Building Energy Models Overestimate Retrofit Savings, Distorting Investment Decisions

Building energy models used to justify retrofit investments systematically overestimate the energy savings that retrofits actually deliver. This "performance gap" — the difference between predicted and measured post-retrofit energy use — routinely ranges from 20% to 60%, meaning buildings frequently save far less energy than their models promised. The problem stems from static modeling approaches that assume idealized occupant behavior, simplified physics, and stable conditions, when in reality building energy performance emerges from complex interactions between envelope, HVAC systems, weather, occupancy patterns, and human behavior that are highly dynamic and difficult to predict.

Retrofit investment decisions — often involving hundreds of thousands to millions of dollars per building — are made on the basis of these energy models. When models overestimate savings, projects deliver poor returns, eroding investor and building owner confidence in energy efficiency as a viable investment class. At the policy level, governments base building energy codes and renovation mandates on modeled performance targets. If those models are systematically optimistic, compliance doesn't actually deliver the expected emissions reductions, creating a hidden gap in national climate commitments. The EU's Energy Performance of Buildings Directive (EPBD) relies heavily on standardized energy performance calculations that multiple studies have shown diverge significantly from actual consumption.

Static calculation methods (e.g., ISO 13790, ASHRAE Standard 90.1 compliance tools) are the dominant approach for retrofit planning because they're fast, standardized, and inexpensive. But they use simplified thermal models, assume fixed occupancy schedules, and cannot capture feedback loops between systems. Dynamic simulation tools like EnergyPlus and TRNSYS offer higher fidelity but require detailed building data that's expensive to collect, expert calibration, and significant computation time — making them impractical for large-scale retrofit planning across building portfolios. Machine learning approaches trained on building operational data show promise, but require years of historical data that is rarely available for individual buildings, and models trained on one building's data don't transfer well to others due to the idiosyncratic nature of each building's construction, systems, and occupancy patterns.

Two complementary advances could close the performance gap. First, hybrid models that combine physics-based building simulation with data-driven calibration from sensors and utility data could capture building-specific dynamics without requiring the full data input of detailed simulation. Second, robust methods for incorporating occupant behavior into energy models — moving beyond idealized schedules to probabilistic occupancy and behavior models — would address one of the largest sources of prediction error. Occupant behavior has been shown to cause 2–3x variation in energy use between identical buildings, yet most models treat occupants as fixed boundary conditions rather than dynamic agents.

A student team could instrument 3–5 campus classrooms or offices with low-cost temperature, CO₂, and occupancy sensors, then compare actual energy consumption against predictions from a standard energy model (e.g., OpenStudio or DesignBuilder). The measured performance gap could be decomposed by source: envelope differences, system efficiency, occupancy variation, and behavioral factors. This would produce a quantitative case study of where prediction models break down. Skills in building science, data analysis, sensor deployment, and HVAC fundamentals would be most relevant.