The coatings failure that wasn’t…

A coatings contractor recently asked me to look at a failure on a job that he had completed within the last year.  The job had been to coat the tube steel legs on a new cooling tower, and the paint was “splitting and separating” as reported by the owner.  The tower was located in an exterior urban environment in Iowa without chemical exposure.  The job seemed fairly simple, and I had specified the products myself.  The coatings system included a moisture cured aluminum primer and an aliphatic polyurethane.  I had delivered the materials to the jobsite myself, so I was familiar with the contractor, the job, and the owner.  The weather was good on the days that the coatings were applied.  The painter was one of the most experienced members of the contractor’s crew.  Prior to visiting the jobsite to see the failure, the contractor and I had only seen photos of the failure.  Since I had been very involved with this job from beginning to end, I agreed to take a look to determine the cause of the failure.  

Although the entire cooling tower had been coated, the only failures present were on the northeast  tube steel support leg.   The failures were occurring primarily in a vertical direction in the center of this support on all four sides.   The coatings on the other three support legs were in excellent condition.  The failures were especially evident on the lower portion of the northeast support leg, and the condition of the coatings improved with greater height.  When failures are not uniformly spread over an entire project, I normally ask myself “What is different about the failure areas?” The answer to this question usually leads to the cause of the failure.  In this case, a close examination of the tube steel revealed that the steel was bowing out on all four sides of the northeast support leg and that the bowing was worse near the ground.  

A theory was developed that water had entered the legs from above and that it had frozen over the winter, causing the steel to bow out.  This condition stressed the coating, similar to a mandrel bend test in a cold weather environment.  

A hole was drilled near the bottom of the northeast support leg to see if there was water inside.  Judging by the amount of water that came out, it appears that the leg had been full of water from bottom to top.  

Due to the concern for compromised structural integrity of leg the that had expanded, it was removed and replaced.  Careful attention was taken to thoroughly seal the top of the leg to avoid the possibility for a reoccurrence of the condition.  For this paint guy, it was an interesting day and a happy end to the coatings failure that wasn’t.