A defensible baseline estimates what the building would likely have consumed under the weather and operating conditions that actually occurred - so the season does not take credit for your team’s work.
A raw year-on-year comparison mixes operational performance with weather, occupancy, opening hours and building changes. It can reward a mild winter, punish a hot summer or hide a genuine improvement.
If heating energy falls by 8% while the winter is 10% milder, the building may not have improved at all. The reverse is equally damaging: a better-controlled building can use more energy during a severe winter and look worse in a simple budget comparison.
A baseline solves a narrower, more useful question: given the conditions in the reporting period, what would the building probably have consumed before the improvement?
Heating and cooling degree days summarise how far outdoor temperature sat from an agreed base. They let the model compare periods by weather demand rather than by calendar label alone.
A heating degree-day calculation needs a stated base temperature, daily weather data and the building’s actual location. Cooling degree days apply the same logic above a cooling base. The source, base and missing-weather treatment must remain visible because each choice changes the result.
Degree days are not automatically the right driver for every load. Servers, lighting and hot water may behave as a base load; hotels may need occupied room-nights; factories may need production volume; offices may need operating hours or occupancy. Use only variables supported by reliable data and a defensible relationship.
A baseline is a documented model built from the building’s own metered history. Its purpose, meters, period, variables, exclusions and acceptance tests should be agreed before anyone sees the claimed saving.
A model becomes evidence only when its inputs trace to source readings, its fit is good enough for the stated use and its approved version cannot move after the improvement is known.
Consider a hypothetical office heating model with a 12,000 kWh monthly base load plus 95 kWh per heating degree day. The numbers below illustrate the method, not a client result.
A real report must also show tariff treatment, model uncertainty, data exclusions and any change in occupancy, schedule or building use. The arithmetic is simple; the governance around it is what makes the result credible.
A baseline should be allowed to fail. Publishing a confident saving from short history, weak fit or an unrecorded building change is worse than admitting that the current data cannot support the claim.
Volts can use the same authorised meter history for performance analysis, operational investigation and reporting, reducing the risk that separate spreadsheets produce separate versions of the saving.
Volts keeps the baseline, the readings behind it and the approval that fixed it on one record.
Software can maintain evidence; it cannot certify an energy-management system or guarantee an auditor’s acceptance. Certification remains the responsibility of the organisation and its auditor.
See energy-management workflows
Enough to represent the seasons and operating states the model must predict. Twelve months is a common starting point for whole-building weather models, but the acceptable minimum depends on interval, data quality and building behaviour.
Yes. Cooling degree days or another temperature relationship can model cooling-sensitive use, provided the data shows a stable and testable connection.
They are calculated from local weather data for the site.
No. A baseline compares a building with its own modelled past. A benchmark compares it with peers, a standard or another reference group.
No single statistic proves validity; residual pattern, data coverage and intended use also matter.
Record a non-routine adjustment for material changes such as floor area, tenant mix, production or operating schedule. Preserve the original model, reason, evidence, approver and effective date.
Not automatically. A whole-building result can show total performance change, but attributing simultaneous measures requires separate action records, submeter evidence or a more specific M&V design.
No. Volts can provide the baseline and the evidence behind it. Certification depends on the organisation’s complete system, documented method and auditor.
Bring one site’s meter history, weather source and known operating changes. We will test whether the available evidence can support a credible performance period - and show the gaps when it cannot.