Proving savings - measurement and verification for BMS improvements

"We tuned the controls and the building uses less energy" is a claim. "Here is the baseline, here is the weather-corrected comparison, and here is the cumulative saving since the change" is evidence. Clients renew service contracts and fund improvement projects on evidence.

Measurement and verification (M&V) is the discipline of showing what an improvement actually achieved. This guide covers the essentials for BMS work: baselines, degree-day normalisation, CUSUM, metering coverage and how to report the result.

Why M&V matters for service work

Energy use moves for many reasons that have nothing to do with the controls: a cold winter, a new tenant, longer opening hours, a change of equipment. Comparing this year's bill with last year's tells you very little on its own.

M&V separates the effect of your change from everything else. Done well, it protects both sides: the client sees what they paid for, and the integrator is not blamed for a rise caused by something else. The International Performance Measurement and Verification Protocol (IPMVP), published by the Efficiency Valuation Organization, is the most widely used framework for this, and its core ideas are worth knowing even for small projects.

Start with a baseline

The baseline is a model of how the building used energy before the change, so you can predict what it would have used afterwards if nothing had been done.

A good baseline has:

Agree the baseline with the client before you make the change. Agreeing it afterwards invites argument.

Normalise for weather with degree days

Heating and cooling energy depend heavily on the weather. Degree days are the usual way to account for this.

A heating degree day figure measures how far, and for how long, the outside temperature was below a base temperature - the temperature below which the building needs heating. Cooling degree days work the same way above a cooling base temperature. Degree day data for UK locations is freely available.

The standard approach is a simple regression: plot consumption for each period against degree days for the same period. For a heating-dominated building this often gives a straight line, with:

The quality of the fit matters. A good straight-line relationship gives you a reliable baseline; a poor one means something else is driving consumption, and that is worth investigating in its own right. ASHRAE Guideline 14 is a widely cited reference for judging model fit. It uses two statistics - the coefficient of variation of the root mean square error, CV(RMSE), and the normalised mean bias error, NMBE - and the figures commonly quoted from it are a CV(RMSE) of no more than 15% and an NMBE within plus or minus 5% for monthly data, loosening to 30% and plus or minus 10% for hourly data. Check the guideline itself for how it applies to your project.

The base temperature itself varies between buildings - the commonly published UK figure of 15.5 °C is a convention, and a well-insulated building with high internal gains may have a lower one. If the fit looks poor, try different base temperatures.

Track savings with CUSUM

CUSUM - cumulative sum of differences - is a simple and powerful way of showing savings over time.

For each period after the change:

Plotted over time, the CUSUM line tells a clear story. During the baseline period it wanders around zero. After a successful improvement it trends steadily in one direction - the cumulative saving. If a fault creeps back in, the slope flattens or reverses, often well before anyone would notice from the bills.

That second use is easy to overlook. CUSUM is not only a way to prove a saving once; it is a way to spot when a saving is being lost, which makes it a natural part of a monitoring-based service contract.

Metering coverage

You can only verify what you can measure. Whole-building meters show the overall effect but can hide the cause, and a small saving on one system can be lost in the noise of the whole building.

CIBSE TM39 is the UK guidance on building energy metering. Its latest revision covers thermal, water and steam metering as well as electricity and gas. It is a good reference when assessing a building's metering and recommending additions. Common gaps in existing buildings include:

Adding or connecting metering is often a sensible early project, because it makes every later improvement easier to prove.

A UK example

Most published evidence on BMS savings comes from North America, so UK examples are worth knowing. CIBSE Journal (November 2023) described a 1967 multi-let office in Croydon of around 17,600 m² where controls changes - ventilation and fan schedules matched to occupancy, simultaneous heating and cooling removed and boilers run only when needed - were reported to cut energy by 28% in nine months, about £171,000 and 232 tonnes of CO2, with no new plant. The savings were verified with a calibrated model under IPMVP. The work also turned up mechanical faults, including seized dampers and stuck valves - a reminder that optimisation and repair go together. It is one building, not a benchmark, but it shows what a well-verified result looks like.

Short-term tests

Not every measure needs a year of data. For a single change - a reset strategy enabled, a schedule tightened, a drive taken out of hand - a before-and-after comparison over a few weeks, using trends of the system concerned, can be enough. Switching the measure off and on for alternate periods is another approach, where the client accepts it.

Be clear about the limits. A short test shows that a change works as intended; it does not prove an annual saving. Say which kind of evidence you are presenting.

Common pitfalls

A few mistakes come up again and again in BMS M&V:

Reporting to the client

A good M&V report is short and honest:

Present savings in energy first, then cost and carbon using stated tariffs and factors. Avoid presenting a single figure as precise when the data does not support it. Published studies of building improvements show wide ranges of results, and a measured, honest figure for one building is worth more than a borrowed headline number.

State the conversion factor year with every carbon figure. The 2026 UK government factors cut the grid electricity factor by about a quarter, so an electrical saving reported this year counts for less carbon than the same saving last year.

The 2026 grid factor shrinks electrical carbon savings on paper - UK grid electricity, and what it does to a saving
Grid electricity, kgCO2e per kWh (generation)
2025 factors0.17700
2026 factors0.13096 (-26%)
Carbon counted for 100,000 kWh of electricity saved
2025 factors17.7 tCO2e
2026 factors13.1 tCO2e

The same kWh saving counts for about a quarter less carbon - say which factor year a report uses.

The same 100,000 kWh saving under the 2025 and 2026 factors. Source: DESNZ greenhouse gas reporting conversion factors 2026, methodology report. Flags: UK, Check against the DESNZ workbook.

How it fits together

M&V closes the loop on a service programme. A health check or service visit finds the opportunities; fault detection keeps finding them; the measures in Where BMS tuning saves energy and the seasonal checks in Seasonal changeover deliver them; and M&V proves the result. Getting the metering and trends right at commissioning makes all of it easier. What the studies really say puts published savings figures in context.

Related tools and training: the CUSUM savings tracker, degree-day normalisation, energy savings estimator and project payback calculators are in tecTools, and the Energy and building control classes course is in tecTraining.

Published guidance mentioned: IPMVP (Efficiency Valuation Organization); ASHRAE Guideline 14; CIBSE TM39 Building energy metering; CIBSE Journal, November 2023 (Croydon office case study). These are described here in our own words - refer to the documents themselves for their methods and requirements.

More articles from 3EC

3EC logo