How HVAC Energy Savings Are Measured

Key takeaways

• Savings cannot be metered directly; they are the difference between a baseline and a reporting period, plus adjustments.
• Raw before-and-after utility bills mix the effect of a project with weather, occupancy, schedules and rate changes.
• IPMVP, ASHRAE Guideline 14, the FEMP M&V Guidelines (Version 5.0, 2024) and DOE's Uniform Methods Project are the main U.S. references.
• Small savings on one piece of equipment are easily lost in whole-building data; measure at the equipment where possible.
• Ask for the baseline, the adjustment method and the measurement boundary before you accept any savings number.

Every efficiency proposal includes a savings number. Before you put that number in a budget, you need to know how it was produced. This guide explains how energy savings from HVAC-R projects are measured in practice, why the obvious method (comparing utility bills) is usually wrong, and what the recognized frameworks require.

Savings are an absence, so they have to be estimated

A meter records energy that was used. It cannot record energy that was not used. The Federal Energy Management Program (FEMP) puts it plainly in its current M&V Guidelines, Version 5.0: savings typically cannot be measured directly because they represent the absence of energy use. The Efficiency Valuation Organization (EVO), which publishes the International Performance Measurement and Verification Protocol (IPMVP), says the same in its Generally Accepted M&V Principles.

Instead, savings are determined by comparing measured use before and after a change and correcting for anything else that changed. Both FEMP and EVO express this as one general equation:

Savings = (Baseline period energy − Reporting period energy) ± Adjustments

• Baseline period: the measured "before" period, with the conditions that applied at the time.
• Reporting period (FEMP also says post-installation or performance period): the measured "after" period.
• Adjustments: corrections that put both periods on the same footing. EVO separates them into routine adjustments for factors expected to vary, such as weather or production, and non-routine adjustments for factors that are supposed to stay fixed but changed anyway, such as floor area, occupancy type or the number of shifts.

Because savings are estimated, every savings figure carries some uncertainty. FEMP describes the goal of M&V as reducing that uncertainty to an acceptable level. It cannot be removed entirely.

Why comparing utility bills misleads

The tempting shortcut is to compare this summer's bill with last summer's. FEMP notes that bill comparison was the most common method in the early days of the energy services industry, and that it often caused problems in buildings with varying energy use. Bills move for many reasons that have nothing to do with the equipment you treated or replaced.

A simple illustration: suppose a building's chillers become more efficient, but the following summer is noticeably hotter. Total cooling load goes up, and the bill may barely move or even rise. A raw comparison says "no savings." A properly adjusted comparison asks what the old equipment would have used under the new, hotter conditions, and measures savings against that. The reverse also happens: a mild summer can make an ineffective product look good. For more on the weather side, see hotter summers and cooling load.

Performance versus use

FEMP describes two factors behind any energy saving. Performance is the rate at which energy is used to do a job. Use is how much of the job is required. For a light fixture, performance is watts and use is operating hours. For a chiller, FEMP explains that performance is the energy required to provide a given amount of cooling, which varies with load (kW per ton), while use is the building's cooling load profile and the total cooling required. Both must be known to calculate savings.

This matters for HVAC-R measures that change efficiency rather than run time. If a treatment or upgrade lowers kW per ton, the energy saved depends on how many ton-hours the system delivers, and that depends on weather and occupancy. A credible method measures performance under comparable conditions and then applies it to a defined load. If kW per ton and related metrics are unfamiliar, see kW/ton, EER and COP explained.

The frameworks that define good practice

These documents fit together rather than compete. FEMP describes its guideline as a specific application of IPMVP that is intended to be fully compatible with it. The FEMP 4.0 edition noted that Guideline 14's approaches closely support the IPMVP options, except that Guideline 14 has no direct parallel to IPMVP Option A. Utility evaluators often use the Uniform Methods Project protocols, which are themselves based on IPMVP options. The IPMVP options guide covers the four options in detail.

The measurement boundary

Before measuring anything, you decide where to draw the line. EVO calls this the measurement boundary. If the goal is to verify the savings from specific equipment, the boundary is drawn around that equipment and its energy flows (a retrofit-isolation approach, IPMVP Options A or B). If the goal is total facility performance, the boundary is the utility meter (Option C). Energy effects outside the boundary are called interactive effects. EVO says they must be estimated or evaluated. They can be ignored only if the M&V plan explains each one and shows it is small compared with the main savings.

For a measure applied to individual HVAC-R units, the boundary choice is often decisive. FEMP 5.0 says whole-facility (Option C) methods should be used only when savings are predicted to exceed about 10% to 15% of the consumption on the meter, and warns that smaller percentage savings risk being "lost in the noise" with monthly data. If a building has one chiller among many loads, even a meaningful improvement at the chiller may be invisible on the utility bill. Measuring at the equipment avoids that problem.

Six principles to test any savings claim against

EVO lists six principles that underpin IPMVP adherence. They make a practical checklist.

EVO also says that an independent reviewer of reported savings should confirm the measure is based on sound scientific principles and that independent evidence supports any claims made before measurement. That is a reasonable standard to apply to every vendor.

How much M&V is enough?

More measurement reduces uncertainty, but it costs money. FEMP's rule of thumb, based on DOE's performance contracting history, puts overall annual M&V costs at about 1.5% to 3% of typical annual guaranteed cost savings for federal energy savings performance contracts. Smaller projects and pilots work differently, but the principle is the same: scale the effort to the value and the risk.

Instrument accuracy matters more than many buyers expect. FEMP's chiller example explains that cooling capacity in tons is calculated from water flow and the temperature difference across the chiller, while power comes from an electrical meter. When the temperature difference is small, small temperature-sensor errors become large percentage errors in kW per ton. FEMP says uncertainty at project level comes from four sources (measurement, sampling, estimation and modeling) and is combined statistically. A report that gives a savings percentage with no word on sensor accuracy is leaving out half the story.

A practical sequence for any HVAC-R efficiency project

1. Define the question. Do you need to know what one measure saves on specific equipment, or what the whole building saves?
2. Draw the boundary and pick the option. Retrofit isolation for equipment-level measures; whole-facility only when the expected savings are large relative to the meter.
3. Write the M&V plan before installation. EVO says that for performance contracts the plan should be agreed before the measures are installed, because the baseline cannot be recreated afterward.
4. Measure the baseline across the range of conditions the equipment normally sees, and record the static factors (set points, schedules, occupancy).
5. Install, then verify operation. EVO calls this operational verification: confirming the measure is installed and working before counting savings.
6. Measure the reporting period with the same instruments and method.
7. Adjust and report. Apply routine and non-routine adjustments, state the uncertainty, and only then convert energy into dollars.

The details of baselines and adjustments are covered in baselines and weather normalization. If you already have a report in hand, use how to read an M&V report. If you need the savings to support a utility incentive, see custom utility rebates and M&V.

Where CryogenX4 fits

CryogenX4 is a one-time treatment injected into existing HVAC-R systems while they run. According to the company, it lifts insulating oil film from internal heat-exchanger surfaces and returns the oil to the sump, which is a change in performance (energy per unit of cooling) rather than in run hours. That makes it a retrofit-isolation question: the effect is best measured at the treated equipment, under comparable conditions, before and after treatment.

The company describes an IPMVP-style field method for doing that: minute-by-minute monitoring of indoor and outdoor temperature, humidity, chilled-water or air flow and energy use over a testing period of usually 3 to 9 months, with a baseline captured before treatment and data compared under similar weather and building load conditions. CryogenX4 reports energy savings of up to 30%, and results vary by equipment condition. Field measurement on your own equipment is how you find out where your systems fall. The method is described in what a CryogenX4 pilot measures.

Next step

Before you approve any efficiency spend, ask the vendor for three things in writing: the measurement boundary, how the baseline will be measured, and how weather and operating changes will be adjusted for. If you would like to see how a CryogenX4 field verification would be set up on your equipment, request a pilot discussion through our contact page.

Frequently asked questions

What is the difference between routine and non-routine adjustments?

Routine adjustments correct for factors that are expected to vary, such as outdoor temperature or production volume, usually through a model. Non-routine adjustments correct for one-off changes to factors that were supposed to stay fixed, such as added floor area, a new shift or changed set points.

Which M&V document should my vendor follow?

IPMVP is the most widely referenced framework. ASHRAE Guideline 14 supplies the technical methods and statistics, and federal projects use the FEMP M&V Guidelines (Version 5.0, 2024). A good plan names the framework and the specific option it uses.

Can whole-building data show savings from treating a few HVAC units?

Usually not on its own. FEMP recommends whole-facility methods only when savings exceed roughly 10% to 15% of the metered consumption, and warns that smaller savings can be lost in the noise. Equipment-level measurement is more suitable.

Does M&V have to be expensive?

It should be scaled to the value at stake. FEMP's rule of thumb for federal performance contracts is about 1.5% to 3% of annual guaranteed cost savings. A short pilot on a few units can use focused, equipment-level monitoring.

Sources

1. M&V Guidelines: Measurement and Verification for Performance-Based Contracts, Version 5.0 (September 2024) — U.S. Department of Energy, Federal Energy Management Program
2. M&V Guidelines: Measurement and Verification for Performance-Based Contracts, Version 4.0 (November 2015) — U.S. Department of Energy, Federal Energy Management Program
3. IPMVP Generally Accepted M&V Principles (October 2018) — Efficiency Valuation Organization (EVO)
4. International Performance Measurement and Verification Protocol (IPMVP) — Efficiency Valuation Organization (EVO)
5. Release of the new IPMVP Core Concepts 2022 — Efficiency Valuation Organization (EVO)
6. Guideline 14-2023: additional material (contents, retrofit-isolation metering and whole-building model forms) — ASHRAE Journal
7. The Uniform Methods Project: Methods for Determining Energy Efficiency Savings for Specific Measures — U.S. Department of Energy