sign in sign up
<<
Home , Glauberite, Hanksite, Nahcolite, Thermonatrite, Trona

Precipitation Sequence During Evaporation of Lake Katwe Brine Joseph Ddumba Lwanyaga1,2, Hillary Kasedde2, John Baptist Kirabira2

1Department of and Water Resources Engineering, Faculty of Engineering, Busitema University, P.O. Box, 236, Tororo, Uganda. 2School of Engineering, College of Engineering, Design, Art, and Technology, Makerere University, P.O.Box, 7062, Kampala, Uganda.

Abstract Traditional salt mining techniques are the dominant salt extraction methods at Lake Katwe in Uganda. In this work, the effect of temperature on the mineral precipitation sequence during evaporation of lake brine was studied to foster the design of a commercial salt extraction process. Isothermal evaporation experiments at different temperatures were undertaken. PHREEQC was used to predict the precipitation sequence. The precipitates were characterised by the XRD and SEM techniques. is the most abundant mineral, with , and Glaserite following, respectively. Halite will be best extracted at higher temperatures whereas the lower temepratures will favour sulphate based . Keywords: Lake Katwe, Evaporation Temperature, PHREEQC, Precipitation, Brine

Introduction A salt plant was established for industrial In Uganda today, salt for both domestic processing of the Lake Katwe brine. The plant and industrial use is mainly imported from never perfomed to its design expectations neighouring countries. The salt import bill is and the components were heavily corroded steadily increasing as the country develops (Driver and Tukahirwa 1990). This led to its which is increasingly straining the country’s abandonment without ever producing any merger resource envelope. To avert this, the meaningful products. Plans for commercial salt deposit at Lake Katwe need to be fully exploitation of the brine commenced with the exploited. This salt lake has been a source characterisation of brine and phase chemistry of salt for a long time. However, the salt experiments. The brines are highly alkaline + + - 2- 2- produced from the lake does not meet the and rich in Na , K , Cl , SO4 , CO3 , and required purity standards for table salt and HCO3- showing an intermediate transition the chlor-alkali industry(Kirabira et al. 2013; between Na-Cl and Na-HCO3 water types Kasedde et al. 2014). (Arad and Morton 1969; Kasedde et al. 2014). Currently at the lake, salt is recovered An isothermal laboratory evaporation by traditional means; brine is channelled experiment at 30°C revealed precipitation from the lake into salt pans. In the salt pan, of commercial salts such as Thenardite the salt crystallizes due to solar evaporation (Na2SO4), (CaSO4), producing two salt grades of different (Na2SO4·10H2O), Burkeite (Na2CO3·2Na2SO4), qualities. Grade I which is averagely about 93% (9Na2SO4·2Na2CO3·KCl), , Halite (NaCl) is the best these rudimentally Trona (Na2CO3·NaHCO3·2H2O), Halite, techniques can produce. Grade II is harvested (NaHCO3), Soda ash (Na2CO3), after Grade I with an average composition of and (Na2CO3·H2O) from the about 65% Halite. The third grade is obtained brine of Lake Katwe. Moreover, the mineral by hacking the lake bed with iron bars; this salts crystallize in the order following the has a composition of about 50% halite. Due sequence starting with , followed to their composition, these three salts attract by chlorides and carbonates, respectively low prices as the demand for them is low (Kasedde et al. 2013) with a rich source (Kasedde et al. 2014). of mineral salts. The present work aims at

Wolkersdorfer, Ch.; Khayrulina, E.; Polyakova, S.; Bogush, A. (Editors) 515 IMWA 2019 “Mine Water: Technological and Ecological Challenges” evaluating possibilities of future salt ex- membrane. The filtrate was then poured traction from the lake deposit. An isothermal into a 50 ml beaker which was then placed evaporation experiment was conducted on in a Labcon thermostatic water bath at 30°C. the lake brines. The precipitated salts were The formed precipitates were filtered off characterized by X-ray diffraction (XRD. using a 5 µm membrane and then left to dry Fractional crsytallisation is often times overnight in open air on an aluminium foil. used to extract most of the aforementioned This procedure was repeated for the other salts from brine. This, therefore necessitates study temperatures (40, 50, 60 & 70°C). investigating the precipitation sequences of minerals from Katwe brine at different Characterization of the precipitates temparatures in order to facilitate the design To determine the of the a commercial salt extraction process. precipitates, the powder X-ray diffraction method was used. A Phillips diffractometer Methods and Materials PW3050 with a X-ray source of Cu Kα Labarotory experiments radiation (λ = 1.540598 Å) was used. The The rainy season surface brine was sampled scan step size was 0.013°., the collection from Lake Katwe and stored in 1.5 L time 1s, and in the range 2θ CuKα from 1° plastic bottle at room temperature prior to 90.5°. The X-ray tube voltage and current to the evaporation study. Using 10 ml of were fixed at 45 kV and 40 mA respectively. the original brine from Lake Katwe, the The Crystallography Open Database (COD) polythermal evaporation experiment was powder diffraction database was searched run at five different temperatures (30, 40, 50, by QUALX2.0 search–match identification 60 & 70°C). These were chosen because they software for phase identification (Altomare et are easily achievable on a commercial scale al. 2015). The morphology of the precipitates with less material selection problems and was determined by a Zeiss Supra 55 VP Field the evaporation process is faster compared Emission Scanning Electron Microscope to lower temperatures. Before evaporation (FEG-SEM); all samples were coated was started, both the biological and organic before SEM analysis in order to make them materials were filtered off using a 5 µm electrically conductive.

Table 1 Brine Composition

pH Brine Density EC Major (g/L) Temp (g/ml) (us/m) Na K Mg Ca Cl Br SO HCO CO (°C) 4 3 3 9.72 25.2 1.15 14109 69.6 11.6 0.0519 0.0048 44.2 0.461 32.7 18.2 38.8

Figure 1 XRD Pattern of mineral precipitates (Gl-Glaserite, Th-Thenardite, Ha-Halite, Tr-Trona, B-Burkeite, Na-, Tn-Thermonatrite, H-Hanksite).

516 Wolkersdorfer, Ch.; Khayrulina, E.; Polyakova, S.; Bogush, A. (Editors) IMWA 2019 “Mine Water: Technological and Ecological Challenges”

Table 2 The mineral composition of the precipitates at the different study temperatures

Mineral Phases Temp Sample 1 Sample 2 Sample 3 Sample 4 °C 30 Calcite, Magnesite, Tychite, Calcite, Calcite, Anhydrite, Anhydrite, Dolomite, Halite Glaserite Thermonatrite, Magnesium Sulphate, Thenardite, Burkeite, Burkeite, Thenardite, Burkeite, Thermonatrite, Hanksite, Trona, Natrite, Halte Halite, Glaserite, Sylvite Halite, Glaserite, Sylvite 40 Calcite, Anhydrite, Calcite, Magnesite, Magnesite, Dolomite, Burkeite, Hanksite, Halite, Magnesite, Magnesium Halite Thenardite, Burkeite, Glauberite, Sylvite Sulphate, Thenardite, Trona, Thermonatrite, Halite, Natron, Halite, Glaserite, Glauberite, Glaserite, Sylvite Arcanite, Sylvite 50 Anhydrite, Magnesite, Calcite, Magnesite, Calcite, Magnesite, Calcite, Anhydrite, Dolomite, Dolomite, Trona, Thenardite, Trona, Dolomite, , Thenardite, Burkeite, Thermonatrite, Halite Halite Burkeite, Hanksite, Trona, Hanksite, Nahcolite, Thermonatrite, Natrite, Thermonatrite, Halite, Halite, Glauberite Glaserite, Sylvite 60 Calcite, Magnesite, Halite, Magnesite, Dolomite, Calcite, Magnesite, Anhydrite, Dolomite, sylvite Halite Dolomite, Trona, Natrite Burkeite, Hanksite, Natron, Halite, Sylvite 70 Calcite, Magnesite, Calcite, Magnesite, Calcite, Anhydrite, Anhydrite, Magnesite, Magnesium Sulphate, Dolomite, Trona, Magnesite, Dolomite, Halite, Dolomite, Gaylussite, Dolomite, Thenardite, Halite, Halite, Sylvite Sylvite Thenardite, Natrite, Halite, Thermonatrite, Glaserite, Glauberite, Sylvite Sylvite

Thermodynamic modelling the carbonates and sulphates of and To model the precipitation sequence during magnesium (Calcite, Tychite, and Dolomite) the evaporation of the Lake Katwe brine, in the first precipitates as predicted by Arad PHREEQC code (Parkhurst and Appelo and Morton (Arad and Morton 1969). This is 2013) version 3.4.0 was used. Due to the high attributed to their low within the concentration and ionic strength of brine, study temperature range. the geochemical calculations were done Table 2 shows the mineral composition using the Pitzer interaction approach of the precipitates at the different study with a modified version of the pitzer.dat temperatures. Four samples are analysed at thermodynamic database distributed with each temperature and these were collected the PHREEQC code. Evaporation was as soon as appeared. At all the study simulated in steps, 51 moles of water in temperatures, Halite emerges as the most 102 steps was removed (92% of the initial abundant mineral as the phase is observed water was evaporated). The sequence of the in almost all the samples. This is also evident modelled mineral precipitation at the study by the dominance of its microstructures in temperatures during the evaporation was Fig.2. noted. The major ions in Lake Katwe brine At 30°C, apart from Halite, Trona is as shown in Table 1 were used as PHREEQC the most abundant mineral followed by input (Kasedde et al. 2014). Thenardite and then Natrite. Halite, Burkeite, Glaserite and then Hanksite is the order Results and Discussion of prominence at 40°C with Thenardite Mineralogical analysis appearing as a trace. The carbonate and bicarbonate minerals dominate the Fig.1 illustrates a representative X-ray precipitates at 50°C with traces of Arcanite diffractogram of evaporation precipitates. and Thenardite. For 60°C as well, the X-ray diffraction results show a dominance of

Wolkersdorfer, Ch.; Khayrulina, E.; Polyakova, S.; Bogush, A. (Editors) 517 IMWA 2019 “Mine Water: Technological and Ecological Challenges” and bicarbonate minerals the latter needs to be minimised if pure salts dominate the precipitates with the sulphates for industrial use are to be produced. Trona and chlorides of sodium precipitating as is observed as blade like (Kasedde et al. Hanksite. Natrite and Hanksite are observed 2014) and columnar massive (Warren 2016) to be the most dominant phases at 70°C. The crystals (Fig. 2 (50°(A)), on top of Thenardite XRD results are supplemented by the SEM (salt cake) (Fig. 2, (70°(C)) and sandwiching results. Halite crystals (Fig. 2 (40°(A), 50°(B) & 50°(C)). Hexagonal prismatic crystals of Microstructual Analysis Hanksite are also observed as one of the Fig. 2 shows the SEM micrographs of the precipitating mineral phases (Fig. 2, (40°(B)) precipitates observed during the brine (Kasedde et al. 2014; Warren 2016). Similar evaporation. Four samples were filterd off microstructures were identified by Kasedde at each study temeperature with Sample and co-workers (Kasedde et al. 2013) in an A first, followed by B and lastly C. The isothermal evaporation study. SEM micrographs reveal distinctive As temperature increases, the mineral microstructures of the dominant mineral phase stratifications become more distinct phases that precipitate out of the brine. At unlike at lower temperatures were phases all the study temperatures, precipitation of are more mingled up. This is attributed Halite is observed. to the increasing difference between the The waxy massive form (Fig. 2, (70° (A)) solubilities of the constituent minerals; as the which is due to the hygroscopic nature of the temperature increases, the difference in the salt (Eswaran et al. 1980; Mees and Tursina saturation with respect to the mother solution 2010) and the sharp cubic crystals (Fig. 2 of the different minerals increases as well. (60°(C))(Abdel-Wahed et al. 2015) are typical Halite morphologies. Massive forms of Halite Thermodynamic modeling are also observed as coatings composed of At 30°C, 40°C, and 50° evaporation anhedral crystals (Fig. 2, (70°(B)) (Mees temperatures, the major salt precipitation and Tursina 2010) on top of salt (rice) cake sequence starts with Thenardite followed normally a fitting description for Thenardite by Trona, Halite and lastly Glaserite (Fig. crystals (Yamamoto and Zhu 1997; Warren 3). This is in agreement with the previous 2016). The aforementioned is evidence of isothermal evaporation study at 30°C mineral stratification and co-precipitation; conducted by Kasedde et al. (Kasedde et al.

Figure 3 Saturation Indices and Moles of solid posted during evaporation at 30°C

518 Wolkersdorfer, Ch.; Khayrulina, E.; Polyakova, S.; Bogush, A. (Editors) IMWA 2019 “Mine Water: Technological and Ecological Challenges”

Figure 2 SEM micrographs of the precipitated minerals from the evaporation experiments.

2013) and Lwanyaga et al. (Lwanyaga et al. When the temperature is increased to 2018). However in this study, Sylvite does 60°C, Glaserite does not equilibrate with the not equilibrate with the brine and therefore solution and therefore does not precipitate. doesn’t precipitate unlike in the previous Further increase temperature to 70°C, results phase developments study by Kirabira and into the precipitation of only Thenardite co-workers (Kirabira et al. 2015). and Halite. At all the study temperatures,

Wolkersdorfer, Ch.; Khayrulina, E.; Polyakova, S.; Bogush, A. (Editors) 519 IMWA 2019 “Mine Water: Technological and Ecological Challenges”

Thenardite precipitates first and later with Acknowledgements increasing evaporation dissolves into the The authors gratefully acknowledge funding solution. The Saturation Index (SI) of Soda from Swedish International Development ash, a mineral of interest in this study, Cooperation Agency (Sida) and the African increases with increasing evaporation Center of Excellence in Materials, Product temperature. This observation predicts Development and Nano-Technology precipitation of this mineral at temperatures (MAPRONANO ACE). higher than 70°C. From the model results as seen in References Fig.3, Halite posted the highest amount of Abdel Wahed MSM, Mohamed EA, El-Sayed MI, precipitate, followed by Thenardite, Trona et al. (2015) sequence during and lastly Glaserite. This is consistent with evaporation of a high concentrated brine previous studies (Arad and Morton 1969; involving the system Na–K–Mg–Cl–SO4- Kasedde et al. 2014; Lwanyaga et al. 2018) H2O. Desalination 355:11–21. doi: 10.1016/j. which ascertained that, at the salt lake, brine desal.2014.10.015 are comprised of more sodium, chloride, Altomare A, Corriero N, Cuocci C, et al. (2015) Carbonate, sulphate, Bicarbonate, & QUALX2.0 : a qualitative phase analysis software salts. using the freely available database POW_COD. There are noticeable differences between J Appl Crystallogr 48:598–603. doi: 10.1107/ the actual precipitated salts from the S1600576715002319 evaporation experiment and the PHREEQC Arad A, Morton WH (1969) Mineral springs and results. The XRD and SEM results confirmed saline lakes of the Western Rift Valley, Uganda. existence of more double salts and hydrates Geochim Cosmochim Acta 33:1169–1181. doi: compared to the PHREEQC model. This 10.1016/0016-7037(69)90039-8 variation is due to the fact that the PHREEQC model mainly considers the thermodynamic Driver P, Tukahirwa EM (1990) Preliminary aspects and not the kinetics, yet, the actual environmental impact assessment of Lake setting considers the kinetics thus some Katwe salt plant, Western Uganda minerals might precipitate from the model Eswaran H, Stoops G, Abtahi A (1980) SEM and not observed by XRD and vice versa. morphologies of halite (NaCl) in soils. J Microsc 120:343–352. doi: 10.1111/j.1365-2818.1980. Conclusions tb04153.x In this work, the effect of temperature on the Kasedde H, Bäbler MU, Kirabira JB, et al. (2013) precipitation and crystallization sequence on Mineral recovery from Lake Katwe brines using the Lake Katwe brine during evaporation was isothermal evaporation. In: Brown A, Figueroa investigated. Increase in temperature fostered L, Wolkersdorfer C (eds) Reliable Mine Water the evaporation rate and therefore it took Technolgy. Publication Printers, Denver, less time to precipitate minerals at higher Colorado, USA, pp 855–860 temperatures than at lower ones. XRD and Kasedde H, Kirabira JB, Bäbler MU, et al. (2014) SEM results confirmed the presences various Characterization of brines and of commercial salts. Halite is the most abundant Lake Katwe, Uganda. J African Earth Sci 91:55– mineral at all the study temperatures. 65. doi: 10.1016/j.jafrearsci.2013.12.004 Other minerals include, Thenardite, Trona, Kirabira J, Kasedde H, Bäbler M, Makumbi T Burkeite, Hanksite and Natrite. carbonates (2015) Phase Developments during Natural and sulphates of calcium and magnesium Evaporation Simulation of Lake Katwe Brine (Calcite, Tychite, and Dolomite) precipitate Based on Pitzer’s Model. Br J Appl Sci Technol first. These are in low quantities and therefore 11:1–7. doi: 10.9734/BJAST/2015/20598 will have to be removed first in the extraction Kirabira JB, Kasedde H, Semukuuttu D (2013) process to avoid contamination of Halite. Towards The Improvement of Salt Extraction At From this study, Halite of table salt grade will Lake Katwe. Int J Sci Technol Res 2:76–81 be best extracted at higher temperatures (60 Lwanyaga JD, Kasedde H, Kirabira JB (2018) & 70°C). Thermodynamic modeling of Lake Katwe brine

520 Wolkersdorfer, Ch.; Khayrulina, E.; Polyakova, S.; Bogush, A. (Editors) IMWA 2019 “Mine Water: Technological and Ecological Challenges”

for industrial salt production. Int J Sci Technol and Inverse Geochemical Calculations. U.S. Res 7:31–37 Geological Survey Techniques and Methods, Mees F, Tursina T V. (2010) Salt Minerals in book 6, chapter A43, 497 p. Saline Soils and Salt Crusts. In: Interpretation Warren JK (2016) Evaporites, Second. Springer of Micromorphological Features of Soils and International Publishing, Cham Regoliths. Elsevier, pp 441–469 Yamamoto S, Zhu Y (1997) Thenardite and Parkhurst DL, Appelo CAJ (2013) Description gypsum precipitated in wall of brick building in of Input and Examples for PHREEQC Version campus of University of the Ryukyus, Okinawa. 3 — A Computer Program for Speciation, J Sedimentol Soc Japan 46:15–21. doi: 10.4096/ Batch-Reaction, One-Dimensional Transport, jssj1995.46.15

Wolkersdorfer, Ch.; Khayrulina, E.; Polyakova, S.; Bogush, A. (Editors) 521