How weather data and operational intelligence are helping energy producers to anticipate and deliver electricity. By its very nature, renewable energy is highly weather-dependent, and the ongoing expansion of renewables is making our global power supply more vulnerable to changing weather conditions.
Countries worldwide are looking to increase their use of renewable energy. Just last year, land-based wind energy accounted for more than a third of all new US generation capacity and the Department of Energy anticipates building a $70 billion offshore wind business pipeline by 2030.
The Clean Energy for all Europeans package sets an ambitious goal of 32% of energy coming from renewable energy sources by 2030. This growth in renewable energies is driving an increased interest in the most accurate and reliable weather data, not just in short-term forecasts, but also long-range and anticipated energy output.
Weather will play a major role in every phase of the renewable energy lifecycle – from site selection, to installation, optimising energy generation, managing supply and demand, routine maintenance and decommissioning sites.
Maintaining a steady flow of energy from nature is very different from using a fuel source where there is more control, like coal or natural gas. With alternative energy, the goal for utility companies is to compensate for source fluctuations, rather than control the flow of the resource. Many utilities depend on private weather companies that use an ensemble of weather models, statistical forecasting and filtering techniques to gain keen insight on the amount of potential power.
Predicting winds for energy use poses a number of challenges. The geography around a wind farm can have a major influence on local wind speeds, so traditional weather prediction models are not useful for accurate
wind forecasts. Wind farms operate at peak efficiency when using hub-height wind forecasts at turbine-level to measure wind speed, direction and gusts.
Solar farms face similar challenges in reliance on accurate forecasts. Direct normal irradiance (DNI) is a key variable that determines the performance of the plant. Wind speed, humidity, and ambient temperature affect the level of thermal losses from receivers, as well as the efficiency of the steam turbines installed at the power block to generate electrical energy.
As these measurements can fluctuate among turbines or panels in a single farm, using data collected on-site using advanced weather stations improves forecast accuracy and maximises performance. For solar farms, the weather station includes an accurate solar tracking system to measure DNI, as well as two pyranometers to measure accurately the global solar and horizontal irradiances.
With the advancement of weather technology, along with site-based information and using an ensemble of weather models, including proprietary and industry-specific models, private weather companies can help operators to maximise power output and better support reliability of service.
Risk to reputation
Accurate weather data and forecasting is critical even before a renewable energy farm is developed. Solar farm developers will analyse the historical solar trends for multiple potential farm locations in order to pick the best location for development. And the more accurate the wind measurements
are for a proposed site, the more likely the wind farm is to deliver on its expected energy yield. Even small miscalculations can have long term effects on the valuation of company stock and profi tability.
Government regulations and utility standards also impact profit by specifying a small error variance from the predicted to the actual amount of energy supply and demand. If actual supply falls outside the variance, the utility company can face financial penalties, lose profitability and have reputational damage.
Some states are proposing legislation to penalise utilities for large outages, even when caused by weather events. To guarantee the stability of a network, it’s important for supply and demand to be balanced.
Weather-informed operational intelligence enables utilities to determine which power plants they need to have available during different parts of the day as both demand and supply from renewable sources change with changing weather conditions.
There is no doubt that the frequency and intensity of extreme weather events continues to increase, which has a direct impact on power plants, but also weather parameters like wind have changed in behaviour.
A study published in the journal Nature reveals global wind speeds since 2010 have recovered to levels last seen in the 1980s. Essentially more wind means more opportunities for power generation, as long as it doesn’t reach safety levels.
Wind and hail forecasts play a big role in maintenance of photo-voltaic (PV) systems. Most systems are tested and certified to withstand hail of up to 1″ (2.54cm) falling at approximately 50mph (80km/h), and to withstand winds of up to approximately 140mph (225km/h).
And, while systems are designed for durability during extreme weather such as hail, high and low temperatures, humidity and solar ultraviolet radiation, awareness of impending weather events can help crews be prepared for potentially damaging weather.
In the case of high winds at land-based wind farms, it may lead to a sudden shut down of the whole farm. For offshore wind farms, strong winds plus another notable change, increased wave heights, threaten output
and operational safety.
These occurrences of extreme weather impact the supply to the grid. Unless generation reserves are ready to take over the power demands under short notice, shutdowns can leave national networks in short supply, meaning the pressure is on grid operators to understand when this will occur and prepare.
Safety and maintenance
Accurate forecasts can also help with downtime and risk assessment. Both weather and risk are data-driven, and by layering information, the level of risk can be reduced significantly. Many components of a weather forecast can be given numerical values, which makes it a useful tool for considering risk. Warning thresholds can be set on forecast parameters, allowing site managers to assign a level of risk to any particular part of a project, which gives risk assessors useful information on potential downtime.
Hindcasting is a useful tool too, as it can give planners the ability of knowing what the best time of year is to conduct routine maintenance or new builds.
Forecasting can help better manage potential damage and repairs to equipment as well, which can have a significant impact on the bottom line; for example, solar PV modules cost more than $300 per kW to replace if damaged. And annual blade inspections to proactively repair minor damage offer significant savings compared to traditional practices of inspecting blades once every three years and reactively addressing more major repairs.
Unpredictability of now COVID-19 has permeated every aspect of life as we know it, and energy demand is no exception. Commercial and industrial demand account for the largest consumption of energy. The National Grid Electricity System Operator predicts the pandemic will suppress peak demand as many people are working from home or in some state of lockdown.
Two other factors contributing to forecasting demand are the long-range winter forecasts for below average cold in several parts of the globe and the Internet of Things, as these can alter consumption behaviour.
As renewable energy companies strive to maintain a steady source of energy, unpredictable fluctuations in supply from those sources is challenging. Weather data and operational intelligence are critical to helping energy producers to confidently anticipate and deliver energy.
Renewable energy may be highly weather-dependent, but with advanced weather models and technology, energy producers can depend on weather data to make confident operational decisions.
About the author
Mike Eilts is the senior vice-president at business intelligence firm DTN, responsible for bringing weather solutions to customers around the world. Prior to DTN, he founded a private weather technology company and held leadership positions within the National Severe Storms Laboratory. He has written more than 75 papers in meteorological journals and conference proceedings and is an American Meteorological Society Fellow. He has an MBA as well as Master and Bachelor of Science degrees in Meteorology, all from the University of Oklahoma.