Daniel Riley

Comparison of PV system performance-model predictions with measured PV system performance

The U.S. Department of Energy has supported development of the Solar Advisor Model (SAM) to provide a common platform for evaluation of the solar energy technologies being developed with the support of the Department. This report describes a detailed comparison of performance-model calculations within SAM to actual measured PV system performance in order to evaluate the ability of the models to accurately predict PV system energy production. This was accomplished by using measured meteorological and irradiance data as an input to the models, and then comparing model predictions of solar and PV system parameters to measured values from co-located PV arrays. The submodels within SAM which were examined include four radiation models, three module performance models, and an inverter model. The PVWATTS and PVMod models were also evaluated.

Introduction to the open source PV LIB for python Photovoltaic system modelling package

The proper modeling of Photovoltaic(PV) systems is critical for their financing, design, and operation. PV_LIB provides a flexible toolbox to perform advanced data analysis and research into the performance modeling and operations of PV assets, and this paper presents the extension of the PV_LIB toolbox into the python programming language. PV_LIB provides a common repository for the release of published modeling algorithms, and thus can also help to improve the quality and frequency of model validation and inter comparison studies. Overall, the goal of PV_LIB is to accelerate the pace of innovation in the PV sector.

Photovoltaic prognostics and heath management using learning algorithms

A novel model-based prognostics and health management (PHM) system has been designed to monitor the health of a photovoltaic (PV) system, measure degradation, and indicate maintenance schedules. Current state-of-the-art PV monitoring systems require module and array topology details or extensive modeling of the PV system. We present a method using an artificial neural network (ANN) which eliminates the need for a priori information by teaching the algorithm “good” performance behavior based on the initial performance of the array. The PHM algorithm was tested on two PV systems under test at the Outdoor Test Facility (OTF) at the National Renewable Energy Laboratory (NREL). The PHM algorithm was trained using two months of AC power production. The model then predicted the output power of the system using irradiance, wind, and temperature data. Based on the deviation in measured AC power from the AC power predicted by the trained ANN model, system outages and other faults causing a reduction in power were detected. Had these been commercial installations, rather than research installations, an alert for maintenance could have been initiated. Further use of the PHM system may be able to indicate degradation, detect module or inverter failures, or detect excessive soiling.

Characterization and modeling of a grid-connected photovoltaic system using a Recurrent Neural Network

Photovoltaic (PV) system modeling is used throughout the photovoltaic industry for the prediction of PV system output under a given set of weather conditions. PV system modeling has a wide range of uses including: prepurchase comparisons of PV system components, system health monitoring, and payback (return on investment) times. In order to adequately model a PV system, the system must be characterized to establish the relationship between given weather inputs (e.g., irradiance, spectrum, temperature) and desired system outputs (e.g., AC power, module temperature). Traditional approaches to system characterization involve characterizing and modeling each component in a PV system and forming a system model by successively using component models. This paper lays the groundwork for using a Recurrent Neural Network (RNN) to characterize and model an entire PV system without the need to characterize or model the individual system components. Input/output relationships are “learned” by the RNN using measured system performance data and correlated weather data. Thus, this method for characterizing and modeling PV systems is useful for existing PV system installations with several weeks of correlated system performance and weather data.

Comparison of a recurrent neural network PV system model with a traditional component-based PV system model

Photovoltaic (PV) system modeling is used throughout the photovoltaic industry for the prediction of PV system output under a given set of weather conditions. PV system modeling has a wide range of uses including: pre-purchase comparisons of PV system components, system health monitoring, and estimation of payback (return on investment) times. In order to adequately model a PV system, the system must be characterized to establish the relationship between given weather inputs (e.g., irradiance, spectrum, temperature) and desired system outputs (e.g., AC power, module temperature). Traditional approaches to system characterization involve characterizing and modeling each component in a PV system and forming a system model by successively using component models. This paper compares a traditional modeling approach using the Sandia Photovoltaic Array Performance Model [1] to a new method of characterization using a recurrent neural network (RNN). The Sandia model predicts system performance from given weather data and individual component characterizations using a defined set of equations, while the RNN “learns” the input/output relationships by training on concurrent weather and performance data. The comparison of a traditional modeling technique and the new RNN method serves to validate the accuracy of the new method in comparison to a widely accepted modeling technique. Modeling using an RNN may be advantageous when component models are not available for the components in a PV system, when the components of a PV system are unknown to the modeler, or when system components are installed or altered in such a fashion that their model parameters are no longer applicable.

HCPV Characterization: Analysis of Fielded System Data.

Sandia and Semprius have partnered to evaluate the operational performance of a 3.5 kW (nominal) R&D system using 40 Semprius modules. Eight months of operational data has been collected and evaluated. Analysis includes determination of Pmp, Imp and Vmp at CSTC conditions, Pmp as a function of DNI, effect of wind speed on module temperature and seasonal variations in performance. As expected, on-sun Pmp and Imp of the installed system were found to be ∼10% lower than the values determined from flash testing at CSTC, while Vmp was found to be nearly identical to the results of flash testing. The differences in the flash test and outdoor data are attributed to string mismatch, soiling, seasonal variation in solar spectrum, discrepancy in the cell temperature model, and uncertainty in the power and current reported by the inverter.. An apparent limitation to the degree of module cooling that can be expected from wind speed was observed. The system was observed to display seasonal variation in performance, li...

Calibration of the Sandia Array Performance Model using indoor measurements

The Sandia Array Performance Model (SAPM) describes the DC output of a PV module under a range of irradiance and temperature conditions. Coefficients for SAPM are normally obtained through a sequence of on-sun tests, which can be expensive and time-consuming. We report progress towards developing test methods and analysis procedures to obtain coefficients for SAPM from indoor testing. We compared module output predictions from SAPM with coefficients extracted from indoor test results, to measured on-sun module output, and found biases in the predicted performance. We hypothesize that these biases result from the uniform cell temperatures during indoor testing, whereas measured cell temperatures vary by up to 10° C among cells during on-sun conditions. However, we also hypothesize other explanations for the observed biases.

Recent advancements in outdoor measurement techniques for angle of incidence effects

Reflection losses from a PV module become increasingly pronounced at solar incident angles >60°. However, accurate measurement in this region can be problematic due to tracker articulation limits and irradiance reference device calibration. We present the results of a measurement method enabling modules to be tested over the full range of 0-90° by articulating the tracker in elevation only. This facilitates the use of a shaded pyranometer to make a direct measurement of the diffuse component, reducing measurement uncertainty. We further present the results of a real-time intercomparison performed by two independent test facilities ~10 km apart.

Testing and characterization of PV modules with integrated microinverters

Photovoltaic (PV) modules with attached microinverters are becoming increasingly popular in PV systems, especially in the residential system market, as such systems offer several benefits not found in PV systems utilizing central inverters. PV modules with fully integrated microinverters are emerging to fill a similar market space. These “AC modules” absorb solar energy and produce AC energy without allowing access to the intermediate DC bus. Existing test procedures and performance models designed for separate DC and AC components are unusable when the inverter is integrated into the module. Sandia National Laboratories is developing a new set of test procedures and performance model designed for AC modules.

Performance Model Assessment for Multi-Junction Concentrating Photovoltaic Systems

Four approaches to modeling multi‐junction concentrating photovoltaic system performance are assessed by comparing modeled performance to measured performance. Measured weather, irradiance, and system performance data were collected on two systems over a one month period. Residual analysis is used to assess the models and to identify opportunities for model improvement.