In December 2023, the COP28 declaration set out to triple renewable energy (RE) capacity and double energy efficiency (EE) improvements by 2030. As an ambition this is not modest; the world has struggled to displace fossil fuels, the contribution of which in the energy basket has stubbornly remained north of 80 percent in the past decade. Indeed, if China, which in recent years has contributed to more than half of the incremental RE capacity, is taken out of the equation, the gains would be very minimal. And it is not lack of intent but the absence of means to transform the complex energy systems that is slowing down the transition. The challenges are many, top 10 of which were highlighted in KPMG International’s pre-COP28 paper titled, Turning the Tide in Scaling Renewables.
Energy transition is a story of growth and transformation, especially in Asia and Africa — the growth hotspots where societies are looking to improve living standards rapidly. This growth would be energy and capital-intensive. The bp World Energy Outlook 2024 has called this “energy addition” in developing nations in contrast to the “energy substitution’ scenario in the developed world. While the capital needs are indeed substantial, the challenges are not necessarily with the availability of capital — they are more to do with how the providers of returnable capital evaluate their prospects. Such risks called out by the capital providers, especially in the developing nations are not insubstantial.
It is in this context that we look at the role of AI and Generative AI (Gen AI) in addressing the climate change issues, specifically the effect of AI in driving radical efficiency in energy systems and enabling rapid renewable energy scale up, especially in emerging economies that are growing and transforming at the same time.
Modern day AI, and especially the combination of AI, Gen AI and High-Performance Computing is transformational. It touches upon every aspect, from materials, equipment design, production and supply chain to project design and development, construction, and operations.
Modern AI can transform the pace of deployment of RE and that gives us a reasonable chance to meet the CoP28 goals on clean energy. AI can radically pace up resource identification, land procurement, permitting, sizing and Interconnection management. Within a farm it can optimise solar and wind asset placement. It is enabling real-time performance improvement, enhancing predictive maintenance capabilities, improving energy yield forecasts, optimizing energy storage. AI is enhancing flexibility and resilience in energy transmission and distribution making integration of variable RE possible at scale. This is bringing down overall costs of energy delivery and enhancing the dependability of RE. Together, these promise a rapid makeover of clean energy production and use.
In addition to the technical functions, modern AI has the potential to bring radical efficiencies in a range of corporate functions, including financial planning, reporting, treasury operations, internal controls, procurement, supply chain management, contract management, learning and development, etc. These can bring in significant predictability, thus reducing the risks considerably. Together, these and other AI applications can substantially help reduce risks and attract finance at acceptable cost points. Also, through the efficiency gains and reduction in project development timelines, the scale of deployment and costs of energy delivery can be significantly improved.
This AI-enabled scenario for clean energy is a promising, yet uncharted land. Creative solutions are needed, but so is discipline to ensure that the promise turns into outcomes. Solutioning must be founded in explicitly expressed problem statements. Security challenges, cost of AI and other concerns must be dealt frontally, with guardrails established. As technologies evolve rapidly, none of us are privy to all the answers. The energy future through AI must be co-created by stakeholders.
But AI is also a double-edged sword. As was the case with automobiles a century ago, AI applications — in particular, data centres — are the new energy guzzlers. That said, just as automobiles have transformed connectivity and thereby human society, AI also has the potential to bring in paradigm-altering changes. It needs to be directed to where it has the most positive impact. There is a real possibility of accelerating clean energy delivery with the wide and appropriate adoption of AI and, in turn, achieve the goals set out at COP28.
