The following is an exert from Network Magazine, November 2019. Redistributed with permission.
Gordon Brown, managing director at Kelvatek, talks about the importance of innovation to meet challenges in the energy sector – and his enthusiasm for the software capabilities he encountered on joining the company.
What has the industry achieved during RIIOED1?
Since 2015, there has been a lot of emphasis on finding automated solutions at high voltage level and that has led to a significant improvement. We have also made progress on the low voltage network but there is much more that can be done to benefit the end consumers here. As a sector we have been dealing with a balancing act, trying to meet increasing standards for security of supply while dealing with growing complexity as we accommodate new technologies and move to decarbonisation of the power supply. We have taken a step forward but there is more that can be done.
What are the key challenges for RIIO-ED2?
As customer behaviour changes and the industry interacts with individuals in more ways, investment in the low voltage network is going to be important. Electric vehicles and low carbon technologies are being applied at customer level and these affect the low voltage network, where there just isn’t the same redundancy built in as there is at high voltage level. We are seeing a major shift in investment plans for ED2 to focus on automating at low voltage level and predicting where problems will arise. We need to find ways of predicting faults before they occur.
How have the Network Innovation Allowance and the Network Innovation Competition aided industry progression?
During ED1, centrally coordinated investment in innovation through the NIA and the NIC have been crucial. We know requirements will evolve and we have to try out new things to keep up. At Kelvatek we invest significantly in research and development but we also rely on working with other parties through collaborative projects. NIA and NIC funding have been crucial to help us help the industry. For example our Weezap low voltage circuit breaker started as an innovation project and is now used by four of the six network operators.
What role can artificial intelligence play in meeting future network challenges?
We are working on an artificial intelligence and machine learning project at the moment, looking to see early signs of fault activity on low voltage cables. We are also looking at high voltage cables using NIA funding, and the ability to see dangerous lowered lines. We are starting to use advanced algorithms to analyse data and look for very early signs of pre-fault activity.
Kelvatek has an excellent reputation for fault management but I joined the company in February and the biggest surprise to me was its existing software capability. Our artificial intelligence and machine learning centre of excellence in Parma, Italy, is second to none in its ability to determine from complex datasets what is happening and predict what will happen next.
We offer services to the industry and, as part of a change I’ve introduced, we will be offering our capabilities directly as a software solution too. Customers can then decide whether they want to build their own in-house data analysis team or use our centralised service.