Four efficiency classes, A to D. Most existing buildings sit in C — not by choice, but because nobody ever asked the question.
SIA 386.111, 2022 edition, is the Swiss adoption of the international standard EN ISO 52120-1. It replaced the edition based on EN 15232-1. Its second part, SIA 386.112 (ISO/TR 52120-2), carries the explanations and justifications of the method.
A class nobody asks for
A building has a building-automation efficiency class much as it has an energy label. The difference is that it is almost never calculated, never written into a contract, and never verified at handover. It exists all the same: it follows from the functions actually present and active in the installation.
What the standard actually classifies
Here is the point that is usually missed: the standard does not classify equipment, it classifies functions.
It goes through the automation functions of a building family by family — control of heat and cooling emission, distribution control, generation control, scheduling and occupancy-based control, ventilation and air conditioning, lighting, blinds — and then the technical building management functions: fault detection, consumption monitoring, reporting. For each function it describes four levels of increasing sophistication. The building's class is the synthesis of those levels.
The direct consequence: there is no “class B BMS” in any catalogue. A controller carries no class. The class belongs to the configured installation, in its actual use.
The four classes
| Class | What the class implies | Typical case |
|---|---|---|
| A | Individual room control, presence and window detection, demand management, optimisation and automatic fault detection | High-performance new build, Minergie target |
| B | Communicating room control, zone schedules, consumption and alarm monitoring | A realistic target in refurbishment |
| C | Zone control, fixed setpoints, central schedules. The BMS displays but does not optimise | Buildings from 2000–2015 |
| D | No communication, manual settings, no monitoring. Not compliant with the state of the art | Unrefurbished installation |
Class C is the method's reference: the standard state, against which the other classes are compared. It is not a design flaw. It becomes a problem only when nobody asked whether it was enough.
Why most buildings sit in class C
It is almost never a decision. It is an accumulation of non-decisions.
- The target class appears in no project document. It is therefore never priced, never compared, never verified.
- Class B functions sit overwhelmingly at floor level — communicating terminal control, zone schedules, metering by use. That is the first item cut when the budget tightens, because it is the least visible.
- Technical management functions are delivered as licence options. Consumption monitoring, fault detection: they are neither enabled nor configured at commissioning, and nobody comes back to them.
- Handover tests operation, not performance. The valve is checked to open; the optimisation function is not checked to exist.
The gap between the designed class and the operated class
This is the most important point in this article, and the one to keep if only one is kept.
A class cannot be bought with hardware. An installation equipped for class B but configured with fixed setpoints, its schedules neutralised by permanent manual commands and its monitoring never enabled, is operated in class C. What counts is the functions that are active, not the ones on the bill of quantities.
A building can therefore be designed as class B, paid for as class B, and operated as class C for twenty years without a single document saying so — because no document ever measured it.
The orders of magnitude
The standard provides reference values that quantify the gap between classes. For office buildings, the gains stated for a change of class are as follows.
| Change | Heating | Cooling | Electricity |
|---|---|---|---|
| D → C | 10% | 8% | 6% |
| C → B | 17% | 13% | 10% |
| C → A | 28% | 24% | 17% |
| B → A | 13% | 13% | 8% |
Two words of caution. These are conventional calculation factors, not measurements: they serve to compare options at design stage, not to promise a bill. And the EN ISO 52120-1 method now modulates them by climate, where the older EN 15232 used fixed factors — a gain calculated for the Swiss plateau and one calculated for a mountain resort are not the same number.
Moving from C to B: what it actually takes
The reflex is to look at the plant room. That is rarely where it is decided.
Moving from C to B almost never requires touching heat generation. What it requires is communication: terminal controllers that talk to the BMS rather than isolated thermostats, zone schedules rather than a single central one, reading and logging the meters already installed, and alarm monitoring that is genuinely used.
The investment therefore falls on floor-level automation and on the network, not on the generator. Which makes it one of the few energy efficiency measures that can be carried out without a shutdown, in stages, and without touching the envelope.
The clause to put in the specification
A class that is not written down cannot be required. Three lines are enough to make it contractual.
- State the target class under SIA 386.111 (SN EN ISO 52120-1), function by function rather than globally — that is the form in which the standard is usable.
- Require the list of corresponding functions, each with the object or data point that delivers it. A function with no data point does not exist.
- Carry the verification of those functions into handover, alongside the functional tests, and record the result. That line is what prevents the gap between the designed class and the operated class.
A client who requires those three points gets something rare: a performance requirement verifiable with no measuring equipment and no campaign, because it is checked function by function, screen in hand.
What it changes
The point of the classification is not the letter. Few clients will rejoice at a B rather than a C.
The point is that it forces you to write down what you expect from a BMS before buying it — function by function, in the language of operation rather than of the catalogue. A building whose target class is written into the specification is a building whose functions have been discussed. In most projects, that discussion simply never happens.
wall-i is an independent engineering office for BMS and building automation, based in Valais, Switzerland. Our specifications are open to every brand and our deliverables follow the SIA standards and the KBOB naming convention. A subject you would like us to cover: hello@wall-i.ch