How to Investigate a Coating Failure
A structured, evidence-first method for finding the root cause of a coating failure — from preserving the scene and reviewing records to field tests, lab analysis and a defensible report.
Key takeaways
- A coating failure investigation aims to identify the root cause so that the repair works and responsibility can be fairly assigned.
- Start by preserving evidence and collecting records; many failures are explained by what the specification, data sheets and daily reports show.
- Combine a systematic site survey and field tests with laboratory analysis of representative samples and controls from sound areas.
- Test hypotheses against all the evidence and report facts, methods and limitations clearly — conclusions must be supported by data.
When a coating fails, the immediate questions are what went wrong, who is responsible and how to fix it. Answering them reliably requires a methodical investigation, not a quick look and an educated guess. The same visible symptom — peeling, blistering, rusting — can have very different causes, and a repair designed for the wrong cause will usually fail again.
This guide outlines a practical framework used by coating consultants, inspectors and owners. Individual failure modes such as blistering and delamination are described in their own articles; here the focus is on the process.
Define the scope and preserve evidence
Begin by agreeing the questions to be answered and who the investigation is for — owner, contractor, manufacturer or a neutral party. In disputes, it is common for all parties to be invited to joint site inspections and sampling, and for agreed protocols to be set before destructive testing.
- Stop further disturbance of failed areas where possible until they are documented.
- Retain samples of coatings from the original batches, if available.
- Photograph everything with scale references, location identifiers and date stamps.
- Establish chain of custody for samples that may be used in litigation or warranty claims.
Review the records
Documents often narrow the possible causes before anyone goes to site. Gather:
- The specification and any approved deviations.
- Product data sheets and safety data sheets for each coat, with batch numbers.
- Daily inspection reports: surface preparation, ambient conditions, dew point, DFT and holiday test results.
- Service history: operating temperatures, chemicals, cleaning methods, cathodic protection data and any incidents.
- Warranty terms and previous inspection reports.
Compare what was specified with what was done and with what the service actually demands. Gaps between these three are fertile ground for root causes.
Conduct the site survey
- Survey broadly first. Walk the whole structure to understand extent and pattern before focusing on details.
- Map the failure. Record location, orientation, elevation, exposure and proximity to edges, welds, water paths or heat.
- Rate the condition. Use standard scales — ASTM D610 or ISO 4628-3 for rusting, ASTM D714 or ISO 4628-2 for blistering — so observations are repeatable.
- Select test locations in failed, transitional and sound areas, so comparisons can be made.
Always test sound areas as controls. A low adhesion value means little until you know whether it is unique to the failed zone or typical of the whole job.
Field testing and sampling
Field tests are chosen to discriminate between competing hypotheses:
- Dry film thickness by magnetic or eddy current gauges per SSPC-PA 2, and destructive measurement of individual coat thicknesses per ASTM D4138.
- Adhesion by knife (ASTM D6677), tape (ASTM D3359) or pull-off (ASTM D4541 on steel, ASTM D7234 on concrete), noting where the failure plane lies. See adhesion testing.
- Soluble salts on exposed substrate and the backs of failed coating.
- pH of blister liquid and substrate surfaces.
- Surface profile by replica tape or depth gauge per ASTM D4417 on exposed steel.
- Solvent rubs to assess cure, for example ASTM D5402.
- Moisture testing on concrete substrates.
Collect samples that include all coats and, where possible, some substrate. Package them to prevent damage, label them clearly and record exactly where each came from.
Laboratory analysis
Laboratories confirm what the field suggests and reveal what cannot be seen on site:
- Optical microscopy of cross-sections shows coat count, thickness, voids, contamination layers and the precise failure plane.
- FTIR spectroscopy identifies binder types, contaminants such as silicones or amine carbamates, and degradation products.
- Differential scanning calorimetry (DSC) measures glass transition temperature and residual cure.
- SEM with EDS provides high-magnification imaging and elemental analysis of interfaces, salts and corrosion products.
- Ion chromatography quantifies chlorides, sulfates and nitrates extracted from samples.
Determine the root cause and report
Evaluate each plausible hypothesis against all the evidence: the pattern, the timing, the field results and the lab data. The correct explanation should account for where the failure occurs and where it does not. Many failures have a primary cause and one or more contributing factors, such as thin film plus salts plus condensation.
| Cause | Clues | How to check | Prevention |
|---|---|---|---|
| Surface contamination (salts) | Blisters, rust at interface, random pattern | Salt tests; IC on samples | Specify and verify salt limits |
| Inadequate profile or cleanliness | Adhesive failure at substrate, smooth steel | Profile replica; visual standards | Inspect prep before priming |
| Wrong film thickness | Early rust (thin) or cracking (thick) | DFT survey; cross-sections | DFT checks per SSPC-PA 2 |
| Intercoat contamination or timing | Failure between coats | Failure plane; FTIR of interface | Respect recoat windows; clean between coats |
| Improper cure | Soft film, low solvent resistance | Solvent rubs; DSC; hardness | Correct mixing, climate control |
| Service mismatch | Failure matches heat, chemicals or abrasion | Service records vs. data sheet limits | Select system for actual service |
The report should state the scope, the information reviewed, methods and standards used, observations, test results, analysis, conclusions and recommendations for repair. Distinguish facts from opinion, acknowledge uncertainties and avoid conclusions the data cannot support.
Do not let the first plausible explanation end the investigation. Confirmation bias — seeing only evidence that fits an early theory — is one of the most common reasons failure reports are later overturned.
Frequently asked questions
Who should pay for a failure investigation?
That depends on contracts and warranties. Often the owner commissions it initially, with costs reallocated once responsibility is established. Agree the arrangement in writing before work starts.
How many samples are enough?
Enough to represent failed, transitional and sound areas across each distinct exposure or surface. Fewer, well-chosen samples with controls are more useful than many random ones.
Can a failure have more than one cause?
Yes, and many do. A good investigation identifies the primary cause and the contributing factors, because the repair may need to address all of them.
Educational reference. Coating performance varies by formulation. Always follow the manufacturer’s product data sheet, safety data sheet and your project specification.
