From AI to emissions: Aligning ASEAN’s digital growth with energy transition goals

This report examines the impact of rapid data centre growth in ASEAN on power sector emissions, highlighting that while these facilities could account for 2% to 30% of national power demand by 2030, approximately 30% of this demand could be met by solar and wind power without battery storage. The authors argue for urgent policy interventions, energy efficiency improvements, and expanded renewable procurement options to prevent the digital boom from derailing regional energy transition goals.

• Data centre electricity demand is projected to grow unevenly across ASEAN, with Malaysia experiencing the most significant surge. Malaysia's demand is expected to rise from 9 TWh in 2024 to 68 TWh by 2030, which would represent 30% of its national power consumption. In contrast, growth is expected to be more gradual in Singapore and Thailand, and minimal in Viet Nam.
• The expansion of data centres risks significantly increasing emissions in countries with fossil-fuel-dependent grids. Malaysia's emissions from data centre power demand could increase sevenfold, from 5.9 MtCO2e in 2024 to 40 MtCO2e by 2030. Indonesia's emissions are expected to quadruple (+13 MtCO2e), and the Philippines' emissions are projected to rise 14 times, from 0.8 MtCO2e in 2024 to 10.5 MtCO2e by 2030.
• Approximately 30% of the electricity demand for data centres in 2030 can be supplied by solar and wind power via the grids without requiring battery storage, suggesting that battery costs are not an immediate barrier to adoption. However, an estimated $45 to 75 billion in solar and wind investments will be needed by 2030 to power these facilities sustainably.
• Energy efficiency is a critical lever for decarbonisation, particularly regarding cooling systems, which can account for 39% of a data centre's total energy consumption. In tropical ASEAN climates, increasing operating temperature set points can save 4.3% to 9.8% of cooling electricity for every one degree Celsius increase. Additionally, immersion cooling technology can reduce power consumption by approximately 40% compared to traditional methods.
• Corporate renewable procurement options in ASEAN are currently limited or inaccessible for smaller companies. While giant tech firms use Power Purchase Agreements (PPAs), unbundled Renewable Energy Certificates (RECs) have been more dominant in Asia. The report suggests that expanding virtual PPAs and green tariffs would improve affordability and access to clean power.
• There is a lack of comprehensive national sustainability frameworks for data centres in most ASEAN countries. While Singapore and Malaysia have introduced frameworks, they were implemented after the industry boom began, and no regulations currently require older, inefficient facilities to be retrofitted to meet energy efficiency or renewable energy standards.

APA

Nadhila, S., & Setyawati, D. (2025). From AI to emissions: Aligning ASEAN’s digital growth with energy transition goals. Ember. https://ember-energy.org/app/uploads/2025/05/Report-From-AI-to-emissions-ASEAN.pdf

Chicago

Nadhila, Shabrina, and Dinita Setyawati. From AI to emissions: Aligning ASEAN’s digital growth with energy transition goals. Ember, 2025. https://ember-energy.org/app/uploads/2025/05/Report-From-AI-to-emissions-ASEAN.pdf.

Wikipedia

{{cite report |last1=Nadhila |first1=Shabrina |last2=Setyawati |first2=Dinita |title=From AI to emissions: Aligning ASEAN’s digital growth with energy transition goals |publisher=Ember |date=27 May 2025 |url=https://ember-energy.org/app/uploads/2025/05/Report-From-AI-to-emissions-ASEAN.pdf |access-date=17 August 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{nadhila2025from, author = {Nadhila, Shabrina and Setyawati, Dinita}, title = {{From AI to emissions: Aligning ASEAN’s digital growth with energy transition goals}}, institution = {Ember}, year = {2025}, month = may, url = {https://ember-energy.org/app/uploads/2025/05/Report-From-AI-to-emissions-ASEAN.pdf}, urldate = {2026-08-17}, note = {Indexed by Climate Insights Directory} }