Contribution of Power Converters in Frequency Stability

of Low-inertia Power Systems

Fadi Kelada, Jérôme Buire

To cite this version:

Fadi Kelada, Jérôme Buire. Contribution of Power Converters in Frequency Stability of Low-inertia
Power Systems. Conférence des Jeunes Chercheurs en Génie Electrique - Journées Couplées du GDR
SEEDS, Jun 2022, Le Croisic, France. �hal-03727208�

HAL Id: hal-03727208

https://hal.archives-ouvertes.fr/hal-03727208
Submitted on 19 Jul 2022

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
 Contribution of Power Converters in Frequency
Stability of Low-inertia Power Systems
Fadi Kelada, Jérôme Buire
Univ. Grenoble Alpes, CNRS, Grenoble INP, G2ELab, 38000 Grenoble, France
{fadi.kelada, jerome.buire}@g2elab.grenoble-inp.fr

ABSTRACT - Integration of low carbon electricity generation technologies such as Renewable Energy Sources (RES) has
been increasingly considered as a potential way to decrease Green House Gases (GHG) emissions from conventional fossil
fuel power plants. With such increased penetration of such RES specially those interfaced with the network through power
electronic converters (RES-CI), new challenges emerge in the way the resulting future power systems are conceived and
operated. In this manuscript, we discuss the challenge of low-inertia prevailing in such systems, we touch down upon the
roots of the problem, and assess a potential solution by considering the possibility of converters contributing to the grid’s
operation, stability and robustness. A small signal dynamic model of a converter outer control emulating inertial response
is utilized to prove the concept. Results show decreased ROCOF and increased nadir values. Additional conclusions related
to tuning, control and stability analysis were also drawn from the results. Finally some future research directions to proceed
in this domain are proposed.

Keywords - Low-Inertia Systems, Inertia emulation, Frequency Stability, Frequency Regulation, Inverter-based Resources
(IBR), Converter-interfaced Generation (CIG), Microgrids.

1. Introduction
Climate change was identified by the United Nations and other environmental organization as the most critical threat
humans might face in the twenty-first century [1]. The UN Sustainable Development Goals (UN-SDG) concerned this
threat with two of its seventeen goals [2], [3]: UN-SDG7 - Affordable and Clean Energy and UN-SDG13 - Climate
Action, setting serious goals for international community to act on [4]. The increase of the RES share is at the heart of
the measures taken to decrease the GHG emissions and help mitigate the effects of Global Warming. Other measures
include also employing low carbon emissions power sources and the respective shift from fossil-based power plants.
Many of the low or 100% carbon free technologies deployed are interfaced to the network through a power electronic
converter. Nevertheless, other portion of these low carbon emissions technologies can also rely on the conventional
synchronous generators to produce electricity and thus are directly connected to the power system. Fig. 1 shows that
many of the technologies that provide 100% Carbon-free emissions are synchronous based while as many are interfaced
using power electronic converters [5]. At the present, power systems mostly utilize power generation units from the
right portion of Fig. 1. These units are mostly of very large capacity and usually centralized. Large centralized facilities
are mostly economic due to their large scale, but they usually have to transmit power to consumers far away, hence
may affect the environment negatively [6].
At a certain instant in the future, the weight of the power-electronic-based and converter-based generation technologies
will exceed that of the more conventional, synchronous generator-based production units. With that new challenges
emerge, such as; the small inertia related to these converter-based units, their intermittency and stability related issues
[6]–[9]. The majority of these CIGs are wind and PV units. The volatility and uncertainty of the primary energy
resources harnessed by these units can be of significant problem to the operation of the power system. Large-scale
penetration of such CIG units into the conventional Synchronous Generator (SG) based power system leads to frequency
stability issues. This is due to the fact that the power electronic converter interfacing such units and the grid prevents
them from having an inertial response to any eventual unbalance. Conventional power plants such as those generation
units at the right-hand-side of Fig. 1 utilize synchronous generators to produce electrical power. These synchronous
generators are directly connected to the grid without any power electronic interface and thus exhibit the natural inertial
response to any eventual unbalance in the network.
CIG units on the other hand does not inherit such natural response. For instance, PV arrays require power electronic
dc–ac inverters to integrate with the grid and do not offer an inertial response, and certain types of wind turbines utilizes
variable frequency ac – dc – ac converters, which decouple the wind turbine inertia from the grid. Consequently, the
inertia of the power system decreases as the penetration of RES-CI/CIG increases. The reduced inertia in the power
system leads to an increase in the rate of change of frequency (ROCOF) and frequency deviations in a very short
time, under power imbalances that substantially affect the frequency stability of the system [7], [10]. Studies such as
that performed on the Eastern US Interconnection system shows that the system’s frequency response will decrease
dramatically when the renewable penetration reaches 60 to 80% [11].
This article examines the challenge of decreasing network’s synchronous inertia, its root causes, and investigates a
proposed solution through the contribution of power converters to the network stability and robustness through emulating

Le Croisic - 14-17 juin 2022 1

 LHS: Modern PS RHS: Conventional PS

100% Carbon-free Technologies

Synchronous Machine Based

Converter-Interfaced Renewable Energy
Generation Sources Pumped-
VS- Nuclear
Batteries Hydro CAES Steam-fired
Pumped -
Fuel Cells Hydro Hydro Synchronous Gas-fired
Flywheels PV Geothermal
Condenser Diesel-fired
SVC/STATCOM - Wind Biomass
Solar-
HVDC
thermal

RES-CI RES-SM

Figure 1: Different power generation plant types. (SVC:Static Var Compensator, STATCOM:Static
synchronous compensator, VS: Variable Speed, CAES: Compressed Air Energy Storage, PS: Power System) [5]

SG’s inertial response in the active power control loop of the converter. Additionally, future research directions are
proposed to further investigate the issue of low-inertia in the future power systems.
The remainder of this paper is organised as follows. Section 2 explores the concept of inertia and the relation between
inertia and frequency regulation and control in conventional power systems. Section 3 first explains the effect of CIG
integration on the previous inertia concept and then assesses the proposed inertia emulation technique using such
converters as a solution to the issue. Finally section 4 concludes with the manuscript most important key-takeaways
and propositions for future steps.

2. The low inertia challenge

Inertia is the term coined to describe the physical behavior of a system to continue in its initial state of motion when an
exterior force acts upon it. A car for example, tends to continue motion in its initial direction when the driver presses
the break pedal. The car doesn’t stop instantaneously, but continues in its initial trajectory for a certain distance until
the friction with the ground dissipates all its Kinetic Energy (KE) and then finally stops. A power plant utilizing an
SG directly connected to the rest of the grid is physically analogous to the previous concept. The famous SG’s swing
equation derived from newton’s second law of motion for rotating objects is shown in equation (1).

X J.ωe d2 θe
P = . (1)
p2 dt2
P
where P is the sum of all active power produced by the generator, e.g., Pe − Pm − Pf for instance. J is the
moment of inertia which physically represents the resistance to the change in the rotational speed of the SG. θe and
ωe are respectively the rotor electrical angular position and the electrical rotational speed. p is the number of rotor
field winding pole pairs.
To further simplify the above second order differential swing equation, first, an important term is introduced, which is
the inertia constant H, which is defined as the the time period in seconds when the system can produce the nominal
power only through the stored KE in its rotor. Thus, the inertia constant H can be expressed as:
J.ωe,0
Ek 2.p2
H= = (2)
S S

Further more, the second order swing equation (1) can be expressed in per unit (P̃ ) and rewritten as two first order
equations. Considering the rated machine power and nominal frequency as base values and assuming only small
deviations from rated frequency (ωe ≈ ωe,0 ) and by substituting by equation (2), we can express the swing equation
simply as:
X dω˜e
P̃ = 2H. (3)
dt
Le Croisic - 14-17 juin 2022 2
The previous swing equation is derived for a single machine. For a power system of (n) SGs, the following generalization
is used. First, the inertia constant for the whole network is calculated as in equation (4). Then, as the frequency at the
different generators’ terminals is not the same at the event of a certain imbalance in the network, the generalized swing
equation represents the aggregated average fluctuations of the electrical frequency at a virtual point in the network
called the ”Center of Inertia” (COI) and is expressed in equation (5) for a certain power imbalance ∆Pdist .

n
X n
X
Ek,i Hi .Si
i=1 i=1
Htotal = = n (4)
Stotal X
Si
i=1

dω̃e,COI
∆P˜dist = 2Htotal . (5)
dt
The total inertia Htotal in a conventional power system can be described as a resistance in the form of KE exchange from
rotating machines, to compensate for the fluctuations in frequency arising from power imbalances. Thus, instantaneous,
short-term energy support under load fluctuations is called Synchronous Inertia (SI). Consequently, increased value of
the total network inertia constant Htotal means that the network exhibits a large amount of KE stored in its rotating
masses and hence, can compensate momentarily for larger power imbalances and result in less severe frequency
fluctuations and drops until the generating units equipped with frequency control governors start reacting to arrest such
frequency fluctuations (primary control) and return the frequency to its nominal value (secondary control).
Frequency ROCOF and nadir are indicators of how severe a frequency fluctuation is following an imbalance event.
The former is an indication of how rapidly the frequency changes after a sudden imbalance and is calculated from
equation (6). The latter determines the lowest frequency reached following the imbalance after the primary frequency
control started acting to prevent further frequency drop. The frequency ROCOF and nadir values are very important,
as they determine the set points of the safe operation relays such as the load shedding relays (UFLS) pre-configured
by the network operators to maintain the safe and reliable operation of the network by arresting increased and further
drop of frequency below limits of operation.
∆Pdist .fb
ROCOF = (6)
2Stotal .Htotal

3. Power converters contribution to network inertia

Power converters don’t intrinsically inherent enough energy storage to provide inertia response such as that provided
by synchronous machines and discussed above. Additionally, power converters interfacing wind energy for instance,
prevent the physical coupling between the KE in the blades and the grid [12], [13]. Thus, the decrease of synchronous
based units and an increase of penetration of such RES-CI units imposes the following challenges to network operators
[7], [8]:
1) Very limited inertial response following an imbalance,
2) Decreased frequency regulation reserves such as Primary Frequency Curtailment Reserves (P-FCR),
3) Increased ROCOF and decreased nadir values,
4) Resulting in triggering of ROCOF relays, activating network security thresholds such as UFLS,
5) Consequently, the overall network frequency stability is jeopardized.
To be able to face these challenges, current research is undergoing in the field of power converters control to compensate
for the previous challenges without jeopardizing the overall grid stability. For that purpose, the classification of power
converters based on control technique has been undergoing a continuous development and adjustment to be more
representative of the requirements of the continuously developing and evolving power system [14]–[16]. Classically in
literature, power converters controls are classified into three distinct categories namely ”Grid-forming”,”Grid-feeding”,
and ”Grid-supporting”. Grid-forming (GFM) control aims to create the network’s nominal frequency and voltage
references, while Grid-feeding/following (GFL) control is most commonly used with PV and wind converters, since
they are designed to work at maximum power point (MPP) and synchronize their output with the network frequency
and voltage. Grid-supporting main objective on the other hand is to participate in the regulation of the ac grid voltage
amplitude and frequency references by controlling the active and reactive power delivered to the grid [17].
In recent literature, GFM technology has grown beyond the simple need to create and maintain the reference network
frequency and voltage to being a generic inverter control that can be tuned and designed to provide a wide variety of
network requirement needed by the network operator, while maintaining its base role as creating and maintaining the
network base frequency and voltage reference. GFM inverter can thus be defined based on its capability and the grid
services it provides to the network [18], [19]. One of these classifications is shown in table I according to [19], [20]. It
can be seen how the distinctive limit between GFL and GFM technologies fades when only classifying the converter

Le Croisic - 14-17 juin 2022 3

 Table I: Classification of power converters according to their contribution to network support and stability
[19], [20]

Strategy/Contribution Strategy A Strategy B Strategy C

Active Power tracking ✓✓ ✓✓ X
Active Power saturation ✓✓ ✓✓ X
Frequency support ✓✓ ✓✓ ✓✓
Inertia contribution X ✓✓ ✓✓
Robustness ✓✓ ✓ ✓

𝑣𝑔
𝑣𝑐 𝑅𝑐 , 𝐿𝑐 𝑖𝑐 𝑖𝑔 𝑅𝑔 , 𝐿𝑔 𝑣𝑒
=
RES-CI

𝑣𝑑𝑐 𝐶𝑑𝑐
SM
≈
Power Converter Hydro-power Plant
𝑃𝑙𝑜𝑎𝑑

(a)

Active Power Control Loop

𝜔
ූ𝑔 1 𝑃𝑔𝑜𝑣𝑒𝑟𝑛𝑜𝑟
−2. 𝑇𝑤 . 𝑠 + 1 ∆𝑃𝑚 ∆𝜔𝑒
𝑃𝑚𝑒𝑎𝑠 - 1 1
𝑃∗ +
𝑠
+ 𝑇𝑦 . 𝑠 + 1 𝑇𝑤 . 𝑠 + 1 + 2𝐻. 𝑠
𝐺𝑎𝑡𝑒 𝑆𝑒𝑟𝑣𝑜 𝐿𝑎𝑔 𝑇𝑢𝑟𝑏𝑖𝑛𝑒 𝑀𝑜𝑑𝑒𝑙

1 - 𝐾 ∆𝜔𝑚 𝜔𝑐 𝜔𝑚
𝜔∗ + + + +
- 𝑅𝑐 𝜔𝑏 𝑠 + 𝜔𝑐 ∆𝑃𝑔

Turbine-governor Model

𝑇𝑢𝑟𝑏𝑖𝑛𝑒 𝐷𝑟𝑜𝑜𝑝

(b) (c)

Figure 2: (a) The network under study, (b) Converter’s active power control loop, (c) Hydro’s
turbine-governor model

control according to the contribution it makes to the grid. It can also be noted how converters can be controlled to
contribute to the network inertia instead of decreasing it.
In order to understand how converters can be controlled to provide inertial response and increase network inertia
constant, let us consider the following network proposed in this paper and shown in Fig. 2a. The network consists
of a power converter connected to a large synchronous machine representing a large hydro-power plant. In order to
analyze the active power and frequency dynamics accompanying the inertial response under study and following a
certain power imbalance, the active power control loop of the converter as well as the hydro turbine-governor models
are considered only in this study. The former and latter models are shown in Fig. 2b-2c respectively. The transfer
function of the converter active power loop (after the droop constant) shown in Fig. 2b is shown in equation (7). The
use of the low pass filter (LPF) with cut-off frequency ωc results in a lag between the grid measured frequency ω̆g
and synchronizing angle used in the control ωm , which emulates the inertial effect.

ωb dωm ωb
p∗ − pmeas = . + .(ωm − ω̆g ) (7)
K.ωc dt K

Equation (7) is analogous to that of virtual synchronous generator (VSG) control concept. It can be reformulated into:
dωm ωb
p∗ − pmeas = 2Hv . + .(ωm − ω̆g ) (8)
dt K
Thus, by comparing equation (7) to equation (8), the emulated inertia constant can be tuned as:
ωb
Hv = (9)
2K.ωc

Developing the small-signal model of the power converter control loop in Fig. 2b would result in a fourth order transfer
function model. Assuming all frequencies are approximately equal, an equivalent decomposition of the fourth order
model in two-second-order model can be deduced [20]. The resulted small-signal model describing the grid frequency
dynamics in response to an imbalance is shown in Fig. 3. An illustration of the effect of the emulated inertia by the

Le Croisic - 14-17 juin 2022 4

 1 𝑃𝑔𝑜𝑣𝑒𝑟𝑛𝑜𝑟 −2. 𝑇𝑤 . 𝑠 + 1 1 ∆𝜔𝑒
1 ∆𝑃𝑚
-+ + 𝑇𝑦 . 𝑠 + 1
+ 2 𝐻 + 𝐻𝑣
𝑃𝑐𝑎𝑝,𝑐
൘𝑃 .𝑠
𝑠 𝑇𝑤 . 𝑠 + 1 𝑐𝑎𝑝,𝐻
𝐺𝑎𝑡𝑒 𝑆𝑒𝑟𝑣𝑜 𝐿𝑎𝑔 𝑇𝑢𝑟𝑏𝑖𝑛𝑒 𝑀𝑜𝑑𝑒𝑙

System Dynamic
Response

∆𝑃𝑙𝑜𝑎𝑑
-+

∆𝑃𝑐 ∆𝑝∗
𝑇𝑢𝑟𝑏𝑖𝑛𝑒 𝐷𝑟𝑜𝑜𝑝
𝐶𝑜𝑛𝑣𝑒𝑟𝑡𝑒𝑟 𝐷𝑟𝑜𝑜𝑝
Hydro-Turbine-Governor
Model Converter Active Power
Control Model

Figure 3: Small-signal dynamic model

(a) (b)

Figure 4: (a) Effect of emulated inertia, (b) converter versus turbine-governor system in regulating active
power

converter in supporting the network stability and robustness to ROCOF and nadir is simulated and some insights are
shown in Fig. 4.
It can concluded from Fig. 4a that a disturbance of 500MW could result in very high ROCOF (6.25 Hz/s) and
jeopardize the network frequency stability in a system with base synchronous inertia constant of 2s (the blue curve).
The contribution of the converter in providing emulated inertial response can result in lower ROCOF and higher
frequency nadir as shown. Additionally, in contrary to conventional synchronous inertia which depends of the generator
parameters as shown from equation (2), the emulated inertia by the converter is tunable as shown from equation (9).
Fig. 4b shows how the turbine-governor model reacts versus the converter active power control loop to this disturbance
of 500MW both at the case where the converter emulates the inertial response and without it. It can be noticed how
the converter control is hundreds of magnitudes faster than the turbine-governor model which is considered a key
advantage for providing frequency regulation support using power converters. It should also be noted that during the
transient period, active power of the converter increased above 500MW which is also his rated maximum power. A
limiter should thus be added to the control to prevent over-loading.

4. Conclusions and future directions

In this paper, we demonstrated the contribution of power converter in the grid robustness and frequency stability
through emulating inertial response using a small-signal dynamic model of a network supplied by a synchronous
generator and a power converter. The validity of the proposed model controllers in response to a certain imbalance
was tested and their inertial response was simulated. The shown configuration appears to be stable in the regime of
small-signals and the inertial response emulated by the power converter’s controls is proven to be effective in arresting
high ROCOF values and limiting the frequency nadir. Another insight brought to the surface by this manuscript is the
effectiveness of the novel power converter classification reported in literature, where power converters are classified
according to their contribution to the network which removes the ambiguity accompanying the previous classification

Le Croisic - 14-17 juin 2022 5

into grid-following/forming and supporting. The authors believe that the recent classification is more representative of
the control architectures.
Recently, many techniques for inertia emulation and frequency regulation using power converters are proposed in the
literature. Evaluating such techniques within different and complex network topologies and investigating the inertia
placement homogeneity across the network is reported to be the next step in assessing potential solutions for the
challenges of low-inertia networks. Time-domain simulations could be handy in evaluating the different controls on
the overall network stability of complex and large networks to analyze the effect of network topology and inertia
placement on the frequency stability of complex and huge power systems and the similarities/differences between such
systems and very small ones such as Microgrids. Possible future investigation of the network’s stability assessment
using Lyapunov’s energy functions could provide interesting insight of the network overall stability. However, the
former despite being a good candidate for evaluating stability of non-linear systems, it requires the good selection of
the energy function which might be a challenging task to perform.

References
[1] Meetings Coverage and Press Releases, Climate Change ‘Biggest Threat Modern Humans Have Ever Faced’,
World-Renowned Naturalist Tells Security Council, Calls for Greater Global Cooperation — Meetings Coverage
and Press Releases. [Online]. Available: https : / / www. un . org / press / en / 2021 / sc14445 . doc . htm (visited on
02/27/2022).
[2] UN DESA, Goal 7 — Department of Economic and Social Affairs, 2020. [Online]. Available: https://sdgs.un.
org/goals/goal7 (visited on 02/27/2022).
[3] ——, Goal 13 — Department of Economic and Social Affairs, 2020. [Online]. Available: https://sdgs.un.org/
goals/goal13 (visited on 02/27/2022).
[4] Springer Nature, Carbon neutrality has become a global goal to mitigate rapid climate change — For Researchers
— Springer Nature. [Online]. Available: https : / / www. springernature . com / gp / researchers / the - source / blog /
blogposts-communicating-research/carbon-neutrality-global-goal-mitigate-climate-change/19685760 (visited on
02/27/2022).
[5] J. Matevosyan, J. Macdowell, N. Miller, et al., “A Future with Inverter-Based Resources: Finding Strength from
Traditional Weakness,” IEEE Power and Energy Magazine, vol. 19, no. 6, pp. 18–28, 2021.
[6] H. Bevrani, B. Francois, and T. Ise, Microgrid Dynamics and Control. 2017.
[7] K. S. Ratnam, K. Palanisamy, and G. Yang, “Future low-inertia power systems: Requirements, issues, and
solutions - A review,” Renewable and Sustainable Energy Reviews, vol. 124, no. February, p. 109 773, 2020.
[8] F. Milano, F. Dorfler, G. Hug, D. J. Hill, and G. Verbič, “Foundations and challenges of low-inertia systems
(Invited Paper),” 20th Power Systems Computation Conference, PSCC 2018, 2018.
[9] P. Tielens, “Operation and control of power systems with low synchronous inertia,” ARENBERG DOCTORAL
SCHOOL Faculty of Engineering Science, no. November, 2017. [Online]. Available: https://lirias.kuleuven.be/
bitstream/123456789/592208/1/phd%7B%5C %7DPieter%7B%5C %7DTielens.pdf.
[10] ENTSO-E, “Inertia and Rate of Change of Frequency ( RoCoF ),” Executive summary, vol. 17, pp. 6–48, 2020.
[Online]. Available: https://eepublicdownloads.azureedge.net/clean- documents/SOC%20documents/Inertia%
20and%20RoCoF%7B%5C %7Dv17%7B%5C %7Dclean.pdf.
[11] S. You, Y. Liu, X. Zhang, et al., “Impact of high PV penetration on U.S. eastern interconnection frequency
response,” IEEE Power and Energy Society General Meeting, vol. 2018-Janua, pp. 1–5, 2018.
[12] Z. Wu, W. Gao, T. Gao, et al., “State-of-the-art review on frequency response of wind power plants in power
systems,” Journal of Modern Power Systems and Clean Energy, vol. 6, no. 1, pp. 1–16, 2018.
[13] E. Muljadi, V. Gevorgian, M. Singh, and S. Santoso, “Understanding inertial and frequency response of wind
power plants,” PEMWA 2012 - 2012 IEEE Power Electronics and Machines in Wind Applications, pp. 1–8, 2012.
[14] B. Kroposki, B. Johnson, Y. Zhang, et al., “Achieving a 100% Renewable Grid: Operating Electric Power Systems
with Extremely High Levels of Variable Renewable Energy,” IEEE Power and Energy Magazine, vol. 15, no. 2,
pp. 61–73, 2017.
[15] U. Markovic, J. Vorwerk, P. Aristidou, and G. Hug, “Stability Analysis of Converter Control Modes in Low-
Inertia Power Systems,” Proceedings - 2018 IEEE PES Innovative Smart Grid Technologies Conference Europe,
ISGT-Europe 2018, no. 2, pp. 11–16, 2018.
[16] U. Markovic, O. Stanojev, E. Vrettos, P. Aristidou, and G. Hug, “Understanding Stability of Low-Inertia Systems,”
2019.
[17] J. Rocabert, A. Luna, F. Blaabjerg, and P. Rodrı́guez, “Control of power converters in AC microgrids,” IEEE
Transactions on Power Electronics, vol. 27, no. 11, pp. 4734–4749, 2012.
[18] EPRI, “Generic Positive Sequence Domain Model of Grid Forming Inverter Based Resource,” 2021.
[19] T. Qoria, T. Prevost, G. Denis, F. Gruson, F. Colas, and X. Guillaud, “Power converters classification and
characterization in power transmission systems,” in 2019 21st European Conference on Power Electronics and
Applications, EPE 2019 ECCE Europe, 2019.
[20] T. Qoria, “Grid-forming control to achieve a 100% power electronics interfaced power transmission systems,”
Electric power, HESAM Université, 2020.

Le Croisic - 14-17 juin 2022 6