The 39th IEEE photovoltaic specialist conference was held between June 16th and 21st at the Tampa bay convention center in Tampa, Florida. It was a congregation of industry experts, and research giants. Researchers from NREL, Sandia National Laboratories and Universities across the globe graced the occasion to present their latest studies on photovoltaic system design, implementation and reliability of on-sun PV modules.. The program was significantly all encompassing. Besides the presentations, social activities and mixer programs were held to allow attendees to interact, network and share knowledge. Of notable interest was the presentation of the cherry award to Keith Emery. Previously unknown to me, I found that he is renowned for his contribution to photovoltaic research for his design, development and implementation of IV characterization methods. He pioneered the first generation of hardware, software and procedures to measure current-vs.-voltage characteristics as a function of temperature, spectrum and intensity for single and multi-junction cells and modules.

Oral and poster presentations at the conference were grouped into eleven categories which ran in parallel beside keynote or plenary sessions. Personally, I attended sessions in the categories of advanced PV module concepts and designs and PV modules and terrestrial systems. From the presentations, I deduced that there is a significant amount of attention being given to system performance evaluation and energy yield assessments of photovoltaic systems. As such, there is a growing interest in research on concepts for data collection which is a necessary input for energy assessments. There were also presentations on the design of experiments for photovoltaic system assessments. Particularly, I found some modeling techniques used to evaluate PV system performance to be of interest. A few of them include:

Validation of the PVLife Model Using 3 Million Module-Years of Live Site Data [1]

In this article, SunPower corporation (the manufacturers of SunPower PV modules) presented their experiments and results on long term system degradation analysis. An interesting fact is that they performed their analysis using a relatively new approach. Rather than using high fidelity diagnosis methods, they settled for noisy large statistical samples that represent records from a large number of installed systems to estimate the median system degradation rate of PV modules. As a key player in the PV module industry, the company aimed to consolidate their understanding and confidence in system degradation trends and hence they’ve developed a model called the “PVLife model” which is used to simulate module degradation characteristics. Their PVLife model operates with inputs of weather data and cell characteristics to determine degradation factors such as UV induced cell degradation, encapsulant browning, bypass diode and solder joint failures.

For comparison, degradation analysis was carried out on a total of 445 systems.  226 systems were comprised of SunPower modules which had an installed capacity of 86MW. These systems had been operating for up to 5.5 years.  There were also included 149 systems of non SunPower modules which were as old as 11.5 years with an installed capacity of 42MW. Altogether, the total fleet-wide modules representing 3.2 million module-years of monitored data were used to determine degradation rates. Following a plethora of statistical analytics, they found the PVLife model to be in very good agreement with the compared module degradation rates. It was further claimed that the model results were used to develop better modules with lower degradation rates. Attention was focused on the relationship between degradation rates and the placement of the module contacts. According to their studies, it was found that front contact modules for a variety of reasons had a higher degradation rate when compared to back contact modules.

Overall, the work by SunPower suggests that they have successfully developed a working system to model the degradation mechanisms due to several factors in PV system operation. Validation of its results against a large dataset of on-sun measurements was shown to be in very good agreement

Simulations of Energy Yield Improvement in Utility-Scale PV Plants Using Distributed Power Point Trackers [2]

Researchers from First Solar Inc. presented their research on the use of distributed maximum power point trackers (MPPT). It was identified that any energy losses in utility scale PV installations decrease the financial value of the system. They aimed to analyze methods that might reduce the losses in utility scale PV installation such as partial cloud cover induced mismatch loss in the system. Since similar loss mechanisms due to building shading has been analyzed in detail, their focus was directed at non-uniform irradiance patterns created by cloud edges on utility-scale PV installations.  A model was developed to simulate the mismatch loss. The model simulates a PV array with a variable number of strings on a mounting structure. It simulates the movement of a cloud edge over the PV array whilst outputting the array IV characteristics based on voltage and current relationships. A study was then conducted on varying string lengths in the arrays for 10 and 15 modules in a string. Results from the analysis using their model showed a decrease in net energy loss when multiple maximum power point trackers were used. The key energy losses were found to be dependent on the length of the cloud edge. Measurement of the cloud edge in correlation the utility –scale system performance was prescribed to be conducted to further assess the impact of distributed MPPT in decreasing net energy losses.

Overall, the conference was a great learning experience. Its success encourages me to look forward to the next annual conference, which is scheduled to be held in Denver, Colorado

Jafaru Mohammed's picture.

-Jafaru Mohammed

M.A.Sc Electrical and Computer Engineering

Department of Electrical Engineering and Computer Science

SUNLAB Research Group

University of Ottawa, Canada.


[1]         E. Hasselbrink, M. Anderson, Z. Defreitas, M. Mikofski, Y. Shen, S. Caldwell, D. Kavulak, Z. Campeau, D. Degraaff, S. Corporation, R. Robles, and S. Jose, “Validation of the PVLife Model Using 3 Million Module-Years of Live Site Data,” in Photovoltaic Specialists Conference, 2013. PVSC’13. 39th IEEE, 2013.

[2]         A. Pope, J. E. Schaar, M. Schenck, F. Solar, and S. Francisco, “Simulations of Energy Yield Improvement in Utility – Scale PV Plants Using Distributed Maximum Power Point Trackers,” in Photovoltaic Specialists Conference, 2013. PVSC’13. 39th IEEE, 2013.