Prepare for painting

Maintenance repaints are easy, he told me. Just wash the old topcoat and apply another coat of the same product and the new coating should be good for a few more years. Such was the belief of the engineer who wrote the specification for recoating the exterior of four large tanks.

Unfortunately, during 10 years in service, the zinc primer had reacted with the topcoat and weakened its bond to the primer, and the coating manufacturer had updated the formulation of the new topcoat material. The plasticizer in the new topcoat softened the old topcoat and further weakened its bond to the primer. The shrinkage stress that developed as the new topcoat cured caused the bond between the old topcoat and the primer to break. After six months, both topcoats were hanging down the sides of the tanks like sheets of wet wallpaper. It was an expensive way to strip a tank for recoating.

The above case might seem extreme, but failures after maintenance recoating occur more frequently than coating failures on new construction, especially when applying new material over old coatings. If multiple coats have built up on old structures, often there are planes of weakness between some of the coats and these are prone to fail under the stress imposed by a freshly applied material.

Test and evaluate

The best way to avoid coating failures is to plan at least one year ahead. Having decided on the best options for repainting, try the proposed procedures and materials by applying test patches. The patches should be at least one square foot. If different types of service environment are involved, use one patch for each environment. For exterior coatings subject to winter freeze, it’s usually sufficient to apply the patches in the fall. Anything that can go wrong usually will have gone wrong by the spring. In climate controlled environments, allow at least a year for failures to show up.

Evaluate the adhesion of the existing coating before applying the test patches. Test the adhesion of the combined coatings about a week after applying the patches and at the end of the test. Evaluate adhesion by using knife cuts and adhesive tape or just by using a sharp knife. The two methods are described in ASTM D 3359-09 and ASTM D 6677-07, respectively. The first test requires cutting either a 45° cross-cut or a cross-hatch pattern into the coating, applying adhesive tape over the cut and peeling the tape off to see how much coating it removes. The second method requires just a 45° cross followed by probing with the point of a knife.

Keep records to plan recoats

You need to know the identity of the existing coating system and its overall condition as well as the substrate type and condition to select an appropriate maintenance recoating procedure. Hopefully you’ll have logs and inspection reports from previous coating operations. You also need to know the characteristics and limitations of the new coatings that are possible candidates for the project.

If you’re fortunate, you’ll find records showing the identity of the existing coating system, but many times this won’t be the case. A school district had the corridors in its schools painted with graffiti-resistant paint, then lost the record. No doubt the janitors knew that nothing would stick to the walls, but not the architect who specified a fresh coat of acrylic latex as part of a redecoration project. Needless to say, it wasn’t a success.

Some coatings are just incompatible with one another. For example, alkyd paints (oil-based) are easily softened by the strong solvents in many two-component coatings such as epoxies and polyurethanes. If these are applied over old alkyd coatings, they might act like paint strippers and peel the old paint off the substrate. Also, many water-based latex paints might not establish good adhesion to old oil-based coatings. Of course, these problems will show up when test patches are applied, but time will be lost.

Knowing the type of existing paint is an important part of choosing a product that can be applied over it successfully. ASTM D 5043-04 describes a procedure for field identification of coatings, but it requires access to a variety of chemicals. Chips of the coating can be identified by a suitably experienced laboratory using infrared spectroscopy, and it might be possible to get the candidate coating supplier’s laboratory to do so. But keeping good records of coating work avoids the problem.

Factory finishes require finesse

An especially difficult problem is maintaining factory-coated siding and roof decking. A galvanized steel substrate has a thin, baked on layer of specialized coating. These coatings include polyesters, silicones and the toughest recoat problem of all, Kynar.

Kynar, a polyvinylidene fluoride, is a close relative of the polymer coating on the inside of non-stick cookware. The manufacturers of these coatings offer special repair products and they should be consulted before trying to repaint this type of siding. Not doing so has lead to some spectacular failures.

Challenges of concrete and cementitious substrates

Failures frequently occur when attempting to recoat and rehabilitate concrete and other similar substrates. Concrete floors are especially problematic. Concrete is porous, which means that contaminants penetrate into it and water permeates through it. An additional complication is laitance, a weak layer of cement and aggregate fines that forms on a concrete surface. This layer can be surprisingly thick if the slab was poured using concrete containing too much water.

It’s possible to determine whether moisture permeation is going to be a problem by taping squares of clear plastic sheeting to the concrete and watching for condensation to form on their lower surfaces (ASTM D 4263 - 83). Moisture permeation rate also can be determined by applying calcium chloride to the concrete and measuring the weight increase of over time (ASTM F 1869 - 09).

This test must be performed close to the time that coating is to be applied. A very large floor coating job on the Gulf Coast failed because a satisfactory moisture permeation rate was measured at the end of a month of drought, but the new coating lifted two days after a tropical storm. By then, the water table was above the top of the slab.

Moisture permeation also can be a problem with concrete walls. A cool, air-conditioned building interior in a warm, humid climate can experience a lot of water vapor movement through the walls. The coating on the interior of a Florida orange juice processing plant was sufficiently permeable to permit water vapor to move through it. The new coat of acrylic latex was a lot less permeable. Water accumulated behind the new paint and caused it and the old coating to delaminate. A test patch would have revealed the wisdom of choosing a different paint for the redecoration.

Ideally, the time to apply a topping to a concrete floor is when it’s new. Many times, we wait until the floor is damaged and dirty before deciding to rehabilitate it with a heavy floor topping, which often consists of multiple coats of an expensive resinous system.

Contaminants soak deep into concrete and aren’t easy to remove. Some, such as acids or foods that promote the growth of acid-producing bacteria, can seriously reduce the strength of the concrete. Removing oily contamination with solvents results in both the solvent and the contaminant soaking farther down into the concrete, which is likely to cause the floor topping to fail later. Also, the concrete surface might have been easily damaged if it consisted of a weak layer of laitance.

It’s a wise precaution to examine some core samples from an old concrete floor before planning to apply a floor topping. The cores show the condition of the concrete; the thickness of the laitance, which must be removed before applying the topping; the depth to which contamination has penetrated; and whether there’s a waterproof membrane below the slab.

Cores taken from an old hangar, where paint had been stripped from aircraft, revealed that the slab was saturated with paint stripper all the way down to the underlying soil. Attempts to take core samples from the floor of a dye manufacturing plant resulted in nothing but piles of aggregate. The cement paste had been completely eaten away. Both projects required completely new floors.

Non-ferrous metals need non-standard coatings

Another substrate that causes a lot of frustration and heartbreak is galvanizing. This is a coating of zinc alloyed to steel. Unlike steel, zinc is a highly reactive metal, which is protected by a thin film of zinc oxide and carbonate, the chemistry of which is complex. Coating galvanized steel as if it were plain steel leads to a lot of failures.

Paints based on natural oils, such as alkyds, react with zinc and a zinc soap forms at the interface between the paint and the metal. This invariably leads to the paint prematurely peeling off the galvanizing. One exception is a composition made from linseed oil mixed with metallic zinc dust and zinc oxide. This is a good primer for repairing failed galvanizing, but because the pigments react with the oil, it has a short shelf-life before it solidifies in the can.

Other coatings, such as acrylics, epoxies and polyurethanes, don’t react with the zinc substrate but can still cause problems. Zinc needs access to oxygen to remain passive and stable. These coatings are more permeable to water vapor than they are to oxygen and, in a damp environment, rapid underfilm corrosion of the zinc can occur. This generates voluminous amounts of “white rust” and blisters the coating off the galvanized surface. The same corrosion occurs when water gets trapped between sheets of coated galvanized siding or decking, but it’s called “wet stain.”

Ideally, prime galvanized steel with a zinc chromate etch or wash primer before applying any other coating. The chromate pigment acts as a corrosion inhibitor and passivates the zinc. Some proprietary alternatives don’t contain hexavalent chromium. The same considerations also apply to aluminum and barrier coatings. Most of the time, galvanizing and aluminum are substrates better left uncoated.

The good news is that most coating work doesn’t fail prematurely. However, most of the failures could have been avoided by better planning and testing the proposed recoat system beforehand.

Geoff Byrnes, CFE, is the president of G.B. Byrnes Consulting, Inc., Houston. Contact him at [email protected] and (713) 460-4130.