The Flexibility Gap: How Building Demand Response Could Cut ASEAN’s Peak Demand Crisis

ASEAN’s power grid is facing an acceleration problem. Peak electricity demand is growing faster than average demand—driven largely by cooling—and the region’s utilities are running out of generation capacity to handle the peaks. But a solution is hiding in plain sight: inside the buildings themselves.

A 2026 Duke University Nicholas Institute study found that a mere 1–2% reduction in data centre peak demand can reduce electricity rates by 0.5–2.8% system-wide and protect grid reliability. For a region facing 60% growth in total energy demand by 2040, that’s not a marginal improvement. It’s structural. Yet most ASEAN commercial buildings remain locked into passive consumption—switching on their cooling systems whenever they choose, oblivious to the grid strain they’re creating.

Why Peak Demand Is Outrunning Average Demand

The International Energy Agency’s Southeast Asia Energy Outlook 2026 is explicit: cooling demand in buildings is the fastest growing segment. Residential air-conditioner stock alone is projected to triple by 2035. Add commercial air-conditioning, industrial chilling, and data centre liquid cooling, and the peak-hour load shape becomes a cliff face.

The problem is temporal concentration. Buildings cool hardest during the same 4–6 hour window—typically 11am–5pm—when outdoor temperatures peak. Utilities must build generation capacity to handle this isolated spike, then run those plants at 30–40% capacity for the rest of the day. For every MW of peak demand, the grid carries inefficient assets.

Data centres compound this. Asia Pacific’s data centre cooling market alone is forecast to reach USD 105.6 billion by 2035, growing at 27.3% per year. Each new facility is a high-density load—50–100 MW for a hyperscale build—arriving with no built-in flexibility.

The Hidden Flexibility Asset

Thermal inertia is building owners’ unmonetized superpower. Modern commercial buildings and data centres, if equipped with smart controls, can pre-cool their spaces or equipment during off-peak hours (when electricity is cheaper and the grid has spare capacity), then reduce mechanical cooling by 10–30% during peak hours—relying on stored thermal mass rather than active mechanical load.

For a 10-storey commercial tower with 20 MW annual demand, shifting just 2 MW out of the peak window cuts system costs dramatically. For a data centre capable of running hot-water loops at variable temperatures, the payoff is even larger: Dell and others have demonstrated that a 1–2°C tolerance adjustment costs nothing operationally but sheds 5–15% of peak cooling load.

This is demand response. And it is almost entirely absent from ASEAN buildings.

Why Buildings Aren’t Responding

Three barriers persist:

  • No price signal. Most ASEAN utilities don’t offer time-of-use tariffs granular enough to reward flexibility. If a building can’t see the peak-hour penalty, it has no financial reason to shift load. Thailand’s decade-long tiering overhaul is an exception—not the rule.
  • No coordination mechanism. Utilities lack the digital infrastructure to contract with individual buildings for flexible load. Aggregators—third parties that bundle small loads into grid-scale resources—barely exist in ASEAN.
  • Control risk. Building managers fear loss of comfort or operational control if utilities can shed their load remotely. Coordination protocols and liability frameworks are undefined.

The Economics of Unlocking Flexibility

The International Energy Agency’s 2026 ASEAN energy roadmap identifies demand response and smart grids as critical to managing renewable integration. But “critical” doesn’t mean self-funding. Early adopters—utilities piloting demand response in Singapore and Vietnam—report contract prices of USD 20–50 per kW per year for commercial and industrial loads willing to reduce peak demand by 10–20%.

For a 2 MW facility capable of shedding 200 kW during peak hours, that’s USD 4,000–10,000 annually. Combined with the time-of-use tariff savings—often 15–25% of the peak-hour increment—the cash case closes in 2–4 years, before any controls hardware wears out.

Scale matters. If ASEAN utilities mobilized half the region’s commercial and industrial stock (roughly 150–200 GW of installed cooling)—conservatively 10% peak shiftability—they could defer 15–20 GW of new thermal generation. At USD 1,200–1,500 per kW to build a gas plant, that’s USD 18–30 billion in avoided capex. Demand response contract payments would consume a fraction of those savings.

What’s Needed

Regional progress requires four steps:

  • Utility adoption of mandatory time-of-use tariffs reflecting true system cost (already legislated in Thailand; pending in Vietnam).
  • Building energy management systems with remote curtailment capability—standard in new constructions, retrofitted through HVAC upgrades.
  • Regulatory clarity: liability caps for grid-shed comfort excursions, insurance frameworks for automated load response.
  • Aggregation platforms—often software, not hardware—that bundle 50–100 buildings into single grid contracts, reaching the 5–10 MW minimum utilities require.

Singapore’s BCA and Thailand’s PEA have both launched pilots. But the region’s response remains fragmented.

The Grid’s Timing Problem

ASEAN’s thermal generation fleet is aging. Phasing out coal while accommodating 100+ GW of new wind and solar capacity requires not just storage, but operational flexibility—the ability to shed load within minutes. Buildings can provide that. Data centres, with their 24/7 operations and predictable thermal dynamics, are particularly well-positioned.

The mismatch is stark: a region with a peak-demand crisis and billions in building assets that could help solve it, yet neither the commercial signals nor the policy framework to activate those assets.

If you’re managing a portfolio of commercial buildings or data centre operations across ASEAN and want to explore how your cooling assets could generate revenue while stabilizing the grid, there’s more to discuss.