Salt Contamination on a Bridge Recoat Over Solvent-Borne Inorganic Zinc
A 1970s steel girder bridge is being recoated. The existing system is a solvent-borne inorganic zinc primer that is largely intact on the webs but has 15-20% rust-through on the bottom flanges and around rivets. The specification calls for SSPC-SP 10/NACE No. 2 near-white blasting followed by a zinc-rich primer and an epoxy intermediate. The contractor blasts a test area, achieves a 2.5 mil anchor profile, and the inspector measures soluble salts by Bresle method at 12 µg/cm² chloride on the blasted web, but only 3 µg/cm² on a freshly blasted bottom flange. Ambient conditions are 38°F with 78% RH; steel temperature reads 40°F. The spec cites an allowable chloride limit of 5 µg/cm². The contractor argues that the web readings are 'just surface residue' and wants to prime the flanges before the end of shift to avoid flash rust overnight.
1. How should the inspector respond to the elevated chloride readings on the web versus the flange, and what does the discrepancy most likely indicate?
Read the worked answer
The inspector cannot average or dismiss the web readings. Soluble salt testing is a pass/fail against the specified limit at each test location; 12 µg/cm² exceeds the 5 µg/cm² limit by more than 2x and must be re-cleaned (typically by fresh-water wash or wet abrasive blasting) and re-tested before priming.
The discrepancy is diagnostic, not random. The web retains intact inorganic zinc over a substrate that has been in service for decades, and zinc-rich coatings on aged bridges commonly concentrate chloride at the coating/steel interface from years of deicing salt exposure and moisture ingress. Blasting removes the coating but can smear or drive contaminated residue into the profile. The flange, where rust-through has already occurred and corrosion products were more readily removed, reads low. This pattern suggests the contamination is substrate/interface-borne rather than airborne fallout.
Practical implications: (1) the web may need a detergent or high-pressure fresh-water wash followed by re-blasting, not just a second blast pass; (2) the inspector should verify the Bresle patch technique (patch size, injection volume, conductivity meter calibration) to rule out test error; (3) the contractor's 'prime before end of shift' request must be refused until re-test passes.
2. Given the ambient and steel temperatures, is the contractor's plan to prime the flanges this shift technically defensible? Explain the reasoning using dew point and coating cure considerations.
Read the worked answer
No, and the inspector should document the refusal. At 38°F dry bulb and 78% RH, the dew point is approximately 34°F (using standard psychrometric tables). The steel at 40°F is only about 6°F above the dew point - a common specification requires the substrate to be at least 5°F above dew point, and this is marginal at best, with the risk that as ambient temperature drops overnight the steel will fall to or below dew point and condensation will form on freshly blasted or freshly primed steel.
Separately, most zinc-rich primers and epoxy intermediates have a minimum application temperature (often 40-50°F for the product) and a minimum substrate temperature. At 40°F steel and falling ambient, cure will be severely retarded, and a slow-curing epoxy over a marginally dry zinc primer invites intercoat adhesion failure and possible amine blush. The contractor's schedule pressure does not override these limits.
The correct call: hold application, monitor steel temperature and dew point continuously (recording at intervals), and either heat the steel/enclosure or wait for conditions within the product data sheet window. Document the hold with time-stamped readings.