The Building Energy Performance Gap and Why It Starts at the Drawing Board

The architecture industry is exceptionally good at predicting how buildings will look. It is remarkably poor at predicting how they will perform. Study after study confirms that buildings routinely consume between two and five times the energy their design models predicted. Some research suggests the gap can stretch to ten times for certain building types. These are not rounding errors. They represent a systematic failure to connect design intent with operational reality, and the cost, in both carbon and dollars, is enormous.

This is a problem the industry has known about for decades and has largely declined to own. The conventional division of responsibility is clear: the architect designs the building, the contractor builds it, and the facilities manager operates it. Performance in operation is somebody else's problem. This separation is no longer tenable, not because of professional idealism, but because the regulatory and commercial environment is rapidly closing the gap between design promises and operational outcomes. Firms that continue to treat post-occupancy performance as outside their scope will find themselves increasingly exposed.

A Gap Nobody Wants to Measure

The building energy performance gap is one of the most well-documented and least-addressed problems in the AEC industry. The pattern is consistent: an energy model is developed during design, often to satisfy certification requirements. The model predicts a certain Energy Use Intensity. The building is certified, the awards are won, the project is handed over. And then, quietly, the building begins consuming significantly more energy than the model said it would.

“Buildings routinely consume 2 to 5 times more energy than their design models predicted, a performance gap the industry has known about for decades.”

The causes are well understood: occupant behaviour that differs from modelling assumptions, HVAC systems that are commissioned to specification but degrade without proper maintenance, control strategies that work in simulation but are overridden in practice, façade performance that changes with orientation and microclimate in ways the energy model simplified, and, perhaps most fundamentally, a design process that optimises for compliance rather than performance. The energy model becomes an artefact of the certification process rather than a genuine prediction of how the building will behave.

In Singapore, the regulatory framework is more rigorous than most. BCA's Green Mark certification requires re-assessment every three years, and Periodic Energy Audits of cooling systems are mandatory for certified buildings. The new Mandatory Energy Improvement regime, expected from late 2025, will require owners of energy-intensive buildings to appoint qualified engineers to audit performance and develop improvement plans targeting at least 10 per cent EUI reduction. These are meaningful interventions. But they are interventions applied in operation, long after the design decisions that most influence performance have been locked in.

Design Decisions, Operational Consequences

“70 to 80 per cent of a building's lifetime energy consumption is determined by decisions made in the first 20 per cent of the design process.”

The inconvenient truth of building performance is that 70 to 80 per cent of a building's lifetime energy consumption is determined by decisions made in the first 20 per cent of the design process. The building's orientation, its massing, its window-to-wall ratio, its structural system, its floor-to-floor height. These foundational choices establish a performance envelope that no amount of operational optimisation can fundamentally alter. A building with an excessive glass curtain wall facing west in a tropical climate will be an expensive building to cool for its entire life, regardless of how sophisticated its BMS is or how diligently its facilities team manages the chiller plant.

Yet these are precisely the decisions that architects make with the least performance feedback. In most practices, the designers who determine a building's massing and orientation have never received post-occupancy data from their previous projects. They do not know whether the west-facing double-height lobby they designed five years ago is costing the building owner an additional S$200,000 per year in cooling energy. They do not know because nobody told them, and nobody told them because the industry's information loop is broken at the point of handover.

At Swan & Maclaren, the firm has begun to confront this directly. When an integrated architecture and engineering practice designs a building's MEP systems and subsequently observes how they perform in operation, the feedback loop is immediate and sometimes uncomfortable. It reveals where modelling assumptions diverged from reality, where control strategies specified during design are not being implemented as intended, and, candidly, where design choices made for aesthetic or programmatic reasons imposed energy costs that a more performance-aware process might have avoided. This kind of honest self-assessment is rare in an industry that celebrates design intent and rarely measures design outcomes.

Closing the Loop

What would it look like for architects and engineers to genuinely own post-occupancy performance? Not as a contractual obligation (the liability implications would be complex) but as a practice discipline that informs how buildings are designed from the outset?

The first requirement is data. Post-occupancy evaluation remains surprisingly uncommon in Southeast Asia. Most firms complete a project, photograph it for their portfolio, and move on. The systematic collection of performance data (actual energy consumption, thermal comfort surveys, daylight measurements, maintenance records) is the exception rather than the norm. Without this data, every new project starts from the same assumptions, and the same performance gaps recur. An integrated practice that designs both the architecture and the engineering has a natural advantage here: the data exists within the firm, if someone is willing to collect and analyse it.

The second requirement is humility. The performance gap persists partly because the industry's incentive structure rewards prediction over verification. Design awards are given for intent, not outcomes. Green certifications are awarded at completion, not after three years of operation. Fee structures do not include post-occupancy review. Until these incentives shift, and they appear to be beginning to, the only firms that will close the loop are those that choose to, not because they are required to, but because they recognise that performance accountability is a competitive advantage.

“Design awards celebrate intent. Green certifications reward compliance. Neither measures whether the building actually works as promised.”

The third requirement is integration. The performance gap is, at its core, a coordination failure between disciplines. The architect who determines the façade does not always coordinate deeply enough with the MEP engineer who sizes the cooling system. The sustainability consultant who builds the energy model may not have full visibility of the architect's design changes after the model was submitted. The commissioning agent who verifies the systems at handover may never speak to the designers who specified them. Each handoff introduces information loss, and each loss widens the gap between what was designed and what is delivered.

Signals Worth Watching

Singapore's new Mandatory Energy Improvement regime represents a meaningful regulatory shift. By requiring building owners to appoint qualified professionals to audit performance and develop improvement plans, the government is signalling that design-stage compliance is no longer sufficient. Operational performance must be demonstrated and improved. This regime is likely to tighten over time, with eventual requirements for disclosure of actual versus designed energy performance. When that happens, the reputational and commercial implications for architects and engineers who consistently over-promise will be significant.

Digital twins and building management systems are maturing to the point where real-time performance data is increasingly available and increasingly affordable. The technical barrier to post-occupancy evaluation is falling rapidly. What remains is the cultural barrier: the willingness of design firms to look at the data, learn from it, and feed it back into their next project. The firms that build this discipline now, while it is still voluntary, will have a substantial head start when it becomes mandatory.

ESG reporting requirements are also converging on actual performance. As institutional investors and corporate tenants face increasing pressure to report Scope 1 and 2 emissions accurately, the performance gap becomes a commercial liability rather than a technical curiosity. A building that was marketed as achieving a certain energy intensity but consistently exceeds it creates reporting problems for its occupants, valuation problems for its owner, and reputational problems for its designer. The market is beginning to care about the gap, even if the profession has been slow to.

Beautiful and Honest

None of this is an argument against beautiful buildings. It is an argument for honest ones: buildings whose design ambitions are matched by their operational performance, and whose architects and engineers are willing to be measured by what the building actually does, not just what the model said it would do.

Swan & Maclaren designs buildings it is proud of. But increasingly, the firm measures that pride not by the render or the award, but by the energy bill, the comfort survey, and the maintenance log. The most beautiful building any firm can design is one that performs exactly as promised. That is a higher standard than the industry typically holds itself to. It is the standard the industry is heading toward, and the firms that arrive early will be better positioned than those forced there later.