UPDATED 1 Sept: The EI library in London is temporarily closed to the public, as a precautionary measure in light of the ongoing COVID-19 situation. The Knowledge Service will still be answering email queries via email , or via live chats during working hours (09:15-17:00 GMT). Our e-library is always open for members here: eLibrary , for full-text access to over 200 e-books and millions of articles. Thank you for your patience.
New Energy World
New Energy World embraces the whole energy industry as it connects and converges to address the decarbonisation challenge. It covers progress being made across the industry, from the dynamics under way to reduce emissions in oil and gas, through improvements to the efficiency of energy conversion and use, to cutting-edge initiatives in renewable and low carbon technologies.
AI and renewables: What’s the potential and reality?
18/10/2023
6 min read
Feature
Artificial intelligence (AI) and machine learning have the potential to support, optimise and accelerate the development of renewables in the energy transition. New Energy World Features Editor Brian Davis offers an insight into a raft of AI applications – and also road tests ChatGPT.
AI offers an Aladdin’s cave of potential applications for optimisation of renewable energy, while also accelerating the transition from conventional fossil fuels and centralised electricity grid systems to a highly decentralised model.
However, reports by the likes of the World Economic Forum (WEF), Bloomberg NEF (BNEF), the US National Renewable Energy Laboratory, International Energy Association and others suggest that despite the promise of AI, it may be limited to pilot projects for predictive maintenance, cost reduction, recommendations for decarbonisation, and energy trading, at least in the short-term.
Nevertheless, as digitalisation and decarbonisation gain momentum, AI will have an important role to play in renewable power generation, distribution and demand, transport and construction. Indeed, as the complexity of managing power systems increases, there will be an inherent need to use AI for advanced analytics, supported by vast amounts of real-time data and automation.
Significant benefits
The benefits are clear. An AI report by the WEF and BNEF on Harnessing AI to accelerate the energy transition maintains that AI algorithms and models will extract insights from large datasets, identify patterns and predict possibilities of potential outcomes of complex, multi-variate situations. AI will be widely adopted across the energy value chain, creating new revenue streams and business models.
Reducing costs through predictive maintenance is a popular first step on the journey, for example. ‘Reducing required investment [eg for maintenance] or shaving energy demand by even a small percentage drives billions of dollars of savings,’ notes the WEF/BNEF report. The analysts forecast that AI could save operators $188bn in power grid equipment replacement costs between now and 2050.
Renewable energy companies including Vestas, E.ON, Duke Energy, Enel, EDF, Iberdrola, Exelon, Schneider Electric, National Grid and SSE Renewables are using AI for predictive maintenance, project analysis, and much else besides.
For example, E.ON utilises self-learning algorithms to predict when medium voltage cables in the grid need to be replaced.
GlobalData estimates that the global market for AI platforms in the power industry will reach about $5.3bn in 2024.
Energy demand forecasting
Forecasting power demand is complex, particularly with the move towards decentralised distribution. Smart grids with AI algorithms can predict energy demand more accurately to help operators balance supply and demand, reducing the need for back-up fossil fuel power generation during peak periods.
Using AI to enable greater energy efficiency and helping flex demand, BNEF estimates that every 1% of demand-side efficiency between now and 2050, could result in $1.3tn of savings.
Major grid operators increasingly use AI to manage the integration of renewable energy sources in a smart grid to predict and balance supply and demand. Research shows that this ‘smart approach’ to grid management can reduce outages by 30% compared to conventional inspection.
Weather forecasting, wind and solar PV efficiency
On the generation side, forecasting wind and solar PV (photovoltaic) power output can use AI with historical weather forecast data and real-time data from sensors and satellite images to optimise production and generation.
SunPower and other solar farm operators use AI to track the sun’s movement and adjust the orientation of solar panels to capture maximum sunlight.
AI-supported systems can also position wind turbines to maximise output while minimising wear and tear. GE, Siemens Gamesa and others use AI to help position turbines and operate wind farms.
Energy storage optimisation
There are also significant benefits to be gained from energy storage optimisation. Tesla’s Powerpack and Powerwall energy storage solutions, for example, use AI to optimise battery efficiency during peak demand or grid outages.
Clean Energy Associates uses AI for resource and environmental impact assessment of solar and wind farms.
Wind farm developer SSE Renewables has trialled a technique for counting puffins on the Isle of May off the coast of Scotland, using AI and image recognition. The project was supported by Microsoft, Avanade and nature reserve NatureScot, and can be used for environmental assessment in new wind projects on and offshore.
Energy trading revolution
AI also has the potential to revolutionise the way renewable energy is traded, improving the efficiency of transactions, for better planning and resource allocation. AI algorithms, such as Power Ledger and Clarity AI, are blockchain-enabled platforms for trading renewable energy, used by hedge funds and traders for analysis of energy markets.
Energy, supply chain and carbon management
There are energy management systems for residential and commercial use which incorporate AI to provide consumers with insights into their energy consumption patterns. Solutions from companies like Nokia, Sense, C3 AI Energy Management and Storm, offer AI-based controllers that shift heat demand to times when energy is cheap.
For supply chain management, Carbon Chain provides an AI-driven carbon accounting solution to analyse a company’s environmental impact. CO2AI is a carbon management software for large corporations, and GE Vernova recently released a pilot AI-based carbon management system called CERius.
For small and medium-sized enterprises, Watershed, Emitwise and Pledge offer AI-driven carbon management solutions for measuring and reporting greenhouse gas emissions. Sweep and Plan A provide AI-based carbon management platforms for large corporates.
Key suppliers
GlobalData estimates that the global market for AI platforms in the power industry will reach about $5.3bn in 2024.
Leading suppliers of AI software in the energy sector include ABB, AutoGrid for AI-driven energy resource management, Bidgeley, Drift Marketplace, C3.ai, Fluence for energy storage, mPrest, SparkCognition, Stem and Uplight for analysis.
More AI in action
Duke Energy, one of the largest US energy holding companies, brought in Avanade, a joint venture between Accenture and Microsoft, to develop an Azure-based cloud platform that monitors methane emissions from natural gas distribution assets using satellite monitoring and AI, resulting in dramatically lower methane emissions.
Another US utility services holding company, Exelon, is using AI, cloud computing and big data to predict failures in power generation, distribution and transmission systems; optimising equipment performance and schedule maintenance, and minimising operating costs. Synthetic data generation is used with drones to automate grid structure inspection, with thousands of photorealistic examples of grid asset defects.
Meanwhile, Spanish utility Iberdrola’s US subsidiary Avangrid has launched a Data Science and Analytics team to develop proprietary AI systems to improve grid reliability.
And this is just a taste of AI treats to come.
Putting ChatGPT to the test
Finally, having completed this article on AI, I decided to put the conversational AI-engine ChatGPT to the test for New Energy World.
The answers to my questions were delivered astoundingly quickly (merely 20 seconds later). But they lacked the deep insight anticipated (even when we purchased a monthly subscription) despite the AI package having access to most Google search information up to September 2021. Although several examples were given, my article took several days of conventional research with more recent material.
The ChatGPT answers below were broadly useful. But they’ve not put this journalist out of a job yet! Judge for yourself.
What are the key challenges for AI being used in renewable energy operations?
ChatGPT: While AI offers tremendous potential for improving renewable energy operations, several key challenges must be addressed to fully harness its benefits:
- Data quality and availability: AI relies on large amounts of data for training and decision-making. Ensuring the availability of accurate and high-quality data from various renewable energy sources can be challenging, especially in remote or less-developed regions.
- Interoperability: Many renewable energy systems and devices use proprietary technologies. Achieving interoperability and standardisation across these systems for seamless AI integration can be complex.
- Data privacy and security: Energy data, particularly for smart grids, can be sensitive. Protecting data privacy and ensuring the security of AI systems against cyber threats is a significant concern.
- Regulatory and policy hurdles: Regulations and policies may lag behind technological advancements. Governments and regulatory bodies need to adapt and set clear guidelines for AI application deployment.
- Model uncertainty: AI models can produce results that are difficult to interpret. This can be problematic in critical energy operations.
- Energy transition challenges: Transitioning from traditional fossil fuel-based energy systems to renewable energy sources introduces complexities in grid management and energy storage. AI systems must adapt to these changes.
- Costs: Implementing AI can be costly, especially for smaller companies or organisations in developing regions. The initial investment in AI infrastructure and expertise can still be a barrier.
- Skills gap: The renewable energy sector may lack the necessary AI expertise and talent to develop, implement and maintain AI solutions effectively.
- Scalability: Ensuring that AI solutions can scale to accommodate the growth of renewable energy installations and increasing demand for clean energy is a significant challenge.
- Ethical concerns: As AI systems become more integrated into renewable energy operations, ethical considerations related to decision-making and bias need to be carefully addressed. AI does not possess the capacity for moral judgement.
- Environmental impact: The carbon footprint of training large AI models can be substantial. Ensuring that AI applications have a net positive environmental impact is crucial.
- Public acceptance: Public perception and acceptance of AI, particularly concerning issues like automation of jobs or potential risks, can impact the deployment of these technologies.
ChatGPT concluded: Addressing these challenges requires collaboration among governments, industries, research institutions and the AI community. Overcoming these hurdles will be essential to maximising the benefits of AI in renewable energy operations and accelerating the transition to sustainable and clean energy sources.
To sum up. There’s still need for the human touch, critical thinking and analysis when it comes to adoption of AI in the context of the renewable energy transition.
