Analysis of Real and Simulated Energy Produced by a Photovoltaic Installations Located in Poland

Abstract

In recent years, the amount of electricity produced by photovoltaic systems in Poland has increased significantly. This paper presents an evaluation of commercial software (PVGIS 5.3, ENERAD, and PVGIS 24) used for simulating energy produced by four photovoltaic installations. The results of the simulation were compared with the real energy production. The installations differ in terms of panel orientation (S, SE, SE-NW), tilt angle (12°, 25°, 37°) and location (roof- or ground-mounted). The average annual electricity production per 1 kW of module power for each installation was as follows: PV1—1104 kWh·kW⁻¹, PV2—1169 kWh·kW⁻¹, PV3—927 kWh·kW⁻¹, and PV4—831 kWh·kW⁻¹. The highest values were recorded for ground-mounted installations facing south. Simulations carried out using computer programs show differences between simulated and real electricity production values of 35–41% for the ENERAD software, 3–13% for the PVGIS 5.3 software, and 3–32% for the PVGIS 24 software. The most accurate forecasts were obtained for the PV2 system in the PVGIS 24 software (MPE 3%, RMSE 12%), and the most unfavorable for the same installation in the ENERAD software (MPE 41%, RMSE 48%).

1. Introduction

The implementation of energy transition is one of the key elements of every country’s policy. It involves developing technologies for obtaining energy from sustainable sources, including renewable energy, thereby contributing to greater climate neutrality by reducing emissions of carbon dioxide and other greenhouse gases [1,2,3,4,5,6]. The European Green Deal regulations adopted by the European Union in 2019 provide for a minimum 55% reduction in net greenhouse gas emissions by 2030 compared to 1990 and for EU countries to achieve climate neutrality by 2050 [7]. In Poland, in 2023, the share of energy obtained from renewable sources in relation to total primary energy was 24.5%, which represented an increase of 5.1% compared to 2019. Solid biofuels (60.1%) and solar energy (7.5%) played a dominant role in energy production from renewable sources. The cumulative annual growth rate of energy production from renewable sources in the period 2019–2023 showed that solar energy recorded the largest increase in share, by 67.1% [8]. This was mainly due to the rapid increase in electricity production from photovoltaic systems (Figure 1). This was driven by the national support program for rooftop photovoltaics and the emergence of larger PV farms [8,9].

The legal acts in force in Poland [10,11] define energy standards for buildings and impose the obligation to consider the possibility of using renewable energy sources in their construction or modernization. The most commonly chosen RES systems are heat pumps and photovoltaic installations. The new EU directive on the energy performance of buildings [12] stipulates that from 2029, all new buildings will be constructed to a zero-emission standard. This means that all new residential, public, and commercial buildings will not be allowed to use fossil fuels for energy production, and 100% of their total annual primary energy consumption will have to come from renewable sources. In addition, the Directive requires the installation of solar energy systems on all new buildings by 2030. Furthermore, from 2030, all existing public and commercial buildings that undergo major renovation will have to be equipped with a photovoltaic system.

In Poland, solar energy is mainly used in households. According to the data contained in the Energy Regulatory Office Report [13], the number of micro-installations using solar radiation energy (PV) exceeded 1.5 million in 2024. Most of them (almost 99%) were prosumer installations. However, it should be noted that as a result of the introduction of a changed method of settling prosumers for energy fed into and withdrawn from the grid on 1 March 2022, the number of new photovoltaic installations decreased significantly [14]. The growth rate of micro-installations decreased from approximately 41% in 2022 to 15% in 2023 and to 10% in 2024 (Figure 2).

Decisions regarding the construction of photovoltaic installations are always preceded by a forecast of the electricity production of a given installation in order to determine whether the power and electricity production of the installation will be adequate for the needs of the building. The yield, i.e., the total amount of energy from a photovoltaic installation that will be produced by the system in a given period of time, is influenced by many factors. These can be divided into climatic and technological factors [15,16,17,18].

Climatic factors are primarily related to the geographical location of the building on which the photovoltaic installation is to be installed. It will affect the results obtained depending on sunlight, shade, ambient temperature, and climate change. It is obvious that the more sunlight falls on the photovoltaic panels, the more energy they will produce. Therefore, in regions with high sunlight exposure, such as southern Europe, the panels will be more effective [19,20]. However, it should be taken into account that photovoltaic panels operate most efficiently at moderate temperatures. Excessively high ambient temperatures can cause a drop in their efficiency [21,22,23]. The occurrence of various atmospheric phenomena, such as cloud cover, rain and snowfall, and even air pollution, also have a significant impact on electricity production. Cloud cover and precipitation are associated with a reduction in the amount of solar radiation reaching photovoltaic panels, which directly affects their efficiency. Snow lying on the panels often completely covers them, causing them to stop producing energy [24,25]. Atmospheric pollution, such as smog or dust, can block sunlight from reaching the panels, which also reduces their efficiency [26]. An important factor in increasing the efficiency of energy generation from PV panels is the type of photovoltaic modules. Modern photovoltaic modules retain at least 80% of their original power even after 25 years. In addition, the use of optimizers, trackers, and energy storage systems makes the installation even more efficient [27,28].

Extremely important factors affecting the amount of energy produced are: the location of the panels (roof, ground, facade), the orientation of the panels, and their angle of inclination [29]. It should be remembered that the best place to install panels is on surfaces that are constantly exposed to sunlight, no shadows fall on them. In combination with the orientation of the panels and their angle of inclination, a higher energy yield is often achieved. This is particularly evident when the panels are installed facing south, with an angle of inclination ranging from approximately 20 to 50 degrees, depending on the location and season. A PV module positioned according to the angle of sunlight provides the highest yields (

Table 1. Summary of solar energy availability depending on orientation and inclination [30].

Slope	W	SW		S	SE		E
	90°	60°	30°	0°	330°	300°	270°
90°	51%	62%	69%	72%	70%	63%	52%
80°	58%	71%	80%	82%	80%	71%	51%
70°	65%	78%	87%	90%	87%	79%	65%
60°	71%	84%	93%	96%	94%	85%	72%
50°	76%	89%	97%	99%	98%	89%	77%
40°	80%	92%	99%	100%	99%	92%	81%
30°	83%	93%	99%	100%	100%	93%	84%
20°	85%	93%	97%	99%	97%	93%	86%
10°	87%	90%	93%	95%	94%	91%	87%
0°	90%	90%	90%	90%	90%	90%	90%

) [30,31].

Due to the influence of so many factors on the production of energy from photovoltaic cells, it is necessary to plan and calculate the profitability of PV installation construction. There are many pieces of software available on the Internet that can be used to estimate the electricity production of photovoltaic systems. Energy production forecasts often depend on the availability of appropriate tools and data from weather stations, satellites, existing PV systems, or numerical weather forecasts. The yield from a PV installation is a function of relevant weather variables and the characteristics of the system itself. The main factors affecting this yield are the amount of sunlight on the module surface and the ambient temperature. The forecast of electricity production S by photovoltaic installations can be calculated using Formula (1) [30]:

$$S={\displaystyle \frac{N·wK·MM·WW}{IR}}\hspace{1em}\left[\mathrm{kWh}\right]$$

(1)

where N—solar radiation on a horizontal surface—data from the PVGIS service, kWh/m²

• wK—coefficient allowing the conversion of sunlight data for an inclined surface of a photovoltaic generator (photovoltaic modules) from sunlight data read from a map, which are for a horizontal surface
• MM—nominal power of modules (PV generator) determined under STC (standard test conditions), i.e., at a temperature of 25 °C and insolation of 1000 W per 1 m² of module for a period of 1 h. This value can be found in the panel data sheet, kW
• WW—efficiency factor—an indicator that takes into account the level of losses in a photovoltaic installation. In a photovoltaic installation, there are: cable losses—1%, inverter losses—3–7%, module losses due to temperature—4–8%, losses due to low solar radiation—1–3%, losses due to shading and dirt—1–5%, losses due to current mismatch of modules—1%, losses on shunt diodes—0.5%
• IR—standard solar radiation intensity at which photovoltaic modules are tested, kW/m².

There are few studies comparing the real electricity production of photovoltaic panels with the production predicted by various types of software [29,32]. This paper presents the results of real electricity production for four photovoltaic installations differing in panel orientation (S, SE-NW, SE), tilt angle (12°, 25°, 37°) and location (roof-mounted, ground-mounted). The results were compared with the results of simulations carried out using PVGIS 5.3, PVGIS 24, and ENERAD software.

2. Materials and Methods

2.1. Research Object

The research object includes four photovoltaic installations located in the Lubelskie voivodeship. The location and technical parameters of the analyzed installations are presented in Figure 3 and

Table 2. Technical parameters of analyzed PV systems.

	PV1	PV2	PV3		PV4
Type of building	public utility	public utility	public utility		public utility
Distance from meteorological station [km]	40	20	20		20
Mounting position	free-standing	free-standing	roof		roof
Slope [°]	25	25	37		12
Orientation	S	S	SE	NW	SE
Type of modules	Jinko Solar MM445-60HLD-MBV	Longi LR4-72HBD-440M	Longi LR4-72HBD-445M	Sharp NU-JC330	Talesun TP660P 270
Number of modules [pcs]	20	22	16	8	45
Power of module [Wp]	445	440	455	330	270
Nominal power of system [kW]	8.9	9.68	9.92		12.15

.

The PV1 installation is located 40 km south-east of the weather station. The installation was fitted in November 2021. There are no elements in the immediate vicinity that could obscure the PV cells. To carry out the analyses the average electricity production obtained from three years of operation of the installation was used.

The PV2 installation is located 20 km south-east of the weather station. The installation was fitted in April 2022. There are no elements in the vicinity of the building that could obscure the PV cells. To carry out the analyses the average electricity production obtained from two years of operation of the installation was used.

The PV3 installation is located 20 km south-east of the weather station. The installation was fitted in September 2021. There are no other elements in the immediate vicinity that could obscure the PV cells. To carry out the analyses the average electricity production obtained from three years of operation of the installation was used.

The PV4 installation is located 20 km east of the weather station. The installation was fitted in September 2019. There are elements in the immediate vicinity that could obscure the PV cells. To carry out the analyses, the average electricity production obtained from five years of operation of the installation was used.

In order to illustrate the climatic conditions prevailing in the vicinity of the analyzed installations,

Table 3. Average monthly weather data in Lublin Radawiec weather station in year 2005–2023.

Month	Tmin	Tmax	V	DS	MS	IR
January	−3.90	−0.20	12.90	2.60	36.10	21.74
February	−3.50	1.4	12.50	3.80	60.50	40.37
March	−0.70	6.80	12.50	5.50	121.50	81.19
April	3.90	13.80	11.50	7.30	172.60	128.30
May	8.40	16.60	9.90	8.80	223.90	162.19
June	12.30	22.90	9.10	9.50	258.90	180.90
July	14.20	25.00	9.30	8.90	250.30	174.20
August	13.70	24.60	8.50	8.50	231.30	152.42
September	9.70	19.10	8.90	6.90	158.70	104.52
October	5.30	13	9.80	5.10	113.50	63.82
November	2.00	6.50	11.40	3.10	49.00	26.99
December	−1.70	1.70	12.70	2.20	24.90	17.85

presents meteorological data from the Lublin Radawiec weather station. This station constitutes a meteorological database used for simulation calculations. The data on average monthly minimum and maximum values of temperature (T) and wind speed (V) were acquired from the Institute of Meteorology and Water Management (IMGW) [33], data on the number of hours of sunshine per day (DS) and month (MS) come from the meteorological service [34] and data on average global horizontal irradiation (IR) from the Sarah 3 database [35]. All data are average values for the years 2005–2023.

The lowest air temperatures are observed in the winter months: December, January, and February. These months also have the fewest hours of sunshine and the lowest solar radiation. The warmest months are June, July, and August, which also experience the greatest solar radiation. During these months, days are significantly longer than in winter, which results in a greater number of sunny hours per day. In May, despite its lower air temperatures than in September, there are more hours of sunshine and greater solar radiation.

2.2. Software Used for Forecasting Electricity Production

Data on real energy production by photovoltaic systems was compared with forecasts obtained through simulations by using three types of software: PVGIS 5.3 [36], the ENERAD photovoltaic calculator [37], and PVGIS 24 [38] software. These programs allow us to obtain data on the predicted energy production of photovoltaic systems anywhere in the world. They are available free of charge via websites. The data on insolation and climatic conditions come from the SARAH 2 (ENERAD software) or SARAH 3 (PVGIS 5.3 and PVGIS 24) databases. In each piece of software, the following data must be entered in order to obtain results:

• location of the photovoltaic system—by selecting a location on the map or entering an address
• mounting position
• slope of PV modules,
• azimuth (orientation) of PV modules
• estimated system losses.

In PVGIS 5.3 and ENERAD software, total losses for the entire system are taken into account. For PV1, PV2, and PV3 installations, losses of 14% were assumed, and for PV4 installations—17%. These losses included: cable losses—1%, inverter losses—2%, module losses due to temperature—5%, losses due to low solar radiation—3%, losses due to shading and dirt—2–5%, losses due to current mismatch of modules—1%.

Losses in the PVGIS 24 software are divided into three types: cable losses—1% (the distance between the inverter and the photovoltaic panels is less than 30 m), inverter losses—2% (according to the inverter manufacturer’s data), and PV losses—1% (according to the software manufacturer’s recommendations). However, regardless of the loss values entered, the program additionally estimates the total system losses. For the PV1 installation, it estimated losses at 9.79%, for PV2—9.79%, for PV3—13.83%, for PV4—10.45%.

In PVGIS 5.3, one must also enter information about the type of photovoltaic panels: crystalline silicon cells, thin film modules made from CIS or CIGS, thin film modules made from Cadmium Telluride CdTe, and the installed peak power of the system. An advantage of the program is that it enables us to perform simulations for tracking PV installations. The program window is shown in Figure 4.

When entering data into the ENERAD software, it is necessary to specify the current or planned electricity consumption. The advantage of the application is that it can calculate the energy storage capacity. The disadvantage of the program is that it can only be used for installations with a maximum power of 10.7 kW. The program window is shown in Figure 5.

The PVGIS24 software is an extended version of PVGIS 5.3. In addition to the data indicated above, the peak power of the installation must also be specified in order to perform the simulation. Its greatest advantage is the ability to simulate installations divided into sections with different azimuths and module tilt angles. This is particularly useful for buildings with small roof areas, where it is necessary to place PV modules in two different orientations. The software window is shown in Figure 6.

The calculation results in all programs include annual and monthly average values for energy production and insolation.

2.3. Assessment of the Correlation of Real and Simulated Results

The assessment of the correlation between the real electricity production results and the values simulated by individual programs was carried out using selected statistical analysis tools. For this purpose, the mean percentage error (MPE) and the root mean square error (RMSE) were used. They were calculated using Equations (2) and (3):

$$MPE\%={\displaystyle \frac{\frac{\mathsf{\Sigma }\left(S_i-R_i\right)}{n}}{\overline{M}}}·100\%$$

(2)

$$RMSE={\displaystyle \frac{\sqrt{\frac{{\sum \left(S_i-R_i\right)}^2}{n}}}{\overline{M}}}$$

(3)

where

• R_i—real electricity production, kWh
• S_i—simulated electricity production, kWh
• n—sample size,
• $\overline{M}$—average real electricity production, kWh

3. Results and Discussion

3.1. Real Electricity Production by the Studied PV Installations

Table 4. Monthly electricity production by PV installations.

PV Installation	Year	Monthly Electricity Production [kWh]												∑
		I	II	III	IV	V	VI	VII	VIII	IX	X	XI	XII
PV1	2022	228	450	1055	934	1483	1589	1268	1284	759	663	188	92	9992
	2023	128	436	771	974	1458	1293	1355	1230	no data	no data	no data	no data	9663 *
	2024	162	315	681	1045	1724	1373	1351	1269	973	623	312	132	9960
	average	172	401	836	984	1555	1418	1325	1261	866	643	250	112	9822
PV2	2023	165	510	887	1175	1620	1511	1521	1355	1251	704	306	124	11,129
	2024	261	361	813	1199	1884	1475	1512	1405	1189	780	388	235	11,502
	average	213	435	850	1187	1752	1493	1517	1380	1220	742	347	180	11,315
PV3	2022	154	404	950	899	1492	1598	1262	1242	727	573	168	74	9543
	2023	120	355	678	891	1360	1326	1334	1136	978	530	210	83	9000
	2024	185	273	609	930	1472	1236	1331	1137	928	557	248	145	9052
	average	153	344	746	906	1441	1386	1309	1172	878	553	209	101	9199
PV4	2020	179	343	937	1628	1568	1651	1566	1554	914	377	185	99	11,001
	2021	49	117	724	1139	1614	1914	1645	1157	867	520	190	70	10,007
	2022	107	286	924	940	1762	1944	1566	1437	807	488	142	50	10,452
	2023	114	324	698	1030	1662	1510	1565	1258	899	412	129	53	9655
	2024	69	266	604	1046	1804	1476	1489	1232	784	350	165	91	9376
	average	104	267	777	1156	1682	1699	1566	1328	854	430	162	73	10,098

no data—value not read due to lack of communication between the inverter and the application. * real total electricity production in 2023.

shows the monthly electricity production of the PV installations analyzed each year. Figure 7 shows the average values for all years.

Energy produced by photovoltaic systems is highly dependent on climate and seasons. The lowest electricity production is observed during the winter months, when air temperatures are lower, when there are only a few sunny days, and snow often covers the panels. The months from May to August are the period of the highest solar radiation, which results in the highest electricity production.

When analyzing the results of monthly energy produced by photovoltaic installations, we can see that May in 2023 and 2024 was much sunnier than the average solar radiation (IR) values indicate. It is during these months that the highest electricity production occurred, while weather data indicate that June and July had the highest solar radiation.

The PV1 installation reached its maximum annual electricity production in 2022. It amounted to 9992 kWh, with an average monthly production of 833 kWh. The highest yield was recorded in May 2024 and amounted to 1724 kWh (17% of total electricity production in 2024). The lowest energy production occurred in December 2022: 92 kWh (0.9% of total production in 2022). Between April and September, the installation produced 75% of the average annual electricity production.

In the case of the PV2 installation, the highest electricity production occurred in 2024 and amounted to 11,502 kWh, with an average monthly production of 958 kWh. The highest yield occurred in May 2024 and amounted to 1884 kWh (16% of electricity production in 2024). The lowest energy production occurred in December 2023—124 kWh (1.1% of total annual production in 2023). Between April and September, the installation produced 76% of the average annual electricity production.

The PV3 installation reached its maximum electricity production in 2022, amounting to 9543 kWh, with an average monthly production of 795 kWh. The highest yield was recorded in June 2022 and amounted to 1598 kWh (17% of electricity production in 2022). The lowest energy production was recorded in December 2022: 74 kWh (0.8% of total annual production in 2022). Between April and September, the installation produced 78% of the average annual energy production.

The PV4 installation reached its maximum electricity production in 2020. It amounted to 11,001 kWh, with an average monthly production of 917 kWh. The highest yield was recorded in May 2024 and amounted to 1804 kWh (19% of total electricity production in 2024). The lowest energy production was recorded in January 2021: 49 kWh (0.5% of total annual production in 2021). Between April and September, the installation produced 82% of the average annual electricity production.

Due to the different total power of PV modules in individual installations, the electricity production was recalculated per 1 kWp of installed PV modules in the installation (

Table 5. Average monthly final yield of photovoltaic installation.

PV installation Year	PV1 2022–2024 [kWh·kW−1]	PV2 2023–2024 [kWh·kW−1]	PV3 2023–2024 [kWh·kW−1]	PV2 2020–2024 [kWh·kW−1]
January	19.38	22.04	15.42	8.52
February	45.01	44.96	34.66	21.98
March	93.88	87.81	75.19	63.99
April	110.59	122.59	91.38	95.18
May	174.70	181.01	145.29	138.44
June	159.36	154.22	139.76	139.83
July	148.82	156.68	131.95	128.91
August	141.69	142.51	118.12	109.26
September	97.26	126.02	88.50	70.31
October	72.26	76.66	55.78	35.36
November	28.13	35.86	21.06	13.35
December	12.54	18.57	10.16	5.99
∑	1103.64	1168.93	927.27	831.12

). This will allow the efficiency of the photovoltaic system to be assessed.

The PV1 and PV2 installations achieved very similar annual yields per 1 kWp of installed PV modules. Both installations are mounted on the ground, at the same angle (25°) and facing the same direction (S). The PV3 installation achieved a slightly lower yield compared to PV1 and PV2. This is due to the installation of some of the panels facing east, which results in less sunlight exposure. The PV4 installation achieved the lowest electricity production per 1 kWp of installed modules because the modules are mounted at a lower angle of inclination (12°) than generally recommended. In addition, there are elements in their vicinity that cause shading of the installation.

3.2. Comparison of Simulation Results with Real Data

For each PV installation analyzed, electricity production simulations were performed using three programs available free of charge on the Internet. The results are presented in

Table 6. Real and simulated values of electricity production by PV installations.

Month	PV1				PV2				PV3				PV4
	REAL	ENERAD	PVGIS 5.3	PVGIS24	REAL	ENERAD	PVGIS 5.3	PVGIS24	REAL	ENERAD	PVGIS 5.3	PVGIS24	REAL	PVGIS 5.3	PVGIS24
January	172.48	99.18	249.85	285.73	213.32	106.42	268.06	302.35	152.93	96.47	221.67	253.93	103.56	257.07	293.15
February	400.55	188.27	422.26	450.31	435.26	205.26	466.32	525.97	343.85	184.41	382.60	441.30	267.11	454.73	521.76
March	835.56	426.64	759.92	855.25	850.02	458.48	836.58	943.59	745.92	409.03	699.75	812.28	777.45	892.87	1032.76
April	984.28	715.66	1058.81	1174.65	1186.64	780.44	1147.00	1293.71	906.48	699.43	986.28	1151.12	1156.42	1289.76	1502.90
May	1554.87	909.70	1204.40	1337.11	1752.17	994.97	1306.65	1473.79	1441.26	895.31	1158.08	1354.10	1682.00	1539.08	1799.18
June	1418.27	1018.23	1266.67	1402.73	1492.85	1110.01	1377.57	1553.77	1386.41	996.51	1235.79	1446.98	1698.99	1653.46	1936.93
July	1324.54	979.70	1234.23	1364.80	1516.64	1060.48	1339.43	1510.75	1308.98	957.03	1196.71	1399.51	1566.26	1611.04	1884.96
August	1261.07	818.44	1166.11	1270.07	1379.53	898.07	1269.93	1432.36	1171.75	797.25	1098.78	1284.96	1327.54	1437.96	1679.61
September	865.65	530.79	919.35	1006.89	1219.88	577.42	998.52	1126.24	877.94	510.63	828.87	966.44	854.25	1062.81	1235.30
October	643.13	291.09	669.75	663.73	742.05	313.26	724.51	817.18	553.33	273.89	575.36	667.24	429.62	700.11	807.69
November	250.37	117.85	318.78	297.13	347.14	127.82	334.28	388.32	208.87	113.07	253.44	312.30	162.23	307.91	352.24
December	111.65	71.33	217.24	217.98	179.77	77.45	229.87	259.28	100.82	71.52	185.42	212.38	72.72	205.69	234.40
∑	9871.46	6166.88	9487.37	10,326.38	11,315.24	6710.08	10,298.72	11,627.31	9198.53	6004.55	8822.75	10,302.54	10,098.16	11,412.49	13,280.88

and Figure 8, Figure 9, Figure 10 and Figure 11. The graphs on the left show the monthly electricity production of individual installations, while the graphs on the right show the electricity production per 1 kWp of installed PV modules. In the case of the PV4 installation, it was not possible to use the ENERAD software because the power of the simulated installations was limited to 10.7 kW.

When analyzing the data, we can see that real electricity production in May was higher than those simulated. This was caused by high solar radiation levels in 2023 and 2024.

In addition, a statistical analysis was performed for each installation to assess the consistency of the measurement results and the forecasts obtained. The results are presented in

Table 7. Values of forecasting errors for the analyzed and simulated installations.

PV Installation	Software	MPE	RMSE
PV1	ENERAD	−37%	42%
	PVGIS 5.3	−3%	16%
	PVGIS 24	5%	13%
PV2	ENERAD	−41%	48%
	PVGIS 5.3	9%	17%
	PVGIS 24	3%	12%
PV3	ENERAD	−35%	40%
	PVGIS 5.3	−4%	14%
	PVGIS 24	12%	15%
PV4	PVGIS 5.3	13%	18%
	PVGIS 24	32%	33%

.

The assessment of the correlation of simulated and real values showed the best match of results for simulations performed by using PVGIS 5.3 for installations PV1, PV3, and PV4, and by using PVGIS 24 for installation PV2. In the case of the forecast made by the ENERAD software, the program significantly underestimated energy production in comparison to real values for all PV installations.

The lowest MPE and RMSE were obtained for simulations by using PVGIS 24 software for the PV2 installation (slope 25°, orientation S). The highest MPE and RMSE were obtained for the same installation when simulating the installation with ENERAD software.

4. Conclusions

• The simulations conducted indicate the greatest consistency between the simulation results and real data obtained by using the PVGIS 5.3 software. For installations PV1. PV3. and PV4. the lowest MPEs were obtained. indicating the smallest deviation of simulated values from real data.
• The ENERAD software significantly underestimated the electricity production for the analyzed PV installations. The differences between the predicted and real values range between 35% and 41%. This may be due to the fact that the program uses the outdated SARAH2 meteorological database or because of the calculation algorithm used. It is questionable whether the values obtained in this program can be used for preliminary analysis of the profitability of installing the systems.
• Based on the results obtained, it can be concluded that the PVGIS 5.3 software best reflects the real conditions of the analyzed PV installations and can be used as a tool for forecasting the annual electricity production of photovoltaic systems. The simulations obtained can be helpful in designing PV installations and in assessing the profitability of their installation.
• The advantage of PVGIS24 software is its versatility. It allows us to conduct simulations for more complex installations whose PV modules face different directions.