*Article* **Determining the Power and Capacity of Electricity Storage in Cooperation with the Microgrid for the Implementation of the Price Arbitration Strategy of Industrial Enterprises Installation**

**Rafał Ku ´zniak, Artur Pawelec , Artur S. Bartosik \* and Marek Pawelczyk**

Department of Production Engineering, Kielce University of Technology, Al. Tysiaclecia P.P. 7, 25-314 Kielce, Poland; rkuzniak@tu.kielce.pl (R.K.); apawelec@tu.kielce.pl (A.P.); m.pawelczyk@tu.kielce.pl (M.P.) **\*** Correspondence: artur.bartosik@tu.kielce.pl

**Abstract:** The growing worldwide costs of energy produced as a result of conventional fuel combustion, the limited capacity of the distribution grid, and the growing number of unstable installations based on renewable energy sources increase the need to implement systems of stabilization and regulate loads for end users. The battery energy storage system (BESS) that operates in the internal microgrid of an enterprise enables the management of the accumulated energy in any time zone of the day. Using a price arbitrage strategy with an electricity storage facility, we can reduce the cost of high electricity prices during peak demand periods. This study aims to determine the most effective method of setting up the capacity and electrical power of an energy storage system operating in a microgrid, in an enterprise to implement a price arbitration strategy. Such a method should include consideration of the characteristics of the demand profile of consumer systems, the charges related to electricity, and electricity storage costs. The proposed deterministic method is based on the use of a defined parameter, "marginal income elasticity". In this study, the size of energy storage refers to the power and electric capacity of BESS that are used for the implementation of the price arbitrage strategy.

**Keywords:** BESS management; price arbitration; shift load; microgrid; energy efficiency

#### **1. Introduction and Review of the Literature Related to the Optimal Power and Capacity of an Electric Energy Storage System**

In recent years, the energy market has seen an increase in interest in electricity storage, resulting in the development of scientific research on various working conditions and the strategies for their operation. Numerous studies have presented reviews of energy storage technologies in terms of their applications in microgrids [1–8]. Researchers presented the main functionalities that can be implemented in microgrids, including the absorption of energy from renewable sources, improvements in the quality parameters of electricity, peak shaving strategies, and price arbitrage and time shifts [9,10]. One of the main goals of the research has been to develop a methodology to achieve the optimal parameters of energy storage from an economic point of view, taking into account the investment and operating costs and the technical and economical parameters of various technologies that have potential for broad usage [11–15].

Research on the battery energy storage system (BESS) that uses deterministic and stochastic methods to determine the cost effectiveness of storage technologies was presented in previous works [16–18]. Adopted models were analyzed, including the costs of individual BESS technologies, the degradation of the capacity over time, and the losses of capacity during the discharge readiness period. The application of the integrated model to define and select energy storage parameters was presented in previous works [19–21]. These works presented models that included the implementation of thermal, electrical, and aging processes, as well as various sources and parameters that characterized the

**Citation:** Ku ´zniak, R.; Pawelec, A.; Bartosik, A.S.; Pawelczyk, M. Determining the Power and Capacity of Electricity Storage in Cooperation with the Microgrid for the Implementation of the Price Arbitration Strategy of Industrial Enterprises Installation. *Energies* **2022**, *15*, 5614. https://doi.org/10.3390/ en15155614

Academic Editor: Tomasz Rokicki

Received: 31 May 2022 Accepted: 23 July 2022 Published: 2 August 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 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/).

production of electricity within the microgrid. Stochastic predictive models, using 24 h wind force forecasting to optimize the power and capacity of energy storage in microgrid systems, were proposed in [22].

The problem of selecting the power and capacity of energy storage to balance microgrids, based on local results with various integrated renewable energy sources and various types of energy storage, was studied [23–26]. These studies presented mathematical models of microgrid systems with sources such as photovoltaic panels and wind turbines. On the basis of actual data characterizing the demand for electric energy, a simulation of the microgrid operation was performed, depending on the variability of electric energy demand in an island system. The use of electricity storage systems to increase the share of energy generated from renewable sources was also considered in previous works [27–30].

The issue of selecting the size of energy storage for households with their own renewable energy production systems for time-shifting functionality was discussed in an earlier study [31]. As a result of that research, it was shown that group energy storage, compensating for the flow of energy transferred to the external grid, is more profitable than individual storage systems. Studies on the maximization of expected daily economic profit, obtained using the time-shifting strategy to postpone the production of renewable energy, were presented in [32].

Korpikiewicz [33] broadly presented the conditions required for the operation of autonomous energy storage to implement a price arbitrage strategy, i.e., the use of variable energy rates throughout the day to reduce energy demand in periods of high energy prices and increase demand in periods of low energy prices. Algorithms describing the logic of determining the BESS charging and discharging cycles to optimize the operation of the system have been presented with the basic technical and operational data of BESS, which were obtained in various energy storage technologies [34].

A very important problem that should be considered when installing BESS in enterprises is the safety of the system. Particular attention should be paid to fire hazards posed by lithium-ion batteries. Therefore, the safety of BESS is the subject of intensive studies conducted by scientists [35], engineering associations [36], territorial units [37], and manufacturers who implement their own fire protection concepts [38]. Despite intensive research, there is still a lack of effective and rapid methods that could be widely used.

In summarizing the literature review, it should be noted that there are several studies on the use of BESS to implement a price arbitrage strategy. Most of the works dedicated to price arbitration focus on separated systems, supporting the distribution network and operating autonomously with constant and fixed charge and discharge values [39–41]. The models of energy storage operation presented in the literature confirm that the operating profit resulting from the use of the storage facility for price arbitrage is proportional to the total storage capacity.

In the available research, there are several studies on the profitability of an energy storage system management strategy that take into account the constraints associated with the actual energy demand and power of microgrids in production plants. Restrictions resulting from legal regulations on billing for the production of energy in a given country or in real microgrid systems may cause the benefits of using storage systems to decrease non-linearly with an increase of BESS capacity and power.

This study aims to determine the most effective method for setting up the capacity and electrical power of an energy storage system operating in an enterprise's microgrid to implement a price arbitration strategy. Our research considered the existing technical and cost limitations in real enterprises that lead to a decrease in the effectiveness of the implementation of a price arbitration strategy. This paper defines the indicators for assessing the effectiveness of this strategy, and on that basis, we propose a determination of the effective boundary for BESS size. The microgrid system of the enterprise is a separate power installation, created from the load devices, active energy storage, or generation of assets with a control-and-regulation system, that is capable of managing the energy and electric power balance within the enterprise, connected to the distribution system

operator's (DSO's) network. In this study, we assumed that the microgrid system managed an electricity storage installation and industrial power load within selected companies that were connected to a medium voltage grid. tor's '

#### **2. BESS Work Strategy, Characteristics of Companies Selected for Research, and the Chosen BESS Model**

The paper analyzes the use of BESS in terms of representative functionality for the electricity market, that is, price arbitrage. Price arbitration is based on the use of daily differences in unit prices of electricity. The essence of this strategy is the storage of energy purchased from the external grid in the price valley and then unloading the battery storage to supply the microgrid loads at times when the unit energy prices are the highest. When examining the conditions of this BESS functionality, one should consider the electricity prices [PLN/kWh] based on the offers of trading companies in the competitive market and the variable rates [PLN/kWh] for the distribution services that are included in the tariff of the appropriate distribution system operator that is approved by the President of the Energy Regulatory Office (ERO). The second option is to consider electricity prices according to the rates of the Polish Power Exchange Stock Market.

For customers who are billed according to the tariffs of energy companies, price arbitrage may be applied by selecting multizone tariff groups. The most diversified prices are in the B23 tariff group. There are three time zones in this group: S1 is the morning peak, S2 is the afternoon peak, and S3 is the the rest of the day. At all hours of the day on Saturdays, Sundays, and public holidays, energy is billed in the S3 zone as the rest of the day. The distribution of hours according to UTC + 1 time (coordinated universal time + 1 h) in tariff group B23 is presented in Table 1. The multi-tariff time zones included in Table 1 are typical for Poland and are applied in the tariffs of the four largest distribution network operators in Poland.


**Table 1.** Distribution of time zones in the B23 tariff group for individual months in UTC+1 time. Yellow background with 1 is zone S1; red background with 2 is zone S2; green background with 3 is zone S3.

In the case of enterprises in the B23 tariff group, the price arbitration strategy is based on avoiding the purchase of electricity from the external grid when the variable unit rates in PLN/kWh are the highest, according to variable fees during the afternoon peak in the S2 zone.

An important element of electricity billing that should be considered when applying the price arbitrage strategy is the capacity fee. In Poland, beginning on 1 January 2021 as part of the implementation of the capacity market, an additional component was introduced to settlements for distribution services: i.e., the capacity fee in PLN/kWh. The rate of the capacity fee is published annually by the President of the Energy Regulatory Office (ERO), along with the designated hours of peak power demand during the day, at which times this component should be added to the consumed kilowatt hours Because the amount of this fee depends on the hours of the day, it increases the daily difference in prices related to electricity consumption and affects the application of the price arbitration strategy [42].

The capacity fee in Poland, after 1 January 2021, applies to all enterprises and is charged from 07:00 to 22:00 on business days. The hours in which the capacity charge applies according to UTC + 1 are presented in Table 2.


**Table 2.** Schedule of hours in UTC+1, time of charging the capacity fee valid from 2021; red background on 1.

15 The price arbitrage strategy research was carried out for three different companies on the basis of the time series of the average 15 min electric load power consumed by the companies and recorded by measuring systems in an annual period. Data recorded in individual 15 min intervals were marked as load power *PL*<sup>15</sup> [kW]. The companies selected as research subjects were marked with letters A, B, and C, to which the year of registration of the time series tested was added (2018 and 2019). The selected companies carry out production activities with the use of various technologies and in various specialties. The enterprises are powered from the medium voltage power grid. Companies A and C are characterized by a constant level of energy consumption on working days; their work is carried out in a three-shift system. Enterprise B works in two shifts, only on working days. The differences in the weekly work organization of the enterprises are visible in Figure 1, which presents the average weekly profiles of power demand in 15 min power-demand intervals.

The characteristics of the organization of the work in the surveyed enterprises and their power demands are also illustrated by the coefficients of variation in the statistics of the 15 min power consumption time series, as shown in Table 3.

**Figure 1.** Comparison of the average weekly power demand profiles in [kW] for an annual period.


**Table 3.** Statistical data of the 15 min load power time series of the enterprises.

Based on the regulations that govern the application of tariff rates by energy companies and the settlement rules on the energy market and the capacity market, simulations of the BESS effect for the "price arbitrage" functionality were carried out. As part of the research, analyses of the time series of parameters characterizing the operating state of BESS, which were created as a result of the simulation of its operation in the microgrid system and the size of settlement data at the point of common coupling (PCC), were carried out.

" " The research consisted of adopting subsequent parameters characterizing the size of the energy storage, increasing them by a fixed value, and simulating their operation for a "price arbitrage" strategy in 15 min intervals for the entire annual measurement period. To investigate the price arbitrage strategy related to electricity, the input was the increasing capacity in kWh. The results of successive "k" simulations at the given BESS capacities were the quantities that described the effects of BESS.

" " " " In the case of microgrids, price arbitrage may be carried out by charging the energy storage from the power grid operated by one of the DSOs in periods when the cost of electricity from internal microgrid sources is lowest. Energy storage is discharged through the receiving systems in periods when the cost of electricity from the DSO grid is highest. As part of the price arbitrage implemented in the microgrid, financial benefits are obtained by taking advantage of the price difference between the avoided purchase of energy from the grid during discharging and the price of energy supplied by BESS. Typical BESS operating states for the implementation of the price arbitrage strategy are presented in Figure 2.

The revenue obtained resulting from the use of BESS for price arbitration, *REVBA*, for one charging and discharging cycle, results from the use of energy stored in the *EBA* energy storage for the company's needs, taking into account the depth of discharge planned for the price arbitration, *DoDA*, and the maximum price, *CESmax*, of energy not taken from the DSO grid in a given zone S, as presented by Equation (1).

$$REV\_{BA} = E\_{BA} \cdot (1 - DoD\_A) \cdot C\_{ES\max} \tag{1}$$

**Figure 2.** Examples of typical BESS operating states with markings of characteristic measures for price arbitrage.

the company's The operating costs of price arbitration with BESS are marked as *O<sup>A</sup>* for one charge and discharge cycle. The costs include charges for energy collected during BESS charging from the DSO network in the S zone at the minimum price, *CESmin*, on a given day. The operating costs include the efficiency of the storage system, ηB, resulting from losses related to the conversion of AC/DC and DC/AC in the charge-and-discharge cycle. This cost is written as follows:

$$O\_A = \frac{1}{\eta} \underset{B}{E\_{BA}} \cdot (1 - DoD\_A) \cdot \mathbb{C}\_{ESmin} \tag{2}$$

The operating income for a single cycle, *INCBA*, with a multi-zone tariff group, can be written as follows:

$$I\text{NC}\_{BA} = \text{REV}\_{BA} - \text{O}\_A = \text{E}\_{BA} \cdot (1 - DoD\_A) \cdot \left(\text{C}\_{\text{ESmax}} - \frac{1}{\eta} \text{ }\_{\text{B}}^{\text{C}} \text{C}\_{\text{ESmin}}\right) \tag{3}$$

 = 1 ∙ (1 − ) ∙ 1 If BESS is used for price arbitrage in the microgrid system, based on the energy supplied by external suppliers from the DSO's grid, the purchase of *O<sup>E</sup>* electricity and the cost of providing the distribution service in the *O<sup>D</sup>* variable part should be considered when calculating revenues and costs. For this reason, the income for a single discharge cycle for microgrids should be calculated by including the separate revenues and costs of electricity, i.e., *REVBE* and *OE*, and for the distribution service, i.e., *REVBD* and *OD*:

$$\text{INC}\_{BA} = [(\text{REV}\_{BE} - O\_E) + (\text{REV}\_{BD} - O\_D)]\_\prime \tag{4}$$

' For a company in the B23 tariff group, the income for a given billing period resulting from the use of the storage system in *n<sup>i</sup>* cycles, assuming one cycle per day and considering the discharge in the peak zones S1 and S2 and the resale of surplus energy at market prices, *CErk* , to the DSO grid, is calculated according to the following relationships:

$$\begin{aligned} \text{INC}\_{BA} &= \eta\_{i} \cdot \text{E}\_{BA} \left\{ \left[ \left( (1 - DoD\_{\text{S1}}) \cdot \text{C}\_{\text{ES1}} + (1 - DoD\_{\text{S2}}) \cdot \text{C}\_{\text{ES2}} + (1 - DoD\_{\text{S2}}) \cdot \text{C}\_{\text{EK}} \right) - \frac{1}{\eta\_{B}} \cdot \text{I}\_{\text{ES1}} \right] \cdot \text{I}\_{\text{ES2}} \cdot \text{I}\_{\text{ES3}} \cdot \text{I}\_{\text{ES5}} \cdot \text{I}\_{\text{ES6}} \cdot \text{I}\_{\text{ES7}} \cdot \text{I}\_{\text{ES8}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES8}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot \text{I}\_{\text{ES9}} \cdot$$

 = ∙ {[[(1 − 1) ∙ 1+(1 − 2) ∙ 2+(1 − ) ∙ ] − 1 ∙ The unit prices of electricity, *CES*1, *CES*2, *CES*3, and *CErk* , are the prices that are accepted for settlements from the offer of electricity trading companies. The unit variable rates for the distribution service, *SZS*1, *SZS*2, *SZS*3, *SZJ* , and *SOze* , *Skog* , *SPcap*, are calculated or adopted by the territorially competent distribution system operators in the form of a tariff approved by the President of the Energy Regulatory Office.

#### 1 *2.1. Assumptions Made in the Simulation Model*

(1 − ) ∙ 3

(2 + + + + ) − ∙ (1 − ) ∙ (S3 + + + )]} The simulations were carried out for enterprises A and B based on data from 2018 and 2019 (the series were marked as A2018, A2019, and B2018, B 2019) and for enterprise C based on data from 2019 (the series was marked as C2019).

] + [(1 − 1) ∙ (1 + + + + ) +(1 − 2) ∙

The following assumptions were made for the simulation model:


$$P\_{\rm BC} \le P\_{\rm II} - P\_{\rm L15} \tag{6}$$

On this basis, a condition was formulated defining the maximum charging power for each of the compartments:

$$
\max \mathbf{E}\_{BC} \le \frac{15[min]}{60[min]} \cdot \mathbf{1}[h] \cdot (P\_{\mathbf{U}} - P\_{\mathbf{L}15}) \tag{7}
$$


$$\begin{aligned} IN\mathbf{C}\_{BA} &= \boldsymbol{\eta}\_{i} \cdot \mathbf{E}\_{BA} \quad \left[ (1 - DoD\_{S2}) \cdot (\mathbf{C}\_{ES2} + \mathbf{S}\_{ZS2} + \mathbf{S}\_{ZI} + \mathbf{S}\_{O2\varepsilon} + \mathbf{S}\_{kog} + \mathbf{S}\_{Pcap}) \\ &- \frac{1}{\eta\_{B}} (1 - DoD\_{A}) \cdot \left( \mathbf{C}\_{ES3} + \mathbf{S}\_{ZS3} + \mathbf{S}\_{ZI} + \mathbf{S}\_{O2\varepsilon} + \mathbf{S}\_{kog} \right) \right] \end{aligned} \tag{8}$$


$$DoD = DoD\_A = DoD\_{\max} = 20\% \tag{9}$$

15. CAPEX costs and the BESS life cycle were not considered in the study. The analyses were limited to the operational economic effects realized by the price arbitration in accordance with the rules of electricity billing law in Poland.

#### **3. Simulations of the Effectiveness of Price Arbitration Implemented in Microgrid Systems with the Use of BESS**

The use of price arbitrage in enterprises entails a complication in programming the BESS operation control system, resulting from the need to include the complex and unpredictable profile of electrical loads. The demand for energy and power in the microgrid varies over time and results from the current demand for electricity by devices, implementing production processes, building infrastructure, servicing of communication routes, transport, social needs, etc. Additionally, it should be noted that each enterprise has a different nature of organizational and technological processes, i.e., each enterprise has its own individual specificities in running a business, which are connected with the demand for energy needs at certain times.

#### *3.1. Indicators of the Effective Selection of Storage Capacity for Price Arbitrage*

To assess the use of various BESS values for the implementation of price arbitration strategies in the enterprises, simulations were carried out. On the basis of the simulations, the implementation of the strategies was assessed. We assumed from the input data that the capacity of the EBA reservoir increases step-by-step by a constant value. Thus, defined parameters were used, which were determined for each tested capacity value based on the annual measurement results. The list of defined parameters is presented in Table 4.

**Table 4.** Measures for the evaluation of the functioning of subsequent BESS figures for the implementation of the price arbitrage strategy.


Along with the increase in BESS capacity, increasingly smaller increases in annual income were observed, which were calculated as the difference between annual revenues and annual OPEX costs. In the case of microgrids, revenue is limited not only by the size of the energy storage, but also by the amount of energy consumed by the load in the price zone in which the storage is discharged and by the maximum value of the contracted capacity. As a result, the revenues do not grow linearly as they do in the case of the classic standalone

BESS operation for price arbitrage, but grow according to a curve with a decreasing slope and a linearly increasing energy storage capacity. The impact of the indicated limitations on the operation of BESS in the enterprise microgrid is illustrated by the graph in Figure 3 that shows the temporal variability of the amount of energy stored in BESS for two arbitrarily selected storage capacities, 1000 kWh and 4500 kWh, in one of the weeks characterized by the highest energy consumption by the A2019 enterprise:

∑ , =35040 =1 ∙∙(1−)

> (, − .−1) , (, − ,−1) ,

BESS capacity utilization =

" "

Marginal income elasticity =

**Figure 3.** Cumulative state of charge BESS for 1000 kWh capacity and 4500 kWh capacity during the week with the highest load for the A2019 enterprise.

enterprise's With an energy capacity of the storage system of 1000 kWh, the charging and discharging cycles were evenly distributed on the working days from 11 March 2019 to 17 March 2019. There were no limitations to this capacity that made it impossible to charge the magazine to a given value. The exception was on 13 March 2019, when restrictions related to the enterprise's microgrid resulted in incomplete recharging of the storage system to the value of 987 kWh, representing 99% of the total capacity. At the same time, virtually all energy stored in BESS (98% to 100% EBA) was used for all cycles. For comparison, with an energy storage capacity of 4500 kWh, incomplete charging occurred every working day and the storage capacity was used only from 63% to 77% on these days *EBA*.

The simulation data for one year, which was obtained using price arbitrage, together with the BESS capacity that increased in successive steps with a constant contracted capacity equal to 1860 kW, are presented in Table 5.



**Table 5.** *Cont.*

The research showed that among the proposed indicators for the use of the price arbitration strategy, the parameter of *marginal income elasticity* was characterized by the greatest volatility. This is illustrated in Figure 4.

In Figure 4a, the *characteristic point* can be determined, beyond which the character of the curve changes from moderately sloping to a curve with a significant decrease. This point, defined by the authors as the *characteristic point* of the curve, determines the value of the BESS capacity, above which its further increase is ineffective. Figure 4a shows that the *BESS capacity utilization* waveform (green) is a less indicative parameter in determining the optimal BESS capacity, as there is no clear *characteristic point* on the curve. Even more difficult is identifying the "characteristic point", which shows the non-linearly decreasing efficiency with increasing energy storage capacity, that is caused by the *annual income parameter*, as presented in Figure 4b.

Figure 4a,b shows that after exceeding the *characteristic point*, the parameter value of the *marginal income elasticity* begins to decrease significantly, along with the constant increase in the capacity of BESS. For the same capacity increases, the parameters of *BESS capacity utilization* (Figure 4a) and *annual income* (Figure 4b) are more linear. On the basis of the simulations, a conclusion can be drawn that the effective operation is increasing the

capacity of BESS to the value for which the internal limitations of the microgrids do not have a significant impact on the effect of the implementation of the price arbitrage strategy.

**Figure 4.** Results of the simulations of the price arbitrage strategy for the A2019 enterprise: (**a**) comparison of the *marginal income elasticity* parameter with the parameters of *annual energy charged, annual energy discharged* in kWh, and *BESS capacity utilization* in %; (**b**) comparison of the *marginal income elasticity* parameter and the *annual income* in PLN.

" " In this study, we arbitrarily assumed that the effective value of the BESS capacity is determined when the parameter of *marginal income elasticity* is equal to 95%. From the results presented in Table 5 for enterprise A2019, the *marginal income elasticity* of 95% was achieved with the BESS capacity equal to 2700 kWh. In the *marginal income elasticity* diagram shown in Figure 4, this point is located before the *characteristic point*. The BESS capacity of 2700 kWh can be considered as the effective size of the energy storage capacity of enterprise A2019. This capacity value corresponds to the maximum charging and discharging powers in the 15 min intervals during the year, considering the work of BESS for price arbitration and the implemented technical limitations. The maximum values of these powers, as presented in Table 5, were calculated as a result of the simulation for the BESS = 2700 kWh capacity. There was also a minimum power size of the inverters for the assumed BESS capacity, as follows:


The differences between the maximum charging and discharging powers are due to the fact that the charging period is 9 h and the discharging period is 5 h to 3 h, depending on the period of the year. Therefore, it follows that the discharging current is significantly higher than the charging current.

#### *3.2. Validation of Indicators Based on Data Obtained from Enterprises (B 2019, C 2019 and A2018)*

To verify the method of determining the optimal BESS dimensions for the implementation of the price arbitration strategy using the *marginal income elasticity* parameter, the B2019 and C2019 time periods were tested in a manner analogous to the method described for the A2019 enterprise. The results are shown in Figure 5.

• • In enterprises A and B, only the *marginal income elasticity* parameter indicates the existence of a *characteristic point* that influences the effectiveness of the price arbitrage strategy, as shown in Figure 5. However, it can be observed in the parameter of *marginal income elasticity* in enterprise B that the *characteristic point* was more difficult to identify than it was in enterprises A and C. This difference was due to the different organization of the working hours in these enterprises. In enterprise B, on the last shift of the working day, the volume of electricity demand in the afternoon peak hours of the S2 zone decreased significantly. In these hours, which were already at low values of BESS capacity, there were cases of incomplete discharges of BESS capacity in the S2 zone. Thus, the decrease in the demand on the microgrid for electricity in the discharge zone, together with the limitation assumed in point 10 in Section 2.1, resulted in the ineffective use of price arbitrage.

**Figure 5.** Results of simulations of the price arbitrage strategy: (**a**) B2019; (**b**) C2019.

The arbitrarily adopted value of 95% for the *marginal income elasticity* parameter clearly indicated the existence of an effective value of the energy storage capacity. The effective value of energy storage capacity was visible in the B2019 and C2019 graphs near or before the *characteristic point* that resulted from the limitations of the microgrid. Table 6 shows the BESS parameters in points for which the *marginal income elasticity* parameter is equal to 95%.


It should be noted that the *characteristic point* in the case of enterprises B and C occurred for various parameters that characterized the use of storage capacity; for the parameter of utilization of the *BESS capacity,* it was approximately 87% for the series B2019 and 95% for the series C2019. In enterprise A, the value of this parameter was 93%. The simulations showed that the parameter of using the storage capacity (i.e., the utilization of the *BESS capacity*), did not change significantly, as evidenced by its flattened characteristics. For these reasons, it can be concluded that this parameter is not very useful in determining the value the of effective use of BESS for a price arbitrage strategy.

The indicator of optimal BESS selection for the same enterprise was also analyzed in relation to the consumption profile from the previous year. The calculation results for the data series A2018 and B2018 are shown in Figure 6.

The results from the simulations again indicated that the *marginal income elasticity* parameter remained the most sensitive. The remaining parameters, which quantified the size of the storage system for the use of the price arbitrage strategy, did not clearly indicate the existence of the *characteristic point* that could be used to determine the optimal size of BESS.

**Figure 6.** The results of the price arbitrage simulation for enterprises (**a**) A2018; (**b**) B2018.

By assuming a *marginal income elasticity* value of 95%, we determined the effective storage size for the arbitrage price strategy. For data series A2018, this was a BESS capacity of 2600 kWh, which is close to the 'characteristic point". For the B2018 data series, this was a BESS capacity of 600 kWh located ahead of the *characteristic point*. A comparison of the results for 2018 and 2019 for enterprises A and B is presented in Table 7.


**Table 7.** Numerical results of the price arbitrage simulation for the A2018 enterprise.

The data in Table 7 show that the individual parameters in 2018 and 2019 were similar. This means that each enterprise maintained its basic nature of demand in 15 min intervals in subsequent years. However, it can be seen that in the case of the same enterprise, the *marginal income elasticity* equal to 95% indicated a higher value of the optimal BESS capacity in 2019. Studies of the load profiles of the same enterprise for 2018 and 2019 showed an increase in optimal storage capacity by only one step of the set capacity in the calculations. This slight difference may be due to the different number of non-working days in the analyzed years, together with the associated Saturdays and Sundays.

For the data series A2018 and C2018, we also examined how the proposed indicator to evaluate the efficiency of selecting the size of the energy storage, defined as the *marginal income elasticity*, behaved for various contractual powers. Figure 7 shows the results of the simulation of the BESS operation, in accordance with the price arbitration strategy for the data series A2018 and B2018, during which the contractual power was increased stepwise by a constant value.

In previous studies, it was assumed that the contractual power was equal to the maximum power of all 15 min power consumptions in an annual period. This was a hypothetical value and, in fact, it was impossible to determine if there were no tools for actively lowering the consumed power. Adopting a certain level of contractual power determines the operation of the energy storage in the event of the implementation of the price arbitration strategy. The higher the contractual power, the greater the possibility of increasing the charging power.

**Figure 7.** The results of the price arbitrage simulation with different contractual powers Pu [kW] for enterprises (**a**) A2018; (**b**) B2018.

For the A2018 data series, with the increase of the contractual power, the point of *marginal income elasticity* equal to 95% as a function of the storage capacity shifts to the right. This means that the contractual power had a significant impact on the effective use of the storage capacity. In the examined enterprise A, the increase in the contractual power resulted in a linear increase in the effective storage capacity, as shown in Figure 8a. Unlike data series A2018, the *marginal income elasticity* curves obtained for enterprise B (data series B2018) showed a weak dependence on the change in contractual power, as shown in Figure 8b). For increasing values of the contractual power, the obtained values of the ratio were the same or increased slightly.

**Figure 8.** Effective BESS capacity with different contractual powers Pu [kW] for enterprises (**a**) A2018; (**b**) B2018.

These differences can be explained by the fact that the limitation of the BESS charging current depends not only on the contractual power, but also on the energy consumption profile during the charging zone hours. Both enterprises, A2018 and B2018, had different work organizations and differed in the level of energy consumption in the adopted Z<sup>C</sup> charging zones. Enterprise A maintained a constant high level of energy consumption during the Z<sup>C</sup> zone hours, and the energy consumption in the Z<sup>C</sup> zone of enterprise B was significantly lower due to the two-shift work organization. It can be assumed that in the case of enterprise B, it was important to limit the amount of energy discharged by BESS to a value not greater than the energy resulting from the demand of internal consumers in zone S2, and that the limitation resulting from the contracted amount of contractual power was insignificant.

#### **4. Discussion and Conclusions**

Studies of real microgrid systems have shown that the nature and variability of electricity consumption by enterprises limit the effective use of price arbitrage strategies. These limitations, which are caused by the rules for billing for electricity and the instantaneous amount of energy load, determine the possibility of charging and discharging the storage system. As a result, the effectiveness of implementing the price arbitrage strategy decreases non-linearly with an increase in the BESS capacity, despite the programming of constant values of charging and discharging energies.

The limitations of the real microgrid systems mean that, for certain BESS capacity values, further increases in the energy storage capacity for the implementation of price arbitrage cease to be effective. To determine this value, the *marginal income elasticity* indicator was used. The curve of this parameter as a function of increasing BESS capacity has a characteristic point, after which the curve begins to significantly decline. Our research showed that the *characteristic point* appears near the value of the *marginal income elasticity* parameter, which is equal to 95%. Our research results showed that the application of the *characteristic point* of the *marginal income elasticity* curve to determine the size of the energy storage capacity establishes the limit of the BESS capacity, which is effective in implementing price arbitrage.

The determination of the effective size of energy storage, based on the *marginal income elasticity* parameter equal to 95%, will indicate the sizes of the effective storage capacity for the same enterprise in the following years. However, in these cases, one should consider the variability in energy and power demand caused by different numbers of days off work, as well as Saturdays and Sundays.

The effective use of energy storage capacity can be influenced by the value of the contractual power reported for settlements to DSOs, especially for enterprises with a continuous nature of production where the intensity of electricity demand does not decrease during BESS charging hours. In enterprises where production is not continuous and the organization of work occurs in one or two shifts, the amount of electricity demand of microgrid loads in the adopted period of energy discharge by BESS is of great importance for the effectiveness of price arbitrage. In cases where this demand is much lower than the maximum load value, the limitation resulting from the amount of contracted power is insignificant and the importance of limiting the amount of discharging energy of the storage system to the amount of energy that is consumed by the microgrid loads increases.

This study undertook simulations aimed at determining the power and capacity of BESS for the functionality of price arbitration. Our research had certain limitations, as outlined below.


The following future work is intended:

1. Research on the possibility of obtaining synergy via the simultaneous use of price arbitrage strategy and strategy peak shaving. These functionalities are representative of two separate markets, i.e., price arbitrage in the electricity market, which is the domain of trading companies, and the peak shaving strategy, which covers activities in the capacity market, a consideration that is important from the point of view of distribution and transmission system operators.


**Author Contributions:** Conceptualization: R.K. and A.S.B.; collected data: R.K.; methodology: R.K., A.S.B., A.P. and M.P.; simulations: R.K.; visualization: R.K.; validation: R.K.; analysis and conclusions: R.K. and A.P.; writing and editing: R.K. and A.S.B.; supervision: A.S.B. and M.P.; funding acquisition: A.S.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research did not receive any specific grant from any funding agencies in the public, commercial, or not-for-profit sectors.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Statements of three enterprises that agreed to shear their 15 min of active power measurements in 2018 and 2019 are attached.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Nomenclature**



#### **References**

