White Paper

We support the Sustainable Development Goals

Virtualizing power systems: how digital

twins will revolutionize the energy sector
Executive summary
A digital twin is a digital representation of a Models and simulations in the power system
target entity with data connections that enable are nothing new. What is new is building highly
convergence between the physical and digital accurate models and sharing those models among
states at an appropriate rate of synchronization [1]1. departments within electric utilities and between
This white paper focuses on the implications of electric utilities in order to support a wide range of
the electrical power grid within the energy sector planning and operations.
as a target entity. Effective use of digital twin
These models come in two basic forms:
technologies can help grid planners and grid
operators in a number of important ways: Equipment models describe the internal functions
of individual devices. They enable analysis and
§ Meet net-zero goals
understanding of individual device behaviours,
§ Handle accelerating load growth which is vital to the utility’s asset management
practices.
§ Support grid resiliency and security
Grid models describe how equipment is
§ Support distribution-connected energy
connected into a cohesive system operating as
resource proliferation
a whole. They enable full-scale grid simulations
§ Overcome infrastructure challenges essential for planning, protecting, and operating
§ Adapt to climate change today’s increasingly complex power system.

As interdependencies grow from a more connected

society, alongside traditional users of power
system grid models, there will be new consumers
of this data. They will include new entities in the
electricity sector, such as suppliers of home
devices that monitor the value and environmental
impacts of electricity usage. In anticipation of new
sector coupling, digital twins can help prepare
for better collaboration between entities with
radically different data needs. For example, the
electricity sector will need to work closely with the
emerging green hydrogen sector, and with other
energy related sectors which will see increased
collaboration, such as the electric vehicle network.
Building digital twins for such complex systems is
not trivial. And building digital twins while the grid
is evolving so rapidly is, as the saying goes, “like
building the plane while flying”.

1 Numbers in square brackets refer to the Bibliography.

3
Executive summary

One of the key areas of modelling that requires This paper introduces the concept of the five “C”
more exploration is the new “grid edge”. Utilities datasets to describe these data: capability data,
and parties connected to the power grid must configuration data, commercial data, controls
share more information so that resources like data, and conditions data. This white paper’s
battery storage and electric vehicles can be recommendations provide tangible actions that
utilized to achieve the goals of a reliable grid that is government agencies, standards bodies, and
economical and supports a green future. digital twin stakeholders can take to unlock the
potential of digital twin technologies and their
Data from the grid edge married to utility grid
revolutionary impact on the energy sector of the
models form the core of the digital twin, including
future.
data covering all phases of the grid lifecycle.

4
Executive summary

Acknowledgments

This white paper has been prepared by a project Amiyu Kumudun, Peak Engineering & Consultancy
team representing a variety of organizations Pte Ltd
working under the IEC Market Strategy Board
Yoshiaki Kushima, Tokyo Electric Power Company
(MSB). The project team included representatives
Power Grid, Inc
from electrical power network businesses,
standards organizations, and equipment vendors Peter J. Lanctot, IEC MSB Secretary
from around the world. Sweta Malik, Ph.D., University College Dublin
The project was co-sponsored by Hiroshi Okamoto, Tatsuro Morino, Tokyo Electric Power Company
Ph.D., from TEPCO Power Grid, Inc. and Maurizio Power Grid, Inc
Bragagni, Ph.D. from Tratos, Ltd., both IEC Market
Matija Naglič, TenneT TSO
Strategy Board members. Coordinating author
and project partner was Scott Coe, Ph.D., from Stefan Niessen, Ph.D., Siemens
GridOptimize. Co-convenors of the project were Taku Noda, Ph.D., Central Research Institute of
Dr Yoshimitsu Umahashi from TEPCO Power Grid, Electric Power Industry (CRIEPI)
Inc, and Mr Peter Mandich from British Standards
Rem Noormohamed, Field Fisher
Institution (BSI).
Jaye Nozarick, Office of Gas and Electricity
Markets (Ofgem)
Project team members (in alphabetical order):
Gopal Ramchurn, Ph.D., University of
Rakesh Bali, Capula Southampton
Pankhuri Bansal, Blockom Consulting Dan Rossiter, British Standards Institution (BSI)
Jay Britton, Britton Consulting Mahmoud Shahbazi, Ph.D., Durham University
Patricia Brown, CIMple Data Chang-hoon Shin, Ph.D., Korean Electric Power
Abbey Dorian, British Standards Institution (BSI) Research Institute (KEPRI)

Ismail Drhorhi, ONEE Branche Electricité Muthu Subramanian, ISO New England

Thomas Duerr, Siemens AG Ximeng Sun, Electric Power Research Institute of

China (SEPRI) CN
Simon Evans, Arup
Chang-Hwan Sung, Korea Electric Power
Sangjin Jeong, Electronics and
Company (KEPCO)
Telecommunications Research Institute (ETRI)
Paul Surin, Trojena
Shivkumar Kalyanaraman, Ph.D., Energy Industry,
Asia, Microsoft Hiroshi Takahashi, Fuji Electric

Tatsuya Kato, Tokyo Electric Power Company Noboru Takao, Tokyo Electric Power Company
Holdings, Inc. Holdings, Inc.

Shunsuke Kawachi, Toshiba Energy Systems & Paul Thompson, National Grid ESO
Solutions Yoichi Tomiki, Tokyo Electric Power Company
Naoki Kobayashi, Tokyo Electric Power Company Holdings, Inc.
Power Grid, Inc.

5
Executive summary

Sebastiaan Van Dort, British Standards Institution

(BSI)

Adrian Vinsome, EA Technology

Peter Winstanley, Ph.D., Semantic Arts

6
Table of contents

Executive summary 3

List of abbreviations 9

Glossary 12

Section 1 Introduction 13
1.1 Mission 13
1.2 Vision 14
1.3 Why an energy sector revolution? 14
1.4 Implications of inaction 15
1.5 What is a digital twin? 15

Section 2 Digital twins today 18

2.1 The origin story 18
2.2 Equipment models 19
2.3 Connectivity models 20
2.4 Organizing digital twin data 21

Section 3 Drivers of radical change 23

3.1 The new grid edge 23
3.2 Energy storage 23
3.3 Electric transport 24

Section 4 Enablers 25
4.1 Adopt data-centric thinking 25
4.2 Improve sensor measurements 25
4.3 Gather data from the grid edge 27
4.4 Encourage manufacturer data exchange 28
4.5 Explore multiple representations 28
4.6 Leverage graph data stores 29
4.7 Consider data sovereignty 30
4.8 Bridge other sectors 30

7
Table of contents

Section 5 Realizing results 32

5.1 Improve grid planning 32
5.2 Enhance testing and staff training 32
5.3 Refine prediction algorithms 33
5.4 Implement flexible demand forecasts 34
5.5 Improve grid reliability 34
5.6 Align economic signals 35

Section 6 Bridging the virtual word 37

6.1 Preparing for revolution 37
6.2 Powering the virtual world 37
6.3 Creating a self-balancing power grid 38
6.4 Accessing digital twins 40
6.5 Anticipating a global energy interconnection 41

Section 7 Crucial standards 43

7.1 Smart energy grid architecture model 43
7.2 The common information model 43
7.3 IEC 61850 45
7.4 Enabling device data exchange 46
7.5 ISO/IEC JTC 1/SC 41 47

Section 8 Recommendations 49
8.1 Gain consensus on standards 50
8.2 Adopt open data practices 51
8.3 Build and operate digital twins 52
8.4 Interact with digital twins 53

Bibliography 54

8
List of abbreviations
Technical and AC alternating current
scientific terms AGC automatic generation control

AI artificial intelligence

AMI advanced metering infrastructure

AMR automatic meter reading

AR augmented reality

BECCS bioenergy with carbon capture and storage

CGMES common grid model exchange standard

CIM common information model

DC direct current

DER distributed energy resource

DLMP distribution locational marginal price

DTI digital twin instance

DTP digital twin prototype

EM electromagnetic

EMS energy management system

EMT electromagnetic transient

ESMP European-style market profile

GEI global energy interconnection

GHG greenhouse gas

HIL hardware-in-the-loop (testing)

IDPP intelligent digital power plant

IED intelligent electronic device

IOP interoperability event

IoT Internet of Things

IT information technology

JTC joint technical committee (ISO/IEC)

LMP locational marginal price

9
List of abbreviations

LTDS long term development statement

NEMF numerical electromagnetic field

OPC-UA open platform communications unified architecture

PLC powerline carrier

PTP Precision Time Protocol (IEEE)

RF radio frequency

RSC regional security coordinator

SC subcommittee

SCADA supervisory control and data acquisition

SDG United Nations Sustainable Development Goal

SDO standards developing organization

SGAM smart grid architecture model

SIL software-in-the-loop (testing)

TC technical committee

TSO transmission system operator

UHV ultra-high voltage

UML unified modelling language

V2G vehicle-to-grid

VR virtual reality

Organizations, ANSI American National Standards Institute

institutions,
BPA Bonneville Power Authority
companies and
organizations ENTSO-E European Network of Transmission System Operators
structures
for Electricity

ERCOT Electric Reliability Council of Texas

IEA International Energy Agency

IEC International Electrotechnical Commission

IEEE Institute of Electrical and Electronics Engineers

IPCC Intergovernmental Panel on Climate Change

10
List of abbreviations

IRENA International Renewable Energy Agency

ISO International Organization for Standardization

KEPCO Korea Electric Power Corporation

KEPRI Korea Electric Power Research Institute

MSB IEC Market Strategy Board

NASA National Aeronautics and Space Administration (US)

Ofgem Office of Gas and Electricity Markets (UK)

SMB IEC Standardization Management Board

TMB ISO Technical Management Board

11
Glossary
deadband

band of input values in the domain of a

transfer function in a control system or signal
processing system where the output is zero

feeder

type of electrical line or wire that carries power from

a main distribution point, such as a substation,
to smaller distribution points or to the end-users

grid edge

area where electricity distribution transitions

between the energy utility and the end user

Kirchhoff’s current law

algebraic sum of currents in a network

of conductors meeting at a point is zero

Kirchhoff’s voltage law

directed sum of the potential differences (voltages)

around any closed loop is zero

12
Section 1
Introduction

1.1 Mission paper demystifies the applicability of digital

twin technology for electrical energy sector
The electrical power sector is undergoing the most
stakeholders. The text emphasizes the unique
radical change in its history, which dates back to
benefits of digital twin technologies in integrating
the deployment of the first power grid nearly 150
renewable energy to deliver maximum benefits for
years ago. A convergence of factors is stretching
the environment, economy, and for communities
power grids to their functional limits, while the
today and into the future.
electrical power industry strives to decarbonize,
promote efficiencies, and prioritize human needs This white paper identifies the key enablers of digital
during the transition. twin technology for the energy sector, dissecting
the core components of virtualized power grids
This white paper reassesses digital twin
through concrete examples. By providing tangible
technologies as key tools to address the sweeping
and actionable standards and frameworks, this
challenges of the net-zero transition in the energy
white paper promotes the capitalization of digital
sector. It suggests initiating a step change in the
twin technology to revolutionize the energy sector.
role of standards to facilitate the decarbonization,
digitalization, and decentralization of the energy Applying digital twins in the power sector has the
sector. potential to advance the United Nations Sustainable
Development Goals (SDGs). Specifically, power
By reassessing the enabling components of
grid digital twins can support:
highly digital power grids of the future, this white

SDG 7 Affordable and clean energy, by optimizing the placement and operation of
renewable energy generation sources

SDG 9 Industry, innovation, and infrastructure, by enabling smart grid technologies,

such as self-balancing distribution networks

SDG 11 Sustainable cities and communities, by enhancing urban energy systems,

such as improved building load management

SDG 13 Climate action, by improving energy ability to host increasing numbers of

distributed and renewable energy sources

13
Introduction

1.2 Vision generation devices like solar and wind generation

on the distribution system. This disrupts the long-
A revolutionized, decarbonized, digitalized, and
standing design of power grids, which facilitate
decentralized energy sector leveraging digital
the flow of energy from large-scale generation
twin technologies will maximize benefits for the
on the transmission system to loads on receiving
environment, economy, and communities today
distribution systems. While the deployment of
and in the future.
distributed energy resources (DERs) is an overall
positive development allowing customers to take
1.3 Why an energy sector part in the net-zero transition, operating power
revolution? grids in this way requires re-engineering and might
necessitate protection scheme reconfigurations
Net-zero goals. There is international consensus
and costly grid upgrades.
to reduce greenhouse gas (GHG) emissions as
they are associated with global warming and Infrastructure challenges. By 2030, renewable
climate change. Net zero is the drive towards power will likely surpass 10 terawatts globally,
balancing human-caused residual GHG emissions nearly quadrupling the current capacity [5]. This
with human-led removals [2] resulting in a net- brings to the forefront the limitations of existing
zero increase in GHG in the atmosphere. The transmission capacity, making infrastructure
Paris Agreement [3] calls for each of the 196 modernization and transmission system expansion
signatory countries to achieve net zero by 2050. a priority. Investment in infrastructure renovation
As ambitious as that goal is, the looming 45% and expansion projects is crucial for facilitating
reduction target for 2030 may be even harder to the energy transition and accommodating DERs.
achieve. According to the International Energy The importance of these efforts will be heightened
Agency (IEA) [4], global energy usage rebounded in the coming decades with the increasing
in 2021 after the COVID-19 pandemic, leading prominence of renewable energy sources, whose
to the largest ever annual increase in global CO2 variability requires robust, flexible grid networks.
production. In other words, infrastructure developments must
be aligned with long-term planning and reflect
Accelerating load growth. An overall accelerating
broader strategies, including introduction of local
growth in electricity usage is stemming from many
and regional energy markets
facets of societal transformation, most notably
digitalization (converting traditional manual Adapting to climate change. In its most recent
processes into computer-based processes), analysis [6], the Intergovernmental Panel on
and electrification (transitioning from fossil-fuel Climate Change (IPCC) has data to show that
consuming devices like heating systems and major tropical cyclones have increased globally
vehicles). Energy justice, which aims to ensure in the last four decades. Severe storm activity
the fair and equitable distribution of the benefits has affected power generation, transmission, and
and burdens associated with energy production, particularly distribution systems that extend to the
distribution, and consumption, also promises to customer’s electricity meter. In short, the more
generate an increase in electricity needs to achieve severe and the more frequent storm activity, the
fairness and inclusiveness. harder it is to keep the power grid both stable and
affordable.
Distribution-connected energy resource.
Another factor driving change is the large
number of customers installing local, small-scale

14
Introduction

In the electricity sector, resilience refers to a grid’s 1.5 What is a digital twin?
reduced vulnerability to multiple failures, whether
In recent years the term “digital twin” has become
due to temporary outages or permanent damages
quite popular in the marketing arena. However, the
to network and control equipment. Extreme
concept is more than just marketing; it describes a
weather and rising temperatures caused by GHG
technology that can bring immense benefits if used
emissions will affect the entire energy sector. With
correctly. The IEC formally defines a digital twin as
climate change, extreme weather, and increasing
a “digital representation of a target entity with data
power outages, critical electricity infrastructure will
connections that enable convergence between the
need to be both resilient and digitalized to facilitate
physical and digital states at an appropriate rate of
rapid response and grid optimization.
synchronization” [1].

In their groundbreaking research entitled Digital

1.4 Implications of inaction Twin: Mitigating Unpredictable, Undesirable
Accelerating load growth, variability of renewable Emergent Behaviour in Complex Systems, authors
generation (particularly in distribution systems), Michael Grieves and John Vickers conceptualized
limitations of existing transmission system the digital twin and distinguished two fundamental
infrastructure, and increasing challenges of types [7]:
extreme weather are together causing a decrease § A digital twin prototype (DTP) is used to
in grid resiliency and complicating efforts to realize model the physical system before it is produced
the full potential of renewable energy generation so that simulations can be performed to
in the electricity sector. Developers proposing improve the design.
renewable energy sources are already facing
§ A digital twin instance (DTI) is used to
interconnection approval delays and higher costs.
describe an instance of the production version
Digital twin technology as a predictive tool to of the physical device throughout the life of
mitigate or eliminate risk is the industry’s best that device.
strategy for counteracting the increasing loss of
Grieves and Vickers applied the concept of
power grid resiliency caused by ever-increasing
digital twin to NASA projects. Accordingly, here
demand, lagging rates of investment in grid
the design of a new rocket engine is used as an
infrastructure, and the increasing physical effects
example where a DTP of the engine is created and
of a changing global climate. Digital twins can
refined during the design process. If seven engines
enable grid designers, grid operators, and grid
were produced based on the shared DTP, each
users to better coordinate the challenges of today
physical engine would have its own DTI tracking
and support the increasing pressures and varying
the data associated with its individual lifecycle.
requirements of the future. If properly designed
with well-organized data ready for use by both In this context, individual pieces of electrical
traditional grid players and society in general, current-conducting equipment used in the
digital twins will help bring certainty and stability to power system can benefit from the digital twin
a future that is currently very uncertain. modelling process. From relatively simple devices,
such as a line switch, to those as complicated
as a hydroelectric or nuclear power plant, the
use of DTPs and DTIs can improve design and
manufacturing. The more impactful case, however,
may be the power grid itself. Power grids are
complex, interconnected systems, made up

15
Introduction

of many components that all work together to data structure and format) the effort can focus
keep the voltage and frequency stable for each on analyzing simulation results rather than on
synchronous system. transforming or manually entering data.

Consider a grid operator who must expand a Features that distinguish this approach from
power grid with a new undersea transmission line historical approaches to managing grid data
to connect a planned offshore wind farm. To study required to support grid planning, design,
different configurations of the new grid, a grid construction, and operation include:
planner uses a digital twin of the existing power
§ A digital twin of the power system describing
system (the grid DTI) with an added model of the
the behaviours of each relevant component of
proposed wind farm (a small grid DTP including
the electrical grid.
multiple turbine string component DTPs and
their relationships), plus several options for the § A digital twin of the power system documenting
transmission line (again, each option would be a how all components are assembled into an
small grid DTP made up of component DTPs and interactive network.
their relationships). The DTI of the existing power § A digital twin of the power system evolving over
system can be combined with the DTP of the farm time with a single history incorporating both
plus one of the transmission line option DTPs to planned changes scheduled by the utility and
create an assembled DTP of one alternative for a unplanned changes resulting from equipment
complete future grid. The various assembled DTPs failures and environmental forces.
can be utilized in system studies throughout the
§ A digital twin of the power system capturing
project planning and the design and construction
many possible future states with different
process, with a single DTP ultimately emerging as
options for upgrades and expansions and
representing the grid design to be energized. Once
with varying predictions for the many energy
energization occurs, the assembled DTP becomes
producers and consumers.
the next version of the DTI, adding to the DTI
lifecycle history. § A digital twin of the power system managed
as a centralized “single source of truth” and
During the process of evaluating and culling
thus capable of being uniformly shared with
alternatives, assembled DTPs are used by multiple
different users.
software tools to perform different types of system
simulation studies. Since all the assembled DTPs § A digital twin of the power system that can be
are expressed in a consistent form (i.e. the same leveraged in different tools to simulate different
scenarios aligned with the goals of those users.

Historical Analysis Operations & Engineering

Maintenance Construction Design Planning

Realized Digital Twin History Many Potential Digital Twin Futures

Now Hours Days Months Years

Figure 1-1 | Time-evolving electrical power system digital twin

16
Introduction

As shown on the left in Figure 1-1, the power

system digital twin (DTI) has a single history
evolving over time. A grid operator's planning
department has the most future-looking system
digital twins (various DTPs), with different planning
scenarios associated with different predictions
that require different countermeasures to reinforce
the grid. Over time, as the future becomes more
certain, the engineering design group and the
grid operator will be able to create construction
plans that address the needs of the grid. This
process typically narrows the number of DTPs to
one. Those plans are then passed to construction
crews who build or reinforce the grid as instructed.
Finally, during a process called “energization”,
the grid operator energizes the new or upgraded
circuits. By that point, it is known which of the
multiple future DTPs has been realized, and the
associated set of grid changes becomes part of
the DTI.

A final note here: each digital twin of a network

owned by a single grid operator can interconnect
both “horizontally” across transmission grids and
“vertically” between transmission and distribution
grids. Taking this philosophy to the extreme, all
grid operators manage their part of the federated,
time-evolving, interconnected electrical power
system digital twin.

17
Section 2
Digital twins today

2.1 The origin story advancements. The first was the inception
of the energy management system (EMS) for
The most unique feature of digital twins for the
transmission operations with a detailed simulation
electrical power grid is the term “digital twin” itself.
capability known as “state estimation”, which took
Throughout the history of the electric grid, engineers
in measurements and fit them to a model of the
have established digital twins leveraging the
present state of the grid. Then, leveraging similar
technologies of the day to build ever-more-complex
methods in an off-line mode, other applications
models of the power systems. Setting aside the
were developed to analyze grid reliability and
many years when calculations were done by hand
other concerns. In the long-term planning domain,
on paper, the first major innovation occurred with
suites of planning analysis applications came to
the advent of computers in the 1950s.
market and were put into use across the globe
The digital era of power system models was allowing grid owners and operators to make better
launched in that decade by Glenn Stagg, Homer decisions on grid design and development.
Brown, Bruce Shipley, Harvey Happ, and other
Over time, punch cards were replaced with file-
pioneers working to explore computer simulation
based systems, making the editing far more
techniques. By the 1960s, early versions of
flexible. This led to neighbouring utilities being able
planning software were utilized to support previous
to share their files to perform wide-area analyses.
manual calculations, and near the end of the
In the 1980s, EMSs with state estimation and other
decade two major developments transformed
simulation applications became the norm; but data
the industry. First, digital supervisory control and
quality and data management lagged.
data acquisition (SCADA) for automatic generation
control (AGC) systems became available to replace By 1990, most transmission control centres had
analogue controls in transmission control centres. assigned sufficient personnel to scrub data and
Second, Bill Tinney, and John Walker at Bonneville get all major simulation applications running in real-
Power Authority (BPA) in the United States time, based on state estimation providing minute-
wrote several landmark grid analysis papers [8] to-minute simulation of the grid. Most EMSs
outlining computational methods that could solve required an engineer dedicated to grid modelling
much larger problems than were feasible by data quality to get a state estimator to “converge”,
prior methods. Jay Britton, contributing author meaning solving the complex mathematics
to this paper, was one of the first to utilize these to create a solution obeying Kirchhoff's laws
techniques when he successfully created a three- across the system. Data management tools also
thousand node tool used for planning power flows. improved, although maintaining accurate models
of constantly evolving neighbouring systems
As every year passed, advances in computer
proved difficult.
science, in both hardware and software, facilitated
development of more sophisticated software In the 1990s, capabilities had progressed so that
tools. Sector trends in the 1960s included several electricity markets could be introduced based on

18
Digital twins today

algorithms like locational marginal pricing, which regional model, performs a power flow analysis,
worked from the state estimation of the grid. and publishes that model back to the TSOs.
With the integration of data quality practices,
To better understand how grid models can be
projections and grid modelling, digital twins
exchanged, it is important to understand what a
became deeply embedded in both grid reliability
grid model is and what it contains. The following
and grid economics. Where in past decades
sections explain how models of equipment and
the grid tended to be overbuilt due to a lack of
models of connectivity are used to build grid
precise knowledge about how it would perform,
models.
now successful operation of the grid was digital-
dependent. Computers became cheaper and thus
a better option than overbuilding the grid, and 2.2 Equipment models
common techniques emerged to create power
The topology of the electrical grid has the key role
system digital twins in uniform ways.
in any power system digital twin, and it is imperative
The 21st century has seen significant transformation to understand this foundation before transactional
in both power systems and their digital twins. data can be introduced. The representation of the
For the first time, common models of the grid electrical grid must be known in its current state
were being exchanged not only within utilities but to efficiently operate the grid. This representation
also between them in non-proprietary formats. is, in fact, the digital twin instance, and is more
Beginning in 2009, the Electric Reliability Council commonly known in the industry as the “as-built”
of Texas (ERCOT), the grid and market operator for grid model. A log of grid modifications is kept
that US state, began collecting grid representations so that the grid can be traced back in time and
from its member transmission utilities using IEC- incidents can be analyzed to determine causes
based standard data exchanges. ERCOT’s first to inform decisions on grid improvements. Grid
model included more than two million individual planners must also map many potential future
conducting equipment elements, and during its states of the grid, or digital twin prototypes, which
first decade of operation, it added more than one accommodate anticipated changes to generation
million new elements, removed half a million old sources, consumer loads, and grid equipment
elements, and made hundreds of thousands of owned and operated by each utility.
parameter changes.
The electrical grid representation is comprised
Europe went further, implementing the common of individual representations of each piece of
grid model exchange standard (CGMES) for all equipment installed on the grid. The equipment
European Union member countries. Like the model tracks the unique characteristics of each
ERCOT implementation (which collects individual type of device, including parameters to track
utility models to assemble a grid model for its entire intrinsic values such as line impedance, and
balancing area), each regional security coordinator operational constraints such as line rating.
(RSC) in Europe assembles models from its
Large or complex equipment, like substation
member transmission system operators (TSOs)
transformers and high-voltage transmission lines,
in order to facilitate grid congestion forecasting.
have dedicated models, either supplied by the
The roughly 40 TSOs in Europe publish their grid
manufacturer or assembled from calculations
models updated every hour of every day for current
based on physical characteristics or measured
conditions including unplanned outages and circuit
in the field. These models can be thought of as
rating changes. Each RSC then assembles its
simplified versions of the manufacturing models,

19
Digital twins today

retaining only the electrical information needed to Problems with transmission models can lead to
perform grid modelling and operations. problems operating the grid that can cascade into
major outages.
Equipment models often incorporate behaviours,
including response functions that emulate how Electricity distribution systems have many
the device will operate under changing conditions more pieces of equipment in their models than
such as voltage or grid frequency fluctuations. The transmissions systems do. The amount of
numerous simple and smaller pieces of equipment, equipment to be tracked from a typical utility can
pole-mounted distribution transformers for number in the millions. Fortunately, the implications
example, have equipment models that are of errors in the model generally lead to smaller grid
“shared”, meaning typical values are used for all impacts. And since sophisticated studies have not
instances. been needed up to now, the challenge of building
highly detailed, highly reliable distribution grid
models has been largely ignored.
2.3 Connectivity models
Digital twins offer a revolutionary approach to
Equipment models are only half the picture. A addressing the complexities and challenges
connectivity model must be created to describe within both electricity transmission and
how the equipment is deployed and wired distribution systems. By creating virtual replicas
together. A key element of the connectivity model or clones, of physical grid components and their
is the terminal. Each piece of equipment in the interconnections, digital twins allow for real-time
field has one or more terminals. A breaker or a simulation, manipulation, and analysis of the entire
segment of conductor installed as a distribution grid's behaviour under various scenarios, including
line, for example, has a terminal on each end. sudden storm conditions, emergencies, failures,
Other devices, a busbar for instance, might have and planned future modifications. This dynamic
more than two terminals, while some simple modelling capability enables more accurate
representations of an “edge-of-the-grid” device prediction of system responses: it facilitates the
might be represented with just one. identification of potential issues before they lead
The terminal connections of each device must to outages, and it assists in planning and testing
be tracked. While much of the equipment can grid upgrades in a virtual environment – all before
be tracked as a series of connected devices, in actual implementation.
many cases the grid is a meshed system. This The management process becomes even more
means the connections are more difficult to track. challenging when the weather creates storm
Indeed, studies utilizing the models need complex damage and field crews must make quick fixes
algorithms to solve for the flow of power. This to the grid to restore normal operations. Looking
complexity comes because the laws of physics to the future, grid planners and engineers will
require a simultaneous balance across the grid, a design a set of proposed changes to the grid
balance based on both the intrinsic parameters in with new elements using equipment models and
the equipment models and the ways in which the a set of proposed changes to the connectivity
devices are wired in the connectivity model. model. Using both predictive models and live data
Electricity transmission systems have relatively on storm conditions, digital twins can help grid
few pieces of equipment and hence the model is operators manage damage to grid equipment,
relatively simple. However, the overall importance guide field crews to make the most effective repairs
of the grid model for transmission is very high. and to help anticipate the consequences on the

20
Digital twins today

immediate grid stability. Ultimately, the adoption that may be as granular as sub-cycle resolution.
of digital twins in power system management Descriptions of the five “C” datasets are provided
promises to produce greater reliability, efficiency, in Table 1.
and adaptability in grid operations. In other words,
digital twins will be a key tool in the modernization
of electrical grids.

2.4 Organizing digital twin data

A digital twin consists of very complex collections
of datasets with relationships among the data.
Unlike some datasets, such as online sales
records or the generation of artificial intelligence
(AI) artwork, the data inside a digital twin are
constrained by the laws of physics. To better
visualize these datasets, one can group them into
categories. The formalization adopted in this white
paper is shown in Figure 2-1.

Figure 2-1 | The five “C” datasets

Datasets closer to the bottom of the figure

represent static or semi-static information which
is essential to understanding equipment and
connectivity of the digital twin. As one moves
higher in the stack, data become more dynamic,
with the highest level representing measurements

21
Digital twins today

Table 1 | The five “C” datasets

Dataset Description

Capability Information related to a given manufacturer's make and model of a device related to
the limits of its operation, typically known as nameplate, datasheet, or catalogue data;
for example, maximum power output
Configuration Information about the device as it is installed and configured, including the electrical
connection and the operational and protection settings; for example, how a transformer
is installed in a substation
Commercial Information about the commercial agreements (contracts) to which the device is
subject, including rates, grid services definitions, and aggregation constituencies; for
example, an agreement to provide backup power
Controls Instructions issued to the device, including real and reactive power profiles and grid
support service schedules effective immediately or for a future time or times; for
example, a battery discharge schedule
Conditions Measurements collected from sensors embedded or attached to devices including
variables such as real, reactive, and apparent power, voltage, and frequency; for
example, the terminal voltage at a meter

22
Section 3
Drivers of radical change

3.1 The new grid edge twin, which historically ended at the customer’s
meter, now must extend farther. Access to such
Historically, power grid design was relatively
information is problematic for many reasons.
straightforward with power flowing in predictable
First, there are privacy issues: until there are clear
ways from large-scale generation through
guidelines from governmental bodies, there will
networks of transmission lines, stepped down to
be only fragments of intelligence. Second, there
distribution systems, to be finally consumed by
are technical challenges: utilities are skilled in the
electricity customers. Historical interconnections
proprietary communication protocols implemented
for generation were large-scale facilities, where
on dedicated networks to monitor their own
each type of facility – whether coal, natural gas,
equipment, but these systems are not well suited
or nuclear – came with decades of historical
to reach into the customer’s equipment. The
performance information. Now, interconnections
final challenge is scale. If standards and privacy
are much more complex. New bulk generation
challenges are resolved, there remains the
often comes primarily from large-scale wind and
significant challenge posed for utilities looking to
solar. While the peak output can be understood
expand their connections from a few hundred to
based on the capabilities of the generation devices,
millions of devices on a decentralized power grid.
knowing when those peaks happen is dependent
on climate and weather patterns.

Loads are, arguably, evolving more quickly than 3.2 Energy storage
generation. In addition to distributed generation – Energy storage refers to the capture and
which is also primarily weather-sensitive – the containment of energy produced at one time
loads themselves are more complicated. Factors for use later. This concept plays a crucial role in
that contribute to this challenge include less balancing supply and demand in the power grid,
regular work schedules; the electrification of more enabling the storage of excess energy generated
devices, including heating systems and industrial during low demand periods to be released during
automation; and perhaps, most critically, increasing periods of high demand. Energy storage is
levels of electric vehicles, which add large and critically important as variable renewables sources
effectively random offsets to the base load. In sum, become a larger fraction of the installed generation
flows are less predictable, and it is now common capacity [9].
to see systems that were designed to flow in one
Battery energy storage has become the
direction reverse with excess generation in places
dominant technology, but it remains expensive to
that historically saw only electricity usage.
manufacture and install. Roundtrip inefficiencies
The effect of load and generation volatility has converting AC power from the grid to DC power
created a greater need for utilities to understand in the battery and back leads to significant losses.
these new features “behind” the customer’s Developers are now installing systems with direct
meter. Thus, for it to be a useful tool, the digital battery charging from renewable generation to

23
Drivers of radical change

reduce the number of conversions from three (DC Connection. An additional modelling challenge is
generation to AC to DC storage and back to AC related to the financial agreements with different
for the grid) to a single conversion (DC generation entities. When an electric vehicle charges its
directly to battery and then a single conversion to battery – or provides grid services – it connects to
AC for the grid). a charging station, and there can be an agreement
with the vehicle owner, the charging station owner,
To properly model these elements in a digital twin,
or, in the most complex case, with both.
utilities must understand how the connections are
made and may need to begin to model DC sub-
circuits. Perhaps the biggest challenge of storage
systems is prediction, as the owners and operators
of storage systems have complete control over
their operations and may not be required to
disclose their plans to the utility. From the utility’s
perspective, these resources may seem to operate
with completely random patterns.

3.3 Electric transport

Electric transport shares many similarities with
battery energy storage systems, and because of
the potential scale of vehicles that will be available
to interact with the grid, electric transport may
ultimately be the most important resource on the
grid. But the management challenge for electric
transport, and hence the grid modelling challenges,
are substantially more difficult than for dedicated
battery energy storage systems.

Mobility. Of course, the first difference between

electric transport and a stationary battery system
is the former’s mobility. This means the modelling
cannot be defined to a specific location and must
be treated as a dynamic resource that can appear
anywhere on the grid.

Purpose. Equally important, the primary purpose

of electric transport is moving people and goods
from place to place. Thus, these resources can
play an important role in the grid, but only ever a
secondary one. Additional capacity from electric
transport is what can reasonably be used without
risking the owner’s ability to use the vehicle without
running out of charge, balancing the need for
transportation with the value earned by using the
vehicle as a grid resource.

24
Section 4
Enablers

4.1 Adopt data-centric thinking electric devices. These sensors measure, among
other things, voltage, current, temperature, and
For many years, and for the most part up to the
phase angle.
present, companies have purchased software
solutions to fit their needs, and those solutions Each typical thermal power plant has roughly
have operated on proprietary data stores. This 20 000 sensor points that stream data, generating
traditional approach is called an application- on the order of two billion measurement values
centric approach: the emphasis is on the each day. Extrapolating to the entire grid, the
features and functions of each application, with number of sensor points is staggering. The number
the representation of the data only a secondary is over three quadrillion (3 × 1015), allocated to the
concern. generation, distribution, transmission (including
substation equipment), and customer (including
As interoperability has become more important,
electric meter) domains, according to a study
software developers have created interfaces to
by the Korea Electric Power Research Institute
allow their solutions to exchange data with other
(KEPRI) [10]. See Figure 4-1.
solutions, in some cases with a partner solution
and in other cases with competing solutions. This
is a step in the right direction, but it is not optimal;
it encourages point-to-point solutions, with each
Transmission, 120
application needing a unique interface for each
dataset.
Distribution, 400
Instead of this traditional application-centric
approach, a better alternative is a data-centric one.
By establishing an enterprise-wide data model
that each utility manages, software solutions can
Customer, 15
operate on the central representation of the data. Generation, 2 700
Of course, the more each software company
can adopt a standard model, rather than its own
proprietary model, the easier it is for utilities to find
solutions that can function on that data without
translation. Figure 4-1 | Power grid sensor readings in
trillions/year

4.2 Improve sensor measurements

Many sensor measurement processes can be
A power grid has a vast number of sensors:
improved; three of the more important ones are
sensors in power plants, sensors in substations,
shown in Figure 4-2. All need attention because
sensors along transmission and distribution lines,
sensor data is one of the key inputs to power
and sensors embedded in customer-owned
system digital twins.

25
Enablers

V PROPRIETARY

PROCESSING
REQUIRES
DELAYED
W MEASUREMENT
SENSORS

HISTORIAN ANALYSIS
COLLECTION
(COMPRESSED) PLATFORM
SYSTEM

°C

Figure 4-2 | Challenges to sensor data processing

Case study 4-1 – Intelligent digital power plant (IDPP)

Leveraging sensor data from Korea Electric Power Corporation (KEPCO) generation facilities, KEPRI
has implemented a pilot system to improve sensor measurements incorporating many of the features
identified here. The project has demonstrated several innovative functions including:

§ AI-friendly, real-time sensor data management platforms

§ SirenX: an early warning system

§ AI-based status diagnosis

§ IoT-based diagnosis

§ Asset performance management

26
Enablers

Standard data formats. Even within a single To resolve perhaps the most prevalent issue
power generation facility, sensors are sourced with sensor data, one should look to IEC 62439.
from different vendors, and the data formats are This series of standards focuses on industrial
inconsistent between products, very often using communication networks and, importantly,
proprietary protocols. In addition to the challenge annexes the IEEE 1588 Precision Time Protocol
of different ways of describing similar data sets, (PTP). Digital twins can assist in providing a
many sensors compress readings, for example, reference model for all the sensor objects, allowing
by only reporting values outside of a “deadband”. measurement to be tied to established identifiers in
Thus understanding measurement data requires the digital twin so that translations to local naming
knowledge of the data format and the logic schemes are avoided.
employed to store the data.
There are many protocols that can be used to
The IEC 62541 series of standards [11] supports communicate controls to devices and to collect
the open platform communications unified measurements from devices. Historically, these
architecture (OPC-UA). Wider adoption of this communications were implemented using the
protocol would likely help reduce inconsistently SCADA protocols; however, SCADA is not well-
formatted data. OPC-UA may also provide an suited to communicate information outside the
effective mechanism to publish other power utility communications infrastructure. IEC 61850
system digital twin datasets, such as equipment was developed and deployed to support the
details and market information. need for high-speed, secure communications
within substations. This standard has been
Timely access. Another challenge is timeliness.
extended to support communications to the field,
Sensor data collected today is often delayed,
including those outside the utility communication
and available only later through access to a data
infrastructure. It has the potential to become the
historian, a software system optimized to collect,
global standard for field communications.
compress, store, and later retrieve large volumes
of time-series data. And while useful for analysis of
events in the past, the data is often not available to
4.3 Gather data from the grid edge
operational systems that might be able to act on
the information in near real-time. While today the percentage of sensors at the
customer domain is minuscule as a relative
Emphasis should be placed on collecting, storing,
percentage (see Figure 4-3), it is likely the area of
and exchanging sensor data closer to real-time.
most growth. There is an explosion of potential
Open-source tools readily available in the space
sensor data available with customer devices.
could supplant vendor solutions that do not
These devices offer enormous potential benefit if
support near real-time data processing.
utilities are able to access them and are willing to
Analysis-ready datasets. To leverage sensor leverage this data, despite not owning the resource
data effectively today, sophisticated labour- or having full trust in it as a data source.
intensive activities are required. Once complete,
Devices can provide information and like traditional
data can be fed into analysis platforms, including
electricity meters can be certified to high accuracy
emerging AI processing engines. While not the
and reliability metrics. These metrics are not
only issues in this space, time alignment and data
always cost prohibitive, and many devices on the
interpolation to create complete “snapshots” of the
market today are already compliant with utility
system are prominent ones.
requirements. With the wall between utility and

27
Enablers

4.4 Encourage manufacturer data

exchange
When a utility company purchases equipment for
the grid from a manufacturer, the utility should also
request a simulation model of the apparatus. The
simulation model should have sufficient accuracy
when integrated in the digital twin of the power
system. With different types of simulations often
necessary, more than one equipment simulation
model might also be required. With appropriate
equipment simulation models provided by the
manufacturers, the utility company can readily
perform simulations using the digital twin as soon
as the equipment is purchased.

This practice, unfortunately, is quite different

Figure 4-3 | Digital twin with interactions
from what is actually being done. Today, experts
for different simulation types (including transient
customer device removed, the knowledge that stability, frequency, voltage, black start, switching,
can be gained is immense. Data can be harvested lightning, inrush current, and protection) must
from devices dedicated to power production build models for the equipment, sometimes
and consumption, such as DC/AC inverters and inserting parameters for the equipment by hand
battery systems, which could, for example, provide into published response models. Building a local
local voltage measurement. Sensors that can aid simulation model can take weeks.
in the management of flexible loads, like smart By focusing on the manufacturers and requesting
thermostats and controllable water heaters, could an official model for each piece of equipment, not
also be more seamlessly integrated. only is the inefficiency of the utility minimized, but
By combining these data with existing sensor the reliability of the models should also increase.
measurements in standard formats, available in If standards are developed for publishing these
near real-time, and ready to be fed into analysis equipment response models, aligned to the
engines, the picture of the power system digital requirements of the grid operators, this will lead
twin comes into focus, as shown in Figure 4-3. not only to more effective studies but when used
Open data practices are essential to success in a real-time digital twin can also improve system
on the grid edge, with options that include both reliability.
regulations mandating exchange of required data
and incentives to customers to open their devices
4.5 Explore multiple
up for access.
representations
A challenge to building a digital twin in the
electrical power system is the need for multiple
visual representations, specifically a geospatial and
an electrical layout. The electrical layout is critical
for studies and is stored as a complex connectivity
model. Very few geospatial characteristics factor

28
Enablers

into the connectivity model and are generally 4.6 Leverage graph data stores
confined to the lengths and line spacing
Electric utilities are not known for their early
characteristics of transmission and distribution
adoption of new technologies. A practice of being
lines which enable the calculation of line
“one version behind” is common, as it allows other
impedance values. However, from the operations
users to find the software developers to eliminate
perspective, the geospatial layout is the critical
not only performance and logic flaws in software
representation: it is essential for guiding crew
but also security weaknesses that if exposed
to the correct locations, estimating renewables
can present major risks to the power grid. This
output, and coordinating with emergency services.
philosophy also extends to data stores. Most utility
What makes the issue of multiple representations systems operate on relational databases. While
more challenging is that each electrical grid needs this technology has served the sector well and has
both representations, though different portions its place, exploring the ideas of graph data stores
of the grid have different requirements for each. has much promise for digital twin enablement.
The connectivity models for transmission systems
Traditional relational databases store data in
and radially operated distribution systems can be
data tables. Each table has a key formed from
derived, for the most part, from their geospatial
one or more columns in the table, which form a
counterparts. Networked distribution systems
unique data record. For example, a customer
typically found in metropolitan areas and especially
account number might be the key to the primary
those found underground are much more difficult
account table. Other tables that need to reference
to manage because lines that cross using a simple
the customer account would use the account
two-dimensional geospatial representation do not
number key from the account table as a “foreign”
necessarily connect electrically.
key. These relationships are established by the
The most complex electrical models exist within database designer and allow users with an
substations. Substations have transformers, understanding of the meaning of the relationships
capacitor banks, circuit breakers, fuses, busses, to query the database by “joining” tables. This
and other equipment, all in a relatively small approach works very well for large datasets with
geographic space. So here the connectivity limited relationships. However, digital twins of the
model, if paired with a geospatial representation, electrical power system include just the opposite:
is likely to require a three-dimensional model. many small datasets with many relationships.
Fortunately, substations constructed today are
This is where graph data stores enter the picture.
often designed using three-dimensional computer-
They use nodes to store data elements and allow
aided design tools, and these representations
a node to have many relationships, each with a
can become part of the digital twin. Augmented
different meaning between itself and other nodes.
reality (AR) technologies, leveraging these models,
This is very different from relational data stores,
allow maintenance crews to perform tasks on the
which have rigid relationships, defined “externally”
physical equipment overlaid with metadata from
in the database design. Graph data stores allow
the digital twin to improve efficiency and increase
for a wide range of relationships to be stored along
safety.
with the nodes and make it easier to build queries.
In conclusion, digital twins have the capability to
hold and present data in different ways and offer
a promising solution to the challenge of managing
multiple representations of an electrical power
system.

29
Enablers

4.7 Consider data sovereignty Similarly, other sectors, such as the natural gas
pipeline sector, had their own systems for planning
Data sovereignty is a concept that describes the
and operations – largely independent of the
idea of any particular entity maintaining control
electrical systems. Thus, each industry became
over its data and of being able to secure its data
oriented toward its data models and its industry-
in accordance with the regulations under which
specific constraints, functions, and visualizations.
the entity operates, even when data is being
Despite similarities in their foundational data
exchanged among different countries with different
constructs, separate development of the data
regulations.
models by different sectors led to divergent
Historically, electric utilities have experienced a systems that are not easily made interoperable.
greater level of sovereignty over their data than
As the electrical system becomes the leading
other sectors have had over their data. This is
energy delivery system for the world, we expect
primarily because much electric utility data is
to see more coupling with other sectors, such
considered critical to a country’s national security
as natural gas, electric transport, and green
and has been exchanged on dedicated networks
hydrogen. Successful coupling in the digital world
owned and operated by the utility or on dedicated
is highly dependent upon strong interoperability
networks rented from telecommunications
standards for data exchange.
companies. Other sectors rely more heavily on
public communications infrastructures. Natural gas network digital twins. Over the past
few decades, the interdependencies between gas
The electrical power system digital twin, however,
delivery networks and electrical power grids have
represents a departure because it not only
become more pronounced. This coupling is based
requires the information collected through the
on factors like higher dependence on natural gas
utility communication infrastructure, it also must
as a fuel for electricity production (often displacing
leverage data collected through public networks.
coal) and the conversion of end-use systems, such
With these new interfaces, there are therefore data
as building heating from gas to electricity, primarily.
sovereignty issues to consider.
In the short term, these interdependencies are
Additionally, should the digital twin data repository likely to increase. But as natural gas is replaced
be hosted outside the utility IT infrastructures, even with renewable sources of both electricity and
more data sovereignty issues arise. The challenge heating systems, such overlap is likely to become
becomes still more critical if the data model for less important.
the digital twin is federated, with different entities
Hydrogen network digital twins. Governments
responsible for different components of the model.
and industries around the world are investing in
the green hydrogen sector. Hydrogen networks
4.8 Bridge other sectors will have a highly coupled two-way dependency
with electrical power systems, as the hydrogen
Historically, electrical power systems around the
networks will use electricity when renewable
world could operate relatively independently of
power generation exceeds demand and produce
other complex systems. Any analysis of a particular
electricity from the stored hydrogen when demand
power system required a good understanding of
exceeds generation.
the grid itself: what the infrastructure would look
like at a particular time with load projections based
on climate and weather prediction.

30
Enablers

Electric transport network digital twins. Even that is important enough to require modelling will
more complex than the interdependence between need to be represented in virtual space. Here,
the power grid and the hydrogen network is that data captures information about those commodity
between the power grid and the emerging electric flows, such as measurements tracking actual flows
transport network. Like the two-way hydrogen in near real-time and predictions about anticipated
network, the electric transport network will be future flows.
able to both consume and provide power. But
Digital twins for these sectors enable
the choice of when vehicles consume or provide
comprehensive modelling of their interactions with
energy will be influenced primarily by their owners’
the power grid, facilitating optimized operation
need to charge or drive their vehicles, and not
and planning. For instance, natural gas network
on the abundance or scarcity of electricity. When
digital twins can simulate the impact of gas supply
electric vehicles are grid-enabled, the door will be
on electricity generation, and hydrogen network
open for electric transport to provide grid support
digital twins can model the dynamic exchange of
services, commonly known as vehicle-to-grid
energy between hydrogen storage and electricity
(V2G) interactions.
production. Similarly, electric transport network
As shown in the bottom portion of Figure 4-4, digital twins allow for the coordination of vehicle
real commodities, including electricity, natural charging with grid conditions, enabling bidirectional
gas and hydrogen, flow among the real physical energy flows and grid support services.
systems. Every real flow or at least every flow

Digital Hydrogen
Storage & Elec
Network tric
Twins Vehicle Networ
k

Natural Gas Electric Generat

ion
Network & Power Grid Data

Physical Hydrogen
Storage & Elec
Network tric
Systems Vehicle Networ
k

Natural Gas Electric Generat

ion Electricity
Network & Power Grid
Hydrogen
Natural Gas

Figure 4-4 | Interdependent physical systems and digital twins

31
Section 5
Realizing results

5.1 Improve grid planning planned maintenance activities. As load and

generation, connected to both transmission and
The 2016 major power blackout in Australia [12]
distribution networks, change over time, planners
is an example of how a system can fail, despite
must study the grid to see where reinforcements
every device on the grid operating as designed.
are needed.
Following major storm activity that damaged
transmission lines in South Australia, several wind As the grid becomes less stable with renewable
farms tripped offline. The loss of generation in turn resources providing a larger portion of the grid’s
triggered a limit violation on a high-capacity import energy and support services, the need for more
transmission line, islanding South Australia from sophisticated studies is increasing. Different kinds
the rest of the grid, which then experienced a wide- of studies are performed to analyze a variety of
area blackout. Importantly, the generation sources different phenomena across different time domains,
performed as designed, with voltage fluctuations as shown in Figure 5-1. These studies include
leading to the sources tripping offline as a safety traditional, long-standing, steady-state simulations,
protocol. However, the system likely would have transient stability and frequency stability studies
fared better had these generation sources, instead that examine sub-second time analysis, and
of tripping offline, “ridden through” the issue until even electromagnetic transient simulation using
the system could recover. sub-cycle analyses. There are research initiatives
exploring numerical electromagnetic field (NEMF)
The point here is not to dwell on this failure but
simulations at even higher resolutions. Generally,
to learn from it. More sophisticated digital twins of
the smaller the timescale of the study, (a) the more
the grid, which could include models of distributed
detail required in modelling the grid representation,
resources and their operational and off-normal
(b) the smaller the sections of the grid that are
settings, can facilitate more varied study types
studied at once, and (c) the larger the computing
that can be relied on with greater confidence. The
power necessary for calculations.
results of such studies can lead to better device
settings to help protect both the equipment and
the system. 5.2 Enhance testing and staff
Studies are not only used to identify the root training
causes of off-normal conditions experienced Conventional approaches to testing protection and
during infrequent reliability events, like those of the automation systems include “open-loop” hardware
Australian blackout, but are an essential function of testing with external test equipment and “closed-
every utility’s short-term and long-range planning. loop”, also known as hardware-in-the-loop (HIL),
Studies are performed before new generation and testing with a near real-time digital simulator. The
new loads are approved to interconnect, when digital twin can provide an alternative because
major switching activities are scheduled and different software-in-the-loop (SIL) stages are
new circuits are energized, as well as for certain possible, starting from open-loop solutions to

32
Realizing results

Simulation by voltages and currents

Simulation by EM fields Simulation by phasors

Time domain
NEMF Stead-state
EMT simulations stability
simulations Three-phase, waveform-level simulations
Electric & magnetic simulations Positive-sequence,
solution Positive-sequence,
field-level solution phasor solution
phasor solution

NEMF: Numerical EMT: Electromagnetic Transient stability and Power flow, fault,
electromagnetic field transient frequency stability voltage stability and
simulations small-signal stability
simulations

1 ns 1 μs 1 ms 1s 10 s Steady
state

Figure 5-1 | Studies as a function of time domain

fully integrated closed-loop simulations. The test Engineer skills development. A digital twin
scenarios can be facilitated with a digital twin can support all interfaces and protocols between
because testing effort can be focused on the different protection devices, making it is easier
algorithms and the logic of the protection system, to learn and scale the protection and automation
rather than on building the grid representations. systems, and in the process supporting the
This can lead to improved processes, including migration to fully digital systems. This virtual
configuration testing, performance testing, and environment thus enables cost-effective training of
engineer skills development. engineers providing a learning environment with a
high number of devices.
Configuration testing. The digital twin approach
has the potential for highly scalable testing,
allowing the test of a single device or an entire
5.3 Refine prediction algorithms
substation. In particular, the configuration of
busbar protection systems is highly complex, with Building on the insights gained from advanced
multiple feeders needing to be configured using an grid planning and reliability studies, digital twins
array of signals from those feeders mapped into in the power sector significantly enhance load and
the simulation. generation prediction algorithms. By leveraging
comprehensive virtual models that mirror physical
Performance testing. Performance testing
power systems, digital twins enable utilities to
includes the verification of the protection system’s
simulate and analyze complex data streams in real-
speed, accuracy, selectivity, and timing across a
time. This capability allows for the integration of
collection of individual protective devices. These
diverse data inputs, such as historical consumption
tests are done to approve different devices for use
trends, operational conditions, and real-time
in the field. The digital twin can provide a highly
environmental factors. The refined accuracy of
accurate model, especially if the original firmware
these simulations empowers better forecasting of
and protection algorithms can be provided directly
both demand and supply variations.
by the manufacturer of the protection devices and
included in the model.

33
Realizing results

Effective load and generation prediction is crucial availability of renewable energy sources, which can
for maintaining grid stability, especially as the vary significantly throughout the day. By accurately
energy landscape evolves with the increasing predicting these fluctuations, utilities can adjust
incorporation of renewable energy sources and their generation and distribution strategies in real-
increasing energy storage capacity systems. time, optimizing energy delivery.
Digital twins facilitate this by allowing for the
Moreover, digital twins allow for testing of
dynamic adaptation of energy systems to
various demand response scenarios in a virtual
changing conditions, reducing the risk of over- or
environment before they are implemented in
under-generation, and ensuring efficient energy
the real world. This ensures that such strategies
distribution. These predictions are vital for routine
are more effective when deployed and helps in
operations and for anticipating and mitigating
customizing energy solutions to meet specific
potential disruptions, thereby enhancing overall
consumer needs without compromising on system
grid resilience.
stability or efficiency.
Moreover, the ability to simulate various operational
scenarios through digital twins can aid utilities
in developing more robust energy generation 5.5 Improve grid reliability
strategies. This leads to optimized scheduling To gain deeper insight into the actual state and
and dispatch of resources, ensuring that energy operational constraints of the power grid, the
supply meets consumer demand in the most digital twin – in this case, the DTI – plays a crucial
efficient manner possible. Through continuous role in enhancing situational awareness. Unlike
improvements in predictive accuracy, digital twins traditional control centre systems where the state
drive advancements in energy management of the power grid is only vaguely captured by
and contribute to the broader goals of energy SCADA measurements, the DTI has the potential
sustainability and reliability. for much more advanced capabilities. Digital twins
are envisioned to continuously fine-tune asset
parameters and automatically verify the response
5.4 Implement flexible demand to grid events.
forecasts
Digital twins can also provide near real-time visibility
Digital twins in the power sector are revolutionizing
into the power system, including both steady state
the way utilities manage and forecast flexible
and dynamic phenomena, enabling not only better
demand. By creating detailed, real-time simulations
situational awareness for the operators but also
of energy networks, digital twins allow for an
realistic initialization of model-based simulations,
unprecedented level of analysis and insight into
like contingency assessment. This unprecedented
consumer behaviour and system performance.
observability of the power grid combined with
This technology facilitates a deeper understanding
refined load and generation predictions (see
of how demand can fluctuate with changes in both
Section 5.3) enables the design of near real-time
external conditions, such as weather patterns, and
and look-ahead stability-aware applications and,
internal factors, such as consumer usage patterns.
with its decision support capabilities, better equips
The enhanced forecasting capabilities provided by operators for effective decision making.
digital twins enable utilities to predict changes in
With the ongoing trend of reducing power grid
demand with greater accuracy and granularity. This
inertia and the increased sensitivity of power
includes the ability to model how demand might
generation patterns to weather changes, as well
respond to different pricing schemes or to the
as the increased complexity to operate new types

34
Realizing results

of assets, power grid operators must respond 5.6 Align economic signals
more quickly to unexpected disturbances to
Beginning in the late 1990s, wholesale electricity
maintain stable and uninterrupted power supply.
markets began to operate in North America, most
The combination of these factors necessitates
of which employed a technique to align energy
the automation of actions that have been typically
prices with true economic signals from the grid.
executed manually by operators. Such automation
The energy price, called the locational marginal
of processes not only reduces the workload
price (LMP), is based on three components.
of power grid operators but also reduces the
likelihood of human error. The “marginal” component is derived from cost of
electricity for the last unit of power needed to meet
Figure 5-2 illustrates how a DTI provides
demand, known as the marginal unit. This marginal
continuous upstream and downstream information
cost increases as more expensive generation
flow, enabling near real-time accurate mirroring
sources are called on to balance the load and
between the actual power grid and its digital replica
generation of the system. There are two “locational”
in both directions. This significant improvement in
components, one reflecting transmission system
conventional power grid operation results in better
constraints, the other transmission system
safeguarding of power grid reliability.
losses. Transmission constraints force more
expensive generation that is closer to the loads
to be dispatched, raising prices for transmission-
constrained load centres. Transmission losses also
incentivize generation sources that are closer to
the loads because the farther generation is from
the loads, the more losses are incurred.

With markets in North America entering their

third decade of successful LMP-based market
operations, other wholesale markets such as
those in Great Britain are considering adopting the
technique [13]. Many regulators around the world
are also considering how the concept may be
applied to retail electricity markets.

While the first component of the LMP, the marginal

cost, is based primarily on well-understood
operational costs (including fuel costs), the second
and third components, congestion and losses,
are highly dependent on tools that can calculate
the prices based on the grid configuration. This
means that a digital twin is essential for accurately
calculating LMPs. Active research is underway to
develop the equations for emerging distribution
locational marginal price (DLMP) calculations, and
Figure 5-2 | DTI for enhanced grid operation
here too digital twins are essential.
Source: www.tennet.eu

35
Realizing results

Case study 5-1 – Standard flexibility services

Great Britain has taken significant strides towards more granular distribution services with the
introduction of the Flexibility Services. Distribution operators in the region are now procuring standard
services, including:

§ Sustain: a change in generation or load to limited power to below firm capacity

§ Secure: a change in generation or load based on network conditions close to real-time

§ Dynamic: the ability to deliver an agreed change in generation following a network abnormality

§ Restore: following a loss of supply, the ability to either remain off supply, reconnect with lower
demand, or reconnect and supply generation

Because these services are procured in advance, distribution operators must study their systems to
determine the appropriate levels of each service needed at different locations.

36
Section 6
Bridging the virtual world

6.1 Preparing for revolution might best be described as the Imagination Age.
Only in the future will we really understand the
There have been three major industrial revolutions
scope of the current transformation, but topics
in our time. The First Industrial Revolution, running
such as the smart phone, the Internet of Things
from the late 18th century through the early
(IoT), augmented reality (AR) and virtual reality
19th century, marked the change from human
(VR), cloud computing, genetic manipulation,
and animal powered machines to machines
nanotechnology, 3D printing, and AI have become
in factories powered by steam and water. The
defining revolutionary forces reshaping our
Second Industrial Revolution, often called the
economic, environmental, and social systems.
Technological Revolution, began in the late 19th
century and was characterized by the beginning of In the First and Second Industrial Revolutions, a
electrification, the advent of mass production, and change in energy source was the driving factor for
the birth of over-the-wire communication. The Third a societal step change. And so the Imagination
Industrial Revolution – variously called the Digital Age may also be coupled to the transition to
Revolution, the Computer Age, or the Energy renewable energy sources. A power grid supported
Revolution – began in the mid-20th century. The by renewable sources to achieve net-zero carbon
shifts in industry caused by these revolutions have emissions looks very little like the power grid that
included the Bessemer steel-making process, the was introduced in the Second Industrial Revolution.
production line and, relevant to this white paper, Every factor listed as part of the Imagination Age
the introduction of more flexible electrical power to will likely play a role in the power grid of the future.
replace the steam and hydro-powered systems of Many components of the Imagination Age have
the past. connections to the power systems that are
During the Third Industrial Revolution, from the obvious such as IoT to supply vast amounts
mid-20th century through the present day, systems of measurement and control points and AI to
began to rely on computing power to understand analyze grid conditions to predict potential issues
the environment (via sensors) and control the (or resolutions to experienced issues). Other
environment (via controllers). If the industrial technologies may play larger or smaller roles that
revolutions are viewed as using mechanical and only time will reveal. But all this computing power
then electrical power to scale beyond human comes at a price.
measures, then this third age marks computer
power systems that can out-think human brains
and source information more accurately and in
6.2 Powering the virtual world
larger volume than any human could. Computing power is growing at an accelerated
pace. Computing energy usage has nearly
The vast number of technological advancements
doubled since 2015, according to the IEA [14].
introduced in recent years implies that we may
This increase is due largely from the introduction
be entering a Fourth Industrial Revolution, which

37
Bridging the virtual world

of cryptocurrency mining, which now accounts for Agency (IRENA). These solutions include direct
nearly half a percent of total energy demand, as supply, electrification, energy efficiency, green
shown in Figure 6-1. Electricity consumption in hydrogen, and bioenergy combined with carbon
data centres has also grown, with energy use split capture and storage (BECCS). By 2050, electricity
roughly equally between computations and cooling is projected to become the primary energy carrier,
the processors. with its share of total final energy consumption
increasing from 21% in 2018 to more than 50%.
The IEA further reports that their base case
This shift is accompanied by changes in sectoral
scenarios suggest that data centre and
boundaries, particularly the electrification of end-
cryptocurrency processing may nearly double
use applications in heating and transportation.
again by 2026, mainly based on AI processing
The rise in electricity's prominence is primarily
forecasts. The AI component itself should
attributed to the substitution of renewable
experience an approximate ten-fold explosion in
electricity for fossil fuels in various end-use
energy usage [15].
applications. Consequently, the annual growth rate
of renewable technologies is expected to increase
eightfold as this transition takes place [14].

With the virtual world inherently tied to the physical

power grid, it is clear that the simulation models
of the past, centralized and locked behind utility
firewalls, must evolve. For the Imagination Age to
be realized, information must flow from the utility to
society at large, and vice-versa.

6.3 Creating a self-balancing

power grid
Power grids of the past had essentially zero control
over when and where energy was used; balance
was provided by the generation side and enough
generation was built to cover any possible load
profile. This system cannot work in the future for
many reasons, but fundamentally because most
renewable resources are not controllable in the
ways that a fossil-fuel generation system can be.

Small amounts of semi-flexible demand under the

moniker of “demand response” appeared in larger
quantities around the turn of the century. This can
Figure 6-1 | Computing global energy use
be seen as the first real move towards an electrical
system with flexibility not only on the generation
More than 90% of the solutions necessary for side but on the load side as well, with load likely
achieving a successful net-zero outcome by to ultimately have more flexibility than generation.
2050 will involve renewable energy, according to
Power grids with flexible demand necessarily
analysis by the International Renewable Energy
require interactions with the customers – the

38
Bridging the virtual world

customers are the ones with the loads, some or must be able to understand the short-term needs
many of which may ultimately become flexible of the customer and the short-term capabilities
loads. Key to the loads being effective in this role is and limitations of the grid.
their visibility into grid conditions. Loads must have
The smarter that loads can become in using
some idea when and where power will be available.
electricity, the more reliable and less expensive the
Whether the decision is relatively simple, like does
system becomes. The avoided costs are not trivial.
it make sense to heat water in advance of when it
Alternatives include overbuilding the renewable
is needed because there is excess generation, or
generation portfolio, perhaps several times over,
the decision is in a complex scenario, like deciding
so that all energy scenarios can be met with less
how to charge an electric vehicle (or even when to
flexible loads.
discharge the vehicle back into the grid), the loads

Case study 6-1 – Locational load flexibility

Traditionally, energy consumption has been rigid, tied to the time and place of the demand. Demand
response programmes have focused primarily on load shedding (reducing usage) or load shifting
(moving usage to a different time). In the face of evolving electricity markets and renewable energy
integration, some loads are capable of time-shifting and space-shifting, that is, moving the load to a
different location that has cheaper or greener power to consume.

Empirical data and models indicate substantial economic and environmental gains from such
practices. For instance, shifting some data centre processing to different geographical locations
based on renewable availability and cost can optimize energy use and reduce carbon footprints by
10-20%.

39
Bridging the virtual world

6.4 Accessing digital twins a geospatial view of the low-voltage network with
available feeder capacity values. Over time, more
Building a digital twin is no simple task. The
tools will become available for different users with
underlying equipment and connectivity models
different needs, expanding beyond today’s typical
require large sets of data and creation of an
focus of connecting new loads or generation to the
evolved data governance process within the
distribution grid.
utility to manage the data. The management task
includes collecting data, validating accuracy, and Direct data access. Online tools are helpful,
tracking changes over time. The governance but to truly realize the benefits of digital twins,
process for the transactional data that sits atop the underlying data must be exposed. Access is
the models can then be streamlined, with little or currently limited for many reasons, chief among
no human intervention in the collection process. them security of the power system and data
confidentiality established by the data owners. But
In the previous section, the topic of how the
the data currently locked inside could be used for
utility can benefit from digital twins was explored.
a wide range of applications, such as providing
However, there are many more users outside the
optimization services for DER owners, giving
utility who could also benefit. How will these users
investment advice on when and where to site local
access the data? The process will be gradual,
generation, and predicting where to charge electric
with the utility exposing more data over time. This
vehicles at the lowest cost.
evolution can be thought of in two major phases.
Ultimately, much of the data should become
Online tools. The easiest way to expose
accessible to the public to achieve the most
information to end-users is through an on-line
societal benefit. This can be achieved, and grid
tool. Many utilities around the world are already
data simultaneously protected, using various data
providing some of their grid model data to the
manipulation techniques, several examples of
public. These tools generally allow access to a
which are listed in Table 2 [16].
small amount of the data, for example, displaying

Table 2 | Data protection techniques

Technique Description
Delay Defer publication for a period of time
Noise Combining with random data
Anonymization Replace identifiers with random values
Pseudonymization Replace identifiers with non-traceable codes
Translation/Rotation Alter time or spatial orientation
Aggregation Combine details and reduce granularity
Differential privacy Obscure individual information while retaining group patterns
Feature extraction Publish key features as opposed to source details

40
Bridging the virtual world

Figure 6-2 | Representation of a hypothetical global power grid. Source: www.terrawatts.com

6.5 Anticipating a global energy The concept of a global power grid was
interconnection explored in the IEC White Paper, Global energy
interconnection, and outlines the potential benefits:
The core issue with a grid powered by renewable
resources is the lack of control over the generation GEI would represent the ultimate stage in the
output levels. Flexible demand and storage evolution of power grids towards greater levels
technologies are solutions; but load can only be of interconnectivity: a global energy network of
shed or time-shifted in limited amounts and during intercontinental and cross-border backbone
limited times, and storage has its own issues, not networks of high and ultra-high voltage (UHV),
least of which are cost and the competition for use as well as smart power grids (transmission
in electric transport. and distribution networks) in all interconnected
countries at various voltage levels. A GEI could
A third option is to increase the capacity and
connect the power grids of all continents and
coverage of the transmission grid across the
take advantage of the diversity of different
globe, a representation of which is shown in
time zones and seasons, thus supporting a
Figure 6-2. This concept has already been realized
balanced coordination of power supply for all
for information delivery with the Internet covering
interconnected countries.
nearly the entire globe. The telecommunications
grid is, however, managed by thousands of Challenges mirror those of any transmission
different entities in almost two hundred countries. upgrade, including long lead times, major capital

41
Bridging the virtual world

investments, land access rights acquisition,

cooperation among myriad political entities, and
environmental concerns – but at a much larger
scale. However, even if the ultimate goal of a
single global grid is not achieved, more broadly
interconnected “supergrid” transmission systems
would deliver positive results. One example would
be adding and/or reinforcing interconnections
among the current four synchronous areas of
North America, into a single supergrid.

The management of supergrids will require

more coordination among grid operators than is
performed today. Both the long-term planning
of the coupled systems and in the near-term
operation of such highly coupled systems will need
accurate models that digital twins can help deliver.

42
Section 7
Crucial standards

7.1 Smart energy grid architecture series of standards are shown, the common
model information model (CIM) and IEC 61850,
both of which are managed by IEC Technical
The smart energy grid architecture model (SGAM)
Committee 57. The scope of TC 57 ends at the
is a three-dimensional architectural framework that
meter, after which other standards, both IEC and
can be used to model interactions among different
non-IEC, are relevant.
entities within the smart energy arena [17]. The
SGAM is shown in Figure 7-1 [18]. Its two-
dimensional model covers the complete set of 7.2 The common information
electric system “domains”, from power generation model
to power consumption at the customer site along
The CIM is comprised of an open-source unified
the bottom, and information scope “zones” from
modelling language (UML) model and a large
local process control through substations to
collection of data exchange protocol standards
enterprise and market along the right. Two core

Figure 7-1 | Smart energy grid architecture model

43
Crucial standards

managed through the IEC. The exchange protocols overarching goal is to establish a standardized
are grouped into different domains under TC 57: communication framework that enhances
interoperability and streamlines interactions in the
§ Grid: IEC 61970 series
dynamic landscape of DERs.
§ Enterprise: IEC 61968 series
A critically important characteristic of the collection
§ Market: IEC 62325 series of CIM data exchange standards is their use of
§ Customer: IEC 62746 series a common underlying information model, which
provides a framework for organizing electric
The standards layer upon one another to provide
utility data and defining data exchanges for both
a comprehensive framework for data exchange
vendor products and utility implementations. The
profiles for different energy market players. The
design and execution of data sharing solutions
IEC 61970 series is a mature set of standards
can thus be rooted in CIM, benefiting greatly from
focused on the integration of applications and
the abundance of established standards, tools,
information exchange within the control centre
artifacts, and practices. This ensures a shared
and between the control centre and grid systems.
understanding of data across various software
This standard serves as the foundation for the
tools and multiple exchanges.
common grid model exchange standard (CGMES)
profile standard defined by ENTSO-E. In parallel, The foundational premise here is that for new
IEC 62325-351 and IEC 62325-451 provide a technologies and integrated data innovations, such
well-established framework for market-related as AI systems and, importantly for this white paper,
data exchanges in European markets, forming digital twin technologies, to deliver value, they
the basis for the European-style market profile must be based on standardized data. The CIM is
(ESMP) exchanges. IEC 62325-452 addresses the increasingly being recognized as the solution to
need for standards to support data exchanges in this challenge.
North American markets, including the concept of Encouragingly, the core modelling on which the
locational pricing. IEC 61968 series complements specifications of IEC CIM standards are based,
this landscape by offering a common way to particularly in electric grid modelling, is stable.
describe utility operations data exchanges, A well-structured IEC process allows local and
including those related to metering, and asset and regional implementation of CIM-based data
outage management. exchanges to contribute improvement suggestions
These standards do not extend into the realm of into the ongoing dialogue driving the evolution of
controllable resources deployed on the distribution the standards. It is recognized that adaptations
grid, particularly those located “behind” a may be necessary for applications that are more
customer’s electricity meter. Proposed standard dynamic, such as the management of power
IEC 62746-4 is expected to play a pivotal role in usage in a smart home. This acknowledgment
bridging this gap by introducing a set of message showcases a nuanced approach to standardization
profiles tailored to convey grid instructions, that accommodates both stability and adaptability
grid conditions, pricing signals, and resource in response to evolving requirements.
capabilities within the emerging DER space. This
specific standard aims to provide a straightforward
set of data exchanges suitable for various simple
DERs, such as stand-alone battery systems and
collections of demand-response resources. The

44
Crucial standards

Case study 7-1 – Great Britain’s CIM based Long Term Development
Statement [19]

The Long Term Development Statement (LTDS) is a regulatory requirement for distribution utilities
in Great Britain to make grid models publicly available, primarily to assist parties interested in
utilizing the electricity distribution network. This typically takes the form of siting a new load or a new
generation source and leverages the interconnection application process.

The regulator in Great Britain, Ofgem, recently announced that the new “form” of the LTDS will
leverage CIM data structures. This cutting-edge improvement will give developers in this region the
ability to load the LTDS data into any CIM-compliant study platform and analyze the system impacts
of developments they are considering. It is anticipated that this improvement will significantly
streamline the connection process by enabling developers to propose better-sited projects, thereby
reducing the evaluation burden faced by distribution utility staff.

7.3 IEC 61850 integrating diverse technologies within the energy

sector by establishing a fit-for-purpose, universally
IEC 61850 stands as a pivotal international
accepted framework for market players to interact
standard, defining communication protocols for
with the power grid in a deeply digital way. Since
intelligent electronic devices (IEDs) within electrical
the CIM is designed for data exchanges between
substations. Historically focused on communicating
systems, adding IEC 61850 capabilities allows
to equipment in the substation, the IEC 61850
for CIM messages to be delivered to devices,
series of standards has grown substantially in
completing Figure 7-2.
recent years with dedicated extensions for hydro
generation controls and interactions with DERs.
The scope of IEC 61850 includes the following.

Engineering. Manipulate and configure models

and exchanges among tools, including IEDs and
substation configuration tools.

Data modelling. Represent the primary devices

like circuit breakers and transformers and
secondary devices such as protection relays and
control units.

Communications. Support data being used “on

the wire” for monitoring and control objectives,
in the form of either client/server architectures
(historically between the substation gateway and
protection relays/IEDs) or multicast with peer-
to-peer communications among multiple station
devices.

Emphasizing interoperability and interchangeability, Figure 7-2 | Digital twin leveraging

IEC 61850 addresses the practical challenges of communication standards

45
Crucial standards

7.4 Enabling device data exchange and they provide a natural pathway for utilities to
communicate with devices other than the utility
The expanding universe of devices to which
meter. The challenge is getting from the meter to
utilities need to send controls data and from which
those devices, a problem that is often surmounted
utilities need to receive conditions data is enabled
with ubiquitous wireless solutions, including Wi-Fi,
by three basic patterns: one that leverages
Bluetooth, Z-Wave, and Zigbee protocols. There
traditional utilities communications infrastructure,
are disadvantages, however: the closed network
and two that leverage public communications
is not easily opened to the customer, and the
infrastructure. The distinction between these last
technology requires management that not every
two is the presence or absence of an intermediary
utility is willing to embrace.
in the communications chain. The three patterns
are shown in Figure 7-3. Aggregator bridge. As demand response
programmes became more popular in recent
decades, another communication channel was
implemented at scale. As opposed to the through-
the-meter approach, which was highly rooted in
standards, this new aggregator-specific approach
was initially unique to each technology, often
with different protocols for each vendor even in
the same technology space. Over time, multiple
manufacturers voluntarily adopted the OpenADR
communication protocol, allowing utilities to
manage devices from different manufacturers in
the same platform. This trend continues today
with DERs. International standards have lagged
behind industry, with the forthcoming IEC 62746‑4
standard potentially filling this gap. Aggregator-
based approaches have a disadvantage opposite
of the meter-based path: utilities have only limited
visibility of information brokered by the aggregator.
That said, many utilities prefer an aggregated view
of these resources, and this perspective naturally
Figure 7-3 | Edge-device communication supports this approach.
options
Direct connect. As DERs, and in particular DERs
with smart DC/AC inverters, become embedded
Through the meter. With the introduction of in the management of voltage and frequency
automatic meter reading (AMR) and then advanced on the grid during normal operations and play
metering infrastructure (AMI), electric utilities were larger roles in the protection schemes during
able to receive and eventually send basic controls off-normal situations, neither the meter-based
to individual customer sites. These communications path nor the aggregator-based paths are optimal
leverage the IEC 62056 series of standards (ANSI solutions. Rather, a standard protocol that the
C12.18 in North America) implemented on a wide utility, the customer, and potentially other parties,
range of technologies, such as radio frequency like aggregators, could leverage is emerging as
(RF), powerline carrier (PLC), and cellular networks, a solution. The standard, which eventually will be

46
Crucial standards

used to support this option, is not established historical and current behaviour of a physical
globally, and, consequently, utilities are considering object or process. Implementation involves an
a variety of options, including IEC 61850-7-420, encapsulated software object or model reflecting
IEEE 2030.5, Sunspec Modbus, and DNP3. a unique physical entity, relying on real-time,
real-world data measurements across various
From a modelling perspective, the CIM and
dimensions.
IEC 61850 have some overlap. Both represent
power system domain assets for the eventual goal The committee is actively developing a range of
of finding an efficient way to represent, integrate, essential foundational standards for digital twin
and utilize electrical devices and complex systems. technologies that have horizontal application
Harmonization efforts are well underway, which will across the foundational, interoperability and
allow for the CIM to have a “path” to devices via applications domains shown in Figure 7-4.
IEC 61850. Strategic approaches to digital twin standardization
adopted by this committee include a focus on
foundational standards such as vocabularies,
7.5 ISO/IEC JTC 1/SC 41 reference architectures, interoperability, and
ISO/IEC JTC 1/SC 41 represents a joint effort trustworthiness.
between IEC and ISO to provide international One key cross-sectoral standard for digital twin
standardization for IoT and digital twin technologies. technology is ISO/IEC 30173. This standard
IoT integrates a multitude of technologies, ranging establishes terminology for digital twins and
from network to cloud computing and AI, that are explores types of digital twins, digital twin system
inherently network-intensive and data-driven. contexts, digital twin lifecycle processes, functional
The joint technical committee emphasizes the views of digital twins, and digital twin stakeholders.
significance of processing this data through The committee advocates systematically collecting
advanced analytics to unlock value and contribute use cases across all application domains to
to the realization of a “smarter” world. As of document standardization requirements. It also
November 2020, the committee's scope expanded proposes an “incubator” to initiate applications for
to encompass standardization in the field of digital various domains and address potential gaps.
twin, along with its related technologies.
If you have a use case of digital twin applications,
Most notably, the committee has developed the committee is currently inviting submissions.
an international consensus position on the
definition of “digital twin” across IEC and ISO. For
ISO/IEC JTC 1/SC 1, a digital twin is defined as [1]:

A digital representation of a target entity with

data connections that facilitate convergence
between the physical and digital states at a
synchronized rate. It possesses capabilities
such as connection, integration, analysis,
simulation, visualization, optimization, and
collaboration.

This definition broadly mirrors wider industry

sentiment. Gartner and Deloitte [20] describe a
digital twin as an evolving digital profile of the

47
Crucial standards

30181 TR 30194
20924 30173
IoT functional architecture IoT and digital twin best
IoT and digital twin Digital twin concepts and
for resource ID practices for use case
vocabulary terminology
interoperability projects

30188 PNW 444 21823-5

Digital twin reference Extraction and transactions Behavioural and policy
architecture of data products interoperability

PNW 440
30186
Guidance on the connection
Digital twin maturity model
to data spaces

Foundational Interoperability Applications

Figure 7-4 | ISO/IEC JTC 1/SC 41 scope related to digital twins

48
Section 8
Recommendations

To realize the vision of a revolutionized, Standards bodies. Standards bodies include

decarbonized, digitalized, and decentralized energy global standards developing organizations (SDOs),
sector that leverages digital twin technologies, it like the IEC, as well as national committees.
is essential that government agencies, standards
Stakeholders. Stakeholders is the largest
bodies and digital twin stakeholders work
category and includes all the entities that might
collaboratively to gain consensus on standards
operate digital twins, such as electric utilities,
and open data practices.
or contribute data to digital twins, such as
As shown in Figure 8-1, each stakeholder group independent power producers, or utilize data,
has unique responsibilities in enabling digital including end consumers and researchers and
twin technologies in the energy sector. The academics. It is important to note that this large
recommendations that this white paper makes category is anticipated to grow over time as more
are addressed to distinct stakeholder groups, parties become interested in the electric power
specifically: system.

Government agencies. Government agencies

include national and regional bodies that set policy,
such as ministries, regulators, and commissions.

Figure 8-1 | Recommendations overview

49
Recommendations

8.1 Gain consensus on standards of core standards for describing foundational

information models, like the physical grid
One of the themes of this white paper is that for
equipment and connectivity models, to ensure that
digital twins to become useful tools, many different
the transactional data that sits “on top” is well-
entities must provide and, critically, also manage
organized and meaningful.
their comprising datasets. This requires adoption

Gain consensus on standards

Recommendations
§ Work with utilities to define digital twin requirements
Government agencies

§ Work with national committees to identify high priority standards to enable power system
digital twins and requirements identified by utilities
§ Mandate, where appropriate, the use of the CIM for machine-to-machine information
exchanges and IEC 61850 to push data to and pull data from grid-edge devices and work
with manufacturers to adopt and implement standards
§ Help stakeholders advance standards to cover requirements and work proactively with
national committees to facilitate agile standards development processes
§ IEC Members to adopt the CIM and IEC 61850 as national standards to provide the foundational
pillars for a digitalized energy sector
§ Provide mechanisms for interested parties to learn about the importance of digital twin
standards in the energy sector to drive their greater use in industry digital twin pilots and
prototypes
§ IEC Standardization Management Board (SMB) to carry out a review of existing digital twin
“horizontal” standards and their applicability to the energy sector
Standards bodies

§ IEC SMB and ISO Technical Management Board (TMB) to request relevant application
technical committee to develop use cases to support the ongoing work within
ISO/IEC JTC 1/SC 41
§ Publicize the critical need for participation and recruit stakeholders from across the data
value chain, i.e. data users alongside data producers; and establish liaison mechanisms with
external technical groups heavily involved in energy data
§ Provide governance for utilizing energy data standards, either in the form of advisory groups,
external engagement activities, or partnerships with relevant government departments and
regulators to mitigate risks of duplication in key energy data standards enabling digital twins
§ When overlapping domains appear, coordinate with relevant SDOs to resolve different
perspectives, as has been achieved with existing IEC/ISO collaborations
§ Get involved with national committees to support the standards development process by
feeding insights on requirements and use cases
Stakeholders

§ Provide financial support and resource commitment for employees to participate in relevant
standards development activities
§ For university stakeholders, teach and sponsor research on digital twins and the foundational
data management concepts that underpin them

50
Recommendations

8.2 Adopt open data practices and digital twin users will need access to as much
of the data in the digital twin as is feasible. These
Open data for power system digital twins is
exchanges must be governed by firm policies that
basically a bilateral activity: utilities will need users’
respect valid privacy and security concerns.
data to build and manage accurate digital twins;

Adopt open data practices

Recommendations
§ Improve awareness of regulatory frameworks, especially around privacy and security
Government agencies

constraints, so that standards can support them

§ Fund initiatives to support the development of common data spaces: platforms that enable
data curation, management, and integration across domains

§ Engage with stakeholders to distil necessary requirements for open data practices and
Standards bodies

translate these requirements into the standards portfolio; these could be done through
technical committee consultation or external industry workshops

§ Get involved as soon as possible to ensure standards support individual data requirements
§ Supply use cases and issues applying existing standards and communicate these needs to
Stakeholders

national committees to enable existing standards to be revised or new standards developed

to meet the requirements of the energy sector
§ Include grid reliability and security in the open data practices through participation from grid
operators in the requirements setting activities

51
Recommendations

8.3 Build and operate digital twins Utilities will need to take the lead in building the
initial models behind digital twins. But there are
Building on the previous recommendations, which
many players who must contribute to this process
set the stage for success, these recommendations
to ensure digital twins deliver real value.
focus on the implementation of digital twins.

Build and operate digital twins

Recommendations
§ Fund digital twin initiatives that leverage consensus-based standards and open data practices
Government agencies

§ Reduce regulatory hurdles and support simplification of the standards landscape by removing
duplicative/conflicting standards
§ Fund pilots for power system digital twins; embedding high priority standards and open data
practices as necessary pre-requisites

§ Develop digital twin toolkits outlining key enabling standards and their applicability to better
connect data producers and data users with the technical content
Standards bodies

§ Provide education to digital twin developers and potential users on standards development
and the useability of available standards. National committees should consider drawing from
their technical expert pool in the development of these educational assets
§ Actively support digital twin interoperability by hosting interoperability events (IOPs) and
providing test data and software tools

§ Digital twin implementers should gather stakeholder input when developing implementation
roadmaps and learn about the standards, making intentional choices to apply those standards
§ Share experiences, both positive and negative, with government bodies and national
Stakeholders

committees
§ Participate actively in national committees’ engagement efforts, whether through the technical
committees or external workshops

52
Recommendations

8.4 Interact with digital twins a step change in the energy sectors’ lagging
adoption of digital twin technologies. Energy
Finally, digital twins must be leveraged. These
sector stakeholders must share more information
recommendations focus on the “end state” as
with one another but also with standards makers
digital twins become a part of normal operations in
to develop the essential information base to cost-
the electric power grid sector.
effectively implement renewable technology and
The world’s energy system’s rapid transition sustainable practices with confidence in their
towards a more flexible, less centralized and efficiency, reliability and security.
renewables dominated model is a welcome
Comprehensive, well-informed, and standards-
pathway toward decarbonization. However, without
based digital twins are crucial for maximizing
the full utilization of digital twin technologies, we
renewables-based digital technologies, products,
face significant challenges balancing these goals
and services, facilitating whole-system integration,
with those of high security and reasonable cost to
and enabling energy sector stakeholders to make
consumers.
informed decisions that will revolutionize the energy
The magnitude, complexity, and urgency of sector. This approach is imperative for achieving
contemporary energy sector challenges – both a future that is decarbonized, decentralized, and
technical and socio-economic – necessitate deeply digital.

Interact with digital twins

Recommendations
§ Educate the stakeholders on the potential benefits to be gained from using digital twins
through use cases and narratives
Government
agencies

§ Monitor digital twins’ implementation processes and publicize successes

§ Educate the stakeholders on the role of standards to support digital twins

Standards bodies

§ National committees, engaging with government agencies, should publish implementation

guides and offer training opportunities

§ Leverage digital twins to optimize real-world operations

Stakeholders

§ Utilize best practices from other sectors to aid its development in the energy sector

§ Leverage the power system digital twins with digital twins from other segments

53
Bibliography
[1] ISO/IEC 30173:2023, Digital twin – Concepts and terminology. Available for purchase: https://
webstore.iec.ch/publication/68081 [Accessed: 29 August 2024]

[2] Intergovernmental Panel on Climate Change, Sixth Assessment Report (IPCC AR6), Working Group
III, Annex 1, 2022 [Online]. Available: https://www.ipcc.ch/assessment-report/ar6/ [Accessed: 29
August 2024] (Definition of net-zero greenhouse gas emissions, modified)

[3] United Nations Climate Action, For a livable climate: Net-zero commitments must be
backed by credible action [Online]. Available: https://www.un.org/en/climatechange/net-zero-
coalition#:~:text=Through%20this%20UN%2Dbacked%20campaign,Off%20Air [Accessed: 29 August
2024]

[4] International Energy Agency (IEA), An updated roadmap to Net Zero Emissions by 2050 [Online].
Available: https://www.iea.org/reports/world-energy-outlook-2022/an-updated-roadmap-to-net-zero-
emissions-by-2050 [Accessed: 29 August 2024]

[5] International Renewable Energy Agency (IRENA), World Energy Transitions Outlook, 2021
[Online]. Available: https://www.irena.org/publications/2021/Jun/World-Energy-Transitions-Outlook/
digitalreport [Accessed: 29 August 2024]

[6] Intergovernmental Panel on Climate Change (IPCC), Sixth Assessment Report (IPCC AR6), Working
Group I, Climate Change 2021: The Physical Science Basis [Online]. Available: https://www.ipcc.ch/
report/sixth-assessment-report-working-group-i/ [Accessed: 29 August 2024]

[7] GRIEVES, M., VICKERS, J., Digital Twin: Mitigating Unpredictable, Undesirable Emergent Behavior
in Complex Systems. In: Kahlen, J., Flumerfelt, S., Alves, A. (eds) Transdisciplinary Perspectives on
Complex Systems. Springer, Cham., 2017 [Online]. Available for purchase: https://link.springer.com/
chapter/10.1007/978-3-319-38756-7_4 [Accessed: 29 August 2024]

[8] TINNEY, W. F., WALKER, J.W., Direct Solutions of Sparse Network Equations by Optimally Ordered
Triangular Factorization. Proceedings of the IEEE, Vol. 55, No. 11, 1967, pp. 1801-1809 [Online].
Available: https://www.engineering.iastate.edu/~JDM/ee553/Tinney1.pdf [Accessed: 29 August
2024]

[9] KOOHI-FAYEGH, S., ROSEN, M.A., A review of energy storage types, applications and recent
developments. Journal of Energy Storage, Volume 27, 2020, 101047, ISSN 2352-152X [Online].
Available: https://www.sciencedirect.com/science/article/abs/pii/S2352152X19306012?via%3Dihub
[Accessed: 29 August 2024]

[10] Overview of IDDP Platform Research and Its Future Plan, KEPRI, 2024 [Online]. Available: https://
www.kepri.re.kr:20808/idpp/Overview%20of%20IDPP%20Platform(www.kepri.re.kr).pdf

[11] IEC TR 62541-1:2020, OPC Unified Architecture – Part 1: Overview and concepts. Available for
purchase: https://webstore.iec.ch/publication/61109 [Accessed: 29 August 2024]

54
Bibliography

[12] The Blackout Report, Biggest Blackouts in History: South Australia 2016, September 28, 2021
[Online]. Available: https://www.theblackoutreport.co.uk/2021/09/28/south-australia-blackout-2016/
[Accessed: 29 August 2024]

[13] Office of Gas and Electricity Markets (Ofgem) Great Britain, Assessment of Locational Wholesale
Pricing for GB, 2023 [Online]. Available: https://www.ofgem.gov.uk/sites/default/files/2023-10/
Ofgem%20Report%20-%20Assessment%20of%20Locational%20Pricing%20in%20GB%20
%28final%29.pdf [Accessed: 29 August 2024]

[14] International Energy agency (IEA), Tracking data centres and data transmission networks, July
2023 [Online]. Available: https://www.iea.org/energy-system/buildings/data-centres-and-data-
transmission-networks#tracking [Accessed: 29 August 2024]

[15] Data Center Frontier, IEA Study Sees AI, Cryptocurrency Doubling Data Center Energy Consumption
by 2026, 2024 [Online]. Available: https://www.datacenterfrontier.com/energy/article/33038469/
iea-study-sees-ai-cryptocurrency-doubling-data-center-energy-consumption-by-2026 [Accessed:
29 August 2024]

[16] National Grid ESO (Great Britain), Virtual Energy System [Online]. Available: https://www.
nationalgrideso.com/future-energy/virtual-energy-system [Accessed: 29 August 2024]

[17] IEC SyC Smart Energy, IEC SRD 63200: SGAM basics [Online]. Available: https://syc-se.iec.ch/
deliveries/sgam-basics/ [Accessed: 29 August 2024]

[18] IEC SyC Smart Energy, CEN-CENELEC-ETSI Smart Grid Coordination Group – Smart Grid Reference
Architecture, 2012 [Online]. Available: https://syc-se.iec.ch/wp-content/uploads/2019/10/Reference_
Architecture_final.pdf [Accessed: 29 August 2024]

[19] Office of Gas and Electricity Markets (Ofgem) Great Britain, Response to Formal Consultation on
the Form of Long Term Development Statement, 2024 [Online]. Available: https://www.ofgem.gov.
uk/sites/default/files/2024-04/LTDS%20Direction%20Letter%20300424.pdf [Accessed: 29 August
2024]

[20] Gartner Information Technology Glossary: Digital Twin [Online]. Available: https://www.gartner.com/
en/information-technology/glossary/digital-twin [Accessed: 29 August 2024]

55
Notes

56
Notes
 ISBN 978-2-8322-9877-0

International
Electrotechnical
Commission
IEC WP Virtualizing power systems:2024-10(en)

CHF 50.-

3 rue de Varembé T +41 22 919 0211

PO Box 131 [email protected]
CH-1211 Geneva 20 www.iec.ch
Switzerland

® Registered trademark of the International Electrotechnical Commission. Copyright © IEC, Geneva, Switzerland 2024