Power generation forecast

FORECASTING OF ELECTRICITY GENERATION BY SOLAR POWER PLANTS

It is known that the fine to solar power plants for settling imbalances will be set starting 2021 (the share of reimbursement of the cost of settling the imbalance to the Guaranteed Buyer for a solar power plant with a capacity of up to 1 MW is 10%, and for a solar power plant over 1 MW is 50%); therefore, it is necessary to collect sufficient data on actual electricity generation as soon as possible and to achieve high forecasting accuracy.

IKNET together with the Slovak company SOLARGIS, which has been a leader in forecasting of electricity generation by solar power plants in Europe and in the world for 10 years, offers high-quality forecasting services in Ukraine.

Why SOLARGIS:

a market leader in forecasting;
a team of the best experts and scientists trusted by thousands of companies around the world (Jinko Solar, SunPower, DNV GL, RTR Rete Rinnovable, ENEL, JGC, Total, etc);
many years of European experience in high-quality, reliable, detailed and accurate forecasting;
possibility of hourly forecasting and forecasting from 7 to 10 days in advance;
free trial of the program from 1 to 3 months;
annual forecasting cost ranging from 960 euros to 1500 euros.

Forecasting methodology
Solar radiation reaching the Earth's surface consists of three components: direct radiation, diffuse radiation and reflected radiation. These three components together create global radiation.
In solar energy applications, the following parameters are commonly used:

Direct Normal Irradiation/Irradiance (DNI);
Global Horizontal Irradiation/Irradiance (GHI) is the sum of direct and diffuse radiation received on a horizontal plane.
Global Tilted Irradiation/Irradiance (GTI) is the sum of direct and reflected radiation received on a surface with defined tilt and azimuth.

SOLARGIS uses a semi-empirical satellite model to obtain radiation data. Data from satellites are used for identification of cloud properties using the most advanced algorithms; numerical weather models are also used for forecasting.
The numerical model is constantly correlated with satellite data to improve prediction accuracy.
The solar radiation retrieval is basically split into three steps:

First, the clear-sky irradiance is calculated using the clear-sky model.
Second, direct normal and global horizontal radiation is calculated based on the satellite data and cloud index to retrieve all-sky irradiance;
Third, direct normal and global horizontal irradiance are used for computing diffuse and global tilted irradiance and/or irradiance corrected for shading effects.

The accuracy of predicting annual global horizontal radiation based on data is ± 4% for most European countries.
At different stages of modeling and calculations, many factors that affect the generation are taken into account. The energy losses can be classified in two groups:

Static: module surface pollution, losses in cables, and mismatch between PV modules;
Dynamic: these losses depend on the irradiance/temperature conditions, which change over the day and over the seasons.

The numerical model takes into account the following losses:

Global irradiation impinging on a tilted plane of PV modules is calculated from Global Horizontal Irradiance (GHI), Direct Normal Irradiance (DNI), terrain albedo, and instantaneous sun position within subhourly time interval.
Shading by terrain features is calculated by disaggregation using SRTM-3 DEM and horizon height.
Losses due to angular reflectivity: The magnitude of effects depends on relative position of the sun and plane of the module. The accuracy of calculations of angular reflectivity losses depends on cleanness and specific properties of the module surface.
Losses due to snow: Losses of solar radiation during winter months are increased by snow on the modules. Typically snow remains on the modules surface from the morning hours to the noon. The highest negative impact on the production has snowfall during cold, cloudy days, when the temperature of the modules is not high enough to melt the snow.
Losses due to dirt and soiling.
Losses due to modules operating outside of standard test conditions (STC): depends mainly on the environmental factors and cleaning of the PV modules surface during the power plant lifetime. Typically the losses at this step are higher for crystalline silicon modules than thin films.
Losses by inter-row shading: Relative spacing leads to electricity losses due to short distance shading.
Power tolerance of modules: From the module power tolerance result bigger or smaller mismatch losses of the modules connected in strings.
Inverter losses.
Inverter losses from conversion of DC to AC.
Simple solar power plant (repair and maintenance).
Long-term degradation of modules.

For more information about the SOLARGIS forecasting methodology go to https://solargis.com/docs/methodology

As part of the annual forecasting package, the customer is regularly provided with the following data parameters:

PV electricity output (PVOUT) – P50, MW/h;
Global horizontal irradiance/irradiation, GHI, kW/m²;
Global tilted irradiance/irradiation, GTI, kW/m²;
Direct normal irradiance/irradiation, DNI, kW/m²;
Air temperature at 2 m, TEMP;
Wind speed at 10 m, WS, m/sec
Wind direction at 10 m, WD;
Relative humidity, RH, %;
Atmospheric pressure, AP, hPa;
Precipitation rate, PREC, mm or kg/m².

All data is processed on SOLARGIS's own servers, no additional equipment is required from the Customer.
In most cases, including Ukraine, the approximate expected error for annual values is from ± 4% to ± 8% for global horizontal radiation and from ± 8% to ± 15% for direct normal radiation.
As part of forecasting electricity generation by solar power plants, IKNET together with SOLARGIS provides the following services:

Providing the solar power plant owner with a questionnaire and an agreement to receive free forecast data in test mode.
Signing of an agreement by the solar power plant owner for obtaining forecast data in test mode with mandatory indication of the free trial period.
Assistance to the solar power plant owner in collecting and generating data to fill out a questionnaire for existing and/or future solar power plants.
Provision by the solar power plant owner of the completed questionnaire to IKNET for further work.
Conclusion by the solar power plant owner of an agreement with IKNET on obtaining data on the annual electricity generation forecasting.
Organization of data transmission for hourly forecasting of electric power generation by a solar power plant to its owner or guaranteed buyer (by agreement).