SYNFOS-wind, Wind Power Generation Output Forecasting Service
Services

Providing high-accuracy wind power generation output forecasts by integrating multiple meteorological models, AI technology, and advanced analytical techniques.

Wind farms are expanding throughout Japan and globally as renewable energy becomes the more sought-after energy resource. Over the last decade we have also seen a shift in the wind power industry, with offshore wind farms gaining popularity and are expected to expand even farther than onshore wind farms.
In 2022, with Feed-in Premium (FIP) succeeding Feed-in Tariff (FIT), renewable electricity producers are now required to write up and submit a daily estimate of their expected energy generation and compare them to actual output values. However, wind energy, which depends entirely on natural wind patterns and often influenced by surrounding terrain, is frequently considered an unstable energy source, making accurate daily energy estimation a complex challenge.
This is where meteorological predictions and forecasts play a crucial role in developing precise energy generation plans.
SYNFOS-wind provide the wind energy industry with high quality wind speed and direction forecasts and highly accurate wind power output predictions.
Utilising multiple weather models, both domestic and international, and applying JWA’s advanced analytical technologies and AI technologies, SYNFOS-wind delivers highly accurate, comprehensive forecasting solutions. Our service supports all types of wind farms by providing essential weather forecasting services tailored to their needs.

● Integrated Forecasting Using Multiple Weather Models

By combining Japan Weather Association’s proprietary meteorological model, SYNFOS, with multiple domestic and international weather models, we have significantly improved forecasting accuracy compared to using a single meteorological model.
Then by adjusting the combination and integration methods of these meteorological models based on forecasting conditions, we provide high-precision forecasts tailored to specific locations and purposes.

● Improving Forecast Accuracy with AI Technology

To meet the need for more advanced forecasting accuracy, JWA has integrated AI technologies, such as surface-layer wind correction for specific wind locations, into its proprietary wind speed and wind power output forecasting models. For forecast accuracy validation on existing wind farms, SYNFOS-wind demonstrated a 10% improvement in annual MAE (Mean Absolute Error) compared to pre-upgrade levels.

●Prediction Equations Using Historical Performance Data

Wind turbines rely solely on natural wind to rotate, and the flow is easily affected by surrounding terrain and structures, making wind characteristics are among the most challenging weather components to predict in the energy industry, especially when compared to factors like solar radiation.
However, by providing historical data on wind speed, wind direction, and wind power output, we can develop location-specific prediction models and incorporate actual performance data for each wind turbine. This also enables forecasting to take the potential wake effects caused by surrounding turbines and structures into account.

* Wind turbines will disturb the natural wind flow, and the wake effect is the trail left behind each turbine where the flow is disturbed, and wind speeds are reduced. On wind farms this can influence on the energy production depending on the distance and angle the turbines are place according to each other.

●Generated Power Output Conversion

By applying the power curve, which represents the relationship between wind speed and wind power generation output, we convert wind speed data into wind power output values. Additionally, by using a performance-based power curve derived from historical wind speed and power output data, we can more accurately reflect the operational status of wind turbines.

●Correction Ratio for Wind Turbine Operational Disruptions

When a portion of a wind farm ceases operation due to wind turbine failure or maintenance, the wind power output decreases in proportion to wind speed.
This discrepancy between forecasted and actual output can result in a persistent overestimation trend.
To address this, we incorporate wind turbine availability data into the forecast, enabling adjusted predictions that accurately reflect the operational status of the wind turbines.

●Real-Time Sequential Correction Using On-Site Data

By integrating real-time wind direction, wind speed, and wind power output data measured at the forecast target location, we can continuously adjust and refine the forecast values (sequential correction).
The effectiveness of sequential correction is particularly significant for forecasts up to several hours ahead, making it highly beneficial for short-term market utilization and battery storage management within power generation facilities, where immediate forecast accuracy is critical.

Service

The wind power output forecasting service offers the following features.
We provide customized forecast information based on specific needs, including frequently updated short-term forecasts, daily forecasts, weekly forecasts, and two-week forecasts.

Features

• Integrated forecasting utilising multiple domestic and international meteorological models including JWA’s proprietary SYNFOS model.
• Enhancing forecast accuracy by incorporating AI-powered models and various analytical techniques such as wind turbine availability adjustment and sequential correction.
• Providing tailored forecast solutions, including frequently updated short-term, weekly, and two-week forecasts based on operational needs.
• Applicable to both onshore and offshore wind power, supporting the growing offshore wind energy sector.

Applicable for:

• General power transmission and distribution companies in creating power supply and demand plans.
• Retail electricity companies in creating economical power procurement plans.
• Wind power companies, power generation balancing groups, aggregators, etc. in creating highly accurate power generation and sales plans.
• Achieving RE100 through self-consignment using wind power.
• Output smoothing at wind power facilities equipped with storage batteries.
• Providing preliminary assessments for transitioning to the FIP system or entering the offshore wind power industry.

FAQ

1. Is it possible to prepare historical forecast data and conduct forecast accuracy verification for preliminary assessments?
>Yes, this is possible. We can also prepare reports, please contact us for more details.

2. What are the specifications of wind power generation equipment required for power output conversion?
> This refers to location information (latitude and longitude), rated output, wind turbine hub height, catalogue value power curve values. Please contact us for more details.

3. Can a forecast model be set up even if historical performance data is unavailable?
>Yes, it is possible to set up a forecast equation for a wind farm prior to the start of operation by using wind direction and wind speed observation data from the surrounding area and JWA’s estimated wind speed. Additionally, forecast models can be developed using only wind power output performance data.

4.Will AI technology always be used regardless of the target location or forecast conditions?
>Depending on the quality and period of the historical performance data used to set the prediction formula, we may propose forecasting methods that do not utilize AI technology. In addition to SYNFOS-wind, we also strive to further improve non-AI-based forecasting methods as well.

5. Can forecast information be provided for overseas wind farms?
>Yes, we provide forecasts for international wind farms. Please contact us for further details.

6. Is any specific software or hardware required to implement the service?
>If the service is provided via WebAPI or email distribution, it can be received through internet connection. If a dedicated data reception server is available, we can integrate data into that server as well.

7. What is the price?
＞We can provide you with an estimate according to the number of forecast locations, forecast type, required elements and other eventual specifications. Please contact us directly for a price quote.

8. Can this service be purchased for personal use?
> Unfortunately, this service is available for corporate customers only.

*”SYNFOS” is a registered trademark of Japan Weather Association (Registration No. 4935276).

 

Related news

Japan Weather Association launches “SYNFOS-wind,” a wind power output forecasting service that improves forecast accuracy by 10% through the introduction of new AI technology (Japan Weather Association news release, January 16, 2024)
https://www.jwa.or.jp/news/2024/01/22153/