A sustainable energy future won’t happen without AI
Steve Kwan, Director of Product Management for Power Generation and Grid Management, Beyond Limits, shares his insights on the role of AI technologies in the future of renewables and how providers are working towards more sustainable practices
When a utility constructs a new power plant facility, they design it with specific operational goals in mind. These design goals in turn also affect the initial equipment chosen for the site, as well as their configuration, construction and operational decisions. Much of the energy infrastructure in place today was designed with traditional fossil fuels in mind for their operations, but that’s quickly changing. Calls for change inspired by data such as the UN’s recent report on climate change and new regulations have prompted strong interest in renewable energy technologies, with global investments expected to reach $3.4 trillion by 2030.
As more renewables become available, there is less power needed from traditional generation units. Existing power plants are increasingly being asked to alter their unit operation so that the total supply on the grid matches demand at the lowest total cost. However, when these units are asked to operate in these unfamiliar operating regions, there’s a lack of historical operating data and operator experience. This can lead to suboptimal operations, typically with high levels of reserve which can be wasteful.
That’s where AI technologies can help. One example of this is Cognitive AI, which combines human expertise and intuition with numerical AI models, running many millions of simulations in a matter of hours to provide users with a better understanding of the capabilities of their generation units under all conditions based on current and historical operating data. These AI solutions can also provide prescriptive recommendations for the operators to optimize their operations amid fluctuating equipment and energy demands before a scenario arises, so teams are equipped with the knowledge and resources required to avoid potential disruptions.
Another factor that operators need to consider during this transition are the potential fluctuations in energy available from renewable resources like solar or wind depending on the regional weather. That means utilities need to be able to look ahead and plan for future events like weather patterns which may change usage demand or affect renewable generation, anticipated plant maintenance outages, or even localized disturbances. Practically this means we need accurate demand forecasts and corresponding generation planning all done with the least cost while maximizing renewable usage and minimizing carbon emissions.
A Cognitive AI solution as described above can be used to run millions of scenarios and provide accurate and optimized recommendations for the operators to predict demand, availability and maintenance up to three or more days in advance, helping to minimize disruptions and keep costs low while still facilitating the shift to renewables.
