Project Data

End-User: East Coast Warehouse

Installer: Forest Electric

Location: Port Elizabeth, New Jersey

Date of install: September 2006

Size of system in kW: 700kW

Energy produced since installation: 1.4GWh

Special Challenges Met: Building Integrated Photovoltaics

Project Brief

There are several PV system performance analysis tools that are routinely used to predict the energy production of photovoltaic power systems.  Possibly the most commonly used tools are Clean Power Estimator and PV Watts.  Of these two, the PV Watts program usually produces a lower estimate, compared to the Clean Power Estimator.  Input variables for these software tools include geographic location, historic weather data bases, tilt angle, azimuth, soiling factors, inverter efficiency and system wiring losses.  However, all of these programs are based on the performance characteristics of crystalline photovoltaic panels.  Therefore, in order to accurately predict the performance of thin-film photovoltaic power systems, adjustments must be made.  Also, aside from the measured characteristics of the crystalline photovoltaic panels and the calculated tilt of the array, the other variables are subject to approximations based on “best guess” estimates.

It is a well-established fact that the triple-junction, thin-film, amorphous silicon photovoltaic panels produce more electrical energy per watt than crystalline photovoltaic panels.  This is primarily due to the following factors:

-Less than ½ the cell temperature de-rating compared to crystalline silicon PV (better performance at high cell temperatures)

-Higher efficiencies at low light levels, diffused light, and low-angle light

-Less affected by soiling and shading factors

Over the past several years, many comparative studies have been done to both verify and quantify this significant difference in energy production performance.  Attached are a series of independent technical papers and associated documents that are examples of the published results of these tests and analyses.

The first is A Real World Examination of PV Systems Design and Performance presented at the 2005 IEEE Photovoltaic Specialist Conference.  This paper documents the results of a year-long test that compares the performance of crystalline and thin-film technologies on a single roof in Santa Cruz, CA.  The data shows that the Uni-Solar Thin-film system (sloped arrays) produced 20% more energy than an identically sized crystalline system.  It also showed that a zero-slope Uni-Solar array produced as much energy as the sloped crystalline array of equal size.

The second paper is Superior Energy Yields of Uni-Solar Triple-Junction Thin Film Silicon Solar Cells compared to Crystalline Silicon Solar Cells under Real Outdoor Conditions in Western Europe, presented at the European Photovoltaic Solar Energy Conference in 2001.  As the title of the paper indicates, the thin-film technology is a superior energy producer when compared to crystalline silicon PV.  In this study, a total of 14 different systems were compared in several western European locations to illustrate, with overwhelming data, that in real-world conditions, even in a less-than-optimum solar environment, the thin-film product was superior.

The third paper is Flat roof integration from the Swiss Federal Office of Energy Annual Report for 2004.  In section 3.3 of this paper (Energy Production), it is noted that the performance of the Uni-Solar PV array was as much as 30% greater than the predicted energy output. 

The fourth paper, Performance Analysis of Large Scale, Amorphous Silicon Photovoltaic Power Systems, presented at the 2005 IEEE Photovoltaic Specialist Conference, examines two years of performance data from the first 500kW thin-film PV power system installed in the US.  The key point of this study is the verification of the consistent, better-than-predicted, performance of a large scale Uni-Solar PV system.  This system was designed and installed by a prominent crystalline PV power system company as a 500kWAC system.  It turned out that the superior performance of the thin-film array has produced measured power outputs of 540kWAC, 8% more than had been expected from the same size crystalline system.

The last set of pages are a collection of charts and graphs that are made up of performance data from crystalline and thin film systems.  Each one illustrates and demonstrates the significant performance differences at different light levels and temperature levels as well as system performance data from actual systems throughout Europe.   The stated conclusion is that for the cases addressed, the Uni-Solar arrays produced from 15% to 35% more energy than crystalline arrays.

The over-all conclusion of these and several other reports is that the predicted energy output in kilowatt-hours, for any crystalline PV system will be 10% to 30% less than a thin-film system of the same rated capacity in watts.