Performance evaluation of 50MWp solar plant under different climatic conditions

Mohamed Saleck Heyine, Ahmed Mohamed Yahya, Daha Hassan Daher, Léon Gaillard, Christophe Menezo, Abdel Kader Mahmoud Département de Physique, Faculté des Sciences et Techniques, Université de Nouakchott Al-Asriya, Nouakchott, Mauritanie Laboratoire des Energies Nouvelles et Renouvelables, Centre d’Etudes et de Recherche de Djibouti, Djibouti, Djibouti Heliocity SAS, 31 rue Gustave Eiffel, Grenoble, France Université Savoie Mont Blanc, Le Bourget-du-Lac, France


INTRODUCTION
Energy is one of the main pillars for any country aiming to increase its socio-economic and technicalindustrial development. This is why the consumption of electrical energy in the world has been increasing for the last decade. Providing energy without intermittency while ensuring resources for future generations is the ultimate challenge for humanity. In this context, the trend to increase the share of renewable energy power plants in electricity production, which contributes to the respect of the environment by reducing the emission of greenhouse gases from fossil fuels [1], [2]. Several types of power plants are based on renewable energy, such as wind farms, solar photovoltaic plants, and others [2]. Solar photovoltaic power plants are systems for producing electrical energy from the sun's radiation, they occupy a prominent place because of their many special features [3]. Solar PV energy has increased significantly over the past decade, from 23 GW in 2009 to 627 GW in 2019 [4]. Global solar PV capacity additions are expected to reach nearly 107 GW in 2020 [5].
In the era of this trend, many countries are currently re-examining their national energy policies intending to switch to a mix of renewable and conventional energy sources. It is in this perspective; Mauritania is oriented to the large-scale integration of renewable energies to achieve a more balanced national energy mix [6]. Thanks to this orientation various projects have seen the light, notably a photovoltaic power plant of 15 MWp connected to the MT network of Nouakchott is put into service since 2013, a wind power plant of 30 MW in Nouakchott is in service since 2014, a photovoltaic power plant of 50 MWp is inaugurated November 2017 in Nouakchott, This plant will be the subject of our study in this paper, a wind power plant of 100 MW is under construction in Boulanouar since 2018, hybrid solar power plants in the interior of the country Kiffa, Atar, Nema, Adel Bagrou, and Aleg provide electrification in these cities since 2017 [7].
Mauritania is one of the aridest Sahelian countries and the most exposed to the effects of desertification, it is located in a geographical axial position between the Sahara and the Sahel. It is characterized by an arid and subtropical climate, constantly hot, dry, and dusty, with a strong influence of the Atlantic Ocean to the west [8]. The Saharan desert space in Mauritania represents ¾ of its surface, the remaining ¼ represents the Sahelian zone [9]. The performance and proper functioning of these energy infrastructures that operate in an arid climate is a subject of research that preoccupies the research units in the fields of renewable energy in Mauritania. Some studies have been conducted on the performance evaluation of solar PV plants installed worldwide. We present in the following the most recent studies on the performance of PV systems in several countries with different metrological characteristics.
Several studies have been caried out to analysis performance of solar PV technologies, [10] presented a study that highlights the state of solar energy in Hungary and analysis the performance of a 9.6 kWp solar installation consisting of two different photovoltaic technologies: polycrystalline silicon photovoltaic technology and amorphous silicon mounted on a roof and connected to the grid. The study of this facility is based on one year of data retrieved from January 2016 to December 2017 with a 10 minutes interval, [11] presented a study of the performance of monocrystalline and polycrystalline solar photovoltaic modules, under the climatic conditions of Manizales-Colombia. Both systems are connected to the electrical grid of the city of Manizales, the first one is a polycrystalline system of 2115 Wp, and the second one is a monocrystalline system of 2040 Wp. Data for this study were collected every 10 minutes for each variable, and the analysis was conducted over four seasons. Zaghba et al. [12] the others examined the impact of seasonal variation and different climatic conditions (sunny, partly cloudy, and cloudy) on the performance of a 2.25 kWp gridconnected PV installation installed on the roof of a parking lot, the installation is based on thin-film silicon technology in the desert environment of the Ghardaïa region, Algeria. The data was recorded with a measurement step of 5 min. The performance evaluation of a grid-connected photovoltaic system of 1.75 kWp based on monocrystalline silicon modules installed in the Saharan zone of southern Algeria (Adrar), in three typical days of the year 2014 under variable climatic conditions (clear, cloudy, and sandstorm) is presented in [12]. Dabou et al. [13] the others examined the impact of seasonal variation and different climatic conditions (sunny, partly cloudy, and cloudy) on the performance of a 2.25 kWp grid-connected PV installation installed on the roof of a parking lot. The measurement data in this study were recorded at an interval of 1 min. Karami et al. [14] presented an experimental study of different silicon-based PV modules (monocrystalline (c-si), Polycrystalline (p-si), and Amorphous (a-si)) during 3 days under varying climatic conditions (clear, cloudy, and rainy).
Real-time measurements were taken every five minutes for different climate parameters. The experimental study was performed to assess the real performance of the selected technologies under real conditions in Casablanca, Morocco. In Ghana the others [15] examined the performance of five solar photovoltaic systems with a total capacity of 20KWp with five different solar cell technologies such as polycrystalline, monocrystalline, copper indium sulfide thin-film, amorphous silicon, and heterojunction incorporating a thin film under a humid tropical climate in Ghana. Daher et al. [16] the others provided experimental results on the performance of a 300 KWp grid-connected photovoltaic power plant using polycrystalline silicon technology operating in dusty maritime and desert climatic conditions in Djibouti using data for the first 4 years of operation of this plant from February 1, 2012, to December 31, 2015, with a minuteby-minute and hour-by-hour step measurement of the electrical performance and environmental conditions respectively. The characteristics, behavior, and sensitivity of a 6.5 KWp grid-connected PV system under varying environmental parameters under tropical climate conditions in Japan is presented in [17], their photovoltaic system using one-year data from August 2009 to July 2010 for two PV technologies mc-SI and CIS. For a large solar power plant, the others in [18] examined a 10 MWp solar power plant using polycrystalline technology in the Indian climate. This study analyzes and compares the performance results of the plant for one year, from April 2014 to March 2015, with the simulation results of the PVsyst and GIS software. Also in Mauritania, the study presented in [19] provided the evaluation and analysis of the performance of a 15 MWp solar photovoltaic power plant installed in Nouakchott using amorphous silicon and micro amorphous silicon technology, their examination was done by exploiting Scada data for one year from September 2014 to August 2015 with a 5 min step.
The various studies presented below were based on the standard established by the international energy agency (IEA), and described in the standard IEC 61724 [20] for the analysis of the performance of different solar photovoltaic installations. The parameters evaluated in the different installations presented above are mainly the energy production, the reference yield, the array yield, the final yield, the capture and system losses, and the PR. However, until now, few researchers have focused on the daily behavior of large grid-connected PV plants to analyze their performance under arid climate with a very small measurement step size and seasonal analysis. The main objective of this work is to analysis the daily performance of the largest grid-connected PV power plant and to appraise the impact of the climatic context in dry and wet seasons, but also in three typical days (clear, cloudy, sandstorm) with a measurement step of ten seconds. The installation was assessed based on various performance parameters, including reference yield, array yield, final yield, system losses, and performance ratio to provide baseline information for the energy evaluation of polycrystalline PV panels. This paper is structured as follows: section 2 describes the PV power plant considered in this work. Section 3 presents the performance analysis methodology. Section 4 presents the results and section 5 summarizes the main conclusions.

DESCRIPTION OF THE SOLAR PHOTOVOLTAIC POWER PLANT
The Toujounine solar power plant has a capacity of 50 MWp located east of the city of Nouakchott, Mauritania (18°4'24.532" North, 15°52'47.101" West). The photovoltaic plant was commissioned in November 2017. The aerial view of the PV plant is shown in Figure 1 and the schematic diagram of the 50 MWp solar power plant connected to the grid is shown in Figure 2. problem on one of the loops. In the following, we present the technical data and specifications of the components of the solar photovoltaic power plant of Toujounine.

PV modules
The Toujounine solar power plant consists of 156,240 Jinko photovoltaic modules. The modules used are based on polycrystalline technology. The PV modules are oriented to the south and tilted at an angle of 12°. The technical characteristics of the PV modules of this plant are shown in Table 1.

Combination boxes
Combiner boxes are used to join together strings of PV modules into a common output and offer protection to the incoming DC strings in case of a fault, they are used to simplify the wiring of the PV panels in the power plant and offer protection against over-voltage and over-current. In the power plant, there are 532 combination boxes. In each box are connected a maximum of fifteen strings of twenty modules in series as shown in Figure 3. Table 2 shows the technical characteristics of the combination box used in the Toujounine power plant.

Inverters and transformers
The solar power plant of Toujounine is composed of 76 DC/AC inverters of type Conext Core XC series 680. The inverters are used to convert the DC power produced by the solar field into an AC power that will be injected into the grid through 38 step-up transformers 400V/33kV with unit power of 1,360 KVA each, manufactured by Schneider. The technical characteristics of the inverters used in this plant are presented in Table 3.

Data acquisition system and weather station
The solar power plant of Toujounine has four weather stations, two on loop A (POT A4 and POT A7), one on loop B (POT B4), and one on loop C (POT C3). Each weather station has an experimental measurement device as shown in Figure 4. Each device is composed of; i) a pyranometer (PYH) on the horizontal plane and a pyranometer (PYP) on the module plane to measure the different values of solar radiation, ii) a thermometer probe (4-wire PT-100) to measure the ambient temperature and the module temperature; and iii) a nanometer (type AIRFLOW TA2), installed at the top of a 2-meter mast, where standard measurements are made to measure wind speed.

Monitoring system
The automatic data acquisition system is required to monitor the performance of a solar installation. The monitoring system is used to control the inverters and communicate the system status and weather data to the user locally or remotely. The solar power plant of Toujounine is controlled by a SCADA system with a storage capacity of 18 TB that works with the MySQL database and has Siemens Wincc type servers that archive the data of 5,000 variables over 1 year. The SCADA system of the plant provides an interface to all the components of the installation and allows it to process in real-time a large number of parameters and to control the installation remotely. Each POT contains an XM408-8C switch which is connected to the previous POT, the next POT, a thermal camera, an IP phone, and an ET200SP PLC. The PLC is connected to the different measurement stations of the metrological parameters and the measurement counters of the electrical parameters.

DATA ACQUISITION AND ANALYSIS METHOD 3.1. Measurement of metrological parameters
The performance analysis of this plant is made based on the data recovered from the SCADA for the period from August 1, 2019, to July 31, 2020. The different measurements are collected with a measurement step of 10 seconds. The metrological parameters collected to be used to analyze the performance are sunshine, ambient air temperature, wind speed, and module temperature. The solar power plant of Toujounine has four measuring stations, each POT contains one measuring station in addition to the metrological station which serves as a test bench. The radiation data are measured in the same plane as the photovoltaic field using the pyranometer (PHY and PYP). The temperature of the ambient air is measured with temperature sensors (PT-100) located in solar shields. The temperature of the PV modules is measured with temperature sensors (PT-100) located on the back surface of one module of each measuring station. The wind speed is measured at a height of 2 m using (AIRFLOW TA2). The location of these sensors in the different weather stations of the installation is representative of the metrological conditions of the plant.

Performance analysis methodology
The International Energy Agency has described in the IEC 611724 standard (IEC-CEI, 1999) [20] the parameters needed to analyze the performance of grid-connected solar PV plants. The performance parameters described by the standard are reference yield, reference PV field, reference system, performance ratio, system losses, and other losses. The performance analysis of photovoltaic system has been the subject of several studies [10], [21] in different countries. The mathematical models of the different performance analysis parameters are presented in Table 4. Inverter efficiency (µinv) µinv = ×100% (%) (11) [10]- [23] Where ( ) the recording interval 10s, (G r ) total in-plane solar irradiation kWh/m 2 , (G STC ) the module's reference inplan irradiance at the standard test condition 1kW/m 2 , (Aa) Area of PV array m 2 .

Meteorological parameters
Mauritania is characterized by a rainy season from July to October and a dry season characterized by a cool period from November to February and a hot period from March to June [8]. Due to the presence of the ocean and the associated wind patterns, the city of Nouakchott enjoys a relatively mild climate compared to the rest of Mauritania. Metrological parameters such as irradiation, module temperature, ambient air temperature and wind speed were collected with a measurement step of 10 seconds during the period from August 1, 2019 to July 31, 2020. Table 5 shows the monthly average of meteorological parameters. Solar irradiance is one of the main parameters for analyzing the performance of PV modules. The average daily irradiance varies from a minimum of 417.183 W/m2 in June 2019 to a maximum of 616.923 W/m2 in March 2020. The average daily ambient temperature varies from a minimum of 18.751 °C in January to a maximum of 29.85 °C in September. The average daily temperature of the module varies from a minimum of 20.368 °C in January to a maximum of 35.068 °C in September. The flow of wind over the PV modules contributes to the decrease in the operating temperature of the modules. Because of the cooling impact on PV modules, higher wind speeds are helpful to their functioning [24]. The monthly variation of the daily average speed presented in Table 5

Seasonal analysis of performance parameters
In order to evaluate the behavior of the 50 MWp power plant in the different metrological conditions of Nouakchott, we will in the following analyze the performance of this power plant in two seasons characterizing the climate of Nouakchott the dry season from March to August and the wet season from September to February. The dry season is characterized by a clear sky with high heat and harmattan, we have selected the month of May 2020 to analyze the plant during this season. The wet season is characterized by cloudy skies and strong winds, it is subdivided into a period of rainy and cool, the month of November 2019 is our choice to analyze the plant in the wet season. The performance analysis of the power plant is done by evaluating the instantaneous parameters that are recorded by the data acquisition system incorporated in the SCADA of the power plant. Figure 5, Figure 6 (a) and Figure 6 (b) shows respectively the irradiance, ambient temperature, and average daily wind speed for May (dry season) and November (wet season) at Toujounine site of the power plant from sunrise to sunset. Although the data for each month is different, the trend is the same. In other words, the average solar irradiance during November (the wet season) was relatively low compared to May with a lot of fluctuation. However, the wind speed in May was always greater than only in November. While the ambient temperature of November was higher in the majority of days than only in May, it can be seen that the module temperature was always higher than the ambient temperature. In Figure 5   The daily average variation of output power is presented in Figure 7 (a) the power delivered by the PV array and Figure 7  We can observe that the power of the PV field and the power injected into the grid follows the instantaneous changes of the irradiation values in both seasons. It can be seen that for the selected month of the dry and wet season the power reaches its maximum at the same time as the measured solar irradiation, this indicates the accuracy in measuring the solar irradiation. The average daily yield of the reference, final, and system in different seasons (dry and wet) are presented in Figure 8. The average daily yield of minimum reference was observed in November at about 5.87% and the maximum yield was recorded in May around 7%. The average daily yield of the final minimum in November 5.80% and the maximum yield is 6.81% was recorded in May. While the average daily yield of the minimum system is 4.50% observed in May against a maximum system yield of 4.63% in November. Table 6 shows the daily average values for reference yield, array yield, final yield, system losses, capture losses, PV efficiency, system efficiency, and inverter efficiency during November 2019 and May 2020, which correspond to the wet season and dry season, respectively.  The ratio of the monthly average daily performance of the plant, the monthly average daily energy produced by the solar field, and the energy injected into the SOMELC power grid are presented in Figure 9. The results show that the wet season has a better performance of 78.9% compared to the dry season of 64.1% due to a lower ambient temperature, which resulted in a lower panel temperature. It can also be seen that during the dry season the amount of power injected into the grid is greater than in the wet season which is justified by the high demand for electricity often recorded in the dry season in Nouakchott because of the high-temperature rise. Figure 9. Daily DC energy, AC energy, and performance ratio in different seasons (dry and wet)

Daily analysis of performance parameters
This part consists of analyzing the performance of the 50 MWp solar power plant of Toujounine in three typical days of the year from August 2019 to July 2020 selected from the data recovered from the SCADA system of the plant. The days selected for this study are; i) clear day 03/01/2020, ii) sandstorm day 02/28/2020 and iii) cloudy day 07/06/2020. Solar irradiance, ambient temperature, module temperature, and wind speed on three typical days (clear, cloudy, and sandstorm) are presented in Figure 10, Figure 11 and Figure 12. Figure 10 shows the daily variation of solar irradiance measured in three days. It can be seen that the clear day is characterized by a higher intensity than the other days with a peak of solar irradiance at about 1 pm. Due to the metrological disturbances of the cloud-covered sky and the dust storm, the irradiance in the other two days is characterized by an irregular profile. Figure 11 shows the daily variation of the ambient temperature and the temperature of the modules measured in three days. There is a significant increase in ambient and module temperature for a clear day compared to the other two days. The module temperature is always higher than the ambient temperature. These results are consistent with the results reported in [14], [26]− [28]. As shown in Figure 12 the measured wind speed is more important on a cloudy day than on other days. The difference between the ambient temperature and only the module temperature is more important on a clear day compared to the other two days where the wind speed is more important, this explains the results presented by [12], [13], [29] that the modules cool down quite quickly thanks to the wind speed. We can observe that the power of the PV array and the power injected into the grid follows the instantaneous changes of the irradiation values in the three days. It can be seen that for the three selected days the power reaches its maximum at the same time as the measured solar irradiation, this indicates the accuracy in measuring the solar irradiation. The fluctuation of solar radiation on a cloudy day and a sandstorm day caused the fluctuation of the power of the system injected into the grid. The output of the Toujounine solar power plant is better on a clear day than on a sandy or cloudy day. This means that along with module temperature, solar irradiation is a key factor in the production and performance of the PV system. The daily average values of ambient temperature, module temperature, wind speed, solar irradiance, DC output power of the PV array, and the AC output power of the power plant are presented on three typical days (clear, cloudy, and sandstorm) in Table 7. The analysis of the grid-connected solar power plant is based on the parameters described by the standard [20] presented in Table 4 of Section 3.2 will be discussed in this section. The daily variation of PV, inverter, and system performance on clear, cloudy, and sandstorm days are presented in Figures 13, 14, and 15, respectively. It is found that the efficiency of the PV module, inverter, and system, is more stable with higher values for a clear day compared to other days, except that the minimum value of efficiency is recorded at the time when the peak irradiation and power are maximum, which can be explained by the fact that the temperature is high. These results affirm the studies [13]- [19], [30], [31] on the effect of temperature on the yield, PV module, inverter, and system.  Figure 13. Daily variation of PV efficiency Figure 14. Daily variation of inverter efficiency In Figures 15, 16, and 17, it can be seen that during the cloudy day, the yield of the PV module, inverter, and system have an irregular variation and with the presence of several peaks, this variation is caused by the rapid succession of irradiation due to the presence of cloud cover. The results also show that on a cloudy day the performance of the PV module, inverter, and system is maximum due to the low temperature and high wind speed that affects the temperature of the module, which is in agreement with the results reported in [13], [29], [30], [32]. The variation in yields and performance ratio is presented in three typical days in Figure 16 and Figure 17 respectively. The results presented in Figure 16 and Figure 17 indicate that the average yield, PV array yield, final yield, system, and capture losses for a clear day were 7.56 kWh/kWp/day, 6.66 kW h/kWp/day, 5.28 kW h/kWp/day, 1.38 h/day, and 0.9 h/day, respectively, while the average performance ratio for a cloudy day was 73.343% due to the low wind speed and low ambient temperature. The minimum of array yield, reference yield, final yeild, and system losses during sandstorm day were 1.33 kW h/kWp/day, 1.43 kW h/kWp/day and 0.8 kW h/kWp/day and 0.64 h/day, respectively, while the minimum capture losses were 0.27 h/day for a cloudy day and the minimum performance ratio was 0.5 h/day on a clear day due to high ambient temperature. During the sandstorm, the performance ratio was lower than during the cloudy day, due to the low daily solar irradiation, which is 2.65 kW h/kWp/day, and was affected by a dust storm (high wind speed with sand dust particles) [26], [28].
The Figure 18 shows the average daily yield of the panels, reference, and final in three typical days. Figure 19 shows the average daily losses of the panels and the system during the three typical days. On a clear day, the electrical energy output of the plant is higher than on other days. On a clear day, the percentage of panel collection is high, as well as the system losses, but the output is low due to weather parameters such as increased panel temperature. The losses of the PV array are higher on clear days than on other days, which shows the effect of module temperature on the performance of the PV array and the performance of the PV system connected to the grid [13], [28], [33]. The Figure 20 shows the daily average variation of the RP, the energy produced by the solar array, and the AC energy injected into the MV grid of Nouakchott on three typical days. The results show that the clear day has a better performance than the other days. It can also be seen that on a clear day, the amount of power injected into the grid is greater than on other days. Table 8 shows the daily average values for array yield, reference yield, final yield, system losses, capture losses, performance ratio, PV efficiency, system efficiency, and inverter efficiency during the monitoring period.

CONCLUSION
This work presented a performance analysis study of the largest grid-connected solar photovoltaic power plant installed in Mauritania with a capacity of 50 MWp. The study analyzed the behavior of the plant under the different climatic conditions of Nouakchott. The seasonal analysis in two months (May and November) which reflect the dry season and the wet season characterizing the climate of Nouakchott and the daily analysis in three typical days (clear, cloudy, and sandstorms) allowed us to better understand the effect of the metrological parameters on the operation of the plant. The performance parameters of this plant were analyzed under actual plant operating conditions using one year of data from July 2019 to August 2020 based on measured data with a ten-second measurement step.
In general, the amount of energy injected into the grid, reference yield, PV field yield, final yield, capture losses, system losses received during a clear day are much higher than in cloudy day or sandstorm day. However, we found that the PR is higher on the peak time on a cloudy day than on clear day due to the low average ambient temperature compared to the clear day. This allowed us to conclude that the performance of a grid-connected PV system is influenced by a variety of environmental factors, including ambient temperature, albeit to a lesser extent than solar radiation. In a Saharan climate, the PV system remains an optimal choice due to the amount of solar irradiation recorded, which contributes to improving the energy mix and decreasing the dependence of Saharan countries with arid climates on fossil fuels characterized as expensive. To improve the performance of solar power plants in arid environments, it is recommended to take care of the different modes of degradation of the panels due to the climatic conditions and the accumulation of dust after long exposure to the plant.