Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Optimal integration of battery energy-storage system with high penetration of renewable energy in radial distribution network

Optimal integration of battery energy-storage system with high penetration of renewable energy in... Abstract Considering the intermittent nature of renewable energy, a storage system to reserve power in off-peak hours and then to supply it during peak hours is necessary. However, if these storage devices in a network are not placed in an appropriate manner, the advantages can never be achieved. In this study, the allocation and sizing strategies of a battery energy-storage system (BESS) in an optimal way are proposed to improve the performance of the radial distribution networks. The test network adopted is a standard IEEE 33 bus network that is integrated with solar power. Simulations are carried out in DIgSILENT PowerFactory for the distribution network model and the process of optimization is done through MATLAB®. The optimization algorithm adopted is the genetic algorithm. The system is studied for 24 hours with a step size of 15 minutes, employing a time-sweep analysis of dynamic load. It is studied under three different scenarios and the results are compared. The results illustrate a considerable reduction in power losses in the case of the optimized BESS. The total losses before the optimization process are 16 365.57 kW. The total losses after the employment of the proposed solution are 10 246.5 kW, which means that the losses are reduced by ≤62%. Open in new tabDownload slide energy and environment, energy storage, solar Introduction These days, great interest has been aroused in reliable and flexible energy integration. It is a great challenge to balance power generation and consumption considering the stochastic nature of renewable energy [1]. Renewable energy is considered a path towards a clean and less polluted environment. It is seen as a possibility to reduce CO2 emissions. The degrees of atmospheric pollution along with the sources of air pollutants have increased essentially over the years [2, 3]. At the same time, numerous research studies have been done on solar energy integration to meet the renewable-energy target [4]. Different policies for renewable energy-generation strategies have also been launched by some governments [5]. Furthermore, it is expected that by the end of 2025, the levelized cost of electricity of solar energy could fall by 43% [6]. Few research studies have also been carried out for loss minimization; however, a renewable energy source was not considered in the network [7]. Despite the environmental and sustainability benefits of renewable energy, one of the most important problems is the management of power. Mainly, the photovoltaic (PV) system generates power depending on the local insolation level according to that area and the time of day. Especially in local electricity grids, in which the percentage of generation from solar is higher than that from other conventional generating technologies, the management of power becomes a challenge. The behaviour of the grid is disturbed due to these non-manageable energy-generating technologies. If the percentage of generated power through PV is low, the disequilibrium between production and consumption is not very problematic but if this percentage increases, the variability of electricity produced and its consumption become considerable matters. This intermittent power becomes the reason for the destabilization of power. Therefore, the management of electricity produced by such intermittent sources is under great consideration these days [8, 9]. One of the possible solutions suggested by the literature is to use hybrid renewable-energy systems for this issue of intermittency, to integrate different renewable-energy resources in an optimal manner [10]. In this system, sources with lower generation during a period can be compensated for by sources having a higher generation level during that period [11]. Thus, system efficiency and supply reliability can be increased [12]. The optimal sizing of generation units is reviewed in detail in [13] and different control strategies for renewable energy in [14]. The other solution suggested by the literature is the battery energy-storage system (BESS) [15]. The energy-storage system (ESS) is the key technology for the reliable integration of renewable energy resources [16]. The surplus amount of energy is stored in the ESS, which is fed when consumption is higher than generation. If the ESSs are not placed in the correct manner in a distribution network, the benefits from the ESS cannot be achieved [17]. As a result, optimal sizing and allocation of the BESS have become crucial areas of research. However, in this study, the allocation and sizing strategies of the BESS in an optimal way are proposed to improve the performance of the radial distribution networks. The problem of the optimal sizing and siting of the BESS in the distribution network has been significantly studied in the literature [18]. In [19], the authors used a modified non-sorting dominated genetic algorithm (GA) for the location and size of distributed generation (DG) optimally in distribution networks (DNs). The objective function is power-loss minimization, which improves the voltage profile of the system. This proposed methodology was tested on an IEEE 33-bus system. Misplacement or misuse of energy-storage devices can lead to several network performance issues in terms of voltage and frequency regulation, power quality, reliability and load controllability [18]. To achieve the optimal operation of an ESS, an appropriate way of placement is required for improvement in voltage and power quality [19, 20], peak-demand reduction [19], relaxation in distribution congestion [18], power-flow control [18], power-loss minimization [21], network security and cost reduction [8, 9], RES integration [10] and system effectiveness. There are several variables that are defined to approach the optimal sizing and siting of a BESS in a distribution network, such as optimizing technique, performance metrics for optimum evaluation, battery technology, and modelling and testing of the network. After a successful analysis of resources and estimation of solar production, a system needs to be developed to maintain the balance between generation and consumption patterns [16]. Intermittent PV production throughout the day can be smoothed by including ESSs and filling in the gaps. The smooth production of renewable energy throughout the day can be achieved through ESSs. They have the ability to handle ramps or the frequently known duck curve [7]. The ability to handle the load ramp cannot be done through traditional conventional generators but it can be handled very quickly through energy-storage devices [22]. The BESS has attracted considerable attention from researchers because of its advantages such as speedy response, controllability, energy management and reliability improvement [5]. Therefore, locating grid-scale ESSs becomes important in improving distribution network performance and minimizing the issues. If not placed in the correct manner in a distribution network, the benefits from the ESSs cannot be achieved [23]. However, in this section, an extensive review of the literature on optimal allocation and sizing of batteries is presented. The term ‘optimization’ means to reach the established goals with the help of available resources through a set of processes. The general optimization process algorithm for optimal sizing and placement has the following steps: (i) Identify the problem to be analysed in a network and the support solution provided by the ESS or we can identify the major objective, such as peak shaving, loss minimization, power quality and reliability, etc. (ii) Formulation of the fitness function and necessary constraints related to the optimal sizing and placement of the BESS. (iii) Evaluate the fitness function using optimization techniques such as analytical or meta-heuristic approaches. (iv) Update the fitness value and current best location and size of the BESS. The general flow chart of the heuristic optimization approach and the utilization of GA in the optimal allocation are shown in Fig. 1. Fig. 1: Open in new tabDownload slide (a) Flow chart of GA optimization approach. (b) Illustration of GA utilization. The objective of this research is to reduce the power losses in a radial distribution network given the constraints on power balance, bus voltage, energy-storage power, charge/discharge balance and remaining capacity of the ESS. The optimal placement and size of the ESS are determined by the GA optimization method. Several GA approaches have been utilized in other studies [15, 17]. These studies highlight the effectiveness of GA compared to other stochastic methods based on the processing duration and error percentage of the solution. In this study, DN data and BESS data were fed as inputs into the GA. The algorithm processes these data through many iterations and the resultant is the optimal place for a BESS that assures the minimum power loss in the distribution network. DIgSILENT PowerFactory (DPF) provides solutions for distribution network problems such as system design, modelling and optimization capabilities, grid interaction skills in a multi-user environment and data handling [24]. MATLAB® is used to control the system models developed in DPF and to facilitate genetic optimization. Furthermore, electricity production from renewable sources such as solar energy varies with changing solar radiation; thus, power generation is a function of time. Performing load-flow simulations during a certain time (in this case, a 24-hour period) in the different nodes of an electric power system requires the use of quasi-dynamic simulation, a time-varying load-flow calculation tool. Different researchers suggest that quasi-dynamic methods, as compared with quasi-static methods, not only improve the fidelity of the simulation of the process, but also reduce the processing time of the dynamic simulations. The main contributions of this paper are summarized as follows: An optimal BESS placement is carried out based on GA to significantly reduce the overall system power losses associated with the integration of high penetration levels of solar PVs. Although a similar investigation was carried out in [25], a unity power factor (p.f.) approach is applied on the ESS dispatch (i.e. the ESSs only inject P to the network). In this study, however, the ESSs inject both the P and Q to the network for better performance improvement with variable p.f. on the dispatch of an ESS. The effectiveness and robustness of the proposed methodology are comparatively tested and validated by different approaches on IEEE 33 bus systems. The first approach studied assumes that only the PV is contributing to the network and the voltage at the farthest node, such as bus no. 18. These values of voltages are later compared with the proposed network, such as after installing the BESS at the optimal place. The second approach is that active power losses are studied with and without the employment of the BESS of the overall DN. Overall, active power losses are reduced after the employment of the proposed strategy. 1 Methodology 1.1 Quasi-dynamic simulation A quasi-dynamic simulation is offered by DPF for the execution of medium-term to long-term electrical studies. It performs several load-flow calculations with user-defined time-step size. This tool is useful for dealing with studies in which long-term generation and demand profiles are defined and network models are developed considering expansion stages and variations [26]. The main benefit of this type of simulation is that it provides faster calculation results because it does not require solving all the mathematical expressions. Other researchers [27, 28] have suggested that quasi-dynamic simulation language offers better accuracy than other methods such as quasi-static. Quasi-dynamic simulations have been used in electrical studies into units such as circuit-interruption devices, PV energy systems, ESSs and distribution network studies. In this work, the ‘Quasi Dynamic’ tool (Fig. 2) is used to analyse the distribution network with varying demand in order to determine the behaviour and changes in system dynamics when some nodes in the system include solar DG. Fig. 2: Open in new tabDownload slide Quasi-dynamic simulation window. By modifying the load, each node in the system has a different demand that varies moment by moment. This new behaviour is analysed under different scenarios. The load flow is calculated every minute for 24 hours in order to identify the changes in the network variables such as power losses and voltage profiles. In order to analyse the simulation results, the IEEE 33 bus system was used as a test system model (Fig. 3). Fig. 3: Open in new tabDownload slide IEEE 33 bus system. 1.2 Load modelling The proposed load model is the IEEE 33 bus test system. It is designed using DPF. The system consists of 33 buses and 32 nodes with a current-carrying capacity of 400 A from Node 1 to Node 9, whereas other lines have a capacity of 200 A. The base voltage is 12.66 kV and the base power is 10 MVA. The total active power load is 3715 kW and the reactive power load is 2300 KVAr. Since the base system is modelled with a firm power load, it is required to simulate the load variations over time to study the dynamic nature of load under a 24-hour period. The quasi-dynamic simulation is used to calculate the set of time-dependent load-flow calculations. The standard Bundesverband der Energie- und Wasserwirtschaft load profile is utilized [30]. The basic system parameters are shown in Table 1. Table 1: 33 Bus system parameters Parameter . Value . Unit . Voltage level 12.66 kV Frequency 60 Hz Rated active power from external grid 3917.7 kW Rated reactive power from external grid 2435.2 kVAr Rated total active power losses on lines 202.7 kW Parameter . Value . Unit . Voltage level 12.66 kV Frequency 60 Hz Rated active power from external grid 3917.7 kW Rated reactive power from external grid 2435.2 kVAr Rated total active power losses on lines 202.7 kW Source: Ref. [29]. Open in new tab Table 1: 33 Bus system parameters Parameter . Value . Unit . Voltage level 12.66 kV Frequency 60 Hz Rated active power from external grid 3917.7 kW Rated reactive power from external grid 2435.2 kVAr Rated total active power losses on lines 202.7 kW Parameter . Value . Unit . Voltage level 12.66 kV Frequency 60 Hz Rated active power from external grid 3917.7 kW Rated reactive power from external grid 2435.2 kVAr Rated total active power losses on lines 202.7 kW Source: Ref. [29]. Open in new tab 1.3 Solar PV modelling There are a few factors that affect the generation from PV models, such as rated power, solar irradiance and temperature. The PV system generates power during the day. The output power of the solar module is shown in Fig. 4. It is generated based on Karachi city solar irradiance measured on the hottest day of the year, which is 21 July. The maximum output power is ~2.2 MW at 1 p.m. The number of parallel inverters is 20 000. The watt/panel is 110 watt so 110 × 20 000 = 2.2 MW. Integration of PV systems in DPF requires basic data for the network model. The type of PV panels is Aleo 150-L. The basic data for the PV model are shown in Table 2. Table 2: Basic data of PV model Name . Aleo 150L . Peak power (MPP) 160 W Rated voltage (MPP) 23.4 V Rated current (MPP) 6.8 A Open-circuit voltage 30.5 V Short-circuit current 7.2 A Material Single-crystalline silicon (mono-Si) NOCT 47℃ Name . Aleo 150L . Peak power (MPP) 160 W Rated voltage (MPP) 23.4 V Rated current (MPP) 6.8 A Open-circuit voltage 30.5 V Short-circuit current 7.2 A Material Single-crystalline silicon (mono-Si) NOCT 47℃ MPP, maximum power point; NOCT, nominal operating cell temperature. Open in new tab Table 2: Basic data of PV model Name . Aleo 150L . Peak power (MPP) 160 W Rated voltage (MPP) 23.4 V Rated current (MPP) 6.8 A Open-circuit voltage 30.5 V Short-circuit current 7.2 A Material Single-crystalline silicon (mono-Si) NOCT 47℃ Name . Aleo 150L . Peak power (MPP) 160 W Rated voltage (MPP) 23.4 V Rated current (MPP) 6.8 A Open-circuit voltage 30.5 V Short-circuit current 7.2 A Material Single-crystalline silicon (mono-Si) NOCT 47℃ MPP, maximum power point; NOCT, nominal operating cell temperature. Open in new tab Fig. 4: Open in new tabDownload slide PV output curve. 1.4 BESS modelling The ESS is simulated using the built-in static generator unit that is used to model any non-rotating power-generation and storage unit that is connected to the AC grid through an inverter. The use of this element is an appropriate approximation and it is also supported by DPF in which ESSs are represented by static generators. To implement this time-varying system, the time-dependent load-flow script ‘time sweep’ is utilized. The objectives achieved by this framework are: appropriate objects on the grid topology such as PV units, energy-storage units and time-dependent loads are identified; specific system parameters are then read to perform the time-dependent load-flow analysis later to extract the results; the adopted load-flow script analyzes the name of storage units and its capacity to calculate the state of charge with user-defined time-step sizes. In this research, it is assumed that the BESS is fully charged at the start of the day and discharges gradually as time goes by[31]. In order to simulate a battery storage unit over a period of time, the unit is modelled considering the following three parameters: (i) maximal charge and discharge power; (ii) charge–discharge efficiency; (iii) storage capacity. The battery storage unit is used for the modelling of a BESS. The active power rating of the battery is 1.1 MW. 1.5 Optimization method GAs are optimization techniques used to solve non-linear or non-differentiable optimization problems. GA uses concepts from evolutionary biology to search for a global minimum. The name ‘genetic algorithm’ comes from the fact that the algorithm is mimicking evolutionary biology techniques. GA works by starting with an initial generation of candidate solutions that are tested against the objective function, then generating subsequent generations of points from the first generation through processes such as selection crossover and mutation. The GA runs its iterations until it converges to the best objective function, i.e. when the fitness function value is no longer changing or it is changing by a really small amount. This paper employs GA as an optimization method to optimally size and site the BESS in a distribution network. This process is done using MATLAB® programming. The system summary from DPF is synchronized with MATLAB® for the process of optimization. It is assumed that the PV system only supplies active power and is placed arbitrarily at buses. The objective function is to reduce line losses, formulated as: Ploss=min∑Ttotalt=1∑nbri,j=1 i≠j g i,j(Ui2 + Uj2 − 2UiUjcoscos (θi − θj))(1) where Δt represents the time interval; ttotal represents the total time period considered, gi,j represents the conductance between buses i and j, nbr represents the total number of branches in the network, Ui and Uj represent the voltage magnitudes of the buses i and j, and θi and θj represent the voltage angles of buses i and j. Equation (1) interprets the expression for total network power losses based on the AC power-flow equation. For each and every evaluation of fitness function, the non-linear AC power flow is calculated. In addition to the above objective function, the optimal network configuration has to satisfy the following operational constraints: Uimin≤Ui≤Uimax(2) Sb≤Sbmax(3) where Uimin represents the minimum voltage magnitude for i buses, Uimax represents the maximum voltage magnitude for i buses, Sb represents the actual power flow and Sbmax represents the maximum actual power flow. Uimin and Uimax represent the voltage limit of each bus and Sbmax represents the maximum current level of each branch b in terms of actual power flow. The whole search process of the proposed method for determining the location and size of the BESS to minimize total active power loss is described in detail as follows: Step 1: Define the input data (Vrated, RL, XL, PV and MVAbase) to simulate the studied distribution network in DPF and evaluate candidate buses to find the best location for the BESS installation. Step 2: Initialize the parameters of the algorithm. Step 3: Select one candidate bus as the location for the BESS installation. Step 4: Execute the algorithm (as in Fig. 1) to solve the optimization problem. Step 5: Conduct a load-flow process at the current candidate bus at which the BESS is installed. Step 6: Evaluate the objective function at that point. Step 7: Update the best solution. Step 8: If the maximum iteration is reached, calculate optimal coefficients (Step 9); otherwise, continue optimization (Step 4). Step 9: Obtain optimal coefficients at the current candidate bus at which the BESS is installed from Step 3. Step 10: If the candidate bus is the last candidate, proceed to the optimal BESS installation (Step 11); otherwise, select the next candidate bus and execute the algorithm (Step 3). Step 11: Determine the optimal siting of the BESS installation, which is the candidate bus providing the minimum value of the objective function, such as reduced power losses. Step 12: Substitute the obtained coefficients in the BESS unit. 2 Results and discussion The results were obtained using DPF for model simulation and MATLAB® for the optimization process. The standard IEEE 33 bus system is taken as the base case and its data are shown in Table 2. The data are modified to create a dynamic load demand and system dynamics are studied for 24 hours with a time-step size of 15 minutes (Fig. 5). There are three cases in this study. The system load is dynamic in all three cases, which is 72.28 MWh. The load is maximum at around 9.30 to 11.30 a.m., as shown in Fig. 5. The first case is the base case that includes no PV and BESS. The total losses in that case were 16 365.57 kW. Fig. 6 shows the total active power losses in a complete day of 32 branches. Fig. 5: Open in new tabDownload slide Load profile of 1 day. Fig. 6: Open in new tabDownload slide Active power losses base case. The total losses after the employment of the proposed solution are 10 246.5 kW, which means that the losses are reduced by ≤62%. The reduction in total active power losses in each case are shown in Figs 7 and 8. Fig. 7: Open in new tabDownload slide Total losses in each case. Fig. 8: Open in new tabDownload slide Losses in kW. The voltage profile of each bus in the base case is shown in Fig. 9. As the voltage reduces at a node with respect to an increase in the distance from the supply side, only the farthest node is studied in this research, which is bus no. 18. The voltage at bus no. 18 in each case is shown in Fig. 9. It is evident that the voltage has improved after the employment of the proposed solution. The voltage between the midnight hours has improved by 0.02 p.u. Hence the overall system voltage has been improved by 97% and it is under the stability limits. Fig. 9: Open in new tabDownload slide Voltage magnitude at bus 18 in each case 3 Conclusion In this research, a GA approach was utilized to find the optimal size and placement of a BESS integrated with PV and a grid in the IEEE 33 bus system. The overall system active power losses are reduced and the voltage is improved after the employment of the proposed solution. Several cases have been established to verify the results compared to the test system. The simulation results reveal that the overall system losses under dynamic load without the usage of BESSs are 16 365.57 kW and after the placement of distributed BESSs they are reduced to 10 246.5 kW. Therefore, losses in the proposed solution are reduced by 62%. The improvement in voltage is also justified in the results by the use of BESSs. Intermittent renewable-energy systems, especially at higher levels of penetration, can also be made stable by this application of BESSs. For future work, the proposed solution can be tested for other renewable-energy integrations as well, such as wind, fuel-cell, etc. The objective function can be expanded as a multi-objective optimization approach and this study will contribute to the efforts. Conflict of interest statement None declared. References [1] Yi W , Zhang Y, Zhao Z, et al. Multiobjective robust scheduling for smart distribution grids: considering renewable energy and demand response uncertainty . IEEE Access , 2018 , 6 : 15 – 24 . Google Scholar OpenURL Placeholder Text WorldCat [2] Colbeck I , Nasir ZA, Ali Z. The state of ambient air quality in Pakistan—a review . Environmental Science and Pollution Research , 2010 , 17 : 49 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat [3] Javed W , Wexler AS, Murtaza G, et al. Spatial, temporal and size distribution of particulate matter and its chemical constituents in Faisalabad, Pakistan . Atmósfera , 2015 , 28 : 99 – 116 . Google Scholar Crossref Search ADS WorldCat [4] Labbé M , Leal M, Puerto J. New models for the location of controversial facilities: a bilevel programming approach . Computers & Operations Research , 2019 , 107 : 95 – 106 . Google Scholar Crossref Search ADS WorldCat [5] Christiansen C , Murray B. A Storage Market Review and Recommendations for Funding and Knowledge Sharing Priorities. Energy Storage Study. Sydney : AECOM , 2015 , 97 – 113 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [6] Taylor M , Ralon P, Ilas A. The Power to Change: Solar and Wind Cost Reduction Potential to 2025. Abu Dhabi : International renewable energy agency (IRENA) , 2016 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [7] Giaccone L , Lazzeroni P, Repetto M, et al. Uncertainty quantification in energy management procedures . Electronics , 2020 , 9 : 1471 . Google Scholar Crossref Search ADS WorldCat [8] Delgado J , Faria R, Moura P, et al. Impacts of plug-in electric vehicles in the Portuguese electrical grid . Transportation Research Part D: Transport and Environment , 2018 , 62 : 372 – 385 . Google Scholar Crossref Search ADS WorldCat [9] Canova A , Giaccone L, Spertino F, et al. Electrical impact of photovoltaic plant in distributed network . IEEE Transactions on Industry Applications , 2009 , 45 : 341 – 347 . Google Scholar Crossref Search ADS WorldCat [10] Denholm P , O’Connell M, Brinkman G, et al. Overgeneration from Solar Energy in California. A Field Guide to the Duck Chart. Golden, CO : National Renewable Energy Laboratory (NREL) , 2015 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [11] Luna-Rubio R , Trejo-Perea M, Vargas-Vázquez D, et al. Optimal sizing of renewable hybrids energy systems: a review of methodologies . Solar Energy , 2012 , 86 : 1077 – 1088 . Google Scholar Crossref Search ADS WorldCat [12] Gupta P , Pandit M, Kothari D. A review on optimal sizing and siting of distributed generation system: integrating distributed generation into the grid . In: 6th IEEE Power India International Conference (PIICON) , Delhi, India , 5 – 7 December 2014 , 669 – 675 . [13] Mekhilef S , Saidur R, Kamalisarvestani M. Effect of dust, humidity and air velocity on efficiency of photovoltaic cells . Renewable and Sustainable Energy Reviews , 2012 , 16 : 2920 – 2925 . Google Scholar Crossref Search ADS WorldCat [14] Arul P , Ramachandaramurthy VK, Rajkumar R. Control strategies for a hybrid renewable energy system: a review . Renewable and Sustainable Energy Reviews , 2015 , 42 : 597 – 608 . Google Scholar Crossref Search ADS WorldCat [15] Chen Z , Mi C, Xiong R, et al. Energy management of a power-split plug-in hybrid electric vehicle based on genetic algorithm and quadratic programming . Journal of Power Sources , 2014 , 248 : 416 – 426 . Google Scholar Crossref Search ADS WorldCat [16] Almas Prakasa M , Subiyanto S. Optimal cost and feasible design for grid-connected microgrid on campus area using the robust-intelligence method . Clean Energy , 2021 , 6 : 823 – 840 . Google Scholar OpenURL Placeholder Text WorldCat [17] Awad AS , El-Fouly TH, Salama MM. Optimal ESS allocation for benefit maximization in distribution networks . IEEE Transactions on Smart Grid , 2015 , 8 : 1668 – 1678 . Google Scholar Crossref Search ADS WorldCat [18] Das CK , Bass O, Kothapalli G, et al. Overview of energy storage systems in distribution networks: placement, sizing, operation, and power quality . Renewable and Sustainable Energy Reviews , 2018 , 91 : 1205 – 1230 . Google Scholar Crossref Search ADS WorldCat [19] Siddique R , Raza S, Mannan A, et al. A modified NSGA approach for optimal sizing and allocation of distributed resources and battery energy storage system in distribution network . Materials Today: Proceedings , 2021 , 47 : 102 – 109 . Google Scholar OpenURL Placeholder Text WorldCat [20] Xiao J , Zhang Z, Bai L, et al. Determination of the optimal installation site and capacity of battery energy storage system in distribution network integrated with distributed generation . IET Generation, Transmission & Distribution , 2016 , 10 : 601 – 607 . Google Scholar Crossref Search ADS WorldCat [21] Jayasekara N , Masoum MA, Wolfs PJ. Optimal operation of distributed energy storage systems to improve distribution network load and generation hosting capability . IEEE Transactions on Sustainable Energy , 2015 , 7 : 250 – 261 . Google Scholar Crossref Search ADS WorldCat [22] Wen S , Lan H, Fu Q, et al. Economic allocation for energy storage system considering wind power distribution . IEEE Transactions on Power Systems , 2014 , 30 : 644 – 652 . Google Scholar Crossref Search ADS WorldCat [23] Gallo A , Simões-Moreira JR, Costa HKM, et al. Energy storage in the energy transition context: a technology review . Renewable and Sustainable Energy Reviews , 2016 , 65 : 800 – 822 . Google Scholar Crossref Search ADS WorldCat [24] Argyrou MC , Christodoulides P, Kalogirou SA. Energy storage for electricity generation and related processes: technologies appraisal and grid scale applications . Renewable and Sustainable Energy Reviews , 2018 , 94 : 804 – 821 . Google Scholar Crossref Search ADS WorldCat [25] Babacan O , Torre W, Kleissl J. Siting and sizing of distributed energy storage to mitigate voltage impact by solar PV in distribution systems . Solar Energy , 2017 , 146 : 199 – 208 . Google Scholar Crossref Search ADS WorldCat [26] Marzano V , Papola A, Simonelli F, et al. A Kalman Filter for quasi-dynamic o-d flow estimation/updating . IEEE Transactions on Intelligent Transportation Systems , 2018 , 19 : 3604 – 3612 . Google Scholar Crossref Search ADS WorldCat [27] Pan Z , Wu J, Abeysekera M. Quasi-dynamic interactions between electricity systems and heating systems in normal operation . In: IEEE Conference on Energy Internet and Energy System Integration (EI2) , Beijing, China , 26–28 November 2017 , 21 – 26 . [28] Qin X , Shen X, Sun H, et al. A quasi-dynamic model and corresponding calculation method for integrated energy system with electricity and heat . Energy Procedia , 2019 , 158 : 6413 – 6418 . Google Scholar Crossref Search ADS WorldCat [29] Starčević V , Zeljković Č, Kitić N, et al. . PV system integration assessment by automated Monte Carlo simulation in DIgSILENT PowerFactory. In: 2021 20th International Symposium INFOTEH-JAHORINA (INFOTEH), East Sarajevo, Bosnia and Herzegovina, 17–19 March 2021 , 1 – 5 . [30] Bitterer R , Schieferdecker B. Repräsentative VDEWLastprofile AktionsplanWettbewerb . Tech. Rep. M32-99. Frankfurt : VDEW , 2001 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [31] Moreno R , Moreira R, Strbac G. A MILP model for optimizing multi-service portfolios of distributed energy storage . Applied Energy , 2015 , 137 : 554 – 566 . Google Scholar Crossref Search ADS WorldCat © The Author(s) 2022. Published by Oxford University Press on behalf of National Institute of Clean-and-Low-Carbon Energy This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com © The Author(s) 2022. Published by Oxford University Press on behalf of National Institute of Clean-and-Low-Carbon Energy http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Clean Energy Oxford University Press

Optimal integration of battery energy-storage system with high penetration of renewable energy in radial distribution network

8 pages

Loading next page...
 
/lp/oxford-university-press/optimal-integration-of-battery-energy-storage-system-with-high-0FSJUr994K
Publisher
Oxford University Press
Copyright
Copyright © 2022 National Institute of Clean-and-Low-Carbon Energy
ISSN
2515-4230
eISSN
2515-396X
DOI
10.1093/ce/zkac027
Publisher site
See Article on Publisher Site

Abstract

Abstract Considering the intermittent nature of renewable energy, a storage system to reserve power in off-peak hours and then to supply it during peak hours is necessary. However, if these storage devices in a network are not placed in an appropriate manner, the advantages can never be achieved. In this study, the allocation and sizing strategies of a battery energy-storage system (BESS) in an optimal way are proposed to improve the performance of the radial distribution networks. The test network adopted is a standard IEEE 33 bus network that is integrated with solar power. Simulations are carried out in DIgSILENT PowerFactory for the distribution network model and the process of optimization is done through MATLAB®. The optimization algorithm adopted is the genetic algorithm. The system is studied for 24 hours with a step size of 15 minutes, employing a time-sweep analysis of dynamic load. It is studied under three different scenarios and the results are compared. The results illustrate a considerable reduction in power losses in the case of the optimized BESS. The total losses before the optimization process are 16 365.57 kW. The total losses after the employment of the proposed solution are 10 246.5 kW, which means that the losses are reduced by ≤62%. Open in new tabDownload slide energy and environment, energy storage, solar Introduction These days, great interest has been aroused in reliable and flexible energy integration. It is a great challenge to balance power generation and consumption considering the stochastic nature of renewable energy [1]. Renewable energy is considered a path towards a clean and less polluted environment. It is seen as a possibility to reduce CO2 emissions. The degrees of atmospheric pollution along with the sources of air pollutants have increased essentially over the years [2, 3]. At the same time, numerous research studies have been done on solar energy integration to meet the renewable-energy target [4]. Different policies for renewable energy-generation strategies have also been launched by some governments [5]. Furthermore, it is expected that by the end of 2025, the levelized cost of electricity of solar energy could fall by 43% [6]. Few research studies have also been carried out for loss minimization; however, a renewable energy source was not considered in the network [7]. Despite the environmental and sustainability benefits of renewable energy, one of the most important problems is the management of power. Mainly, the photovoltaic (PV) system generates power depending on the local insolation level according to that area and the time of day. Especially in local electricity grids, in which the percentage of generation from solar is higher than that from other conventional generating technologies, the management of power becomes a challenge. The behaviour of the grid is disturbed due to these non-manageable energy-generating technologies. If the percentage of generated power through PV is low, the disequilibrium between production and consumption is not very problematic but if this percentage increases, the variability of electricity produced and its consumption become considerable matters. This intermittent power becomes the reason for the destabilization of power. Therefore, the management of electricity produced by such intermittent sources is under great consideration these days [8, 9]. One of the possible solutions suggested by the literature is to use hybrid renewable-energy systems for this issue of intermittency, to integrate different renewable-energy resources in an optimal manner [10]. In this system, sources with lower generation during a period can be compensated for by sources having a higher generation level during that period [11]. Thus, system efficiency and supply reliability can be increased [12]. The optimal sizing of generation units is reviewed in detail in [13] and different control strategies for renewable energy in [14]. The other solution suggested by the literature is the battery energy-storage system (BESS) [15]. The energy-storage system (ESS) is the key technology for the reliable integration of renewable energy resources [16]. The surplus amount of energy is stored in the ESS, which is fed when consumption is higher than generation. If the ESSs are not placed in the correct manner in a distribution network, the benefits from the ESS cannot be achieved [17]. As a result, optimal sizing and allocation of the BESS have become crucial areas of research. However, in this study, the allocation and sizing strategies of the BESS in an optimal way are proposed to improve the performance of the radial distribution networks. The problem of the optimal sizing and siting of the BESS in the distribution network has been significantly studied in the literature [18]. In [19], the authors used a modified non-sorting dominated genetic algorithm (GA) for the location and size of distributed generation (DG) optimally in distribution networks (DNs). The objective function is power-loss minimization, which improves the voltage profile of the system. This proposed methodology was tested on an IEEE 33-bus system. Misplacement or misuse of energy-storage devices can lead to several network performance issues in terms of voltage and frequency regulation, power quality, reliability and load controllability [18]. To achieve the optimal operation of an ESS, an appropriate way of placement is required for improvement in voltage and power quality [19, 20], peak-demand reduction [19], relaxation in distribution congestion [18], power-flow control [18], power-loss minimization [21], network security and cost reduction [8, 9], RES integration [10] and system effectiveness. There are several variables that are defined to approach the optimal sizing and siting of a BESS in a distribution network, such as optimizing technique, performance metrics for optimum evaluation, battery technology, and modelling and testing of the network. After a successful analysis of resources and estimation of solar production, a system needs to be developed to maintain the balance between generation and consumption patterns [16]. Intermittent PV production throughout the day can be smoothed by including ESSs and filling in the gaps. The smooth production of renewable energy throughout the day can be achieved through ESSs. They have the ability to handle ramps or the frequently known duck curve [7]. The ability to handle the load ramp cannot be done through traditional conventional generators but it can be handled very quickly through energy-storage devices [22]. The BESS has attracted considerable attention from researchers because of its advantages such as speedy response, controllability, energy management and reliability improvement [5]. Therefore, locating grid-scale ESSs becomes important in improving distribution network performance and minimizing the issues. If not placed in the correct manner in a distribution network, the benefits from the ESSs cannot be achieved [23]. However, in this section, an extensive review of the literature on optimal allocation and sizing of batteries is presented. The term ‘optimization’ means to reach the established goals with the help of available resources through a set of processes. The general optimization process algorithm for optimal sizing and placement has the following steps: (i) Identify the problem to be analysed in a network and the support solution provided by the ESS or we can identify the major objective, such as peak shaving, loss minimization, power quality and reliability, etc. (ii) Formulation of the fitness function and necessary constraints related to the optimal sizing and placement of the BESS. (iii) Evaluate the fitness function using optimization techniques such as analytical or meta-heuristic approaches. (iv) Update the fitness value and current best location and size of the BESS. The general flow chart of the heuristic optimization approach and the utilization of GA in the optimal allocation are shown in Fig. 1. Fig. 1: Open in new tabDownload slide (a) Flow chart of GA optimization approach. (b) Illustration of GA utilization. The objective of this research is to reduce the power losses in a radial distribution network given the constraints on power balance, bus voltage, energy-storage power, charge/discharge balance and remaining capacity of the ESS. The optimal placement and size of the ESS are determined by the GA optimization method. Several GA approaches have been utilized in other studies [15, 17]. These studies highlight the effectiveness of GA compared to other stochastic methods based on the processing duration and error percentage of the solution. In this study, DN data and BESS data were fed as inputs into the GA. The algorithm processes these data through many iterations and the resultant is the optimal place for a BESS that assures the minimum power loss in the distribution network. DIgSILENT PowerFactory (DPF) provides solutions for distribution network problems such as system design, modelling and optimization capabilities, grid interaction skills in a multi-user environment and data handling [24]. MATLAB® is used to control the system models developed in DPF and to facilitate genetic optimization. Furthermore, electricity production from renewable sources such as solar energy varies with changing solar radiation; thus, power generation is a function of time. Performing load-flow simulations during a certain time (in this case, a 24-hour period) in the different nodes of an electric power system requires the use of quasi-dynamic simulation, a time-varying load-flow calculation tool. Different researchers suggest that quasi-dynamic methods, as compared with quasi-static methods, not only improve the fidelity of the simulation of the process, but also reduce the processing time of the dynamic simulations. The main contributions of this paper are summarized as follows: An optimal BESS placement is carried out based on GA to significantly reduce the overall system power losses associated with the integration of high penetration levels of solar PVs. Although a similar investigation was carried out in [25], a unity power factor (p.f.) approach is applied on the ESS dispatch (i.e. the ESSs only inject P to the network). In this study, however, the ESSs inject both the P and Q to the network for better performance improvement with variable p.f. on the dispatch of an ESS. The effectiveness and robustness of the proposed methodology are comparatively tested and validated by different approaches on IEEE 33 bus systems. The first approach studied assumes that only the PV is contributing to the network and the voltage at the farthest node, such as bus no. 18. These values of voltages are later compared with the proposed network, such as after installing the BESS at the optimal place. The second approach is that active power losses are studied with and without the employment of the BESS of the overall DN. Overall, active power losses are reduced after the employment of the proposed strategy. 1 Methodology 1.1 Quasi-dynamic simulation A quasi-dynamic simulation is offered by DPF for the execution of medium-term to long-term electrical studies. It performs several load-flow calculations with user-defined time-step size. This tool is useful for dealing with studies in which long-term generation and demand profiles are defined and network models are developed considering expansion stages and variations [26]. The main benefit of this type of simulation is that it provides faster calculation results because it does not require solving all the mathematical expressions. Other researchers [27, 28] have suggested that quasi-dynamic simulation language offers better accuracy than other methods such as quasi-static. Quasi-dynamic simulations have been used in electrical studies into units such as circuit-interruption devices, PV energy systems, ESSs and distribution network studies. In this work, the ‘Quasi Dynamic’ tool (Fig. 2) is used to analyse the distribution network with varying demand in order to determine the behaviour and changes in system dynamics when some nodes in the system include solar DG. Fig. 2: Open in new tabDownload slide Quasi-dynamic simulation window. By modifying the load, each node in the system has a different demand that varies moment by moment. This new behaviour is analysed under different scenarios. The load flow is calculated every minute for 24 hours in order to identify the changes in the network variables such as power losses and voltage profiles. In order to analyse the simulation results, the IEEE 33 bus system was used as a test system model (Fig. 3). Fig. 3: Open in new tabDownload slide IEEE 33 bus system. 1.2 Load modelling The proposed load model is the IEEE 33 bus test system. It is designed using DPF. The system consists of 33 buses and 32 nodes with a current-carrying capacity of 400 A from Node 1 to Node 9, whereas other lines have a capacity of 200 A. The base voltage is 12.66 kV and the base power is 10 MVA. The total active power load is 3715 kW and the reactive power load is 2300 KVAr. Since the base system is modelled with a firm power load, it is required to simulate the load variations over time to study the dynamic nature of load under a 24-hour period. The quasi-dynamic simulation is used to calculate the set of time-dependent load-flow calculations. The standard Bundesverband der Energie- und Wasserwirtschaft load profile is utilized [30]. The basic system parameters are shown in Table 1. Table 1: 33 Bus system parameters Parameter . Value . Unit . Voltage level 12.66 kV Frequency 60 Hz Rated active power from external grid 3917.7 kW Rated reactive power from external grid 2435.2 kVAr Rated total active power losses on lines 202.7 kW Parameter . Value . Unit . Voltage level 12.66 kV Frequency 60 Hz Rated active power from external grid 3917.7 kW Rated reactive power from external grid 2435.2 kVAr Rated total active power losses on lines 202.7 kW Source: Ref. [29]. Open in new tab Table 1: 33 Bus system parameters Parameter . Value . Unit . Voltage level 12.66 kV Frequency 60 Hz Rated active power from external grid 3917.7 kW Rated reactive power from external grid 2435.2 kVAr Rated total active power losses on lines 202.7 kW Parameter . Value . Unit . Voltage level 12.66 kV Frequency 60 Hz Rated active power from external grid 3917.7 kW Rated reactive power from external grid 2435.2 kVAr Rated total active power losses on lines 202.7 kW Source: Ref. [29]. Open in new tab 1.3 Solar PV modelling There are a few factors that affect the generation from PV models, such as rated power, solar irradiance and temperature. The PV system generates power during the day. The output power of the solar module is shown in Fig. 4. It is generated based on Karachi city solar irradiance measured on the hottest day of the year, which is 21 July. The maximum output power is ~2.2 MW at 1 p.m. The number of parallel inverters is 20 000. The watt/panel is 110 watt so 110 × 20 000 = 2.2 MW. Integration of PV systems in DPF requires basic data for the network model. The type of PV panels is Aleo 150-L. The basic data for the PV model are shown in Table 2. Table 2: Basic data of PV model Name . Aleo 150L . Peak power (MPP) 160 W Rated voltage (MPP) 23.4 V Rated current (MPP) 6.8 A Open-circuit voltage 30.5 V Short-circuit current 7.2 A Material Single-crystalline silicon (mono-Si) NOCT 47℃ Name . Aleo 150L . Peak power (MPP) 160 W Rated voltage (MPP) 23.4 V Rated current (MPP) 6.8 A Open-circuit voltage 30.5 V Short-circuit current 7.2 A Material Single-crystalline silicon (mono-Si) NOCT 47℃ MPP, maximum power point; NOCT, nominal operating cell temperature. Open in new tab Table 2: Basic data of PV model Name . Aleo 150L . Peak power (MPP) 160 W Rated voltage (MPP) 23.4 V Rated current (MPP) 6.8 A Open-circuit voltage 30.5 V Short-circuit current 7.2 A Material Single-crystalline silicon (mono-Si) NOCT 47℃ Name . Aleo 150L . Peak power (MPP) 160 W Rated voltage (MPP) 23.4 V Rated current (MPP) 6.8 A Open-circuit voltage 30.5 V Short-circuit current 7.2 A Material Single-crystalline silicon (mono-Si) NOCT 47℃ MPP, maximum power point; NOCT, nominal operating cell temperature. Open in new tab Fig. 4: Open in new tabDownload slide PV output curve. 1.4 BESS modelling The ESS is simulated using the built-in static generator unit that is used to model any non-rotating power-generation and storage unit that is connected to the AC grid through an inverter. The use of this element is an appropriate approximation and it is also supported by DPF in which ESSs are represented by static generators. To implement this time-varying system, the time-dependent load-flow script ‘time sweep’ is utilized. The objectives achieved by this framework are: appropriate objects on the grid topology such as PV units, energy-storage units and time-dependent loads are identified; specific system parameters are then read to perform the time-dependent load-flow analysis later to extract the results; the adopted load-flow script analyzes the name of storage units and its capacity to calculate the state of charge with user-defined time-step sizes. In this research, it is assumed that the BESS is fully charged at the start of the day and discharges gradually as time goes by[31]. In order to simulate a battery storage unit over a period of time, the unit is modelled considering the following three parameters: (i) maximal charge and discharge power; (ii) charge–discharge efficiency; (iii) storage capacity. The battery storage unit is used for the modelling of a BESS. The active power rating of the battery is 1.1 MW. 1.5 Optimization method GAs are optimization techniques used to solve non-linear or non-differentiable optimization problems. GA uses concepts from evolutionary biology to search for a global minimum. The name ‘genetic algorithm’ comes from the fact that the algorithm is mimicking evolutionary biology techniques. GA works by starting with an initial generation of candidate solutions that are tested against the objective function, then generating subsequent generations of points from the first generation through processes such as selection crossover and mutation. The GA runs its iterations until it converges to the best objective function, i.e. when the fitness function value is no longer changing or it is changing by a really small amount. This paper employs GA as an optimization method to optimally size and site the BESS in a distribution network. This process is done using MATLAB® programming. The system summary from DPF is synchronized with MATLAB® for the process of optimization. It is assumed that the PV system only supplies active power and is placed arbitrarily at buses. The objective function is to reduce line losses, formulated as: Ploss=min∑Ttotalt=1∑nbri,j=1 i≠j g i,j(Ui2 + Uj2 − 2UiUjcoscos (θi − θj))(1) where Δt represents the time interval; ttotal represents the total time period considered, gi,j represents the conductance between buses i and j, nbr represents the total number of branches in the network, Ui and Uj represent the voltage magnitudes of the buses i and j, and θi and θj represent the voltage angles of buses i and j. Equation (1) interprets the expression for total network power losses based on the AC power-flow equation. For each and every evaluation of fitness function, the non-linear AC power flow is calculated. In addition to the above objective function, the optimal network configuration has to satisfy the following operational constraints: Uimin≤Ui≤Uimax(2) Sb≤Sbmax(3) where Uimin represents the minimum voltage magnitude for i buses, Uimax represents the maximum voltage magnitude for i buses, Sb represents the actual power flow and Sbmax represents the maximum actual power flow. Uimin and Uimax represent the voltage limit of each bus and Sbmax represents the maximum current level of each branch b in terms of actual power flow. The whole search process of the proposed method for determining the location and size of the BESS to minimize total active power loss is described in detail as follows: Step 1: Define the input data (Vrated, RL, XL, PV and MVAbase) to simulate the studied distribution network in DPF and evaluate candidate buses to find the best location for the BESS installation. Step 2: Initialize the parameters of the algorithm. Step 3: Select one candidate bus as the location for the BESS installation. Step 4: Execute the algorithm (as in Fig. 1) to solve the optimization problem. Step 5: Conduct a load-flow process at the current candidate bus at which the BESS is installed. Step 6: Evaluate the objective function at that point. Step 7: Update the best solution. Step 8: If the maximum iteration is reached, calculate optimal coefficients (Step 9); otherwise, continue optimization (Step 4). Step 9: Obtain optimal coefficients at the current candidate bus at which the BESS is installed from Step 3. Step 10: If the candidate bus is the last candidate, proceed to the optimal BESS installation (Step 11); otherwise, select the next candidate bus and execute the algorithm (Step 3). Step 11: Determine the optimal siting of the BESS installation, which is the candidate bus providing the minimum value of the objective function, such as reduced power losses. Step 12: Substitute the obtained coefficients in the BESS unit. 2 Results and discussion The results were obtained using DPF for model simulation and MATLAB® for the optimization process. The standard IEEE 33 bus system is taken as the base case and its data are shown in Table 2. The data are modified to create a dynamic load demand and system dynamics are studied for 24 hours with a time-step size of 15 minutes (Fig. 5). There are three cases in this study. The system load is dynamic in all three cases, which is 72.28 MWh. The load is maximum at around 9.30 to 11.30 a.m., as shown in Fig. 5. The first case is the base case that includes no PV and BESS. The total losses in that case were 16 365.57 kW. Fig. 6 shows the total active power losses in a complete day of 32 branches. Fig. 5: Open in new tabDownload slide Load profile of 1 day. Fig. 6: Open in new tabDownload slide Active power losses base case. The total losses after the employment of the proposed solution are 10 246.5 kW, which means that the losses are reduced by ≤62%. The reduction in total active power losses in each case are shown in Figs 7 and 8. Fig. 7: Open in new tabDownload slide Total losses in each case. Fig. 8: Open in new tabDownload slide Losses in kW. The voltage profile of each bus in the base case is shown in Fig. 9. As the voltage reduces at a node with respect to an increase in the distance from the supply side, only the farthest node is studied in this research, which is bus no. 18. The voltage at bus no. 18 in each case is shown in Fig. 9. It is evident that the voltage has improved after the employment of the proposed solution. The voltage between the midnight hours has improved by 0.02 p.u. Hence the overall system voltage has been improved by 97% and it is under the stability limits. Fig. 9: Open in new tabDownload slide Voltage magnitude at bus 18 in each case 3 Conclusion In this research, a GA approach was utilized to find the optimal size and placement of a BESS integrated with PV and a grid in the IEEE 33 bus system. The overall system active power losses are reduced and the voltage is improved after the employment of the proposed solution. Several cases have been established to verify the results compared to the test system. The simulation results reveal that the overall system losses under dynamic load without the usage of BESSs are 16 365.57 kW and after the placement of distributed BESSs they are reduced to 10 246.5 kW. Therefore, losses in the proposed solution are reduced by 62%. The improvement in voltage is also justified in the results by the use of BESSs. Intermittent renewable-energy systems, especially at higher levels of penetration, can also be made stable by this application of BESSs. For future work, the proposed solution can be tested for other renewable-energy integrations as well, such as wind, fuel-cell, etc. The objective function can be expanded as a multi-objective optimization approach and this study will contribute to the efforts. Conflict of interest statement None declared. References [1] Yi W , Zhang Y, Zhao Z, et al. Multiobjective robust scheduling for smart distribution grids: considering renewable energy and demand response uncertainty . IEEE Access , 2018 , 6 : 15 – 24 . Google Scholar OpenURL Placeholder Text WorldCat [2] Colbeck I , Nasir ZA, Ali Z. The state of ambient air quality in Pakistan—a review . Environmental Science and Pollution Research , 2010 , 17 : 49 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat [3] Javed W , Wexler AS, Murtaza G, et al. Spatial, temporal and size distribution of particulate matter and its chemical constituents in Faisalabad, Pakistan . Atmósfera , 2015 , 28 : 99 – 116 . Google Scholar Crossref Search ADS WorldCat [4] Labbé M , Leal M, Puerto J. New models for the location of controversial facilities: a bilevel programming approach . Computers & Operations Research , 2019 , 107 : 95 – 106 . Google Scholar Crossref Search ADS WorldCat [5] Christiansen C , Murray B. A Storage Market Review and Recommendations for Funding and Knowledge Sharing Priorities. Energy Storage Study. Sydney : AECOM , 2015 , 97 – 113 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [6] Taylor M , Ralon P, Ilas A. The Power to Change: Solar and Wind Cost Reduction Potential to 2025. Abu Dhabi : International renewable energy agency (IRENA) , 2016 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [7] Giaccone L , Lazzeroni P, Repetto M, et al. Uncertainty quantification in energy management procedures . Electronics , 2020 , 9 : 1471 . Google Scholar Crossref Search ADS WorldCat [8] Delgado J , Faria R, Moura P, et al. Impacts of plug-in electric vehicles in the Portuguese electrical grid . Transportation Research Part D: Transport and Environment , 2018 , 62 : 372 – 385 . Google Scholar Crossref Search ADS WorldCat [9] Canova A , Giaccone L, Spertino F, et al. Electrical impact of photovoltaic plant in distributed network . IEEE Transactions on Industry Applications , 2009 , 45 : 341 – 347 . Google Scholar Crossref Search ADS WorldCat [10] Denholm P , O’Connell M, Brinkman G, et al. Overgeneration from Solar Energy in California. A Field Guide to the Duck Chart. Golden, CO : National Renewable Energy Laboratory (NREL) , 2015 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [11] Luna-Rubio R , Trejo-Perea M, Vargas-Vázquez D, et al. Optimal sizing of renewable hybrids energy systems: a review of methodologies . Solar Energy , 2012 , 86 : 1077 – 1088 . Google Scholar Crossref Search ADS WorldCat [12] Gupta P , Pandit M, Kothari D. A review on optimal sizing and siting of distributed generation system: integrating distributed generation into the grid . In: 6th IEEE Power India International Conference (PIICON) , Delhi, India , 5 – 7 December 2014 , 669 – 675 . [13] Mekhilef S , Saidur R, Kamalisarvestani M. Effect of dust, humidity and air velocity on efficiency of photovoltaic cells . Renewable and Sustainable Energy Reviews , 2012 , 16 : 2920 – 2925 . Google Scholar Crossref Search ADS WorldCat [14] Arul P , Ramachandaramurthy VK, Rajkumar R. Control strategies for a hybrid renewable energy system: a review . Renewable and Sustainable Energy Reviews , 2015 , 42 : 597 – 608 . Google Scholar Crossref Search ADS WorldCat [15] Chen Z , Mi C, Xiong R, et al. Energy management of a power-split plug-in hybrid electric vehicle based on genetic algorithm and quadratic programming . Journal of Power Sources , 2014 , 248 : 416 – 426 . Google Scholar Crossref Search ADS WorldCat [16] Almas Prakasa M , Subiyanto S. Optimal cost and feasible design for grid-connected microgrid on campus area using the robust-intelligence method . Clean Energy , 2021 , 6 : 823 – 840 . Google Scholar OpenURL Placeholder Text WorldCat [17] Awad AS , El-Fouly TH, Salama MM. Optimal ESS allocation for benefit maximization in distribution networks . IEEE Transactions on Smart Grid , 2015 , 8 : 1668 – 1678 . Google Scholar Crossref Search ADS WorldCat [18] Das CK , Bass O, Kothapalli G, et al. Overview of energy storage systems in distribution networks: placement, sizing, operation, and power quality . Renewable and Sustainable Energy Reviews , 2018 , 91 : 1205 – 1230 . Google Scholar Crossref Search ADS WorldCat [19] Siddique R , Raza S, Mannan A, et al. A modified NSGA approach for optimal sizing and allocation of distributed resources and battery energy storage system in distribution network . Materials Today: Proceedings , 2021 , 47 : 102 – 109 . Google Scholar OpenURL Placeholder Text WorldCat [20] Xiao J , Zhang Z, Bai L, et al. Determination of the optimal installation site and capacity of battery energy storage system in distribution network integrated with distributed generation . IET Generation, Transmission & Distribution , 2016 , 10 : 601 – 607 . Google Scholar Crossref Search ADS WorldCat [21] Jayasekara N , Masoum MA, Wolfs PJ. Optimal operation of distributed energy storage systems to improve distribution network load and generation hosting capability . IEEE Transactions on Sustainable Energy , 2015 , 7 : 250 – 261 . Google Scholar Crossref Search ADS WorldCat [22] Wen S , Lan H, Fu Q, et al. Economic allocation for energy storage system considering wind power distribution . IEEE Transactions on Power Systems , 2014 , 30 : 644 – 652 . Google Scholar Crossref Search ADS WorldCat [23] Gallo A , Simões-Moreira JR, Costa HKM, et al. Energy storage in the energy transition context: a technology review . Renewable and Sustainable Energy Reviews , 2016 , 65 : 800 – 822 . Google Scholar Crossref Search ADS WorldCat [24] Argyrou MC , Christodoulides P, Kalogirou SA. Energy storage for electricity generation and related processes: technologies appraisal and grid scale applications . Renewable and Sustainable Energy Reviews , 2018 , 94 : 804 – 821 . Google Scholar Crossref Search ADS WorldCat [25] Babacan O , Torre W, Kleissl J. Siting and sizing of distributed energy storage to mitigate voltage impact by solar PV in distribution systems . Solar Energy , 2017 , 146 : 199 – 208 . Google Scholar Crossref Search ADS WorldCat [26] Marzano V , Papola A, Simonelli F, et al. A Kalman Filter for quasi-dynamic o-d flow estimation/updating . IEEE Transactions on Intelligent Transportation Systems , 2018 , 19 : 3604 – 3612 . Google Scholar Crossref Search ADS WorldCat [27] Pan Z , Wu J, Abeysekera M. Quasi-dynamic interactions between electricity systems and heating systems in normal operation . In: IEEE Conference on Energy Internet and Energy System Integration (EI2) , Beijing, China , 26–28 November 2017 , 21 – 26 . [28] Qin X , Shen X, Sun H, et al. A quasi-dynamic model and corresponding calculation method for integrated energy system with electricity and heat . Energy Procedia , 2019 , 158 : 6413 – 6418 . Google Scholar Crossref Search ADS WorldCat [29] Starčević V , Zeljković Č, Kitić N, et al. . PV system integration assessment by automated Monte Carlo simulation in DIgSILENT PowerFactory. In: 2021 20th International Symposium INFOTEH-JAHORINA (INFOTEH), East Sarajevo, Bosnia and Herzegovina, 17–19 March 2021 , 1 – 5 . [30] Bitterer R , Schieferdecker B. Repräsentative VDEWLastprofile AktionsplanWettbewerb . Tech. Rep. M32-99. Frankfurt : VDEW , 2001 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [31] Moreno R , Moreira R, Strbac G. A MILP model for optimizing multi-service portfolios of distributed energy storage . Applied Energy , 2015 , 137 : 554 – 566 . Google Scholar Crossref Search ADS WorldCat © The Author(s) 2022. Published by Oxford University Press on behalf of National Institute of Clean-and-Low-Carbon Energy This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com © The Author(s) 2022. Published by Oxford University Press on behalf of National Institute of Clean-and-Low-Carbon Energy

Journal

Clean EnergyOxford University Press

Published: Jun 1, 2022

References