Michael G. Deceglie

Thermal and Electrical Effects of Partial Shade in Monolithic Thin-Film Photovoltaic Modules

Photovoltaic cells can be damaged by reverse bias stress, which arises during service when a monolithically integrated thin-film module is partially shaded. We introduce a model for describing a modules internal thermal and electrical state, which cannot normally be measured. Using this model and experimental measurements, we present several results with relevance for reliability testing and module engineering: Modules with a small breakdown voltage experience less stress than those with a large breakdown voltage, with some exceptions for modules having light-enhanced reverse breakdown. Masks leaving a small part of the masked cells illuminated can lead to very high temperature and current density compared to masks covering entire cells.

Outdoor performance of a thin-film gallium-arsenide photovoltaic module

We deployed a 855 cm2 thin-film, single-junction gallium arsenide (GaAs) photovoltaic (PV) module outdoors. Due to its fundamentally different cell technology compared to silicon (Si), the module responds differently to outdoor conditions. On average during the test, the GaAs module produced more power when its temperature was higher. We show that its maximum-power temperature coefficient, while actually negative, is several times smaller in magnitude than that of a Si module used for comparison. The positive correlation of power with temperature in GaAs is due to temperature-correlated changes in the incident spectrum. We show that a simple correction based on precipitable water vapor (PWV) brings the photocurrent temperature coefficient into agreement with that measured by other methods and predicted by theory. The low operating temperature and small temperature coefficient of GaAs give it an energy production advantage in warm weather.

Partial shade stress test for thin-film photovoltaic modules

Partial shade of monolithic thin-film PV modules can cause reverse-bias conditions leading to permanent damage. In this work, we introduce a partial shade stress test for thin-film PV modules that quantifies permanent performance loss. The test reproduces shading and loading conditions that may occur in the field. It accounts for reversible light-induced performance changes and for the effects of light-enhanced reverse breakdown. We simulated the test procedure using a computer model that predicts the local voltage, current and temperature stress resulting from partial shade. We also performed the test on three commercial module types. Each module type we tested suffered permanent damage during masked ash testing totaling < 2 s of light exposure. During the subsequent stress test these module types lost 4%{11% in Pmp due to widespread formation of new shunts. One module type showed a substantial worsening of the Pmp loss upon light stabilization, underscoring the importance of this practice for proper quantification of damage.

Metastable changes to the temperature coefficients of thin-film photovoltaic modules

Transient changes in the performance of thin-film modules with light exposure are a well-known and widely reported phenomenon. These changes are often the result of reversible metastabilities rather than irreversible changes. Here we consider how these metastable changes affect the temperature dependence of photovoltaic performance. We find that in CIGS modules exhibiting a metastable increase in performance with light exposure, the light exposure also induces an increase in the magnitude of the temperature coefficient. It is important to understand such changes when characterizing temperature coefficients and when analyzing the outdoor performance of newly installed modules.

Evaluation of PV module field performance

This paper describes an effort to inspect and evaluate PV modules in order to determine what failure or degradation modes are occurring in field installations. This paper will report on the results of six site visits, including the Sacramento Municipal Utility District (SMUD) Hedge Array, Tucson Electric Power (TEP) Springerville, Central Florida Utility, Florida Solar Energy Center (FSEC), the TEP Solar Test Yard, and University of Toledo installations. The effort here makes use of a recently developed field inspection data collection protocol, and the results were input into a corresponding database. The results of this work have also been used to develop a draft of the IEC standard for climate and application specific accelerated stress testing beyond module qualification.

A scalable method for extracting soiling rates from PV production data

We present a method for analyzing time series production data from photovoltaic systems to extract the rate at which energy yield is affected by the accumulation of dust, dirt, and other forms of soiling. We describe an approach that is based on prevailing methods, which consider the change in energy production during dry periods. The method described here builds upon these methods by considering a statistical sample of soiling intervals from each site under consideration and utilizing the robust Theil-Sen estimator for slope extraction from these intervals. The method enables straightforward application to a large number of sites with minimal parameterization or data-filtering requirements. Furthermore, it enables statistical confidence intervals and comparisons between sites.

Real-Time Series Resistance Monitoring in PV Systems Without the Need for I–V Curves

We apply the physical principles of a familiar method, suns-Voc, to a new application: the real-time detection of series resistance changes in modules and systems operating outside. The real-time series resistance (RTSR) method that we describe avoids the need for collecting I-V curves or constructing full series resistance-free I-V curves. RTSR is most readily deployable at the module level on microinverters or module-integrated electronics, but it can also be extended to full strings. Automated detection of series resistance increases can provide early warnings of some of the most common reliability issues, which also pose fire risks, including broken ribbons, broken solder bonds, and contact problems in the junction or combiner box. We describe the method in detail and describe a sample application to data collected from modules operating in the field.We apply the physical principles of a familiar method, suns-Voc, to a new application: the real-time detection of series resistance changes in modules and systems operating outside. The real-time series resistance (RTSR) method that we describe avoids the need for collecting IV curves or constructing full series-resistance-free IV curves. RTSR is most readily deployable at the module level on micro-inverters or module-integrated electronics, but it can also be extended to full strings. Automated detection of series resistance increases can provide early warnings of some of the most common reliability issues, which also pose fire risks, including broken ribbons, broken solder bonds, and contact problems in the junction or combiner box. We describe the method in detail and describe a sample application to data collected from modules operating in the field.

Robust measurement of thin-film photovoltaic modules exhibiting light-induced transients

Light-induced changes to the current-voltage characteristic of thin-film photovoltaic modules (i.e. light-soaking effects) frustrate the repeatable measurement of their operating power. We describe best practices for mitigating, or stabilizing, light-soaking effects for both CdTe and CIGS modules to enable robust, repeatable, and relevant power measurements. We motivate the practices by detailing how modules react to changes in different stabilization methods. We also describe and demonstrate a method for validating alternative stabilization procedures, such as those relying on forward bias in the dark. Reliable measurements of module power are critical for qualification testing, reliability testing, and power rating.

Performance stabilization of CdTe PV modules using bias and light

Reversible performance changes due to light exposure frustrate repeatable performance measurements on CdTe photovoltaic modules. It is common to use extended light exposure to ensure that measurements are representative of outdoor performance. We quantify the extent to which such a light-exposed state depends on module temperature and consider bias in the dark to aid in stabilization. We evaluate the use of dark forward bias to bring about a performance state equivalent to that obtained with light exposure, and to maintain a light-exposed state prior to standard test condition (STC) performance measurement. Our results indicate that the most promising method for measuring a light-exposed state is to use light exposure at controlled temperature followed by prompt STC measurement with a repeatable time interval between exposure and the STC measurement.

Optical cell temperature measurements of multiple CPV technologies in outdoor conditions

It is well known that photovoltaic performance is dependent on cell temperature. Although various methods have been explored to determine outdoor concentrating photovoltaic (CPV) cell temperature, no method has proven to work across all module technologies and result in desirable uncertainties. Menard (2012) has recently published results claiming accurate measurements of cell temperature using the wavelength shift of light emitted from the sub-cells of a Semprius CPV module. This work focuses on efforts to verify Menards results using additional CPV technologies that are on-sun at NREL. Baseline electro-luminescence emission is recorded for modules under a low level forward bias and under isothermal conditions using thermal chambers. The same modules or sister modules are then placed on NRELs high accuracy two-axis tracker for outdoor measurements. Photo-luminescence emission peaks are measured for multiple modules at stable wind and irradiance conditions. Emission results from the sub-cells are compared to what is documented in the literature for the given semiconductor material. The signal to background ratio is analyzed and the possible broad applicability of this procedure is discussed.