KTA-Tator, Inc.

About Us

Services

Careers

Trade Shows

News & Articles

KTA-TATOR, INC.
Corporate Headquarters
115 Technology Drive
Pittsburgh, PA 15275
Phone: 412.788.1300
Fax: 412.788.1306
E-mail: info@kta.com

Instrument Sales
1-800-KTA-GAGE
Lead Hotline
1-800-LEAD-OFF

Home

KTA Challenge #15
Answer - Chloride


Photograph 1	Photograph 2

It is clear from the information provided that the specified thickness was applied throughout the project, and the degree of surface preparation in non-failing areas was good. These areas were free of corrosion and the coating was well-adherent. The failure typically involved areas where the new coating was applied directly to the bare metal (the SSPC-SP15 areas), and was associated with pits in the steel. The specification for SSPC-SP15 allows for residues of rust and paint in the pits, so that fact that rust was present beneath the coating in these areas doesn't necessarily mean that the degree of power tool cleaning did not meet the specification requirements

The data from the field inspection of the failure that was not disclosed in the clues involves chloride testing. The surfaces were wiped clean and washed to remove surface contamination, then the coating was chipped away from both non-failing and failing pitted areas. The exposed substrate was tested using the Bresle Cell. Although the specification clearly dictated that swabbing methods be used, you can not adequately control the extraction procedure on an overhead surface using swabbing, so the Bresle Cell was used for the investigation. Even if you can collect the water from swabbing, the cell methods are preferred. The results in three failing areas were 16, 45, and 102 g/cm². It is likely that the actual values were even higher because some of the corrosion products (which likely contained chloride) were removed during the mechanical removal of the coating. The results in non-failing areas after removal of the coating were below the detection limit of the test (<4 g/cm²).

When the painting work was done, because of problems encountered when trying to test the underside of the flanges, the chloride tests were only taken on the top of the top flange and on the web after cleaning. Assumptions were made that these values would be representative of the underside of the flanges. There were a few problems with this assumption. First, the web and top of the flanges exhibited less pitting and contained much more intact paint that was being overcoated. These surfaces were also readily accessible for pressure washing. In contrast, the undersides of the flanges were primarily bare pitted steel and were located within a few feet of the floor of the containment, making it much more difficult to thoroughly wash.

Another problem with washing prior to mechanical cleaning is that depending on the thoroughness of the washing and the thickness of the corrosion, the washing may only remove the chloride from the surface of the corrosion. Once the corrosion is removed by mechanical cleaning, chloride can still remain on the steel. Unless additional washing is performed, enough chloride may remain to cause problems. In this case, the bottom flange was never tested, so there was no way of knowing if the cleaning was adequate, or if additional washing should have been performed after power tool cleaning and prior to painting.

The problem was compounded by the pinholing that was present in some of the areas. Extra care must be taken when applying moisture cure urethanes in pitted areas as outgassing can occur, leading to pinholes through the coating. The surface should be thoroughly inspected for the presence of pinholes, and the pinholes should be repaired prior to the application of subsequent coats. However, even with the pinholing, chloride was the major factor in the rust-through of the film in this case. Failing pitted areas were present that showed no evidence of pinholes through the film. The underlying levels of chloride were high. In contrast, similar patches of pitted steel on the top of the flange were not corroding. The chloride beneath the film was non-detectible.

As an aside, in the field it is very difficult to accurately measure the thickness of coatings in pitted areas. The Tooke Gage was used, but note that this instrument requires making a scribe at a known angle through the coating. The scribe is then examined through a 50x microscope. For the Tooke Gage readings to be accurate, the angle of the scribe must be controlled. Unfortunately, the design of the gage requires 3 points of contact with the surface when making the scribe (2 legs and the cutting tip). When dragging the cutting tip over pitted steel, if one of the legs drops into a pitted area, the angle of the cutting tip is changed and as a result, the values are inaccurate. In many cases, the best that can be done in the field is to measure as close to the pitting as possible, making sure that all 3 contact points of the gage remain on the same plane while making the scribe.

KTA can assist you with specification language related to chlorides, methods of chloride remediation, or in assessing the cause of coatings failures.

Sixty five percent of the answers correctly concluded that chlorides remaining on the surface prior to painting were the culprit. Other responses included inadequate power tool cleaning, incomplete application of the primer to the pitted steel, too much moisture at the time of application, metal fragments created by power tool cleaning that became embedded in the coating, and poor design (abrasive blast cleaning should have been specified). The winner, by luck of the draw is Pieter van der Poel of The Eastern Seaboard Industrial Estate (Thailand), who also provided a very thorough response.