Peter Gotseff

An Extensive Comparison of Commercial Pyrheliometers under a Wide Range of Routine Observing Conditions

In the most comprehensive pyrheliometer comparison known to date, 33 instruments were deployed to measure direct normal solar radiation over a 10-month period in Golden, Colorado. The goal was to determine their performance relative to four electrical-substitution cavity radiometers that were calibrated against the World Radiometric Reference (WRR) that is maintained at the World Radiation Center in Davos, Switzerland. Because of intermittentcabling problems with one of the cavity radiometers, the average of three windowed, electrical-substitution cavity radiometers served as the reference irradiance for 29 test instruments during the 10-month study. To keep the size of this work manageable, comparisons are limited to stable sunny conditions, passing clouds, calm and windy conditions, and hot and cold temperatures. Other variables could have been analyzed, or the conditions analyzed could have employed higher resolution. A more complete study should be possible now that the instruments are identified; note that this analysis was performed without any knowledge on the part of the analyst of the instruments’ manufacturers or models. Apart from the windowed cavities that provided the best measurements, two categories of performance emerged during the comparison. All instruments exceeded expectations in that they measured with lower uncertainties than the manufacturers’ own specifications. Operational 95% uncertainties for the three classes of instruments, which include the uncertainties of the open cavities used for calibration, were about 0.5%, 0.8%,and 1.4%.The open cavitiesthat wereused for calibrationof allpyrheliometers havean estimated 95% uncertainty of 0.4%‐0.45%, which includes the conservative estimate of 0.3% uncertainty for the WRR.

Accurate power prediction of spatially distributed PV systems using localized irradiance measurements

In this paper, a method for estimating power injected into an electrical distribution system from spatially distributed residential PV systems using data from ground-based weather stations is described. This method was developed as part of the High Penetration PV Deployment Project at the Arizona Public Service (APS). Verification of this predictive method is also described in this paper. Finally, correlation statistics and power production ramp rates are calculated for irradiance based on measured power within the APS study area using two different sets of weather station measurements. The method has potential applications to real-time or forecasted power estimation on distribution circuits in addition to the development of broader guidelines on the optimal number and location of irradiance sensors for power estimation in distribution circuits with high penetrations of PV.

Evaluation of Clear Sky Models for Satellite-Based Irradiance Estimates

This report describes an intercomparison of three popular broadband clear sky solar irradiance model results with measured data, as well as satellite-based model clear sky results compared to measured clear sky data. The authors conclude that one of the popular clear sky models (the Bird clear sky model developed by Richard Bird and Roland Hulstrom) could serve as a more accurate replacement for current satellite-model clear sky estimations. Additionally, the analysis of the model results with respect to model input parameters indicates that rather than climatological, annual, or monthly mean input data, higher-time-resolution input parameters improve the general clear sky model performance.

Technologies to increase PV hosting capacity in distribution feeders

This paper studies the distributed photovoltaic (PV) hosting capacity in distribution feeders by using the stochastic analysis approach. Multiple scenario simulations are conducted to analyze several factors that affect PV hosting capacity, including the existence of voltage regulator, PV location, the power factor of PV inverter and Volt/VAR control. Based on the conclusions obtained from simulation results, three approaches are then proposed to increase distributed PV hosting capacity, which can be formulated as the optimization problem to obtain the optimal solution. All technologies investigated in this paper utilize only existing assets in the feeder and therefore are implementable for a low cost. Additionally, the tool developed for these studies is described.

Voltage support study of smart PV inverters on a high-photovoltaic penetration utility distribution feeder

This paper reports on tools and methodologies developed to study the impact of adding rooftop photovoltaic (PV) systems, with and without the ability to provide reactive power, on the primary voltage profile of a distribution feeder. Simulation results are provided from a study of a SDG&E utility distribution feeder. The simulation model of the feeder was built in OpenDSS and verified by comparing the simulated voltages to field measurements. First, all PV inverters were set to operate at unity power factor, and the impact on feeder voltages was analyzed. Then simulations were conducted with voltage support activated for all the smart PV inverters. These included different constant power factor settings and Volt/VAr controls. Modeling results quantify and illustrate the ability of these smart inverters to provide reactive power that impacts the primary voltage.

Alternatives to the 15% Rule

The third solicitation of the California Solar Initiative (CSI) Research, Development, Demonstration and Deployment (RD&D) Program established by the California Public Utility Commission (CPUC) is supporting the Electric Power Research Institute (EPRI), National Renewable Energy Laboratory (NREL), and Sandia National Laboratories (SNL) with collaboration from Pacific Gas and Electric (PG&E), Southern California Edison (SCE), and San

Quantifying the impact of incidence-angle dependence on solar radiometric calibration

Evaluating photovoltaic cells, modules, arrays, and system performance relies on accurate measurements of the solar radiation resources available for power conversion. Measuring solar resources accurately can lead to a reduction in the investment risks associated with installing and operating solar energy systems. The National Renewable Energy Laboratorys Solar Radiation Research Laboratory collects and disseminates solar irradiance data and provides calibrations of broadband radiometers that are traceable to the international standards. It is essential that radiometric data are traceable to the international system of units, e.g., through the World Radiometer Reference and World Infrared Standard Group. This paper demonstrates the importance and application of an existing approach that ultimately reduces the uncertainty of radiometric measurements. Almost all commercially available broadband radiometers use a single responsivity value that is generated at a 45° solar zenith angle (incident angle) based on outdoor calibrations or transfers between radiometers inside integrating spheres or that responsivity is generated using normal incident radiation based on indoor calibrations using lamps and comparisons to reference radiometers to compute measured irradiance data. However, based on our experience and that of other experts in the radiometric science community, this method introduces increased uncertainty to the data. If a single responsivity value is used, the radiometer will overestimate or underestimate the irradiance data compared to the reference irradiance. This was demonstrated in Myers [1], Reda [2], and Reda et al. [3]. Further, by using responsivity as a function of solar zenith angle, the uncertainty for some instruments in the responsivity value can be reduced by as much as 50% compared to using a single responsivity calculated at 45° [2, 3].

Performance testing using silicon devices - Analysis of accuracy

Accurately determining PV module performance in the field requires accurate measurements of solar irradiance reaching the PV panel (i.e., Plane-of-Array - POA Irradiance) with known measurement uncertainty. Pyranometers are commonly based on thermopile or silicon photodiode detectors. Silicon detectors, including PV reference cells, are an attractive choice for reasons that include faster time response (10 microseconds (μs)) than thermopile detectors (1 s to 5 s) and lower cost and maintenance. The main drawback of silicon detectors is their limited spectral response. Therefore, to determine broadband POA solar irradiance, a pyranometer calibration factor that converts the narrowband response to broadband is required. Normally, this calibration factor is a single number determined under clear-sky conditions with respect to a broadband reference radiometer. The pyranometer is then used for various scenarios including varying airmass, panel orientation, and atmospheric conditions. This would not be an issue if all irradiance wavelengths that form the broadband spectrum responded uniformly to atmospheric constituents. Unfortunately, the scattering and absorption signature varies widely with wavelength and the calibration factor for the silicon photodiode pyranometer is not appropriate for other conditions. This paper reviews the issues that will arise from the use of silicon detectors for PV performance measurement in the field based on measurements from a group of pyranometers mounted on a 1-axis solar tracker. We also present a comparison of simultaneous spectral and broadband measurements from silicon and thermopile detectors and estimated measurement errors when using silicon devices for both array performance and resource assessment.

Representative day selection using statistical bootstrapping for accelerating annual distribution simulations

Capturing technical and economic impacts of solar photovoltaics (PV) and other distributed energy resources (DERs) on electric distribution systems can require high-time resolution (e.g. 1 minute), long-duration (e.g. 1 year) simulations. However, such simulations can be computationally prohibitive, particularly when including complex control schemes in quasi-steady-state time series (QSTS) simulation. Various approaches have been used in the literature to down select representative time segments (e.g. days), but typically these are best suited for lower time resolutions or consider only a single data stream (e.g. PV production) for selection. We present a statistical approach that combines stratified sampling and bootstrapping to select representative days while also providing a simple method to reassemble annual results. We describe the approach in the context of a recent study with a utility partner. This approach enables much faster QSTS analysis by simulating only a subset of days, while maintaining accurate annual estimates.

A high-speed, real-time visualization and state estimation platform for monitoring and control of electric distribution systems: Implementation and field results

Continued deployment of renewable and distributed energy resources is fundamentally changing the way that electric distribution systems are controlled and operated; more sophisticated active system control and greater situational awareness are needed. Real-time measurements and distribution system state estimation (DSSE) techniques enable more sophisticated system control and, when combined with visualization applications, greater situational awareness. This paper presents a novel demonstration of a high-speed, real-time DSSE platform and related control and visualization functionalities, implemented using existing open-source software and distribution system monitoring hardware. Live scrolling strip charts of meter data and intuitive annotated map visualizations of the entire state (obtained via DSSE) of a real-world distribution circuit are shown. The DSSE implementation is validated to demonstrate provision of accurate voltage data. This platform allows for enhanced control and situational awareness using only a minimum quantity of distribution system measurement units and modest data and software infrastructure.