Salt Contamination Showdown on a Ballast Tank Recoat
A yard is recoating the internal ballast tank of a 12-year-old bulk carrier. After abrasive blasting to Sa 2.5, the surface is a uniform grey-white. The specification requires soluble salt testing with a conductivity meter and a maximum of 50 mg/m² as chloride. The inspector performs three Bresle-style patch extractions at 20-minute intervals on the same area of the tank floor. Results: 38, 61, and 74 mg/m². Ambient conditions are 31 °C and 78% RH. The blaster claims the surface is 'visibly clean' and wants to start priming before the afternoon shift. The coating is a two-coat epoxy system with a maximum overcoating interval of 24 hours at 20 °C.
1. Interpret the three salt readings. What is the most defensible conclusion about surface condition, and what should the inspector do before any primer is applied?
Read the worked answer
The readings are not three independent samples of a stable surface - they are trending upward (38 → 61 → 74 mg/m²) on essentially the same spot. That trend is the key signal. It is consistent with salt migrating to the surface as residual moisture evaporates from the substrate at 31 °C, or with the extraction method progressively pulling salt out of pores and pits rather than a genuinely clean surface. Either way, the surface fails the 50 mg/m² limit at the second and third readings.
Reasoning steps:
- Visible cleanliness (Sa 2.5) says nothing about soluble salts; the blaster's 'looks clean' argument is not evidence.
- Three readings on one area are not a valid average. The highest reading governs, because salt is not uniformly distributed and a passing average can hide a failing pit.
- The upward trend indicates the test area may not be representative; the inspector should re-blast or fresh-water wash and re-test different, previously untested locations, not keep re-testing the same spot.
Action: reject the surface for priming, document the three readings with locations and times, require re-cleaning (typically high-pressure fresh water wash and re-blast or a proprietary salt remover), and re-test at multiple locations before accepting. Do not allow primer over a surface reading 74 mg/m².
2. The blaster offers to 'average the three readings to 57.7 and call it close enough.' How should the inspector respond, and what does the specification's intent tell you?
Read the worked answer
Decline. Averaging is not a valid acceptance method for a maximum-limit contamination test. A maximum limit means every tested location must be at or below 50 mg/m²; a location reading 74 fails regardless of what other locations read. Averaging would convert a pass/fail criterion into a statistical one the specification never authorized.
Also note the arithmetic: the mean of 38, 61, and 74 is 57.7, which is itself above 50 - so even the improper averaging argument fails on its own terms. That is worth pointing out because it closes the discussion quickly.
The specification's intent is to prevent osmotic blistering and premature disbondment under immersion. Soluble chloride at the coating/substrate interface draws water through the film osmotically; a localized high-salt pit can blister even if the surrounding area is clean. That is precisely why the limit is a maximum per location, not an average.
Document the refusal, the readings, and the re-cleaning requirement in the inspection report.