Strengthening Long-term Forecasting Capabilities for Renewable Energy Generation

As the share of new energy in the power system continues to rise, the interconnection between energy and climate issues has become a prominent focus of global scientific research and engineering practices. Currently, predictions for new energy generation capabilities are primarily concentrated on short-term forecasts, while monthly and annual predictions are lacking. This gap poses significant challenges for the improvement of the electricity market and the planning of power systems.

The Global Energy Interconnection Development and Cooperation Organization (GEIDCO) and the National Climate Center recently held a press conference to release the findings of the “Global New Energy Generation Annual Forecast 2025”. This report emphasizes the importance of annual generation capacity predictions for assessing interannual variations in new energy generation and conducting annual energy balance analyses. It also provides new perspectives for medium- to long-term energy strategic decision-making, energy system planning, and risk management.

Addressing the Long-Term Forecasting Gap

In 2024, the global installed capacity of wind and solar power is expected to reach approximately 300 gigawatts, accounting for about 30% of the world’s total installed capacity. However, various climatic factors have led to a decline in the actual generation capacity of wind and solar energy. For example, while Germany saw a 4% increase in wind power capacity year-on-year, its generation decreased by 2%. In India, excessive rainfall during the monsoon season resulted in a 3% to 10% reduction in solar radiation across many regions.

Liu Zehong, Vice Chairman of GEIDCO, stated that climate significantly impacts all aspects of the power system, including generation, grid, and load. As the contribution of new energy increases, the challenges in balancing the power system also grow. Therefore, there is an urgent need to incorporate climate factors into power system planning and operations, enhance meteorological forecasting for electricity, and improve the climate resilience and safety of power systems to better adapt to the impacts of climate change. “Currently, global new energy generation forecasts focus mainly on short-term and ultra-short-term predictions, with long-term forecasts, especially those on a monthly or annual scale, still absent,” Liu noted.

To accurately gauge the long-term trends in new energy generation, several challenges must be addressed. First, the world should be divided into multiple regions to make precise predictions for wind, solar, temperature, and other meteorological factors for each location over the coming year. Additionally, it is essential to accurately identify the locations and scales of all wind and solar power stations globally. Furthermore, research is needed to develop precise conversion models that link atmospheric dynamics to electricity output.

Xiong Shaoyuan, Deputy Director of the China Meteorological Administration, pointed out that the new power system has an increasing demand for climate resource assessments and warnings regarding extreme weather events. The China Meteorological Administration is accelerating the establishment of an energy meteorological service system, expanding the scope of these services, and enhancing collaboration and innovation within the industry to improve energy meteorological service capabilities.

Currently, GEIDCO and the National Climate Center have initiated technical efforts to conduct medium- and long-term forecasting studies on new energy generation capacity from both energy and meteorological perspectives. Research teams are establishing frameworks for forecasting meteorological elements, identifying wind and solar power sites, and predicting generation output. They aim to create an annual forecast for new energy generation capacity that covers the globe with a spatial resolution of 28 kilometers and a temporal resolution of one month.

Projected Increase in Generation Capacity

This year, the Annual Forecast indicates two major trends in global new energy generation. First, overall generation capacity is expected to see a slight increase year-on-year, although atmospheric responses will differ from previous years, impacting global wind and solar generation capacity. The forecast shows that global wind power generation capacity will remain roughly stable compared to last year, with an average utilization of 2,300 hours. In contrast, solar power generation capacity is expected to increase, averaging 1,350 hours of utilization.

This year’s regional variations in new energy generation will be more pronounced. Under the backdrop of global warming, the changes in the atmospheric system are becoming more complex and unstable. Various climatic factors will lead to significant differences in new energy generation capacity across different countries and regions. For instance, it is anticipated that southern Africa will experience abnormally high precipitation, resulting in a decline in both wind and solar generation capacity. However, certain areas in the north and south may see significant increases in wind power generation capacity due to changing temperature differences between land and sea and the periodic strengthening of the subtropical monsoon system.

Yang Fang, Deputy Director of the Economic and Technical Research Institute at GEIDCO, projected that the average hours of electricity generation for wind power in China will decrease by about 2% compared to last year. Considering the development of new energy, the overall generation capacity is expected to grow by around 20%. This year, the average wind power generation hours nationwide are projected to be approximately 2,100 hours, characterized by higher generation in winter and spring and lower generation in summer and autumn. Months like June, July, and September may see noticeable declines. Provinces such as Ningxia and Shandong are expected to see increases in wind power generation hours of 2% to 8%, whereas provinces such as Henan, Xinjiang, and Hebei may experience declines of 5% to 9%.

Furthermore, the average solar power generation hours in China are expected to increase by about 2.5% compared to last year, with generation capacity projected to grow by about 40%. The national average solar generation hours this year are expected to be around 1,300 hours, with the second half of the year potentially surpassing levels seen during the same period last year. The primary reason for this is that last autumn, many northern regions and coastal areas in the east and south experienced significantly increased rainfall, leading to lower solar generation hours compared to previous years. In regions with high solar installations, provinces like Zhejiang and Hebei are expected to see increases in solar generation hours of approximately 8% and 6%, respectively, while Guizhou and Shandong may see declines of 2% and 1%.

Significance of Forecasting

“The incorporation of climate prediction information into the new energy annual forecasting domain is a significant advancement that provides stable and reliable climate services for the energy industry,” stated Xiao Chan, Deputy Director of the National Climate Center. He emphasized that to better serve the development of new energy, climate predictions need to combine traditional numerical forecasting models with artificial intelligence techniques to enhance the accuracy and timeliness of meteorological forecasts. Future efforts will focus on producing forecasts on various timescales, including monthly, quarterly, and annual predictions, to meet the needs of electricity market annual trading and intra-month scheduling.

“The uncertainties and unpredictability associated with new energy’s participation in power system operations and market operations over long periods are key challenges that need to be addressed,” noted Liu Shuo, Director of the New Energy Department at the Beijing Power Trading Center. He added, “Currently, new energy generation forecasts primarily focus on short-term scales of 7 to 3 days. In market design, the lack of monthly and annual data creates unclear market boundaries. The annual prediction of new energy generation provides more accurate information to define these boundaries. Forecasting annual generation capacity also helps enhance the ability of new energy enterprises and market participants to engage in the market, improving decision-making at the long-cycle level. Considering the current operational cycle of the electricity market, annual trading typically occurs near the end of the year, so it is recommended that annual forecast results be published around October each year to better guide the formulation of new energy generation plans for the following year.”

“Longer prediction time scales, from monthly to quarterly to annual, provide more time for disaster preparedness in the power grid,” said Lu Zongxiang, a professor and Executive Vice Dean of the Tsinghua Sichuan Energy Internet Research Institute. He highlighted that climate is increasingly becoming a significant factor affecting power system disasters. If planning is effectively executed at the annual level, resources can be prepared in advance, significantly enhancing the grid’s disaster prevention and mitigation capabilities. Additionally, expanding forecasting capabilities from short-term to long-term will greatly enhance the controllability of new energy sources, allowing them to better match the demands of the power system. He concluded, “The intersection of meteorology and electricity is a current hot topic, but there is still a lack of communication between the knowledge systems, discourse systems, and technological frameworks of the two fields. Further cross-disciplinary integration requires building a new, comprehensive mechanism that promotes data fusion and innovative technical methodologies simultaneously to establish effective cross-industry collaboration models.”