Performance assessment of 20.4 kW eco-house grid-connected PV plant in Oman

ABSTRACT

The aim of this article is to analyse the performance of the Photovoltaic (PV) and to study the effect of soiling on the energy generation under Muscat environmental conditions. The generated energy is consumed in the eco-house, and the access energy is fed back to the electric grid. Two-ways metre was used to record the energy taken from the grid or fed back to the grid. The system was monitored from December 2014 until today. Various performance parameters of the plant were evaluated which include system efficiency, performance ratio (PR), capacity factor (CF) and different types of PV system losses and yields. The annual average values of recorded ambient temperature and solar radiation were 32°C and 482 W/m², respectively.

The total annual energy produced was 23,595 kWh, whereas the annual average daily reference yield, array yield and final yield were 5.59, 3.78 and 3.64 kWh/kW_P/day, respectively. The annual average daily PR and CF were 0.67 and 0.15, respectively, while the annual average system efficiency was 10.3%. Furthermore, the annual average daily capture, system and cell temperature losses were 1.82, 0.14 and 2.95 h/day,respectively, with a small reduction in energy production when modules left uncleaned.

KEYWORDS:

• Grid connected PV
• dust
• Oman climate
• eco-house

1. Nomenclature

	=	PV array area, m2
CF	=	Capacity factor
	=	ambient temperature, °C
	=	module temperature, °C
	=	measured PV module temperature,, °C
NOCT	=	nominal operating cell temperature = 45°C
G	=	solar irradiance, kW/m2
	=	solar irradiation, W/m2
	=	total generated daily energy, kWh
	=	total generated monthly energy, kWh
	=	AC output energy, kWh
	=	AC energy output temperature corrected, kWh
	=	DC energy output, kWh
	=	global in-plan solar insolation,
	=	array capture losses,
	=	system losses,
	=	Cell temperature losses,
	=	number of days in the month
PR	=	performance ratio
	=	AC output power, W
	=	inverter power capacity, kW
	=	inverter output AC power at each instant of time , kW
	=	PV array full load power at STC, kW
STC	=	standard test conditions
	=	array yield,
	=	final yield,
	=	final yield,
	=	temperature coefficient of the PV module maximum power, °C

2. Introduction

Owing to the importance of renewable energy resources to reduce environmental issues, many countries have set target for this resource that should be achieved. Photovoltaic (PV) has fast technological development with declining in cost as well as has special characteristics such as modularity, simplicities of installation and can be integrated within building structure. Global electricity generation capacity of PV has raised from 40 GW in 2010 to 402 GW in 2017 (Renewable Global Status Report 2018).

The performance of the PV system depends on several factors such as the type of technology used, temperature, plan of array irradiation, degradation, soiling and spectral irradiance distribution. Many studies have been conducted on the PV power plant performance in different geographical locations and climate conditions and were reported in Adaramola (2015); Lima, Ferreira, and Morais (2017); Micheli et al. (2014), to name a few.

The energy performance analysis of on-grid three different PV technologies was presented in Başoğlu et al. (2015). The cadmium-telluride module resulted in higher value of performance ratio (PR) and mean array efficiency than other under weather conditions of Northwest of Turkey. The performance analysis of a 2.2-kW_p grid-connected PV system in northeastern of Brazil was discussed in De Lima, Ferreira, and Morais (2017). The PR and capacity factor (CF) were found to be 0.829 and 0.192, respectively. The results obtained from monitoring a 1.72 kW_p rooftop grid-connected PV in Ireland for 1 year were presented in Ayompe et al. (2011). The annual average daily system losses and capture losses were 0.23 and 0.22 h/day, respectively. The results of a 11.2-kW_p roof top grid-connected PV system in Eastern India, monitored for 1 year, were discussed in Sharma and Goel (2017). The final yield of the system was 3.67 h/day, PR was 0.78 and system efficiency was 12.05%. A techno-economic study of a 1-kW_p grid-connected PV system in Indonesia found that the system technically can meet the basic energy need of a household with a payback period of 17.6 years (Tarigan and Kartikasari 2015). Padmavathi and Daniel (2013) reported the performance analysis of a 3-MW grid-connected PV plant for 1 year located in India. Daily and seasonal variations in the plant output were presented using monitored data at 5-minute intervals. The annual average CFPR and system efficiency of 190-kW_p grid-connected PV system in India were found to be 0.0927, 0.74 and 8.3%, respectively (Sharma and Chandel 2013)

Oman has very hot and humid climate. The summer season is extremely hot (May–August) and rainfall is very low throughout the year. Solar energy in Oman is considered as one of the highest in the world, as well as good wind speed is also available mainly in the southern part of Oman (Al-Badi, 2011a). Several papers about Oman’s solar and wind energy resource assessment and utilisation studies were published (Al-Badi, 2011b, Kazem et al. 2014). Oman PV market is growing especially because of the recently introduced law that encourages the homeowners to install PVs on their rooftop and sell their excess solar energy back to the grid at market rates which can reach during the peak times five times as much as they pay for consuming electrical energy. The goal of this paper is to analyse 20.4-kW_P polycrystalline rooftop PV performance under Oman’s environmental conditions. The system was installed in December 2014. In this paper, the most recent data for 1 year (May 2017–April 2018) are presented. The modules were left unclean. Different performance parameters were calculated from the measured results. The parameters include the following: annual average energy generated, the annual average daily yields (array yield, final yield and reference yield), annual average efficiency of the system, PR, CF and losses (array capture, system and cell temperature), as well as the effect of the dust on the performance is discussed.

3. System configuration

3.1 Rooftop PV system

The system consists of 80 PV modules with an overall power of 20.4 kW_p, 20 kW inverter, an eco-house and a grid. Each 20 modules are connected in series which make one string and each two strings are connected in parallel at the junction box to form one array. The two arrays are feeding the DC side of the inverter. The schematic diagram for the whole system is illustrated in Figure 1. The modules are mounted on metal structure tilted 23.5° south above the flat roof of the eco-house as shown in Figure 2. Table 1 shows the specification of the inverter, while Table 2 displays the electrical and mechanical properties of PV modules. The eco-house is located in Sultan Qaboos University (23.6 N, 58.2 E), Oman. The PV system was installed in December 2014 and far away from any obstacles which can cause shading on the PV cells. The measurements were started in December 2014 and continue till today. The modules were left unclean.

Figure 1. Schematic layout of the system.

Figure 2. Pictures of photovoltaic panels on the rooftop of the eco-house.

Table 1. SMA Sunny tripower 20000TL inverter specification.

Inverter	Specification
Input Max. DC power Max. input voltage PV voltage range, MPPT Max. input current	20.44 kW 1000 V 320–800 V 33 A
Output Nominal ac Voltage range Rated power Rated grid voltage frequency Max. output current Maximum Efficiency	160–280 V 20 kW 230 V 50 Hz 29 A 98.4%
Weight	61 kg

Table 2. PV module specifications.

Item	Specifications
Type	Polycrystalline
Number of cells per module	60
Dimensions	1.665 m X 0.999 m X 0.05 m
Number of modules	80
Rated power (Pmax)	255 W
Voltage at Pmax	30.9 V
Current at Pmax	8.27 A
Open circuit voltage	38 V
Short circuit current	8.84 A
Module efficiency	15.33%
Temp coefficient of power	−0.429%/℃
Company/country of origin	S-ENERGY/Korea

3.2. Data logging system

This system consisted of a PC, data logger, pyranometer to measure the global solar radiation, temperature sensor to measure the ambient temperature and sensor for measuring the wind speed. The following are some of the data that are recorded by the data logger and saved on a daily basis: produced energy by the PV, voltage, current, energy consumed by the house, energy taken/fed back from/to the grid, ambient temperature, each room temperature, humidity, wind speed and the illumination in each room. The measurements were taken every 20 s since December 2014 until today. The results discussed in this paper are for one whole year (May 2017–April 2018). The modules were left unclean. The evaluation of uncertainty of the measured results is done by calculating the minimum, maximum and average values of each parameter. For solar irradiance, any value exceeds 1600 W/m² or less than 0 was disregarded, and with respect to the current, any value exceeds the rated current by large percentage also was omitted.

The load of the house in May 2017 together with the total energy consumed and the amount of energy generated by PV and fed back to grid are presented in Figure 3.

Figure 3. The load of the house in May 2017 and the amount of energy generated by PV and fed back to grid.

4. Meteorological data

All results presented in this paper are the monthly average daily for the one whole year. Figure 4 presents the measured total solar radiation and wind speed from May 2017 to April 2018. These values varied from 384 W/m² in December to 538 W/m² in May. From the measured results, it is clear that the solar radiation during the whole year in Oman is high with annual average value of 486 W/m². The solar radiation during Winter is lower compared with the other seasons. The wind speed ranges between 1.2 m/s in January to 2.5 m/s in May; in general, the wind speed in Muscat governorate is low compared with other sites in Oman. High wind speeds are beneficial to PV due to the cooling effect on the modules. The 5 years measured average values for solar irradiations in Muscat are depicted in Figure 5, these were obtained from Public Authority for Civil Aviation (PACA), Directorate General of Meteorology. The general trend of solar irradiations presented in Figure 4 is similar to what is shown in Figure 5.

Figure 4. Measured monthly global plane-of-array irradiation at tilt angle of 23.5° south and wind speed from May 2017 to April 2018.

Figure 5. Measured average monthly total plane-of-array irradiation obtained from PACA for 5 years.

4.1 Modelling of the module temperature

The module temperature ( can be calculated by applying the following equation (Lugue 2003)

(1)

where G is the solar irradiance in kW/m², T_a is the ambient temperature in °C, NOCT is the nominal operating cell temperature (45 ^oC as specified by manufacturer’s data sheet).

Figure 6 displays the measured average monthly ambient temperature, the sun-shine duration over 1 year and the calculated module temperature. The ambient temperature (daytime hours) varied between 22.9°C in January to 38.7^oC in May, whereas the sun-shine duration varied between 9.67 h in December to 12.9 h in June. The worst effect of the module temperature on the power production is during summer, the module temperature reached 55.5°C in May.

Figure 6. Measured monthly average ambient temperature and the sun-shine duration over the monitored period, and the calculated module temperature.

5. System analysis

The grid-connected PV system performance is normally evaluated using IEC 61,724 Standard (International Electrotechnical Commission 1998). The parameters include the following: energy produced, yields (array yield, final yield and reference yield), system efficiency, PR, CF and losses (array capture, system and cell temperature). Energy quantities are calculated and normalised to full load PV power and denoted to as yields, this represents the actual PV operation relative to its full load capacity (Padmavathi and Daniel 2013). Efficiencies are normalised to PV module area. These normalised parameters provide a basis for comparison for grid-connected PV system working under different operating conditions (Adaramola & Vagnes, 2015).

5.1 AC energy

The total monthly alternating current (AC) energy generated by the PV system can be calculated as:

(2)

where is the number of days in the month and is the total generated daily energy.

The instantaneous AC energy output from the inverter was measured on 20-s intervals. The total energy produced per month is illustrated in Figure 7. The total energy output per month ranged between 1465 kWh in November to 3473 kWh in March. The total annual energy produced was 27,067 kWh. The most productive period of the PV system is from March to May during which relatively higher solar radiations were recorded that increased the generated output energy.

Figure 7. Measured total energy generated per month over the monitored period.

Although Muscat has low annual rainfall, however rain fall is mostly from December to April. The total AC power generated from the PV modules, for three random consecutive days in December 2017, after the inverter is depicted in Figure 8.

The variations of power at each instant in time can be calculated by taking the difference between two consecutive samples (Santiago et al. 2015):

(3)

where and are the inverter output AC power at each instant of time and at the following one, , respectively. is the inverter power capacity.

Furthermore, the variations in power at each instant of time (20 s) are also represented in the Figure. As shown, the PV output power has almost the same changes through the year but with a change in the peak power. These variations make the PV energy forecast straightforward which should allow extensive integration of PV into Oman future power network with the presence of Energy Management System. To cover the load during the night and to stabilise the system energy, storage will be required.

Figure 8. The measured PV production and variation through three consecutive days in December 2017.

5.2 System yields

These consist of array, final and reference yields. The yields represent the actual PV operation relative to its full load capacity.

5.2.1 Array yield

It is defined as the ratio of the total direct current (DC) energy output from a PV array over a defined period (year, month or day) to the PV array full load power at standard test conditions (STC) and is given as (International Electrotechnical Commission, 1998; Ayompe et al. 2011):

(4)

The STC are done at irradiation of 1000 W/m², module temperature of 25°C and air mass (AM) of 1.5.

5.2.2 Final yield

It is the ratio between the total AC energy generated of the PV system and the full load power at STC and is given as [38] [34]:

(5)

5.2.3 Reference yield

It is the ratio between the global in-plan solar insolation (kWh/m²) and the array reference irradiance (1 kW/m²). It represents the solar energy available at a given location for a certain period and is given as (International Electrotechnical Commission, 1998; Ayompe et al. 2011):

(6)

Figure 9 shows the reference, array and final yields over one whole year. The final, reference and array varied between 2.4 and 5.5 kWh/kWp-day, 3.7 and 6.9 kWh/kWp-day and 2.5 and 5.6 kWh/kWp-day, respectively. The annual average final, reference and array yields were 3.64 kWh/kWp-day, 5.59 kWh/kWp-day and 3.78 kWh/kWp-day, respectively. Relatively lower level of solar radiation in November resulted in low final yield. The higher value of reference yield during summer indicating more solar radiation is available during summer. However, the monthly average value of the final yield is small during summer owing to the excessive ambient temperature. The difference between the array yield and final yield is owing to the inverter losses. The final yields previously reported are: Oman, Sohar city, 5.1 kWh/kWp-day (Kazem et al. 2014), Kuwait, 4.5 kWh/kWp-day (Al-Otaibi 2015), Morocco, 4.45 kWh/kWp-day (Attari, Elyaakoubi, and Asselman 2016), India, 3.8 kWh/kWp-day (De Lima, Ferreira, and Morais 2017), Singapore, 3.12 kWh/kWp-day (Wittkopf et al. 2012), Germany, 1.8 kWh/kWp-day (Jahn and Nasse 2004). The final yield for Muscat city is 3.47 kWh/kWp-day, Oman is lower compared with Sohar city, Kuwait, Morocco, and close to India but higher than other places.

Figure 9. Monthly average final, reference and array yields over 1 year.

5.3 System efficiency

It is defined as inverter AC output power divided by the multiplication of total in-plan solar irradiation (W/m²) and PV array area (m²), and is given as:

(7)

where is the PV array area (m²) and is the global irradiation on the array plane (W/m²).

5.4 Capacity factor (CF)

It is the actual total energy generated per year divided by the amount of energy generated by the PV system when operated at full load for a year. The CF would be unity, if the generation system supplied full load power continuously and is given as (International Electrotechnical Commission 1998; Kymakis, Kalykakis, and Papazoglou 2009):

(8)

5.5 Performance ratio (PR)

It is the ratio between the actual and theoretical energy production of PV plant on a given time basis. It represents the effect of losses on a PV module’s normal power output and allows comparison of PV systems independent of geographical location, tilt angle, orientation and plant capacity. It is given as the final energy yield of the PV system divided by the reference yield (International Electrotechnical Commission 1998; Shiva Kumar and Sudhakar 2015).

(9)

Figure 10 displays the PV CF and PR over 1 year. The CF varied between 0.11 in June and 0.23 in March, and with an annual average of 0.15. The average annual value of CF (0.15) is almost similar to CF of a typical plant in Morocco (0.1484) (Attari, Elyaakoubi, and Asselman 2016), and in India (0.1569) (Padmavathi and Daniel 2013) but higher compared to that in Norway (0.1058) (Adaramola & Vagnes, 2015) and in Serbia (0.1288)(Dragana and Tomislav 2015). The PR varied between 0.37 in June and 0.96 in March, with an average of 0.67 per year which is similar to what was reported in (Al-Badi 2018). The PR reported in southern Algeria was varied from 0.67 to 0.86 (Necaibia et al. 2018), whereas the annual average reported in northern Ghana and in Algerian Sahara were 0.7 (Mensah, Yamoah, and Adaramola 2019) and 0.72 (Sahouane et al. 2019), respectively. Their values are not far compared to the results presented in this paper. The values of the PR are relatively low, which could be explained due to the fact that the high temperatures during summer increase the module temperatures, thus resulting in high capture and cell temperature losses which cause a reduction in system efficiency as shown in Figure 10. PR values and system efficiency are higher in winter than in summer owing to the cooling of module surface in winter.

Figure 10. Performance ratio and capacity factor per month over 1 year.

5.6 PV energy losses

There are a variety of losses occurred in the PV system and these are: array capture, system and cell temperature losses. There are additional losses that can occur in the PV system such as (Krauter 2006) reflection losses, soiling, degradation and thermal losses due to the increase in cell temperature, mismatch and losses due to shadowed cell and wiring losses. The energy output for the PV modules can degrade gradually (system’s age) owing to the climate factors effect such as temperature and humidity.

5.6.1 Array capture losses

They are owing to PV array losses, it indicates the losses that result from PV operation since the PV is not able to fully utilise the available radiation (due to such as dust accumulation, partial shading and inhomogeneous irradiance) and is given as (International Electrotechnical Commission 1998; Ayompe et al. 2011):

(10)

5.6.2 System losses

They are due to losses in the system components (such as wiring losses, diodes losses and system ageing) and are given as (International Electrotechnical Commission 1998; Ayompe et al. 2011):

(11)

5.6.3 Cell temperature losses

They are caused by operating PV modules above the standard test condition (irradiance 1000 W/m², module temperature 25°C and air mass 1.5). Based on Table 2, the peak power of the PV module decreases by 0.429% for every 1°C increase in the temperature of the cell above standard test conditions. The temperature losses can be calculated as (Sidi et al. 2016; Ueda et al. 2009):

(12)

where (kWh) is the AC output energy; (kWh) is the AC energy output temperature corrected and calculated as (Ueda et al. 2009):

(13)

where is the temperature coefficient of the PV module maximum power (−0.458%/℃) and is the measured PV module temperature (°C).

The capture, system and cell temperature losses over 1 year are shown in Figure 11. The system losses varied between 0.1 h/day in November and 0.2 h/day in April with an average value of 0.14 h/day, the capture losses varied between 0.38 h/day in February and 4.2 h/day in June with an average value of 1.82 h/day, and the cell temperature losses varied between 0.22 h/day in January and 5.8 h/day in May with an average value of 2.95 h/day. The cell temperature losses reached small value during the winterbut had a large increase during summer due to the increase in ambient and module temperatures. The capture losses have higher values in summer compared with winter which indicates that although solar radiation is high during summer, but due to high temperature, the modules are not able to generate the expected energy. The system losses are almost the same during the whole year with a little increase during summer time since this represents the losses in different system components.

Figure 11. The average capture, system and cell temperature losses per month over 1 year.

5.6.4 Factors affecting the cell temperature losses

The measured solar irradiation, ambient temperature and wind speed in an arbitrary week for 3 days in summer and winter are illustrated in Figure 12. All these factors (high solar irradiation, increased in ambient temperature and low wind speed) can increase the module temperature, as a result increases the cell temperature losses as shown in Figure 11. Furthermore, other factors may affect the module temperature such as type of PV cell and type of mounting structure.

Figure 12. Effects of ambient temperature, wind speed and solar irradiation on cell temperature losses during (a) winter; (b) summer.

5.7 Effect of dust on PV performance

PV modules must be cleaned regularly to avoid dust accumulation that reduces the transmittance of the glass of the PV modules, as a result energy production is affected. Dust pollution deposition is affected by several factors such as weather conditions and different site-specific environments (Darwish et al. 2015). The impact of dust accumulation, humidity level and wind speed on the efficiency of PV were discussed in Mekhilefa, Saidurb, and Kamalisarvestani (2012). Kaldellis and Kokala (2010) reported that Solar radiation is the main factor which affects PV power output, but soil and dirt can cause degradation up to 15% of the PV efficiency. Ibrahim et al. (2009) stated that a dusty module produced 25% and 35% lower energy compared to a cleaned one after a period of 3 months and 1 year, respectively. A study on a solar PV system near Riyadh, Saudi Arabia, indicated a 32% reduction in power, after 8 months, in comparison with an identical PV cleaned daily (Salim, Huraib, and Eugenio 1988). Bird dropping makes a hot spot which leads to local overheating of the module (Ghazi, Sayigh, and Kenneth 2014). The heat produced instead of electrical power decreases both the efficiency and the lifetime of the module. Power losses due to dust depend on the cleaning schedule of PVs, type of dust accumulated and length of time since the last rainfall. To investigate the effect of dust on the eco-house rooftop PV, the modules were left uncleaned.

Figure 13. Picture of unclean photovoltaic panels.

The picture of unclean PV panels is depicted in Figure 13. The presence of few bird droppings and little dirt on the modules is obvious in the picture. It should be noted that the amount of dust accumulated on the PV in this area is limited owing to the presence of mountains, and the PV panels were tilted. The percentage of reduction is small because of the Sultan Qaboos University-specific environment which is characterised by low dust accumulation. The instantaneous power generated during an arbitrary day in May 2017 and May 2018 are presented in Figure 14. The reduction in the peak power is clear Figure 14(b). The peak power was close to 15 kW in May 2017, however it was reduced to 13 kW in May 2018.

Figure 14. The instantaneous power generated in (a) May 2017 and (b) May 2018.

6. Conclusions

In this research paper, the measured results of 20.4 kW_P grid-connected PV plant, under Muscat governorate environmental conditions, are discussed. The system was monitored since December 2014 until today. The facility is placed on the rooftop of Sultan Qaboos University eco-house, Oman. The produced energy is consumed in the eco-house and the access energy is fed back to the electric grid. Two-ways metre was used to record the energy taken from the grid or fed back to the grid.

The total annual energy produced was 23,595 kWh, whereas the annual average daily reference yield and array yield were 5.59 and 3.78 kWh/kW_P/day respectively. The annual average daily CF was 0.15, while the annual average system efficiency was 10.3%. Furthermore, the annual average daily capture losses, system losses and cell temperature losses were 1.82, 0.14 and 2.95 h/day, respectively.

The annual average value of the PR for the rooftop PV was 0.67 and the final yield for this system was 3.64 kWh/kW_P/day. These values fluctuate over the year which could be attributed to the high temperatures during summer that increase the module temperatures, thus resulting in high capture and cell temperature losses which cause a reduction in system efficiency.

To investigate the effect of dust on the eco-house rooftop PV, the modules were left uncleaned. The percentage of reduction of energy generated is small because of the Sultan Qaboos University-specific environment which is characterised by low dust accumulation.

The results of this paper are compared with results obtained from other researchers which internationally revealed that the monthly average daily CF reached 15% which is either higher or similar to other systems installed in different locations worldwide, whereas the annual average daily final yield reached 3.64 kWh/kWp-day which is either lower or higher compared with other systems around the world. The average system efficiency over 1 year reached 10.3% compared with the name plate PV module efficiency (15.33%) at STC.

Acknowledgements

The author would like to acknowledge The Research Council, Oman and Sultan Qaboos University for providing the financial support for construction of the Eco House and providing all required equipment, sensors and tools to do research.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was funded by The Research Council, Oman and Sultan Qaboos University.

Notes on contributors

Abdullah Al-Badi

Abdullah Al-Badi obtained the B.Sc. in Electrical Engineering from SQU, Oman, in 1991. He received the degree of M.Sc. and Ph.D from UMIST, UK, in 1993 and 1998, respectively. He has published more than 160 publications in many well-known international journals, proceedings of refereed international conferences and technical reports. He carried out  44 research projects and consultancy (Main or co-investigator), with total attracted fund of + US$2.2 million. His research interests include renewable energy, distributed generation, power quality, power system analysis and power electronics and drives. He is a Senior Member of the Institute of Electrical Engineering and Electronics, IEEE, USA and a consultant in Oman Society of Engineers. He is IEEE program evaluator for Accreditation Board for Engineering and Technology (ABET) and External Reviewer for Oman Academic Accreditation Authority (OAAA). In 2009 he was appointed as a Dean for the Deanship of Admissions and Registration and in Feb. 2014 was appointed as a Dean for the College of Engineering, he is currently a Professor in Electrical & Computer Engineering Department in Sultan Qaboos University.