United States Environmental Protection Agency Office of Air Quality Planning and Standards Research Triangle Park NC 27711 EPA-450/3-79-038 November 1979 Air Review of New Source Performance Standards for Phosphate Industry ------- EPA-450/3-79-038 Review of New Source Performance Standards for Phosphate Fertilizer Industry by William Herring Emission Standards and Engineering Division U.S. ENVIRONMENTAL PROTECTION AGENCY Office of Air, Noise, and Radiation Office of Air Quality Planning and Standards Research Triangle Park, North Carolina 27711 November 1979 ------- This report has been reviewed by the Emission Standards and Engineering Division of the Office of Air Quality Planning and Standards, EPA, and approved for publication. Mention of trade names or commercial products is not intended to consitute endorsement or recommendation for use. Copies of this report are available through the Library Services Office (MD-35) , U.S. Environmental Protection Agency, Research Triangle Park, N.C. 27711, or from National Technical Information Services, 5285 Port Royal Road, Springfield, Virginia 22161. 11 ------- TABLE OF CONTENTS Page LIST OF ILLUSTRATIONS vi LIST OF TABLES v1 1. INTRODUCTION 1-1 1.1 PURPOSE AND SCOPE 1-1 1.2 STUDY RESULTS 1-2 2. INDUSTRY GROWTH AND PROJECTIONS 2-1 2.1 PRODUCTION CAPACITY 2-1 2.2 REFERENCES FOR SECTION 2. 2-3 3. PRODUCTION PROCESS CHANGES 3-1 3.1 WET PROCESS PHOSPHORIC ACID 3-1 3.1.1 Dihydrate Processes 3-1 3.1.2 Hemihydrate (Hemidihydrate) Processes 3-4 3.2 SUPERPHOSPHORIC ACID 3-5 3.3 DIAMMONIUM PHOSPHATE 3-6 3.4 TRIPLE SUPERPHOSPHATE 3-9 3.4.1 TSP from Phosphate Rock Treatment 3-9 3.4.2 GTSP from Limestone Treatment 3-14 3.5 PHOSPHATE ROCK DRYING AND GRINDING 3-14 3.6 PRODUCTION PROCESS CHANGES SINCE NSPS PUBLICATION 3-15 3.6.1 Production Processes in Use at Time of NSPS Publication 3-15 3.6.2 Production Process Changes in the U.S. 3-15 3.6.3 Production Process Changes Outside the U.S. 3-16 3.6.4 Discussion of Limestone Treatment (Simplot) Process for GTSP Production 3-17 3.6.5 Discussion of Wet-Rock Grinding 3-18 3.6.6 Discussion of Hemihvdrate (Hemidihydrate) Processes for WPPA Production 3-20 3.7 STATUS OF POTENTIAL REGULATION OF RADIOACTIVE EMISSIONS FROM PHOSPHATE PLANTS UNDER THE CLEAN AIR ACT 3-21 3.8 REFERENCES FOR SECTION 3. 3-23 iii ------- 4. FLUORIDE EMISSION CONTROL CHANGES 4-1 4.1 CONTROL SYSTEMS PREVIOUSLY AVAILABLE 4-1 4.2 NEW CONTROL SYSTEMS 4-3 4.2.1 Modified Scrubbing Systems for WPPA and TSP Production 4-4 4.2.2 Modified Scrubbing System for DAP Production 4-6 4.2.3 Discussion of the Modified Scrubbing Systems 4-7 4.3 REFERENCES FOR SECTION 4. 4-10 5. ADVANCES IN GYPSUM POND FLUORIDE CONTROL 5-1 5.1 PREVIOUS OPERATION AND CONSTRUCTION OF POND SYSTEMS 5-1 5.2 PRESENT OPERATION AND CONSTRUCTION OF POND SYSTEMS 5-7 5.2.1 Treatment of Pond Water for Discharge 5-7 5.2.2 Pond Lining 5-7 5.2.3 Pond Area Reduction by Cooling Towers and Fluoride Recovery 5-8 5.2.4 Gypsum Pond Emission Reduction by Fluoride Recovery from WPPA Evaporator Gas 5-9 5.2.5 Gypsum Pond Emission Reduction by Fluoride Recovery from WPPA Process Water 5-11 5.2.6 Gypsum Pond Reduction by Hemihydrate WPPA Production 5-13 5.3 REFERENCES FOR SECTION 5. 5-14 v 6. NSPS ENFORCEMENT PROBLEMS AND COMPLAINTS 6-1 6.1 NEW SOURCES SINCE NSPS PUBLICATION 6-1 6.2 COMPLAINTS OF PHOSPHATE PLANT EMISSIONS AND SUGGESTED IMPROVEMENTS OF REGULATIONS 6-2 6.2.1 Florida 6-2 6.2.2 Idaho 6-4 6.3 STATUS OF DEVELOPMENT OF FLUORIDE MONITORING 6-4 6.4 STATE PLANS FOR REGULATING EXISTING PHOSPHATE PLANTS 6-5 6.5 REFERENCES FOR SECTION 6. 6-8 7. CONCLUSIONS 7-1 7.1 HEMIHYDRATE PROCESS FOR WET PROCESS PHOSPHORIC ACID (WPPA) 7-1 7.2 SIMPLOT PROCESS FOR GRANULATED TRIPLE SUPERPHOSPHATE (GTSP) 7-2 7.3 WET GRINDING OF PHOSPHATE ROCK 7-2 7.4 SCRUBBING SYSTEMS, COOLING TOWERS, AND GYPSUM PONDS 7-3 7.4.1 Fluoride Recovery 7-3 7.4.2 Cooling Towers 7-3 ------- 7.5 FLUORIDE EMISSION MONITORING 7-5 7.6 RADIOACTIVE EMISSIONS 7-5 7.7 AREAS FOR RESEARCH AND DEVELOPMENT 7-6 7.8 INDUSTRY GROWTH 7-6 7.9 REGULATION OF EXISTING SOURCES 7-7 7.10 REFERENCES FOR SECTION 7. 7-7 8. RECOMMENDATIONS 8-1 8.1 HEMIHYDRATE PROCESS FOR WET PROCESS PHOSPHORIC ACID (WPPA) 8-1 8.2 SIMPLOT PROCESS FOR GRANULATED TRIPLE SUPERPHOSPHATE (GTSP) 8-1 8.3 WET GRINDING OF PHOSPHATE ROCK 8-1 8.4 SCRUBBING SYSTEMS, COOLING TOWERS, AND GYPSUM PONDS 8-1 8.4.1 Fluoride Recovery 8-1 8.4.2 Cooling Towers 8-1 8.5 RADIOACTIVE EMISSIONS 8-2 8.6 RECOMMENDATION ON NSPS REVISION STUDY 8-2 ------- LIST OF ILLUSTRATIONS Figure Number Page 3-1 Wet-Process Phosphoric Acid Process 3-2 3-2 Stauffer Superphosphoric Acid Process 3-7 3-3 TVA Diammonium Phosphate Process 3-8 3-4 Run-of-Pile Triple Superphosphate Process 3-10 3-5 TVA One-step Granular Triple Superphosphate Process 3-12 4-1 Spray-crossflow Packed Bed Scrubber 4-2 5-1 Typical Gypsum Pond Servicing a 1000 TPD-PJ),- Plant * a 5-3 LIST OF TABLES Table Number Page j 2-1 Phosphate Production Capacity in the U.S. 2-2 5-1 Typical Material Balance of Fluoride in Manufacture of Wet-Process Phosphoric Acid 5-4 5-2 Fluoride Emission Factors for Selected Gypsum Ponds at 90°F; Ib /acre day 5-6 VI ------- 1. INTRODUCTION 1.1 PURPOSE AND SCOPE On August 6, 1975, the Environmental Protection Agency promulgated New Source Performance Standards (NSPS) for Five Categories of Sources in the Phosphate Fertilizer Industry (40 FR 33152). These standards establish emissions limits and require emission testing and reporting for total fluorides from wet process phosphoric acid (WPPA) plants, superphosphoric acid (SPA) plants, diammonium phosphate (DAP) plants, triple superphosphate (TSP) plants, and granular triple superphosphate (GTSP) storage facilities. The Clean Air Act Amendments of 1977 require that the Administrator of the EPA review and, if appropriate, revise established standards of performance for new stationary sources at least every four years (Section 111(b)(1)(B)). This report includes reviews of recent and projected growth of the phosphate fertilizer industry. Changes in process techno- logy since 1975 are presented, and advances in the control of fluorides from gypsum ponds are discussed. Inquiries were made on enforcement problems of the current NSPS and on continuous monitoring of fluoride control devices, and the results are presented. Compliance test results; information from the literature; and discussions with representatives of industry, control equipment vendors, phosphate fertilizer plant designers, EPA regional offices, state agencies, and the Tennessee Valley Authority form the basis for these reviews. The information obtained was analyzed to determine whether potential benefits exist that would warrant NSPS revision now. 1-1 ------- 1.2 STUDY RESULTS Based on results of this investigation, conclusions were drawn on significant developments in the phosphate industry since NSPS promulga- tion, and recommendations were formulated on action to be taken pertaining to NSPS revision. These conclusions and recommendations are shown in Section 7 of this report. 1-2 ------- 2. INDUSTRY GROWTH AND PROJECTIONS 2.1 PRODUCTION CAPACITY The total capacity of plants in the U.S. for production of wet process phosphoric acid (WPPA), superphosphoric acid (SPA), ammonium phosphate (AP), and triple superphosphate (TSP), for the years beginning 1974 and projected through 1985, is shown in Table 2-1. The Table 2-1 values were derived from TVA compilations. This data source was sup- 2 ported by recommendation of the Fertilizer Institute. Table 2-1 shows that production capacity increased for WPPA, SPA, AP, and TSP, during the years 1974 through 1979, by 3071, 496, 955, and 313 thousand tons P2^5' respectively. This amounts to an average annual growth of 614, 99, 191 and 63 thousand tons P205 for each product, respectively, over those 5 years; Table 2-1 also shows that the projected capacity increases for those products from 1979 to 1980 are 461, 350, 0, and 0, respectively. For 1980 through 1985, the projected capacity increase is zero for all aforementioned products. Recent literature indicates that U.S.-located phosphate plants are now producing at nearly full capacity due largely to high export demand; but that further expansion would be contingent on continued current export levels and increased domestic demand - conditions not assured. Exports might be reduced because of potential high production outside 345 the U.S. ' ' Pertaining to domestic demand, U.S. and Puerto Rican phosphate use decreased 9% in 1978; and there are strong indications that some soils have been overferti1ized with phosphate. * These factors support the expectation of little or no expansion of phosphate production capacity in the U.S. between 1980 and 1985. 2-1 ------- TABLE 2-1. PHOSPHATE PRODUCTION CAPACITY IN THE U.S. (THOUSAND SHORT TONS P^)1 Product 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 Phosphoric acid (WPPA) Superphosphoric acid Ammonium phosphate9 Concentrated (triple) Superphosphate6 . 6680 747 3806 2127 8518 877 4548 2367 9001 1012 4721 2596 9346 1037 5267 *, 2440 9601 893 4794 2440 9751 1243 4761 2440 10212 1593 4761 2440 10212 1593 4761 2440 10212 1593 4761 2440 10212 1593 4761 2440 10212 1593 4761 2440 10212 1593 4761 2440 ro ro Includes monoammonium phosphate and diammonium phosphate. Includes run-of-pile and granulated triple superphosphate. ------- 2.2 REFERENCES FOR SECTION 2. 1. Tennessee Valley Authority. World Fertilizer Capacity: "Phos Acid WP," TVA-05/10/79; "Super Acid W," TVA-05/21/79; "AM Phosphate," TVA-05/21/79; "Cone Super," TVA-05/21/79. Muscle Shoals, Alabama. 22 p. 2. Letter from K.T. Johnson, The Fertilizer Institute, to W.O. Herring, U.S. EPA. April 18, 1979. Subject: Phosphate Industry in U.S. 3. Key Chemicals, Phosphoric Acid. Chemical and Engineering News. p. 18. April 30, 1979. 4. 1979 Fertilizer Situation. U.S. Dept. of Aqriculture. Washington, D.C. Publication FS-9. Dec. 1978. D. 23. 5. Harris, G.T., and E.A. Harre. World Fertilizer Situation and Outlook - 1978-85. International Fertilizer Development Center and Tennessee Valley Authority. Muscle Shoals, Alabama 35660. Bulletin IFDC-T-13, March 1979. pp. 8-13. 6. Reference 4, p. 11. 7. Situation 78 - TVA Fertilizer Conference. Tennessee Valley Authority. Muscle Shoals, Alabama 35660. Bulletin Y-131. Aug. 1978. p. 54. 2-3 ------- 3. PRODUCTION PROCESS CHANGES All processes used commercially to make wet process phosphoric acid, superphosphoric acid, diammonium phosphate, and triple super- phosphate are summarized in Sections 3.1 through 3.5, below. Changes of production process application since NSPS publication in August 1975, are identified and discussed in Section 3.6. The status of potential regulation of radioactive emissions from phosphate plants is given in Section 3.7. 3.1 WET PROCESS PHOSPHORIC ACID 3.1.1 Pi hydrate Processes Wet process phosphoric acid (WPPA) is made by reacting sulfuric acid (H2S04) with fluorapatite (Ca10(P04)6F2) in PnosPnate rock- In tne dihydrate processes, calcium sulfate, as the dihydrate, gypsum (CaS04 2H20), is also formed. The overall reaction is described by the following equation: 3 Ca1Q (P04)6 F2 + 30H2S04 + Si02 + 58H20 - 30CaS04 ' 2 H20 + 18H3P04 + H Calcium sulfate precipitates, and the liquid phosphoric acid is separated by filtration. A modern dihydrate, WPPA plant flow diagram is shown in Figure 3-1. Finely ground phosphate rock is continuously metered into the reactor and sulfuric acid is added. The single-tank reactor (Dorr- Oliver design) illustrated in Figure 3-1 consists of two concentric cylinders. Reactants are added to the annul us and digestion occurs in this outer compartment. The second (central) compartment provides retention time for gypsum crystal growth and prevents short-circuiting of rock. 3-1 ------- GO I ro WASH . WATER GYPSUM POND WATER GROUND PHOSPHATE ROCK 'HEATER SI FAN SUCK DRY SL9.TH CAKE., I SUCK 3 WASH WASH REMOVE I PR I 1 WASH I 2 WASH 1 WASH I LIQUOR -». VACUUM .TO VACUUM "*• AND HOT WELL Si ft. PARTICULATE. SO,. ODOR GYPSUM SLURRY, TO POND •TO SCRUBBER , HYDROFLUOSILICIC ACID Fiqure 3-1. Flow diaqram illustratinq dihydrate wet-process phosphoric acid process. ------- Concentrated sulfuric acid is usually fed to the reactor. The only other water entering the reactor is the filter-wash water. Considerable heat of reaction is generated in the reactor and must be removed. This heat removal is done in modern plants by vacuum-flash- cooling part of the slurry and sending it back into the reactor. The reaction slurry is held in the reactor for up to 8 hours before being sent to the.filter. The most common filter design is the rotary horizontal tilting-pan vacuum filter denoted in Figure 3-1. It consists of a series of individual filter cells mounted on a revolving annular frame. Product slurry from the reactor is introduced into a filter cell and vacuum is applied. After a dewatering period, the filter cake / undergoes two or three stages of washing with progressively weaker solutions of phosphoric acid. The wash-water flow is countercurrent to the rotation of the filter cake with heated fresh water, or barometric condenser water, used for the last wash; the filtrate from this step is used as the washing liquor for the preceding stage, etc. After the last washing, the cell is subjected to a cake dewatering step and then inverted to discharge the gypsum. Cleaning of the filter media occurs at this time. The cell is then returned to its upright position and begins a new cycle. The 32 percent P205 acid obtained from the filter generally is concentrated to 54$ in a two- or three-stage vacuum evaporator system. In the evaporator, illustrated 1n Figure 3-1, provision is made for recovery of fluoride as fluosilicic acid. Inclusion of this recovery feature depends on economics. Many evaporation plants do not have this device. 3-3 ------- 3.1.2 Hemihydrate (Henrid1 hydrate) Processes Commercial hemihydrate, or hemidihydrate (HDH), processes for making phosphoric acid are distinguished from dihydrate processes by inclusion of two main stages - the hemihydrate stage, or first stage; and the dihydrate stage, or second stage. Fluoride recovery can be co- 2 installed with the HDH process. Hemihydrate Stage In the hemihydrate stage, phosphate rock is reacted with a mixture of 98% sulfuric acid and return phosphoric acid (from the hemihydrate 2 filter) to yield phosphoric acid and calcium sulfate hemihydrate. The overall reaction is represented by the equation: 3Ca1Q(P04)6F2 + 30H2S04 + Si02 + 13H20 -> 30CaS04 ' 1/2H20 + 18H3P04 + H Relatively high temperatures and P205 concentrations are required to obtain calcium sulfate as hemihydrate, instead of dihydrate, in the reaction of fluorapatite with H2S04. The resulting slurry is filtered to obtain the hemihydrate residue and phosphoric acid filtrate. Hemihydrate crystallization permits easier filtration, compared with dihydrate crystallization, requiring less water to dilute the acid for satisfactory filtration. Therefore, the acid obtained directly from filtration in the hemihydrate stage contains 40 to 54% P205, and is suitable as the finished product WPPA, 2 ' without further concentration by evaporation. 3-4 ------- Pi hydrate Stage In the di hydrate stage, with further H^SO- addition, the hemi hydrate is recrys tali zed as the dihydrate. Co-precipitated lattice P205 is released from the hemihydrate precipitate during recrystallization. The dihydrate slurry is filtered with one or two wash stages. Filtrate, from the first wash stage, is recycled to the hemihydrate filter to 2 recover the released PoOq- Fluoride Recovery Fluorides are evolved from the reactors in both the hemihydrate and dihydrate stages, and from the flash cooler, in the hemihydrate stage. These fluorides can be recovered by means of a co- ins tailed scrubbing 2 system as byproduct 3.2 SUPERPHOSPHORIC ACID Superphosphoric acid (SPA) is produced by submerged combustion or vacuum evaporation of clarified WPPA (containing 54% PoOc) to a p?°5 concentration between 72 and 76%. Vacuum evaporation is, commercially, much the more important method. There are two commercial processes for the production of super- phosphoric acid by vacuum evaporation: 1. The falling film evaporation process (Stauffer Chemical Co.) 2. The forced circulation evaporation process (Swenson Evaporator Co.) Feed acid clarification in both processes is usually accomplished by settling or by a combination of aging and settling. In general, both processes are also similar in other operations. 3-5 ------- Figure 3-2 shows the Stauffer process. Feed acid is added to the evaporator recycle tank, where it mixes with concentrated product acid to maintain a highly concentrated process acid for lower corrosion rates. This mixture is pumped to the top of the evaporator and distributed to the inside wall of the evaporator tubes. The acid film moves down along the inside wall of the tubes receiving heat from the steam on the outside. Evaporation occurs and the concentrated acid is separated from the water vapor in a flash chamber located at the bottom of the evaporator. Product acid flows to the evaporator recycle tank and vapors to the barometric condenser. To minimize P205 loss, the separator section contains a mist eliminator to reduce carryover to the condenser. 3.3 DIAMMONIUM PHOSPHATE Diammonium phosphate (DAP) is made in the reaction: 2 NH3 + H3P04 * (NH4)2 HP04 / It contains 18 percent nitrogen and 46 percent available 1>2®5' Tlie ™ process for DAP production is the one in widest use, and there are several variations of the original now in use. A flow diagram of the 5 basic process is shown in Figure 3-3. Anhydrous ammonia and phosphoric acid (about 40 percent PJ^S^ are reacted in the preneutralizer using a NH3 / H3P04 mole ratio of 1.35, which allows evaporation to a water content of 18 to 22 percent without thickening the DAP slurry to a nonflowing state. The slurry flows into the ammoniator-granulator and is distributed over a bed of recycled 3-6 ------- steam from package boiler FALLING-FILM EVAPORATOR Condensole, „ to package Steam boiler :Wct -pro cess (phosphoric Concent locid(54%P2Oj) FE ; CO EO TANK EVAPORATOR RECYCLE TANK f * rated ocid 1 \. , p* i i . ./ Vapors Freih end recycle acid Hr- c — T~" ^ L. • r d .OOLIN: TAN* f~ C03i< To ejectors BAROMETRIC CONDENSER ?SiF4, HF Hot V.CH r^iF/i. HF «~» i Water = C_| — . i coolant. C ci' ml Superphosphoric Superphosphoric discharge ocid (68-72%P2OsJ Ftqure 3-2. Stauffer Superphosphoric acid process. 3-7 ------- CO I CO PHOSPHORIC ACID- J PR AMMHK.T- CT UP TO AIR POLLUTION _ AMMONIA.JjF^, _H£ „ CONTROL SYSTEM AMMONIA, PARTICULATE, SiF^ HF ^ GRANULATOR DRYER CYCLONES COOLER CYCLONES rVf _N ! 1 " DRYER I 1 4J> AMMONIA PRENEUTRALIZER FEED -^ PARTICULATE OVERSIZE MILL . t COOLER PRODUCT Figure 3-3. TVA diammonium phosphate process. ------- fines. Ammoniation to the required mole ratio of 2.0 takes place in the granulator by injecting ammonia under the rolling bed of solids. It is necessary to feed excess ammonia to the granulator to achieve a 2.0 mole ratio. Excess ammonia and water vapor driven off by the heat of reaction are directed to a scrubber, which uses phosphoric acid as the scrubbing liquid. The ammonia is almost completely recovered by the phosphoric acid scrubbing liquid and is recycled to the preneutralizer. Solidification occurs rapidly once the mole ratio has reached 2.0, making a low solids recycle ratio feasible. Granulated DAP is next sent to the drier, then screened. Undersized and crushed oversized material are recycled to the granulator. Product sized material is cooled and sent to storage. r 3.4 TRIPLE SUPERPHOSPHATE (TSP) 3.4.1 TSP from Phosphate Rock Treatment Triple superphosphate (TSP), also called concentrated superphosphate, is made from phosphate rock in the reaction: Ca]0 (P04)6 F2 + 14H3P04 + 10H20 •* 10 CaH4(P04)2 ' H20 + 2HF The product contains from 44-47 percent available 1*2®$' 3.4.1.1 Run-of-Pile Triple Superphosphate - Figure 3-4 is a schematic diagram of the den process for the manufacture of run-of-pile (ROP) TSP Phosphoric acid containing 52 - 54 percent P205 is mixed at ambient temperature with ground phosphate rock, usually in a TVA cone mixer. Mixing is accomplished by the swirling action of rock and acid streams introduced simultaneously into the cone. The chemical reaction begins during mixing. 3-9 ------- PHOSPHATE ROCK—? _V PHOSPHORIC ACIO O CONTROLS to I SIP,. PARTICULATE Finure 3-4. Run-of-pilc triple superphosphate production and storaqe. ------- After mixing, the slurry goes to a "den," where solidification occurs. One of the most popular den designs is the Broadfield, a linear horizontal slat belt conveyor mounted on rollers with a long stationary box mounted over it and a revolving cutter at the end. The sides of the stationary box serve as retainers for the slurry until it sets. The solidified slurry from the den must be cured - usually for 3 weeks or more - to allow the reactions to approach completion. Final curing occurs during sheltered storage, illustrated in Figure 3-4. 3.4.1.2 Granular Triple Superphosphate - Granular triple superphosphate (GTSP) is made from cured ROP TSP by treatment with water and steam in a rotary drum, then drying and screening. The TVA one-step granular process, shown in Figure 3-5, makes GTSP directly. Ground phosphate rock and recycled process fines are fed into the acidulation drum along with concentrated phosphoric acid and steam. The use of steam helps accelerate the reaction and ensure an even distri- bution of moisture in the mix. The mixture is discharged into the granulator, where it solidifies, passes through a rotary cooler, and is screened. Over-sized material is crushed and returned with undersized material to the process. 3-11 ------- A 1 . . | RECYCLED »nr.K — «-T N/V; STEAM-J PHOSPHORIC ME/I AGIO 1 udjb METER I 1 *{ r >fl ' L 1 - 1 *c TER I— CONO. KD UMP S r r « ^ iom ATIO DHUM u WEENS >TEAM C X. 1 FINES I 1 1 1 nr ) RANULATOR COOLER || 1" ^7 i --"-• FINE OVERSIZE ] 1 1 ^ S _L M ROLL CRUSHER r i •i AGE ILL 1 _ 1 PARTICULATE »SiFA, PARTICULATE _ >SiF4> PARTICULATE PRODUCT STORA6K Figure 3-5. TVA one-step process for qranular triple superphosphate, 3-12 ------- The Dorr-Oliver slurry granulation process also produces GTSP directly. Ground phosphate rock is mixed with phosphoric acid (39% PpOc) in a series of mixing tanks. A thin slurry is continuously removed, mixed with a large quantity of dried, recycled fines in a pugmill mixer (blunger), where it coats out on the granule surfaces and builds up the granule size. The granules are dried, screened, and mostly (about 80 percent) recycled back into the process. Emissions from the drier and screening operations are sent to separate cyclones for dust removal and collected material is returned to the process. After manufacture, GTSP is stored for a short curing period. After storage of usually 3 to 5 days, during which some fluorides evolve from the storage pile, the product is .considered cured. Front-end loaders move the GTSP to elevators or hoppers, where it is conveyed to screens for size separation. Oversize material is rejected, pulverized, and returned to the screen. Undersize material is returned to the GTSP production plant. Material within specification is shipped as product. 3-13 ------- 3.4.2 GTSP from Limestone Treatment GTSP is made from limestone in the reaction 2 H3P04 + CaC03 -»• CaH4(P04)2 ' H20 + C02 The product contains 45 to 50% available P2°5- Dilute, 22-38% P205 phosphoric acid and limestone are reacted in an agitated vessel system. Fines from the granulation circuit can be added, and reacted slurry can be recycled to disperse the reactants. The final slurry goes to a granulator, where pellets are formed and then dried and screened. The reaction appears complete in the liquid phase, and the resultant stable granulator product does not continue to emit fluorides. The only fluoride entering this process is that contained in the WPPA fed. 3.5 PHOSPHATE ROCK DRYING AND GRINDING / Wet rock from mining and beneficiation contains from 7 to 20 percent moisture. Some customer contracts require 3 percent maximum moisture. Rock is dried to comply with that requirement. The two types of dryers used in the field are rotary and fluid bed dryers. The rotary dryers are inclined parallel flow units equipped with steel lifting flights, and are 2.1 or 2.4 meters (7 or 8 feet) in diameter. The fluid bed dryers operate by passing hot gases up through a perforated grating, which agitates and dries the rock. The dryers are primarily fired by oil, but use natural gas on standby service. The rock is discharged from the dryers at a temperature range of 100°C (212°F) to 140°C (284°F) o onto conveyor belts feeding dry storage bins. 3-14 ------- Some phosphate rock is pulverized at the mines and shipped to the customer ready for acidulation. To accomplish this, the rock is conveyed to the grinder after it has been dried. There, it is ground, then conveyed to the storage bin or loaded directly into the car. Some phosphate rock is also pulverized for direct soil application, to a fineness of 85 percent less than 0.07 millimeter (0.003 inch) particle size. This rock is bulk-loaded into cars, or is conveyed to a machine o where it is bagged, normally in 45-kilogram (100-pound) bags. Drying is often eliminated. Rock, containing 6 to 20% moisture, is transported from the beneficiation plant to the chemical plant where it 910 is ground and chemically processed wet. ' 3.6 PRODUCTION PROCESS CHANGES SINCE NSPS PUBLICATION .- 3.6.1 Production Processes in Use at Time of NSPS Publication When the present NSPS was published in August 1975, WPPA was made by dihydrate processes, SPA by submerged combustion or vacuum evaporation, DAP by the TVA and similar processes, run-of-pile TSP by phosphate rock treatment with phosphoric acid in the den process, and GTSP by treatment of run-of-p1le TSP and by phosphate rock treatments with phosphoric acid that produce GTSP directly.12 3.6.2 Production Process Changes in the U.S. In the U.S., there has been no substantial change in commercial production of WPPA, SPA, DAP, and run-of-pile TSP since NSPS publication.13"30 The production of GTSP by treating limestone with phosphoric acid is a new process introduced by the J.R. Simplot Co. at its Pocatello, Idaho plant. 3-15 ------- The elimination of phosphate rock drying by wet-rock grinding, prior to chemical processing is an important change of production practice. Wet-rock grinding facilities are in full-scale operation in phosphoric acid plants at the Plant City complex of C.F. Industries (Florida)» at Agrico Chemical's Faustina (Louisiana) facility, and at Q both the old and new plants of W.R. Grace's Bartow complex (Florida). IMC is also converting its dry-rock grinding operations to wet grinding 32 at its New Wales phosphate chemicals complex (Florida). There have been no reports of commercial WPPA production by hemihydrate or hemidihydrate (HDH) processes in the United States; however, the Heyward-Robinson Company of New York is now marketing the hemihydrate 33 process developed by Nisson Chemicals of Japan. Occidental Research Corporation claims development of a proprietary method to make WPPA by the hemihydrate process. Also, TVA has demonstrated pilot plant 4 35 production of WPPA by its "hemihydrate foam" process. 3.6.3 Production Process Changes Outside the U.S. A commercial plant using the HDH process, developed by Fisons Ltd. in the United Kingdom (UK), is now producing WPPA in Trepca, Yugoslavia. The plant was designed and built by Lurgi Chemie and Huettentechnik GmbH of West Germany. A similar plant for Albright & Wilson at Whitehaven, 2 UK, is under construction. HDH plants are also in commercial use in oo oc Japan. The HDH process has been used commercially since 1974. ' 3-16 ------- 3.6.4 Discussion of Limestone Treatment (Simplot) Process for GTSP Production In the Simplot process, when H-PO, in WPPA is reacted with CaCO^ to make GTSP, fluorides are present only in the WPPA. Thus the only fluorides that might be emitted are the residual fluorides (H2SiF6 or HF and SiF4) in the WPPA. In the old GTSP process, when fluorapatite (Ca(P04)gF2) in phosphate rock is reacted with H^PO,, the fluorides present are comprised of both the fluorapatite fluoride content and the WPPA residual fluorides. It follows from the above that the new Simplot process would emit much less fluoride (uncontrolled) than the old GTSP process, for each mole of monocalcium phosphate (CaHJPOJg ' H20) in GTSP produced. However, the phosphate content of each CaH4(P04)2 * 2H20 molecule, produced in either process, is derived, ultimately, from fluorapatite. Therefore, the emittable fluorides from the total processing of fluorapa- tite to make one mole of CaH.(P(h)2 ' H20 is the same regardless of whether the old process or the Simplot process is used. Potential fluoride emissions from the Simplot process are lower only in the absence of the WPPA plant. The Simplot process provides a means of reducing local fluoride emissions at the plant where it is applied. Fluorapatite is not used as a feed material, so this fluoride source is eliminated. Excess limestone can also be fed to precipitate the fluoride in the WPPA as insoluble calcium fluoride (CaF«). In contrast, the old process emits more fluorides which require more scrubbing water and therefore larger pond areas. The amount and complexity of emission control equipment is reduced by the Simplot process. The company has found that the only gas cleaning 3-17 ------- devices necessary are a scrubber on the dryer and conventional (uncoated) 38 baghouses on other system components, e.g. screens and elevators. Results of the NSPS compliance test at the Simplot plant showed 31 average emissions of 0.187 Ib F (fluoride)/ton P205 fed. The require- ment for NSPS compliance in GTSP production is 0.20 Ib F/ton P205 fed. 3.6.5 Discussion of Wet-Rock Grinding Drying and grinding of phosphate rock can cause particulate emissions to the atmosphere. The particles can escape as dust if dryer gases are uncontrolled, as well as during grinding, especially if the materials are very dry. Dust can also be a significant problem at each point of transfer of the materials. The size of the dry, ground phosphate rock 39 particle is less than 0.07 millimeter (0.003 inch). The world's primary phosphate occurrences are of sedimentary origin and contain radioactive materials, predominantly uranium and its decay / products. Uranium, along with a very small amount of thorium, is thought to have been deposited contemporaneously with the phosphate. Phosphate rock in central Florida contains uranium concentrations of between 0.01 and 0.02 percent; despite the fact that uranium ore mined in the western United States solely for its uranium content contains 10 to 20 times this concentration, the phosphate industry currently mines slightly more 40 total uranium than does the uranium industry. 3-18 ------- Under the Environmental Impact Statement for the Central Florida Phosphate Industry, proposed action will eliminate rock drying with certain exceptions where approved dryers will be used. This will decrease sulfur dioxide (SCL) and dust emissions in Polk County caused by drying, grinding, and transportation as mines in that county are depleted and new mines are opened elsewhere. Since the new mines will ship wet rock, an estimated 1140 metric (1250 short) tons per year of dust and 7090 metric (7800 short) tons per year of S0? emissions from dryers in Polk County will not migrate into adjoining areas. The emissions from existing 41 rock drying will decrease as the dryers are phased out. Reduction in radiation levels will occur as dry-rock grinding is replaced with wet-rock grinding and dryers are eliminated. This will lower fugitive dust levels, thus lowering escaping radionuclides, and also result in lower radiation levels in the immediate vicinity of the 41 grinders and the eliminated dryers. Wet-rock grinding has four advantages: (1) Reduces by about half the capital expense - from receipt of unground wet rock through the point of feeding it into the 42 acid processing system. 42 (2) Eliminates dry-rock dust pollution. (3) Improves fuel economy by 2.5 gallons per ton of phosphate rock ground, which combines with electrical power savings'to reduce 4? operating costs by $3.00-$4.25 per ton of Po°5' (4) Improves reliability, thus, reducing the required amount of surge of ground rock. If a plant is located near a mine, a rock slurry can be pumped directly to the plant from the mine, 42 eliminating rail transportation and belt conveyors. 3-19 ------- By converting from dry to wet grinding, IMC expects to save 8 million gallons of fuel oil and 18 million kWh of electricity per year in producing WPPA at its New Wales complex. Conversion will cost about $11.3 million.32 3.6.6 Discussion of Hemihydrate (Hemidihydrate) Processes for WPPA Production Solutions to three WPPA production problems are claimed for the two-stage HDH process: (1) energy consumption, (2) gypsum disposal, and 2 (3) fluoride emission. Energy consumption is reduced by avoiding the need for evaporative 2 concentration of product acid. The gypsum produced is reported to be of sufficient purity for use 2 in building material. Gypsum from conventional dihydrate processes cannot be so used because of its radioactivity level. Fluoride emissions are controlled by recovery reported to be greater / than 99% in a co-installed, on-line system. The recovered fluorides may be concentrated to 20 to 24% H2SiFg.2 Capital cost savings are reported for the HDH process (compared with the dihydrate process) in rock grinding, steam used for acid concen- tration, weak acid intermediate storage, and product acid clarification. These savings are partially offset by a larger reaction volume and filter area requirement. The capital cost for recrystallization and dihydrate filtration approximately equals that for acid concentration in dihydrate processes. An overall 5 to 10% capital cost reduction for the HDH process, compared with dihydrate processes, has been reported. Operating costs are also reported lower for the HDH process, mainly 2 because of P205 recovery exceeding 98%, and low steam consumption. 3-20 ------- In the HDH process, recrystallization depends largely on phosphate rock composition. Morocco rock is used at the Trepco, Yugoslavia, plant. The planned Whitehaven, UK, plant is designed to use Morocco rock, but is reported capable of processing other rocks such as Florida 2 rock. Reasons given for not using HDH processes in the U.S. include: 1. Uranium recovery is practiced by many plants in the U.S., and is reported to be greater from the more dilute phosphoric acid produced in dihydrate processes. About 1 Ib uranium/ton P^Og is recoverable. At about $40 per pound of uranium, it is more economical to use the current dihydrate process, recover the uranium, and finally concentrate the acid from 28 percent.43 2. Florida rock is not of sufficiently high quality to allow its use.36 3. Problems are encountered in the filtration operation when using Florida rock. 4. The relatively large amount of clay fines and other impurities 36 make this process difficult to control. 3.7 STATUS OF POTENTIAL REGULATION OF RADIOACTIVE EMISSIONS FROM PHOSPHATE PLANTS UNDER THE CLEAN AIR ACT The Office of Radiation Programs (ORP) of EPA has been assigned the responsibility for potential standards for radioactive emissions. ORP has furnished the following information regarding the status of that 45 work as it relates to phosphate plants. 3-21 ------- "Phosphate ore deposits contain radioactive materials primarily in the form of uranium-238 and its decay products. The concentrations of these radioactive materials in phosphate ores range up to 100 times greater than the concentrations of these materials in normal soils and rocks. Therefore, the mining and processing of phosphate ore has the potential for releasing quantities of radioactive materials into the atmosphere which may be of public health concern. "The Environmental Protection Agency, as required by Section 122 of the Clean Air Act as amended in 1977, is presently reviewing available information to determine whether emissions of radioactive materials into the environment will cause or contribute to air pollution which may reasonably be anticipated to endanger public health. If the Agency makes a positive determination, then radioactive materials most probably will be listed as hazardous pollutants under Section 112 of the Clean Air Act and emission standards established for sources emitting important quantities of radioactive materials. "Studies are now in progress to assess the health impact from emissions of radioactive materials from phosphate processing activities. These studies include development of information on the emission levels, pathways of human exposure and health risks to the exposed populations. The results of these studies will be evaluated to determine if a need exists to establish standards under Section 112 of the Clean Air Act for phosphate processing plants. Some preliminary results of these studies have already been published in the following EPA reports: (1) "Radiation Dose Estimates Due to Air Particulates Emissions from Selected Phosphate Industry Operations," Technical Note, ORP/EERF-78-1, June 1978. 3-22 ------- (2) Radiological Surveys of Idaho Phosphate Ore Processing the Thermal Process Plant, Technical Note, ORP/LV-73-3, November 1977. (3) Radiological Surveys of Idaho Phosphate Ore Processing the Wet Process Plant - Technical Note, ORP/LV-78-1, April 1978." 3-23 ------- 3.8 REFERENCES FOR SECTION 3. 1. Final Guidelines Document: Control of Fluoride Emissions from Existing Phosphate Fertilizer Plants. U.S. EPA. Research Triangle Park, N.C. Publication No. EPA-450/2-77-005. March 1977. pp. 4-3 to 4-10. 2. Blumrich, W.E., H.J. Koening, and E.W. Schwehr. The Fisons HDH Phosphoric Acid Process. Chemical Engineering Progress. 74_: 58-61. Nov. 1978. 3. Slack, A.V. Fertilizers. Kirk-Othmer Encyclopedia of Chemical Technology, 2nd edition, Vol. 9, Stanton A. ed. New York, John Wiley & Sons. 1966. p. 93. 4. Reference 1. pp. 4-11 to 4-17. 5. Reference 1. pp. 4-17 to 4-20. 6. Reference 1. pp. 4-21 to 4-28. 7. U.S. Patent. Direct Granulation Process for Triple Superphosphate. Assignee: J.R. Simplot Co. No. 4,101,637. July 18, 1978. 8. Central Floride Phosphate Industry Areawide Impact Assessment Program, Vol. 1: Description of Program and Industry. U.S. EPA Region 4. Atlanta, GA. Project No. 08-01-4196. Sept. 1978. pp. 25-26. 9. Environmental Impact Statement, Central Florida Phosphate Industry, Vol. 1. U.S. EPA Region 4. Atlanta, GA. Publication No. EPA 904/9-78-026a. Nov. 1978. p. 2.1. 10. Environmental Impact Statement, Central Florida Phosphate Industry, Vol. 2. U.S. EPA Region 4. Atlanta, GA. Publication No. EPA 904/9-78-026b. Nov. 1978. p. 2.21 11. Federal Register. Volume 40, p. 33152. 12. Reference 1. pp. 4-1 to 4-30. 13. Telecon. W. Herring, U.S. EPA OAQPS, to W. Aronson, U.S. EPA Region 4. April 17, 1979. Subject: Phosphate Industry in Region 4. 14. Letter from K.T. Johnson, The Fertilizer Institute, to W.O. Herring, U.S. EPA. April 18, 1979. Subject: Phosphate Industry in U.S. 15. Telecon. W. Herring to G. McNeil!, U.-S. EPA Region 4. April 18, 1979. Subject: Phosphate Industry in Region 4. 16. Telecon. W. Herring to Derr Leonhardt, State of N.C. April 20, 1979. Subject: Phosphate Industry in N.C. 17. Telecon. J. Rom, U.S. EPA Region 4, to W. Herring. April 23, 1979. Subject: Phosphate Industry in Region 4. 3-2'4 ------- 18. Telecon. W. Herring to J. Symes, State of Florida. April 23, 1979. Subject: Phosphate Industry in Florida. 19. Telecon. W. Herring to J. Pfander, U.S. EPA Region 10. May 21, 1979. Subject: Phosphate Industry in Region 10. 20. Telecon. W. Herring to R. Stenner, State of Idaho. April 19, 1979. Subject: Phosphate Industry in Idaho. 21. Telecon. W. Herring to B. Varner, U.S. EPA Region 5. April 24, 1979. Subject: Phosphate Industry in Region 5. 22. Telecon. W. Herring to L. Szempruch, U.S. EPA Region 5. April 24, 1979. Subject: Phosphate Industry in Region 5. 23. Telecon. W. Herring to J. Reed, State of Illinois. April 24, 1979. Subject: Phosphate Industry in Illinois. 24. Telecon. W. Herring to J. Hepola, U.S. EPA Region 6. May 21, 1979. Subject: Phosphate Industry in Region 6. 25. Telecon. W. Herring to S. Spruiell, U.S. EPA Region 6. June 5, 1979. Subject: Phosphate Industry in Region 6. / 26. Telecon. G. Wallin, State of Texas, to W. Herring. April 30, 1979. Subject: Phosphate Industry in Texas. 27. Telecon. W. Herring to 0. Tanner, State of Louisiana. May 1, 1979. Subject: Phosphate Industry in Louisiana. 28. Telecon. W. Herring to F. Collins, U.S. EPA Region 4. May 21, 1979. Subject: Phosphate Industry in Region 4. 29. Telecon. W. Herring to B. Swan, U.S. EPA Region 10. May 21, 1979. Subject: Phosphate Industry in Region 10. 30. Telecon. M. Johnson, U.S. EPA Region 10. May 22, 1979. Subject: Phosphate Industry in Region 10. 31. Letter and attachments from J.F. Cochrane, J.R. Simplot Co., to D.P. Dubois, EPA Region 10. Dec. 30, 1977. Subject: TSP Performance Test. 32. Phosphate Process Saves on Energy. Chemical Marketing Reporter. pp. 4,27. Sept. 11, 1978. 33. Phosphoric Acid Process Cuts Costs. The Journal of Commerce. p. 5. Dec. 22, 1978. 34. Farm Chemicals, p. 138. Jan. 1979. 3-25 ------- 35. TVA Displays Energy-Saving Phosphoric Acid Process. Chemical & Engineering News. pp. 32, 38-41. Nov. 13, 1978. 36. Linero, A.A. and R. A. Baker. Evaluation of Emissions and Control Techniques for Reducing Fluoride Emissions from Gypsum Ponds in the Phosphoric Acid Industry. U.S. EPA. Washington, D.C. Publication No. EPA-600/2-78-124. June 1978. pp. 193-200. 37. ' Memo from G.B. Crane, EPA OAQPS, to S.T. Cuffe, EPA OAQPS. April 30, 1979. Subject: Phosphate Fertilizer NSPS Review. New Low Emission Process for GTSP. 38. Telecon. W. Herring to J. Smith, J.R. Simplot Co. April 25, 1979. Subject: Simplot GTSP process. 39. Reference 8. p. 35. 40. Reference 10. pp. 1.48-1.50. 41. Reference 9. pp. 21., 3.1, 3.8. 42. Reference 10. p. 2.21. 43. Telecon. G.B. Crane, U.S. EPA OAQPS, to E. Wyatt and D. Leyshon, Jacobs Engineering Group. April 17, 1979. Subject: Phosphate Fertilizer Industry. 44. Telecon. W. Herring to A. Richardson, EPA Office of Radiation Programs. June 6, 1979. Subject: Radioactive Emissions. 45. Memo from P.J. Magno, EPA (ORP) to G.B. Crane, EPA (OAQPS). June 20, 1979. Subject: Regulation of Radioactive Emissions from Phosphate Plants. 3-26 ------- 4. FLUORIDE EMISSION CONTROL CHANGES 4.1 CONTROL SYSTEMS PREVIOUSLY AVAILABLE Systems available to control fluoride emissions (hydrogen fluoride (HF) and silicon tetrafluoride (SiFj), when the NSPS for phosphate plants was published in August 1975, consisted of scrubbers. Gypsum pond or coolinq pond water was used as scrubbing media. Spent scrubbing water was returned to the plant pond system without chemical treatment. The spray-crossflow packed bed scrubber, shown in Figure 4-1, was generally used in the more effective scrubbing systems. A venturi scrubber was used in series with and before the spray-crossflow packed bed unit for gas streams having high solids content, and where solids might precipitate from scrubbing solution, as in granulated triple superphosphate (GTSP) and diammonium phosphate (DAP) processes. r The spray-crossflow packed bed scrubber consists of two sections - a spray chamber and a packed bed - separated by a series of irrigated baffles. Both the spray and the packed sections are equipped with a gas inlet. Effluent streams with relatively high fluoride concentrations - particularly those rich in SiF^ - are treated in the spray chamber before entering the packing. This preliminary scrubbing removes SiF^ thereby reducing the danger of plugging the bed. At the same time, it reduces the loading on the packed stage and provides some solids handling capacity. Gases low in SiF^ can be introduced directly to the packed section. The spray section accounts for approximately 40 to 50 percent of the total length of the scrubber. It consists of a series of counter- current spray manifolds with each pair of spray manifolds followed by a system of irrigated baffles. The irrigated baffles remove precipitated silica and prevent the formation of scale in the spray chamber. 4-1 ------- GAS INLET ro POND WATER to LJJ CD Q LU SECONDARY GAS INLET GAS FLOW FIGURE 4-1. SPRAY-CROSSFLOH PACKED BED SCRUBBER. ------- In spray-crossflow packed bed scrubbers the gas stream moves hori- zontally through the bed while the scrubbing liquid flows vertically through the packing. Solids tend to deposit near the front of the bed where they can be washed off by a cleaning spray. This design also allows the use of a higher, irrigation rate at the front of the bed to aid in solids removal. The back portion of the bed is usually operated dry to provide mist elimination. Pressure loss through the scrubber is usually 4 to 6 inches of water. Filters in the water lines ahead of the spray nozzles prevent / plugging by suspended solids. The ratio of scrubbing liquid to gas ranges from 0.02 to 0.07 gpm/acfm depending upon the fluoride content - especially the SiF^ content - of the gas stream. Approximately one third of this water is used in the spray section while the remaining two thirds is used in the packing. The packed bed is designed for a scrubbing liquid inlet pressure of about 4 or 5 pounds per square inch (gauge). Water at this pressure is available from the pond water recycle system. The spray section requires an inlet pressure of 20 to 30 pounds per square inch (gauge). This normally requires the use of a booster pump. Spent scrubbing water is collected in a sump at the bottom of the scrubber and pumped to a pond. 4.2 NEW CONTROL SYSTEMS Since NSPS publication, scrubbers have remained the principal means of controlling fluoride emissions from phosphate plants. A few modifications of scrubbing system design have been introduced. These involve scrubbing to recover fluorides as saleable fluosilisic acid ) byproduct, and removing the remaining fluorides from the gas 4-3 ------- by scrubbing with water that is cooled in closed-loop, cooling tower 14-19 systems. In some of these systems, scrubbing water is caustic treated, and a slip stream is limed to precipitate the fluoride as CaF2 and regenerate the caustic. One phosphate production process modification has been introduced since NSPS publication, as a means of controlling fluoride emissions. That modification consists of making GTSP by treating limestone, instead of phosphate rock, with wet process phosphoric acid (WPPA). This process and its emissions were described in Section 3, PRODUCTION PROCESS CHANGES. A modification of phosphate rock processing has been introduced. Emissions, including radioactive constituents, evolved during rock drying and grinding, are reduced by eliminating drying and by grinding the rock wet. This modification was also described in Section 3. 4.2.1 Modified Scrubbing Systems for WPPA and TSP Production A modified scrubbing system is used at one plant complex in Mississippi to control fluoride emissions from WPPA and triple superphosphate (TSP) production. In the WPPA process at this plant complex, gases from the phosphate rock acidulation reactor are treated in a spray-crossflow packed bed scrubber to collect fluorides. Water from gypsum pond overflow is used as the scrubbing medium. Gases from the reactor vacuum cooler and from four vacuum evaporators used for acid concentration go to a series of low-pressure-drop cascade scrubbers. Each cascade scrubber receives gas from one vacuum cooler or evaporator. The scrubbing water from the reactor scrubber is used in the cascade scrubbers, removing additional 4-4 ------- fluoride from the gases in each, finally leaving the scrubber train as byproduct 23% H2SiFg. To avoid deposition of gelatinous Si(OH)4, which might plug equipment, the arrangement of gas and liquid streams flowing to each cascade scrubber must be in an order that effects a satisfactory F/Si ratio in the scrubbing liquid. This system is reported to recover 99% of the gaseous fluorides emitted during acidulation and acid concen- tration as saleable H2$iF6.14"18 Gases from each cascade scrubber go to one of the series of plant barometric condensers where the remaining fluorides are collected. The water from the barometric condensers, at pH 3-3.5 and 120°F, goes to a cooling tower where sodium hydroxide (NaOH) is added to the stream to increase its pH to greater than 4, reduce fluoride vapor pressure, and prevent fluoride emissions from the tower. Make-up water is also added. The resulting liquid stream leaves the cooling tower at 90°F, then returns to the barometric condensers, forming a closed loop. A slip stream from the cooling tower is treated with calcium hydroxide (CaCOH^) to obtain pH 11.5-12 and precipitate the fluorides in a crystalline product composed of fluoride and silicate salts of calcium. This product is presently landfilled but has potential use as glass raw material. Fluorides emitted from TSP production and storage facilities at the above plant complex are also collected in scrubbers. The TSP scrubbing system is incorporated as a portion of the WPPA closed-loop 20 cooling tower system. Quantities of fluoride emissions from TSP production are smaller than those from WPPA production; and emissions from production of both products can be handled compatably since they are derived from the same basic raw materials and therefore, consist of similar constituents. 4-5 ------- In addition to the use of cooling towers with caustic treatment of water, as just described, cooling towers are used or planned for use at several other phosphate plants in closed-loop water handling systems, but without any reported water treatment to reduce fluoride vapor pressure. In some of these systems, process gas is treated in Swift spray-chamber 19 scrubbers' to recover fluoride as byproduct HSiF. In the usual plant arrangement, the Swift scrubbers are positioned between the acid-concen- tration vacuum evaporators and the barometric condensers, and provide 65 18 19 to 90% fluoride recovery. ' In at least one plant, the process gas is not treated to recover fluorides prior to final scrubbing; and the resulting scrubbing water is cooled in a closed-loop cooling tower 21 without caustic treatment. 4.2.2 Modified Scrubbing System for DAP Production A modified scrubbing system is used to control fluoride emissions from DAP production at three NPK (nitrogen, phosphorous, potassium mixed 15 fertilizer) plants at the previously mentioned Mississippi plant complex. Gases from the preneutralizer and granulator first go to a venturi scrubber where most of the ammonia (NH3) in the gas is absorbed in dilute phosphoric acid. Residual ammonia leaves the venturi, in the outlet gas, as gaseous NH3 or as submicron particulate ammonium fluoride (NH4F) or ammonium bi fluoride (NH4F*HF). Phosphoric acid might be •I r -i p contained in the venturi outlet gas in entrained liquid particles. The gas from the venturi goes to the first stage (nitrogen-phosphorous removal section) of a two-stage spray-crossflow packed bed scrubber where residual ammonia and phosphates are removed in additional dilute phosphoric acid scrubbing liquid. ' 4-6 ------- The scrubbing solution leaving the venturi and the spray-crossflow packed bed first stage goes to the sump of the venturi scrubber. It is treated with make-up dilute phosphoric acid and then recycled to both the venturi and the spray-crossflow packed bed first-stage scrubbers. Another stream from the venturi sump returns part of the solution to the DAP process in the preneutralizer feed stream, to recover the NH3 and phosphates.16'17 The remaining fluorides in the exit gas from the spray-crossflow packed bed scrubber first stage are removed in the second stage of that scrubber. The scrubbing water for the second stage is cooled in a cooling tower and treated with NaOH in a closed-loop system to increase its pH and limit fluoride emissions from the tower. A slip stream of water from the tower is limed to precipitate the fluoride and regenerate the NaOH.15'16'17 Gases from the DAP dryer are treated in a scrubbing system similar to that just described for the DAP preneutralizer and granulator gases. Gases from the cooler may also be similarly treated or might be sent to the inlet of the spray-crossflow packed bed scrubber treating the DAP dryer gas, thus being treated as a part of that gas stream. The system for handling gas from DAP production, including the cooling tower, is segregated from that used for WPPA and TSP process 20 gases to keep nitrogen compounds out of the WPPA-TSP system. 4.2.3 Discussion of the Modified Scrubbing Systems By employing NaOH to increase the pH of scrubbing solution and reduce fluoride vapor pressure, modified scrubbing systems provide for a potential fluoride emission reduction estimated as much as 75% below 18 present achievement levels. Maintenance problems, however, are reported 4-7 ------- to restrict scrubbing efficiency. Applicable standards are met when packed beds are in good order, but maintaining that condition is reported to be difficult.20 A plant representative states that the system has not yet been perfected. Refinements are still being made, and more time is needed to develop improved utilization of the system. Packed beds become clogged. Various types of packing and washing techniques are being tried to over- come clogging. Deterioration of rubber scrubber lining has necessitated relining in less than one year of usage. Some lined components have had to be replaced with stainless steel. Maintaining service is generally 20 difficult because of corrosion of steel parts. The closed-loop scrubbing system, employing cooling towers and alkali water treatment, was installed mainly to improve water utilization and prevent water pollution. This purpose has been met satisfactorily. Because of high rainfall in its Mississippi location, the plant complex cannot use all of the water that falls on its property. The plant gypsum pond system collects the rain water, and pond overflow feeds all 20 plant processes. Any excess water must be treated before release. The plant land area is too limited to .provide for cooling ponds of sufficient size to accept the cooling duty for the WPPA and TSP processes. The closed-loop system avoids the unacceptable alternative of taking water into the plant from the nearby bayou, using it for scrubbing, then 20 returning it to the bayou. It also reduces the total amount of plant water that must be treated before leaving the plant. By eliminating the cooling pond, the closed-loop scrubbing system with alkali treatment is estimated to reduce plant pond area and fluoride emissions from the ponds by at least 50%.14'17'18 4-8 ------- Phosphate plants converting from cooling ponds to cooling towers are located in Louisiana. As in the Mississippi plant complex just discussed, rain added to water from plant processes causes water levels in ponds to increase. Therefore, water must be recycled for disposal. Cooling towers effect evaporative cooling, while pond cooling is largely 19 by convection. Thus, cooling towers appear advantageous for reducing quantities of water that must be handled in recycling and disposal. They seem, therefore, to be coming into increased use in states with high rainfall. 4-9 ------- 4.3 REFERENCES FOR SECTION 4. 1. Final Guidelines Document: Control of Fluoride Emissions from Existing Phosphate Fertilizer Plants. U.S. EPA. Research Triangle Park, N.C. Publication No. EPA-450/2-77-005. March 1977. pp. 6-1 to 6-87. 2. Pflaum, C.A. Practical Design of Cross-flow Scrubbers in the Phosphate Industry. International Minerals & Chemical Corp. Tampa, FL. (The Fertilizer Institute Environmental Symposium. Washington. March 8, 1978.) 11 p. 3. Letter from K.T. Johnson, The Fertilizer Institute, to W.O. Herring, U.S. EPA. April 18, 1979. Subject: Phosphate Industry in U.S. 4. Telecon. G.B. Crane, U.S. EPA OAQPS, to E. Wyatt and D.Leyshon, Jacobs Engineering Group. April 17, 1979. Subject: Phosphate Fertilizer Industry. 5. Telecon. W. Herring, U.S. EPA OAQPS, to W. Aronson, U.S. EPA Region 4. April 17, 1979. Subject: Phosphate Industry in Region 4. 6. Telecon. W. Herring to G. McNeill, U.S. EPA Region 4. April 18, 1979. Subject: Phosphate Industry in Region 4. 7. Telecon. W. Herring to Derr Leonhardt, State of N.C. April 20, 1979. Subject: Phosphate Industry in N.C. 8. Telecon. J. Rom, U.S. EPA Region 4, to W. Herring. April 23, 1979. Subject: Phosphate Industry in Region 4. 9. Telecon. W. Herring to J. Symes, State of Florida. April 23, 1979. Subject: Phosphate Industry in Florida. 10. Telecon. W. Herring to R. Stenner, State of Idaho. April 19, 1979. Subject: Phosphate Industry in Idaho. 11. Telecon. W. Herring to J. Reed, State of Illinois. April 24, 1979. Subject: Phosphate Industry in Illinois. 12. Telecon. G. Wall in, State of Texas, to W. Herring. April 30, 1979. Subject: Phosphate Industry in Texas. 13. Telecon. W. Herring to 0. Tanner, State of Louisiana. May 1, 1979. Subject: Phosphate Industry in Louisiana. 14. Telecon. G.B. Crane and W. Herring to A.J. Teller, Teller Environmental Systems, Inc., May 14, 1979. Subject: Review of Phosphate Fertilizer NSPS. 15. Plant Closes Loop on its Wastewater Treatment. Environmental Science & Technology. ]2: 260-263. March 1978. 4-10 ------- 16. Teller, A.J. New Technologies in Fertilizer Emission Control. Teller Environmental Systems, Inc. Worcester, MA. (Meeting, Central Florida Section, American Institute of Chemical Engineers. Daytona. May 14-16, 1976.) 11 p. 17. Teller, A.J. New Technologies in Control of Fertilizer Plant Emissions, Pond Control - Fluoride Products. (Meeting, Fertilizer Round Table. Nov. 1975.) 25 p. 18. Teller, A.J. Scrubbers in the Fertilizer Industry, Their Success, Near Future, And Eventual Replacement. (Meeting, Fertilizer Round Table. Washington. Nov. 1973.) pp. 150-158. 19. Telecon. J. Dehn, EPA Region 6, to W. Herring. April 27. 1979. Subject: Phosphate Industry in Region 6. 20. Telecon. W. Herring to A. Gentry, Mississippi Chemicals Co., May 16, 1979. Subject: Emission Control at Pascagoula Plant. 21. Telecon. W. Herring to J. Murray, 01 in Co. July 27, 1979. Subject: Emission Control at 01 in Co. Plant. 4-11 ------- 5. ADVANCES IN GYPSUM POND FLUORIDE CONTROL 5.1 PREVIOUS OPERATION AND CONSTRUCTION OF POND SYSTEMS When the phosphate industry NSPS was published in August 1975, liquids entering gypsum ponds at plants in the U.S. consisted princi- pally of rainfall plus the process streams listed below. These process streams were formed from recycled gypsum pond water plus fluorides and other substances generated in the processes. 5.1.1. Discharge streams from dihydrate processes making wet process phosphoric acid (WPPA). 1. Gypsum slurry (filter cake slurried with pond water). Approximately 2.5 pounds of gypsum are produced per pound of pure phos- 2 3 phoric acid , or 1.2 pounds of gypsum per pound of 30% phosphoric acid. 2. Liquid from barometric condensers that treat gas from (a) the reactor vacuum cooler, and (b) the vacuum evaporators that concen- trate the WPPA. At some plants, the gas from the WPPA evaporators might have been treated in scrubbers for flousilisic acid (H2SiFg) recovery prior to entering the barometric condensers. 3. Liquid discharged from the scrubber that treats gases from the acidulation reactor, filters, hot wells, and filtrate seal tanks. 5.1.2 Superphosphoric acid (SPA) process discharge streams. 1. Liquid from the barometric condenser that treats gas from the vacuum evaporator. 2. Liquid from heat-exchanger cooling tank. .5-1 ------- 5.1.3 Diammonium phosphate (DAP) process discharge streams. 1. Liquid from the scrubbers that treat gases from the reactor, granulator, dryer, and cooler. 5.1.4 Run-of-pile (ROP) triple superphosphate (TSP) process discharge streams. 1. Liquid discharged from the scrubbers that treat gases from the mixer, den, and storage building. 5.1.5 Granular triple superphosphate (GTSP) process discharge streams. 1. Liquid discharged from the scrubbers that treat gas from the reactor, granulator, dryer, cooler, and screens. 4 Gypsum ponds are generally dyked areas. In the past, they have 3 usually not been lined to prevent seepage. The gypsum pond serves two purposes: (1) as a settling and storage area for waste gypsum, and (2) as an area for cooling process water prior to reuse. Figure 5-1 is a simplified representation of a typical gypsum pond serving a 1000-ton/day-P205 WPPA plant. This pond, handling both slurry and process water, would have about 350 acres of wet area. Water depth would be about 10 feet. Most likely it would be located adjacent to the plant and surrounded by mined-out land of sparse vegeta- tion or swamp. Assuming that the pond is used for both gypsum settling and cooling, there is a region where the stream from the sluicing operation joins the pond. This area, known as the gypsum flats, is where the gypsum settles. It is constantly worked by draglines, which remove settled wet gypsum and transfer it onto an active gypsum pile 5-2 ------- GYPSUM PILE rgROCESS HATEB.g-1151E. i en i CO / I I I \ FERTILIZER PROCESSES f Jp_PROCESS_ 0 9~5~07 FRESH ~ MAKE-UP WATER DOUBLE LIMING t EFFLUENT figure 5-1. Typical gypsum pond servicing a 1000-ton/day PJD(- plant. 2U5 ------- to dry. The gypsum pile would be about 80 feet high on about 150 acres adjacent to the wet pond. Table 5-1 shows a fluoride material balance for WPPA production. TABLE 5-1 TYPICAL MATERIAL BALANCE OF FLUORIDE IN MANUFACTURE OF WET-PROCESS PHOSPHORIC ACID BASIS: 100 LB PHOSPHATE ROCK Fluoride-bearing material or source Fluoride, Ib Phosphate rock 3.9 Product wet process acid 1.0 Gypsum 1.2 Barometric condensers 1.67 Air 0.03 Total fluoride output 3.9 Fluorides in the gypsum slurry and in the water from the barometric condensers and the scrubber that treats process emissions to air go to the gypsum pond. It therefore follows from Table 5-1 that over 70 percent of the fluorine content of the rock used in the wet-acid process may pass to that pond. If the same plant also produces DAP or TSP, a large part of the fluorine content of the phosphoric acid will also pass to the gypsum pond through the water scrubbers in these additional processes 5-4 ------- Thus, 85 percent or more of the fluorine originally present in the phosphate rock may find its way to the gypsum pond. Fluoride associated with the gypsum is, however, in insoluble form, probably as calcium fluoride (CaF2), before being sent to the pond.6'7 It is believed that fluorides from the barometric condensers are the primary source of pond . . 8 emissions. In gypsum ponds, approximately one acre foot of disposal volume is 4 required per year for each daily ton of PpOg produced by the plant. Based on wet process phosphoric acid production, plants have gypsum ponds of surface areas in the range of 0.1-0.4 acre per daily ton of P2°5- Thus a large plant may have a gypsum pond with a surface area of 200 acres or more. The water of pond systems is normally acid, having a pH around 1.5. This acidity is probably due mainly to inclusion of phosphoric acid in the washed gypsum from the gypsum filter. Gypsum pond water can be expected to contain from 0.2 to 1.5 percent fluosilicic acid (2000-12,500 ppm F) or most often, 5000-6000 ppm F. The fluosilicic acid decomposes to silicon tetrafluoride and hydrogen fluoride, resulting in a vapor-liquid equilibrium. The fluoride concentra- tion of a given pond does not continue rising as fluorides are added, but tends to stabilize. This action may be due to precipitation of complex calcium silicofluorides in the pond water. 5-5 ------- Published emission factors from gypsum ponds range from 0.2 to 10 Ib F/acre-day. Table 5-2 shows the emission factors obtained from one comprehensive investigation. Table 5-2 FLUORIDE EMISSION FACTORS FOR SELECTED GYPSUM PONDS AT 90°F (Ib/acre-day) Wind velocity at 16 ft elevation, m/sec Pond 10 (6,400 ppm F) Pond 20 (12,000 ppm F) 1 0.8 0.8 2 1.3 1.3 4 2.3 2.3 6 - 3.2 The most recent measurements of fluoride emissions from gypsum ponds indicated fluoride concentration above the pond of 18 to 46 parts per billion (ppb), consisting almost entirely of hydrogen fluoride (HF), as measured by the Remote Optical Sensing of Emissions (ROSE) system. Emission rates of 0.2 to 7.3 Ib F/acre-day were indicated by concurrent 12 wet sampling and analysis. 13 Gypsum ponds also contain radioactive material. Dried areas of 14 ponds are a possible source of radioactive emissions to the atmosphere. Radioactive emissions from phosphate plants are being investigated by 15 the Office of Radiation Programs, EPA, for possible regulation. The status of this investigation is reported in Section 3, "PRODUCTION PROCESS CHANGES." 5-6 ------- 5.2 PRESENT OPERATION AND CONSTRUCTION OF POND SYSTEMS Since NSPS publication, infprmation on ponds not previously reported, including a few changes in gypsum pond operation and construction, has been published. This information is shown below. Apart from the changes noted, present pond operation and construction is the same as described above. 5.2.1 Treatment of Pond Water for Discharge When rainfall exceeds evaporation, water is discharged from the 5 pond. This water has been treated prior to entering a natural water course. Successful treatment of discharges has been demonstrated by the use of a two-stage process; in the first stage the discharge passes through limestone (calcium carbonate), which effectively increases the pH to approximately 4.5 and in the second stage hydrated lime in used to increase the pH to more than 7. This process reduces fluoride to less than 10 milligrams per liter, phosphate to approximately 30 milligrams o per liter and radium to less than 1 picocurie per liter. 5.2.2 Pond Lining The Environmental Impact Statement (EIS) for the Central Florida Phosphate Industry includes the proposed requirement: "Line gypsum ponds with an impervious material unless it can be demonstrated in the site-specific EIS that such lining is unnecessary in protecting ground water from chemical and radiological contamination." Since Florida plants account for most WPPA production capacity in the U.S., and would, therefore, have the majority of gypsum ponds, most new ponds constructed in the U.S. will probably have impervious linings. 5-7 ------- 5.2.3 Pond Area Reduction by Cooling Towers and Fluoride Recovery Phosphoric acid plants utilize a wide variety of gypsum cooling pond arrangements. In most cases process and gypsum sluicing waters are transported to a common pond allowing these waters - which are vastly different in properties to mix, with the ultimate result that both process and gypsum pond waters become highly contaminated with phos- phoric acid, h^SCL, and hLSiF,-. In some cases, separate cooling and gypsum ponds are utilized. All process waters except gypsum sluicing water are sent to cooling ponds. Gypsum slurry is pumped from the filtration operation to a gypsum pile where the gypsum settles. The supernatant water is subsequently re- cycled through the cooling pond, thus contaminating it with phosphoric acid, HLSO,, and fluorides from the filtered gypsum. The required size of the gypsum slurry pond is small (about 5 acres) since no area is required for cooling. This water would be the most contaminated and acidic water in the plant because of the presence of phosphoric acid, H^SO,, iron and aluminum complexes, and fluorides from the filtration operation. The pond area required for the barometric condensers is determined by the cooling duty requirements. This area is estimated to be 0.1 acre per ton of P205 per day. Since the cooling pond receives condensed vapors from the flash cooler and evaporators, entrained phosphoric acid could be present as a contaminant. This, nowever, can be minimized by entrainment separators so that the main contaminant entering the cooling pond could be limited to fluorides. 5-8 ------- Since NSPS publication, cooling towers have been introduced, in place of cooling pond area, at phosphate plants as described in Section 4, "FLUORIDE EMISSION CONTROL CHANGES." In these applications, fluorides in the gases from the WPPA vacuum cooler and vacuum evaporators are scrubbed to recover the fluorides as H2SiFg prior to final scrubbing in the process barometric condensers. Water from the WPPA-reactor scrubber is used to scrub the vacuum cooler/evaporator gases, thus effecting further fluoride recovery. Recovery efficiencies as high as 99% are reported. The water from the barometric condensers is cooled in closed- loop cooling tower systems. In some of these systems, the scrubbing water in the closed loop is caustic treated and a slip stream is limed to precipitate the fluorides for disposal, and to regenerate the caustic. With cooling pond area thus mostly eliminated, pond systems are reduced 18 to the principal function of handling gypsum slurry. From /Table 5-1, most of the soluble (potentially emittable) fluorides are contained in the gases routed to the barometric condensers. Additional fluorides are contained in the gases from the WPPA reactor. Since the fluorides from both of these sources are efficiently recovered and converted to the saleable H2SiFg byproduct, and since pond area is reduced by removal of pond thermal load by cooling towers, pond area and 1R pond fluoride emissions can be reduced at least one-half. 5.2.4 Gypsum Pond Emission Reduction by Fluoride Recovery from WPPA Evaporator Gas Fluoride emissions from the pond system at the W.R. Grace WPPA plant are reduced by the process described in the following excerpt from a letter to the EPA: 5-9 ------- "The fluoride recovery system in the W.R. Grace phosphoric acid plant is the Swift & Co. process licensed to Grace by Swift. Fluorides are recovered in the form of hydrofluosilicic acid (HLSiFg) of approximate 25 percent strength. The process recovers approximately 65 percent of the fluorine in these vapors as 25 percent hydrofluosilicic acid, which calculates to approximately 25 percent fluorine recovery from the total coming in with the phosphate rock. The vapors leaving the phosphoric acid vacuum evaporators are scrubbed under vacuum with a recirculating solution of hydrofluosilicic acid whose temperature is approximately that of the vapors. Little or no water is condensed while the SiF. and HF are absorbed in the fluosilicic acid solution. The lean vapors from the fluorine scrubber are then passed to the usual barometric condenser for total condensation. A specific gravity controller activates a valve which discharges the H«SiFg to storage when up to the desired strength. Makeup water then flows into the hot well 13 through a float control valve." Pertaining to fluoride recovery as a means of reducing pond emissions, the Central Florida EIS includes the proposed requirement: "Provide for recovery of fluorine compounds from phosphoric-acid evaporators unless it is determined at the time of permit applica- tion that market conditions are such that the cost of operation (not including amortization of initial capital cost) of the re- covery process exceeds the market value of the product. If there is an exception, the site-specific EIS is to contain an estimate of pond-water fluoride concentrations to be attained and levels of 5-10 ------- fluorine emission. Estimated fluorine emissions from new- source gyp ponds should not cause the plant complex to exceed the total allowable point-source fluorine emissions within the plant complex if a permit is to be issued." Also, from the Central Florida EIS, "Reported costs of fluoride recovery operations exceed prices by 25 to 36 percent. Fluoride is marketed for the fluoridation of water supplies and for use as an industrial chemical."13 5.2.5 Gyspum Pond Emission Reduction by Fluoride Recovery from WPPA Process Water Fluoride emissions from the pond system at the USS Agri-Chemicals WPPA plant are reduced by the process described in the following excerpt from a letter to the EPA: "As a result of development engineering in the late 1960s to generate a means of diluting concentrated sulfuric acid with wet- process phosphoric acid pond water and thereby eliminate production problems with respect to conventional sulfuric acid dilution coolers and simultaneously create a negative water balance in the phosphoric acid pond water system, an offshoot of this work was devised and a process resulted which would recover fluosilicic acid (FSA) from phosphoric acid. "Our (USS Agri-Chemicals) process uses 22 percent of P205 recycle phosphoric acid from the filtration wash steps as a diluent for diluting 98 percent sulfuric acid. The heat of dilution drives off silica (sic) tetrafluoride (SiF.) from the 22 percent recycle stream, along with water vapor. The hydration product is a 25 5-11 ------- percent H^SiFg fluosilicic acid of extremely low P^Oc content. Conventional processes for stripping FSA during phosphoric acid evaporation produce FSA containing between 0.5 and 2 percent PO^C- Our process consistently produces FSA with a content well below 0.3 and frequently under 0.1 percent POr (100 percent H^SiF basis). "We installed a commercial unit which became operational in May 1970 as a first of its kind at a total capital cost of $1.5MM. In addition to reclaiming FSA through the recycle acid dilution route, we built certain flexibilities in the commercial plant to innovate recovery of FSA from fluorine scrubber systems from the triple super phosphate (TSP) manufacture and from steam stripping of the final concentrated phosphoric acid. The latter two process adjuncts did not prove satisfactory and have since been abandoned. I cannot project accurately the capital costs of a unit designed and built for recycle acid service only on today's cost basis but would guess at least $2.5MM. "Since this was an entirely new concept and the process conditions were extremely severe, the operating costs and profit benefits have been considerable less than satisfactory. Mechanical and corrosion problems resulting from the stringent operating conditions have led to prohibitive maintenance costs. The process deals with high- temperature mixes of sulfuric acid and phosphoric acid in an atmosphere of fluoride compounds, all of which contain abrasive, precipitated SiO« and calcium sulphate (gypsum). The high temperatures tend to promote the anhydride formation of calcium sulphate, and scaling is 5-12 ------- a serious problem. We have learned to live with the process and, 19 though it now performs well, it is a costly operation." 5.2.6 Gypsum Pond Reduction by Hemihydrate WPPA Production As discussed in Section 3, "PRODUCTION PROCESS CHANGES", gypsum produced in the hemihydrate (hemidihydrate or HDH) process for WPPA is reported suitable as building material. Sale of this material would 20 greatly reduce its accumulation in the gypsum pile and in ponds. The HDH process has not, however, been commercially demonstrated in 20 the U.S.; commercial plants in Europe and Japan have been reported. 5-13 ------- 5.3 REFERENCES FOR SECTION 5. 1. Final Guidelines Document: Control of Fluoride Emissions from Existing Phosphate Fertilizer Plants. U.S. EPA. Research Triangle Park, N.C. Pulbication No. EPA-450/2-77-005. March 1977. pp. 5-15 to 6-82. 2. Linero, A.A. and R.A. Baker. Evaluation of Emissions and Control Techniques for Reducing Fluoride Emissions from Gypsum Ponds in the Phosphoric Acid Industry. U.S. EPA. Washington, D.C. Publication No. EPA-600/2-78-124. June 1978. p. 3. 3. Central Florida Phosphate Industry Areawide Impact Assessment Program, Vol. 1: Description of Program and Industry. U.S. EPA Region 4. Atlanta, GA. Project No. 08-01-4196. Sept. 1978. p. 37. 4. Reference 2. pp. 6-7. 5. Reference 2. pp. 143-6. 6. Reference 1. pp. 5-6, 5-15. 7- Reference 2. p. 184. 8. Reference 2. p. 159. 9. Reference 1. p. 5-16. 10. Reference 1. p. 5-15. 11. Reference 1. pp. 5-16, 5-17. 12. Boscak, V.G., N.E. Browne, and N. Ostojic. Measurement of Fluoride Emissions from Gypsum Ponds, Draft Final Report. U.S. E.P.A. Washington, D.C. Contract 69-01-4145, Task 10. Sept. 1978. 13. Environmental Impact Statement, Central Florida Phosphate Industry, Vol. 2. U.S. EPA Region 4. Atlanta, GA. Publication No. EPA 904/9-78-026b. Nov. 1978. p. 2.25. 14. Telecon. W. Herring, EPA (OAQPS) to P.J. Magno, EPA (ORP). June 7, 1979. Subject: Radioactive Emissions from Phosphate Plants. 15. Memo from P.J. Magno, EPA (ORP) to G.B. Crane, EPA (OAQPS). June 20, 1979. Subject: Regulation of Radioactive Emissions from Phosphate Plants. 16. Environmental Impact Statement, Central Florida Phosphate Industry, Vol. 1. U.^. EPA Region 4. Atlanta, GA. Publication No. EPA 904/9-78-026a. Nov. 1978. p. 2.3. 5-14 ------- 17. Reference 2. pp. 154-5. 18. Telecon. G.B. Crane and W. Herring to A.J. Teller, Teller Environmental Systems, Inc., May 14, 1979. Subject: Review of Phosphate Fertilizer NSPS. 19. Reference 13. p. 2.24. 20. Blumrich, W.E., H.J. Koening, and E.W. Schwehr. The Fisons HDH Phosphoric Acid Process. Chemical Engineering Progress. 74: 58-61 Nov. 1978. 5-15 ------- 6. NSPS ENFORCEMENT PROBLEMS AND COMPLAINTS 6.1 NEW SOURCES SINCE NSPS PUBLICATION Based on inquiries to EPA Regions 4, 5, 6, and 10, and the States of Florida, North Carolina, Illinois, Texas, Louisiana, and Idaho, only one phosphate production plant was identified as a new, modified, or reconstructed source. That source is the J.R. Simplot Co. plant at Pocatello, Idaho, that makes granular triple superphosphate (GTSP). That plant began production of GTSP in March 1977, using a process modification that consists of treating limestone with wet process phos- phoric acid (WPPA). This modified process is described in Section 3, "PRODUCTION PROCESS CHANGES." Results of the NSPS compliance test at the Simplot plant, submitted to the EPA Regional Administrator on Dec. 30, 1977, show average fluoride 21 emissions of 1.11 Ib/hr, or 0.187 Ib/ton P2°5- Tne requirement for NSPS compliance in GTSP production is 0.20 Ib F (fluoride)/ton P205. According to a State of Idaho estimate, the above emission rate compares with fluoride emissions from the old process as great as 43 Ib/hr. The State also notes that the old process utilized the ROP/GTSP method where run-of-pile triple superphosphate (ROP) was manufactured and cured, and then GTSP was made from the cured ROP. The State notes further that the new process eliminates the ROP step and the new GTSP product thus goes to storage in the final cured state. High 19 fugitive emissions are therefore eliminated in the storage area. 6-1 ------- 6.2 COMPLAINTS OF PHOSPHATE PLANT EMISSIONS AND SUGGESTED IMPROVEMENTS OF REGULATIONS Inquiries about complaints were made to representatives of EPA Regions 4, 5, 6, and 10, the states of Florida and North Carolina in Region 4, Illinois in Region 5, Texas and Louisiana in Region 6, and Idaho in Region 10. Responses indicated that public complaints applying to emissions from phosphate production facilities were limited to those 1-18 summarized below. 6.2.1 Florida Region 4 has had recent complaints about phosphate plants in central Florida: 1. Heavy emissions are alleged to occur at night. The complaints indicate that control equipment is not maintained satisfactorily and not operated continuously. 2. Excursions above allowable limitations of Florida regulations are alleged to affect blooms on orange trees. 3. Plant neighbors report heavy fugitive emissions from GTSP storage facilities. These emissions have not occurred during Region 4 1 22 visits, but Region 4 representatives have seen pictures of them. ' The complaints on the central Florida plants were received from a total of five individuals during 1974 to 1979. They included five 22 complaints from one individual in 1977. A Region 4 representative commented that there is currently no way of checking whether or when control equipment has been on or off. He suggested that NSPS should include continuous monitoring of controls 6-2 ------- to show when equipment has been in service and operated per design specifications, and when not in service. The following public comments recorded in Volume III of the Environmental Impact Statement (EIS) for the Central Florida Phosphate Industry also appear pertinent: 1. "With respect to fluoride and uranium, recovery of these two compounds, based solely on economic consideration ... is totally unacceptable in view of the documented damage caused by these compounds. "They should be removed from the waste water as a simple matter of health protection. If a profit results, so much the better for the industry. But placing such an emphasis on profit is neither justifiable 23 nor desirable from the public or environmental health standpoint." 2. "We suggest you include the information that Florida has the highest rate of lung cancer in the United States, the statistics weighted so incidences are not attributable to age. "It does not seem unreasonable to assume, based on studies by Goffman and others, that these lung cancers are related to the high levels of radioactivity in our region. Additionally, statistics readily available from the Department of MRS for last year show that Florida led 24. the nation in deaths from cancer." 3. "Without adequate, proper and timely monitoring by the federal and state environmental agencies, all proposed actions and existing rules, regulations and requirements are worthless and ineffective. "Therefore, we request that adequate surveillance programs be instigated with an appropriate funding level in order to accomplish those goals which should be of the highest priority."25 6-3 ------- Pertaining to the above comments on uranium and radioactivity, radioactive emissions from phosphate plants are being investigated by the Office of Radiation Programs, EPA, for possible regulation. The status of this investigation is reported in Section 3, "PRODUCTION PROCESS CHANGES." 6.2.2 Idaho A State of Idaho representative stated that farmers have complained about contamination of cattle feed by fluoride emissions from the Simplot plant, mentioned above. He also said that private legal action against the company was taken by the farmers, and that the company has since reduced the emissions. Since the aforementioned legal action was taken privately, the State was not involved, and has not retained a file on this matter. The representative of the State commented, however, that it is too soon to know whether the emission reduction has been 18 19 sufficient to stop harm to the cattle. ' 6.3 STATUS OF DEVELOPMENT OF FLUORIDE MONITORING Pertaining to the Region 4 suggestion that emission control systems be monitored, information was obtained on the current status of systems for monitoring fluoride emissions. A continuous monitor for fluoride emissions was recently evaluated by the EPA Environmental Sciences Research Laboratory, which furnished the following comment: The monitor is designed to detect HF gas. Fluoride particulate is not included in the measurement. Should any particulate get 6-4 ------- into the analyzer past the particulate thimble filter, the water- soluble portion of the particulate fluoride will be included in the 27 selective ion electrode measurement. Pertaining to the same investigation, the EPA Emission Measurement Branch commented: These instruments have been around since initial promulgation [of the phosphate plant NSPS] and have not proven to be adequate for continuously monitoring total fluoride emissions. The instrument is designed to measure only gaseous fluoride and in field appli- cations, has proven to be unreliable and produce erratic results. 28 John Nader's report also points out these problems. 6.4 STATE PLANS FOR REGULATING EXISTING PHOSPHATE PLANTS The status of state plans for controlling fluoride emissions from existing phosphate plants, as of October 1979, under Section lll(d) of the Clean Air Act, is summarized as follows: A. EPA Region 4 1. Negative declarations (no phosphate plants in state) were 29 received from Alabama, Kentucky, Tennessee, South Carolina, and Georgia. 2. North Carolina. Plan not yet submitted. State has one phosphate plant. State held a public hearing June 1978, and intends to submit lll(d) plan about December 1979. Requirements of plan are expected to be less stringent than lll(d) guidelines regarding granulated triple OQ superphosphate (GTSP) storage. 3. Mississippi. Plan not yet submitted. State has one phosphate plant. State requested extension to July 1979. 29 6-5 ------- 4. Florida. Plan has been submitted. It is less stringent than lll(d) guidelines. State must have public hearings before Region can act on plan. State plans to hold and certify hearings in FY 80. pQ State has 9 or 10 plants in 2 counties. B. EPA Region 6 on 1. Negative declaration received from New Mexico. plants. 2. Texas. Plan not submitted. State has several phosphate 30 3. Oklahoma. Plan not submitted. Region does not have 30 information on whether State has phosphate plants. 4. Arkansas. Draft plan submitted in July 1979. Goes to 30 hearing in November 1979. 5. Louisiana. Plan has been submitted. Plan generally met lll(d) guidelines except would allow emissions about 100 times guideline requirements for triple superphosphate (TSP). The reason given for allowing the higher TSP-plant emissions is that, because of hot weather, pond temperature is too high for more effective fluoride removal in scrubbers. The Region is considering suggesting that the State employ a contractor to study the claim that guidelines cannot be met, but has 30 not yet acted. C. EPA Region 10 1. Idaho. Plan not submitted. An extension to July 1979 had 31 been discussed, but was not formally requested. 6-6 ------- 2. Idaho 1s the only state 1n Region 10 that has phosphate plants. 32 D. No state plan has yet been approved by EPA. EPA guidelines for development of State emission standards are specified 1n the publication, "Final Guidelines Document: Control of Fluoride Emissions from Existing Phosphate Fertilizer Plants," U.S. EPA, Research Triangle Park, N.C., EPA-450/2-77-005, March 1977. 6-7 ------- 6.C, REFERENCES FOR SECTION 6. 1. Telecon. W. Herring, U.S. EPA OAQPS, to W. Aronson, U.S. EPA Region 4. April 17, 1979. Subject: Phosphate Industry in Region 4. 2. Telecon. W. Herring to G. McNeil!, U.S. EPA Region 4. April 18, 1979. Subject: Phosphate Industry in Region 4. 3. Telecon. J. Rom, U.S. EPA Region 4, to W. Herring. April 23, 1979. Subject: Phosphate Industry in Region 4. 4. Telecon. W. Herring to J. Hund, EPA Region 4. May 3, 1979. Subject: Phosphate Industry in Region 4. 5. Telecon. W. Herring to F. Collins, U.S. EPA Region 4. May 21, 1979. Subject: Phosphate Industry in Region 4. 6. Telecon. W. Herring to B. Varner, U.S. EPA Region 5. April 24, 1979. Subject: Phosphate Industry in Region 5. 7. Telecon. W. Herring to L. Szempruch, U.S. EPA Region 5. April 24, 1979. Subject: Phosphate Industry in Region 5. 8. Telecon. W. Herring to J. Hepola, U.S. EPA Region 6. May 21, 1979. Subject: Phosphate Industry in Region 6. 9. Telecon. W. Herring to S. Spruiell, U.S. EPA Region 6. June 5, 1979. Subject: Phosphate Industry in Region 6. 10. Telecon. W. Herring to J. Pfander, U.S. EPA Region 10. May 21, 1979. Subject: Phosphate Industry in Region 10. 11. Telecon. W. Herring to B. Swan, U.S. EPA Region 10. May 21, 1979. Subject: Phosphate Industry in Region 10. 12. Telecon. M. Johnson, U.S. EPA Region 10. May 22, 1979. Subject: Phosphate Industry in Region 10. 13. Telecon. W. Herring to J. Symes, State of Florida. April 23, 1979. Subject: Phosphate Industry in Florida. 14. Telecon. W. Herring to Derr Leonhardt, State of N.C. April 20, 1979. Subject: Phosphate Industry in N.C. 15. Telecon. W. Herring to J. Reed, State of Illinois. April 24, 1979. Subject: Phosphate Industry in Illinois. 16. Telecon. G. Wallin, State of Texas, to W. Herring. April 30, 1979. Subject: Phosphate Industry in Texas. 6-8 ------- 17. Telecon. W. Herring to 0. Tanner, State of Louisiana. May 1, 1979. Subject: Phosphate Industry in Louisiana. 18. Telecon. W. Herring to R. Stenner, State of Idaho. April 19, 1979. Subject: Phosphate Industry in Idaho. 19. Letter and attachments from R. Stenner, State of Idaho, to W. Herring. April 20, 1979. Subject: Phosphate Industry in Idaho. 20. U.S. Patent. Direct Granulation Process for Triple Superphosphate. Assignee: J.R. Simplot Co. No. 4,101,637. July 18, 1978. 21. Letter and attachments from J.F. Cochrane, J.R. Simplot Co., to D.P. Dubois, EPA Region 10. December 30, 1977. Subject: TSP Performance Test. 22. Telecon. W. Herring, U.S. EPA OAQPS, to W. Aronson, U.S. EPA Region 4. April 18, 1979. Subject: Phosphate Industry in Region 4. 23. Environmental Impact Statement, Central Florida Phosphate Industry, Vol. 3. U.S. EPA Region 4. Atlanta, GA. Publication No. EPA 904/9-78-026C. Nov. 1978. p. 1-63. 24. Reference 23. p. 1-46. 25. Reference 23. p. 1-222. 26. Telecon. W. Herring to R. Stenner, State of Idaho. April 25, 1979. Subject: Phosphate Industry in Idaho. 27. Memo from J.S. Nader, EPA Environmental Sciences Research Laboratory, to G.B. Crane, EPA Emission Standards and Engineering Division. June 25, 1979. Subject: Status of Development of Continuous Monitor for Fluoride Emissions. 28. Memo from E. McCarley, EPA Emission Measurement Branch, to G.B. Crane. June 26, 1979. Subject: Report, Continuous Monitoring System for Fluoride Emissions from Stationary Sources. 29. Telecon. W. Bishop, EPA Region 4, to W. Herring. October 10, 1979. Subject: State Plans for Existing Phosphate Plants. 30. Telecon. W. Herring to J. Stubberfield, EPA Region 6. October 9, 1979. Subject: State Plans for Existing Phosphate Plants. 31. Telecon. K. Lepic, EPA Region 10, to W. Herring. October 11, 1979. Subject: State Plans for Existing Phosphate Plants. 32. Discussion. W. Herring with R. Shell, EPA Control Programs Development Division. October 9, 1979. Subject: State Plans for Existing Phosphate Plants. 6-9 ------- 7. CONCLUSIONS 7.1 HEMIHYDRATE PROCESS FOR WET PROCESS PHOSPHORIC ACID (WPPA) The hemihydrate (hemidihydrate or HDH) process for WPPA production appears to have certain potential advantages for reducing fluoride air pollutant emissions, compared with the conventional dihydrate processes. Fluorides evolved in process gases are recovered as byproduct fluosilicic acid (H^SiFg) at an efficiency reported to be greater than 99%. (Possible fluoride recovery between 65 and 99% is reported for the dihydrate process). Also, P205 recovery is reported to exceed 98%, which is alleged to be greater than in the dihydrate process. Greater P«0c recovery would reduce the amount of phosphoric acid entering pond water. This reduction would tend to increase pond pH, and thereby further reduce fluoride emissions from ponds and improve scrubber efficiency. Gypsum produced in the HDH process is reported to be of sufficient purity to be used in building material. This indicates that the gypsum may be substantially free of radioactive constituents and saleable. This is in contrast to dihydrate-process gypsum, which has radioactive constituents, and therefore is not saleable and must be retained at the plant in gypsum piles. The HDH process therefore might reduce gypsum piles and the attendant radioactive emissions and fugitive dust generated in handling the piles. 7-1 ------- Possible disadvantages of the HDH process are that radioactive constituents removed from the gypsum would apparently go into the product WPPA or pond water. Also, recovery of valuable uranium from HDH product WPPA is reported less satisfactory than recovery from the more dilute WPPA produced in the dihydrate processes. The HDH process has been used commercially in Europe and Japan but not in the U.S. 7.2 SIMPLOT PROCESS FOR GRANULATED TRIPLE SUPERPHOSPHATE (GTSP) The new Simplot process, in which GTSP is made by treating lime- stone with WPPA, provides a means for reducing local fluoride emissions at the plant where it is applied. However, the fluoride emissions from the Simplot process added to those from the production of the WPPA, used in the Simplot process, might not be less than the corresponding total for the old process. Emission test results for the Simplot process (0.187 Ib F/ton P2°5) indicate that fluoride emissions comply with - but are not substantially less than - requirements of the present NSPS (0.20 Ib F/ton P2°5)- The Simplot process appears to reduce or eliminate fugitive emissions and evolved fluorides during storage by eliminating the curing period required when GTSP is made from run-of-pile triple superphosphate (ROP). 7.3 WET GRINDING OF PHOSPHATE ROCK Emissions of sulfur dioxide (S0«) and dust with radioactive content, that occur during drying, grinding, and transport of phosphate rock, can be substantially reduced by eliminating rock drying prior to grinding and chemical processing. 7-2 ------- Companies are converting from dry to wet grinding of phosphate rock to save energy. 7.4 SCRUBBING SYSTEMS, COOLING TOWERS, AND GYPSUM PONDS Scrubbing systems remain the principal means of controlling fluoride emissions from phosphate plants. A few modifications have been introduced since NSPS publication. 7.4.1 Fluoride Recovery Scrubbers that recover fluorides as saleable H2S1Fg have been incorporated at a substantial number of WPPA plants. These scrubbers are reported to recover between 65 and 99% of the gaseous fluorides emitted during acidulation and acid concentration. Fluoride removal by this means substantially reduces the fluorides emitted from plant pond systems or cooling towers. Costs of fluoride recovery as H2SiFg might exceed H2SiFg market prices, but not to an extent that precludes its commercial applicability for reducing emissions. 7.4.2 Cooling Towers Cooling towers are being used at a substantial number of plants in place of cooling ponds, or instead of the gypsum pond area required to cool water for scrubbers and barometric condensers. The cooling towers are used in closed-loop systems. Here, water from the scrubbers or barometric condensers goes to the cooling tower, where it is cooled - mainly by evaporation - to the required scrubbing temperature. The water is then returned to the scrubbers or barometric condensers, thus closing the cool ing-tower loop. Make-up water is added to the closed loop from the gypsum pond water overflow caused by rainfall, or from the plant storm water collection system. Four variations of this cooling- tower closed-loop system were identified in this investigation: 7-3 ------- 1. Cooling Tower without Fluoride Recovery and without Caustic Treatment In a WPPA plant system, process gases are scrubbed, without prior removal of fluorides as byproduct hLSiF,.. The resulting scrubbing water is cooled in a cooling tower, then returned to the scrubbers. The water in this closed loop is not caustic treated to increase its pH and limit fluoride emissions from the tower. Substantial fluoride emissions from the cooling tower probably occur in this system. 2. Cooling Tower with Fluoride Recovery but without Caustic Treatment In a WPPA plant system, process gases are first scrubbed in a scrubber designed to recover fluorides as byproduct hLSiFg. The gases are then scrubbed again in the plant barometric condensers. The water from the barometric condensers is cooled in a cooling tower and returned to those condensers. This water is not caustic treated to limit fluoride emissions from the tower. Fluoride emissions from the cooling tower would be less in this system than in the system described in paragraph 1, above. 3. Cooling Towers with Fluoride Recovery and Caustic Treatment A WPPA plant system that employs a cooling tower with fluoride recovery and caustic treatment is similar to that described in paragraph 2, above, except that the water in the cooling tower is treated with sodium hydroxide (NaOH) to increase its pH and limit fluoride emissions from the tower. Also, a slip stream of water from the tower is limed to precipitate the fluoride and regenerate the NaOH. Fluoride emissions 7-4 ------- from the cooling tower would be less in this system than in either of the systems described in paragraphs 1 or 2, above. 4. Cooling Tower with Nitrogen and Phosphorous Recovery and with Caustic Treatment In a diammounium phosphate (DAP) plant system, process gases are first scrubbed with dilute phosphoric acid to recover ammonia and phos- phorous compounds. To remove fluorides, the gases are then scrubbed again with water that is treated with NaOH in a cooling-tower closed- loop system. A slip stream of water from the tower is limed to preci- pitate the fluoride and regenerate the NaOH. Fluoride emissions from the cooling tower would be reduced to a very low rate in this system. 7.5 FLUORIDE EMISSION MONITORING A system that would satisfactorily monitor fluoride emissions has not been developed. However, the present NSPS for phosphate plants requires continuous monitoring of total pressure drop across the process scrubbing system. It is thus required that records be kept of when emission control systems are operated and not operated. 7.6 RADIOACTIVE EMISSIONS The health impact of radioactive emissions from phosphate plants is being studied by the EPA Office of Radiation Programs (ORP). Results of this study will determine whether those emissions will be listed as hazardous pollutants under Section 112 of the Clean Air Act. If these emissions are so listed, standards for their control will be developed by ORP, under Section 112. 7-5 ------- 1.1 AREAS FOR RESEARCH AND DEVELOPMENT As noted above, phosphate plant pond area and pond fluoride emissions may bo reduced by a WPPA plant system that Includes: fluoride recovery ar, byproduct H^MFg, and a closed-loop cooling tower system where scrubbing water \c> caustic treated and a slip stream 1s Hmed to recover the caustic and precipitate the fluorides. This system 1s used, however, at only one plant, and Its user reports difficulties, which that company 1s working to overcome. Furthermore, the developer of this system, Dr. Aaron Teller, advocates a different system for future application. Dr. Teller's preferred system would use a baghouse with continuous injection of a dry additive that would remove fluorides by adsorption and filtration. Advantages claimed for this dry system Include reduced pond area, reduced pond emissions, reduced plumbing costs, and reduced corrosion. Dr. Teller estimates pond cost at $2000 to $2500 per acre vear and plumbing costs at *1000 per foot of piping. Tl'f? dry Teller system has not yet been used at a commercial phosphate plant. Two such units are being furnished to the IMC New Wales, Fla., plant. These units will be applied to GTSP and DAP processes. They are expected to be on line by 1980. The dry continuous-Injection baghouse described above might provide an Improved emission control system for phosphate plants, compared with scrubbing systems now 1n use. 7.8 INDUSTRY GROWTH Some processes are being modified and there 1s some new construction *t phosphate plants 1n the U.S. Particularly notable are (1) conversions 7-6 ------- from processing dry phosphate rock to processing wet rock and (2) conversions from process cooling water ponds to cooling towers. These modifications and new construction do not apply to the affected facilities as defined In the present NSPS. The responses obtained In this Investigation to enquiries to EPA Regional offices and State agencies Indicate that the extent of new and modified source construction 1s small. Furthermore, there 1s negligible projected Industry growth to 1985. 7.9 REGULATION OF EXISTING SOURCES Only two states - Florida and Louisiana - have submitted formal Section lll(d) plans for controlling fluoride emissions from existing phosphate plants. Both of these plans have less stringent requirements than those of the EPA guidelines. One state - Arkansas - has submitted a Section 111(d) draft plan. No state plan has been approved by EPA. 7.10 REFERENCES FOR SECTION 7. 1. Memo from G.B. Crane, EPA Emission Standards and Engineering Division, to File. 5/15/79. Subject: Review of Phosphate Fertilizer NSPS: Conversation with Dr. Aaron Teller. 7-7 ------- 8. RECOMMENDATIONS 8.1 HEMIHYDRATE PROCESS FOR WET PROCESS PHOSPHORIC ACID (WPPA) The henrihydrate (HDH) process is not recommended for study as a possible basis for NSPS revision because there are no commercial HDH plants in the U.S. and because no data showing reduced emissions were found in this investigation. However, reported reduced energy consumption and reduced capital cost might lead to commercial use of this process in the U.S. In the event of commercial adoption, this matter should be reconsidered. 8.2 SIMPLOT PROCESS FOR GRANULATED TRIPLE SUPERPHOSPHATE (GTSP) The Simplot process is not recommended for study as a possible basis for NSPS revision because available data do not show that it effects a substantial reduction of process emissions below the present NSPS requirement. This process should, however, be included in any future study concerned with regulating evolved-fluoride or fugitive emissions from GTSP storage facilities. 8.3 WET GRINDING OF PHOSPHATE ROCK Wet grinding is not recommended for study as a possible basis for NSPS revision because the energy-saving advantage of this process modification is presently inducing companies to adopt it. 8.4 SCRUBBING SYSTEMS, COOLING TOWERS, AND GYPSUM PONDS 8.4.1 Fluoride Recovery Fluoride recovery, as byproduct H2SiFg, is recommended for inclusion if a study is undertaken to determine a basis for NSPS revision. 8.4.2 Cooling Towers Cooling towers are recommended for inclusion if a study is undertaken to determine a basis for NSPS revision. 8-1 ------- 8.5 RADIOACTIVE EMISSIONS Findings of the Office of Radiation Programs study of radioactive phosphate plant emissions should be obtained if a study is undertaken to determine a basis for NSPS revision. This information would be needed to show impacts of proposed NSPS revisions on radioactive emissions. 8.6 RECOMMENDATION ON NSPS REVISION STUDY In view of the low projected growth of this industry, a study to establish a basis for NSPS revision is not recommended now. Possible revision should be reconsidered in four years. 8-2 ------- TECHNICAL REPORT DATA (Please read Instructions on the reverse before completing) i. REPORT NO. EPA-450/3-79-038 3. RECIPIENT'S ACCESSION NO. 4. TITLE AND SUBTITLE Review of New Source Performance Standards for Phosphate Fertilizer Industry 5. REPORT DATE November 1979 6. PERFORMING ORGANIZATION CODE 7. AUTHOR(S) William Herring 8. PERFORMING ORGANIZATION REPORT NO. 9. PERFORMING ORGANIZATION NAME AND ADDRESS EPA, Office of Air Quality Planning and Standards Emission Standards and Engineering Division Research Triangle Park, N.C. 27711 10. PROGRAM ELEMENT NO. 11. CONTRACT/GRANT NO. 12. SPONSORING AGENCY NAME AND ADDRESS Same 13. TYPE OF REPORT AND PERIOD COVERED 14. SPONSORING AGENCY CODE EPA 200/04 15. SUPPLEMENTARY NOTES 16. ABSTRACT Since promulgation of new source performance standards (NSPS) for the "phosphate fertilizer industry, in 1975, commercial applications of a few new systems that reduce air pollution from phosphate plants have been reported. These include scrubbing system modifications that reduce the size of ponds used to cool process water, and reduce pond fluoride emissions. Also, a proprietary new process produces a stable granular triple superphosphate (GTSP) directly. This process reduces or eliminates the emissions of fluorides and fugitive particulate during the curing, storage period. The new GTSP process also eliminates the scrubbing of certain process gas, thus reducing required cooling pond area and pond fluoride emissions. Wet grinding of phosphate rock, in place of dry grinding, has been introduced, which eliminates dryers and reduces emissions that include particulate with radioactive content. NSPS revision, however, would probably not have significant impact now because the extent of new and modified source construction in this industry is small, and because industry growth projected to 1985 is negligible. The recommendation is, therefore, made not to develop NSPS revisions now, but to assess additional developments in this industry in four years, and then reconsider possible revision. 17. KEY WORDS AND DOCUMENT ANALYSIS DESCRIPTORS b.lDENTIFIERS/OPEN ENDED TERMS c. COS AT I Field/Group Air Pollution Phosphate Industry Fluorides Air Pollution Control Stationary Sources Particulate New Source Peformance Standards Industry Study 13B 18. DISTRIBUTION STATEMENT Unlimited 19. SECURITY CLASS (This Report) Unclassified 21. NO. OF PAGES 81 20. SECURITY CLASS (This page) Unclassified 22. PRICE EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE ------- |