Solar Power Forecasting
SolarFor is a software solution for solar power forecasting. SolarFor delivers highly accurate predictions of solar power production for the operational horizon (ranging from a few minutes ahead in time, up to a couple of weeks). SolarFor is very flexible and has a long track record of producing accurate forecasts.
Why Do You Need SolarFor?
Intermittent power production from renewables, like solar farms, has transformed the electricity sector in many countries. Solar power forecasting is a necessity in such markets in order to plan accurately and operate the power system efficiently. This applies to both commercial players in liberalized power markets and system operators who need to understand the impact of renewable energy production in their portfolio and on the electricity system as a whole.
Accurate solar power forecasting is needed by asset owners and electricity traders in order to nominate and trade the power production efficiently. By increasing forecast accuracy, asset owners and traders can reduce costs of imbalance fees and penalties.
Transmission system operators need accurate solar power forecasts to maintain system stability. Intermittent power production from solar power can cause system instability and increase the cost of balancing the electricity system (e.g. from standby capacity). In order to increase system stability, reduce cost and minimize curtailment of solar power, it is essential for the system operator to be able to plan and manage the total power production based on accurate solar power forecast.
The highly accurate solar power forecasts delivered by SolarFor help asset owners, traders and system operators around the world to manage and optimize their portfolio every day.
- Market leading solar power forecast accuracy
- Proven operational track record
- Reliable, stable and with high availability with a proven track record from clients requiring availability of 99,9% and above
- Robust, low maintenance system with minimal interference for the client
- Highly flexible and configurable to almost any conditions. SolarFor has been deployed and used over a large geographic area
- Supports asset owners and traders optimize their power production portfolio for accurate nomination and trading of electricity, minimizing imbalance fees and penalties
- Supports system operators manage overall electricity grid, increase system stability and minimize cost of balancing the system and reserve cost
- Self-learning algorithms which continuously adapt and re-configure power forecasts List of special tailored modules available, which can be deployed to handle complex requirements for specific and challenging solar farms or complex regulatory or commercial requirements in specific markets
- Efficient error handling, fall-back procedures and issue warning system
- Comparisons between a large number of solar power forecasting services have shown that SolarFor™ delivers very accurate state-of-the-art predictions, making it the preferred choice for customers.
How Does SolarFor Work?
SolarFor is a self-learning and self-calibrating software system based on a combination of physical models and advanced machine learning. This combines the best of artificial intelligence with solar power domain knowledge in order to produce the most accurate solar power forecasts available.
SolarFor is initialized using historical weather and production data to train the models or relevant data describing the power curve from the design of the solar farm.
After initialization, forecasts are produced every time the system receives new data, which can be either updated weather forecasts or new production data. SolarFor can run in either online mode and continuously receive real-time production data or in off-line mode where historical data are retrieved monthly, or any other time interval. By integrating SolarFor directly with the SCADA system and thereby providing real-time production data, very accurate short-term forecasts can be achieved.
Weather forecasts (incl. ensemble forecasts) are typically supplied as an integrated part of the solution. The system can use one or more weather forecast providers as input and automatically detects the optimal prioritization of the different weather forecasts for each solar farm and for different forecast horizons. Weather data are also made available to the client such that the client can compare power forecasts and weather inputs.
The self-learning and self-calibrating algorithms will continuously learn about the solar farm characteristics and will adapt to changing conditions, seasonal variations, and as the photovoltaic module ages, such that forecasts stay accurate over time without the need for manual configuration.
By delivering schedules of availability and curtailments, the client can make sure that SolarFor takes these planned events into consideration. Furthermore, if the client provides real time availability and curtailment information from the solar farm, the data will be used to train models and adjust short term forecasts which significantly increase accuracy.
SolarFor can deliver power forecasts in almost any file format and can be integrated directly into the operational IT-platform of the client, such that data are retrieved and delivered seamlessly to and from relevant systems.
SolarFor is available as a software package installed locally on the client’s servers, or as a service hosted on servers operated and maintained by ENFOR.
SolarFor is supplied with various support, maintenance and license packages, which can be tailor-made to client specifications to provide a cost/performance ratio which fits the needs of the individual client.
The following list of key features are provided with the standard SolarFor module:
- Integrates with weather forecast from all major weather forecast providers
- Data flow and input validation automatically issue warnings in the event of errors
- Efficient fall-back procedures and estimation of substitute values in case of errors or missing input values
- Highly configurable browser-based graphical user interface for easy access and display of all relevant information
- On-line mode or off-line mode, using either online production data or historical production data
- Configurable forecast horizons. Furthermore; different weather forecasts providers can be used for different forecast horizons for optimal accuracy
- Configurable time resolution in forecasts. Different intervals can be defined for short term intraday forecasts and day-head and week-ahead forecasts
- Fixed, single and dual axis tracking
- Limitations imposed by inverter setup
- Temperature dependent panel efficiency
- Calculation and tracking of maximal (clear sky) production
- Configurable forecast update frequency
- Configurable performance reports to monitor and track system performance
- Data integration interfaces based on FTP, SFTP or Web Services supporting numerous formats and file types (CSV, XML, SOAP, JSON etc.)
Key Features Provided from Special Modules
- Module for forecasting of uncertainty bands (quantiles) which can be used for trading/bidding strategies and risk assessment
- Module for forecast scenario generation
- Support for multiple weather forecast providers/ numerical weather predictions(NWP)
- Combination module. Combines multiple internal forecasts (based on different weather forecasts) and/or external forecasts. Calculates optimal weighing of individual forecasts, and produce high accuracy combined forecast
- Upscaling module for using online measurement from some solar farms improve forecast for other solar farms without on-line measurements
- Ensemble weather forecast module. Use ensemble forecast as input for improving forecast accuracy on both short term and long-term horizons
- High resolution forecasting module. Forecasting of time resolution of 5 minutes or less
- Automatic shadow detection
- NWP error correction models based on satellite data and online measurements
Wind power prediction for wind farms
Electricity load prediction of electrical power systems
Load forecasting, temperature control, and optimization of district heating networks