Some offers promise 5 to 20% savings through energy monitoring[1][1]Press and announcementsActemium, “Monitoring et energy management” (Monitoring and energy management), offer page announcing savings of 5 to 20%See in sources ↓. None of them says how that saving will be measured. In the field, the meters report data, each dashboard is added to the previous one and consumption does not go down. Here we follow the order that changes the outcome. It comes in four steps: scope the consumption tracking, make the data reliable, measure against an adjusted baseline, then prove.
What is energy monitoring and where does it stop compared with the BMS?
Collect, analyze, alert: what energy monitoring covers
Energy monitoring means collecting, analyzing and reporting the energy consumption of a commercial building or an industrial site. Its measurement interval ranges from the monthly reading to near real time. A monitoring system observes and alerts. It cannot control equipment: that is the job of the BMS (building management system).
This tool is part of energy management. Energy management is defined as an approach, the EnMS (energy management system, called SMÉ in France) within the meaning of ISO 50001[2][2]Standards and protocolsISO, “ISO 50001:2018, Energy management systems”, revision of the version published in 2011See in sources ↓, equipped with software, the EMS (energy management system), which is only as good as its data, i.e. the metering. Setting it up is therefore a matter of the company's energy management before it is the purchase of a screen.
From monthly readings to alerts: which time interval to choose?
Each time interval tells a different story. The monthly bill observes. The load curve shows the shape of the demand. The hourly interval reveals the night baseload, what keeps running in an empty building.
At a 10 to 15-minute interval, the measurement becomes fine enough to trigger an alert. The finer the interval, the earlier consumption tracking spots a drift.
Energy monitoring or energy audit: what is the difference?
An audit is a photo, monitoring is a film. Carrying out an audit records a state, an analysis of energy uses and an action plan on a given date. Energy tracking then checks, month after month, that the energy performance promised by that plan really holds.
Monitoring or supervision: which one looks at what?
Supervision displays BMS points in real time: a valve, a setpoint, an alarm. Monitoring tracks consumption over months and years. Energy supervision remains a screen sitting on top of the BMS or of a supervision-only system (called GTC in France), two terms used almost interchangeably in French quotes. The hypervisor, for its part, aggregates several sites and several BMS brands.
A dashboard observes, the BMS corrects
The EMS detects the drift, the BMS corrects it: an EMS that does not write to the BMS remains a dashboard. The distinction carries a lot of weight when signing a quote for a platform.
An energy management system measures and alerts. The BMS acts on each piece of equipment. What remains is to pin down exactly what each of the systems competing for your quote does.
Monitoring, BMS control and predictive control compared
Monitoring, control and predictive control often end up in the same quote. Yet they do not act at the same point in the system: the table below compares them on six criteria, from the data source to what they make it possible to prove.
| Criterion | Monitoring (dashboard) | BMS control (setpoints, sequences, schedules) | Predictive control |
|---|---|---|---|
| What the system does | It measures, compares and alerts. | It applies the programmed setpoints, sequences and schedules. | It recalculates setpoints from a forecast. |
| Data source | Bills, meters, sub-meters, trend-logged BMS points. | Sensors and equipment states, read in real time. | BMS history, weather, solar gain, occupancy, building thermal inertia. |
| Frequency | From the monthly reading to a 10 to 15-minute interval. | Continuous, according to the programming. | Setpoints recalculated every 15 minutes, 24/7. |
| Who acts and when | A human, after reading the alert. | The controller, according to the rules written at commissioning. | The algorithm, which writes to the BMS within comfort constraints. |
| Direct effect on consumption | None The drift becomes visible, it is not corrected. |
Yes If the sequences are consistent and free of manual overrides. |
Yes On the hot water and chilled water loops and on supply air. |
| What it makes it possible to prove | That a drift exists. | Compliance with setpoints and schedules, if the points are trend-logged. | A gap, only against an adjusted baseline (HDD, occupancy). |
Look at the "direct effect on consumption" row: only control and predictive control act to reduce consumption. The latter is useless on a BMS or supervision-only system that does not already hold its setpoints. Hence the order we apply: make reliable first, optimize next, AI only if the complexity of the portfolio justifies it.
What tracking energy consumption really makes possible
Spotting a drift before it reaches the bill
Energy drift gives no warning. A night baseload creeps up. Manual overrides stay active after a maintenance visit. An AHU (air handling unit) overconsumes a little more each day, until the filter-change alert.
The benefits of monitoring lie there: spotting these gaps before the bill reveals them. Energy savings come afterwards, from the correction made in the BMS.
What is the link between energy monitoring and the Décret tertiaire?
The Décret tertiaire (France's tertiary decree on energy savings in commercial buildings) requires an annual declaration of consumption on OPERAT, the French Observatory of energy performance, renovation and actions in the tertiary sector. Monitoring provides the material needed for this declaration, provided the meters are reliable.
The Décret tertiaire sets the targets, the BACS decree No. 2020-887 (French decree on building automation and control systems) imposes the automation tool: it has covered commercial buildings whose heating or air conditioning exceeds 290 kW of rated output since 1 January 2025[3][3]RegulationsFrench Republic, “Decree No. 2020-887 of 20 July 2020 on building automation and control systems in non-residential buildings and automatic heat regulation”, known as the BACS decree, codified in Article R. 175-2 of the French Construction and Housing Code (Code de la construction et de l’habitation), LégifranceSee in sources ↓. For existing buildings of 70 to 290 kW, decree No. 2025-1343 postponed the deadline from 2027 to 1 January 2030 at the latest: the obligation applies earlier if the system is replaced before then[4][4]RegulationsFrench Republic, “Decree No. 2025-1343 of 26 December 2025”, postponing from 2027 to 1 January 2030 the deadline for existing buildings of 70 to 290 kW, LégifranceSee in sources ↓. ISO 50001[2][2]Standards and protocolsISO, “ISO 50001:2018, Energy management systems”, revision of the version published in 2011See in sources ↓, the ISO management standard, frames the approach: its ISO certification applies to a company, never to software.
Setting up energy monitoring in 6 steps
Step 1: inventory the meters, BMS points and trend logs and make them reliable
Before buying a screen, you need to know which meters are reliable. Setting up monitoring therefore starts with an inventory of the meters, sub-meters, BMS points and trend logs actually kept. This survey is the basis of the metering plan. Here is the path a piece of data follows, link by link, with the place where it gets lost and the test that shows it.
Each fault in the diagram can be tested on the existing supervision system, before any purchase. A gap in the trend log discovered after signing distorts the entire baseline year. The question of which data an EMS needs is therefore settled upstream, with your operations team.
Have this chain checked before signing for a platform. An audit of the meters and trend logs tells you what a dashboard will really be able to show.
Step 2: choose the indicators and the metering plan
Four indicators are enough to start: kWh/m², the night baseload, the occupancy ratio and kWh per HDD (heating degree days, called DJU in France). Each one is tied to a specific meter in the metering plan.
In the ISO 50001 approach[2][2]Standards and protocolsISO, “ISO 50001:2018, Energy management systems”, revision of the version published in 2011See in sources ↓, the EnPI, or energy performance indicator (IPÉ in French), is a component of the EnMS. The rule to lay down remains simple: an indicator without a reliable meter behind it is worthless.
Step 3: collect data from the existing installation
Data collection starts from the bills, meters and BMS points already in place, which cover most needs once the step 1 inventory is done. An IoT solution on a wireless network only comes as a complement, for what the installation does not yet measure.
Step 4: transmit and log trends without gaps (BACnet, LonWorks, archiving interval)
Transmission goes through the protocols already present on the installation: BACnet or LonWorks, often shortened to LON. At commissioning, three tests are needed. Does each point report? Is the unit correct? Is the archiving interval actually kept?
Trend logging not tested at handover becomes a gap in the energy data, discovered a year later. No later technical check can rebuild a trend log that was never recorded.
Step 5: analyze the consumption curves
Analyzing a consumption curve starts with the hourly profiles, then with the night baseload, the residual consumption when the building is empty.
Take a weekend curve: a plateau that does not come down on Saturday often signals forgotten manual overrides on an AHU. Dashboards do not name the faulty equipment. Consumption tracking gives the time, the operator traces it back to the BMS point.
Step 6: what to do with an alert?
An alert that triggers no action in the BMS is noise. Acting on it takes four steps:
- Qualify the alert: bad data or real drift.
- Identify the equipment involved.
- Correct it in the BMS: setpoint, schedule or sequence.
- Check the effect on the next curve.
Without this loop, energy monitoring remains an observation.
Tracking tools and energy costs: prove the saving before announcing it
Which sensors to add when the BMS already has hundreds?
An IoT solution is justified for an isolated meter that was never connected, such as an industrial site feeder left outside the BMS. It is not justified for a quantity the BMS already measures. IoT technologies and their wireless network complement data collection. They do not replace the temperature, presence or flow sensors that already control the equipment.
Can you do monitoring with Enedis data alone?
The Enedis load curve (Enedis is the French electricity distribution network operator) shows the shape of the electricity demand. It says nothing about end uses, nor about gas, district heating or chilled water. Monitoring limited to this curve sees that consumption is drifting, without knowing what or where: it is electricity consumption tracking, not a diagnosis.
Consumption tracking software: the functions to require
Platform, EMS or energy management system: the same software under three names. Consumption tracking software must provide HDD normalization, drift detection, alerts and export for OPERAT. The overall view of several buildings is the job of the hypervisor. An application that stacks dashboards leaves the real question untouched: is the displayed drop real?
Why does a drop on the dashboard prove nothing without an adjusted baseline?
A mild winter is enough to manufacture a false saving. A visible drop proves nothing until it is compared with a baseline adjusted for weather and occupancy. That is the principle of the International Performance Measurement and Verification Protocol (IPMVP)[5][5]Standards and protocolsEfficiency Valuation Organization, “International Performance Measurement and Verification Protocol (IPMVP)”, measurement and verification of energy savingsSee in sources ↓. The dashboard, for its part, only knows raw kWh. The calculator below applies heating degree day (HDD) correction to your own heating figures.
If measured consumption falls below expected consumption, the gap becomes defensible. If it stays above, the raw drop displayed on the screen comes from the weather, not from an action. Only an adjusted baseline distinguishes real energy savings from a seasonal effect.
What Foobot does with your consumption data
BMS advisory: make the meters reliable before buying yet another screen
Our BMS advisory is aimed at the company's Procurement, Works and Operations teams: it stands on its own, with no mandatory AI phase afterwards. The Meter Reliability deliverable tracks down disconnections, data gaps, duplicates, invisible drifts and consumption peaks. We go over the metering plan point by point, like a second commissioning.
Make reliable, measure, act: the order that makes monitoring useful
Remember three verbs, in this order: make reliable, measure, act. Make the data chain reliable first, otherwise every alert argues over a faulty meter. Then measure against an adjusted baseline, otherwise the weather takes credit for your savings. Finally act in the BMS, the only place where a setpoint really changes consumption.
Useful monitoring is judged on one thing only: the share of its alerts that end in a corrective action in the BMS, then in an improvement measured month after month. To get started on your own data, our EMS is free: it provides tracking and alerts, without portfolio ESG reporting or multi-country support.
FAQ - energy monitoring
Why set up energy monitoring?
To see an energy drift before the bill does: a baseload creeping up, forgotten manual overrides, an AHU overconsuming. Setting it up is only worthwhile if each alert leads to a correction in the BMS.
How do you configure effective energy monitoring?
Start from a verified metering plan: which meters report without gaps, with which unit. Then choose an hourly interval to read the baseload, and a 10 to 15-minute interval to raise alerts. Test trend logging as soon as the system is commissioned.
Which tools are used to control energy consumption?
Three tools, three roles. EMS-type software collects, normalizes and alerts. The BMS acts on the equipment. IoT technologies only complement meters that are not connected. An EMS that does not write to the BMS remains a dashboard.
What is an energy monitoring system?
It is the chain used to collect, analyze and alert on consumption: meters and sensors, controllers, transmission, trend logging and then analysis software. Its value depends on the weakest link, often a forgotten meter or a trend log full of gaps.
How do you draw up an energy balance?
Gather twelve months of bills and meter readings, by energy and by end use where the metering allows. Correct heating for HDD to compare two years. This consumption tracking then becomes your baseline period.
What are energy performance indicators?
EnPIs, a component of the EnMS within the meaning of ISO 50001[2][2]Standards and protocolsISO, “ISO 50001:2018, Energy management systems”, revision of the version published in 2011See in sources ↓, measure the energy performance of an end use. The most common in commercial buildings: kWh/m², night baseload, occupancy ratio and kWh per HDD. Each one is only as good as its meter.
Which indicators should be used for energy control?
To control, monitor what triggers an action: the night baseload, manual overrides visible outside occupied hours and the gap to the adjusted baseline. Read every 10 to 15 minutes, they feed alerts that the BMS handles.
Sources and references
5 sources.- [3]French Republic, “Decree No. 2020-887 of 20 July 2020 on building automation and control systems in non-residential buildings and automatic heat regulation”, known as the BACS decree, codified in Article R. 175-2 of the French Construction and Housing Code (Code de la construction et de l’habitation), Légifrance20 July 2020↑
- [4]French Republic, “Decree No. 2025-1343 of 26 December 2025”, postponing from 2027 to 1 January 2030 the deadline for existing buildings of 70 to 290 kW, Légifrance26 December 2025↑