Girth-weld coating after overnight blasting: salt, rust, and a closing weather window

Reject. Four independent nonconformances exist, any one of which is disqualifying:

• Dew point margin. Steel at 6 C with dew point 5 C gives a 1 C margin. ISO 8502-4 and virtually every pipeline coating specification require at least 3 C. Condensation on a 6 C surface at 85% RH is imminent, and moisture on the substrate ruins epoxy adhesion.
• Re-oxidation and bloom. Overnight rust staining means the surface is no longer Sa 2 1/2 to ISO 8501-1; it is at best a B/C condition and must be re-blasted. A pale bloom on a blasted surface is the classic visual signature of soluble salt contamination, which is not removed by dry brushing.
• Chloride status unverified. No valid measurement exists. ISO 8502-6/8502-9 Bresle extraction is required; recycled abrasive can itself carry chloride and must be checked to ASTM D4940 (conductivity) and to SSPC-AB 1 cleanliness.
• Profile out of specification. 45 micrometres is below the 50 micrometre minimum. Under-profile gives insufficient mechanical keying; the peaks and valleys must be re-established, not topped up.

'Coat now, document later' converts a correctable surface condition into a permanent adhesion failure buried at every girth weld on the spread.

Corrective sequence

1. Stop and control the environment first. Erect a windbreak/enclosure, dehumidify, and heat the steel until surface temperature is at least 3 C (preferably 5 C) above dew point measured with a calibrated dew-point meter - not a pocket hygrometer alone. Record the reading at the joint, not at the site office.
2. Replace the abrasive. Reject the recycled grit until a conductivity/chloride check passes; use clean, dry, specification-compliant abrasive.
3. Re-blast to Sa 2 1/2, removing rust and bloom completely; pay attention to the weld cap, the heat-affected zone, and the cutback edge where salts collect.
4. Bresle-test for chloride to ISO 8502-6/8502-9 using an in-date kit, at a minimum on each suspect joint and on a statistical sample across the spread. Do not proceed above 3 micrograms per square centimetre.
5. Verify profile with replica tape or an ISO 8503 comparator: 50-90 micrometres, with a documented reading.
6. Dust check to ISO 8502-3 (rating 2 or better).
7. Apply within the recoat window, with a documented dew point check immediately before the first coat, and record steel temperature, RH, dew point, batch, and thickness for every joint.

Every step is recorded on the joint report before the coating is mixed.

Handle it as a technical decision with a documented trail, not a schedule decision.

• Establish the basis. Ask for the product data sheet and the manufacturer's written position on minimum surface temperature, maximum RH, dew point margin, and salt tolerance. Most high-performance epoxies are explicitly not surface-tolerant; if the manufacturer confirms a wider envelope in writing, that becomes the technical basis for evaluation.
• Do not concede on salt or rust. No epoxy qualification allows coating over an unmeasured chloride level or over visible bloom. These are measurable, non-negotiable conditions with a known failure mechanism: osmotic blistering and disbondment that will not appear for months and cannot be detected by a holiday test.
• Consider a partial concession only where evidence supports it. If chloride is measured below the limit after re-blasting and the only shortfall is a 2 C dew-point margin, a manufacturer-approved heated enclosure may close the gap - the specification is then met, not waived.
• If schedule pressure is real, the correct routes are: extend the window with heating and dehumidification, mobilise a second crew, or use a formally qualified alternative product with data at the actual conditions - all engineering-approved and documented before work resumes.

A concession signed to protect a schedule, without measurement data, is the document that will be produced in the failure investigation.