Coating Delamination on Water Treatment Tanks Free

A visual assessment of the coating condition of the tank exteriors was made at each building, identified as Building #1 and Building #2. At each building, there were relatively small areas or spots where the darker gray topcoat was delaminating from the underlying light gray coating on one of the two tanks. For tanks with failure, the delamination was present on the support legs and tank shell. There were also some small spots of coating damage and rust that appeared to be caused by mechanical damage.

At Building #1, the enclosure was completed with the tanks under cover. Tank 1A had areas of topcoat delamination on three of the support legs (particularly the front legs) and the tank shell. The back right leg had pinpoint rusting over the lower portions. The largest areas of delamination were on the front left leg, along most of the length at one edge, and on the upper tank shell near a pipe connection. The failing area on the upper shell was the largest and was approximately 12 inches by 18 inches (30.5 to 45.7 cm) in size. Another common area of delamination was on the base plate of the support legs. Except for the delamination, the coatings were in good condition. Tank 1B did not have any visible coating failure, and the coatings were in good condition.

At Building #2, the enclosure was not completed, and the tanks were still exposed to exterior conditions. Tank 2B had small areas of topcoat delamination on the support legs and tank shell. The failing areas on the shell were relatively small, ranging in size from approximately 1 inch by 2 inches (2.5 cm by 5.1 cm) in size to 2 inches by 2 inches. The delamination on the support legs primarily consisted of small spots on the base plates. Except for the delamination, the coatings were in good condition. Tank 2A did not have any visible coating failure, and the coatings were in good condition.

The total coating DFT was determined using an electronic gage. Spot measurements were recorded for all of the tanks. The average coating thickness for the tanks, including those without failure, ranged from 13 to 23 mils (330.2 to 584.2 microns). For the tank with pinpoint rusting on the back support leg, the DFT was only 3 to 4 mils in that area. Overall, the coating thickness was similar between the tanks and did not appear to correlate with the failing areas.

The number of coats present was determined using a Tooke gage, which destructively measures the thickness of each coat in multi-coat systems.

This gage allows for the observation of a coating cross-section created with a precision cutting tip of a known angle, and it shows the coating thickness in addition to possible contamination, voids, underlying rust, or mill scale, if present. The Tooke readings showed 8 to 10 mils (203.2 to 254.2 microns) of the light gray coating and 3 to 5 mils of the dark gray topcoat. The readings could not distinguish between multiple coats of the light gray layer (primer/intermediate). The steel surface had a blast cleaned appearance, and no contamination or defects were observed.

The coating adhesion was assessed according to ASTM D6677, “Standard Test Method for Evaluating Adhesion by Knife.” This method involves scribing an “X” through the coating with a razor knife and then probing the intersection with the knife to rate adhesion based on the amount of force required to disbond the coating and the size of the coating chip disbonded. A numeric rating scale (progressively ranging from 0 to 10 in increments of 2) is provided in the standard. A rating of 0 indicates that the coating was easily peeled from the substrate in pieces > ¼ inch (0.6 cm), and a 10 rating indicates that the coating was extremely difficult to remove, and fragments were no larger than approximately ⅓2 inch (0.07 cm). Typically, ratings of 8 to 10 are considered to represent good adhesion, 4 to 6 represent fair adhesion, and 0 to 2 represent poor adhesion.

Adhesion tests were performed on the tanks at locations where the coating was intact and not failing, including adjacent to areas of delamination on the tanks with failure. The adhesion of the topcoat to the underlying coating layer ranged from fair (4) to good (10) on these tanks, with the typical rating from 6 to 8. The adhesion of the exposed light gray coating (primer/intermediate) at areas where the topcoat had delaminated was good (rating of 10). For the two tanks without failure, Tank 1B had good adhesion (ratings 8 to 10), and Tank 2A had fair to good adhesion (ratings 6 to 10), with a typical rating of 8.

Samples from all tanks were obtained during the field investigation, including from failing and non-failing areas for the tanks with coating issues.

The intended laboratory analysis included microscopic examination and infrared spectroscopy. The microscopic examination of the samples was conducted using a digital microscope to characterize the coating characteristics and identify any obvious defects. The samples generally displayed three coating layers as specified, with coating thicknesses consistent with the field measurements. Nothing unusual was observed with the coating layers.

The infrared spectroscopic analysis was performed to determine if amine blush was present. Amine blush is a chemical reaction that occurs between water, carbon dioxide, and the amine component of epoxy coatings. This most often occurs when there are cooler and/or damp conditions present during the curing period of the coating. The result of amine blush is the formation of an oily or waxy film on the surface of the epoxy coating, which inhibits adhesion of a subsequently applied coating. Since all of the failure was between the epoxy intermediate and polyurethane finish, amine blush was a possible cause. Although the application conditions were within the manufacturer’s requirements, the air and surface temperatures were often near the minimum recommended temperature of 50 °F.

The testing involved rinsing any residue from the bottom surfaces of the delaminated samples and analyzing them with the infrared spectrometer. Surface rinsings were conducted using deionized water and a disposable glass pipette. The surface was rinsed repeatedly, and then the water was evaporated at room conditions. The resulting spectra did not contain any bands associated with an amine blush.

The presence of amine blush was also evaluated using an amine blush test kit. The test kit works by evaluating the pH of the surface. In theory, if an amine blush is present, the surface would be alkaline. The kit is provided with fluid in a spray bottle. The fluid is sprayed on the surface, and the color, after interaction with the surface, indicates the pH. If the surface has a neutral pH, the color of the fluid remains yellow. If the pH of the surface is alkaline, the color shifts to a blue/purple color.

To determine if the test fluid is working as designed, pre-treated cotton swabs are provided in the kit. The solution is sprayed on the swab, and if the solution is functioning, the swab turns blue. The fluid used for testing was verified by using one of the swabs provided with the kit. When the fluid was applied to samples, the color remained yellow, indicating a negative result for amine blush.

The field investigation found that small areas of the polyurethane topcoat were delaminating from the underlying epoxy coat on one of the tanks at each building. The adhesion of the topcoat to the underlying coating was fair to good on tanks with and without delamination. Overall, the coating adhesion was considered acceptable. The coating thickness was in general accordance with the specified thickness for the system and did not appear to be related to the problem.

Since the laboratory analysis did not find issues with the coatings or that amine blush occurred, attention focused on the application conditions in the shop and exposure conditions at the project site. As previously noted, the application conditions met requirements; however, there was a question of whether the minimum recoat times for the coatings were met. The recoat times for a temperature as low as 50 °F were 36 and 32 hours, respectively, for the primer and intermediate coats. If temperatures had remained near 50 °F between the application of each coat of the system, there was a possibility that the recoat time was too short.

The tanks were delivered in January and exposed to the winter conditions for several weeks. Weather records for the area showed that low temperatures in January and into February were typically below freezing and had gone below 10 °F (–12.2 °C) on multiple days. There had also been periodic snowfall of several inches over the period before the tanks were installed. These exposure conditions soon after the coating system had been applied appears to have initiated the failure. The thermal stresses from cold temperatures caused the topcoat to delaminate in limited areas where there was weaker coating adhesion. The questionable recoat time may have been a contributing factor in the topcoat not achieving good adhesion over all surfaces.

For the tanks with coating failure, the problem was relatively limited, making spot field repair the best option. The failing areas were to be prepared according to SSPC-SP 2, “Hand Tool Cleaning” and/or SSPC-SP 3, “Power Tool Cleaning.” The areas of coating delamination were cleaned, roughened, and feathered, followed by reapplication of the polyurethane topcoat. For areas of rusting, an epoxy primer was also applied.

Jay Helsel is a senior consultant with KTA-Tator. He holds an MS in Chemical Engineering from the University of Michigan, is a licensed professional engineer in multiple states, an AMPP Senior Certified Coatings Inspector, Protective Coatings Specialist, and Certified Concrete Coatings Inspector. At KTA, Helsel manages coating projects, performs coating failure investigations and coating and corrosion assessments, writes coating specifications, and is a regular instructor for KTA coating inspection courses.