United States Environmental Protection Agency Industrial Environmental Research Laboratory Cincinnati OH 4'5268 Research and Development EPA-600/S2-81-081 Mar. 1982 Project Summary Characterization of Boliden's Sulfide-Lime Precipitation Plant D. Bhattacharyya, C. Sund-Hagelberg, K. Schwitzgebel, G. M. Blythe, J. C. Terry, and F. B. Craig Sulfide precipitation is an effective process for the treatment of industrial wastes containing highly toxic heavy metals. The attractive features of the sulfide precipitation process are: attain- ment of high degree of metal removals over a broad pH range, effective pre- cipitation of certain metals (such as: As, Cu, Cd, Hg) even at very low pH, low detention time requirement in the reaction tank because'of the high reac- tivity of sulfides, and the feasibility of selective metal recovery. With sulfide precipitation, the high reactivity of sulfides (S2-, HS-, H2S) with heavy metal ions and the very low solubili- ties of the heavy metal sulfides over a broad pH range are features not found with the hydroxide precipitation pro- cesses. Sulfide precipitation processes to remove heavy metals have gained con- siderable importance. I1~61 Bhattachary- ya, et al.I34) have done extensive bench- scale sulfide precipitation work at the University of Kentucky. High degree of separation of heavy metal cations and arsenic from actual smelter efflu- ents was obtained with a combination lime-sulfide precipitation process. Dur- ing the second phase of the project a joint work (by University of Kentucky, Radian Corporation, and Boliden Metall Corporation, Sweden) was undertaken to obtain full-scale sulfide precipita- tion data from a unique system (to remove arsenic, heavy metals and flu- oride) recently developed by a Swed- ish nonferrous metal production (cop- per and lead smelter) company. The full-scale plant (200 mVhour capac- ity) was,put into operation in 1978 and was designed to precipitate (at pH 3-5) As, Zn, Cu, Pb, Cd, and Hg as sulfide for possible recycle to "roast- er," and remove fluoride separately by lime (at pH > 10) as CaF2 (uncontami- nated with heavy metals) for landfill. The of an investigation involving both full-scale results and laboratory-scale precipitation of metals (from nonfer- rous metal production operations) are presented. This Project Summary was devel- oped by EPA's Industrial Environmen- tal Research Laboratory. Cincinnati, OH, to announce key findings of the research project that is fully docu- mented in a separate report of the same title (see Project Report order- ing information at back). Experimental The full-scale data was obtained at Boliden Metall Corporation, Skellefte- hamn, Sweden The system was de- signed to treat process water, and rain and wash waters as a single waste stream to maximize containment of heavy metals at a reasonable cost. The process water was typically 30 to 40% of the total flow. Plant Description A schematic dia- gram of the process is shown in Figure 1. The wastewaters contained heavy metals and were normally acidic. The major constituents were arsenic (100- 500 mg/l) and zinc (25-200 mg/l). Cop- ------- NaOH Rain and Wash Water Polymer Lime Source: • Floor Washings • Arsenic Plant • Electrostatic Precipitator Washing Reaction Tanks (AszSs. HgS, CdS, ZnS, etc. Formation] * Samp/ing Points Process Water Source: • Central Gas Scrubber Water (Cu & Pb Smelter Gases Con- tainment SO?) • ZnO Plant Gas Washer • Scrubber from Cu Matte Tapping figure 1. Schematic diagram of the full-scale sulfide-lime treatment process. To Drumfilter (intermittent operation) Metal Sulfide Cake Fluoride Precipitation Discharge per, lead, mercury, and selenium were present in smaller amounts (a total of 40 mg/l). The incoming water was first partly neutralized (with NaOH) to pH 2.5-3.0, then sodium sulfide was added. The amount of reagent added was con- trolled by monitoring the pH, i.e., the addition of sodium sulfide was stopped when a predetermined pH value was reached. Sodium sulfide was added as a 1 5 percent solution to the first reaction tank, which had a 55 m3 capacity. This resulted in a wastewater residence time of between 20 and 30 minutes at flowratesof 110to170mVh Thethird reaction tank was also equipped to allow the addition of more sulfide. This was sometimes necessary in order to com- pensate for secondary reactions The wastewater-precipitate slurry was pumped from the third reaction tank to a thickener where flocculant was added to enhance sedimentation properties. The sludge from the thickener under- flow was further concentrated by using a belt drum filter. The filtrate was returned to the reaction tanks. The over- flow from the thickener was polished in one of two parallel multilayer filters. The effluent from the polishing step was fed to the fluoride treatment plant. A 10 percent lime slurry was added to adjust the pH to 10 or greater. The CaF2 slurry was pumped to a settling pond. The overflow from this settling pond was fed to an inter mediate holding pond for discharge to the smelter effluent receiver. The 15 percent Na2S solution and the 10 percent lime slurry were prepared within the plant. The 40 percent sodium hydroxide solution was purchased com- mercially. The mixing tanks were instru- mented with level controls. Occasiona ------- overflows from these mix tanks were directed to the exit stream of the CaF2 precipitation tank. Although, the full-scale plant had been in operationforapproximatelyone year, the plant was still considered to be in a startup mode, due to the opera- tional problems related to a higher than anticipated arsenic level in the waters treated. Process Evaluation. The full-scale process was evaluated in 1979. A test plan was devised to determine the heavy metal and arsenic removal of the overall process in three short tests (ST-01, ST- 02, ST-03), and to determine the per- formance of the individual process ves- sels in two more detailed or "long" tests (LT-01 and LT-02). The important sam- pling points are shown in Figure 1. For the short tests only the incoming streams (Points 1 and 2) and the effluent stream (Points 7 and 8) were sampled Laboratory Studies. Bench-scale studies were also conducted at the University of Kentucky to identify the effects of pH and sulfide dosage on arsenic and zinc separations. In the full-scale process considerable variations of reagent dos- age and pH values were observed. The effects of dissolved S02 (present in smelter effluents) on metal sulfide pre- cipitation were also established. Process Chemistry The extent of metal sulfide precipita- tion is expected to be a function of pH, type of metal, sulfide dosage, and other interfering ions (such as dissolved S02. SO2") that might be present. The follow- ing types of reactions take place in the sulfide reaction tanks: M2+ + S2" = M S (Solid) M = Hg, Cu, Cd, Pb, Zn With arsenic (III) the primary reactions are- As2O3 + 3 S2" + 3H20 ~ As2S3 (Solid) + 6 OH" If the pH is allowed to rise above 6, AsaS3 solubilizes as: 2 As2S3 (Solid)+ 2H20 = 3H+ + As3Si" + HAs02 HgS, As2S3, CuS, CdS, and PbS can be completely precipitated even at pH 2, whereas ZnS precipitation would be incomplete at pH < 5 Figure 2 shows \ I .c 0 -2 -4 -6 S-a -14 -16 CuzS 02 4 6 8 10 12 14 pH Figure 2. Calculated solubilities of metal sul fides as a function of pH. the theoretical solubilities (calculated with a, computer program which included all possible reactions) of various metal sulfides. Except arsenic, the solubilities of other metals decrease with an in- crease in pH values. Arsenic precipi- tates only in acidic pH. In the presence of dissolved sulfur dioxide, side reac- tions between sulfite (SO2, S03) and sulfide will also consume some Na2S reagent to form elemental sulfur and thiosulfate (H2S03 + 2H2S - 3S + 3H2); 4HSO3 + 2HS~ - 3S2O3" + 3H20). These reactions are particularly impor- tant for metals with higher solubility (such as, ZnS) in the acidic pH values Results and Discussions For the five full-scale test runs the combined (process water + wash water) inlet concentrations (in mg/l) of metals were As = 130-450, Cu = 3-5, Pb = 20-40, Cd = 3-16, Zn = 30-60, Hg = 2-4, Fe - 5-20, Sulfite = 600-1000, Sulfate = 1500-2000, and Fluoride = 90-130. The five tests represented typi- cal operation on five separate days, with no deliberate attempt to vary operating conditions Table 1 summarizes the oper- ating conditions for the five tests The sulfide precipitation pH was in the range of 3 7 to 4 8 and the CaF2 precipitation pH was in the range of 11.2 to 11 5 The approximate sodium sulfide addition rate (in mg/l of wastewater) ranged between 400 to 900 mg/l. This corresponds to Na2S dosage fluctuations of 0.8 to 3 times the theoretical stoichiometric dosages Figures 3 and 4 show the results of the two long-test runs at various sam- pling points. Point 7 corresponds to the sulfide precipitation effluent, whereas Point 8 corresponds to the effI uent from the entire treatment process The multi- layer filter effluent contained less than 2 mg/l suspended solids, hence the soluble and total concentrations at Point 7 were approximately the same The thickener solid removal efficiency was about 92%. Figures 3 and 4 show that during sulfide precipitation no signifi- cant improvement in soluble metal con- centrations is observed after reaction tank II. Hg, Cu, Cd, and Pb removal by sulfide precipitation was always excel- lent, whereas As an Zn separation was complete only after lime precipitation. The range of separations and treated effluent concentrations (for all five tests) from the overall sulfide-lime treatment process is shown in Table 2. The removal of all heavy metals (except As and Zn) was excellent (> 99%) even only with Na2S precipitation atlowpH Table 3 shows the As and Zn Table 1. Operating Data Summary of Heavy Metal Removal Tests Total Inlet Rate (m3/hr) Process Water Flow Rate (m3/hr) Runoff Water Flow Rate (m3/hrl Runoff Water pH Reaction Tank 1 pH Reaction Tank II pH Reaction Tank III pH Thickener Overflow pH Lime Treatment pH Test 1 ST-01 114 40 74 2.4 3.7 3.4 3.7 4.0 11.3 Test 2 ST-02 128 60 68 2.7 3.7 4.2 4.1 4.0 11.3 Test 3 LT-01 108 31 77 2.3 3.9 4.3 4.0 4.0 11.2 Test 4 ST-03 143 43 100 2.3 4.9 4.8 3.9 3.8 11.5 Test 5 LT-02 131 39 92 2.9 4.2 4.3 4.1 3.8 11.4 ------- woo I .o 2 0) -Q " \ \ \ Test Run: LJ-01 IBoliden Metal I) Sulfide Precipitation pH = 3 9 to 4.3 Precipitation pH= 11.2 Solid Symbols: Indicate Less Than The Concentration Shown 345678 Sampling Point Number Figure 3. Residual metal and fluoride concentrations at various sampling points (full-scale process) for test run LT-01 separations by the sulfide process (prior to lime precipitation step) were consid- erably lower. The poorer separations were due to the pH and sulfide dosage variations, and consumption of a por- tion of the sulfide reagent by 863 pres- ent in the inlet wastewater. The zero separation of zinc and poor separation of As for the ST-01 run was primarily due to the insufficient addition of NazS reagent. Several observations indicated the partial disappearance of sulfide from solution through a pathway otherthan heavy metal precipitation. These observations were based on the facts that: (a) no H2S gas loss from the solutions occurred, in spite of apparent stoichiometric over- doses at acidic pH values, (b) non- closure of sulfide material balance without the consideration of thiosu If ate and elemental sulfur formation; and (c) the reduction of sulfite concentration in the sulfide reaction tanks The inlet wastewater to the sulfide precipitation plant contained high concentration (600-1000 mg/l) of sulfite. Bhattachary- ya and Sun'71 have conducted extensive bench-scale stojdies with synthetic and Boliden Plant wastewaters to establish the effects of sulfite-sulfide side reac- tion, and pH variation on arsenic, zmc, and other heavy metals precipitation. Bench-scale studies conducted at the University of Kentucky showed that proper control of pH and sulfide reagent addition are necessary for effective pre- cipitation of arsenic and zinc. With arsenic precipitation the pH effect is more complicated because the reaction of arsenic and sulfide produces OI-T ions (by the reaction 2HAs02 + 3S2~ = As2S3(S) +60H~), thus instantaneous pH rise will occur unless maintained constant by acid addition. Studies with 1000 mg/l As (III) at 1.0 x Na2S dosage showed that at pH 3 (if not maintained constant) if pH was allowed to rise before adjusting to pH 3, the residual arsenic concentration was ]5 times higher (30 mg/l instead of residual As of 2 mg/l with constant pH). Figure 5 (conducted with actual Boliden waste- water) shows that arsenic precipitation is best at low pH, whereas Zn precipita- tion is best above pH 5 Hence a two- stage precipitation (with As2S3 removal in the first stage) would be most ideal The removal of heavy metals in the presence of high sulfite is required in various industrial wastewaters Exten- sive bench-scale studies conducted with As, Zn, and other heavy metals in the presence of 0-1000 mg/l S03 showed that at pH < 5, side reaction between sulfite and sulfide consumed a signifi- cant portion of the Na2S reagent partic- ularly during precipitation of metals of higher solubility (such as ZnS, NiS) Figure 6 shows the reduction of SO3(by sulfide reaction) concentration and the resultant formation of thiosulfate dur- ing ZnS precipitation at pH 3 Table 4 shows (experiments with single salts) that for metals with low solubility pro- duct (such as, CuS) M2+ -S2~ reaction predominates over S03 -S2~ reaction. Higher reaction pH also reduced the importance of sulfite-sulfide reaction. Results with actual Boliden wastes showed that if the reaction was carried out at pH 4 with 1 OX stoichiometric dosage Zn residuals can be reduced to 5-7 mg/l (instead of 15-36 mg/l Zn obtained with full-scale tests). The con- sumption of sulfide by sulfite was also verified in two special full-scale tests, which showed the formation of thiosul- fate and elemental sulfur in the reaction ------- 1000 400 200 till Test Hun: L T-02 (Boliden Metal I) Sulfide Precipitation pH = 3.8to4.2 ,_[L / imr> I Precipitation pH= 11.4 Fluoride F O As A In V Pb D Cd O Cu Hg Solid Symbols: Indicate Less Than The Concentration Shown 001 1+23 4 5 6 7 8 Samp/ing Point Number Figure 4. Residual metal and fluoride concentrations at various sampling points (full-scale process) for test run L T-02. Table 2. Overall Removal of Metals by Sulfide-Lime Precipitation Process (Full- Scale Process) Component As Cu Pb Cd Zn Hg F % Removal 96 to 99 >90 97 to 99 99 99 99 70- 78 Effluent Concentration, mg/l 1 8 to 4.1 <0.1 0. 1 to 02. 0.01 0.1 to 03 0.008-0.01 20 to 34 products. Laboratory results also showed that with excess Na2S dosages (>1 4X) the sulfide not consumed by metal pre- cipitation reacted with sulfite to form primarily elemental sulfur Summary and Conclusions The full-scale data provided the fol- lowing results, (a) the combined suIfide- hme precipitation process provided ex- cellent removals for all heavy metals and arsenic, (b) with sulfide precipita- tion (no lime treatment) process at pH 3 5-5 0 although Cu, Cd, Hg, and Pb removals were excellent (98-99%), ar- senic and zinc removals were not always satisfactory due to improper operating conditions andsuIf ite-suIf ide reactions; (c) excess sulfide dosage caused no H2S odor problem because of the presence of dissolved SOs in wastewater (excess sulfide reacts with sulfite to form thio- sulfate and sulfur), (d) anionic polymer provided excellent flocculation of sul- fide precipitates, and (e) sulfide sludge was easy to dewater, vacuum filtration provided 25-30% solids Laboratory studies (bench-scale) showed that proper control of pH (particularlyfor As precipi- tation) and sulfide reagent addition improved arsenicandzmc precipitation. The side reaction between sulfite-sulfide consumed a significant portion of the Na2S reagent (to form thiosulfate and sulfur) particularly during precipitation of metals of higher solubility (such as ZnS). References 1 "Control and Treatment Technol- ogy for the Metal Finishing Indus- try: Sulfide Precipitation," EPA Technology Transfer Report No. EPA-625/8-80-003 (April 1980). 2 Lantz, J. B., "Evaluation of a Developmental Heavy Metal Waste Treatment System," Naval Con- struction Battalion Center Report No. ER-314-40-6(1 979). 3. Bhattacharyya, D ,etal., "Separa- tion of Toxic Heavy Metals by Sul- fide Precipitation," Sep. Sci. and Technology. 14, 441 (1979) 4. Bhattacharyya, D , era/., "Sulfide Precipitation of Heavy Metalsfrom Non-Ferrous Metal Production Wastes," 51st Annual Conference - Water Pollution Control Federa- tion, California (1978). ------- Table 3. Arsenic and Zinc Removals by Sulfide Process (Full-Scale Process) Test Number LT-01 LT-02 ST-01 ST-02 ST-03 /Va25 Dosage Reaction pH ~2.7X* 3.9-4.3 ~3.7X* 4.7-4.3 ~0.8X* 3.4-3.7 ~2.0X* 3.7-4.2 =O.9X* 3.9-4.9 % Removal As Zn 78 82 67 87 78 51 75 0 69 55 *Even with excess Na^S no H2S gas was formed because of sulfite-sulfide reactions. t .0 «J c u o o N 0) cc 30 20 10 I 0 65 60 .O 50 40 O 30 c CD 20 10 Figure 5. Raw Feed 4 pH i NazS dosage = 1 .OX Q Boliden ST-03 Feed A Boliden LT-01 Feed D Boliden ST-01 Feed 3456 pH Residual arsenic and zinc concentrations after sulfide precipitation (bench-scale process). 5. Larson, H. P., and Ross, L. W., "Two-Stage Process Chemically Treats Mine Drainage to Remove Dissolved Metals," Operating Hand- book of Mineral Processing, 349 (February 1976). 6. Schlauch, R. M., and Epstein, A. C., "Treatment of Metal Finishing Wastes by Sulfide Precipitation," EPA-600/2-77-049 (1979). 7. Bhattacharyya, D., and Sun, G., "Precipitation of Heavy Metals and Arsenic with Sodium Sulfide: Sulfite- Sulfide Interaction," In Prepara- tion. (1980). 6 ------- 16 oo 5 a} 1 00 o 3 CO 0 c Q) O O CJ 5 Q) I System: \ Zn, = 500 mg/l (7.7 mM) (S03),= 1000 mg/1(12.5 mM) NazS Dosage =1.0X(7.7 mM) 10 15 20 Time, Minutes 25 30 Figure 6. Consumption of su/fite by sulfide during ZnS precipitation in the pres- ence of Na2SO3(bench-scale process). Table 4. Consumption of Sulfide by Sulfite During Metal- Sulfide Precipitation Initial metal concentration =100 mg/l Initial NaiSOyConcentration = 1000 mg/l NazS dosage = 1 .OX System Reaction PH Na2S03 As(lll)-Na2S- Na2S03 Na2S03 % Sulfide Consumed by Sulfite 3 4 3 33.6 12.8 2.1 D. Bhattacharyya is with the University of Kentucky, Lexington, KY; C. Sund- Hagelbergis with BolidenMetallCorporation, Sweden; K. Schwitzgebel, G. M. Blythe, and J. C. Terry are with Radian Corporation. Austin, TX; F. B. Craig (also the EPA Project Officer, see below) is with the Industrial Environmental Research Laboratory, Cincinnati, OH 45268. The complete report, entitled "Characterization of Boliden's Sulfide-Lime Pre- cipitation Plant," (Order No. PB 81 -209 2 72; Cost: $10.50 (subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Industrial Environmental Research Laboratory U.S. Environmental Protection Agency Cincinnati, OH 45268 S. GOVERNMENT PRINTING OFFICE: 1982/559 -092/338I ------- United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 Postage and Fees Paid Environmental Protection Agency EPA 335 Official Business Penalty for Private Use $300 RETURN POSTAGE GUARANTEED IL ------- |