Industrial Application of Photovoltaic Systems with Storage for Peak Shaving: Ecuador Case Study ^†

Abstract

Decentralized generation has gained importance in the energy industry, since self-consumption with renewable resources presents attractive costs and allows load management actions. In this sense, photovoltaic generation systems are a promising technology. This work presents a proposal for a peak shaving system using solar photovoltaic (PV) energy and a battery storage system, known as battery energy storage systems (BESS), to be installed by an industrial customer to reduce energy consumption during peak hours. For the study, a hybrid approach is presented, starting from deterministic variables, such as the demand curve of the industry under study, and the generation of stochastic variables, such as the energy production of the photovoltaic system. For the analysis of the proposed peak shaving system, the design and sizing of the photovoltaic systems are developed in a base case of self-generation and an optimized system for the system to cover the energy demand generated during peak hours. The technical–economic study carried out in the research allowed us to determine the optimal power of the photovoltaic system with the storage system. The proposed system allows the integration of a peak shaving strategy from a certain power limit, in order to cover the peak demand over this power limit, allowing the system to be profitable under the current regulations and standards in Ecuador.

1. Introduction

In recent years, seeking a balance between demand and generation is ideal for properly functioning the power electric system (PES). The gradual increase in electricity demand worldwide makes the installation of alternative energy generation systems indispensable [1]. Technological development has allowed a large penetration of renewable generation systems to displace fossil resources [2,3]. Following the important objectives of the Paris Agreement, a massive reduction in CO₂ emissions should be achieved by reducing energy consumption by implementing frugality measures, energy efficiency, rapid diffusion, and the optimal use of renewable energies [4]. On the other hand, in the last decade, traditional electrical grids have undergone a migration to a more modern network in which a complex communications network is included that allows the monitoring, control, and management of electrical systems, which is known as a Smart Grid [4].

The increasing integration of renewable energy resources (RES) in distribution systems with high and inherent variable production, with the inclusion of new assets, such as electric vehicle charging stations, and the incorporation of new industries worldwide, is leading to increasing challenges in grid management. The loads presented today in the country’s different sectors, in addition to having a volatile demand profile, can also require large amounts of power from the grid in various and short periods of the day, to which a traditional power grid is not prepared to respond [5].

Electric power is one of the main inputs for economic and social development in the different regions of the world as it drives the growth of industrial production. To ensure the best conditions for productivity and competitiveness, it is necessary to propose guidelines allowing scenarios of greater energy efficiency in which the participation of renewable energies is more actively involved [6]. Industry is considered one of the sectors with the highest energy consumption. It has been one of the main axes for the application of energy efficiency improvements, including integrating renewable energies.

The different consumption patterns of industries, and the forms of billing considered in the different countries of the region and the world, are one of the main problems for the increase in production in industries [7,8]. The competitiveness of industrial companies depends, among other important factors, on low energy costs. Industrial customers use energy-intensive equipment at short time intervals during the day, so the additional costs of keeping up with peak demand are passed onto customers as demand charges. The problem of peak demand reduction has also been studied at the local level; thus, in 2017, in reference [9], a study of the different methods of peak demand reduction and their feasibility in Ecuador was presented. A study of the different peak shaving strategies developed at the international and regional levels is presented in ref. [9]. A regulatory analysis for implementing peak shaving strategies in Ecuador is also presented. The implementation of demand response as a peak shaving strategy is evaluated on the demand curve of the National Interconnected System and a typical load profile of the industrial sector. The displacement of demand from the peak period to the base period allows a 20% load reduction, which can be achieved using load controllers.

2. Proposal Methodology

This paper proposes optimizing a battery storage system as a peak shaving strategy for a commercial medium voltage, MV, customer. Figure 1 presents the network connection schematic of the study network. As seen in Figure 1, the schematic presents a DC system coupling connection, an industrial load, and a grid connection point [10].

The evaluation methodology proposed for optimizing the photovoltaic generation system with battery storage as a peak shaving self-consumption system for large consumers is presented in the following paragraphs.

The methodology starts with the analysis of the demand profile of an industrial consumer in order to determine the optimal power of the photovoltaic generation system with battery storage. Once the average daily demand profile of the industry is analyzed and defined, we proceed to the sizing of the PV system with storage. The sizing of the system is carried out by checking the coupling criteria of the PV generator–inverter.

To make the system economically viable, an optimization of the photovoltaic system with storage is proposed by determining a limit power of the network, which allows for reducing the size of the installation so that the system covers the peaks of the demand in case of exceeding the limit power.

A diagram of the methodology proposed in this work is shown in Figure 2.

2.1. Step 1: Determine the Demand Profile

To determine the average demand profile, a classification of working days, weekends, and holidays was made to have a total reading of the industry’s behavior. This allows us to have the average demand profile as a starting point and identify the maximum demand for the sizing of the photovoltaic system with BESS.

Once the classes were determined, the maximum demand curve for the most critical day was identified, allowing the photovoltaic installation’s initial sizing (Figure 3). The energy consumed in the industry on the most critical day is 2.7 MWh/day.

2.2. Step 2: Sizing the Solar Photovoltaic System

Once the maximum daily demand curve has been obtained, the optimum power of the installation is calculated using the following Equation (1) [11,12]:

$$P_{Otp}=\frac{\sum _{i=mes 1}^{month 12}{Emonth}_i}{{Plant }_{Factor}\times 8760}$$

(1)

where the Emonthly is calculated with the maximum daily demand to supply the most critical day of demand.

With the P_Otp data, we proceeded to simulate the photovoltaic installation. The solar system sizing was carried out using the coupling criteria of the generated PV-inverter [13,14]. To calculate the number of strings in parallel, as well as the number of panels per string for this case study, the technical characteristics of the 340 Wp Jinko polycrystalline photovoltaic panel and the 500 kW Hyundai HPC 500HL inverter are used.

3. Case Studies

Once the simulation has been performed in the PVsyst software (Version 7.2), the energy production data for all the days of the simulation year are imported with an hourly interval for which the most critical day of the year is analyzed, determined as June 21, in which the radiation and temperature data presented in Figure 4a,b, respectively, have been obtained.

Based on the results obtained, two-scenario system analysis is proposed:

-

Scenario 1 (Self-consumption): Peak shaving system with renewable self-generation and battery storage, limiting the maximum consumption of the grid and the injection of energy into the grid.

-

Scenario 2 (Optimized): Peak shaving system with renewable self-generation and battery storage, limiting grid consumption to a power limit so that the photovoltaic system and storage can supply the energy needed during peak hours if the grid power limit is exceeded.

3.1. Scenario 1

For Scenario 1, self-consumption, the energy performance of the installation was evaluated based on the data presented in

Table 1. Simulation data for Scenario 1.

Photovoltaic Modules	Jinkosolar-JKM 340PP-72
Number of Modules	1656
Modules in series	18
Modules in parallel	92
Total area of modules	3213 m2
Inverter	Hyundai HPC-500HL-EU
Number of Inverters	1
Nominal Power of the Installation	563 kWp
Rated AC power	500 kWCA
Battery	LG Chem Rack R800
Battery Type	Li NMC
Battery in series	1
Parallel Cells	29
Depth of discharge	95%
Stored Energy	1287 kWh
Overall System Capacity	1867 Ah
Annual Industry Requirement	818 MWh/year
Industry Average Load	93.3 kW
Maximum Industry Load	143.1 kW

.

A photovoltaic system with battery energy storage is simulated using the self-consumption strategy. This strategy maximizes self-consumption by prioritizing the user’s needs. The photovoltaic energy is used first to feed the load and then to charge the battery, which will later be used to cover the demand outside the hours of radiation. The objective of this scenario is to minimize the energy consumed by the grid; it is worth mentioning that, based on the proposed scheme, the storage system is not charged from the grid. The demand profile and the energy generated by the system are presented in Figure 5.

Based on the data obtained from the simulation, as well as the industry demand measurement data, the following equation can obtain the battery capacity (2) [15,16]:

$$C_{Batería}=\frac{\frac{E_{Surplus }}{Inverter Efficiency }}{System Voltage}$$

(2)

The batteries are sized with the help of Equation (2) and the data obtained from the equipment specification sheets. Based on the maximum and minimum DC input voltage of the inverter, which are between 450 and 850, respectively, a commercial battery voltage of 725 VDC is established.

Figure 6 shows the temporal evolution of the power balance of the system for an average week in July, the most critical month of the year. As can be seen, the most critical day of the week is when there is a participation of energy from the grid, while the other days are covered entirely by the photovoltaic system and the batteries.

Once the annual evaluation of the installation has been carried out, it can be observed that the system stores approximately 44% of the surplus energy, so 20.7% of the annual energy demand will have to be covered by the grid. This consumption of 142 MWh per year represents a saving of 82% of the total energy consumed annually by the industry under study.

3.2. Scenario 2

Once Scenario 1 of the solar photovoltaic system with storage has been obtained, the system optimization is carried out as a peak shaving strategy in the system under study. It is important to determine a power limit that allows the calculation of the photovoltaic system’s appropriate power and storage capacity.

The PV system’s optimal power should supply energy above the determined limiting power (Figure 7). This will allow the PV system to be sized in such a way that it can store only the energy needed to reduce the industry’s peak demand that is within the peak hours above the determined power limit. The grid input power limit P_lim is defined based on billable demand, which indicates that it cannot be less than 60% of the maximum demand recorded in the previous 12 months, so the maximum demand is calculated with Equation (3) [17].

$$P_{lim}={MaximumDemand}_{\mathrm{A}\mathrm{n}\mathrm{n}\mathrm{u}\mathrm{a}\mathrm{l} }\times 60\% \mathrm{kW}$$

(3)

On the other hand, in Scenario 2 (optimized), as a peak shaving strategy, the system was optimally sized so that the system limits the power input from the grid to 60% of the maximum generated demand. Therefore, the PV system with storage overseas covers the peaks generated above the determined power limit. The simulation data of the optimized case are presented in

Table 2. Simulation data for Scenario 2.

Photovoltaic Modules	Jinkosolar-JKM 340PP-72
Number of Modules	864
Modules in series	18
Modules in parallel	48
Total area of modules	1676 m2
Inverter	Hyundai HPC-500HL-EU
Number of Inverters	1
Nominal Power of the Installation	297 kWp
Rated AC power	250 kWCA
Battery	LG Chem Rack R800
Battery Type	Li NMC
Battery in series	1
Parallel Cells	5
Depth of discharge	95%
Stored Energy	222 kWh
Overall System Capacity	322 Ah
Annual Industry Requirement	818 MWh/year
Industry Average Load	93.3 kW
Maximum Industry Load	143.1 kW

.

Based on the optimization of the solar photovoltaic system’s capacity, the system’s optimal capacity is 242 kW, so the evaluation of the most critical day was performed with an installation of 250 kWp based on the existing commercial equipment (Figure 7).

Figure 8 presents the time evolution of the power balance of the optimized system for an average week of July, the most critical month of the year [18]. In the figure, it is possible to appreciate the energy peak shaving from the grid solved by the storage system if the established power limit is exceeded.

Once the system’s annual performance has been evaluated, it can be observed that the system stores approximately 14% of the energy generated by the PV system. Therefore, the grid system is predominant.

4. Results

As can be seen in Figure 9a for Scenario 1 (self-consumption), the PV system with storage was sized to cover the entire demand of the industry. This means that the PV system will generate a total of 807 MWh/year to cover the annual demand of the industry, which is equal to 818 MWh/year. The designed system provides approximately 87% of the industry’s total demand. The photovoltaic system can cover the demand during the hours when there is radiation and, in turn, generate surplus energy that is stored in the batteries to cover peak demand throughout the day [19].

In case two, presented in Figure 9b, the PV system can cover the demand in the hours when there is radiation and, in turn, generate surplus energy that is stored in the batteries to cover peak demand throughout the day, when the determined power limit is exceeded.

The results obtained for the two cases evaluated are shown in

Table 3. Energy results of the evaluated cases.

Scenarios Evaluated	Power Produced [MW/year]	Stored Energy [MW/year]	Direct Use Energy [MW/year]	Power Injected [MW/year]	Power Consumed from the Grid [MW/year]
Without Peak Shaving System	0	0		0	817
Case 1 (Self-consumption)	870	355	360	92	140
Case 2 (Optimized)	421.9	58.6	303	59	459

. As can be seen in the table, even though the self-consumption system reduces energy consumption from the grid by 80% compared to the optimized system, which only reduces energy consumption by 43%, when evaluating the system according to the current ARCENNER 001/21 regulation, which specifies a net energy consumption, it is extremely important to evaluate the financial aspect of the installation [20].

The most relevant results obtained in the financial analysis are presented in

Table 4. Most relevant results of the evaluated cases.

Scenarios Evaluated	Investment [USD]	Savings due to Reduced Purchase of Energy [USD/year]	Payback Period [years]	ROI [%]	LCOE [USD/kWh]
Without Peak Shaving System	0	0	0	0	0
Case 1 (Self-consumption)	1,009,328.00	55,746.06	23	−9.2	0.32
Case 2 (Optimized)	424,198.00	16,360.00	16	22.7%	0.09

. The photovoltaic system with battery storage for self-consumption as a peak shaving strategy, in Scenario 1, allows for greater savings in the purchase of energy; however, it requires a higher investment and, by managing an energy netting system, it is not profitable, presenting a negative ROI. However, case two, with a lower power of the photovoltaic system as well as the storage system, presents a better solution with IER 22.7% in approximately 16 years.

5. Conclusions

One of the key points for the design of photovoltaic stations with storage is the quality of the demand profile data that can be obtained, as well as the characterization of these data to obtain a correct starting point. Likewise, the use of computer tools specialized in the design and sizing of photovoltaic installations allows the validation of mathematical models and provides support in terms of meteorological data information for the sizing of these. One of the most critical points in the development of these systems is the great variety of manufacturers of elements that exist today, so having a reference of validated elements in operation is a great solution. Even though the manufacturing costs of photovoltaic systems have been decreasing year by year, storage systems, such as lithium technology, still do not reach competitive prices in the countries of the region due to the subsidies that each country has. However, the integration of the two technologies can be economically attractive for the development of projects smaller than 2 MW for commercial and industrial consumers.

As has been evaluated in this work, a high storage capacity can double the cost of the project, due to the high prices of the elements that make up the installation. Proposing a photovoltaic system with storage as self-generation is not profitable in commercial and industrial sectors due to the energy costs in Ecuador. Despite the existence of a tariff scheme with hourly demand consumption, electricity costs are not comparable to the costs of energy produced. Therefore, the proposal of “peak shaving” with a photovoltaic system with optimal storage to reduce peak demand from limited power is a viable strategy.

The optimization of the system reduces the initial investment of the project by 50% and, since it is carried out within the regulatory framework of Ecuador, where there is a netting of energy, the proposal presents an economic feasibility from year 16 onwards. The results obtained indicate that one of the best solutions for the industry is the installation of 250 kWp and battery storage with a capacity of 222 kWh.