Riventa AI solar aim to offset electricity use on pumping stations

An artificial intelligence (AI)-based approach to maximise solar photovoltaic system (PV) system efficiency for water pumping stations has won Riventa a place as a finalist in the £4M Water Discovery Challenge competition. A two-year event that brings together financial and non-financial support to help solutions launch and succeed in the water sector, Riventa’s method of offsetting electricity use on pumping stations by installation of solar PV systems is one of 20 new solutions to be chosen.

To help offset electricity use on pumping stations, Riventa is developing an AI-based approach to maximise solar efficiency

Encouraging innovation that can help solve challenges facing the water sector, Discovery aims to facilitate engagement for suppliers with water companies, backed by a £200M Innovation Fund that has been established by Ofwat, the water services regulation authority for England and Wales.

The 20 finalists will initially share £1M in funding, and then, up to 10 companies will go on to be awarded up to £450,000 each as well as additional support to further develop and test their ideas.

Speaking for Riventa, Dr Tom Clifford, who has worldwide experience in optimising the performance of pumps, blowers and turbines, said, “Water Companies want to offset electricity use on pumping stations by installing solar PV systems, so that they can maximise the internal rate of return (IRR). However, pumping networks are subject to multiple tariffs throughout the day, varying in cost where the current lowest charges are during the night, with the network also containing water storage to decouple supply and demand. The receipt of solar energy will be at zero charge throughout the ‘sunshine’ hours, thus altering the possible optimum pumping schedule by effective use of the storage.”

He added, “The technical challenge is to determine how the pumping schedule will change as PV install capacity varies, at each of the sites.  It is complex to find the solution because the pumping stations are linked in the network, meaning any change in load profile will impact the operating schedule at other sites. Varying array size will also alter solar capture, changing the daily tariff structure, requiring re-optimisation of the pump schedule. Solar capture will vary throughout the year, impacting the IRR result by array size selections. [As] all three points are linked, the multifaceted problem will be optimised by an AI algorithm. Over a 25 years lifespan, a design will be calculated [to] achieve the greatest discounted net present value (NPV).”