Data Centre Energy Efficiency in Singapore: Meeting IMDA's New PUE Standards

Singapore is one of the most concentrated data centre markets in Asia, and also one of the hardest places on earth to run a data centre efficiently. The tropical climate works against every cooling system, while demand keeps climbing. In response, the Infocomm Media Development Authority (IMDA) released its Green Data Centre Roadmap on 30 May 2024, setting a clear efficiency benchmark: a Power Usage Effectiveness (PUE) of 1.3 or lower at 100% IT load over the next ten years. For operators, that target reshapes how a facility must be measured and managed.

What IMDA's PUE benchmark actually requires

PUE is the ratio of total facility energy to the energy delivered to IT equipment. A PUE of 2.0 means half the power goes to cooling, distribution losses, and overheads rather than computing. IMDA's roadmap aim of 1.3 or lower is demanding for a tropical site, where ambient heat makes cooling the single largest non-IT load.

The roadmap does not sit alone. It is reinforced by a stack of Singapore standards: the Tropical Data Centre Standard (SS 697:2023), which guides operators in raising server-room operating temperatures to 26 degrees Celsius and above; SS 564, the Green Data Centres energy and environmental management standard modelled on ISO 50001; and the refreshed BCA-IMDA Green Mark for Data Centres (GMDC:2024), launched in October 2024. Together they signal a direction of travel: efficiency is becoming the licence to grow capacity, not an optional sustainability extra.

That direction has a commercial edge. Data centre capacity in Singapore is deliberately scarce, and the roadmap pairs new capacity growth with green and efficiency commitments. In practice, the operators best positioned to win additional allocation are those that can demonstrate strong, measured energy performance rather than merely promise it. A credible PUE programme is therefore as much a business case as a compliance exercise: it protects both the right to expand and the operating margin on power, which is one of the largest line items a colocation or hyperscale facility carries.

Why the tropical climate makes PUE a measurement problem

In a temperate market, free cooling does much of the work. In Singapore, it cannot. Cooling energy dominates the PUE equation, so the largest efficiency gains come from operating cooling systems smarter, not just buying more of them.

IMDA notes that for every 1 degree Celsius increase in operating temperature, a data centre can secure cooling energy savings of between 2% and 5%. Raising setpoints sounds simple, but no operations team will do it blind. Pushing temperatures up without dense, real-time visibility into rack inlet conditions, chiller performance, and airflow risks thermal excursions on customer equipment. The constraint is not willingness; it is confidence in the data.

That is the core insight behind every credible PUE programme: you cannot improve, or defend, a number you cannot measure to a high resolution.

Granular metering and DCIM: the foundation of a defensible PUE

A PUE figure is only as trustworthy as the metering behind it. Calculating it from utility bills and nameplate ratings produces a number that collapses under scrutiny, particularly when an enterprise customer's procurement team asks how it was derived.

A defensible programme measures both sides of the ratio directly. Revenue-grade meters at the main intakes and the UPS input and output points establish the denominator, while submeters on mechanical cooling, lighting, and auxiliary feeders break down where the overhead actually goes. Data Centre Infrastructure Management (DCIM) then unifies that meter data with breaker status, temperature, and humidity in a single operator view, rather than leaving it scattered across vendor-specific consoles. A PUE engine that aggregates from one-minute to annual windows turns raw readings into the rolling metric that GMDC:2024 and enterprise audits expect.

This is precisely the architecture EcoXplore deployed at Singtel's DC West facility, where Hall 11 was commissioned with an integrated energy-monitoring and DCIM stack from day one. Measuring PUE at meter-level granularity, with the numerator and denominator both captured to revenue-grade accuracy, means the figure stands up in front of a customer rather than needing a footnote.

From measurement to efficiency: the practical levers

Once a facility can see its energy flows clearly, the route to a lower PUE follows a familiar order.

Optimise cooling before adding capacity

Cooling is where the largest savings sit. Higher operating setpoints under the SS 697:2023 approach, hot- and cold-aisle containment, variable-speed drives on pumps and fans, and right-sized chiller staging all reduce the non-IT load. Each change is validated against the live PUE trend rather than assumed.

Catch degradation before it costs efficiency

A fouling chiller or a failing fan bearing quietly erodes efficiency long before it triggers an alarm. Condition monitoring of rotating plant such as chillers, pumps, and CRAC units flags performance drift early, so a maintenance team intervenes before energy waste, or an outage, sets in. Efficiency and uptime are the same discipline viewed from two angles.

Manage energy as an ongoing system

Meeting a PUE target is not a one-time commissioning task. An energy management system such as EcoXplore's PecStar® iEMS provides the continuous reporting, benchmarking, and exception management that keep a facility on target as IT load shifts and equipment ages. This is also the layer that produces the audit-ready evidence for GMDC:2024 and SS 564 alignment.

How EcoXplore supports Singapore data centre operators

Meeting IMDA's PUE benchmark is an engineering problem first and a reporting problem second. It calls for accurate metering, integrated monitoring, and the field experience to optimise cooling without compromising the IT environment.

EcoXplore delivers that as an end-to-end, engineered platform rather than a collection of disconnected tools. Our work spans revenue-grade metering, DCIM integration, condition monitoring, and energy management under one PecStar® iEMS architecture, backed by ISO 9001:2015 and ISO 45001:2018 certification, a BCA ME02 L5 licence, BizSAFE Star status, and GeBIZ listing. With engineering teams across Singapore, Thailand, Malaysia, Indonesia, and Vietnam, we support multi-site operators who need consistent PUE measurement across the region, not just one facility.

If your data centre needs a PUE figure that holds up to an enterprise audit and a roadmap to 1.3, talk to the EcoXplore engineering team about an integrated monitoring and energy management assessment.