The Meval Tool for Estimating and Verifying Energy Savings from HVAC System Upgrades

Energy efficiency has long been recognized as a critical pillar of the energy transition, yet its potential remains under-utilized as a tradable and financeable resource. A central barrier lies in the inherent fragility of energy savings estimates, which are highly sensitive to changes in building usage patterns, occupancy behavior, and HVAC system operation. These uncertainties complicate M&V, limiting confidence in reported savings and hindering the development of scalable business models and financing mechanisms for energy efficiency upgrades.

Traditional M&V approaches predominantly rely on predictive models that estimate baseline energy consumption and compare it with post-retrofit performance. While effective under stable conditions, these methods struggle to maintain accuracy when non-routine events or operational changes occur. As buildings are dynamic systems, such changes are the norm rather than the exception, causing M&V to deviate from a pure prediction problem and exposing fundamental limitations in conventional modeling techniques.

This white paper introduces a novel methodology developed by HEBES Intelligence and implemented in the Meval tool. The approach reframes M&V as a problem of mapping comparable operational states between pre- and post-retrofit periods, rather than solely predicting a counterfactual baseline. By identifying and categorizing the evolving states and conditions within a building, the methodology enables continuous adaptation to changes such as activity shifts, operational adjustments, and other non-routine events.

At the core of the Meval method is a dynamic modeling framework that detects relevant system states across time and establishes correspondences between them. Energy savings are then quantified as differences in consumption between matched states, ensuring that comparisons remain valid even as building behavior evolves. This state-based approach enhances robustness, reduces bias, and improves the reliability of savings estimates in real-world conditions.

The paper further examines the limitations of prediction-based M&V, explores causal-informed redesign options, and positions Meval within the broader landscape of M&V methodologies. By addressing key sources of uncertainty and enabling more trustworthy quantification of energy savings, the proposed approach represents a significant step toward making energy efficiency a credible, transactable resource.

Ultimately, Meval provides both a practical tool and a conceptual advancement for M&V, supporting more resilient business models and unlocking new opportunities for financing energy efficiency at scale.