Solar power forecasting: Difference between revisions

Revision as of 20:55, 12 August 2019

Latest revision as of 06:48, 2 June 2025

(41 intermediate revisions by 25 users not shown)

Line 1:

{{short description|Power forecasting}}

{{Multiple issues|

{{cleanup|reason=unencyclopedic, incomprehensible and unverified information|date=October 2017}}

{{underlinked|date=January 2013}}

{{more footnotes|date=July 2012}}

}}

{{Sustainable energy}}

'''Solar power forecasting''' is the process of gathering and analyzing data in order to predict [[solar power]] generation on various time horizons with the goal to mitigate the impact of solar intermittency. Solar power forecasts are used for efficient management of the [[Electrical grid|electric grid]] and for power trading.<ref>{{Cite journal|date=2016-06-01|title=Day-ahead forecasting of solar power output from photovoltaic plants in the American Southwest|url=https://www.sciencedirect.com/science/article/abs/pii/S0960148116300398|journal=Renewable Energy|language=en|volume=91|pages=11–20|doi=10.1016/j.renene.2016.01.039|issn=0960-1481|last1=Larson |first1=David P. |last2=Nonnenmacher |first2=Lukas |last3=Coimbra |first3=Carlos F.M. |bibcode=2016REne...91...11L |url-access=subscription}}</ref>

'''Solar power forecasting''' involves knowledge of the [[Sun]]´s path, the [[atmosphere]]'s condition, the scattering processes and the characteristics of a [[solar energy]] plant which utilizes the Sun's energy to create [[solar power]]. Solar [[photovoltaic system]]s transform solar energy into electric power. The power output depends on the incoming radiation and on the solar panel characteristics. Photovoltaic power production is increasing nowadays. Forecast information is essential for an efficient use, the management of the electricity grid and for solar energy trading. Common solar forecasting method include stochastic learning method, local and remote sensing method, and hybrid method (Chu et al. 2016).

As major barriers to solar energy implementation, such as materials cost and low conversion efficiency, continue to fall, issues of intermittency and reliability have come to the fore.<ref>{{Cite book|title=Solar Energy Forecasting and Resource Assessment - 1st Edition|url=https://www.elsevier.com/books/solar-energy-forecasting-and-resource-assessment/kleissl/978-0-12-397177-7|access-date=2021-06-29|website=www.elsevier.com|date=25 June 2013 |isbn=978-0-12-397177-7 }}</ref> The intermittency issue has been successfully addressed and mitigated by solar forecasting in many cases.<ref>{{Cite journal|date=2016-02-01|title=Benefits of solar forecasting for energy imbalance markets|url=https://www.sciencedirect.com/science/article/abs/pii/S0960148115302901|journal=Renewable Energy|language=en|volume=86|pages=819–830|doi=10.1016/j.renene.2015.09.011|issn=0960-1481|last1=Kaur |first1=Amanpreet |last2=Nonnenmacher |first2=Lukas |last3=Pedro |first3=Hugo T.C. |last4=Coimbra |first4=Carlos F.M. |bibcode=2016REne...86..819K |url-access=subscription}}</ref><ref>{{Cite journal|date=2019-10-01|title=Operational solar forecasting for the real-time market|url=https://www.sciencedirect.com/science/article/abs/pii/S0169207019300755|journal=International Journal of Forecasting|language=en|volume=35|issue=4|pages=1499–1519|doi=10.1016/j.ijforecast.2019.03.009|issn=0169-2070|last1=Yang |first1=Dazhi |last2=Wu |first2=Elynn |last3=Kleissl |first3=Jan |s2cid=195463551 |doi-access=free}}</ref><ref>{{Cite journal|date=2018-01-15|title=Solar photovoltaic generation forecasting methods: A review|url=https://www.sciencedirect.com/science/article/abs/pii/S0196890417310622|journal=Energy Conversion and Management|language=en|volume=156|pages=459–497|doi=10.1016/j.enconman.2017.11.019|issn=0196-8904|last1=Sobri |first1=Sobrina |last2=Koohi-Kamali |first2=Sam |last3=Rahim |first3=Nasrudin Abd. |bibcode=2018ECM...156..459S |url-access=subscription}}</ref>

==Generation forecasting==

Information used for the solar power forecast usually includes the [[Sun]]´s path, the [[atmosphere|atmospheric]] conditions, the scattering of light and the characteristics of the [[solar energy]] plant.

The energy generation forecasting problem is closely linked to the problem of ''weather variables forecasting''. Indeed, this problem is usually split into two parts, on one hand focusing on the forecasting of solar PV or any other meteorological variable and on the other hand estimating the amount of energy that a concrete power plant will produce with the estimated meteorological resource.

In general, the way to deal with this difficult problem is usually related to the spatial and temporal scales we are interested in, which yields to different approaches that can be found in the literature. In this sense, it is useful to classify these techniques depending on the forecasting horizon, so it is possible to distinguish between ''now-casting'' (forecasting 3–4 hours ahead), ''short-term forecasting'' (up to 7 days ahead) and ''long-term forecasting'' (months, years...)

Solar radiation closely follows the physical and biological development of the earth. Its spatial and sequential heterogeneity powerfully influence the forcing of environmental and hydrological organisms by manipulating air temperature, soil moisture and vapor transpiration, snow cover and lots of photochemical procedures. Therefore, solar radiation drives place efficiency and plant life allotment, organism a key feature in undeveloped and forestry sciences that be obliged to be known precisely. The quantity of solar radiation obtainable at the earth’ surface is at the outset controlled at worldwide balance, organism above all precious by the Sun Earth geometry and the atmosphere. On the other hand, a complete explanation of its freedom time unpredictability require deliberation of limited procedure which frequently turn out to be also applicable, as is the casing in mountainous region. Predominantly, limited territory adjust the inward bound solar radiation by shadow casts, slope of elevation, surface gradient and compass reading, as a result, precise spatial model of inward bound solar radiation be supposed to regard as the pressure of the terrain surface. In the final time, more than a few events to consist of the confined terrain special effects in the solar radiation countryside have been projected, such as the use of Geographical Information Systems (GIS), artificial intelligence or post dispensation of satellite stand technique. Solar radiation can be also evaluated using numerical weather forecast (NWP) models. Nevertheless, the space and time balance determined with them and the incomplete computational ability frequently avoid the deliberation of terrain connected property.

Generally, the solar forecasting techniques depend on the forecasting horizon

Otherwise, exclamation technique agree to us to acquire spatially persistent database from data evidence at inaccessible station greater than wide region. Even though their dependability is powerfully needy on the opening coldness between position, they eventually rely on experiential statistics, which have a superior precision than extra method. Therefore, while an adequate footage spatial thickness is accessible, disturbance method are preferred. Conventionally, solar radiation has not been as densely example as additional variables as temperature or rainfall, therefore the ease of use of capacity is frequently in short supply. Though, the number of experimental system which record solar radiation has developed and interruption has been converted into an appropriate technique for solar radiation evaluation. Nevertheless, radiometric stations are frequently come together approximately farmland or occupied region, typically during basin and plane area, while mountains at rest require enough footage thickness. This truth is particularly applicable afford the tall spatial unpredictability of solar radiation in these province. As an outcome, particular interruption method that tolerate include outdoor foundation should be used to make clear this extra spatial unpredictability. Several diverse spatial interruption techniques can be established. On the other hand, data ease of use in mountainous region is often extremely restricted. As a result, it is hard to construct a precise solar radiation climatology in hilly area to be used in environmental science, climate change.

* ''Nowcasting'' (forecasting 3–4 hours ahead),

Solar radiation is a hardly illustration changeable with reverence to supplementary ecological variables such as temperature or precipitation, in fraction payable to the high maintenance price of the necessary radiometric sensors. It is extremely sensitive to ecological feature on or after local to limited balance. Predominantly, terrain surface confronts the conventional interruption method while forecast through far above the ground spatial decision are required, particularly due to the lack of measurement stations in mountainous areas. Geo-statistics front a stochastic move toward to resolve the spatial forecast difficulty that stop dependence on before imagine deterministic models and permit us to consist of the consequence of outside in sequence foundation stand on investigation data-sets.

* ''Short-term forecasting'' (up to seven days ahead) and

* ''Long-term forecasting'' (weeks, months, years)

Many solar resource forecasting methodologies were proposed since the 1970 and most authors agree that different forecast horizons require different methodologies. Forecast horizons below 1 hour typically require ground based sky imagery and sophisticated time series and machine learning models. Intra-day horizons, normally forecasting irradiance values up to 4 or 6 hours ahead, require satellite images and irradiance models. Forecast horizons exceeding 6 hours usually rely on outputs from numerical weather prediction (NWP) models.<ref>{{Cite thesis|title=Solar Energy Resourcing and Forecasting for Optimized Grid Integration|url=https://escholarship.org/uc/item/8kd3b038|publisher=UC San Diego|date=2015|language=en|first=Lukas|last=Nonnenmacher}}</ref>

==Nowcasting==

The term "Nowcasting" in the context of solar power forecasting, generally refers to same spatial and temporal scales as [[Nowcasting (meteorology)|meteorological Nowcasting]], which focuses on forecast horizons from a few minutes ahead, out to 4–6 hours ahead. In general, the 'Nowcasting' forecast horizon refers to those not well-served by global numerical weather prediction models, which produce outputs at up to hourly resolutions and only update every 6 hours. [[Solar power]] nowcasting then refers to the prediction of solar power output (or energy generation) over time horizons of tens to hundreds of minutes ahead of time with up to 90% predictability.<ref>{{cite web |last1=Vorrath |first1=Sophie |title=New APVI solar tool shows daily, time-based forecast for each state |url=https://reneweconomy.com.au/new-apvi-solar-tool-shows-daily-time-based-forecast-for-each-state-52445/ |website=RenewEconomy |language=en-AU |date=31 May 2019}}</ref> It has historically been very important for [[Electrical grid|electrical grid operators]] in order to guarantee the matching of supply and demand on energy markets. Such solarSolar power nowcasting services are usually related to temporal resolutions of 5 to 15 minutes, with updates as frequent as every 5 minutesminute. The regular updates and relatively high resolutions required from these methods require automatic weather data acquisition and processing techniques, which are chiefly accomplished by two primary means:<ref>{{Cite web|url= https://www.elsevier.com/books/solar-energy-forecasting-and-resource-assessment/kleissl/978-0-12-397177-7 |title=Solar Energy Forecasting and Resource Assessment - 1st Edition|website=www.elsevier.com |access-date=2019-05-08}}</ref>

The high resolution required for accurate nowcast techniques require high resolution data input including ground imagery, as well as fast data acquisition form irradiance sensors and fast processing speeds.

#'''Statistical techniques.''' These are usually based on [[time series]] processing of measurement data, including [[Surface weather observation|meteorological observations]] and power output measurements from a solar power facility. What then follows is the creation of a [[Training, validation, and test sets|training dataset]] to tune the parameters of a model (I. Espino eta al, 2011), before evaluation of model performance against a separate testing dataset. This class of techniques includes the use of any kind of statistical approach, such as [[Autoregressive–moving-average model|autoregressive moving averages]] (ARMA, ARIMA, etc.), as well as machine learning techniques such as [[Artificial neural network|neural networks]], [[support vector machine]]s (etc.). These approaches are usually benchmarked to a persistence approach in order to evaluate their improvements. This persistence approach just assumes that any variable at time step t is the value it took in a previous time.

#[[File:Satellite Based Solar Nowcasting.gif|thumb|An example of satellite-based cloud cover nowcasting, which is used to generate predication of solar power outputs. Credit: [http://solcastglobal.com Solcast]]]'''Satellite based methods.''' These methods leverage the several [[Geostationary orbit|geostationary]] Earth observing [[weather satellite]]s (such as [[Meteosat|Meteosat Second Generation (MSG) fleet]]'')'' to detect, characterise, track and predict the future locations of cloud cover. These satellites make it possible to generate solar power forecasts over broad regions through the application of image processing and forecasting algorithms. Key forecasting algorithms include cloud motion vectors (CMVs).<ref>{{Cite web|url= http://glossary.ametsoc.org/wiki/Cloud_motion_vector |title=Cloud motion vector - AMS Glossary|website=glossary.ametsoc.org |access-date=2019-05-08}}</ref> Relevant methods for applying physical models based on satellite image processing techniques provide an estimation of future atmospheric values can be found in ''Alvarez et al.'', 2010.

#'''StatisticalThe techniquesactual nowcast is then frequently enhanced by e.g.''' These[[statistical technique]]s. In the case of nowcasting, these techniques are usually based on [[time series]] processing of measurement data, including [[Surface weather observation|meteorological observations]] and power output measurements from a solar power facility. What then follows is the creation of a [[Training, validation, and test sets|training dataset]] to tune the parameters of a model (I. Espino eta al, 2011), before evaluation of model performance against a separate testing dataset. This class of techniques includes the use of any kind of statistical approach, such as [[Autoregressive–moving-average model|autoregressive moving averages]] (ARMA, ARIMA, etc.), as well as machine learning techniques such as [[Artificial neural network|neural networks]], [[support vector machine]]s (etc.).<ref>{{cite Thesejournal|last1=Sanjari|first1=M.J.|last2=Gooi|first2=H.B.|date=2016|title=Probabilistic approachesForecast areof usuallyPV benchmarkedPower toGeneration a persistencebased approach inon Higher-order toMarkov evaluateChain|journal=IEEE theirTransactions improvements.on ThisPower persistenceSystems|volume=32|issue=4|pages=2942–2952|doi=10.1109/TPWRS.2016.2616902|s2cid=43911568 approach just assumes that any variable at time step t is the value it took in a previous time.|hdl=10072/409742|hdl-access=free}}</ref>

An important element of nowcasting solar power are ground based sky observations and basically all intra-day forecasts.<ref>{{Cite journal|date=2011-11-01|title=Intra-hour forecasting with a total sky imager at the UC San Diego solar energy testbed|url=https://www.sciencedirect.com/science/article/abs/pii/S0038092X11002982|journal=Solar Energy|language=en|volume=85|issue=11|pages=2881–2893|doi=10.1016/j.solener.2011.08.025|issn=0038-092X|last1=Chow |first1=Chi Wai |last2=Urquhart |first2=Bryan |last3=Lave |first3=Matthew |last4=Dominguez |first4=Anthony |last5=Kleissl |first5=Jan |last6=Shields |first6=Janet |last7=Washom |first7=Byron |bibcode=2011SoEn...85.2881C }}</ref>

==Short-term solar power forecasting==

[[File:Diagram-of-the-UCSD-sky-imager-USI-and-related-solar-and-sky-geometries-th-0-is-the.jpg|alt=|thumb|An example of sky-imager used for detecting, tracking and predicting cloud cover conditions in the vicinity of a solar energy facility of interest. Most often, these devices are used to make estimates of solar irradiance from the images using local calibration by a pyranometer. The solar irradiance short-term forecasts are then fed into PV power modelling routines to generate a solar power forecast. <br />Credit: [https://topex.ucsd.edu/rs/presentations/cortes.pdf UC San Diego]]]''Short-term'' forecasting provides predictions up to seven days ahead. Due to the power market regulation in many jurisdictions, intra-day forecasts and day-ahead solar power forecasts are the most important time horizons in this category. Basically all highly accurate short term forecasting methods leverage several data input streams such as meteorological variables, local weather phenomena and ground observations along with complex mathematical models.

=== Ground based sky observations ===

''Short-term'' forecasting provides predictions up to 7 days ahead. This kind of forecast is also valuable for grid operators in order to make decisions of grid operation, as well as, for electric market operators.<ref>{{cite journal|last1=Sanjari|first1=M.J.|last2=Gooi|first2=H.B.|title=Probabilistic Forecast of PV Power Generation based on Higher-order Markov Chain|journal=IEEE Transactions on Power Systems|date=2016|doi=10.1109/TPWRS.2016.2616902}}</ref>

For intra-day forecasts, local cloud information is acquired by one or several ground-based sky imagers at high frequency (1 minute or less). The combination of these images and local weather measurement information are processed to simulate cloud motion vectors and [[optical depth]] to obtain forecasts up to 30 minutes ahead.<ref>{{Cite journal|date=2011-11-01|title=Intra-hour forecasting with a total sky imager at the UC San Diego solar energy testbed|url=https://www.sciencedirect.com/science/article/abs/pii/S0038092X11002982|journal=Solar Energy|language=en|volume=85|issue=11|pages=2881–2893|doi=10.1016/j.solener.2011.08.025|issn=0038-092X|last1=Chow |first1=Chi Wai |last2=Urquhart |first2=Bryan |last3=Lave |first3=Matthew |last4=Dominguez |first4=Anthony |last5=Kleissl |first5=Jan |last6=Shields |first6=Janet |last7=Washom |first7=Byron |bibcode=2011SoEn...85.2881C }}</ref>

Under this perspective, the meteorological resources are estimated at a different temporal and spatial resolution. This implies that meteorological variables and phenomena are looked from a more general perspective, not as local as nowcasting services do. In this sense, most of the approaches make use of different numerical weather prediction models (NWP) that provide an initial estimation of weather variables. Currently, several models are available for this purpose, such as [[Global Forecast System]] (GFS) or data provided by the European Center for Medium Range Weather Forecasting ([[ECMWF]]). These two models are considered the state of the art of global forecast models, which provide meteorological forecasts all over the world.

In order to increase spatial and temporal resolution of these models, other models have been developed which are generally called mesoscale models. Among others, [[HIRLAM]], [[Weather Research and Forecasting Model|WRF]] or [[MM5 (weather model)|MM5]] are the most representative of these models since they are widely used by different communities.

=== Satellite based methods ===

To run these models a wide expertise is needed in order to obtain accurate results, due to the wide variety of parameters that can be configured in the models. In addition, sophisticated techniques such as data assimilation might be used in order to produce more realistic simulations.

These methods leverage the several [[Geostationary orbit|geostationary]] Earth observing [[weather satellite]]s (such as [[Meteosat|Meteosat Second Generation (MSG) fleet]]'')'' to detect, characterise, track and predict the future locations of [[cloud cover]]. These satellites make it possible to generate solar power forecasts over broad regions through the application of [[image processing]] and forecasting [[algorithm]]s. Some satellite based forecasting algorithms include cloud motion vectors (CMVs)<ref>{{Cite web|title=Cloud motion vector - AMS Glossary|url=http://glossary.ametsoc.org/wiki/Cloud_motion_vector|access-date=2019-05-08|website=glossary.ametsoc.org}}</ref> or [[Streamlines, streaklines, and pathlines|streamline]] based approaches.<ref>{{Cite journal|date=2014-10-01|title=Streamline-based method for intra-day solar forecasting through remote sensing|url=https://www.sciencedirect.com/science/article/abs/pii/S0038092X14003752|journal=Solar Energy|language=en|volume=108|pages=447–459|doi=10.1016/j.solener.2014.07.026|issn=0038-092X|last1=Nonnenmacher |first1=Lukas |last2=Coimbra |first2=Carlos F.M. |bibcode=2014SoEn..108..447N |url-access=subscription}}</ref>

Finally, some communities argue for the use of post-processing techniques, once the models’ output is obtained, in order to obtain a probabilistic point of view of the accuracy of the output. This is usually done with ensemble techniques that mix different outputs of different models perturbed in strategic meteorological values and finally provide a better estimate of those variables and a degree of uncertainty, like in the model proposed by Bacher et al. (2009)

=== Numerical weather prediction ===

UnderMost this perspective,of the meteorologicalshort resourcesterm are estimated at a different temporal and spatial resolution. This implies that meteorological variables and phenomena are looked from a more general perspective, not as local as nowcasting services do. In this sense, most of theforecast approaches make use of different [[numerical weather prediction]] models (NWP) that provide an initialimportant estimation of the development of weather variables. Currently, severalThe models areused availableincluded for this purpose, such asthe [[Global Forecast System]] (GFS) or data provided by the European Center for Medium Range Weather Forecasting ([[ECMWF]]). These two models are considered the state of the art of global forecast models, which provide meteorological forecasts all over the world.

In order to increase spatial and temporal resolution of these models, other models have been developed which are generally called mesoscale models. Among others, [[HIRLAM]], [[Weather Research and Forecasting Model|WRF]] or [[MM5 (weather model)|MM5]]. Since these NWP models are thehighly mostcomplex representativeand ofdifficult to run on local computers, these variables are usually considered as exogeneous inputs to [[solar irradiance]] models sinceand theyingested areform widelythe respective data provider. Best forecasting results are usedachieved bywith different[[data communitiesassimilation]].

Finally, someSome communitiesresearchers argue for the use of post-processing techniques, once the models’ output is obtained, in order to obtain a [[Probability|probabilistic]] point of view of the accuracy of the output. This is usually done with ensemble techniques that mix different outputs of different models perturbed in strategic meteorological values and finally provide a better estimate of those variables and a degree of uncertainty, like in the model proposed by Bacher et al. (2009).

==Long-term solar power forecasting==

''Long-term'' forecasting usually refers to forecasting oftechniques applied to time horizons on the annualorder orof monthlyweeks availableto resourceyears. ThisThese istime usefulhorizons can be relevant for energy producers and to negotiate contracts with financial entities or [[Public utility|utilities]] that distribute the generated energy.

In general, these long-term forecasting is usually done at a lower scale than any of the two previous approaches. Hence, most of these models are run with mesoscale models fed with reanalysis data as input and whose output is postprocessed with statistical approaches based on measured data.

In general, these long-term forecasting horizons usually rely on [[Numerical weather prediction|NWP]] and [[Climatology|climatological]] models. Additionally, most of the forecasting methods are based on [[Mesoscale meteorology|mesoscale]] models fed with reanalysis data as input. Output can also be postprocessed with [[Statistics|statistical]] approaches based on measured data. Due to the fact that this time horizon is less relevant from an operational perspective and much harder to model and validate, only about 5% of solar forecasting publications consider this horizon.

==Energetic models==

Any output from anya model described above must then be converted to the electric energy that a particular solar PV plant will produce. This step is usually done with statistical approaches that try to correlate the amount of available resource with the metered power output. The main advantage of these methods is that the meteorological prediction error, which is the main component of the global error, might be reduced taking into account the uncertainty of the prediction.

As it was mentioned before and detailed in ''Heinemann et al.'', these statistical approaches comprises from ARMA models, neural networks, support vector machines, etc.

On the other hand, there also exist theoretical models that describe how a power plant converts the meteorological resource into electric energy, as described in Alonso et al. The main advantage of this type of models is that when they are fitted, they are really accurate, although they are too sensitive to the meteorological prediction error, which is usually amplified by these models.

Hybrid models, finally, are a combination of these two models and they seem to be a promising approach that can outperform each of them individually.

Line 53 ⟶ 63:

{{reflist}}

* Y. Chu, M. Li and C.F.M. Coimbra (2016) “Sun-Tracking Imaging System for Intra-Hour DNI Forecasts” Renewable Energy (96), Part A, pp.&nbsp;792–799.

* Luis Martín, Luis F. Zarzalejo, Jesús Polo, Ana Navarro, Ruth Marchante, Marco Cony, Prediction of global solar irradiance based on time series analysis: Application to solar thermal power plants energy production planning, Solar Energy, Volume 84, Issue 10, October 2010, Pages 1772-1781, {{ISSN|0038-092X}}, {{DOI|10.1016/j.solener.2010.07.002}}.

* Heinemann, D., Lorenz E., Girodo M. Forecasting of solar radiation. Oldenburg University, Institute of Physics, Energy Meteorology Group.

* Alonso, M, Chenlo F. Estimación de la energía generada por un sistema fotovoltaico conectado a red. CIEMAT. Laboratorio de sistemas fotovoltaicos.

* Alvarez, L., Castaño, C.A., Martín, J. A computer vision approach for solar radiation nowcasting using MSG images. EMS Annual Meeting Abstracts. Vol. 7, EMS2010-495, 2010. 10th EMS/8th ECAC.

* Espino, I., Hernández, M.. Nowcasting of wind speed using support vector regression. Experiments with Time Series from Gran Canaria. Renewable Energy and Power Quality Journal, {{ISSN|2172-038X}}, N9, 12 May 2011.

* Bacher, P., Madsen, H., Nielsen H.A. Online short-term solar power forecasting. Solar Energy. Vol 83, Issue 10, October 2009: 1772-1783.

* Diagne, H.M., David, M., Lauret, P., Boland, J. Solar irradiation forecasting: state-of-the-art and proposition for future developments for small-scale insular grids. In Proceedings of the World Renewable Energy Forum 2012 (WREF 2012), Denver, USA, May 2012.

==External links==

* [https://www.nrel.gov/grid/solar-wind-forecasting.html Solar and Wind Forecasting projects], by National Renewable Energy Laboratory (NREL).

* [http://www.solarelectricpower.org/media/147876/SEPA-ForecastReport-2014-ExecSummary.pdf SEPA – Predicting Solar Power Production]

[[Category:Photovoltaics]]