CNST 201 Corrosion Report

Mr. Edifice:

The writer has completed the authorized inspection of the corrosion at the referenced site. The purpose of this inspection was to determine the cause of the corrosion in this metal building. This report presents our findings and our engineering evaluation.

Site Inspection

On November 12, 1996, the writer performed a site inspection of the referenced site. This inspection was observed by Ms. Fayld Structure. The inspection was limited to only accessible areas (i.e. wall panels and ceiling tiles were not removed to expose the interior of the building).

Ms. Structure informed the writer that the rust was first noted on the south side of the building. The rust has continued to develop and is now apparent at several locations.

During our inspection the following items were noted:

1. Exterior of the Building
   1. The majority of the corrosion was at the bottom of the metal sheeting near the fasteners on the north and south sides of the building. These fasteners were nailed into wood that was attached directly to the footing.
   2. Corrosion was noted at several doors and windows. The writer did not observe caulking or flashing at these locations.
   3. Corrosion was noted on the awning covering the entrance to the building.
   4. Our inspection was performed after a rain. Several low areas that had ponded water were noted within ten (10) feet of the footing on the north and south sides. Areas of footing near corroded metal were observed as wet.

2. Interior of Building
   1. Corrosion was noted at several isolated locations at fasteners and seams within the metal roofing. In most cases, the corrosion appeared to be shallow.
   2. Most of the corrosion on the metal roof was located at fasteners attached to wood stamped Fire-X.
   3. This corrosion was noted to be concentrated over the garage area.
   4. An exhaust fan was on one wall of the attic storage area. This fan was observed to be working.
   5. The writer did not observe any damp or wet areas within the building. Also the writer did not observe any evidence of previous water damage.
   6. The smoke stack appears to be fairly new. There was no evidence of deterioration.
   7. Several of the trusses were checked. There was no evidence of deterioration or distress except for minor flaking of the coating and very isolated areas with surface rust.

3. Sampling for Laboratory Testing: During our inspection, the writer removed the following samples for laboratory testing.
   1. Metal sheeting that had no evidence of corrosion.
   2. Metal sheeting that had evidence of corrosion.
   3. A piece of wood that was attached to the footing and to the metal siding that had corroded.
   4. A grey-white powder found on the above wood (referenced in #3c).
   5. Wood samples from trusses and purlins marked Fire-X.

The writer was informed by the contractor that the following materials were used on this project:

1. Side wall metal - "Regal Rib", manufactured by Cincinnati Sheet Metal and Roofing Co.
2. Roof Metal - "Strongpanel", manufactured by Granite City Steel.
3. Fire treated lumber - purchased from Southern Illinois Lumber Company (this lumber was marked Fire-X and was treated by Dixie Wood Preserving Company).

The writer called a representative with Cincinnati Sheet Metal and Roofing Company. We were informed that Regal Rib was dipped in zinc to provide a protective coating. A bonding treatment and an epoxy primer was also applied to both sides of the sheet. A siliconized polyester baked enamel finish was applied to the exposed surface.

General Review of Corrosion

Since this report is concentrated on the corrosion damage, it is the writer's opinion that a general review of corrosion is necessary. Most metals when exposed to oxygen will form a metal oxide. These metal oxides are corrosion product. For metals, such as zinc, aluminum and tin, the metal oxides will form a protective layer that will prevent additional chemical reactions. Therefore, corrosion damage would be limited to the surface oxide layer and further damage would not likely occur.

When iron is exposed to oxygen, it does not form this protective layer but forms a porous ironoxide. This will allow oxygen to penetrate through the layer; corrosion will continue unless preventive measures are undertaken.

One of the methods used to prevent corrosion of iron or steel is to use a zinc coating or to galvanize the steel. The zinc coating when exposed to oxygen would normally form a protective zinc oxide layer. This would prevent further corrosion of either the zinc or the steel.

For corrosion of a zinc-coated steel to occur, the protective zinc oxide coating must be continuously removed so that fresh zinc is exposed. This reaction would continue until the zinc is exhausted. The corrosion of the iron would begin. Under normal atmospheric conditions, corrosion to zinc-coated metal would be extremely low. This results because the zinc oxide layer would not be eroded away. Since zinc oxides have a low solubility in pure water (or relatively pure water such as rain water), this condition would provide a relatively slow rate of corrosion.

The solubility of zinc oxide increases dramatically with acidic water or in water that contains dissolved salts. This water must be in liquid form (not high humidity) so that the zinc oxides can go into solution. Thus, if a zinc-coated iron has an environment of acidic water and /or water with dissolved salts existed, the protective zinc oxides would be removed and corrosion would be expected.

General Review of Fire-Retardant Treatments

Normal fire retardant treatments by chemical impregnation require that the wood retain from 2 ½ to 5 pounds of dry salts per cubic foot of wood. The salts that are principally used are monammonium and diammonium phosphates, ammonium sulfates, zinc chloride, sodiumtetraborate and boric acid.

Fire-retardant wood has two partial problems that can influence corrosion. The first problem is that many of the fire-retarded salts are quite corrosive to metal. However, combinations of those salts can be used to provide neutral formulation and thus a somewhat less corrosive treatment. Also if corrosion inhibitors, such as sodium dichromate, have been added, corrosion can be reduced to insignificant levels.

Second, fire retardants are usually more hydroscopic than untreated wood. At high humidities the fire-retarded salts may be extruded from the wood and be available to enter into a water solution. Therefore if a treatment does not include corrosion inhibitors and the salts are extruded into a water solution, a corrosive environment would be expected.

Discussion of the Finding

The two samples of sheet metals were analyzed to determine the zinc and iron contents. The results of the tests are listed in the appendix.

The sample with visible corrosion was completely void of the interior surface coating sand the zinc coating. This sample also contained approximately one-half the amount of zinc as found in the non-corroded sample.

From these results, it is apparent that a corrosive environment was found directly on the interior of the metal sheeting. To create this corrosive environment, the interior of the sheet metal was likely to have been in contact with water in liquid form that contained acids or dissolved salts.

Analysis of Wood Sample

Five wood samples were analyzed to determine the amounts of zinc, iron and chromates. These results are shown in the appendix.

All the wood samples showed very low levels of chromates. This amount would be considered similar to the amount in untreated wood. This indicates that sodium dichromate, a normal corrosion inhibitor, was not present in the fire treatment formulation. This observation was also confirmed by the Dixie Wood Preserving Company. They stated that the treatment formulation was clear; a treatment formulation that contained chromates would be bright yellow in color.

The sample of wood from the footing, taken near extensive corrosion, indicates a substantially higher iron and zinc content than the other samples. It is apparent that the corrosion products were transported form the metal siding into the treated wood. Since the zinc and iron would only be slightly soluble in pure water, an acid or salt water would be needed for the transport. Also, since iron and zinc would be soluble only in water in liquid form, the transport of the corrosion product from the corroded metal siding into the wood sample would likely occur in a liquid solution of aid or salt water. This transport of the corrosion into the wood would not be expected from a high humidity.

The grey-white powder that was found on the wood attached to the footing was analyzed and found to be essentially calcium salts with minor amounts of iron (8.9%) and zinc (1.9%).

A review of the analysis and the inspection findings indicate that for corrosion to happen water inliquid form containing acid and/or salts must come into contact with the metal. The writer observed low areas near the foundations and uncaulked joints near windows, doors, and awnings that would trap liquid water. This water would be in contact with the metal siding and fire-retardant wood. Water can also enter the wood by the wicking action or capillary action of the wood. This action would draw water into the wood at the concrete foundation. Water would also be drawn into the wood due to the hydroscopic characteristic of fire-treated wood. This water would be available from the surface and ground water found near the building.

The leached salts from the fire-retarded wood could create a highly corrosive environment that would deteriorate the steel siding. The corrosion on only the inside of the metal siding plus deposits of the corrosion products in the wood again suggest the presence of acid and/or salt water in liquid form.

Summary and Conclusions

1. For corrosion to occur to a zinc-coated metal siding, water in liquid form that contains salts and/or acids must come in contact with the metal.
2. The field inspection indicated that water in liquid for can easily enter the system.
3. The presence of zinc and iron in the wood near corrosion indicates that water in liquid form has existed and has transported the corrosion product into the wood.
4. Fire-retarded wood normally has salts that can be extruded at high humidities. These salts could create an acid and/or salt solution with water. Our test indicates that the Fire-X formulation did not contain sodium dichromate, and normal corrosion inhibitor.
5. Corrosion had begun on the inside surface where a direct wood-metal siding contact could occur.
6. Corrosion of the windows, doors, and awnings would indicate that the salt/acids would not come from ground water. In these areas, the water in liquid form would likely be trapped rain water.
7. It is the writer's opinion that the corrosion is a result of the water leaching salts from the fire-retarded wood and coming into contact with the metal siding. This has resulted in removal of the zinc coating and has caused corrosion to the metal siding.

It has been a pleasure to serve as your consultant on this project. If you have any questions with the report, please call.

Very truly yours,

Losten T. Dark, P.E.
Consultant

Appendix

Table I - Metallic Composition of Metal Siding
Sample	%Iron	%Zinc
Sample Without Corrosion	80.28	5.20
Sample With Corrosion	90.44	3.47

Table II -- Analysis of Metals From Wood
Sample Location	%Iron	%Zinc	%Chromite
Purlin	0.009	0.17	Less than 0.01
Truss - Stamped Fire X	0.011	0.42	Less than 0.01
Truss - Stamped Fire X	0.017	0.002	Less than 0.01
Wood Nailed to Footing	5.07	0.72	Less than 0.01
Reference Piece - Untreated Pine not from thisproject	--	0.003	Less than 0.01