The problem with self-reported savings
Most distributed energy programs today rely on deemed savings: pre-calculated estimates of how much energy a measure should save, based on engineering models and assumptions. Install a heat pump with a rated efficiency of 10 HSPF replacing an 8 HSPF unit, and the program credits you with a fixed number of kilowatt-hours saved per year.
This approach has a fundamental limitation: it tells you what should happen, not what actually happened.
Real-world performance varies. Installation quality, occupant behavior, local climate patterns, building characteristics, and equipment degradation all affect actual savings. Studies consistently find gaps between deemed and measured savings, sometimes by 30% or more in either direction.
When the stakes are low, rough estimates are adequate. But as distributed energy resources take on larger roles in grid planning, carbon accounting, and financial markets, the gap between estimated and actual performance becomes a real problem.
What independent M&V changes
Measurement and verification (M&V) replaces estimates with evidence. Instead of calculating what a resource should save, M&V measures what it actually saves by analyzing real meter data.
Independent M&V adds a critical layer: the entity verifying performance is not the same entity that installed the equipment or runs the program. This separation matters for the same reason that financial audits are performed by independent firms, not by the company being audited. When the verifier has no stake in the outcome, the numbers are more credible.
The process works like this:
-
Baseline modeling. A statistical model is built from pre-intervention meter data, capturing the relationship between energy consumption and variables like weather, time of day, and day of week.
-
Counterfactual estimation. After the resource is installed or activated, the model projects what consumption would have been without the intervention, adjusted for current conditions.
-
Savings calculation. The difference between the counterfactual (what would have happened) and actual consumption (what did happen) is the verified savings, calculated at hourly granularity.
-
Continuous updates. This is not a one-time audit. The model runs continuously against incoming meter data, producing updated savings figures as conditions change.
Why it matters now
Several trends are converging to make independent M&V essential rather than optional.
Grid planning depends on DER performance. Utilities and grid operators are increasingly counting distributed resources toward resource adequacy requirements. If a utility plans to meet peak demand partly through aggregated DERs, it needs confidence that those resources actually perform during peaks. Deemed savings cannot provide that confidence. Verified, meter-based performance data can.
Carbon markets require credible accounting. As carbon reporting moves from voluntary to mandatory (SEC climate disclosure rules, state-level requirements), the bar for substantiating emissions reduction claims rises. “We installed efficient equipment” is not sufficient. Verified hourly savings data, combined with marginal emissions factors, produces defensible carbon accounting.
Financial instruments need reliable data. Energy efficiency and DER investments are increasingly financed through structures that depend on performance data: pay-for-performance contracts, efficiency-as-a-service models, and securitized portfolios. Investors and counterparties in these structures need independent verification, not self-reported numbers from the party getting paid.
Program administrators face accountability pressure. Ratepayer-funded energy efficiency programs are under growing scrutiny. Regulators and consumer advocates want to know whether programs deliver what they promise. Independent M&V provides the evidence base for that accountability.
What good M&V looks like
Not all verification is equal. The quality of M&V depends on several factors:
Granularity. Monthly or annual M&V catches gross trends but misses important patterns. Hourly M&V reveals when savings occur, enabling capacity and carbon attribution. A resource that saves energy at 3 AM is very different from one that saves during a summer afternoon peak, and only hourly data captures this.
Statistical rigor. The baseline model should account for weather, occupancy patterns, and other relevant variables. It should report uncertainty bounds, not just point estimates. A claimed savings of 500 kWh is meaningless without knowing whether the confidence interval is plus or minus 50 kWh or plus or minus 400 kWh.
Independence. The verifier should have no financial relationship with the installer, program administrator, or asset owner that would create incentive bias. Independence is what makes verification trustworthy.
Transparency. The methodology should be published and auditable. Stakeholders should be able to understand how savings are calculated, what data inputs are used, and what assumptions underlie the models.
Continuity. One-time verification tells you how a resource performed during a specific period. Continuous verification tells you how it performs over its lifetime, catching degradation, behavioral changes, and seasonal patterns that a single audit would miss.
How Aristotle handles this
Aristotle is WattCarbon’s M&V engine. It ingests meter data from enrolled assets, builds statistical baseline models, and calculates verified savings at hourly granularity on a continuous basis.
The verified savings feed directly into two scoring systems: GridScore, which measures capacity value (when savings occur relative to peak demand), and CarbonScore, which measures carbon impact (how much CO2 is avoided given local grid emissions).
These verified results are then issued as Energy Attribute Certificates through WEATS, WattCarbon’s registry. Each certificate carries the verified energy, capacity, and carbon attributes, with a complete audit trail back to the underlying meter data.
The methodology Aristotle uses is published as open methodology through the OpenEAC Alliance, where it is subject to peer review and public comment.