To achieve net-zero energy buildings, renewable energy generation is required to meet the energy demands remaining after load reduction and efficiency efforts have been implemented. Solar photovoltaic (PV) arrays typically offer the best means for providing this energy source. The decisions of which photovoltaic product is selected and how each system is designed, operated, and maintained depends, in large part, on the electrical performance information provided to the decision makers (e.g., the PV array owner, facilities manager, financier). This project will improve the measurement science associated with characterizing solar cells and calibrating reference cells and so directly impacts the quality of the available electrical performance information of the photovoltaic product that is ultimately purchased. FY14 work that leads to the creation of a solar cell calibration service at NIST will decrease the measurement uncertainty in electrical performance ratings of solar devices, thus giving more confidence to those specifying systems. Planned work to collect, archive, mine, analyze, and share high quality data from four well instrumented solar field arrays will help address shortcomings identified by the solar PV modeling community to improve the existing computer models, to get more consistency among the execution of the models by users, and to establish best practices for the PV field monitoring. These model predictions play a huge role in determining whether a project gets approved for construction, what design and PV product are used, and who is awarded the contract. Getting the modeling right is especially important to the program objective because an accurate model is needed to determine the least expensive system design that will also provide enough annual output to realize net-zero energy.
Objective: To develop and improve the measurement science by FY2016 needed: (1) to more accurately characterize the electrical performance of solar photovoltaic cells, and (2) to more accurately and consistently predict the electrical performance of installed arrays by improving both the modeling algorithms and the practices used to efficiently collect and use field verification data.
What is the new technical idea? NIST will seek to better both laboratory characterization of PV cells and model predictions of PV system output. NIST has been successful in developing a hybrid monochrometer + light-emitting diode (LED) based spectral response measurement technique and applying it to single-junction, monocrystalline silicon (mono-Si) PV cells. Although a prevalent technology, single-junction mono-Si cells are the easiest cell type to characterize. The application of the technique to non mono-Si cells and to cells having multiple junctions will be thoroughly evaluated and steps will be taken to minimize the measurement uncertainty (e.g., by using different light biasing techniques, adding voltage biasing). In addition to spectral response, quantifying the degree of linearity of a cell’s current output with respect to irradiance is an important component to cell characterization. NIST will explore using LEDs and a flux addition framework. This approach should be less onerous than the approach specified in the authoritative ASTM standard.
With regard to PV system output, the new technical idea is to quantify the benefits of high quality field data, obtained from NIST testbeds, for: (1) comparing the predictive capabilities of different PV system modeling algorithms, (2) making improvements to existing and developing new prediction algorithms, and (3) quantifying the impact from using progressively lower quality field measurements (e.g., less time resolution, fewer field instruments/measurements, instruments with higher uncertainties, a poorer equipment maintenance schedule) for model comparisons and for model inputs.
What is the research plan? The plan for the solar cell electrical performance characterization effort is to obtain at least four unique, non mono-Si single-junction test samples by partnering with National Laboratories and various commercial PV companies. The hybrid spectral response technique will be evaluated to identify the best implementation of the technique for each of the obtained PV technologies. Using those samples that are single-junction (but not mono-Si), a flux addition approach that uses a minimum of two LEDs will be explored for the purpose of quantifying a cell’s (non) linearity. To complement the cell characterization effort by helping to explain observed macro performance differences, a test method for measuring charge carrier lifetimes within the cell will be investigated. In looking ahead to offering a calibration service, the researchers will participate in a Round 2 World Intercomparison of filtered reference cells organized by the Bureau International des Poids et Mesures (BIPM). This comparison will aid NIST in evaluating its progress towards optimizing its hybrid technique. The researchers will also identify and establish agreements to leverage other measurement services, private and public, in accordance with the NIST “Checklist for the Establishment of a NIST Calibration Service.” Finally, cell characterizations and equipment calibrations will be conducted periodically for the sister project in the program entitled “Measurement Science for Service Life Prediction of Polymers Used in Photovoltaic Systems.”
To assist the PV community’s efforts to improve modeling of system output, NIST will start a multi-year effort to use data from four field-sites and a rooftop weather and PV test station to evaluate discrepancies between model predictions and measured data. NIST will automate the process of collecting, conducting quality checks, and archiving the large amount of data that are continuously generated at the field sites and then used, initially, by NIST researchers. Separate instrumentation to track and measure the sun’s spectral and broadband irradiance will be deployed to develop best approaches for collecting data that are required as input to some models. NIST will then utilize data from the facilities in developing detailed computer models of each of the four field sites using the industry’s most widely-used program, PVsyst. The resulting high-quality data covering a minimum of two years of operation and the lessons learned in comparing the measured output with the model predictions will be used to establish best practices for PV system modeling and for creating and maintaining outdoor test stations. These best practices are planned to be collaboratively developed by the PV Performance Modeling Collaborative (http://pvpmc.org), with one or both evolving into a formal guideline.
Start Date:October 1, 2011
Lead Organizational Unit:el
Project Leader: Brian P. Dougherty
Related Programs and Projects:
100 Bureau Drive, M/S 8632