EN 54-5:2017+A1:2018 defines eight point heat detector classes, A1 through G, each pinned to a specific typical application temperature, maximum application temperature, and minimum/maximum static response temperature, with classes A1 and A2 both rated for 25°C typical and 50°C maximum application use [S1][S3].
Class B is the next step up the temperature ladder: 40°C typical, 65°C maximum application, with a 69°C minimum static response and 85°C maximum static response, making it the class of choice for spaces that already run warm under normal operation [S1][S3]. A free-suffix A1 or A2 detector may still carry an internal rate-of-rise function, but the suffix "R" or "S" must be marked on the product to make that behaviour contractually visible [S3].
How the three classes break down on temperature and response
The Hochiki application note reproducing the EN 54-5:2001/A1:2002 table gives the exact static-response windows that the 2017+A1:2018 revision carried forward: A1 covers 54–65°C, A2 widens the upper end to 54–70°C, and B jumps to 69–85°C [S3]. The same source shows class C at 84–100°C and class D at 99–115°C, with E, F and G extending up to 144–160°C for the highest-temperature industrial spaces [S3].
On the rate-of-rise test rig, class A1 detectors must respond within tighter windows than A2 at low rates of rise. At 1 K/min from typical application temperature, A1 has an upper limit of 40 min 20 s, while A2 allows 46 min 0 s; at 3 K/min the A1 upper limit is 13 min 40 s against 16 min 0 s for A2; and at 5 K/min A1 must trip by 8 min 20 s while A2 can take up to 10 min 0 s [S3][S5]. Class B shares the wider A2-style windows, not the tighter A1 windows, so B is a slower-to-alarm class by design [S3].
Suffix logic: what R and S actually add to the class letter
Appending "S" to a class letter (A1S, A2S, BS) means the detector is fixed-temperature only, set to trip at or above the class's minimum static response temperature, with no rate-of-rise element contributing to the decision [S3]. The S variant is the right pick for boiler rooms, commercial kitchens, and any space where fast ambient swings (door opening, process start-up) would otherwise generate false alarms on a rate-of-rise unit [S3].
Appending "R" (A1R, A2R, BR) means the detector also reacts to a defined rate of rise in air temperature, triggering before the fixed threshold is reached when the rise rate is high enough [S3]. The standard's test table shows A1R has a lower response time limit of 29 min 0 s at 1 K/min from typical application temperature, tightening to 1 min 0 s at 10 K/min and 30 s at 20 K/min, while at 30 K/min the upper limit collapses to just 1 min 40 s [S3][S5]. A1R is therefore the fastest-responding combination in the A-tier, A2R sits slightly behind, and BR trades speed for higher ambient tolerance [S3].
Detection envelope: where each class is the right pick
Class A1 (with or without R/S suffix) targets the cleanest indoor environments: offices, corridors, bedrooms, hospital rooms, and any space where the ambient stays close to 25°C and the maximum never crosses 50°C under non-fire conditions [S1][S3]. The 25°C typical / 50°C ceiling is the EN 54-5 baseline; anything above that envelope is a non-A1 application [S1].
Class A2 covers the same 25–50°C envelope on typical and maximum application temperature but allows a slightly wider 54–70°C static-response band, which lets the detector tolerate small headroom above the 50°C ceiling without nuisance tripping [S1][S3][S5]. For practical purposes, A1 and A2 are often dual-certified on the same product, as the ABUS RM-1100-2 certificate illustrates: a single detector body carries both class A1 and class A2 ratings on the same datasheet [S5].
Class B is the right step when the space itself runs hot: boiler rooms, industrial laundries, certain manufacturing cells, and ceilings above production equipment where ambient routinely sits between 40°C and 65°C [S1][S3]. Specifying A1 or A2 in a 40°C-typical space would generate constant false alarms because the static response threshold (54°C) is too close to the normal operating temperature; B lifts that floor to 69°C [S3].
Comparison: A1 vs A2 vs B on the four decision criteria
On typical application temperature, A1 and A2 are tied at 25°C with a 50°C maximum; B moves both up to 40°C typical and 65°C maximum, the class boundary that defines "warm space" under EN 54-5 [S1][S3]. On static response, A1 must trip between 54°C and 65°C, A2 between 54°C and 70°C, and B between 69°C and 85°C, so the headroom above maximum application temperature is roughly 4°C, 5°C, and 4°C respectively [S3].
On rate-of-rise response speed, A1 (and A1R) is the fastest, A2 (and A2R) is close behind with slightly looser upper limits, and B trails because the standard grants B the wider A2-style rate-of-rise windows [S3][S5]. On false-alarm tolerance, A2 and any "S"-suffix variant are more tolerant than A1, while B is the most tolerant of the three because its higher static floor simply does not see the 54–65°C regime at all [S3]. The JBE-2106 detector from Jade Bird Fire exemplifies the design freedom the standard allows: a single product ships with three sensitivity profiles (A2R, A2, A2S) selectable in the field [S6].
Standards and certification trail
The current EN 54-5 edition in force across CEN member states is EN 54-5:2017+A1:2018, published August 2018, which superseded the earlier EN 54-5:2017 and the much older EN 54-5:2000/2001+ A1:2002 referenced in the Hochiki application note [S2][S3]. Compliance with the 2017+A1:2018 edition is what shows up on a CPR Certificate of Constancy of Performance, such as the DBI Certification 2531-CPR-232.1485 issued for the ABUS RM-1100-2 family on 2021-11-09 [S5].
The standard splits the requirements into four families: operational reliability (clause 4.2), nominal activation conditions including static response temperature and rate-of-rise response (clause 4.3), response delay and sensitivity tolerance, and a separate test suite for the suffixes R and S [S2]. Detection is only one part of the picture; for broader fire-alarm system design and detector spacing, BS 5839-1 caps a single heat detector at roughly 30 m² of coverage while NFPA 72 permits up to 37 m², a difference that matters when laying out a B-class space in a UK versus US project [S4].
Product and application footprint in 2026
The 630/670 intrinsically safe range from FFE / Safety Technology International carries EN 54-5 classes A1R, A1S, A2S, BR, BS, CR and CS, a 2025-05-14 datasheet that shows the full A-to-C spectrum still in active production for hazardous-area use [S8]. Each variant is tested and shipped with a single class designation; selecting the right one is a design decision, not a feature toggle [S8].
For a working specifier the decision tree is short: pick A1 (or A1R) for the cleanest indoor spaces with 25°C typical ambient; pick A2 (or A2S) for the same envelope but with slightly higher headroom or where ambient can spike briefly; pick B (or BR) when the space itself runs at 40°C or higher under normal operation, which covers most boiler rooms, commercial kitchens, and a long list of industrial cells that the standard's class table exists to serve [S1][S3][S4]. For detection technology beyond heat, the same engineering logic of class-based thresholds shows up in toxic-gas detector performance standards, which shifted to a class-based system under EN IEC 62990-1:2022.
A more practical note for procurement: the next data point to track is the publication of any CEN amendment to EN 54-5:2017+A1:2018 (the 2018 amendment is the most recent in the public record at the time of writing), and any move by notified bodies to retire A1/A2 dual-class declarations on addressable detector datasheets in favour of single-class certifications, which would force specifiers to commit to a class on the order rather than leaving it to commissioning [S2][S5].
For component-level specifications, see heat detector, rtd pt100, and heat treatment furnace.