Murdoch University Research Repository

Welcome to the Murdoch University Research Repository

The Murdoch University Research Repository is an open access digital collection of research
created by Murdoch University staff, researchers and postgraduate students.

Learn more

Simulation and performance evaluation of a photovoltaic system for remote area applications

Lim, Alexander Alamil (1993) Simulation and performance evaluation of a photovoltaic system for remote area applications. PhD thesis, Murdoch University.

PDF - Whole Thesis
Available Upon Request


The main objective of the work described in this thesis is to simulate and evaluate the performance of a transportable photovoltaic system, the Solar Pack, that is currently deployed to serve the Aboriginal communities in several remote areas of Australia and which could possibly be adopted in many developing countries. This study also addresses the necessary meteorological and PV system parameters and methods needed to predict the Solar Pack's energy output.

Realizing the fact that the diffuse and/or direct beam components of the global radiation are very seldom measured, especially in many developing countries, a correlation model is developed to estimate the hourly diffuse fraction in terms of the measured hourly global irradiance and clearness index for Perth, Western Australia. Half-hourly diffuse and global irradiation data from the Australian Bureau of Meteorology were integrated and used in the development of a linear model. The validity of the model is tested against similar well-known models and actual meteorological data and found to perform better statistically in terms of the root mean square error, mean bias error and absolute mean percentage error. The diffuse or direct irradiance values estimated using the model agree quite well with the measured values with slight underestimation particularly for times of day with high clearness index values. Overestimation is observed for low to intermediate clearness index values. This tends to indicate that the model could be further improved if other predictive parameters such as information on the air mass or solar altitude is incorporated in the model.

The computer simulation program, PVFORM, is used to predict the hourly plane of array insolation and the Solar Pack PV performance characteristics. The result of a mean bias error test (MBE = +0.16) on the comparison between the measured and estimated plane of array (POA) insolation indicates that PVFORM tends to slightly overestimate the POA insolation particularly during the summer months. Slight underestimation is observed for the winter months of June and July. The average overall accuracy of PVFORM in estimating the POA insolation based on the mean percentage error test is 4.31%.

The correlation model developed in this study to estimate the hourly diffuse fraction of the global radiation is also used to generate the hourly direct radiation values as required by the PVFORM insolation model. A mean percentage error of 4.99% is obtained for 1991-1992 period. The results indicated that the correlation model can also be employed to generate direct radiation values that can be used as inputs by the PVFORM insolation model.

The measured hourly performance of the Solar Pack PV system is compared with that predicted by the PVFORM simulation program. Initial results show that the simulated values overestimated the measured values of PV energy. A high mean percent error of 71.75% was obtained. This was found to be due to the cutting-off function of the voltage controller which isolates the main PV array of the Solar Pack when the battery voltage reaches about 28 volts. The introduction of additional load has greatly reduced the difference between the measured and estimated PV energy where the mean percent error decreased to 10.53%. Overestimation, however, is still observed based on the mean bias test and this is primarily due to the absence of a maximum power point tracker (MPPT) on the Solar Pack. The PVFORM simulation program operates on the assumption that the PV system being simulated is maximum power-tracked. Other factors that cause these discrepancies are investigated. When the PVFORM dc power calculation code was modified to include of the effect of the cutting-off function of the electronic controller, the average overall accuracy of PVFORM in estimating PV energy was found to be 4.3%.

The installation of a maximum power point tracker in 1993 reversed the comparative trend between the measured and estimated values. The mean bias error test shows that the PV energy underestimation is more pronounced during the summer months. The observed trend in the energy estimation tends to indicate that for a PV system which is maximum power-tracked, the PVFORM simulation program would provide slightly lower values than those measured particularly during the summer months.

Sensitivity analysis is performed to determine the effect of increasing the ambient temperature on the calculated PV energy. It was found that a 1°C change in ambient temperature would result in average 0.39% variation in calculated PV energy.

PVFORM's use of the battery state of charge (SOC) to describe the distribution of energy of the Solar Pack PV system is investigated. Possible incompatibilities between the SOC usage and Solar Pack system control are addressed. This has important implications for predicting the reliability and effectiveness of the PV system being simulated.

The application of the loss of load hours (LOLH) and loss of module hours (LOMH) parameters based on the PVFORM-calculated SOC values and the measured PV energy and battery voltage indicated that the reliability and availability of the Solar Pack cannot be accurately described in this way.

Due to the daily and seasonal variation in the POA insolation, the Solar Pack is found to provide an average of 3 - 8 kWh/day of PV energy. The minimum and maximum amount of energy occurring during the months of June and January, respectively. Calculation of the loss of load hours based on the measured hourly battery voltage indicated that a reliability of 100% is attained by the Solar Pack for the duration of this study. The average daily conversion efficiency of the Solar Pack with and without the MPPT is determined to be 8% and 8.6%, respectively.

This study shows that much of the energy produced by Solar Pack is dumped and there is a need for proper load management. Facilities that would allow intermittent operation should be added so that the system can be kept fully loaded.

Item Type: Thesis (PhD)
Murdoch Affiliation: School of Mathematical and Physical Sciences
Notes: Note to the author: If you would like to make your thesis openly available on Murdoch University Library's Research Repository, please contact: Thank you.
Supervisor(s): Jennings, Philip and Pryor, Trevor
Item Control Page Item Control Page