Protecting Secondary Containment from Pesticides and Fertilizer · 2016. 10. 24. · Protecting Secondary Containment from Pesticides and Fertilizer C by Duy T. Nguyen, Michael F

Copyright ©1995, Technology Publishing Company54 / Journal of Protective Coatings & Linings

oncrete has been the material ofchoice for building secondarycontainment structures because it is relatively inexpensive and has structural properties that make it

ideal for supporting the loads of vehicles and large tanks. The chemical properties of concrete, however, male it susceptible to corrosion by common fertilizers. Though fairly impervious to water penetration, con-crete is easily penetrated by vapors and or-ganic solvents. This article describes re-search requested by the Environmental Protection Agency’s (EPA) Office of Pesti-cide Program to help identify coatings for concrete that could reduce the migration of pesticides through containment structures. EPA’s concern with concrete is due in part to the poor performance of drip pads at some lumber treating facilities. At these sites, wood preserving chemicals have passed through concrete pads and contami-nated underlying soil and ground water. A preliminary study1 revealed that herbicides penetrated concrete samples after 5 simu-lated spills despite being rinsed shortly after each spill.

Editor’s Note: This article is based on a paper given at SSPC 94, held November 11- 17, 1994, in Atlanta, GA, and published in the proceedings entitled Managing Costs and Risks for Effective and Durable Protec-tion, SSPC 94-19.

Protecting SecondaryContainment from

Pesticides and Fertilizer

Cby Duy T. Nguyen, Michael F. Broder,

Cathy L. McDonald, andd Eugene A. Zarate,TVA Enviromenntal Research Center

Some states require concrete sec-ondary containment to be sealed or pro-tected with a coating when exposed to fer-tilizers and pesticides. Many retailers haveused penetrating sealants to satisfy theselaws, even though sealers provide littleprotection from chemical attack, and theydo not prevent pesticides from penetratingthe surface. Other retailers, who have ap-plied thick film coatings that were properlyselected for the intended service, have haddisastrous results because the coatingswere applied under poor ambient conditions. Consequently, there is muchskepticism regarding the performance andbenefit of protective coatings. The primaryobjective of the study described here wasto determine the performance of concreteand organic coatings in containment struc-tures designed for bulk agricultural pesti-cides. Bulk products are a fraction of thepesticides in common use for agriculturalpurposes; most are herbicides. The coatingstested in this project were evaluated forcost, ease of application, and performance.

Selecting Test Materials

Resource constraints made it impossible totest the hundreds of available commercialcoating products and bulk pesticides. Totest many coatings and pesticides, a two-tiered test program was used. Screening

Copyright ©1995, Technology Publishing Company

Testing Coatings for Pesticides

JULY 1995 / 55

Fig. 1 - An exampleof the dimensionalexpansion of a freefilm after 21 days ofexposure to a solventbased pesticide. Photos courtesy ofTVA EnvironmentalResearch Center

tests identified products for more extensivetesting. Pesticides that were nonreactive with many coatings were eliminated fromsubsequent tests, as were coatings that were sensitive to chemical attack from sev-eral pesticides.

In the first series of tests, 12 coatingswere screened in 10 pesticides and urea-ammonium nitrate (UAN) fertilizer solution(32-0-0 UAN grade). Free films of coatingswere used in screening tests for 2 reasons:first, coatings could be evaluated with ex-posures as short as 21 days, and second,physical properties of coatings were easier to measure in free films than on concrete.

In the second series of tests, the 4 most corrosion-resistant coatings, 3 most corrosive pesticides, and the UAN solutionwere tested. Concrete specimens were coated to determine the ability of coatings to protect the concrete in test chemicals.

Pesticides were selected from a list of27 of the most common bulk products usedin the U.S. Of these, 13 were combinationsthat included a product selected or weresimilar to a product selected. Ten of the 27products were selected by eliminating aqueous solutions, mixtures, and productswith duplicate chemistry (Table 1). Someaqueous solutions were omitted becausethey are less reactive than solvent-borne materials. Atrazine was selected because itis the most common pesticide in groundwater. An insecticide, dimethoate, was added at EPA’s request. Dimethoate was chosen because it easily penetrates and softens certain plastics.

Coatings were selected from a list ofcandidates supplied by manufacturers thatmarket products for coating concrete at re-tail agri-chemical facilities of 20 companiescontacted. Most offered 1 or 2 products toprotect concrete from immersion in pesti-cides containing organic solvents such asxylene and chlorobenzene. At least 1 of each of the generic coatings commonly used in secondary containment was tested:

urethane, epoxy phenolic, polysulfide, ureaurethane, vinyl ester, and epoxy novolac.Table 2 lists coatings, resin cost per area,and minimum dry film thickness (DFT).

Test Procedures

Screening TestsThe test procedures, including fabrication of specimens and pre- and post-test evalua-tions, were designed to conform as closelyas possible to 5 ASTM standards2: D 823, D1005, D 3924, D 2370, and D 3363.

Specimen PreparationFabrication of the test specimens was basedon ASTM D 823. The coatings were pre-pared in accordance with the manufactur-ers’ instructions. After thorough mixing andstabilizing, the coating mixture was pouredon a thin, flat sheet coated with polyte-trafluoroethylene and spread uniformly andevenly using a rod coated with the samematerial. Since most of the coatings werehigh solids formulations, only 1 coat was required to obtain the desired thickness.

After drying for a minimum of 94hours, the test specimens were cut from thesheets using a double-blade precision spec-imen cutter or a cutting tool made for acry-late resin. Special care was taken to insurethat no chipping, shearing, or other defectswere introduced. Each specimen was 1 in.



56 / Journal of Protective Coatings & Linings

x 6 in. (25 mm x 150 mm). The DFT of thespecimens was measured using a digitalcoating thickness gage if coating thicknesswas less than 50 mils (1.3 mm), or a digitalcaliper if coating thickness was 50 mils (1.3mm) or greater. The average DFT was de-termined from at least 3 measurementsacross the length of the specimen.

Following thickness measurementsand visual examination, the test specimenswere washed in soapy water for 1-2 min-utes to remove dirt and oils, and then de-greased in alcohol for 2-3 minutes. After the specimens were allowed to stabilize forat least 4 hours, the mass of the test speci-mens was determined to within 0.1 mgusing a calibrated digital balance. The datawere recorded in a computer database.

For each coating, 1 specimen was se-lected for pencil hardness tests in accor-dance with ASTM D 3363, and 5 were se-lected for tensile testing.

Tensile tests were conducted using 1in. x 2 in. (25 mm x 50 mm) specimens.The tests were performed on a computer-controlled universal test machine. Testspeeds were based on the rigidity of thecoating and ranged from 5 to 55 percentelongation per minute. Test data, includingultimate tensile strength, percent of totalelongation, and elastic or Young modulus,

were determined by averaging the results of the 5 test specimens.

Test solutions were full strength for-mulations of pesticides purchased from alocal agri-chemical retailer. Table 1 listscommon names of these chemicals. Urea-ammonium nitrate fertilizer (32-0-0 UAN)was prepared using 35 weight percent commercial granular urea, 45 weight per-cent commercial granular ammonium ni-trate, and 20 weight percent tap water.

Six specimens of each coating wereimmersed in approximately 1 L (0.95 qt) ofeach test solution. The test was conductedfor 21 days without agitation. The test spec-imens were observed every 24 hours. After21 days, the test specimens were removed,cleaned, weighed once more, and exam-ined for pattern deterioration.

Post-Test EvaluationCoating evaluations were based on com-parison of the data obtained before andafter the test. Included were changes incolor and appearance, expansion, soften-ing, swelling, chemical attack, aging, per-centage of weight gained or lost, and dete-rioration of mechanical properties, such asultimate tensile strength, percent of elonga-tion, elastic modulus, and hardness.

The volumetric absorption rate (VR)was determined by the weight change perunit of volume. The weight change consist-ed of the absorption of the test chemicalinto the body of the test specimen per unitof time or a weight loss due to the coatingdeterioration per unit of time. This quantitywas calculated using the following equa-tion:

VR = 365 * (Mf – Mi) / (t * v) where VR is volumetric absorption or dete-rioration rate in mg/cm3/yr; Mi is initialweight of test specimen in mg; Mf is finalweight of the test specimen in mg; v is volume of test specimen in cm3; and t istotal exposure time in days.

The percent of weight gain clue to ab-



JULY 1995 / 57

sorption of corrosive species or loss due tocoating deterioration was calculated usingthe following equation.

P = (Mf - Mi) * 100 / Mi A positive value for P indicates a weightgain; a negative P indicates a weight loss.

Evaluation CriteriaASTM standards2 offer several guidelinesfor evaluating the deterioration of organiccoatings exposed to corrosive environ-ments. Among these are evaluations of blis-tering (ASTM D 714), rusting (ASTM D610), and cracking (ASTM D 661). Howev-er, these standards and criteria apply onlyto coatings on steel panels.

NACE International has publishedguidelines on evaluating elastomer materi-als in sour gas environments (NACE,1991).3 Similarly, ASTM2 has publishedstandards for evaluating the resistance ofplastic to chemical reagents (ASTM D 543)and for evaluating rubber in immersiontests (ASTM D 471).

Papers by Fisher4 and Fisher and Car-penter5 include criteria for determining theacceptability of lining materials for immer-sion service. These papers state that an ac-ceptable lining neither gains more than 10percent by weight or volume nor suffersany weight loss after a 30-day exposure.Ideally, the lining should sustain no surfaceattack, embrittlement, softening, blistering,or color change. The basic criteria for as-sessing a coating’s suitability for pesticidesecondary containment were selected fromboth ASTM and Fisher and Carpenter. How-ever, a slight modification of these criteriawas needed because the test duration wasonly 21 days instead of 30 days. Despitebeing designed for immersion service,these weight change criteria were chosenfor these evaluations because most pesti-cide secondary containment floors are ex-posed continuously to spilled materials.Most secondary containment floors arelevel (poorly drained), and spills are not al-

ways cleaned up immediately. Thus, in thisseries of screening tests, the test coatingswere classified by the following criteria.• Coatings are considered satisfactory forhandling chemical media if weight gain isless than 7 percent or weight loss is in-significant in 21-day exposure tests. (A value of less than 2 percent was consideredinsignificant in the test program.)• Coatings are considered satisfactory forhandling chemical media if there is no sig-nificant change in ultimate tensile strength(SUTS), percentage of elongation (e), elasticor Young modulus (E), or pencil hardness.The changes in SUTS, e , and E must havelittle effect on the mechanical performanceof the coating during service.• Coatings are considered satisfactory forhandling chemical media if there is no sur-face attack, softening or swelling, cracking




or embrittlement, orcolor change testing.

Coated Concrete TestsConcrete SpecimenFabrication Concrete specimenswere cylindrical inshape (2 in. [50 mm] indiameter and 4 in. [100

mm] high) and were fabricated in accor-dance with ASTM C 192-90.6 Type I Port-land cement was selected for testing. Theconcrete mix was designed to obtain an av-erage compressive strength of 3,000 psi (200 MPa), a slump of 3 in. (75 mm), an aircontent of 3 percent by volume, and awater-to-cement ratio of 0.5. Cement, sandaggregate, and tap water were mixed for 10minutes. Slump was then checked and ad-justed to approximately 3 in. (75 mm) byusing water reducer and super plasticizer.Mixing continued for an additional 10 min-utes. Test specimens were prepared by fill-ing plastic molds with the concrete mix, vi-brating for 1 minute, curing 28 days beforeremoval, and abrasive blasting. Surface de-fects and cracks were patched using a ready-mix vinyl concrete patching material.

Coating ApplicationSurfaces of the concrete specimens wereprepared in accordance with ASTM C 811-907 and NACE RP0376-76.8 Surfaces wereabrasive blasted using 20/40 grit sand,etched and scrubbed using 20 percent byvolume HCl solution, rinsed, and dried. Test specimens were stored at laboratory conditions (80 F [27 C] and 50 percent rela-tive humidity) before coating application.

Coatings were formulated and mixedaccording to manufacturers’ recommenda-tions. All coatings were brush applied (ex-

cept for parts of the polysiloxane system)to a desired DFT and cured for 4 days be-fore exposure. For the epoxy phenoliccoatings, the concrete surface was coatedwith 2 coats of 100 percent solids epoxy-blend filler sealer to seal pores and cracks.Then, 4 coats of epoxy phenolic topcoatwere applied to obtain 25-mil (635-micron)minimum DFT. The epoxy novolac coatingsystem consisted of 3 coats: a primer coat,a base coat, and a topcoat. Total DFT wasapproximately 25 mils (635 microns). Thevinyl ester resin coating system consisted of3 coats: a primer coating applied using abrush, a base coat containing aggregatefiller applied using a putty knife, and a top-coat applied using a putty knife. Total DFTwas approximately 35 mils (889 microns).

Chemical ExposureEach specimen was immersed in 500milliliters (17 fluid ounces) of the testchemical. Six specimens per chemical wereused for each type of coating. Test durationwas 90 days. Upon completion, test speci-mens were removed, cleaned, dried, pho-tographed, and examined for evidence ofdeterioration. Three specimens were sub-jected to compressive tests (ASTM C 873-939), and the other 3 were exposed tofreeze/thaw cycle tests (ASTM C 666-9210)for 7 days at temperatures ranging from 20to 55 F (-7 to 13 C) and approximately 7cycles per day before compressive testing.For control purposes, uncoated specimenswere also subjected to chemical exposure,freeze/thaw cycling, and compressive tests.

Abrasion TestsAbrasion tests were performed only on top-coats in accordance with ASTM D 4060-9011 using a digital abrader. Coating sam-ples for abrasion testing were prepared byspreading the coating on a sheet coatedwith polytetrafluoroethylene, and cuttingfour-inch (100-millimeter) square panels.After a four-day cure, panels were exposed

Fig.2 - (left)Coated concrete

specimens prior to90-day immersion.

Fig. 3 - (right)Yellow color

illustrates the depthof penetration of

imazaquin,monochlorobenzene,

and pendimethalinmixture after

90-day exposure.



JULY 1995 / 59

to test chemicals for 3 months and thentested for abrasion. Abrasion resistance wasdetermined using weight loss after 500 cy-cles under a 1000-gram (2.2-pound) load.Abrasion index (AI) is presented inmg/1000 cycles and calculated using thefollowing formula:

AI (%) = (Mi – Mf) * 103/Cwhere Mi is initial weight in mg, Mf is finalweight in mg, and C is the number of testcycles.

Evaluation and Performance CriteriaCoating performance and chemical effectson concrete were evaluated based on visu-al examination, in accordance with ASTM D6608712, D 661-8613, and D 4449-90.14

Performance and effects were also evaluat-ed on the basis of AI (ASTM D 4060-90)and compressive strength of test speci-mens. Degree of coating deterioration wasdetermined by comparing coated speci-mens subjected to chemical exposure tospecimens that were not exposed. For acoating to provide acceptable protection,the following criteria must be satisfied.• Coating must be impermeable to testchemicals with no evidence of chemicalpenetration into concrete after prolongedexposure. It must not show signs of deteri-oration, including chalking, swelling, de-lamination, flaking, and expansion.• Compressive strength of exposed coatedconcrete must retain at least 80 percent ofthe average value of uncoated and unex-posed concrete.• For abrasion service, AI of the test coat-ing must be less than 150 mg/1000 cyclesafter chemical exposure.15

Results

Screening TestsValues for mechanical properties and aver-age DFT of coatings before chemical expo-sure are shown in Table 3. Percentage of

weight gain or loss, volumetric rate of ab-sorption (VR), and mechanical properties ofcoating after exposure to atrazine anddimethoate are listed in Tables 4 and 5. Data for the other pesticides tested are in the original SSPC paper (SSPC 94-19).

Urethane ResinBefore chemical exposure, the urethane had a pencil hardness of approximately 5B(Table 3). After the 21-day exposure to pes-ticides, tests on the urethane coating produced mixed results. In atrazine, anaqueous solution, urethane showed no sig-nificant weight gain or loss (less than 1 per-cent). In bentazon, the coating weight in-creased slightly (approximately 2-3 percent). However, the weight gain of thecoating was greater than 8 percent when exposed to the solvent-borne pesticides.Also, severe weight gains of 25, 27, and 103 percent occurred when the coating wasexposed to clomazone, rnetolachlor, anddimethoate, respectively.

The visual examination showed thatthe coating failed in most test solutions dueto softening, dimensional expansion andcontraction, and cracking. The coating wasdissolved or converted back to the liquidstate when exposed to dimethoate. Due tounreliable mechanical test results, no post-test evaluations were conducted based onthe ultimate tensile strength, total elonga-tion, or elastic modulus.

Epoxy Phenolic ResinThe mechanical properties of the epoxyphenolic before exposure are given in Table 3. In addition to good mechanicalproperties, the epoxy phenolic coating ex-hibited good chemical resistance. Exposureto the pesticides did not damage the coat-ing. The coating had a slight weight loss(less than 2 percent) after the 21-day expo-sure to atrazine, bentazon, clomazone,metolachlor, alachlor, and trifluralin. How-ever, a slight weight gain was observed




when the coating was exposed topendimethalin, imazaquin+monochloroben-zene+pendimethalin (IMP), and dimethoate. The average weight gain inpendimethalin and IMP was about 2 per-cent, and in dimethoate the weight gain was about 6 to 8 percent.

Inspection of the coating’s mechanicalproperties after exposure to atrazine andbentazon indicated that the pencil hardnessof the coating decreased only 1 unit of hardness, from 3H to 2H. In clomazone, the coating hardness decreased about 2units, from 3H to 1H. The hardness in-creased 1 unit in alachlor, trifluralin, anddimethoate, but was unchanged in meto-lachlor, EPTC-thiocarbamate, and IMP.

Contrary to the hardness test results,

the results of the tensile tests showed thatthe coating aged significantly when ex-posed to most of the test chemicals, withthe exception of dimethoate. The SUTS in-creased 21 to 37 percent in atrazine; 42 to58 percent in bentazon; 57 to 66 percent inclomazone; 59 to 65 percent in meto-lachlor; and more than 100 percent inEPTC-thiocarbamate, alachlor, IMP, and tri-fluralin. The total elongation decreased 50to 57 percent in atrazine, 21 to 43 percentin bentazon, 22 to 57 percent in cloma-zone, up to 29 percent in rnetolachlor, 21percent in EPTC-thiocarbamate, 7 percentin alachlor, and 21 percent in trifluralin.Total elongation increased inpendimethalin, IMP, and dimethoate by 29,14, and 43 percent, respectively. In



JULY 1995 / 61

dimethoate, the coating was softened. TheSUTS decreased 36 to 41 percent.

Visual examinations showed no signof deterioration on any test specimens. Aslight discoloration was observed on thespecimens exposed to pendimethalin andIMP. However, this was attributed to thecolor of chemical reaction between the testchemical and the coating.

Polysulfide ResinResults of mechanical tests performed be-fore chemical exposure indicated that thepolysulfide was strong, very elastic, andsoft (Table 3). After exposure to most testsolutions, test results indicated unsatisfacto-ry performance. The weight change wasless than the 7 percent criterion in atrazineand bentazon. However, the coating weightincreased significantly in EPTC-thiocarba-mate (21 to 22 percent) and trifluralin(about 11 percent) and increased greatly in

clomazone (51 to 64 percent), metolachlor(82 to 88 percent), alachlor (55 to 59 per-cent), pendimethalin (42 to 49 percent),and IMP (51 to 58 percent). In dimethoate,all test specimens failed completely due toswelling and softening and were not recov-ered. Visual examination showed no discol-oration of the test specimens. Expansionand swelling due to the accumulation ofsolvent were the primary causes of coatingfailure. Due to the severe weight change ofthe coating specimens, no post-test me-chanical evaluations were performed.

Urea Urethane ResinThe urea urethane is soft, but strong andelastic. The coating had an SUTS similar tothat of epoxy phenolic but was approxi-mately 4 to 5 times more elastic (Table 3).Similar to the urethane and polysulfidecoatings, urea urethane did not deterioratewhen exposed to water-borne herbicides,




atrazine and bentazon, but lacked resis-tance to solvent-borne products. Coatingweight increased only 1 percent or less inatrazine and bentazon, approximately 5-8percent in metolachlor, and 3-5 percent intrifluralin. However, the weight gain of thecoating exceeded the 7 percent criterion inthe other pesticides.

Visual inspection showed no sign ofchemical attack on specimens exposed toatrazine and bentazon. Also, no dimension-al changes were observed. These speci-mens were in excellent condition after the21-day exposure. However, in solvent-borne materials, urea urethane failed dueto softening and expansion caused by ab-sorption of the chemicals. The coating alsofailed by cracking when exposed to cloma-zone and dimethoate. Due to the severeweight change of the specimens, no me-chanical evaluations were performed afterchemical exposure.

Vinyl Ester ResinThe vinyl ester showed excellent resistanceto all test chemicals. The weight changewas minimal in all test conditions. Indimethoate, the most severe chemical, thecoating weight increased only 0.4 percent.The greatest weight gain was caused byIMP, but it was less than 2 percent. Allweight losses were less than 1 percent. Me-chanical examination before chemical ex-posure revealed that the vinyl ester resintested is a strong but brittle coating (Table3). Hardness tests performed after chemicalexposure showed only minor changes inthe hardness values. The coating hardnessdecreased slightly in metolachlor anddimethoate to 5H and 3H, respectively. Re-sults of tension tests showed slight changesis SUTS and elastic modulus, but not in thetotal elongation. The total elongation of thecoating (e) was still about 1 percent afterexposure to most of the chemicals. In



JULY 1995 / 63

atrazine and alachlor, the SUTS ranged from5,000 to 5,400 psi (35 to 37 MPa), corre-sponding to increases of 4 to 13 percent. Inother test chemicals, the SUTS decreased 1-30 percent. However, in trifluralin, the SUTS measured was over 3,500 psi (24.1MPa). The elastic modulus of the coatingdecreased in all test solutions. The greatestreduction (43-50 percent) was observed inpendimethalin. In other test chemicals, thereduction of the elastic modulus rangedfrom 8-42 percent.

The post-test visual examination ofthe specimens revealed no signs of chemi-cal attack. A slight discoloration, similar tothat found on the epoxy phenolic resin,was observed on the specimens exposed topendimethalin and trifluralin. However, itwas caused by the color of chemicalresidues retained on the specimen surface,not by any chemical reactions between thetest chemical and the coating.

Epoxy Novolac ResinSimilar to vinyl ester, epoxy novolac wasstrong, but inelastic (Table 3). Upon expo-sure to pesticides, epoxy novolac sufferedno obvious effects of chemical attack. Theweight gain or loss in each case was lessthan 2 percent and therefore not significant.

Mechanical evaluation indicated nochange in coating hardness in any of thetest chemicals, except dimethoate, in whichthe hardness decreased to 3H. No signifi-cant change was observed in total elonga-tion. The SUTS changed slightly in most testchemicals, but still ranged from 3500-4500psi (24 to 31 MPa). The elastic modulus de-creased in most test chemicals, but in-creased slightly in trifluralin. Decreasesranged from 5 to 43 percent; increasesranged from 8 to 29 percent.

Visual examination showed no evi-dence of chemical attack or deteriorationafter the 21-day exposure. A slight discol-oration was found only on the specimensexposed to pendimethalin and IMP.

100 Percent Solids Polyamide EpoxyThe 100 percent solids polyamide-curedepoxy had a milky color and was strongand brittle (Table 3). After the 21-day expo-sure, atrazine had little effect on the coat-ing films. Weight gains were 1 percent orless. No discoloration, cracking, or swellingwas observed on the test specimens.

Exposure to solvent-borne herbicidesproduced mixed results. Metolachlor wasnot corrosive to this coating. The weightgain was only 0.3 percent. The hardness,SUTS, and elongation of the coating did notchange significantly; however, the elasticmodulus (E) decreased slightly. In cloma-zone, the coating weight increased 4-6 per-cent, but the increase did not exceed thetest criterion of 7 percent. Mechanical testsrevealed no significant decrease in the me-chanical properties of the coating. In con-trast to metolachlor and clomazone, IMPand dimethoate caused excessive weightincreases: 11 and 14 percent, respectively.Softening also occurred during the expo-sure of the coating to these test chemicals.The SUTS and hardness decreased while theelongation increased.

Amino-Amine-Cured EpoxyThe mechanical properties of the amino-amine-cured epoxy resin determined be-fore and after exposure are given in Tables3, 4, and 5. Like the 100 percent solidspolyamide-cured epoxy, the amino-amine-cured epoxy deteriorated in the solvent-borne herbicides (clomazone, IMP, anddimethoate) but not in atrazine. The weightgain of the coating films was less than 2percent in atrazine. Metolachlor also hadlittle effect on the coating; weight gain wasless than 1 percent. Clomazone anddimethoate caused excessive weight gainsto the amino-amine-cured epoxy coating.Weight gains ranged from 26-44 percentduring exposure to these chemicals.

The mechanical properties of thecoating did not change significantly after




exposure to atrazine and metolachlor. However, the coating did soften when ex-posed to clomazone and dimethoate. The hard-ness decreased from HB to 6B, whilethe tensile strength decreased about 60 per-cent in clomazone and 90 percent indimethoate. The total elongation increasedafter exposure to both chemicals.

Polysiloxane EpoxyThe polysiloxane epoxy exhibited excellentcorrosion resistance to both water- and sol-vent-borne herbicides. The coating gainedvery little weight, less than 2 percent, whenexposed to atrazine and IMP. The coatingsuffered minor weight loss, less than 2 per-cent, when exposed to clomazone, meto-lachlor, and dimethoate. Since the coatingwas very brittle, only hardness tests wereperformed. No other mechanical tests wereconducted on this coating.

100 Percent Solids Epoxy PhenolicIn addition to good mechanical properties,the 100 percent solids epoxy phenolic coat-ing exhibited good corrosion resistance.The highest weight gain, approximately 7

percent, was observed in IMP. The weightgain was about 6 percent in both cloma-zone and dimethoate, but only 1 percent orless is atrazine and metolachlor.

Inspection of the coating’s mechanicalproperties after chemical exposure indicat-ed that the pencil hardness of the coatingdecreased only 1 to 2 units of hardness inmost of the test chemicals. The SUTS in-creased in most test chemicals, with the ex-ception of IMP. The elongation increasedslightly, 2 to 5 percent, in most solutions.

90 Percent Solids EpoxyBefore chemical exposure, the SUTS of the90 percent solids epoxy coating was only628 psi (4.33 MPa), and the films were veryductile. The coating had a hardness of Band a total elongation of about 12 percent.The coating hardened during chemical ex-posure. That is, the SUTS and hardness in-creased, and the total elongation decreased.The SUTS increased by approximately 400percent, and the hardness increased by 5units, while the total elongation decreasedfrom 12 percent to 1 or 2 percent after ex-posure to atrazine, metolachlor, and IMP.



JULY 1995 / 65

During exposure to clomazone anddimethoate, the coating continued to age,but the hardening was offset by softeningdue to chemical absorption. As a result, theSUTS increased only slightly (from 628 to777 and 928 psi [4.33, 5.36, and 6.40 MPa],respectively, in clomazone anddimethoate), and the hardness was un-changed. Total elongation changed from 12percent to 16 percent in clomazone andfrom 12 percent to 18 percent indimethoate. The coating failed to meet thetest criteria due to excessive weight loss inatrazine and metolachlor. In dimethoateand clomazone, the coating gained weight,approximately 6 and 12 percent, respective-ly. In IMP, the coating lost less than 4 per-cent of its initial weight.

Polysulfide EpoxyThe polysulfide-epoxy coating had bothstrength and ductility before chemical ex-posure (Table 3). However, during thechemical exposure, polysulfide epoxy resindeteriorated substantially. In clomazoneand dimethoate, the coating failed com-pletely. Four of the 5 test specimens ex-posed to each of the chemicals failed dueto cracking and softening and could not beretrieved. One test specimen in each chemi-cal remained intact but had a large weightgain, 51 percent in clomazone and 84 per-cent in dimethoate. The coating weight alsoincreased about 24 percent in IMP.

Atrazine and metolachlor did notharm the coating. Weight loss in atrazinewas 0.3 percent. In metolachlor, the coatingweight increased less than 2 percent.

Corrosivity of UAN to CoatingsResults indicated that UAN had little effecton the coatings. With the exception of the90 percent solids epoxy resin, the coatingsperformed very well in UAN The 90percent solids epoxy resin had a weightloss of 2.7 percent, failing the weight losscriterion. The other coatings exhibited only

minor weight gains (less than 1 percent).The fact that UAN solution has little effecton coatings is important for applicationswhere pesticides and UAN solution areboth used. Coating selection can then bebased on the pesticides involved.

Tests on Uncoated ConcreteTable 6 gives the compressive strengths ofuncoated concrete with and without chemi-cal exposure. Compressive test results indi-cated that the unexposed and uncoated testconcrete had an average compressivestrength of 3,650 psi (25.2 MPa) after a 28-day cure. These specimens withstood thefreeze-thaw cycle tests. No evidence ofcracking or spading was observed on anyof these specimens. The average compres-sive strength after freeze-thaw cycle testswas 3,108 psi (21.4 MPa).

Clomazone penetrated the entirespecimen. Concrete compressive strengthdecreased significantly, as much as 37 per-cent. Average compressive strength of 3 ex-posed specimens was 2,300 psi (15.9 MPa).Freeze-thaw cycle tests caused no addition-al damage with compressive strength re-maining approximately the same. Discol-oration due to precipitation and chemicalpenetration occurred when uncoated con-crete was exposed to IMP.

Similar to clomazone, IMP penetratedthe entire thickness of the Uncoated con-crete specimens. The concrete was alsosoftened in IMP. Compressive strength de-creased approximately 45 percent. Freeze-thaw tests resulted in an additional de-crease in compressive strength (7 percent).Average compressive strengths of concretetested in IMP were 1,770 and 2,000 psi(12.2 and 13.8 MPa) with and withoutfreeze-thaw testing, respectively.

Tests indicated dimethoate did not af-fect uncoated concrete. Compressivestrength decreased slightly (1 percent orless). Average compressive strength was3,610 psi (24.9 MPa) Freeze-thaw tests re-




sulted in additional decreases in concretecompressive strength. However, the aver-age decrease was only 7 percent and theaverage compressive strength was 3,390 psi(23.4 MPa). Dimethoate was detected inspecimen surface defects such as voids orcavities. However, no evidence of chemicalpenetration, tunneling, or absorption wasobserved on concrete specimens exposedto dimethoate, although a minor odor wasdetected after compression tests.

UAN solution was extremely corrosiveto uncoated concrete. Compressive strengthdecreased approximately 37 percent to 1,970 psi (13.6 MPa). UAN completelypenetrated the test specimens. Other specimens exposed to freeze-thaw testsfailed and could not be retrieved for compressive tests.

Tests on Coated ConcreteTable 7 gives abrasion test results on coat-ings before and after chemical exposure.Tables 8 and 9 give compressive strengthsof coated concrete with and without chemi-cal exposure (dimethoate). Data on IMP,clomazone, and 32-0-0 UAN can be foundin the original paper in SSPC 94-19.

Epoxy Phenolic ResinBefore chemical exposure, the epoxy phe-nolic had a shiny finish and was mediumgray. The coating had good abrasion resis-tance with an average AI of 72 mg/1,000cycles (Table 7). In clomazone, IMP, andUAN, the coating retained its gloss after 3months of exposure. No evidence of deteri-oration, such as delamination, spalling,chalking, or softening, was observed whenthe coating was exposed to clomazone andUAN solution. Discoloration was observedonly on specimens exposed to IMP due tostaining by the pesticide. In addition to dis-coloration, the topcoat delaminated whenexposed to IMP, but the primer coat didnot. The epoxy phenolic coating providedan effective barrier against clomazone, IMP,

and UAN penetration. No evidence ofchemical absorption or tunneling was de-tected. The epoxy phenolic coating was at-tacked by dimethoate. Softening and etch-ing were the primary causes ofdeterioration. Although the chemical affect-ed the coating, there was little effect onconcrete strength. The decrease in com-pressive strength, to approximately 3,500psi (24.1 MPa), was less than 6 percent(Table 8). Odor detection of dimethoate inconcrete specimens after compressive testsindicated chemical penetration and tunnel-ing into the concrete. Thus, with exposureto dimethoate, potential contamination ismore serious than loss of structural integri-ty.

Mixed results were obtained fromabrasion tests. In IMP and UAN, the abra-sion resistance of the epoxy phenolic coat-ing was still excellent with average valuesof 70 and 78 mg/1000 cycles when ex-posed to IMP and UAN, respectively (Table7). However, abrasion testing could not beconducted on coating panels exposed toclomazone and dimethoate, because speci-mens were soft and swollen.

Epoxy Novolac ResinBefore chemical exposure, the epoxy no-volac had a shiny finish and was mediumgray in color. This coating exhibited excel-lent abrasion resistance with an average AIof 45 mg/1000 cycles, the lowest among 4coatings tested (Table 7).

Discoloration due to pesticide de-posits was the main problem caused by ex-posure to IMP. Otherwise, the coating ex-hibited excellent performance in all testsolutions. The coating retained its glossafter 3 months of exposure. No sign of de-terioration, such as delamination, spalling,chalking, discoloration, or softening due tochemical penetration or tunneling was ob-served. No significant decrease in compres-sive strength with and without freeze-thawtesting was observed. The average com-



JULY 1995 / 67

pressive strength of coated concrete rangedfrom 3,900-3,740 psi (26.9-25.8 MPa). Thedecrease ranged from 0-6 percent (Tables 8and 9). Abrasion tests also showed that thechemicals had no effect on the coating. Av-erage AI ranged from 47 to 62 mg/1000 cy-cles after exposure to the test chemicals(Table 7).

Vinyl Ester ResinBefore chemical exposure, the vinyl esterexhibited a dull light gray finish. The AIwas 89 mg/1000 cycles, the highest of the 4coatings tested (Table 7). No sign of chemi-cal attack or deterioration was observed oncoated concrete when tested in UAN. How-ever, the coating lost its gloss when ex-posed to the pesticides. There was evi-dence of etching and thinning on thetopcoat, but no swelling or softening. Anabundance of filler crystals was found onthe exposed surface during visual examina-tion. Chemical penetration by clomazoneand dimethoate was indicated by a strongodor emitted after compressive tests. Com-pressive test results indicated that mostdamage was due to chemical attack of thecoating. Concrete compressive strength de-creased 3 percent or less (Table 8).

No significant loss in abrasion resis-tance was found when the coating was ex-posed to clomazone, IMP, or UAN. Thesespecimens have AI values of 102, 91, and85 mg/1000 cycles, respectively. AI of the

coating decreased significantly to 171mg/1000 cycles when tested in dimethoate.Therefore, this coating may be susceptibleto abrasion when exposed to dimethoateand heavy traffic.

Polysiloxane ResinBased on the manufacturer's recommenda-tion, this coating was used with sand fillerto increase solids content and abrasion re-sistance for concrete floors. The coatingwas dark gray in color. Like other coatingsexposed to IMP, the color turned yellowdue to pesticide deposit. Otherwise, thecoating exhibited excellent performance inall test chemicals and provided excellentprotection. No evidence of chemical pene-tration or tunneling was observed on theconcrete specimens. Decreases in compres-sive strength were 5 percent or less includ-ing those measured after freeze-thaw test-ing (Table 8).

Conclusions

Because urethanes and epoxies can be for-mulated in a number of ways, these con-clusions are not necessarily applicable toother types of epoxies and urethanes. Testssimilar to those described here or tests ofcoatings applied to concrete specimensmay be necessary to determine a coating’ssuitability for secondary containment ser-




vice. The following conclusions have beendrawn from the test results (Table 10).• Urethane, polysulfide, urea urethane, 100 percent solids polyamide-cured epoxy,amino-amine-cured epoxy, and polysulfideepoxy performed well in atrazine and ben-tazon but unsatisfactory in solvent-borne liquids. They failed due to extensive expan-sion, softening, cracking, and dissolution.• The 100 percent solids epoxy phenolicand 90 percent solids phenolic performedwell except for 2 specimens with weightgains in dimethoate slightly above the 7 percent criterion. The 90 percent solidsepoxy did not receive a good rating be-cause it had weight losses between 2 and 3percent in atrazine, metolachlor, and UAN.• Epoxy phenolic, epoxy novolac,

polysiloxane epoxy, and vinyl ester per-formed well in all test chemicals. Thepolysiloxane epoxy, however, was verybrittle, and its mechanical properties couldnot be tested. Although the mechanicalproperties of the other 3 coatings changed,the test specimens were still in good condi-tion overall after 21 days of exposure.• Water-borne pesticides and UAN had lit-tle effect on coatings. Only the 90 percentsolids epoxy coating failed due to excessiveweight loss. Atrazine and bentazon werethe least destructive herbicides. The sol-vent-borne products were destructive toseveral coatings. The most destructive pes-ticides in decreasing order weredimethoate, IMP, and clomazone.• Test results indicated that thermosetting



JULY 1995 / 69

resins (epoxy and vinyl-ester coatings) pro-vide better corrosion resistance than ther-moplastic resins when exposed to solvent-borne pesticides. These thermosetting resintypes provide good resistance to water-borne herbicides.• Uncoated concrete was rapidly penetrat-ed when exposed to clomazone, IMP,dimethoate, and UAN. Absorption of clo-mazone, IMP, and UAN caused the concreteto soften. However, absorption ofdimethoate did not cause softening. Freeze-thaw cycling caused serious damage to concrete exposed to UAN solution but hadno effect on concrete exposed to cloma-zone, IMP, or dimethoate.• Without proper barrier protection, con-crete is penetrated and attacked by UAN

solution. All 4 coatings tested (epoxy phe-nolic, epoxy novolac, vinyl ester, andpolysiloxane) provided excellent protectionfor concrete exposed to UAN solution.• When exposed to clomazone, epoxy no-volac and polysiloxane provide excellentprotection for concrete (resist chemicalpenetration or softening). Epoxy phenolicand vinyl ester coatings provide limitedprotection from abrasion.• When exposed to IMP, all test coatingsdiscolored because of suspension deposits.Epoxy phenolic delaminates because IMPpenetrates the topcoat and accumulates atthe primer/topcoat interface. Epoxy no-volac and other epoxy coatings performedadequately in pesticides tested and in UAN.The 100 percent solids epoxy vinyl ester




and polysiloxane coatings demonstratedexcellent resistance to the pesticides.• Organic coatings are only required forprotection of concrete from dimethoatewhen contamination is a consideration. Toprevent dimethoate from penetrating theconcrete substrate, both epoxy novolac andpolysiloxane coatings provide excellent barriers. Epoxy phenolic, however, is not acceptable because of its low abrasion re-sistance. Vinyl ester is not acceptable be-cause of the possibilities of reducing filmthickness due to abrasion.• The emulsifiable concentrates, whichcontain organic solvents, were the most de-structive to the coatings because of theirsolvents. Generally the inert ingredientsand not the pesticides soften or destroy thecoatings. Atrazine and bentazon, the 2water-borne pesticides, had little effect onthe coatings tested.

With laboratory tests completed, fieldtrials of 7 coatings from 4 companies areunder way at a retail agrichemical facility inwestern Tennessee. JPCL

References

1. M.F. Broder, D.T Nguyen, A. L. Harner,“Effects of Fertilizer arid Pesticides onConcrete,” ASAE Paper No. 921090, St.Joseph, MI: American Society of Agricul-tural Engineers, 1992.

2. Annual Book of ASTM Standards, Sec-tion 6, Volume 06.01; Section 8, Volume08.01; Section 9, Volume 09.01,Philadelphia, PA; ASTM: 1992.

3. NACE Book of Standards, “EvaluatingElastomeric Materials in Sour Gas Envi-ronments,” NACE Standard TM0187-87Surface, Houston, TX: NACE Interna-tional, 1987.

4. A.O. Fisher, “Tools and Techniques forLaboratory Corrosion Testing,” Houston,TX: NACE International, Corrosion/64.

5. A.O. Fisher and C.N. Carpenter, “Ad-

vances in the Chemical Testing of Elas-tomers,” Materials Performance, May1981.

6. Annual Book of ASTM Standards, “Rec-ommended Practice of Making/CuringConcrete Test Specimens in the Labora-tory,” ASTM C 192-90, Volume 04.0.

7. Annual Book of ASTM Standards, “Rec-ommended Practice, for Surface Prepa-ration for Application of Chemical-Re-sistant Resin Monolithic Surfacings,”ASTM C 811-90, Volume 04.05.

8. NACE Book of Standards, “MonolithicOrganic Corrosion Resistant Floor Sur-face,” NACE Standard RP0376-76,Houston, TX: NACE International,1991.

9. Annual Book of ASTM Standards, “Rec-ommended Test Method for CompressiveStrength of Concrete Cylinders Cast inPlace in Cylindrical Molds,” ASTM C873-93, Volume 04.02.

10. Annual Book of ASTM Standards, “Stan-dard Test Method for Resistance of Con-crete to Rapid Freezing and Thawing,”ASTM C 666-92, Volume 04.02.

11. Annual Book of ASTM Standards, “Stan-dard Test Method for Abrasion Resis-tance of Organic Coatings by the TaberAbraser,” ASTM D 4060-90, Volume06.01.

12. Annual Book of ASTM Standards, “Rec-ommended Practice for Evaluating De-gree of Checking of Exterior Paints,”ASTM D 660-87, Volume 06.01.

13. Annual Book of ASTM Standards, “Rec-ommended Practice, for Evaluating De-gree of Cracking of Exterior Paints,”ASTM D 661-86 Volume 06.01.

14. Annual Book of ASTM Standards, “Rec-ommended Practice for Visual Evalua-tion of Gloss Differences Between Sur-faces of Similar Appearance,” ASTM D4449-90, Volume 06.01.

15. D. Weldon, “Understanding Test Datafrom Coatings Manufacturers’ ProductData Sheets,” JPCL, May 1993, p. 52

Duy T. Nguyen, PhD,works for the TennesseeValley Authority (TVA)

in the Division ofNuclear Engineering at

Chattanooga, TN. Whenthis paper was written,be was a metallurgical

engineer at TVA’sEnvironmental Research

Center (ERC), MuscleShoals, AL. He holds a

PhD from the Universityof Oklahoma in

metallurgicalengineering. He is a

member of ASTM,NACE, and the

American Societyfor Metals.

Michael F. Broder hasbeen an agricultural

engineer at TVA’s ERCfor the last 16 years.

In this position, be hasbeen a consultant to

fertilizer andagrichemical

manufacturers andretailers. He is a member

of SSPC, the NationalSociety of Professional

Engineers, and theAmerican Society of Agri-

cultural Engineers.He has an MS in

agricultural engineeringfrom Purdue University.

and is a professionalengineer, registered

in Alabama.

Cathy L. McDonald isa chemical engineer

at TVA’s ERC. She hasover I7 years’ experiencein corrosion studies andtesting, plant chemistry,

and plant testing.She is a graduate of theUniversity of Mississippi

with a BS in chemicalengineering. Cathy is a

member of NACEInternational.

Eugene A. Zarate,PhD, currently manages

the inorganic section ofthe Analytical Laboratory

at TVA’s ERC. He is theprincipal investigator

on a team working onmercury detection.

He has a PhD ininorganic/

organometallicchemistry from

Case Western ReserveUniversity. Zarate

is a member ofthe American

Chemical Society.

Documents

Protecting Secondary Containment from Pesticides and Fertilizer · 2016. 10. 24. · Protecting Secondary Containment from Pesticides and Fertilizer C by Duy T. Nguyen, Michael F