1

Fast Frequency Response from Energy Storage

Systems – A Review of Grid Standards, Projects
and Technical Issues
Lexuan Meng, Jawwad Zafar, Shafi K. Khadem, Alan Collinson,

Kyle C. Murchie, Federico Coffele, Graeme Burt

1 OLTC On-Load-Tap-Changers
Abstract—Electric power systems foresee challenges in stability PFR Primary Frequency Response
due to the high penetration of power electronics interfaced PPM Power Park Module
renewable energy sources. The value of energy storage systems RegA Regulation A of PJM
(ESS) to provide fast frequency response has been more and more RegD Regulation D of PJM
recognized. Although the development of energy storage ROCOF Rate-Of-Change-Of-Frequency
technologies has made ESSs technically feasible to be integrated in
SIR Synchronous Inertia Response
larger scale with required performance, the policies, grid codes
and economic issues are still presenting barriers for wider SNS Smarter Network Storage
application and investment. Recent years, a few regions and SOF System Operability Framework
countries have designed new services to meet the upcoming grid SOP System Operation Practice
challenges. A number of grid-scale ESS projects are also SoC State-of-Charge
implemented aiming to trial performance, demonstrate values, STATCOM Static Compensator
and gain experience. This paper makes a review on the above UKPN UK Power Networks
mentioned aspects, including the emerging frequency regulation
services, updated grid codes and grid-scale ESS projects. Some key II. INTRODUCTION
technical issues are also discussed and prospects are outlined.

Index Terms--frequency response, energy storage, grid code. E XCESSIVE carbon emissions and fossil fuels based energy
resources are the main concerns for energy system
development since 90s of last century. Power sector is certainly
I. NOMENCLATURE the major consumer of such resources. Expectations come to the
ACE Area Control Error mass utilization of renewable energy resources (RES) to
ACS Area Correction Signal substitute conventional generation.
AGC Automatic Generation Control However, the fluctuating and intermittent characteristics of
BA Balancing Authority most RES cause critical issues in electric power system (EPS)
DG Distributed Generator for power balancing, stability and system level regulation.
DNO Distribution Network Operator Moreover, the electrification of other energy sectors, e.g.
EFR Enhanced Frequency Response transport and heating, brings potential of congestion in EPS in
ENTSO-E European Network of Transmission System near future. Enhancing the flexibility of the power sector can
Operators for Electricity significantly reduce overall cost and improve system
EPS Electric Power System performance, thus providing possibility to accommodate more
ESS Energy Storage System RES and consumers. Major flexibility solutions include [1]–[3]:
FERC Federal Energy Regulatory Commission i) flexible generation to ensure back-up capacity, ii) greater
FFR Fast Frequency Response interconnection to systems beyond the region, iii) enabling
FFR-AUS Fast Frequency Response of Australia demand to respond more to short-term price signals, and iv)
FFR-IR Fast Frequency Response of Ireland
increased electrical energy storage systems (ESS).
GIR Grid Interconnection Requirement
From grid stability point of view, frequency dynamics and
HVDC High Voltage Direct Current
stability are the key measures which indicate the strength of the
IEEE Institute of Electrical and Electronics Engineers
MMC Modular Multi-level Converter grid as well as the balance condition between generation and
NERC North America Electric Reliability Corporation demand. Grid frequency control is facing key challenges under
NC-RfG Network Code on Requirements for Grid high penetration of non-synchronous generation [4]. Although
Connection of Generators few large international jurisdictions are experiencing high rate-

Lexuan Meng, Jawwad Zafar, Federico Coffele and Graeme Burt are with Power Networks Demonstration Center, University of Strathclyde, Glasgow, UK
(Emails: [email protected], [email protected], [email protected], and [email protected]).
Shafi K. Khadem is with International Energy Research Centre, Tyndall National Institute, Ireland (Email: [email protected]).
Alan Collinson is with Scottish Power Energy Networks, UK (Email: [email protected]).
Kyle C. Murchie is with Scottish and Southern Electricity Networks, UK. (Email: [email protected])
 2

of-change-of-frequency (ROCOF) issues, Ireland and UK are a. Deadband

exceptions due to their relative smaller size and limited The deadband is an operating region centred around nominal
interconnection with continental grids. Both Ireland and UK frequency where the plant controller will not adjust its power in
have emerging concerns for high ROCOF (>0.5 Hz/s) [5]. The response to frequency deviations, as shown in Fig. 1. This
System Operability Framework (SOF) in UK assessed the deadband is a natural feature in conventional generators due to
evolving system inertia trend which indicated an increasing risk mechanical imperfections. Nowadays, the improvement in
of high ROCOF especially under the “Gone Green” future control accuracy and usage of power converters (especially for
energy scenario [6]. These challenges require the grid to equip nonsynchronous generators) can largely reduce the deadband,
with new capabilities, such as ROCOF mitigation, inertia while intentional deadband is still widely used to relieve system
enhancement, over-frequency generation shedding scheme, from continuous adjustment and to reduce wear&tear.
frequency response to large disturbances, and so on. From the grid point of view, a narrower deadband and a
Driven by the challenges mentioned above, grid-scale ESS smaller droop gain can increase the contribution from
is a promising solution finding applications in domestic, generation units and improve the power grid frequency stability
industrial and commercial circumstances [7]–[9]. Although [12], [13]. However, on the other hand, they can cause undue
there still are economic, regulatory and technical barriers wear&tear of rotating machines and faster degradation of
preventing the wide adoption of grid-scale ESS, there has been electro-chemical storage systems. Droop settings of 3-5% and
active development and grid code adaptation appear for ESS deadband no more than ±17-36 mHz are widely used as a good
application in recent years [7], [9]–[11]. New frequency compromise, as shown in Table I [14]. The exceptions are:
regulation services are emerging aiming to take full utilization • UK, Ireland and New Zealand use a narrower deadband
of the ESS advantages. The major task of this paper is to review due to their relative smaller system sizes and low inertia
the existing grid connection requirements applicable to ESSs, features;
as well as the emerging frequency response services demanding • Ontario, Finland, Brazil, New Zealand and Singapore use
fast responses, with a special focus on transmission level a wider range for droop and/or deadband since in these regions
applications. the frequency response is procured by market and a wider range
gives the generator more flexibility to adjust their participation.
III. FREQUENCY RESPONSE AND DESIGN CONSIDERATIONS
A nominal frequency is set in AC electric power systems, i.e.
60Hz in North America and 50Hz in Europe and China. The
frequency has to be maintained within a limited range by
keeping the balance between consumption and generation at all
times. The grid balancing services can be carried out in three
categories according to time scales: primary frequency response
(PFR), secondary frequency response and tertiary frequency
response. Recently, due to the concern of decreasing inertia, a
number of new frequency services are designed which are
featured with fast responding requirement, namely fast
Fig. 1. Droop based frequency response
frequency response (FFR). They are considered as one of the
solutions to handle low system inertia and high ROCOF
challenges. It is worth mentioning that FFR is used in this paper
as a general term for frequency response services with fast
responding requirement. It covers new services e.g. Enhanced
Frequency Response (EFR) of UK, Fast Frequency Response
of Ireland (FFR-IR), FFR of Australia (FFR-AUS) and
Dynamic Regulation Signal (RegD) of PJM, although they
differ in names and types.
A. The Key Parameters in Frequency Response Services
A typical droop-based frequency response setting is shown Fig. 2. Response time settings
in Fig. 1, which essentially requires adaptive power with regard
to frequency deviation. It is also a natural response from a b. Droop Coefficient
conventional synchronous generator. Deadband, droop The droop coefficient characterizes the generator’s power
coefficient, response speed and duration time are the major profile according to frequency deviations. It is defined by [14]:
parameters in frequency response services. A summary and ∆𝑓𝑝.𝑢.
𝐷𝑟𝑜𝑜𝑝(%) = 100 ∙ ( ) (1)
comparison of those parameters in different regions are given ∆𝑃𝑝.𝑢.
in Table I and II for conventional PFR and emerging FFR
services, respectively. In addition, the response time setting is
shown in Fig. 2 and introduced in the following part.
 3

TABLE I COMPARISON OF INTERNATIONAL CASES – CONVENTIONAL SERVICES

Deadband Response Time Minimum Procurement
Region Document Droop
(±mHz) Requirement Duration (Mandatory/ Market)
36 5%
Global IEEE 1547-2018 [77] 5s (0.2-10s) N/A N/A
(17-1000 mHz) (3-5%)
United FERC Order 842
<=36 <=5% No undue delay Sustained Mandatory
States NERC BAL-003-1 [79]
<=34 (steam/hydro)
Texas, US Nodal Operating Guides [87] 4-5% 14-16 30 sec Mandatory
<=17 (all other)
New
Operating Procedure No.14
England, <=36 4-5% No undue delay Sustained Mandatory
[88]
US
PJM, US M-14D [90] <=36 <=5% No undue delay Sustained Mandatory
Ontario,
Market Rules IESO [89] <=36 2-7% <1 s (activation time) 10 sec Market
Canada
Europe ENTSO-E RfG (GIR) [11] 0-500 2-12% 2-30s N/A N/A
UK Grid Code [83] <=15 3-5% 10 s 30 sec Market/Mandatory
Ireland The Grid Code [78] <=15 3-5% 2 -10 s 30 sec Mandatory
Denmark TR 3.3.1 for Battery Plants [84] <=10 2-12% 15 s N/A Market
Finland VJV2013 [91] 0-100 2-12% 5s for half & 30s for full 2 min Market
15 s for half & 30s for
Italy TERNA Italian Grid Code [93] 10 or 20 2-5% 15 min Mandatory
full
9 s reach 90% of full
Brazil Procedimentos de Rede [92] <=40 2-8% 1 min Market
response
New Electricity Industry
0 (no deadband) 0-7% 6s 60 sec Market
Zealand Participation Code [94]
Singapore Transmission Code [95] 50 3-5% 9s 10min Market
The Network Code [96] <=150 0-10% 10 s 10min Contracted
South
Renewable Power Plant Agreed with System Agree with System
Africa 0-10% 10 s 10min
Connection Code [97] Operator Operator

China Q/GDW 696-2011 [86] 200 N/A ≤ 0.2 s ≥120 s Mandatory

TABLE II COMPARISON OF INTERNATIONAL CASES – NEW SERVICES

Deadband Response
Region Service Type Droop Power Requirement Duration
(±mHz) Speed (s)
Min. power-step 1MW
FFR – Dynamic 15-200
Post-fault Max. power-step 5MW
Ireland N/A 2 8-10 s
contingency Min. power-step 1MW
FFR – Static 200-700
Max. power-step 75MW
Post-fault
FFR1 50-150
contingency
Australia N/A N/A 0.5-1 6s
Post-fault
FFR2 N/A
emergency
Pre-fault 15 (high inertia) Min. power 1MW
UK EFR Envelope 1 15 min
continuous 50 (low inertia) Max. power 50MW
Frequency
PJM, US RegD N/A N/A Min. 0.1MW 2 Sustained
recovering

where ∆fp.u. is the measured change of system frequency as a is especially critical for the power grids with relatively smaller
percentage with respect to nominal frequency, and ∆Pp.u. is the capacity and higher penetration of non-synchronous generators
corresponding active power change with respect to the such as Ireland, UK and New Zealand. It can be clearly noticed
generator’s power rating. that in Table I these three grids require shorter response time
The droop coefficient range is commonly designed by TSO. (full response delivery in 2~10s compare to 30s in Italy and
A consistent droop across all generators ensures that they Finland).
contribute with respect to their capacities to the PFR. In a The response speed of a frequency response is majorly
market environment, the droop coefficient can be adjusted defined by the time delay (Tdelay) and ramp-up rate (Kp), as
(within a specified range) by the generator owner/operator shown in Fig.2. The time delay includes measurement time,
based on the willing to provide more or less reserves to the communication delay and device activation time. The effect of
service. these parameters on the usefulness of the service has been
studied in [15]. It reveals that a high ramp-up rate is of great
c. Response Speed
importance for the usefulness of the service, consequently
A fast responding service can effectively reduce the power converter interfaced energy storage systems are highly
frequency deviation and ROCOF during contingency event, suitable providers for FFR. In addition, it is also concluded that
thus enhancing the usefulness of the service [13]. This feature
 4

a time delay within 1s can be acceptable for ROCOF lower than ROCOF.
0.25Hz/s as long as the ramp up rate and total FFR capacity Thus, solutions such as converter interfaced ESSs can be
(shown as PT in Fig. 2) are sufficient. beneficial to the grid stability, safety and reliability, by
Based on the above understanding, the grid inertia and the providing FFR type services. A number of energy storage
necessity of fast responding services are explained in detailed technologies are listed in Table III that are potential candidates
in the following part. for providing such services, thanks to their considerable
power/energy size and fast response time [5], [18].
B. The Necessity of FFR in a Low Inertia Grid
System inertia is defined as the resistance to changes in the
system frequency by storing/injecting kinetic energy from/to
the system during power imbalance, which comes from the
synchronously connected rotating machine.
The kinetic energy of a system (Ek) is calculated as [16], [17]:
𝑛
1 2
𝐸𝑘 = ∑ ( 𝐽𝑖 𝜔𝑚𝑖 )
2
𝑖=1
where J and ωmi are the moment of inertia (kg.m2) and the
angular speed of the rotor of the i-th rotating machine
respectively, and n is the total number of machines. The system
total inertia constant is the ratio of total stored kinetic energy in Fig. 3 Comparison of frequency performance with /without FFR service.
MJ, at synchronous speed to the MVA rating of the base system:
𝑛 TABLE III. EXAMPLES OF ES TECH SUITABLE FOR FAST RESPONSE SERVICES
2
𝐸𝑘 𝜔𝑠𝑚
𝐻𝑠𝑦𝑠 = = ∑(𝐽𝑖 ) Power Energy
𝑆𝑏𝑎𝑠𝑒 2𝑆𝑏𝑎𝑠𝑒 Response Discharge
𝑖=1 Technology Range Ratings
Time Time
where Sbase is the MVA base of the system and ωsm is the (MW) (MWh)
synchronous speed of the system (rad/s). It can be seen that the Flywheel
0.1-400 Up to 5 <= 4 ms Up to 15 min
(converter)
total system inertia depends on the number of connected Lithium up to 50 or minutes to
rotating machines as well as the kinetic energy stored in their Up to 20 Few to 40 ms
Batteries more hours
rotating mass. In comparison, the instantaneous physical Flow Batteries
~3, 50
Up to 60 Few to 40 ms
seconds to
storage of a power converter (no matter what is connected at the possible hours
Advanced
source side) is the energy stored in its DC-side capacitor, which Lead-Acid up to 40 Up to 40 Few to 40 ms
seconds to
hours
is negligible compared to the rotational inertia of synchronous Batteries
machines. Super-
milliseconds to
capacitors up to 50 Up to 0.28 10-20 ms
The initial ROCOF that is determined by the size of the (high voltage)
1 hour
power imbalance and the system inertia can be expressed as:
𝑓0 ∆𝑃 𝑓0 ∆𝑃
𝑅𝑂𝐶𝑂𝐹 = = IV. CHARACTERISTICS OF FFR SERVICES IN SELECTED AREAS
2𝐻𝑠𝑦𝑠 𝑆𝑏𝑎𝑠𝑒 2𝐸𝑘
The typical examples of FFR services that are pondered
where ΔP is the disturbance size and f0 is the nominal system
here: RegD signal from PJM (US) [19], EFR from National
frequency. Obviously, the initial ROCOF is larger if the power
Grid (UK) [20] and FFR-IR from EirGrid (IRE) [21]. The
disturbance is bigger or the system stored energy is smaller. As
design of these services also considers the feasibility and cost
more conventional generators are substituted by converter
for ESSs, such as energy neutral design from PJM, US [19] and
interfaced generations, the Hsys and Ek become smaller.
envelope design from National Grid, UK [20]. This section
Moreover, the increasing power consumption and the stochastic
performs a review on these selected services.
feature of renewable generation cause more frequent and higher
power disturbances. The combination of the two factors can A. Enhanced Frequency Response, National Grid, UK
result in larger ROCOF and thus worse the grid stability. National Grid, UK has recently procured resources (ESS) for
Conventionally, the frequency deviation caused by power a new PFR service, namely EFR [20]. EFR differs from
variation and contingency events is compensated by PFR conventional PFR in its fast response speed. The characteristic
majorly from rotational generators. However, their inherent of EFR are summarized as follows:
deadband and slow response usually result in a full response  Service delay includes the time to detect a frequency
power delivery in around 10-20s which is too slow for a low deviation plus the time for response instructing and the
inertia system. With the help of a fast responding energy source, time for ESS to deliver output change. The total delay has
such as a converter interfaced ESS, power can be delivered to to be no greater than 1s, with the time delay for detection
effectively mitigate the frequency nadir and ROCOF [15]. An and instructing response no greater than 500ms.
example is given in Fig. 3, which simulates a frequency loss  The service delivery envelops (wide and narrow band
event for a big generator in a low inertia network and compares services) are shown in Fig. 4. The ESS output must always
the frequency behavior with and without an FFR. It can be seen be within the upper and lower envelopes.
the frequency nadir is significantly reduced as well as the  The deadband is defined as the frequency range between
 5

reference point C and D where the ESS gives zero MW B. Fast Frequency Response, EirGrid/SONI, Ireland
output. This area can be used by the ESS to manage its EirGrid/SONI, Ireland is developing a multi-year program-
state-of-charge for later services. Delivering a Secure, Sustainable Electricity System (DS3) [4]
 The power limit and droop coefficient can be calculated [21], which defines two additional fast frequency services: FFR
by the reference points given in two tables in Fig. 4. and SIR. These services aim to enforce grid security and
stability.
SIR is the active power output and synchronizing torque that
a unit can provide following disturbances. Synchronous
machines based generators, condensers and loads are naturally
available for providing this service, and ESSs, with proper
measurement and control, can also be a significant player. An
example SIR is shown in Fig. 5 (a).
FFR is defined as the additional increase in MW output from
a generator or reduction in demand following a frequency event
that is available within 2 seconds of the start of the event and is
sustained for at least 8 seconds. The extra energy provided in 2
to 10 second timeframe by the increase in MW output must be
Fig. 4 Enhance frequency response envelope [18].
greater than the any loss of energy in the 10 to 20 second
 The capacity of high and low response is symmetrical.
timeframe due to a reduction in MW output below the initial
 ESS connection should meet grid code minimum
MW output (i.e. the hatched blue area must be greater than the
requirements which also give guidelines on voltage and
hatched green area in Fig.5 (b)).
reactive power control.
 During EFR provision, it is not allowed to provide any C. RegD Signal, PJM, US
other services that could impair the EFR service quality. PJM defines two types of frequency regulation services:
But it is allowed to participate in the provision of other RegA and RegD. RegA refers to traditional generators which
services outside the EFR service period. have limited ramp rates; RegD is designed for fast ramping
In addition, the design principles are also briefly resources, such as batteries.
summarized here: The control scheme of RegD is shown in Fig. 6. The filtered
 Frequencies outside 50±0.25Hz are considered post-fault, area control error (ACE) is first sent to a PID controller to
and the service provider must follow the exact profile generate an ACE correction signal (ACS) as a regulation
defined in Fig. 4; frequencies that are within 50±0.25Hz reference. The ACS is then separated into two signals by a low-
are considered pre-fault, for which an envelope is pass filter and a high-pass filter respectively. The high-pass
designed so that the service provider can have room to filtered signal, is filtered again by a low-pass filter to remove
manage its operating condition. noise and RegD regulation signal is generated accordingly. The
 According to [22], the maximum delivery is specified to low-pass filtered signal is the RegA regulation signal.
be 50±0.5Hz, which set a 0.5Hz frequency deviation limit. In addition, a new Conditional Neutrality Controller is
Increasing the figure will make the service less useful designed and implemented since 2017, as the feedback loop
while lowering the figure will ask service providers to shown in Fig. 6 [19]. RegA signal for ramp-limited resources,
contribute more energy (enhance the cost and make the is also used to balance energy to zero neutral when available to
service hard to deliver). do so in order to manage the State-of-Charge (SoC) of RegD
 The output power of the provider can vary within ±9% of service providers.
capacity allowing some flexibility and lowering the cost. In the feedback loop, the cumulative sum of the RegD signal,
However, increase this figure will limit the service namely the total absorbed or generated energy by RegD
provision capability. resources, is calculated, and a simulated SoC (the percentage of
 Frequency insensitivity band has two settings: 50±0.05Hz used energy in total RegD capacity) is generated and used to
and 50±0.015Hz. When grid inertia is high, the ±0.015Hz determine the payback gain. Three payback gains can be set
band is used. ESSs use this deadband in high inertia according to the simulated SoC:
conditions so that they can provide useful services when 1. Simulated SoC between -25%~25%, the payback gain is
the frequency is relatively stable. ±0.05Hz is used when set to a value to reset the SoC of RegD resources back to neutral
the grid is in low inertia and frequency is less stable within 15 min;
experiencing high rate of change. 2. Similarly, for SoC between 25%~50% or -50%~-25%, this
 The ESS must be able to deliver at 100% its EFR capacity has to be done within 7 min;
for 15 min minimum. Lowering this figure, the service 3. For SoC between 50%~100% or -100%~-50%, it is 3 min.
delivery time may not be sufficient to allow the various The selected payback gain is then added into the total
reserve services to react; on the other hand, increasing this regulation reference to adjust the output of the RegA resources.
time is not necessary and will result in higher cost. The more the RegD resources are charged or discharged, the
 6

(a) (b)
Fig. 5 DS3 new services: (a) Synchronous Inertia Response; (b) Fast Frequency Response.

Fig. 6 Conditional Neutral Controller, PJM.

more efforts the RegA resources will make to bring them back frequency, or a big spike in frequency) [24]. Thus, the FFR in
to neutral. EirGrid and NEM do not use a droop control, they are static
Furthermore, a priority switch is also used in case that if the frequency response triggered at a certain frequency level. The
ACE is very large, the controller places extra priority on ACE EFR service is more similar to the PJM fast regulation service
control by switching the payback loop off and using the ACE model, although in the PJM case a central signal (AGC)
signal instead. instructs the battery on what to do (rather than requiring a
response based upon locally measured frequency, as is the case
D. Considerations for ESS
for EFR in Great Britain) [24], [25].
For ESSs, it is commonly recognized that they can face more
challenges if they asked to follow the same requirements as V. CRITICAL REVIEW OF TECHNICAL INVESTIGATIONS
other generators. As a result, special settings have been applied
This section gives a high-level review on the technical
in the emerging services mentioned above to ease the ESS
investigations for ESS applications in FFR service. For real
integration. For example, the FERC order 842 [23] has stated
world applications there is still lack of understanding on several
that in order to alleviate the potential of over-charge/discharge
essential issues, e.g.:
and excessive wear&tear, a SoC (state-of-charge) range shall be
 Which is the better storage solution for FFR in terms of
identified before the connection to provide timely and sustained
cost and performance?
PFR. Furthermore, it is also widely agreed that the ESS can
 How much total FFR capacity is needed to maintain the
have more contribution to the grid stability if featured services
grid stability?
or requirements are pre-designed, resulting in the emerging
 How to optimally position the FFR resources?
services as FFR-IR and EFR in order to take advantage of the
 Which is the better way of controlling and coordinating
accurate and fast response of converter interfaced ESS.
FFR resources?
Notably, EFR is as a continuous control service which is
 What are the challenges of using ESS for multiple grid
provided via a droop response to frequency. The aim of the
services?
service is to manage pre-fault system frequency (ie. to maintain
Hence, this section is divided into three parts: i) ESS planning
system frequency closer to 50Hz under normal operation), but
which includes cost evaluation, sizing and placement methods;
not designed to arrest frequency decline post-fault [24], [25].
ii) ESS control, and (iii) challenges of multi-service provision.
This makes it fundamentally different to the FFR services
specified in EirGrid, Ireland and NEM, Australia, which are A. ESS Planning
targeted at post fault controls (ie. managing a big drop in  Cost Evaluation for FFR
 7

The high cost of storage systems has been a major obstacle it comes to the techno-economic benefits, very few are found.
for wider application in power grids, although their benefits are In [36], a co-optimization approach is investigated to solve the
obvious. It is of great importance to select the cost-efficient storage investment problem such as finding the type, placement
solution for a certain application. For FFR use purpose, it can and sizing of ESS in a transmission-constrained network.
be seen from Table II that, the FFR-IR and FFR-AUS are power Regarding the ESS planning, in [37], authors prove that the
critical services which require high power capability within minimum operating cost is a decreasing convex function of the
short duration time (6~10s). On the other hand, the EFR and ESS energy capacity. This leads to the optimal sizing of ESS
RegD are energy critical services asking for a service duration that strikes a balance between the capital investment and
of 15min or more. operating cost.
The studies in [26] and [27] have tried to evaluate the cost  Placement of FFR
and suitability of storage solutions in different applications. The The state-of-the-art review reveals different applications of
two studies have common recognition that flywheel is the most energy storage such as virtual inertia response and inter-area
cost-effective solution for high power and short duration oscillation damping provision [38]–[40], optimal power flow
primary frequency service due to the low power capital cost, [41], placement in the network and control tuning. It is
long lifetime and mature technology. Super-capacitor is recognized that apart from the total inertia response (IR)
comparable with flywheel for FFR purpose although lifetime is required in the system, its placement is also very important.
the major limiting factor. In addition, it is worth noting that the Both small-signal system linearization based [38], [39] and
no-load power loss is also an importance factor since the typical non-linear system, simulation based [40] techniques have been
FFR services are designed for contingency event which has low used in the process of location optimization. The computation
occurrence rate. No-load losses are omitted in both studies [26] effort required to achieve the result of the optimisation exercise
and [27], and super-capacitor typically has much lower no-load is also addressed in [38], [39]. The constraints, within which the
losses compared to flywheel. For EFR and RegD, as they also optimisation is carried out, includes the characteristics
require considerable amount of energy, Li-ion batteries are associated with the battery such as its power limit and SOC, and
naturally preferred from cost and efficiency point of view [28]. the line limits of the power system which also impacts the
In summary, in order to have a complete evaluation and placement [38]–[41]. Finally, it is suggested that in order to
comparison of potential solutions, it is necessary to take into maximise the value from the multiple service provision
consideration the following aspects: capability of energy storage devices, tools are required that take
- Capital cost of major system components, i.e. storage into account the different application time horizons from
systems, converter, control systems, mechanical seconds to minutes for simulation durations of a year, which
systems, cooling and environmental control, etc. will inform the overall best location for the energy storage
- Footprint of the solution, indoor, outdoor and special device [42].
construction requirements, etc
B. ESS Control
- Operational and maintenance costs
- Power losses and equivalent cost The operation of the ESSs participating in frequency
- Use mode and technical specification response services has to follow the technical specifications
 Sizing of FFR given by respective grid standards. In general, the control
Sizing of ESS is one of the key issues to achieve the techno- methods for fast responding frequency services can be
economic sustainability of ESS for grid services. It is to be categorized into four types: P-f (active power-frequency) droop,
noted that compare to inertia response services, the FFR/PFR triggered static response, AGC signal, and emulated inertia
service requires much higher energy ratings. Though this field control:
is promising and relatively new in operation, very few  P-f Droop Control
researches have been done on the development of method for P-f droop control (or dynamic frequency response as defined
estimating/optimizing the size of ESS for IR and FFR services. in National Grid, UK) is used in most conventional and
A critically brief review is done here as well, to understand emerging frequency response services as can be seen in Table I
the progress of the ESS sizing for these services. In most of the and II [13], [43]–[45]. The active power output of the generator
cases, sizing of ESS is done arbitrarily [29], or a number of is in proportion to the frequency deviation. It is an inherent
simulation for various ESS sizes is executed and the optimal feature of conventional synchronous generators guaranteeing
size is chosen according to desired results [30]. A probabilistic automatic frequency recovering and power sharing. Nowadays
approach with high number of simulations is also applied when P-f droop characteristic (with a deadband) is also applied to
the power network is highly penetrated from renewable energy converter interfaced energy resources as a requirement for grid
such as solar [31]. The impact of sizing upon the overall interconnection.
frequency dynamics is further extended in [32] where a method  Static Frequency Response
for estimating the ESS sizing is presented targeting the National Grid, UK and EirGrid/SONI, Ireland also define
power/frequency characteristics of the required services. static frequency response services (see Table I and Table II),
Sizing of ESS was optimized in few cases and mostly done which are triggered at a certain frequency level instead of using
to maximize the financial benefits from the grid ancillary a proportional droop [21], [24], [43], [46]. A constant power or
service market [33], [34] or in microgrid application [35]. When discrete step of power is generated according to the frequency
 8

level. Although the large majority of frequency response voltage stiff feature are the major challenges of applying this
requirements in UK are met by dynamic providers, static method.
providers can also be of use especially during a large loss of Synchronverter control [56] implements a swing equation
generation. The static service is provided by fast acting based inertia emulation control and a reactive power-voltage
resources or demand side cut-off. droop to generate the angle and voltage amplitude references
On one hand, the fast and full power response given by static respectively. These two references are fed into modulation
providers can be more effective to save the post-fault frequency directly without inner voltage or current control loops. However,
drop from too high rate-of-change and too low nadir. In addition, the major risk is the difficulty of integrating voltage and current
as demonstrated in the Smarter Network Storage (SNS) project, saturation due to absence of inner voltage and current loops.
the higher aggregated active power contribution by static Power synchronization control (PSC) [57] combines the
response can potentially gain higher income compared to the P- swing equation based inertia emulation control with vector
f droop based dynamic services [47]. On the other hand, it has current control and realizes stable operation in weak grid
also been concluded that the static response service is less condition by supporting both inertia and stiff voltage feature.
beneficial to the pre-fault frequency stabilization compared The current limitation is also guaranteed by the current control
with the continuous dynamic response service. It is more loop.
suitable for contingency frequency response, such as the FFR In summary, inertia emulation controls show promising
services in Ireland and Australia. performance for low inertia power grids especially during the
 Automatic Generation Control transient of frequency disturbances. The major challenges are
PJM, US uses a centralized automatic generation control still the emulation of a stiff voltage source feature and the
signal to perform frequency regulation [19]. Although this proper integration of current limitation during transient.
frequency service is indeed a secondary frequency regulation
C. Challenges for Multi-Service Provision from ESS
which is not comparable with the primary frequency response,
the authors decide to involve it since a dynamic regulation The maximized utilization of ESS by providing multiple
(RegD) is designed which also eases the service provision from services could improve the business cases of grid-scale ESS,
energy storage systems and it is also a fast responding service and encourage more investment and applications [47]. Thus,
which can take full advantage of ESS [19]. The RegD signal is some of the ESS operational scheduling problems considering
generated according to the area control error and sent to service multiple services are formulated in [58], [59]. However, the
providers. The regulation speed of RegD is comparable with the challenges exist in several aspects:
emerging FFR and EFR services and provides more accurate • Complicated control
regulation of the frequency level. The conditional neutral The concurrent and multiple service provision requires the
control design also provides proper coordination of the service ESS converter to be able to operate in different modes e.g.
providers. However, the communication delay, communication automatic voltage control mode for voltage support, automatic
failure and cyber-attack can jeopardize the safety of the utility frequency response mode for frequency service. Such services
grid. should be coupled to match the best energy and power ratings
 Inertia Emulation Control of the batteries. At this stage, good coordinated control is very
This type of control regulates the interfacing converter to important as it requires the battery to make charging and
emulate the mechanical inertia, e.g. mimic the behavior of a discharging decisions at different time scales while accounting
synchronous machine. The main objectives are to provide for the stochastic information such as load demand, electricity
virtual inertia, integrate droop control, enable smooth islanding prices, and regulation signals. Discussion on different control
transition, and provide voltage support. It can be applied in strategies has been left to analyze in another research paper.
converter interfaced energy resources, e.g. wind turbine, PV Analytical approach shows that solving the problem for even
and ESS [48]–[54]. The active power output is in proportion to a single-day operation would increase the computational
the rate-of-change-of-frequency by, in most cases, complexity due to the large state space and the number of time
implementing a swing equation, thus can effectively mitigate steps. In that case, dynamic programming approach could
ROCOF. The following part gives a brief review of typical reduce this complexity [60]. The other problem is the multiple
inertial emulation control methods. service provision has huge impact on the storage life time
P-df/dt based droop control [49], [52] regulates the power in degradation [61]. Part of these issues can be overcome by
proportion to the ROCOF so as to provide virtual inertia to the introducing hybrid ESS. Integration of electrochemical
system. However, this type of control cannot avoid PLL in the batteries and super-capacitor with a coordination of current and
control loop thus can be unstable in case of grid fault or weak voltage control loop in the converter could be a good solution
grid conditions. for this, which has been validated in 50kW test-bench [60].
VIrtual Synchronous MAchine (VISMA) control [55] Implementing ESS for multiple service provision in
integrates a high order synchronous machine model in transmission and High Voltage DC (HVDC) networks could
converter control scheme and generates current reference for also be achieved by introducing HVDC-scale modular
the cascaded current controller. It is a natural and multilevel converter (MMC) [60]. In general, the combination
straightforward implementation of machine model. However, of such issues largely complicates the control, operation of ESS,
the complexity due to the high order model and the lack of increases the development cost considerably and hence require
 9

extensive research and grid-scale demonstration as well. thereby improving life span.
• Coordination with existing grid facilities In addition, it has also been indicated in the project report
The existing grid facilities, such as On-Load Tap Changers that the concurrent service provision feature can potentially
(OLTC), distributed generators (DG) with grid forming maximize the utilization and business value, however, the
converters, synchronous condensers, and so on, may have combination of services, if not well coordinated, may cause
unexpected interactions with ESS operating in different modes. voltage step magnitude exceedance. Further study has to be
The concurrent service provision certainly puts more load on carried out in this regard.
the grid side asking for enhanced operation strategies. In such
B. Flywheel-Battery Hybrid Grid Storage, Ireland
case, harmonious integration of fast acting ESS with
coordinated control of SoC and a proposed unified frequency Under the background of DS3 as well as the renewable
controller [62], and co-located ESS and DGs with local voltage penetration target, a number of new grid scale energy storage
controller [63] could maximize the active power production systems have been trialed in Ireland. The example illustrated in
during the high penetration of DGs when voltage limits are this section is a 480 kVA rated flywheel/battery hybrid energy
violated. storage solution developed by Schwungrad Energie Limited. A
demonstration project was carried out to show how the hybrid
VI. GRID-SCALE ENERGY STORAGE SYSTEM PROJECTS solution responded to real frequency events over a period of 9
months. The goal was to provide system services (without
This section introduces three real-world projects: 1) example participating in the market) up to 20 min with full power
from UK shows the effectiveness of Li-ion (the most popular achieved in the timeframe of 500 ms (measured from the time
and matured ESS technology) for frequency response service; of system frequency falling through 49.80Hz). The metrics used
2) the hybrid (Flywheel and Pb-acid battery) solution in Ireland to classify the quality of performance were the response time
demonstrates the possible responses for FFR; 3) PJM and sustainability of power output. For a commercial
demonstrated the value of another hybrid (Ultra-capacitor and installation the target would be to deliver the following DS3
Pb-acid battery) solution UltraBattery® for RegD service along products: FFR, SIR, as well as primary, secondary and tertiary
with some demand side management strategy. operating reserves [4].
A. Smarter Network Storage, UK Four different control modes were tested as shown in Fig. 7.
In 2012, UKPN designed the SNS project to explore the The deadband of the frequency response is assumed 50mHz,
technical and commercial feasibility as well as regulatory within which range the storage can manipulate its SoC freely.
barriers to multiple applications of large-scale battery storage Frequency response is triggered at the threshold of 49.8Hz
system [47]. The storage site was commissioned in December requiring that the storage systems reach their full power output
2014, comprising a 6MW, 7.5MVA, 10MWh Li-ion battery. Its within 500ms and sustained for 5s. The performance of the four
primary applications were peak shaving, frequency response, control methods is compared:
electricity reserve and tolling. Concurrent provision of 1) Static response by ROCOF predicted triggering (Mode
frequency response and reactive power support were also I). In this control mode, when the ROCOF exceeds the
tested. threshold a frequency event is predicted. As a result, the control
Firm frequency response is one of the main tests in this system generates a “full blast” command before the frequency
project including both dynamic and static frequency response. reaches the frequency threshold value of 49.8Hz. It can be seen
Firm frequency response requires: i) delivery of full service that both the battery and the flywheel reach their full response
within 10s; delivery of service continuously for 30min; service power before the frequency reaches 49.8Hz and sustained for
type can be dynamic or static. Dynamic firm frequency 5s. The advantage of this control mode is that due to the
response (DFFR) demands active power output from ESS prediction of the event, the response is delivered ultra-fast.
following a droop curve with a 0.05 Hz deadband (50 mHz for However, the issues related with this control mode include: i)
wide band, 15 mHz for narrow band) and a ±0.3 Hz frequency accuracy of prediction is questionable, may cause false
sensitive mode range. Static firm frequency response (SFFR) response, ii) arrested of response before the frequency returning
contracts ESS with a fixed output power level and activates the back into deadband can reduce the usefulness of the service. In
service when frequency exceeds the upper or lower boundary addition, the hybrid system use-mode is not optimized. The
[47]. Besides, the multi-service provision capability of ESS was ideal use of hybrid battery with flywheel is to use flywheel as a
also tested [47], in which active and reactive power can be fast and high power responding device and use battery as a low
delivered concurrently. Both above applications have power and sustained responding resource to maximize the
demonstrated the effectiveness of SNS in delivering required frequency support.
services. Reactive power support service from ESS may reduce 2) Static feedback response by ROCOF predicted
distribution network losses by optimizing the reactive power triggering (Mode II). Similar to Mode I, the frequency event is
flow, while the power capacity of the connecting converter
(reactive power capability) decides the impact level of such
service provision. Voltage control service, majorly managed by
controlling reactive power in inductive grids, can improve
voltage stability and reduce the stress of substation’s OLTC
 10

Fig. 7 Flywheel-Battery Hybrid Grid Storage performance in FFR.

predicted by ROCOF measurement and the frequency response flywheel can be achieved to maximize the grid support, reduce
is triggered before reaching the threshold. The service is operational cost, and improve owner profit.
sustained for 5s, and different from Mode I, the storage systems 2) ROCOF prediction based triggering method needs to be
gradually reduces its power based on a droop gain instead of a further evaluated before real world application, since the
sudden arrest of service. The advantage of this mode is the frequency behavior is becoming more dynamic.
smooth reduction of power according to frequency condition 3) 500ms response time is much shorter than the FFR
which can potentially benefit the grid stability. However, the requirement of 2s, and in the case of Mode I and Mode II the
accuracy of the prediction is still uncertain and more energy response time is even shorter. False triggering could incur.
needs to be delivered in order to have the smooth reduction of 4) The four control modes can be compared from the
power. usefulness and economy point of view.
3) Frequency triggered static response (Mode III). In this
C. East Penn Smart Grid Demonstration Project, US
mode, the initial ROCOF of the event is not high enough to
predict the event, accordingly the frequency response is This smart grid project of PJM demonstrates distributed
triggered near the threshold of 49.8Hz. Both flywheel and energy storage for grid support, in particular the economic and
battery reach their full response after the triggering event. The technical viability of a grid-scale, advanced energy storage
response time is recorded as within 500ms according to the system using UltraBattery® technology with a total regulation
report. The response is sustained until the frequency reach capability of 3MW for frequency regulation ancillary services
49.9Hz. The event is triggered again at around 20s. Compared and demand management services [64]. This demonstration
with Mode I and Mode II, the time to reach full response after system follows PJM’s RegD signal. Over the course of the
the threshold is longer but still within the 500ms requirement. demonstration, the system has been operating at various power
In addition, taking a solid threshold as the triggering event can levels and durations (typically 2~2.4MW for maximum one
be considered more reliable. hour duration).
4) Dynamic droop response (Mode IV). The dynamic The testing results are shown in Fig. 8 [64]. The response
droop mode emulates the behavior of a conventional generator time of the system to a 3MW command was determined to be
which continuously deliver power according to frequency less than 4 seconds, or 0.75MW/s. The perfect following of the
deviation. In summary, the trial results have demonstrated the regulation signal and fast response result in high performance
capability of battery and flywheel storage systems to deliver score and increased payment compensation. The averaged
fast frequency response service. Several control modes have signal following accuracy is 94.2% during the operating time.
been tested which provide very fine reference for future The SoC is maintained within 45~55% validating the energy
implementation of such devices. The operation of a hybrid neutral control of RegD as was introduced in Section IV.C.
energy storage system is also validated showing satisfying In addition, a comparison is also given in Fig. 9 showing the
results. Nevertheless, a number of key issues are still unclear, accuracy difference between different technologies in
including: participating PJM regulation service [65]. The results clearly
1) The best use-mode of a hybrid energy storage system is indicate the advantage of using battery as ESS can provide more
not explored. A better coordination between battery and efficient and useful service to the grid.
 11

However, the design of the services need further

improvement considering, for example, the following aspects:
 There is still lack of understanding of grid behavior with
high penetration of power electronic devices, especially
their influence over grid stability, responses to fault
conditions, the stronger coupling between frequency and
voltage, and so on. A good understanding of the above
questions is the key to design a useful service.
 The frequency dynamics is becoming local, as a result the
definition, measurement and detection of frequency event
need to be extended and refined.
 Frequency services can be further classified to enable
application of versatile technology solutions and expand
the service market. An example is the EFR in UK, which
require both fast response (1s) and long duration (15min).
It limits the application of potentially promising solutions
like super-capacitors which are energy critical.
Fig. 8 System’s response to 3MW regulation signal.
B. The Role and Definition of ESS
To expand the application of grid-scale ESSs, improvement
is still required in the following aspects:
 Clear definition and differentiation of ESS with other
generation facilities, relieve over levy and regulation of
ESS [66]. Treat ESS as a subset of flexibility assets, for
the grid, rather than a subset of generation assets due to
the diverse applications of ESS.
 Re-value ESS from system operators when considering
system expansion and planning, update system planning
tools with regard to ESS benefits and grid impact.
 By recognizing the benefits to grid stability, safety and
system operational cost reduction, new market incentives
and increased ESS service payment are to be implemented
for fast cost recovery.
 Continuous efforts on grid-scale ESS trials and projects to
develop a mature service market, clarify the business
Fig. 9 System’s SoC range for regulation profile
model and identify key challenges.
Recent research also throw emphasis on the design and
VII. KEY FINDINGS AND PROSPECTS
control issues of ESS with multi-sources [48], [67], [68],
A. Frequency Regulation and Response techno-economic benefits [69][81] for these new services.
It has been commonly recognized that with the increasing C. ROCOF and Inertia Response
penetration of power electronic converters and renewable
The new trend of grid services includes inertia response
energy sources, system stability has become a critical issue
aiming to mitigate the high ROCOF caused by increasing
requiring advanced grid services. Fast frequency responses and
penetration of non-synchronous generators and low grid inertia.
inertia response are the two major solutions with regard to
To mitigate high ROCOF, a faster response is needed.
enhancing grid stability and reliability. The emerging new
However, the ROCOF measurement and detection has an
services include EFR from UK, FFR from Ireland and
inherent trade-off between measurement length and
Australia, RegD signal from PJM, US. These services have
measurement accuracy. A longer period of measurement
been designed with consideration for wider grid-scale ESS
window (2-100 cycles, 50-500ms) can reduce the false
adoption, e.g.:
detection of frequency event, while the faster response
 The EFR has specialized envelope design in order to
requirement asks for shorter response time.
reduce ESS operational cost and enable ESS self-
In addition, false triggering of frequency response can also
management (SoC balancing).
be caused by short-circuit fault or power quality issues, since
 The FERC Order 755 and 784 have opened the ancillary
distorted voltage waveform affects the accuracy of frequency
market to ESS by involving additional performance score
measurement. Further investigation and coordination of various
and compensation.
grid services is required in order to deliver a reliable service to
 The conditional neutral design of RegD signal helps the
the grid.
overall energy balance within a certain period.
From the control perspective, VSM type of controls have
 12

promising features for ROCOF mitigation and grid Response from EirGrid Ireland and Dynamic Regulation Signal
stabilization. However, there are still challenges on the voltage- from PJM US are introduced in detail as example projects. The
stiff control and transient current limitation. From the planning major findings are summarized and prospects are discussed.
perspective, the cost evaluation, sizing and positioning of ESSs
need further investigations. IX. APPENDIX
D. New Solutions Typically, there exist various requirements related to ESS
grid connection and frequency service provision, commonly
Besides energy storage, other inverter interfaced energy
contained within an international standard or national/regional
resources are also potential providers for frequency and inertia
grid code, where generators must comply with—both prior to
response. Thus, on one hand, the grid codes and regulations are
initial connection, and while operational, namely grid
trying to design new services to accommodate mass energy
interconnection requirements (GIR) and system operation
storage applications, while on the other hand, these services
practice (SOP), respectively [14]. The entities responsible for
have to be technology neutral in order to minimize the cost and
creating and enforcing these standards can vary across regions,
encourage a versatile market.
e.g. Institute of Electrical and Electronics Engineers (IEEE),
Static Compensator (STATCOM) with power intensive
North America Electric Reliability Corporation (NERC) and
super-capacitors is considered one of the promising devices to
European Network of Transmission System Operators for
provide inertia for frequency and voltage support e.g. Siemens
Electricity (ENTSO-E), etc. They usually provide a system
SVC PLUS ES [70]. ABB along with SP Energy Networks is
wide minimum standard applicable to all regions. In addition to
implementing a 4-year project Phoenix to demonstrate a
that, the regional TSOs can also have dedicated grid codes
sustainable design, deployment and operational control of a
which meet the minimum standards while take into account
synchronous condenser with a static compensator. The
regional features. The objective of this section is to provide an
objective is to mitigate the issues UK National Grid is facing,
overview and comparison of all the grid standards applicable to
including reduced inertia, low short-circuit power and limited
ESS interconnection and fast frequency service provision.
voltage control. The synchronous condenser, thanks to its
rotating mass and overloading capability, is able to boost A. IEEE 1547-2018 [75]
system inertia (although quite limited compared to converter IEEE 1547 establishes criteria and requirements for
interfaced storage) and enhance system short-circuit power interconnection of distributed resources with EPS. It involves
level; the STATCOM, on the other hand, reacts fast in case of requirements on distributed resources performance, operation,
voltage and power quality issues. testing, safety and maintenance. These requirements are
Furthermore, the design, sizing and positioning philosophy applicable to all technologies including fuel cells, photovoltaics,
of FFR or SIR type of devices need to be further investigated. dispersed generation and energy storage.
Apart from power and energy requirements, other criteria The previous version of IEEE 1547 (IEEE 1547-2003) had
should also be considered e.g. network inertia, network power limited content related to voltage regulation, voltage and
level, network topology, maximum ROCOF, frequency frequency responses to area EPS abnormal conditions. Also had
dynamical feature, number of measurable event (pre-/post- no clause on frequency response or active power support, and
fault) per day/week/month, and so on. only the response to abnormal frequency conditions is stated. In
From the review of energy storage devices, it is also found comparison, IEEE 1547-2018 has clearly defined the
that devices are capable to response within few milliseconds but frequency-droop (frequency response) specifications as given
do not have high power density which is required for fast in Table I (row 1).
frequency response. Thus, the hybrid ESS with a combination
of high power and high energy density devices, operating in B. ENTSO-E Network Code, Europe
active control mode could be very good solution for the new The ENTSO-E Network Code for grid connection
grid services [71], [72]. Power sharing methods in hybrid ESS requirements (NC-RfG) [11] aims to set out clear and objective
play an important role in enhancing the battery lifetime as well requirements for Power Generating Modules, including both
[73]. The solutions can also be operated in multiple service Synchronous Power Generating Modules and Power Park
mode [74]. Modules (a unit or ensemble of units non-synchronously
connected or connected through power electronics to a system),
VIII. CONCLUDING REMARKS for grid connection in order to contribute to non-discrimination,
This paper provides an overview of the connection effective competition and the efficient functioning of the
requirements, design considerations, service characteristics and internal electricity market and to ensure system security.
real-world implementation of grid-scale ESS for frequency NC-RfG classifies the type and significance of generators
response provision. Although a number of barriers still exist, according to their maximum capacity. The detailed categories
system operators in many countries are making effort to design are not repeated here but can be found in [11]. Although there
new frequency services, market rule adjustment and project is no classification for ESS, it can be considered as a Power
trials in order to encourage business investment and Park Module following the general requirements classified
applications. Three representative examples, Enhanced according to power size, such the ones defined in Irish Grid
Frequency Response from National Grid UK, Fast Frequency Code [76].
NC-RfG also defines a set of frequency regulation
 13

requirements. A frequency sensitive mode is set, and within the Frequency Sensitive Mode-Overfrequency (LFSM-O) and
range of which generators must respond to frequency deviation Limited Frequency Sensitive Mode-Underfrequency (LFSM-
according to a preset droop profile. The frequency response U). For battery plants smaller than 1.5 MW, they are obliged to
shall be activated as fast as possible. provide LFSM-O, while for those above 1.5 MW, they must
The full active power frequency response requires 1.5-10% provide both LFSM-O and LFSM-U. The details are collected
ramp, 2-sec reaction and full service provision within 30-sec. in Table I [82].
The detailed requirements from ENTSO-E are also collected in
F. State Grid Corporation of China, China
Table I (row 7).
The State Grid Corporation of China has defined dedicated
C. NERC and FERC, United States Technical Guideline (Q/GDW 564-2010) [83] and Operation &
The most critical standard regarding PFR in US is NERC Control Specification (Q/GDW 696-2011) [84] for grid
BAL-003-1 [77]. This standard requires that each Balancing connection of ESS. Q/GDW 564-2010 requires that 1) the ESS
Authority (BA) annually satisfies a minimum Frequency should respond to frequency event according to specified
Response Measure in order to ensure that it is capable of frequency ranges, and 2) the ESS is able to control its reactive
avoiding Under Frequency Load Shedding for a specified loss power (power factor between 0.98 lead to 0.98 lag) following
of generation. The specific details regarding how BAs meet this set-points from operators in order to regulate grid voltage.
target is left to the BAs themselves rather than directed by Q/GDW 696-2011 defines the operational requirements of
NERC. NERC has, however, outlined recommended speed ESS in providing frequency response in normal operation range
governor settings for each of the three U.S. interconnections (grid frequency between 49.5~50.2 Hz):
that support compliance with BAL-003-1.  Following the preset operating curve approved by system
The Order 755 [78] is proposed by the commission to operator with control error less than 10%
remedy undue discrimination in the procurement of frequency  Active power response time within 200ms
regulation in the organized wholesale electricity markets and  Provide frequency response such that: i) 49.5~49.8Hz,
ensure that frequency regulation service providers receive just ESS discharges with response time less than 200ms; ii)
and reasonable and not unduly discriminatory or preferential frequency higher than 50.2Hz, ESS charges with response
rates. This was extended by Federal Energy Regulatory time less than 200ms; iii) full power continuous operation
Commission (FERC) Order 784 [79] on 18 July 2013, which no less than 2 minutes.
required each public utility transmission provider to take into Q/GDW 696-2011 also specifies that when active power
account the speed and accuracy of regulation resources in its (frequency response) and reactive power (voltage response) are
determination of reserve requirements for regulation and required at the same time, the frequency response has higher
frequency response service. Furthermore, the FERC Order 842 priority.
requires the all the PFR participants have to deliver timely
G. EirGrid, Ireland and SONI, Northern Ireland
response and sustained until the frequency is back into their
deadband range. No specific number is given for either the The Grid Codes of Ireland [76] has been modified to include
response time or duration, but a minimum 12-month rolling PPMs. Battery storage units are treated as PPM units. They
average initial PFR performance and sustained PFR must provide all capabilities as required in Grid Code. Battery
performance of 0.75 is required [80]. generation will be considered like any other generator, and
Conventional services for some of the US regions are given battery demand will be treated like other demand customers and
in Table I (row 2-5). therefore additional definitions are required for this PPM
technology type. This type of storage is similar to Pumped
D. National Grid, United Kingdom Storage Plants where there are two modes of operation. So the
The National Grid in UK is a member of ENTSO-E and terminology for Energy Storage devices has been kept
mandates that all grid connected generators subject to the Grid consistent with that for pumped storage plants.
Code, including asynchronously connected resources, are The generators connection and operating capabilities are
capable of providing PFR. Connection requirements [81] are also specified in the Grid Codes of Ireland, requiring generators
given in Table II (row 7). A market mechanism is adopted based to provide support in different conditions and frequency ranges.
on which the generators (or demand response resources in Firm Dispatch strategies are designed according to the size at the
Frequency Response) submit mandatory bids for capacity point of connection [76].
reservation. Recently, a new Enhanced Frequency Response Besides, the main TSOs in Ireland (EIRGRID and SONI)
(EFR) service is procured by National Grid requiring service have created a DS3 program [4] as mentioned before. A number
tenders provide a modulated frequency response activated of new services are proposed, among which the Synchronous
within 1s [20]. EFR is designed with an ultrafast response in Inertia Response (SIR) and FFR are closely related with
order to reduce ROCOF and maintain grid stability. frequency regulation and ESS could play a significant role. SIR
asks for immediate provision of active power following
E. ENERGINET, Denmark
frequency disturbances in order to limit the ROCOF. In addition
The TSO of Denmark, ENERGINET, has defined the to SIR, the FFR service requires a fast (within 2s) and
operational rules for grid connected battery plants in 2017 [82]. continuous supply of power (at least 8s) after frequency
Two types of frequency responses are used namely Limited deviation. Important parameters for this new service are also
 14

given in Table II. Details about these two services are given in [29] G. Delille, B. Francois, and G. Malarange, “Dynamic Frequency Control
Support by Energy Storage to Reduce the Impact of Wind and Solar
Section IV. Generation on Isolated Power System’s Inertia,” IEEE Trans. Sustain.
Along with these, the important values for some other active Energy, vol. 3, no. 4, pp. 931–939, Oct. 2012.
regions and countries such as Texas (US) [85], New England [30] V. Knap, R. Sinha, M. Swierczynski, D.-I. Stroe, and S. Chaudhary,
“Grid inertial response with Lithium-ion battery energy storage
(US) [86], Ontario (Canada) [87], PJM (US) [88], Finland [89],
systems,” in 2014 IEEE 23rd International Symposium on Industrial
Brazil [90], Italy [91], New Zealand [92], Singapore [93] and Electronics (ISIE), 2014, pp. 1817–1822.
South Africa [94][95] are also given in Table I. [31] M. Yue and X. Wang, “Grid Inertial Response-Based Probabilistic
Determination of Energy Storage System Capacity Under High Solar
Penetration,” IEEE Trans. Sustain. Energy, vol. 6, no. 3, pp. 1039–1049,
REFERENCES Jul. 2015.
[1] U. National Grid, “Future Energy Scenarios,” 2017. [32] V. Knap, S. K. Chaudhary, D.-I. Stroe, M. Swierczynski, B.-I. Craciun,
[2] S. Bradbury, “STORAGE BUSINESS MODELS IN THE GB and R. Teodorescu, “Sizing of an Energy Storage System for Grid
MARKET,” 2014. Inertial Response and Primary Frequency Reserve,” IEEE Trans. Power
[3] M. Bell et al., “Power sector scenarios for the fifth carbon budget Other Syst., vol. 31, no. 5, pp. 3447–3456, Sep. 2016.
members of the Secretariat who contributed to this report,” 2015. [33] A. Oudalov, D. Chartouni, and C. Ohler, “Optimizing a Battery Energy
[4] Eirgrid and SONI, “DS3: Frequency Control Workstream 2015,” 2015. Storage System for Primary Frequency Control,” IEEE Trans. Power
[5] D. Lew et al., “Technology Capabilities for Fast Frequency Response,” Syst., vol. 22, no. 3, pp. 1259–1266, Aug. 2007.
2017. [34] M. Kintner-Meyer et al., “National Assessment of Energy Storage for
[6] “Enhanced Frequency Control Capability (EFCC) National Grid Battery Grid Balancing and Arbitrage: Phase 1, WECC,” 2012.
Storage Investigation Report-November 2015.” [35] M. R. Aghamohammadi and H. Abdolahinia, “A new approach for
[7] The Electricity Storage Network, “DEVELOPMENT OF optimal sizing of battery energy storage system for primary frequency
ELECTRICITY STORAGE IN THE NATIONAL INTEREST,” 2014. control of islanded Microgrid,” Int. J. Electr. Power Energy Syst., vol.
[8] Renewable Energy Association, “Energy Storage in the UK - An 54, pp. 325–333, Jan. 2014.
Overview,” 2015. [36] S. Wogrin and D. F. Gayme, “Optimizing Storage Siting, Sizing, and
[9] Bird&Bird, “The role of energy storage in the UK electricity system,” Technology Portfolios in Transmission-Constrained Networks,” IEEE
2014. Trans. Power Syst., vol. 30, no. 6, pp. 3304–3313, Nov. 2015.
[10] Workstream 6 of the Smart Grid Forum, “The customer-focused smart [37] Y. J. A. Zhang, C. Zhao, W. Tang, and S. H. Low, “Profit-Maximizing
grid: Next steps for regulatory policy and commercial issues in GB Planning and Control of Battery Energy Storage Systems for Primary
Report of Workstream Six of the Smart Grid Forum, 2015,” 2015. Frequency Control,” IEEE Trans. Smart Grid, vol. 9, no. 2, pp. 712–723,
[11] “COMMISSION REGULATION (EU) 2016/ 631 - of 14 April Mar. 2018.
2016 - establishing a network code on requirements for grid [38] B. K. Poolla, D. Gross, and F. Dorfler, “Placement and Implementation
connection of generators,” 2016. of Grid-Forming and Grid-Following Virtual Inertia and Fast Frequency
[12] G. Kou, P. Markham, S. Hadley, T. King, and Y. Liu, “Impact of Response,” IEEE Trans. Power Syst., pp. 1–1, 2019.
Governor Deadband on Frequency Response of the U.S. Eastern [39] B. K. Poolla, D. Groß, T. Borsche, S. Bolognani, and F. Dörfler,
Interconnection,” IEEE Trans. Smart Grid, vol. 7, no. 3, pp. 1368–1377, “Virtual Inertia Placement in Electric Power Grids,” Springer, New
May 2016. York, NY, 2018, pp. 281–305.
[13] J. Undrill, “Primary Frequency Response and Control of Power System [40] Y. Zhu, C. Liu, K. Sun, D. Shi, and Z. Wang, “Optimization of Battery
Frequency,” 2018. Energy Storage to Improve Power System Oscillation Damping,” IEEE
[14] C. Roberts, “Review of International Grid Codes,” 2018. Trans. Sustain. Energy, pp. 1–1, 2018.
[15] Q. Hong et al., “Fast frequency response for effective frequency control [41] S. Bose, D. F. Gayme, U. Topcu, and K. M. Chandy, “Optimal
in power systems with low inertia,” 2018. placement of energy storage in the grid,” in 2012 IEEE 51st IEEE
[16] H. Thi, “General rights Frequency Characterization and Control for Conference on Decision and Control (CDC), 2012, pp. 5605–5612.
Future Low Inertia Systems.” [42] M. G. Hoffman, A. Sadovsky, M. C. Kintner-Meyer, and J. G. Desteese,
[17] F. Milano, F. Dorfler, G. Hug, D. J. Hill, and G. Verbic, “Foundations “Analysis Tools for Sizing and Placement of Energy Storage in Grid
and Challenges of Low-Inertia Systems (Invited Paper),” in 2018 Power Applications A Literature Review,” 2010.
Systems Computation Conference (PSCC), 2018, pp. 1–25. [43] Y. G. Rebours, D. S. Kirschen, M. Trotignon, and S. Rossignol, “A
[18] X. Luo et al., “Review of Voltage and Frequency Grid Code Survey of Frequency and Voltage Control Ancillary Services—Part I:
Specifications for Electrical Energy Storage Applications,” Energies, Technical Features,” IEEE Trans. Power Syst., vol. 22, no. 1, pp. 350–
vol. 11, no. 5, p. 1070, Apr. 2018. 357, Feb. 2007.
[19] PJM, “Implementation and Rationale for PJM’s Conditional Neutrality [44] “P1-Policy 1: Load-Frequency Control and Performance [C].”
Regulation Signals,” 2017. [45] “Network Code on Load-Frequency Control and Reserves in line with
[20] “ENHANCED FREQUENCY RESPONSE Invitation to tender for pre- the ACER Framework Guidelines on Electricity System Operation,”
qualified parties,” 2016. 2013.
[21] “DS3 System Services Protocol-Regulated Arrangements DS3 System [46] “FAST FREQUENCY RESPONSE IN THE NEM,” 2017.
Services Implementation Project,” 2017. [47] S. Bradbury, J. Hayling, P. Papadopoulos, and N. Heyward, “Smarter
[22] “National Electricity Transmission System Security and Quality of Network Storage Electricity Storage in GB: SNS 4.7 Recommendations
Supply Standard,” 2019. for regulatory and legal framework (SDRC 9.5).”
[23] J. Richardson and M. Bennett, “RM16-6- Essential Reliability Services [48] F. Sanchez, J. Cayenne, F. Gonzalez-Longatt, and J. L. Rueda,
and the Evolving Bulk-Power System—Primary Frequency Response.” “Controller to enable the enhanced frequency response services from a
[24] “Enhanced Frequency Response Frequently Asked Questions.” multi-electrical energy storage system,” IET Gener. Transm. Distrib.,
[25] J. Riesz and J. Palermo, “INTERNATIONAL REVIEW OF vol. 13, no. 2, pp. 258–265, Jan. 2019.
FREQUENCY CONTROL ADAPTATION Australian Energy Market [49] M. Dreidy, H. Mokhlis, and S. Mekhilef, “Inertia response and
Operator,” 2016. frequency control techniques for renewable energy sources: A review,”
[26] H. Thiesen, C. Jauch, and A. Gloe, “Design of a System Substituting Renew. Sustain. Energy Rev., vol. 69, pp. 144–155, Mar. 2017.
Today’s Inherent Inertia in the European Continental Synchronous [50] G. Delille, B. François, S. Member, and G. Malarange, “Dynamic
Area,” Energies, vol. 9, no. 8, p. 582, Jul. 2016. Frequency Control Support by Energy Storage to Reduce the Impact of
[27] M. Farhadi and O. Mohammed, “Energy Storage Technologies for High- Wind and Solar Generation on Isolated Power System’s Inertia,” IEEE
Power Applications,” IEEE Trans. Ind. Appl., vol. 52, no. 3, pp. 1953– Trans. Sustain. ENERGY, vol. 3, no. 4, 2012.
1961, May 2016. [51] M. ; Altin, J. C. Kuhlmann, K. ; Das, and A. D. Hansen, “Optimization
[28] K. Li and K. J. Tseng, “Energy efficiency of lithium-ion battery used as of Synthetic Inertial Response from Wind Power Plants,” Citation, vol.
energy storage devices in micro-grid,” in IECON 2015 - 41st Annual 11, no. 5, 2019.
Conference of the IEEE Industrial Electronics Society, 2015, pp. [52] R. Eriksson, N. Modig, and K. Elkington, “Synthetic inertia versus fast
005235–005240. frequency response: a definition,” IET Renew. Power Gener., vol. 12,
 15

no. 5, pp. 507–514, Apr. 2018. pp. 1–6.

[53] F. M. Gonzalez-Longatt and S. M. Alhejaj, “Enabling inertial response [75] “IEEE Standard for Interconnection and Interoperability of Distributed
in utility-scale battery energy storage system,” in 2016 IEEE Innovative Energy Resources with Associated Electric Power Systems Interfaces,”
Smart Grid Technologies - Asia (ISGT-Asia), 2016, pp. 605–610. IEEE Std 1547-2018, pp. 1–138.
[54] A. Bonfiglio, F. Gonzalez-Longatt, and A. R. Procopio, “Integrated [76] EIRGRID, “Grid Code v6 EirGrid Grid Code,” 2015.
Inertial and Droop Frequency Controller for Variable Speed Wind [77] NERC, “BAL-003-1.1 Frequency Response and Frequency Bias
Generators.” Setting,” 2017.
[55] H.-P. Beck and R. Hesse, “Virtual synchronous machine,” in 2007 9th [78] FERC, “RM11-7-000 Frequency Regulation Compensation in the,”
International Conference on Electrical Power Quality and Utilisation, 2011.
2007, pp. 1–6. [79] FERC, “Third-Party Provision of Ancillary Services; Accounting and
[56] Q.-C. Zhong and G. Weiss, “Synchronverters: Inverters That Mimic Financial Reporting for New Electric Storage Technologies,” 2013.
Synchronous Generators,” IEEE Trans. Ind. Electron., vol. 58, no. 4, pp. [80] NERC, “BAL-001-TRE-1-Primary Frequency Response in the ERCOT
1259–1267, Apr. 2011. Region.”
[57] L. Zhang, L. Harnefors, and H.-P. Nee, “Power-Synchronization Control [81] National Grid (UK), “THE GRID CODE (UK),” 2017.
of Grid-Connected Voltage-Source Converters,” IEEE Trans. Power [82] ENERGINET, “Technical regulation 3.3.1 for battery plants,” 2017.
Syst., vol. 25, no. 2, pp. 809–820, May 2010. [83] State Grid Corporation of China, Q/GDW 564-2011, Technical
[58] M. Kazemi, H. Zareipour, N. Amjady, W. D. Rosehart, and M. Ehsan, guideline for electric energy storage system interconnecting with
“Operation Scheduling of Battery Storage Systems in Joint Energy and distribution network. 2012.
Ancillary Services Markets,” IEEE Trans. Sustain. Energy, vol. 8, no. 4, [84] State Grid Corporation of China, “Q/GDW 696-2011, Operation and
pp. 1726–1735, Oct. 2017. control specification for electric energy storage system interconnecting
[59] B. Olek and M. Wierzbowski, “Local Energy Balancing and Ancillary with distribution network,” 2012.
Services in Low-Voltage Networks With Distributed Generation, Energy [85] ERCOT, “Nodal Operating Guide,” 2018. .
Storage, and Active Loads,” IEEE Trans. Ind. Electron., vol. 62, no. 4, [86] ISO-NE, “Operating procedure,” 2017. .
pp. 2499–2508, Apr. 2015. [87] IESO Canada, “Market Rules.” .
[60] B. Cheng and W. Powell, “Co-optimizing Battery Storage for the [88] PJM, “PJM Generators Operational Requirement,” 2017. .
Frequency Regulation and Energy Arbitrage Using Multi-Scale [89] M. Kuivaniemi, N. Modig, and R. Eriksson, “FCR-D design of
Dynamic Programming,” IEEE Trans. Smart Grid, pp. 1–1, 2016. requirements,” 2017.
[61] A. Perez, R. Moreno, R. Moreira, M. Orchard, and G. Strbac, “Effect of [90] Operador Nacional do Sistema Elétrico, “PROCEDIMENTOS DE
Battery Degradation on Multi-Service Portfolios of Energy Storage,” REDE (Networking Procedures),” 2018. .
IEEE Trans. Sustain. Energy, vol. 7, no. 4, pp. 1718–1729, Oct. 2016. [91] Terna, “The code for transmission, dispatching, developing and security
[62] J. W. Shim, G. Verbic, N. Zhang, and K. Hur, “Harmonious Integration of the grid (Annex A15 and A73),” 2014. .
of Faster-Acting Energy Storage Systems Into Frequency Control [92] Electricity Authority Te Mano Hiko, “Normal Frequency Management
Reserves in Power Grid With High Renewable Generation,” IEEE Strategic Review,” 2017.
Trans. Power Syst., vol. 33, no. 6, pp. 6193–6205, Nov. 2018. [93] Energy Market Authority, “Singapore Electricity Market Rules
[63] V. Calderaro, V. Galdi, F. Lamberti, and A. Piccolo, “Co-located storage Appendix 6J,” 2017.
systems with renewable energy sources for voltage support in [94] “The South African Grid Code-Network Code Rev The South African
distribution networks,” in 2015 IEEE Power & Energy Society General Grid Code The Network Code Version 7.0 2 The South African Grid
Meeting, 2015, pp. 1–5. Code-Network Code Rev.”
[64] East Penn Manufacturing Co, “Final Technical Performance Report [95] “GRID CONNECTION CODE FOR RENEWABLE POWER PLANTS
Grid-Scale Energy Storage Demonstration of Ancillary Services Using (RPPs) CONNECTED TO THE ELECTRICITY TRANSMISSION
the UltraBattery,” 2015. SYSTEM (TS) OR THE DISTRIBUTION SYSTEM (DS) IN SOUTH
[65] T. Lee, “ENERGY STORAGE IN PJM EXPLORING FREQUENCY AFRICA.”
REGULATION MARKET TRANSFORMATION,” 2017.
[66] “EUROPEAN ENERGY STORAGE TECHNOLOGY
DEVELOPMENT ROADMAP 2017 UPDATE Joint EASE/EERA
recommendations for a.”
X. BIOGRAPHIES
[67] M. Bahloul and S. K. Khadem, “Design and control of energy storage
system for enhanced frequency response grid service,” in 2018 IEEE Lexuan Meng (S’2012, M’2015) received the
International Conference on Industrial Technology (ICIT), 2018, pp. B.S. degree in Electrical Engineering and M.S.
1189–1194. degree in Electrical Machine and Apparatus from
[68] V. Knap, S. K. Chaudhary, D.-I. Stroe, M. Swierczynski, B.-I. Craciun, Nanjing University of Aeronautics and
and R. Teodorescu, “Sizing of an Energy Storage System for Grid Astronautics (NUAA), Nanjing, China, in 2009
Inertial Response and Primary Frequency Reserve,” IEEE Trans. Power and 2012, respectively. In 2015, he received the
Syst., vol. 31, no. 5, pp. 3447–3456, Sep. 2016. Ph.D. degree in Power Electronic Systems from
[69] D. M. Greenwood, K. Y. Lim, C. Patsios, P. F. Lyons, Y. S. Lim, and P. Institute of Energy Technology, Aalborg
C. Taylor, “Frequency response services designed for energy storage,” University, Denmark. 2015-2017, he was a
Appl. Energy, vol. 203, pp. 115–127, Oct. 2017. postdoc in Aalborg University working with
[70] E. Spahic, C. P. Susai Sakkanna Reddy, M. Pieschel, and R. Alvarez, Maersk Drilling on flywheel energy storage
“Multilevel STATCOM with power intensive energy storage for system for marine applications. 2017-2018, he
dynamic grid stability - frequency and voltage support,” in 2015 IEEE was a research engineer in Power Network
Electrical Power and Energy Conference (EPEC), 2015, pp. 73–80. Demonstration Centre, Glasgow, UK responsible for projects on grid scale
[71] Y. Kim, V. Raghunathan, and A. Raghunathan, “Design and energy storage systems. He is now an R&D engineer in FACTS, Power Grid
Management of Battery-Supercapacitor Hybrid Electrical Energy division, ABB, Sweden, with main responsibility on FACTS controllers and
Storage Systems for Regulation Services,” IEEE Trans. Multi-Scale grid energy storage systems. His research interests are in power electronic
Comput. Syst., vol. 3, no. 1, pp. 12–24, Jan. 2017. systems and devices for future grid applications.
[72] D. M. Greenwood, K. Y. Lim, C. Patsios, P. F. Lyons, Y. S. Lim, and P.
C. Taylor, “Frequency response services designed for energy storage,”
Appl. Energy, vol. 203, pp. 115–127, Oct. 2017.
[73] M. Bahloul and S. K. Khadem, “Impact of power sharing method on
battery life extension in HESS for grid ancillary services,” IEEE Trans.
Energy Convers., pp. 1–1, 2018.
[74] R. Capo-Misut, R. S. Munoz-Aguilar, J. Rocabert, J. I. Candela, and P.
Rodriguez, “Control of energy storage system integrating
electrochemical batteries and SC for grid-connected applications,” in
2016 IEEE Energy Conversion Congress and Exposition (ECCE), 2016,
 16

Jawwad Zafar was born in Pakistan. He range of projects including but not limited to new NIC, NIA, IRM and Losses
graduated with a Bachelor’s degree in Industrial Discretionary Reward. Responsibilities include the successful delivery of the
Electronics from the NED University of distribution Innovation Strategy and the securing of funding for new
Engineering and Technology, Karachi. transmission and distribution innovation concepts. In addition to Innovation
Dr Zafar received a Masters degree in Power Strategy, he is one of the SSEN representatives on the ENA Open Networks
Engineering from Chalmers University of Project, leading the TRANSITION, EFFS and FUSION (T.E.F.) interface with
Technology, Sweden in Electric Power the relevant workstreams and National Grid ESO. In February 2019 this role
Engineering and a PhD from Universite Libre de was extended to include all significant GB industry trials. He presently supports
Bruxelles, Belgium. the shaping of TRANSITION and its interface with the LEO Project partners to
His employment experience includes public ensure alignment with industry direction and that the Oxfordshire Projects
R&D centres, universities and consultancy. His collectively achieve more than their set deliverables, maximising the benefits
interest is in the control and impact of converter interfaced devices on the realised by our stakeholders which most importantly includes consumers.
electricity network.
Federico Coffele received the B.Sc. in
Electrical Engineering and M.Eng. in
Shafiuzzaman K Khadem (S’09-M’13) Electrical Power Systems from the
received the B.Sc and M.Sc degrees in Applied University of Padova, Italy in 2004 and
Physics and Electronics from University of 2007, respectively. He worked for ABB
Dhaka, Bangladesh, and PhD in Electrical and Power Products in Italy, in the R&D
Electronic Engineering from Technology department, and then he joined the
University Dublin, Ireland. University of Strathclyde, in Glasgow,
At present, he is a Senior Researcher and lead the Scotland, working in the area of protection
Embedded and Distributed Generation systems and control of power distribution networks,
(EDGE) research group in the International power system modelling and simulation.
Energy Research Centre (IERC). Prior to that, he During this period he also worked toward his PhD, which he received from the
also served in different research and academic positions in universities, University of Strathclyde in 2012.
international organizations and industries in multi-disciplinary teams. He also Currently he is the Research and Development Manager of the Power
led the Electricity Transmission System Leadership Training Course for top Network Demonstration Centre (PNDC) at the University of Strathclyde, which
management staff of the newly created Electricity Sector Regulatory Agency is a world-class facility with dedicated staff to accelerate the adoption of new,
(ARSEL) – Cameroon (World Bank sponsored). He is research active in smart smart technologies within present and future power grids, supporting the
grid, intelligent energy systems integration, such as microgrids, virtual power increased accommodation of renewable energy, electric vehicles and demand
plant, converter design, active power filtering. side management.
Dr Khadem received recognitions from different institutions including HEA
(IE), ICTP (IT), EU Erasmus (UK), DIT (IE), IRCSET (IE), IEEE (USA). He
is a member of the research grant evaluation panel H2020 MSCA (IF). He is Graeme M. Burt (M’95) received the B.Eng.
also a member of the IEEE power and energy society and smart grid degree in electrical and electronic engineering,
community. and the Ph.D. degree in fault diagnostics in
power system networks from the University of
Alan Collinson For the past four years Alan Strathclyde, Glasgow, U.K., in 1988 and 1992,
has been part of the Commercial and respectively. He is currently a Professor of
Innovation Group at SP Energy Networks. electrical power systems at the University of
He has worked on several innovation projects, Strathclyde where he co-directs the Institute for
focusing on Smart Grid technologies, Energy and Environment, directs the Rolls-Royce
including Network Visibility, Community University Technology Centre in Electrical
Energy, Energy Storage and the DSO Power Systems, and is lead academic for the
transition. Power Networks Demonstration Centre (PNDC).
Alan previously worked in the SP Manweb In addition, he serves as spokesperson for the
Network Design and Planning department, board of DERlab e.V., the association of
responsible for 33kV and 132kV network distributed energy laboratories. His research interests include the areas of
reinforcement, including the major regeneration of the Liverpool City Centre power system protection and control, distributed energy, hybrid propulsion
and the connection of many large on-shore wind farms in North and Mid Wales. and experimental validation.
Alan’s first involvement in utility-scale electrical energy storage stretches back
to the 1990’s, with his work with the International Energy Agency and the
formation of the UK National Team on Energy Storage, which included the
inaugural international energy storage conference – EESAT ’98, in Chester. He
subsequently became involved in the DNO-led Energy Storage Operators
Forum.
Alan has been involved in many industry forums and ENA working groups and
is currently working within the ENA’s “Open Networks” DSO project.

Kyle Murchie was born in Glasgow in

Scotland, United Kingdom, on February 25,
1990. He graduated from Mearns Castle High
School, Newton Mearns, and studied at the
University of Strathclyde.
His employment experience includes
Scottish and Southern Electricity Networks,
Electricity North West, TNEI Group and UK
Power Networks (formerly EDF Energy
Networks). His special fields of interest
included Power System Economics and the
wider transition to a Low Carbon Economy.
Kyle Murchie is currently Situated within the DSO and Innovation team at
Scottish and Southern Electricity Networks, in a role encompassing a broad