Clear-sky Index Research Articles

Effects of temporal averaging on short-term irradiance variability under mixed sky conditions

Abstract. Characterizations of short-term variability in solar radiation are required to successfully integrate large numbers of photovoltaic power systems into the electrical grid. Previous studies have used ground-based irradiance observations with a range of different temporal resolutions and a systematic analysis of the effects of temporal averaging on the representation of variability is lacking. Using high-resolution surface irradiance data with original temporal resolutions between 0.01 and 1 s from six different locations in the Northern Hemisphere, we characterize the changes in representation of temporal variability resulting from time averaging. In this analysis, we condition all data to states of mixed skies, which are the most potentially problematic in terms of local PV power volatility. Statistics of clear-sky index k* and its increments Δk*τ (i.e., normalized surface irradiance and changes therein over specified intervals of time) are considered separately. Our results indicate that a temporal averaging time scale of around 1 s marks a transition in representing single-point irradiance variability, such that longer averages result in substantial underestimates of variability. Higher-resolution data increase the complexity of data management and quality control without appreciably improving the representation of variability. The results do not show any substantial discrepancies between locations or seasons.

A Markov-chain probability distribution mixture approach to the clear-sky index

This paper presents a Markov-chain probability distribution mixture approach to the clear-sky index (CSI). The main assumption is that the temporal variability of the state of clear and the state of cloudy can be described by a two-state Markov-chain, and the variability within each state can be approximated by a probability distribution, unique for each state. Measurables such as the mean clear-sky index, fraction of bright sunshine, expected duration of clearness and expected duration of cloudiness events are shown to be related to the parameters of the method. Additionally, the Ångström equation, which relates mean normalized solar irradiance to the fraction of bright sunshine, is shown to arise as the expectation of the method. In order to numerically verify the method, a simulation model is constructed based on data sets for two different climatic regions: Norrköping, Sweden and Oahu, Hawaii, USA. Results from the simulation model based on training data shows good agreement with testing data, and when comparing the results to existing models in the literature it is comparable to the state of the art. It is shown that the simulation model generates a non-trivial, generally non-zero, autocorrelation function. Finally, challenges with the method and open problems are discussed.

Comparison of intraday probabilistic forecasting of solar irradiance using only endogenous data

Accurate solar forecasts are necessary to improve the integration of solar renewables into the energy grid. In recent years, numerous methods have been developed for predicting the solar irradiance or the output of solar renewables. By definition, a forecast is uncertain. Thus, the models developed predict the mean and the associated uncertainty. Comparisons are therefore necessary and useful for assessing the skill and accuracy of these new methods in the field of solar energy.The aim of this paper is to present a comparison of various models that provide probabilistic forecasts of the solar irradiance within a very strict framework. Indeed, we consider focusing on intraday forecasts, with lead times ranging from 1 to 6 h. The models selected use only endogenous inputs for generating the forecasts. In other words, the only inputs of the models are the past solar irradiance data. In this context, the most common way of generating the forecasts is to combine point forecasting methods with probabilistic approaches in order to provide prediction intervals for the solar irradiance forecasts. For this task, we selected from the literature three point forecasting models (recursive autoregressive and moving average (ARMA), coupled autoregressive and dynamical system (CARDS), and neural network (NN)), and seven methods for assessing the distribution of their error (linear model in quantile regression (LMQR), weighted quantile regression (WQR), quantile regression neural network (QRNN), recursive generalized autoregressive conditional heteroskedasticity (GARCHrls), sieve bootstrap (SB), quantile regression forest (QRF), and gradient boosting decision trees (GBDT)), leading to a comparison of 20 combinations of models.None of the model combinations clearly outperform the others; nevertheless, some trends emerge from the comparison. First, the use of the clear sky index ensures the accuracy of the forecasts. This derived parameter permits time series to be deseasonalized with missing data, and is also a good explanatory variable of the distribution of the forecasting errors. Second, regardless of the point forecasting method used, linear models in quantile regression, weighted quantile regression and gradient boosting decision trees are able to forecast the prediction intervals accurately.

Using Shortwave Radiation to Evaluate the HARMONIE-AROME Weather Model

Evaluation of global shortwave irradiance forecasts from the HARMONIE-AROME weather prediction model is presented in this paper. We give examples of how such an evaluation can be used when testing a weather model or reanalysis product. We specifically use the non-dimensional clear sky and variability indices. We have tested seven months of HARMONIE-AROME 40h1.1 output against Danish global irradiance stations and 35 years of the Irish Met Éireann reanalysis (MÉRA) simulations. MÉRA, which is run with HARMONIE-AROME 38h1.2, is shown to have a significantly lower bias than the previously available global horizontal irradiance (GHI) reanalysis data from the ERA-Interim dataset. The Danish HARMONIE-AROME 40h1.1 has a negative bias during the summer months that is not seen in the Irish HARMONIE-AROME 38h1.2. For both model runs, we find a negative bias in the shortwave irradiance forecasts on days with thick clouds. This suggest that the model has too much cloud water in thick clouds.

A Method for Robust Estimation of Vegetation Seasonality from Landsat and Sentinel-2 Time Series Data

Time series from Landsat and Sentinel-2 satellites have great potential for modeling vegetation seasonality. However, irregular time sampling and frequent data loss due to clouds, snow, and short growing seasons, makes this modeling a challenge. We describe a new method for modeling seasonal vegetation index dynamics from satellite time series data. The method is based on box constrained separable least squares fits to logistic model functions combined with seasonal shape priors. To enable robust estimates, we extract a base level (i.e., the minimum dormant season value) from the frequency distribution of clear-sky vegetation index values. A seasonal shape prior is computed from several years of data, and in the final fits local parameters are box constrained. More specifically, if enough data values exist in a certain time period, the corresponding local parameters determining the shape of the model function over this period are relaxed and allowed to vary freely. If there are no observations in a period, the corresponding local parameters are locked to the parameters of the shape prior. The method is flexible enough to model interannual variations, yet robust enough when data are sparse. We test the method with Landsat, Sentinel-2, and MODIS data over a forested site in Sweden, demonstrating the feasibility and potential of the method for operational modeling of growing seasons.

Towards the intrahour forecasting of direct normal irradiance using sky-imaging data

Increasing power plant efficiency through improved operation is key in the development of Concentrating Solar Power (CSP) technologies. To this end, one of the most challenging topics remains accurately forecasting the solar resource at a short-term horizon. Indeed, in CSP plants, production is directly impacted by both the availability and variability of the solar resource and, more specifically, by Direct Normal Irradiance (DNI). The present paper deals with a new approach to the intrahour forecasting (the forecast horizon Δtf is up to 30min ahead) of DNI, taking advantage of the fact that this quantity can be split into two terms, i.e. clear-sky DNI and the clear sky index. Clear-sky DNI is forecasted from DNI measurements, using an empirical model (Ineichen and Perez, 2002) combined with a persistence of atmospheric turbidity. Moreover, in the framework of the CSPIMP (Concentrating Solar Power plant efficiency IMProvement) research project, PROMES-CNRS has developed a sky imager able to provide High Dynamic Range (HDR) images. So, regarding the clear-sky index, it is forecasted from sky-imaging data, using an Adaptive Network-based Fuzzy Inference System (ANFIS). A hybrid algorithm that takes inspiration from the classification algorithm proposed by Ghonima et al. (2012) when clear-sky anisotropy is known and from the hybrid thresholding algorithm proposed by Li et al. (2011) in the opposite case has been developed to the detection of clouds. Performance is evaluated via a comparative study in which persistence models – either a persistence of DNI or a persistence of the clear-sky index – are included. Preliminary results highlight that the proposed approach has the potential to outperform these models (both persistence models achieve similar performance) in terms of forecasting accuracy: over the test data used, RMSE (the Root Mean Square Error) is reduced of about 20Wm−2, with Δtf=15min, and 40Wm−2, with Δtf=30min.

Stochastic Model for Generation of High-Resolution Irradiance Data and Estimation of Power Output of Photovoltaic Plants

This paper presents a method for the generation of synthetic time series of solar radiation with a one-minute time resolution using an automatic cloud classification procedure. Solar radiation data from ground-based measurements are used to determine the correlation between the irradiance and the prevailing cloud classes extracted from satellite images. A specific set of Markov chains for each cloud class is adjusted empirically and used for a stochastic simulation of clear sky index time series. The different models are then combined to convert the irradiance into power output time series of photovoltaic power plants of different sizes. Tests to determine the methodology's performance showed positive results in reproducing the statistical characteristics of observed time series data. Estimates of a selected set of metrics were obtained for several sites in Brazil and allowed the characterization of the solar resource for photovoltaic plants of various sizes and mounting methods considering the generation potential and short-term variability.

Conditional summertime day-ahead solar irradiance forecast

We investigated the accuracy of numerical weather prediction (NWP)-based global horizontal irradiance (GHI) and clear-sky index forecasting over southern Nevada. Accurate forecasts of solar irradiance are required for electric utilities to economically integrate substantial amounts of solar power into their power generation portfolios. Solar irradiance forecasting can enhance the value of renewable energy by anticipating fluctuations in these variable resources. Summertime cloud variability depends largely on the combination of tropical and extratropical synoptic-scale forcing, most of which is observable, predictable, and highly related to the North American Monsoon moisture surge events. We used high-resolution realtime NWP output based on the weather and research forecasting (WRF) model to study the ability of the model to deliver day-ahead GHI and clear-sky index forecasts for a the National Renewable Energy Laboratory (NREL)-University of Nevada site, located in Las Vegas, Nevada. High-resolution forecast products were obtained from the Desert Research Institute (DRI) archived real-time numerical weather forecasting products. Results showed the importance of developing a site-specific seasonal and weather-dependent model output statistics (MOS) approach to improving forecast accuracy, which removes the bias and reduces the overall relative root-mean – square error (rRMSE) of GHI by as much as 6%, when compared to the uncorrected model output; improving forecast accuracy is obtained by adding information that relates regional-scale circulation patterns driving cloudiness, hence irradiance variability to the target area. We show the seasonal dependence of the NWP forecast accuracy and demonstrate that intelligent weather functions provide a pathway to improve accuracy of solar forecasts further.

Assimilating observations to simulate marine layer stratocumulus for solar forecasting

Integration of solar energy forecasts into the electric network is becoming essential because of the continually increasing penetration level of solar energy. Three-dimensional numerical weather prediction (NWP) models predict the weather based on the current weather conditions (called initialization) and simulate the ensuing atmospheric processes. The accuracy of forecasts therefore depends, in part, on the accuracy of the model initializations. Data assimilation is recognized as the most widely used technique to improve the initialization into NWP models. In this study, meteorological observations from the surface and upper-air in-situ networks over the southern California coast are assimilated into the advanced research version of the Weather Research and Forecasting (WRF) model using a three dimensional variational data assimilation technique (3DVAR). A single observation test was conducted to tune-up the length scale and variance scale along with the regional domain-dependent background error statistics. A customized version of 3DVAR data assimilation was deployed with two sets of cyclic data assimilation with 6-h and 1-h assimilation windows along with the cold-start mode. The cyclic data assimilation experiments consistently outperformed the cold-start data assimilation and WRF for intra-day Global Horizontal Irradiance (GHI) and Clear Sky Index (CSI) forecast. Hourly cyclic assimilation showed the highest forecast skill score against ground measurements and satellite measurements. Even at the coastal stations with more challenging meteorological conditions, the hourly cyclic assimilation consistently outperformed the 24-h persistence forecast. The average (mean of four case studies) hourly cyclic data assimilation showed the highest forecast skill score in GHI and CSI intra-day forecast with reference to 24-h persistence forecast up to 39.4% and 40.7% respectively at the coastal stations. The spatial distributions of GHI biases estimated against SolarAnywhere satellite measurements showed that the hourly cyclic assimilation consistently improved the stratocumulus cloud coverage, thickness, and life time over the coastal region, but biases are still present further inland.

A tuning routine to correct systematic influences in reference PV systems’ power outputs

Power output measurements from PV systems are subject to a wide variety of systematic external and internal influences, such as shading, soiling, degradation, module and inverter quality issues and other system-level losses. All of these influences upon PV power measurements make the use of PV power output datasets for higher-level analysis problematic, particularly in their use as reference PV systems for estimating the power of a regional portfolio. To address these issues, we present a three-step method. Firstly, a parameterisation and quality control of power measurements is performed, which also corrects for consistent inefficiencies by a loss factor LF. Secondly, the detection of systematic de-ratings affecting PV system power output differently for each time step of the day (predominantly due to shading) together with the implementation of a subsequent “re-rating” of the power output measurements in a process referred to as tuning. The pivotal element of this approach is a 30-day running 90th percentile of the clear-sky index for photovoltaics kpv and the computation of a daily de-rating profile. Lastly, high kpv related variance in the early morning and evening is detected and filtered. Whilst these three methods are independent of each other, we suggest applying them in combination following the same order as in our paper. Cross-validations of these methods demonstrate significant improvements to the PV power measurement profiles, particularly in their use as reference PV systems for upscaling approaches. The RMSE falls from 0.174 to 0.09W/Wp,rRMSE from 46.5% to 21.9%,MAPE from 47.9% to 20.8% and the correlation r increases from 0.767 to 0.919. Hence, we report overall improvements to RMSE,rRMSE,MAPE and r by 48%,53%,57% and 20%, respectively.

Short-term peer-to-peer solar forecasting in a network of photovoltaic systems

Solar forecasting is a necessary component of economical realization of high penetration levels of photovoltaic (PV) systems. This paper presents a short term, intra-hour solar forecasting method. This “peer-to-peer” (P2P) forecasting method is based on the cross-correlation time lag between clear-sky index time series of pairs of PV-systems that are influenced by the (assumed) same cloud sequentially, with the feature that the forecast horizon (FH) can be set at a fixed value. The P2P forecasting algorithm was evaluated for 11 central PV-systems (out of 202) over a half year period from the 1st of March through the 31st of August 2015 using the forecast skill (FS) metric. Positive FS means improvement over reference clear-sky index persistence forecasting. The P2P forecasting method was evaluated over a subset of days with either high, all or low irradiance variability. The average forecast skill (avgFS) concerning forecast horizons between 5 and 8min was 5.99%, −1.61% and −16.0% over these periods respectively, indicating the superior performance of the P2P method over persistence during the highly variable days, which are most interesting from the perspective of electricity grid management.

An autocorrelation-based copula model for generating realistic clear-sky index time-series

This study presents a method for using copulas to model the temporal variability of the clear-sky index, which in turn can be used to produce realistic time-series of photovoltaic power generation. The method utilizes the autocorrelation function of a clear-sky index time-series, and based on that a correlation matrix is set up for the dependency between clear-sky indices at N time-steps. With the use of this correlation matrix an N-dimensional copula function is configured so that correlated samples for these N time-steps can be obtained. Results from the copula model are compared with the original data set and an uncorrelated model based on zero correlated clear-sky index data in terms of distribution, autocorrelation, step changes and distribution. The copula model is shown to be superior to the uncorrelated model in these aspects. As a validation the model is tested with solar irradiance for two different geographical regions: Norrköping, Sweden and Hawaii, USA. The copula model is also applied to a set of bins of daily mean clear-sky index and the use of bins is shown to improve the results.

On the properties of aggregate clear-sky index distributions and an improved model for spatially correlated instantaneous solar irradiance

An important factor in grid integration of solar power is the so-called dispersion-smoothing effect, i.e., that differences in cloudiness over dispersed systems make the aggregate output less variable. This effect has been studied for irradiance step-changes on different time horizons, but not so much for instantaneous irradiance. In this paper, an improved probabilistic model is proposed for how instantaneous solar irradiance is correlated and aggregated over a network with arbitrary number of sites and dispersion. The model is fitted to irradiance data with a 1-s resolution from a network with 17 pyranometers in Hawai’i. A previously proposed three-state model of the instantaneous clear-sky index is partly confirmed by showing that clear and cloudy states can be separated and modeled by independent distribution models. It is also shown that the station-pair correlations for the instantaneous clear-sky index, as well as the shape of the distribution for the cloudy states, depend characteristically on the average degree of cloudiness, represented by the daily clear-sky index. For dispersed sites within the studied network, separated by distances up to 1km, and for daily clear-sky indices above approximately 0.5, the model performs better in reproducing the aggregate clear-sky index than non-spatial data. The proposed model could assist distribution system operators (DSOs) in grid planning and operation, as shown in a case study.

Analytic model for correlations of cloud induced fluctuations of clear-sky index

Abstract We have derived an analytic model to describe the correlation of cloud-induced fluctuations on photovoltaic (PV) system power generation versus distance or time lag in terms of the parameters that describe the clouds (duration, opacity, ramp smoothness and lag). Clouds are modeled as clear-sky index deviations with sigmoidal ramps. The found auto-correlation function is suited to deterministically describe the correlation versus distance or time lag over a short period of recent time (up to several hours depending on the amount of clouds and the temporal resolution of the data) to a very high level of detail, as opposed to other methods that use a statistical approach. The sensor-pair cross-correlation of a cloud with variable width (and/or specified to a more complex or two-dimensional localized shape) can also be fully understood. Application of the model is investigated in terms of relative aggregate variability of an ensemble of (virtual) PV-systems for one, two or three subsequent clouds. This analysis explicitly relates the shape of clouds to the dimensions of a dense PV-network which is beneficial for understanding variability of PV-systems in a large plant or dense urban region.

Simulating clear-sky index increment correlations under mixed sky conditions using a fractal cloud model

Modelling clear-sky index increments (i.e. changes in normalized surface irradiance over specified intervals of time) and their spatial autocorrelation structures is important for the reliable grid integration of photovoltaic power systems. In order to capture increment correlation structures under mixed sky conditions, we apply a fractal cloud model. First, we use thousands of fish-eye sky camera and satellite images to estimate cloud edge fractal dimension D on scales between tens of meters and hundreds of kilometers. Box-counting analyses of cloud edges extracted from these images result in best-fit estimates of fractal dimension between 1.4 and 1.6, with satellite approximations being consistently higher than sky camera estimates. Contrary to previous studies, we find no evidence of any scale break and attribute the systematic discrepancy between the two estimates of D to intrinsic differences in the instrument-specific techniques of cloud detection. Next, we synthesize small-scale fractal cloud fields, with D=1.5 and satellite-derived cloud images as input. Using cloud motion vectors from the same satellite images, we then translate the small-scale cloud fields across a set of model pyranometer locations to obtain corresponding clear-sky index time series. From the simulated time series we calculate increment correlation structures for distances between tens of meters and about ten kilometers, and compare them to observation-based results from 1Hz measurements of up to 99 pyranometers. The observed isotropic, along-wind, and across-wind spatial autocorrelation structures of clear-sky index increments are captured well by the model, both in terms of overall value and shape. While there are some systematic differences between modelled and observed structures (e.g. underevaluation for very small time scales of a few seconds), the differences are essentially small. In general, the simulated correlations are not strongly sensitive to variations in the fractal model parameters.

Improving time series prediction of solar irradiance after sunrise: Comparison among three methods for time series prediction

Short-term prediction for renewable energy outputs up to 6h is important especially for preparing backup power plants such as thermal power plants and hydro power plants to keep the voltage and the frequency in a power grid constant. However, short-term prediction of solar irradiance for the morning is especially difficult because measurements of solar irradiance before sunrise are zeros and useless for the prediction. Here we propose to use recently derived infinite-dimensional delay coordinates for predicting solar irradiance after sunrise based on a time series before the sunrise. As a result, one can make time series prediction by taking into account the long history of previous changes of solar irradiance. We demonstrate that the examined short-term time series prediction has effectively predictive skills because its prediction errors are smaller by about 85% best compared with the 24h clear sky index persistence.

Generating synthetic five-minute solar irradiance values from hourly observations

We have developed a method for generating synthetic five-minute pairs of global horizontal irradiance (GHI) and direct normal irradiance (DNI), interpolated from hourly mean values.Our method develops a model based on five-minute pairs of GHI and DNI from a location where real five-minute data is available. Five-minute GHI and DNI clear sky index values are calculated for each five-minute time period, and hourly values are calculated for each hourly period. The hourly values identify a clear sky category into which the twelve five-minute GHI and DNI clear sky index values are placed. The result is clear sky categories each containing five-minute clear sky index values that capture how five-minute GHI and DNI fluctuates for each clear sky category.The process for generating synthetically interpolated five-minute pairs of GHI and DNI from hourly mean values is straightforward. Hourly GHI and DNI clear sky index values are used to identify the clear sky category from which twelve five-minute synthetic pairs of GHI and DNI clear sky index values are bootstrapped. The index values are multiplied by the modelled perfect clear sky GHI and DNI to give synthetically interpolated five-minute pairs of GHI and DNI.Our method performs well. The hourly total Normalised Root Mean Square Deviation (NRMSD) results for our method are 1.4–1.8% for GHI and 2.2–3.0% for DNI. The hourly total Mean Bias Deviation (MBD) results for our method are −0.6% to 0.2% for GHI and −1.5% to −0.2% for DNI. The R2 values between observed and synthetic hourly totals are all greater than 0.99. The monthly Kolmogorov-Smirnov Integral (KSI) values were good for 97% of all months for DNI and for all months for GHI.Our method can be applied to locations where only hourly data is available. For example, it could be applied to the Australian Bureau of Meteorology’s hourly solar irradiance gridded satellite data, to capture the spatial variation in solar irradiance for multiple locations. The result would be particularly useful, from a solar energy perspective, for designing a reliable renewable energy electricity system where its performance could be evaluated for any location in Australia.

A synthetic, spatially decorrelating solar irradiance generator and application to a LV grid model with high PV penetration

Residential photovoltaic (PV) technology is expected to have mass global deployment. With widespread PV in the electricity distribution grids, the variable nature of the solar resource must be understood to facilitate reliable operation. This research demonstrates that synthetic, 1-min resolution irradiance time series that vary on a spatial dimension can be generated based on the following inputs: mean hourly meteorological observations of okta, wind speed, cloud height and atmospheric pressure.The synthetic time series temporally validate against observed 1-min irradiance data for four locations—Cambourne, UK; Lerwick, UK; San Diego, CA USA; and Oahu, HI USA—when analysing 4 metrics of variability indices, ramp-rate size, irradiance magnitude frequency and clear-sky index frequency. Each metric is calculated for the modelled and observed data at each location and CDF profile correlation compared as well as applying the Kolmogorov-Smirnov (K–S) test with 99% confidence limits. CDF correlation coefficients of each metric are all above R⩾0.908, and a minimum of 90.96% of daily irradiance time series passed the K–S test. A spatial validation was performed comparing the model outputs to real observation data. The spatial correlation coefficient regression with site separation was successfully recreated with MAPE=0.865%, RMSE=0.01 and R=0.955. The spatial instantaneous correlation was shown to behave anisotropically when using fixed cloud direction, with different correlation in along and cross wind directions. Cloud cover states of 40–60% showed the most spatial decorrelation while 0% and 100% had the least.The model outputs are applied to a distribution grid impact model using the IEEE-8500 node test feeder. PV scenarios of 25%,50%, and 75% uptake were modelled across a 1.5×1.5km grid. The magnitude and frequency of severe tap changing events are found to be significantly higher when using a single irradiance time series for all PV systems versus individually assigning spatially decorrelating time series.

Cloud cover effect of clear-sky index distributions and differences between human and automatic cloud observations

The statistics of clear-sky index can be used to determine solar irradiance when the theoretical clear sky irradiance and the cloud cover are known. In this paper, observations of hourly clear-sky index for the years of 2010–2013 at 63 locations in the UK are analysed for over 1 million data hours. The aggregated distribution of clear-sky index is bimodal, with strong contributions from mostly-cloudy and mostly-clear hours, as well as a lower number of intermediate hours. The clear-sky index exhibits a distribution of values for each cloud cover bin, measured in eighths of the sky covered (oktas), and also depends on solar elevation angle. Cloud cover is measured either by a human observer or automatically with a cloud ceilometer. Irradiation (time-integrated irradiance) values corresponding to human observations of “cloudless” skies (0 oktas) tend to agree better with theoretical clear-sky values, which are calculated with a radiative transfer model, than irradiation values corresponding to automated observations of 0 oktas. It is apparent that the cloud ceilometers incorrectly categorise more non-cloudless hours as cloudless than human observers do. This leads to notable differences in the distributions of clear-sky index for each okta class, and between human and automated observations. Two probability density functions—the Burr (type III) for mostly-clear situations, and generalised gamma for mostly-cloudy situations—are suggested as analytical fits for each cloud coverage, observation type, and solar elevation angle bin. For human observations of overcast skies (8 oktas) where solar elevation angle exceeds 10°, there is no significant difference between the observed clear-sky indices and the generalised gamma distribution fits.

QCPV: A quality control algorithm for distributed photovoltaic array power output

Distributed PV power output measurements and their metadata are subject to significant errors and uncertainty. However, no standard mechanism for quality controlling these data is currently available within the literature. For this purpose, we present a two part approach to parameterize PV system metadata and quality control their power output measurements. These methods are based on open-source solver routines and require only one exogenous input (ambient temperature), and therefore are widely applicable. The method allows for PV array orientation to be determined to within ∼4°. Furthermore a loss factor LF is derived which captures overall PV system losses and correctly detects a mean degradation of the modules of ∼0.5% per annum over time. The central routine, entitled QCPV, imposes system specific limits on the measured data mainly through the use of the extraterrestrial irradiance and the clear sky index for photovoltaics kpv. It also enables, amongst other things, the detection of cloud enhancement events and spurious power output reporting through the application of across systems statistics.