Artificial intelligence could become a critical tool for helping Asia Pacific’s overstretched power grids cope with climate change, according to a new report by London-based energy think tank Ember.
To continue reading, subscribe to Eco‑Business.
There's something for everyone. We offer a range of subscription plans.
- Access our stories and receive our Insights Weekly newsletter with the free EB Member plan.
- Unlock unlimited access to our content and archive with EB Circle.
- Publish your content with EB Premium.
The report warned that power systems across the region are juggling two kinds of volatility at once, including the daily variability of fast growing wind and solar power, and the deeper climate shocks from more frequent heatwaves, droughts and floods.
Asia-Pacific now produces almost half of the world’s renewable power, and is expected to account for about 60 per cent of the growth in global electricity use up to 2050, scientists found. India and major Southeast Asian economies alone are projected to generate around 6,100 Terawatt-hours (TWh) of renewable electricity annually by 2050, exceeding Europe’s total electricity generation today.
At the same time, Southeast Asia alone has been getting steadily hotter, with average temperatures rising a small but steady amount every decade since the 1960s. Climate scientists say extreme heat events that used to happen only about once every 10 years will occur more than four times as often even in a 1.5°C warmer world, stretching the region’s grids beyond their original operations.
Most governments have focused on flexible investments like adding batteries, interconnectors and demand response to providing options in case of power fluctuations, but Said Muyi Yang, senior anergy analyst at Ember, argued that this addresses “only half of the challenge.”
“
AI provides the coordination capabilities to make systemic adaptation scalable, not as a silver bullet, but as the connective tissue that links fragmented data, siloed models, and complex decisions
Said Muyi Yang, senior anergy analyst, Ember
AI could provide the “coordination layer” needed to make Asia Pacific’s power grids make this shift, which will allow understanding of how failures can cascade across power, water, transport and other critical sectors, and redesigning grids accordingly, Muyi said.
“Climate adaptation is no longer an optional add-on. It is a core design requirement for system reliability. AI provides the coordination capabilities to make systemic adaptation scalable, not as a silver bullet, but as the connective tissue that links fragmented data, siloed models, and complex decisions,” he said.
Countries are shifting towards “system level adaptation”, which goes from just strengthening individual pieces of infrastructure to redesigning and operating the whole power system with its wider network of dependencies in mind, noted the study.
For instance, the report cited how China has put up a city-scale AI orchestration platform, which uses AI to help manage and optimise the power grid in real time, with a strong focus on supporting low‑carbon, flexible operations.
But it is not the case for many other countries in the region, especially in South and Southeast Asia where the transition is being held back by fragmented data, siloed models and growing complexity, said the study.
Being able to integrate diverse datasets, including mapping climate risks, the vulnerabilities of different power plants, and the various ways the systems rely on each other, remains “uneven and weak” across the region, said researchers, with climate information scarce and scattered.
Incomplete weather records, sparse monitoring and limited climate projections make it hard to assess specific and cascading risks, where a shifting climate means past patterns no longer reliably predict the future, the analysis said .
As a “fusion layer”, AI can pull together information that comes in different formats, levels of detail or quality and turn it into one clear picture of the system, researchers said.
Unlike traditional tools, which need data to be cleaned and put into a fixed format first, modern AI models can learn how different data sources relate to each other and automatically line them up, even if they are unstructured or only partly complete.
These models can also intelligently fill in the gaps in patchy datasets by estimating missing values or creating realistic “stand-in” data, which helps give better coverage in places where measurements are scarce.
“To use AI effectively for system adaptation, countries need the capabilities to integrate it into the planning and operation of power grids: high-quality data systems, technical expertise and computing infrastructure. Without these foundations, AI risks becoming another layer of complexity rather than a tool for improving resilience,” it said.
