You are on page 1of 8


Performance of Coating Systems
in Industrial Atmosphere
011 the Arabian Gulf
J. Carew, A. AI-Haslzem, and w.I. Riad
Materials Application Department, Kuwait Institule for Scientific Research, p.o. Box 24885, Safat 13109, Kuwait

M. Othman
Shuaiba Area Authority, P.CJ. Box 4690, Sa"lt 13047, KUlcait

M. Islam
COI'test Cnlumbus Technologies Inc., 2704 Sunbury Blvd., Columbus. Off 43235

The performance and durability of 11 industrial coating systems were studied for two Other studies have dealt with a
and a half years at five weathering sites in the industrial belt of Shuaiba Area, KU'wait, range of building materials, includ-
in the Arabian Gulf. Coating performance was related to prevailing industrial atmos- ing plastics, sealants, and surface coat-
pheric conditions and compared with their behavior in laboratory accelerated tests. ings on concrete mortar exposed to
Experience has shown that under conditions like those in Kuwaiti industrial areas, weather conditions in Dubai, Saudi
degradation of coatings has tended to be faster than in the Western countries for which Arabia, for more than seven years. 2,3
most of these coating systems were developed. In the high mean and extreme tem-
peratures at Dubai, most of the mate-
he need to establish proper comparisons remained largely un- rials studied degraded more rapidly
selection criteria for industrial published. One paper presented ex- than when exposed at a site in the
coatings used in the industrial tensive data on the performance of United Kingdom. However, no single
belt of Kuwait is growing. Many coat- several marine, industrial, and resi- factor of acceleration could be defined
ing systems developed and manu- dential coatings under atmospheric between the two sites.
factured for temperate regions have conditions in Saudi Arabia over a 10- The current test program was
been used in hot and harsh climates, year period. 1 The paper also covered undertaken to fulfill the need for in-
but they performed inadeql]ately the service performance of aluminum formation and to provide guidance
there. (It should be noted that some alkyd, alkyd enamel, chlorinated rub- to decision makers in the local indus-
failures have been related to paints ber, epoxy, and vinyl mastic coating tries on paint selection for specific
applied during difficult weather con- systems under mild, refinery, and ma- areas in the industrial belt. This ar-
ditions.) However, little information rine atmospheres over a period of 20 ticle describes the performance results
is available on the performance or years. All the coatings appeared to obtained for commercial coating sys-
results from exposure tests of indus- have performed well under the con- tems exposed at different test sites
trial coatings in the Arabian Gulf. ditions of exposure. The paper con- for a period of two and a half years.
What information is available has cluded that the most critical factor Because of the long-term exposure
been based on tests conducted by in- affecting service performance of most required for the outdoor tests, the
dividual companies on their own of these coatings systems was the sur- coating systems were evaluated first
products, and specific results and face preparation. in accelerated laboratory tests.

24 MP/December 1994

dured by industrial structures along posure results. radiation/ condensation cycling. a more the Arabian Gulf seashore of Kuwait.te primer (based orrlon~"oil 40 alkyd and ureti1il-ne alkyd) " 1 to 2 Alkyd enamel ' 30 3 Zinc-rich epoxy 50 Epoxy polyamide/micaceous iron oxide 100 250 8 to 10 Epoxy polyamide/micaceous iron oxide 100 4 Chlorina)fJd r~.13 cleaned with distilled water wt% sodium chloride dissolved in MP/December 1994 25 . dium chloride plus sulfate. recent publication concluded that for Salt Spray/UV-Light Radiation/ a medium-oil alkyd primer contain. was measured after curing for two simulate outdoor weathering condi.bber primer 75 Chlori natedrubberundercoal 75 190 2 to 3 Chlorinate(tfubbertopcoat 40 5 Epoxy (two component) 50 200 3 to 4 Polyurethane 150 6 Zinc epoxy 50 Epoxy polyamide sealer 25 225 5 to 7 Epoxy polyamide (high build) 100 Polyurethane 50 7 Inorganic zinc silicate 50 200 10 to 12 Epoxy polyamide 100 Polyurethane 50 8 Acrylic wate'r-base~' 40 Unknown (new AcryliC enarnelwi'!ler-based 30 system) 9 Zinc chromate (based on oil-modified 80 alkyd) 40 60 2 to 3 Alkyd undercoat 40 Alkyd gloss 10 Zinc ethyl silicate :. 4-11 These workers conduded cycle represented a very severe ser. It has long been argued that continuous UV-light/ condensation. undercoat. and allowed to dry at coatings make use of either a con. the laboratory tests made use 2 1/2. The wet! dry salt spray test used in applied. After the topcoat cured. The exposures appeared to be identical. 3D-degree angle from the vertical. However. This test measured at the same locations doors. dard test chamber. Experimental Condensation Cycle ing zinc hydroxyl phosphate and iron (Wet/Dry Cycle) oxide pigments.6 panels were exposed inside a stan- than cyclic wet/ dry immersion and mm) used for this study were pre. tween continuous wet and cyclic ex. the tions and they create failure modes used a higher concentration of so. and and topcoat) making up the different ronment. were measured. degradation such as a wet/ dry cycle 23°C and 45% relative humidity. The tinuous wet test (for instance.. coated mersion was a more severe exposure The test panels (150 by IDD by 1. or a cyclic wet/ dry envi. vice environment similar to that en. thickness of the coated panels was similar to those observed out. scribed elsewhere? The present test days. 12 The surface of each panel was sand panels were sprayed indirectly with Nevertheless. the salt of salt spray. coating systems (Table 1) were spray cyclic exposures provide more realis. ultraviolet (UV)-light individual paints (primer. and the coating thickness tic results because they more closely this study is a variation on that de. Pane! Preparation In wet/dry cycle tests. for the present blasted to Swedish standard SA an atomized salt spray based on 5 work. primer and undercoilt thicknesses that a relationship does not exist be. continuous wet im.l 50 11 Zinc epoxy 50 Epoxy polyamide (high build) 125 225 5 to 7 Polyurethane 50 Most accelerated aging tests for of factors of importance in coatings and ethanol. COATINGS & LININGS TABLE 1 Coating Systems Studied Dry Film Thickness life Expectancy Under Severe (flm) Paint Conditions System Generic Type Individual Tolal (years) Inorganic zinc silicate 75 epoxy polyamide 25 135 5 to 7 Acrylic enamel 35 2 Zinc phosplla.50 Epoxy polYainide·~ealer 25 Epoxy polyamide (high bui Id) 1Rq> 10 to 12 Polyurethi'!rJl. supported at a that the failure mechanism in both pared from a hot-rolled carbon steel. spray test).

tde 8 (0. and frequency ratings are on a scale 10 to 0. cycle were incorporated into the wet/ densation cycle in test apparatus The tests panels were inspected dry salt spray cycle corrosion test. was 5. 4. (AI Rust grade rating is on a scale 10 to 0. 0 = dense.000 ppm The spraying rate of the salt mist hours of radiation at 60°C followed sulfate. where 10= none. 16 The conditions used were: 12 dence of coating breakdown every 26 [VIP/December 1994 .4 mL/hour. checking Rustgra9Els. trolyte when located around the The coated panels were exposed rosion in an industrial atmos. followed exposure of the coated total of 3. and all coated pan. MD = medium dense. as operating to ASTM G 53-84 specifica.yeIl6wing. One of each du- UV-light radiation and condensation panels to a UV -light radiation/ con. TABLE 3 Performance Data and Defect Rating for Atmospheric Coating Systems Exposed to Salt Spray/UV Radiation/Condensation Cycle (Wet/Dry Cycle) (AIRust visually and microscopically for evi- by other workers. M = medium. chFLkj~g.7 tions. F = few.000 hours. COATINGS & LININGS TABLE :2 Performance Data and Defect Rating for Atmospheric Coating Systems Exposed to Continuous UV Radiation/Condensation Cycle Rust. reported to be unreliable tests for cor.H.yellowing Slight ch~lkinll:i~i[1g . where 10= none.3%). The wet/ to the wet/dry cycle corrosion test phere.. potable water containing 3. then cycled through has been attributed to the absence of 35 ± 2°C followed by drying for 16 the UV / condensation test chamber ammonium and sulfate species as hours under ambient conditions) was for 200 hours on a rotating basis for a well as the effects of wet/ dry cycling. by 12 hours of condensation (using dium chloride spray alone have been els received a similar quantity of elec. deionized water) at 40°C.15 The observed unreliability dry cycle (l00 hours of salt spray at for 200 hours. Salt spray tests based on so. blister. plica te test panel was cross-scribed. periphery of the chamber.

Site E ( 15 km from sea)-Mild pol. els 'Nere rated in accordance Such changes in title surface diatbn. . The fest panels were in. and general leaving crazing 8117 This test procedure is intended pearance. after 3. The location of five test sites are various coating depended the alkyds and gloss) showed as follows: fundamentally on the test conditions a relatively decline in gloss. . The polyurethane ap- at five test sites.12 Continuous UV at 40°C for 180 moni:h for defects such as dccsation cycling degraded the to ASTM D 336"1.000 v\'cre also measured. and emanated from the scribe. exhibited significant degradati. and types of defects produced by the ing visible on the surfaces. 19 This Acceffer:@1'ed Te~ii::s may allow moisture retention The performance of the various by the and lead to rusting of the Triplicate test panels coated with coating systems exposed under ac. un- specifications. Site D (3 km froIn the sea)-Mild in the test centered on tion/ condensation in a weathering pollution area). chalking.5 km from the the lvvo accelerated test Cycle pollution (upwind fYO:TI appear to be very The continuous wetting test in. Coated panels were lution (rural area). as was evidence of blisters and chlor-soda [chlorine. COATII\iGS & LINII\IGS Point System Point System o 1 @ 7 8 [ll 9 10 o 11 130 70 SO 90 Time (Weeks) I'IGUI'!IE 1 Coatings at test site A.2 km from the sca). to which the coatings were subjected. and chamber operating to ASTM G 53-84 .uld the gloss retention obvJc).':es of these U\I-Light Radiation/Condensation Site C (1. dium-oil in the wetl dry cyclic 12 hours of condensation (using The were examined c~\'ej('y S.. 200 hCiUrs. each of the 1 J systems were exposed celerilted test conditions are surnrno. . hours of exposure. substrate.s result (Table 2) recluc- to simulate deterioration caused by measured at a 60-degree angle. The diffeHT. Gloss values and weight pollution (next to of the test cycle. Similar exposed to 12 hours of illumination The exposure at the time results have been recorded for a me- under UV light at 60°C followed by of this report was two and a half years.J.on in suggest that the barrier o Site B km from the sea).. ment to which it is exposed.bse- spected and rated for performance as quent interactions with the environ- Results ~rrlld DisclL~ssion described earlier. chalking and yellowing were ob- Heavy pollution (downwind from The alkyd (enamel and gloss) and served. was marked by corrosion creep. Site A (0. uf this coating SYSlClfl bccarnc MP/December 1994 27 . However. in the reveal that the modes of degradation no blisters and very little or no rust- Shuaiba industrial area. der-rusting. in racks facing south rized in Tables 2 and 3. and protective properties. chbcinated rubber coating and rusting under the These sodium hydroxide] plant). next to cement clinker the degradation on volved tile exposure of the coated (enamel and gloss panels to continuous UV light radia.. re.. and rust creep. The tables peared to be in good condition with at an angle of 45 degrees. Pan. 18 condition of paint can cause su. ASTM D 6'10-43. gloss and yelloi/olil'g of dew and the effect of natural UV ra. b.

gloss retention observed for alkyd wet/ dry salt spray /UV radiation/ sure.19 the test In contrast. vironment parameters of the indus- grade and frequency) observed for istic deterioration effect. COATINGS & LININGS FIGURE 2 Coatings at test site B. and E (Figures 1 through 4). FIGURE 3 Coatings at test site D. as well as distances from the Arabian Gulf were paint systems in natural atmospheric loss of gloss. In concurrence with reported From Table 5. with some trial area (Table 4) and at varyious the alkyd and chlorinated rubber rusting in the scribe lines. ep- nation did not appear to be closely oxy micaceous iron oxide. produced by the three-factor combi. The results Test site C is located in the vicinity of tion/ condensation environment. 10. (enamel and gloss) and chlorinated condensation environments. The observations for coatings exposed to polyurethane (systems 5. and acrylic Outdoor Exposure representative of the degradation (in water-based systems. a refinery. 6.000 hours of exposure appears to produce results that are spray exposure to the UV / condensa. the wet/dry Five sites with the different en- terms of gloss retention and rust spot cycle exposure test gave a more real. For the polyurethanes. it can be seen that the B. D. from tests in a continuous UV radia- degradation processes. similar to that observed for the paint degradation processes than results tered the nature of the corrosion/ systems exposed at site C (Figure 5). selected for outdoor panel exposures. ineffective after 500 hours of expo. 7. the addition of salt rubber after 3. in the accelerated tests were quite more comparable to natural exposure tion test (Table 3) fundamentally al. and 28 MP/December 1994 . service environments at test sites A.

FIGURES Coatings at test site C. and 11) and micaceous. 7. based paints. COATINGS & LININGS Point System Point System • . 5. ber [4]) exposed at the test sites. 20 Also. sual observations of the coated test sites.10. It has been reported that in oil- ering. 10.8 :II. as a function of time curves (Figures systems at sites D and E and grade 7 tem 3) and acrylic water-based (sys. a comparison of gloss retention and epoxy-micaceous-iron oxide with age. the other paint systems. 3 . even in the iron-oxide epoxy system (3) showed The chlorinated rubber (system heavily polluted environment of site that these paints retained their gloss 4) and the alkyd gloss (system 9) coat- B. In gen. tween grade 10 for the polyurethane ments making the film become brittle eral. Vi. (fine cracks which did not penetrate ated laboratory tests. andB. at all test sites. 6. This trend in performance was also better than the other coating systems ing systems showed checking defects apparent in the results of the acceler. • 4 1119 10 FIGURE 4 Coatings at test site E. most coatings failed to retain panels after two and half years of ex.2 • 7' . 1 to 5) for the polyurethane systems for most of the systems at sites A tem 8) paint systems outperformed (5.. ex. water and aging can gradu- ceptions were systems 3. in chlorinated MP/December 1994 29 . 7. the topcoat) after exposure at all test In the course of outdoor weath. ally leach out the plasticizing ele- 11 exposed at sites D and E. and coated test panels to be limited to be. such as the alkyd gloss their gloss values. 11) epoxy-micaceous-iron oxide (sys.. posure detected rust spots on the system. (acrylics [2 and 9] and chlorinated rub.

Chalki'~g is of failure because surface prepara. chalking is collectE':I for the five exposure sites compared considered the most acceptable form are in Table 4. is ing tested have provided cal'. with the slightly higher rela- changed for long periods. for epoxies C. 0 = dense. COATINGS 8. Furthermore. . phenomenon only.O. conditions at sites A cmd B appear to the formation of a friable. These two systems proved un. = not detected. nia.l caS2S. Fur- sequence of exposure to the photo. before rnc:intenance is than those observed for the other sites. The exposure the other systens tested. are not expected to survive more than thermore. tern 3) react quite rapidly. In terms of corrosion protection seashore. . where 10= no rust spots. powdery tion for subsequent maintenance gen. LININGS [i. /ip~rticulate ()lg~m"): •. migrate under the influence of generally considered to be a surface varying protection levels depending heat. . of sulfur ammo- C:1Used by of the binder loose material. TABLE is degrees) Retention of Coating Systems After Two Years' Exposure at the Five Test Sites ° (AIDefect rating is on a scale of 1 to 0. Further. F = few. the alkyd systems oxide species in the atmosphere. M = medium. age (1988 to 1989) atmospheric data the suffered In aJ. The nearness of sites A and B to the and polyurethanes can remain un. the concentrations of am- chemically active rays of the sun and one to three years under severe at.end nitrogen by weathering. be the most severe.. ance. ••. ditions of the site. monium and chloride ions in the cD:l1Ciensate from dew. of the underlying steel. would result in a 30 MP/Oecember 1994 . N/A = not available. all the coat. and therefore is on the predominant atmospheric con- suitable for outdoor application not harmfut except to the appear. sulfide. tive humidity. The annual aver- under these test conditions. mospheric conditions such as those particulates in the site A Different binders ::-eact at differ.BlE 4 Annual Averages of Atmospheric Parameters Obtained at the Five Test Sites Su Ifur dioxide (ppm) Ammonia Total sulfur(ppm) !ilYc:lrogen sullide(ppm) Nijrjc oxides (pPrtll Tetal suspended .•. Ammonia (inparticulale)· Chlorine (in particulate) flir wind speed (m/s) Horizontal Vertical i\J. acrylics necessary.elle plasticizer elemer·Is of epoxies. MD = medium dense. with higher con- on the of a paint erally consists o'11y of the centratiom. particularly as a con. at test sii:2s and to S0111e extent and B atmospheres are much higher ent rates. rubber paints..

F. performed better (both aesthetically and in corrosion resistance) at all test More infonnation may be available in paper no. Moore. J. 15 expected. Coatings Techno!. Lacquer.341. M. 1985). 2 (1981): p. E.748 (1987): p.K. Lindgren. Hempels Marine Paints. 5. PA: ASTM).I. panel surfaces.. 17. PA: ASTM). Prog. 89. Coatings Techno!. 138.N. Am. 7. and 6. 23 (1983): p. Martin. The polyurethane and epoxy. the dust showed the presence of 38% j. D. Soc. and GTC Paints. The conventional coating systems based on alkyd. Maryland. 20. Ali. Corros. B.W. 76. sig.I.G. and chlorinated rubber performed mod- erately compared with the polyure- thane systems.A Chandier.B.A Bayliss.J. p. Scantlcbury. J.B. Rhodes. 1. 6. 60. B. 4.W. F. Crowder. 417. J. Tickle. SIS 05 5900-1967 nificant amounts of clinker dust have (Stockholm. 45. ASTM D 610-43. PA:ASTM). Analysis of 571. J. IX: NACE.s. 61. Techno!. McKnight. Org. 445. conditions prevalent in Kuwaiti in. and Water-Exposure Apparatus (Unfiltered Car- ing to protect a steel substrate under bon-Arc Type) for Testing Paint. AG. Nguyen. OCCA 48 (1965): p. 59. Corrosion Control by Organic Coatings. to a salt spray /UV radiation/ conden. Durability of Building Materials 3 (1985):p. Lcidheiser (Houston. 97. 2.c. J. posed at site C was faster than 11. 9. Corrosion Protec- tion of Steel Structures (Essex. Systems topcoated with polyure- thane offer extremely good corrosion protection for steel. flected by gloss loss. Campbell. Skerry. OCCA 62 (1979): p. Sci. Although they are expensive initially. acrylic. 76'5 (1988): p. "Practice for Operating Light- sulfate. 97.S.55% 16. COATINGS & LININGS prolonged time of wetness on the Techno!. England: Elsevier micaceous-iron oxide-based systems Applied Science Publishers. "Practice for Operating Light- The projected ability of a coat. BS Skerry. 39. K. I. 269. J.A Ali. The authors acknowledge the Shuaiba Area Authority and the Ku- wait Institute for Scientific Research.E. the nearby cement clinker factory. 12. J. This can be attributed to (1987): p. 54.M. Timmins. Ho.D. A Alavi.A Simms. ASTM G 53-84. Proc. 317. 39. 10. Kendig. K. J.L Lindgren. 4 (1979): p. Sweden: Swedish National Standards Institute). Degradation. J. H. 19. of coatings ex.A. PJ®~®rrences 1.K Appleman. 772 (1989): p. been removed periodically from the 14. 1. calcium 20% silica. Thanks are also expressed to Interna- tional Paints. 65 (1988): p. and Water-Exposure Apparatus (Fluorescent Lv- Condensation Type) for Exposure of Nonmetallic C~u11C~usions Materials" (Philadelphia. Polymer Mate- rials Science Engineering 58 (1988): p. ed. Div. 3. Swedish Standard 2 1/2. Skerry. M.A Eden. Barraclough. Harrison. Construction 6. "Method for Evaluating De- cated in the laboratory by exposure gree of Rusting on Painted Steel Surfaces" (Phila- delphia. OCCA 62. Smith. Over the two years of exposure. Harrison. and Related Products Using the Dew Cycle" (Phila- delphia. H. K. 8. AST:v1 3361-81. OCCA 45 (1962): surface of the coatings. Coatings MP/December 1994 CirGI . P. as re. presented at CORROSlON/94. locations in the Shuaiba area than the Baltimore. 13. 115. K. 686 (1982): p. Mansfeld. 18. other coating systems studied. B. A Alavi. M. 1311. Coatings Techno!. 1981). Varnish. Coat. J.R. M. D. their effective life is long. J. 17. p. K. dustrial areas can be largely dupli. Coatings sation test environment. 60. E. J. J. Tsai. D. Embree.