In this month’s Cases from the F-Files, rust spots became evident on the interior roof deck of a newly constructed, unoccupied warehouse. Were the rust spots a result of what was going on inside or outside of the warehouse?
Background of the Project
As a large warehouse building in the north-east U.S. that was being constructed for an automotive supplier was nearing completion, a problem was observed on the interior roof deck and the steel support structure for the roof deck. The roof structure was comprised of steel deck and associated steel joists supported by steel beams and columns. Various steel/iron piping was suspended from the roof. The large building was open from the floor to the roof except for a second-level mezzanine along one side of the building. The steel was specified to be shop-primed as follows:
- Roof deck: polyester primer, dry film thickness (DFT) of 0.20 to 0.30 mils
- Joists: alkyd dip primer, DFT of 0.8 to 1.2 mils
- Piping: no primer specified (reported to have a lacquer coating)
After erection, the roof structure was finish coated with a waterborne interior dry fall coating. Product information stated that the coating featured “good adhesion and resistance to flash rusting when applied to most types of interior ceilings and overhead surfaces in commercial and industrial buildings.” The recommended DFT was 1.5 to 2 mils. Painting of the roof started at the south end of the building and progressed to the north.
Application of the coating to the roof structure began in late summer and continued into November. The polyester primer was shop-applied. In early November, following a reported period of heavy rain in late October, rust spots were observed on the roof deck (Fig. 1). The building had not been conditioned through at least October, and there was no control of conditions until the building’s HVAC system was eventually in operation. The interior environment was reported to be humid following the wet period. An independent investigation into the cause of the rust spotting was requested.
Fig. 1: Typical corrosion/rusting along edge of corrugated roof deck. All figures are courtesy of KTA-Tator. |
The Site Investigation
The roof deck structure was accessed in various locations throughout the facility. At each location, visual observations, adhesion evaluations, and coating thickness measurements were made. The finish coat applied to the roof deck and structure was white throughout the building.
Visual Examination
Although some rusting of the roof deck could be observed from the floor of the building, the extent of the problem was not obvious until the roof structure was examined from a close distance. Small rust spots were observed scattered over the roof deck in most locations. The rust spots typically had formed along the overlapping angled edges of adjacent deck pieces. When the finish coating was scraped from the deck at rust spots, little corrosion of the steel deck was found, but dark, black spots were evident, indicating the source of the corrosion (Fig. 2).
Rust spots were present on the roof deck to some degree in all locations examined, although the rusting was generally more prevalent toward the south end of the building. The white finish coat on the deck was typically glossy and uniform in coverage; however, some limited areas of incomplete coverage were found toward the north end of the roof. At these locations, gray deck primer could be seen beneath the white finish coat. Consistent with the north/south pattern described above, less rusting was present where the finish coat did not achieve complete coverage. Although painting the roof structure in the facility had been completed, the finish coat had not been applied on some locations along the west side because of limited access to these areas. The gray primer of the roof deck was examined in two such locations and appeared to be in relatively good condition. But when small darker spots were examined under magnification, it was obvious that some small rust spots were present in limited access areas where the finish coat had not been applied.
In addition to the north/south differences in the degree of rust spots observed on the roof deck, a lesser amount of rusting was observed on the mezzanine levels of the building along the eastern side. The lower roof deck, underneath the mezzanine deck, in particular, displayed less (but some) rust spots (Fig. 3). The main roof deck, or the upper level roof at the mezzanine, had some of the typical rust spot formation, but less overall than for the roof over the open portion of the building. Again, rust spots were most often present along the angled edges of overlapping deck pieces at the upper level of the mezzanine.
Fig. 2: Rusting on roof deck where coating was removed showing corrosion (black color) in the steel. |
Fig. 3: Area of lesser corrosion on the roof deck. |
Adhesion and Thickness
The adhesion of the finish coat, as assessed by the tape test method (ASTM D3359, Method A), was good (ratings of 4A or 5A) in all locations evaluated. The total thickness of the coating system was measured non-destructively using an electronic gage. The measurements were relatively consistent and showed a greater thickness on piping and roof structure components as compared to the roof deck. The total coating thickness for the roof deck ranged from 3 to 5 mils, with the thickness on the joist structure considerably higher at 5 to 15 mils (noting that the total thickness measurements included any existing primer layer on the component). The primer thickness of the roof deck was measured at 0.3 to 0.5 mils in the locations where the finish coat had not been applied. Representative samples were collected throughout the facility for laboratory analysis.
Laboratory Investigation
Laboratory analysis included visual and microscopic examination and infrared spectroscopic analysis of the roof deck primer. The results of the microscopy supported the field observations and described corrosion on the backside of coating samples removed from areas where rust spots were present on the roof deck and piping. The infrared spectroscopic analysis identified the roof deck primer as a polyester coating consistent with the product information for this primer. The finish coat thickness was in general accordance with the manufacturer’s product recommendations and was typically greater than the recommended 2 mils. The primer thickness was thin but consistent with the specified range.
Pulling the Site and Laboratory Investigation Together
In reviewing all of the details related to work at the job site, the reported period of wet weather appeared to be a significant event. In fact, when records for a nearby weather station were reviewed, it was discovered that two significant rain events had occurred. The records showed a period of heavy rainfall (0.84 inches) in early October, including nearly a three-day period where the dew point temperature remained at or near the ambient air temperature (Fig. 4). Because the interior of the building was not conditioned, it was assumed that the interior temperatures were similar (or worse) than the exterior condition. A similar period of heavy rain also occurred in late October, with 0.75 inches of rain falling and another extended period where the dew point temperature remained very close to the ambient air temperature (Fig. 5). Rust spots on the roof structure were reported shortly after the late October wet period, but likely began to form after the earlier period of rainfall.
Fig. 4: Early October weather data |
Fig. 5: Late October weather data |
The investigation revealed that the formation of rust spots over the roof deck was caused by three factors:
- The use of a thin primer on the roof deck without substantial corrosion resistance properties
- The introduction of moisture inside the building from October rain events
- The use of a water-based acrylic finish coat in conjunction with the thin deck primer and moisture in the building
The polyester deck primer did not provide much corrosion resistance to the steel deck. The thin layer could not prevent corrosion of the steel deck after moisture was introduced in the building interior. Further, the use of the water-based acrylic finish coat could not prevent the formation of rust spots, given the thin roof deck primer and introduction of moisture in the building. Although the product information for the finish coat stated that the coating was resistant to flash rusting, water-based acrylic coatings generally do not provide optimal corrosion resistance. Lack of corrosion resistance allowed rust to come through the finish coat (Fig. 6). Significant moisture trapped in an interior environment can eventually permeate the coating to reach the steel substrate and initiate corrosion. Once this occurs, the coating film tends to hold the moisture, allowing corrosion to continue until corrosion and corrosion staining from the steel migrates through the coating, forming rust spots.
Fig. 6: Typical corrosion/rusting along edge of corrugated roof deck. |
The degree of corrosion was not expected to worsen if the interior environment remained dry and environmentally controlled. The recommended coating repair was a new coat of the alkyd repair coating applied to the roof deck to sufficiently seal the surface and prevent any continued formation of rust spots. Since the building interior was conditioned at this point, extra control of the interior conditions was not specifically recommended.
Conclusion
The beginning of this article asked, “Were the rust spots a result of what was going on inside or outside of the warehouse?” The answer is both. Since the inside of the building was not conditioned, what happened outside—the significant rain and a dew point temperature that remained close to the ambient air temperature—affected the conditions inside, therefore causing a coating failure.
Article From The Journal Of Protective Coatings & Linings ©2012 Technology Publishing Company
Jay Helsel is a Senior Coatings Consultant with KTA where he has been employed for over 10 years. He is a registered Professional Engineer in numerous states, and he is also a NACE Certified Coatings Inspector Level 3 (Peer Review). Jay has extensive marine and shipboard experience having served 11 years in the Coast Guard, most recently as a Lieutenant Commander in the area of Marine Vessel Inspection. www.env-ktacom-ktastaging.kinsta.cloud