Ever since the Babylonians began to predict weather from cloud patterns and astrology, humans have tried to build a solid method to predict short-term, and increasingly of late, longer-term weather patterns.
The science of weather forecasting dates back to the careers of Francis Beaufort, an Irish hydrographer who served as Britain’s Hydrographer of the Navy in the mid-19th century, and his contemporary Robert FitzRoy, chief of weather data at the country’s Board of Trade.
Their scientific rigor and interest in all the earth sciences led them to be called the fathers of modern meteorology, though by the early 20th century advances in atmospheric physics enabled greater accuracy. By 1955 computers were able to produce practical forecasts.
Historically, there has always been a relevant role in weather forecasting in our energy industry. Temperature changes signal increases or drops in heating or cooling demand, as businesses and consumers turn to air conditioners or space heaters.
Similarly, rainfall forecasts are critical in managing hydropower generation assets: by building up databases of historical rainfall patterns and combining them with current weather forecasts, asset managers are able to determine how to optimize their generation.
The first fossil-fuel-fired power plant (Thomas Edison’s plant in London) and the first hydro dam (Fox River, Wisconsin) both began operating in the same year of 1882. Obviously, these pioneers did not enjoy the benefit of accurate weather forecasts and so “brown-outs” were a common phenomenon when demand overwhelmed the available supply.
The advent of scientific weather forecasting made it possible to manage supply to meet fluctuations in demand when weather conditions change. But it’s really in the last 30 years that technological advances have made possible new areas of weather forecasting. With the rise of solar and wind generation has emerged a demand for predictions of how much sunshine and how much wind are to come over different periods into the future.
Typical weather data now available includes forecasts for minimum and maximum temperatures, wind speed and direction, rainfall and “net radiation” (sunlight), all of which carry great relevance for the energy sector.
As an example, the European Centre for Medium-Range Weather Forecasts (ECMWF) and the Swedish Meteorological and Hydrological Institute (SMHI) provide forecasts for wind and solar power generation by the country for the next 12, 24 and 36 hours, and compares its forecasts with observed generation.
Weather forecasts act as a sort of risk management tool for renewable asset managers by helping to reduce uncertainty over the availability of primary energy sources (wind, sun, and rain). By careful analysis of forecasts, plant operators can decide how much energy they might have available to sell over a set period, enabling them to make the best use of assets.
It’s even possible to take risk management one step further now and buy or sell weather derivatives as a hedge against the loss of generation. A typical weather derivative is based on a specific index that can measure any aspect of the weather, be it rainfall, wind or hours of sunshine.
The upsurge in demand for accurate weather data – both forecasts and historical data – has led to a proliferation in the supply of market-relevant, actionable information from a huge variety of sources. Whereas in years past market participants may have relied on established, often state-owned providers, the last 20 years have seen numerous organizations enter the field.
This offers utility operators a vast choice of data to choose from, to compare and even to integrate into their operations. Combined with other risk management tools, it means that the new range of (interruptible) renewable power generation assets can now integrate seamlessly with legacy plants to offer the security of supply on a scale nobody could have expected 30 years ago.