The input of a solar inverter depends on multiple factors: the solar resource, weather conditions, and control strategies. Traditional design calculations specify the maximum current either as 125% of the rated module current or as the maximum 3 h average current from hourly simulations over a typical year, neglecting extreme irradiance conditions: cloud enhancement events that usually last minutes. Inverter power-limiting control strategies usually prevent extreme events to cause strong currents at the inverter, but in some cases, they can fail, leading to high currents. In this study, we aim to report how frequent and strong these high currents could be. We use 10 years of 1 min data from seven stations across the United States to estimate the photovoltaic string output through modeling the short-circuit current Isc, and the maximum-power point current Imp, and compare them to traditional inverter design values. We consider different configurations: minutely to hourly resolution; 5 min to 3 h averaging time intervals; monofacial and bifacial modules (with a case of enhanced albedo); and 3 fixed-tilt angles and horizontal single-axis tracking. The bifacial modules with enhanced albedo lead to the highest currents for 1 min data, exceeding 3 h averages by 53% for Isc and 38% for Imp. The 3 h average maxima surpass the conservative 125% design rule for bifacial modules. Inverter ratings at either a 200% of the rated current or 1.55 times the 3 h maximum could withstand all events regardless of control strategies. In summary, for some locations it is prudent to compare current design rules to subhourly simulations to guarantee the fault-free operation of solar PV plants.

Siting a solar power plant requires accurate information on the historic solar resource to determine the ideal location and assure that the project will be financially sound. This paper compares the existing and popular method of using a Typical GHI Year (TGY) as a form of historic solar resource assessment, against the method of Typical Power Year.

In this article we show how a state-of-the-art satellite irradiance model - SolarAnywhere - is capable of undetected calibration issues at a trusted reference ground truth irradiance measurement station. This evidence suggests that the best satellite models have now achieved a degree of accuracy and versatility that makes them an acceptable, if not a preferred choice, for solar energy engineering applications ranging from long-term site characterization and system monitoring.

Commercial PV is an appealing way to reduce electricity costs, however due to the variable nature of PV power and the larger power needs of most commercial applications, demand charges induced by variable load and PV generation can offset savings of generating electricity on site. This paper will explore options to reduce these demand charge costs through forecast improvements by downscaling variability and smart persistence using real time site observations. This improved forecast will be used by a load control system which will employ strategies such as load shifting to attempt to reduce demand charges.

In this article, we present a state-of-the-art longtern energy prediction methodology to minimize risk in the project investment. We applied the algorithm to the SolarAnywhere data over 19 years period. he algorithm systematically creates synthetic years from the period of dataset to draw distribution that best describes the data. We conducted the experiment for projects under different climatic conditions. We compared the result with several theoretical distribution models. The results show that the method is so robust that it can predict the real production data at a site with high level of accuracy and provide confidence in the investment.

Four major research objectives were completed over the course of this study. Three of the objectives were to evaluate three, new, state-of-the-art solar irradiance forecasting models. The fourth objective was to improve the California Independent System Operator’s (ISO) load forecasts by integrating behind-the-meter (BTM) PV forecasts. The three, new, state-of-the-art solar irradiance forecasting models included: the infrared (IR) satellite-based cloud motion vector (CMV) model; the WRF-SolarCA model and variants; and the Optimized Deep Machine Learning (ODML)-training model. The first two forecasting models targeted known weaknesses in current operational solar forecasts. They were benchmarked against existing operational numerical weather prediction (NWP) forecasts, visible satellite CMV forecasts, and measured PV plant power production. IR CMV, WRF-SolarCA, and ODML-training forecasting models all improved the forecast to a significant degree. Improvements varied depending on time of day, cloudiness index, and geographic location. The fourth objective was to demonstrate that the California ISO’s load forecasts could be improved by integrating BTM PV forecasts. This objective represented the project’s most exciting and applicable gains. Operational BTM forecasts consisting of 200,000+ individual rooftop PV forecasts were delivered into the California ISO’s real-time automated load forecasting (ALFS) environment. They were then evaluated side-by-side with operational load forecasts with no BTM-treatment. Overall, ALFS-BTM day-ahead (DA) forecasts performed better than baseline ALFS forecasts when compared to actual load data. Specifically, ALFS-BTM DA forecasts were observed to have the largest reduction of error during the afternoon on cloudy days. Shorter term 30 minute-ahead ALFS-BTM forecasts were shown to have less error under all sky conditions, especially during the morning time periods when traditional load forecasts often experience their largest uncertainties. This work culminated in a GO decision being made by the California ISO to include zonal BTM forecasts into its operational load forecasting system. The California ISO’s Manager of Short Term Forecasting, Jim Blatchford, summarized the research performed in this project with the following quote: “The behind-the-meter (BTM) California ISO region forecasting research performed by Clean Power Research and sponsored by the Department of Energy’s SUNRISE program was an opportunity to verify value and demonstrate improved load forecast capability. In 2016, the California ISO will be incorporating the BTM forecast into the Hour Ahead and Day Ahead load models to look for improvements in the overall load forecast accuracy as BTM PV capacity continues to grow.”

This article evaluates the accuracy of solar energy forecasts as a function of geographic footprint ranging from a single point to regions spanning several hundred km. The forecast models that are evaluated include SolarAnywhere®, ECMWF, GFS, HRRR, NDFD and satellite-based cloud motion. The forecast time horizons range from one hour ahead to 2 days ahead. In addition, a new accuracy metric is introduced: this metric quantifies the cost of remedying forecast errors with backup generation if the forecasts overpredict, or with curtailment in case of underprediction.

In this paper we address the use of satellite-based irradiance data as a proxy for the ground irradiance data to determine the intraday solar variability as a function of a finite difference of hourly clear sky index, here after called nominal variability. The satellite data based nominal variability is compared to that of the ground data based variability to determine the use of satellite data as replacement for ground data for this application. A mathematical relationship has been developed to predict nominal variability as a function of the day's clear sky index. The article also demonstrates the application of the intraday variability to predict day ahead hourly forecast variability range as a function of the day's clear sky index. The results show that the intraday solar irradiance variability can be calculated using historical satellite data and provides a similar result to that of variability computed using quality historical ground data. The results also show the potential of intraday solar variability to characterize day ahead forecast variability.

This article presents and validates the latest version of the SUNY satellite model. In the future this new model will be deployed operationally as part of SolarAnywhere. The new version includes an improved treatment of clear sky, the ingestion of now - casting numerical weather predictions, and a more effective treatment of the model's dynamic range to better represent extreme (clear and overcast) conditions. This new version results in substantial performance improvement across a diversity of climates.

We present and evaluate a new operational solar radiation forecast model to be deployed on a prototype basis as part of the SolarAnywhere (SA) data service. The SA service covers North America and provides seamless access to historical, real time and forecasted solar irradiance data with a maximum possible geographical resolution of 1 km and a maximum time resolution of 1 minute for historical and real time data. The new forecast product presented in this article pertains to SA's intermediate spatial and temporal resolution data, respectively 10 km and hourly. The forecast time horizon ranges from one hour ahead to five days ahead.

Validation and operational improvements to the existing SUNY satellite-to-solar irradiance model through incorporation of four of the geostationary satellite infrared (IR) channels are presented herein. The SUNY model is the gridded data set used by NREL in the National Solar Radiation Database (NSRDB) and is available commercially through the Clean Power Research software, SolarAnywhere® Data. This improved model addresses the present model's limitations when representing the irradiance conditions in circumstances of snow cover, high ground reflectance and persistent cloud cover. Improvements in the satellite-to-solar irradiance model have been realized using the IR channels to detect snow conditions and modulate the model background to more accurately reflect irradiance conditions.