Issue link: https://fhpublishing.uberflip.com/i/808182
NETWORK / 28 / APRIL 2017 W hen National Grid's dis- cussions with technol- ogy provider DeepMind hit the national head- lines earlier this year, it became clear that Arti- ficial Intelligence (AI) and cognitive com- puting have the potential to transform the energy industry. AI technology's ability to crunch huge amounts of data promises, at least on paper, to help networks in the tran- sition to a low-carbon future, while offering the opportunity to make the distribution of energy, cheaper, better and more manage- able. Techniques developed in current AI tech- nology will inform and have an increas- ing role to play in enabling the stability of next-generation power grids by selecting and using control algorithms for individual assets to perform the best under specific sce- narios and conditions on the grid. Distribution networks and transmission operators such as National Grid are already investigating AI technology's ability to bal- ance the flow of energy by predicting when peaks in generation will occur, and allow problems to be solved in real time. As demand response continues to emerge as a key technology there will be an increased need for smart control strategies that enable distributed energy storage assets to perform their desired functions. Here are five ways cognitive technology applications could help networks meet the demands of the future. 1. Predicting demand DeepMind is a British technology company bought by Google in 2014. It uses a system of neural networks trained on different oper- ating scenarios and parameters within the operating framework. It has the potential to allow a system oper- ator to develop a more efficient and adaptive framework to understand and adapt to fluc- tuating renewable energy generation and optimise efficiency. A DeepMind spokesperson said there was a huge potential for predictive machine learning technology to help energy systems reduce their environmental impact. National Grid is working with DeepMind to explore how the technology could accu- rately predict demand patterns across the UK to ensure it makes the best use of renewable energy, and help save money for bill payers. 2. Managing big data Finland's main electricity transmission grid operator, Fingrid, has used AI technology to infuse its network with intelligence, making it more resilient, reliable and secure. ArtificiAl intelligence Fingrid used the information gathered from AI to combine near real-time big data analytics with external factors, such as weather, to increase the reliability of the grid. Using networks of sensors and IBM's advanced analytics, Fingrid pioneered a solution called ELVIS (Electricity Verkko Information System), which provides sys- tem operators with a consolidated view of the entire electricity transmission grid, from long-term plans to the day-to-day manage- ment and maintenance of assets and infra- structure. Previously Fingrid collected data from disparate systems and databases manually, which could take days or even weeks to col- late for root cause analysis. The system's modules were connected to one another in a flexible manner ensuring the integrity of data moving between them. The ELVIS solution combines IBM Max- imo so–ware with other mission-critical sys- tems, including ArcGIS mapping so–ware from ESRI, Primavera Enterprise Project Port- folio Management from Oracle, and, for the mobile-user interface, SAP Work Manager. 3. Adding value with algorithms UK start-up Upside Energy, in partnership with Heriot-Watt University in Edinburgh, will use machine learning and AI methods to manage a portfolio of energy storage assets to provide real-time energy reserves to the grid. Machine learning will be fundamental to how Upside evolves its Energy Advanced Algorithmic Platform and delivers growing value to the energy system and wider society. The company has been awarded grant funding to develop AI for its innovative cloud-based Virtual Energy Store and a knowledge transfer partnership (KTP) grant by Innovate UK to maximise the opportu- nities presented by the emerging energy demand response market. In 2015, Upside was part of a consortium that secured Innovate UK funding to develop the Virtual Energy Store, which co-ordinates energy stored in existing assets including uninterruptible power supplies (UPS), bat- tery, solar PV systems and electric vehicles. This store can be used by the grid as a back-up power source, with the potential to reduce the UK's reliance on traditional power stations. Upside aims to sell a range of balancing services to National Grid, and in the longer term sell similar services to distribution net- work operators and energy suppliers. 4. Capturing expert knowledge National Grid is looking to the latest advances in technology to test how cognitive technology can capture the expert knowl- edge, expertise and experience of its engi- neers to help make faster, more informed decisions. National Grid Gas Transmission (NGGT) announced at the beginning of the year it had launched a six-month project using IBM's Watson cognitive so–ware technology. The project will use IBM's Watson to test how cognitive technology can be used to capture expert knowledge from NGGT's ageing workforce and reduce its reliance on technical experts for common enquiries. It is hoped the technology will help National Grid make better use of its diverse sources of knowledge, information and capabilities and help capture the years of skills and insights from experienced engi- neers to enable the next generation to apply this knowledge when making decisions in the future. Fingrid uses the information gathered from AI to combine near real-time big data analytics with external factors to increase the reliability of the grid.