nternational Journal of Applied Science and Technology ol. No. 3; March Performance Assessment of Polycrystalline Silicon Pv Modules in Low Latitude Regions as A Function of Temperature Abstract P.E. Ugwuoke, PhD National Centre for Energy Research and Development University of Nigeria Nsukka. C.E. Okeke, PhD Department of Physics and Astronomy University of Nigeria Nsukka. The performance of any P module is directly proportional to the solar irradiance impinging on its surface and having enough energy capable of releasing electrons from the valence level to the conduction level of the cells of the module. t has been observed that module performance is significantly hampered by thermal stress due to temperature variations during operation under natural environment. A one-year performance characteristics of a 5-Watt polycrystalline silicon P module as a function of temperature at Nsukka in Nigeria (Lat. 6 5 ) was evaluated and reported in this paper. Significant decrease in efficiency Eff, imum voltage, open-circuit voltage oc and fill factor FF was noticed with increase in the module temperature at irradiances above 6 W/m while imum current and short-circuit current sc showed no noticeable effect of the module temperature rise. At an irradiance of W/m and a module temperature of 45 C, the imum power output of the module was observed to be 8.84 Watts, which is approximately 58 % of the manufacturer s specification. The module performance ratio (MPR) at irradiance of 6W/m, defined as the ratio of effective efficiency to the efficiency at standard test condition (STC), was observed to be 76.75% and decreased appreciably to 6.7% at W/m. Key Words: Performance ariation, Temperature effects, Polycrystalline Silicon, Photovoltaic modules, Low Latitude.. ntroduction The P module output ratings and operation specifications are usually determined by solar P module manufacturers in their cold and temperate environments under simulated and steady-state sunlight conditions. These specifications are designated Standard Test Conditions (W/m irradiance, 5 C cell temperature and air mass of.5 global spectrum). However, when deployed in the field, the performances are bound to vary from the stipulated specifications because the solar intensity is both location and time dependent and is usually highly modulated due to the rapidly changing cloud cover and dust haze. This leads to variations in ambient and module temperatures []. Temperature has a non-negligible effect on the P module characteristics. The short-circuit current will increase slightly with increasing module temperature T mod. This increase in the short-circuit current arises from a decrease in the band-gap energy E g of the material as the temperature increases according to []. E T E T ----------------------------------------------------- () g g T b Band gap energy of silicon, E g () =.6 e, α = 7 x -4 ek - and b = K. But short-circuit current exp q C -------------------------- () SC KT Where = saturation current = AT 3 exp E g --------------------------- (3) KT 95
Centre for Promoting deas, USA www.ijastnet.com The open-circuit voltage oc would decrease linearly with increase in module temperature because of the exponential increase in saturation current[].. Experimental Procedures The 5-Watt polycrystalline silicon P module and the temperature sensors were installed on an elevated support structure at 46 metres above sea level [3]. The module was tilted at approximately to the horizontal and south-facing. The investigation was performed for a period of one year, spanning from November to October at the grounds of Energy Research Centre, University of Nigeria, Nsukka. The module has the following nominal parameters at STC: 96 No. of cells per module = 36 cells of 4 parallel and 9 series strings Maximum rated power = 5 W p Maximum rated voltage = 7. Maximum rated current =.9 A Open-circuit voltage =. Short-circuit current = 3. A Surface Area =.38 m Make = Tessag Solar ndustries, USA The sensors are connected directly to a software-based CRX Campbell Scientific data logger, while the module is connected to the logger via a voltage divider. nstantaneous data collection were performed by the logger at intervals of 5 minutes and averaged over minutes. Data download at the data acquisition site was performed every days to ensure effective and accurate monitoring of the data acquisition system (DAS). The global solar radiation was monitored using a solar radiation sensor called SENSOL-MONOKRSTALLN, manufactured by KS PHOTOOLTAK Company in Western Germany, with calibration of 6.6m/Wm -. The ambient temperature was monitored using a Campbell Scientific temperature sensor. The back-surface temperature of the module was monitored using copper-constantan thermocouples. 3. Results and Discussion Module performance response to temperature variations was monitored in terms of open-circuit voltage oc, short-circuit current sc, voltage at imum power, current at imum power, imum power P and efficiency E ff. n order to further determine the rate of variation of module response variables with module temperature, a linear statistical model Y = a + b (T mod 5) was fitted to the observed data to predict module performance at varying module temperatures, where the coefficient b is the rate of variation with respect to module temperature [4]. The temperature coefficients for currents α and for voltages β of the module were subsequently evaluated as shown in figures 4 7. These temperature coefficients are simply the values of the coefficients of their fitting terms [5, 6]. Table shows the annual average variation of open-circuit voltage, shortcircuit current and module temperature under different meteorological conditions at different times of the day. t is observed here that the short-circuit current peaks at between hours and 4 hours when the average irradiance and module temperatures are imum, while the open-circuit voltage peaks around hours when the average irradiance is relatively moderate and average module temperature is close to ambient. Figure shows the characteristics of the module as a function of module temperature. Here the increase in module temperature reduces the open-circuit voltage while the short-circuit current is noticed to be unaffected by the temperature rise. Figures and 3 show the average variation of open-circuit voltage and short-circuit current respectively with module temperature at different months of the year. The imum power, efficiency and fill factor were evaluated using the following expressions [7]: Max. Power P = -------------------------------------- (4) Fill factor FF = Efficiency Eff = Pin SC OC ------------------------------------- (5) OC P SC in FF OC A E SC e FF ---- (6)
nternational Journal of Applied Science and Technology ol. No. 3; March Where A is the module area. On the basis of the data obtained and analysis performed, the module temperature was responsible for a significant percentage of the variations in all response variables of the module. However, the module temperature has a linear relationship with ambient temperature. t is the temperature difference between the module temperature and ambient temperature that influences the module performances significantly. The module temperature is noticed to be primarily dictated by the ambient temperature and irradiance, as could be seen in the strong correlation of module temperature with ambient temperature and irradiance. The ambient temperature sets the base temperature of the module while the irradiance predominantly sets the temperature rise of the module which is observed to be about.5 C per Wm -. n order to comprehensively evaluate the module performances over a reasonable time frame, it is important to predict the module temperature as a function of ambient temperature, wind speed, global irradiance and relative humidity. A mathematical model by Osterwald et al [8] to predict the module temperature, T mod for the module as a function of ambient temperature Ta, wind speed, WS, global irradiance Hg and relative humidity RH, was fitted to the field monitored real data of these ambient parameters. The intended equation for the module temperature is given by [8]: T mod = K H g + K T a + K 3 RH + K 4 WS + C ------------------------------- (7) Using the NLREG data analysis software developed by P.H Sherrod [9], an empirical equation for the module temperature T mod, was established as T mod ( C) =.355H g +.75T a +.83RH +.87WS 4.83 --- (8) Where K, K, K 3 and K 4 are coefficients which show the rate of variation of T mod with H g, T a, RH and WS. Table is a correlation matrix showing the correlation of T mod with H g, T a, RH and WS while table 3 shows the summary of average performance response of the module as a function of different irradiance and module temperature levels. 4. Conclusion From the data analysis and performance evaluation of the polycrystalline module in our local environment, it was discovered that the imum power output and efficiency of the module were significantly lower than their rated performances. At irradiance of W/m, the power output reduced by about 4.3% of the manufacturer s specification. Maximum efficiency of 9.67% was achieved at irradiance of 6W/m. Module temperature was noticed to be significantly influential to the general performance of the modules. Open-circuit voltage oc, imum voltage, efficiency Eff and fill factor FF, were observed to be imum in the mornings when module temperatures are low, while the short-circuit current sc and imum current were observed to be imum in the afternoons when the insolation and module temperatures are highest. References. Atiku A.T. and Sambo A.S (99) Experimental Characterization of Photovoltaic Solar Cells Nig. J. Solar Energy ol.9 pp 85 97.. Del Cueto J.A. () Closed-form Solutions and Parameterization of the Problem of Current-oltage Performance of Polycrystalline Photovoltaic Modules Deployed at Fixed Latitude Tilt, Paper presented at the 3 st EEE Photovoltaic Specialists Conference and Exhibitions, Florida. 3. Emery K (996) Temperature dependence of Photovoltaic Cells, Modules and Systems, Proceedings of the 5 th EEE Photovoltaic Specialists Conference, pp75 78. 4. King D.L. (997) Temperature Coefficients for P Modules and Arrays: Measurements, Methods, Difficulties and Results, Paper presented at the 6 th EEE Photovoltaic Specialists Conference, California. 5. Lasnier F. and Gan Ang T. (99) Photovoltaic Engineering Handbook, published under the Adam Hilger mprint by OP Publishing Company Ltd, England. 6. Nsukka North-East Topographic Map, : 5,; Sheet 87 NE (963), published by the Federal Survey, Federal Republic of Nigeria. 7. Onyegegbu S.O. (989) Performance of Photovoltaic Cells in an Equatorial Climate, Solar and Wind Technology, ol. 6 (3) pp 75 8 8. Osterwald C. (3) Photovoltaic Module Thermal/Wind Performance: Long-term Monitoring and Model Development for Energy Rating NCP and Solar Program Review, NREL pp 936 939. 9. Sherrod P.H. (99) NLREG Non-linear Regression Analysis Program ersion 6.3 (www.nlreg.com) 97
Current (A) Centre for Promoting deas, USA www.ijastnet.com Table. Annual Hourly Average Performance of the module. Polycrystalline Module T amb ( O C) Diffuse rradiance (W/m ) Global rradiance (W/m ) Wind Speed (m/s) Time oc sc T mod T A H d H W.S. 8 8.9.6 5.48 4. 7.68 6.6.97 9 8.87.9 8.8 4.79 3.5 59.3.48 9..45 3.44 5.85 37.54 4.65.449 8.96.6 34.54 6.58 4.4 479.37.636 8.83.9 36.59 7.5 46.8 5..58 3 8.77.98 37.9 8.3 45.5 54.33.946 4 8.74.84 38.5 9.9 43.3 5.5.86 5 8.54.55 37.4 9.39 38.38 438..747 6 8.43. 34.59 9.37 33.47 38.44.73 7 8.5.67 3. 9.4 3.75 7.68.593 8 6.53.8 9.8 8.58 9.6 5.3.9 Table. Correlation Matrix of T mod with Ambient Parameters for the polycrystalline silicon module Polycrystalline Tmod T mod. H g.999 Ta.9 RH -.4945 WS.35.5 44 C.5 39 C 37 C 3 C.5 4 6 8 4 6 8 oltage () Fig.. - Characteristics as a function of module temperature for the Polycrystalline Silicon P Module 98
Open-circuit oltage () Module Temp. ( O C) Short-circuit current (A) Module temp. ( O C) Open-circuit voltage () nternational Journal of Applied Science and Technology ol. No. 3; March 4 Tmod oc 9. 35 9 3 8.8 5 5 8.6 8.4 8. 8 7.8 5 7.6 Oct. Nov. Dec. Jan. Feb. Mar. Apr. May. Jun. Jul. Aug. Sep. 7.4 Time of the year (months) Fig.. Monthly Average variation of open-circuit voltage and module temperature for the Polycrystalline module 4 35 Tmod sc.6.4 3. 5.8 5.6.4 5. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May. Jun. Jul. Aug. Sep. Time of the year (months) Fig. 3 Monthly Average variation of short-circuit current and module temperature for the Polycrystalline module 9. 9. oc measured 9 8.9 8.8 oc predicted 8.7 8.6 8.5 8.4 β =.36 8.3 8. 3.79 3.9 35.8 36.9 38.9 43.5 43.53 45.7 Fig. 4 Measured and predicted oc as a function of module temperature for the polycrystalline module 99
(A) () Short-circuit current (A) Centre for Promoting deas, USA www.ijastnet.com 3.5.5 SC predicted SC measured.5 α =.97 Fig. 5 3.79 3.9 35.8 36.9 38.9 43.5 43.53 45.7 Measured and predicted sc as a function of module temperature for the polycrystalline module 5 4.5 4 predicted measured 3.5 3.5 β =.643.5 Fig. 6 3.79 3.9 35.8 36.9 38.9 43.5 43.53 45.7 measured vs predicted as a function of module temperature for the polycrystalline module.5.5 predicted measured.5 α =.7883 3.79 3.9 35.8 36.9 38.9 43.5 43.53 45.7 Fig. 7 measured vs predicted as a function of module temperature for the polycrystalline module 3
nternational Journal of Applied Science and Technology ol. No. 3; March Glob. rrad. H (W/m ) Table 3. Summary of Average Performance Response for the Polycrystalline P module at different rradiance and module temperature levels. Diffuse rrad. H d (w/m ) T amb (o c ) Wind Speed W.S. (m/s) T mod (o c ) oc () 3 38.3 5.94.334 3.79 8.95. 4.8.55 8.5 7. 38.75 4 37.4 6.74.4 3.9 9..33 4.7.87.33 8. 48.49 5 47.5 7.98.4 35.8 8.94.4 4.4. 7. 9.7 63.6 6 45.5 8.5.64 36.9 8.85.5 4..55.89 9.67 76.4 7 54.78 9.6.77 38.9 8.8.9 4.9.69 3.8 9. 6.6 8 57.6 9.94.8 43.5 8.75.38 3.65.85 5.5 8.36 56.58 9 59.9 3.5.645 43.53 8.64.7 3.3.4 8.3 8.34 55.83 53.35 3.38.3 45.7 8.56.84.76.6 8.84 7.65 54.7 sc (A) () (A) P (W) (%) FF (%) 3