Publication No. 01-008-026

USE OF FLORIDA PHOSPHOGYPSUM IN SYNTHETICCONSTRUCTION AGGREGATE

Prepared by

U.S. Department of the Interior

Bureau of Mines

under a grant sponsored by the

Florida lnstitute of Phosphate Research

Bartow, Florida

September, 1983

FLORIDA INSTITUTE OF PHOSPHATE RESEARCH

USE OF FLORIDA PHOSPHOGYPSUM IN SYNTHETICCONSTRUCTION AGGREGATE

Research Project FIPR #81-01-008Memorandum of Agreement #14-09-0070-855, Bureau of Mines

Final Report, September 1983

Prepared byU.S. Department of the Interior

Bureau of MinesTuscaloosa Research Center

P.O. Box LUniversity, AL 35486

Principal InvestigatorsAlexander May, John W. Sweeney, and James R. Cobble

Prepared forFlorida Institute of Phosphate Research

1855 West Main Street

Bartow, Florida 33830

FIPR Program ManagerG. Michael Lloyd, Jr.

DISCLAIMER

The contents of this report are reproduced herein as receivedfrom the contractor.

The opinions, findings, and conclusions expressed herein are notnecessarily those of the Florida Institute of Phosphate Research,nor does mention of company names or products constitute endorse-ment by the Florida Institute of Phosphate Research.

CONTENTS

Abstract .............................................................

Introduction .........................................................

Acknowledgments ......................................................

Materials used .......................................................

Criteria for evaluating samples ......................................

Asphalt-concrete .....................................................

Cement, cement-kiln-dust, and silica powder ..........................

Clays - fired samples ................................................

Calcium oxychlorides .................................................

Fly ash ..............................................................

Discussion ...........................................................

Conclusions ..........................................................

References ...........................................................

1.

2.

3.

4.

5.

6.

7.

8.

9.

TABLES

Analyses of fly ash samples, percent .............................

Typical specifications for compressive strengths .................

Asphalt-concrete containing phosphogypsum........................

Cement, cement-kiln-dust, and silica powder mixtures containingphosphogypsum. .................................................

Clay mixtures containing phosphogypsum - fired samples ...........

Calcium oxychlorides containing phosphogypsum....................

Compressive strengths of Alabama and Florida fly ashcompositions ...................................................

3:l fly ash:lime mixtures containing phosphogypsum ...............

5:l fly ash:lime mixtures containing phosphogypsum ...............

10. 1O:l fly ash:lime mixtures containing phosphogypsum..............

PAGE

1

1

2

2

4

5

6

7

7

8

18

21

21

9

9

9

13

TABLES CONT.

PAGE

11. Times for lime, fly ash phosphogypsum compositions to reach1,000 psi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

12. Thirty-five percent phosphogypsum mixtures containing 0 to 25percent lime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

13. Range of compositions of lime, fly ash, phosphogypsum toproduce aggregate strengths of 1,000 psi...................... 18

ILLUSTRATIONS

1. Particle size distribution of phosphogypsum dried at 45°C ....... 3

2. Compressive strength versus time 3:l fly ash:lime ratio ......... 10

3. Compressive strength versus time 5:l fly ash:lime ratio ......... 11

4. Compressive strength versus time 1O:l fly ash:lime ratio........ 12

5. Compressive strength versus percent phosphogypsum............... 14

6. Compressive strength versus percent fly ash..................... 15

7. Compressive strength versus percent lime.,...................... 16

8. Compressive strength versus fly ash:lime ratio.................. 17

9. Compressive strength versus time 35 percent phosphogypsum....... 19

10. Compressive strength versus percent lime, with 35 percentphosphogypsum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

1

Abstract

In its role to provide technology topromote the efficient use of mineralsand mineral process waste, the U.S.Bureau of Mines has conducted researchto identify and develop high volume usesfor phosphogypsum.

An asphalt-concrete containing 20percent phosphogypsum met the FloridaDepartment of Transportationspecifications for their Type II, SAHMand ABC-l asphalt concretes.

Mixtures containing up to 50 percentphosphogypsum, and the remainder clays,produced aggregate when fired at 1,000°C. Their compressive strengths exceeded1,000 psi.

Mixtures containing a maximum of 50percent phosphogypsum, 6 to 10 percentlime and the remainder fly ash hadcompressive strengths as high as 4,800psi. The effects of the amounts ofphosphogypsum, lime and fly ash wereinvestigated. Also, the effects of flyash to lime ratios were studied.

Mixtures of phosphogypsum with cement,cement-kiln-dust, silica powders andcalcium chloride were also investigated.These were not considered useful forroad construction.

Introduction

The Bureau of Mines and the FloridaInstitute of Phosphate Research enteredinto a memorandum of agreement toco-fund research to develop a syntheticaggregate made from indigenous Floridawaste materials, primarilyphosphogypsum. The research addressedthe mission of both agencies; from theBureau of Mines perspective it developedtechnology with wide generic applicationto economically recycle waste materials,and from the Florida Institute ofPhosphate Research objectives ofdeveloping technology to utilizephosphogypsum generated by the Floridaphosphate industry.

The phosphate industry in Florida is avital segment of the Nation's economyand provides a critical mineral requiredfor fertilizer production. In the 1980Bureau of Mines' Mineral Yearbook,J. W. Pressler (22)' reported thatFlorida supplied 83 percent of thedomestic and 34 percent of the world'sphosphate requirements. About 82percent of the Florida phosphate isconverted into phosphoric acid, which isused to make fertilizers.

Phosphogypsum and gypsum are bothcalcium sulfate dihydrate. The namephosphogypsum is used to designate thebyproduct of wet-process phosphoricproduction, while gypsum refers to thenatural mineral. Phosphate rock, whichis composed of apatite minerals,(calcium phosphates containing varyingamounts of carbonate and fluoride), isdigested with sulfuric acid and water toproduce phosphoric acid, phosphogypsum,and minor quantities of hydrofluosilicicacid.

A. May and J. W. Sweeney (20) reportedthat 335 million tons of phosphogypsumhad accumulated in Florida in 17phosphogypsum stockpiles or stacks.These piles occupied an average of 227acres each and ranged from 30 to 140 ftin height. More phosphogypsum is beinggenerated at a rate of about 33 milliontons a year. By the year 2000, theprojected accumulation will be over abillion tons of phosphogypsum.

The Bureau of Mines has completed a2-year research program assessingpotential environmental effectsassociated with phosphogypsum inFlorida. A. May and J. W. Sweeney(20-21) developed baseline data toassess whether there were anyenvironmental problems associated withthe storage of phosphogypsum; and

'Underlined numbers in parenthesesrefer to items in the list of referencesat the end of this report.

2

determined that phosphogypsum was notcorrosive or toxic by the EnvironmentalProtection Agency criteria, reported inthe Federal Register (14), and thattoxic elements and radium were not beingleached from the stockpiles.

A. May and J. W. Sweeney (20) reportedthat the radium content of Floridaphosphogypsum averages 21 pCi/g. OnJanuary 5, 1983, EPA issued its finalstandards to govern residualradioactivity at inactive uraniumprocessing sites in the Federal Register(15). These restrict the radiumactivity to 5 pCi/g in the top 15 cm ofsurface layer and 15 pCi/g in layersbelow 15 cm. Similar standards may beextended to phosphogypsum. Thus,compositions containing up to 23.8percent phosphogypsum may be used insurface layers and compositionscontaining up to 71.4 percentphosphogypsum may be used below 15 cmdepth.

Gypsum used in the United States in1980 for cement retarders, agriculturalapplications, fillers, plasters, andprefabricated products was 19.5 milliontons; which includes only 0.66 milliontons of phosphogypsum used inagriculture as reported by J. W.Pressler (22). Erlenstadt (13)discussed the use of phosphogypsum incement retarders, fillers, plasters andprefabricated products. Floridaphosphogypsum contains traces ofphosphate and fluoride, which must beremoved in order to use thephosphogypsum in these applications.Wheelock (26) reported on themanufacture of sulfuric acid made fromphosphogypsum, Another use forphosphogypsum is in road construction,especially with lime and fly ash. Flyash has been extensively used as aconstruction material. Hester (18) gaveconstruction and performance data forapproximately 200 miles of concretepavement containing a small amount offly ash constructed in south Alabama.The engineering properties,compositions, and performance of.

lime-fly ash mixtures were considered byBarenberg (11), and by theTransportation Research Board (24), inresearch sponsored by the AmericanAssociation of State Highway andTransportation officials in cooperationwith the Federal Highway Administration.Brink (12) gave details on a 200-acreparking lot paved with lime-flyash-synthetic gypsum compositions,containing about 20 percent syntheticgypsum. The synthetic gypsum resultedfrom the manufacture of hydrofluoricacid, from acid mine drainage sludge,and from sulfur dioxide scrubber sludge.No phosphogypsum was used in theseapplications.

This investigation reports the use ofphosphogypsum to make aggregate suitablefor road construction. Its uses withasphalt, cement, cement-kiln-dust,silica, clays, oxycholorides, lime andfly ash are considered.

Acknowledgments

The authors wish to acknowledge theassistance of Dr. David P. Borris,Executive Director, Florida Institute ofPhosphate Research for his advice andassistance and also to acknowledge theassistance of the Florida Department ofTransportation for evaluating thesynthetic phosphogypsum aggregate.

Materials Used

The phosphogypsum used was a sampleobtained from a USS Agri-Chemicals'stockpile near Bartow, Florida. Theanalysis and characteristics of Floridaphosphogypsum are fully described byA. May and J. W. Sweeney (20). Thesize distribution of phosphogypsum,dried at 45° C, is shown in figure 1, aplot of percent passing versus U.S.Standard sieve numbers. The scale usedto locate the sieve numbers is (4/log 2)[log (size, mm)] shown at the top of thegraph. This scale was derived from thesize openings of U.S. Standard sieves,which are the fourth root of 2 raised toan integer power specified in ASTM E-11

3

Figure 1. - Particle size distribution of phosphogypsum dried at 45" C.

70 (5). The scale is the integer. Thisplot has the same form as plottingagainst log (size, mm), but has theadvantages that the scale is linear, theunits are small integers, and theintegers correspond exactly to the sievenumbers.

Two fly ash samples were used, onefrom Alabama and the other from Florida.The Alabama fly ash was from theWilsonville, Alabama, steam plant andits production methods, quality controland characteristics were presented byStyron (23). The Florida fly ash wasfrom Tampa Electric Company's Big Bendpower plant. Analyses of these flyashes are shown in table 1. Acommercial, high calcium, lime was alsoused.

Some tests were made using commercialball clays, clay solids from phosphaterock beneficiation, cement-kiln-dust,commercial calcium chloride, and reagentgrade calcium oxide, calcium hydroxide,and calcium chloride.

4

Criteria for Evaluating Samples

The criteria for evaluating samples ofasphalt-concrete were free fromambiguity. The requirements, listed inthe Florida Department of TransportationSpecifications (16), were applied tophosphogypsum, sand, aggregate, asphaltmixtures, Requirements for aggregatesare also listed in the Floridaspecifications.

Tests, such as the Los Angelesabrasion test, are needed for compliancewith specifications, but are notsuitable for screening large numbers ofcompositions, because of the quantitiesof materials needed. Unconfinedcompression strength was used toevaluate the samples because smallsamples sufficed and quantitativeresults were obtained which related touseful applications of aggregates.However, a specific value of compressivestrength was needed as a criterion toindicate potentially useful products.

A study of 16 lime-fly ash pavementsby the Transportation Research Board(24) gave the initial compressivestrengths of the pozzolanic basematerials; and the strengths andperformance evaluations of the pavementsafter 1 to 8 years of service. If thepavements were poorly constructed,severely overloaded at an early age, orif the subbase was unstable; thepavements cracked and deteriorated underrepeated loading. The over-all dataindicated that an initial compressivestrength of 1,000 psi gave a highquality pavement. Thus, a minimumstrength of 1,000 psi was adopted inthis investigation as the criterion tojudge the phosphogypsum compositions.For comparison, the strengthrequirements of various mortars cured at23° C, 95 percent relative humidity areshown in table 2.

5

'American Society for Testing andMaterials.

In addition to the compressivestrength criterion, practicalconsiderations were also employed.Thus, uses of expensive reagents, highenergy requirements, or factors whichwould cause the final product to benon-competitive were also used toevaluate the potential usefulness of theproducts.

Asphalt-Concrete

These tests were performed todetermine if phosphogypsum could be usedto replace sand or fine aggregate inasphalt-concrete. Twenty mixes ofvarious combinations of asphalt,phosphogypsum, sand, and coarseaggregate were tested.

The tests run were: stability,resistance to plastic deformation, andflow of asphalt-concrete by the Marshallapparatus, American Society for Testingand Materials (ASTM), standard methodD1559 (9). Also bulk density, ASTMD1188 (6), maximum specific gravity,ASTM D2041 (l0), and the percent airvoids, ASTM D3203 (8) were determined.The optimum asphalt content and thevoids in the mineral filler were derivedfrom the test data.

The supplement to the FloridaDepartment of TransportationSpecifications (17) list requirementsfor 12 types of asphalt-concrete. Thereare four surface course mixes, S-I,S-II, Type II, and Type III; onepatching mix, sand-asphalt hot mix,SAHM; three asphalt base course mixes,ABC-l, ABC-2, and ABC-3; and fourwearing surface or friction course.mixes, FC-1, FC-2, FC-3, and FC-4.

A mix containing 7.0 percent gradeAC-20 asphalt and 93 percent mineralfiller; composed of 20 percentphosphogypsum, dried at 150° C, 30percent sand, and 50 percent limerockaggregate; met the requirements forasphalt-concrete for Type III, SAHM, andABC-1 mixes. Thus, phosphogypsum couldbe used as a mineral filler inasphalt-concrete mixes for surface orbase courses or for patching. However,none of the mixes met the requirementsfor wearing or friction courses. Table3 shows the test results for the 20percent phosphogypsum mix, and therequirements for Type III, SAHM, andABC-l mixes.

The maximum amount of phosphogypsumthat proved satisfactory was only 20percent.2 When higher percentages weretested, coating of all the aggregateparticles with asphalt was incomplete,and the asphalt-concrete test resultsand the aggregate particle sizedistribution did not meetspecifications. The particle sizes ofthe phosphogypsum are much smaller thanthe size requirements for aggregate inasphalt-concrete. This can be seen bycomparing figure 1 and table 3. Forexample, 20 to 45 percent of theaggregate is required to pass the number

The 20 percent refers tophosphogypsum dried at 150° C, thetemperature normally used to dryaggregate in asphalt plants. Underthese conditions phosphogypsum maydehydrate to the hemihydrate; and 20percent of hemihydrate is equivalent to23.7 percent of phosphogypsum.

6

40 sieve for Type III, shown in table 3.Figure 1 shows that almost 80 percent ofphosphogypsum passes the number 40sieve. Therefore, only part of the sandcan be replaced by phosphogypsum andretain the proper aggregate sizedistribution to produce a satisfactoryasphalt-concrete.

'Asphalt-concrete: 7 percent gradeAC-20 asphalt, 93 percent aggregate.Aggregate: 20 percent phosphogypsum, 30percent sand, 50 percent limerock stone.

2Florida Department of Transportationspecifications (16).

Cement, Cement-Kiln-Dust andSilica Powder

These tests were performed to deter-mine if the phosphogypsum could be usedto replace sand in cement mortars, of if

phosphogypsum would produce strong, hardmaterials with cement-kiln-dust or withsilica powder. Kiln dust is essentiallywaste cement, so its use with byproductphosphogypsum would be advantageous.The mixtures with silica powder wereused in attempts to reproduce claims ina patent by R. C. Vickery (25). Heclaimed that a building material, havingcompressive strengths of 3,000 to 5,000psi, could be made from compositions of1 to 25 percent siliceous material and 2to 10 percent lime with the remainderconsisting of phosphogypsum; andvariations and modifications of theformulation.

Various mixtures of phosphogypsum,cement, cement-kiln-dust, hemihydratemade from phosphogypsum, silica powder,lime and water were made into standard2-inch cubes and tested by the ASTMmethod Cl09 (7) for compressionstrength. Typical results are shown intable 4. The compositions of the cementand the cement-kiln-dust mixtures werethe same as that of Portland cementmortar used in the standard compressiontest; but with phosphogypsum replacingsand and kiln dust replacing cement.The composition using silica powder wasone listed in the patent cited. Thesecompositons are shown in table 4.

7

To determine if an acceleratedreaction would occur betweenphosphogypsum and cement or cement-kiln-dust, 2-inch cubes, of the samecompositions mentioned, were autoclavedat 200" C and 225 psi for 16 hours. Thecompressive strength of the cementmortar dropped to 290 psi, and that ofthe cement-kiln-dust to 200 psi. Theautoclave contained more water after thesamples were treated than the amount ofwater initially placed in the autoclave.This indicated that the phosphogypsumhad dehydrated to hemihydrate with therelease of water of hydration. Othermixes on autoclaving gave the sameresults - lowering of their strengthsand release of water.

The strengths of the cement and silicapowder compositions met the criterion,1,000 psi, adopted for a usefulaggregate. Cement-kiln-dustcompositions failed this test. However,the strength of the cement-phosphogypsummortar was less than half that requiredfor Portland cement-sand mortars.Considering the relatively low strengthsobtained, and the cost of cement andsilica powder, further investigationsusing cement, cement-kiln-dust, orsilica powders were curtailed.

Clays - Fired Samples

Briquettes, balls, and cylinders wereprepared containing 25 percent, 50percent, and 75 percent phosphogypsumand the remainder either commercial ballclay or clay from phosphate rockbeneficiation. A minimum amount ofwater was added to form a plastic mass.The briquettes were hand formed in amold 1.5- by 0.3- by l-inch dimensions.The balls were hand formed about 0.5 to0.75 inch diameter. The cylinders, 1inch diameter by 1.3 inch length, wereformed in a press under 8,000 psi andthen extruded. All samples were driedat least 24 hours before firing.

The briquettes and cylinders wereplaced in a cold furnace, brought up to600" C, 800" C, or 1,000" C and kept at

these temperatures 1 hour, after whichthe furnace was turned off and thesamples allowed to cool in the furnace.The balls were placed in a hot furnaceat 600° C, 800° C, or 1,000° C for 15minutes and then removed. Somebriquettes were heated to 1,300° C.

The briquettes were used for visualdetermination of structural soundnessand measurement of dimensionalstability. Samples were sound at 600°C, 800° C, and 1,000° C anddisintegrated at 1,300° C, probablybecause gypsum decomposes at the highertemperature. The higher temperaturesproduced lighter colors and gradualincrease in shrinkage.

The balls were used to estimate thetolerances of the mixes to thermalshock. All samples remained intact whensuddenly placed in the hot furnace asdescribed. These specimens were hardand could be dropped from 6 to 10 feeton a hard surface with no damage.

The cylinders were used to measure thecompressive strengths of the mixtures.These strengths are shown in table 5.The commercial ball clays with 25percent phosphogypsum, fired at 600° C,800° C, or 1,000° C, or with 50 percentphosphogypsum fired at 1,000° C may besuitable as an aggregate. Clay fromphosphate rock beneficiation with 25percent phosphogypsum fired at 800° C or1,000° C or with 50 percentphosphogypsum fired at 1,000° C may alsobe suitable as an aggregate. The claysfrom the phosphate rock beneficiationand the phosphogypsum are bothbyproducts of the phosphate industry.Their use, in proportions up to 50percent of each would utilize bothmaterials. Energy considerations, toheat the materials, may be a detrimentto their use to make an aggregate.

Calcium Oxychlorides

Solutions of soluble alkaline earthchlorides, sulfates, or nitrates reactwith calcium or magnesium oxide or

8

hydroxide to form oxy-compounds, such ascalcium oxychloride, 3CaO*CaC12*nH20.R. S. Kalyoncu, (19) reported on theapplications of these types of compoundsfor refractories. The initial reactionoccurs in minutes, followed by a slowerreaction to form hard cementitiousmaterials.

Numerous mixtures of reagent grade The objective of this investigationcalcium hydroxide, calcium cloride, and was the practical development of anwater were made to select a uniform aggregate from phosphogypsum. Thus, weprocedure to produce the oxychlorides. did not investigate why commercialOf these mixtures, four were machined chemicals gave products whose strengthsinto cylinders l-inch diameter and were so much lower than the productsl-inch height for compression tests. obtained with reagent grade chemicals.The strengths of these samples averaged Because of the low strengths and the2,200 psi. An oxychloride, containing instability of the products in water,20 percent phosphogypsum, had a further work on oxycholorides wascompressive strength of 2,100 psi. discontinued.

Three different compositions,containing 10 percent, 30 percent, and50 percent phosphogypsum, and commerciallime and commercial calcium chloridewere investigated. All compositions hada 3 to 1 molar ratio of lime to calciumchloride. Four batches of eachcomposition were made, and the sampleswere cast into molds 1.5 inch diameterby 1.5 inch height. The samples werestored at approximately 25° C and 75percent relative humidity until testedfor their compressive strengths, whichare shown in table 6. Each value intable 6 is the average of four batches.

The results showed that the strengths A systematic investigation of thewere very low, and decreased with age effects of lime, fly ash and

and with increased phosphogypsumcontent. The samples prepared from bothreagent grade and commercial chemicalswere also tested for their stability inwater. The samples cracked anddeteriorated extensively within severalmonths.

Fly Ash

As mentioned in the introduction; alime-fly ash-synthetic gypsum mixturewas used to pave a 200 acre parking lot.This lot was used for only 1 week duringan exposition, and insufficient time wasavailable to adequately prepare thesurface, design the mixture, or toevaluate the results. Although thepavement served its purpose much of itwas damaged by traffic. The project didnot clearly establish the factorsinvolved in using synthetic gypsum withlime and fly ash.

9

phosphogypsum on the strengths ofaggregate mixtures was pursued in thisresearch effort. Standard 2-inch cubesof various compositions were tested bythe ASTM method Cl09 (7) for compressivestrengths.

As reported in the materials section,two fly ash samples were available.These were compared by measuring thestrengths of the mixtures, containing 40percent phosphogypsum, 10 percent limeand 50 percent of each fly ash, on a drybasis; plus a minimum amount of water toform mortars. The average results areshown in table 7.

The strengths of the mixturescontaining either fly ash would beadequate for aggregate. However, theremaining tests were made using theFlorida fly ash because of its proximityto Florida sources.

A series of test were performed onmixtures with fly ash:lime ratios of3:1, 5:l and 10:l; each containing 20 to80 percent phosphogypsum. Anotherseries of test were run on mixturescontaining 35 percent phosphogypsum,0-25 percent lime, and the remainder flyash. The fly ash:lime ratiosnecessarily differed for each of thesemixtures. Compressive strengths weredetermined on each specimen after curingfor 7, 28 and 84 days at 23° C and 95percent relative humidity.

Compressive strengths of the mixtureswith the 3:l fly ash: lime ratio areshown in table 8 and shown in figure 2.Figure 2 shows that the compressive

strength increases with time and with adecrease in the percentage ofphosphogypsum.

Compressive strengths of the 5:l flyash:lime mixtures are shown in table 9and shown in figure 3; and the resultsfor the 10:l ratio are shown in table 10and figure 4. Results for both of theseratios, show the same trends as infigure 2--an increase in strength withtime and with a decrease in thephosphogypsum content.

12

Figure 4. - Compressive strength versus time 10:1 fly ash:lime ratio.

Table 10. - 1O:l fly ash:lime mixturescontaining phosphogypsum

CompressiveTest strength, psi

Curing time,No. G L F days'

12 2013 3514 50 4.55 45.45 11015 65 3.18 31.82 8016 80

‘At 23° C and 95 percent relativehumidity.

G - percent phosphogypsum, dry basis.L- percent lime, dry basis.F - percent fly ash, dry basis.

The change in strength with time isindicated by the slopes of the lines infigures 2, 3, and 4. In figure 2, at 80percent phosphogypsum; in figure 3, at65 and 80 percent phosphogypsum; and infigure 5, at 35, 50, 65 and 80 percentphosphogypsum; the slopes are smallbetween 28 and 84 days.

This data indicated small changes instrength that occurred after 28 days.Each of these "flat" lines representmixtures having less than 6 percentlime.

The time for the mixtures to reach1,000 psi is about the same for each flyash:lime ratio for the same percentphosphogypsum. From the intersectionsof the lines with 1,000 psi on figures2, 3, 4, we obtain times shown intable 11.

The curing time ranges from 2 to 7weeks for compositions achieving greaterthan 1,000 psi in 28 days.

The change in compressive strength asa function of phosphogypsum content isillustrated in figure 5; as a functionof fly ash content in figure 6 and as afunction of lime content in figure 7.These figures also indicate effects offly ash:lime ratios.

13

Table 11. - Times for lime, fly ashphosphogypsum compositionsto reach 1,000 psi

80 NAPNAp Not applicable, does not reach

1,000 psi.'Averages for compositions whose

strengths reached 1,000 psi.

The trends in compressive strengthsappear to be well defined from figures5, 6, and 7. For example, figure 5shows that the 5:l ratio is almost astraight line from 20 to 65 percentphosphogypsum. The correlationcoefficient for this line is -0.998.However, the percent phosphogypsum, pluslime, plus fly ash, equals 100. Thesecomponents are not independentvariables. Higher phosphogypsum contentnecessarily requires lower lime and flyash content. Thus, it is not certain ifthe lower strengths with increasingphosphogypsum content, shown in figure5, were due to phosphogypsum, lime, orfly ash.

The 3:1, 5:1, and 1O:l fly ash:limeratios gave different compressivestrengths. To show how these ratiosaffected the strengths, they arecompared in figure 8. The strengths aregreatest for 20 percent phosphogypsumand least for the 80 percentphosphogypsum for all ratios.

Figure 8 shows a distinctive featureof the data. The 20 and 35 percentphosphogypsum mixtures increased instrength from the 3:l to the 5:l ratioswhile the 50, 65, and 80 percentphosphogypsum mixtures decreased instrength. Overall consideration of thedata indicated that this drastic changewas due to the lime content.

14

Figure 5. - Compressive strength versus percent phosphogypsum.

Figure 8. - Compressive strength versus fly ash:lime ratio.

18

Compressive strengths, as a functionof lime content, with a constantphosphogypsum content, are shown intable 12. The changes in the strengthsof these mixtures with time are shown infigure 9, which indicated that the limerequirements are between 5 and 10percent.

Strengths as a function of percentlime is shown in figure 10 for mixturescontaining 35 percent phosphogypsum.Figure 10 contains data from tables 8,9, 10, and 12; and also indicates flyash:lime ratios. Maximum strength wasobtained with a 5:l fly ash:lime ratioand 10.8 percent lime. The fly ash:limeratio and percent lime for maximumstrength differs from these values whenthe percent phosphogypsum is differentfrom 35 percent.

Figure 10 indicates that to developadequate strength, the fly ash:limeratio should be between 3 and 10 and thepercent lime should be between 6 and 11percent. Mixtures containing more than65 percent phosphogypsum would beexcluded, since all samples with 6percent or more of phosphogypsum had lowstrengths. The range of compositionsfor adequate strength are shown in table13.

A mixture containing 10 percent lime,40 percent phosphogypsum and 50 percentfly ash was tested for abrasion by theLos Angeles abrasion test. The loss wasless than 50 percent, meeting therequirement of the Florida Department ofTransportation Specifications. Unitweight of the aggregate was 106 lb/cuft, also meeting the specifications.

Tests were conducted to evaluate theswelling properties of lime, fly ash,phosphogypsum compositions when immersedin water. The expansion of specimens,immersed in water 28 days, ranged from0.31 to 4.18 percent, with greaterexpansion corresponding to higher limeor phosphogypsum contents. Preliminaryresults indicate that the lime contentmust be 10 percent or less, and thephosphogypsum content must be less than20 percent to limit swelling to lessthan 1 percent.

Discussion

This investigation showed that the useof phosphogypsum with cement,cement-kiln-dust, silica powder, orcalcium oxychlorides was not practicalto make aggregates. The productsobtained had low strengths, orstability, or would be expensive toproduce.

Phosphogypsum can be used as a mineralfiller in asphalt-concrete. However,only about 20 percent phosphogypsum canbe used because its particle size is toosmall to replace all the fine aggregatein these mixtures.

19

Figure 9. - Compressive strength versus time 35 percent phosphogypsum.

20

Figure 10. - Compressive strength versus percent lime, with 35 percentphosphogypsum.

Mixtures of clays and phosphogypsumproduce aggregate when fired at800-1,000° C. The cost of firing themixes may be a detriment to this use.However, clays from phosphatebeneficiation dnd phosphogypsum are bothwaste products of the phosphateindustry, and both could be utilized inthis application.

Mixtures of fly ash, lime andphosphogypsum produced aggregates havingcompressive strengths near 5,000 psi,requiring no energy other than mixing.These mixtures may contain as much as 50percent phosphogypsum. However, theyare slow curing, requiring from 2 to 7weeks to reach a strength of 1,000 psiand about 3 months to developconsiderable strength.

The strength decreases with anincrease in phosphogypsum content andincreases with an increase in fly ashcontent. For adequate strength the limecontent should be about 6 to 11 percent.Below 6 percent lime, strengths are low,and above 11 percent lime, strengthsremain about the same. The fly ash:limeratios may vary from 3:l to 1O:l. Thestrengths decrease when this ratio isgreater than 5:l. However, the higherratios require less lime and may be moreeconomical than the lower ratios.

The swelling characteristics of thelime, fly ash, phosphogypsumcompositions may limit the amount ofphosphogypsum that can be used in thesemixtures.

Conclusions

Asphalt-concrete containing 7 percentgrade AC-20 asphalt and 93 percentaggregate; consisting of 20 percentphosphogypsum, 30 percent sand and 50percent limerock stone; met the FloridaDepartment of TransportationSpecifications (17) for Type III, SAHM,-and ABC-l asphalt-concrete.

Mixtures containing 25 and 50 percentphosphogypsum and the remainder clayfrom phosphate rock beneficiationproduced aggregate when fired at 1,000°C for 1 hour. The compressive strengthsexceeded 1,000 psi.

Coarse aggregate made from mixtures oflime, fly ash and phosphogypsum,obtained maximum compressive strengthsof 4,800 psi. Mixtures containing amaximum of 50 percent phosphogypsum 6to 10 percent lime, 3:l to 1O:l flyash:lime ratio and a minimum of 38percent fly ash produced aggregate with28-day compressive strengths exceeding1,000 psi.

1.

2.

3.

4.

5.

6.

References

American Society for Testing andMaterials. Standard Specificationsfor Blended Hyraulic Cements C595-76 in 1977 Annual Book of ASTMStandards. Part 13 Cement; Lime;Ceilings and Walls, Philadelphia, PA1977, pp. 369-377.

. Standard Specifications forFly Ash and Other Pozzolans for Usewith Lime C 593-76a in 1977 AnnualBook of ASTM Standards. Part 13,1977, pp. 363-368.

. Standard Specifications for-----Natural Cement C l0-76 in 1977Annual Book of ASTM Standards. Part13, 1977 pp. 4-6.

. Standard Specifications for-----Portland Cement C 150-77 in 1977Annual Book of ASTM Standards. Part13, 1977, pp. 140-146.

________ . Standard Specifications forWire-Cloth Sieves for TestingPurposes E 11-70 in 1977 Annual Bookof ASTM Standards. Part 13, 1977,pp. 507-511.

Standard Test Method for-----.Bulk Specific Gravity of CompactedBituminous Mixtures usingParaffin-Coated Specimens D 1188-71

22

in 1978 Annual Book of ASTMStandards. Part 15 Road and PavingMaterials; Bituminous Material forHighway Construction, Waterproofingand Roofing, and Pipe: SkidResistance Philadelphia, PA. 1978,pp. 372-373.

7. . Standard Test Method forCompressive Strength of HydraulicCement Mortars (Using 2-in or 50-mmcube specimens) C 109-77 in 1977Annual Book of ASTM Standards. Part13, 1977, pp. 62-67.

8. . Standard Test Method forPercent Air Voids in CompactedDense and Open Bituminous PavingMixtures D 3203-75 in 1978 AnnualBook of ASTM Standards. Part 15,1978, pp. 803-804.

9. . Standard Test Method forResistance to Plastic Flow ofBituminous Mixtures Using MarshallApparatus D1559-76 in 1978 AnnualBook of ASTM Standards. Part 15,1978, pp. 425-433.

1 0 . . Standard Test Method for-___-Theoretical Maximum SpecificCravity of Bituminous PavingMixtures D2041-78 in 1978 AnnualBook of ASTM Standards. Part 15,1978, pp. 522-530.

11. Barenberg, E. J. Utilization of Ashin Stabilized Base Construction.Paper in Ash Utilization.Proceedings: Third InternationalAsh Utilization Symposium.Sponsored by National CoalAssociation, Edison ElectricInstitute, American Public PowerAssociation, National AshAssociation, and Bureau of Mines,Pittsburgh, PA. Compiled by J. H.Faber, W. E. Eckard and J. D.Spencer. BuMines IC 8640, 1973,pp. 180-196.

12. Brink, R. H. Use of Waste Sulfateon Transportation '72 Parking Lot.Paper in Ash Utilization

13.

14.

15.

16.

17.

18.

19.

Proceedings: Third InternationalAsh Utilization Symposium.Sponsored by National CoalAssociation, Edison ElectricInstitute, American Public PowerAssociation, National AshAssociation, and Bureau of Mines,Pittsburgh, PA. Compiled by J. H.Faber, W. E. Eckard and J. D.Spencer. BuMines IC 8640, 1973,pp. 197-207.

Erlenstadt, G. Upgrading ofPhosphogypsum for the ConstructionIndustry. Paper in Phosphogypsum.Proceedings of the InternationalSymposium on Phosphogypsum. LakeBuena Vista, Florida, Nov. 5-7,1980. Editors D. P. Borris and P.W. Boody, pp. 284-293.

Federal Register, v. 45, No. 98,May 19, 1980, p. 33122.

Federal Register, v. 48, No. 3,Jan. 5, 1983, p. 592.

Florida Department ofTransportation StandardSpecifications for Road and BridgeConstruction 1977, pp. 535-540.

Florida Department ofTransportation SupplementalSpecifications to the 1977 StandardSpecifications for Road and BridgeConstruction, July 1979, pp. 44-45.

Hester, J. A. Fly Ash in RoadwayConstruction Paper in Fly AshUtilization Proceedings: EdisonElectric Institute-National CoalAssociation-Bureau of MinesSymposium, Pittsburgh, PA.,compiled by J. H. Faber, J. P.Capp, and J. D. Spencer. BuMinesIC 8348, 1967, pp. 87-100.

Kalyoncu, R. S. Chemically BondedRefractories-A Review of theState-of-the-Art. BuMines RI 8878,1982, pp. 11-13.

23

20. May, A. and J. W. Sweeney.Assessment of Environmental ImpactsAssociated with Phosphogypsum inFlorida. BuMines RI 8639, 1982,19 PP.

21. May, A. and J. W. Sweeney.Evaluation of Radium and ToxicElement Leaching Characteristics ofFlorida Phosphogypsum Stockpiles.BuMines RI 8776, 1983, 23 pp.

22. Pressler, J. W. Gypsum. Articlein U.S. Bureau of Mines MineralsYearbook, 1980, v. I, Metals andMinerals, pp. 385-396.

23. Styron, R. W. Quality Control andBeneficiation of Fly Ash. Paper inAsh Utilization. Proceedings:Second Ash Utilization Symposium.Sponsored by National CoalAssociation, Edison ElectricInstitute, American Public PowerAssociation, National AshAssociation and Bureau of Mines,Pittsburgh, PA., compiled by J. H.Faber, N. H. Coates, and J. D.Spencer. BuMines IC 8488, 1970,pp. 151-164.

24. Transportation Research Board.National Cooperative HighwayResearch Program Synthesis ofHighway Practice 37. Lime-FlyAsh-Stabilized Bases and Subbasis,1976, 66 pp.

25. Vickery, R. C. SyntheticLightweight Building Material.U.S. Patent 3,847,634, November 12,1974.

26. Wheelock, T. D. Desulfurization ofPhosphogypsum. Paper inPhosphogypsum. Proceedings of theInternational Symposium onPhosphogypsum. Lake Buena Vista,Florida, November 5-7, 1980.Editors D. P. Borris and P. W.Boody, pp. 377-400.

APPENDIX

Reviewers' Comments and USBM Rebuttal

Comments on the Final Report on

“Use of Florida Phosphogypsum in Synthetic Construction Aggregate”

1. This report is a good preliminary study on the use of Florida phosphogypsum

in synthetic construction aggregates. The findings of the report indicated

that phosphogypsum mixtures can be successful ly used as synthet ic

construction aggregates.

2. A minimum compression strength of 1000 psi, which was adopted in this study

as the criterion to judge the phosphogypsum compositions, tends to limit the

scope of the study and the use of phosphogypsum in highway construction. It

may be desirable for the highway base course to have the compressive

strength of 1000 psi. The requirement for compressive strength for subbase

course materials is certainly much lower than 1000 psi.

3. A follow-up study of the mixture of phosphogypsum with fly ash and lime is

necessary to implement the use of the mixture in highway construction.

Optimum water content and fatigue testing should be studied for the phospho-

gypsum mixture.

4. One of the findings as stated in in the report, "...For adequate strength the lime

content should be about 6 to 11 percent. Below 6 percent lime, strengths are

low....” is contradictury to Dr. Don Saylak’s findings as reported in the ASTM

meeting on Gypsum held in Atlanta, Georgia, on April 14-15, 1983, and the

results obtained at the University of Miami. Judging from the scope of this

study, the reviewer believes that it is premature to make such a statement.

Bureau of Materials and ResearchP. 0. Box 1029Gainesville, FL 32602

July 19, 1983

Dr. David Borris, Executive DirectorFlorida Institute of Phosphate Research1855 West Main StreetBartow, Florida 33830

Subject: FIPR #81-01-008

Dear Dr. Borris:

We have reviewed the Research Project FIPR #81-01-008 on"Use of Florida Phosphogypsum in Synthetic Construction Aggregate."

The statements made in the report regarding the use of thissynthetic aggregate in asphalt concrete are misleading. Specific-ically, on Page 1, the second paragraph of the Abstract, althoughthe mix may have met the Marshall design requirements, there isno indication the aggregate itself met the requirements of theFlorida Department of Transportation Standard Specifications.Similar synthetic aggregates had been evaluated by this Bureauand were found to degrade in water. Attached is a copy of theletter to Alexander May indicating the findings.

Strengths sufficient for base structures can be attainedusing phosphogypsum in combination with lime and fly ash. However,expansion seems to be a problem. Durability was not considered,thus considerably more development work in the lab, as well as inthe field, is required before the usefulness of this material as abase structure can be properly evaluated.

We appreciate the opportunity to comment, since road construc-tion materials are under constant review by this Bureau for theDepartment of Transportation, and we trust that this review willbe of some interest to your organization.

Very truly yours,

j..lj&Gy,& dl,-

L. L. Smith, P.E.Deputy State Materials andResearch Engineer

LLS:aycc: Dr. K. H. HoAttachment

Florid Department of Transportation

GOVERNOR

(904) 372-5304

PAUL N. PAPPAS

Bureau of Materials and ResearchP. 0. Box 1029Gainesville, Florida 32602

November 4, 1982

Mr. Alexander MayBureau of MinesTuscaloosa Research CenterP. 0. Box LUniversity, Alabama 35486-9777

Subject: Phospho-gypsum Synthetic Aggregate

Dear Mr. May;

We have reviewed the test results sent to us and based on additionaltests in our laboratory, we find that the aggregate sent to us forevaluation is not suitable for use in asphalt concrete due to its moisturesusceptibility.

Attached is a summary of the tests performed. It should be noted, asdiscussed with you over the phone, that the water immersion of theMarshall specimens is not a standard test. Additionally, samples of thecrushed synthetic aggregate were immersed in water and degradation of theaggregate was observed.

It is possible that your tests using cubes of the synthetic are notrepresentative of the surface of the crushed material, The cubes couldhave a "glazed" surface preventing the intrusion of water where thecrushed surface is open and porous appearing.

We hope that this information will assist you in your development-process.

Very truly yours,

G. C. PageBituminous Materials andPavement Evaluation Engineer

GCP:cjg

Attachment

CC: Mr. C. F. PottsMr. L. L. SmithMr. K. H. HoMr. K. H. Murphy

REPLIES TO REVIEW OF MANUSCRIPT

Manuscript: "Use of Florida Phosphogypsum in Synthetic ConstructionAggregate"

Comment:

Reply :

Comment:

Reply :

Comment:

Reply :

Comment:

1. This report is a good preliminary study on the use of Floridaphosphogypsum in synthetic construction aggregates. The findings ofthe report indicated that phosphogypsum mixtures can be successfullyused as synthetic construction aggregates.

The report does indicate that phosphogypsum mixtures have apotential for use as a construction aggregate. However, additionalresearch, development and field testing would be required.

2. A minimum compression strength of 1000 psi, which was adopted inthis study as the criterion to judge the phosphogypsum compositions,tends to limit the scope of the study and the use of phosphogypsumin highway construction. It may be desirable for the highway basecourse to have the compressive strength of 1000 psi. Therequirement for compressive strength for subbase course materials iscertainly much lower that 1000 psi.

The comment is valid, the criterion of a minimum 1,000 psi tends tolimit the scope of the study for evaluating subbase or basematerials. The criterion was adopted as a working guide based onthe strengths reported to give a high quality pavement as reportedon page 4 of the manuscript. We were principally interested infinding a substitute for hard-rock aggregate.

3. A follow-up study of the mixture of phosphogypsum with fly ashand lime is necessary to implement the use of the mixture in highwayconstruction. Optimum water content and fatigue testing should bestudied for the phosphogypsum mixture.

We agree that more testing would be required. The Bureau of Minesdid investigate the swelling properties of these mixes, butpriorities within the Bureau prevented further research in thisarea.

4. One of the findings as stated in the report, For adequatestrength the lime content should be about 6 to 11 percent. Below 6percent lime, strengths are low...." is contradictury [sic] toDr. Don Saylak's findings as reported in the ASTM meeting on Gypsumheld in Atlanta, Georgia, on April 14-15, 1983, and the resultsobtained at the University of Miami. Judging from the scope of thisstudy, the reviewer believes that it is premature to make such astatement.

2

Reply : The statement cited was based on the results obtained and is valid.The authors discussed the seeming contradiction with Dr. Saylak, atthe meeting cited, and during his visit to the Bureau inNovember, 1983. The fly ash used by Dr. Saylak contained 24.0 to25.4 percent total calcium oxide (lime) while the Bureau of Minesused fly ash containing only 4.1 percent calcium oxide. Thus, Dr.Saylak did not have to add lime since it was already present in hisfly ash.

Dr. Saylak was also seeking strengths to meet Texas highwayspecifications of 100 psi for stabilized subbases or 650 psi forcement stabilized bases, while the Bureau was seeking strengths inexcess of 1000 psi. Actually, there is no contradiction involvedwhen reviewing the data.

Comment: The statements made in the report regarding the use of thissynthetic aggregate in asphalt concrete are misleading.Specifically, on Page 1, the second paragraph of the Abstract,although the mix may have met the Marshall design requirements,there is no indication the aggregate itself met the requirements ofthe Florida Department of Transportation Standard Specifications.Similar synthetic aggregates had been evaluated by this Bureau andwere found to degrade in water. Attached is a copy of the letter toAlexander May indicating the findings.

Reply : The authors appreciate the cooperation of the Florida Department ofTransportation (FDT) and the letter cited in the comment. TheBureau of Mines agreed with the FDT conclusions that the Bureau'ssynthetic aggregate "is not suitable for use in asphalt concrete dueto it moisture susceptability." This preliminary evaluation by theFDT assisted us in redirecting our research efforts.

The reviewers' comment states "The statements made in the reportregarding the use of this synthetic aggregate in asphalt concreteare misleading." However, there are NO statements in the report onusing the Bureau of Mines' synthetic aggregate in asphalt concrete!

The statement referred to in the abstract and also in theconclusions, referred to the asphalt concrete containingphosphogypsum not synthetic aggregate. This asphalt concrete wasmade using Florida limerock and sand, commercially used in Floridaroad construction plus phosphogypsum. This mixture met FDTspecifications for Type 11, SAHM, and ARC-1 asphalt concrete. Thephosphogypsum was substituted for part of the fine aggregate in themix. The Bureau of Mines also immersed this Marshall test specimenin water and it showed no swelling, bleeding or detrimental effects.The Bureau of Mines will not print misleading or inaccuratestatements in its reports, only statements of fact, based onresearch evidence presented.

3

Comment:

Reply :

Strengths sufficient for base structures can be attained usingphosphogypsum in combination with lime and fly ash. However,expansion seems to be a problem. Durability was not considered,thus considerably more development work in the lab, as well as inthe field, is required before the usefulness of this material as abase structure can be properly evaluated.

We agree, as previously stated (3rd reply). We did investigateswelling of these mixes but were unable to continue further researchbecause of the project being terminated, due to a change inresearch priorities made by our Headquarters Office.

These replies are respectfully submitted to help clarify any misunderstandingsor statements in the manuscript.

Alexander May