Direct Normal Irradiance Measurements Research Articles

Intrahour Direct Normal Irradiance Forecasting Based on Sky Image Processing and Time-Series Analysis

The present paper exhibits a hybrid model for intrahour forecasting of direct normal irradiance (DNI). It combines a knowledge-based model, which is used for clear-sky DNI forecasting from DNI measurements, with a machine-learning-based model, that evaluates the impact of atmospheric disturbances on the solar resource, through the processing of high dynamic range sky images provided by a ground-based camera. The performance of the hybrid model is compared with that of two machine learning models based on past DNI observations only. The results highlight the pertinence of combining knowledge-based models with data-driven models, and of integrating sky-imaging data in the DNI forecasting process. Parts of this paper were published as journal articleKarout, Y.; Thil, S.; Eynard, J.; Guillot, E.; Grieu, S. Hybrid intrahour DNI forecast model based on DNI measurements and sky-imaging data. Solar Energy. 2023, 249, 541-558. https://doi.org/10.1016/j.solener.2022.11.032

Hybrid intrahour DNI forecast model based on DNI measurements and sky-imaging data

Short-term forecasting of direct normal irradiance (DNI), which refers to photons that did not interact with the atmosphere on their way to the observer, is of great interest in the advanced management of concentrated solar power systems. So, this paper exhibits a hybrid model dedicated to the intrahour forecasting of DNI (horizons are 5, 10, and 15 min). This hybrid model combines a knowledge-based model with a machine learning model: the knowledge-based model is used for clear-sky DNI forecasting from DNI measurements; the machine learning model evaluates the impact of the atmospheric disturbances on the solar resource, through the processing of high dynamic range sky images provided by a ground-based camera. The hybrid model is evaluated by comparing its performance, for different sky conditions, with that of two machine learning models based on past DNI observations only. The results highlight the pertinence of combining knowledge-based models with data-driven models, and of integrating sky-imaging data in the DNI forecasting process: the proposed hybrid model successfully manages clear-sky, overcast, and mixed situations. Regardless of the forecast horizon, the hybrid model outperforms the reference models both in normalized root mean squared error and in DNI ramp detection.

Development of a Simple Methodology Using Meteorological Data to Evaluate Concentrating Solar Power Production Capacity

Evaluation of the Concentrating Solar Power capacity factor is critical to support decision making on possible regional energy investments. For such evaluations, the System Advisor Model is used to perform capacity factor assessments. Among the required data, information concerning direct normal irradiance is fundamental. In this context, the Engerer model is used to estimate direct normal irradiance hourly values out of global horizontal irradiance ground measurements and other observed meteorological variables. Model parameters were calibrated for direct normal irradiance measurements in Évora (Southern Portugal), being then applied to a network of 90 stations, part of the Portuguese Meteorological Service. From the modelled direct normal irradiance, and for stations that comprise 20 years of data, typical meteorological years were determined. Finally, to identify locations of interest for possible installations of Concentrating Solar Power systems, annual direct normal irradiance availabilities and the respective capacity factor, for a predefined power plant using the System Advisor Model, were produced. Results show annual direct normal irradiance availabilities and capacity factors of up to ~2310 kWh/m2 and ~36.2% in Castro Marim and in Faro, respectively. Moreover, this study supports energy policies that would promote Concentrating Solar Power investments in Southern Portugal (Alentejo and Algarve regions) and eastern centre Portugal (Beira Interior region), which have capacity factors above 30%.

Transient simulation of a control strategy for solar receivers based on mass flow valves adjustments and heliostats aiming

Solar central receivers have inherently large time constants and time delays caused by their thermal inertia are due to the transport of the heat transfer fluid alongside the panels. This problem is even accentuated if these systems are continuously exposed to direct normal irradiance (DNI) fluctuations. Ensuring an adequate operation of this kind of process implies the adaptation of control strategies. This paper proposes a strategy that couples the temperature control of the outlet fluid while considering the heliostats' aiming points location to maintain the receiver operating below its design limit and simulates its performance under several scenarios. These tests are carried out on a non-steady-state solar plant model that considers its optical and thermal characteristics. The structure of the proposed control system is based on using intermediate flow valves along the heating flow path to reduce time delay. Hence, it renders the control system more responsive to either a DNI variation affecting the solar field or moving its operation from one state to another. The control strategy structure contains feedback and feedforward loops to determine the near-future process response to DNI measurements. It allows adding more information to the loop used to maintain the system's desired steady-state operation. The overall strategy can take the process to a steady-state operation regardless of the hour or day of operation. It produces a low variation of the outlet temperature, with a standard deviation of 3 °C, while the receiver is under a transient DNI and maintains the heliostats aimed toward the receiver without exceeding its heat flux limits. Results indicate that the proposed strategy achieves over 80% improvement based on the integral absolute error (IAE) performance index compared to conventional single-loop temperature control. Besides the simulated numerical results, this research also presents the advantages of considering intermediate mass-flow valves for control purposes of the receiver. This work should be considered a step towards improving multivariable control algorithms for the safe operation of solar central receivers.

Advances in aerosol optical depth evaluation from broadband direct normal irradiance measurements

Aerosols are part of the attenuation processes that impact solar radiation within the atmosphere. They influence the availability and spectrum of the solar resource for each location at the earth’s surface. The present study presents advances in the development of a methodology intended to estimate the aerosol optical depth (AOD) at a given location from broadband direct normal irradiance (DNI) measurements and an appropriate radiative transfer model (RTM) operated backwards. For this purpose, databases provided by AERONET and BSRN at 16 stations throughout the world are jointly employed as inputs to the proposed methodology. The validation of two RTMs (SMARTS and SOLIS) is first undertaken to estimate DNI under clear-sky conditions at each station, assuming both AOD and additional atmospheric inputs are known from sunphotometric measurements. Results indicate that both models achieve good performance, characterized by a relative rRMSE of 3.2% for SMARTS and 3.8% for SOLIS. In the second, and most important stage, the AOD at 550 nm (AOD550) is derived using these models again, but in an iterative mode, now using the 1-minute DNI measurements as inputs. Periods of clear line of sight to the sun first need to be selected from the irradiance measurement record. This, along with other difficulties, make this operation prone to errors when only DNI measurements are available. In spite of this, the results show that AOD can be estimated with a 16-site average mean bias error of only between −0.024 and 0.015 AOD unit and an absolute RMSE between 0.025 and 0.050 AOD unit (compared to the AERONET ground truth), depending on model. Notable improvements are obtained if secondary atmospheric variables are extracted from the MERRA-2 reanalysis and are included as inputs for local computations. The present results suggest that the method is able to compare favorably with AOD estimates from MERRA-2 predictions or MODIS observations, for instance.

Soiling effect in second-surface CSP mirror and improved cleaning strategies

Abstract Deployment of future CSP plants is foreseen to take place in southern Portugal, a region with great potential for solar energy harvesting. For the operation and maintenance of such power plants, it is important to know in advance the potential influence that soiling can have in second-surface mirrors, particularly in the specular reflection. For this reason, a study regarding the effects of soiling in second-surface mirrors’ specular reflection is carried out during a period of one year and a half. Moreover, and since soiling particle deposition can have a crucial economic impact for the energy management of CSP plants, it is important to analyze and establish a method to improve the cleaning operations of a CSP plant considering local conditions. The applied methodology is based in programming artificial cleaning profiles based on ground soiling effect and direct normal irradiance measurements, which are then introduced in a plant simulator to estimate the yearly annual energy production. The generated energy is then used to compare different profiles in face of any possible energy selling price and cleaning costs. The presented method is unique in comparison with current literature, due to its versatility, and can be used for any region.

Recalibration of SPN1 pyranometers against pyrheliometer and its relevance for the evaluation of concentrating solar process heat plants

Accurate determination of direct normal irradiance (DNI) is essential for evaluating and monitoring of concentrating solar power (CSP) and solar process heat (SPH) plants. Currently, a bi-axially tracked Pyrheliometer (PHM) is the most accurate measuring device, but requires high maintenance effort and is cost-intensive. As an alternative, the cost-effective pyranometer SPN1 was investigated by comparing highly resolved minute mean DNI measurements of two SPN1 devices to a reference PHM at the Institute for Solar Technology SPF in Rapperswil (Switzerland). To reduce systematic errors, a linear and a quadratic correction function have been applied to four years of measurements. They reduced the root mean square error RMSE from 48.8 W/m2 (11.5% of mean PHM DNI = 426 W/m2) to 37.9 W/m2 (8.9%) and 29.6 W/m2 (6.9%), respectively. It is also shown that even recalibrations based on only one month of data already reduced the RMSE to maximum 33.4 W/m2 (7.8%), if conducted between April and September. Considering the impact of solar elevation and DNI range on the SPN1 accuracy by bifurcating the correction functions only resulted in a negligibly small improvement of the RMSE. Finally, the influence of SPN1’s DNI measurement accuracy on the performance analysis of a real SPH plant in Switzerland was evaluated. After recalibration and correction of DNI values, the evaluation of monitoring data yielded annual efficiencies that were 15% and 12% higher in 2014 and 2015, respectively. It is recommended to recalibrate SPN1 devices over a minimum of one month against a PHM according to the methods presented in this paper.

Method of identifying the lengths of equivalent clear-sky periods in the time series of DNI measurements based on generalized atmospheric turbidity

In photovoltaic (PV) power plant construction, solar resources need to be evaluated using some criteria. Besides considering the classical total amount of solar resources at the considered location, the lengths and distributions of equivalent clear-sky periods have also been taken into account in recent years to consider the stability of both the PV system output power and other connected grids. We present a method of identifying the lengths and distributions of equivalent clear-sky periods using time series of direct normal irradiance (DNI) measurements based on generalized atmospheric turbidity. This method can be employed not only to calculate the lengths and distributions of historical equivalent clear-sky periods, but also for real-time equivalent clear-sky condition detection. In this report, we demonstrate the validity our algorithm using data from the Solar Radiation Research Laboratory (39.74°N, 105.18°W) and compare our results with those obtained using the Reno and Hansen algorithm.

A new approach to the real-time assessment and intraday forecasting of clear-sky direct normal irradiance

Clear-sky Direct Normal Irradiance (DNI) is the solar power received at ground level per unit of area, at a specific location, under cloud-free conditions. Regarding Concentrating Solar Power (CSP) technologies, such conditions mean that there is no cloud between the Sun and the observer, i.e. the solar field. Since clear sky defines the nominal operating conditions of CSP plants, real-time estimates and forecasts of clear-sky DNI are key information for power plant operators tasked with the management of those plants. So, the present paper focuses first on a new algorithm for the real-time detection of clear-sky situations from DNI measurements. This algorithm makes use of the last-known clear-sky situation and requires the maximum speed at which the atmosphere becomes opaque to be evaluated. The paper also focuses on an efficient approach to the real-time assessment of clear-sky DNI. This approach combines the model developed by Ineichen and Perez with a persistence of atmospheric turbidity, taking advantage of the fact that changes in this quantity are relatively small throughout the day in comparison to changes in DNI, even when the sky is free of clouds. Performance is evaluated via a comparative study, in which empirical models are included, using one-minute data from two sites (Golden, USA, and Perpignan, France). MAE and RMSE are lower than 10Wm-2 and 21Wm-2, respectively. The same approach is capable of providing accurate intraday forecasts of clear-sky DNI. It has proven to be the best compromise between accuracy and complexity (reference is a persistence of DNI) among the considered approaches, including neuro-fuzzy approaches. For a forecast horizon of 5h,MAE≃30Wm-2 and RMSE≃37Wm-2.

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.

Intra-hour direct normal irradiance forecasting through adaptive clear-sky modelling and cloud tracking

Hybrid fossil-solar power cycles possess high solar-to-electric conversion efficiency and the potential to provide stable power output without the need for costly storage systems. However, the power output of the concentrating solar power (CSP) component of the plant can fluctuate in sync with intermittent direct normal irradiance (DNI). For hybrid plants, the controllable fossil-fuel unit can be used to compensate for shortfalls in CSP output during periods of light DNI intermittency or provide the entire plant output during periods of high DNI intermittency. This fossil-fuel unit requires 5–10 min to ramp its power output up or down to perform these balancing functions. Thus, by accurately predicting future DNI, and hence CSP output, the fossil-fuel unit can guarantee stable plant-wide power output during periods of intermittency.This paper develops an intra-hour DNI prediction system using a ground-based cloud motion vector (CMV) framework and real-time DNI measurements. This system comprises a clear-sky DNI model and a cloud fraction prediction algorithm. The presented CSM is based on the Ineichen model (Ineichen, 2008), where the model parameters are adaptively estimated from identified clear-sky DNI measurements over a moving window. For the cloud fraction prediction model, this paper presents an enhanced “sector-ladder” method (Quesada-Ruiz et al., 2014) that uses the weighted mean of circular quantities (Fisher, 1995) and autoregressive filtering to improve cloud flow predictions. Furthermore, a method to forewarn against periods of high DNI intermittency using the generated DNI predictions is presented.The proposed DNI prediction system is evaluated using 37 days of sky-camera images and DNI data collected over the summer of 2014/2015 at the University of Queensland. Over all test days, the adaptive CSM has an average root mean square error of 3.06%, which represents a 19% improvement over a CSM that uses the optimal model parameters from the previous day’s data. Additionally, the modifications to the cloud flow prediction algorithm (the sector-ladder method) are shown to improve the cloud velocity prediction accuracy by a factor of seven over a period of visually determined constant cloud velocity. We find the overall prediction accuracy of the DNI prediction system to be statistically similar to the accepted short-term benchmark of persistence; however, it performs more consistently over a range of weather conditions and is able to forewarn against periods of impending intermittency with 93% accuracy. The latency from data collection to prediction is less than 30 s, making the method eminently suitable for real-time applications.

Modeling beam attenuation in solar tower plants using common DNI measurements

Solar radiation reflected by concentrating mirrors is attenuated due to atmospheric extinction as it travels to the receiver of a solar tower plant. The lack of information on the magnitude of extinction increases the uncertainties in yield analysis and tower plant design. In-situ measurements of atmospheric extinction as well as measurement correction methods have been recently performed and developed (Hanrieder et al., 2012, 2015), but specific information is unavailable for individual plant projects. It is well known though that the extinction varies significantly with site and time.To overcome this absence of information a model to derive the attenuation loss between heliostat and receiver from common direct normal irradiance (DNI) measurements was developed by Sengupta and Wagner (2011) (SW2011 model). We present an updated version of that model and a comparison between the performance of the models using extinction measurements.In the new approach presented here, different precipitable water vapor (PWV) amounts are considered and the model is adjusted to the elevation of the investigated site.The strongest assumption in this approach is the assumption about the aerosol extinction height profile. Three different height profiles are tested for the Plataforma Solar de Almería (PSA) resulting in three different new transmittance models.The SW2011 as well as the three new models are evaluated with one year of corrected extinction data derived with the ABC (absorption and broadband correction) method of Hanrieder et al. (2015) and a Vaisala FS11 scatterometer at PSA. The new models show a mean difference to the reference data set of 0.01, 0.05 and 0.03 and a root mean square error (RMSE) of 0.052, 0.056 and 0.049 (compared to a mean bias of −0.08 and RMSE of 0.095 for the SW2011 model for transmittances through a 1km slant range). These results indicate the importance of adequate assumptions for the aerosol height profile. Testing the developed TM with the LIVAS height profile (Amiridis et al., 2015) for PSA shows satisfying results and this motivates testing the approach for other sites. By applying an additional correction for the Linke turbidity (TL) derived as in Ineichen and Perez (2002) the mean bias can be further lowered.An uncertainty analysis shows that the absolute uncertainty coincide with the RMSE levels of the evaluation. Performing the additional TL correction promises an improvement of the overall performance of the model.The new models outperform the SW2011 model due to the PWV and elevation adjustments. The approach can be applied for different sites and incorporated in already existing ray-tracing or plant optimization tools. It is expected to be valuable for reducing uncertainty in power tower design and operations.

Using multi-pyranometer arrays and neural networks to estimate direct normal irradiance

Direct Normal Irradiance (DNI) is a critical component of solar irradiation for estimating Plane of Array (POA) irradiance on flat plate systems and for estimating photovoltaic and concentrating system power output. Current approaches to measuring or estimating DNI suffer from either high equipment costs or low precision and may require detailed environmental data. An alternative approach, using artificial neural networks to estimate DNI from the irradiance measurements of multiple pyranometers, is studied. We consider various neural network topologies and study the resulting errors. The neural network-based estimators are found to have higher accuracy than those obtained from empirical correlations of GHI measurements alone. Additionally, the use of a different GHI sensor than the one used to obtain the neural network training data does not induce significant errors. The ability of this method to be used as a quality-control instrument for pyrheliometer measurements is also discussed. We find that the proposed methodology is capable of detecting many instances of unreliable DNI measurements by considering the deviation between the predicted DNI and measured DNI. A more detailed analysis can be conducted by taking advantage of the data streams from the individual pyranometers.

Validation of the Surface Downwelling Solar Irradiance Estimates of the HelioClim-3 Database in Egypt

HelioClim-3 (HC3) is a database providing time series of the surface downwelling solar irradiance that are computed from images of the Meteosat satellites. This paper presents the validation results of the hourly global horizontal irradiance (GHI) and direct normal irradiance (DNI), i.e., beam irradiance at normal incidence, of versions four and five of HC3 at seven Egyptian sites. The validation is performed for all-sky conditions, as well as cloud-free conditions. Both versions of HC3 provide similar performances whatever the conditions. Another comparison is made with the estimates provided by the McClear database that is restricted to cloud-free conditions. All databases capture well the temporal variability of the GHI in all conditions, McClear being superior for cloud-free cases. In cloud-free conditions for the GHI, the relative root mean square error (RMSE) are fairly similar, ranging from 6% to 15%; both HC3 databases exhibit a smaller bias than McClear. McClear offers an overall better performance for the cloud-free DNI estimates. For all-sky conditions, the relative RMSE for GHI ranges from 10% to 22%, except one station, while, for the DNI, the results are not so good for the two stations with DNI measurements.

Evaluation of decomposition models of various complexity to estimate the direct solar irradiance over Belgium

Solar energy production is directly correlated to the amount of radiation received at a given location. Appropriate information on solar resources is therefore very important for designing and sizing solar energy systems. Concentrated solar power projects and photovoltaic tracking systems rely predominantly on direct normal irradiance (DNI). However, the availability of DNI measurements from surface observation stations has proven to be spatially too sparse to quantify solar resources at most potential sites. Satellite data can be used to calculate estimates of direct solar radiation where ground measurements do not exist. Performance of decomposition models of various complexity have been evaluated against one year of in situ observations recorded on the roof of the radiometric tower of the Royal Meteorological Institute of Belgium in Uccle, Brussels. Models were first evaluated on a hourly and sub-hourly basis using measurements of global horizontal irradiance (GHI) as input. Second, the best performing ground-based decomposition models were used to extract the direct component of the global radiation retrieved from Meteosat Second Generation (MSG) images. Results were then compared to direct beam estimations provided by satellite-based diffuse fraction models and evaluated against direct solar radiation data measured at Uccle. Our analysis indicates that valuable DNI estimation can be derived from MSG images over Belgium regardless of the satellite retrieved GHI accuracy. Moreover, the DNI retrieval from MSG data can be implemented on an operational basis.

Analyzing the contribution of aerosols to an observed increase in direct normal irradiance in Oregon

Annual average total irradiance increases by 1–2% per decade at three monitoring stations in Oregon over the period from 1980 to 2007. Direct normal irradiance measurements increase by 5% per decade over the same time period. The measurements show no sign of a dimming before 1990. The impact of high concentrations of stratospheric aerosols following the volcanic eruptions of El Chichón and Mount Pinatubo are clearly seen in the measurements. Removing these years from the annual average all‐sky time series reduces the trends in both total and direct normal irradiance. Clear‐sky periods from this long direct normal time series are used in conjunction with radiative transfer calculations to test whether part of the increase could be caused by anthropogenic aerosols. All three sites show relatively low clear‐sky measurements before the eruption of El Chichón in 1982, suggesting higher aerosol loads during this period. After removing the periods most strongly impacted by volcanic eruptions, two of the sites show statistically significant increases in clear‐sky direct normal irradiance from 1987 to 2007. Radiative transfer calculations of the impact of volcanic aerosols and tropospheric water vapor indicate that only about 20% of that clear‐sky increase between background aerosol periods before and after the eruption of Mount Pinatubo can be explained by these two factors. Thus a statistically significant clear‐sky trend remains between 1987 and 2007 that is consistent with the hypothesis that at least some of the increase in surface irradiance could be caused by a reduction of anthropogenic aerosols.

Comparative study on different models for estimation of direct normal irradiance (DNI) over Egypt atmosphere

The results obtained by using seven-parameterization broadband models to estimate Direct Normal Irradiance (DNI) along with two spectral models for four sites in Egypt atmosphere were compared with ground DNI measurements. Some statistical indicators (MBE, RMSE and R 2) have been used to measure the performance of the used models. MBE for all dataset is <1% to both spectral models (SPCTRAL2, SMARTS2) and broadband models (MLWT1, MLWT2 and REST) while is equal to 1.2% to YANG model. However, RMSE are around 2% for spectral models and 3% to the broadband models. The error in prediction of DNI to such models is below experimental errors a part from the big number of observations. On the other hand, Louche, Dogniaux and Rodgers models provide relatively bad performance, RMSE are at most cases >4%. Determination coefficient ( R 2) results to all models are near 1.0. If we excluded spectral models, the broadband models MLWT1, MLWT2 and REST along with YANG models provide the best performance in all tests, therefore, those models can be used in Egypt atmosphere.

Evaluation of the estimation of diffuse irradiance from global and direct normal irradiance measurements

As part of a radiometer intercomparison experiment, different combinations of pyranometer and pyrheliometers were used to estimate diffuse solar irradiance on a horizontal surface. Sixteen combinations were possible with four pyranometers and four pyrheliometers. The values were intercompared and then compared to the measured values, obtained with a solar tracking/occulting disc system. The difference in estimated values using different radiometer combinations varied from 1 to 21 W/m 2. It was found that uncertainty in measuring the global irradiance accounted for most of the difference in estimated values. In the worst case, a 2.1 per cent difference in the global irradiance as measured by two different pyranometers caused a 12 per cent difference in estimates of the diffuse irradiance. It is shown that, if the estimated and measured values are analyzed statistically, agreement to within 1 per cent is possible.