Singapore Green Mark Version 7: An AI-Ready Energy Retrofit Playbook for Existing Buildings

A practical path for Singapore buildings to baseline energy use, prioritise retrofits and prepare stronger digital evidence.

Singapore’s Green Mark framework is moving towards a more practical, performance-led approach for existing buildings. At the International Built Environment Week on 2 September 2026, BCA announced Green Mark Version 7, including a new BCA Energy Rating pathway, the SLE70 tier, recognition for alternative cooling technologies and a longer certification validity period for qualifying buildings.

For commercial buildings, industrial sites, warehouses and SMEs, the message is clear: energy improvement does not have to begin with a major redevelopment. It can begin with a reliable baseline, better operational data and a retrofit plan that connects engineering decisions with measurable outcomes.

What Green Mark Version 7 means for existing buildings

Green Mark Version 7 is intended to make it easier for existing buildings to start with energy-performance improvements. This is important for properties where replacing every major system at once may not be practical because of cost, tenant needs, operational continuity or limited plant-room space.

The new BCA Energy Rating pathway is particularly relevant to owners who want a clearer way to demonstrate building energy performance. The SLE70 tier also signals a higher-efficiency ambition for buildings that can achieve the applicable performance level. However, owners should confirm the latest BCA criteria, submission requirements and applicability for their building type before making formal certification commitments.

The framework’s recognition of alternative cooling technologies is another useful development. It encourages project teams to consider more than conventional equipment replacement when assessing cooling demand, system efficiency, operational constraints and whole-life value.

Step 1: Establish an energy baseline before selecting solutions

A retrofit programme should begin by answering a basic question: where is the building using energy, and when?

Collect at least 12 months of available utility information where possible, together with operating hours, occupancy patterns, production schedules, weather context and major equipment changes. For warehouses and industrial facilities, record refrigeration, process loads, loading-bay activity, ventilation and storage conditions. For offices and mixed-use buildings, separate tenant, common-area, lift, lighting and ACMV-related consumption where the metering arrangement allows.

The baseline should identify more than total kilowatt-hours. Useful indicators may include energy use by floor area, operating hour, temperature zone, production output or storage volume. The right metric depends on the building’s function. A warehouse should not be assessed in exactly the same way as an office tower or a light-manufacturing facility.

Where data is incomplete, begin with a data-quality review. Check meter intervals, missing readings, inconsistent naming, manual spreadsheets and changes in tariff or operating conditions. AI cannot compensate for poor data without creating misleading recommendations.

Step 2: Improve ACMV performance and control

Air-conditioning and mechanical ventilation systems are often a major energy focus, but optimisation is not limited to changing chillers or air-handling units. The first review should consider operating schedules, temperature setpoints, outside-air requirements, fan speeds, filters, sensors, dampers, valves, controls and zone-level demand.

• Compare equipment schedules with actual occupancy and operating hours.
• Check whether sensors are correctly located, calibrated and producing usable readings.
• Review simultaneous heating and cooling, excessive outside air and over-ventilation.
• Trend supply-air temperature, chilled-water conditions, fan speed and zone comfort.
• Investigate abnormal pressure drops, fouled coils, blocked filters and valve behaviour.

For warehouses and industrial sites, the assessment should also account for door openings, dock activity, heat-generating equipment, high-bay spaces and areas with different environmental requirements. Comfort, product protection and workplace conditions remain operational priorities; energy reduction should not be pursued by weakening necessary controls.

Step 3: Evaluate alternative cooling technologies properly

Green Mark Version 7’s recognition of alternative cooling technologies creates an opportunity to widen the options considered during retrofit planning. Depending on site conditions, alternatives may include different heat-rejection arrangements, thermal storage, improved ventilation strategies, heat recovery or other lower-energy approaches.

The correct choice depends on the building’s load profile, available space, water considerations, maintenance capability, noise constraints, resilience requirements and operating schedule. A technology should not be selected because it is new or highly visible. It should be assessed against measured demand, installation risk, maintainability and the evidence required for the intended Green Mark pathway.

A feasibility review should therefore compare options using a consistent decision matrix. Include energy impact, capital work, disruption, controls integration, spare-parts requirements, maintenance skills and expected operating behaviour. Engineering judgement remains essential, even when AI is used to analyse the data.

Step 4: Build a smart-meter and data foundation

Smart meters and sub-metering can turn a one-time audit into an ongoing management process. The objective is not to collect every possible data point. It is to collect the right information at a useful interval and make it accessible to the people responsible for action.

A practical data architecture may combine main incoming meters, ACMV plant meters, distribution-board meters, temperature and humidity sensors, equipment run status, building-management-system points and maintenance records. Establish consistent asset names, timestamps, locations and ownership of each data stream.

Data should be presented through dashboards that answer operational questions: Is night-time consumption higher than expected? Which air-handling unit is drifting from normal performance? Did a schedule change increase demand? Is a refrigeration system operating longer than its baseline? Can the facilities team verify whether an intervention delivered the expected result?

Step 5: Apply AI where it supports decisions

AI is most useful when it reduces the time required to detect, investigate and prioritise issues. Relevant applications may include:

• Anomaly detection: identify energy or equipment behaviour outside an established operating pattern.
• Predictive maintenance: flag early signs of abnormal run time, temperature, vibration or control behaviour for human review.
• Load analysis: compare demand across time periods, zones and operating conditions.
• Fault diagnostics: help narrow possible causes of inefficient ACMV or refrigeration performance.
• Workflow automation: convert alerts into inspection tasks, escalation steps and follow-up records.

These tools should support, not replace, competent facilities and engineering teams. Every alert needs a sensible threshold, a responsible owner and a way to confirm whether the recommended action was appropriate. Start with a small number of high-value use cases instead of deploying an overly complex platform with no operational adoption.

Step 6: Prioritise retrofits using evidence and risk

Once the baseline and data foundation are in place, rank retrofit opportunities using more than estimated energy savings. Consider safety, statutory obligations, equipment condition, business continuity, tenant impact, maintenance risk, carbon objectives, ease of implementation and the quality of available evidence.

A useful sequence is:

1. Correct data, scheduling and control issues that require limited physical work.
2. Address maintenance and equipment-condition problems affecting efficiency or reliability.
3. Implement targeted controls, sub-metering and sensor improvements.
4. Evaluate major ACMV, refrigeration, lighting or cooling-system replacements.
5. Measure results and update the investment plan using actual performance.

This staged approach helps owners avoid spending the entire budget on a single replacement while leaving operational waste unresolved. It also creates a clearer record of what changed and what improved.

Step 7: Create digital evidence for the energy-rating pathway

Retrofit evidence should be organised from the beginning, not assembled at the end. Maintain a central record of utility data, meter schedules, equipment inventories, commissioning information, maintenance activities, control changes, invoices, drawings, test results and before-and-after performance.

Use version-controlled files and consistent naming. Record the date, location, responsible person and reason for each major change. Where AI produces an alert or recommendation, retain the relevant input data, human review and action outcome. This improves operational accountability and can make later verification more efficient.

Because Green Mark Version 7 requirements and submission procedures should be checked against current BCA guidance, treat the evidence register as a living project document. Confirm which records are required for the chosen pathway before finalising the retrofit scope.

A practical starting point for Singapore businesses

For a building owner or facility manager, the first 30 to 60 days can focus on three deliverables: an energy and asset baseline, a ranked list of operational and engineering opportunities, and a digital evidence structure that can grow with the project.

ISS can support organisations that need to connect engineering review, facility-management workflows and AI automation. The right starting point may be an energy-data assessment, ACMV optimisation review, smart-meter plan, predictive-maintenance use case or broader retrofit roadmap.

Contact ISS to discuss your engineering, facility management or AI automation requirements for an existing Singapore building.