An energy saving cannot be measured. It is consumption that did not happen, so it is a calculation. Whoever writes the calculation controls the result. In an energy performance project, the measurement and verification plan annexed to the contract often has a single author: the party selling the saving. Here is what this document must contain for your building and what to refuse before signing.
The measurement and verification plan: the rules of the game for a performance commitment
What an M&V plan is and where it fits in the IPMVP protocol
The measurement and verification plan (M&V plan, called PMV in France) sets, before the works or commissioning, the method for calculating an energy saving. It defines the measurement boundary, the baseline period, the adjustments, the meters and the report format. It is the document on which any M&V plan-based measurement and verification compliant with the IPMVP protocol rests.
IPMVP, the International Performance Measurement and Verification Protocol, has existed since 1997 and is maintained by EVO, the Efficiency Valuation Organization[1][1]Studies and reportsEfficiency Valuation Organization (EVO), “International Performance Measurement and Verification Protocol (IPMVP)”See in sources ↓. Its fundamental principles boil down to one idea: a saving is obtained as the difference between an adjusted baseline and a measured consumption[1][1]Studies and reportsEfficiency Valuation Organization (EVO), “International Performance Measurement and Verification Protocol (IPMVP)”See in sources ↓.
The ISO 50015 and ISO 50001 standards complement this framework without replacing it[2][2]Studies and reportsISO, “ISO 50015:2014, Measurement and verification of energy performance of organizations” and “ISO 50001:2018, Energy management systems”See in sources ↓. An ISO 50001 ISO certification attests to an energy management system. It does not guarantee the accuracy of a claimed saving.
What is an M&V plan for in an EPC or a performance guarantee?
In an energy performance contract or a performance guarantee, the M&V plan of an EPC turns a promise into a contractually enforceable saving. Its purpose: to underpin the bonuses, the penalties and the return on investment of your works.
An energy efficiency project without an M&V plan guarantees a figure that nobody can recalculate. A specification for a global performance contract (marché global de performance, a French public procurement vehicle) illustrates this: its consumption cap is only as good as the measurement and verification method that comes with it[3][3]PDFStudies and reportsEGF BTP, “Cahier des charges d’un marché global de performance” (specification for a global performance contract)See in sources ↓.
Did you know? The Éco Énergie Tertiaire scheme (the Décret tertiaire, France's tertiary decree on energy savings in commercial buildings) tracks the annual consumption of every commercial building of 1,000 m² or more on the OPERAT platform[4][4]RegulationsLégifrance, “Decree No. 2019-771 of 23 July 2019 on obligations to reduce final energy consumption in commercial buildings” (Décret tertiaire)See in sources ↓. It does not say which action produced the reduction. Nor does the decree require IPMVP.
Who writes the M&V plan, and at what stage of the project?
The usual author is the contract holder or an engineering firm it appoints. The building owner rarely writes it. The document is annexed to the contract, as any serious EPC template provides, before the works or before the control system goes live.
So your work happens before signing. Past that point, implementation follows a method that is already locked in.
The author's competence matters as much as the service delivered. EVO's PMVA and PMVE credentials[1][1]Studies and reportsEfficiency Valuation Organization (EVO), “International Performance Measurement and Verification Protocol (IPMVP)”See in sources ↓, the AEE's CMVP and the certificate from AFNOR Certification, part of the AFNOR group[7][7]Studies and reportsAFNOR Certification, “Certification de personnes Référent en mesure et vérification de la performance énergétique” (personnel certification as a measurement and verification of energy performance referent)See in sources ↓, validate training in the international measurement protocol. We come back to this below.
Four links, a single author of the calculation
Scroll to see the four links →- Link 1 Building owner You, who pay for the saving WritesRarely the plan. Signs the contract annex. MeasuresIts bills, rarely more. Can challengeThe baseline and the adjustments, if it holds the raw data.
- Link 2 Contract holder EPC, performance guarantee or AI control WritesUsually the plan, then the M&V report. MeasuresWith its own meters and tool. Can challengeOperational drifts that eat into its saving.
- Link 3 O&M contractor Day-to-day operation of the installations WritesSetpoints and overrides (manual overrides). MeasuresMeter readings and BMS trend logs. Can challengeA baseline that ignores operating constraints.
- Link 4 Third-party verifier No saving to sell WritesA reasoned opinion on the calculation. MeasuresRedoes the calculation from the raw data. Can challengeAny result it cannot reproduce.
- Who set the baseline?
- Who decides on non-routine adjustments?
- Who holds the raw data?
- Who pays for verification?
- Who settles a disagreement?
The content of an IPMVP-compliant M&V plan, section by section
Sections 1 and 2: the measured action, the IPMVP option and the measurement boundary
An IPMVP-compliant M&V plan is organized into thirteen sections in the 2012 and 2016 editions; the 2022 edition adds two, the facility overview and the operational verification requirements[1][1]Studies and reportsEfficiency Valuation Organization (EVO), “International Performance Measurement and Verification Protocol (IPMVP)”See in sources ↓:
- The energy conservation measure
- The IPMVP option and the measurement boundary
- The baseline period
- The reporting period
- The basis for adjustment
- The analysis procedure
- The energy price used
- Meter specifications
- Monitoring responsibilities
- Expected accuracy
- The M&V budget
- Report format
- Quality assurance
The choice of option, in section 2, determines everything else. Controls tuning without works is not measured like a generator replacement. The following table breaks down the IPMVP option choice across four common actions in a commercial building project.
| Action | Most common option | Measurement boundary | Main adjustments | Baseline duration | Watch point |
|---|---|---|---|---|---|
| Generator replacement With works |
A B | Energy meter on the generator output. | Heating degree days (HDD). | One full heating season. | The gap between the measured efficiency and the consumption read at the utility meter. |
| Envelope insulation With works |
C or even D | Main building meters. | HDD, occupancy, floor areas. | 12 months minimum. | A small effect, drowned out by the building's other end uses. |
| AHU replacement With works |
B | Fan meter, coil energy. | Operating hours, occupancy. | A full annual cycle as soon as the coils depend on the weather. | Overrides (manual overrides) that distort the baseline. |
| HVAC controls tuning or AI control on the existing BMS No works |
B if HVAC is sub-metered, otherwise C | Meters and BMS trend logs. | HDD, solar radiation, occupancy. | A full annual cycle. | Whether the saving holds over time, long after commissioning. |
The last row of the table directly concerns control projects: option B or C requires a full operating cycle, in practice twelve months, for the baseline and then for the reporting period. A controls saving weighs little against the building's other end uses. Over too short a period, it disappears into the meter noise.
Sections 8 to 10: meters, monitoring responsibilities and expected accuracy
The M&V tools are well known: meters, BMS trend logging and energy management software, often called GTE in France (gestion technique énergétique). Section 8 specifies their technical specifications and the time step, hourly or daily. Section 9 designates who reads what and who checks the quality of each data point.
Section 10 sets the expected accuracy, in other words the accepted statistical uncertainty. Without it, setting up measurement and verification becomes a matter of interpretation, with everyone using their own tool.
What most M&V plans leave out, and how to review the operator's plan
A baseline built on data that nobody has qualified
Would your baseline stand up to a third party?
Answer criterion by criterion. The verdict is indicative only: it always has to be confirmed on the building's actual historical data.- 1 Do you have at least 12 months of continuous history, including a heating season and a cooling season?
- 2 Do the meters within the boundary record at least at hourly intervals?
- 3 Have data gaps and duplicates been identified and handled?
- 4 Is each meter assigned with certainty to the end use or zone being measured?
- 5 Did the meters stay connected and unchanged throughout the period?
- 6 Is the period free of any active override (manual override)?
- 7 Are HDD, occupancy and floor areas available for the same period?
- 8 Can the building owner access the raw data outside the operator's tool?
A baseline calculated on patchy historical data, on wrongly assigned meters or over a period with overrides (manual overrides) distorts every saving derived from it. Yet few plans require this data to be qualified before it is locked in.
Regulations often provide the raw material. The BACS decree No. 2020-887 (French decree on building automation and control systems) requires hourly monitoring of consumption by functional zone[5][5]RegulationsLégifrance, “Decree No. 2020-887 of 20 July 2020 on building automation and control systems in non-residential buildings and automatic heat regulation” (BACS decree)See in sources ↓. Buildings above 290 kW have been subject to it since 1 January 2025, those from 70 to 290 kW from 1 January 2030 following a postponement[6][6]RegulationsLégifrance, “Decree No. 2025-1343 of 26 December 2025 amending the application dates of obligations concerning building automation and control systems in commercial buildings”, postponing the deadline for 70 to 290 kW buildings to 1 January 2030 | JORF of 27/12/2025See in sources ↓. Here are the eight questions to ask yourself before approving the baseline, whether the history comes from the BMS or from an energy management tool.
Every “No” or “I don't know” deserves a written reservation in the contract. Also demand access to the raw data: a measurement plan that refers back to the operator's tool cannot be verified.
A plan silent on how the saving holds when operations drift
A control saving does not hold on its own. Overrides (manual overrides) pile up, a setpoint goes back up after a VIP complaint (the top-floor director who demands his 24 °C all year round, seen many times) and energy drift sets in. None of this calls for works.
A plan that stops measuring after handover certifies a saving it will never track. A poorly chosen IPMVP option or too short a reporting period makes this drift invisible. On a control project, the promised energy efficiency must be tracked for the entire term of the contract.
Have the plan checked by a certified practitioner
A review is only as good as the reviewer. The CMVP certification (Certified Measurement and Verification Professional) is issued by the AEE, the Association of Energy Engineers. EVO stopped co-signing it in 2022 and issues its own credentials, PMVA and PMVE[1][1]Studies and reportsEfficiency Valuation Organization (EVO), “International Performance Measurement and Verification Protocol (IPMVP)”See in sources ↓. AFNOR Certification, for its part, certifies individuals as “Référent en mesure et vérification de la performance énergétique” (measurement and verification of energy performance referent)[7][7]Studies and reportsAFNOR Certification, “Certification de personnes Référent en mesure et vérification de la performance énergétique” (personnel certification as a measurement and verification of energy performance referent)See in sources ↓. These credentials attest to individual training and competence. So ask for the name of the certified person who signs the plan, not just the company name.
What Foobot brings to a measurement and verification plan
BMS advisory: make the meters reliable before locking in the baseline
A baseline is no better than the meters that feed it. With BMS advisory, we first track down disconnections, data gaps, duplicates and invisible drifts in the building's historical data. Setting up usable, trend-logged metering comes next. This work is valuable in itself, whether or not a performance commitment follows.
AI Control: savings measured according to IPMVP with HDD correction
Proof before promise. At Valeo, AI Control shows -28.6% in heating and electricity consumption, measured according to the IPMVP protocol[8][8]Field feedbackFoobot, Valeo case study, HVAC consumption measured according to IPMVPSee in sources ↓. This is a measured case, not a commitment transferable to your building. The control loop runs every 15 minutes, 24/7, through the existing BMS, with no major works. The energy performance gains remain auditable by a third party.
An M&V plan is negotiated before the contract is signed
An M&V plan signed as the operator wrote it sets the rules of a game you no longer control. Demand the fifteen sections of IPMVP 2022 and a baseline built on qualified data. Refuse one-sided clauses: adjustments left to the contract holder, captive raw data, a report without sources. Then have the document reviewed by someone with no saving to sell. The right time is before signing, not after the first verification. For a review of your draft M&V plan, you can contact us.
have your M&V plan reviewed or written
Further reading:
FAQ - measurement and verification plan
What is a Measurement and Verification Plan?
A measurement plan sets, before the works or commissioning, the method for calculating a saving under IPMVP: measurement boundary, baseline, adjustments, meters and reporting. Annexed to the contract, it makes the saving recalculable by a third party.
What is at stake with Measurement and Verification?
Measurement and verification makes a saving contractually enforceable. It underpins the bonuses, the penalties and the return on investment of a performance guarantee. Without it, the contract promises a figure that nobody can recalculate.
How do you build this topic effectively into an energy performance project?
Include the M&V plan from the tender stage, before the works, then annex it to the contract. Implementing it requires reliable meters and a qualified baseline. An ISO 50001 approach structures the monitoring, without replacing the plan.
What are the principles of the IPMVP protocol?
IPMVP rests on six fundamental principles: accuracy, completeness, conservativeness, consistency, relevance and transparency. They bind the author of the plan as much as the verifier. An organization-level ISO certification does not cover them: they apply to the savings calculation itself.
How do you assess energy savings?
You compare the adjusted baseline consumption with the consumption measured during the reporting period, non-routine adjustments included. The chosen IPMVP option sets the boundary: isolated measure, whole building or calibrated simulation. A saving is always calculated.
What tools are used for measurement and verification?
The tools are energy meters, BMS trend logging and energy management software (GTE in France). The best tool is not enough: measurement and verification requires raw data accessible at hourly intervals, over the whole period.
Sources and references
8 sources.- [1]Efficiency Valuation Organization (EVO), “International Performance Measurement and Verification Protocol (IPMVP)”↑ 1 2 3 4 5
- [2]ISO, “ISO 50015:2014, Measurement and verification of energy performance of organizations” and “ISO 50001:2018, Energy management systems”↑
- [3]EGF BTP, “Cahier des charges d’un marché global de performance” (specification for a global performance contract)PDF↑
- [7]AFNOR Certification, “Certification de personnes Référent en mesure et vérification de la performance énergétique” (personnel certification as a measurement and verification of energy performance referent)↑ 1 2
- [4]Légifrance, “Decree No. 2019-771 of 23 July 2019 on obligations to reduce final energy consumption in commercial buildings” (Décret tertiaire)↑
- [5]Légifrance, “Decree No. 2020-887 of 20 July 2020 on building automation and control systems in non-residential buildings and automatic heat regulation” (BACS decree)↑
- [6]Légifrance, “Decree No. 2025-1343 of 26 December 2025 amending the application dates of obligations concerning building automation and control systems in commercial buildings”, postponing the deadline for 70 to 290 kW buildings to 1 January 2030 | JORF of 27/12/2025↑
- [8]Foobot, Valeo case study, HVAC consumption measured according to IPMVP↑