ISO 12944:2018 Part 2 sets six atmospheric corrosivity categories C1, C2, C3, C4, C5, and CX, with carbon-steel thickness loss after the first year of exposure quantified as ≤1.3 µm for C1, 1.3–25 µm for C2, 25–50 µm for C3, 50–80 µm for C4, 80–200 µm for C5, and 200–700 µm for CX [S4].
Each atmospheric category is paired with one of four durability bands introduced in the 2018 revision: Low (L) up to 7 years, Medium (M) 7–15 years, High (H) 15–25 years, and Very High (VH) more than 25 years, and the combination drives total dry film thickness, primer chemistry, and the laboratory test set that qualifies the system [S3][S4].
How C1 through CX map to real environments
C1 covers heated interiors with clean atmospheres, where the standard effectively allows uncoated steel for a typical 50-year building design life; examples include offices, schools, and shops where dry internal air suppresses corrosion [S4].
C2 describes rural atmospheres and unheated depots or sports halls where condensation cycles begin to act on the substrate, with first-year carbon-steel loss between 1.3 and 25 µm [S4].
C3 covers urban and industrial atmospheres with moderate sulphur dioxide loading plus coastal low-salinity exposure, and interior food-processing, laundry, brewery, and dairy spaces; first-year loss is 25–50 µm [S3][S4].
C4 adds industrial and moderate-salinity coastal zones, interior chemical plants, swimming pools, and boatyards, and raises first-year loss to 50–80 µm on unprotected carbon steel [S4].
C5 covers industrial areas with high humidity and aggressive atmospheres and coastal zones with high salinity, with first-year loss 80–200 µm, while CX covers offshore areas, extreme humidity industrial zones, and subtropical or tropical atmospheres with first-year loss 200–700 µm; CX is governed by the dedicated Part 9 framework [S1][S4].
What changed in the 2018 revision versus 1998
The 1998 C5-I (industrial very high) and C5-M (marine very high) sub-categories have been retired and replaced by a single C5 for the harshest onshore environments plus CX for offshore; the corrosion-protection scope now also adds IM4, sea or brackish water with cathodic protection, alongside the original Im1, Im2, and Im3 categories [S1][S2].
Durability bands were re-cut: the old 2–5 year Low and 5–15 year Medium were extended, Low now reaches 7 years, Medium sits at 7–15 years, High at 15–25 years, and a new Very High band covers anything over 25 years, which the 2018 guidance treats as the practical ceiling for most land-based specs [S1][S3].
For C2 and C3, the laboratory regime stays the old ISO 6270-1 water-condensation and ISO 9227 neutral salt-spray hours; from C4 High upward, ISO 12944-6:2018 introduces phased cyclic ageing under Annex B to better replicate in-service wet/dry cycling, and CX/Part 9 only accepts cyclic-tested systems [S1][S2].
Coating-system selection by category and durability
A C2 Low spec can be qualified with as little as 48 h ISO 6270-1 condensation and no salt-spray requirement, while C3 Low needs 48 h condensation plus 120 h neutral salt spray; a C5 Very High system must survive 2688 h of Annex B cyclic ageing, the most aggressive atmospheric test set in the standard [S2][S3].
For C4 High, the 2018 tests demand 480 h ISO 6270-1, 720 h ISO 9227, and no Annex B cyclic test, whereas C4 Very High steps up to 720 h condensation, 1440 h salt spray, and 1680 h cyclic ageing; CX has its own row in Part 9 with 4200 h cyclic ageing at the extreme band [S2].
Surface preparation stays anchored to ISO 8501-1 grades, with Sa 2½ (near-white metal, comparable to SSPC-SP10) as the workhorse for most atmospheric systems and Sa 3 (white metal, SSPC-SP5) reserved for the most severe CX and immersion duty where residual contamination cannot be tolerated [S2].
What durability actually means in a specification
Durability in ISO 12944 is the expected interval to first major maintenance, not a warranty of the coating or a guarantee against substrate loss; a C4 High system targeted at 15–25 years still needs the right primer chemistry, compatible topcoat, and the documented surface profile to actually hit that window [S3][S5].
A practical implication: specifiers who treat Very High as a 25-year no-maintenance promise often find the real first-maintenance interval shortened by under-specifying surface preparation, applying outside the dew-point window, or letting DFT drift below the minimum nominal thickness for the chosen category [S5].
Conversely, stepping a project from C3 Medium up to C4 High purely to be cautious can more than double the total DFT requirement and add a third coat, which is why a careful atmospheric survey against the C1 to CX descriptions is usually cheaper than over-specifying the industrial coating system [S5].
Comparison of corrosivity categories against decision criteria
On the four criteria that drive coating selection, the categories line up as follows: first-year carbon-steel thickness loss on unprotected substrate (C1 ≤1.3 µm, C2 1.3–25 µm, C3 25–50 µm, C4 50–80 µm, C5 80–200 µm, CX 200–700 µm), typical DFT envelope (C1/C2 often below 120 µm, C3 around 120–200 µm, C4 200–320 µm, C5 320–500 µm plus, CX/Part 9 typically above 500 µm), qualifying laboratory test (C2/C3 use ISO 6270-1 + ISO 9227, C4 High+ and C5 add ISO 12944-6 Annex B cyclic ageing, CX is cyclic-only), and standard part used (C1–C5 systems sit in Part 5 tables, CX systems sit in Part 9 tables) [S1][S2][S3][S4].
For designers, the practical takeaway is that a C4 Very High system, 1680 h cyclic ageing plus 720 h condensation plus 1440 h salt spray, is roughly the most aggressive onshore system that a land-based fabricator will encounter, and any spec that crosses that line is signalling CX duty and a waterproof coating selection process that has to follow Part 9 rather than Part 5 [S2][S5].
Where ISO 12944 stops and other standards take over
ISO 12944 covers atmospheric corrosivity and four immersion categories (Im1 fresh water, Im2 sea or brackish water without cathodic protection, Im3 soil, Im4 sea or brackish water with cathodic protection) but does not quantify chemical resistance on its own; chemical plant service usually pulls in supplementary immersion testing against reagents such as 10% NaOH, 10% H2SO4, and mineral spirits at 18% aromatics, which Cloverdale's guide lists as the three baseline immersion chemicals for ISO 12944-6 testing [S2].
On coating thickness verification, ISO 12944 sets the nominal DFT and minimum number of coats per system but relies on ISO 19840 for the DFT acceptance rules and on a calibrated coating thickness gauge for the actual measurement, a separation that often confuses first-time spec writers [S5].
For substrate and steel-prep questions, ISO 12944 Part 4 cross-references ISO 8501-1 for visual surface preparation grades, and structural design choices that affect corrosion (crevices, water traps, lap joints) sit in Part 3, so durability is treated as a system property rather than a paint-only attribute [S3].
Trackable signals and the next specification node
Forthcoming revisions to ISO 12944 are not confirmed in any of the public material surveyed, so the safe engineering node to watch is the maintenance of published test durations: any change to the C5 Very High 2688 h Annex B figure or to the CX 4200 h cyclic-ageing figure in Part 9 would immediately invalidate a large number of existing specifications that have been qualified against those exact numbers [S2][S3].
Two further signals are worth tracking: the extension of the Annex B cyclic regime downwards into C3 and C4, which would push more lower-tier systems into wet/dry cycling rather than steady-state salt spray, and the gradual retirement of legacy SSPC-SP6 references in favour of ISO 8501-1 Sa 2 across European structural steelwork, particularly relevant to projects where the same fabrication shop supplies both ISO-spec and regional-spec jobs, including those routed through construction machinery and equipment supply chains that increasingly demand EN 1090-2 execution class evidence [S1][S2].
Background reading: Medium Die Casting Machine Footprint and Crane Hook Height for Die Changes.