Three ways to harness the power of generative AI for the energy transition
World Economic Forum
June 21, 2024 07:00 MYT
June 21, 2024 07:00 MYT
THE energy transition is one of the most urgent challenges facing the world today. As energy businesses work to reach their sustainability goals, reinvent their enterprises and succeed in the energy transition to net zero, they have a powerful new ally: generative AI.
AI Brief
- Generative AI is transforming the energy sector by enhancing project efficiency and reducing costs.
- Companies must ensure a robust and complete data foundation, establish strong responsible AI programmes and make sustainable choices in their operations, going beyond mere technology adoption.
- A new report by the World Economic Forum, Fostering Effective Energy Transition 2024, highlights the importance of digital innovations, such as generative AI, in enhancing energy equity and security.
This revolution in the power of AI has arrived at a time of great need for innovation. We must reduce the cost premium of lower carbon solutions, scale the technologies required, renew and reskill the workforce and attract and deploy as much as $4 trillion a year in investment.
Accenture research shows that generative AI holds incredible potential as a catalyst for reinvention – it can improve productivity in nearly half of the activities in the energy industry. We estimate that, by 2030, the industry’s investment in generative AI will more than triple, from approximately $40 billion a year to over $140 billion.
Leading companies are already realizing value across their value chains: exploration, development and production, and reinventing some of the most critical workflows. A national oil company, for example, uses generative AI, through a large language model (LLM) and a dedicated search engine, to allow employees to 'chat' in real-time with a growing knowledge base of over a quarter of a million documents. In practical terms, this means a recent graduate can immediately access the knowledge of an industry veteran, significantly increasing efficiency, productivity, upskilling, de-risking execution and putting knowledge where it matters: right at the front line.
We see three key opportunities for generative AI to drive the energy transition:
1. Speed and cost in the delivery and execution of capital projects
Generative AI enables better forecasting of the project schedule, reduction in delays, cost overruns and other project risks by proposing effective mitigation actions. It can cut the time necessary to perform the upfront concept, engineering and detailed design work, compressing the review and approval processes by as much as half.
2. Enhanced asset efficiency and productivity
By leveraging operational data, generative AI can improve the maintenance, operations and efficiency of key assets. It can, for example, adjust the angle of solar panels or the pitch of wind turbine blades in real time to maximize energy capture based on weather conditions, ensuring that power is available to the grid at forecasted times of greatest demand and optimal pricing.
3. Strengthened supply-demand management and trading
Generative AI can handle vast quantities of structured and unstructured data, enabling new solutions that can predict or automatically suggest or respond to energy demand. Ultimately, this could flatten the energy demand curve, lower the capital expenditure required in physical infrastructure and improve overall use rates.
To realize value from generative AI, companies need to take some important actions.
First, most companies still need to access the right data and a strong digital core. This means ensuring a robust and complete data foundation, using energy-efficient and powerful cloud data platforms and modernizing applications to harness their full power.
Second, companies must also ensure that they establish strong responsible AI programmes. This commitment is critical, given the imperatives of energy security and the continued advances in AI and government policies around its responsible use.
Third, companies must make sustainable choices in the ways in which they work beyond the adoption of technology. The carbon footprint of LLMs, for example, can be decreased through a strategic selection of algorithms, tailored hardware and energy-efficient cloud data centres. An experiment by Accenture showed that modifying pre-trained models for new tasks, instead of building models from scratch, retained the same accuracy level and used nearly three times less energy.
By responsibly and sustainably adopting next-level technologies and specifically generative AI, the industry can reinvent its core, while accelerating and de-risking the energy transition.