Reveal an Sustaaeble Ee Reviews 692017) 144-155, Contents lists available at ScionoeDirect Renewable and Sustainable Energy Reviews journal homepage: www.elsevier.comilocate/rser Inertia response and frequency control techniques for renewable energy sources: A review Mohammad Dreidy, H. Mokhlis’, Saad Mekhilef Departmen of Hc! Bapnerng, Pcty of ginerng, University of Malays, 0603 Kula Lap, Malaysia ARTICLE INFO ABSTRACT ergy storage nite (ESS) ‘requseyreplaton weseed cota Rae of ange of gure (ROCOM Renewable ergy soe (ES) Preservation ofthe eavironment has become the main motivation to integrate more renewable energy sources (WES2) in electra networks However, several lechnieal eres ate prevalent a high lve RES peneteation. The ‘most important echnical sve ithe dfs i achieving the frequency stably of here new sstems, a they ‘contain less generation units that provide reserve power. Mozeover, new power systems have smal inertia ‘onstant due tothe decoupling of the RESG fom the AC grid using power converters. Therefore, the RES in normal operation cannot participate with other conventional generation street in frequency regulation. This paper reviews several inertia and frequency contro techniques proposed for variable speed wind turbines and ola PV generators, Generally, the inertia and frequency relation techniques were divided into two rain groups. The fist group inludes the deloaing technigue, whieh allow the RESs to Keep a certain amount of Teserve power, wile the second group includes inertia exalation, fat power reserve, and droop techniques, ‘whi is used Io celeate the RES® reserve power at under frequency event 1. Introduction Recently, air pollutants generated by fossil fuel power plants, such as carbon dioxide, nitrogen oxide, and Sulphur dioxide are causing serious environmental problems (2). Acid rain and global warming are regarded a8 major causes of the environmental polation (2.3). Inthe United States, fossil fuel power plants emit about 2.2 billion tons of cathon dioxide (CO;) annually [4]. These problems forced governments and other agencies around the world to set targets ia increasing the application of RESS inthe generation of eletrical power (5). China, for terample, has seta target of generating over 15% of ie total power from renewable energy by 2020 with 420GW of hydro, 50 GW solar, 200 GW of wind, 30GW of biomass. As shown in Fig. 1, several counties set diferent future prospective targets in increasing power generation from RESs. These plans are crucial in order to address the ‘tremendous increase in world energy demand while simultaneously reducing the amount of pollations, Generally, integrated RESs in @ power system decreases depen- ‘dence on fossil fuel, improve voltage profile, and increase the reliability ‘of power system [7—10]. However, the high penetration of RESs can lead to critical frequency stability challenges (17). First, the RESS typically have low or non-existent inertial responses 12]. For example, the variable epeed wind turbines aze usually connected tothe network. by power electronic converter, which effetively decouple the wind turbine inertia from mitigating system transients. Ferthermore, solar wep do ores10103 photovoltai plants do not provide any inertia response to the power system. Therefore, replacing conventional sourees with RESS will reduce the inertia of the whole power system. This fact is supported by [13.14] both of which predicted that the increasing number of RESs in the UK could reduce the inertia constant by up to 70% between, 2013/14 and 2033/34. Due to this inertia reduction, the Rate of Change of Frequency (ROCOF) of the power system will be high enough to activate the load-shedding controller, even at a small magnitudes of imbalance, In (15), different penetration levels of RESs were used with a Synchronous Generator (SG) to cover 3.8 MW load demand, As reported in (15) and shown in Tig. 2, the ROCOF of the power system increase whenever the percentage-intaled capacity of the RES increases Second, an increase in the penetration level ofthe RESe decreases the number of generation units providing reserve power for primary and secondary control. For thie reason, the frequency deviation will be Increased, as reported in (16) and shown in Fig. % To overcome the frequency stability challenges represented by small inertia response and reserve power, RESS must create new frequency control techniques to allow them to participate in frequency regulation operations. This paper presents comprehensive review of Inertia and frequency control techniques for solar PV and wind turbines. These techniques enable the RESs to ineveasingly stabilize the power system, This paper is organized in the following order. Section 2 wil discuss the frequency response of conventional power 164032) 2016 Heir i, lies reserved2M. Dridy ea Renewable Energy Tans LOPS TESS ’ Fig 1. Renemble neay tages by eouney 6 2020 (5.5, enowae and Sustainable nergy Revie $9 2017) 1-155 rimy teat 2 tre, Ee Fos f Fig. 4 Timeframes inahed in sstem frequen respene [21 frequency start to decline with the frequency rates, depending on the total system inertia and the amount of unbalanced power, ae given by the swing equation [17,18 GmtsG OPV ase ae 30 g Pe P . 4 ta o a where dt isthe rate offequeney change, Hyys isthe total system gon inertia constant, Sis the rating power of the generator, Py, Pe ate the Bos mechanical power and electrical power, respectively, and fo is the ; system frequency. Prior to any controller activation and due to inertia poo response, the synchronous generator releases the kinetic energy stored 3 as in its rotating mass, which lats for ~10 (18.18), After tha, ifthe se frequency deviation surpasses specific value, the primary ffequency om a ‘0 1s controller will be immediately activated, This controller use the Installed capacity of RES ig 2, Mosimom ROCOF of the Moe for to types of RES spp 5 ed 6, — Hots a ffati] 2» 30 Tie) ig Frequency distin or mo diferent nrc 6). soureas, Section 3 will present several inertia and frequency control techniques for RESe without ESS. Section 3.1 will present several inertia and frequency control techniques for RES with ESS, Section 5.11 will explain the dierent soft computing approaches used with frequency regulation control. Section ‘1.1.1 wall discuss the advantage and disadvantages ofeach control methods, The conclusions and future research will be presented in Section 3.1.12. 2. Frequency response of conventional power sources ‘The general frequency response with operation limite correspond= ing to England and Wales are shown in Hig. % During normal ‘operations, the system frequency i close to 50 Hz, However, when fan even ovcurs that causes generation-demand unbalance, the system generator governor to return the frequency to save values within 30 s [19,20}. After 30, a new control called the secondary control wil be activated to return the system frequency to its nominal value. ‘As shown in Fig 4, the secondary controller needs several minutes, to recover the system frequency to its nominal value. Therefore, a reeerve power should be available to cover the increase in power demands during this period. Finally, the remaining power deviation activates the tertiary frequency control. Differing from primary and secondary controllers tertiary controller requires manual adjusting in the dispatching of generators or changes of the schedule periods. This paper does not deal with this type of controller [22). Generally, Inertia ‘and frequency control techniques for RESs is commonly divided into two main categories; control techniques for RESs without any support from ESS and control techniques for RESs with ESS. ig. 5 illustrates the dtferent techniques that fall under each eategory: 3. Control techniques designed for RESs without energy storage systems In order to minimize the negative impact of high RESs penetration, ferent inertia and frequency control techniques for RESs with and without ESS can be considered, These techniques enable RES, such at ‘wind turbine and colar PV plants, to contribute to the frequency regulation, 3.1. Wind turbine Wind energy is one of the most applied renewable sources throughout the world. Many countries that have wind energy potential started replacing conventional power plants with wind energy plants Statistics show that future wind penetration in the US. and Europe will exceed 20% within the next two decades (23), ‘There are two main categories of wind turbine; fixed speed and variable speed [25]. A fixed speed wind turbine generally usee an induction generator that is connected ditectly to the grid and can provide an inertial responce to the frequency deviation, even though this inertia is small compare to the synchronous generator. variable speed wind turbine mainly uses a Permanent Magnet Synchronous Generator (PMSG), or DFIG. The PMSG is fully decoupled from the2M. Dridy ea enowae and Sustainable nergy Revie $9 2017) 1-155 Teagus for RES ith I | Sole PY Pls Wot Tine Boxy Ste Syste | | EneeySerae Spo) + ¥ ¥ Delong || teeta Rapes Desading etoime |] resto ||P |] ost { | ian erin] [ Pastpower | [oversee ish Angle ig 5 net 0d egueney contr ecige desig or RES ig. 6. Power asian rotting speed carci (Pte age a) 23). sid; this i because the stator ofthis type of generator is connected to the power electronic converter in order to inject the power into the rid, The DFIG is similar to the PMSG, except for the fact that this generator is connected to the grid via a rotor circuit. The power electronic converter used in a variable speed wind turbine enables the ‘wind turbine to regalate the oulpat power over a wide range of wind speeds (25). However, this coupling isolates the wind turbine from the frequency response under disturbance. Additionally, traditional wind turbines follow the maximum power plot, as shown in Vig. 6. Therefore, they do not have reserve power to support frequency control, The sum output power from a wind turbine, defined asa fonction of ‘rotor speed, is given by [25,271 Parr = Kyo? @ where eis the rotor speed and Kp isthe constant (controller gain) for the tracking of the maximum pawer curve, and obtained by: @ where p isthe air density, isthe radius of the turbine wheel, CPope is the maximum power coefficient, Aype is the optimum tip speed. The maximum power point conteoller determines the operating point along, the power load line. This operation was conducted using the speed regulation of wind turbine within the speed limits and piteh regulation aller the rated speed Researchers have investigated two main techniques to support frequency control using a variable speed wind turbine, inertia response, and power reserve control. Inertia contr enables the wind turbine to release the kinetic energy stored in the rotating blades within 10's to arrest the frequency deviation, while reserve control technique ses the pitch angle controller, speed controller, or a combination ofthe two to enhance the power reserve margin during unbalanced power events 3.11. Inertia response control Unlike conventional generators that can automatically release the kinetic energy stored in their rotating mass, wind turbines do not have the same ability to release the kinetic energy stored in rotating blades For this reason, the wind turbine need a suitable contoller to provide inertia response. Generally, there are two control techniques that deal with inertia response; inertia emulation and fast power reserve. Inertia emulation is the frst technique; i proposes new control loops to release the kinetic energy stored in rotating blades of wind turbine Thie addtional power is uted to terminate the frequency deviation uring unbalance events. Fast power reserve is the second technique,2M. Dridy ea enowae and Sustainable nergy Revie $9 2017) 1-155 Fg. 7 Iberia eration fr vale ped wind tein (201 ‘which can also be used to terminate the frequency deviation. However, this technique responds to frequency deviations by releasing constant power for a set amount of time. 3.1.1. Inertia emulation, Using a power electronic converter with a suitable controller enables the variable speed wind turbines to release the kinetic energy stored in their rotating blades. Tis kinetic energy is ‘used as an inertia respons inthe range 2-6 s [2°]. Generally, there are ‘two types of inertia response; one loop inertia response and two-loop inertia response. In the former, one loop control based on the ROCOF is used to release the kinetic energy stored in the rotating blades, while ‘the latter uses two loops based on ROCOF and frequency deviations, In [21,30=32] the one loop inertia response is added to the speed control system to enable the wind turbine to respond to the ROCOF. This control loop is called inertia emulation, which exactly emulates the inertia response of conventional power plants as shown in Fs. 7, ‘The output power from wind turbine Praeas determine the refer ence rotor speed 2, ry; Which is compared to the measuring rotor speed ay, mens and used by the PI controler to provide maximum power, During normal operations, the reference power transferzed to ‘the converter is equal to the maximum power without any contribution from the inertia conteol loop. After a power deficit, a certain amount of power Py, based on the value of ROCOF and virtual inertia constant Hy, will added to the Pggnipr. Due to the power increment, the generator will slow down, and the kinetic energy stored in the rotating wind turbine blades wil be released. The additional power Pin comes from ‘the inertia response loop, which depends on ROCOF, and is given by [33 doy, Pe IHX tng x te re © Due to the constant aduitonal power resulting from the inertial control loop, thie type of contrl has two disadvantages. First, the rotor speed is rapidly reduced, leading to big losses in aerodynamic power. Second, the controller takes time to bring back the energy during rotor speed recovery. These disadvantages can be avoided, a8 per 3, where a new inertia response constant is formulated, This constant is called the effective inertia response, which is based on frequency value Generally, the inertia constant fora wind turbine is defined by: 2s Ss n o ‘where Fsin i the kinetic energy’ stored in the rotating mass ofthe wind turbine, $B isthe rated power, and J is the moment of inertia. Ea. (5) can be rewritten by substituting the corresponding power from Ea. (2). Then, the effective inertia constant will be ew)» 201 an'Cyop « © ‘The main ides isto increase the valu ofthe inertia constant as long asthe sysem frequency continues to decline, Consequently, the torque transfer to the converter is reduced, a shown in Tis. 8. ‘The principle of the one loop inertia response diseussed carlir is, aged Text Fig 8, Tongue desod do to nei spose) that it provides a decelerating torque signal proportional to ROCOF. This decelerating torque lasts until frequency is restored, Consequently, without support from another controller, the overall reference torque injected into the converte T.rec" wil be decreased by the maximum power point, which is working to bring back the system to the optimum curve. Asa result, the power injected into the grid will be reduced directly and bring back the frequency support immediately. In order to avoid this reacceleration of wind turbine, [35,18][18,35] proposed a two loop control inertia reeponee, at hown in ig. 9, Thie control provides an additional torque AT, which is proportional to frequency deviation and lasts until nominal frequency is recovered. The treo loop inertia response contrl system with two additional modifia- tions is presented in [36], A new block called delay speed recovery is added to recover the turbine speed as soon as possible. A wave iter is the other modification, which is adapted in the af loop to avoid constant value. In this paper, the author also discusses the effect of Giferent values of Ky and K, on system stability BLL. Fast power reserve. Generally, the inertia response ean be temulated, asthe control signal depends on the frequency deviation or ROCOF, as pointed out previously. It can also be defined as a constant 10% ofthe nominal active power for 10's, despite various wind speeds [37]. The shortterm constant power, which is called the fat power reserve, is released from the kinetic energy stored in the rotating mass of the wind turbine. This fast power reserve ean be achieved by controling the rotor speed set point. This is given by Raunt = bare o where Poonce, i the constant amount of active power, ¢ ie the time duration forthe fast power reserve, i te ital rotational speed, and o,¢ is the rotational speed atthe end of inertial response. Thus, the reference rotational speed can be obtained by: Pom or. ref= on = Jano — 2! y @1M. Dridy ea FL s pa 1 = ar aH, [ = come +f Fig 11, Hoc igri fat power seve conc Different works in literature have discussed the priniple operation ‘of the fast power reerve. References [38,29] diseused the capability of variable speed wind turbines to provide short term overproduction power for different wind speeds. In fact, these two papers did not provide any controller design for fast power reserve, However, (30) proposed a fast power reserve controller system for a wind turbine, as hom in Fig. 10. The amount of constant power and the time duration determines the rotor speed in accordance with Ea. (8). Ref. [40] proposed an architecture for a fast power reserve controlet, a showm in Tig. Ll, This figure contains a detecting and \ciggering scheme, power shaping, and an MPPT controller. ‘The operation of the fast power reserve controller begins once the frequency devition exceeds a certain threshold; a control signal is sent from the detecting and triggering scheme to bypass the maximum ‘power point tracking and enables power shaping, as shown in Fg. 12. ‘This scheme continues providing extra power during overproduction, However, when the kinetic energy discharge is complete, the speed recovery Tunetion brings the rotor speed back to its pre-event value, and restores the maximum power. Ths restoration often leads to the under-production phase, in which power is withdrawn from the grid to return the rotor speed tite desired value, To avoid an immediate drop in the output power, the transition from overproduction to under- production should follow a sloped transition Different strategies for fast power reserve for wind farms were proposed in [41]. The author discussed the operation of a centralized controler, which is responsible for frequency regulation. This central controller has two main tacks, the iret tack being to determine the amount of addtional power for each wind turbine, while the second taskis to determine the appropriate time to recover kinetic energy after ‘over production is finished Fig. 12, Powe characteris fo ft poets antl |). enowae and Sustainable nergy Revie $9 2017) 1-155 ae a ig 13, Paquencyropor schon with drop sped eet (4) 8.12, Droop contral The droop control scheme shown in Fig. 13 regulate the active power output from a wind turbine proportional to frequency change Thie controller greatly improves the frequency nadir ae well az the frequency recovery process following disturbances. The active power is adjusted according to linear characteristics, and ie given by [12-09] anu = Sa 4P= Pi Po o where Ris the droop constant, fneas and Py are the new frequency and wind turbine output power, respectively, while fnon and Py are the Initial operating points The linear relation between frequency and the active power of the wind turbine is ilustrated in Fig. 1. When the frequency falls from Snom 0 frneas the wind turbine increases the output of power from P, oP to compensate for frequency devitions 45. 5.1.3, Deloading control From an economic point of vie, wind turbines are designed to ‘operate on an optimum power extraction curve. As a result, they do not participate in frequency regulation. For this reason, sutficient reserve capacity must be available in the system to address any frequency deviation. Deloading is a new technique to ensure a reserve margin by shifting the wind turbine's operating point from its optimal power extraction curve loa reduced power level. Based on the wind turbines aerodynamic behaviour, the mechanical output povser captured by the wind turbine willbe ACP B) a) where pis the air density, is the rotor sweep are, vis the wind speed, Gp is the power cooffcient, fis the piteh angle, and Ais the tip speed ratio, which ie given by: ¥ ay From Eq. (10), itis clear thatthe ourput power of the wind turbine i dependent on the tip speed ratio A and pitch angle. In general, the Geloading technique has two types of control system; speed control and piteh angle control P, Powerpu) BR Fig, 14, Wind wie droop Arcee,1M. Dridy ea « enowae and Sustainable nergy Revie $9 2017) 1-155 xe 6 J page pie é 8 vn 2 — ern M16 tome cr ie vine pag LS oe crear aaa utr speal el) Fig 15.0) MPFT ap doeded power uve of wind whine (6) () Calelation 4.1.3.1. Deloading by speed control. Speed contro is suggested to change the value ofthe tip speed ratio Ay shifting the operating point towards the lft or the right of the maximum power point, as shown in Fig. 15 (a). This figure illustrates the deloading funetion of a 1.5 MW. DFIG-based wind turbine by (1-x) of the maximum power under definite wind velocity (Vi). The wind turbine running at point A ean bbe deloaded by the under-speed or over-speed control. For the under- peed contro, the operating point ofthe wind turbine moves towards point C, while for the overspeed contro, the operating point of the wind turbine moves towards point B, which is preferable. Referring to Fig 15(b), when the system frequency drops, the wind turbine releases a deGnite amount of active power proportional to ‘frequency deviation. Then, the operating point will be located between, A.and B with Preg given by: ui ~ Wen ay where Prnas is the maximum power (pu), Paris the deloaded power (PU), Or mane 35 the rotor speed at maximum power, ayaa is the rotor speed at de-loaded power, ois the rotor speed corresponding to reference power. Generally, deloading using overspeed control is preferably used at medium wind speeds. 3.1.32. Deloading by pitch angle control. Pitch angle is the second controller used to deload the wind turbine by inereasing the blade's angle. It is preferred that this controller is activated when the wind turbine generator arrives at the rated speed and when the over-speed controller fails to perform this operation Fig. 16 shows the power-rotor speed curve for a DFIG wind turbine under different pitch angles, This igure illustrates the deloading technique for a wind turbine running at point A; in this case, the controller fils to inerease the rotation speed over the rated speed. ‘The, the pitch angle controller begins to increase the angle of the wind turbine blades and shifte the operating point from point Ato point B without any change inthe rotor's speed, Generally, several different works inthe literature have dealt with the deloading technique, as itis used with a variable speed wind turbine, as per [7 The deloading technique supports primary frequency control under two operating conditions, as shown in 5g. 17. In normal conditions, the variable speed wind turbine works at the optimal power curve, extracting the operating point from the look-up table. However, when the deloading switch is turned on, the eloading mode will be activated, In this ease, the speed and pitch angle controllers will cooperate to allow the wind turbine to reserve some power under diferent modes. Eq, (12) determines the reference power for speed and pitch control to provide a 10% reserve power. To release the active power stored in the rotating maze due to deloading control, droop control is also presented in this work. The amount of releasing power is proportional to frequency deviations, and is limited to 10% of wind turbine rated power. [18] present the inertia response and primary frequency for DFIG- ‘based wind turbines. The inertia contreller is emulated to release the kinetic energy stored in the wind turbine rotating blades for a few seconds. ILis proposed that a deloading strategy with 90% sub-optimal power functions asthe primary frequency control. This strategy, bazed fn cooperation between the speed and pitch controller, provides the wind turbine with relatively long-term reserve power. Fis. 18 shows the eloading technique used with a wind turbine in three operating modes. Inthe first operating mode, the over-speed control is used to deload the wind turbine. For example, the deloading of the wind turbine running at point F by 90% sup-optimal power was done by ‘increasing the generator rotor speed towards point C. In the second ‘operating mode, the overspeed and pitch angle controller vere combined to achieve a certain sup-optimal power, For example, to deload the wind turbine running on point 8 with 90% sup-optimal power, the over-speed controller needs to shift the operating point towards point D. However, the over-speed controller increased the speed until the wind turbine arrives at point G, After that, the over speed controller will nat be able to increase the rotation speed any Tonger. Asa result ofthis, the pitch angle controller increases the Blade piteh angle, which shifts the operating point towards point A. In the {hird region, the piteh controller is used by itself to achieve the target deloading value ‘The cooperation between pitch angle and speed controller for a1M. Dridy ea enowae and Sustainable nergy Revie $9 2017) 1-155 Pitch Aue Cavole a4 Laakap ble Ped ompttion Troi & droop & ig 17. Pinay fnqveney glo for» DELC-bvd wind tcbne with deloaing one 57 fa, VaETan| rm Fig, 18.90% sub-optinal operon curve wi he oor ped into 89) variable speed wind turbine is also presented in (18). This paper ‘proposed three operating modes depending onthe range of wind speed. ‘The author proposed a decision algorithm to manage the cooperation between pitch angle and the over-speed controller. This algorithm determines the power set value for the pitch angle controller and the power margin for the over-speed controller. “Another study [50] used cooperation between pitch angle and over- speed controller to allow the wind turbine to participate in frequency regulation. However, this time, the controllers made their decision based on reserve power value depending on the network operator Fequest. In (91), the same frequency regulation controllers used for the _DFIG wind turbine were redesigned and implemented in the PMSG to enable this type of wind turbine to contribute to primary frequency contra Tm (52,53) the pitch and aver-speed conttllers, coordinated with Aroop control, were proposed. These controllers were activated based ‘on wind speed ranges to enable the DFIG-based wind turbine to contribute in frequency regulation. Moreover the over-speed control strategy using wind speed measure to determine the sub-optimal poser based on the deloading tracking curve is saved into the look up table 3.2, Solar photovoltaic array contributions in frequency regulation Recently, the penetration of solar photovoltaic PV into distribution networks has significantly increased. Asa result of this, reserve power ‘rom the remaining conventional source unit is insufficient to regulate system frequency under island conditions. Moreaver, due to the high cost of solar photovoltaic systems, different MPPT techniques have been introduced to extract the maximum power from ths souree [= 56). However, the use of MPPT techniques enable the solar photo- voltaic PV to operate without any reserve power. For these reasons, diferent modifications have been made to the design of controllers tused with a PV converter to allow them to effectively participate in frequency regulation “According to [57 smart photovoltaic inverters do not have the fall commercial control capability to change the output power fom PV systems, even if they have the ability to provide frequency down- regulation by curtailing power. Moreover, research related to this type ‘of control is stil in its early stages, and mainly rely on two types of controller. The frst type uses solar photovellaic PV supported by ESS to regulate frequency; this type will be discussed in Section TV (B), while the second type proposes the deloading technique for solar photovoltaic PV without ESS, as presented in [5850]. These papers present a comprehensive control stheme, which allows the photovoltaic system to regulate frequency. ig. 19 shows the deloading technique, which is achieved by increasing the PV voltage beyond the MPP voltage. Ths is achieved by inereasing the value from Varrr by voltage Vaetonds Which allows the PV array to maintain some reserve power. This reserve power will not be released until the system frequency deviates. Under these conditions, a conteol signal proportional to frequency deviation Vgcqris added to the de reference voltage Tis clerly shown in ig. 19 thatthe change in output power from the PV will not only depend on the Vsqr- value, but also depends on the frequency deviation, ae given by Bq. (13). Wéeref = Vurr + Vans ~ Yas as) The operation of the deloaded controller is ustrated in Pig. 20 where PV is working at point 3 to maintain some reserve povrer. The situation continues until the system frequency begins to decline, at ‘which point, a control signal related to frequency deviation will reduce the PV voltage, and make the PV work at point 2 Var Vasant 7 Fig 19, Center or dele slr PY (0)1M. Dridy ea ower a ears | Fig, 20, Solar PV with denang tla (6. ig. 21. Improved contre for dednaded PV which oni th amount of able reser 6 In fact, the controller disused in Fig. 19 has a big problem in that it does not take into consideration the remaining reserve power for ‘each PV unit. For this reazon, all PV unite will release same amounts of tetive power needed for frequency regulation, even of the reserve power of each unit will not be equal. As a result ofthis, some PV units with lese reserves vil reach the MPP faster and wil not be able to contribute any forther to frequency regulation. This will lad to the rnon-niform distribution of frequency regulation, Ref, [61] proposed a ‘new modification to the previous controller by adding « new contrl signal representing the remaining reserve power AVrererve: 88 show in Fig. 21. The reference voltage of the new controller is given by Ea. (14), which clearly shows that the output power released from the PV units i not equal, and depends on the reserve power available foreach Vaere ure + Vand — Yay) ~ (AP AV reserve x Kp) an Another technique proposed in (62) enables a solar PV plant to regulate frequency. Two algorithms were implemented; the first was the traditional MPPT conteoller, which ie responsible for operating the PY plant on MPP during normal operations. For transient conditions, a control signal would activate the deloading algorithm, which uses madiied fractional open circuit voltage. Thie modification proposed the usage ofthe ratio K asa controlled vatiable, which determines the amount of reserve power for PV plants limited tothe range (0.8-0 95), ‘The main finding ofthis paper shoves theta PV generator has the ability to regulate the frequency and follow load changes, In addition, (63) Aiscussed a control scheme designed for a PV panel to regulate the frequency of an islanded micro-grid. The main objective ofthis paper ‘was to use a tracking algorithm to follow a command signal, hich changes according 0 the frequenty deviation of the micro-grid Following this, the control system continues until the controller reaches the PMPP, after which the PV panel stays running on the PMP. Generally, all control techniques discussed in this paper were designed to provide the solar PV system with reliable control to regulate the frequency in grid connected or off-grid mode. These techniques were mainly based on the MPPT controller running, the enowae and Sustainable nergy Revie $9 2017) 1-155 Pig 22, Sole PV fngueey repute. PV array in the deloading mode. In contrast, [64] proposed a frequency regulator consisting of an adaptive frequency scheme using nonlinear control to caleulate the active power signal P*, depending on the frequency deviation, as shown in Fig. 22. This signal is needed to update the reference power Pres used by the power controller to determine the output power of the solar PV by regulating the duty ycle (D) of the power converter. 4. Control techniques designed for RESs supported by energy storage systems According to previous sections, diferent control techniques have been proposed to provide RESs with the abilty to regulate system frequency during disturbaneas. However, these techniques have some iscues in terms of reliability, as the nature of the RESs depends on variable factors. Therefore, ESSs are needed to work with variable speed wind turbines or glar photovoltaic PVs to inerease the reliability of frequency regulation, 4.1. Control techniques used for wind turbines supported by ESS In 65, cooperation between frequency control techniques and ESS was proposed for the DFIG wind turbine. This cooperation helps overcome problems of frequency control techniques, such as frequency ‘osellation and the second frequency drop. Tn fact, the ESS has two main functions in supporting frequency regulation in all wind speed ranges. In the fie function, the ESS provides the active power required for speed recovery to prevent the second frequency drop: while in the second, the ESS is regarded as a backup system for the provision of power during deficits. In [56], primary frequency control was used in wind power plants to maintain a certain level of power reserve. The wind power plant is supported by flywheel storage to fulfil the power reserve requirements tet by the network operator. In steady state conditions, a central ‘controller distributes the poster reserve requirement between the wind turbines and the flywheel storage system, as shown in Fig, 23. The power reserve margin x (in pu) is determined depending on the wind speed range, and is given by: Carnal sal ten oT Dinan ts powsT eS eens) yw wind urine. Oo fm Fig 25, Schur dagom of Gegueney eUltn of wind thine abd Abed,2M. Dridy ea Bib, < vgrated Es sty, > vyrated as) ‘where Pope is the maximum power extracted from the wind turbine and Petit the wind turbine power under deloading conditions 1n (67), a virtual inertia technique was proposed for the DFIG wind turbine to provide short-term frequency regulation. Since this techni- ‘ave focuses on short-term oscillation, there is no need for long power ‘regulation. For this reason, 2 super-capacitor is connected to the DC- link of the DFIG wind turbine inverter via a BC-DC converter. A comparison study done in this work showed that using the DFIG rotating mass or super-eapacitor asthe virtual inertia source improves system stability, However, each type has diferent impacts. Although the rotating mass virtual inertia does not require additional compo- nents, its performance is highly dependent on fluctuated wind speed. On the other hand, super-capacitor virtual inertia, which ean enhance system stability and is independent of wind speed, require new components 42. Control techniques used for a solar PV plant supported by an ESS In [68), the power modulation technique used for PV generation ‘output was described as using a double layer super-eapactor in the manner shown in Fig. 2%, This figure shows a PV generation system, consisting of a PV array, an inverter, and a euper-capactor. The arr _generates de power Ps, After that, the inverter converts the de power ta ae power P, and transmits this power to the uly via a service line. “The super-| Date Syren nea PA pw Pv amay Tee2M. Dridy ea ‘Tablet Adrntge and advantage of ier and frqaeney cote ehnigue for RS, ESS Tpeot— Teenique—Adntiges—Dindantges Can provide Depend on frend od enronoental Ireursey cones! eonkone Vasile util Poweretaced Second eqnmey ped wind Repose diel rom—_ep may be Cine Teemique —Foaing mae appeoed Deluding Proves primary Lovee wean Technique fequebey contol percentage of negy we soln PY Delong -iuelabe—Higheowt duet pian sen te price and Toes etic per miate the Pals te abaor power Dower fluctuations fom gig be ‘ery fl sare Vacsble etal Prevent scond High eat duet seed wind response feqeney op —_atery pice ad fuibine Belding ighlyrebale. some energy technique study should be done to find cheap and effective storage systems ‘such as hydro pumping storage, superconducting magnetic energy ‘orage, and electrical vehicle storage [78 7. Conclusions and future research In the near future, power systems networks will experience 2 significant inerease in the penetration level of RESe. For example, by 2020, renewable capacity willbe enough to supply an estimated 26% of flobal electricity needs [79]. However, integrating more RESs into [power systems wil educe the number of conventional generation units that provide primary frequency control and inertia response. For ‘example, the inertia response of NORDIC power aystem will be reduced bby 35% between 2010 and 2020 [80], For this reason, a new control scheme must be integrated with RES* to provide inertia and primary frequency control, This paper reviewed several ineitia and frequency regulation controllers for wind turbine and solar photovoltaic plants with and without ESS. Whether supported by ESS or not, the inertia and frequency regulation controller discussed in this work is suitable for the present penetration level of RESs, However, for future power system with high penetration level of RESs, futher studies must be conducted to develop an effective, intelligent, and robust primary frequency regulation schemes. A coordination between the primary frequency contol and frequency protection controller, such as load shedding controller, will be required, Furthermore, advanced comput ing methods and fast communication technologies wil be needed to realize the adaptive frequency control scheme, Acknowledgement ‘The authors would like to thank the Ministry of Higher Education of Malaysia for providing financial support under the research Grant NO: HIR-MOHE 00004-16001). References 11 Dong J, X46 6, bang, X8 reey-ving power seneation pati in hin resists pat projec an ply tecomnndtons-a tere. Reaoe {2} Suidur & Rabie N, tla M, Solangi K. Environmental impact of wind energy {3} Kyoto Protacl te the United Nations framework convention on eh ange enowae and Sustainable nergy Revie $9 2017) 1-155 Arn gsi 1 Ban hp it dn sed cn pon Sane dean a semen SSeS nity Seriado SES Sehr Iu ty ans ane eae Scumaes Snes ctoses ase "Sot e wr i fp NEL ae ei a ot ‘eprnretncesttoceonioe net ya tec) SS inlA Asal i neon sy 0, hans cnn of ap sted sn ‘sist naa stringy he ae he iE ma ays he Gan too) Sr ie dh Rare i gy pil ing deere te Re Gan Ee iy ut TE) Arie a |r ‘cc Se acm 9) 11 Gee yes ibe he tne. Weichert eta og) “oes yp Pe Rous Bae ai Srnec tes tne nad ess SEES melts hs roc mei She bc dan pt of ttn nt pve ‘pei loon we get ea prvi sage Co ee aa REST sda ian easier "5 Dy Bs, Re 2h. Re of nt a Panaby tage monn aeaeos espana te eee Gant en tet ce Simtel rc be tpn craps te ph Wa Toe gran veces [onrbon ci ske but donot se pene Fe 05 ‘Se et i nil de nd wind gion SR SEG ERA Spee me Kn an natn the medi of ind hin Wind win fu an ond in cs SEV TSE Rete olny ct oper te i alpen ny ER Ne eal See rowincon top 8 Sen er Cpr tn pn oh ed eta Gonzaler-LongattF, Chik K., Rashay: K. Kfets ofthe synthetic inecta from ipa ait seg re Frcs tod an Gon 68) lop ae peti alps BIST iover er ane we m © w 9} bn bey bs) bs} ol ey eal bs bal st es) bea ks) v9) (30) bn 2) (33)2M. Dridy ea [34] Wat, Inia 06. Tord an aecement of power ate esueney eippar fa Morren J, De Haan SW, Kling WE, Frees J Wind tries emulating neta and ‘based wind trbine. in: Proceedings of IEEE HCCE Asia Downunder (ECCE Asa); ‘Wael Beco: A Contblon of wind ene converts wi ea frlatos to eqn contd and reguesey Aub in poner sete ‘Proceedings af eB ntereainal worksop 8 eg see ategrain of wind wernt poner seme ae wal at 08 ochore wing fr, Brees, eran Soe {Usb NR. tire 7, KatoosD.Tenporary pena teguney concel poe, Srl ed ed iets ad appease Pons ‘erpeducn capably of are spend wind abies, Renee Eads ‘Hani, Anmalage UD, Jot Shorten eueney supper sing inertial Keng PAC Li, Basin H, Ol BT Kinetic ence of wndterbine goer fr fee equa sugpet IEEE Power St 2009:24279-87. Misha Sy Zaron Py SdbarP evel ental fr gurney reguin in « sistem wi igh PY penetration: Proeeingo KEE Power nner Sve General eetig (PES 2011 Soephine Rua tinting ASG win bins inet an dou cont sehen we iateliget fy eostle in Idan De Het Eg Teh (4s) Yao We Lee KY. conte conigraon af wind farm or london wd Frequency upper by eosin the ets eI Proceeds of SEE Power ry Genera Meeting 2011p BM Rah Re Sahar) Rat Al. Cnt mahade and objects for flees coupled eau energy rsutes a icroged rven ERE Vidvanandan K, Senoy Prosar requacy regulation by dlonded wind rbins ng Sin Un Corned age rainy ely |W, Gao W, Wang J Gu A cordate pena eguency regulation oe Permanent magn! syachonoa end trine generation Ls Proceedings of THEE Power Hecton ab Mach Wind pp Con (DEMWAY 2012 p= ‘hangie 6 Xuors W, nF Catrel sategy of urge DEC wed am for power gi eqn etn Ls IEEE Convl Conte (CCC) 2012518, Chien, 2012 9.885540 [50] Teens FD Res Seva Rer , Macnopals A Dae J regency ‘por by wind power plat inated pew varing enersion mI ‘roeedngs a IH Power nd ery Soc Conral Newig; 2017p. I= in Lave 3 Ener combate o MPPT chun fo PY ree [WSEAS Tan Power St 208564555, De Sto AC, Cabo 1, Sampaio LP, de Area Melo, Csnesin CA. “rakaton of the mas MPT techniques fer pote appctns I Proceedings a EHH Trans nd Heer, 2015 60 p1156067 Huse, shen ¢ dy of macro power ado edie ncn af OE envetelorphtooinie pee sen In Prcedags of TEBE Power Hcton ‘Specs Ceferene 1990 PEC 98 Recerd 29 Anal IEE, 1998p 0-92 Hoke A, baksmareD. Ace over costal fphotvale pone tema In Prowedings athe It ae Coneence HEH Technl ar Sasso Sure) Confrence, 2015, 2018p 70-7 81 es) 7 st bs} (0) ta (2) 1 81 6) wn ts) 1 (30) su (321 41 (551 bse 37) enowae and Sustainable nergy Revie $9 2017) 1-155 (58) Zaria, Mish Star. Decing inert respons rm enna PV ‘ye for ragheny regain In Proc IE nt Brie and Ear Se Cot (GPRDES) on EEE Power Seton, 2012, 1-5. otimans€, Castile A Fost teguecy response cps ofphotovalepover “ZarinaP, Misr, Sekhar P Photovoltaic sytem based transient mitigation and ‘Requney elton ns Preceding of he 2012 Anal THEE India Conternee NbIcoN 2012 p> ‘hn flr. alors Goteeyonlarbiy a FF te, Pappu VAR, ChowuryB, at plement aque regulation epaiy Intsela piotvle pone pln Im Proceedings of North Ame Powe ep (oat (NAPS) 2080p. Toe Frc! fk Ap ower Hon Gn nd tptin APE) 0 ‘koa Alico, BegunaneR Tarbouch M Nena onl seeing Soe on ET inant Cnfeenee on Pos Ess ad snd Dees (Gu 2012-1 2012p. 1. Mae, Wen J, Xw Ve © Lae W-, Corde ence rate of wid ‘ucine eter an eer sarge eqns oF egac ppot I Ap See Aaa etn: 2014 9 2-7 Teed aber ad yaar pinay eine cnt ogee Arai MPM, El-Seedany BF lmplerenting virus iets io DFIG-based wind Bhat R Chowdhury B. Gi frequency and voltage supper using PY systems with sty strge a: Proeeigsaf Noth er Power Sep Ctl NAPS) 2011. (hamana Me, Chomabury BH. DroopSaed eta in a potoraaiccenri mero wth Baer ery Storage Proeesigs of Nath Anes Powe Sp (oat CAPs); 2005 p.I-e Sha 1 F Courdated Vand #Q cota of sala photic generators, Se-ngawene Ny Nearoet Optinal uy lgi-based adaptive controller equipped ‘st DIG wind tine freq oul a sand done powet mse Proceeding of he 2015 IEE Inova Sse Go Techlogeae SET ‘Ada 2018 Nov 1p es Sent God 201239) 1955-4, [AL SQ, Herces #1, Preuency conte of oated network with wind and ee {eects by wing adap acl neural neta enrol Internatio ‘requeney conte. Energy Syst 201561) 109-2 Data erp, Yone A, Runaua Ki CH A eqn cont approach by Convers 201262) 359-71 ‘le teinforemsents for deeper ecxwsble generation lx ule bere sas “smart gr: a review on vebiele to id technologies ané optimization technsqus. evevabeHtegySedlamTery Marka Report. Avaale ea: seven ot ethene pl: 201 aetened 05.07 6] utr pte inertia pert European aeeck of amin ate pears for ele Ava ane ns esc s neces) Piblietins/SOC/Noctie Node port _Putre yee. etapa 203 (veesed 050716) (91 (ea te (ee) fest (eal us) test len tes) (61 0) ru val re val vs) 8) vs) v9) (90)