PV generation forecasting tool to save UK money

Researchers at the University of Sheffield have developed a service to forecast energy generation levels from PV up to three days in advance.
Published: Tue 22 Aug 2017

Sheffield Solar, based at the University of Sheffield’s Department of Physics, has launched a service which provides forecasts for energy generation from photovoltaic (PV) systems in the UK for up to 72 hours ahead.

While the research unit is making the tool available for free for a trial period, it is requesting feedback from traders and generators so that the product can be further refined.

Balance and optimise

The service builds on two years’ work with National Grid and is aimed at assisting traders, generators and grid operators to balance and optimise the electric system and portfolios. By providing a clearer picture of how much power is likely to be generated from the UK’s 12GW of solar PV capacity, total system costs can be reduced. This could result in reduced costs for all stakeholders, including the 30m UK households.

The improved data will provide a more accurate forecasting which will help generators and traders reduce the penalties they face for getting their supply/demand balance wrong. These costs are often carried by their customers. The solution could also help the UK’s National Grid to spend less on procuring balancing services by incentivising consumers to adjust their consumption. This could mean lower bills for the utility as well its customers.

Accurate estimates to improve efficiency

National Grid already uses Sheffield’s real time PV forecasting tool, but the new service will enable them, and other stakeholders in the industry, to plan ahead by combining live output data with weather forecasts. Through this new solution, it shouldbe possible to accurately estimate future output.

Updated frequently, it combines weather forecast data with the data from live generating systems to provide a forecast for the next 72 hours.

The service is currently being trialled on the University of Sheffield’s ’Sheffield Solar’ website. The group has plans to develop the service with researchers initially releasing a half hourly forecast. This will be followed by a regional forecast and finally they intend to provide forecasts for individual systems around the UK.

Business development manager Aldous Everard says that the service has the potential to go beyond half hourly data and into more granular detail.

With growing penetration of renewables, the need for such a tool should not be underestimated. In Vermont, Vermont Electric Power Company has demonstrated improved network management and renewables dispatchability with a hyperlocal (1kmx1km) weather forecasting model.


Related Webinar