Design of Three Electric Vehicle Charging Tariff Systems to Improve Photovoltaic Self-Consumption

Citation: Etxegarai, G.; Camblong, H.; Ezeiza, A.; Lie, T.T. Design of Three Electric Vehicle Charging Tariff Systems to Improve Photovoltaic Self-Consumption. Energies 2024,17, 1806. https://doi.org/ 10.3390/en17081806 Academic Editors: Marianna Jacyna, Emilian Szczepa´nski and Mariusz Izdebski Received: 27 February 2024 Revised: 3 April 2024 Accepted: 4 April 2024 Published: 9 April 2024 Copyright: © 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). energies Article Design of Three Electric Vehicle Charging Tariff Systems to Improve Photovoltaic Self-Consumption Garazi Etxegarai 1,2,* , Haritza Camblong 1,3 , Aitzol Ezeiza 1and Tek Tjing Lie 3 1 Department of Systems Engineering & Control, Faculty of Engineering of Gipuzkoa, University of the Basque Country (UPV/EHU), Europa Plaza 1, E-20018 Donostia, Spain; [email protected] (H.C.); [email protected] (A.E.) 2ESTIA Institute of Technology, University of Bordeaux, 64210 Bidart, France 3Department of Electrical and Electronic Engineering, Auckland University of Technology, Auckland 1010, New Zealand; [email protected] *Correspondence: [email protected] Abstract: Electric vehicles (EVs) are emerging as one of the pillars for achieving climate neutrality. They represent both a threat and an opportunity for the operation of the network. Used as flexible loads, they can favor the self-consumption of photovoltaic (PV) energy. This paper presents three EV charging tariff systems (TSs) based on the self-consumption of excess PV energy. The TS objectives are to increase the self-consumption rate (SCR) and thus indirectly decrease the charging cost of the EV users. Two of the proposed TSs correspond to an indirect control of EV charging. The third TS is a hybrid system where the charging power is controlled. The TS is designed using a series of rules that consider the momentary PV surplus and the charging power of each EV. The influence of the TS is simulated by considering real data from a PV collective self-consumption project in the Basque Country (Spain). The TS simulations performed with 6 months of data show a 13.1% increase in the SCR when applying the third TS, reaching an average of 93.09% for the SCR. In addition, the cost of EV charging is reduced by 25%. Keywords: electric vehicles; charging system control; pricing schemes; PV energy; collective self-consumption; self-consumption rate 1. Introduction The transport sector accounts for a large contribution to greenhouse gas emissions. In this context, electric vehicles (EVs) are emerging as one of the solutions that will help the European Union (EU) achieve its climate neutrality targets [ 1 ]. Thus, the EU has taken the measure of banning the sale of new petrol and diesel cars from 2035 [ 2 , 3 ]. In the first quarter of 2023, over 2.3 million EVs were sold worldwide, 25% more than in the same period the precedent year [ 4 ]. To cope with this growth, EV chargers will have to be installed every 60 km by 2026 [5]. According to [ 6 ], uncontrolled deployment of EVs would increase peak electricity demand by 35% to 51%. Without proper control, the high penetration of EVs will widen the gap between peak and off-peak loads on the grid [ 7 ]. This fact could overload distribution lines and transformers, leading to higher grid losses and reduced equipment lifetime [8]. However, coordinating the charging of EVs could also offer significant flexibility [ 9 ]. One solution involves using renewable energy sources (RES) for EV charging. By scheduling charging during times when RES is available, more RES is integrated into the grid and the gap between peak and off-peak power is reduced [ 9 ]. This way, local solar energy self-consumption holds the potential to emerge as a multifaceted solution. When using EVs as flexible loads to help reduce grid operation issues, it is essential to manage the charging pattern in a controlled manner. EV control systems that intend to ensure an acceptable charging profile can be classified into two main types: direct Energies 2024,17, 1806. https://doi.org/10.3390/en17081806 https://www.mdpi.com/journal/energies Energies 2024,17, 1806 2 of 23 and indirect control [ 10 ]. Direct control involves acting on the load patterns directly by controlling the charging power. One of the drawbacks of direct control systems is the high communication and computational infrastructure required. On the other hand, indirect control is based on encouraging EV users to adapt their charging patterns. The most common method of influence is through charging tariffs. There are several charging tariff systems, such as real-time pricing (RTP), critical peak pricing (CPP), and time-of-use (ToU) charging. In the RTP solution, prices for each time slot are announced shortly before the interval starts. Although this scheme is very efficient due to its dynamic nature, it requires a large information and communication technology (ICT) infrastructure. Furthermore, it requires substantial user participation [ 11 ]. In the case of CPP, extremely high prices are charged for occasions where peak demand is very high [ 12 ]. This type of pricing can be associated with others, whereas in ToU strategies, different fixed windows in the day are related to different fixed rates [ 13 ]. ToU charging system is the most widely used because of the simplicity of implementation [10,14]. Several papers have studied how EVs can benefit the electricity grid [ 15 ]. In [ 16 ], pricing mechanisms to incentivize EV users to shift their charging to fill the off-peak zones are proposed. The authors highlight that by using ToU-based charging, EV users would choose to charge the vehicle at the beginning of the cheapest zone, causing a new consumption peak. For this reason, the authors propose two charging mechanisms: one non-cooperative and the other cooperative, which consider the charging schedules of the vehicles that have already arrived. In the non-cooperative scenario, each EV schedules its own charging without cooperating with the other EVs, while in the cooperative scenario, all EVs are controlled by an aggregator. In [ 17 ], a day-ahead dispatch strategy is proposed for EVs considering the carbon quota. In addition, this strategy can provide peak shaving and valley-filling services. Other works have studied the synergy between photovoltaic (PV) energy and EV charging. In [ 18 , 19 ], an indirect control approach presents a dynamic pricing system based on the Stackelberg game for an EV charging station associated with a PV system. The Stackelberg equilibrium seeks a win–win situation, reducing the cost for users and increasing the profit of the charging station. In [ 19 ], the charging station sets an appropriate selling price to maximise its profit. Then, the Stackelberg equilibrium is solved considering the user’s criteria, and the prices are determined by the charging station. There are also studies examining the role of EVs in increasing the PV self-consumption rate (SCR). Studies in [ 20 , 21 ] present improvements in PV self-consumption and selfsufficiency using batteries and EVs without applying any control. For instance, in [ 20 ], households with PV generation exhibited a SCR of 26% without storage, 59% with an EV, and 31% with batteries sized for that case study. Conversely, works [ 22 , 23 ] implement a direct control with the objective of maximising the SCR. In [ 22 ], direct control is used to define the charging pattern of EVs considering them as a flexible charging source, and in the case of vehicle to grid (V2G) as a storage device. In this work, three charging algorithms are proposed. The first algorithm uses real-time information, the second incorporates V2G technology, and the third is an optimisation algorithm using predictions for both demand and production, taking into account V2G technology. The results present a SCR of 49% for the uncontrolled case, 62% with the first algorithm, 79% with the second, and 87% with the third. The study conducted in [ 23 ] presents a combination of smart metering and smart charging that helps local energy communities increase self-consumption. A case with four consumers and a prosumer is studied. In the scenario without EVs, one-fifth of the PV energy is consumed, while with smart metering and EVs, the SCR is increased by 45%. In [ 24 ], a distributed and centralised smart charging scheme for EVs in residential buildings with PV systems is presented. The aim of smart charging is to minimise net load variability, thereby increasing self-consumption and reducing consumption peaks. In the centralised charging scheme, a central unit determines the charging time and power for a fleet of EVs, whereas, in the distributed charging approach, charging decisions are made at Energies 2024,17, 1806 3 of 23 the user level. The proposed smart charging schemes consider EV energy demand, arrival and departure times, and predictions of building consumption and PV production. Considering the state-of-the-art study conducted, there are no studies that assess the improvement of PV SCR through the use of indirect control systems in EV charging. The work presented in this paper proposes three charging tariff schemes based on temporary PV surpluses in a real PV collective self-consumption (CSC) project in Aduna (Basque Autonomous Community, Spain). All three tariff systems (TS) share the same objective: to influence user behaviour in shifting the timing of EV charging to increase the consumption of local PV energy, and thus indirectly decrease the charging cost of the EV users. The following hypotheses were considered: I. EV users prefer to charge their vehicles when prices are lower. Therefore, if reduced prices are offered during PV surplus hours, users will adjust their charging times. II. The resolution in the control of EV charging power considered in this article is ideal. The two main contributions of the research study are as follows: • The design of indirect control EV charging is based on temporary PV surplus, with the main objective of increasing the SCR of a real PV CSC project. • A detailed description of the design of three TSs for EV charging, which can be easily replicated and adjusted to any case. It is also worth highlighting that these TSs would promote the development of EVs, which is the aim of the Aduna town council, the owner of the PV panels and EV chargers. The rest of this document is organised as follows. Section 2presents the case study and the data used. In Section 3, the proposed three TSs are explained. Numerical results and discussion are provided in Section 4. Finally, the last section concludes this article. 2. Case Study The three TSs proposed in this work were simulated with real historical data collected from the PV CSC project. This project takes place in the municipality of Aduna, located in the Basque Country, Spain. There are eight consumption points associated with the CSC project. One of them is a public EV charging point. For administrative reasons, the PV panels with a capacity of 62.4 kWp were not yet installed during this study. The mentioned data can be divided into three groups: the data related to the eight consumption points, the data on the public EV charging station obtained from the Charge and Parking application, and the data related to PV power generation. The consumption of the seven points (all except the EV charger) is considered to be the basic consumption for this case study. This consumption is used to obtain the PV surplus by calculating the difference between PV production and basic consumption. All proposed TSs are based on the PV surpluses. In this document, these surpluses are directly calculated by subtracting the basic consumption from the PV generation. However, when implementing the TSs in real time, the energy management system (EMS) will use the day-ahead forecasts of both PV production and consumption (as carried out in [25,26]). The consumption historical data were recorded hourly and were available for the last 3 years. However, since the public charging point was not installed until November 2021, the simulations considered data obtained between November 2021 and April 2022, the same period as for the EV charging data. On the other hand, the sampling time was reduced from 1 h to 10 min (the sample time corresponding to PV generation), with the objective of producing more accurate simulations. In this way, the EV charging time was more closely adjusted to the real charging time. To perform the change in the sampling time, the value of the power consumed over one hour was used in the six intervals that constitute one hour. The public EV charging station is a 22 kW three-phase charger with two connectors. Regarding the EV charging data, the following information was obtained using the Charge and Parking application: the time at which an EV was connected and disconnected, the Energies 2024,17, 1806 4 of 23 EV brand and model, and the total energy consumed for each charge. By knowing the EV brand and model, it was possible to check and record the maximum charging power for each EV. It should be noted that not all EV models charge at the same power. Among the 23 EV models registered in the Charge and Parking application, the most common maximum charging power values were 3.7 kW, 7.2 kW, and 11 kW. These maximum power values were used in the design of TS1 and 2. Regarding PV data, they were obtained using local historical solar irradiance data for the simulation period, from the Basque Meteorological Agency, Euskalmet [ 27 ], with a sampling interval of 10 min. PV production was estimated by multiplying the irradiation by the peak power of the PV panels and applying a correction factor to account for efficiency. 3. Proposed Pricing Methods The following section describes the three proposed TSs to influence the EV user’s charging pattern. Although the three TS algorithms share the same objective, their rules are different. TS1 and 2 are indirect controls. TS1 is the least complex: when the PV surplus exceeds 7 kW, a lower price is offered to EV users to encourage charging at that time. On the other hand, TS2 is more personalised and is offered when the PV surplus exceeds different levels corresponding to the different charging power of EVs. Finally, TS3 is based on a hybrid control. In this TS, a cheaper charging price is always offered when there is a PV surplus, regardless of the amount, and the EV on-board battery management system (BMS) and the charger modify the charging power depending on the surplus at any given time, without consuming energy from the grid. The three proposed TSs offer a cheaper price than the market price to encourage charging at the most opportune times. However, when there is no surplus, the market price is offered. 3.1. Market Tariff Used as Base Tariff The market tariff used as the base tariff for the three TSs is one of the GoiEner cooperatives [ 28 ], specifically the 3.0 TDVE tariff, which is tailored for public charging points. The 3.0 TDVE tariff has different prices per hourly slot, with modifications each month, as described in Figure 1. There are six periods with different energy and power rates used in the various hourly slots. Regarding the GoiEner’s self-consumption compensation price, it is 0.0839 EUR/kWh. The compensation price is the price at which the PV panel owner is compensated for injecting locally generated PV energy into the grid that has not been instantaneously consumed. Charging and compensation prices are reviewed every trimester. Energies 2024, 17, x FOR PEER REVIEW 5 of 27 Figure 1. Different power and energy prices per hourly slot of the GoiEner’s 3.0 TDVE tariff. 3.2. Tariff System 1 The first TS is the simplest one. Whenever the PV surplus exceeds 7 kW, the reduced price is offered. There are two main reasons why the threshold is set at 7 kW. On the one hand, if the threshold is very low, by offering the reduced price, EVs would consume more from the grid and less from the surplus, as the users would charge their EVs when the surplus is lower. On the other hand, the average maximum charging power of all registered EVs is around 7 kW. The main disadvantage of this TS is that there are occasions when the surplus is less than 7 kW and is not used. To calculate SCRs that would be obtained with the implementation of TS1, a simulation was carried out based on the historical data presented in Section 2. The simulator was coded in MATLAB ® software 9.13.0.2166757 (R2022b) Update 4. The simulation was performed for each day where EVs were charged during the above-mentioned six months. The EVs considered in the simulation on a given day were processed in the order of original arrival at the charging point. Additionally, since charging system 1 is an indirect control, once the EV had started to charge in the simulation (when the reduced price was offered), the EV was charged at its maximum power until it was fully charged, even though after a while the reduced price was no longer offered. This process was considered to be the most realistic. The flow chart in Figure 2 describes the simulation process for one day under the influence of TS1. The diagram can be divided into three phases. The first phase is composed of the two main loops, which ensure that all EVs and all sampling periods of the day are analysed. It starts by detecting how many EVs were charged that day (variable Num_EV_Charge). The simulation starts by analysing the first EV of the day, so the index y takes the value 1 (y = 1). When the charge of the first EV is simulated, index y is incremented to 2 to consider the next EV of the day. When y is greater than Num_EV_Charge, it means that all EVs have been simulated and therefore the simulation of the EV charging paern ends, leading to Figure 7′s flow chart (linked to connector A) where the charging cost is calculated. Figure 1. Different power and energy prices per hourly slot of the GoiEner’s 3.0 TDVE tariff. Energies 2024,17, 1806 5 of 23 To conclude, the three proposed TSs offer the compensation price (0.0839 EUR/kWh) when the specific conditions of each TS are fulfilled. When the conditions are not met, the charging price is the market price for the corresponding time slot and month. 3.2. Tariff System 1 The first TS is the simplest one. Whenever the PV surplus exceeds 7 kW, the reduced price is offered. There are two main reasons why the threshold is set at 7 kW. On the one hand, if the threshold is very low, by offering the reduced price, EVs would consume more from the grid and less from the surplus, as the users would charge their EVs when the surplus is lower. On the other hand, the average maximum charging power of all registered EVs is around 7 kW. The main disadvantage of this TS is that there are occasions when the surplus is less than 7 kW and is not used. To calculate SCRs that would be obtained with the implementation of TS1, a simulation was carried out based on the historical data presented in Section 2. The simulator was coded in MATLAB ® software 9.13.0.2166757 (R2022b) Update 4. The simulation was performed for each day where EVs were charged during the above-mentioned six months. The EVs considered in the simulation on a given day were processed in the order of original arrival at the charging point. Additionally, since charging system 1 is an indirect control, once the EV had started to charge in the simulation (when the reduced price was offered), the EV was charged at its maximum power until it was fully charged, even though after a while the reduced price was no longer offered. This process was considered to be the most realistic. The flow chart in Figure 2describes the simulation process for one day under the influence of TS1. The diagram can be divided into three phases. The first phase is composed of the two main loops, which ensure that all EVs and all sampling periods of the day are analysed. It starts by detecting how many EVs were charged that day (variable Num_EV_Charge). The simulation starts by analysing the first EV of the day, so the index y takes the value 1 (y= 1). When the charge of the first EV is simulated, index yis incremented to 2 to consider the next EV of the day. When yis greater than Num_EV_Charge, it means that all EVs have been simulated and therefore the simulation of the EV charging pattern ends, leading to Figure 7’s flow chart (linked to connector A) where the charging cost is calculated. Next, the first sampling period (t= 1) is analysed. Given that in the simulation the sample time is 10 min, the first sampling period, t= 1, starts at 00:10. The index tis incremented every new period until it reaches a value of Num_samples, 144 (00:00). When t is greater than Num_samples, the simulation ends, leading to Figure 7’s flow chart. The second phase analyses whether the conditions are met to offer the reduced price according to TS1. For this purpose, the difference between the consumption and the PV production at time t(diff(t)) is calculated. If diff (t) exceeds 7 kW, the reduced price is offered, and therefore EV(y) (y th EV) charging is carried out. On the contrary, if the difference does not exceed 7 kW, the variable tis incremented, and the difference is recalculated and compared with the threshold of 7 kW. Whenever diff(t) is greater than 7 kW, it is checked if there is free space at the charging point. If the variable EV_Connected is less than 2, this means that there is space available, and the simulation of the EV(y) charging pattern starts. In the third phase, the charging pattern of the vehicles is simulated. First, index t y is created. This index represents how many time intervals the EV(y) has been connected to. It is necessary to (i) know how long EV(y) has needed to complete its charge and (ii) consider that a connector of the charging point has been occupied during the time interval t y . Index t y increases each time EV(y) spends a sample period charging. The variable EV_rem_en oversees how much energy the EV(y) must consume to complete its charge. These data are known from the registers of the Charge and Parking application. Whenever EV_rem_en is greater than 0, it means that the EV(y) has not yet completed its charge. The variable EV_char_pat records the EV(y) charge pattern, i.e., the energy consumed in each tinterval. During an interval, an EV charges the maximum amount of energy. This Energies 2024,17, 1806 6 of 23 value is obtained by multiplying the maximum charging power of the EV by the sampling time. In the flow chart, this maximum energy is defined as EV_max_en_SP. Whenever the variable EV_rem_en is greater than EV_max_en_SP, the charge pattern for that interval is the maximum energy that can be charged. After recording in the variable EV_char_pat how much energy has been consumed for that interval, EV_rem_en is updated, as shown in Equation (1). In addition, the fact that a connector has been occupied by that interval is also recorded (Equation (2)). Energies 2024, 17, x FOR PEER REVIEW 6 of 27 Figure 2. Flow chart related to TS1. Next, the first sampling period (t = 1) is analysed. Given that in the simulation the sample time is 10 min, the first sampling period, t = 1, starts at 00:10. The index t is incremented every new period until it reaches a value of Num_samples, 144 (00:00). When t is greater than Num_samples, the simulation ends, leading to Figure 7′s flow chart. Figure 2. Flow chart related to TS1. Energies 2024,17, 1806 7 of 23 EV_rem_en =EV_rem_ent+ty−1−EV_char_patt+ty, (1) EV_Connectedt+ty=EV_Connectedt+ty+1. (2) Afterwards, EV_rem_en is checked again to ensure that it is still greater than 0. t y is increased by one, and EV(y) continues to be charged. In the last interval before the end of the charge, EV_rem_en is lower than the maximum power it can charge in one interval. Then, the charge pattern for that interval is the amount of remaining energy (EV_rem_en). Next, the variable EV_rem_en becomes 0, and the variable EV_Connected registers one last time that the connector is occupied. Finally, as the charging of EV(y) is completed, its charging pattern is saved, and the variable yis incremented. The simulation starts again from the first phase analysing the charge of the next EV of the day. Although the charging of a single EV is simulated at each round of the flow chart, when starting again from the beginning of the flow chart with the index y= 2, the index tis updated to 1. Continuing with the aforementioned steps, it is checked whether the conditions for charging the vehicle are met and whether there is free space. This means that, although only one EV is charged in each round of the flow chart, two EVs can be charged in the same time interval t. The simulation continues to analyse all EVs of the day until yis greater than Num_EV_Charge. 3.3. Tariff System 2 The second TS is similar to the first one. However, a more customised approach is considered. As described in Section 2, each EV has a maximum charging power. Three power levels are mainly found: one at 3.4 kW, another one at around 7 kW, and the last one at 11 kW. In TS2, the reduced price is offered when the PV surplus exceeds one of these 3 values. Three different tariffs are published, customised to the maximum charging power of each EV. The aim is to make the most of all the time slots where there is sufficient PV surplus to charge each EV. Thus, EVs that consume less are charged when there is less surplus, and EVs that consume more are charged when there is more surplus, thereby allowing them to consume less energy from the grid. This charging scheme is also indirect. As with TS1, the simulation of the charging patterns was processed in the recorded historical order of EV charging. In addition, once the EV connected, its charging continued at its maximum power until it was fully charged. Figure 3shows the flow chart that describes the simulation process of the charging for one day under the influence of TS2. Phases 1 and 3 of the flow charts are the same as those of Figure 2. The difference between TS1 and 2 lies in the conditions under which the reduced price is offered in the second phase. In this phase, first, the maximum charging power (EV_max_pot) of the y th EV is detected. Depending on the value of the maximum charging power of the EV, a different tariff is offered. Hence, once the maximum charging power of the EV is known, it is classified into one of these three groups: charging power (a) lower than 7 kW, (b) between 7 and 11 kW, and (c) greater than or equal to 11 kW. After classifying the y th EV within the corresponding group, the difference between the consumption and the PV production at time t(diff(t)) is calculated. If the variable diff (t) exceeds the threshold of the corresponding group, y th EV charging is carried out. For group a), for EVs with maximum charging power under 7 kW, diff(t) must exceed 3 kW. In group b), for EVs with charging power between 7 kW and 11 kW, diff(t) must exceed 7 kW. And finally, in group c), for EV maximum charging power greater or equal to 11 kW, diff(t) must exceed 11 kW. On the contrary, if the difference does not exceed the corresponding threshold, tis incremented, and the difference is recalculated and compared with the corresponding threshold for each group. Whenever diff(t) is greater than the threshold, the availability of free space at the charging point is checked. If the variable EV_Connected is less than 2, this means that there is space available, and the simulation of the EV(y) charging pattern starts. Energies 2024,17, 1806 8 of 23 Energies 2024, 17, x FOR PEER REVIEW 9 of 27 Figure 3. Flow chart related to TS2. 3.4. Tariff System 3 TS3 operates as a hybrid configuration. In addition to offering reduced prices to motivate the user to switch the charging to a more convenient time, it also acts on the charging power of EVs. Therefore, it avoids consuming energy from the grid and maximises the use of PV surplus. However, in cases where there is no more PV surplus for the rest of the Figure 3. Flow chart related to TS2. 3.4. Tariff System 3 TS3 operates as a hybrid configuration. In addition to offering reduced prices to motivate the user to switch the charging to a more convenient time, it also acts on the charging power of EVs. Therefore, it avoids consuming energy from the grid and maximises the use of PV surplus. However, in cases where there is no more PV surplus for the rest of the day, only those EVs that are still connected finish their charge by consuming energy from the grid. This process was considered to be more realistic. Energies 2024,17, 1806 9 of 23 The control of the EV charging power is carried out by the public charger that communicates with the EV batteries’ BMS. Figure 4shows the flow chart related to TS3. The three phases are different from the diagrams in Figures 2and 3. This is due to the hybrid nature of the system, in which the charging power is controlled. Energies 2024, 17, x FOR PEER REVIEW 12 of 27 Figure 4. Flow chart related to TS3. The leers A-C are connectors linking the existing flowchart to the flowchart in Figures 7, 5 and 6 respectively. Figure 4. Flow chart related to TS3. The letters A-C are connectors linking the existing flowchart to the flowchart in Figures 5–7respectively. Energies 2024,17, 1806 16 of 23 although Figure 11 shows that there are more PV surpluses, a part is consumed from the grid. This is due to the fact that TS1 is an indirect control where the charging power is not controlled. Figures 12 and 13 are related to TS2. TS2 offers a reduced price depending on the maximum charging power of each EV. In this case, the first EV with a charging power of less than 7 kW benefits from the reduced tariff when the PV surplus exceeds 3.7 kW. The second EV, with a charging power of 11 kW, benefits from the reduced tariff when the surplus exceeds 11 kW. With TS2, Figure 12 illustrates how the blue area has expanded further into the yellow curve (PV surplus). Nonetheless, a small part is still consumed from the grid. Energies 2024, 17, x FOR PEER REVIEW 19 of 27 Figure 13 shows that by offering tariffs customised to each EV’s charging power, the charging time of each EV is better distributed depending on its power and surplus. Figure 12. Basic consumption, PV production, and EV consumption under TS2 influence for 19 March 2022. Figure 13. Grid consumption, PV energy surplus, and EV consumption under TS2 influence for 19 March 2022. Finally, Figures 14 and 15 present the consumption under the influence of TS3, where, whenever surpluses occur, the reduced charging price is offered. In addition, this system operates with direct control, modifying the charging power of the EVs. As illustrated in Figure 14, the EV charging consumption (blue area) is perfectly aligned with the yellow Figure 12. Basic consumption, PV production, and EV consumption under TS2 influence for 19 March 2022. Energies 2024, 17, x FOR PEER REVIEW 19 of 27 Figure 13 shows that by offering tariffs customised to each EV’s charging power, the charging time of each EV is better distributed depending on its power and surplus. Figure 12. Basic consumption, PV production, and EV consumption under TS2 influence for 19 March 2022. Figure 13. Grid consumption, PV energy surplus, and EV consumption under TS2 influence for 19 March 2022. Finally, Figures 14 and 15 present the consumption under the influence of TS3, where, whenever surpluses occur, the reduced charging price is offered. In addition, this system operates with direct control, modifying the charging power of the EVs. As illustrated in Figure 14, the EV charging consumption (blue area) is perfectly aligned with the yellow Figure 13. Grid consumption, PV energy surplus, and EV consumption under TS2 influence for 19 March 2022. Energies 2024,17, 1806 17 of 23 Figure 13 shows that by offering tariffs customised to each EV’s charging power, the charging time of each EV is better distributed depending on its power and surplus. Finally, Figures 14 and 15 present the consumption under the influence of TS3, where, whenever surpluses occur, the reduced charging price is offered. In addition, this system operates with direct control, modifying the charging power of the EVs. As illustrated in Figure 14, the EV charging consumption (blue area) is perfectly aligned with the yellow curve (PV surplus), without consuming from the grid. The same result can be seen in Figure 15, where the EVs start charging from the first instant where PV surpluses occur, and by modifying EV charging power, no electricity is consumed from the grid. Energies 2024, 17, x FOR PEER REVIEW 20 of 27 curve (PV surplus), without consuming from the grid. The same result can be seen in Figure 15, where the EVs start charging from the first instant where PV surpluses occur, and by modifying EV charging power, no electricity is consumed from the grid. Figure 14. Basic consumption. PV production and EV charging consumption under TS3 influence for 19 March 2022. Figure 15. Grid consumption, PV energy surplus, and EV consumption under TS3 influence for 19 March 2022. It is appropriate to mention that the simulations performed are made under the assumption that human behaviour is ideal. That is, all EV users would be willing to modify their vehicle charging schedule. However, after conducting a survey among the members Figure 14. Basic consumption. PV production and EV charging consumption under TS3 influence for 19 March 2022. Energies 2024, 17, x FOR PEER REVIEW 20 of 27 curve (PV surplus), without consuming from the grid. The same result can be seen in Figure 15, where the EVs start charging from the first instant where PV surpluses occur, and by modifying EV charging power, no electricity is consumed from the grid. Figure 14. Basic consumption. PV production and EV charging consumption under TS3 influence for 19 March 2022. Figure 15. Grid consumption, PV energy surplus, and EV consumption under TS3 influence for 19 March 2022. It is appropriate to mention that the simulations performed are made under the assumption that human behaviour is ideal. That is, all EV users would be willing to modify their vehicle charging schedule. However, after conducting a survey among the members Figure 15. Grid consumption, PV energy surplus, and EV consumption under TS3 influence for 19 March 2022. It is appropriate to mention that the simulations performed are made under the assumption that human behaviour is ideal. That is, all EV users would be willing to modify Energies 2024,17, 1806 18 of 23 their vehicle charging schedule. However, after conducting a survey among the members of the CSC project, the opinion of the users and their likelihood to modify their charging schedule will be considered. Finally, Table 1compiles the SCR values of all the cases and the cost of the EV charges. Regarding the SCR values, firstly, it should be noted that on 19 March 2022, a considerable PV surplus was recorded, reaching 30 kW. Furthermore, only two EVs were registered at the charging point that day. Considering this fact, it was clearly impossible to achieve a 100% SCR. The remaining PV surplus, i.e., that not consumed by the EV, could be consumed, at least partially, by the remaining flexible loads of the other seven consumption points of the CSC. Future work will analyse how the SCR could be further increased. Table 1. SCR and charge cost values for 19 March 2022. SCR Charging Price EV1 EV2 Basic consumption 51.13% - - Original EV charge 56.54% EUR 3.19 EUR 20.34 EV charge with TS1 76.32% EUR 2.18 EUR 11.38 EV charge with TS2 77.96% EUR 1.99 EUR 10.99 EV charge with TS3 78.90% EUR 1.77 EUR 10.72 4.1.2. Different Characteristic Days Analysis Two other days with different characteristics are analysed. Figure 16 compiles the charging patterns under the influence of the three tariffs on 9 February 2022, where a single EV with a power of 11 kW was charged. The EV is charged with TS1 as the surplus exceeds the 7 kW threshold at some times. Regarding TS2, it is not applied to this EV since the surplus does not reach 11 kW. Finally, with TS3, the EV is charged as soon as PV surplus energy is available, without the need to consume from the grid. As far as SCR is concerned, with TS1, the SCR increases from 90.49% to 93.27%, while with TS3, it increases to 94.18%. Energies 2024, 17, x FOR PEER REVIEW 22 of 27 Figure 16. Charging paern and charging information under the influence of the three proposed tariffs for 9 February 2022. Figure 17. Charging paern and charging information under the influence of the three proposed tariffs for 6 April 2022. 4.2. Analysis of the Effect of the Three TSs on the SCR over One and Six Months Figure 18 presents the SCR for all days where charging occurs during the month of March 2022, as long as the basic SCR is below 100% (i.e., there are PV surpluses). SCR values are depicted for the basic consumption, the original charging recorded by the Charge and Parking manager, and the three designed TSs. TS3 always obtains the highest SCR values. On the other hand, most of the time, TS2 obtains higher SCRs than TS1. In Figure 16. Charging pattern and charging information under the influence of the three proposed tariffs for 9 February 2022. Energies 2024,17, 1806 19 of 23 Figure 17 examines the situation for 6 April 2022, where there are almost no PV surpluses, and the basic SCR is 98.66%. That day, as the surpluses do not reach the 7 kW threshold required for TS1, this TS remains unapplied. Under TS2, a reduced price is offered to the first two EVs that have a charging power of 6.6 kW. The third EV with a power of 7.4 kW is not charged. Indeed, as in TS1, the surplus does not reach 7 kW. With TS2, the SCR increases to 99.25%. Lastly, with TS3, all surpluses are consumed, reaching a SCR of 100%. Energies 2024, 17, x FOR PEER REVIEW 22 of 27 Figure 16. Charging paern and charging information under the influence of the three proposed tariffs for 9 February 2022. Figure 17. Charging paern and charging information under the influence of the three proposed tariffs for 6 April 2022. 4.2. Analysis of the Effect of the Three TSs on the SCR over One and Six Months Figure 18 presents the SCR for all days where charging occurs during the month of March 2022, as long as the basic SCR is below 100% (i.e., there are PV surpluses). SCR values are depicted for the basic consumption, the original charging recorded by the Charge and Parking manager, and the three designed TSs. TS3 always obtains the highest SCR values. On the other hand, most of the time, TS2 obtains higher SCRs than TS1. In Figure 17. Charging pattern and charging information under the influence of the three proposed tariffs for 6 April 2022. 4.2. Analysis of the Effect of the Three TSs on the SCR over One and Six Months Figure 18 presents the SCR for all days where charging occurs during the month of March 2022, as long as the basic SCR is below 100% (i.e., there are PV surpluses). SCR values are depicted for the basic consumption, the original charging recorded by the Charge and Parking manager, and the three designed TSs. TS3 always obtains the highest SCR values. On the other hand, most of the time, TS2 obtains higher SCRs than TS1. In addition, it is noticeable that on days with a high basic SCR, the increase in SCR is not as significant as on the other days. Energies 2024, 17, x FOR PEER REVIEW 23 of 27 addition, it is noticeable that on days with a high basic SCR, the increase in SCR is not as significant as on the other days. Figure 18. SCR for basic consumption, original charge, and three TS. Table 2 presents the average SCR values for each month and the overall average for all simulated days over the 6 months. The overall average basic SCR is 83.82%. The SCR obtained with the original charge of the EVs is 86.54%. The average SCR stands at 90.86% with TS1, 91.04% with TS2, and 93.09% with TS3. It can be concluded that these TSs clearly show an improvement in self-consumption, achieving an increase in SCR of 8.8% when applying TS3, for instance. Table 2. Average SCR values and SCR increase for each month. Month SCR Values (%) SCR Increase / Basic Consumption Basic Original Charge TS1 TS2 TS3 Original Charge TS1 TS2 TS3 November 92.09 94.20 95.07 95.54 98.01 1.2% 1.8% 4.3% 1.2% December 89.90 92.04 94.10 94.10 96.71 2.4% 2.4% 5.3% 2.4% January 84.42 87.58 95.27 96.51 97.73 10.4% 11.8% 13.3% 10.4% February 84.41 87.51 91.16 90.96 92.98 5.6% 5.3% 7.6% 5.6% March 79.87 82.43 87.90 88.07 89.32 8.4% 8.5% 10.3% 8.4% April 72.25 75.47 81.66 81.08 83.77 9.0% 8.3% 11.9% 9.0% Mean 83.82 86.54 90.86 91.04 93.09 6.2% 6.4% 8.8% 6.2% 4.3. Comparison of the SCR Increase in the Three TSs with the Original Charge Figure 19 presents four pie charts divided into ranges based on the SCR values. For instance, the yellow range shows all days when the SCR is between 90% and 100%. Figure 19a displays SCR values related to the original charge registered by the Charge and Parking manager. Results obtained with TS1, 2, and 3 are shown in Figure 19b–d, respectively. It is noticeable that in 21.7% of the days, SCR values below 70% are obtained with uncontrolled charging. Conversely, with TS1, these values are reduced to 10.8%, 12% with TS2, and 10.8% with TS3. On the other hand, when examining higher SCR, without TS, a SCR higher than 90% is obtained only 48.6% of the time, while these data increase to 60.8% with TS1, 62.2% with TS2, and 68.9% with TS3. Figure 18. SCR for basic consumption, original charge, and three TS. Energies 2024,17, 1806 20 of 23 Table 2presents the average SCR values for each month and the overall average for all simulated days over the 6 months. The overall average basic SCR is 83.82%. The SCR obtained with the original charge of the EVs is 86.54%. The average SCR stands at 90.86% with TS1, 91.04% with TS2, and 93.09% with TS3. It can be concluded that these TSs clearly show an improvement in self-consumption, achieving an increase in SCR of 8.8% when applying TS3, for instance. Table 2. Average SCR values and SCR increase for each month. Month SCR Values (%) SCR Increase/Basic Consumption Basic Original Charge TS1 TS2 TS3 Original Charge TS1 TS2 TS3 November 92.09 94.20 95.07 95.54 98.01 1.2% 1.8% 4.3% 1.2% December 89.90 92.04 94.10 94.10 96.71 2.4% 2.4% 5.3% 2.4% January 84.42 87.58 95.27 96.51 97.73 10.4% 11.8% 13.3% 10.4% February 84.41 87.51 91.16 90.96 92.98 5.6% 5.3% 7.6% 5.6% March 79.87 82.43 87.90 88.07 89.32 8.4% 8.5% 10.3% 8.4% April 72.25 75.47 81.66 81.08 83.77 9.0% 8.3% 11.9% 9.0% Mean 83.82 86.54 90.86 91.04 93.09 6.2% 6.4% 8.8% 6.2% 4.3. Comparison of the SCR Increase in the Three TSs with the Original Charge Figure 19 presents four pie charts divided into ranges based on the SCR values. For instance, the yellow range shows all days when the SCR is between 90% and 100%. Figure 19a displays SCR values related to the original charge registered by the Charge and Parking manager. Results obtained with TS1, 2, and 3 are shown in Figure 19b–d, respectively. It is noticeable that in 21.7% of the days, SCR values below 70% are obtained with uncontrolled charging. Conversely, with TS1, these values are reduced to 10.8%, 12% with TS2, and 10.8% with TS3. On the other hand, when examining higher SCR, without TS, a SCR higher than 90% is obtained only 48.6% of the time, while these data increase to 60.8% with TS1, 62.2% with TS2, and 68.9% with TS3. Energies 2024, 17, x FOR PEER REVIEW 24 of 27 (a) (b) (c) (d) Figure 19. (a) Percentage of SCRs by ranges for the original charge; (b) percentage of SCRs by ranges with the TS1; (c) percentage of SCRs by ranges with the TS2; (d) percentage of SCRs by ranges with the TS3. 4.4. Analysis of the Economic Savings of the Three TS Table 3 compiles the average percentage savings for each month from the EV user’s point of view. With TS3, more energy is consumed from PV surplus and less from the grid. Therefore, it is also the TS that provides the most economic savings, achieving savings of 32%. Finally, the values of the global average over the six months indicate that both TS1 and TS2 achieve the same savings of 22%. As for TS3, an average of 25% savings is obtained for each EV charge. Table 3. Monthly average savings for the three proposed tariffs. Month Mean Savings w/TS1 Mean Savings w/TS2 Mean Savings w/TS3 November 2021 10% 13% 11% December 2021 14% 24% 26% January 2022 31% 26% 27% February 2022 29% 23% 25% March 2022 30% 26% 32% April 2022 21% 21% 30% Mean 22% 22% 25% 5. Conclusions and Future Works The research study presented in this paper analyses three TSs for EV charging based on PV surplus within a CSC project. Contrary to most articles found in the scientific literature, the main objective of this work is to increase the SCR rather than to increase the Figure 19. (a) Percentage of SCRs by ranges for the original charge; (b) percentage of SCRs by ranges with the TS1; (c) percentage of SCRs by ranges with the TS2; (d) percentage of SCRs by ranges with the TS3. Energies 2024,17, 1806 21 of 23 4.4. Analysis of the Economic Savings of the Three TS Table 3compiles the average percentage savings for each month from the EV user’s point of view. With TS3, more energy is consumed from PV surplus and less from the grid. Therefore, it is also the TS that provides the most economic savings, achieving savings of 32%. Finally, the values of the global average over the six months indicate that both TS1 and TS2 achieve the same savings of 22%. As for TS3, an average of 25% savings is obtained for each EV charge. Table 3. Monthly average savings for the three proposed tariffs. Month Mean Savings w/TS1 Mean Savings w/TS2 Mean Savings w/TS3 November 2021 10% 13% 11% December 2021 14% 24% 26% January 2022 31% 26% 27% February 2022 29% 23% 25% March 2022 30% 26% 32% April 2022 21% 21% 30% Mean 22% 22% 25% 5. Conclusions and Future Works The research study presented in this paper analyses three TSs for EV charging based on PV surplus within a CSC project. Contrary to most articles found in the scientific literature, the main objective of this work is to increase the SCR rather than to increase the economic benefits of a charging station owner. In addition to the increase in the SCR, EV users also benefit from a reduction in the charging cost. As three different tariffs are designed and analysed, stakeholders can assess which case suits them best, besides having easy replicability. Overall, the behaviour of the TSs can be divided into two cases: when PV surpluses are low and when they are high. When they are low, there are often scenarios where TS1 and TS2 are not applied, as the surpluses do not reach the set threshold. TS3 is always offered. However, it is usually necessary to consume from the grid to complete the charging of the EVs. In these cases, the SCR is usually very high, reaching 100%. Regarding days where there is a significant amount of surplus, generally, all three TSs are offered. With a large surplus of PV, the SCR tends to be lower, but the increase in the SCR thanks to the TS is highlighted. As for which TS is the best, all of them have their advantages and disadvantages. TS1 is the simplest to implement. Additionally, if the surplus exceeds the 7 kW threshold, all EVs are charged. As a drawback, TS1 may lead to underutilisation of PV surplus. Considering the 6-month simulations, TS1 results in a 6.2% increase in SCR compared to the original charge, while the cost of EV charge is reduced by 22%. Concerning TS2, its main advantage lies in its significant potential on days with substantial PV surplus. However, if there are no high surplus levels, it could result in not offering the tariff to all EVs. Overall, with TS2, the SCR increase is 6.4%, and the EV charging cost is 22% lower. Finally, TS3 proves to be the most beneficial system regarding the SCR and the economic savings of the users, with a SCR increase of 8.8% and a reduction of the cost of 25%. Its advantages lie in offering a reduced price regardless of surplus levels and controlling the charging power, thereby taking advantage of all the potential for self-consumption. On the other hand, its implementation complexity and related costs stand as a disadvantage. It is important to note that direct control of EV charging is not always possible, especially when charging points are public. Therefore, although it has been shown that direct control leads to higher SCR increases, it is also essential to analyse ways in which higher SCR can be achieved through indirect control of EV charging. In addition, indirect controls are less costly to implement. Energies 2024,17, 1806 22 of 23 This study is carried out as part of a CSC project, involving eight consumption points and PV production. To conduct the calculations presented in this article, historical recorded data were used. However, in the future, the real-time EMS will use predictions of both PV consumption and production to calculate the PV surplus. Regarding other future studies, different actions are planned. On the one hand, it is important to better represent the uncertainty of human behaviour. For this purpose, a survey of CSC members conducted by the retailer will be considered. These members will be asked about their willingness to change the EV charging time. Additionally, an analysis of the scientific literature will be carried out to determine how EV users react to charging tariffs. According to the results of these studies, probability factors that depend on the level of change in the charging time will be considered in the simulation. Finally, the confidence interval of the PV production and the consumption predictions will be considered. Author Contributions: Conceptualisation, G.E. and H.C.; methodology, G.E. and H.C.; software, G.E.; validation, H.C., A.E. and T.T.L.; formal analysis, G.E.; investigation, G.E.; data curation, G.E.; writing—original draft preparation, G.E.; writing—review and editing, G.E., H.C., A.E., and T.T.L.; visualisation, G.E.; supervision, H.C.; project administration, H.C. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Data Availability Statement: Data related to this study will be available upon request. Acknowledgments: We would like to thank the renewable energy generation and consumption cooperative GoiEner for their assistance in providing us with information about the CSC project of Aduna and for the historical data collected. Conflicts of Interest: The authors declare no conflicts of interest. References 1. CORDIS EU Research Results. Enabling a Successful Transition towards Electric-Powered Road Transport. Available online: https: //cordis.europa.eu/article/id/443730-enabling-a-successful-transition-towards-electric-powered-road-transport (accessed on 17 November 2023). 2. European Parliament. EU Ban on the Sale of New Petrol and Diesel Cars from 2035 Explained. Available online: https://www.europarl.europa.eu/news/en/headlines/economy/20221019STO44572/eu-ban-on-sale-of-new-petrol-anddiesel-cars-from-2035-explained (accessed on 17 November 2023). 3. European Parliament. Reducing Car Emissions: New CO 2 Targets for Cars and Vans Explained. Available online: https://www.europarl.europa.eu/news/en/headlines/society/20180920STO14027/reducing-car-emissions-new-co2-targetsfor-cars-and-vans-explained (accessed on 17 November 2023). 4. IEA, International Energy Agency. Global EV Outlook 2023. Available online: https://www.iea.org/reports/global-ev-outlook2023 (accessed on 17 November 2023). 5. European Parliament. Car-Recharging Stations Should Be Available Every 60 km, Say MEPs. Available online: https: //www.europarl.europa.eu/news/en/press-room/20221014IPR43206/car-recharging-stations-should-be-available-every-60 -km-say-meps (accessed on 17 November 2023). 6. Mangipinto, A.; Lombardi, F.; Sanvito, F.D.; Paviˇcevi´c, M.; Quoilin, S.; Colombo, E. Impact of mass-scale deployment of electric vehicles and benefits of smart charging across all European countries. Appl. Energy 2022,312, 118676. [CrossRef] 7. Su, J.; Lie, T.T.; Zamora, R. Modelling of large-scale electric vehicles charging demand: A New Zealand case study. Electr. Power Syst. Res. 2019,167, 171–182. [CrossRef] 8. Salah, F.; Ilg, J.P.; Flath, C.M.; Basse, H.; van Dinther, C. Impact of electric vehicles on distribution substations: A Swiss case study. Appl. Energy 2015,137, 88–96. [CrossRef] 9. Guthoff, F.; Klempp, N.; Hufendiek, K. Quantification of the Flexibility Potential through Smart Charging of Battery Electric Vehicles and the Effects on the Future Electricity Supply System in Germany. Energies 2021,14, 2383. [CrossRef] 10. Venkatraman, A.; Hug, G.; Schaffner, C.; Vayá, M.G. Optimal Design of Time-of-Use Tariffs using Bilevel Optimization. In Proceedings of the IEEE PES Innovative Smart Grid Technologies Conference Europe (ISGT-Europe), Novi Sad, Serbia, 10–12 October 2022; pp. 1–5. [CrossRef] 11. Vardakas, J.S.; Zorba, N.; Verikoukis, C.V. A Survey on Demand Response Programs in Smart Grids: Pricing Methods and Optimization Algorithms. IEEE Commun. Surv. Tutor. 2015,17, 152–178. [CrossRef] 12. Herter, K. Residential implementation of critical-peak pricing of electricity. Energy Policy 2007,35, 2121–2130. [CrossRef] 13. Li, R.; Wang, Z.; Gu, C.; Li, F.; Wu, H. A novel time-of-use tariff design based on Gaussian Mixture Model. Appl. Energy 2016,162, 1530–1536. [CrossRef] Energies 2024,17, 1806 23 of 23 14. Rahman, A.; Aziz, T.; Masood, N.-A.; Deeba, S.R. A time of use tariff scheme for demand side management of residential energy consumers in Bangladesh. Energy Rep. 2021,7, 3189–3198. [CrossRef] 15. Kandpal, B.; Pareek, P.; Verma, A. A robust day-ahead scheduling strategy for EV charging stations in unbalanced distribution grid. Energy 2022,249, 123737. [CrossRef] 16. Hu, Z.; Zhan, K.; Zhang, H.; Song, Y. Pricing mechanisms design for guiding electric vehicle charging to fill load valley. Appl. Energy 2016,178, 155–163. [CrossRef] 17. Zhang, L.; Yin, Q.; Zhu, W.; Lyu, L.; Jiang, L.; Koh, L.; Cai, G. Research on the orderly charging and discharging mechanism of electric vehicles considering travel characteristics and carbon quota. IEEE Trans. Transp. Electrif. 2023. [CrossRef] 18. Tao, J.; Huang, D.; Li, D.; Yang, X.; Ling, C. Pricing strategy and charging management for PV-assisted electric vehicle charging station. In Proceedings of the 2018 13th IEEE Conference on Industrial Electronics and Applications (ICIEA), Wuhan, China, 31 May–2 June 2018; pp. 577–581. [CrossRef] 19. Dai, Y.; Qi, Y.; Li, L.; Wang, B.; Gao, H. A dynamic pricing scheme for electric vehicle in photovoltaic charging station based on Stackelberg game considering user satisfaction. Comput. Ind. Eng. 2021,154, 107117. [CrossRef] 20. Gudmunds, D.; Nyholm, E.; Taljegard, M.; Odenberger, M. Self-consumption and self-sufficiency for household solar producers when introducing an electric vehicle. Renew. Energy 2020,148, 1200–1215. [CrossRef] 21. Roselli, C.; Sasso, M. Integration between electric vehicle charging and PV system to increase self-consumption of an office application. Energy Convers. Manag. 2016,130, 130–140. [CrossRef] 22. van der Kam, M.; van Sark, W. Smart charging of electric vehicles with photovoltaic power and vehicle-to-grid technology in a microgrid; a case study. Appl. Energy 2015,152, 20–30. [CrossRef] 23. Barone, G.; Brusco, G.; Menniti, D.; Pinnarelli, A.; Polizzi, G.; Sorrentino, N.; Vizza, P.; Burgio, A. How Smart Metering and Smart Charging may Help a Local Energy Community in Collective Self-Consumption in Presence of Electric Vehicles. Energies 2020,13, 4163. [CrossRef] 24. Fachrizal, R.; Munkhammar, J. Improved Photovoltaic Self-Consumption in Residential Buildings with Distributed and Centralized Smart Charging of Electric Vehicles. Energies 2020,13, 1153. [CrossRef] 25. Zapirain, I.; Etxegarai, G.; Hernández, J.; Boussaada, Z.; Aginako, N.; Camblong, H. Short-term electricity consumption forecasting with NARX, LSTM, and SVR for a single building: Small data set approach. Energy Sources Part A Recovery Util. Environ. Eff. 2022, 44, 6898–6908. [CrossRef] 26. Etxegarai, G.; Zapirain, I.; Camblong, H.; Ugartemendia, J.; Hernandez, J.O. Curea. Photovoltaic Energy Production Forecasting in a Short Term Horizon: Comparison between Analytical and Machine Learning Models. Appl. Sci. 2022,12, 12171. [CrossRef] 27. Euskalmet. Available online: https://www.euskalmet.euskadi.eus/inicio/ (accessed on 17 November 2023). 28. GoiEner. Renewable Energy Generation and Consumption Cooperative. Available online: https://www.goiener.com/es/ (accessed on 17 November 2023). Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.