Assessment of the Synthetic Inertial Response of an Actual Solar PV Power Plant

Electrical Power and Energy Systems 157 (2024) 109875 Available online 21 February 2024 0142-0615/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Assessment of the synthetic inertial response of an actual solar PV power plant Raquel Villena-Ruiz a , * , Andr´ es Honrubia-Escribano a , Jesús C. Hern´ andez b , Emilio G´ omezL´ azaro a a Renewable Energy Research Institute and DIEEAC-ETSII-AB, Universidad de Castilla-La Mancha, 02071 Albacete, Spain b Department of Electrical Engineering, Universidad de Ja´ en, 23071 Ja´ en, Spain ARTICLE INFO Keywords: Grid code Inverter Renewable energy Solar PV Synthetic inertia ABSTRACT As the large-scale integration of renewable energy grows, inertia in power systems is reduced, which causes the rate of change of frequency to increase and worsens the frequency nadir during disturbances. There is, therefore, an urgent need to accelerate the development of international standards and regulations that require solar PV systems and other renewable installations to also contribute to stability in the network by emulating inertia. Spain is a pioneer in this field: it is currently the only country that includes the synthetic inertia technical requirement in its grid code. In this context, the present paper analyzes the synthetic inertial response of a PV system model that forms part of an actual PV power plant in Spain under two scenarios, according to its grid code: i) when measurements are conducted at the power conversion PV system level and hence only its control algorithms are executed; and ii) when measurements are conducted at the power plant level and the power plant controller also influences the response of the entire installation. Among its aims, this paper seeks to thoroughly assess the response of an actual PV power conversion system that has the capability of actively contributing to frequency regulation in a network and to critically analyze the suitability of having a power plant controller when the PV facility contributes to power-frequency control by means of an inertial response. The results reveal, at the power conversion PV system level, that the inertia emulation module significantly reduces the response times of the installation to changes in frequency -by more than 68% in the worst case under frequency increases, and by 73% under frequency decreases, thus contributing to more rapid power-frequency regulation in the network. However, at the power plant level, the power plant controller disturbs the PV power conversion system’s behavior and causes the opposite effect: response times increase -up to more than 7% in the most unfavorable case when the inertia module is enabled. 1. Introduction The introduction section is organized as follows. Section 1.1 describes the current situation regarding the presence of Variable Renewable Energy (VRE) in power systems. It also addresses the main challenges -with a particular emphasis on the loss of inertia that countries are facing to correctly integrate these renewable resources to maintain the reliability of power systems, focusing on the case of Spain. Section 1.2 details the main scientific contributions on the emulation of inertia in Photovoltaic (PV) systems, and finally, Section 1.3 summarizes the contribution and novelty of the paper and defines its structure. 1.1. Motivation and incitement. Development of the new Spanish grid code According to the International Energy Agency (IEA), the integration of VRE into countries’ power systems can be divided into different phases, characterized by penetration level and by technical, regulatory, market and institutional integration challenges. These phases range from phase 1, in which the share of VRE in the system is insignificant, with effects being at a very local level, to phase 6, in which large energy storage systems will be required to cope with the very high share of VRE [1]. In phase 4, the share of VRE is already high and significant operational and regulatory modifications are required. In countries at phase 4, * Corresponding author. E-mail addresses: [email protected] (R. Villena-Ruiz), [email protected] (A. Honrubia-Escribano), [email protected] (J.C. Hern´ andez), emilio.gomez@ uclm.es (E. G´ omez-L´ azaro). Contents lists available at ScienceDirect International Journal of Electrical Power and Energy Systems journal homepage: www.elsevier.com/locate/ijepes https://doi.org/10.1016/j.ijepes.2024.109875 Received 12 July 2023; Received in revised form 19 December 2023; Accepted 12 February 2024 International Journal of Electrical Power and Energy Systems 157 (2024) 109875 2 it is urgent to adapt the regulatory documents for VRE-based technologies to provide system services and respond appropriately to changes in generation and demand. Although most countries are still at phase 1 or 2, in others, such as Denmark or Spain, the penetration level of VRE is already very high, and hence immediate substantial modifications in the regulatory frames that govern the operation of renewable power plants are required. At technical and operational level, the requirements demanded for grid connection of generators are typically established in documents known as grid codes. The measures to enable the achievement of greater power system flexibility as the share of VRE increases will, therefore, have to be reflected, first and foremost, in these normative documents. In the European Union (EU), a set of three grid codes, known as Network Connection Codes (NCC), were released in 2016:i) Regulation EU 2016/ 631 [2], establishing a network code on requirements for grid connection of generators; ii) Regulation EU 2016/1388 [3], establishing a network code on demand connection; and iii) Regulation EU 2016/1447 [4], establishing a network code on requirements for grid connection of high voltage direct current systems and direct current-connected power park modules. Drawing on these NCCs, the member countries have developed their own grid codes. In Spain, Royal Decree (RD) 647/2020 of July 7 [5] and Ministerial Order (MO) TED/749/2020 of July 16 [6] are the official documents resulting from the national adaptation of the NCCs. This grid code includes the requirements for grid connection of synchronous power-generating modules, power park modules and offshore power park modules; i.e., all types of electricity generation facilities. RD 647/2020 [5] regulates certain aspects required for the implementation of the grid codes, while MO TED/749/2020 [6] defines the technical requirements with which generation facilities must comply. However, these documents do not contain information on the compliance process evaluation, nor on the certification methods of the power plants. Therefore, an additional document was released to coordinate the process of compliance monitoring of the NCCs, in particular, to monitor compliance with the technical requirements defined in MO TED/749/2020 [6]. This document is the Spanish technical supervision standard, ‘Norma T´ ecnica de Supervisi´ on’ (NTS) [7], the latest edition of which was published in July 2021. MO TED/749/2020 and the NTS contain information on how the electricity generation facilities that are to be connected to the network must behave under different types of disturbances. The requirements to be evaluated in the installations include power regulation under both over-frequency and sub-frequency periods, reactive power control through different methods, Fault-Ride Through (FRT) capability and reactive power capability. Moreover, there is a specific requirement for power-generating modules that do not have an inherent capability to resist or slow down frequency deviations, i.e., for those nonsynchronous modules lacking ‘natural’ inertia. This requirement is the capability of providing synthetic inertia. Synthetic inertia, also known as virtual inertia or hidden inertia, is the artificial inertial behavior emulated by renewable generation units that are decoupled from the grid by electronic converters [8]. More technically, inertia emulation consists of the electronic converters responding with increases or decreases in active power proportional to the time derivative of grid frequency [9]. The Spanish grid code includes not only the synthetic inertia requisite but also the methodology to be followed by PV power plants in Spain in order to comply with it. Nonetheless, rather than being a mandatory requirement, it is currently of voluntary application. In this sense, it can be stated that Spain is a pioneer in the inclusion of the synthetic inertia technical requirement in its grid code, since the EU, in Regulation EU 2016/631 [2], refers to the requirement but the choice and discretion on how to apply it and under what conditions is left to each country. In other countries, such as Germany, Italy or the UK, where the penetration of renewable energy is also very high, discussions have started to take place about the synthetic inertia requirement. However, it has not yet been included as a requisite in their grid codes, not even as an ‘optional requirement’. Until now, frequency control in a power system was typically conducted by synchronous generators, which, in addition, provided inertia to the system because of their rotor masses. Nevertheless, as the integration of VRE grows, power systems’ inertia is reduced, and the rate of change of frequency increases when the system is subjected to disturbances such as increase or loss in the generation or demand. This makes it more difficult to control synchronism of the power system and results in having a less robust system to cope with those sudden large imbalances between generation and demand. This is, therefore, one of the main reasons why additional frequency control is required and VRE based technologies should also contribute to this issue and provide inertia to the system. In this new scenario, in which the share of VRE in power systems is already reaching significant proportions, and with a future that is shaping up to be even more ambitious in terms of the presence of clean generation sources, it is necessary for these technologies to also contribute to the regulation of power frequency. In Spain, wind and solar PV power are the renewable technologies with the greatest presence in the electricity system. In the particular case of PV power, its global market grew significantly in 2022. The total cumulative installed PV power capacity increased from 942 GW at the end of 2021 [10] to 1185 GW at the end of 2022 [11]. In this context, China is the global leader, with 414.5 GW installed, accounting for more than one third of the world PV power capacity. In the EU, which continued to grow strongly, Spain was the largest market for PV power in 2022, with an estimated 8.1 GW annual installed capacity, and is ranked first worldwide in PV penetration, contributing to the electricity demand with a penetration level of 19.1% in 2022 [11]. In this context, the present paper analyzes the synthetic Inertial Response (IR) of a specific PV power conversion system model that forms part of an actual PV power plant in Spain. It does so in two different scenarios: i) analyzing its response when only the control algorithms of the PV power conversion system itself are considered; and ii) analyzing the IR response of the complete PV power plant when also considering the control actions of the Power Plant Controller (PPC). The simulations required to assess the compliance of the facility with this ‘inertia emulation’ requirement are conducted, and the methodology that must be followed is described, as well as the main control algorithms that govern the operation of the PV system when providing the inertial response. In this work, the synthetic inertia support is provided by a PV system with no Energy Storage System (ESS) -or battery bank. This means that, in the face of a frequency decrease, if the generating unit is to be able to increase the active power injection, it will be necessary to “cap” its operation, or, in other words, to have it operating in a de-loaded control mode. Otherwise, if the system is already operating at its rated power -or at a setpoint near the available power, this increase in power will not be possible -or it will be possible only up to the maximum available power. In order to increase the theoretical basis of the paper, however, this study also shows a typical configuration of solar PV system equipped with an ESS, as will be seen later in the document. The PV installation analyzed has, therefore, the capability to contribute to power-frequency regulation in a network with an IR. However, as things stand, it is up to the Transmission System Operator (TSO) to give the final go-ahead for the PV facility to make a real contribution to the power-frequency inertial regulation in the Spanish market. In this sense, a major concern arises: currently, as the contribution to power-frequency regulation of PV power plants is not remunerated, what are the incentives for owners to have power conversion PV systems equipped with this technology? Indeed, in Spain, unlike in other European countries, such as France, Belgium, Netherlands, Germany, the United Kingdom or Denmark, and as detailed in [12], the actions carried out by the power plants to contribute to primary frequency control are not remunerated, although it is mandatory for them to have that capability. R. Villena-Ruiz et al. International Journal of Electrical Power and Energy Systems 157 (2024) 109875 3 Therefore, the answer to the question raised is not simple. Although not mandatory for the time being, there is no doubt that the inertia emulation requirement will be included in the new Spanish grid code in the near future. This is prompting PV power converters manufacturers to step up their efforts and be prepared to respond appropriately when this requirement becomes mandatory. 1.2. Literature review The scientific literature includes several contributions related to the synthetic-or virtual inertia requirement in PV installations or in other renewable energy power plants. For example, in [13], the authors conduct a literature review of the most up-to-date virtual inertia implementation techniques in Renewable Energy Source (RES) units. Primarily, the paper analyzes these techniques descriptively and at the control algorithm level. In [14], the authors also conduct a review, although, in this case, they do so by focusing on the inertia values of electricity systems in different countries, including also the damping factor evolution and a description of the most common frequency control strategies implemented in RES-based power plants, mainly in wind power plants. The work in [15] conducts, like the previous studies, a review of the strategies followed by power conversion PV systems to emulate inertia, presenting a clear summary of the main virtual inertia techniques implemented in these power electronic converters and their functioning. A number of papers have also been published on the scheduling of the operation of power systems with a high presence of renewable energy generation plants, which must be able to provide the system, as effectively as conventional power plants, with synthetic inertia, as well as ensure an adequate response during power-frequency control. In [16], the uncertainty in power generation by wind and solar PV installations is estimated using a deep learning approach. Furthermore, in the article, this uncertainty is taken into account when scheduling grid operation, and all this with the overall objective that these RES-based power plants are also able to provide inertia to the system, as well as frequency response support. The results of the article show how the operation of the test system analyzed, in terms of safety related to frequency control, is improved, and how, in this way, costs are reduced. Other interesting articles that address the topic of synthetic inertia in renewable installations include [17,18]. Both propose models to estimate the synthetic inertia and power-frequency control responses to be provided by the generating units analyzed, although in the case of [17] the authors focus on wind power. Thus, in [17], the inertia and powerfrequency requirements of an IEEE test system under different wind penetration levels are estimated in a daily framework -wind generation uncertainty is also characterized, as well as the synthetic inertia and power frequency responses that, under these scenarios, the wind power plants that form part of the test system have to provide. In [18], taking a microgrid as a basis for the study, the authors propose a model, which has the capability of providing synthetic inertia, that optimizes the operation of different distributed energy resources to meet the powerfrequency regulation requirements in the grid. The results show how the rate of change of frequency and the frequency nadir during frequency disturbances improve. The rest of the scientific articles related to the study of synthetic inertia in PV systems deal, for the most part, with the design of control strategies aimed at providing this type of IR, although they do so from different perspectives. The control dynamics of the Direct Current (DC)/DC converter and the DC/Alternating Current (AC) inverter that form part of a PV system, as well as the dynamics of the virtual inertial controller implemented, are analyzed in detail in [19]. The whole system is simulated using the MATLAB/Simscape Power System toolbox. The design parameters are modified to study their impact on the stability of the system. The main parameters established in different grid codes to assess frequency stability are considered in [20] to analyze the performance of a combined strategy for the virtual inertia and frequency damping control in PV systems. The validation of the control strategy is then validated by means of dynamic simulations. Other articles, also based on the design of control strategies in PV systems focus more, however, on the management of the power available in ESSs coupled to these renewable systems to provide an IR for network frequency support. For example, a synthetic inertia control in PV systems is proposed in [21] to deliver the required power through the use of a hybrid ESS, with the results being validated through simulations. The authors in [22] present a control strategy, based on the use of a supercapacitor as an ESS, to emulate inertia, and with the aim of also providing a primary frequency response. Meanwhile, in [23], the researchers propose a control strategy of supercapacitors to improve the stability of PV-based systems in standalone operation, analyzing the operation of the strategy theoretically through the use of the multidisciplinary and educational software tool MATLAB/Simulink. Lastly, the work in [24] also aims to design a control method to provide virtual inertia, although, in this case, the strategy is based on the use of the power available in the DC-link of the PV inverter. Other works, such as [25,26], analyze the operation of PV systems, focusing on the management of power reserves. The main goal of [25] differs from the aim of the present study, since PV systems are used, in that case, to design a frequency control strategy aimed at reserving power for grid frequency support, all of this through the prediction of the system frequency using a machine learning model. The results primarily focus on analyzing the frequency response of the system under different scenarios, and not on the active power response of the system itself. The work in [26] is in the same line as the previous one, since it focuses on the design of a power reserve control in PV systems for grid frequency support, modeling, as well, a specific strategy to allow the system to operate in different modes. The authors of the present work also have extensive experience in analyzing the response of solar PV systems to different types of disturbances, as well as studying how these types of systems provide primary frequency control and dynamic grid support. In [27], they implemented a positive-sequence model representing a utility-scale PV system -based on the generic model developed by the Western Electricity Coordinating Council (WECC), which is integrated into the IEEE 39-bus system. Both the PV system and the network are then simulated when subject to different contingencies under a variety of scenarios. In [28], the authors of the present contribution also implemented a utility-scale PV system, together with a hybrid ESS, to provide primary frequency control, including this time, however, both a statism response and an IR. The PV system modeled, also based on the generic model developed by the WECC, albeit including a number of new functionalities, is analyzed by means of transient stability analyses when coupled to the IEEE 39-bus transmission system. The work conducted by the authors of the present paper in [29,30] is focused on the evaluation of actual PV power plants in compliance with the technical requirements established in a national grid code; in particular, in the Spanish grid code. On the one hand, the compliance of a real solar PV power plant with a number of technical requirements -such as power-frequency, reactive power control and reactive power capability requirements included in the Spanish grid code is analyzed in [29]. The paper also delves deeper into the process of compliance of the renewable facility with the Spanish grid code. On the other hand, in [30], the authors propose the use of an aggregated PV power plant simulation model as a new tool to be used during the process of evaluating the compliance of new solar PV installations with the requirements established in the Spanish grid code. The aggregated model proposed is almost ten times faster than the one described, and allows the certification and commissioning processes of new solar PV power plants in Spain to be sped up. Nevertheless, there are noteworthy differences between the previous studies conducted by the authors and the present contribution. In the first two articles [27,28], a generic PV system, instead of a real one, is R. Villena-Ruiz et al. International Journal of Electrical Power and Energy Systems 157 (2024) 109875 4 simulated, and, additionally, only in [28] is the synthetic IR of the system analyzed. Moreover, the PV system is not simulated, in either of the two articles, according to any requirement included in any national grid code. In the second two articles cited above, [29,30], although the dynamic simulations are conducted according to the provisions of the Spanish grid code, the emulation of inertia is not a feature of the PV system under study. The present paper serves to continue the work already conducted by the authors in [27–30]. Thus, it widens the scope of [27,28] because the contribution seeks to analyze the response of a PV system from, on this occasion, a realistic perspective, and details how the inertia emulation requirement is considered in current national regulations. Finally, it can be stated that the present work also complements the work carried out by the researchers in [29,30], since it goes beyond the mere analysis of mandatory requirements and analyzes one of the most advanced requirements that could be demanded of PV power plants, which is not actually mandatory yet in any network code worldwide. 1.3. Contribution and organization of the study In light of the information discussed in the previous sections, there is clearly an urgent need to accelerate the development of international standards and regulations that require PV systems and other renewables to contribute to power frequency regulation in the network by emulating inertia. Additionally, it is also clear that the scientific articles published on the topic of synthetic inertia in PV installations mainly focus on the review of control strategies or the proposal of specific control methods –validated, in many cases employing a simulation tool. Hence, these papers have a high theoretical component, and the control approaches designed are not usually developed on the basis of any grid code, nor are the responses of the PV systems evaluated according to these national normative documents. Moreover, most of the papers published focus on analyzing how the frequency of the simulated network is affected by the IR provided, and not on the power response of the control device itself. Additionally, because of the very high practical applicability of the topics addressed in the study, this is a contribution of special interest to industry. Under this framework, the aims of the present paper are as follows: i) to provide the scientific world with a work that provides a real perspective on inertia emulation by renewable installations, that serves to complement existing studies in this field and that demonstrates that this is a critical requirement which, in the very near future, will be included in the grid codes of many countries; ii) to mitigate the lack of information about real experiences in the compliance evaluation process of PV systems with the synthetic inertia requirement established in a national grid code; iii) to present the methodology followed and the specific set of analyses required for an actual PV power plant to comply with this requirement and have the capability of actively contributing to frequency regulation in a power system by emulating inertia -based on the Spanish grid code; iv) to evaluate the current status of the synthetic inertia capability requirement included in grid codes worldwide, as well as the existence of mandatory or non-mandatory compliance; v) to highlight the challenges currently faced by renewable facilities’ owners and to identify future trends regarding compliance with the virtual inertia requirement included in grid codes; vi) to analyze the simulation responses of an actual PV system equipped with the necessary control strategy to provide an IR; vii) analyzing the response of the PV power plant as a whole when taking into account, in addition to the power conversion PV systems designed with the capacity to provide an IR, the PPC of the installation; and viii) to critically analyze the suitability of having a PPC when the power plant contributes to the power-frequency control by means of an IR. The paper is structured as follows: Section 2 describes how load variations affect frequency in a power system, introduces the concept of inertia and summarizes how frequency control is currently being carried out in power systems by both conventional and renewable energy power plants. Section 3 presents the methods followed to conduct the analyses performed in this contribution, presenting, as well, the configuration of the PV system under study and its control algorithms. Section 4 describes the results obtained when submitting the PV system to the synthetic inertia requirement included in the Spanish grid code. Finally, Section 5 summarizes the main conclusions obtained. 2. Power-frequency control and inertia in PV power plants In a system comprising a shaft connecting a turbine and a generator, with a constant mechanical torque applied, two situations could occur: i) if the load (or demand) were to increase, the electrical torque would increase, the system would begin to slow down and the electrical frequency would decrease; and ii) if the load were to decrease, the electrical torque would decrease, the system would begin to accelerate and the electrical frequency would progressively increase [31]. Under this framework, given that there are a large number of generators that contribute to electricity generation in a network, frequency control must be addressed globally throughout the power system. There is, thus, a need for a control system to regulate the mechanical power applied to the turbines connected to the synchronous generators in order to maintain the electrical frequency of the system stable, at all times, in response to any variation in demand. Hence, two main variables are intended to be controlled through the power production of the generators that are part of a power system: frequency and power flow through lines. In addition to balancing generation and demand, the power generated by each power plant must also be based on the functioning of the electricity market, which primarily establishes how much power each generating unit must inject and how much power should be exchanged between neighboring control areas [32]. Nevertheless, as can be deduced from a Primary Frequency Control (PFC) action [32] -also usually referred to as Frequency Containment Reserve (FCR), the power generated by a conventional generating unit is not immediately adapted to the new situation in which the load has changed. Therefore, in the event of an increase in the load, the amount of energy required instantly to compensate for this load increase comes from the kinetic energy of the rotating masses of the generators, the speed of which starts to decrease. This is the so-called inertial response. In this regard, the inertia constant, usually represented by the letter H, is a key parameter in electric power systems. In a generator, H is defined as the kinetic energy stored in the rotating mass at synchronous speed, divided by the rated power of the machine, as shown in Eq. (1) [33,34]. H=J ω 2 r 2Sr (1) where J is the moment of inertia (kg ⋅ m 2 ), ω r is the rated mechanical angular speed (rad/s), and S r is the rated apparent power of the generator (V A). H can be expressed in MWs/MV A, although if a change of base is made and the inertia constant is defined as a function of the base power of the system, it can be expressed in seconds (s) or in the per unit system (p.u.), as explained in [35,36]. The inertia constant indicates, intuitively, the inertia of a generator regardless of its rated power. Turning to the subject of the PFC, in conventional power plants, to contribute to this control action, the governors of the turbines act to control flow in the inlet valve and thus control output power. However, in most renewable energy power plants, which are electronically interfaced, there are neither turbines nor large synchronous generators. PFC is, therefore, not carried out through the regulation of an input flow rate, but through the operation of the power electronic converters that form part of the renewable generation devices. In the case of PV power plants, to contribute to power-frequency regulation, the droop control mode is followed, as in conventional power plants [34]. A droop –or statism– constant (R) must be configured in each PV power conversion system according to the applicable grid code. R is defined as the speed increase of the machine (Δ ω ) divided by the power output increase (ΔP), and R. Villena-Ruiz et al. International Journal of Electrical Power and Energy Systems 157 (2024) 109875 5 with a minus sign so that it is always positive, as can be observed in Eq. (2) [34]. R= − Δ ω ΔP(2) R is typically expressed in percentage. For instance, a statism value of 3% means that a decrease in frequency of 3% causes a 100% increase in generated power. In this way, the power-generating modules will be capable of activating the provision of active power frequency response when changes in this last variable occur [29], in the same way as conventional power plants do. Thus, when frequency rises occur in the network, PV power plants must guarantee a reduction of their active power output. On the other hand, when frequency drops occur, these facilities must guarantee an increase in their active power output, thus seeking to constantly balance generation and demand. In the case of Spain, as indicated in Section 1, MO TED/749/2020 [6] defines, among other things, the characteristics of the power-frequency technical requirements that renewable power plants must comply with, while the technical supervision standard [7] details the compliance process evaluation of the requirements and includes information on the parameters settings. In this regard, conducting dynamic simulation analyses is one of the tools allowed to check whether a PV power plant in Spain complies with the power-frequency technical requirements [30]. This is done as a step prior to the commissioning of the facility. Therefore, thus far, we have discussed the remarkable growth of VRE-based power plants over the past few years and the consequent reduction of inertia in power systems, highlighting the necessity for these renewable installations to contribute to grid stability in the same way as other power plants do. Moreover, the control mode followed by conventional generation facilities to contribute to power-frequency regulation has been explained, as well as the mechanism based on which PV power plants are intended to provide power-frequency support. In this regard, the concept of statism –or droop– has also been described, in addition to briefly discussing the importance of the IR of generators. Nevertheless, just as there is a clear distinction between the inertial response of conventional power plants and their PFC action -concepts that have been discussed in the previous paragraphs, it is also necessary to make this distinction in the context of renewable installations. In other words, in PV generation, it is also necessary to differentiate the change in the PV power conversion system output power due to the established droop, which causes the power to increase or decrease at a constant rate in the face of a load variation, from the IR that attempts to be emulated through the synthetic inertia requirement. This is designed so that the PV power conversion systems respond with increases or decreases in active power proportional to the time derivative of grid frequency. The difference between these two concepts will be shown clearly in Section 4, while the theoretical background behind the synthetic inertia supply capability of a PV system is discussed at greater length in Section 3.2. Summarizing, as inertia in power systems decreases due to the closure of conventional power plants, changes in demand have a higher impact, in the sense that frequency variations are reflected in the network more quickly; i.e., the ‘dynamics’ of the power system are faster. This forces the generation units to react more quickly to adapt to the new situation. All this highlights the importance of inertia in networks, as well as the need for new renewable generation plants to also be able to contribute inertia to the system. 3. Methodology The present section describes the methods used to conduct the analyses in this work. In particular, Section 3.1 reviews the main characteristics of the analyses to be conducted when assessing the synthetic inertia capability requirement in PV power plants according to the Spanish grid code, and includes the specific values that are set to carry out the simulations in the two scenarios analyzed, when the IR is studied at the power conversion PV system level and when studied at the Point of Common Coupling (PCC) and the plant is equipped with a PPC. Section 3.2 describes the configuration of the PV system and the control algorithms that allow both the PFC and the IR to be provided by the power conversion PV systems. Finally, Section 3.3 details the layout of the PV power plant –formed by PV systems, transformers and lines and equipped with the aforementioned PPC, the capability of which to contribute to power frequency regulation in the network by providing a synthetic IR is studied. 3.1. Synthetic inertia capability requirement of the PV system according to the Spanish grid code The NTS, which was released to monitor compliance with the technical requirements defined in Spanish MO TED/749/2020, also released a specific document with guidelines on drafting the inertia emulation report in those cases where applicable. Among other aspects, this document states that the report will only be delivered to the TSO by the generating units that have the capability of emulating this nonmandatory requirement. In this sense, the capability of providing synthetic inertia will become a mandatory requirement as soon as it is regulated as part of the balancing services of the power system. The compliance of the generating units with this requirement can be evaluated through dynamic simulations, which are required to be conducted only at the Power Generation Unit (PGU) level, unless the Power Park Module (PPM) owner declares the existence of higher order hierarchical control at the PPM level that may impact inertia emulation control. Thus, in this case, as the PPC also has an integrated inertia module -as will be explained further on, simulations will also be required at the PPM level. It should be noted at this point that in order to carry out the aforementioned ‘dynamic simulations’, a validated PV power conversion system model was used. That is to say, a model that faithfully represents reality with minimum errors is used; a model the behavior of which has been ratified and accepted by the Spanish Transmission System Operator (TSO), Red El´ ectrica de Espa˜ na (REE). PGUs refer to the main generation unit that forms part of the power plant, while PPMs refer to the power plant as a whole [29]. The simulation analyses are required to show the improvement of the unit response with the inertia emulation control enabled compared to the same simulations without it enabled. In no case should the powerfrequency control mode be deactivated. In particular, when conducting the dynamic simulations, the NTS states that two paths may be followed: i) performing the simulations based on the provisions of standard IEC 61400-21-1 [37]; and ii) performing an alternative set of simulations, proposed by the owner of the PPM and to be agreed with the TSO in advance. In our case, the first path has been followed. Standard IEC 61400-21-1 describes the synthetic inertia requirement as the ability of the PGU to support the grid by providing additional active power in the case of frequency events. This capability must be tested at the PGU while operating during continuous operation at different load levels of active power (P). The three following conditions are set while conducting the different tests: 1. At a partial load between 0.25 and 0.50 times the nominal power (P N ) of the PGU. 2. At a load level above 0.80 times the nominal power of the PGU. 3. At a load level below the nominal power of the PGU. The nominal power must be equal, in turn, to the available active power (P av ), i.e., to the power that could be generated depending on the primary resource available. From the above conditions it follows, therefore, that only in the third case the PGU is operating with a reduced active power reference or, in R. Villena-Ruiz et al. International Journal of Electrical Power and Energy Systems 157 (2024) 109875 6 other words, in a de-rated or de-loaded control mode. Standard IEC 61400-21-1 [37] also states that the previous load levels must be set to analyze the PV system response under two conditions, frequency increases and frequency decreases. It should be noted, at this point, that none of the normative documents states what the value of the frequency variation should be, but rather that it should be a sufficiently sharp variation to allow differences between having and not having the inertia emulation module enabled to be analyzed. Therefore, during the test cases, different inertia parameters must be set, as well as the test conditions. Then, the measured values from the simulation results must be evaluated according to the criteria included in Standard IEC 61400-21-1. Standard IEC 61400-21-1 regulates aspects related to the measurement and evaluation of electrical parameters in wind energy generation systems [38]. However, this standard is followed because no specific methodology has been developed for PV systems. Moreover, to conduct the simulations, the PGU is connected to a fictitious network that emulates the characteristics of the actual network. The power system simulation software tool used is DIgSILENT PowerFactory version 2022. More specifically, the synthetic inertia capability requirement is evaluated through the analysis of the time series of active power and frequency reference at the PGU terminals. Then, on the basis of the simulation results obtained, a number of parameters must be considered to assess the compliance of the PGU with the inertia emulation requirement. The most important ones are listed as follows, and are shown graphically in Fig. 1: - Reaction time. Defined as the time elapsed from the time the control signal for activation of the PGU response is sent until the amplitude change reaches 10% of the step height of the measured output variable. - Response time. Defined as the time that elapses from the beginning of the event -i.e., the change in the power response of the PGU until its value enters the predefined tolerance band of the target value. - Settling time. Defined as the time that elapses from the beginning of the event -i.e., the change in the power response of the PGU until its value remains continuously within the predefined value of the tolerance band of the active power response. - Predefined tolerance band. The tolerance band considered in the evaluation of the settling time is defined as 5% of the active power variation with which the PGU responds to the frequency change. Once the parameter has been properly set in the PGU and all the tests have been conducted, the assessment of some of the parameters listed above determines the compliance of the PV system under study with the synthetic inertia technical requirement. Thus far, no specific criteria or limits have been developed or set for the acceptance of the results. Thus, as is currently established in the Spanish grid code, only two parameters need to be taken into account for the compliance evaluation of the system with the synthetic inertia technical requirement: the system will be considered to comply if the reaction and response times are lower with the inertia module activated. In addition to the variables listed above, there are others the value of which will also be shown in the results section, once the simulations are performed. These are: i) default active power boost (ΔP), defined as the boost of active power when the frequency change occurs; ii) gradient of active power boost (dP/dt), defined as the rate of change in active power when the frequency change occurs; iii) steady-state time, defined as the time the system operates in steady state once the settling time has expired; iv) ramp-down time, defined as the time during which the active power value decreases from 90% to 10% of the target value; and v) recovery time, defined as the time from the end of the event after which the active power is continuously within the predefined tolerance band around the value of active power prior to the event. The time parameters just defined are also shown, graphically, in Fig. 1. Based on the general description of the synthetic inertia tests included in the Spanish grid code and explained in the present section, four parameters must be specifically set to conduct the simulations. These are the change and gradient of frequency applied, and the inertia trigger and recovery values, defined as the frequency thresholds where the PV power conversion system will start boosting active power and will stop boosting active power, respectively. In the case where the inertial response of the PV power conversion system is analyzed, frequency increases and frequency decreases are set to 1 Hz each. In addition, the gradient of frequency applied is set to 0.50 Hz/s in the tests under frequency increases, and to −0.16 Hz/s in the tests under frequency decreases. Finally, the values of frequency trigger and frequency recovery to assess the IR are set to 50.01 Hz and 49.99 Hz under the frequency increase and decrease tests, respectively. Table 1 summarizes the value of all these parameters (note that, in test cases 3 and 6, the system is operating in a de-loaded control mode, as mentioned previously). On the other hand, in the case where the inertial response is analyzed at the power plant level -and in this case it is necessary because the plant controller also has an integrated inertia module that can influence the overall behavior of the installation, the load level conditions established Fig. 1. Time parameters for characterization of the synthetic inertia response. Adapted from [37] R. Villena-Ruiz et al. International Journal of Electrical Power and Energy Systems 157 (2024) 109875 7 to conduct the tests have also been described in the present section. Moreover, the same types of simulations, with both the inertia module activated and deactivated, must be conducted. In this way, it will be verified whether the response time is lower or not with the inertial module enabled. In this case, the frequency increases and decreases are of 0.60 Hz each, the gradient of frequency applied is set to 0.30 Hz/s in the tests under frequency increases, and to −0.25 Hz/s in the tests under frequency decreases. The values of frequency trigger and frequency recovery to assess the IR are set to 50.01 Hz and 49.99 Hz under frequency increases and decrease tests, respectively. Table 2 summarizes the value of all these parameters (note that, as in the case of test cases 3 and 6, in test cases 9 and 12 the system is also operating in a de-loaded control mode, as mentioned previously). 3.2. Synthetic inertia supply capability: Control algorithms of the PV system The configuration of the PV system analyzed is schematically shown in the upper part of Fig. 2 (only the area with a white background); a system which has the capability of providing synthetic inertia. It should be noted, therefore, that the system studied in the present contribution is not a hybrid system, i.e., it does not have any ESS -or battery bank. However, the real converter tested is a two-stage power conversion system (more information about this type of systems can be found in [39]), since coupling an ESS is one of the most suitable solutions that are outlined in the future for solar installations to contribute inertia to the system, and because there are real prospects of this system being coupled to the power plant in the future. Moreover, this ‘typical’ ESS is also shown to provide the study with a greater theoretical background. Therefore, the system consists mainly of a PV array connected with the DC link through a PV converter, and a bidirectional Voltage Source Converter (VSC), through which the DC link is interfaced with the network. The dashed lines represent the control signals entering and leaving the control algorithms of the elements on which they act. As can be seen, and as it was explained above, in this case, the device tested is a two-stage PV power conversion system instead of a one-stage PV power conversion system [40–42]. The main reason for having this topology is so that in the future an ESS such as the one mentioned above can be coupled to it. It should be noted that this fact does not affect the final response of the system, nor does it modify the implementation or the control functions that are part of the inertial response algorithm. The configuration has two independent control subsystems (shown in the area with white background of Fig. 2). The two control algorithms are as follows: i) the first one (the PV Converter Control Algorithm) controls the PV converter -or boost converter to track the maximum power coming from the PV array and inject it into the DC link, and is also responsible not only for contributing to PFC but also for providing an IR; and ii) the second one (the VSC Control Algorithm) controls the DC link voltage and the reactive power flow between the DC link and the network, and it does so by controlling the VSC -or PV inverter. In particular, regarding the control algorithm implemented in the PV converter, it is worth noting that the device is not only able to contribute to PFC, but also to provide an IR. Thus, as mentioned, both the PFC and IR of the PV system are governed by the control algorithm implemented in the PV converter. This is shown graphically in Fig. 2 (please note the asterisk and the place where the acronyms IR and PFC appear). On the one hand, in reference to the PFC capability, the algorithm implemented complies with the requirements established in the Spanish grid code, and allows the PV system to respond to changes in frequency by increasing or decreasing its output power according to the statism –or droop– established. Eq. (3) shows the active power command provided by the PFC algorithm (ΔP P F C−cmd ) [28]. This control is shown, graphically, in the top part of Fig. 3, in dark gray. ΔPPFC−cmd = − |Δf| − |ΔfRI | R⋅fn ⋅Pref (3) where Δf is the frequency deviation, Δf RI is the frequency response insensitivity, P ref is the actual active power output at the moment the frequency sensitive mode (FSM) is triggered -i.e., at the moment the PGU’s capability to contribute to the PFC is activated, R is the statism and f n is the nominal frequency. Regarding the synthetic IR capability of the PV system, the algorithm is defined based on a derivative control, the performance of which is similar to those that conventional generators carry out [43]. Eq. (4) shows the active power command provided by the synthetic IR algorithm (ΔP IR−cmd ). This control is also shown, graphically, in the bottom part of Fig. 3, in light gray. ΔP1R−cmd = − 2H*⋅f*⋅df * dt ⋅Pref f2 n (4) where H* is the synthetic inertia constant –let us recall that the letter used to represent the inertia constant in conventional generators is H, as indicated in Section 2, with f * being the measured value of frequency. In Fig. 3, Δf, estimated as the difference between the reference frequency (f ref ) and f *, goes through a series of control blocks through which ΔP P F C−cmd is ultimately estimated, as previously indicated in Eq. (3). Moreover, within this algorithm, i.e., within the PFC capability algorithm, T P establishes the frequency measurement accuracy, Δf SS−max represents the maximum steady state frequency deviation, and LRC F SM is the limit rate of the change in active power due to the FSM. On the other hand, within the IR capability algorithm, T D establishes the frequency derivative measurement accuracy, T F 1 is the time constant of a low pass filter used to filter noise from the frequency data, and Table 1 Setting of parameters for test cases 1, 2, 3, 4, 5 & 6 to analyze the IR of the PV system. Frequency Increases Frequency Decreases Test Case 1 2 3 4 5 6 P (pu) 0.30 0.85 0.85 (P av =P N ) 0.30 0.85 0.85 (P av =P N ) Change in f Applied (Hz) 50–51 50–51 50–51 50–49 50–49 50–49 Gradient of f Applied (Hz/ s) 0.50 0.50 0.50 −0.16 −0.16 −0.16 f Inertia, Trigger (Hz) 50.01 50.01 50.01 49.99 49.99 49.99 f Inertia, Recovery (Hz) 50.01 50.01 50.01 49.99 49.99 49.99 Table 2 Setting of parameters for test cases 7, 8, 9, 10, 11 & 12 to analyze the IR of the PV power plant at the PCC. Frequency Increases Frequency Decreases Test Case 7 8 9 10 11 12 P (pu) 0.35 0.80 0.90 (P av =P N ) 0.35 0.80 0.90 (P av =P N ) Change in f Applied (Hz) 50–50.6 50–50.6 50–50.6 50–49.4 50–49.4 50–49.4 Gradient of f Applied (Hz/s) 0.30 0.30 0.30 −0.25 −0.25 −0.25 f Inertia, Trigger (Hz) 50.01 50.01 50.01 49.99 49.99 49.99 f Inertia, Recovery (Hz) 50.01 50.01 50.01 49.99 49.99 49.99 R. Villena-Ruiz et al. International Journal of Electrical Power and Energy Systems 157 (2024) 109875 8 LRC IR is the limit rate of the change in active power due to the IR control mode. Finally, as shown in the right-hand part of Fig. 3, both the ΔP P F C−cmd and the ΔP IR−cmd command signals are added, and the resulting value is the FCR active power command (P F CR−cmd ), which provides both a PFC and an IR as a consequence of a change in frequency [28,43]. Most of the power conversion PV systems currently only have the capability of contributing to PFC in the power system, and therefore they only have the PFC algorithm implemented. Thus, in this case, the power conversion PV system manufacturer ‘added’ a new functionality, which is the capability of providing an IR, thus making the power conversion PV system behave with increases or decreases in active power proportional to the time derivative of grid frequency. In the present contribution, the change in the output power of the PV system is determined, precisely, by the availability of active power, especially with regard to ‘active power to be raised’. In other words, when a frequency rise occurs, the PV system will be required to decrease its active power injection. In this case, where the active power response is ‘to be down’, it is self-evident that there will be no major problem, since it is always possible to reduce the active power setpoint of an electricity generation unit. In contrast, when a decrease in frequency occurs, the PV system will be required to increase its active power injection. However, if the generation unit is already operating at its rated power, a problem may arise, since no ‘extra’ active power will be available to comply with the requirements demanded. This means there may be times when the PV installation will have to be ‘capped’ to be able Fig. 2. Configuration of one of the PV systems analyzed (white colored area). The gray colored area represents how the system could be enhanced with an ESS. Fig. 3. Control algorithm in the PV converter. PFC and IR capability. R. Villena-Ruiz et al. International Journal of Electrical Power and Energy Systems 157 (2024) 109875 9 to contribute to PFC in the cases where frequency decreases. This is why it would be desirable to use ESSs in combination with PV systems equipped with PFC and IR capability: they would allow energy to be stored and would be ready to inject it into the network when frequency decreases occur. Therefore, the trend with regard to the participation of PV power plants in the PFC of a power system is towards the use of the state-of-theart converters in combination with ESSs, i.e., the trend is towards the deployment and operation of hybrid PV power plants. In this sense, and as mentioned above, Fig. 2 also shows, in the gray colored area, an ESS, which could be configured alongside the PV facility. The ESS is formed by a Li-ion battery bank and a supercapacitor (SC) bank. This is done to provide high energy storage and fast response time. The implementation of these hybrid systems would mitigate the effects of reduced inertia in power systems, offering power ramp rates higher than those usually required. In this case, in which both systems would operate jointly, there would be another algorithm that would control the ESS converters, managing the available active power of the ESS at the DC link and, hence, at the PCC, thus controlling network frequency and being responsible for the PFC of the hybrid system. Overall, this approach would operate the ESS while keeping the PV system in the Maximum Power Point Tracking mode (MPPT). 3.3. PV power plant with synthetic inertia supply capability and power plant controller In addition to analyzing the response of the PV power conversion system model that forms part of the PGU, the IR of this power electronic device is also analyzed at PPM level, i.e., considering, furthermore, the PPC of the PV power plant. The functions of the PPC are to record the network conditions, such as frequency, voltage, or active and reactive power demand, and, once these magnitudes have been recorded, its function is also to give instructions to the PV systems to act accordingly and generate the active and reactive power required. The behavior of the PPC is governed by a series of parameters set by each manufacturer, and their value is critical in establishing the resulting dynamics. In this case, the PPC also has an integrated inertia module, although not all do. Thus, the PPC, in the event of a change in frequency, in addition to indicating to the PV systems the new power to be injected, tells them how to inject it; that is, it establishes that the power injection must vary proportionally to the time derivative of the frequency. Therefore, measurements are taken at PPM level. The main objective of these tests is to assess the suitability of having a PPC equipped with the capability of sending control signals to the PV power conversion systems during frequency deviations in the network to contribute to PFC. The PV power plant, shown in Fig. 4, is formed by 12 PV systems of 1.637 MVA each -with IR capability, 12 Medium Voltage (MV) - Low Voltage (LV) transformers, and 1 High Voltage (HV) - MV transformer. 4. Results The present section is divided into two subsections. In Section 4.1, the simulation responses of the PV system model are analyzed when the inertia emulation module is enabled and disabled. In Section 4.2 the behavior of the PV power plant -equipped with a PPC at the PCC level is Fig. 4. Layout of the actual PV power plant tested. R. Villena-Ruiz et al. International Journal of Electrical Power and Energy Systems 157 (2024) 109875 16 [6] Ministerio para la Transici´ on Ecol´ ogica y el Reto Demogr´ afico, Orden TED/749/ 2020, de 16 de julio, por la que se establecen los requisitos t´ ecnicos para la conexi´ on a la red necesarios para la implementacion de los c´ odigos de red de conexi´ on. Tech. rep., Ministerio para la Transici´ on Ecol´ ogica y el Reto Demogr´ afico 2020. [7] Red El´ ectrica de Espana. Norma t´ ecnica de supervisi´ on de la conformidad de los m´ odulos de generaci´ on de electricidad segun el Reglamento UE 2016/631. Tech. rep., Red El´ ectrica de Espana 2021. [8] Sarasua JI, Martínez-Lucas G, García-Pereira H, Navarro-Soriano G, MolinaGarcía ´ A, Fern´ andez-Guillam´ on A. Hybrid frequency control strategies based on hydro-power, wind, and energy storage systems: application to 100% renewable scenarios. IET Renew Power Gener 2022;16(6):1107–20. [9] Sarojini RK, Palanisamy K, De Tuglie E. A fuzzy logic-based emulated inertia control to a supercapacitor system to improve inertia in a low inertia grid with renewables. Energies 2022;15(4):1333. [10] International Energy Agency. Photovoltaic Power Systems Programme, Snapshot of Global PV Markets 2022, Tech. rep., International Energy Agency; 2022. [11] International Energy Agency. Photovoltaic Power Systems Programme, Snapshot of Global PV Markets 2023, Tech. rep., International Energy Agency; 2023. [12] Lopez A, Ogayar B, Hern´ andez J, Sutil F. Survey and assessment of technical and economic features for the provision of frequency control services by householdprosumers. Energy Policy 2020;146:111739. [13] Tamrakar U, Shrestha D, Maharjan M, Bhattarai BP, Hansen TM, Tonkoski R. Virtual inertia: current trends and future directions. Appl Sci 2017;7(7):654. [14] Fern´ andez-Guillam´ on A, G´ omez-L´ azaro E, Muljadi E, Molina-García ´ A. Power systems with high renewable energy sources: a review of inertia and frequency control strategies over time. Renew Sustain Energy Rev 2019;115. [15] Yap KY, Sarimuthu CR, Lim J-M-Y. Virtual inertia-based inverters for mitigating frequency instability in grid-connected renewable energy system: a review. Appl Sci 2019;9(24):5300. [16] Kushwaha P, Prakash V, Bhakar R, Yaragatti UR. Synthetic inertia and frequency support assessment from renewable plants in low carbon grids. Electr Pow Syst Res 2022;209:107977. [17] Prakash V, Kushwaha P, Sharma KC, Bhakar R. Frequency response support assessment from uncertain wind generation. Int J Electr Power Energy Syst 2022; 134:107465. [18] Nema S, Prakash V, Pandˇ zi´ c H. Adaptive synthetic inertia control framework for distributed energy resources in low-inertia microgrid. IEEE Access 2022;10: 54969–79. [19] Khazaei J, Tu Z, Liu W. Small-signal modeling and analysis of virtual inertia-based pv systems. IEEE Trans Energy Convers 2020;35(2):1129–38. [20] Peng Q, Yang Y, Liu T, Blaabjerg F. Coordination of virtual inertia control and frequency damping in pv systems for optimal frequency support. CPSS Trans Power Electron Appl 2020;5(4):305–16. [21] Sarojini RK, Kaliannan P, Teekaraman Y, Nikolovski S, Baghaee HR. An enhanced emulated inertia control for grid-connected pv systems with hess in a weak grid. Energies 2021;14(6):1721. [22] Karpana S, Batzelis E, Maiti S, Chakraborty C. Pv-supercapacitor cascaded topology for primary frequency responses and dynamic inertia emulation. Energies 2021;14 (24):8347. [23] Yang L, Hu Z, Xie S, Kong S, Lin W. Adjustable virtual inertia control of supercapacitors in pv-based ac microgrid cluster. Electr Pow Syst Res 2019;173: 71–85. [24] Ramos JG, Araujo RE, Virtual inertia and droop control using dc link in a two-stage pv inverter. In: 2020 IEEE 14th International Conference on Compatibility, Power Electronics and Power Engineering (CPEPOWERENG), vol. 1, IEEE; 2020. p. 55–60. [25] Su Y, Li H, Cui Y, You S, Ma Y, Wang J, et al. An adaptive pv frequency control strategy based on real-time inertia estimation. IEEE Trans Smart Grid 2020;12(3): 2355–64. [26] Peng Q, Tang Z, Yang Y, Liu T, Blaabjerg F. Event-triggering virtual inertia control of pv systems with power reserve. IEEE Trans Ind Appl 2021;57(4):4059–70. [27] Bueno PG, Hern´ andez JC, Ruiz-Rodriguez FJ. Stability assessment for transmission systems with large utility-scale photovoltaic units. IET Renew Power Gener 2016; 10(5):584–97. [28] Hern´ andez JC, Bueno PG, S´ anchez-Sutil F. Enhanced utility-scale photovoltaic units with frequency support functions and dynamic grid support for transmission systems. IET Renew Power Gener 2017;11(3):361–72. [29] Martínez-Lavín M, Villena-Ruiz R, Honrubia-Escribano A, Hern´ andez JC, G´ omezL´ azaro E. Evaluation of the latest spanish grid code requirements from a pv power plant perspective. Energy Rep 2022;8:8589–604. [30] Martínez-Lavín M, Villena-Ruiz R, Honrubia-Escribano A, Hern´ andez JC, G´ omezL´ azaro E. Proposal for an aggregated solar pv power plant simulation model for grid code compliance. Electr Pow Syst Res 2022;213. [31] Ledesma Larrea P. Analisis dinamico y control de sistemas el´ ectricos; 2020. [32] ENTSO-E. 2021 Load-Frequency Control Annual Report, Tech. rep., European Network of Transmission System Operators for Electricity (ENTSOE); 2022. [33] NREL. Inertia and the Power Grid: A Guide Without the Spin, Tech. rep., National Renewable Energy Laboratory (NREL); 2022. [34] Zhang W, Cantarellas AM, Rocabert J, Luna A, Rodriguez P. Synchronous power controller with flexible droop characteristics for renewable power generation systems. IEEE Trans Sustainable Energy 2016;7(4):1572–82. [35] Kimbark E. Power system stability. John Wiley and Sons; 1948. [36] Vittal V, McCalley JD, Anderson PM, Fouad A. Power system control and stability. John Wiley & Sons; 2019. [37] IEC 61400-21-1. Wind energy generation systems - Part 21-1: measurement and assessment of electrical characteristics - Wind turbines; 2019. [38] Villena-Ruiz R, Honrubia-Escribano A, Jim´ enez-Buendía F, Sosa Avenda˜ no J, Frahm S, Gartmann P, Fortmann J, Sørensen P, G´ omez-L´ azaro E. Extensive model validation for generic IEC 61400-27-1 wind turbine models. Int J Electr Power Energy Syst 2022;134:107331. [39] Nanou SI, Papakonstantinou AG, Papathanassiou SA. A generic model of two-stage grid-connected pv systems with primary frequency response and inertia emulation. Electr Pow Syst Res 2015;127:186–96. [40] Cabrera-Tobar A, Bullich-Massagu´ e E, Aragü´ es-Pe˜ nalba M, Gomis-Bellmunt O. Topologies for large scale photovoltaic power plants. Renew Sustain Energy Rev 2016;59:309–19. [41] Xiao W. Photovoltaic power system: modeling, design, and control. John Wiley & Sons; 2017. [42] Hassaine L, OLias E, Quintero J, Salas V. Overview of power inverter topologies and control structures for grid connected photovoltaic systems. Renew Sustain Energy Rev 2014;30:796–807. [43] ENTSO-E. Network code on requirements for grid connection applicable to all generators, Tech. rep., European Network of Transmission System Operators for Electricity (ENTSO-E); 2015. R. Villena-Ruiz et al.