Repainting a failed coating without establishing why it failed buys a second failure on the same schedule. The pattern on the steel is evidence, and the standards that read it were written precisely so the argument can be settled with measurements rather than opinions.
Grade the Failure Before Explaining It
The first discipline in a coating investigation is to describe what is there without reaching for a cause. That sounds obvious and is routinely skipped, because the person looking usually arrives with a theory and a repaint quotation already in draft.
ISO 4628 exists to impose that discipline. The series designates the quantity and size of defects and the intensity of uniform changes in appearance, with Part 1 providing the general introduction and designation system and each subsequent part handling one defect type on its own scale. Part 2 assesses the degree of blistering, Part 3 the degree of rusting, Part 4 cracking, Part 5 flaking, Part 8 delamination and corrosion around a scribe or other artificial defect, and Part 10 the degree of filiform corrosion.
Worth noting that the series is live. Part 3 reached a new edition with the 2016 version withdrawn in December 2024, Part 5 was revised in 2022, and Part 10 was reissued in 2024 replacing the 2016 edition, with the newer part carrying a slightly changed series title. A report citing a decade-old edition is not wrong about the mechanism, but it may be grading against a superseded scale.
A failure described as “the coating has broken down” cannot be argued about usefully. A failure graded by part and scale can be compared against a previous survey, against a sister vessel, and against a contractual requirement.
Part 10 shows how specific these methods are. Filiform corrosion is rated from two measurements, the length of the longest filament and the most frequent filament length, and its annex of twelve reference pictures explicitly does not substitute for the numerical assessment. The reference images support the measurement; they do not replace it.
What Each Pattern Tends to Mean
Once the condition is graded, the pattern points at a mechanism, and the mechanism points at who or what is responsible.
Blistering. Liquid or gas accumulating between layers or at the substrate. Where the cause is osmotic, water-soluble contamination left under the film draws moisture through the coating, which makes blistering a surface preparation question before it is a product question. Solvent entrapment and overcoating outside the stated window produce their own versions.
Disbondment. Adhesion lost across an area rather than at a point. Where cathodic protection is present, the alkaline environment generated at the steel surface is a recognised mechanism, which is why coating and cathodic protection have to be designed as one system rather than two.
Undercutting. Corrosion spreading laterally beneath intact-looking film from a holiday, a scribe or a mechanical defect. This is exactly what ISO 4628-8 was written to measure, and its presence says the film has a discontinuity somewhere upstream of the visible damage.
Pinpoint rusting. Discrete spots across an otherwise sound area. The pattern suggests many small discontinuities rather than one large failure, which points at film thickness over the profile, pinholes, or holidays that were never detected.
The distinguishing question: is the failure distributed with the geometry, with the work sequence, or with the environment? Geometry implicates design and access. Sequence implicates application. Environment implicates the specification.
That last test does most of the work in practice. Failure concentrated at edges, welds and corners usually indicates film thickness and edge treatment rather than product performance. Failure that follows a day’s work boundary or a particular bay points at application conditions on a particular shift. Failure distributed uniformly across a large area, in a system that was applied and recorded properly, points at a system specified for a less demanding environment than the one it is in.
The Measurements That Settle It
A diagnosis stands or falls on the evidence gathered before anything is cleaned off, and the useful set is small.
The soluble salt measurement needs a caution that is written into the standard itself. ISO 8502-6 covers extraction of water-soluble contaminants using the Bresle method, with flexible adhesive patches or sleeves that attach to a surface of any shape or orientation, and it is explicit that extraction may miss soluble material and return a false negative. Named causes include material held in crevices or under metal folds, and material lying beneath corrosion layers, inhibitor passivation layers, oil, grease or other thin films.
ISO 8502-9 then provides the field method for conductometric determination of water-soluble salts, estimating the surface density of salts before or after surface preparation. It too has a stated limit: it cannot determine the individual amounts of specific salts such as chlorides, sulphates or sodium, and it assesses only contaminants that form ions in contact with water.
A clean salt reading does not prove the surface is clean. The standard lists the circumstances in which the method misses what is there, and a failure investigation is full of exactly those circumstances.
Together those two limits explain a recurring dispute. A contractor produces salt test results showing an acceptable surface, a failure follows that looks osmotic, and both sides are working from the same paperwork. The results may have been correctly obtained and still have missed contamination under a corrosion layer or in a fold, which is why the readings belong alongside the records rather than standing as a verdict on their own.
Why the Repaint Decision Comes Last
The repair scope follows from the diagnosis, and the four mechanisms demand different scopes.
Contamination-driven blistering is not repaired by a thicker coat over the same surface; the contamination has to be removed, which is a preparation scope rather than a painting one. Disbondment where cathodic protection is present is a system design question, and recoating without addressing the interaction reproduces the failure. Undercutting means a discontinuity exists beyond the visible damage, so the removal boundary has to be established by investigation rather than by where the rust stops. Pinpoint rusting over a wide area suggests a systemic thickness or detection shortfall, which spot repair will not cure.
What all four have in common is that the cheapest version of the repair is the one chosen when nobody established the cause. Grading the condition, measuring the few things that discriminate between mechanisms, and reading the original application records costs a fraction of a repaint and routinely changes what the repaint should be.
It also changes who pays. A failure traced to application conditions, overcoating intervals or film thickness sits with the execution of the work. A failure traced to a system specified for a lower corrosivity category sits with the specification. Neither conversation can be had without the measurements, which is the practical argument for taking them before the surface is disturbed. Our industrial painting and coating work includes that assessment and the preparation scope that follows from it, because the repair that lasts is the one aimed at the mechanism rather than at the appearance.
Frequently Asked Questions
How should a coating failure be recorded?
Graded against the ISO 4628 series, which designates the quantity and size of defects and the intensity of uniform changes in appearance. Part 1 gives the general introduction and designation system, with blistering in Part 2, rusting in Part 3, cracking in Part 4, flaking in Part 5, delamination and corrosion around a scribe or artificial defect in Part 8 and filiform corrosion in Part 10. Check the edition, since several parts have been revised recently.
What does blistering usually indicate?
Accumulation of liquid or gas at an interface. Where the mechanism is osmotic, water-soluble contamination remaining under the film draws moisture through the coating, which makes it a surface preparation issue rather than a product issue. Solvent entrapment and overcoating outside the stated interval produce their own forms, so the distribution of the blistering and the original application records are needed to separate them.
Can a salt test prove the surface was clean?
Not on its own. ISO 8502-6 states that extraction by the Bresle method may miss soluble material and give a false negative, naming material in crevices or under metal folds and material beneath corrosion layers, inhibitor passivation layers, oil, grease or other thin films. ISO 8502-9, the field conductometric method, estimates total surface density of water-soluble salts but cannot determine amounts of individual salts such as chlorides, sulphates or sodium.
How do you tell an application fault from a specification fault?
By how the failure is distributed. Concentration at edges, welds and corners points at film thickness and edge treatment. Alignment with a work boundary or a particular area or shift points at application conditions. Uniform distribution across a large area in a properly applied and recorded system points at a system specified for a less demanding environment than the one it is serving.
Sources: ISO 4628 series, Paints and varnishes, Evaluation of degradation of coatings, Designation of quantity and size of defects and of intensity of uniform changes in appearance, comprising Part 1 general introduction and designation system, Part 2 assessment of degree of blistering, Part 3 assessment of degree of rusting with the 2016 edition withdrawn in December 2024, Part 4 assessment of degree of cracking, Part 5 assessment of degree of flaking revised 2022, Part 8 assessment of degree of delamination and corrosion around a scribe or other artificial defect, 2012, which describes rating delamination and corrosion spreading from a scribe or artificial defect on a coated specimen exposed to a corrosive environment and excludes pitting corrosion and pit depth, and Part 10 assessment of degree of filiform corrosion, 2024 edition replacing 2016, rating from the length of the longest filament and the most frequent filament length with twelve reference pictures at Annex A that do not substitute for the numerical assessment · ISO 8502-6:2020, Preparation of steel substrates before application of paints and related products, Tests for the assessment of surface cleanliness, Part 6, Extraction of water soluble contaminants for analysis, Bresle method, describing extraction using flexible adhesive patches or sleeves suitable for field use on surfaces of any shape or orientation, and stating that extraction may miss soluble material or give false negatives where material is held in crevices or under metal folds or lies beneath corrosion layers, corrosion inhibitor passivation layers, oil, grease or other thin films · ISO 8502-9:2020, Part 9, Field method for the conductometric determination of water-soluble salts, estimating the surface density of water-soluble salts on steel before or after surface preparation, stating that it cannot determine individual amounts of specific salts such as chlorides, sulphates or sodium and assesses only contaminants forming ions in contact with water · ISO 4624, pull-off test for adhesion
MarineCraft Journal
Get the next briefing in your inbox
Safety protocols, offshore training, regulation and market analysis, written for Malaysia’s maritime and offshore industry. No sales pitches.
We use your address only to send MarineCraft Journal. Unsubscribe any time. See our Privacy Policy.