Behaviors of Lead and Zinc in Top Submerged Lance (TSL) Plant at Sukpo Zinc Refinery

Size: px

Start display at page:

Download "Behaviors of Lead and Zinc in Top Submerged Lance (TSL) Plant at Sukpo Zinc Refinery"

Douglas Sparks
5 years ago
Views:

1 Materials Transactions, Vol. 53, No. 5 (212) pp. 985 to The Japan Institute of Metals EXPRESS REGULAR ARTICLE Behaviors of Lead and Zinc in Top Submerged Lance (TSL) Plant at Sukpo Zinc Refinery Byung-Su Kim 1,+, Soo-Buck Jeong 1, Jae-chun Lee 1, Doyun Shin 1 and Nam-Il Moon 2 1 Mineral Resources Research Division, Korea Institute of Geoscience and Mineral Resources, 124 Gwahang-no, Yuseong-gu, Daejeon 35-35, Korea 2 Sukpo Zinc Refinery Young Poong Cooperation, Bonghwa-gun, Kyoungbuk , Korea The Sukpo Zinc Refinery of Young Poong Cooperation (YPC) has recovered zinc and lead from zinc residue by using top submerged lance (TSL) technology since 27. This recycling operation, named the TSL Plant, was designed to operate 2 furnaces each equipped with one top submerged lance. The operation performance shows good metal recoveries. The plant is currently measuring oxygen pressure in its slag bath to respond to the varying slag bath conditions instantly. This paper describes the behaviors of lead and zinc acquired by determining the oxygen pressure in the slag bath and the results of measurements are compared with those of the thermodynamic calculations. [doi:1.232/matertrans.m2126] (Received January 6, 212; Accepted February 15, 212; Published April 25, 212) Keywords: zinc, lead, zinc residue, top submerged lance (TSL) technology, recycling 1. Introduction The Sukpo Zinc Refinery of Young Poong Cooperation (YPC) in Korea has produced 38, ton per year for zinc at the present. The zinc plant utilizes the roasting and conventional neutral leaching process to produce zinc from its concentrate. In the zinc hydrometallurgy process, iron in the leach solution is removed by the ph control. 1) Thus, around 24, ton per year of zinc residue that contains 15 2 mass% zinc and 3 5 mass% lead and so on is generated from the zinc process. Thus, the treatment of the zinc residue has become an important issue because of the environmental problems and the limited space available at the plant as well as the resource recovery perspective. With this in mind, Sukpo Zinc Refinery investigated available technologies for recovering valuable metals like lead and zinc from the zinc residue and finally, top submerged lance (TSL) technology was selected. The advantage of the process is to recover lead and zinc as an oxide form at the dust collector with producing an inert material met the limits of pollution restrictions. The operation is called the TSL Plant. The TSL Plant largely consists of smelting furnace and cleaning furnace with one top submerged lance, respectively. The smelting furnace is operated under reducing atmospheres at around 1553 K with supplying extra oxygen to easily melt the residue. Otherwise, the cleaning furnace is operated under the higher reducing atmospheres than the smelting furnace at around 1492 K without supplying extra oxygen to enhance zinc and lead recoveries. In present time, the plant has treated 24, ton per year of lead and zinc containing residue, from which about 38, ton per year for zinc and about 9, ton per year for lead has been produced. In the process, controlling the reducing atmosphere in the furnaces is one of the most important factors in determining the operation efficiency. In general, it was very well known that the atmosphere in furnace could be identified by + Corresponding author, bskim@kigam.re.kr measurement of the oxygen pressure by assaying the Fe 2+ and Fe 3+ concentrations in the slag bath. 2) During the early stages of the TSL operation in Sukpo Zinc Refinery, the oxygen pressure was measured by assaying the Fe 2+ and Fe 3+ concentrations in the slag bath. However, this method had some drawback in rapidity. Thus, to overcome the problem by instantaneous reading of the oxygen pressure, a portable oxygen pressure measurement system has been used in Sukpo Zinc Refinery. Therefore, this paper describes the behaviors of lead and zinc in the TSL Plant determined by measuring the oxygen pressures in the slag bath and the results of measurements are compared to those of thermodynamic calculations. 2. Description of the TSL Plant The TSL Plant adopts a continuous operation with two furnaces, i.e., a smelting furnace and a cleaning furnace. The inner diameters of the furnaces are 4.6 m and 4.6 m, respectively. The bath height of the both furnaces is 1 m. One top submerged lance is positioned at the center of each furnace and is fired with coal and stoichiometric oxygen. The oxygen is supplied by compressed air, pure oxygen, and afterburning air. Figure 1 shows the flowsheet of the plant. The raw materials are charged into the smelting furnace via hoppers, conveyors, and a port at the furnace top. The waste gas travels with the produced fume oxide through a waste heat boiler (W.H.B), a dust electrostatic precipitator (D.E.P), a ZnO scrubber, a waste gas fan (W.G. Fan), and a stack to air. The fume oxides are removed from W.H.B and D.E.P, and pumped with water to the scrubbers for capturing SO 2 in the waste gas. And, the molten slag is continuously supplied to the cleaning furnace through a launder. Reductant coal is only charged into the cleaning furnace through a feed port. The second furnace oxides are collected at a D.E.P to be sent to the existing leaching plant. Therefore, Sukpo Zinc Refinery has been used an oxygen pressure measurement device to measure the oxygen pressure and temperature simultaneously at the slag bath in the TSL

2 986 B.-S. Kim, S.-B. Jeong, J. Lee, D. Shin and N.-I. Moon Raw Materials Fuel Coal with Carrier Air Oxygen Combustion Air After Buming Air W.H.B. D.E.P. ZnO Scrubber Mist Remove Stack Item Table 2 Rate (kg/h) Material balance of the cleaning furnce. Zn Pb Cu Sb mass% kg mass% kg mass% kg mass% kg <Input> Conveyors Fume Oxide (I) W.G. Fan SFS 13, Fuel coal 9 Reductant Fuel Coal with Carrier Air Combustion Air After Buming Air To Heat recovery Lump coal 2 Total 14, <Output> Slag D.E.P Fume oxide Smelting Furnace W.G. Fan Cu speiss Fume Oxide * CFS 12, Total * CFS: Cleaning furnace slag. Fig. 1 Cleaning Furnace Discarded Slag Flowsheet of the TSL Plant at Sukpo Zinc Refinery. Reference electrode Thermocouple Mo lead wire Polyproplene connector Item Table 1 Rate (kg/h) Material balance of the smelting furnace. Zn Pb Cu Sb mass% kg mass% kg mass% kg mass% kg 4 mm Zinc residue <Input> 16, , Electrolyte Cap (Corrugated paper) Al 2 O 3 cement Mo lead wire Ceramic body (Housing) Fuel coal 4,93 14 mm Oxygen probe Lump coal 1,2 stopper Silica sand 2, Total 24,21 2, <Output> Fume oxide 3, , * SFS 13, m Probe fitting part 2 m Holder (steel) Fig. 2 Construction of oxygen pressure measurement device. Total 2, * SFS: Smelting furnace slag. Plant. This instantaneous measurement has enabled Sukpo Zinc Refinery to control the furnace operation promptly. 3. Results of Recent Operation Tables 1 and 2 show the material balances of a typical campaign operated in the TSL Plant at Sukpo Zinc Refinery. This campaign treated zinc residue currently produced from the conventional neutral leaching process. This Zinc Residue used during the campaign is from a material stockpile produced from the previous operation. All feed materials were charged into the furnace without feed preparation. However, to avoid difficulties in processing and charging the feed the moisture content of residue should not exceed 3%. Zinc, lead and antimony were recovered as fume oxide, copper recovered as speiss along with antimony. The recoveries of zinc, lead, antimony and copper were about 83, 93, 68 and 53% in the campaign, respectively. 4. Measurement of Oxygen Pressure in Slag Bath The measurement system of oxygen pressure currently used at the site is a portable device and consists of (1) a consumable oxygen probe used for only one measurement, (2) a probe holder, and (3) a voltage recorder, as shown in Fig. 2. The oxygen probe consists of a stabilized zirconiamagnesia electrolyte, a reference electrode made of molybdenum/molybdenum oxide as anode and molybdenum wire connectors for the anode and the cathode. The cell s configuration is expressed as:

3 Behaviors of Lead and Zinc in Top Submerged Lance (TSL) Plant at Sukpo Zinc Refinery 987 mv mv Temp Time, t / s Fig. 3 A typical recorder trace for the emf and temperature measurements of slag. Mo=O 2 ðmo MoO 2 Þ=ZrO 2 MgO=slag=Mo The probe also has a R-type (Pt Rh 13%/Pt) thermocouple. The probe mounted at the tip of holder was submerged 5 1 cm into a slag bath for 2 5 s to read the emf and bath temperature of slag simultaneously. Figure 3 shows a typical recorder traces for the emf and temperature measurements of slag using the probe. The emf was measured ten times at the smelting and cleaning furnaces of the TSL Plant, respectively. Both molybdenum and molybdenum oxide are stable as solids under the operating temperature, so their activities in slag are 1. Therefore, using the equilibrium oxygen pressure (P ; Mo MoO 2 ) over the reference electrode, the oxygen pressure (P ; in slag) in a slag bath can be calculated by the measured emf and the standard Gibbs free energy of reference electrode, which is represented by 3) RT ln P ; in slag ¼ 4FE þ RT ln P ; Mo MoO 2 : Here, RT ln P ; Mo MoO 2 can be calculated from the standard Gibbs free energy (G Mo MoO 2 ) of the reaction (3). Mo þ O 2 ðgþ ¼MoO 2 ð3þ In this investigation the following equation was used for the above reaction (3). G Mo MoO 2 ¼ 58;2 þ 169:66T ðj=molþ 4Þ ð4þ Thus, eq. (5) could be induced from eqs. (2) and (4), which can be used to calculate the oxygen pressure in slag bath. ln P ; in slag ¼ ð4ef þ G Mo MoO 2 Þ ð5þ RT Here, E represents the emf (V), F is Faraday s constant (C), T is the temperature of slag bath (K), and R is the gas constant. Oxygen pressures in slag bath were also measured through assaying Fe 2+ and Fe 3+ concentrations in the slag bath for comparing those measured by the oxygen probe. Reaction (6) is an equilibrium reaction between FeO and FeO 1.5 in a slag bath, and eq. (7) is the standard Gibbs free energy of reaction (6). Temperature, T / K ð1þ ð2þ 4ðFeOÞ slag þ O 2 ðgþ ¼4ðFeO 1:5 Þ slag ð6þ G FeO Fe 2 O 3 ¼ 57;3 þ 212:T ðj=molþ 4Þ ¼ RT lnða 4 FeO 1:5 =a 4 FeO P ; in slagþ ð7þ Here, a FeO and a FeO1:5 are the activities of FeO and FeO 1.5. So, eq. (8) could be induced from eq. (7) when the activity coefficients of FeO 1.5 and FeO in a slag bath are assumed to be equal in a certain temperature range, which can be used to calculate the oxygen pressure in the slag bath. Thus, through assaying Fe 2+ and Fe 3+ concentrations in a slag sample taken from a certain slag bath the oxygen pressure in the bath can be calculated, which is a simple way to determine the oxygen pressure in the bath without using any oxygen probe device. ln P ; in slag ¼ G FeO FeO 1:5 þ 4ln mass% of Fe3þ in slag RT mass% of Fe 2þ ð8þ in slag On the other hand, slag samples were taken to be analyzed chemically when the oxygen probe device dipped into the slag bath. The total iron and divalent iron contents in the slag specimens were determined by titration with K 2 Cr 2 O 7, the SiO 2 content by a gravimetric analysis method and other components in the slag and metal phases were analyzed by the inductively coupled plasma (ICP) method (JY-38 plus, Horiba Ltd., Kyoto, Japan). All the oxygen pressures measured must be compared at the same temperature because oxygen pressure is dependent on temperature. However, it is impossible to keep the experimental temperature constant in the actual furnaces. Thus, the measured oxygen pressures were all normalized to 1523 K according to the equilibrium reaction of FeO and FeO 1.5 in the slag bath. Kemori etc. have confirmed that oxygen pressures in the slag of flash furnace are regulated by the equilibrium reaction of FeO and FeO 1.5 which is represented as the reaction (6). 3) Thus, the measured oxygen pressure (p ; in slag) at the temperature of T 1 can be normalized to that measured at the temperature of T 2 by eq. (9), assuming that the slag bath is regular solution, i.e., the activity ratio of FeO and FeO 1.5 is not influenced by a small temperature change from T 1 to T 2. ln P ;in slag at T 2 ¼ ln P ; in slag at T 1 þ G Fe FeO 1:5 at T 2 RT 2 G Fe FeO 1:5 at T 1 ð9þ RT 1 Table 3 shows the chemical compositions of the corresponding slag samples, and Table 4 does the measured and normalized oxygen pressures at each slag bath. 5. Interpretation of Results To interpret the behaviors of lead and zinc in the TSL Plant the contents of zinc and lead in slag bath were calculated using the oxygen pressures measured by the oxygen probe at the slag bath and assaying the Fe 2+ and Fe 3+ concentrations in it. Reaction (1) is an equilibrium reaction between metal lead and lead oxide in a slag bath, and eq. (11) is the standard Gibbs free energy of reaction (1).

4 988 B.-S. Kim, S.-B. Jeong, J. Lee, D. Shin and N.-I. Moon Table 3 Slag compositions in the smelting and cleaning furnaces (mass%). Sample Zn Pb T.Fe Fe 2+ Fe 3+ SiO 2 CaO Al 2 O 3 MgO Smelting furnace slag Cleaning furnace slag Table 4 Measured and normalized oxygen pressures of slags. Sample *2 SFS *3 CFS Temp. (K) By assaying Fe 2+ and Fe 3+ concentrations Fe 2+ Fe 3+ *1 EMF (mv) By the oxygen probe * ¹7.51 ¹ ¹7.92 ¹ ¹6.69 ¹ ¹6.6 ¹ ¹6.95 ¹ ¹7.42 ¹ ¹7.4 ¹ ¹7.28 ¹ ¹7.55 ¹ ¹8.5 ¹ ¹7.5 ¹ ¹7.72 ¹ ¹7.2 ¹ ¹6.98 ¹ ¹7.3 ¹ ¹6.78 ¹ ¹8.63 ¹8.4 6 ¹8.49 ¹ ¹8.54 ¹ ¹7.52 ¹ ¹11.67 ¹ ¹8.55 ¹ ¹9.95 ¹ ¹8.47 ¹ ¹9.85 ¹ ¹9.13 ¹ ¹8.72 ¹ ¹8.3 ¹ ¹9.66 ¹9.7 4 ¹9.15 ¹ ¹1.4 ¹ ¹9.15 ¹ ¹1 ¹ ¹9.9 ¹ ¹11.2 ¹ ¹9.29 ¹ ¹11.6 ¹ ¹9.16 ¹ ¹12.4 ¹ ¹9.15 ¹8.38 *1 Oxygen pressure normalized at 1523 K. *2 SFS: Smelting furnace slag. *3 CFS: Cleaning furnace slag.

5 Behaviors of Lead and Zinc in Top Submerged Lance (TSL) Plant at Sukpo Zinc Refinery 989 Pb content in slag (mass %) Calculated by equ. (12) Experimental data , in slag / kpa Fig. 4 Relationship between oxygen pressure and lead content at slag in the smelting furnace at 1523 K. 2Pb þ O 2 ðgþ ¼2ðPbOÞ slag G Pb PbO ¼ 39;2 þ 155:4T ðj=molþ 4Þ ¼ RT lnða 2 PbO=a 2 Pb P ; in slagþ ð1þ ð11þ So, eq. (12) can be obtained by rearranging eq. (11) with mass% Pb in the slag when the activity of metal lead, a Pb was assumed to unity. ln mass% of Pb in slag ¼ 1 2 ln P O 2 ; in slag þ ln n t; in slag þ ln M Pb ln PbO G Pb PbO 2RT ð12þ Here, G Pb PbO is the standard Gibbs free energy of the reaction (1) (J/mol), a PbO is the activity of PbO, M Pb is the mole weight of Pb (g), PbO is the activity coefficient of PbO, and n t is the total moles per 1 g slag. In the study, PbO was assumed to 1.25 by the previous papers. 5,6) Also, the total moles were analyzed to be 1.42 in the smelting furnace and 1.48 in the cleaning furnace, which were obtained by averaging the total moles per 1 g slag for all the campaigns tested, respectively. Thus, the lead content in a slag bath can be calculated by eq. (12) using the above data with the oxygen pressures measured as explained previously. Figure 4 shows the relationships between the oxygen pressures and lead contents at the slag in the smelting furnace at 1523 K, and Fig. 5 does the relationships between the oxygen pressures and lead contents at the slag in the cleaning furnace at 1523 K. The solid curves in these figures were the calculated based on eq. (12). Also, the open and dark circles present the experimental data obtained by the oxygen probe and assaying Fe 2+ and Fe 3+ concentrations. Examination of these figures reveals that the lead content in the slag of the smelting furnace relatively agrees well with the oxygen pressures calculated based on eqs. (5) and (8), which indicates that the both oxygen pressure measurements could be reliable for estimating the lead content in the slag of the smelting furnace. However, it is seen in the cleaning furnace that the oxygen pressure measured by assaying Fe 2+ and concentrations is relatively much lower than that measured by the oxygen probe. The difference in the oxygen pressure might be because metallic iron in the slag of the clean furnace is produced. Also, the difference in the oxygen pressure might be due to some errors in the procedure of slag sampling and analysis to measure the oxygen pressure by assaying Fe 2+ and Fe 3+ concentrations in the slag since the oxygen pressure is sensitively affected by the Fe 2+ and Fe 3+ concentrations at the range of low oxygen pressure of below ¹9 kpa (1 ¹11 atm). However, the detail reason was not clearly known in this work. Also, the figures show that the lead content in the slag of the smelting and cleaning furnaces relatively agrees well with the oxygen pressures calculated based on eqs. (5) and (12). The zinc content in slag also can be calculated by eq. (15) which can be obtained by rearranging eq. (14) since the equilibrium vapor pressure of metal zinc is relatively high at the operation temperature of the TSL Plant. Equation (13) is an equilibrium reaction between metal zinc in gas phase and zinc oxide in a slag bath, and eq. (14) is the standard Gibbs free energy of reaction (13). 2ZnðgÞþO 2 ðgþ ¼2ðZnOÞ slag ð13þ G Zn ZnO ¼ 94;89 þ 379:36T ðj=molþ 4Þ Fe 3+ Pb content in slag (mass %) Calculated by equ. (12) Experimenta data ¼ RT lnða 2 ZnO=P 2 Zn P ; in slagþ ln mass% of Zn in slag, in slag / kpa Fig. 5 Relationship between oxygen pressure and lead content at slag in the cleaning furnace at 1523 K. ¼ ln P Zn þ 1 2 ln P O 2 ; in slag þ ln n t; in slag ð14þ þ ln M Zn ln ZnO G Zn ZnO ð15þ 2RT Here, G Zn ZnO is the standard Gibbs free energy of the reaction (13) (J/mol), a ZnO is the activity of ZnO in the slag, P Zn is the partial pressure of zinc (atm) that is assumed to be.7 (7.9 kpa) by the previous paper, 7) M Zn is the mole weight of Zn (g), and ZnO is the activity coefficient of ZnO that is assumed to be ,8) And, the total moles (n t ) are 1.42 in the smelting furnace and 1.48 in the cleaning furnace as explained previously. By using the above data and the measured oxygen pressures, the zinc content in a slag bath can be calculated by eq. (15).

6 99 B.-S. Kim, S.-B. Jeong, J. Lee, D. Shin and N.-I. Moon Zn content in slag (mass %) Calculated by equ. (15) Experimental data , in slag / kpa Fig. 6 Relationship between oxygen pressure and zinc content at slag in the smelting furnace at 1523 K. Zn content in slag (mass %) Calculated by equ. (15) Experimental data , in slag / kpa Fig. 7 Relationship between oxygen pressure and zinc content at slag in the cleaning furnace at 1523 K. Figures 6 and 7 present the relationships between the oxygen pressure and zinc content at the slags in the smelting and cleaning furnaces at 1523 K. The solid curves in these figures were the calculated based on eq. (15). The open and dark circles also present the experimental data obtained by the oxygen probe and assaying Fe 2+ and Fe 3+ concentrations. As shown in these figures, the zinc contents in the slag of the smelting furnace relatively agree well with the oxygen pressures calculated based on eqs. (5) and (8). However, it is seen in the cleaning furnace that the oxygen pressure measured by assaying Fe 2+ and Fe 3+ concentrations is relatively much lower than that measured by the oxygen probe, as explained previously. The phenomenon means that the method to measure oxygen pressures by assaying Fe 2+ and Fe 3+ concentrations at the slag bath is only useful to interpret the behavior of zinc in the smelting furnace. Also, the figures show that the zinc content in the slag of the smelting and cleaning furnaces relatively agrees well with the oxygen pressures calculated based on eqs. (5) and (15). Based on these observations, it was thus thought that in the slag bath of the TSL Plant at Sukpo Zinc Refinery, the oxygen pressure measured by assaying Fe 2+ and Fe 3+ concentrations at the slag bath is only useful for responding to the varying slag bath of the smelting furnace. However, a portable oxygen pressure measurement system enabled YPC operators to respond promptly to control the both furnace operations. 6. Conclusion The TSL Plant at Sukpo Zinc Refinery is one of the best processes to treat the lead and zinc containing residue generated from the zinc hydrometallurgy process, and for producing clean and inert waste materials. It was investigated in the study that the behaviors of lead and zinc in the TSL Plant determined by measuring the oxygen pressures in the slag bath and the results of measurements are compared to those of thermodynamic calculations. From the investigation, it was found that the oxygen pressure measured by the oxygen probe is useful for responding to the varying slag bath of smelting and cleaning furnaces. However, the method to measure oxygen pressures by assaying Fe 2+ and Fe 3+ concentrations at the slag bath was only useful to interpret the behaviors of lead and zinc in the smelting furnace and had a drawback in rapidity. Also, by using an portable oxygen probe that has a Mo/MoO 2 electrode, reliable measurements of the oxygen pressure in the slag bath of TSL Plant could be obtained rapidly. These rapid measurements enabled YPC operators to respond promptly to control the furnace operation. Acknowledgements This study was supported by the R&D Center for Valuable Recycling (Global-Top Environmental Technology Development Program) funded by the Ministry of Environment. (Project No.: 11-C2-IR) REFERENCES 1) W. A. Jankola and H. Salomon-de-Friedberg: Iron control Technologies, Third Int. Symp. on Iron Control in Hydrometallurgy, ed. by J. E. Dutrizac and P. A. Riveros, (CIM, Montreal, Quebec, 26) pp ) Y. Lee, N. Moon and C. Choi: Proc. 2nd Int. Symp. on Quality in Non- Ferrous Pyrometallurgy, ed. by M. A. Kozlowski, S. A. Argyropoulos and R. W. McBean, (CIM, Montreal, Quebec, 1995) pp ) N. Kemori, H. Kurokawa and Z. Kozuka: J. Min. Metall. Inst. Japan 12 (1986) ) A. Yazawa: Nonferrous Metallurgy, (The Japan Institute of Metals, Sendai, Japan, 198) pp (in Japanese). 5) N. Moon, M. Hino, Y. Lee and K. Itagaki: J. Min. Metall. Inst. Japan 113 (1997) ) N. Moon, M. Hino, Y. Lee and K. Itagaki: Proc. 5th Int. Conf. on Molten Slags, Fluxes and Salts 97, ed. by ISS, (Iron & Steel Society, Warrendale, PA, 1997) pp ) Y. Takeda: Zinc & Lead 95, An International Symposium on the Extraction and Application of Zinc and Lead, ed. by T. Azakami, N. Masuko, J. E. Dutrizac and E. Ozberk, (MMIJ, Tokyo, 1995) pp ) S. Surapunt, Y. Takeda and K. Itagaki: J. Min. Mater. Process. Inst. Japan 111 (1995)

Kinetics of the Volatilization Removal of Zinc from Manganese Dust

Materials Transactions, Vol. 51, No. 7 (21) pp. 1313 to 1318 #21 The Japan Institute of Metals EXPRESS REGULAR ARTICLE Kinetics of the Volatilization Removal of Zinc from Manganese Dust Byung-Su Kim 1;