Postponing the BACS decree (French decree 2020-887 on building automation and control systems) to 1 January 2030 for existing commercial buildings of 70 to 290 kW is a delay, not an exemption[1][1]RegulationLégifrance, “Décret n° 2025-1343 du 26 décembre 2025” (Decree No. 2025-1343 of 26 December 2025), postponing to 1 January 2030 the BACS deadline for existing commercial buildings of 70 to 290 kW (Article R. 175-2 of the French Construction and Housing Code)See in sources ↓. Yet a dashboard does not cut a single kWh until the BMS (building management system) acts.
Serious energy management answers three questions about the energy consumed by each of the company's sites. Who reads the consumption data? Who changes the settings? How do you prove the saving achieved?
Energy management in a commercial building: definition, scope and stakes
Energy management in one sentence: monitoring is not enough, you have to act
Energy management is the approach that measures a commercial building's energy consumption, detects its drifts and gets equipment settings changed. Practitioners in France also call it GTE (gestion technique énergétique, technical energy management) or energy management.
The energy management system (EMS, called SGE in French) is only its tool: software that displays data in real time, without controlling a single valve. Energy management as a discipline then organizes the approach over time, around an action plan. Without control of the equipment, the best solution is still a dashboard. Energy management starts when someone changes a setpoint.
Energy management, efficiency and performance: three distinct notions
Energy management is an organization: roles, a routine, decisions. Energy efficiency refers to a ratio between the service delivered and the energy consumed. This concept measures a yield.
Energy performance, finally, is a measured result, tracked by an EnPI (energy performance indicator). An energy saving only exists once this result has been compared with a reference consumption.
Energy management: an approach, a tool and data
Energy management is a continuous improvement approach, not software. The approach (EnMS) is equipped with software (EMS) and is only worth as much as its data: metering. The EnMS (energy management system in the ISO sense) is based on the ISO 50001 standard, published in 2011 and revised in 2018[2][2]Standards and protocolsISO, “ISO 50001:2018 Energy management systems”, revision of the version published in 2011See in sources ↓. It organizes a policy, an energy review, significant energy uses, an EnPI, a baseline and an action plan, in a PDCA cycle (plan, do, check, act). No software is required. Between operators and management, the energy manager keeps this approach and its tool alive.
Commercial building and industrial site: two scopes not to be mixed
Offices, care homes, hospitals, local authority buildings: all of this falls under the commercial (tertiary) sector. Production on an industrial site is outside our scope, as is residential.
Across a real estate portfolio, building energy management is not decided at headquarters. It is energy management building by building: each asset has its BMS, its meters and its drifts.
Décret tertiaire, BACS decree and Directive 2023/1791: what the texts require
Three texts govern the management of a commercial asset. The Éco Énergie Tertiaire scheme (the Décret tertiaire, France's tertiary decree on energy savings in commercial buildings) requires a 40% cut in final energy consumption by 2030 compared with a reference year. A threshold set in absolute value can also serve as the target[3][3]RegulationLégifrance, “Décret n° 2019-771 du 23 juillet 2019” (Decree No. 2019-771 of 23 July 2019), obligations to reduce final energy consumption in commercial buildingsSee in sources ↓. The BACS (building automation and control system) obligation requires four functions. The system monitors and analyzes consumption hourly, detects losses of efficiency, remains interoperable and allows both manual shutdown and autonomous management of equipment[4][4]RegulationLégifrance, “Décret n° 2020-887 du 20 juillet 2020” (Decree No. 2020-887 of 20 July 2020), building automation and control systems in non-residential buildings (BACS decree), amended by Decree No. 2023-259 of 7 April 2023 and codified in Articles R. 175-1 et seq. of the French Construction and Housing CodeSee in sources ↓. It has applied above 290 kW since 2025. Between 70 and 290 kW, existing buildings have until 1 January 2030. The obligation applies earlier if the heating or air conditioning is replaced before that date[1][1]RegulationLégifrance, “Décret n° 2025-1343 du 26 décembre 2025” (Decree No. 2025-1343 of 26 December 2025), postponing to 1 January 2030 the BACS deadline for existing commercial buildings of 70 to 290 kW (Article R. 175-2 of the French Construction and Housing Code)See in sources ↓.
Directive (EU) 2023/1791 makes an energy management system mandatory above a consumption threshold[5][5]RegulationEuropean Union, “Directive (EU) 2023/1791 on energy efficiency”, energy management system mandatory above a consumption thresholdSee in sources ↓. This regulatory framework imposes, it does not encourage: implementation remains the responsibility of those subject to it.
Did you know? The BACS decree does not impose any class of NF EN ISO 52120-1[6][6]Standards and protocolsAFNOR, “NF EN ISO 52120-1”, performance classes of building automation and control systemsSee in sources ↓. It sets its own functional requirements[4][4]RegulationLégifrance, “Décret n° 2020-887 du 20 juillet 2020” (Decree No. 2020-887 of 20 July 2020), building automation and control systems in non-residential buildings (BACS decree), amended by Decree No. 2023-259 of 7 April 2023 and codified in Articles R. 175-1 et seq. of the French Construction and Housing CodeSee in sources ↓. Another often forgotten point: a building's class is that of its weakest function.
Investors and tenants: consumption figures you can defend
Investors and tenants expect figures that a third party can review. Asset value, tCO2e (tonnes of CO2 equivalent), carbon footprint: everything ends up in a report or ESG reporting. For a company managing a portfolio, the stakes lie in defending these figures more than in publishing them. Without reliable metering, a low-carbon transition pathway or falling greenhouse gas emissions will not survive an audit.
Energy management tools and software: what each one measures, what it changes, how to choose
The EMS collects, normalizes and alerts without touching the equipment
The energy management system (EMS) is monitoring software. The acronym also refers to tools for the power grid or charging stations. Only buildings concern us here. The EMS collects meter and sensor readings, normalizes the data by heating degree days (HDD), detects drifts and raises alerts.
Platform, EMS, software solution: these three words refer to the same object, detailed in our guide to the energy management system. Its dashboards are used for monitoring, never to change a setpoint.
The hypervisor aggregates several sites and several BMS brands
Hypervision means multi-site supervision. The building hypervisor brings BMS points from different brands into a single platform and aligns their dashboards. Neither EMS nor BMS, it aggregates without controlling. Sales pitches readily present it as an all-in-one technology solution. In the field, the hypervisor depends on each site's metering infrastructure and BMS: it is worth no more than they are.
Predictive control and AI: useful in the face of real complexity
Predictive tools rely on artificial intelligence, often on machine learning. Their contribution lies in taking thermal inertia and solar gain into account. They also factor in warm-up time, zone by zone. This optimization is justified when a fixed sequence can no longer keep up.
AI is not an end in itself. For lighting, a presence detector is enough.
Measure, alert, control, prove: the grid of tool families
Sorting tools by brand settles nothing: the catalogs all look alike. The grid below sorts them by function, to judge energy management software on what it actually changes. It asks each family four questions before giving its limit. What does the tool measure? What does it change on the HVAC? What data does it require? Who runs it day to day?
| Tool family | What the tool measures | What it changes on the HVAC | Data required | Who runs it day to day | Typical limit |
|---|---|---|---|---|---|
| Meters and monitoring software (EMS) Measures and alerts |
Consumption per meter and IoT sensors, drifts corrected for HDD. | Changes nothing. | Reliable meter readings with trend logging. | Energy manager, energy officer. | Alerts without correcting if nobody acts. |
| BMS / supervision-only system (called GTC in France) Acts |
Equipment statuses, temperatures and alarms. | Setpoints, schedules, valves. | Points brought up and correctly addressed. | Operator, maintenance contractor. | Poor long-term trend logging and poor comparison between sites. |
| Multi-site hypervisor Aggregates |
Points aggregated from several BMSs and several brands. | Aggregates, does not control. | Access to each site's supervision. | Real estate department, FM. | Worth no more than the BMSs it reads. |
| Predictive control If real complexity |
Thermal response of the building, weather, occupancy. | Setpoints and optimal start times, via the BMS. | BMS history, weather, write access. | Control service provider together with the operator. | Useless if a simple sequence is enough. |
| BMS advisory Makes reliable |
Gaps between the specification and actual operation. | No direct write: action plan, clearing overrides (manual overrides), making meters reliable. | Points list, history, documentation. | Building owner. | Does not control on its own: relies on the operator. |
The last column carries the essentials and the reading rule fits on one line: make reliable first, optimize next, then artificial intelligence in the face of real complexity. Monitoring software connected to dubious meters confidently displays a wrong dashboard. A hypervision platform changes nothing about that. Nor does predictive control, since it learns from the same data.
Can an EMS control an AHU?
No, unless it writes a setpoint to the BMS. The EMS spots an air handling unit running at night and raises the alert. To correct it, the system must automatically send a setpoint to the controller. Otherwise, energy management stops at the observation: an EMS that does not write to the BMS remains a dashboard. Energy management happens in the controller.
Who changes the setpoint after an alert?
The alert goes out in real time. And then? The operator waits for an order. The facility manager (FM) refers it to the integrator, who quotes for out-of-contract work. These diluted responsibilities come from the contracts, not from the people.
Does your operation and maintenance contract state who handles an alert and within what time? Before connecting an EMS to the BMS, write this implementation into the expected service level.
Do you need an EMS if you have a BMS?
Often, yes. A BMS keeps poor trend logs over several years and compares sites poorly. It also rarely corrects for the effect of the weather. Monitoring software, EMS or SGE, fills these three gaps on one condition: reliable meters. An energy management system built on patchy data draws clean, wrong curves.
Start from the scope: sites, end uses and decisions to support
To choose energy management software, start from the decision to be informed. It may be deciding on capital expenditure (CAPEX), staying on a tertiary decree trajectory or tracking down consumption outside occupied hours. Each decision calls for a different tool. A single site is often satisfied with a simple software solution. A multi-site real estate portfolio requires a platform able to help compare heterogeneous assets.
Data required, reversibility, commitment: the questions to ask before signing
Before signing, ask the vendor this list of four questions. It complements the specification of an EMS and targets the characteristics the demo avoids:
- How many months of history do you require at startup?
- Which points do you need to read and write?
- Which export format guarantees the reversibility of the contract?
- What is the minimum commitment period for the service?
Energy management software that refuses to export your data locks you in.
The 5 key steps to implementing an energy management system
Take stock of end uses and available BMS points
The first of the five key steps is to inventory three things: significant energy uses, meters and the BMS points actually brought up to supervision. An installation's drawing often shows points that never come up. Implementing an energy management system starts with this gap, not with choosing a tool.
Make the data chain reliable, from sensor to dashboard
This second step is the one most often skipped. An indicator is only as good as the weakest link in the data chain. The diagram below follows the information along its five links and gives the typical failure of each. Clicking a link highlights the question to ask your operator.
Five links from sensor to decision: the typical failure of each
An indicator inherits every failure located upstream. Select a link to find the question that checks it. Scroll right to follow the five linksRead from right to left, this diagram changes how you look at alerts. A dashboard in the red may reveal a duplicated meter rather than a drift. Overrides (manual overrides) are more insidious: the displayed setpoint is correct but the equipment does not follow it. No system, not even monitoring software, corrects a wrongly addressed point.
Establish the baseline, then prioritize the action plan
The third step builds the baseline on one year of reliable, HDD-corrected history. Without this correction, it mostly measures the weather. The analysis then ranks each action by its measurable gain and its cost. The action plan is then judged on verified energy performance, not on its number of lines.
Act on setpoints, optimal start and water loops
The fourth step moves to action. To optimize consumption, the levers are well known: setpoints, schedules, optimal start times and the temperatures of the hot and chilled water loops. The first gain often comes from overrides (manual overrides) set after a complaint and then forgotten.
Log every changed setting. Without a record, next month's energy saving cannot be attributed to anything.
Monitor drifts and adjust settings over time
The fifth step is the only one that never stops. Monitoring relies on a monthly review of consumption by end use, corrected for HDD. Real-time monitoring detects the drift. The review designates who corrects it.
This principle of continuous improvement stands in contrast to the one-off audit filed away in a drawer. Its implementation depends less on dashboards than on a fixed slot in the operator's calendar.
Benefits of an energy management system: savings to be proven
Bills and comfort: what well-tuned HVAC control changes
Well-tuned HVAC control acts on two items: the bill and the complaints. In our Valeo case study, HVAC control produced -28.6% HVAC consumption, measured according to IPMVP. The market readily promises double-digit energy savings. We take these promises seriously when they can be verified. Cost reduction then comes with a lower carbon footprint and lower greenhouse gas emissions, demonstrable on a low-carbon pathway.
HDD weather correction: a mild winter is not a saving
A raw difference in the bill mixes weather and settings. Heating degree days (HDD, called DJU in France) bring heating consumption back to a reference climate, because a mild winter is not a saving. Enter your two years below: this calculator helps you do the operation with your own data.
Bring your consumption back to the climate of the reference year
HDD calculated on an 18 °C baseIf the corrected difference is close to zero while the bill falls, the weather did the work. If it remains clearly negative, your settings have really reduced the energy consumed. It is this corrected figure that energy management presents to an investment committee, not the raw bill.
Return on investment: why the payback period lengthens on small sites
The return on investment depends first on the size of the site. The audit, integration and monitoring form a fixed cost, whereas the avoided expense follows consumption. On a small commercial site, the payback period lengthens markedly. The CEE operation sheet BAT-TH-116 (CEE energy-savings certificates), open to classes A or B, follows a logic distinct from compliance[7][7]RegulationFrench ministry in charge of energy, “Fiche d’opération standardisée CEE BAT-TH-116” (standardized CEE operation sheet BAT-TH-116), building management system, classes A or BSee in sources ↓. A company does not choose a solution to collect a subsidy.
What Foobot does on your existing BMS, without works
BMS advisory: getting meters, sequences and alarms back in order
Our first pillar, BMS advisory, stands on its own. It often starts with the “obsolete” BMS audit, which gives a second opinion before any replacement. On a new installation, handover and the warranty year (GPA, garantie de parfait achèvement, the French one-year defects liability period) then go through operational tests. Once the building is in service, Operations advisory clears overrides (manual overrides) and restores consistent sequences and useful alarms. That leaves making meters reliable, which tracks down data gaps and duplicates. This service also helps frame maintenance before any tool is implemented.
AI Control: regulating HVAC every 15 minutes through the BMS
When faced with real complexity, AI Control takes over. We train a digital twin of each building on its BMS history, weather and occupancy. Artificial intelligence then adjusts the hot and chilled water loops, supply air and peak shaving to optimize consumption.
Control goes through the existing BMS, over BACnet or LonWorks (LON). It passes through a multi-protocol controller or through the supervision, with no added hardware. After three steps (audit, learn, control), it runs every 15 minutes, 24/7. Gains are measured according to IPMVP and the commitment lasts 12 months.
AI Optimal Start: recalculating the start time every day, zone by zone
Our AI Optimal Start recalculates the HVAC restart time every day. The calculation includes thermal inertia, weather, occupancy and the warm-up time of each zone. Neither too early nor too late.
Conventional optimizers reason at the scale of the whole building. Zone-by-zone optimal start is a matter of verifiable smart management. The service is run by our teams or available as a self-service version for integrators.
Energy management: a dashboard does not cut a single kWh until the BMS acts
Remember the division of roles: the EMS detects, the BMS corrects, the hypervisor aggregates. An energy management system that stops at the observation reduces no consumption. Your next move is not buying software but checking the data chain.
To find out what your installation can really do, start with the building's BMS. Then check the difference between a BMS and a supervision-only system that your quotes often hide. Our free version of the EMS tracks your energy. It covers neither portfolio ESG reporting nor multi-country. To go beyond monitoring, ask us for a quote.
FAQ - energy management
How does a smart energy management system work?
A smart energy management system collects data from meters and sensors, corrects it for weather and spots deviations. It then calculates a setpoint and writes it to the BMS. Without that write, it only observes.
What is smart energy management?
Smart energy management anticipates instead of reacting. It takes weather, thermal inertia and occupancy into account to adjust settings. Artificial intelligence only has a place there when facing real complexity. For lighting, a presence detector is enough.
Energy management, energy efficiency, energy performance: what are the differences?
Management is an approach that sets who measures, who decides and who corrects. Efficiency is a ratio between the service delivered and the energy consumed. Performance is a result measured over time, tracked by an EnPI.
What is energy management?
It is energy management applied to a building. It consists of measuring consumption by end use, detecting drifts and acting on equipment settings. Monitoring without corrective action remains reporting, however precise it may be.
How do you choose energy management software?
Start from the decision to be informed, not from the features. Energy management software is judged on its access to BMS points, read-only or read-write. Reversibility also counts: getting your data back at the end of the contract.
What are the steps of energy management?
Five key steps follow in this order. Inventory end uses and points, make the data chain reliable, establish the baseline and the action plan. Then come action on settings and the monthly monitoring of drifts.
How do you optimize energy consumption?
To optimize consumption, start by clearing forgotten overrides (manual overrides). Then adjust setpoints, schedules and optimal start times. Every changed setting is logged and checked after HDD correction, otherwise the saving cannot be verified.
What are the benefits of an energy management system?
It delivers measurable cost reduction, fewer comfort complaints and figures you can defend in front of an investor. The return on investment varies with the size of the site. Only an IPMVP measurement can demonstrate it beyond dispute.
How do you implement an energy management system?
Implementation follows the logic of the ISO 50001 standard, which requires no software: policy, energy review, baseline, action plan. In the field, meter reliability and access to BMS points come before the tool.
Which tools for energy management?
Monitoring software measures, alerts and produces the dashboard. The BMS acts on the equipment. The hypervisor aggregates several sites. Predictive control is only added when there is real complexity, never before the data has been made reliable.
Does the hypervisor replace the EMS?
No. The hypervisor aggregates multi-site data and several BMS brands in a single view. It is neither an EMS nor a BMS: it does not normalize by HDD, does not detect drifts by end use and does not control any equipment.
Are a BMS and an EMS redundant?
No, the two complement each other: the EMS detects, the BMS corrects. A BMS keeps poor trend logs over several years and compares sites poorly. The EMS fills this trend-logging gap, provided it relies on reliable meters.
Sources and references
7 sources.- [1]Légifrance, “Décret n° 2025-1343 du 26 décembre 2025” (Decree No. 2025-1343 of 26 December 2025), postponing to 1 January 2030 the BACS deadline for existing commercial buildings of 70 to 290 kW (Article R. 175-2 of the French Construction and Housing Code)26/12/2025↑ 1 2
- [3]Légifrance, “Décret n° 2019-771 du 23 juillet 2019” (Decree No. 2019-771 of 23 July 2019), obligations to reduce final energy consumption in commercial buildings23/07/2019↑
- [4]Légifrance, “Décret n° 2020-887 du 20 juillet 2020” (Decree No. 2020-887 of 20 July 2020), building automation and control systems in non-residential buildings (BACS decree), amended by Decree No. 2023-259 of 7 April 2023 and codified in Articles R. 175-1 et seq. of the French Construction and Housing Code20/07/2020↑ 1 2
- [5]European Union, “Directive (EU) 2023/1791 on energy efficiency”, energy management system mandatory above a consumption threshold09/2023↑
- [7]French ministry in charge of energy, “Fiche d’opération standardisée CEE BAT-TH-116” (standardized CEE operation sheet BAT-TH-116), building management system, classes A or B↑