis shows that digitalization has the potential to save around 80 billion United States dollars (USD) per year, or about 5% of total annual

power generation costs. This can be achieved by reducing operation and maintenance

costs, improving power plant and network efficiency, reducing unplanned outages and

downtime, and extending the operational lifetime of assets. One example of this is the use of drones to cheaply monitor thousands of kilometres of transmission lines over rough terrain.

Digitally interconnected systems could fundamentally transform electricity markets

The greatest transformational potential for digitalization is its ability to break down boundaries between energy sectors, increasing flexibility and enabling integration

across entire systems. The electricity sector is at the heart of this transformation, where digitalization is blurring the distinction between generation and consumption, and enabling four interrelated opportunities:

• “Smart demand response” could provide 185 gigawatts (GW) of system

flexibility, roughly equivalent to the currently installed electricity supply

capacity of Australia and Italy combined. This could save USD 270 billion of

investment in new electricity infrastructure that would have otherwise been

needed. In the residential sector alone, 1 billion households and 11 billion smart

appliances could actively participate in interconnected electricity systems, allowing these households and devices to alter when they draw electricity from the grid.

better match energy demand to times when the sun is shining and the wind is

blowing. In the European Union alone, increased storage and digitally-enabled Executive summary

17

© OECD/IEA, 2017

• Digitalization can help integrate variable renewables by enabling grids to

demand response could reduce curtailment of solar photovoltaics (PV) and

wind power from 7% to 1.6% in 2040, avoiding 30 million tonnes of carbon dioxide emissions in 2040.

• Rolling out smart charging technologies for electric vehicles could help shift charging to periods when electricity demand is low and supply is abundant. This would provide further flexibility to the grid while saving between

USD 100 billion and USD 280 billion in avoided investment in new electricity infrastructure between 2016 and 2040.

• Digitalization can facilitate the development of distributed energy resources, such as household solar PV panels and storage, by creating better incentives and making it easier for producers to store and sell surplus electricity to the

grid. New tools such as blockchain could help to facilitate peer-to-peer

electricity trade within local energy communities. Direct energy consumption of digital technologies

Digital technologies that make all these potential benefits possible also use energy. As billions of new devices become connected over the coming years, they will draw

electricity at the plug while driving growth in demand for – and energy use by – data

centres and network services. However, sustained gains in energy efficiency could keep overall energy demand growth largely in check for data centres and networks over the next five years.

Data centres worldwide consumed around 194 terawatt hours (TWh) of electricity in 2014, or about 1% of total demand. Although data centre workload is forecast to triple by 2020, related energy demand is expected to grow by only 3% thanks to continued efficiency gains.

Data networks, which form the backbone of the digital world, consumed around

185 TWh globally in 2015, or another 1% of total demand, with mobile networks accounting for around two-thirds of the total. Depending on future efficiency trends, by

2021 electricity consumption from data networks could increase by as much as 70% or

fall by up to 15%. This large range highlights the critical role of policy in driving efficiency gains.

Beyond the next five years, providing credible assessments of energy use by digital technologies is extremely difficult. Direct energy use over the long run will continue to

be a battle between data demand growth versus the continuation of efficiency

18

Executive summary

© OECD/IEA, 2017

improvements.

Building digital resilience to prepare for inevitable cyber-attacks While digitalization can bring many positive benefits, it can also make energy systems

more vulnerable to cyber-attacks. To date, the disruptions caused to energy systems by reported cyber-attacks have been relatively small. However, cyber-attacks are

becoming easier and cheaper to organise. Moreover, the growth of the Internet of Things (IoT) is increasing the potential “cyber-attack surface” in energy systems.

Full prevention of cyber-attacks is impossible, but their impact can be limited if countries and companies are well-prepared. Building system-wide resilience depends

on all actors and stakeholders first being aware of the risks. Digital resilience also needs to be included in technology research and development efforts as well as built into policy and market frameworks.

International efforts can also help governments, companies and others to build up digital resilience capabilities. A variety of organisations are involved, each contributing

its comparative strengths, including to share best practices and policies as well as to

help mainstream digital resilience in energy policy making. Managing privacy concerns and impacts to jobs

Privacy and data ownership are also major concerns for consumers, especially as more detailed data are collected from a growing number of connected devices and appliances. For instance, data on energy use in households collected by smart meters

can be used to tell when someone is home, using the shower, or making tea. At the same time, aggregated and anonymised individual energy use data can improve

understanding of energy systems, such as load profiles, and help lower costs for individual consumers. Policy makers will need to balance privacy concerns with these

other objectives, including promoting innovation and the operational needs of utilities.

Digitalization is also affecting jobs and skills in a variety of energy sectors, changing

work patterns and tasks. This is creating new job opportunities in some areas while causing losses in others. Policy makers in the energy field should participate in

broader government-wide deliberations about these effects and how to respond to them.

Government policy design is critical Policy and market design are vital to steering digitally enhanced energy systems onto

an efficient, secure, accessible and sustainable path. For example, digitalization can assist in providing electricity to the 1.1 billion people who still lack access to it. New emissions and technologies to track air pollution at the neighbourhood level.

Executive summary

19

© OECD/IEA, 2017

digital tools can promote sustainability, including satellites to verify greenhouse gas

Policy-making processes can also benefit from more timely and sophisticated

collection and publication of energy data that greater access to digital data could facilitate. Emerging low-cost digital tools, such as online registries, web-crawled data and quick response codes, can lead to more targeted and responsive policy regimes.

While there is no simple roadmap to show how an increasingly digitalized energy world will look in the future, the IEA recommends ten no-regrets policy actions that governments can take to prepare:

• Build digital expertise within their staff. • Ensure appropriate access to timely, robust, and verifiable data. • Build flexibility into policies to accommodate new technologies and developments.

• Experiment, including through “learning by doing” pilot projects. • Participate in broader inter-agency discussions on digitalization. • Focus on the broader, overall system benefits. • Monitor the energy impacts of digitalization on overall energy demand. • Incorporate digital resilience by design into research, development and product manufacturing.

• Provide a level playing field to allow a variety of companies to compete and serve consumers better.

• Learn from others, including both positive case studies as well as more

20

Executive summary

© OECD/IEA, 2017

cautionary tales.

Chapter 1

Digitalization: A new era in energy? Highlights •

Digitalization describes the growing application of information and

communications technologies (ICT) across the economy, including energy systems.



The trend toward greater digitalization is enabled by advances in data, analytics and connectivity: •

increasing volumes of data thanks to declining costs of sensors and data storage



rapid progress in advanced analytics, such as machine learning



greater connectivity of people and devices as well as faster and cheaper data transmission.



Digitalization encompasses a range of digital technologies, concepts and trends, such as artificial intelligence, the “Internet of Things” (IoT) and the Fourth Industrial Revolution.



Some of these digitalization trends are truly astounding – 90% of the data in the world today were created in just the past two years, and there are now more mobile phone subscriptions in the world than there are people.



Digital technologies have been helping to improve energy systems for decades, but the pace of their adoption is accelerating. For example,

global investment in digital electricity infrastructure and software has been increasing by 20% annually in recent years.



These digitalization trends raise the fundamental question of whether we

Digitalization: A new era in energy?

21

© OECD/IEA, 2017

are entering a new era of digitalization in energy.

What is digitalization? Digitalization describes the growing application of ICT across the economy, including energy systems. Consider, for instance, the many digital technologies in play when

hailing a ride through an app – your smartphone and its mobile internet connection;

the global positioning system (GPS) that pinpoints your location; and the real-time traffic data and advanced analytics that direct your driver.

Digitalization can be thought of as the increasing interaction and convergence between the digital and physical worlds. The digital world has three fundamental elements:

• Data: digital information • Analytics: the use of data to produce useful information and insights • Connectivity: exchange of data between humans, devices and machines (including machine-to-machine), through digital communications networks.

The trend toward greater digitalization is enabled by advances in all three of these

areas: increasing volumes of data thanks to the declining costs of sensors and data storage, rapid progress in advanced analytics and computing capabilities, and greater connectivity with faster and cheaper data transmission.

How pervasive is digitalization? Around 90% of the data in the world today were created over the past two years (IBM,

2017). This exponential growth has led to the use of increasingly large units of measurement. For example, global annual internet traffic surpassed the exabyte

threshold in 2001 and is expected to pass the zettabyte threshold by 2017. (One exabyte is 1 000 000 000 000 000 000 bytes, or 1018 bytes, and one zettabyte equals

1 000 exabytes, or 1021 bytes [Figure 1.1].) Internet traffic has tripled over the past five

years (Cisco, 2017a).

Advances in computing power and efficiency have enabled more powerful and

sophisticated analytics, such as artificial intelligence (AI) and automation. For example, AI and the ability for machines to process large and complex data are bringing us closer to fully automated cars and trucks. Advanced analytics also enable

the creation of digital replicas of physical assets (“digital twins”) that can be used to

simulate and optimise industrial design and oil and gas drilling. Analytics also enable

sophisticated control of building and industrial process equipment, and are ushering in

22

Digitalization: A new era in energy?

© OECD/IEA, 2017

a new wave of automation in manufacturing, including robotics and 3D printing.

Figure 1.1 Global internet traffic KB MB GB TB PB EB ZB YB

kilobyte megabyte gigabyte terabyte petabyte exabyte zettabyte yottabyte

1987 2 TB

103 bytes 106 bytes 109 bytes 1012 bytes 1015 bytes 1018 bytes 1021 bytes 1024 bytes

1997 60 PB

2007 54 EB

2017 1.1 ZB

Key message: The world is witnessing an explosion in data, with internet traffic expected to enter the zettabyte era in 2017.

Sources: IEA calculations based on Cisco (2017b) and Cisco (2015).

People and devices are also becoming connected in ever-increasing numbers. More

than 3.5 billion people, or nearly half the global population, now use the internet – up

from only 500 million in 2001 (ITU, 2017). About 54% of households now have internet

access at home (Figure 1.2, left). In the last five years, global mobile broadband

subscriptions increased threefold and surpassed 4 billion active subscriptions in 2017,

while mobile phone subscriptions reached 7.7 billion. The developing world is leading

the more recent growth in connectivity, accounting for almost 90% of the total growth in mobile broadband subscriptions over the past five years (Figure 1.2, right). Figure 1.2 Global trends in connectivity

Access rates 100%

5

80%

4

60%

3

40%

2

20%

1

0% 2007

2009

2011

2013

Electricity

2015 Internet

2017*

0

Mobile broadband subscriptions

billions

2007

2009

2011

Developed countries

2013

2015

2017*

Developing countries

Notes: * denotes estimate for 2017; “Internet access” is defined as households with internet access at home; developed/developing country classifications are based on the UN M49. Sources: ITU (2017), ICT Facts and Figures 2017; IEA (2017), Energy Access Outlook: From Poverty to Prosperity.

Digitalization: A new era in energy?

23

© OECD/IEA, 2017

Key message: Connectivity is increasing rapidly, particularly in the developing world.

Everyday objects such as watches, home appliances and cars are being connected to

communications networks – the “Internet of Things” (IoT)1 – to provide a range of

services and applications, such as personal healthcare, smart electricity grids,

surveillance, home automation and intelligent transport. The number of connected IoT devices is forecast to grow from 8.4 billion in 2017 to over 20 billion by 2020 (Gartner, 2017).

Financial markets, investment trends and digital disruption in other sectors also signal

the pervasiveness of digitalization and the greater interactions between digital and

energy worlds. The five largest publicly traded companies by market capitalisation currently are all ICT companies with businesses that either depend on digital technologies or provide related services (Figure 1.3). Energy companies, on the other

hand, remain global leaders by revenue: six of the top 10 companies are in the energy sector, with the sole digital company – Apple – in 9th place (Fortune, 2017). Figure 1.3 Largest companies by market capitalisation

2002

ICT Energy Other

2007

2012

2017

Key message: Digital technology companies have become global leaders by market capitalisation, though energy companies still lead in revenues.

Notes: Rankings are for publicly traded companies; market capitalisations calculated at the end of Q2; circle sizes are relative to market capitalisation.

1

At the most basic level, IoT is the concept of connecting everyday objects to networks to provide a range of services or

applications. IoT encompasses both machine-to-machine (M2M) communication (where devices interact and share data without the direct involvement of people) and connecting “things” to networks to enable people to remotely control processes or manage

24

Digitalization: A new era in energy?

© OECD/IEA, 2017

their devices (IEA, 2014).

Investments and acquisitions by ICT companies in energy-related companies signal

the sector’s increasing interest in energy (e.g. Google/Nest; Oracle/Opower). ICT

companies are also investing more in energy, notably renewables, accounting for more

than half of total corporate power purchase agreements (PPAs) for renewables over the past two years (see Box 5.2).

While the rapid rise of Airbnb and Uber might show how quickly digital disruption could occur, there is considerable variability in disruption potential across sectors. For instance, some have less of a “tether” to the physical world (e.g. media), while others

(e.g. agriculture) fundamentally involve the production and supply of physical goods, even though digital technologies can have a significant impact on the way business is conducted.

A new era of digitalization in energy? Digital technologies have been helping to improve energy systems for decades. In fact, the energy sector has often been an early adopter of large information technology (IT)

systems. In the 1970s, power utilities were digital pioneers, using IT to facilitate management and operation of the grid. Electricity markets are now monitored and

controlled in real time over vast geographical areas serving large numbers of customers.

Oil and gas companies have long used digital technologies to model exploration and

production assets, including reservoirs and pipelines. The industrial sector has used

process controls and automation for decades, particularly in heavy industry, to maximise quality and yields while minimising energy use. Intelligent transport systems

are using digital technologies in all modes of transport to improve safety and efficiency.

The pace of digitalization in energy is increasing. Investment in digital technologies by

energy companies has risen sharply over the last few years. For example, global

investment in digital electricity infrastructure and software has grown by over 20%

annually since 2014, reaching USD 47 billion in 2016 (Figure 1.4). This digital investment in 2016 was almost 40% higher than investment in gas-fired power

generation worldwide (USD 34 billion) and almost equal to total investment in India’s electricity sector (USD 55 billion).

The astounding advances in digitalization and their rapid deployment across the

energy landscape raise the fundamental question of whether we are on the cusp of a new digital era in energy. The remainder of this report attempts to answer this

Digitalization: A new era in energy?

25

© OECD/IEA, 2017

fundamental question.

Figure 1.4 Investments in digital electricity infrastructure and software 80

USD (2016) billion Electricity systems software Industrial energy management software

60

Building energy controls EV chargers

40

Smart grid infrastructure Smart meters

20 0

Other 2014

2015

2016

Global gas power generation

India Video game power industry sector revenues

Key message: Investment in digital electricity infrastructure and software grew over 20% annually between 2014 and 2016, overtaking global investment in gas-fired power generation.

Notes: Global gas power generation and India power sector are 2016 investments; EV = electric vehicle. Sources: Calculations for investment in digital infrastructure and software based on MarketsandMarkets (2016), Internet of Things in Utility Market; BNEF (2016), Digital Energy Market Outlook.

Purpose and structure of report This report describes the status of digitalization in energy, how it is affecting energy

systems, what might happen in the future and what all this means for policy makers,

companies and consumers. Digitalization and energy is a complex and constantly

evolving topic. As such, this report is not intended to be a definitive, exhaustive analysis. Rather, it seeks to shed light on how the energy and digital worlds interact

and to serve as a springboard for further analysis by the International Energy Agency (IEA).

The rest of this report is structured as follows:

• Chapter 2 analyses the current and potential impact of digitalization on energy use in the three main demand sectors – transport, buildings and industry.

• Chapter 3 looks at the impact of digital technologies on energy supply, focusing on oil and gas, coal and electricity.

• Chapter 4 explores the more transformational potential of digitalization to break down individual energy silos, to blur the boundaries between energy demand and supply, and to help create a highly interconnected electricity system.

• Chapter 5 assesses the direct use of energy by digital technologies themselves, • Chapter 6 considers the main cross-cutting risks of digitalization – cybersecurity, privacy and economic disruption.

26

Digitalization: A new era in energy?

© OECD/IEA, 2017

including data centres, data networks and connected devices.

• Chapter 7 focuses on the critical role of government policy – how digitalization

can help achieve policy goals and improve the policy-making process, and the importance of integrated policy frameworks and market design. It concludes with a set of no-regrets recommendations for policy makers.

Due to the cross-cutting nature of digitalization, certain subjects are covered in multiple chapters. For instance, the impact of digitalization on electricity systems is discussed in both Chapters 3 and 4. Similarly, policy implications related to particular

energy sectors (e.g. transport) are addressed, respectively, in Chapters 2, 3 and 4, while cross-cutting policy issues are discussed in Chapters 6 and 7.

This report is also complemented by a new, interactive website – www.iea.org/digital – which will provide a dynamic means to access the information in this report.

Many terms and concepts are introduced and discussed in this report, some of which

may be unfamiliar to either energy sector or digital experts, respectively. While concepts are explained as they arise, a full list of terms is also available in the glossary at the end of this report.

This report utilises a variety of IEA analytical tools, including both existing and new IEA analysis. It draws on various IEA models, including the World Energy Model, the Energy

Technology Perspectives Model and the Mobility Model (MoMo). The report contains

references to an IEA Central Scenario, which is used in this report to describe the

pathway for energy markets and technological progress based on the continuation of existing energy and climate policies and measures, and to a certain extent announced

commitments and plans. It is broadly in line with the World Energy Outlook New

Policies Scenario (NPS) as well as the 2017 Energy Technology Perspectives Reference Technology Scenario (RTS). The Central Scenario should not be interpreted

Digitalization: A new era in energy?

27

© OECD/IEA, 2017

as a forecast.

References BNEF (Bloomberg New Energy Finance) (2016). Digital Energy Market Outlook. Cisco (2017a). Cisco Visual Networking Index: Forecast and Methodology, 2016-2021. Cisco (2017b). The Zettabyte Era: Trends and Analysis June 2017.

www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-index-vni/vnihyperconnectivity-wp.pdf.

Cisco (2015). The History and Future of Internet Traffic. https://blogs.cisco.com/sp/the-historyand-future-of-internet-traffic.

Fortune (2017). Fortune Global 500 List 2017. http://fortune.com/global500/. Gartner (2017). Gartner Says 8.4 Billion Connected “Things” Will Be in Use in 2017, Up 31 Percent

From 2016. www.gartner.com/newsroom/id/3598917.

IBM (2017). 10 Key Marketing Trends for 2017. IBM Marketing Cloud. www01.ibm.com/common/ssi/cgi-bin/ssialias?htmlfid=WRL12345USEN.

IEA (International Energy Agency) (2017). Energy Access Outlook: From Poverty to Prosperity. IEA (2014). More Data, Less Energy: Making Network Standby More Efficient in Billions of

Connected Devices.

ITU (International Telecommunication Union) (2017). ICT Facts and Figures 2017. www.itu.int/en/ITU-D/Statistics/Pages/facts/default.aspx.

28

Digitalization: A new era in energy?

© OECD/IEA, 2017

MarketsandMarkets (2016). Internet of Things in Utility Market.

Chapter 2

Impact of digitalization on energy demand in transport, buildings and industry Highlights •

Digitalization is having a major impact on transport, buildings and industry. How significant this will turn out to be in the future will differ for each sector and particular application.



Transport is becoming smarter and more connected, improving safety and efficiency. In road transport, connectivity is enabling new mobility sharing services. Combined with advances in vehicle automation and

electrification, digitalization could result in substantial but uncertain energy and emissions impacts. Over the longer term, road transport

energy use could either drop by about half or more than double, depending on the interplay between technology, policy and behaviour.



Digitalization could cut total energy use in residential and commercial

buildings by around 10% to 2040. These efficiency gains are largest in heating and cooling, particularly through the use of smart thermostats

and sensors. Smart lighting allows for potentially substantial cuts in lighting electricity demand. However, new services and comforts brought

about by digitalization – as well as greater use of standby power by idle devices and appliances – could offset potential savings.



Industry has been using digital technologies for a long time to improve

safety and productivity. Digitalization could lead to further significant energy savings with short payback periods through improved process controls

within

industrial

plants

and

beyond

the

plant

fence.

Three-dimensional (3D) printing, machine learning and enhanced connectivity could have an even bigger impact.



Sector-specific and cross-sectoral policies are needed to maximise benefits made possible by digital technologies and to address

Impact of digitalization on energy demand in transport, buildings and industry

29

© OECD/IEA, 2017

challenges, including cybersecurity, data privacy and job losses.

Introduction Digitalization can be a powerful means for increasing efficiency, productivity and

energy savings in transport, buildings and industry. This chapter explores specific

impacts of digitalization within each of these demand sectors. 1 The magnitude of potential impacts – and associated barriers – varies greatly depending on the particular application (Figure 2.1).

Broader systemic implications, including digitalization’s ability to break down

boundaries between various demand and supply sectors leading to even more

potentially transformative impacts, are discussed in Chapter 4. Cross-cutting risks

such as cybersecurity, data privacy and job losses are discussed in Chapter 6.

High

Autonomous cars/trucks

Mobility as a Service

Smart thermostats

User programming Occupancy sensors

Optimised process control

Low

Magnitude of potential change to energy demand

Figure 2.1 Digitalization’s potential impact on transport, buildings, and industry

Additive manufacturing Accelerating innovation in industry

Industrial automation

Daylighting sensors

Big data in aviation

Low

Unmanned shipping

Beyond the plant fence

Transport Buildings

Remote control

Industry

Drone delivery

Barriers to digitalization

High

Key message: Digital technologies and applications face a variety of barriers to adoption and use, and their impacts on energy use differ across demand sectors.

Notes: The digitalization trends/strategies included in this figure are not intended to be exhaustive. “Magnitude of potential change to energy demand” indicates the potential impact of digitalization on energy demand in absolute terms, which may be positive or negative. “Barriers to digitalization” include technological, regulatory and public perception components. The quadrants are illustrative only and intended to give a sense of relative magnitude.

1

Digitalization outside the energy sector, such as e-commerce, e-materialisation (e.g. e-books, DVDs to streaming video), and

teleworking, could also change energy use patterns through a range of efficiency, substitution, and rebound effects. Relevant in quantifying the direct, indirect, structural, and behavioural effects of digitalized products and services.

30

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

examples are discussed in this chapter. See Horner, Shehabi, and Azevedo (2016) for a detailed discussion around the challenges

Transport Cars, trucks, planes, ships, trains and their supporting infrastructure are becoming

“smarter” and more connected, improving safety and efficiency across the transport system. The most revolutionary changes could come in road transport, where

ubiquitous connectivity and automation technologies could fundamentally transform how people and goods are moved. Thus, this chapter focuses mostly on road transport, with passing reference to other transport subsectors.

Deployment of digital technologies in transportation The digitalization of transport essentially describes the concept and development of

“intelligent transportation systems” (ITS). ITS involve: the deployment of sensors for data collection; use of communications technologies to enable remote control; and application of advanced analytics to improve system operations, safety, efficiency and service, as well as to lower costs (US Department of Transportation, 2016).

Everyday examples of ITS include in-road traffic detectors to control traffic lights,

radio-frequency identification (RFID) to automatically collect tolls, and the use of the global positioning system (GPS) and telecommunications for roadside assistance. The three main trends shaping the future of ITS are connectivity, shared mobility and automation.

Smarter and more connected vehicles and infrastructure Smarter and more connected vehicles and infrastructure are collecting and analysing

large volumes of data in order to make individual vehicles as well as the broader system operate more efficiently. Connected vehicles, equipped with internet access,

cellular radio, dedicated short-range communications (DSRC)2 and other means of

sensing and connectivity, are increasingly able to interact with nearby vehicles, infrastructure or both (Li et al., 2012; Narla, 2013).

Across all transport modes, digital technologies are helping to improve energy

efficiency and reduce maintenance costs. In aviation, the latest commercial aircraft are equipped with thousands of sensors, generating almost a terabyte of data on an average flight (Airbus, 2017). Big data analytics optimise route planning and can help

pilots make in-flight decisions and reduce fuel use. Ships are also being equipped with

2

DSRC enables wireless communication using signals in the very high to super high frequency range (75 megahertz to

vehicle-to-infrastructure [V2I]). For more on DSRC, see Dedicated Short Range Communications (DSRC); The Future of Safer Driving, US Department of Transportation, 2016, www.its.dot.gov/factsheets/pdf/JPO-034_DSRC.pdf.

Impact of digitalization on energy demand in transport, buildings and industry

31

© OECD/IEA, 2017

5.9 gigahertz in the United States) for vehicle safety and mobility applications (including vehicle-to-vehicle [V2V] and

more sensors, helping crew take actions to optimise routes, while advances in satellite communications are enabling greater connectivity. Better communication between

ships and ports could optimise ship speed to meet requirements on port arrival timing,

yielding significant fuel savings from slow steaming. In rail, sensors are being used to

monitor everything from engine temperature to vibrations, while cameras collect visual data on the condition of the tracks.

In road transport, real-time information on location and routing can help optimise vehicle and fleet operations. Connectivity also enables the execution of a wide range

of other operations, from checking the status of battery charging for electric vehicles (EVs) to platooning of freight trucks.3 Connected transport systems also foster the

development and adoption of shared mobility services and platforms. Shared mobility

Shared mobility emerged in major cities in the early 2000s with the arrival of

car-sharing services from providers such as ZipCar, whereby members could borrow cars on a short-term (hourly) basis (Clewlow, 2016). Many of these services are now

accessed through smartphone apps, allowing users to locate and unlock vehicles, and

making one-way journeys possible.4 Worldwide membership of car-sharing services

could grow from more than 7 million globally in 2015 to around 36 million by 2025 (Frost and Sullivan, 2016).

Bike-sharing services based on similar technologies and platforms are now available in over 1 000 cities worldwide. Such services have exploded in popularity in the People’s

Republic of China (hereafter, “China”). The number of active users has more than

doubled within only one year (2016) and is expected to top 50 million by the end of

2017. Shanghai alone has an estimated 450 000 shared bikes, compared with 21 000 in Paris. “Dockless” systems (widely deployed in China) and electric bicycle systems (e.g. Copenhagen, Madrid) could further increase the popularity of these services globally.

App-based ride-hailing services are also beginning to have a major effect on urban

mobility. With the widespread adoption of smartphones equipped with GPS and the

availability of digital road maps through APIs (application program interface), 5 the uptake of such services is growing rapidly, including in emerging economies. Didi Chuxing operates in over 400 cities in China, with 20 million rides per day.

3

Platooning refers to the practice of driving vehicles (primarily heavy-duty tractor-trailers or rigid trucks) in a single line with small

gaps between them to reduce drag and thereby save fuel during highway operations. V2V and V2I communication technologies can One-way car-sharing systems like car2go allow users to pick up a vehicle at one location and drop it off at another.

5

An API is a list of commands that allows software programs to communicate with each other and use each other’s functions.

32

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

enable trucks to drive in very close proximity without sacrificing safety or manoeuvrability. 4

Singapore-based GrabTaxi operates in 50 cities across Southeast Asia. With growing

internet access and a large population of young people who are concerned about

pollution and open to ride-sharing, some analysts believe India could soon become one of the largest markets for shared mobility in the world (Morgan Stanley, 2017). App-based ride-hailing services are beginning to apply advanced algorithms to

co-ordinate and match users along routes, making possible truly “shared” mobility services, such as UberPool and LyftLine.

Mobility as a Service (MaaS) platforms aim to simplify the range of shared mobility

services by offering a unified routing and payment platform.6 MaaS platforms allow

users to subscribe to an all-inclusive, multi-modal “mobility package” and access a

variety of shared mobility services,7 including bicycles, buses, trains, cars, taxis and

ride-hailing services. The world’s first such service, Whim, was introduced in Finland in 2016.

Automated driving technologies Automated driving technologies aim to improve safety and driving convenience through

advanced sensing and automated decision-making capabilities that can assist or

replace human control. While recent attention has focused on passenger cars,

automated driving technologies are already widely deployed in the railway business,

primarily on rapid transit and urban lines, and in aviation, where unmanned flight is already common in military applications. The use of drones for delivery services are being tested in some cities (Box 2.1).

Despite a host of technical, regulatory and other policy-related hurdles, prospects for the deployment of connected automated vehicles in road transport are promising due to rapid advances in key technologies8 as well as their large potential cost savings and

safety benefits. 9 Commercial applications, particularly where labour costs are

proportionately high (e.g. buses and ride hailing) or where automation could enable higher vehicle utilisation (e.g. trucks), are well-suited to be targeted first (Wadud, 2017). Major automakers have announced plans to introduce highly automated

6

Various names and acronyms have been applied to describe MaaS. In the United States, the Federal Transit Administration has

adopted the term “Mobility on Demand”: www.transit.dot.gov/research-innovation/mobility-demand-mod-sandbox-program.html. 7

Such platforms could integrate services from both public and private operators, as well as public-private partnerships and other

hybrid business models, relying on data sharing via APIs. 8

Including machine vision and 3D cameras, laser-imaging detection and ranging (LIDAR) and advanced GPS, and artificial

intelligence software. 9

With over 90% of crashes attributable to human error or choice (US Department of Transportation, 2015), an automated,

also improving mobility access for the growing number of seniors, as well as children and the disabled.

Impact of digitalization on energy demand in transport, buildings and industry

33

© OECD/IEA, 2017

connected, electric and shared (ACES) mobility system could significantly reduce the number of road fatalities and injuries, while

passenger vehicles as early as 2020,10 with some experts predicting its widespread

adoption in the period 2025-40 (Arbib and Seba, 2017; Fulton, Mason and Meroux, 2017; Gartner, 2016; Milakis et al., 2015; Wadud, 2017).

Box 2.1

Drones

The use of unmanned aerial vehicles, or drones, for civilian and commercial purposes is at an early

stage. Delivery by drones is being tested in cities in China, India and California, but prospects for mass deployment are uncertain. Capacity limitations are expected to restrict the use of drones as

substitutes of road vehicles to deliver the massive volume of commodities circulating throughout cities and the economy (IEA, 2017a). Under reasonable assumptions for loading, unloading and flight times, it would take about 15 drones operating around the clock to deliver the same number of

products as a single light commercial vehicle does in a typical eight-hour shift (McKinnon, 2015). In addition, the per-kilometre costs of drone delivery are likely to remain higher than those of conventional trucks in the medium term.

Life-cycle analysis shows that the potential energy and emissions advantages of drones are likely to

be limited and highly context dependent (Goodchild and Toy, 2017). Safety, security, liability and noise considerations are further barriers to the widespread use of drones, and their viability in goods delivery markets is likely to depend on the regulatory context. Nonetheless, drones have proved their

utility in a variety of specific applications, such as delivering high-value goods in cities, vaccines to remote communities, or essential supplies for disaster relief. In the energy supply sector, drones carrying cameras or multiple sensors (e.g. GPS, radar, sonar)

have an array of applications such as performing inspections of corrosion on oil rigs and wind

turbine blades. They can also conduct surveys of environmental conditions and wildlife at potential offshore wind sites, which may be difficult or expensive to reach by other means. Drones equipped with thermal imaging cameras have been used to conduct aerial thermography to detect power

losses and failure risks in solar photovoltaic plants. Specially equipped drones can also perform

basic maintenance of electricity grids; for instance, a flame-throwing drone can burn off rubbish entangled around wires.* * “Flame-throwing drone removes net entangled in China power line”: https://youtu.be/iOqWfLZT8OM.

The investment and financial implications of the roll-out of connected automated

vehicles and mobility services are enormous. With the market for automated driving technologies expected to grow to almost USD 200 billion by 2030 (Archambault et al.,

10

Here “highly automated” refers to automated monitoring and control of the vehicle, corresponding to Levels 4 and 5 of the

hierarchy defined by the Society of Automotive Engineers (SAE) International Standard J3016 (SAE International, 2016). The SAE Standard defines six levels of automation, from no automation (Level 0) to full automation (Level 5). A key distinction of levels occurs between Levels 2 and 3, depending on whether a human (Levels 0-2) or car (Levels 3-5) monitors the environment. (although Level 4 allows “driverless” mode only under certain conditions, e.g. highway).

34

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

Vehicles with high or full automation (Levels 4-5) are commonly referred to as “autonomous”, “driverless”, or “self-driving”

2015), 11 automakers, technology companies and app-based ride-hailing service providers are all vying to control the key technologies that will underpin the market,

pouring billions into research, strategic partnerships, and acquisitions.12 High-profile

legal battles (e.g. Waymo v. Uber) over proprietary automated driving technologies have ensued (Abuelsamid, Alexander and Jerram, 2017). Tesla, whose vehicles are

now equipped with the necessary hardware for fully automated driving (Tesla, 2016), has announced plans to release an electric truck with automated capabilities. Tesla’s market capitalisation recently surpassed that of Ford, GM and, most recently, BMW,

despite selling fewer than 80 000 vehicles and spending USD 770 million more than it earned in 2016 (Mitchell, 2017).

Implications for energy use and emissions: Focus on road transport Transport currently accounts for 28% of global final energy demand and 23% of global

CO2 emissions from fuel combustion (IEA, 2017b). In the IEA Central Scenario,13 final energy consumption for transport grows by almost half to 165 exajoules in 2060, with most of the demand coming from road freight vehicles (36%) and passenger light-duty

vehicles (28%). The dynamics and net effects of automated, connected, electric and shared (ACES) mobility will play a key role in shaping the future energy and emissions trajectory of the overall transport sector.

Digitalization can affect road transport energy demand in many ways. Highly

automated vehicles reduce driver stress and allow for more productive use of travel time, making private car travel more attractive. Automation will also make road freight

transport cheaper. Both of these factors could encourage more travel activity, resulting in increased congestion and energy demand. On the other hand, shared and autonomous transport could facilitate vehicle right-sizing

11

14

and accelerate EV

This market refers to “sales of autonomous vehicle components for road vehicles”. By comparison, this figure is slightly less than

Toyota’s consolidated global net revenues in fiscal 2016, which totalled about USD 240 billion (Toyota, 2016). 12

Recent acquisitions of automated driving companies include Intel (Mobileye, USD 15.3 billion), Ford (ArgoAI, USD 1 billion), GM

(Cruise Automation, 1 billion) and Uber (Otto, USD 680 million). Major investments in shared mobility platforms include Apple (Didi Chuxing, USD 1 billion), GM (Lyft, USD 500 million) and Volkswagen (Gett, USD 300 million). For a more complete listing of

automobile manufacturer investments in connected and automated vehicle technologies and shared mobility businesses, see Table 3 of Slowik and Kamakaté (2017). 13

As defined in Chapter 1, the Central Scenario describes the pathway for energy markets and technological progress based on the

continuation of existing energy and climate policies and measures, and to a certain extent announced commitments and plans. It is broadly in line with the World Energy Outlook New Policies Scenario (NPS) as well as the 2017 Energy Technology Perspectives

Reference Technology Scenario (RTS). This should not be interpreted as a forecast.

Increased use of shared and autonomous transport could help optimise the size of vehicles to match utilisation patterns,

resulting in improved energy efficiency.

Impact of digitalization on energy demand in transport, buildings and industry

35

© OECD/IEA, 2017

14

adoption, reducing energy use and emissions.15 High utilisation rates of automated and

shared vehicles spurring more rapid vehicle (and fleet) turnover could favour and accelerate the uptake of highly efficient technologies including EVs, reducing the emissions intensity of travel (Johnson and Walker, 2016).

The successful integration of shared and automated mobility services with mass public transit, walking and cycling could also help to reduce energy use.16 For instance, feeder

services provided by shared and automated vehicles could foster the use of

high-capacity public transport. In cities with high population density and good public transport networks, digitalization could contribute to a shift away from the traditional

paradigm of vehicle ownership towards the provision of MaaS. The challenge for local authorities is to find the right mix of fiscal instruments and transport and urban planning to ensure that autonomous and shared vehicles strengthen, rather than take over, public transport.

The consequences of ACES mobility for energy and emissions are highly uncertain.

They will depend on the combined effect of changes in consumer behaviour, policy intervention, technological progress and vehicle technology. Analyses of extreme

scenarios developed for the US Department of Energy find that automated vehicles in some cases could reduce fuel consumption by more than 90% or increase it threefold under a “perfect storm” of inefficient behavioural response compared with central projections (Brown, Gonder and Repac, 2014).

Several recent studies consider a range of ACES mobility development scenarios, most

of which focus on passenger vehicles (Brown, Gonder and Repac, 2014; Fulton, Mason

and Meroux, 2017; Greenblatt and Saxena, 2015; Stephens et al., 2016; Wadud, MacKenzie, and Leiby, 2016). Two of these studies, in particular, demonstrate the range of implications of possible developments in vehicle autonomy, sharing and electrification for energy use and emissions:

• Wadud, MacKenzie and Leiby (2016) examine the mechanisms by which

connected and automated vehicles may lead to a wide range of outcomes with

respect to energy use. The study explores a range of possible technological and

societal developments in the United States, and finds that road transport

15

The coupling of connectivity and automation with EVs could help diversify energy use, reduce pollutant emissions and, if the

electrical grid is progressively decarbonised, also result in lower greenhouse gas emissions. 16

Public transport authorities and transit agencies across several municipalities in the United States have integrated app-based

ride sharing into their portfolio of services, promoting and in some cases even subsiding Uber, Lyft and conventional taxi rides that connect to and from public transit stations, or that serve underprivileged populations (Slowik and Kamakaté, 2017). Nevertheless, early evidence from New York City suggests that Uber and Lyft may be displacing mobility that would otherwise have taken place Chicago and Los Angeles (Clewlow and Mishra, 2017).

36

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

on public transport (Schaller, 2017). Similar results were found in a recent survey in seven major US cities including New York,

energy demand could be reduced by almost half under an optimistic scenario, or more than double in a pessimistic scenario (Figure 2.2). The study also finds

that even at low levels of automation, prioritising system-level connectivity and

co-ordination among vehicles over individual vehicle automation could result in significant gains in energy efficiency.

• Fulton, Mason and Meroux (2017) explore the global energy and carbon dioxide

(CO2) emissions implications of automation, electrification and shared mobility.

The findings suggest that technological innovation in electrification and

automation alone may result in increased greenhouse gas emissions. Sharing of

trips and vehicles, MaaS and strong support for public transit are needed to steer the transformation of urban mobility in a way that leads to lower emissions.

Figure 2.2 Range of possible energy impacts from vehicle automation in the United States I) Optimistic scenario: "Have our cake and eat it too"

II) Pessimistic scenario: "Dystopian nightmare" Heavy-duty vehicles

Energy intensity

Light-duty vehicles

Travel demand Energy demand Total road transport energy -80%

-40%

0%

40%

80%

120%

-80%

-40%

0%

40%

80%

120%

Key message: US road transport energy demand is reduced by almost half under an optimistic scenario, but more than doubles in a pessimistic scenario.

I) In this scenario, virtually all of the potential benefits of automation are realised, including smoother traffic flow, fewer accidents (vehicles become smaller/lighter) and the widespread adoption of eco-driving and platooning.

II) In this scenario, broad adoption of higher levels of automation is coupled with strong increases in demand for travel, and no efficiency benefits materialise (e.g. no platooning or eco-driving). Note: This figure only shows two of the four scenarios from the study. One of the other scenarios describes a world with limited automation and minimal impact on energy use, while the other assumes large improvement in fuel efficiency as well as increased travel demand, resulting in a moderate net reduction in energy use.

Impact of digitalization on energy demand in transport, buildings and industry

37

© OECD/IEA, 2017

Source: Wadud, MacKenzie and Leiby (2016), “Help or hindrance? The travel, energy and carbon impacts of highly automated vehicles”.

The IEA recently considered the energy and CO2 emissions implications of

digitalization in freight transport (IEA, 2017a). In a Modern Truck Scenario, digitally

enabled changes17 to improve supply chain and logistics operations contribute about

one-third of the cumulative greenhouse gas reduction potential in the road freight

sector to 2050 relative to the IEA Central Scenario (Figure 2.3). Such efforts could culminate in new paradigms for goods delivery, such as the “physical internet”, 18

reducing the number of trucks and traffic needed to deliver goods. Connectivity and automation can improve the operational efficiency of road haulage through

eco-driving, shifting operations to less congested times and routes, and platooning.

Finally, smarter charging of plug-in electric trucks can both improve the economics of electrification and reduce the carbon intensity of electricity sourced.19

Figure 2.3 Digitalization’s impact on energy use and emissions reductions in road freight 0%

Energy demand

GHG emissions Systemic measures

-10%

Vehicle efficiency

Net change

-20%

Fuel switching

-30% -40%

Approximate contribution of digital technologies

-50% -60% -70% -80%

Key message: Digital technologies can play a key role in reducing energy demand and greenhouse

gas emissions in road freight, notably by facilitating horizontal and vertical collaborations among companies to streamline supply chains and logistics.

Notes: Greenhouse gas emissions are measured on a well-to-wheels basis; systemic measures include platooning, the retiming of urban deliveries and driver training, among others; fuel switching includes both the use of advanced biofuels and the adoption of alternative powertrains running on low-carbon energy carriers, such as electricity and hydrogen; GHG = greenhouse gas. Source: IEA (2017a), The Future of Trucks – Implications for Energy and the Environment.

17

Examples include: GPS coupled with real-time traffic information for route optimisation; on-board monitoring and feedback that

enhances eco-driving performance; vehicle connectivity that can safely reduce gaps between platooning trucks to improve fuel

efficiency; and data sharing between companies across the supply chain to ship more goods with fewer trips. 18

The “physical internet” describes an open, shared global logistics system inspired by the movement of data on the internet, in

contrast with the proprietary logistics systems that are common today. Currently, nearly all logistics service providers and carriers

maintain proprietary assets, both physical (e.g. warehouses and trucks) and operational (e.g. information on routes, customers and markets) (Wible, Mervis and Wigginton, 2014). 19

For example, charging at times when renewable electricity generation exceeds demand, supplying power to the grid during peak

and the grid as a result of digitalization.

38

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

demand times and providing other ancillary grid services. See Chapter 4 for a discussion on the potential interactions between EVs

Energy use and emissions in the road transport sector and other transport modes

could also be affected by digitalization beyond the sector itself.20 For instance, lower

transport demand as a result of more teleworking could reduce travel and energy

demand, although empirical studies suggest that the actual impact on mobility and energy use are mixed. In fact, teleworking may result in an overall increase in energy

use in many contexts, for instance, by increasing energy use at home and changing overall travel decisions (Choo, Mokhtarian and Salomon, 2005; Kim, Choo and Mokhtarian, 2015; Larson and Zhao, 2017; Melo and de Abreu e Silva, 2017; Ory and Mokhtarian, 2006).

Additive manufacturing, commonly known as 3D printing, could reduce demand for long-distance shipping of goods. 3D-printed components could also lead to lighter

aircraft (Huang et al., 2016) and improve turbine efficiency, thereby reducing in-flight fuel use (see Box 2.3). Digitalization could also pave the way for a variety of systemic

cross-sectoral interactions and opportunities. For instance, digitally-enabled smart

charging and vehicle-to-grid (V2G) technologies could play an important role in integrating higher shares of variable renewables into electricity supply (see Chapter 4).

Barriers to digitalization and policy considerations Government policies and regulations will play an important role in the deployment of digital technologies in transport, as well as steering developments towards lower energy and emissions pathways. Addressing barriers and risks As transport becomes increasingly digitalized, questions about vehicle and software

certification, liability, cybersecurity,21 data privacy, and employment (see Box 6.5) will

need to be addressed (Hern, 2017).

Connected and automated vehicles, in particular, face a variety of technical, regulatory

and other 22 hurdles that could impede their widespread deployment. Real-world

testing, learning and data collection are vital to improving technology and instilling public confidence in automated driving technologies. Regulators should consider

20

For example, the move from DVDs to streaming video could cut energy use from DVD production and transport, while increasing

energy use by data centres and networks. The convenience of such e-services could also result in rebound effects that encourage greater use of these services, resulting in higher energy use overall. See Horner, Shehabi, and Azevedo (2016) for other examples. 21

Ensuring cybersecurity becomes more difficult once vehicles are connected to each other (V2V) and to infrastructure (V2I)

(Perlroth, 2017).

For example, ethical and moral considerations. The German government recently released a report on ethical aspects of

automated cars (BMVi, 2017).

Impact of digitalization on energy demand in transport, buildings and industry

39

© OECD/IEA, 2017

22

getting rid of outdated regulations and introduce new ones to encourage competition and facilitate trials.

Harmonising and standardising communications and data protocols will also be

important in permitting cross-border travel of connected automated vehicles. For example, the European Union is working with member states, automakers, telecom companies

and

interoperability.

other

industrial

stakeholders

toward

legislation

to

ensure

23

Public policy can also help promote innovative mobility services and foster

multi-modality 24 by establishing platforms for sharing and maintaining high-quality databases to be easily queried and integrated with APIs (Shaheen et al., 2016).25 Such

publicly accessible databases would need to be well maintained to ensure the data are accurate and the data made available in such a way as to ensure personal privacy.

In road freight, sharing of data, assets and services across the supply chain can enable

dramatic improvements in freight logistics (IEA, 2017a). Governments can also promote

the

reporting

of

aggregated

information

(such

as

vehicle-

and

tonne-kilometres taking place in a given area and at a given time) making it possible for authorities to monitor the impact of public policies without compromising confidentiality. All datasets need to meet high standards of cybersecurity (see Chapter 6).

Toward smart and sustainable mobility Policy can help steer developments in automated and connected mobility towards lower energy use and emissions. For example, the gradual introduction of distance-26

and congestion-based pricing could moderate potential rebound effects27 stemming from high levels of automation. Other measures could help to direct consumer interest

23

For more on the European Commission’s priorities in shaping legislation on “co-operative, connected and automated mobility”

under the framework of its Digital Single Market strategy, see: https://ec.europa.eu/digital-single-market/en/cooperativeconnected-and-automated-mobility-europe.

24

“Multimodality” refers to seamless trip-making across multiple transport modes (e.g. walking, cycling, bus, urban rail and cars).

25

For example, the approach being pursued under the Finnish Transport Code is to establish public entities as repositories of

mobility databases and data exchange hubs (Ministry of Transport and Communications, 2016), making data easily accessible to app developers and to the public at large. Shared mobility operators in Washington, DC, Boston and San Francisco have shared data with public agencies, either voluntarily or by mandate (Shaheen, Cohen and Zohdy, 2016). Collaboration between Rio de

Janeiro’s public transport app Moovit and Google Waze resulted in the integration of real-time road traffic and public transport data (Olson, 2014).

26

Distance-based pricing should be differentiated on the basis of vehicles’ fuel efficiency and/or pollutant emissions performance.

27

Rebound effects occur when energy savings resulting from improved efficiency (resulting in lower energy costs) can stimulate

of automation could improve energy efficiency, but could also result in greater overall driving activity due to reduced costs.

40

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

increased energy consumption and general expenditures that counteract the technical potential savings. For example, high levels

and spending towards shared mobility to curb energy use and emissions. This might

be especially effective in markets with rising car ownership, such as China, India and other emerging economies. In road freight, global standardisation of truck sizes and pallets could boost fuel efficiency in the same way that the adoption of standardised containers enabled vast operational efficiency gains in maritime freight.

Policy can also adapt to leverage the synergies between sharing, automation, and electrification as a means to accelerate EV adoption. The European Commission’s

“Europe on the Move” strategy seeks to overhaul existing legislation and set up new

initiatives to promote “clean, competitive and connected mobility” (European

Commission, 2017). App-based ride-hailing service providers such as Lyft have already

set goals to provide services using only electric, autonomous vehicles powered by 100% renewable energy (Zimmer, 2016; Zimmer and Green, 2017).

Buildings Digitalization of the provision of energy services There is significant potential for digitalization to improve energy services and user comfort in buildings, while also reducing overall energy use. Smart energy management can:

• Help ensure that energy is consumed when and where it is needed, by improving the responsiveness of energy services (e.g. by using lighting sensors) and predictively with respect to user behaviour (e.g. through learning algorithms that auto-programme heating and cooling services).

• Enable demand response to reduce peak loads (e.g. shifting the time of use of a washing machine), to shed loads (e.g. adjusting temperature settings to lower

energy demand at a particular time) and to store energy (e.g. in thermal smart

grids) in response to real-time energy prices or other conditions specified by the user.

• Predict, measure and monitor in real time the energy performance of buildings, allowing

consumers,

building

managers,

network

operators

and

other

stakeholders to identify where and when maintenance is needed, when investments are not performing as expected or where energy savings can be achieved.

The energy loads of a building can be displaced or turned on and off using active control systems that collect, process and adapt to real-time data using sensors, and app). Active controls can also integrate and intelligently link building energy services with information from the grid, allowing for better management of supply and demand. Impact of digitalization on energy demand in transport, buildings and industry

41

© OECD/IEA, 2017

which can be managed using a single front-end dashboard (e.g. smart phone or tablet

Such integration, paired with algorithms and market information, can help to drive down transaction costs between electricity generators and consumers. In retail markets, aggregating distributed energy loads across multiple end-users enable

demand shaping, both reactively (e.g. using price signals) or proactively (e.g. using input on consumer preferences) (see Chapter 5).

Energy implications of digitalization in buildings Buildings account for nearly one-third of global final energy consumption and 55% of

global electricity demand. Electricity demand growth in buildings has been particularly rapid over the last 25 years, accounting for nearly 60% of total growth in global

electricity consumption. In some rapidly emerging economies, including China and

India, electricity demand in buildings grew on average by more than 8% per year over the last decade. In the IEA Central Scenario, electricity use in buildings is set to nearly

double from 11 petawatt hours (PWh) in 2014 to around 20 PWh in 2040, requiring large increases in power-generation and network capacity.

Improving the operational efficiency of buildings by using real-time data could lower

total energy consumption between 2017 and 2040 by as much as 10% compared with the Central Scenario, assuming limited rebound effects in consumer energy demand

(Figure 2.4). Cumulative energy savings over the period to 2040 would amount to 65 PWh – equal to the total final energy consumed in non-OECD countries in 2015. Figure 2.4 Cumulative energy savings in buildings from widespread digitalization 70

PWh

Non-residential Residential

60 50

Others

40

Appliances

30

Lighting

20

Water heating

10 0

Space cooling United States

China

India

EU

ASEAN

Other

World

Space heating

Key message: The widespread deployment of active controls, assuming limited rebound effects, would save up to 65 PWh cumulatively to 2040, or twice the energy consumed by the entire buildings sector in 2017.

42

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

Notes: Non-residential includes buildings used for wholesale and retail trade, storage, education, health, hospitality and commercial activities, as well as offices and public buildings; EU = European Union; ASEAN = Association of Southeast Asia Nations.

The largest potential savings are in heating, cooling and lighting, which together

represented more than 60% of total final energy demand in buildings in 2015. For

instance, smart thermostats can improve management of heating and cooling loads, allowing for improved and even remote control of temperatures throughout a building. This can ensure thermal comfort when and where it is needed while also maintaining

and increasing energy savings when heating and cooling are not needed. Increasingly, learning algorithms in smart thermostats are taking this one step further, automatically pre-heating or pre-cooling a building space relative to expected occupant presence, user preferences, forecast weather conditions and other input information (e.g. energy prices).

Smart controls and connected devices, including even simple occupancy and photo sensors, consume energy in order to maintain connectivity, even when on standby (see Chapter 5 for a discussion on the direct use of energy by digital technologies, including

connected devices). Standby power in buildings applications alone can be significant.

For example, it ranges from 0.15 watts (W) to over 2.71 W for smart lighting, potentially

consuming as much as 25 kilowatt hours (kWh) per light fixture or bulb if running 24

hours per day year-round (Kofod, 2016). Consequently, some connected lamps can

consume more energy per year in standby mode than when actually in use, cutting their net energy efficiency by more than half.

The potential for energy demand growth from connected devices in buildings, whether

they are smart or not, is already noticeable in many markets. For instance, the number

of household appliances that are connected to communication networks, ranging from televisions and washing machines to doorbells and security cameras, has taken off in recent years, with their overall energy use amounting to an estimated 360 terawatt hours (TWh), or 13% of total household electricity consumption for appliances and

other small plug loads in 2015 (Figure 2.5). In the IEA Central Scenario, 50% of

household electricity demand for appliances by 2040 is expected to come from connected devices, presenting opportunities for smart demand response28 but also

increasing the need for standby power.

Economies of scale and continued product improvement are expected to halve the

energy intensity of active control devices over the next 25 years, from an average of around 2 kWh per square metre (m2) per year in 2010 to 1 kWh/m2/year in 2040

(Figure 2.6). Globally, active controls are projected to consume 275 TWh in 2040 –

accounting for 0.7% of buildings energy use in the IEA Central Scenario. This is far less than the 4 650 TWh potentially saved by those smart controls in 2040 (IEA, 2017b).

Smart demand response refers to the possibility for consumers, smart controls and connected devices to adjust electricity

consumption in response to time-based rates or other forms of financial incentive.

Impact of digitalization on energy demand in transport, buildings and industry

43

© OECD/IEA, 2017

28

Figure 2.5 Household electricity consumption of appliances and other small plug loads 6 000

TWh

5 000 4 000 3 000 2 000

Network-enabled

1 000 0

Not connected 2010

2015

2020

2025

2030

2035

2040

Key message: The share of connected “network-enabled” appliances in total household electricity

consumption is set to grow rapidly, presenting opportunities for smart demand response but also increasing the need for standby power consumption.

Note: In the case of televisions, only “smart TVs” are considered to be network-enabled; those with connections to cable television or other broadcasting networks are not included.

300

2.4

250

2.0

200

1.6

150

1.2

100

0.8

50

0.4

0

2010

2015

2020

2025

2030

2035

2040

Annual energy intensity (kWh/m2/year)

Energy consumed (TWh/year)

Figure 2.6 Global energy use and average energy intensity of active controls in buildings

Sensors Actuators and displays Central controllers Controllers

Energy intensity

0.0

Key message: The energy intensity of active controls is expected to improve to 2040, consuming far less than the potential energy savings achieved by smart controls in buildings.

Notes: Sensors include occupancy and daylighting sensors; central controllers collect data from an entire dwelling or building and monitor operating units; controllers only operate on a specific zone of the building.

In addition to the energy efficiency savings, connected devices and active controls in

buildings can offer greater comfort to consumers and bring important social and health-related benefits. For instance, around 11% of the population in the European Union is not able to heat their households adequately at an affordable cost (Pye and

energy poverty (IEA, 2017c). Optimisation of heating and cooling loads through active controls (with appropriate financing tools and support schemes for technology 44

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

Dobbins, 2015). As many as 15% of households in IEA member countries suffer from

deployment) would help to reduce buildings energy demand, allowing healthier

temperatures when and where occupants are present and reducing overall household spending on heating and cooling. Smart thermostats Smart thermostats are programmable connected devices that can help households monitor and regulate their heating and cooling loads. Potential savings range between

15% and 50% depending on the building and control technology (Grözinger et al., 2017). Smart thermostats can also automatically learn, enabling improved energy demand management in a building (or even by zone or room) using information on presence,

occupant routine and even weather forecasts through online data. Better prediction of

heating and cooling loads can enable improved equipment and/or network operation (e.g. optimised compressor working cycles and greater utilisation of heat exchangers).

At the same time, smart thermostats can enable utilities to deploy demand-side

response in buildings, where peak heating and cooling loads can have significant impact on the grid.

In the United States, purchase of smart thermostats doubled from 3% of total residential thermostat sales in 2014 to 6% in 2016 (Wilczynski, 2017). Globally, revenue

from the sale of connected smart thermostats and their software and services is projected to increase from USD 1.1 billion in 2016 to USD 4.4 billion in 2025 (including

USD 1.6 billion in North America, USD 1.3 billion in Asia-Pacific and USD 1.1 billion in Europe) (Navigant Research, 2016).

Major heating and cooling equipment providers, such as Johnson Controls, Danfoss and Honeywell, are offering products that can be scheduled and controlled remotely by smartphone apps or Bluetooth. Nest Labs, one of several newcomers to the market,

offers Wi-Fi-enabled thermostats that use self-learning algorithms and sensors to

improve heating, cooling and hot-water management in households. Based on field

studies, Nest claims its thermostats achieve average energy savings of 10-12% for

heating and 15% for cooling (Nest, 2015).

Some utility companies, including British Gas, Engie and Scottish and Southern Energy, are also offering their own smart thermostats. The entry of these non-traditional product suppliers into the smart thermostat market reflects a growing business

opportunity, which can also help to secure competitive advantage for utility providers offering better controls to customers. Encouraging the use of smart thermostats could also help to reduce their need to invest in additional power generation capacity.

Certain utilities in the United States are exploiting the potential for demand-side

temperature settings during peak heating or cooling periods (National Grid, 2017).

Data collected by smart devices could similarly be used by utilities to identify Impact of digitalization on energy demand in transport, buildings and industry

45

© OECD/IEA, 2017

management that smart thermostats present by rewarding customers who adjust their

opportunities for launching energy-efficiency programmes or even to offer new services (e.g. through an energy service company [ESCO]), driving up spending on energy-efficiency

improvements

such

as

insulation

or

the

replacement

of

heating/cooling appliances with more efficient ones. Other business models, such as WattTime, incorporate software in certain smart thermostats that adjusts the amount

of electric heating and cooling (e.g. on-off cycles of air conditioners) according to the availability of renewable energy using real-time data from power grid operators.

Governments are starting to play an important advisory role in the roll-out of smart

thermostats. For example, the United States Environmental Protection Agency is

endorsing devices that demonstrate verified minimum savings for heating and cooling through the Energy Star Program, with seven thermostats recognised to date (Energy Star, 2017). Smart lighting The global lighting market is changing rapidly with the growing adoption of

light-emitting diode (LED) technology. LED sales reached around 30% of the global

residential market alone in 2016 (IEA, 2017b). This shift could reshape lighting services

and energy demand, particularly as the luminous efficacy of LED technology

(measured in lumens per Watt [lm/W]) continues to improve. New LED products

coming onto the market are now able to produce 150 lm/W (compared with a standard

compact fluorescent lamp at 50-70 lm/W) for residential use and up to 200 lm/W for

commercial applications.

Smart lighting (e.g. high-performance LEDs connected to building control systems)

can process user preferences along with information such as daylight conditions and

building occupancy. This offers an opportunity to provide greater lighting quality and building energy services while also achieving substantial energy savings. The IEA estimates that smart lighting could save nearly 8 PWh of cumulative electricity demand

between 2017 and 2040 compared with the IEA Central Scenario, or 14% of total final lighting energy consumption during that period, in addition to the savings already

brought by expected shifts to LEDs. More than three-quarters of the additional savings from smart lighting come from commercial buildings (Figure 2.7).

Smart lighting in buildings can also help reduce energy consumption for other end

uses, such as heating and cooling, while also helping to increase the quality and

value-added of other building services. For instance, LEDs equipped with sensors can

directly interact with a building’s energy management system, improving information on heating and cooling needs relative to occupancy, activity or daylight. Recent LED technologies can also be directly integrated into the building’s ethernet, turning that point to a specific room or position within a building) (Box 2.2).

46

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

lighting into a building services and information tool (e.g. using location-based data

Figure 2.7 Potential electricity savings from smart lighting in buildings to 2040 2 800

Cumulative to 2040

Annual

TWh

3%

2 400

11%

2 000 1 600

51%

1 200

Savings

800

Non-residential

400 0

35%

Residential 2015

2020

2025

2030

2035

2040

Key message: Smart lighting could cut global electricity needs for lighting in buildings by more than 20% in 2040.

Note: The electricity savings potential from smart lighting, including sensors and digital controls, is calculated as the energy use reductions beyond the estimated savings from expected deployment of LED lighting already estimated in the IEA Central Scenario.

Box 2.2

LEDs as a value-added tool for energy services

Recent developments in LED technologies are rapidly reshaping the lighting market as a potential tool for value-added services in buildings. For instance, French retail giant Carrefour and the United

Arab Emirates mall Aswaaq are using LEDs to make shopping easier for customers using

light-positioning technology (Philips, 2015). Light fidelity (or Li-Fi) modulates light at frequencies

imperceptible to the human eye, allowing a unique digital signal from an LED. This signal, captured by a smart phone, can be used for product searches or store mapping to provide location-based information for shoppers. Light-positioning data can also be used to provide energy services tailored to the needs of

individuals, for example, by allowing individualised heating, cooling or light settings for a particular space in an office building. The Edge building in Amsterdam, which incorporates “Power-overEthernet” (PoE, using a single cable to provide both data connection and electric power) and LED

technology, is one example of this type of service. Integrated light sensors connected to the building’s information technology network provide information on occupancy patterns and energy usage, which can help the building manager to provide services more efficiently and effectively, such as by optimising cleaning schedules, use of space and ventilation. LED lighting is also adding value beyond in buildings, such as in street lighting. For instance, Energy Efficiency Services Limited (EESL) has installed over 3.5 million LED street lights in India. Those lights, beyond the energy efficiency improvements and providing better visibility for drivers and

improved safety for pedestrians, also communicate real-time individual data to help the electricity

Impact of digitalization on energy demand in transport, buildings and industry

47

© OECD/IEA, 2017

network monitor and control performance, address problems and ensure utility service delivery to the public.

While more efficient LEDs and smarter lighting services are expected to lead to overall energy savings, there may be some rebound effects, whereby the lower cost and

improved quality of lighting increase demand for it. This may be particularly the case in the residential sector, where users will be offered a multitude of new lighting options

ranging from customised ambiances to fully automated systems orientated more

toward user control than energy management. By contrast, smart lighting services in commercial buildings appear to have much greater potential for saving energy, as

lighting services tend to be more clear-cut (e.g. basic illumination of work spaces and meeting rooms) and there is a stronger incentive to lower operating and maintenance

costs. Similarly, smart lighting in public places, notably street lighting, may also cut energy use beyond the direct energy savings from the use of LED lamps, for example

by connecting streetlights to traffic lights and other traffic management tools. LEDs for public lighting could also improve local safety.

Barriers to further digitalization and policy considerations Many obstacles lie in the path to realising the benefits of widespread digitalization in

buildings. These range from concerns about privacy to technical and economic considerations, all of which affect the choice of whether to install and use sensors and

devices that provide data and smart control to the building management system. There

may be a case for utilities to offer financial incentives and introduce innovative tariff

schemes to encourage building owners and occupants to adopt digital technologies, given the potential network cost savings from optimised energy use in buildings.

Greater effort is also needed to communicate the benefits of digitalization to end users in the form of improved comfort and cost savings.

Standards for connected devices will be crucial to the prospects for digitalizing

buildings. Policy makers and companies need to ensure that devices are able to

provide and receive information using open source or compatible software to allow for interoperability across technologies. Common technical standards for connected

devices will help ensure their interoperability at different levels (e.g. with other devices, with building management systems and with the grid). Standards could also help ensure user friendliness and product operability; design, interface and ergonomics can

influence how well the devices are used and the amount of energy that can be saved as a result.

New business models for energy services could also help overcome barriers to

digitalization in buildings. Large-scale deployment of smart building energy

management tools and devices will call for a different approach to the provision of energy. Traditionally, building owners, operators and occupants purchased appliances

such as heat, in the same way that they purchased energy from a utility provider, typically through volume-based contracting. In the digital future, as building energy 48

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

and equipment such as boilers and light bulbs to provide a specific energy service,

systems become more complex (and perhaps more expensive to install), new business models that provide a given set of energy services rather than an amount of energy

may make more sense. Those new business models could also allow for broader

energy efficiency delivery, perhaps requiring from the service provider (often easier to regulate) a minimum energy performance of the building.

The emergence of ESCOs or similar business models could also create opportunities for the provision of comprehensive energy packages, such as smart controls combined

with heat pumping technologies and appropriate building renovation measures, aimed

at delivering energy savings across a range of devices and end uses. Supportive policy frameworks, such as bulk procurement of energy-efficient technologies and white

certificates,29 can help in this regard by driving down costs of products and ensuring

those technologies actually deliver on savings.

Industry Industry has a long history of using digital technologies, originally to improve safety and increase production through automation. Additional benefits include less

downtime, lower operating costs, reduced energy consumption and better product quality. This section focuses on manufacturing, construction and minerals extraction industries. Although industry contains many different subsectors, processes and outputs, many of the benefits from digitalization are similar. For example, increased

data collection and analysis to optimise production process, to improve energy efficiency and to reduce waste apply to all production processes. Nonetheless, the

usefulness of some technologies, such as 3D printing, where objects are created

through a computer-controlled system by depositing consecutive layers of material, is greater in certain sub-sectors.

In 2014, industry was responsible for 38% of global final energy consumption and 24%

of total CO2 emissions (IEA, 2017a). With the expected continuing expansion of

industrial production over the coming decades, particularly in emerging economies, the

value of digitalization in improving the efficiency of energy and material use will only increase. While it is expected that digitalization in industry will continue in an

incremental manner in the near term, some digital technologies may have far-reaching

effects on energy use in certain areas, especially when they are applied in combination.

The degree to which a particular company or sector embraces or is affected by

digitalization depends on several factors: the complexity of the production process (i.e. A tradeable instrument issued by an authorised body guaranteeing that a specified amount of energy savings has been achieved,

Impact of digitalization on energy demand in transport, buildings and industry

49

© OECD/IEA, 2017

29

usually combined with an obligation on a utility to achieve a certain overall amount of energy savings.

the extent to which the process can be automated and how sensitive the process is to changes in operating variables); the financial capacity of the firm; its exposure to volatile energy prices; market competition; and the flexibility of its supply chains.

Cultural factors, particularly the willingness of a firm to take on greater levels of risk by embracing a new technology or to change established operating practices, also influence the extent and pace of digitalization.

The human impact of digitalization, particularly the potential for digital technologies to

reduce the number of employees required to operate and maintain a production

process, is another factor which may limit its uptake. In addition, cybersecurity and

data privacy concerns may also inhibit extensive digitalization in at least some industries (see Chapter 6).

The impact of digitalization on industry can be separated into the changes that take place within a particular plant and those that have implications outside the plant, or

“beyond the plant fence” (Figure 2.8). These are each explored in turn. Figure 2.8 Application of digital technologies and strategies in industry Connected industrial equipment

Beyond the plant fence

Connected supply value chains

Cloud-connected workers Remote-controlled operations

Digital twin

Inside the plant

Artificial intelligence

Decision support from advanced data analytics

Advanced process control systems

Industrial robotics

Smart sensors Low

3D printing

Readiness level for wide adoption Industrial equipment

Analytics and enabled workforce

High

Connectivity

Key message: Digitalization in industry can take diverse forms, ranging from automated equipment to connecting industrial operations based in different locations.

Notes: A digital twin results from virtually replicating a real industrial plant; enabled workforce refers to the enhancement Sources: Based on World Economic Forum (2017a), Digital Transformation Initiative: Chemistry and Advanced Materials Industry, and World Economic Forum (2017b), Digital Transformation Initiative: Mining and Metals Industry.

50

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

of workers’ skills through the use of digital technologies.

Digitalization within industrial plants The extent to which industrial process control is already digitalized varies across

industries, reflecting the degree of automation (Figure 2.9). The introduction of more and smarter sensors to monitor various parameters, ranging from operating conditions to equipment status, allows for the identification and diagnosis of system

inefficiencies and the development of operation and maintenance schedules that

reduce downtime by predicting equipment failure. Current trends point to the development of “plug and play” and interoperable monitoring and optimisation software, which can be easily implemented and used in different industrial processes

and firm sizes. Advanced data analytics can also support operators in making highly complex decisions by analysing vast amounts of data in short periods of time, helping troubleshoot equipment or even take corrective actions in emergency events. Figure 2.9 Stages of digitalization in industrial process control Offline

open-loop control

Online

open-loop control

Autonomous

closed-loop control

Digital meters and sensors collect data on system

Digital meters and sensors collect data on system

Digital meters and sensors collect data on system

human operators.

algorithms, but implemented

algorithms and automatically

performance, but control and optimisation is undertaken by

performance, with optimisation actions determined by control manually by human operators.

performance, with optimisation actions determined by control implemented by digital systems.

Key message: Increased digitalization enables greater autonomy of process controls. Source: DRET (2013), Case Studies in Systems Optimisation to Improve Energy Productivity.

It is difficult to present a single figure for the energy savings that digitalization can

yield in industry, as potential savings vary according to the type of activity, management systems, culture and the degree of integration along supply value chains. That said, real-plant data show that energy efficiency gains from the application of

advanced digital process controls can yield significant savings at little or no net cost.

Two examples illustrate this potential. In the United States, improved process controls produced estimated energy savings of over USD 330 million in small and

medium-sized manufacturers at a total investment cost of USD 235 million over the period 1987-2015 (Figure 2.10). The average payback from these measures was

typically less than a year. Recommended measures and associated energy savings increased markedly during the late 1990s and early 2000s with advances in computing

technology. The divergence between cumulative energy savings and investment cost it easier and cheaper to optimise process controls to improve energy efficiency.

Impact of digitalization on energy demand in transport, buildings and industry

51

© OECD/IEA, 2017

that can be seen from the late 1990s reflects how technological advances have made

30

1.2

20

0.8

10

0.4

0 0 1987 1991 1995 1999 2003 2007 2011 2015 Energy savings (left axis)

Payback (right axis)

400

400

300

300

200

200

100

100

USD (2010) million

1.6

Cumulative savings USD (2010) millions

40

Years

Annual savings USD (2010) millions

Figure 2.10 Energy savings, payback and investment in digitally enabled optimisation of process controls in the United States

0 0 1987 1991 1995 1999 2003 2007 2011 2015 Cumulative investment cost (right axis)

Key message: Improved process controls enhanced by digitalization can deliver substantial energy and cost benefits.

Notes: These measures were identified by audits funded by the US Department of Energy and undertaken by Industrial Assessment Centers (IACs). The energy savings identified are dependent on the number of audits undertaken in a given year, the number of recommendations identified and the companies in which the audits were conducted. Source: IAC (2017), IAC Database (database).

A similar picture emerges from the results of an analysis of the potential energy savings from compliance with national industrial energy efficiency policies by

optimising process control in Australia (Figure 2.11). The majority of the potential savings that were identified had payback periods of less than two years, with greater savings possible over longer periods.

Energy savings

(% total process or technology use)

Figure 2.11 Potential energy savings from improvements in process control enabled by digitalization by subsector in Australia, 2010-11 12% 8% 0 - 2 years 4% 0%

2 - 4 years 4+ years Food product manufacturing

Beverage, tobacco and textile

Wood, pulp, paper and printing

Basic chemical Mineral product, Fabricated Other and chemical primary metal metal product manufacturing product and metal manufacturing manufacturing product manufacturing

Key message: All industrial sectors analysed present opportunities with payback periods of less Source: ClimateWorks Australia (2013), Industrial Energy Efficiency Data Analysis Project.

52

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

than two years for energy savings from improved process controls enabled by digitalization.

Digital technology has also had an impact on the way products are manufactured.

Technologies such as industrial robots and 3D printing are becoming standard practice in certain industrial applications. These technologies can help increase accuracy and reduce industrial scrap.

The worldwide supply of industrial robots has grown considerably in recent years (Figure 2.12). Since 2010, sales of industrial robots have grown on average by 16% per

year, driven mainly by emerging economies, particularly China. China was responsible for 27% of the supply of industrial robots worldwide in 2015, with the automotive and

electronics industries by far the largest sectors for deployment globally. Deployment of

industrial robots is expected to continue to grow rapidly, with the total stock of robots rising from around 1.6 million units at the end of 2015 to just under 2.6 million at the end of 2019 (IFR, 2016).

Figure 2.12 Number of industrial robots worldwide by sector Automotive industry Electrical/electronics Metal Chemical, rubber and plastics Food

End 2015

Others

End 2014

Unspecified

End 2013 0

20 000

40 000

60 000

80 000

100 000

120 000

Key message: The number of industrial robots in use continues to rise, driving productivity gains. Source: IFR (2016), “Executive summary world robotics 2016 industrial robots”.

As industries continue to digitalize, 3D printing has emerged as one of the most visible

process technologies and one most likely to transform certain industrial operations

(Cotteller and Joyce, 2014). 3D printing can produce both plastic and metal parts in layer-by-layer fashion, on demand and directly from digital 3D files. It has several

advantages compared with conventional manufacturing, including reductions in lead time, reduction of scrap materials, lower inventory costs, less manufacturing

complexity, reduced floor space and the ability to deliver manufactured pieces with complex shapes and geometries (Huang, 2016). It can yield significant energy and

resource savings under the right conditions. As an electricity-driven process, it also promotes the electrification of thermal forming processes such as metal casting and

forging. This could lead to lower CO2 emissions to the extent that power generation is

Impact of digitalization on energy demand in transport, buildings and industry

53

© OECD/IEA, 2017

decarbonised.

3D printing could also lead to the production of new objects with different or novel shapes, which could improve the function of final products and lead to energy savings

beyond the industrial sector. For example, it is already being applied in the production of lightweight aircraft components by some aircraft manufacturers to reduce fuel consumption (Airbus, 2016).

3D printing markets have grown rapidly in recent years. In 2016, the worldwide

industry grew by 17.4% to USD 6.1 billion and is expected to exceed USD 21 billion by 2020 (Wohlers Associates, 2017). However, the technology currently faces barriers that may limit widespread adoption, including high production costs, low throughput rates

and difficulties in meeting the technical requirements for certain products (Huang,

2016; Huang et al., 2017). For these reasons, its use so far has been limited mainly to high-value applications in the aerospace (Box 2.3), medical and transport industries. Box 2.3

3D printing in US aircraft manufacturing

Assessing the net energy and resource savings that can be achieved by 3D printing requires a life-cycle assessment approach. One recent study quantified the energy and resource impacts of selected lightweight metallic additive manufacturing components in the US aircraft fleet, under

different adoption scenarios to 2050 (Huang et al, 2016). The assessment found that 9% to 17% of

total typical aircraft mass could be replaced by lighter 3D printed components in the near term. This could deliver two environmental benefits. First, the reduced materials intensity of these alternative

components could avoid nearly 20 000 tonnes/year of metal demand in 2050. Second, reduced aircraft mass could reduce the overall fuel use of the US aircraft fleet by up to 6.4% in 2050 if those

components were fully adopted. Under a rapid adoption scenario of the selected 3D printed components, the resulting cumulative fuel savings by 2050 would be equivalent to 75% of the fuel consumption of US domestic aviation in 2015 (Figure 2.13). Figure 2.13

25 000

Potential reductions in metals demand and fuel use in the US commercial aircraft fleet from the use of 3D printing

Tons

Metal demand in 2050 1 750

Million GJ

Cumulative aircraft fuel savings

1 500

20 000

1 250

15 000

1 000 750

10 000

500

5 000

250

0

Conventional AM components components Aluminium alloys Nickel alloys Titanium alloys

0

2030 2040 2050 2030 2040 2050 2030 2040 2050 Slow adoption

Mid-range adoption

Rapid adoption

Key message: The use of 3D printed components in commercial aircraft could lead to significant material demand and fuel savings.

Source: Huang et al. (2016), Energy and emissions saving potential of additive manufacturing: the case of lightweight

aircraft components.

54

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

Notes: GJ = gigajoule; AM = additive manufacturing.

Digitalization can also accelerate innovation in industry. By virtually replicating real

industrial plants – through the creation of a “digital twin” that accurately models the

impact of changes to an existing production process – new manufactured products can

reach the market earlier by saving time and resources in product design and in the

development and optimisation of new production processes. Enhanced data collection and analytics can improve industry’s understanding of consumer patterns and societal

needs. When also combined with artificial intelligence or machine learning, automation

and smart robots, virtual and autonomous experimentation becomes possible. Such experimentation greatly enhances industry’s ability to innovate and to do so in a more rapid manner, as it avoids the costs associated with using physical industrial plants for

testing, which requires downtime as well as the use of large amounts of energy and other resources. For example, chemicals manufacturers have used simulated

experiments and optimisation techniques to reduce the batch time30 for producing expanded polystyrene by as much as 30%. This has led to significant energy savings and substantially less production time and cost (World Economic Forum, 2017a).

The learning-by-doing phase of the innovation cycle, where industry improves operating practices and product design by capturing lessons from actual experience,

can also be significantly shortened and made more economical through digitalization. Through continuous data collection and exchange, learning from plant operations can be integrated into new and revamped engineering designs in real time. Accelerating

this learning-by-doing phase can translate into energy savings, as emerging industrial processes could reach their optimal energy performance level more rapidly.

Digitalization beyond the industrial plant fence The impact of digitalization in industry is also moving beyond the plant fence. Digitalization opens up a wide range of opportunities when connecting a particular

industrial facility to its surroundings. For example, digitalization can better connect

producers along product value chains, thereby facilitating the reuse and recycling of materials.

Cloud-based collaborative platforms already exist to let manufacturers exchange

information on available surplus raw materials, industrial by-products and waste. An example is the Materials Marketplace – an online platform that allows producers to

find outlets for their traditional and non-traditional industrial waste streams.31 Supply

chains could be optimised by using inventories informed by real-time demand data from virtual suppliers’ portals to maximise asset utilisation and minimise freight energy

30

“Batch time” refers to the duration of a chemical process reaction.

31

http://materialsmarketplace.org/.

Impact of digitalization on energy demand in transport, buildings and industry

55

© OECD/IEA, 2017

costs.

By connecting industrial equipment to the network, companies can also benefit from identifying and providing real-time information on the availability of local waste

streams, which can be captured and used to displace other forms of energy. Examples include excess heat, organic waste and off-gases, which can be captured and used to

produce heat or electricity, thereby reducing both the cost of purchasing other forms of energy and the environmental footprint of the plant.

Within a firm, data analytics that benchmark energy performance across sites can be used

to

identify

bottlenecks

and

opportunities

for

energy

savings,

while

cloud-connected workers can benefit from a more dynamic exchange of lessons learned and experience. Minimising energy costs reduces a firm’s exposure to price volatility and can provide a competitive advantage.

Autonomous closed-loop control (see Figure 2.9) can allow industrial processes at

different facilities to be remotely controlled to adapt in real time to market needs and

potential energy supply constraints, thereby reducing system inefficiencies. There are

examples of large industrial complexes already being highly digitalized, involving firm-wide operations management systems that gather data from the production process, in real time, and update the plant status database remotely (World Economic Forum, 2017a).

Barriers to further digitalization and policy considerations Lack of accurate information, shortcomings in technical expertise and cultural barriers

may limit the potential for further digitalization within various industries. Informing

stakeholders about digitalization can help overcome these problems. So far,

technology and service providers have been the main source of digitalization

information. Companies that lack technical expertise in the area can struggle to make

sense of it and understand how they can take advantage of the new technologies available.

Governments may have a role to play by becoming more involved in providing advice, particularly for small and medium-sized enterprises that may not have had as much exposure to digital technology. Governments can also promote energy management systems that embrace advanced process controls to identify and monitor energy

efficiency improvements. The adoption of energy management systems such as ISO

50001 – the global standard for energy management – is driven by government policies or incentives in many countries.

Government support for research and development (R&D) can also mitigate

investment risks and boost private-sector industrial investment in digital technologies. increasing the capacity of data analytics to solve complex problems and enable

56

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

Potential R&D opportunities include expanding the applicability of 3D printing and

autonomous machine learning. Public-private partnerships can also be an effective

framework to support demonstration and pilot projects involving the application of digital technologies in industry to reduce perceived risks and accelerate adoption.

Policies also need to encourage the participation of industry in electricity demand

response, which digitalization makes possible through improved process controls and

greater connectivity. Policy makers and regulators can make it easier for industrial firms to participate in demand response by reducing barriers for industrial sites to connect to electricity grids and district heating networks and harnessing innovative business models (see Chapter 4).

While improved connectivity can increase productivity across firms, supply chains and

industrial sectors, to fully realise these benefits also requires policies and efforts

related to cybersecurity and data privacy. Digitalization’s impacts on jobs and skills,

Impact of digitalization on energy demand in transport, buildings and industry

57

© OECD/IEA, 2017

especially in the industrial sector, can also cause challenges (see Chapter 6).

References Abuelsamid, S., Alexander, D., and Jerram, L. (2017). Navigant Research Leaderboard Report:

Automated Driving. www.navigantresearch.com/wp-assets/brochures/LB-NGBC-16-ExecutiveSummary.pdf.

Airbus (2017). “Planes can talk, let’s listen”. http://airbus-xo.com/planes-can-talk-lets-listen/ (accessed 22 June 2017).

Airbus (2016). “Innovative 3D printing solutions are ‘taking shape’ within Airbus”, Airbus press release. www.aircraft.airbus.com/newsevents/news-events-single/detail/innovative-3dprinting-solutions-are-taking-shape-within-airbus/ (accessed 24 July 2017). Arbib, J. and Seba, T. (2017). Rethinking Transportation 2020-2030. www.rethinkx.com/transportation.

Archambault, P., et al. (2015). “Monetizing the rise of Autonomous Vehicles”. Cars 2025 (Vol. 3). Goldman Sachs. http://pg.jrj.com.cn/acc/Res/CN_RES/INVEST/2015/9/17/f70472c6-f4ad4942-8eab-3c01f3c717a7.pdf. Brown, A., Gonder, J. and Repac, B. (2014). “An Analysis of Possible Energy Impacts of Automated Vehicle.” Road Vehicle Automation, 137–153. http://doi.org/10.1007/978-3-319-05990-7.

BMVi (Bundesministerium für Verkehr und digitale Infrastruktur) (2017). Ethics Commission:

Automated and Connected Driving. www.bmvi.de/SharedDocs/EN/publications/report-ethics-

commission.pdf?__blob=publicationFile.

Choo, S., Mokhtarian, P. L., and Salomon, I. (2005). “Does telecommuting reduce vehicle-miles travelled? An aggregate analysis for the US? Transportation, 32, 37–64.

Clewlow, R. R. (2016). “Shared-use mobility in the United States: current adoption and potential impacts on travel behavior”. Transportation Research Board.

www.reginaclewlow.com/pubs/TRB2016_SharedUseMobility_final.pdf. Clewlow, R. R. and Mishra, G. S. (2017). Disruptive Transportation: The Adoption, Utilization, and

Impacts of Ride-Hailing in the United States. Institute of Transportation Studies, University of

California, Davis, Research Report UCD-ITS-RR-17-07.

ClimateWorks Australia (2013). “Industrial Energy Efficiency Data Analysis Project”.

www.climateworksaustralia.org/project/industrial-energy-efficiency/industrial-energyefficiency-data-analysis (accessed 24 July 2017).

Cotteller, M., and Joyce, J. (2014). “3D opportunity: Additive manufacturing paths to performance, innovation, and growth”. https://dupress.deloitte.com/dup-us-en/deloitte-review/issue-

14/dr14-3d-opportunity.html (accessed 24 July 2017).

DRET (Department of Resources, Energy and Tourism (Australia)) (2013). “Case Studies in Systems Optimisation to Improve Energy Productivity”. 2013.pdf (accessed 24 July 2017).

58

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

www.eex.gov.au/sites/g/files/net1896/f/files/2014/08/Systems-Optimisation-Case-Study-

Energy Star (2017). Certified Smart Thermostats.

www.energystar.gov/productfinder/product/certified-connected-thermostats/results (accessed 24 July 2017).

European Commission (2017). “Europe on the Move - An agenda for a socially fair transition

towards clean, competitive and connected mobility for all”. Commission Staff Working Document, accompanying the document Communication from the Commission to the European Parliament,

the Council, the European Economic and Social Committee and the Committee of the Regions,

Brussels, 31.5.2017, SWD(2017) 177 final. http://eur-lex.europa.eu/legalcontent/EN/TXT/PDF/?uri=CELEX:52017SC0177&from=EN.

Frost and Sullivan (2016). Future of Carsharing Market to 2025; Technology Advancements, Market Consolidation and Government Initiatives to Influence Market Growth Over the Next Decade, Report #MB4D-01, August 2016. Fulton, L., Mason, J., and Meroux, D. (2017). Three Revolutions in Urban Transportation. Davis, CA. https://steps.ucdavis.edu/wp-content/uploads/2017/05/ITDP-3R-Report-v6.pdf.

Gartner (2016). Gartner’s 2016 Hype Cycle for Emerging Technologies. www.gartner.com/newsroom/id/3412017 (accessed 21 June 2017).

Goodchild, A. and Toy, J. (2017). “Delivery by drone: An evaluation of unmanned aerial vehicle

technology in reducing CO2 emissions in the delivery service industry”. Transportation Research,

Part D: Transport and Environment.

Greenblatt, J. B., and Saxena, S. (2015). “Autonomous taxis could greatly reduce greenhouse-gas emissions of US light-duty vehicles”. Nature Climate Change, 5(9), 860–863.

http://doi.org/10.1038/nclimate2685.

Grözinger, J., et al. (2017). “Optimising the energy use of technical building systems – unleashing the power of the EPBD’s Article 8”. Ecofys, www.ecofys.com/files/files/ecofys-2017-optimisingthe-energy-use-of-tbs-final-report.pdf.

Hern, A. (2017). “Assume self-driving cars are a hacker’s dream? Think again”. The Guardian,

30 August 2017. www.theguardian.com/technology/2017/aug/30/self-driving-cars-hackerssecurity.

Horner, N. C., Shehabi, A., and Azevedo, I. L. (2016). “Known unknowns: Indirect energy effects of

information and communication technology.” Environmental Research Letters , 11(10), 103001.

Huang, R. (2016). A Multi-Scale Life Cycle Framework for the Net Impact Assessment of Additive

Manufacturing in the United States. Northwestern University.

Huang, R., et al. (2017). “Cost minimization in metal additive manufacturing using concurrent structure and process optimization”. Proceedings of the ASME 2017 International Design

Engineering Technical Conference & Computers and Information in Engineering Conference IDETC2017. http://vdel.me.cmu.edu/publications/2017dac/paper.pdf.

of lightweight aircraft components”. http://doi.org/10.1016/j.jclepro.2015.04.109. Impact of digitalization on energy demand in transport, buildings and industry

59

© OECD/IEA, 2017

Huang, R., et al. (2016). “Energy and emissions saving potential of additive manufacturing: the case

IAC (Industrial Assessment Centers) (2017). IAC Database (database). https://iac.university/download.

IEA (International Energy Agency) (2017a). The Future of Trucks: Implications for Energy and the

Environment. OECD/IEA, Paris.

IEA (2017b). Energy Technology Perspectives 2017: Catalysing Energy Technology Transformations. OECD/IEA, Paris.

IEA (2017c). Energy Access Outlook 2017: From Poverty to Prosperity. World Energy Outlook

Special Report. OECD/IEA, Paris.

IFR (International Federation of Robotics) (2016). “Executive summary world robotics 2016 industrial robots”.

https://ifr.org/img/uploads/Executive_Summary_WR_Industrial_Robots_20161.pdf (accessed 24 July 2017). Johnson, C. and Walker, J. (2016). “Peak Car Ownership: The Market Opportunity for Electric

Automated Mobility Services”. Rocky Mountain Institute. www.rmi.org/peak_car_ownership.

Kim, S. N., Choo, S., and Mokhtarian, P. L. (2015). “Home-based telecommuting and intra-

household interactions in work and non-work travel: A seemingly unrelated censored regression approach”. Transportation Research Part A: Policy and Practice, 80, 197–214.

http://doi.org/10.1016/j.tra.2015.07.018.

Kofod, C. (2016). Smart Lighting – New Features Impacting Energy Consumption First Status Report. IEA 4E Solid State Lighting Annex Task 7. http://ssl.iea-

4e.org/files/otherfiles/0000/0085/SSL_Annex_Task_7_-_First_Report_-_6_Sept_2016.pdf. Larson, W., and Zhao, W. (2017). “Telework: Urban form, energy consumption, and greenhouse gas implications”. Economic Inquiry, 55(2), 714–735. http://doi.org/10.1111/ecin.12399.

Li, K. et al. (2012). “Intelligent environment-friendly vehicles: Concept and case studies”. IEEE

Transactions on Intelligent Transportation Systems, 13(1), 318–328.

http://doi.org/10.1109/TITS.2011.2170680.

McKinnon, A. C. (2015). “Postscript 1 – Distribution by drone”. In M. Piecyk, et al. (eds.), Green

Logistics (3rd ed.). Safari Books Online. www.safaribooksonline.com/library/view/green-

logistics-3rd/9780749471859/C19.xhtml.

Melo, P. C., and de Abreu e Silva, J. (2017). “Home telework and household commuting patterns in Great Britain”. Transportation Research Part A: Policy and Practice, 103, 1–24. http://doi.org/10.1016/j.tra.2017.05.011.

Milakis, D., et al. (2015). Development of automated vehicles in the Netherlands: Scenarios for 2030

and 2050. https://repository.tudelft.nl/islandora/object/uuid:c22db456-b61a-4908-b2f6-

51d16d5708f8?collection=research.

September, 2016. www.lvm.fi/documents/20181/880492/3-

60

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

Ministry of Transport and Communications (2016). Transport Code Fact Sheet, 3/2016, 20

2016+Transport+Code.pdf/930dbb2c-d4c2-4248-a61c-3cb4fed8eaf9, (accessed 28 August 2017).

Mitchell, R. (2017). “Tesla reports fourth-quarter and full-year results, says Model 3 on track”. Los

Angeles Times, 22 February 2017. www.latimes.com/business/autos/la-fi-hy-tesla-earnings-4q2016-story.html.

Morgan Stanley (2017). Shared Mobility on the Road of the Future.

www.morganstanley.com/ideas/car-of-future-is-autonomous-electric-shared-mobility.

Narla, S. R. K. (2013). “The evolution of connected vehicle technology: From smart drivers to smart cars to... self-driving cars”. ITE Journal (Institute of Transportation Engineers), 83(7), 22–26.

National Grid (2017). “The Smart Platform for Your Smart Home”. www.nationalgridus.com/MAHome/Energy-Saving-Programs/ConnectedSolutions (accessed 24 July 2017).

Navigant Research (2016). Communicating Thermostats, Smart Thermostats, and Associated

Software and Services: Global Market Analysis and Forecasts, Navigant Research,

www.navigantresearch.com/research/smart-thermostats.

Nest (2015). “Energy Savings from the Next Learning Thermostat: Energy Bill Analysis Results”.

Nest Labs, February 2015. http://downloads.nest.com/press/documents/energy-savings-whitepaper.pdf.

Olson, P. (2014). Why Google's waze is trading user data with local governments. Forbes, 7 July 2014. www.forbes.com/sites/parmyolson/2014/07/07/why-googlewaze-helps-localgovernments-track-its-users/.

Ory, D. T., and Mokhtarian, P. L. (2006). “Which came first, the telecommuting or the residential relocation? An empirical analysis of causality”. Urban Geography, 27(7), 590–609. http://doi.org/10.2747/0272-3638.27.7.590.

Perlroth, N. (2017), “Electronic Setups of Driverless Cars Vulnerable to Hackers”. New York Times, 7 June 2017. www.nytimes.com/2017/06/07/technology/electronic-setups-of-driverless-carsvulnerable-to-hackers.html?mcubz=0. Philips (2015). “Where are the discounts? Carrefour’s LED supermarket lighting from Philips will guide you”. www.newsroom.lighting.philips.com/news/2015/20150521-Where-are-the-

discounts-Carrefours-LED-supermarket-lighting-from-Philips-will-guide-you (accessed 24 July 2017).

Pye, S. and Dobbins, A (2015). “Energy poverty and vulnerable consumers in the energy sector across the EU: Analysis of policies and measures”.

https://ec.europa.eu/energy/sites/ener/files/documents/INSIGHT_E_Energy%20Poverty%20%20Main%20Report_FINAL.pdf.

SAE International (2016). J3016: Taxonomy and Definitions for Terms Related to Driving Automation Schaller, B. (2017). Unsustainable? The Growth of App-Based Ride Services and Traffic, Travel and

the Future of New York City. Brooklyn, NY. www.schallerconsult.com.

Impact of digitalization on energy demand in transport, buildings and industry

61

© OECD/IEA, 2017

Systems for On-Road Motor Vehicles. http://standards.sae.org/j3016_201401/.

Shaheen, S., Cohen, A. and Zohdy, I. (2016). Shared Mobility: Current Practices and Guiding

Principles. No. FHWA-HOP-16-022. 2016,

https://ops.fhwa.dot.gov/publications/fhwahop16022/fhwahop16022.pdf. Shaheen, S., et al. (2016). Smartphone Applications to Influence Travel Choices: Practices and

Policies. No. FHWA-HOP-16-023. 2016.

https://ops.fhwa.dot.gov/publications/fhwahop16023/fhwahop16023.pdf. Slowik, P. and Kamakaté, F. (2017), “New mobility: Today’s technology and policy landscape”.

International Council for Clean Transportation.

www.theicct.org/sites/default/files/publications/New-mobility-landscape_ICCT-whitepaper_27072017_vF.pdf. Stephens, T. S. et al. (2016). “Estimated Bounds and Important Factors for Fuel Use and Consumer Costs of Connected and Automated Vehicles”. National Renewable Energy Laboratory. Golden,

CO. www.nrel.gov/docs/fy17osti/67216.pdf.

Tesla (2016). “All Tesla cars being produced now have full self-driving hardware”. www.tesla.com/blog/all-tesla-cars-being-produced-now-have-full-self-driving-hardware (accessed 21 June 2017).

Toyota (2016). “Toyota Motor Corporation (TMC) announces financial results for fiscal year ended March 31, 2016”, http://corporatenews.pressroom.toyota.com/releases/toyota-april-march2016-financial-results.htm. US Department of Transportation (2016). History of Intelligent Transportation Systems. Washington, DC. https://ntl.bts.gov/lib/59000/59200/59263/download1.pdf.

US Department of Transportation (2015). “Critical reasons for crashes investigated in the National Motor Vehicle Crash Causation Survey”.

https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/812115. Wadud, Z., MacKenzie, D., and Leiby, P. (2016). “Help or hindrance? The travel, energy and carbon

impacts of highly automated vehicles”. Transportation Research Part A: Policy and Practice, 86,

1–18. http://doi.org/10.1016/j.tra.2015.12.001.

Wadud, Z. (2017). “Fully automated vehicles: A cost of ownership analysis to inform early adoption”.

Transportation Research Part A: Policy and Practice, 101, 163-176.

www.sciencedirect.com/science/article/pii/S096585641630739X.

Wible, B., Mervis, J., and Wigginton, N. S. (2014). “Rethinking the global supply chain”. Science, 344(6188), 1100–1103. http://doi.org/10.1126/science.344.6188.1100.

Wilczynski, E. (2017). “Turning up the heat: The rapid surge in smart thermostat programs”.

www.esource.com/Blog/ESource/ES-Blog-3-6-17-Smart-Thermostats (accessed 24 July 2017).

Wohlers Associates (2017). “Wohlers Report 2017 shows vibrant new business activity in 3D printing with softened growth worldwide. https://wohlersassociates.com/press72.html

62

Impact of digitalization on energy demand in transport, buildings and industry

© OECD/IEA, 2017

(accessed 24 July 2017).

World Economic Forum (2017a). Digital Transformation Initiative: Chemistry and Advanced

Materials Industry. http://reports.weforum.org/digital-transformation/wp-

content/blogs.dir/94/mp/files/pages/files/dti-chemistry-and-advanced-materials-industrywhite-paper.pdf.

World Economic Forum (2017b). Digital Transformation Initiative: Mining and Metals Industry. http://reports.weforum.org/digital-transformation/wp-

content/blogs.dir/94/mp/files/pages/files/wef-dti-mining-and-metals-white-paper.pdf.

Zimmer, J. (2016). The Third Transportation Revolution Lyft’s Vision for the Next Ten Years and

Beyond. https://medium.com/@johnzimmer/the-third-transportation-revolution-27860f05fa91 (accessed 21 June 2017).

Zimmer, J., and Green, L. (2017). Lyft Climate Impact Goals.

Impact of digitalization on energy demand in transport, buildings and industry

63

© OECD/IEA, 2017

https://blog.lyft.com/posts/2017/6/14/lyft-climate-impact-goals (accessed 21 June 2017).

© OECD/IEA, 2017

Chapter 3

Impact of digitalization on oil and gas, coal, and power supply Highlights •

Energy companies have been adopting digital technologies for years, helping to increase the recovery of fossil resources, improve production processes, reduce costs and improve safety.



The oil and gas sector has a relatively long history with digital technologies, notably in upstream, and significant potential remains for digitalization to enhance operations further. The widespread use of

existing digital technologies could decrease production costs by between

10% and 20%. With the use of both existing and emerging digital technologies, technically recoverable oil and gas resources could be

boosted by around 5% globally. The potential impact of digitalization is likely to be greatest for tight oil and shale gas resources.



In the coal sector, digitalization can further improve geological modelling, mining optimisation and other related processes, automation, predictive maintenance, and worker health and safety. The overall impact, however, may be more modest than in other sectors.



Digitalization within the power sector has the potential to save around

USD 80 billion per year, or about 5% of total annual power generation costs, based on the current system design and enhanced global deployment of available digital technologies to all power plants and

network infrastructure. This can be achieved by reducing operation and maintenance costs, improving power plant and network efficiency, unplanned

outages

operational lifetime of assets.

Impact of digitalization on oil and gas, coal, and power supply

and

downtime,

and

extending

the

65

© OECD/IEA, 2017

reducing

Introduction Digitalization can improve safety, increase productivity and reduce costs in oil and gas, coal and power. This chapter explores digitalization’s impact within each of these

supply sectors, as well as sector-specific barriers to further digitalization. The

magnitude of these potential impacts – and associated barriers – varies greatly

depending on the particular application (Figure 3.1).

Broad systemic implications are discussed in Chapter 4, including digitalization’s

ability to break down barriers between various demand and supply sectors, leading to

even greater potentially transformative impacts. Cross-sectoral policy challenges such as cybersecurity, data privacy and job losses are discussed in Chapter 6.

Figure 3.1 Digitalization’s potential impact in oil and gas, coal, and power

High

Unplanned outages Unconventional upstream Extending lifetimes Optimisation / efficiency Reducing O&M costs

Oil and gas Coal

Digital automation

Geological modelling

Low

Magnitude of potential impact

Connected power system (Chapter 4)

Power plant efficiency

Power

Conventional upstream Downstream

Low

Barriers to digitalization

High

Key message: The magnitude of impacts on productivity and efficiency differ across applications in supply sectors.

Notes: The list of applications in the figure is not intended to be exhaustive; “Magnitude of potential impact” indicates the

66

Impact of digitalization on oil and gas, coal, and power supply

© OECD/IEA, 2017

overall potential impact of digitalization on productivity and efficiency; “Barriers to digitalization” include technological, financial, regulatory and public perception components; the quadrants are illustrative only and intended to give a sense of relative magnitude; O&M = operation and maintenance.

Oil and gas The oil and gas industry is well accustomed to pushing the boundaries of technology, especially in the upstream sector. About 40 years ago, oil and gas platforms in the Gulf of Mexico ventured into what was then considered deep water, building solid structures grounded on the sea floor about 350 metres below sea level. Today, floating

dynamically positioned offshore platforms operate equipment at depths of up to

3 000 metres and drill into reservoirs several kilometres below the sea floor.

Technological advances have allowed the exploitation of oil and gas fields in remote

locations, whether in deep water or harsh land environments, which were once considered technically inaccessible. All of these advances would not have been possible without state-of-the-art digital technologies.

Potential for further digitalization in the upstream sector Exploration and production have traditionally been the most profitable part of the oil and gas sector, and this is the area where digital technologies have tended to have the

largest impact. Technologies that can unlock more volumes of oil or natural gas or

make production processes more efficient are prime candidates for research, development and deployment. The upstream industry has pioneered the complex task

of processing extremely large datasets generated by seismic surveys of land and oceans, for example, in order to delineate the outline and structure of reservoirs to

help optimise their development. Processing this data requires some of the world’s most powerful computers. Other digital applications include the real-time, dynamic

steering of drill bits from remote operations centres, or the use of highly sophisticated

sensors to optimise where wellbores should be placed to maximise oil and gas recovery.

The initial focus of digitalization in the upstream oil and gas industry in the future is most likely to be on expanding and refining the range of existing digital applications already in use. For example, miniaturised sensors and fibre optic sensors in the

production system could be used to boost production or increase the overall recovery of oil and gas from a reservoir. Such sensors could also be employed to measure

environmental performance, such as the efficiency or emissions intensity of operations. Other examples are the use of automated drilling rigs and robots to inspect

and repair subsea infrastructure and to monitor transmission pipelines and tanks. Drones could also be used to inspect pipelines (which are often spread over extended

areas) and hard-to-reach equipment such as flare stacks and remote, unmanned

offshore facilities. As with all new technologies, as their use becomes more frequent

and widespread, the associated costs will fall, creating a snowball effect. This should more frequent inspections and more effective preventative maintenance).

Impact of digitalization on oil and gas, coal, and power supply

67

© OECD/IEA, 2017

lead to increased safety, lower labour costs and increased equipment reliability (due to

In the longer term, the potential exists to improve the analysis and processing of data,

such as the large, unstructured datasets generated by seismic studies. Rapid analysis of data can lead to faster decisions and increase the operating time of drilling rigs,

wells and facilities, thereby reducing delays when executing new projects. As a result, this should also lead to lower costs and the more efficient use of capital. The use of

more sophisticated processing algorithms could also assist with the finding of new oil and gas fields, the generation of development plans, and the ranking of exploration portfolio options for upstream operators.

A further option is the use of artificial intelligence (AI), which is still in its infancy but

offers a great deal of promise for upstream operations. AI could be used to analyse

well performance, troubleshoot underperforming fields, suggest corrective actions and even deploy robots to carry out tasks. It could also enhance reservoir modelling, and

thus aid operations by rapidly detecting and correcting suboptimal production behaviour (Figure 3.2).

Figure 3.2 Example of digital visualisation of an oil well and reservoir

Key message: Enhanced modelling of the subsurface through the use of digital technologies that accurately delineate the reservoir, internal geological features, fluid distribution and flow, helps optimise production.

68

Impact of digitalization on oil and gas, coal, and power supply

© OECD/IEA, 2017

Source: © Schlumberger, 2017.

One particular area of interest for the use of digitalization is tight oil and shale gas

production. Coupling AI with sensors and sophisticated data management tools could allow companies to operate and

maximise production from thousands of

unconventional tight oil and shale gas wells with only a handful of engineers and

technicians. Tight oil and shale gas production is likely to be especially suitable for the application of new digital technologies given the shorter-term nature of investment

cycles (compared with conventional fields), which should favour the introduction and

implementation of new technologies in a more continuous manner. Widely shared

inventories of equipment suppliers could also shorten procurement times and reduce

costs, creating a more efficient “just-in-time” supply chain across the unconventional oil and gas industry.

Enhanced recovery of oil and gas resources The volume of oil and gas resources that can be produced is a critical factor in understanding the future trajectory of oil and gas prices. Remaining technically

recoverable oil and gas resources around the world are currently estimated to be around 1.4 trillion tonnes of oil equivalent (toe) (IEA, 2016). Through the widespread use of existing and emerging digital technologies across the global resource base, the IEA estimates that this could be increased by up to 75 billion toe, or around 5%, equal to over 10 years of current world oil and gas consumption (Figure 3.3).

Figure 3.3 Impact of digitalization on global technically recoverable oil and gas resources 500 400

billion toe 3%

3%

3%

% Increase in resources 15%

300

Absolute increase in resources

200 100 0

Oil

Gas

Oil

Conventional

Gas Unconventional

Key message: Widespread digitalization could boost global technically recoverable oil and gas resources by around 5%.

Impact of digitalization on oil and gas, coal, and power supply

69

© OECD/IEA, 2017

Note: It is assumed that both existing and emerging digital technologies are deployed across the entire world oil and gas resource base.

The impact of digitalization on the volumes of oil and gas that can be recovered varies according to the resource in question, and its potential to boost recoverable resources

is likely to be largest for tight oil and shale gas resources. For both of them the

widespread use of digital technologies could increase recovery by around 15%.1 This is because current recovery factors for unconventional reservoirs are much lower than

those for conventional reservoirs. For example, while up to 90% of the gas trapped in a conventional reservoir can usually be recovered with existing technologies, the

recovery factor for an unconventional shale gas reservoir is often only between 15% and 35%.

The digital technologies that could achieve an increase in recovery factors are wide-

ranging and include advanced processing for seismic data to yield much more reliable and high-resolution digital images of the reservoir, enhanced modelling and tracking of

fluids within the reservoir, and automated production control of wells. By improving reservoir simulations, digital technology can help operators optimise the spacing between wells, the lateral length of horizontal wells, and the amount of proppant used during hydraulic fracturing. All of these efforts are aimed at maintaining maximum output while minimising the required capital investment in wells and facilities. Optimised production processes In addition to expanding the recoverable resource base, production processes could be optimised through enhanced connectivity and monitoring to help reduce both capital

and operating costs (Figure 3.4). There are few detailed estimates of the potential

magnitude of these cost reductions, however, as many of the most promising digital

technologies are tailored to specific subsectors and niche applications, and are still in

the early stages of development. Based on discussions with operators and service providers, the IEA estimates that the widespread use of existing digital technologies could help reduce costs by around 10-20%. Even greater cost reductions could be possible if emerging technologies prove to be successful and are adopted widely.

Monitoring of methane emissions Another important area where digitalization could improve oil and gas operations is in

detecting, measuring and avoiding methane emissions. The oil and gas sector is a major source of anthropogenic methane emissions, with continuous or sporadic

1

The resulting increase for all unconventional oil is only 3%, as indicated in Figure 3.3. This is a result of the fact that tight oil

represents the smallest share of unconventional oil resource; extra-heavy, bitumen and kerogen oil that compose the largest share recovery determines a corresponding increase for all unconventional gas.

70

Impact of digitalization on oil and gas, coal, and power supply

© OECD/IEA, 2017

do not benefit much from digitalization. Shale gas has the largest share in unconventional gas resources and a 15% increase in its

emissions potentially occurring at any stage during production, processing and transport of oil and gas.2

The urgent need is for systems that can provide effective monitoring and

quantification of emission levels at low cost. Digitalization could help here, either by lowering the direct cost of detection (for example, using drones in localised facilities)

or by helping to understand better the data that are collected to develop parametric and predictive monitoring systems. Once a source of emissions is identified, stopping or reducing them is relatively simple.

Figure 3.4 Digitally-connected remote operations in oil and gas fields

Real-Time Optimization

Permanent 4D/reservoir monitoring

Frac Monitoring Smart Mobile Worker GBC

Leak detection pipelines

Smart Wells

Smart Waterflood Fiber Optic Sensing

Key message: Many oil and gas operations are digitally connected and can be monitored and operated remotely to help optimise production.

2

See the forthcoming IEA World Energy Outlook 2017 for a detailed discussion on oil and gas methane emissions.

Impact of digitalization on oil and gas, coal, and power supply

71

© OECD/IEA, 2017

Source: Courtesy of Shell Global Solutions International B.V.

Improved downstream operations Downstream oil and gas operations could also benefit from more widespread use of digital technologies. Profit margins in this sector are often low and therefore any small improvements would be greatly beneficial to operators.

One current example is companies that are exploring ways to customise their retail services by offering to refuel cars at home or work by bringing fuel to customers rather

than customers needing to go to a refuelling station. This could be made possible by

the automatic communication of data on the amount of fuel left in the vehicle’s tank to a local fuel station. Another example is the use of smart gas meters (similar to smart

electricity meters) that can help match gas demand and supply.

Barriers to further digitalization Looking at the oil and gas sector, the adoption of digital technologies to date has been

uneven. Overall, it trails other sectors, such as finance, retail, medicine and automobiles. Several factors explain this lag:

• Timing: The oil and gas industry is very capital intensive and large projects

typically take many years to develop. Digital technologies, in contrast, evolve quickly. Once a large, multi-billion-dollar project has been designed and

sanctioned, the industry’s focus is generally on effective execution of the

project; design changes are kept to a minimum, impeding the incorporation of emerging innovations.

• Age of infrastructure: Many oil and gas facilities around the world were installed some time ago, and do not necessarily have the appropriate

infrastructure to accommodate new digital technologies. Retrofitting these facilities would carry additional costs, making the application of digital technologies less attractive.

• Internal focus: Since oil and gas are commodities, it is hard for companies to

differentiate between products and services to consumers. As a result, digital technologies have tended to be used more to enhance safety, operational reliability and reduce costs.

• Small mistakes have big consequences: The oil and gas industry has developed a relatively risk-averse management perspective that can slow down

the adoption rate of new technologies, regardless of their potential. The deployment of new equipment, including new digital options, often requires high-level management approval, which can lead to delays and add to costs.

chain. The international oil companies apart, few companies are vertically

integrated with interests in upstream, refining, transport and retailing. 72

Impact of digitalization on oil and gas, coal, and power supply

© OECD/IEA, 2017

• Fragmentation: The oil and gas industry is highly fragmented along the supply

Digitalization remains largely tailored to the needs of individual subsectors, such that companies may not be able to take full advantage of potential

cross-cutting benefits (see Chapter 4). Nonetheless, large oilfield service

providers, a key source of technological innovation in the industry, will undoubtedly continue to expand their digital services on offer.

• Long-term demand trends: Oil and gas resources may be larger than the total

amount that will be consumed given current trends in the deployment of low-carbon technologies and energy efficiency. While most reservoirs could

technically benefit from the use of sophisticated digital applications, it may not always make sense economically to do so.

• Information technology (IT) support infrastructure: While many digital technologies are already available, many operators are not well placed to

exploit them, as their use requires well-developed IT infrastructure as well as a well-trained workforce.

• Conservative management culture: The oil and gas industry’s capital-

intensive nature and operational hazards have historically forged a relatively conservative management culture. Oil and gas companies may look to service

companies and third-party vendors to develop digital technologies rather than

bear the risks and cost of large-scale in-house research and development programmes.

Coal Historically, the main technological advances in underground coal mining were achieved through mechanisation. This significantly reduced the number of coal miners

required, while increasing productivity. In surface mining, the most significant development was the use of large-scale equipment, allowing the more efficient and

rapid movement of large volumes of coal and overburden3 involving fewer miners. The application of digital technologies, including data sensors and advanced computers,

has already started and is set to grow, which will lead to significant changes in some operations in the coal sector.

Deployment of digital technologies Digital technologies are being used throughout the coal supply chain to reduce

production and maintenance costs, and enhance workers’ safety. Examples include

3

Overburden is the material that lies above a coal seam and needs to be removed to allow access to the coal seam.

Impact of digitalization on oil and gas, coal, and power supply

73

© OECD/IEA, 2017

semi- or fully-automated systems, robotic mining, remote mining, operation

automations, mine modelling and simulations, and the use of global positioning system (GPS) and geographic information system (GIS) tools.

The increased availability of low-cost sensors and computer-aided simulations will bring new opportunities for coal operations. For example, sensors can provide the exact status of various components of the essential equipment in real time and

analytics can compare the actual configuration with the “optimal” situation as designed so that the process can be optimised. This could also improve management

of constraints, such as water availability, and minimise the amount of surface land

affected by mining. Digital technologies, data analytics and automation will be increasingly adopted to improve productivity while enhancing safety and environmental performance through multiple applications. Specific applications include:

• Geological modelling: Modern data management techniques, new tools to

simulate and define coal deposits and clearer result visualisation may help better assess reserves and their qualities, as well as enhance understanding of

geological conditions. Drones are already being used to define ground profile

and survey operations. The capacity to obtain and process a growing volume of data could mean that updates to the geological model can be produced continuously.

• Process

optimisation:

The

combination

of

low-cost

sensors

and

computer-aided simulations can help to optimise processes in the coal supply

chain. For example, sensors can detect deviations from optimal operational

conditions, which can be assessed and corrected. Digital optimisation of

open-pit mines, especially with overburden removal, transportation and

disposal, could help reduce the use of materials such as explosives, leading to significant reductions in operating costs. Materials currently represent the largest operating cost component in open-pit mining.

• Process automation: In some parts of the coal supply chain, automation is already common practice, notably coal transportation and washing, but there is potential for further use. Tele-remote equipment as well as driverless trucks

and shovels are two possibilities. Tele-remote equipment allows mining

equipment to be operated from a control room, helped by computerised tools that can simulate geological and environmental conditions. This can raise productivity as well as improve safety by removing workers from risky or

hazardous working environments. While the death toll in coal mining has been reduced dramatically in recent years, it is still higher than in other industries. All

major mining regions in the People’s Republic of China (hereafter, “China”) open-pit mining, driverless trucks and shovels are already operating and full

74

Impact of digitalization on oil and gas, coal, and power supply

© OECD/IEA, 2017

have targets for workerless underground mines by 2030 (NDRC/NEA, 2016). In

automation of the whole chain (drilling, blasting, shovelling, dumping, washing and transportation) appears to be feasible in the medium term.

• Predictive maintenance: Despite the fact that mining equipment operates in a

challenging environment, digitally enabled predictive maintenance 4 in coal

mining is not significantly different from that in other sectors and industries. With abundant availability of low-cost sensors and analytics, maintenance practices can be based on the actual state of components, which can lead to lower equipment unavailability, higher performance and reduced costs.

Potential impacts of digitalization Digitalization will continue to boost the productivity, improve the safety and reduce the

environmental impact of coal mining. According to IEA communication with experts in

the field, in some cases the application of data analysis and monitoring has already

improved electric shovel performance by over 10% at zero cost by minimising wait time and truck queuing. In others, increasing fill factors, and hence reduced cycle times,

have increased shovel productivity (i.e. the volume that a shovel loads per day) by over 10%. Data analysis, monitoring and computer-aided modelling have increased

continuous miner productivity by over 20% in some mines,5 while data analysis and

modelling have reduced offline maintenance time by over 35% in certain underground operations.

Among the benefits of better monitoring and data processing is a better work

environment. For example, noise and dust can be monitored in real time and computers can help to understand where and why high levels of emissions occur. This information can help in the timely implementation of prevention and mitigation of these issues.

Barriers to further digitalization Digitalization of coal will undoubtedly progress in the coming years, but may not revolutionise the way the industry operates. This is due, at least in part, to the expectation that, absent deployment of carbon capture and storage (CCS, see

Box 7.1), coal demand may fall in the long term as a result of concerns about air quality and climate change. The current debate about coal in certain regions is more about stranded and risky assets rather than about increasing the recoverable resource base. Without CCS, the existing recoverable resource base for coal is greater than the

4

Targeted maintenance to address a potential issue before it worsens, causes collateral damage or leads to downtime. Predictive

5

A continuous miner is a common piece of equipment used in underground mining.

Impact of digitalization on oil and gas, coal, and power supply

75

© OECD/IEA, 2017

maintenance is done based on the actual condition of the equipment rather than based on working hours, time lapsed, etc.

“carbon budget” compatible with the current internationally agreed climate targets under the Paris Agreement.

The other main barriers to digitalization of the coal industry relate to the availability of

financial resources and public opinion. Perceptions of coal as a risky asset make it very difficult to channel money to the sector in certain regions, which can hamper

wide-scale deployment of new technologies, including digitalization. In addition, in

many countries the public is highly sensitive to threats to employment in the sector. As coal mining and, in particular, underground coal mining, is a labour-intensive business, digital technologies can be perceived as a threat to jobs.

Nonetheless, a number of factors may help overcome these barriers. Cost inflation, worsening geological conditions and increasing competition from other energy sources

will most likely push coal producers into seeking ways of lowering costs, including

through the application of digital technologies. In any case, their deployment is anticipated in indirect ways in the sector, as the use of digital technologies spreads

throughout the broader economy. In several countries, especially China and India, coal

is likely to remain an important resource, and will lead the wide-scale application of the most advanced digital technologies in the industry. Furthermore, mining of other resources besides coal will be needed for the foreseeable future, so it is likely that any general progress achieved in mining will also be applied to coal.

Power sector Technological innovation has always been at the core of power sector development,

starting with early competition between alternating current and direct current. Over time, the range of available power generation technologies has expanded from a small

set of designs, most of which burn fossil fuels, to nuclear, hydropower, bioenergy, solar, wind and geothermal.

Digitalization of power sector assets represents another step in the process of technological innovation. Digital technologies are widely available and their costs have fallen dramatically in recent years, particularly for sensors and supporting software. Applied to the current structure and operation of power systems, the data and

analytics components of digitalization can provide a series of improvements, helping to reduce costs for existing and new projects across all types of power generation, improving their technical performance and competitiveness.

The role of renewable energy is expanding in many power systems around the world. A key opportunity related to the deployment of digital technologies is the enhanced

ability of power systems to integrate increasing shares of variable renewables (see

76

Impact of digitalization on oil and gas, coal, and power supply

© OECD/IEA, 2017

Chapter 4).

For all types of power plants, as well as for transmission and distribution networks,

digital technologies offer an array of opportunities to improve performance for the

benefit of individual companies, the system as a whole, energy consumers and the environment (Figure 3.5). The connectivity component of digitalization has the potential to reshape the power sector by connecting power supply with key demand sectors such as transport, buildings and industry (see Chapter 4). Figure 3.5 Impact of digitalization on electricity sector assets Digitalization in the power sector

Connectivity:

Data and analytics:

Enables broad structural change (see Chapter 4)

Provide for predictive maintenance, planning and operational changes

Reduced O&M costs

Reduced investment needs

Improved efficiencies

Reduced fuel consumption and costs

Reduced CO2 emissions

Reduction of unplanned outages

Extended asset lifetimes

Improved system stability

Reduced investment needs

Key message: Digitalization in the power sector has the potential to bring benefits to the owners of power sector assets, the wider electricity system, consumers and the environment.

Notes: green : financial benefits for asset owner; red : system benefits, consumer benefits; blue : global environmental benefits; CO2 = carbon dioxide.

Data and analytics in power plants and electricity networks Digitalization in the power sector is wide-ranging, including the gathering of digital data about the state and performance of power sector assets, processing the

information through software platforms and, ultimately, altering activities. It provides data and analytics that influence real-time actions by the owners and operators of the

assets, including minor operational changes, allowing them to alter activities to avoid

excessive stress on the asset. This, in turn, leads to improved system efficiency and lower costs.

In power generation, digital sensors – which can number several thousand for a single

power plant – can be added to an existing plant or incorporated into the design of new components of power plants (e.g. temperature readings), as well as input flows (e.g. fuel, air or cooling water) and output flows of electricity or emissions. Impact of digitalization on oil and gas, coal, and power supply

77

© OECD/IEA, 2017

ones. Sensors provide real-time information concerning the state of various

In electricity networks, sensors provide information on the state of transmission and distribution lines at various points (e.g. temperature, voltage or current). This

information can be stored or transmitted to relevant parties for use in managing the

grid and operating generating plant. While collecting information on the state of the

power system is not new, digital sensors can provide more information on a continuous and real-time basis, something that would have been too costly or difficult to collect previously.

Sensors can also collect complementary information, such as ambient temperature,

which, combined with the electrical information, can enable better management of the

system. For example, high ambient temperatures may increase the physical strain on power plants, a factor that needs to be taken into consideration in operational decisions about ramping up or down the output of the power plant.

Potential benefits of digital data and analytics Digital data and analytics can reduce power system costs in at least four ways: by

reducing O&M costs; improving power plant and network efficiency; reducing unplanned outages and downtime; and extending the operational lifetime of assets.

The IEA estimates that the overall savings from these digitally enabled measures could be in the order of USD 80 billion per year over 2016-40, or about 5% of total annual

power generation costs based on the enhanced global deployment of available digital

technologies to all power plants and network infrastructure (Figure 3.6).

Figure 3.6 Illustrative potential worldwide cost savings from enhanced digitalization in power plants and electricity networks over 2016-40 40

USD billion (2016)

30 20 OPEX

10 0

CAPEX 5% lower O&M costs

Efficiency: 5% more electricity output per unit of fuel

Efficiency: 5% lower total network losses

5-yr life extension for power plants

5-yr life extension for networks

Key message: Cumulative savings from the widespread use of digital data and analytics in power plants and electricity networks could average around USD 80 billion per year.

Notes: Assumes the enhanced global deployment of existing digital technologies to all power plants and network

78

Impact of digitalization on oil and gas, coal, and power supply

© OECD/IEA, 2017

infrastructure; CAPEX = capital expenditure; OPEX = operational expenditure; yr = year.

Reduced O&M costs Digital data and analytics can reduce O&M costs, enabling predictive maintenance,6

which can lower costs for the owner of plants and networks and ultimately the price of

electricity for end users. This type of maintenance is only possible with detailed component-level information delivered in real time, a prohibitively difficult and expensive task without digital sensors.

Globally, the IEA estimates that O&M costs in power generation and electricity networks were just over USD 300 billion in 2016. Over the period to 2040, a 5%

reduction in O&M costs achieved through digitalization could save companies, and

ultimately consumers, an average of close to USD 20 billion per year. For renewable energy technologies, improved maintenance also supports better performance throughout the life of the project. Improved efficiencies Digital data and analytics can help achieve greater efficiencies through improved planning, improved efficiency of combustion in power plants and lower loss rates in networks, as well as better project design throughout the overall power system. In

conventional power plants, efficiency gains reduce the amount of fuel consumed and

CO2 emitted per unit of output. This can be made possible through better maintenance

and other improvements, such as optimising the mix of fuel and air in the combustion

process. Consider the example of a 5% increase in the electricity output per unit of fuel input (equal to a 2 percentage point gain in power plant efficiency) for all subcritical

and supercritical coal-fired power plants built over the past 20 years (over 900 gigawatts worldwide). In such a case, these power plants – which generated 4 500

terawatt hours in 2015 – would have consumed 70 million tonnes less coal, emitted 200 million tonnes less CO2 less (0.6% of global energy-related CO2 emissions) and

saved more than USD 4 billion in fuel costs. Industry estimates suggest that higher

efficiency gains are possible for individual power plants using digital technologies (Annunziata et al., 2016).

In electricity networks, efficiency gains can be achieved by lowering the rate of losses

in the delivery of power to consumers, for example through remote monitoring that

allows equipment to be operated more efficiently and closer to its optimal conditions, and flows and bottlenecks to be better managed by grid operators. Additional gains are possible through enhanced connectivity (see Chapter 4). In addition, losses from theft

can be identified more accurately through monitoring enabled by smart metering.

Predictive maintenance in the power sector refers to the ability to perform targeted maintenance of power plants and electricity

networks in order to address a potential issue before it worsens, causes collateral damage or leads to downtime for the asset.

Impact of digitalization on oil and gas, coal, and power supply

79

© OECD/IEA, 2017

6

Reduced losses mean less electricity needs to be generated to meet demand. As a

result, fuel consumption and associated costs are reduced, along with CO2 and other

emissions and, ultimately, electricity prices to end users. Globally, losses in the

transmission and distribution of electricity account for 8% of total electricity generation, equivalent to the amount of electricity consumed in global iron and steel

production, lighting and cooking combined today. Losses in transmission and

distribution are very large in many developing countries, exceeding 15% of total

generation in Africa, Latin America and India.

Data and analytics can improve planning, as well as project design throughout the overall power system, making them more efficient and less costly. Investment in new

power plants and electricity networks can be better co-ordinated, strategically located

to supplement existing infrastructure and meet the dynamic needs of consumers at

minimum cost. In addition, data and analytics can also be applied to the design of individual projects in order to maximise the value of the project to the system. For example, sophisticated designs for new wind power projects are now possible based

on digital data, enabling the optimal choice of turbine technology and distribution of

turbines to maximise the potential of resources while minimising integration challenges.

Reduced frequency of unplanned outages Digital data and analytics can also reduce the frequency of unplanned outages through

better monitoring and predictive maintenance, as well as limit the duration of downtime by rapidly identifying the point of failure. This reduces costs and increases

the resilience and reliability of supply. Network failures are expensive, both for the utility and for the economy. For example, power supply interruptions in the United

States alone have been estimated to cost around USD 100 billion per year

(LaCommare and Eto, 2006). Emerging economies generally suffer most from frequent power cuts.

Extension of operational lifetimes In the long term, one of the most important potential benefits of digitalization in the power sector is likely to be the possibility of extending the operational lifetime of power plants and network components, through improved maintenance and reduced physical stresses on the equipment. Longer lifetimes would increase revenues to the owners of the assets while reducing the investment requirements of the power system

as a whole. This would lower prices to end users. A potential downside of extending the lifetime of fossil-fuelled power plants is that it could slow a transition to cleaner

sources of electricity, raising an additional challenge to curbing CO2 and local pollutant

80

Impact of digitalization on oil and gas, coal, and power supply

© OECD/IEA, 2017

emissions.

While the full extent of the lifetime extensions that could be achieved through

digitalization is not yet known, especially as many technologies are just emerging, the

potential cost savings are substantial. Electricity supply infrastructure is generally long-lived, with expected lifetimes ranging from 20 or 25 years for wind turbines and

solar photovoltaics, to 40 to 50 years for fossil-fuelled power plants and network

infrastructure, and up to 70 years for hydropower assets. Were the lifetime of all the power assets in the world to be extended by five years,7 the IEA estimates that close to

USD 1.3 trillion of cumulative investment could be deferred over 2016-40, or about 7%

of total power sector investment in the Central Scenario.8 On average, investment in

power plants would be reduced by USD 34 billion per year and that in networks by USD 20 billion per year.

Barriers to further digitalization There are few barriers to the deployment of digital technologies in the electricity sector compared with most other energy sectors. Companies that own and operate power

plants have direct financial incentives to invest, as they benefit directly from fuel cost

savings, reduced O&M spending and the competitive advantage of lower overall operating costs in the wholesale market.

Nonetheless, misaligned financial incentives can present a significant barrier for fully

exploiting the potential of data and analytics in electricity networks. This is especially

so in regulated markets where total investment in physical assets is the basis of revenues, while investments in digital technologies are not incentivised.

Difficulties in obtaining data may also present a significant barrier to using digital

technologies to improve the planning of power systems. Information concerning individual power plants and network infrastructure is essential, but asset owners and operators may not be willing to share it for reasons of commercial confidentiality. One way to solve this problem would be for regulators to introduce disclosure requirements

that protect confidentiality.

7

In this estimate, assets must operate for at least 25 years in order to gain the full benefit of digitalization. For those with fewer

than 25 operational years remaining, their operational lives are extended in proportion, e.g. where a plant has ten more years of expected lifetime, its operations are extended by two years.

8

As defined in Chapter 1, the IEA Central Scenario describes the pathway for energy markets and technological progress based on

It is broadly in line with the New Policies Scenario of the IEA World Energy Outlook and with the Reference Technology Scenario in Energy Technology Perspectives 2017. This should not be interpreted as a forecast.

Impact of digitalization on oil and gas, coal, and power supply

81

© OECD/IEA, 2017

the continuation of existing energy and climate policies and measures, and to a certain extent announced commitments and plans.

References Annunziata, M. et al. (2016). Powering the Future: Leading the Digital Transformation of the Power

Industry. GE.

IEA (International Energy Agency) (2016). World Energy Outlook. OECD/IEA, Paris 2016. LaCommare, K. H., & Eto, J. H. (2006). Cost of power interruptions to electricity consumers in the United States (US). Energy, 31(12), 1845-1855.

82

Impact of digitalization on oil and gas, coal, and power supply

© OECD/IEA, 2017

NDRC/NEA (2016). Energy Technology Innovation Action Plan 2016-2030 of China.

Chapter 4

System-wide impacts: From energy silos to digitally interconnected systems Highlights •

The greatest transformational potential for digitalization in energy is its ability to break down boundaries between energy sectors, increasing flexibility and enabling integration across entire systems.



The electricity sector is at the heart of this transformation. Together with the progressive electrification of the energy system and the growth of decentralised sources of power, digitalization is blurring the distinction between supply and demand, and creating opportunities for consumers

to interact directly in balancing demand with supply in real time. Throughout this process, centralised transmission networks will continue to be the backbone supporting the transition, balancing the overall system.



Digitalization can enable the active participation of consumers from all demand sectors in energy system operations. By 2040, 1 billion households and 11 billion smart appliances could actively participate in interconnected electricity systems, allowing them to alter when they

draw electricity from the grid. This smart demand response could provide 185 GW of system flexibility – comparable to the currently installed

electricity supply capacity of Italy and Australia combined. This could save USD 270 billion of investment in new electricity infrastructure that would otherwise have been needed to ensure security of supply.



Digitalization can help integrate variable renewables by enabling grids to better match energy demand to when the sun is shining and the wind is blowing. In the European Union alone, increased storage and demand response could reduce the curtailment of solar photovoltaic (PV) and

wind power from 7% to 1.6% in 2040, and avoid about 30 million tonnes (Mt) of carbon dioxide (CO2) emissions in 2040.

The roll-out of “smart charging” of electric vehicles (EVs), which shifts demand to off-peak times, could save between USD 100 billion and

System-wide impacts: From energy silos to digitally interconnected systems

83

© OECD/IEA, 2017



USD 280 billion (depending on the number of EVs deployed) in avoided investment in new electricity infrastructure over 2016-40.



Digitalization can facilitate larger shares of distributed energy resources, turning consumers into “prosumers”; new tools such as blockchain may facilitate such local energy trading systems.



Fundamental changes to policy and regulation will be required to ensure that the benefits of the digital transformation of electricity are fully realised and the risks minimised.

Digitalization is transforming the way the electricity system functions Beyond the direct efficiency gains and cost savings described in the previous two chapters, digitalization holds the potential to catalyse more fundamental, system-wide

changes. Electricity is likely to be the first energy sector to see the impact of this deeper transformation and the one that will ultimately be most affected. Traditionally, electricity is generated in large power plants, transferred through transmission and

distribution networks and delivered to end users in the residential, commercial, industrial and transport sectors (Figure 4.1). This model is set to change dramatically.

By matching demand to the needs of the overall system in real time, digitalization

opens up the opportunity for millions of consumers as well as producers to sell

electricity or provide valuable services to the grid. Connectivity is the key factor. It permits the linking, monitoring, aggregation and control of large numbers of individual energy-producing units and pieces of consuming equipment. These assets can be big or small, e.g. a rooftop solar PV system in a home, a boiler on an industrial site or an EV.

As digitalization advances, a highly interconnected system can emerge, blurring the

distinction between traditional suppliers and consumers, with increasing opportunities

for more local trade of energy and grid services (Figure 4.2). As this physical infrastructure evolves and stakeholders’ roles change, centralised grids and the

owners and operators of transmission networks will continue to provide the backbone

84

System-wide impacts: From energy silos to digitally interconnected systems

© OECD/IEA, 2017

that balances the overall electricity system.

Figure 4.1 Traditional structure of electricity sector

Transport

Power generation transmission and distribution

Fossil fuels

Buildings

Data/analytics Electricity Fuels

Industry

Key message: Data and analytics can improve performance and enable cost savings, but, without connectivity, do not fundamentally change the way the electricity sector functions.

Figure 4.2 The role of digitalization in reshaping the electricity sector

Homes

Transport

Connectivity/data Energy flow

Power generation transmission and distribution

Industry

Key message: Connectivity, combined with electrification and decentralisation, holds the potential to create a highly interconnected system, transforming the way electricity is supplied and consumed. System-wide impacts: From energy silos to digitally interconnected systems

85

© OECD/IEA, 2017

Businesses

Digitalization is one part of this process. The continuing electrification of energy

services across all end-use sectors, notably transport, and the growth of decentralised

sources of power – which would happen even without digital technologies – are the other drivers. But digitalization – particularly the growth in connectivity among

producers, grid operators and end users – is supporting these trends and helping to accelerate the transformation of the electricity system and the establishment of new

business models. By allowing for the exchange of operational information in real time between equipment anywhere in the energy system, inefficiencies within each sector

are removed, improving reliability and lowering costs, as consumers and producers respond instantaneously to changing market conditions (Figure 4.3).

Figure 4.3 Possible steps in the digital transformation of the electricity system

Demand

From energy silos

Inflexible demand

To digitally interconnected systems

Demand response large consumers

Trends

Supply

Sector integration

Centralised supply

Industry

Centralised supply

(Competitive markets)

Aggregators

Aggregators

(low share of demand) Buildings

Transport

Distributed sources (low share of supply)

Demand integration

Prosumers Smart grid

Transactive energy Vehicle to grid

Blockchain Decentralised supply

High connectivity - big data and analytics

Low connectivity Low electrification

Electrification of buildings, electric vehicles

Centralised system

Decentralised sources

Key message: The deployment of digital technologies is creating a more interconnected and responsive electricity system, with the potential to help increase flexibility, efficiency and reliability.

The low-voltage electricity distribution network – the section of the grid closest to

consumers – will be central to realising the potential of digital technologies to

transform the electricity system. Distribution grids will increasingly have to take on the

role of balancing supply and demand, including hosting more heterogeneous, distributed energy resources, including EVs, solar PV and battery storage. Distribution

grids, however, are nested within broader networks and the overall transmission grid, which will continue to provide overall balancing of the system, and form the backbone

for the transformations described in this chapter. Transmission networks themselves will benefit from the increased capability to manage operations brought about by

digitalization, and will have to monitor and interact more strongly with distribution grids. The interface between operators of transmission and distribution grids will thus have to be strengthened. This will require technical, market and institutional changes

the years to come as digital technology improves, costs fall and opportunities for costeffective deployment expand. 86

System-wide impacts: From energy silos to digitally interconnected systems

© OECD/IEA, 2017

(Box 4.1). The process is at an early stage in most countries, but is set to accelerate in

Box 4.1

From one-way electricity grids to smart energy systems

The regulatory, financial and institutional structures in place around the world today allow grid owners and operators to plan investments and operate in a way that meets demand and service quality requirements while recovering costs. Traditional grids are based on a technical paradigm where the low- and medium-voltage parts – the distribution grids – traditionally pass on power generated elsewhere in one direction to consumers, who are passive recipients. Impending transformation in the way the electricity system functions will call for fundamental changes at technical, economic and institutional levels. All of the deeper cross-sectoral changes shaped by digitalization will be hosted and supported by the physical grid. On the demand side, digitalization enables consumers to become direct participants in energy markets through demand response, helping to balance supply and demand locally. On the supply side, consumers can become local producers and periodic exporters of energy to the grid. With more distributed production (electricity generated by small-scale plant and injected into the local distribution grid) and local trade, distribution networks will have to take on responsibility for balancing supply and demand locally, as well as providing security and reliability to the overall system. This role is currently performed by large-scale transmission network operators. The latter will continue to have a fundamental role in providing the physical backbone of the overall system and participating actively in efficient data exchange and market design. For instance, data exchange platforms – hubs where information can be made available to interested parties – will be necessary to maximise the use of distributed or centralised energy where it is most valuable and to devise new business models to make the new arrangements work. Both transmission and distribution network operators will have to strengthen co-operation and information exchange. Ultimately, the role of networks, and particularly local distribution grids, will have to evolve from the traditional one of delivering energy one-way into providing, on the one hand, local and regional hubs for active trade in electricity produced in centralised and distributed locations, and on the other, services such as reliability, security and flexibility.

The continuing electrification of the global economy makes the transformation of the

electricity system all the more important and opens up greater opportunities for exploiting new digital technologies. As economies mature, the share of electricity in

the overall energy mix typically increases across all end-use sectors. Electrification of end uses, such as EVs, heat pumps and electricity-based production of metals, can

also be a means of decarbonising (to the extent that power is generated with clean energy sources) and curbing local air pollution. In fact, low-carbon technologies represent almost two-thirds of electricity generation in advanced economies in 2040 under the Central Scenario (IEA, 2016). 1

As defined in Chapter 1, the IEA Central Scenario describes the pathway for energy markets and technological progress based on

the continuation of existing energy and climate policies and measures, and to a certain extent announced commitments and plans.

System-wide impacts: From energy silos to digitally interconnected systems

87

© OECD/IEA, 2017

1

Digitalization of electricity is part of a broader process of technological change, which

– along with changes in regulatory and market design – is both driving and responding

to the underlying trends in electricity demand and supply. This is reflected in the cost reductions that have been achieved across a range of electricity-related technologies

over the last decade. For example, the unit cost of small-scale PV has dropped by a factor of five since 2008, sensors by more than 95% and battery storage by more than

two-thirds (mostly thanks to the deployment of EVs) (Figure 4.4). The average cost of

a smart meter has dropped by about one-quarter, with nearly 600 million smart meters being deployed globally.

Figure 4.4 Unit costs of key emerging electricity technologies 100

Cost reduction (2008 = 100)

90 80 70

Smart meters

60

Smart appliances

50 40

EV batteries

30 Utility-scale PV

20 10

Sensors

0 2008

2010

2012

2014

2016

Key message: Technology cost reduction is a key driver enhancing connectivity throughout the electricity sector.

Sources: IEA analysis based on Bloomberg New Energy Finance (2017); Holdowsky et al. (2015); IEA (2017a; 2017b; 2017c); Navigant Research (2017).

The four main elements of the transformation of the electricity system that digital

technologies can make possible or support are: 1) smart demand response; 2) the integration of variable renewable energy sources; 3) the implementation of smart charging for EVs; and 4) the emergence of small-scale distributed electricity resources

such as household solar PV. They are interlinked as, for example, demand response

will be critical to providing the flexibility needed to integrate more generation from variable renewables. All four of these elements are expected to contribute significantly to the emergence of digitally interconnected electricity systems (Figure 4.5).

Energy Technology Perspectives 2017. This should not be interpreted as a forecast.

88

System-wide impacts: From energy silos to digitally interconnected systems

© OECD/IEA, 2017

It is broadly in line with the New Policies Scenario of the IEA World Energy Outlook and with the Reference Technology Scenario in

Figure 4.5 Four key elements in the creation of a digitally interconnected electricity system Worldwide

Capacity generation requirement 300 GW

6 900 TWh 3 900 TWh

140 GW

Standard charging

190 GW

Smart charging

75 GW

Demand response potential

European Union

2040

2015

EV smart charging

150 million EVs

Curtailment of wind and solar PV

500 million EVs

Small-scale solar PV

7%

1000 TWh

1.6%

Variable renewables integration

2040

2040 with flexibility measures

Distributed generation and prosumers

2015

2040

Key message: Digitalization is set to greatly enhance demand flexibility, the integration of variable renewables, smart changing for EVs and distributed generation. Sources: Analysis based IEA (2016; 2017d).

Smart demand response Digitalization allows for a greater number of electricity consumers to respond flexibly

to signals from the system – a mechanism known as demand response. The aim of

demand response is to maintain security of supply at least cost. Digital connectivity allows appliances and equipment to be monitored continuously and connected to the

grid. The information is collected and then used to shape demand to optimally match it with available supply.

During the hours when supply is scarce or electricity networks are congested, connected devices such as smart electric heaters and air conditioners, industrial

boilers and smart home appliances can be switched off or run at lower load automatically. These connected devices can reduce or shift consumption to other

periods when supply is abundant, for example, when the sun shines, the wind blows or

when there are no technical problems with the electricity grid (Figure 4.6). Sophisticated new digital technologies allow this to be done in a way that does not affect the comfort of the consumer. The end user is typically compensated for the break in service through a price incentive.

Demand response has been in operation in some regions for many years, but has remained very limited in scale and largely restricted to large industrial consumers.

Today, only 1% of demand globally, or about 40 GW of capacity, is able to directly group of large consumers, typically energy-intensive industries, are offered financial incentives in return for accepting the possibility of their supply being interrupted at

System-wide impacts: From energy silos to digitally interconnected systems

89

© OECD/IEA, 2017

respond to shortages or excess supply (Navigant Research, 2017). Generally, a small

short notice when the grid operator deems it necessary to ensure the security of the electricity system (Box 4.2).

Figure 4.6 Impact of demand response on the daily load curve Inflexible demand

Demand response Excess generation Demand response Variable renewables Dispatchable generation Total demand

0h

8h

16h

24h 0h

8h

16h

24h

Key message: Demand response, facilitated by digital technology, shifts consumption to those hours with a surplus of electricity supply.

Box 4.2

Interruption of large electricity consumers as the first form of demand response

The most basic form of demand response is through programmes focused on interrupting demand

from large consumers, directed mainly at large industries. This type of demand response – also called interruptible service – is a demand-side management tool aimed at providing flexible and

rapid response to the needs of the grid operator in situations of imbalance between generation and

demand. For example, an exceptional peak in consumption due to extreme weather conditions, unplanned outages in supply, or a combination of the two could compromise the security and regular operation of the grid and may even result in an electricity blackout.

As a preventive measure, large electricity consumers, such as energy-intensive industries, in

response to an order issued by the system operator, reduce their consumption to help maintain the balance between generation and demand, so that other consumers are not affected. These large energy consumers receive financial rewards in return for providing this service. In most mature electricity markets, the grid operator allocates the interruptibility services through competitive auction mechanisms.

A combination of technologies at the intersection between digitalization and energy

can help demand response penetrate to a larger share of consumption. Greater

automation, the diffusion of Internet of things (IoT) devices in the residential and

commercial sector (e.g. smart thermostats directly connected to the power market and systems will all allow further integration across demand and supply, and unlock

greater cost savings for individual consumers and the system overall (Gils, 2014; 90

System-wide impacts: From energy silos to digitally interconnected systems

© OECD/IEA, 2017

to weather forecast providers) and higher deployment of EVs and smart charging

O’Connell et al., 2015. Artificial intelligence (AI) could potentially greatly enhance the

efficiency and effectiveness of demand response in the longer term.

Globally, the IEA estimates that about 3 900 TWh of current electricity consumption is technically available today for demand response, and it is expected to almost double by 2040 to about 6 900 TWh, or nearly 20% of electricity consumption worldwide. The

potential for demand response varies by region and sector, but in all regions, most of

the current and future technical potential at lower cost lies in the buildings sector, especially in space and water heating and cooling. Electricity demand for space heating and cooling can be shifted over a certain number of hours, the extent

depending on the thermal inertia. Most of the remaining potential in buildings is

related to electricity used for large appliances, such as washing machines,

refrigerators, dishwashers and clothes dryers. EVs are expected to become participants in demand response programmes over time (IEA, 2016).

The benefits of increased demand response are enormous. In the IEA Central

Scenario, the implementation of the full technical potential of demand response (6 900 TWh) results in about 185 GW of additional flexibility for the electricity system

globally in 2040 – roughly equivalent to the currently installed electricity supply

capacity of Italy and Australia combined. This amount of flexibility would avoid a cumulative USD 270 billion (in 2016 dollars) of investment in new electricity infrastructure (new power-generation capacity and transmission and distribution). As

the bulk of demand response potential is in the buildings sector, almost 1 billion households and 11 billion connected appliances participate in demand response programmes by 2040 under the Central Scenario (Figure 4.7). Large commercial

buildings, such as supermarkets, hotels and offices, industry and EVs can also play a

significant role.

Figure 4.7 Global potential of demand response and its implications Theoretical potential and flexibility

6 900 TWh 2040 today

185 GW

Implications, 2040 1 billion participating

11 billion

Total

3 900 TWh

Demand response

Participating

System flexibility

Households

Connected devices

Key message: The largest potential for demand response lies in the buildings sector, with 1 billion Sources: Analysis based IEA (2016; 2017d).

System-wide impacts: From energy silos to digitally interconnected systems

91

© OECD/IEA, 2017

households and 11 billion smart appliances expected to be contributing by 2040.

While the potential benefits to the entire system are enormous, the financial savings to

individuals may not be sufficient to persuade a large proportion of consumers to

participate in demand response on their own, as they typically account for a small

percentage of the average electricity bill in developed countries. This highlights the

need for regulatory and policy frameworks that distribute the costs and benefits adequately through incentives for both the system operator and consumers, while ensuring the cybersecurity of the grid.2

The role of aggregators may also prove to be key. Aggregators – also known as

demand response providers – gather consumer demand of any type, as well as the

supply of distributed producers such as renewables-based power plants, to provide balancing services to the grid by adjusting power demand and/or shifting loads at short notice. The “pool” of aggregated load is managed as a single flexible consumption unit – equivalent to a virtual power plant – and sold to the markets or to

the grid operator. In this way, the aggregator provides an interface between large

numbers of individual consumers/producers and power markets/grid operators (Box 4.3). Box 4.3

Demand response market and aggregators

The demand response market has witnessed steady growth in recent years and is expected to

further increase into the future. Globally, the market is worth more than USD 5 billion in 2014 and is expected to grow fivefold by 2022 (Grand View Research, 2016). As of today, North America remains the dominant market – mainly concentrated in California. The Asia-Pacific region currently accounts for more than 10%, with increasing penetration of smart meters being an important driver for further growth.

In 2014, over 50% of total revenues from demand response came from industrial applications. The

residential segment is poised to grow at double-digit percentages, with an increasing number of consumers involved. Despite the growing interest from traditional utilities and investors, the demand

response market is still highly fragmented and particularly region-oriented, with some aggregators having a strong position over a particular region but little influence globally. That said, key participants have started expanding globally through mergers and acquisitions3 and upcoming smart

grid projects in countries such as Australia, India, the People’s Republic of China (hereafter,

“China”) and Japan could attract more global players. The aggregator business model depends on a regulatory framework that allows the grid operator to buy demand flexibility from a third party. The

relatively underdeveloped state of the market and its geographic distribution fully reflect this circumstance. The further development and stability of such regulatory frameworks will thus be key for the future growth of the aggregator model.

2

The risk of intentional and unintentional threats to system infrastructure has to be assessed and managed in order to maintain

3

For example, in June 2017 EnerNOC, one of the main demand response providers, was acquired by Enel Group.

92

System-wide impacts: From energy silos to digitally interconnected systems

© OECD/IEA, 2017

the stability and the reliability of the electricity system (see Chapter 6).

Integration of variable renewables Variable renewables, such as solar PV and wind power, are inherently intermittent,

with output that can be predicted accurately only a few hours to days in advance. These technical properties make variable renewables less predictable than traditional

power plants. Under the IEA 450 Scenario,4 more than one-quarter of global electricity is generated by wind and solar PV by 2040 (IEA, 2016). In many regions such as the European Union, this share is much higher in 2040 under this scenario (Box 4.4).

Beyond a certain threshold of market penetration by variable renewables, having

enough flexibility in the power system becomes crucial for maintaining reliability and cost-effectiveness, affecting how infrastructure is planned and operated. While other

sources of flexibility are available (e.g. ramping up and down production from conventional energy sources), there is a vast untapped potential for digitally enabled

demand response that, together with storage, can cost-effectively help to

accommodate

a

higher

share

of

decarbonisation of the electricity sector. Box 4.4

variable

renewables

and

accelerate

the

Digitalization can help to decarbonise the EU electricity sector

In the European Union, variable renewables are playing a pivotal role in decarbonising the power

sector. Under the 450 Scenario, over 570 GW of wind and solar PV installations will need to come online by 2040 (IEA, 2016). In this scenario, renewables make up more than 60% of total electricity

generation, reaching almost 2 100 TWh in 2040, more than twice today’s level. Wind and solar PV account for almost 60% (1 250 TWh) of the overall renewable generation, making the European

Union one of the regions with the largest combined share of wind and solar PV in total power generation. Improving cross-border interconnections within and between countries is one of the main actions

undertaken by the European Union to give flexibility to the power system and accommodate a greater share of variable renewables.5 Nevertheless, additional flexibility measures are needed as soon as the share of variable renewables in the generation mix increases beyond 27%, on average,

across the continent. These are required to avoid periodic curtailment of solar PV and wind power

generation at times when total supply from those sources exceeds demand. In the absence of additional flexibility measures, total curtailment would exceed 85 TWh in 2040, which is equivalent to nearly 7% of the combined generation from wind and solar PV. The introduction of measures to prevent curtailment will be critical to increased deployment of wind

and solar PV. The IEA projects that in 2040 digitally enabled demand response measures can reduce

4

The 450 Scenario sets out an energy pathway consistent with the goal of limiting the global increase in temperature to 2oC by

limiting concentration of greenhouse gases in the atmosphere to around 450 parts per million of CO2.

Cross-border power interconnections between countries are equivalent to 11% of the installed generation capacity across the

European Union.

System-wide impacts: From energy silos to digitally interconnected systems

93

© OECD/IEA, 2017

5

curtailment by some 22 TWh, and a further 45 TWh can be saved through additional energy storage,

a substantial share of which would be digitally enabled battery storage, limiting curtailment to 1.6% of total wind and solar PV generation. This allows the system to accommodate 67 TWh of additional generation from variable renewables and to avoid about 30 Mt of CO2 emissions (Figure 4.8). Figure 4.8

Flexibility measures and their implications in the European Union in 2040

Flexibility measures

Implications

Curtailment Demand response

22 TWh

Storage technologies

85 TWh

No flexibility

Key message:

67 TWh

30 Mt

45 TWh

With flexibility

Additional solar and wind generation

Avoided CO2 emissions

Digitally enabled demand response and storage technologies are expected to

drastically reduce curtailments of solar and wind, boosting their share of power generation and reducing CO2 emissions.

Source: Analysis based on IEA (2016).

The growing role of EVs EVs have the potential to revolutionise passenger car transport, with far-reaching

implications. EVs start from a very low base, however: in 2016, the global fleet of EVs –

including plug-in hybrids and battery EVs – doubled to two million, or around a tenth of

1% of the total vehicle fleet (IEA, 2017d). The way in which EV charging infrastructure

is deployed and the technologies that are used could have significant impacts on energy systems.

Providing enough power to fully top up the battery of a typical EV in four hours would

require nearly 9 kilowatts (kW) of capacity today6 – equivalent to the current peak demand of an average Californian household. Shifting the charging of EVs to when

demand is low and there is abundant low-cost generation (such as wind and solar power) could reduce the need for additional generation capacity to meet EV demand, delivering significant savings for the system.

The deployment of EVs requires new investment to build charging infrastructure. EVs

6

Assuming a 35 kilowatt hour EV battery.

94

System-wide impacts: From energy silos to digitally interconnected systems

© OECD/IEA, 2017

represent mobile demand for electricity, and as such can theoretically be charged at

any point in the day, at low recharging rates either at home or the office, or via faster charging using public charging infrastructure (Eurelectric, 2015; Fitzgerald and Nelder,

2017). When, where and how much power EVs draw from the grid will vary significantly

and will depend not only on drivers’ behaviours, but also on incentives and signals that

could be provided to consumers in the form of monetary savings and to the entire system in the form of greater flexibility and reduced investment needs.

Co-ordination of charging strategies through digital technologies (so-called “smart charging”) will be required to take full advantage of this opportunity. With smart

charging, price and control signals provide incentives for connected EVs to charge

when there is abundant production of low-cost, low-carbon electricity or to stand by when the network is congested. Smart charging requires digital infrastructure to permit

communication between charging points and back-end systems, so as to allow grid operators to send requests to increase or reduce demand at certain times.

In the IEA Central Scenario, global electricity capacity reaches 12 000 GW. The

increase in capacity needs due to EV penetration depends heavily on the penetration of EVs themselves. Due to the uncertainties surrounding EV deployment, assessment is made of a range of impacts at different levels of EV penetration.

In a scenario where 150 million EVs are deployed by 2040, capacity needs for standard

EV charging reaches 140 GW in 2040. This compares with a 75 GW charging capacity requirement were smart charging to be implemented, resulting in a reduction of 65 GW

by 2040. In a more ambitious scenario where the global EV fleet reaches 500 million by 2040, capacity savings increase to 110 GW.

This flexibility provided by smart charging avoids, respectively, USD 100 billion and

USD 280 billion of investment in electricity infrastructure (new power generation

capacity and transmission and distribution), otherwise required to cover the peak in electricity demand from EVs (Figure 4.9). The higher the penetration of EVs, the more significant the savings from deploying them together with digitalized infrastructure. In

the case of 500 million EVs, the savings are equivalent to about two-thirds of the projected investment in all EV charging infrastructure out to 2040.

Smart charging can also provide services to the grid to enhance power quality and reliability, which could increase its value further. Beyond smart charging, deeper

flexibility can also be unlocked through vehicle-to-grid (V2G) technologies that allow for bidirectional charging (Box 4.5).

Where and when the charging of EVs will occur is also critical, especially from the perspective of utilities and distributors. EVs can be charged “behind the meter”,

charging is optimised to occur either at home or at work during the day. Leading

manufacturers of EVs, batteries and PV systems are already aiming to exploit this System-wide impacts: From energy silos to digitally interconnected systems

95

© OECD/IEA, 2017

whereby the additional load is met by rooftop PV systems and/or batteries, and the

potential through mergers and acquisitions. EVs can also be charged on the distribution network at public charging stations. In this case, the revenues of traditional utilities/grid operators would be higher and flexibility services from delayed charging would be more readily available.

Figure 4.9 Impact of smart versus standard charging of EVs EV fleet consumption profile (illustrative)

EVs standard vs. smart charging Capacity requirement 300 GW

140 GW

Standard charging Smart charging

190 GW

75 GW Smart charging

Standard charging

150 million EVs

500 million EVs

Key message: Smart charging infrastructure can reduce peak demand from EV charging by at least 65 GW and result in system savings of USD 100 billion in 2040. Sources: Analysis based IEA (2016; 2017d).

Box 4.5

Vehicle-to-grid: EVs as mobile batteries

As a roaming battery on wheels, an EV can both withdraw electricity from and inject electricity into the grid, thus being able to sell balancing services to the grid operator, or to meet demand within the home when necessary. While the benefits from flexibility provided to the grid could be enormous, technological issues remain, mainly linked to a faster reduction of EV battery lifetimes as a

consequence of the number of charging-discharging cycles that would result from increased use by V2G. Currently, V2G technology is being experimented through pilot projects and small-scale commercial

initiatives. Recent studies have shown that using V2G only to provide valuable grid services such as frequency regulation does not affect EV battery lifetimes. Using V2G, however, to balance residential

energy consumption (also called V2Home) may not be economically attractive compared to other residential storage solutions. The economic use cases for V2G remain uncertain. However, unlocking V2G as an option requires

specific investments and measures, from charging infrastructure to interoperability requirements,

96

System-wide impacts: From energy silos to digitally interconnected systems

© OECD/IEA, 2017

that need to be built into EV deployment from the outset. Early assessment and eventual action will be required to avoid having to retrofit future systems to enable V2G at a cost premium.

Distributed generation, mini-grids and “prosumers” The electricity sector is becoming more decentralised with the proliferation of distributed

energy

resources

connected

directly

to

local

distribution

grids.

Digitalization is allowing customers (in most cases through aggregation) to become more active in adapting their own electricity production (largely from solar PV), use and storage.

The rise of small-scale, distributed generation allows consumers to increasingly have

the choice to buy electricity from a retailer or to produce at least part of it themselves,

becoming “prosumers”. As the costs of solar PV and batteries continue to decline and the share of consumers participating in demand response programmes grows, the

volumes of energy produced and stored “behind” the electricity meter could substantially increase. Withdrawals and injections of electricity from and into the grid can be actively managed using new digital technology.

These developments could transform the way the electricity supply functions. With the growth in small-scale distributed generation, the electricity generated and distributed

through the grid is set to fall as a result of lower net consumption, potentially driving

down the revenues obtained by traditional utilities and grid operators unless welldesigned regulated use-of-system charges are raised. This, in turn, may constrain the ability to invest in maintaining or upgrading infrastructure, or generate distortions whereby consumers fully reliant on the grid could end up subsidising prosumers, who

take less energy from the power network but still need the grid infrastructure to be maintained. To some extent, electrification – particularly from new loads such as EVs –

could help compensate for the reduction in grid-based electricity consumption, but sweeping regulatory changes will be required to maximise these opportunities at

minimum cost (MIT, 2016). Local communities and cities may gradually begin to

leverage some of the opportunities from transactive and community energy (Box 4.6). Box 4.6

Blockchain, transactive energy and peer-to-peer trading

Blockchain – also called distributed ledger technology – first drew attention ten years ago as the basis for the cryptocurrency Bitcoin. Blockchain is a decentralised data structure in which a digital

record of events (such as a transaction, or the generation of a unit of solar power) is collected and

linked by cryptography into a time-stamped “block” together with other events. This block is then

stored collectively as a “chain” on distributed computers. Any participant to a blockchain can read it or add new data.

As no single computer system that could fail or be compromised is relied upon, data written to the

blockchain is very secure against hacking (see Chapter 6). Because blockchains are transparent and trustworthy, they facilitate direct exchanges of value between parties, peer to peer, without the need

transactions can be faster and cheaper than transfers sent through an intermediary (such as an energy exchange). Blockchain transactions can also be automated using “smart contracts” that System-wide impacts: From energy silos to digitally interconnected systems

97

© OECD/IEA, 2017

for a third-party intermediary institution or service provider. In principle, these peer-to-peer

instruct machines to sell or buy among themselves: self-initiating and self-verifying according to pre-determined conditions and preferences, and transferring funds. Both start-up companies and utilities see potential for blockchain to help solve key energy sector challenges, including co-ordination between increasing numbers of heterogeneous devices, owners and operators in smart grids, and the need for low-friction, automated trading to enable flexibility.

Projects testing uses for blockchain in the energy sector increased rapidly in 2015-16. Many focus on customer markets and enabling micro-trading among solar power prosumers.*

LO3 Energy in New York is using blockchain and a microgrid to enable a Brooklyn community to buy and sell locally generated renewable electricity peer to peer within a small neighbourhood. German

start-up StromDAO uses blockchain to create a “virtual power plant” where participants can

self-supply by investing in off-site renewable capacity and reselling this production in a spot market.

Grid tariffs and taxes are built in. Utility company Innogy (formerly RWE) is creating a blockchain

electronic wallet to manage billing for EV charging, also in Germany. Car owners will be able to pay using smart contracts for electric charging at different points, for parking fees and highway tolls. They would also benefit from receiving car-sharing fees.

Although still early-stage and small-scale, projects of this kind suggest that decentralised energy, flexibility from transactive energy and blockchain could develop together to positive effect. Several

factors, however, might limit uptake of energy blockchains. Fully independent peer-to-peer systems such as Brooklyn may not be scalable, not least because larger and more complex blockchains require increasing investments in computing power. Geographically more extensive projects such as StromDAO rely on utility networks, which constrains the negotiability of the prices implemented

within their additional layer of transaction. While the range of applications is huge, the realisable potential of blockchain in energy sector applications is highly uncertain, and solutions can often be

delivered equally effectively using conventional information and communications technology, e.g. transactive energy and associated services such as those proposed by the Innogy project. Note: * This is not all that blockchain could be applied to. Other uses are being tested throughout the energy chain. Peer-to-peer trading and settlement in wholesale power and natural gas markets is being trialled in Europe by Enerchain, backed by companies Enel, Iberdrola, RWE, Total, and Vattenfall (https://enerchain.ponton.de). Start-up Grid Singularity is using blockchain to collect energy generation and grid equipment performance data (http://gridsingularity.com). Volt Markets in the United States uses blockchain to track Renewable Energy Certificates (RECs) (https://voltmarkets.com/blockchain). Millions of solar facilities have joined a project to post live solar data to blockchain for use by scientists and researchers at www.electricchain.org/. In 2014, BAS Nederland was the first energy company in the world to accept Bitcoin for bill payment, since followed by companies in Germany (Enercity), Belgium (Elegant) and Japan (Marubeni).

In more mature markets, the implementation of such microgrids could lead to benefits for the local community: increasing the amount of renewable energy generated, traded

and distributed within their members; developing a connected network of distributed

energy resources that will improve the overall resiliency and efficiency of the grid; and creating financial incentives and business models that encourage community

Despite the expected growth in decentralised generation and storage in more developed energy markets, the majority of electricity systems are likely to remain 98

System-wide impacts: From energy silos to digitally interconnected systems

© OECD/IEA, 2017

investment in renewable sources and energy efficiency.

largely based on centralised generation and a robust transmission and distribution network for the foreseeable future. Moreover, many of the technologies highlighted in

this chapter can be deployed in a more centralised fashion as well. Indeed, tracking solutions such as blockchain applied to transactive energy in distribution grids will need supportive centralised infrastructure.

Decentralised generation and storage may also be a highly attractive method for bringing energy access to those areas of the world currently without sufficient energy

services (see Chapter 7). In fact, it may be that digitally enabled decentralised

solutions are able to provide energy access first, with the grid connection only coming later in time, if ever.

The prospects for centralised versus decentralised solutions will differ significantly from country to country. How much infrastructure is already in place in a given country,

how much new investment is required to meet demand, the existing energy mix and

resource endowments, as well as the current policy framework, all enhance or curb the

potential for digitally enabled decentralisation. In many systems, the role of centralised infrastructure is likely to evolve into a complementary but essential function of

delivering services that decentralised infrastructure cannot provide – reliability and

other services beyond energy-only delivery. In China, for example, large-scale

investments in the transmission network have become necessary to connect remote centralised generation to main demand centres.

Towards ever-smarter energy systems The electricity sector in each country has a unique set of characteristics. How much

infrastructure has already been deployed (and how), the fuel mix in power generation, the abundance of renewable and conventional energy resources, patterns of demand

and the structure of the broader economy all vary across and within countries.

Consequently, there is no unique end point to the digital transformation of the sector.

In fact, no jurisdiction will ever reach an optimal end point, as technologies will continue to develop and policy understanding will continue to evolve.

The cornerstone of the transformation of the electricity sector will be the emergence of new business models that monetise the cross-sectoral linkages explored in this chapter. These are set to reshape the experience of energy consumers, as

digitalization redefines their interaction with energy suppliers. The fundamental shift may be away from energy-only, asset-intensive business models to platforms that

enable the exchange of services (WEF, 2017b). For instance, the sale of electricity, energy management technologies that allow consumers to optimise their in-house

System-wide impacts: From energy silos to digitally interconnected systems

99

© OECD/IEA, 2017

consumption, EV charging and other services can be packaged together.

Traditional utilities, network operators and third parties are already developing joint platforms for decentralised energy that include installing rooftop solar PV and batteries behind the meter, and operating them as “virtual power plants”. Community

energy, peer-to-peer trading or virtual marketplaces are already being tested in a number of pilot projects at varying scales in France, Denmark, the United States, Korea and Japan.

Policy will be key to the pace and success of this transformation. Deployment of digital technologies and decentralised resources, as well as the electrification of transport and other end uses, is already being supported by governments in many countries. This is helping to drive down technology costs and encourage faster diffusion.

Besides physical assets, many countries are also redesigning their electricity market to

create a coherent and well-structured regulatory framework that can unlock the full potential of digitalization while ensuring energy security and a proper functioning of

the market. For example, the European Commission has proposed a series of measures – from a Digital Single Market to “Clean Energy for All Europeans” – to allow

consumers, utilities and other stakeholders to fully benefit from digitalization. Policy co-ordination across sectors and ministries will be fundamental to achieve the full

100

System-wide impacts: From energy silos to digitally interconnected systems

© OECD/IEA, 2017

potential of digitalization and to effectively manage the challenges it engenders.

References Bloomberg New Energy Finance (2017). Utilities, Smart Thermostats and the Connected Home

Opportunity. Bloomberg Finance L.P., New York.

Eurelectric (2015). Smart Charging: Steering the Charge, Driving the Change. A Eurelectric paper, March 2015.

www.eurelectric.org/media/169888/20032015_paper_on_smart_charging_of_electric_vehicles_fi nalpsf-2015-2301-0001-01-e.pdf. Fitzgerald, G. and C. Nelder (2017). From Gas to Grid: Building a Charging Infrastructure to Power

Electric Vehicle Demand. Rocky Mountain Institute, Colorado.

Gils, H.C. (2014). Assessment of the theoretical demand response potential in Europe. Energy, 67, 1-18. Elsevier, https://doi.org/10.1016/j.energy.2014.02.019.

Grand View Research (2016). Smart Demand Response Market Analysis by Application (Residential,

Commercial, Industrial) and Segment Forecasts To 2022. www.grandviewresearch.com/industryanalysis/smart-demand-response-market.

Holdowsky, J. et al. (2015). Inside the Internet of Things. Deloitte University Press.

https://dupress.deloitte.com/content/dam/dup-us-en/articles/iot-primer-iot-technologiesapplications/DUP_1102_InsideTheInternetOfThings.pdf.

IEA (International Energy Agency) (2017a). Renewables 2017. OECD/IEA, Paris. IEA (2017b). Tracking Clean Energy Progress 2017. OECD/IEA, Paris. IEA (2017c). World Energy Investment 2017. OECD/IEA, Paris. IEA (2017d). Energy Technology Perspectives 2017. OECD/IEA, Paris. IEA (2016). World Energy Outlook 2016. OECD/IEA, Paris. MIT (2016). Utility of the Future: An MIT Energy Initiative Response to an Industry in Transition. Massachusettes Institute of Technology, Cambridge, MA. http://energy.mit.edu/wpcontent/uploads/2016/12/Utility-of-the-Future-Full-Report.pdf. Navigant Research (2017). Market data: Demand Response. Global Capacity, Sites, Spending and

Revenue Forecasts. Navigant Consulting, Inc., www.navigantresearch.com/research/market-

data-demand-response.

O'Connell, N., et al. (2015). “On the inclusion of energy-shifting demand response in production

cost models: Methodology and a case study”. National Renewable Energy Laboratory (NREL),

U.S. Department of Energy Office of Energy Efficiency & Renewable Energy, Washington, D.C.,

System-wide impacts: From energy silos to digitally interconnected systems

101

© OECD/IEA, 2017

http://orbit.dtu.dk/files/118476222/NREL_Official_Report.pdf.

© OECD/IEA, 2017

Chapter 5

Energy use by information and communications technologies Highlights •

Information and communications technologies (ICT) – including data centres, data networks and connected devices – have emerged as an important source of energy demand.



As billions more devices and machines are connected over the coming years, they will draw electricity at the plug while driving growth in

demand for – and energy use by – data centres and network services. However, sustained gains in energy efficiency could keep overall energy

demand growth largely in check for data centres and networks over the next five years.



Data centres worldwide consumed around 194 terawatt hours (TWh) of electricity in 2014, or about 1% of total demand. Despite a forecasted

tripling in data centre workloads by 2020, their electricity use is projected to grow by only 3% thanks to continued efficiency improvements.



Data networks consumed around 185 TWh globally in 2015, or around 1% of global electricity use. Mobile networks accounted for around two-thirds of the total. Depending on future efficiency trends, electricity consumption of networks could increase by 70% or fall by 15% by 2021.



The rapid proliferation of connected devices, giving rise to the “Internet of Things” (IoT), is creating opportunities for more efficient energy use. But these devices also use electricity. By 2020, more than 20 billion connected IoT devices and nearly 6 billion smartphones are expected to be online. Policies to improve device efficiency and to reduce standby power consumption will be critical to limit energy demand growth.



Providing credible forecasts of ICT energy use beyond the next five years is extremely challenging. Alongside the expected huge data demand gains will continue, or whether they will slow or stall.

Energy use by information and communications technologies

103

© OECD/IEA, 2017

growth, the key uncertainty over the longer term is whether efficiency

Introduction As the world becomes increasingly digitalized, information and communications technologies (ICT) are emerging as an important source of energy demand in their own right. This chapter looks at the direct1 energy use of three key ICT segments:

• Data centres, which are facilities used to house networked computer servers that store, process and distribute large amounts of data. Data centres use

energy to power both the information technology (IT) hardware (e.g. servers, drives and network devices) as well as the supporting infrastructure (e.g. cooling equipment).

• Data transmission networks, which transmit data between two or more

connected devices. Data networks use energy to transmit data through fixed and mobile networks.

• Connected devices, 2 which are consumer electronics, appliances and other

devices that can be connected to networks and interact with the network or other devices.

As billions more devices and machines are connected over the coming years, they will

draw electricity at the plug while also driving growth in demand for data centres and network services. Between 2007 and 2012, worldwide electricity use by ICT grew at an

estimated 7% per year, compared with 3% per year for electricity use overall (Van Heddeghem et al., 2014).3

Improvements in energy efficiency could help to limit the growth in energy demand in all three segments. Major improvements in the efficiency of computing (described by

Koomey’s Law4) coupled with the short lifespans of devices and equipment, which

hastens turnover, is improving the efficiency of the overall stock of devices, data

1

This chapter focuses primarily on the direct operational (use-phase) energy use of ICT. From a life cycle perspective, operational

energy use dominates for data centres and networks, while for connected devices, energy used to manufacture and dispose of

devices is generally more important. This is particularly true for mobile, battery-powered devices that are already very efficient,

tend to have shorter lifetimes and are used less intensively than, for instance, data centres, which are used continuously and have longer lifetimes (Hischier et al., 2014). 2

Connected devices are also often referred to as “networked”, “end”, or “edge” devices (i.e. at the edge of the network).

3

ICT included data centres, data networks and personal computers in this study.

4

Koomey’s Law describes the doubling of peak-output efficiency every 1.57 years for computing hardware (Koomey et al., 2011).

Peak-output efficiency is the number of computations that can be performed per kilowatt hour of electricity consumed during peak output. More recent analysis shows a slowing of this trend to every 2.7 years since 2000 (Koomey and Naffziger, 2015, 2016) due

to the end of Dennard Scaling (described in Bohr [2007]). That said, most computers only run a small fraction of the time at peak output (e.g. about 1% for mobile devices and laptops; 10% for enterprise data servers). Encouragingly, “typical-use efficiency”, a double every 1.5 years to 2020 (Koomey and Naffziger, 2015).

104

Energy use by information and communications technologies

© OECD/IEA, 2017

metric that considers average efficiency across a year and is perhaps more appropriate for devices, is expected to continue to

centres and networks. Devices are becoming smaller and more efficient, as evident, for example, in shifts from CRT to LCD screens, and from personal computers to tablets

and smartphones (Malmodin and Lundén, 2016; Stobbe et al., 2015; Urban et al., 2014). In the United States, data centre energy use has already levelled off since 2010

(Shehabi et al., 2016), despite strong growth in demand for data centre services. The future direction of ICT energy use will depend on how these trends play out over time.

Current energy use and near-term outlook Data centres Most of the world’s Internet Protocol (IP)5 traffic goes through data centres. Increasing

connectivity is driving demand for data centre services and energy use (mostly

electricity), with multiplying effects. For every bit of data that travels the network from data centre to end users, another 5 bits of data are transmitted within and among data

centres (Cisco, 2016a).

The IEA estimates that global data centre electricity demand in 2014 amounted to

around 194 TWh, or about 1% of global final demand for electricity. Despite large increases in demand for data centre services, concerted efforts to improve energy

efficiency have limited electricity demand growth in recent years. Koomey (2011)

estimated that data centres consumed 203-272 TWh in 2010, or 1.1-1.5% of total

global electricity use for that year.6

The United States remains the world’s largest data centre market measured by

workload7 (Cisco, 2016a) and by number of data centre servers (Shehabi et al., 2016).

The electricity consumption of US data centres remained essentially flat over 2010-14,

accounting for around 1.8% of overall US electricity use in 2014 (Shehabi et al., 2016).

Without efficiency improvements, energy use would have nearly doubled over the

same period. The Asia-Pacific region, the second-largest market, is expected to

account for more than one-third of global data centre workloads in 2020 (Cisco, 2016a).

5

IP traffic includes fixed and mobile internet traffic (IP traffic that crosses an internet backbone), corporate IP wide area network

(WAN) traffic, and IP transport of TV and video on demand (VoD). 6

There have been few estimates of global data centre energy use since Koomey (2011). One example is Van Heddeghem et al.

(2014), who estimated worldwide consumption of 268 TWh in 2012. Studies estimating energy use of data centres vary widely in their methodology and scope, making comparisons across studies difficult. The IEA global estimates are derived using the

Lawrence Berkeley National Laboratory (LBNL) data centre model and technology assumptions (Shehabi et al., 2016), and global

server and storage stock assumptions derived from Cisco’s Global Cloud Index (GCI) series (Cisco, 2016a), and is compatible with the approach in Koomey (2011). IEA believes this approach to be sound, as the LBNL methodology and data were thoroughly

A workload is a unit of measurement quantifying the amount of processing that a computer can do in a given amount of time.

Energy use by information and communications technologies

105

© OECD/IEA, 2017

vetted by the IT industry, and data from Cisco’s GCI on global server and storage stocks allow for generalisation globally. 7

Based on current trends in the efficiency of hardware and data centre infrastructure,

the IEA expects global data centre energy demand to grow by about 3% to 200 TWh in

2020 (Figure 5.1).8 This is despite a projected tripling of data centre workloads, a 22% increase in the number of servers and a 46% increase in storage drives (Cisco, 2016a;

Shehabi et al., 2016).9 The strong growth in demand for data centre services is offset

by continued improvements in the efficiency of servers, storage devices, network switches and data centre infrastructure,10 as well as a shift to much greater shares of

cloud and hyperscale data centres. Hyperscale data centres are very efficient, large-scale public cloud data centres operated by companies such as Alibaba, Amazon, and Google (Cisco, 2016a).

Figure 5.1 Global data centre energy demand by end use and data centre type End use 250

Data centre type

TWh

250 200

200 Infrastructure

150

Network

100

Storage

50 0

TWh

Servers 2014

2020

Hyperscale data centres

150

Cloud data centres (non-hyperscale)

100 50 0

Traditional data centres 2014

2020

Key message: Despite an expected tripling of data centre workloads, global data centre electricity

demand is expected to grow by only 3% to 2020 thanks to efficiency improvements in IT hardware and data centre infrastructure, and a shift to hyperscale data centres.

Note: Data centre infrastructure refers to energy consumed by non-IT equipment such as cooling. Sources: IEA analysis based on Cisco (2011a, 2012a, 2013a, 2014a, 2015a, 2016a, 2016b, 2016c); Cook et al. (2017); Shehabi et al. (2016).

The shift away from small inefficient data centres towards much larger cloud and

hyperscale data centres, often through outsourcing data centre services, is a major and growing source of energy efficiency gains. This trend is evident in the shrinking

share of data centre infrastructure in total energy demand (Figure 5.1, left) given the

8

This analysis uses data centre workload projections from the Cisco GCI; at the time of analysis, the latest available projections

run to 2020. 9

Cisco (2016a) projects that the global data centre installed server base will rise from around 39 million in 2014 to around 48

million in 2020, and that global data centre installed storage capacity will rise from around 380 exabytes in 2015 to over Data centre infrastructure includes cooling, uninterruptible power supply, lighting, etc. (Shehabi et al., 2016).

106

Energy use by information and communications technologies

© OECD/IEA, 2017

1 800 exabytes in 2020.

10

very low power usage effectiveness (PUE)11 of large data centres. Hyperscale data

centres have highly efficient IT equipment and run at high capacity, thanks in part to virtualisation software which enables data centre operators to deliver higher work

outputs with fewer servers (Cisco, 2016a). However, hyperscale data centres are only

viable for operations where latency – the time delay in data transmission – is not critical, given that hyperscale data centres are typically located further away from the end user. Hyperscale data centres could represent 47% of all installed data centre servers by 2020, up from 21% in 2015 (Cisco, 2016a).12

Near-term trends in global energy consumption by data centres will largely be determined by the efforts of the ICT industry to continue to improve energy efficiency,

as well as government policies and programmes to promote efficient data centre

operations. An analysis of the prospects for energy use in data centres in India

illustrates the technical potential for saving energy, and the risk of significant increased energy use if that potential is not realised (Box 5.1). Box 5.1

Two futures of data centre energy use in India

Despite its large population and role as a hub for the technology industry, India is still a relatively small player in the global data centre industry. As of 2016, India accounted for only around 1.5% of global server sales, or around 150 000 server units (IDC, 2016, 2017). This compares with estimated sales of around 3 million servers in the United States alone (Shehabi et al., 2016). However, the Indian data centre market is expected to grow by more than 20% per year over the next several years, with much of this growth coming from large cloud data centres (Alchemy Research and Analytics, 2015). This growth is consistent with Cisco’s GCI forecast of workload growth in the overall Asia-Pacific region (Cisco, 2016a). IEA analysis of data from the Manufacturers’ Association for Information Technology (MAIT, an ICT industry association in India) and Cisco GCI workload growth suggests that the installed base of servers in India will grow from around 460 000 in 2014 to around 820 000 in 2020 (MAIT, 2016). The IEA estimates that 2014 data centre energy demand in India amounted to around 2.2 TWh, or about 0.2% of India’s total electricity consumption (Figure 5.2). Under current trends, data centre energy demand could grow by more than 50% by 2020, driven mostly by servers and growth in the largest data centre types (cloud and hyperscale). This growth rate is higher than that expected at the global level or in the United States, due to more rapid growth in workload demand occurring in the Asia-Pacific region. Under an improved management scenario (modelled after US improved management practices in Shehabi et al. [2016]), energy demand could be reduced by around 15% compared with the current trends scenario in 2020. This represents a significant slowing of energy demand growth, despite strong growth in the numbers of IT devices and data centres in the Indian market.

11

PUE is a measure of how efficiently a data centre uses energy; the very best hyperscale data centres can have PUE values of

around 1.1 (meaning 0.1 kilowatt hours [kWh] used for cooling/power provision for every 1 kWh used for IT equipment).

Over the same time period, Cisco (2016a) projects that the percentage of global servers located in traditional (i.e. non-cloud)

data centres will decrease from 52% to 23%.

Energy use by information and communications technologies

107

© OECD/IEA, 2017

12

Figure 5.2

4.0

Data centre energy demand in India by end use and data centre type

End use

TWh

4.0

3.5

TWh

Data centre type

3.5 Infrastructure

3.0 2.5

Network

2.0 1.5

Storage

1.0 0.5

Servers

0.0

Current trends 2014

Improved management 2020

3.0

Hyperscale

2.5 2.0

Cloud (nonhyperscale)

1.5 1.0 0.5

Traditional

0.0

Current trends 2014

Improved management 2020

Key message: Improved management practices could help slow energy demand growth from data centres in India, reducing energy demand by around 15% compared to current trends in 2020.

Note: The improved management scenario (Shehabi et al., 2016) includes reductions in the number of installed servers due to removal of inactive servers and improvements in PUE to reflect global best practice values by data centre class. Sources: IEA analysis based on Cisco (2011a, 2012a, 2013a, 2014a, 2015a, 2015b, 2016a, 2016d); Confederation of Indian Industry (2013); Confederation of Indian Industry and LBNL (2016a, 2016b); Cook et al. (2017); MAIT (2016); Shehabi et al. (2016).

Data transmission networks The IEA estimates that global electricity use by internet data transmission networks13

in 2015 amounted to around 185 TWh, or around 1% of total worldwide electricity demand (Figure 5.3).14 Mobile data networks accounted for around two-thirds of this total. As with data centres, future energy demand depends on the growth of demand for data and the pace of further efficiency improvements.

The range of possible outcomes is wide. The IEA projects electricity demand in 202115

under two energy efficiency improvement scenarios: one assuming a moderate rate of

improvement of 10% per year, which is close to conservative estimates of historical

13

Internet data transmission networks include core, metro, edge and access (wired, WiFi and mobile). Traditional fixed telephony

systems and equipment on customer premises are excluded from the energy totals. 14

Other recent estimates of global electricity use attributable to transmission networks typically amount to around 1-1.5% of total

worldwide electricity demand, with about half from mobile networks (Andrae and Edler, 2015; Corcoran and Andrae, 2013; Ericsson, 2015; Malmodin, Bergmark, and Lundén, 2013). However, comparing global energy estimates is difficult due to

differences in data sources, system boundaries and assumptions for data traffic. Therefore, there are large uncertainty ranges around the data network energy use estimates in this report (see Figure 5.3).

IEA analysis uses IP traffic projections from the Cisco Visual Networking Index; at the time of this analysis, the latest available

projections run to 2021.

108

Energy use by information and communications technologies

© OECD/IEA, 2017

15

improvements; and one assuming a more rapid rate of improvement of 20% per year,

based on the historic rates achieved in well-managed networks in developed countries

with high capacity utilisation. In the moderate efficiency-improvement scenario, the

midpoint of the electricity demand range in 2021 rises by over 70% to about 320 TWh.

Under the high efficiency-improvement scenario, the midpoint drops by 15% to about 160 TWh (Figure 5.3).

Figure 5.3 Electricity use by internet data transmission networks 500

TWh

400 300 200 100 0

Fixed

Mobile

Total

Fixed

Mobile

Total

Fixed

Moderate rate of efficiency improvement 2015

Mobile

Total

High rate of efficiency improvement 2021

Key message: The prospects for future electricity demand from data transmission networks hinge critically on the pace of efficiency improvements.

Note: Points are the midpoint of the estimated range, which is denoted by vertical bars. Sources: IEA analysis based on Cisco projections for global IP traffic (Cisco, 2016e, 2017a) and network energy intensity estimates from Andrae and Edler (2015); Aslan et al. (2017); Fehske et al. (2011); GSMA (2012); Malmodin (2017); Malmodin et al. (2014); Schien and Preist (2014); Schien et al. (2015).

Several relevant trends are shaping the future of data network electricity use. Global IP traffic increased more than threefold over 2011-16 and is projected to grow by a similar rate between 2016 and 2021 (Cisco, 2017a). Global average fixed broadband

speeds are projected to nearly double from 2016 levels to 53 megabits per second

(Mbps) by 2021 – a thousand times faster than 56k modems that were widely used 15-20 years ago (Cisco, 2017a). Data demand growth over the next five years will be

driven strongly by consumer markets, which currently account for 80% of all data demand – mostly for video (Cisco, 2017a). The nature of data transmission is changing

rapidly, with traffic from wireless and mobile devices expected to account for more

Energy use by information and communications technologies

109

© OECD/IEA, 2017

than 63% of total IP traffic by 2021, up from 49% in 2016 (Figure 5.4).

Figure 5.4 Global IP traffic by access mode 3 500

Exabytes per year

3 000 2 500 2 000

Mobile

1 500

Fixed/WiFi from mobile devices

1 000

Fixed/WiFi from WiFi-only devices

500 0

Fixed/wired 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021

Key message: The nature of data access is changing rapidly from wired to wireless devices. Sources: IEA calculations based on Cisco (2011b, 2012b, 2013b, 2014b, 2014c, 2014d, 2015b, 2015c, 2016e, 2016f, 2017a, 2017b, 2017c).

This shift toward greater use of mobile networks may have significant implications for the energy use of data transmission networks, given the considerably higher electricity intensities (kWh per gigabyte [GB]) of mobile networks compared with fixed-line

networks at current traffic rates. 2G networks can be roughly two orders of magnitude

(i.e. over 100 times) more energy intensive than fixed-line networks, 3G roughly one

order of magnitude (i.e. over 10 times) and 4G roughly four times (Figure 5.5).16 While the latest mobile telecommunications technologies are much less energy intensive

than older technologies (e.g. 4G can be more than 50 times more energy efficient than 2G), their higher speeds may also allow for greater usage and traffic volumes.

One of the fastest-growing areas of mobile connections and traffic is machine-tomachine (M2M). This includes GPS for logistics and vehicles, smart metering and other IoT technologies, which enable the many energy efficiency and demand

response opportunities discussed in Chapters 2 and 4. It is projected that there will be

over 13 billion M2M connections by 2021, up from around 6 billion in 2016 (Cisco, 2017c). A shift in some M2M connections to a technology called low power wide area

(LPWA), which has very low energy use, may mitigate some of the energy demand

effects of the huge projected growth in M2M connections. At the same time,

associated security and privacy risks of LPWA also need to be considered (GSMA, 2017a). 16

The average energy intensity of mobile networks can also vary greatly depending on their capacity utilisation. As traffic within a

given access mode (e.g. 2G, 3G or 4G) increases, its overall average energy intensity can decrease, which means precise here can be generalised globally.

110

Energy use by information and communications technologies

© OECD/IEA, 2017

comparisons with fixed-line energy intensities are highly case-specific. Still, the overall rank order of energy intensities implied

Figure 5.5 Electricity intensity of network transmission by access type in 2015 100

kWh/GB

10

1

0.1

0.01

2G Fixed

3G

4G

Mobile

Key message: Mobile networks typically have higher electricity intensities (kWh/GB) than fixedline networks, but newer generation mobile networks and increased mobile network capacity utilisation are closing the gap.

Notes: Fixed and mobile network energy intensities are inclusive of core, metro, edge and access networks; due to large differences in electricity intensities, a log scale is used to illustrate orders of magnitude differences; the energy intensities of fixed and mobile networks are highly dependent upon assumed traffic rates. Sources: IEA analysis based on Andrae and Edler (2015); Aslan et al. (2017); Fehske et al. (2011); GSMA (2012); Malmodin (2017); Malmodin et al. (2014); Schien and Preist (2014); Schien et al. (2015).

Despite strong growth in data demand and shifts to mobile transmission, three

important trends can help counteract these growth trends and limit overall growth in

energy demand:

• Data transmission network technologies are rapidly becoming more efficient, meaning more data can be sent using less energy. Fixed-line network energy intensity has halved every two years since 2000 in developed countries (Aslan

et al., 2017). Mobile access network energy efficiency has in recent years improved at annual rates of around of 10-20% (Fehske et al., 2011; GSMA, 2012; Verizon, 2012).

• Capacity utilisation of networks is increasing, driving down energy use per byte sent, even with existing equipment.

• Mobile networks are shifting rapidly away from older networks towards more

efficient 4G. By 2021, 4G is expected to cover around 80% of mobile traffic,

Energy use by information and communications technologies

111

© OECD/IEA, 2017

while 2G is expected to cover less than 1% (Cisco, 2017b).

Connected devices Connected devices – also referred to as networked, edge or end devices – are

consumer electronics, appliances and other devices that can be connected to networks and interact with the network or other devices. Until recently, only a few devices were typically connected to communications networks, primarily computers, televisions, routers and modems. With the widespread diffusion of broadband internet,

as well as wireless and mobile access, an increasing variety of consumer devices, appliances and infrastructure across all sectors are being connecting to the internet and to each other.

Billions of new connected devices are expected to be connected over the next few years. The number of smartphones is expected to increase from 3.8 billion in 2016 to

almost 6 billion by 2020 (GSMA, 2017b),17 while the number of connected IoT devices

is expected to triple from about 6 billion in 2016 to over 20 billion by 2020 (Gartner,

2017). Over the longer term, it is conceivable that most electrical devices – and even some consumer items such as clothing – could become connected IoT devices, using energy to collect, process, store, transmit and receive data.

In discussing the energy use of connected devices (as a segment of ICT), it is helpful to distinguish between two types of connected devices: “electronic edge devices”, whose primary function is data storage/use, such as laptops and smartphones, and

“other edge devices”, whose primary functions are not data-related, such as

networked kitchen appliances and cars. Only data-related energy use is considered in this section, specifically energy use by “electronic edge devices” as well as any networked standby energy use of “other edge devices”.18

Several studies have estimated the energy use of connected devices (e.g. Andrae and

Edler, 2015; Corcoran and Andrae, 2013; IEA, 2014; Malmodin et al., 2010; Malmodin et

al.,

2013;

Van

Heddeghem

et

al.,

2014).

However,

differences

in

time,

scope/boundaries, assumptions and data sources make any direct comparisons across studies difficult.19

It is important to note that the direct energy use of devices at the plug is just one part

of the energy story for connected devices. Energy is also used to manufacture and

17

Smartphones are already playing an important role in driving efficiency and new business models across all sectors, including

energy. In transport, for instance, smartphones are the key enabler of app-based ride-hailing services (see Chapter 2). 18

While “other edge devices” do consume energy to provide their primary service, these energy demands should be considered a

part of their respective end-use sectors (e.g. a connected fridge is a home appliance, which is part of the buildings sector). 19

For example, Van Heddeghem et al. (2014) estimated total consumption of 307 TWh in 2012 from personal computers (including

equipment would consume around 380 TWh in 2015.

112

Energy use by information and communications technologies

© OECD/IEA, 2017

desktops, laptops and monitors), while Malmodin, Bergmark, and Lundén (2013) estimated that user, home and office ICT

dispose of these devices, which is more important in smaller, highly efficient and shorter-lived devices such as tablet computers and smartphones. 20 More than

three-quarters of lifecycle21 energy use for a tablet is associated with production, while

for desktops, use-phase energy accounts for more than half (Hischier et al., 2014).

The growth in the number of connected devices will continue to drive up the volume of data being generated and transmitted, along with the energy used in data centres and

data networks. Again, these indirect impacts are proportionally more significant for smaller devices given their high operational efficiency. For tablets, the energy used to

provide its internet services is estimated to be over 10 times greater than the energy needed for the production, use and disposal of the device (Hischier et al., 2014).

Long-term energy demand outlook Given the rapid pace of technological progress and change, providing credible forecasts of ICT energy use beyond the next five years is extremely challenging. This section briefly explores the key drivers of future ICT energy demand.

Demand for data centre services is expected to continue to grow strongly after 2020.

How this affects energy use will continue to be largely determined by the pace of energy efficiency gains. The continued shift to efficient cloud and hyperscale data

centres will reduce the energy intensity of data centre services. The use of artificial intelligence (AI) and machine learning may also help. In one recent case, machine learning was applied to Google data centres, which reduced their energy use for cooling by 40% (DeepMind, 2016). If demand for data centre services outpaces

efficiency gains, powering data centres with renewable energy will become increasingly important to curbing greenhouse gas emissions (Box 5.2).

For connected devices, standby power consumption is a particular concern, with

devices such as smart TVs and connected appliances using energy continuously to maintain connectivity. Inefficient networked standby could waste around 740 TWh per year by 2025, equivalent to the current annual electricity consumption of France and

the United Kingdom combined (IEA, 2014). The standby power consumption of IoT devices that are plugged in22 (excluding televisions and computers) is projected to

grow to 46 TWh by 2025, with 36 TWh coming from home automation (4E TCP, 2017).

20

From a life-cycle perspective, operational energy use dominates for data centres and networks, while for connected devices,

energy used to manufacture and dispose of devices is generally more important (Hischier et al., 2014). 21

Includes energy used for production, use, and end of life (i.e. disposal). Refer to the Hischier et al. (2014) for a full discussion of

22

Also known as “mains-connected” IoT devices, i.e. not battery- or self-powered and need to be plugged in to operate.

Energy use by information and communications technologies

113

© OECD/IEA, 2017

assumptions.

Box 5.2

Data centres, renewable energy and demand response potential

ICT companies, particularly hyperscale data centre operators, are already investing heavily in

renewable energy to protect themselves from the volatility of power prices, reduce their environmental impact and improve brand reputation (Cook et al., 2017; World Resources Institute,

2017). Over 20 of the top 50 ICT companies by revenue have quantified renewable energy

procurement goals, including some committing to 100%, primarily through power purchase agreements (PPAs) with electricity producers. Overall, ICT companies have been responsible for

more than half of total corporate renewable energy PPAs over the past three years (Figure 5.6). While ICT PPAs remain small as a proportion of overall new renewables capacity, accounting for just

1.9 GW out of over 160 GW brought online in 2016, procurement by ICT companies could still be a significant driver of renewables investment in various regions due to the size and influence of these companies (Miller et al., 2015). Figure 5.6 4

Corporate renewable power purchase agreements by type of buyer

GW

3 2

Commercial and public Industry

1

ICT 0

2010

2011

2012

2013

2014

2015

2016

Key message: Corporate buying via PPAs has accounted for about 10 GW of new utility-scale

renewables-based generating capacity over the past three years, led by ICT companies looking to hedge price volatility, diversify supply and meet sustainability goals.

Notes: Includes projects based on utility green tariffs in the United States; shares by buyer type are calculated based on contracted capacity. Sources: Calculations based on BNEF (2017); Platts (2017); Rocky Mountain Institute (2017).

Digitally enabled demand response could yield considerable cost savings for the grid as a whole while also facilitating greater integration of variable renewable energy (see Chapter 4). As large and

growing electricity users with a high degree of flexibility, data centres could play an increasingly

important role in demand response (Wierman et al., 2014). Data centres are highly automated and monitored, making them significantly more flexible and responsive compared to conventional

114

Energy use by information and communications technologies

© OECD/IEA, 2017

industrial facilities. For instance, 5% of load can typically be shed in five minutes without changes to how the IT workload is handled (Ghatikar et al., 2012).

Battery-powered connected devices are already highly energy efficient, and emerging

“energy harvesting” technologies could eventually help to further reduce energy use or even make some low-power electronics battery-free.23 Energy harvesting describes the

process by which energy is harvested or scavenged from sources in the environment, such as solar, wind, movement and sound, to power a device.

The nascent IT infrastructure for blockchain and cryptocurrencies is evolving rapidly, and its implications for global electricity use are not yet well understood. Early estimates suggest that the electricity use of Bitcoin data miners – one prominent

example of the emerging blockchain IT infrastructure – may currently be on the order

of less than 1/40th of 1% of global electricity use (Bevand, 2017; Deetman, 2016). However, as blockchain applications grow, understanding and managing its energy use

implications may become increasingly important for the energy analysis and policy communities.

Efforts by the ICT industry to improve energy efficiency and government policies to promote best practices could help to continue to hold down ICT energy demand

growth over an extended period. For instance, in data networks, policies to accelerate

the early phase-out of energy-intensive legacy networks could be particularly important. With data demand growth expected to be strongest in Asia-Pacific and

North America (Cisco, 2017a), continued efforts toward more efficient networks and data centres in these regions will be critical.

Nonetheless, there is considerable uncertainty about how long current ICT efficiency trends can continue. As an upper bound, Koomey, Matthews and Williams (2013)

estimated that, should current trends continue, processor efficiency limits may be

reached by around 2040 based on the physical efficiency limits of transistors.24 Other

experts estimate that theoretical limits to Koomey’s Law may be reached by around

2060 as a result of Landauer’s Principle (Demaine et al., 2016; Lääkkölä, 2015).25 It is also possible that the rate of energy efficiency improvements will diminish well before these theoretical limits are reached.

26

Complicating matters further, emerging

23

Researchers at the University of Washington recently demonstrated a battery-free cellphone (Talla et al., 2017).

24

In 1985, physicist Richard Feynman estimated that improvement by a factor of 1011 would be possible compared to computer

technology at the time. Efficiency improved by 4x104 from 1985 to 2009, meaning that if long-term trends of Koomey’s Law

continue, Feynman’s limit would be reached in 2041 (Koomey, Matthews and Williams, 2013). Feynman assumed a three-atom transistor to calculate his limit, but smaller transistors (such as the one-atom transistor demonstrated in 2012 [Fuechsle et al., 2012]) could push these limits out even further.

25

Laundauer’s Principle is the minimal amount of energy that must be consumed by erasing one bit of information.

26

Although gains in peak-output efficiency have already slowed as processors face the physical limitations of shrinking transistors,

(Koomey, 2015; Koomey and Naffziger, 2015, 2016).

Energy use by information and communications technologies

115

© OECD/IEA, 2017

ongoing improvements in chip design and power management have continued to deliver overall computing efficiency gains

computing technologies such as quantum computing 27 may change the nature of energy use for computing, although their application may not become widespread.

Policy considerations Actions and commitments by device manufacturers as well as network and data centre

operators will be central to driving further efficiencies and moderating overall ICT energy use. Government policies can also continue to play a key role, including through:

• Regulations, policies and programmes for more efficient devices, such as minimum energy performance standards, information campaigns, and energy labelling. 28 Such efforts should include a strong focus on reducing standby

losses

of

networked

EU 801/2013),

devices

improvements

energy-harvesting technologies.

to

(e.g.

European

power

Commission

management

and

Regulation

the

use

of

• Policies and programmes to encourage more efficient and sustainable device manufacturing practices, such as product eco-labels (e.g. Electronic Product

Environmental Assessment Tool [EPEAT]) and incentives for energy-efficient

industrial processes and renewable energy.

• Incentives and guidance for efficient data centre operations (e.g. see Huang and Masanet [2015] for a summary of best practices and how to calculate savings for incentives programmes).

• Policies and programmes for more efficient data transmission networks,

including network device energy efficiency standards, improved metrics and incentives for efficient network operations, and support of international

technology protocols (e.g. Institute of Electrical and Electronics Engineers [IEEE]) and standards development for low-energy networks.

• Better national data systems to collect data on ICT devices and their energy use characteristics29 to inform energy analysis and policy making – e.g. the US

Energy Information Administration has collected data on connected devices in homes and commercial buildings (Residential Energy Consumption Survey

27

While conventional computers use bits represented by 0s or 1s, quantum computers use quantum bits (qubits) to encode

information as 0s, 1s, or both at the same time, allowing them to solve problems faster, while potentially using less power to perform calculations. 28

For example, EU regulations covering a wide range of connected devices; ENERGY STAR labelling in the United States to

encourage low-energy connectivity.

The lack of data for this important end use severely limits analysts’ ability to conduct robust analysis of ICT energy use and

provide sound policy advice.

116

Energy use by information and communications technologies

© OECD/IEA, 2017

29

[RECS], Commercial Buildings Energy Consumption Survey [CBECS]), and has started to collect data on servers in data centres (CBECS).

• Governments leading by example through purchasing only efficient ICT and running their own data centres in the most efficient ways.

• Voluntary agreements with companies and industries on efficiency and CO2

emissions targets – e.g. the European Union and the United States have

adopted voluntary agreements to improve the efficiency of connected set-top

boxes; Verizon has stated public efficiency goals for their network operations, for which they publish indicators (kWh/GB) and progress annually.

• Uptake of other voluntary initiatives, such as the Connected Devices Alliance (CDA) Voluntary Principles for Energy Efficient Connected Devices.30

The policy challenges in addressing energy use from connected devices are inherently

different from those related to other historically important plug-loads such as refrigerators, freezers, air conditioners and televisions, which have been on the market

for decades. With these traditional products, policy makers were able to study a relatively stable market and implement informed energy performance standards. In contrast, the market and scope of connected devices are evolving at an unprecedented

pace, with new types of devices with complex operating modes coming into use almost

daily. In this new reality, policy makers will need to make creative use of available policy tools and levers.

Policy approaches to address the direct energy impacts of greater digitalization also

need to be developed as part of an overall strategy that addresses wider aspects

associated with the role of connectivity in the energy system and beyond.31 For digital

technologies to play their role in delivering energy and environmental benefits, favourable market conditions will be needed to stimulate the take up of smart appliances and equipment, as well as the associated infrastructure.

30

The CDA Voluntary Principles for Energy Efficient Connected Devices have two focuses: 1) Design Principles provide guidance

on the key features of energy-efficient connected devices, networks and communications protocols – for use by designers,

manufacturers and protocol authors; and 2) Policy Principles encourage a common global framework for the development of 31

See recent reports from the Global e-Sustainability Initiative (GeSI), e.g. #SMARTer2030 (Global e-Sustainability Initiative, 2015).

Energy use by information and communications technologies

117

© OECD/IEA, 2017

government policies and measures (CDA, 2016). See: https://cda.iea-4e.org/cda-principles.

References 4E TCP (2017). Energy Efficiency of the Internet of Things. www.iea-4e.org/document/384/energyefficiency-of-the-internet-of-things-technology-and-energy-assessment-report.

Alchemy Research and Analytics (2015). Data Centres Market in India. Andrae, A., and Edler, T. (2015). “On global electricity usage of communication technology: Trends to 2030”. Challenges, 6(1), 117–157. http://doi.org/10.3390/challe6010117.

Aslan, J., et al. (2017). “Electricity intensity of Internet data transmission: Untangling the estimates”.

Journal of Industrial Ecology. http://doi.org/10.1111/jiec.12630.

Bevand, M. (2017). “Electricity consumption of Bitcoin: a market-based and technical analysis”. http://blog.zorinaq.com/bitcoin-electricity-consumption/ accessed 27 September 2017).

BNEF (2017). EMEA Corporate PPA Database. Bohr, M. (2007). “A 30 year retrospective on Dennard’s MOSFET scaling paper”. IEEE Solid-State

Circuits Newsletter, 12(1), 11–13. http://doi.org/10.1109/N-SSC.2007.4785534.

Cisco (2017a). Cisco Visual Networking Index: Forecast and Methodology, 2016-2021.

https://www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-indexvni/complete-white-paper-c11-481360.html.

Cisco (2017b). Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update, 2016-

2021. www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-indexvni/mobile-white-paper-c11-520862.pdf.

Cisco (2017c). The Zettabyte Era: Trends and Analysis June 2017.

www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-index-vni/vnihyperconnectivity-wp.pdf.

Cisco (2016a). Cisco Global Cloud Index: Forecast and Methodology, 2015-2020.

www.cisco.com/en/US/solutions/collateral/ns341/ns525/ns537/ns705/ns1175/Cloud_Index_W hite_Paper.html#wp9000816.

Cisco (2016b). Cisco Global Cloud Index 2015-2020 - Cisco Knowledge Network Session.

www.cisco.com/c/dam/m/en_us/service-provider/ciscoknowledgenetwork/files/622_11_15-16Cisco_GCI_CKN_2015-2020_AMER_EMEAR_NOV2016.pdf.

Cisco (2016c). Cisco Global Cloud Index Q&A.

www.cisco.com/c/dam/en/us/solutions/collateral/service-provider/global-cloud-index-gci/qand-a-c67-738065.pdf.

Cisco (2016d). Cisco Global Cloud Index 2015-2020 - Cisco Knowledge Network Session. Cisco (2016e). Cisco Visual Networking Index: Forecast and Methodology, 2015-2020. Cisco (2016f). The Zettabyte Era: Trends and Analysis. Cisco (2015b). Cisco Visual Networking Index: Forecast and Methodology, 2014-2019.

118

Energy use by information and communications technologies

© OECD/IEA, 2017

Cisco (2015a). Cisco Global Cloud Index: Forecast and Methodology, 2014–2019.

Cisco (2015c). The Zettabyte Era: Trends and Analysis. Cisco (2014a). Cisco Global Cloud Index: Forecast and Methodology, 2013-2018. Cisco (2014b). Cisco Visual Networking Index: Forecast and Methodology, 2013-2018. Cisco (2014c). Cisco Visual Networking Index (VNI): Global and North America (NA) Mobile Data Traffic Forecast Update 2013-2018.

Cisco (2014d). The Zettabyte Era: Trends and Analysis. Cisco (2013a). Cisco Global Cloud Index: Forecast and Methodology, 2012-2017. Cisco (2013b). Cisco Visual Networking Index (VNI) Global Mobile Data Traffic Forecast Update. Mobile World Congress.

Cisco (2012a). Cisco Global Cloud Index: Forecast and Methodology, 2011-2016. Cisco (2012b). Cisco Visual Networking Index: Forecast and Methodology, 2011-2016. Cisco (2011a). Cisco Global Cloud Index: Forecast and Methodology, 2010-2015. Cisco (2011b). Cisco Visual Networking Index: Forecast and Methodology, 2010-2015. Confederation of Indian Industry (2013). Energy Efficiency in Indian Data Centers.

https://datacenters.lbl.gov/resources/energy-efficiency-indian-data-centers-present-trendsfuture-opportunities.

Confederation of Indian Industry, and LBNL (Laurence Berkeley National Laboratory) (2016a).

Accelerating Energy Efficiency in Indian Data Centers: Final Report for Phase I Activities. https://datacenters.lbl.gov/sites/all/files/IndiaFinalPhase1Report_LBNL2016.pdf.

Confederation of Indian Industry, and LBNL. (2016b). Accelerating Energy Efficiency in Indian Data

Centers: Report for Phase II Activities.

https://datacenters.lbl.gov/sites/all/files/IndiaPhaseIIFinalReport_CIILBNL2016.pdf. CDA (Connected Devices Alliance) (2016). “CDA Design & Policy Principles for Energy Efficient Connected Devices”. https://cda.iea-4e.org/cda-principles. Cook, G., Kong, A., and Bach, N. (2017). Clicking Clean: Who Is Winning the Race to Build a Green

Internet? Greenpeace.

Corcoran, P., and Andrae, A. (2013). “Emerging trends in electricity consumption for consumer ICT.

Electrical and Electronic Engineering (Reports), 1–56.

Coroama, V. C., et al. (2013). “The direct energy demand of internet data flows”. Journal of

Industrial Ecology, 17(5), 680–688. http://doi.org/10.1111/jiec.12048.

DeepMind (2016). “DeepMind AI reduces Google data centre cooling bill by 40%.

https://deepmind.com/blog/deepmind-ai-reduces-google-data-centre-cooling-bill-40/.

Deetman, S. (2016). “Bitcoin could consume as much electricity as Denmark by 2020.

https://motherboard.vice.com/en_us/article/aek3za/bitcoin-could-consume-as-much-

Energy use by information and communications technologies

119

© OECD/IEA, 2017

electricity-as-denmark-by-2020 (accessed 27 September 2017).

Demaine, E. D., et al. (2016). “Energy-efficient algorithms”. In Proceedings of the 2016 ACM

Conference on Innovations in Theoretical Computer Science. http://doi.org/http://dx.doi.org/10.1145/2840728.2840756.

Ericsson (2015). “Decoupling Energy from Traffic Growth”. In Ericsson Mobility Report November

2015. https://www.ericsson.com/assets/local/news/2016/03/ericsson-mobility-report-nov2015.pdf

Fehske, A., et al. (2011). “The global footprint of mobile communications: The ecological and economic perspective”. IEEE Communications Magazine, 49(8), 55–62. http://doi.org/10.1109/MCOM.2011.5978416.

Fuechsle, M., et al. (2012). “A single-atom transistor”. Nature Nanotechnology, 7(4), 242–246. http://doi.org/10.1038/nnano.2012.21.

Gartner (2017). “Gartner says 8.4 billion connected ‘things’ will be in use in 2017, up 31 percent from 2016”. www.gartner.com/newsroom/id/3598917 (accessed 12 July 2017).

Ghatikar, G., et al. (2012). Demand Response Opportunities and Enabling Technologies for Data

Centers: Findings from Field Studies. Lawrence Berkeley National Laboratory.

Global e-Sustainability Initiative (2015). #SMARTer2030-ICT Solutions for 21st Century Challenges. http://smarter2030.gesi.org

GSMA (2017a). “New report outlines security considerations for LPWA technology”. www.gsma.com/iot/news/new-report-outlines-security-considerations-lpwa-technology/ (accessed 9 September 2017).

GSMA (2017b). The Mobile Economy: 2017.

www.gsmaintelligence.com/research/?file=9e927fd6896724e7b26f33f61db5b9d5&download.

GSMA (2012). “Mobile’s Green Manifesto 2012”. https://www.gsma.com/publicpolicy/mobilesgreenmanifesto. Hischier, R., et al. “Grey energy and environmental impacts of ICT hardware”. In L. M. Hilty and B. Aebischer (eds.), ICT Innovations for Sustainability. Advances in Intelligent Systems and

Computing (Vol. 310). http://doi.org/10.1007/978-3-319-09228-7.

Huang, R., and Masanet, E. (2015). “Data Center IT Efficiency Measures”. In the Uniform Methods

Project: Methods for Determining Energy Efficiency Savings for Specific Measures. National Renewable Energy Laboratory.

IDC (2017). “India x86 server market declined by 16.8 percent in Q4 2016; but government remains growth enabler: IDC India”. www.idc.com/getdoc.jsp?containerId=prAP42433017 (accessed 24 July 2017). IDC (2016). “India x86 server market declines by 16.9 percent in Q2 2016: IDC India”. www.idc.com/getdoc.jsp?containerId=prAP41864916 (accessed July 24 2017).

Efficient in Billions of Connected Devices. OECD/IEA, Paris.

Koomey, J. (2011). Growth in data center electricity use 2005 to 2010. Analytics Press. 120

Energy use by information and communications technologies

© OECD/IEA, 2017

IEA (International Energy Agency) (2014). More Data, Less Energy: Making Network Standby More

Koomey, J. (2015). “A primer on the energy efficiency of computing”. In R. H. J. Knapp, B. G. Levi, and D. M. Kammen (eds.), Physics of Sustainable Energy III: Using Energy Efficiently and

Producing It Renewably, 1652, 82–89. Melville, NY: American Institute of Physics. http://doi.org/10.1063/1.4916171.

Koomey, J., et al. (2011). “Implications of historical trends in the electrical efficiency of computing”.

IEEE Annals of the History of Computing, 33(3), 46–54. http://doi.org/10.1109/MAHC.2010.28.

Koomey, J., Matthews, H. S., and Williams, E. (2013). “Smart everything: Will intelligent systems reduce resource use?”. Annual Review of Environment and Resources, 38(1), 311–343. http://doi.org/10.1146/annurev-environ-021512-110549.

Koomey, J., and Naffziger, S. (2016). “Energy efficiency of computing: what’s next?”. Electronic

Design, 28. http://electronicdesign.com/microprocessors/energy-efficiency-computing-what-s-

next.

Koomey, J., and Naffziger, S. (2015). Moore’s Law Might Be Slowing Down, But Not Energy

Efficiency - IEEE Spectrum. https://spectrum.ieee.org/computing/hardware/moores-law-might-

be-slowing-down-but-not-energy-efficiency (accessed 1 September 2017).

Lääkkölä, R. (2015). Data Center Degrowth – an Experimental Study. Aalto University. MAIT (2016). Energy Efficiency in Public Procurement of IT Hardware.

www.mait.com/assets/energy-efficiency-in-it-procurement_report_final__cc__fnl-for-print.pdf.

Malmodin, J. (2017). Personal communication, 21 June 2017. Malmodin, J., and Lundén, D. (2016). “The energy and carbon footprint of the ICT and E&M sector in Sweden 1990-2015 and beyond”. 4th International Conference on ICT for Sustainability,

(Ict4s), 209–218. www.ericsson.com/assets/local/hidden-pages/our-publications/conferencepapers/paper25_ict4s16.pdf.

Malmodin, J., Bergmark, P., and Lundén, D. (2013). “The future carbon footprint of the ICT and E&M sectors”. International Conference on Information and Communication Technologies for

Sustainability, 12–20. http://doi.org/10.3929/ethz-a-007337628.

Malmodin, J., et al. (2014). “Life cycle assessment of ICT: Carbon footprint and operational

electricity use from the operator, national, and subscriber perspective in Sweden”. Journal of

Industrial Ecology, 18(6), 829–845. http://doi.org/10.1111/jiec.12145.

Malmodin, J., et al. (2010). “Greenhouse gas emissions and operational electricity use in the ICT and entertainment and Media sectors”. Journal of Industrial Ecology, 14(5), 770–790. http://doi.org/10.1111/j.1530-9290.2010.00278.x.

Miller, J., et al. (2015). Renewable Electricity Use by the U.S. Information and Communication

Technology (ICT) Industry. NREL/TP-6A20-64011.

Platts (2017). World Electric Power Plants Database. https://www.platts.com/products/worldRocky Mountain Institute (2017). “Business Renewables Center”. https://www.rmi.org/ourwork/electricity/brc-business-renewables-center/. Energy use by information and communications technologies

121

© OECD/IEA, 2017

electric-power-plants-database.

Schien, D., et al. (2015). “The energy intensity of the internet: Edge and core networks”. In H. L. and A. B. (eds.), ICT Innovations for Sustainability. Advances in Intelligent Systems and Computing 310, 157–170. Springer, Cham. http://doi.org/10.1007/978-3-319-09228-7_9.

Schien, D., and Preist, C. (2014). “Approaches to energy intensity of the internet”. IEEE

Communications Magazine, 52(11), 130–137. http://doi.org/10.1109/MCOM.2014.6957153.

Shehabi, A., et al. (2016). United States Data Center Energy Usage Report. Berkeley, CA. https://eta.lbl.gov/publications/united-states-data-center-energy.

Stobbe, L., et al. (2015). Development of ICT-related electricity demand in Germany.

www.bmwi.de/Redaktion/EN/Pressemitteilungen/2015/20151210-gabriel-studie-strombedarfikt.html.

Talla, V., et al. (2017). “Battery-free cellphone”. Proceedings of the ACM on Interactive, Mobile,

Wearable and Ubiquitous Technologies, 1(2), 1–20. http://doi.org/10.1145/3090090.

Urban, B., et al. (2014). Energy Consumption of Consumer Electronics in U.S. Homes in 2013. Fraunhofer USA Center for Sustainable Energy Systems, (June), 1–158.

www.ce.org/CorporateSite/media/environment/Energy-Consumption-of-ConsumerElectronics.pdf.

Van Heddeghem, W., et al. (2014). “Trends in worldwide ICT electricity consumption from 2007 to 2012”. Computer Communications, 50, 64–76. http://doi.org/10.1016/j.comcom.2014.02.008.

Verizon. (2012). 2012 Corporate Responsibility Supplement.

https://www.verizon.com/about/sites/default/files/2012-verizon-corporate-responsibilitysupplement.pdf.

Wierman, A., et al. (2014). “Opportunities and challenges for data center demand response”.

International Green Computing Conference, 1–10. http://doi.org/10.1109/IGCC.2014.7039172.

World Resources Institute (2017). Corporate Renewable Energy Buyers’ Principles: Increasing

Access to Renewable Energy.

122

Energy use by information and communications technologies

© OECD/IEA, 2017

http://www.wri.org/sites/default/files/Corporate_RE_buyers_guide-JAN2017.pdf.

Chapter 6

Cross-cutting risks: Cybersecurity, privacy and economic disruption Highlights •

Digitalization carries three main cross-cutting risks which, while not unique to the energy sector, need to be assessed and managed: cybersecurity, privacy and economic disruption.



Digitalization can make energy systems more vulnerable to digital risks, such as geomagnetic storms and cyber-attacks. Governments and

companies need to work together to manage the increased complexity of threats to cybersecurity. Full prevention of cyber-attacks is impossible,

but their impact can be limited if countries and companies are well prepared. Digital resilience needs to be integrated into technology research and development, and policy and market frameworks.



Privacy and data ownership are becoming a major concern as more and more detailed data are collected, in particular information from smart meters about household energy use. Policy makers need to balance

privacy concerns with other objectives, which include promoting

innovation in markets, operational needs of utilities and the wide-ranging potential of the digital transformation of electricity.



Digitalization is leading to major disruptions in the energy sector and the broader economy, affecting jobs and skill requirements. It is changing

work patterns and tasks, causing job losses in some areas and creating new jobs in others. The impact varies across the different parts of the

energy sector. Policy makers in the energy field should participate fully in relevant government-wide decision-making on digital matters. Digitalization entails a number of wider challenges which, while not unique to the energy sector, need to be assessed and managed. This chapter will explore the three

most important: cybersecurity, privacy and economic disruption. Energy policy makers for energy operations and business, and to contribute to well-considered policies.

Cross-cutting risks: Cybersecurity, privacy and economic disruption

123

© OECD/IEA, 2017

will need to work with colleagues across their governments to track the implications

Digital security Cyber threats to energy security Digitalization brings many benefits, but it also opens the door to increasing risks for

energy security, both from natural hazards such as geomagnetic storms and from unintended cyber incidents and intentional cyber-attacks (Box 6.1). For example, in

1989, a severe geomagnetic storm caused a nine-hour outage in transmission grids in Quebec. In December 2015, hackers illegally breached the computer and control systems of a Ukrainian regional electricity distribution company and managed to

disconnect substations, resulting in the loss of power for 225 000 customers. This was

followed by a second attack in December 2016 that appears to have used malware specifically designed to take over control of electric grid circuit breakers (ESET, 2017).

Generic cyber-attacks, such as the worldwide “WannaCry” and “NotPetya”

ransomware incidents in May and June 2017 respectively, have also impacted the energy sector by damaging information technology (IT) business systems.

Box 6.1 •

Malware, ransomware, phishing and bots

Malware is software (e.g. a virus, worm, rootkit, or botnet) specifically designed to disrupt,

damage, or gain unauthorised access to an information and communications technology (ICT) system. •

Ransomware is a type of malware that encrypts user data, asking victims to pay a ransom in order to obtain a decryption key.



Phishing/whaling is trying to obtain sensitive information, such as usernames and passwords, by means of an email (or other electronic communication) that is disguised as a trustworthy

communication, but which, when opened or when a link is clicked on, allows the sender a point of access. •

Botnets (short for robot networks) are automated programs that run over the internet. Some botnets run automatically, while others only execute commands when they receive specific input. Not all botnets are malicious.

The disruption caused to energy supply by reported cyber-attacks to date (Table 6.1) has been small compared to major power outages from storms and mechanical

equipment failures, or from geopolitical disruptions to oil and natural gas supply. However, cyber-attacks – whether aimed at sabotage, taking control of energy

systems, industrial espionage, or ransom – are becoming easier and cheaper for

hostile actors to organise. These range from amateurs to professional hackers-for-

124

Cross-cutting risks: Cybersecurity, privacy and economic disruption

© OECD/IEA, 2017

hire, to other states and state use of proxies (HMG, 2016a).

Table 6.1

Open source information regarding cyber-attacks affecting energy infrastructure

Incident

Description (from open-source information)

Shamoon 1 and 2

“Shamoon 1” virus carried out cyber-sabotage and destroyed over 30 000

(Saudi Arabia, 2012

computers at Saudi Aramco. There was no direct impact on oil production, but

and 2016)

the company was forced to revert to traditional paper and telephone trading for several weeks. Qatari natural gas company, RasGas, was also impacted. The virus was set to execute after working hours in order to minimise detection.

“Shamoon 2” virus targeted similar vulnerabilities and was used to overwrite parts of computer hard discs. Western Ukraine

The first confirmed cyber-attack specifically against an electricity network.

(2015)

(SCADA) and firmware with a combination of malware, personnel credentials

power grid

Attackers accessed substations’ supervisory control and data acquisition

obtained by means of email phishing, and Denial of Service (DoS) to prevent customers from obtaining call centre information about the blackout.

Investigators concluded that a large well-co-ordinated team had prepared the attack over several months. The Mirai Botnet

“Mirai” malware exploited low security in connected smart devices, such as

(2016)

cameras, to use a botnet (a network of devices under simultaneous command by the attacker to overload the victim by continuously sending data) to deliver the largest DoS attack to date. This attack did not target or impact energy

infrastructure, but illustrates the vulnerability of the Internet of Things (IoT). Industroyer/

A second brief but significant attack on the Ukrainian electricity system,

Crash Override

thought to have been a test run for malware “Industroyer” (also known as

(Ukraine, December

“Crash Override”). This versatile malware enables attackers to view, block,

2017)

suggests expert knowledge of several standardised industrial communication

2016 – reported May

control or destroy grid control equipment such as circuit breakers. Its design protocols widely used to control infrastructure – not only electricity grids –

throughout Europe, Asia and the Middle East. This was an example of a cyber intrusion into the control systems of critical infrastructure. Nuclear plant spear

This incident occurred in the United States. It used targeted email messages

(2017)

exposing recipients’ credentials for the control engineering network. The

phishing attack

containing fake Microsoft Word résumés for engineering jobs, potentially

hackers also compromised legitimate external websites that they knew their victims frequented (known as a watering hole attack).

WannaCry (2017)

“WannaCry” ransomware hit hundreds of thousands of computers in thousands of organisations in some 150 countries, taking advantage of an access point in Microsoft operating systems for which some users had failed to install the

secure update (or “patch”). These attacks did not target energy infrastructure, China (hereafter, “China”), over 20 000 China National Petroleum Corporation (CNPC) petrol stations went offline. Cross-cutting risks: Cybersecurity, privacy and economic disruption

125

© OECD/IEA, 2017

but several energy companies reported problems. In the People’s Republic of

Some cyber-attacks target operational technology (OT): the computers, software and networks used to control, monitor, manage and protect energy delivery systems. Other

attacks might target only the IT business systems of energy companies that do not control the physical process of energy delivery, but result in administrative

interruptions (as in the case of Shamoon). Unintentional cyber incidents linked to the increasing complexity of “systems of systems” that combine many layers of ICT and

OT are also becoming more frequent, for example when an update in one type of equipment causes malfunctions in other equipment.

The growth of the IoT and changes in digital technologies are increasing the potential “cyber-attack surface” in energy systems. The expansion of the IoT, combined with the

diversification and decentralisation of energy technologies, will link millions of new small-scale prosumers and billions of devices into the electricity system. Industry

forecasts estimate that total numbers of connected IoT devices could more than

double to over 20 billion by 2020 (Gartner, 2017).1 If there is one suspect device at the

edge of a network, this can be a weak point for the whole system. A recent study by the European Parliament concluded: “The development of smart energy has also led to

exponential growth of networked intelligence throughout the energy grids and also

consumer premises. The result is that a massively expanding ‘attack surface’ now forms the operational foundation of the energy ecosystem. As the energy system is

also fundamentally interconnected with every other critical infrastructure network, the cybersecurity threat to the energy sector impacts every aspect of our modern society” (European Parliament 2016; see also Global Smart Grid Federation, 2016).

Digital technologies used in centralised energy systems are also changing. Early

digitalization often relied on IT and OT that were proprietary or vendor-specific. Electricity substations typically might use several generations of equipment, assembled piecemeal over time. This older infrastructure often pre-dates embedded

security standards, but can benefit from an element of “security by obscurity”, so that hacking it requires specialised knowledge. Today, with more connectivity between system components, more automation, a shift to cloud computing, and the

replacement of energy-specific IT by sophisticated open-protocol industry standards, newer systems have higher security, but lose the protection of obscurity and the level of specialised energy system knowledge needed for attack diminishes.2

1

Machine-to-machine (M2M) connections, which include energy sector applications such as smart meters and process sensors in

power plants, will be the fastest-growing category, growing at around 20% a year (Cisco, 2017). See also Chapter 5. 2

There can be advantages to maintaining “old-school” (manual) control-of-key infrastructure as a backup to remote digital

management. In the case of the December 2015 incident in Ukraine, operators were quickly able manually to switch circuit

126

Cross-cutting risks: Cybersecurity, privacy and economic disruption

© OECD/IEA, 2017

breakers back on and override the cyber-attack. Many power grids in other countries are now fully automated, where a similar

attack could be more difficult to counteract.

Cyber-attack techniques can target personnel, products (both data and physical infrastructure) and processes (system data flow). This means that there is greater

need to protect the integrity and confidentiality of information, and to identify trustable

sources and authorised recipients, also in supply chains and procurement. Meanwhile, the global scope of the internet and the geographical reach of multinational energy

companies mean that an attack at one location can instantly spread worldwide. Risk analysis, therefore, needs to span products, people and processes.

Taking all of this into account, some experts in strategic risk assessment see credible

scenarios for “low-probability, high-risk” attacks that could shut down the electricity grids of a major economic region for a period of days or – on a rolling basis – weeks

(Lloyd’s, 2015; Cambridge Centre for Risk Studies, 2016; National Academies of

Sciences, Engineering, and Medicine, 2017; Madnick, 2017). Energy professionals and

oversight authorities also foresee a future of many small-scale nuisance attacks from bots (numbering thousands per day).

While cyber-attacks can affect a variety of parts of the global economy, attacks on energy systems are likely to be particularly disruptive. Unlike most IT systems,

electricity OT systems must operate in real-time and cannot simply install patches or updates, or shut down and reboot as in the typical response to digital failures or breaches. Special security considerations and precautions are therefore needed.

Building resilience Full prevention of cyber-attacks is impossible, but their impact can be limited if countries and companies are well prepared and have built-in resilience. This is particularly important for critical infrastructure: the physical and institutional assets

that are essential for an economy to function, such as large-scale energy systems. For example, Mexico has around 3 000 “strategic installations”, half of them owned by the

national oil company PEMEX and another 13% by the Federal Electricity Commission (Government of Mexico, 2014). Any infrastructure in Germany on which more than 500 000 people depend is considered critical. This includes all gas power plants and electricity transmission grids.

Building system-wide resilience depends on all actors and stakeholders first being

aware of the risks. The success of any one attack will depend not only on the capability of the attacker, but also on the vulnerability of the target and its

preparedness to respond. For example, a recent study finds that few oil and gas

companies today acknowledge cyber breach as a major risk. Some categorise cyber risk together with other risks such as labour disputes; others do not mention cyber risk in their company strategies. The same report finds that three-quarters of oil and gas

Furthermore, older upstream equipment was not built with security in mind and lacks

Cross-cutting risks: Cybersecurity, privacy and economic disruption

127

© OECD/IEA, 2017

companies were victims of some kind of cyber-attack at least once in the last year.

monitoring tools, while only 14% of companies have fully operational security monitoring centres on their networks (Mittal, Slaughter and Zonneveld, 2017). Digital energy security should be built around three key concepts:

• Resilience, i.e. the ability of a nation, system or institution to adapt to changing contexts, to withstand shocks, and to quickly recover or adapt to a desired level

of stability, while preserving the continuity of critical infrastructure operations (Larkin et al., 2015).

• Cyber hygiene, i.e. the basic set of precautions and monitoring that all ICT users should undertake. This includes first and foremost building awareness. Other key elements are secure configuration of equipment and networks, keeping software up to date, avoiding giving staff and users unnecessary

system privileges or data access rights, and training to establish a

security-conscious culture throughout an organisation. For example, staff may need to be trained on the secure use of their personal computers at home and

mobile devices.

• Security by design, i.e. the incorporation of security objectives and standards as a core part of the research and design process; security should not be a later

add-on after a product has been built or supplied to users. Security by design can be an efficient way to reduce overall risk.

The choice of technology and system architecture is an important factor for resilience.

Cybersecurity and resilience need to be made a central part of energy research, development and deployment. For example, the National Renewable Energy Laboratory in the United States has developed a cybersecurity testbed that imitates

the power and communication network of a typical distribution utility. It incorporates a

nine-layer security architecture that has real-time transactions between different actors (end users and/or systems).

Microgrid electricity networks may have an advantage in that they can be islanded

(temporarily segregated from the rest of the grid). Blockchain technology is part of the cyber problem as cryptocurrencies can be used to collect ransomware payments. But

blockchain can also be a potential cybersecurity solution at the grid edge, for example

if used to validate whether a device is running up-to-date firmware and has not been

tampered with (see Chapter 4). The Pacific Northwest National Laboratory in the United States is developing blockchain solutions to strengthen the cybersecurity of transactive energy systems, including within microgrids and building-to-grid

128

Cross-cutting risks: Cybersecurity, privacy and economic disruption

© OECD/IEA, 2017

connections, as well as validate that a device is running up-to-date firmware, and

confirm that it has not been tampered with.3 Countries that are still building their

energy systems have the advantage of being able to design into new infrastructure both high security and networks that can be rapidly isolated.

Establishing clear and accepted definitions is also crucial, both for analytic concepts

such as “threat” and “risk” (Marchese and Linkov, 2017), and for more technical specifications, such as varying definitions of “air-gapping” (the physical separation of

ICT devices from other devices/networks). For example, cybersecurity specialists highlight that IT and OT personnel often have different security priorities. In IT, these

are the confidentiality, integrity and availability of a system. In OT, safety is first, followed by availability. Confusion can result.

In relation to technical specifications, the International Electrotechnical Commission

(IEC) has an important role. IEC develops worldwide standards and conformity

assessments for equipment and business processes, and is currently incorporating

cybersecurity. To date, it has identified around 650 electrotechnical standards from 40 different standard-setting organisations that have applicability to cybersecurity. A categorisation of cybersecurity standards is being prepared for 2018.

Another critical need is clarity about the division of responsibilities for security, preparedness and response among market players, including for-profit energy system

operators, and governing bodies, including government agencies and regulators. A company or local operator is likely to be able to handle attacks by botnets or amateurs

(the “script kiddie” teenage hacker), but will not be able to manage the impacts of a major systemic attack (Figure 6.1).

Uncertainty about risks makes it difficult to justify large expenditure on staff or on cyber-insurance policies. Setting regulatory requirements can help ensure necessary

investments are made, for example, adding cybersecurity criteria to the rate base for electricity grids.

At the same time, companies acknowledge that compliance with regulatory standards

does not, on its own, guarantee infrastructure will be secure. In general, regulatory

standards will struggle to keep up with rapid technological changes and new vulnerabilities.

Governments and energy companies, therefore, need to be both proactive and adaptive. In some cases, innovative solutions might be found when energy companies

work together. In an interesting experiment, Austrian energy market actors have

Several other US national laboratories are also doing research into cybersecurity for energy delivery systems:

www.energy.gov/oe/activities/cybersecurity-critical-energy-infrastructure.

Cross-cutting risks: Cybersecurity, privacy and economic disruption

129

© OECD/IEA, 2017

3

decided to set up a joint team on cybersecurity, with the aim of saving costs, pooling expertise and improving early visibility of attacks and risks (see also Carr, 2016).4 A

wider group of European companies have formed a European Network for Cyber Security to collaborate on training.5 Examples like this suggest that companies may

find benefits in enhancing information sharing among themselves, beyond any requirements to report to regulators. Wider consultation with experts in the university sector6 and with other industry sectors can also be valuable. Figure 6.1 Capabilities of companies to handle cyber-attacks

Top 10 cyber forces

Beyond capabilities of most companies

State actors Professional cyber attackers

Capability level of large, well-prepared companies

Most companies can handle

Ambitious hacker Amateur hacker

Key message: The handling of some attacks falls within the capability of companies themselves, while larger-scale attacks by sophisticated actors may require more active government responses. Source: Presentation by Swissgrid at Florence School of Regulation workshop on Cybersecurity in the Energy Sector, 24 March 2017 (adapted by IEA).

Best practices and policy Much can be learned by examining how a variety of governments are currently

structured and organised to deal with cyber threats. In many countries, cybersecurity is

part of governance frameworks that are broader than energy, but which include measures specific to energy.

In the United States, the federal Department of Energy (DOE) is the lead for cybersecurity in the energy sector. DOE has facilitated the development of a series of

4

https://cert.at/reports/report_2016_chap05/content.html.

5

www.encs.eu.

6

For example, the CERT Division of the Software Engineering Institute at Carnegie Mellon University (www.cert.org/about/) and

the Cyber Security Centre at Warwick University (https://www2.warwick.ac.uk/fac/sci/wmg/research/csc/).

130

Cross-cutting risks: Cybersecurity, privacy and economic disruption

© OECD/IEA, 2017

industry-led roadmaps to provide a strategic framework to co-ordinate public and

private initiatives for resilient energy delivery systems, most recently the 2011

Roadmap to Achieve Energy Delivery Systems Cybersecurity (DOE, 2011). In 2013, DOE launched a public-private partnership Cybersecurity Risk Information Sharing Program (CRISP) to provide electricity sector organisations with near-real-time cyber threat information and analysis.7

In addition, the US Federal Energy Regulatory Commission works jointly with Canada

within the North American Electric Reliability Corporation (NERC) to set and maintain

critical infrastructure protection standards. 8 NERC also operates the Electricity

Information Sharing and Analysis Center – a secure forum for the sharing of cybersecurity threat information within the electricity sector. For example, NERC issued an alert and recommendations in February 2016 upon findings from the December 2015 cyber-attack in Ukraine.

In Europe, the 2016 Network Information Security Directive (NIS)9 requires EU member

states to adopt a national cybersecurity strategy.10 NIS created a co-operation group and a computer emergency response team, and establishes security and notification

requirements for operators of essential services and digital service providers. At EU level, proposals for a strengthened cyber resilience strategy were published by the European Commission in September 2017, 11 while detailed work on electric grid

security codes is under way in the Smart Grids Task Force.12

The EU cybersecurity strategy is also closely linked to a wider critical infrastructure

protection programme.13 Under this framework, the German federal network regulator,

Bundesnetzagentur, published a catalogue of minimum IT security standards in August 2015.14 The Cyber Essentials Scheme in the United Kingdom plays a similar role, although its mandate is broader than the energy sector (HMG, 2016b). In France,

the National Agency for the Security of Information Systems (Agence nationale de la securité des systèmes d’information, ANSSI) formulates and supervises cybersecurity 7

https://energy.gov/oe/energy-sector-cybersecurity-preparedness-0.

8

www.nerc.com/pa/CI/Comp/Pages/default.aspx.

9

Directive (EU) 2016/1148 of the European Parliament and of the Council of 6 July 2016 concerning measures for a high common

level of security of network and information systems across the Union, http://eur-lex.europa.eu/legalcontent/EN/TXT/PDF/?uri=CELEX:32016L1148&from=EN. 10

At the time of writing, all EU countries except two have finished drafting their cybersecurity strategies. Fifteen out of the

26 strategies do not specifically mention the energy sector. 11

https://ec.europa.eu/digital-single-market/en/policies/cybersecurity.

12

https://ec.europa.eu/energy/en/topics/markets-and-consumers/smart-grids-and-meters/smart-grids-task-force.

13

https://ec.europa.eu/home-affairs/sites/homeaffairs/files/what-we-do/policies/crisis-and-terrorism/critical-

infrastructure/docs/swd_2013_318_on_epcip_en.pdf.

Energy suppliers must appoint an IT security officer registered with the regulator, and they are obliged to introduce and certify

an information security management system (ISMS), www.ids.de/en/news/article/newsletter-dezember-2015.html.

Cross-cutting risks: Cybersecurity, privacy and economic disruption

131

© OECD/IEA, 2017

14

standards and regulations for “operators of vital importance.” ANSSI also collects and provides information about incidents and cyber-attacks.

Japan adopted a cybersecurity framework law in 2014. This requires the government to establish clear responsibilities for operators of critical infrastructure and strengthens

the role of the National Center of Incident Readiness and Strategy for Cybersecurity.15

An updated cybersecurity strategy adopted in 2015 highlights the need for “a policy

platform to build a common understanding on security measures required for the entire IoT system as well as for the individual components of IoT systems” (Government of Japan, 2015).

In March 2017, the Japan Electricity Information Sharing and Analysis Center (ISAC)

was set up to share and analyse cyber-risk information and build capability in the electricity industry. Japan also specifically calls for strengthening international

co-operation among ISAC groups and developing international rules and norms regarding cyberspace, as well as capacity building in other countries.16

Many countries also organise large-scale exercises to test emergency preparedness, providing an important opportunity to identify vulnerabilities and to strengthen

public-private co-operation. Leading examples of country and regional cyber-attack preparedness include the following:

• NERC organises a biennial electricity grid exercise, GridEx, simulating both cyber and physical attacks. GridEx 2015 involved more than 300 organisations across North America, including industry, law enforcement, and government agencies (NERC, 2016a).

• The EU Agency for Network and Information Security has co-ordinated major cybersecurity exercises since 2010. The 2014 exercise included an energy

infrastructure scenario in which several operators were severely impacted by a cyber-attack in the middle of a harsh winter. The exercise involved over

200 organisations (27 from the energy sector) from the 28 EU member countries. Among its objectives was to provide an opportunity to individual member states to evaluate their internal contingency plans and to explore the

public affairs handling of large-scale cyber incidents. The 2016 exercise

introduced a focus on the IoT, drones, cloud computing and ransomware (ENISA, 2017).

15

State of Japan, Cybersecurity Basic Act, No. 104 of 2014,

www.shugiin.go.jp/internet/itdb_gian.nsf/html/gian/honbun/houan/g18601035.htm. The Initiative for Cyber Security Information-Sharing Partnership of Japan (J-CSIP) promotes information sharing in seven

priority Special Interest Groups, including electricity, oil and gas industries.

132

Cross-cutting risks: Cybersecurity, privacy and economic disruption

© OECD/IEA, 2017

16

• Various exercises are also held at national or regional level; in 2015, for

example, Nordic countries organised an exercise on co-ordinated handling of a potential disruption in the regional power system.17

International efforts can also play a critical role in helping governments, companies and others to build up digital resilience. A variety of organisations are involved in this

effort in one form or another, including the Group of Seven (G7), the IEA (see Box 6.2),

the Organisation for Economic Co-operation and Development (OECD), and national security organisations such as the North Atlantic Treaty Organization (NATO). For

example, work by the OECD Committee on Digital Economy, Policy Working Party on Security and Privacy in the Digital Economy includes a Recommendation on the

Protection of Critical Information Infrastructures (OECD, 2008; currently being updated).

Ensuring that each international organisation focuses on its own comparative strengths, complements others’ efforts, and establishes open lines of communication

(as appropriate) will be critical. For example, the NATO Warsaw Summit in July 2016 highlighted a “need to enhance strategic awareness on energy security, including through sharing intelligence and through expanding our links with other international organisations such as the IEA” (NATO, 2016). Box 6.2

IEA digital resilience efforts

Digital resilience is a growing challenge for the IEA family of countries and corporate partners. Identification of digital risks and appropriate resilience measures needs to be core to technology

policy development and the design of various regulatory frameworks for energy markets. Governments and energy system operators need to work together proactively to manage the increasing complexity of risks and threats.

As the world’s premier energy security organisation, the IEA stands ready to help its members by

supporting efforts to strengthen digital resilience. Certain aspects of digital resilience are

appropriate for IEA analysis and multi-party collaboration. These include: raising awareness; helping to establish the business case for utilities to become cyber-prepared; helping to mainstream digital resilience into a variety of government and company plans and strategies; and developing and

sharing best practices and policies. Operational aspects of digital security, such as threat

assessment and monitoring, and incident management and response, are sensitive matters of national security and outside of the scope of IEA analysis. The IEA work on energy security raises awareness among countries on new digital risks within a wider context of planning for resilience over the short, medium and long term. The IEA conducts

emergency response reviews of its members; these reviews investigate, among other topics,

NordBER meeting 9-10 September 2015 in Reykjavik, “Energy shortage: Coordinated handling of a potential disturbance in the

Nordic power system”.

Cross-cutting risks: Cybersecurity, privacy and economic disruption

133

© OECD/IEA, 2017

17

whether members have robust national governance arrangements for internal co-ordination and

information sharing on resilience to a wide variety of risks. In the future, these reviews will also include governance of digital resilience. The IEA also organises emergency response exercises on a regular basis to test the preparedness of

the IEA and its member countries to respond to oil, natural gas and electricity disruptions. These exercises expose participants to analysis conducted by the IEA Secretariat in assessing various disruption scenarios, which may include potential high-level impacts on energy markets resulting from a cyber-attack. In a follow-up to this report, the IEA will undertake a more thorough overview of policies and

practices on digital resilience among its members. This will then be used to develop best policy recommendations that can be applied in reviews and exercises.

Data privacy and ownership As in other sectors of the economy, the growing use of digital technologies in the

energy sector raises questions about data privacy and ownership. Smart grids and demand response technologies rely on vast quantities of consumer-specific, real-time

electricity usage data. In the case of active management by utilities or aggregators of behind-the-meter smart home appliances, this will include records of personal energy

use events, such as heating water for a shower or opening a refrigerator, giving anyone who has access to those data a snapshot of householders’ personal daily routines and activities (Figure 6.2).

How much information people are comfortable sharing with service providers, how confidentiality can best be protected, who owns this type of consumer-specific data

(the customer/prosumer who generates it or the service provider who collects it), and who can use or share data and for what purposes, are becoming critical questions.

Many household consumers want to protect their anonymity; they may worry that energy data could be linked to other personal information, such as health records, to create a profile or score. Such data could be used for commercial data mining that

might expose them to intrusive targeted marketing techniques. This kind of detailed

behind-the-meter energy data could also be used to find out when buildings are empty, and could be used for criminal activity (National Geographic News, 2012).

For industrial and commercial consumers, demand response infrastructure that tracks

energy use might reveal information about business practices and operations that that

might be considered as proprietary. Data breaches might impact the willingness of

customers to use new services, something that operators that collect energy data – for

134

Cross-cutting risks: Cybersecurity, privacy and economic disruption

© OECD/IEA, 2017

example, utilities and energy management service providers – must protect against.

Figure 6.2 Behind-the-meter energy use data hob heaters

power, kW

peak = 7.18 kW mean = 0.49 kW daily load factor = 0.7 energy consumption = 11.8 kWh

oven preheating

toaster

kettle

washing machine

oven cycling

kettle

refrigerator

0

2

4

6

8

10

12

14

16

18

20

22

0

time of day, h

Key message: Some energy consumers are concerned that detailed behind-the-meter data could reveal personal daily routines.

Notes: Data from the customer’s side of the utility meter; h = hour; kW = kilowatt; kWh = kilowatt hour. Source: Newborough and Augood (1999), “Demand-side management opportunities for the UK domestic sector” (reproduced courtesy of the Institution of Engineering and Technology).

At the same time, consumers and companies will sometimes want to share their

energy data with third parties, for example to a marketplace for energy efficiency services. Utilities see opportunities to monetise energy use datasets, establishing a

new source of revenues. Utilities in New York have proposed that regulators could allow a fee to be charged for making granular “value-added” data (such as aggregated

customer usage information) available to local distributed energy resource (DER) providers.18

A key step to enabling equipment and devices to communicate and make decisions in digitalized energy and transport systems is to establish their interoperability (Box 6.3).

18

Joint Utilities, Supplemental Distributed System Implementation Plan, Case 16-M-0411, November 2016. “Data access is also

important in the context of developing new market-based revenue streams that are tied to providing value-added services. One example of a potential value-added service is a data analysis service that makes available more granular and customized

information to developers and other market participants. The Joint Utilities distinguish between basic data that is available at no incremental cost and value-added data that will be available for a fee. Examples of value-added system data may include data.”

Cross-cutting risks: Cybersecurity, privacy and economic disruption

135

© OECD/IEA, 2017

forecasted load data, circuit voltage profiles, and power quality data. An example of value-added customer data is aggregated

Box 6.3

Interoperability and standardisation

Interoperability of different data formats, technologies and applications allows systems to function and interact with one another in a reliable, safe, secure and user-friendly manner. This smooth

interface may require the alignment of physical, semantic and organisational elements. Routing protocols for distributed energy resources in smart grids are an example of the role of standard setting, and can help to enable the digitalization and automation of demand response. The lack of interoperability of technology standards can be a barrier to the effective use of new technologies; overly narrow one-size-fits-all standards or standards that might prove burdensome or

conflicting, conversely, could slow down innovation and technology deployment. Optimally,

standardisation should build on synergies between the various players rather than lead to fragmentation of the market and duplication of efforts (OECD, 2017).* The digital sector in particular is characterised by rapid development of technologies, and often

technical solutions to interoperability are developed within fora and consortia (e.g. the Industrial Internet Consortium) rather than in formal standard-making bodies. But policy makers also have a

role to play in encouraging or guiding sufficient standardisation so that consumers retain a wide market choice, and are not locked into using products that are only interoperable with other products from the same manufacturer or that are based on the same protocols (European Smart Grids Task Force, 2016).

The European Union is particularly active in standard setting, especially because of the impact that standards can have on the EU single market. A key participant in European work on standards for digital interoperability in energy is the European Committee for Standardization-European

Committee for Electrotechnical Standardization (CEN-CENELEC), which brings together the national

standardisation bodies of 34 European countries and includes as stakeholders industry, consumer rights representatives, trade unions and environmental groups. Recent areas of work by CEN-

CENELEC include the development of standards for electricity and telecommunications networks, energy management systems, data formats for electronic invoicing, and for the proficiency levels to be reached in order to obtain professional qualifications in digital skills. Some of this work builds on

legislation, while some anticipates future policy decisions by European governments. The

European Union also participates in international standard setting, so that around 80% of CENELEC

standards in electrotechnology are adoptions of international standards developed under parallel procedures in the International Electrotechnical Commission (CEN-CENELEC, 2017).

Regarding future challenges for interoperability in smart grids, an interesting resource is the roadmaps and frameworks issued by the National Institute of Standards and Technology (NIST) in the United States. * Standard setting can also help to establish design priorities, as in the case of “(cyber)security by design”.

Balancing privacy and other policy objectives overall systems perspective, policy makers might consider that there is a wide public

interest in making aggregate data available publicly. For instance, such data could 136

Cross-cutting risks: Cybersecurity, privacy and economic disruption

© OECD/IEA, 2017

Policy makers are often caught in the middle of these privacy debates. Viewed from an

provide insights for urban planners or allow researchers to investigate aggregate

efficiency opportunities that cannot be realised by individual consumers on their own. At the same time, policy makers must also take into account the privacy concerns of their constituents.

Some governments may seek arrangements to access data from private firms in

exchange for access to markets, as was the case in a deal struck between the city of

Boston and the ride-sharing company Uber. At the same time, policy makers are also increasingly concerned that the concentration of data collection and processing in the

hands of a few large companies could have possible competition and anti-trust impacts. In this regard, the OECD is undertaking work on how data-driven network

effects, reinforced by user feedback loops, and the great economies of scale associated with IT infrastructure may lead to situations of dominant market power (OECD, 2016a; The Economist, 2017).

Some possible solutions are technical. Data can be made anonymous by means of aggregation so that private information could not be attributed to a specific household. Another solution is to limit data granularity by adding a time lag. In Germany, the law on smart meter data allows transmission of household data only every 15 minutes; in France, 10 minutes is considered sufficient for the purpose of smart grid operations.19

Other solutions involve policy choices about how to balance different societal objectives. Policy makers wanting to encourage the development of demand response

markets will need to strike an appropriate balance between consumers’ privacy concerns, fostering innovation in demand response markets, and the operational needs of utilities.

A key question for policy makers is whether regulation should take an opt-in or an opt-out approach to customer authorisation. Opt-in programmes give maximum protection and require affirmative customer authorisation for certain data to be shared.

Opt-out programmes, however, are more likely to favour mass participation in demand response markets.

Alternatively, customers could be given a range of confidentiality options. “Minimum”

customers might choose only to participate in sufficient data collection to enable “core” smart grid operations, such as load balancing and price formation. “Maximum” customers might agree to detailed data being made available for marketing purposes

to commercial energy efficiency providers, with the aim of learning about possible

19

Federal Republic of Germany, Messstellenbetriebsgesetz vom 29 August 2016; for France, see www.cnil.fr/fr. But such is the

demand (see Chapter 7).

Cross-cutting risks: Cybersecurity, privacy and economic disruption

137

© OECD/IEA, 2017

potential power of these data that time lags of three to six months for detailed data could still transform understanding of energy

savings. One example of this approach is a voluntary code of conduct developed by the US DOE and the Federal Smart Grid Task Force in 2015, which distinguishes between

primary-purpose use of data that can be “reasonably expected,” and more extensive secondary-purpose data sharing that requires an opt-in consent (DOE, 2015).

While each jurisdiction is different and faces different opportunities, cultural context

and circumstances, there are several real-world examples that may be useful to policy makers (Box 6.4).

Box 6.4 •

Examples of specific approaches to privacy regulation

California: Since 2011, California has adopted some of the world’s first rules specifically

addressing smart meter data protection. State laws prohibit utilities from sharing or otherwise

disclosing a customer’s consumption data and patterns to a third party without the customer’s prior consent. The law also requires utilities to provide security, including encryption for usage data gathered through smart meters or otherwise. In 2014 these provisions were extended to internet service providers, financial institutions and other businesses that might handle or receive smart meter data.* •

South Korea: From 2011 to 2015, South Korea adopted a set of strong general-purpose

protections for personal information, specifying limits on data collection, use, outsourcing,

disclosure, editing, searching, storage and destruction. The 2011 Smart Grid Construction and Utilization Promotion Act also emphasises data protection, requiring operators to obtain individual consent for the collection of personal data.** •

European Union: The General Data Protection Regulation (GDPR) was adopted in 2016 and came into full force in May 2018. It establishes an updated framework for personal data

protection in the European Union. The GDPR requires data privacy and customer consent for any data collection or use to be designed into all business processes for products and services, including where foreign-based companies collect or process the data of EU residents. All data

breaches must be reported to the relevant national authorities. A principle of “data portability” establishes a right to transfer personal data from one service provider to another. The GDPR means significant changes to business practices for many companies. Some of the implications for implementation in the energy sector have been analysed by the European electricity sector association Eurelectric (Eurelectric, 2016). •

France: In 2016, French legislation established the concept of “public interest data”, enabling the government to request commercial entities to give access to data that they hold for the purpose of establishing public statistics.***

** Smart Grid Act 2011, 지능형전력망의 구축 및 이용촉진에 관한 법률/知能型電力網法;

*** Article 19 of Loi N°2016-1321 du 7 octobre 2016 pour une République numérique.

138

Cross-cutting risks: Cybersecurity, privacy and economic disruption

© OECD/IEA, 2017

* California Civil Code, Division 3: Part 4: Obligations arising from particular transactions, and draft bill SB-356 on energy data transparency, currently in process of adoption by the legislature, http://leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=201720180SB356; see also, for example, California utility PG&E’s smart meter data collection opt-out program, www.pge.com/en_US/residential/save-energy-money/analyze-yourusage/your-usage/view-and-share-your-data-with-smartmeter/smartmeter-updates/smart-meter-opt-out-program.page;

In many cases, the legal frameworks in force today regarding privacy, consumer

protection and electronic communications are not specifically designed for energy.

Moreover, many of these general frameworks do not adequately establish up-to-date technical definitions. For example, is transmission of consumer energy usage data over

the smart grid an “electronic communication”, or should a smart meter network be

classified as a remote computing device? Lack of clarity or robust protections could hinder participation in sharing digital energy information, and therefore limit overall system efficiency and possibly lead to a backlash from consumers.

As policy makers, regulators, companies and consumers grapple with these problems, they will need to consider implementing a number of actions, including:

• Physical, administrative and technical security measures to protect and, where possible, anonymise customer information.

• Procedures to maintain data quality and integrity, including means for consumers to access and correct any errors in stored personal data.

• Transparency about the purposes for which data will be collected and used, as well as clear limitations on their use.

• Provisions for consumers to give informed consent prior to any release of usage data, particularly to third parties.

• Provisions for consumers to choose to share their data with third parties when the release of such information is beneficial for them, or to freely transfer their data from one service provider to the next.

Economic disruption: Jobs and skills in an increasingly digitalized energy world Throughout the broader global economy, digitalization is changing the nature of existing jobs, creating entirely new jobs and threatening job losses. There will no doubt be both winners and losers as a result of increasing digitalization, and policy makers must play a proactive role in preparing for an uncertain future.

Significant research is under way to explore the impact of digitalization across the broader economy, including the OECD’s Going Digital project. This OECD project

includes a number of cross-cutting initiatives as well as in-depth analysis of economic disruption, including further analysis of jobs and skills, and the potential societal

impacts of job polarisation and income inequality. OECD reports during 2018-19 will put forward specific policy recommendations to help a range of policy makers to

Cross-cutting risks: Cybersecurity, privacy and economic disruption

139

© OECD/IEA, 2017

navigate and manage the digital transition.

Building on IEA analysis of the impact of investment on energy employment (IEA,

2017), a review of key energy sectors demonstrates the many – and varied – ways in

which digital technologies can affect jobs in the energy sector.20 Overall, digitalization is likely to lead to further efficiencies along the supply chain, but is less likely to replace still-sizeable labour needs for major engineering and construction activity related to physical infrastructure.

Workers supporting digital infrastructure will need specialised ICT skills, such as coding and cybersecurity, while across the energy sector, all workers will need generic

ICT skills to operate digital technologies. Complementary “soft” skills such as leadership, communication and teamwork skills will also be required for the expanding number of opportunities for ICT-enabled collaborative work (OECD, 2016b).

While digitalization may lead to lower labour intensity across energy systems as a whole, the impacts of digitalization are highly sector-specific:

• In the life cycle of a power plant, digitalization has the greatest impact on equipment – its manufacturing, siting, and operation and maintenance – raising

the productivity and reliability of the plant and therefore potentially reducing

labour intensity. In thermal generation, digitalization may change the tasks of

existing operation and maintenance, while creating new jobs in data science. In the renewables sector, the use of robots to clean solar panels and drones to monitor wind turbines could reduce the need for some employees.

• In upstream oil and gas, a large portion of employment is associated with initial

field development. Digitalization and other innovations have helped to lower

costs and raise productivity, although reductions in employment are difficult to disaggregate from the wider effects of the lower oil price environment. The widespread use of 3D seismic analysis has reduced drilling needs, but created

new jobs in ICT and data science roles, which require very different skill sets and often are located in a different region than drilling operations.

• In energy efficiency of buildings, almost all jobs are upfront by nature (e.g. construction and refurbishment), that is, fewer jobs are associated with

the operation of the energy-efficient equipment once it installed. That said,

those working on the operations and maintenance of buildings may need

additional skills to deal with new technologies through training and development, as the lack of new skills may be a barrier to energy efficiency.

20

Empirical studies of energy employment variously measure employment effects in up to three categories: Direct employment

(construction workers and facility operators); Indirect employment (persons employed to produce inputs for the new facility, in employment.

140

Cross-cutting risks: Cybersecurity, privacy and economic disruption

© OECD/IEA, 2017

manufacturing industry or service sectors); Induced employment (jobs generated by consumption as the wages earned in direct

• In manufacturing, some jobs involve predictable, routine and repetitive physical tasks, putting them at a higher risk of automation. While the use of robotics and

automation has historically yielded substantial operational benefits (including safety and productivity), the deployment of such technologies requires careful consideration of its impact on jobs, in particular job loss for low-skilled

workers. At the same time, the increased use of data, sensors and 3D printing could provide new job opportunities in advanced manufacturing.

• In transport, enhanced connectivity has created new employment opportunities

on ride-hailing app platforms, but fully autonomous cars could threaten to eliminate the jobs of substantial numbers of human drivers (Box 6.5).

Box 6.5

Automation and jobs: Focus on autonomous cars and trucks

The potential impact of automation on jobs has been the focus of a series of studies from academia, international organisations and consulting firms over the past five years (Arntz, Gregory and Zierahn, 2016; Bughin et al., 2017; Frey and Osborne, 2013; PWC, 2017). This work highlights that jobs

composed of a high share of automatable tasks – such as those involving predictable, routine and repetitive physical activities, and the collection and processing of data – are at higher risk of

automation than less routine activities. In the energy sector, significant attention has focused on the potential impact of autonomous cars on jobs.

Fully autonomous cars and trucks* could threaten the livelihood of commercial drivers around the

world. The United States alone has around 3.5 million truck drivers, 665 000 bus drivers, 230 000 taxi drivers and at least 500 000 active drivers with ride-sharing companies Uber and Lyft. Digitalization

could therefore have immense impacts on jobs even in this single sector, especially if technological change and adoption is rapid. Fears of widespread job losses, however, need to be tempered with a closer look at the labour

market and demographics, as well as technical feasibility, public acceptance and the regulatory

environment. For example, the US trucking industry already has a shortage of around 50 000 drivers,

growing to 800 000 by 2030 (ITF, 2017), and the turnover rate of drivers is very high due to the nature of the work (long working hours, extended periods of time away from home, safety and risk

issues). Moderate to high levels of automation in trucks, especially over the near term, could be

deployed to complement human drivers, helping bridge the gap of driver shortages and job creation. But even with fully autonomous or connected platooning trucks on highways, human drivers may be needed for last-mile, in-city driving where decision-making is more complex.

The pace and extent of automation and its impacts on jobs in the energy system remain highly uncertain, and will depend on a number factors that will vary across regional and sectoral contexts

(Bughin et al., 2017). Facing an uncertain future, governments must prepare for a variety of scenarios, including a rapid transition to fully autonomous cars and trucks. Detailed analysis and key

policy recommendations specific to potential labour impacts resulting from autonomous trucks are available in a recent report released by the International Transport Forum: Managing the Transition * Autonomous driving technologies and other digitalization trends in transport are discussed in Chapter 2.

Cross-cutting risks: Cybersecurity, privacy and economic disruption

141

© OECD/IEA, 2017

to Driverless Road Freight Transport (ITF, 2017).

References Arntz, M., Gregory, T., and Zierahn, U. (2016). “The Risk of Automation for Jobs in OECD Countries: A Comparative Analysis”. OECD Social, Employment and Migration Working Papers, 2(189), 47– 54. http://doi.org/10.1787/5jlz9h56dvq7-en.

Bughin, J. et al. (2017). “A Future That Works: Automation, Employment, and Productivity”.

McKinsey Global Institute.

Cambridge Centre for Risk Studies (2016). Cyber Risk Framework for Critical Infrastructure Threat

Scenario: Mapping the Consequences of an Interconnected Digital Economy.

www.itrc.org.uk/integrated-infrastructure-cyber-resiliency-in-society-mapping-theconsequences-of-an-interconnected-digital-economy/#.WXNAbXkUmUk.

Carr, M. (2016). “Public–private partnerships in national cyber-security strategies”. International

Affairs, 92: 1, 2016.

www.chathamhouse.org/sites/files/chathamhouse/publications/ia/INTA92_1_03_Carr.pdf.

CEN-CENELEC (2017). Annual Report 2016.

www.cencenelec.eu/news/publications/Publications/Annual_Report_2016_Tome_1_accesibility. pdf.

ENISA (European Union Agency for Network and Information Security) (2017). Cyber Europe 2016:

After Action Report. June 30, 2017. www.enisa.europa.eu/publications/ce2016-after-action-

report. Published on February 01, 2017.

ESET (2017). Win32/Industroyer: A New Threat for Industrial Control Systems.

www.welivesecurity.com/wp-content/uploads/2017/06/Win32_Industroyer.pdf.

Eurelectric (2016). The Power Sector Goes Digital: Next Generation Data Management for Energy

Consumers. www.eurelectric.org/media/278067/joint_retail_dso_data_report_final_11may_as2016-030-0258-01-e.pdf.

European Parliament (2016). Cyber Security Strategy for the Energy Sector. ITRE Committee.

www.europarl.europa.eu/RegData/etudes/STUD/2016/587333/IPOL_STU(2016)587333_EN.pdf.

European Smart Grids Task Force (2016). My Energy Data. Report by European Smart Grids Task

Force, Expert Group 1: Standards and Interoperability.

https://ec.europa.eu/energy/sites/ener/files/documents/report_final_eg1_my_energy_data_15_ november_2016.pdf. Frey, C. B., and Osborne, M. (2013). The Future of Employment: How Susceptible are Jobs to Computerisation? Oxford, UK. Gartner. (2017). Gartner Says 8.4 Billion Connected “Things” Will Be in Use in 2017, Up 31 Percent From 2016. Retrieved July 12, 2017, from www.gartner.com/newsroom/id/3598917 (accessed 12

142

Cross-cutting risks: Cybersecurity, privacy and economic disruption

© OECD/IEA, 2017

July 2017).

Global Smart Grid Federation (2016). Smart Meter Security Survey, August 2016. www.globalsmartgridfederation.org/wp-

content/uploads/2016/08/smart_meter_security_survey.pdf. Government of Japan (2015). Cyber Security Strategy, provisional translation. Cabinet decision of 4 September 2015. www.nisc.go.jp/eng/pdf/cs-strategy-en.pdf.

Government of Mexico (2014). Programa para la Seguridad Nacional 2014-2018: Una Política

Multidimensional para México en el Siglo XXI [National Security Programme 2014-2018: A

Multidimensional Policy for Mexico in the 21st Century].

http://cdn.presidencia.gob.mx/programa-para-la-seguridad-nacional.pdf.

HMG (Her Majesty’s Government) (2016a). National Cyber Security Strategy 2016-2021. H M Government.

www.gov.uk/government/uploads/system/uploads/attachment_data/file/567242/national_cybe r_security_strategy_2016.pdf.

HMG (2016b). Cyber Essentials Scheme: Requirements for Basic Technical Protection from Cyber

Attacks. UK National Centre for Cyber Security. www.ncsc.gov.uk/scheme/cyber-essentials.

IEA (International Energy Agency) (2017). World Energy Investment 2017. OECD/IEA, Paris, France. ITF (International Transport Forum) (2017). Managing the Transition to Driverless Road Freight

Transport. OECD/ITF, Paris, France.

Larkin, S. et al. (2015). “Benchmarking agency and organizational practices in resilience decision making”. Environment Systems and Decisions, 35(2), 185-195.

Lloyd’s (2015). Emerging Risk Report. www.lloyds.com/news-and-insight/riskinsight/library/society-and-security/business-blackout.

Madnick, S. (2017). “Preparing for the cyberattack that will knock out U.S. power grids”. Harvard

Business Review. https://hbr.org/2017/05/preparing-for-the-cyberattack-that-will-knock-out-us-power-grids.

Marchese, D. and Linkov, I. (2017). “Can you be smart and resilient at the same time?”

Environmental Science & Technology/Research Gate, May 2017.

www.researchgate.net/publication/316849102_Can_You_Be_Smart_and_Resilient_at_the_Same _Time. Mittal, A., Slaughter, A. and Zonneveld, P. (2017). Protecting the Connected Barrels:

https://dupress.deloitte.com/content/dam/dup-us-en/articles/3960-connected-

barrels/DUP_Protecting-the-connected-barrels.pdf. National Academies of Sciences,

Engineering, and Medicine (2017). Enhancing the Resilience of the Nation's Electricity System. www.nap.edu/catalog/24836/enhancing-the-resilience-of-the-nations-electricity-system.

National Geographic News (2012). Who’s Watching? Privacy Concerns Persist as Smart Meters Roll

Out. (14 December 2012). http://news.nationalgeographic.com/news/energy/2012/12/121212-

Cross-cutting risks: Cybersecurity, privacy and economic disruption

143

© OECD/IEA, 2017

smart-meter-privacy/.

NATO (North Atlantic Treaty Organization) (2016). Warsaw Summit Communiqué. Issued by the Heads of State and Government participating in the meeting of the North Atlantic Council in Warsaw 8-9 July 2016. www.nato.int/cps/en/natohq/official_texts_133169.htm. NERC (North American Electric Reliability Corporation) (2016a). Grid Security Exercise: GridEx III

Public Report. www.nerc.com/pa/CI/CIPOutreach/GridEX/NERC%20GridEx%20III%20Report.pdf.

Newborough, M. and Augood, P. (1999). “Demand-side management opportunities for the UK domestic sector”. In IEE Proceedings – Generation, Transmission and Distribution, 146: 3.

OECD (Organisation for Economic Co-operation and Development) (2017). Key Issues for Digital

Transformation in the G20. Report prepared for a joint G20 German Presidency/OECD

conference. www.oecd.org/G20/key-issues-for-digital-transformation-in-the-G20.pdf. OECD (2016a). Big Data: Bringing Competition Policy to the Digital Era.

www.oecd.org/competition/big-data-bringing-competition-policy-to-the-digital-era.htm.

OECD (2016b). Skills for a Digital World. www.oecd.org/els/emp/Skills-for-a-Digital-World.pdf. OECD (2008). Recommendation on the Protection of Critical Information Infrastructures. www.oecd.org/sti/40825404.pdf

PWC (2017). UK Economic Outlook. Retrieved from www.pwc.co.uk/economic-services/ukeo/pwcuk-economic-outlook-full-report-march-2017-v2.pdf. The Economist (2017). “Fuel of the future: Data is giving rise to a new economy. How is it shaping up?” Edition 6 May 2017. www.economist.com/news/briefing/21721634-how-it-shaping-updata-giving-rise-new-economy.

US DOE (United States Department of Energy) (2015). Data Privacy and the Smart Grid: A

Voluntary Code of Conduct. Final Concepts and Principles.

www.energy.gov/sites/prod/files/2015/01/f19/VCC%20Concepts%20and%20Principles%202015 _01_08%20FINAL.pdf. US DOE (2011). Roadmap to Achieve Energy Delivery Systems Cybersecurity. Energy Sector Control Systems Working Group.

https://energy.gov/sites/prod/files/Energy%20Delivery%20Systems%20Cybersecurity%20Road

144

Cross-cutting risks: Cybersecurity, privacy and economic disruption

© OECD/IEA, 2017

map_finalweb.pdf.

Chapter 7

Policy Highlights •

Technological advances and falling costs are driving the digital

transformation of energy systems, but policy and market design are critical to help steer the digital transformation onto a secure and sustainable path.



Fully harnessing digital tools can help achieve specific policy objectives. For example, enabling policy regimes and new business models could assist in providing electricity to the 1.1 billion people who still lack

access to it. New digital tools can also promote sustainability, for example the use of satellites to monitor and verify greenhouse gas emissions, and cutting-edge tools to track air pollution at the neighbourhood level.



Policy-making processes can also benefit from more timely and sophisticated collection and publication of key energy data. Emerging low-cost digital tools, such as online registries, web-crawled data and

quick response codes can lead to more targeted and responsive programmes.



While there is no simple roadmap for what an increasingly digitalized

energy world will look like in the future, this chapter recommends ten no-regrets policy actions that governments can take to prepare. Technological advances and falling costs are driving the digital transformation of energy systems, but policy and market design are critical to help steer the digital

transformation onto a secure and sustainable path. This chapter considers how digitalization can help achieve specific policy objectives, for example, energy access,

sustainability and energy security. It then examines how digital tools can improve the policy-making process itself, and summarises lessons learned from the entire report about implementing well-considered, integrated policy frameworks. Finally, this chapter – and this report – conclude with a series of no-regrets recommendations to

digitalization in energy.

Policy

145

© OECD/IEA, 2017

help policy makers best navigate the inherently undefinable future of increasing

Promoting energy access, sustainability and security As discussed throughout this report, while the digitalization of energy may not be a policy end in itself, it can be a means to help achieve a variety of energy policy

objectives. These include increased productivity and efficiency, improved safety, enhanced revenue collection, and accelerating the pace of innovation. Three specific policy objectives are now explored in greater depth: energy access, sustainability, and security.

Energy access In sub-Saharan Africa and other developing countries, decentralised energy sources,

such as off-grid solutions, can be an effective and rapid option to bring electricity

access to the population. Receptive policy environments and new business models are tapping the opportunities presented by digital connectivity, including smart payments and mobile phones.

Access to modern and reliable forms of energy is a critical enabler of human and

economic development. According to the latest analysis from Energy Access Outlook:

from Poverty to Prosperity, 1 around 1.1 billion people still lack access to electricity,

although momentum is building towards the Sustainable Development Goal of ensuring access to affordable, reliable, sustainable and modern energy for all by 2030. Digitalization, especially when guided by policy, can help accelerate the pace.

In certain countries in sub-Saharan Africa, mobile phones are more prevalent in homes

than electricity, and mobile phones and the associated infrastructure, such as cell

towers, may be able to help facilitate access to a large array of energy services (Figure 7.1). Companies such as BBOXX, M-Kopa, Off-Grid Electric and Mobisol have developed new business models for the provision of energy access that target areas

covered by mobile networks but not electricity grids. There are two main business

models that utilise pay-as-you-go (PAYG) mobile financing and payment schemes to provide access to renewable off-grid systems:

• Rent-to-own, where consumers can use their phones to pay a fixed amount up-front for an off-grid device, such as a solar home system, and then pay for the rest of it in affordable instalments.

• Solar-as-a-service, where consumers pay a regular fee for the energy service, but never actually own the device.

For more on energy access, see Energy Access Outlook: from Poverty to Prosperity (IEA, 2017), part of the World Energy Outlook

146

Policy

© OECD/IEA, 2017

1

series, at www.iea.org/energyaccess.

Figure 7.1 Mobile phone ownership and electricity access in selected sub-Saharan African countries 100% 80% 60% 40%

Mobile phone ownership in rural households (2015)

20% 0%

Rural electrification rate (2016) Uganda

Tanzania

Nigeria

Kenya

Ghana

Key message: Mobile phones provide an opportunity to accelerate electricity access. Sources: IEA (2017), Energy Access Outlook 2017: From Poverty to Prosperity; USAID (2017), STATcompiler.

Crucially, mobile networks enable companies to remotely monitor products and collect usage data, disable a device when a customer misses a payment, and turn the device

back on when the payment has been made. The availability of well-established telecoms, widely used mobile money systems and relatively friendly business markets

has been a strong driver of the location of these businesses. At the end of 2016,

East Africa accounted for 55% of the registered mobile money accounts in sub-Saharan Africa, followed by West Africa at around 30% (GSMA, 2017a). As a result, PAYG energy service business models are most active in four countries: Kenya

(where the PAYG model started, see Table 7.1), Tanzania, Rwanda and Uganda. Other markets are opening up, however, especially in Ethiopia, Ghana, and Nigeria. Illustrative electricity offers from PAYG market leaders in Kenya

Company M-KOPA

Products offered 8 W solar home system (4 x 1W LED, flashlight, radio)

BBOXX

15 W solar home system (4 x 1W LED, radio) 50 W solar home system (4 x 1W LED, radio, LED TV)

Cost (USD) Upfront deposit: USD 33.50

Per day: USD 0.48 (for 365 days) Total: USD 208

Upfront deposit: USD 9

Per month: USD 9 (for 36 months) Total: USD 336

Upfront deposit: USD 21

Per month: USD 21 (for 36 months) Total: USD 761

Notes: Values as of 31 January 2016 and are described here in USD, although customers pay in their local currency; LED = light-emitting diode; W = watt. Source: Adapted from Sanyal et al. (2016)

Policy

147

© OECD/IEA, 2017

Table 7.1

Paired with PAYG financing, solar home systems (consisting of a solar module and a

battery) are the most popular and successful off-grid solutions, often substituting for kerosene and diesel. On average, these systems provide customers with basic energy access at Tier 1 and Tier 2 (GSMA, 2016a and 2016b).2

Increasingly, these systems are bundled with energy-efficient appliances, which can

offer higher levels of modern energy services at lower cost and are able to be

supported by the off-grid system. Over the past four years, over USD 360 million has been invested in PAYG solar (GSMA, 2017a and 2017b). Despite the investment, to date it remains a relatively small market.

The emergence of this new model does not need to be in direct competition with existing access efforts, especially as it represents an inflow of additional capital from

private investors. Some governments are entering into partnerships with companies to

distribute solar home systems, such as the partnership recently forged between the Republic of Togo and BBOXX, where the goal is to distribute over 300 000 solar home

systems in Togo over the next five years (ESI Africa, 2017). In some cases, digitally enabled access models are causing some existing players, such as utilities, to rethink

their role and approaches to providing access. Some utilities are exploring how they might use the electricity provided to telecom towers, which provide the load and

security of payment necessary to justify greater investment, as an anchor for the electrification of nearby households. In this way, beyond just the use of information

and communications technology (ICT) for solar home systems, the potential exists for a mutually beneficial expansion of ICT-related infrastructure and electricity access.

In addition, increasing attention has focused on how mobile technology beyond just

digital payment, such as cloud-based metering and software platforms, can also be paired with larger systems such as minigrids. These grids could offer customers

additional services and support commercial activities and agriculture, contributing to economic growth.

There are challenges, however, especially related to financing and how to scale up the

products offered by PAYG providers to offer a higher level of energy services to consumers. Policy uncertainty and a lack of transparency can create the perception of risk, discouraging investment and halting the progress of decentralised systems and

non-state actors. Co-ordinated planning is important and should take account of approaches to upgrading existing systems and integrating decentralised systems into

the grid when or if it arrives. Additionally, while some of these new business models are competitive without aid or subsidy, they have much greater effect if government

2

Tier 1 access means enough energy for a light and phone charging. Tier 2 can run general lighting, a TV and a fan.

148

Policy

© OECD/IEA, 2017

policies set appropriate tariff structures and simplify the permitting and licensing

processes. Fundamentally, for these new models to be effective, they must offer

services that are affordable to consumers. Rural customers may pay more for their electricity than urban users or be unable to use higher-powered appliances, because

the unit cost of electricity is typically higher for decentralised systems than a grid tariff

unless the government decides in favour of cross-subsidisation to even costs out, as has been the case historically.

Environmental sustainability This report identifies a number of ways in which digitalization can help policy makers

to improve energy systems’ environmental impact. These include energy use

improvements in transport, buildings and industry that can help reduce air pollution

and greenhouse gas emissions. For example, in industry – a sector responsible for 24% of total carbon dioxide (CO2) emissions in 2014 – there is clear value in digitalization to

improve the efficiency of energy and materials use (Chapter 2). Drones and data

analysis can improve the detection and measurement of methane emissions, at low cost, in oil and gas operations (Chapter 3). Digitalization can help better integrate

variable renewables such as wind and solar, highlighting the potential to avoid 240

million tonnes of CO2 emissions cumulatively to 2040 in the European Union alone (Chapter 4). Meanwhile ICT companies are playing a strong role in renewables investment, including some companies committing to 100% renewable energy procurement goals (Chapter 5).

However, in certain circumstances digitalization could actually increase energy use

and emissions. For example, self-driving vehicles might encourage increased travel.

Digitally enabled efficiency gains will not only lower the cost of wind generation, but also coal. Smart controls and connected devices consume energy in order to maintain

connectivity, even when on standby. Connected devices also have energy and

environmental impacts from manufacture and disposal (e.g. electronic waste). It is thus important that policy makers bear in mind that digitalization’s impact may not always align with environmental protection and that well-thought-through policies will need to take all of these dynamics into account.

Specific technologies that are critical for sustainability might also receive a significant boost from digitalization. Take the ability of carbon capture and storage (CCS) to reduce CO2 emissions from power and industry. Each element in the CCS value chain

could significantly benefit from advances in digital technologies (Box 7.1).

Wide scope also exists for more creative use of digital technologies, both by

companies and policy makers, to help promote environmental sustainability. For example, many cities measure air pollution using static monitoring stations, but high

Policy

149

© OECD/IEA, 2017

costs can limit deployment. Pilot projects are extending this network into the Internet

of Things (IoT) by means of additional sensors added to vehicles. Google even has a new initiative to track air pollution in its Streetview map.

Box 7.1

Digitalization and CCS

Digital technology applications for CO2 capture are similar in nature and benefit to digitalization in industry and power generation. Specifically, optimisation of control processes through automation

and enhanced data collection and analytics are likely to reduce overall costs. Digitalization can be increasingly important in the future when CO2 emissions from smaller and more dilute sources will need to be captured and stored.

Digitalization of CO2 transport and storage has been evolving along the same lines as the digitalization of upstream oil and gas (see Chapter 3). In fact, much of the digital transformation and innovation from the oil and gas industry appears to be transferable to CO2 storage assessment and development. As in the oil and gas industry, the nascent CO2 storage industry will be dealing with

large volumes of time-series information/data captured from instruments that monitor and control plant processes. CCS will also face similar “big data” challenges as oil and gas operations, with both structured and unstructured data for assets spread over wide geographies. Future CO2 storage projects should be able to benefit from key technology innovations that are revolutionising the oil and gas industry (Managi et al, 2004, 2005). Such technologies include 3D

seismic (Bohi, 1999) as well as smart drilling and other developments related to directional, multi-stage hydraulic fracturing and unconventional gas resources (Wood Mackenzie, 2014). It is also necessary to integrate multi-dimensional and multi-disciplinary models and tools to

understand the performance of reservoirs where CO2 is injected and assess impacts on other resources and the environment over large areas. Real-time leakage detection and CO2 monitoring

using state-of-the-art technologies over different domains (subsurface, surface and atmospheric)

and over time (prior to, during and post-injection) will be essential to assure both the public and regulators that CO2 storage is safe, to estimate how much CO2 is stored and to minimise risks of leakage.

Another example will be the use of drones (see Box 2.1) to monitor remotely the CO2 storage sites, both offshore (www.stemm-ccs.eu/) and onshore. Technology could improve safety with unmanned

CO2 injection at offshore platforms, which could be operated remotely thanks to smart sensors and controls.

The world’s largest supercomputer is being used by digital energy company Envision Energy in China to design and optimise sites for wind farms.3 In the solar sector, GTM

Research’s Solar Data Hub tracks and forecasts solar installation markets across the

3

www.envision-energy.com/2016/10/18/envision-energy-launches-energy-analytics-platform-ensight/.

4

www2.greentechmedia.com/solardatahub.

150

Policy

© OECD/IEA, 2017

United States.4 And start-up company BioCarbon Engineering is using drones to plant

trees by firing up to 100 000 germinating seed pods per day into the ground from the air in an effort it calls “industrial-scale reforestation” (ABC News, 2017; WEF, 2017).

Another creative example of how digital technologies can advance sustainability policy objectives is work by national space agencies to improve the monitoring, reporting and

verification of greenhouse gas emissions. More precise accounting is critical for verification schemes and towards ensuring integrity in carbon certification schemes such as carbon markets.

Today, such accounting is usually calculated using data on fossil fuel consumption,

land-use, and/or economic trends. An alternative method is being pioneered by the

Japan Aerospace Exploration Agency and NASA, using satellites that track CO2 and methane from space by measuring concentrations of gases in a column of air, combined with results from a network of ground-based monitoring stations (Kornei, 2017).

The technology is complex and satellite launches are expensive and difficult to schedule, but by 2030 several satellites are expected to be operational, forming a

co-ordinated fleet of monitoring stations shared by several space agencies. This effort builds on a joint declaration by space agencies in 2015 that “satellite observations are

the key element of a global measuring system aimed at verifying the reality of commitments taken in line with the United Nations Framework Convention on Climate Change” (International Academy of Astronautics, 2015).

Both the United Nations and the World Bank have supported programmes exploring the potential of big data to provide new insights in monitoring, analysing and adapting

to climate change. The UN initiative “Data for Climate Action,” launched in March 2017, challenges data-rich organisations, both public and private, to make large

amounts of data available for free. Scientists, researchers and entrepreneurs, in turn,

could then use these data flows to identify novel insights and means to influence human behaviour and its effect on climate change.

Policy makers should be aware of these types of new digitally enabled opportunities to

promote sustainability, including more cutting-edge options.5 This, in turn, will help

policy makers to be able to implement more cost-effective policies, including additional public-private partnerships.

See also recent reports from the Global e-Sustainability Initiative (GeSI), which highlight opportunities from the energy sector and

beyond, e.g. SMARTer2030: http://smarter2030.gesi.org/.

Policy

151

© OECD/IEA, 2017

5

Energy security While digitalization can increase cybersecurity risks (see Chapter 6), it can also bring

positive security benefits. For example, among the supply-side effects of digitalization (Chapter 3) are improvements in the management of energy infrastructure

maintenance. Sensors and near real-time analytics are being used to monitor and predict early-stage problems with individual items of equipment before a mechanical

failure can interrupt operations. In this way, digitalization can help to reduce unplanned outages. Close monitoring of equipment can also help to spot human error

and can quickly identify the exact location of damage due to external factors such as extreme weather.

Other digitalization elements that can help energy security are on the demand side (Chapter 2). Energy saving has underpinned energy security efforts since the 1970s,

and digitalization offers a significant chance to further strengthen these efforts, notably in residential and commercial buildings, where a 10% reduction in total energy

demand to 2040 could be achieved by means of smart controls on heating and cooling, lighting, and appliances. Better matching of energy demand to supply fluctuations by means of smart demand response also improves energy security, and is of increasing

importance as electricity systems integrate larger shares of variable wind and solar (Chapter 4).

Digital tools to improve the policy-making process Digitalization is not only a powerful tool to achieve a myriad of policy objectives, it can

also help improve the policy-making process itself. This section discusses the impact of digitalization on energy statistics, followed by specific examples of how new digital tools can assist policy makers with standards for energy-efficient appliances.

Better energy statistics Collecting and harnessing digital data (Box 7.2) has the potential to revolutionise our understanding of energy and shed new and vital light on how and why energy is used. With access to digital data and new methods for collecting and combining data, the quality, timeliness and availability of energy data can be transformed. This, in turn, can

help improve the decision-making process of policy makers, companies and other

actors.

Take the example of electricity production and consumption. If power generators

152

Policy

© OECD/IEA, 2017

collect real-time information in a digital form, potentially this could be supplied, under

the right conditions, to official energy statistical institutions. This could help make possible the ability to have near real-time official generation and fuel input data.6 Box 7.2

Types of digital data

Data is a widely used term with many meanings. In the context of digitalization, it is worth considering three interlinked but separate definitions: •

Statistics are facts such as a country’s annual energy balance, or more formally, “numerical statements of facts in any department of inquiry placed in relation to each other.”*



Structured data are organised large datasets, usually generated by state administrations, for

example, information about feed-in-tariffs or energy audits. Structured data can also come from the energy industry, for example from electricity meter readings. •

Big data are large amounts of data gathered from a range of diverse sources, often in near real time. Because the data may be collected through business or control processes that have not

been specifically designed for data analysis, ICT consultancy Gartner defines big data as “high volume, high velocity, and/or high variety information assets that require new forms of processing to enable enhanced decision making, insight discovery and process optimization.” Note: * Lyon Bowley, Sir Arthur (1915), An Elementary Manual of Statistics.

Digital data can also revolutionise an evolving process known as “data fusion”, in

which datasets are created which are far more powerful than a simple sum of their parts. Currently there are two main trends in data fusion with relevance to energy

analysis: geographical information systems (GIS) and data aggregation. GIS allows

different data to be combined as layers into a single map.7 The map can be used to display average wind speeds at different heights together with solar radiation, the plan

of the electricity grid, transport infrastructure, cadastre information, and protected areas. Combining these data into a single map means that the evaluation of a potential

new energy project takes much less time, with very precise results, facilitating investment and decreasing risks.

6

However, this supply-side data would still be insufficient to track electricity demand trends, as data from grid operators and

electricity retailers does not capture energy use by self-consumers – such as households which have solar photovoltaic (PV)

panels or batteries installed behind-the-meter (see Chapter 4). Current data collection procedures need to be adjusted to be able to cover this market segment.

7

Examples of energy-related GIS include: the Swiss Confederation Geocatalog (https://map.geo.admin.ch); the Australia

Renewable Energy Mapping Infrastructure (AREMI, http://nationalmap.gov.au/renewables/); the Netherlands National Energy Atlas (www.nationaleenergieatlas.nl/en/kaarten); Natural Resources Canada’s GeoApp

(http://geoappext.nrcan.gc.ca/GeoCanViz/map/nacei-cnaie/en/index.html); Indonesia’s ESDM One Map

https://irena.masdar.ac.ae/gallery/#gallery).

Policy

153

© OECD/IEA, 2017

(http://geoportal.esdm.go.id/monaresia/home/); and the International Renewable Energy Agency’s Global Atlas (IRENA,

Data aggregation, as the name suggests, aims to merge and manage different data sources and use them to provide an enhanced understanding of the energy system.

One example is work in the United Kingdom to combine data for local areas about annual consumption of electricity and/or gas with information on building stocks (type

of buildings, floor area, age of buildings), energy audits, and socio-economic indicators. The resulting analysis can both be commercially useful and help understand the impact and targeting of energy efficiency policies, as shown by the UK

National Energy Efficiency Data framework (NEED).8 A second example are data hubs that are being established in the Nordic countries with the objective of enhancing

competition in electricity retail markets by bringing all data to one access point in order to make supplier switching and other market processes easier for consumers and other market players. Denmark’s electricity grid operator, EnergiNet.dk, is currently

securely storing anonymised highly granular near real-time data on their Energi Data

Service. This is then made available as aggregated open data on electricity consumption, production by source, trade, CO2 emissions, transmission line capacity

and prices on wholesale markets with the goal of supporting new product development and statistical analysis.9

An essential part of any data fusion is a rigorous validation process, which in turn generates high-quality data for further use. Datasets created through the integration of different sources of data need to be repeatable, and statistics of this sort require metadata to explain how they have been constructed and where there is uncertainty. Attention to the representativeness and completeness of data remains vital.

The opportunities provided by digitalization to improve energy statistics can only be realised if policy makers and industry enable access to these new sources of data.

Much work still needs to be done to develop the guidelines and mechanisms that will

allow this (Masanet et al., 2017). Recent policy papers from the European Commission provide a good overview of key issues and some possible ways forward (European Commission, 2013; 2017a; 2017b).

Data on energy efficiency standards for appliances Minimum energy performance standards for appliances, buildings and vehicles have proven to be among the most cost-effective measures used by governments around

the world to improve energy efficiency. Yet these standards are sometimes less effective than they might be because of information asymmetries between

8

www.gov.uk/government/collections/national-energy-efficiency-data-need-framework.

9

www.energidataservice.dk/.

154

Policy

© OECD/IEA, 2017

manufacturers and regulators, as well as a lack of effective compliance regimes.

Three low-cost digital technologies and techniques can significantly improve data collection and analysis about appliance energy performance, providing better information to consumers and greatly improving the policy-making process:

• Online registration systems are web-based facilities where the manufacturers

or importers of equipment and appliances register eligible products with a regulatory authority before they can be sold in a market. Typically, registries

require the submission of technical documentation to demonstrate product

compliance with programme requirements. This database also allows the regulator to track changes in the market over time. Overall, registration systems can facilitate programme development, reduce the costs and resources needed

to develop compliance mechanisms, provide a more streamlined and

transparent system for manufacturers to interface with, and provide information

that

can

help

consumers

make

informed

energy-efficient

purchasing decisions. A number of countries have successfully implemented

these online systems, including Australia, Brazil, Canada, the People’s Republic of

China

(hereafter,

“China”),

India,

Japan,

New Zealand, Viet Nam and the United States.

Saudi

Arabia,

Thailand,

• Web crawling (also known as web harvesting or web scraping) is a technique that uses search algorithms to automatically collect information from websites.

Crawling can be used to collect online data on sales, prices, efficiency levels and types of products. Easy and cheap access to this data can be helpful to policy makers as they develop standards or evaluate the impact of energy

labelling programmes. Swedish and US policy makers have achieved promising results when compared to traditional forms of market research. This is still a new technique, but web crawl data is becoming increasingly relevant as more

appliances and equipment are sold online. The sharing of international

knowledge and experience on web crawling, as the technology and methodologies develop further, would be a valuable exercise for policy makers worldwide.

• Quick response (QR) codes are a type of machine-readable optical label, resembling a two-dimensional barcode (Figure 7.2). The code contains

information about the item to which it is attached. Applications include product tracking, identification, document management, and general marketing. QR codes enable information to be accessed in a consistent format. Their use

improves the quality of registration systems. For example, imported appliances and equipment can be easily checked by customs officials by simply scanning

labels, so that non-compliant manufacturers are easily detected when cross-

Policy

155

© OECD/IEA, 2017

checked against the registration database.

Figure 7.2 Example of a QR code on an appliance energy efficiency label in China

Key message: QR codes contain an efficient way to access additional information about the labelled item.

Source: China National Institute of Standardization (2015). QR Codes for Appliances and Equipment – Enabling Access to

Information and Facilitating Compliance.

China is at the forefront of using new technologies for managing mandatory energy

performance standards and labelling programmes. In particular, China has led the way

in using QR codes to guide consumers to information that can support energy efficient purchasing decisions.

Observing that ordinary (non-digital) energy labels were having a limited impact on

consumers, who tend to rely on retailers to advise on purchase decisions, the China

National Institute of Standardization in 2016 introduced QR codes on 16 types of

appliances.10 These QR codes allow consumers to easily access relevant information via smart phone apps. The information available is updated daily based on the product registry, and includes comparisons of different appliance prices and running costs,

manuals, repair options and end-of-life recycling options. Additionally, the scannable codes are expected to enhance compliance checks without needing large numbers of

technical experts; local compliance authorities can see whether a product or manufacturer has performed badly or well in tests in other parts of the country.

QR codes are a good fit for China, where about 1.2 billion people have access to a

mobile phone, of which more than half are smartphones with some form of scanning

capability. In addition, QR codes are not patented, are easy to use and low cost. To

China National Institute of Standardization, 2017. Best Practice of QR Code Application in China Energy Label. Presentation by

Dr. Yujuan Xia. Resources and Environment Branch. CNIS.

156

Policy

© OECD/IEA, 2017

10

date, more than 4 000 appliance models have QR codes and in a nine-month period

there have been 85 million scans.11

There is a growing risk that different countries might opt to design their own different

QR code protocols. International collaboration to develop common data requirements

or a QR code protocol would enable wider use and, in the future, international adoption by regulators and manufacturers.

Policy frameworks and market design In many parts of the energy system, digitalization can facilitate positive change, but only if policy makers undertake efforts to understand, channel and harness

digitalization’s impacts and to minimise its risks. It should first be acknowledged that policy makers have different starting points – country by country and jurisdiction by jurisdiction – with respect to energy resources, digital and energy infrastructure, and market designs. Furthermore, certain countries have already adopted general policy goals seeking rapid digitalization of their economies and societies as a whole. The

Singapore Smart Nation Vision can be viewed as a useful case study of a holistic digitalization strategy which encompasses energy and transport (Box 7.3), although

this specific model may not be replicable or suitable in other countries. Other leaders in e-government include Estonia.

A key common theme emerging from this report is that many potential benefits of

digitalization need policy support and frameworks in order to be fully realised.

Transport is a key example. Smart, connected EVs and shared mobility could potentially transform cities, but a host of technical, legal, regulatory and other policyrelated hurdles first need to be overcome (see Chapter 3, and in particular, Figure 3.1).

Similarly, as Chapter 4 discusses, digitalization is contributing to a complex transition in the structure of electricity systems and in business models. Here also, governments

may want to consider how far standard-setting (see Box 6.3), market design and regulatory frameworks can help to shape the process of change toward preferred

outcomes. For example, digitalization of grids might require a revision of the rate-basing rules to facilitate ICT investment and purchase of software services on an equal basis with copper-wire investment. Digitalization of utility companies might be

supported using a range of measures, including regulatory pilot schemes, subsidies for

11

China National Institute of Standardization, Personal communication in 2017.

Policy

157

© OECD/IEA, 2017

demonstration projects and standard-setting.

Box 7.3

Singapore’s Smart Nation Vision

Launched in 2014, Singapore’s Smart Nation Vision aims to use digital technology to make Singapore “the world’s first smart nation”. While Singapore differs in many ways from other

governments – including the fact that Singapore’s status as both a city and a country allows its government to bring to bear a full range of both municipal and national policy levers – its integrated efforts offer an interesting case study and raise many policy issues other governments may face more of in the future. Singapore’s digitalization efforts focus on a range of infrastructure, regulation and innovation support: •

Singapore has invested heavily in a robust digital backbone, including a high-speed broadband

network to improve connectivity rates. Singapore has constructed a common platform, joining up

various sensors distributed across the island with a data-sharing gateway and video and data

analytics capabilities, allowing real-time data to be quickly collected and processed for analysis, incident response and policy making. •

To facilitate interoperability of digital devices, Singapore is working to develop standards for network system architecture, communication and security protocols for sensors and the IoT. In

the financial sector, Singapore has introduced a regulatory sandbox to allow fintech solutions and business models to be tested outside existing regulations for a limited duration. •

To build up human resource capabilities in data analytics and cybersecurity, the Singapore government is training 10 000 public servants in data science, while allowing military service conscripts to serve in cybersecurity functions, learning skills that they can carry over to their future careers.



Singapore has committed approximately USD 14 billion in funds through to 2020 to support private- and public-sector R&D, predominantly in developing technological and digital capabilities and sustainable urban solutions.



More public data and maps are being made available to the private sector. For example, realtime public transport data available online were used to create “Bus Uncle”, a Facebook

Messenger chatbot that tells the waiting time for the next bus in the distinct local creole Singlish.

Singapore’s integrated approach can be particularly seen in the multitude of efforts undertaken in energy and transport. Singapore is pioneering advanced urban congestion pricing with the

introduction of a global navigation satellite system, allowing more flexibility to shift charges and applicable zones to respond to changes in traffic flow, and better refinement in charging for exact distances travelled on congested roads. An initial pilot project co-ordinated by various government

agencies helped assess usage patterns of electric vehicles (EVs), different charging technologies

and their impact on the grid, and cybersecurity concerns. It culminated in new larger-scale pilots for EV car-sharing and taxis, and hybrid and electric buses, testing out viable business models. Various

158

Policy

© OECD/IEA, 2017

autonomous vehicle testbeds are being explored and deployed, ranging from on-demand cars and buses, to truck platoons, to multi-purpose vehicles for road cleaning and refuse collection.

Likewise, digitalization can technically achieve significant gains in demand response,

providing useful flexibility as well as energy savings. However, market reforms may be

needed in order to provide the incentive for this to happen. For example, solar PV-with-battery prosumers could be useful participants in demand response and grid

balancing at local level. But if charging and discharging batteries into the grid entails

large tariffs or is taxed at each transaction, then prosumers will not to want to join in

demand response, and will prefer to keep the energy that they generate and store for use behind-the-meter in their own buildings.

Meanwhile, regulators will need to be alert to the impact of rapidly increasing volumes

of automated transactions as these are carried out within grids, putting pressure on physical system controls (Vasconcelos, 2017). Decisions about these issues will also have implications for the emergence of new business models in the energy sector.

Digitalization has also already resulted in large upheavals in many sectors and markets

(see Chapters 1 and 6). In energy, some uses of digital technologies are long-standing

– for example, mainframe computing in oil and gas and electric utilities’ operations. Other uses of ICT deployed at scale more recently have uncertain impacts.

Consequently, flexibility in policy frameworks will be needed to accommodate hard-topredict future developments. With the enormous speed of change in business models

and the way services are provided, governments with overly rigid frameworks may risk

being slow to react to, and to take advantage of, the opportunities that digitalization offers.

Chapter 6 has also highlighted how digitalization carries a number of potential risks as

well, in particular digital security, privacy, and changes in jobs and skills. Further

challenges will certainly emerge. Balancing an open approach – open data, openness

to new market entrants – with the management of these issues will require both foresight and collaboration with experts from beyond the energy sector.

No-regrets policy recommendations Based on the analysis presented throughout this report, a preliminary set of no-regrets

recommendations has been drawn up to help policy makers navigate the opportunities

and challenges brought about by greater digitalization in the energy sector. This list is not intended to be exhaustive or definitive, and recognises that national circumstances

and contexts vary between countries.12 It is hoped it will foster further discussion

12

For instance, in energy resources and infrastructure; digital readiness, and economic development.

Policy

159

© OECD/IEA, 2017

among governments, companies and other stakeholders.

Governments should:

• Build expertise. Energy policy makers need to make sure they are well-informed about the latest developments in the digital world, its nomenclature, trends, and ability to impact a variety of energy systems (both in the near and longer term). A major part of this endeavour consists of ensuring that energy policy makers have access to staff with digital expertise. Education policies and technical training to ensure an adequate pool of relevant expertise for both the private and public sectors will also be critical. Conferences, workshops and exercises can also help.

• Ensure appropriate access to data. Opportunities provided by digitalization to improve energy statistics can only be realised with access to data. For example, these could include: electricity consumption data at a high level of detail in both space and time; information on installed distributed energy resources; and data about energy infrastructure. Ensuring timely and robust, verifiable and secure access to the necessary data, from business and across government, while protecting privacy, is critical. Policy makers should consider how guidelines and mechanisms can enable sharing of data.

• Build flexibility into policies. While energy infrastructure can be expected to last 50 years or more in many instances, software, applications, and even ICT hardware turns over quickly. As policy makers design a range of energy policies, they should ensure appropriate flexibility to deal with new developments in digital and communication technologies, while these continue to rapidly evolve, often in hard-to-predict ways.

• Experiment. As explored throughout this report, there is no way to predict with certainty how particular digital technologies will interact with specific energy system applications, especially in complex real-world situations that involve multiple policy

objectives and

uncertain (and

sometimes unintended)

feedbacks. Accordingly, governments should consider setting up and exploring a wide variety of real-world experiments that can yield “learning by doing”. California’s programme of pilot projects in electricity demand response and smart grids is a good example. Governments may also consider setting up equivalent digital “sandboxes”13 along the lines of fintech test zones developed

13

A sandbox is a type of “safe house” where pilots of specified types of innovative products, services and market arrangements

otherwise prevent the pilot taking place.

160

Policy

© OECD/IEA, 2017

can take place, with the support of sector regulators. In the sandbox, there may be relief from regulatory requirements which would

in Australia, Indonesia and Singapore. Such sandboxes, for example, could be set up to enable testing of peer-to-peer transactive energy markets or autonomous vehicle experimental zones.

• Participate in broad inter-agency discussions. Many jurisdictions around the world are developing digital strategies for their whole economies. For example, since May 2015, the European Commission has delivered 35 legislative proposals and policy initiatives in its Digital Single Market strategy. Energy policy makers should be active in these inter-agency discussions to ensure energy sector perspectives and equities are taken into account.

• Focus on overall system benefits. In line with broader IEA recommendations, the costs and benefits of digitalization in energy should be considered not only per component or per individual consumer, but also in terms of overall net benefits to the security, sustainability and affordability of the system as a whole. This approach is particularly important in electricity where the transition to smart energy systems may require significant changes in market design.

• Monitor the energy impacts of digitalization on overall energy demand. Policy makers should be aware of the possibility that new digital devices and services have the potential to increase energy consumption, for example, as a result of growing quantities of smart household and consumer electronics. Understanding consumer behaviour and being always up-to-date in monitoring the energy efficiency of new energy-using devices will be increasingly important.

• Incorporate security by design. As an efficient way to reduce overall digital security risks, policy makers should include security considerations in all publicly supported technology research and design programmes, and in product manufacturing through standard-setting.

• Provide a level playing field. Governments should strive to provide technology-neutral and delivery-route-neutral policies and platforms for digital energy (for example in relation to the role of smart meters or other energy management systems), to allow a variety of companies to compete to find new business models and to serve consumers better. Considerations of security, privacy, economic disruption and other concerns will also need to be taken into account.

• Learn from others. It is acknowledged that each country is different in many nonetheless, there are lessons to be learned from the experiences of other Policy

161

© OECD/IEA, 2017

ways that are relevant to digitalization’s increasing impact on energy systems;

governments and jurisdictions. These lessons can include both positive case studies as well as more cautionary tales. Useful collaborations and best policy sharing can take place in a variety of fora, including the Connected Devices

14

See www.iea.org/tcp/.

162

Policy

© OECD/IEA, 2017

Alliance and a wide range of IEA Technology Collaboration Programmes.14

References ABC News (2017). “Climate change in drones' sights with ambitious plan to remotely plant nearly

100,000 trees a day”, ABC News, 25 June 2017. www.abc.net.au/news/2017-06-25/the-plan-to-

plant-nearly-100,000-trees-a-day-with-drones/8642766 (accessed on 9 October 2017). Bohi, D. R. 1999. Technological Improvement in Petroleum Exploration and Development. In

Productivity in Natural Resource Industries: Improvement through Innovation. Simpson [Ed.].

RFF Press (June 2, 1999). ISBN 978-0915707997.

China National Institute of Standardization, 2017. Best Practice of QR Code Application in China

Energy Label. Presentation by Dr. Yujuan Xia. Resources and Environment Branch. CNIS.

China National Institute of Standardization (2015). QR Codes for Appliances and Equipment –

Enabling Access to Information and Facilitating Compliance. Lin Haoxin. CNIS. 21st Century Energy Efficiency Standards and Labelling Programmes Workshop. December 2015. www.iea.org/media/workshops/2015/productsdec15-

16/2.4_LinHaoxin_QRCodesforAppliancesandEquipmentEnablingAceesstoInformationandFacilita tingComplianceLinHaoxinCNIS.pdf

European Commission (2017a). Communication on Building a European Data Economy.

https://ec.europa.eu/digital-single-market/en/news/communication-building-european-dataeconomy.

European Commission (2017b). Staff Working Paper on Building a European Data Economy.

https://ec.europa.eu/digital-single-market/en/news/staff-working-document-free-flow-dataand-emerging-issues-european-data-economy.

European Commission (2013). Scheveningen Memorandum on Big Data and Official Statistics.

http://ec.europa.eu/eurostat/documents/42577/43315/Scheveningen-memorandum-27-09-13.

ESI Africa (2017). “BBOXX: solar innovation to light up Togo”, ESI Africa, www.esiafrica.com/news/bboxx-solar-innovation-to-light-up-togo/, accessed 13 July 2017. GSMA (2016a). Mobile for Development Utilities: Unlocking Access to Utility Services: The

Transformation Value of Mobile. www.gsma.com/mobilefordevelopment/wp-

content/uploads/2016/07/Mobile-for-Development-Utilities-Annual-Report.pdf GSMA (2016b). The Mobile Economy: Africa 2016.

www.gsmaintelligence.com/research/2016/07/the-mobile-economy-africa-2016/569/.

GSMA (2017a). Global Mobile Money Dataset.

www.gsma.com/mobilefordevelopment/programme/mobile-money/global-mobile-moneydataset.

GSMA (2017b). Mobile for Development Utilities: Lessons from the use of mobile in utility pay-as-

you-go-models. https://www.gsma.com/mobilefordevelopment/wpmodels.pdf.

Policy

163

© OECD/IEA, 2017

content/uploads/2017/01/Lessons-from-the-use-of-mobile-in-utility-pay-as-you-go-

International Academy of Astronautics (2015). Summit Declaration, Friday 18 September 2015,

Mexico City, Mexico. http://iaaweb.org/iaa/Scientific%20Activity/declarationmexico.pdf.

IEA (International Energy Agency) (2017). Energy Access Outlook 2017: From Poverty to Prosperity.

World Energy Outlook Special Report.

Kornei, K. (2017). “Satellite quantifies carbon dioxide from coal-fired power plants”. Eos, 98, 9 October 2017. https://doi.org/10.1029/2017EO084189. Lyon Bowley, Sir Arthur (1915). An Elementary Manual of Statistics. P.S. King & Son. Managi S., Opaluch J.J., Jin D. and Grigalunas T.A., 2004. “Technological change and depletion in

offshore oil and gas”. Journal of Environmental Economics and Management, 47, pp. 388-409.

Managi S., Opaluch J.J., Jin D. and Grigalunas T.A., 2005. “Technological change and petroleum exploration in the Gulf of Mexico”. Energy Policy, 33, pp. 619-632.

Masanet, E., et al. (2017). “Leveraging smart system technologies in national energy data systems:

Challenges and opportunities”. Energy and Economic Growth State-of-Knowledge Paper Series;

working paper.

Sanyal, S. et al. (2016). Stimulating Pay-As-You-Go Energy Access in Kenya and Tanzania: The

Role of Development Finance. World Resources Institute.

www.gogla.org/sites/default/files/recource_docs/stimulating_pay-as-you-

go_energy_access_in_kenya_and_tanzania_the_role_of_development_finance.pdf.

USAID (United States Agency for International Development) (2017). Demographic and Health

Surveys Programme; STATcompiler (database), www.statcompiler.com/en/.

Vasconcelos, J. (2017). Chapter [check] in López-Ibor Mayor, Vincente (ed.), Clean Energy Law and

Regulation: Climate Change, Energy Union and International Governance. Wildy, Simmonds and

Hill Publishing.

Wood Mackenzie (2014). Presentation by William Durbin to the Asia Petrochemical Industry

Conference (APIC). 15-16 May, 2014, Thailand. www.woodmac.com/reports/chemical-marketsapic-2014-the-changing-energy-landscape-and-its-impact-on-feedstock-dynamics-forchemicals-21947039/.

WEF (World Economic Forum) (2017). “These drones can plant 100,000 trees a day”, blog 29 June 2017. www.weforum.org/agenda/2017/06/drones-plant-100000-trees-a-day (accessed 9

164

Policy

© OECD/IEA, 2017

October 2017).

Glossary 3D printing

3D (three dimensional) printing, or additive manufacturing, is a

computer-controlled technology that builds objects by depositing consecutive layers of material.

4G

The fourth generation of mobile telecommunications, succeeding 3G. A 4G system provides faster broadband internet access to mobile

devices, with peak speed requirements defined by the International Telecommunications Union (ITU) of 100 Mbps for high-mobility

communication (such as from trains and cars) and 1 Gbps for

low-mobility communication (such as pedestrians and stationary users). 5G

The proposed fifth generation of mobile telecommunications,

succeeding the current 4G standard. 5G is expected to feature faster

speeds, increased reliability and improved security, underpinning key technologies such as connected cars and the IoT. 5G is expected to

be deployed from 2020. Additive

See 3D printing.

ACES

Automated, connected, electric and shared (referring to vehicles and

Active controls

Automated devices or hardware systems which collect, process and

manufacturing

mobility).

adapt to real-time data using sensors, for example in a building, and

which can be automated or managed from a single front-end

dashboard (e.g. a smart phone app or building energy management system). Active controls can also integrate and intelligently link

building energy services with information from the grid, allowing for

Glossary

165

© OECD/IEA, 2017

better management across supply and demand.

Aggregator

Aggregators, also known as demand response providers, gather consumer demand of any type, as well as energy supply from

distributed producers such as renewables-based power plants, in

order to provide balancing services to the grid by adjusting power

demand and/or shifting loads at short notice. The aggregated load is managed as a single flexible consumption unit and sold to the

markets or to the grid operator. In this way, the aggregator provides an interface between large numbers of individual

consumers/producers and power markets/grid operators. Analytics

The use of data to produce useful information and insights.

Application

An API is a list of commands that allows software programs to

program

interface (API) App-based ride

hailing

communicate with each other and use each other’s functions.

Prearranged transport services that use an online-enabled

application or platform (such as smart phone apps) to connect drivers using their personal vehicles with passengers. Companies that provide these services are known as transportation network

companies (TNCs), such as Uber, Lyft, Didi Chuxing and Grab. Artificial

intelligence (AI)

The simulation of human intelligence processes by machines,

especially computer systems. These processes include learning,

reasoning and self-correction. One subfield or application of AI is machine learning. (See also learning algorithm.)

Automation

The use of various control systems to allow equipment, a process, or

a system to operate automatically, with minimal or no need for human input.

Autonomous vehicle

Often referred to as “driverless” or “self-driving”, an autonomous

vehicle is capable of sensing its environment and navigating safely

and effectively with no or minimal human intervention. In this report, “autonomous vehicles” refers to vehicles with high levels of

automation, specifically levels 4 and 5 in the Society for Automotive Engineers International Standard J3016. Autonomous

Where digital meters and sensors collect data on system

industrial control

algorithms and automatically implemented by digital systems.

166

performance, with optimisation actions determined by control

Glossary

© OECD/IEA, 2017

closed-loop

Bandwidth

The volume of data that can be sent through a network connection, typically measured in bits per second (bps). Greater bandwidth supports faster transfer of data.

Behind the

On the customer’s side of the utility meter.

Big data

Large amounts of data gathered from a range of diverse sources,

Blockchain

Also called distributed ledger technology. A decentralised data

meter

often in near real time.

structure in which a digital record of events (such as a transaction) is collected and linked by cryptography into a time-stamped “block” together with other events.

Botnet

Automated programs that run over the internet, short for

(ro)botnet(work). Some botnets run automatically, others only

execute commands when they receive specific input. Often used to launch spam email campaigns, denial-of-service attacks or online fraud schemes. Not all botnets are malicious.

Broadband

A term applied to high-speed telecommunications systems (i.e. those capable of simultaneously supporting multiple information formats such as voice, high-speed data services and video services on

demand). Central Scenario

The IEA Central Scenario describes the pathway for energy markets and technological progress based on the continuation of existing

energy and climate policies and measures, and to a certain extent

announced commitments and plans. It is broadly in line with the New Policies Scenario of the IEA World Energy Outlook and with the

Reference Technology Scenario in Energy Technology Perspectives

2017. The Central Scenario should not be interpreted as a forecast. Cloud computing

The practice of using a network of remote servers hosted on the

internet to store, manage and process data, rather than a local server

Glossary

167

© OECD/IEA, 2017

or a personal computer.

Connected device

Connected devices – also referred to as networked, edge or end

devices – are consumer electronics, appliances and other devices

that can be connected to networks and interact with the network or

other devices. Examples include smartphones and tablets, as well as other consumer electronics, household appliances, machines, and objects (e.g. televisions, washing machines, security cameras,

industrial equipment, cars, clothing). The primary function of some

connected devices is data storage or use (“electronic edge devices”, e.g. smartphone; smart TV), while the primary function for others is not data‑related (“other edge devices”, e.g. home appliances, lighting). (See also the Internet of Things.) Connectivity

Exchange of data between humans, devices and machines through

Computer

Organisation devoted to ensuring that appropriate technology and

response team

networked systems and to limiting damage; it also ensures continuity

Curtailment

The practice of temporarily decreasing electricity supply due to grid

emergency (CERT)

digital communications networks.

systems management practices are used to resist cyber-attacks on

of critical services following successful attacks, accidents or failures.

constraints or other causes, for example decreased electricity supply from a variable renewable energy (VRE) generator at times when

total supply from those sources exceeds demand. Critical

infrastructure

An asset, system or part of a system which is essential for the

maintenance of vital societal functions, such as the health, safety,

security, economic or social well-being of people in a country, and the

disruption or destruction of which would have a significant impact as a result of the failure to maintain those functions. Cryptocurrency

A digital asset designed to function as a medium of exchange, e.g. Bitcoin. Cryptography is used to secure the transactions and to control the creation of additional units of the currency.

Cyberspace

A global domain within the information environment consisting of the interdependent network of information system infrastructures

including the internet, telecommunications networks, computer

168

Glossary

© OECD/IEA, 2017

systems and embedded processors and controllers.

Cybersecurity

The ability to protect or defend the use of cyberspace from

cyber-attacks and cyber incidents, preserving the availability and

integrity of networks and infrastructure and the confidentiality of the information these contain. Commonly also refers to the safeguards and actions available to do this. Cyber incident

Actions taken through the use of computer networks that result in an actual or potentially adverse effect on an information system and/or the information residing therein. A violation or imminent threat of

violation of computer security policies, acceptable use policies, or standard security practices. Cyber-attack

An attack, via cyberspace, for the purpose of disrupting, disabling, destroying or maliciously controlling a computing

environment/infrastructure, or stealing controlled information. Cyber-physical systems

Machines that sense their environment, collect, analyse and act on data, and collaborate with each other and with humans. System

featuring a tight combination of, and co-ordination between, the

system’s computational and physical elements. Today, a pre-cursor

generation of cyber-physical systems can be found in areas as

diverse as aerospace, automotive, chemical processes, civil

infrastructure, energy, healthcare, manufacturing, transport,

entertainment and consumer appliances. This generation is often referred to as embedded systems. In embedded systems, the

emphasis tends to be more on the computational elements, and less

on an intense link between the computational and physical elements. Data and text

Automated research in the digital environment for the purpose of

Data protection

Data protection law or rules control how your personal information is

mining

and data privacy

discovering and extracting knowledge from unstructured data.

used by organisations which collect, process or disseminate data about which there may be an expectation of privacy.

Data

A subset of information in an electronic format that allows it to be

Data centre

Facilities designed to house information technology (IT) equipment.

Data traffic

The amount and characteristics of data being transmitted on a

retrieved or transmitted.

network, e.g. quantity, timing, packet size, content. Also called

Glossary

169

© OECD/IEA, 2017

network traffic.

Demand

response

Demand response or demand-side response refers to the possibility

for consumers to adjust their electricity consumption during periods

of peak demand, when power supply is scarce or electricity networks are congested, in response to time-based financial incentives.

Demand response can consist of interrupting demand for a short

duration, or adjusting the intensity of demand for a certain amount of time by reducing or shifting loads, or storing energy. For connected

devices, demand response functionality might enable a power utility or aggregator to remotely turn off air‑conditioning units in customer

homes to avoid peak load issues. Digital

Signal transmission that conveys information through a series of

Digitization

The conversion of analogue to digital.

Digitalization

Digitalization describes the increasing application of digital

coded pulses representing 0s and 1s (binary code).

technologies (i.e. ICT) across the economy, including energy, to

achieve desired outcomes such as improved safety, efficiency and productivity. The trend toward greater digitalization is enabled by

advances in data, analytics and connectivity: increasing volumes of

data thanks to the declining costs of sensors and data storage, rapid progress in advanced analytics and computing capabilities, and greater connectivity with faster and cheaper data transmission. Digital

See ICT.

Digital twins

Digital replicas of physical assets that can be used to simulate and

Distributed

Small-scale energy resources, such as solar PV, wind power or

technologies

energy

resources (DER)

optimise industrial design and oil and gas drilling.

batteries, generating or storing power at or near the point of

consumption. DER are usually connected to the distribution level

electricity grid. Also sometimes referred to as decentralised energy resources.

Drone

Also known as an unmanned aerial vehicle (UAV), a drone can either be piloted remotely by a human or else can be a fully autonomous vehicle.

170

See connected devices.

Glossary

© OECD/IEA, 2017

Edge device

eGovernment

Use of ICT tools and systems to provide better public services to

Energy

Something is more energy efficient if it delivers more services for the

efficiency

citizens and businesses.

same energy input, or the same services for less energy input. For

example, when a light-emitting diode (LED) uses less energy than a

compact florescent lamp (CFL) or incandescent bulb to produce the same amount of light, the LED is considered to be more energy efficient. Energy intensity

A comparative measure of total energy consumption relative to a specific indicator, such as primary energy use per unit of gross

domestic product or final energy consumption per square metre. Energy

management system

A computer-based system that defines the practices, methods and

structure for industrial or commercial firms, for example, to monitor

their energy consumption and establish practices to improve

efficiency within a defined perimeter, such as an industrial site or a building. A building energy management system, using computer

software and equipment hardware (e.g. sensors and active controls), can monitor and control heating, cooling, ventilation and lighting

systems, along with other buildings services such as fire and security systems. Energy poverty

A lack of access to modern energy services. These services are

defined as household access to electricity and clean cooking facilities (e.g. fuels and stoves that do not cause air pollution in houses).

Energy saving

The reduction or avoidance of energy use with respect to demand for energy services. Energy savings may result from energy efficiency

measures (e.g. insulation of a building to reduce heating and cooling loads), improved energy management (e.g. use of sensors that turn off equipment when it is not being used) or user decisions, such as the choice to ride a bicycle rather than use a car.

Glossary

The uninterrupted availability of energy sources at an affordable price.

171

© OECD/IEA, 2017

Energy security

Electric vehicle (EV)

A vehicle whose powertrain includes both a battery (which can be recharged via an external power source) and an electric motor.

Electric vehicles include battery electric vehicles (BEVs), plug-in

hybrid electric vehicles (PHEVs) and fuel cell electric vehicles

(FCEVs). BEVs have no internal combustion engine. PHEVs combine a battery and an internal combustion engine. Charging of BEVs and PHEVs can be achieved by plugs, or by (stationary or dynamic) conductive or inductive power transfer. FCEVs have hybrid

powertrains that use fuel cells, optionally together with a battery and supercapacitor, and an electric motor. Firmware

Permanent software programmed into a digital device’s read-only

memory.

Flexibility

The capability of the electricity system to respond to upward or

systems)

manner over a timescale ranging from a few minutes to several hours.

(in electricity

downward changes in the supply/demand balance in a cost-effective

Flexibility is often associated with the ramping capability of

dispatchable power plants in the system, but it also refers to other resources including storage, demand-side management and grid infrastructure. GW

Gigawatt (109 Watts). 1 Watt = 1 joule per second.

Hardware

The physical components of an information system. (See also

Home or

Refers to systems that integrate diverse electrical devices and

automation

on site or remotely, e.g. through internet access or automatically in

building

software and firmware.)

energy-consuming equipment in a house or building, allowing control

accordance with selected settings. Such automation has

convenience, energy efficiency and safety benefits. (See also energy management system.) Hyperscale

Hyperscale data centres are very efficient, large-scale public cloud

(data centre)

data centres operated by companies such as Alibaba, Amazon and

ICT

Information and communications technology. Used in this publication

Google.

in the broadest sense (i.e. digital technologies) to include all types of digital equipment such as sensors, connected devices, network

172

Glossary

© OECD/IEA, 2017

equipment and infrastructure (e.g. data centres and network cables).

Internet of

Things (IoT)

Where everyday objects are connected to networks to provide a range of services or applications in areas such as cars, home automation and smart grids. The IoT encompasses both machine-to-machine

(M2M) connection, where devices interact and share data without the direct involvement of people, and things connected to networks to

enable people to remotely control processes or manage their devices. Intelligent

Intelligent transport systems include the use of sensors,

systems (ITS)

system operations, with the aim of improving safety, efficiency and

transport

communications technologies and advanced analytics to improve

service, as well as lowering costs. Examples of ITS include in-road

detectors to control traffic lights, RFID to automatically collect tolls, and the use of the GPS and telecommunications for roadside assistance.

Internet

The internet is the single, interconnected, worldwide system of

commercial, governmental, educational and other computer networks

that share (a) the protocol suite specified by the Internet Architecture Board (IAB), and (b) the name and address spaces managed by the Internet Corporation for Assigned Names and Numbers (ICANN). Internet Protocol (IP) traffic

IP traffic includes fixed and mobile internet traffic (IP traffic that crosses an internet backbone), corporate IP wide area network

(WAN) traffic, and IP transport of TV and video on demand (VoD).

Interoperability

The ability of different computer systems or software to exchange

Last mile

In transport, the “last mile” refers to the stage of a trip or (in supply

and make use of information.

chains) delivery that is typically the most difficult to address in terms of perceived inconvenience or cost. In the case of goods delivery, getting merchandise from a final delivery hub to the customer is generally the most expensive and least efficient leg of product

shipping. In the case of passenger services, the term refers to the

perceived inconvenience of going from a public transit station to the final destination (often home or work). In the case of passenger

transport, the term has an analogue: the “first mile”, which refers to

the leg of a trip from home to the nearest public transport hub (and the difficulties in making this leg of a trip comfortable, convenient

Glossary

173

© OECD/IEA, 2017

and cheap).

Learning

algorithm

A process or method used to extract patterns from data collection (e.g. from building sensors and controls) to identify and adapt appropriate solutions or applications for a device or system.

Examples include smart thermostats in buildings, that collect and

process data on occupant presence, routine and preferences to adapt building energy services, such as lighting, heating, cooling and ventilation, in order to reduce total energy consumption while

maintaining or improving the energy service and user comfort. (See

also smart controls.) Load

Low power wide

The level of network interaction of a network‑enabled device, which influences its power consumption.

LPWA networks provide low power draw and wide area coverage, and

area (LPWA)

are designed for IoT and M2M applications that provide low data

Machine

Machine learning, a subfield of artificial intelligence, is the science of

networks

learning

rates, long battery lives, and at low cost.

getting computers to act without being explicitly programmed.

Machines are given access to large data sets and allowed to learn for themselves. (See also artificial intelligence.)

Machine-to-

machine (M2M)

Malware

M2M connections include energy sector applications such as smart meters and process sensors in power plants, GPS for logistics and vehicles, smart metering, and other IoT technologies.

A software (e.g. a virus, worm, rootkit, botnet or other code-based

malicious entity) specifically designed to disrupt, damage, or gain unauthorised access to an ICT system.

Microgrid

Small electric grid systems linking a number of households or other

Mobility as a

MaaS identifies mobility solutions that are consumed as a service.

Service (MaaS)

consumers. Also sometimes called a mini-grid.

This is enabled by platforms that integrate public and private,

motorised and non-motorised mobility options and offer a unified

trip-making and payment platform. MaaS platforms allow users to subscribe to an all-inclusive, multi-modal “mobility package” and

access a variety of shared mobility services, including bicycles, buses,

trains, cars, taxis and ride-hailing services. MaaS typically comes with

174

Glossary

© OECD/IEA, 2017

a shift away from the conventional vehicle ownership paradigm.

Network (digital)

A digital structure that allows the transmission of data or information between two or more connected devices; networks can interconnect with other networks and contain subnetworks. Different types of networks include local-area networks (LANs) and wide-area networks (WANs), e.g. the internet.

Network

(electricity or

Electrical grid or gas pipeline.

gas)

Networked

See connected device.

Node

An electronic device that can send, receive or forward information;

device

nodes can be network devices (such as a modem) or edge devices (such as a digital telephone handset, a printer or a computer).

Open data

Free and widely available data for consultation and reuse, including reuse for commercial purposes, with a view to increasing transparency and stimulating economic activity.

Operation and

Refers to monitoring and repair of equipment, utilities or other

(O&M)

performance, functionality and asset value. O&M may include

maintenance

property (e.g. buildings and vehicles) to ensure operational

scheduled or unscheduled service checks and repairs, where

increasingly connected devices and sensors are helping to monitor and predict O&M needs. (See also predictive maintenance.) Optical network

Optical networks transmit information in the form of light pulses through thin glass or plastic optical fibre, offering much higher

transmission capacity than conventional copper-wire networks. Patch

An update to an operating system, application or other software,

Peer-to-peer

In energy, connecting system users and market participants with each

Glossary

other to enable direct trading.

175

© OECD/IEA, 2017

(P2P)

issued specifically to correct particular problems with the software.

Phishing / whaling

Trying to obtain sensitive information such as usernames and

passwords by means of an email (or other electronic communication) which is disguised as a trustworthy communication, but which when opened or when a link is clicked on allows the sender a point of access.

Photovoltaic (PV)

Physical internet

One of the main technologies for producing solar power. PV cells

directly convert solar energy into electricity; this is a semiconductor device.

An open, modular and shared global logistics system, inspired by the movement of data on the internet, in contrast to the proprietary logistics systems that are common today. Currently, nearly all

logistics service providers and carriers maintain proprietary assets,

both physical (e.g. warehouses and trucks) and operational (e.g. information on routes, customers and markets). Platform

A layer of software that combines different kinds of devices, data and services, on top of which other firms can build their own offerings,

e.g. search engines, social media, e-commerce platforms, app stores, price comparison websites. Platooning

Platooning refers to the practice of driving vehicles (primarily heavyduty tractor-trailers or rigid trucks) in a single line with small gaps

between them to reduce drag and thereby save fuel during highway operations. Vehicle-to-vehicle (V2V) and vehicle-to-infrastructure

(V2I) communication technologies can enable trucks to drive in very close proximity without sacrificing safety or manoeuvrability. Plug-and-play

Software or devices that are intended to work perfectly when first

used or connected, without need for installation, reconfiguration or adjustment by the user.

Power

In the context of energy, refers to electricity (e.g. power plant).

Power purchase

A financial agreement/contract between two parties: one that

(PPA)

electricity.

Power usage

A measure of how efficiently a data centre uses energy; the best

(PUE)

used for cooling/power provision for every 1 kWh used for IT

effectiveness

176

generates electricity and one that is looking to purchase the

global data centres have PUE values of around 1.1 (meaning 0.1 kWh equipment).

Glossary

© OECD/IEA, 2017

agreement

Predictive

maintenance

Predictive maintenance refers to the ability to perform targeted

maintenance in order to address a potential issue before it happens

or worsens, causing damage or downtime. Predictive maintenance is

based on direct measurement of the real-time status of components. Prosumers

Pro(ducer-con)sumers. Usually refers to small-scale, distributed

generation which allows consumers to increasingly have the choice to buy electricity from a retailer or to produce at least part of it themselves.

Radio-frequency

RFID uses electromagnetic fields to automatically identify and track

identification

tags attached to objects. The tags contain electronically stored

Ransomware

A type of malware which encrypts user data, asking victims to pay a

Real time

Information is available simultaneous with an event, or immediately

(RFID)

information.

ransom in order to obtain a decryption key.

after collection. There is no delay in the timeliness of available information.

Rebound effect

The reduction in expected gains from the introduction of new

technologies or policies that increase the efficiency of resource use,

because of behavioural or other systemic responses, for example the increased consumption of energy following improvement in the efficiency of an energy-consuming product or service. Ride-sharing

Ride-sharing refers to multiple passengers sharing a single vehicle to take a trip. These passengers may be making trips with different

origins and destinations, or they may be starting and arriving at the same places, but use the same vehicle to economise on costs,

resources or energy consumption. More recent innovations include “dynamic” and “app-based” ride-sharing. Right-sizing

Improved resource and energy efficiency of service provision can be

achieved by optimising the size of equipment, for example vehicles, to match utilisation patterns, for instance to match the number of passengers.

Sandbox

A type of “safe house” where pilots of specified types of innovative

products, services and market arrangements can take place, with the

Glossary

177

© OECD/IEA, 2017

support of sector regulators.

Self-

consumption

When an energy producer, e.g. of solar power from PV, consumes or stores this directly, behind the meter, rather than feeding it into the grid.

Sensor

A device which detects or measures some type of input from the

Server

A server is a computer that provides functionality and services to

Shared mobility

A transport strategy that enables users to gain short-term access to

physical environment (e.g. daylight, temperature, motion or pressure).

other devices through network connections.

vehicles, bicycles and other transport modes on an as-needed basis.

Shared mobility services include car-sharing services, app-based

shared ride-hailing services like LyftLine and UberPool, long-distance

ride-sharing platforms like BlaBlaCar, and bike-sharing services. Simulation

The use of computing to model the dynamic responses of a system to

Smart charging

A charging strategy for electric vehicles that uses connectivity and

exogenous events.

other digital technologies to automatically shift battery charging to

times when electricity prices are low and/or when overall electricity demand is low. Smart cities

A city becomes smart by virtue of strategically leveraging ICT

infrastructures and applications towards better delivery of benefits to the citizens. These might include making a city more sustainable and greener through less energy consumption and more of it from renewable sources, or improving the efficiency of transport.

Smart controls

ICT-enabled devices, such as building systems controls, that can take

(including smart

decisions in an adaptive or predictive manner based on sensors and

thermostats)

smart LED lighting is connected to and can interact with building

lighting, smart

data and possibly with the use of learning algorithms. For instance, control systems to predict and control lighting services relative to occupant presence, routine and daylight conditions.

appliance

178

Network-enabled devices or appliances (e.g. washing machines and

televisions) with integrated information and communication functions that allow them to transmit and receive information.

Glossary

© OECD/IEA, 2017

Smart device /

Smart grid

An integrated electricity system that uses advanced software

applications and communication network infrastructure, together with sophisticated sensing and monitoring technologies, to optimise the

management of energy supply, demand and transmission. Smart grids use ICT to gather and act on data, so as to co-ordinate the needs and

capabilities of all generators, grid operators, end users and electricity market stakeholders to operate all parts of the system as efficiently as possible, minimising costs and environmental impacts while maximising system reliability, resilience and stability. Smart meter

A meter that records electricity consumption in intervals of an hour or less, and communicates that information at least daily back to the utility for monitoring and billing purposes. This type of advanced metering infrastructure differs from traditional automatic meter

reading in that it enables two‑way communication between the meter and the central system. Smart meter functionality includes remote reading, two-way communication, support for advanced tariff and payment systems, and remote disablement and enablement of supply. Software

The programmes and other operating information used by a

Solar PV

See photovoltaics.

Standardisation

The process of implementing and developing technical standards

computer.

based on the consensus of different parties that include firms, users, interest groups, standards organisations and governments.

Standardisation can help to maximise compatibility, interoperability, safety and repeatability. Standby power

The power consumed by an appliance or device when it is not actively

Subcritical/

Refers to the efficiency of fuel conversion by coal power plants.

supercritical

in use but is ready to be rapidly put into use.

Supercritical (SC) refers to those thermal plants in which steam is raised above the critical point of water (221 bar and 374oC). Ultra-

supercritical (USC) usually refers to those plants in which steam is

Glossary

179

© OECD/IEA, 2017

raised at temperatures higher than 593oC (1 100oF).

System operator

The organisation responsible for operating part or all of the power

system. Originally at high voltage only, active operation of low-voltage grids is emerging in order to manage a growing amount of distributed (mainly solar) power plants. System operation is ideally separate(d) from ownership of transmission and generation assets

Recovery

Remaining recoverable resources of oil and gas are comprised of

resources)

in known fields) and as yet undiscovered resources that are judged

(recoverable

proven reserves, reserves growth (the projected increase in reserves likely to be ultimately producible using current technology. Resources can be defined either as technically recoverable (i.e. producible with current technology) or as technically and economically recoverable, meaning that they are exploitable at current oil prices.

Robotics

The science of design, engineering and use of increasingly intelligent machines that sense, act purposefully and perform work

autonomously, and their control and information processing systems. Test bed

In engineering, an area equipped with instruments used for testing

Total final

Is the sum of consumption by the different end-use sectors. TFC is

consumption (TFC)

machinery, engines, etc. under working conditions.

broken down into energy demand in the following sectors: industry

(including manufacturing and mining), transport, buildings (including residential and services) and other (including agriculture and non-

energy use). It excludes international marine and aviation bunkers, except at world level where it is included in the transport sector.

Transactive energy

The use of a combination of economic and control techniques to

improve grid reliability and efficiency, in particular in a world with

increasing numbers of independent prosumers. Often discussed in

the context of market-based transactive exchanges between small prosumers.

Upstream oil and gas

The upstream sector includes production of oil and gas from all

onshore and offshore oil and gas facilities (from either conventional or unconventional reservoirs) and gathering and processing of the

180

Glossary

© OECD/IEA, 2017

produced hydrocarbons.

Unconventional reservoirs

Unconventional reservoirs include a variety of oil and gas resources

that cannot easily be accessed using standard drilling technologies, where the hydrocarbons are trapped by the nature of the rock itself

rather than by the geometrical arrangement of the rock layers (as in conventional structural or stratigraphic traps). Unconventional

resources include extra-heavy oil and bitumen (oil sands), shale gas, and coal-bed methane. Value chain

The complete range of activities comprising the development of

network‑enabled devices, which includes device and component manufacturing, software development, network design, network architecture, communication protocol development, technical standardisation processes and service providers. The

network‑enabled device value chain spans the computing and

electronic sector, the appliance and equipment manufacturing sector, and the media and telecommunications sector. Variable

Renewable energy technologies – including solar PV, onshore and

energy (VRE)

river hydropower – whose maximum output at any time depends on

renewable

offshore wind, concentrating solar power without storage and run-ofthe availability of fluctuating renewable energy resources, such as wind or solar insolation.

plant

units such as wind farms and solar parks, as well as flexible power consumers and batteries.

Vehicle-to-grid

Technology by which an EV can both draw electricity from the grid

Web crawling

A technique that uses algorithms to automatically collect information

Workload (data

In computing, the workload is the amount of processing that the

Zettabyte

A measurement of data equivalent to 1021 bytes, or a billion terabytes.

(V2G)

centre)

Glossary

A network of decentralised, small or medium-scale power generating

and supply it back into the grid.

from websites. (Also known as web harvesting or web scraping.)

computer has been given to do at a given time.

181

© OECD/IEA, 2017

Virtual power

© OECD/IEA, 2017

Abbreviations and acronyms ACES

automated, connected, electric and shared

ANSSI

Agence nationale de la securité des systèmes d’information

artificial intelligence (France)

API

application program interface

BEV

battery electric vehicle

ASEAN CAPEX

CBECS CCS

CDA

CEN-CENELEC

Association of Southeast Asian Nations capital expenditure

Commercial Buildings Energy Consumption Survey (United States) carbon capture and storage

Connected Devices Alliance

European Committee for Standardization-European Committee for Electrotechnical Standardization

CERT

computer emergency response team

CNPC

China National Petroleum Corporation

CFL

CO2

CRISP

DER

compact florescent lamp carbon dioxide

Cybersecurity Risk Information Sharing Program

distributed energy resources

DOE

Department of Energy (United States)

DSRC

dedicated short-range communications

DoS

EPEAT ESCO

Denial of Service

Electronic Product Environmental Assessment Tool energy service company

EV

electric vehicle

GCI

Global Cloud Index

FCEV GDPR GIS

GPS G7

fuel cell electric vehicle General Data Protection Regulation (European Union) geographical information system global positioning system Group of Seven

IAC

Industrial Assessment Center (United States)

IEA

International Energy Agency

ICT

information and communications technology

IEC

International Electrotechnical Commission

IoT

Internet of Things

IEEE IP

Abbreviations and acronyms

Institute of Electrical and Electronics Engineers Internet Protocol

183

© OECD/IEA, 2017

AI

ISAC

Japan Electricity Information Sharing and Analysis Center

IT

information technology

ITS

ITU

intelligent transport systems

International Telecommunications Union

J-CSIP

Cyber Security Information-sharing Partnership of Japan

LED

light-emitting diode

Li-Fi

light fidelity

LBNL

LIDAR LPWA

Lawrence Berkeley National Laboratory laser-imaging detection and ranging low power wide area

MaaS

Mobility as a Service

M2M

machine-to-machine

MAIT

NATO NEED

Manufacturers’ Association for Information Technology (India) North Atlantic Treaty Organization

National Energy Efficiency Data Framework (United Kingdom)

NERC

North American Electric Reliability Corporation

NIST

National Institute of Standards and Technology (United States)

NIS

NPS

Network Information Security Directive (European Union) New Policies Scenario

O&M

operation and maintenance

OPEX

operating expenditure

OECD

Organisation for Economic Co-operation and Development

OT

operational technology

PHEV

plug-in hybrid electric vehicle

PAYG PoE

PPA

PUE PV

QR

pay-as-you-go

Power-over-Ethernet

power purchase agreement power usage effectiveness photovoltaic

quick response

R&D

research and development

RFID

radio-frequency identification

SAE

Society of Automotive Engineers

RECS

RTS

Residential Energy Consumption Survey (United States)

Reference Technology Scenario

SCADA

supervisory control and data acquisition

V2G

vehicle-to-grid

UAV V2I

V2V yr

3D 184

information security management system

unmanned aerial vehicle

vehicle-to-infrastructure vehicle-to-vehicle year

three-dimensional Abbreviations and acronyms

© OECD/IEA, 2017

ISMS

Units of measure GB

Gbps

bits per second gigabyte

gigabits per second

GJ

gigajoule

h

hour

kWh

kilowatt hour

GW kW

lm/W

gigawatt

kilowatt

lumens per watt

Mbps

megabits per second

m2

square metre

Mt

PWh toe

Twh W

Abbreviations and acronyms

million tonnes petawatt hour

tonne of oil equivalent terawatt hour watt

185

© OECD/IEA, 2017

bps

.

y Agency

ab

le

To ge t her

Internationa l

iea

E-mai

w

erg En

PD

l: bo

oks

S e c u re

ta Sus

in

0% dis c o u s at 2 nt io n er s

or

o o ks b / g

Fv

Onl in w w.

e

s k h o o o p b

@ iea

g . or Global Gas Security series

Energy Policies Beyond IEA Countries series

Energy Technology Perspectives series

Energy Statistics series

World Energy Outlook series

Energy Policies of IEA Countries series

Oil

Gas

Coal

Renewable Energy

World Energy Investment series

Market Report Series Energy Efficiency

This document and any map included herein are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. This publication reflects the views of the IEA Secretariat but does not necessarily reflect those of individual IEA member countries. The IEA makes no representation or warranty, express or implied, in respect of the publication’s contents (including its completeness or accuracy) and shall not be responsible for any use of, or reliance on, the publication. Unless otherwise indicated, all material presented in figures and tables is derived from IEA data and analysis.

IEA PUBLICATIONS, International Energy Agency Website: www.iea.org Contact information: www.iea.org/aboutus/contactus

Cover design: IEA. Photo credits: © Getty Images Layout in France by IEA, October 2017 IEA/OECD possible corrigenda on: www.oecd.org/about/publishing/corrigenda.htm

Digitalization & Energy Digital technologies are everywhere, affecting the way we live, work, travel, and play. Digitalization is helping improve the safety, productivity, accessibility, and sustainability of energy systems around the world. But it is also raising new security and privacy risks, while disrupting markets, businesses, and workers. Digitalization & Energy is the International Energy Agency’s first comprehensive effort to depict how digitalization could transform the world’s energy systems. The report examines the impact of digital technologies on energy demand sectors, looks at how energy suppliers can use digital tools to improve operations, and explores the transformational potential of digitalization to help create a highly interconnected energy system. This report seeks to provide greater clarity to decision makers in government and industry on what digitalization means for energy, shining a light on its enormous potential and most pressing challenges. It also lays out no-regret recommendations to help steer the world towards a more secure, sustainable, and smarter energy future.