HUNGARIAN JOURNAL OF INDUSTRIAL CHEMISTRY VESZPREM Vol. 30. pp. 127 ~ 130 (2002) DEVELOPMENT FOR SELECTIVE CESIUM REMOVAL FROM THE EVAPORATION CONCENTRATES OF THE PWR PAKS S. SZOKE, G. PATZAYandL. WEISER (Department of Chemical Technology, Budapest University of Technology and Economics, H-1521 Budapest, HUNGARY) Received: January 14, 2002 At the Pressurized Water Reactor Paks the diluted radioactive wastewater is converted to concentrate by evaporation and · d k h · d. I"d · th 1 · 134C 137c d 60c the evaporator bottom 1s store in tan s. T e most Important ra tonne 1 es m ese so ut10ns are s, s an o. The volume reduction technologies like selective ion exchange for cesium and the ultrafiltration for cobalt may reduce the volume of the liquid waste and the burial costs. In our research we studied our granular potassium nickel hexacyanoferrate(li) as a Cs-selective ion-exchanger. Its capacity was depending on the preparation method, tem~erature of pretreatment and age of the ion exchanger. We investigated also the effect of metal ions ( Fe2+, Fe3+, MJ?+, Mg +, Ca2+, Co2+, NP+, Cu2+) on the Cs-capacity in the presence of complex forming compounds citrate, oxalate and EDT A. The cesium ion exchange capacity was increased in case of addition of inactive cobalt or nickel salts. Additionally, we studied filtration, adsorption and ultrafiltration separation processes for cobalt removal. The results showed that only adsorption by active carbon could successfully be used for the cobalt removal from the evaporation concentrates. Experiments w_ere also performed in our laboratory and at the PWR Paks. Keywords: cesium and cobalt removal, ultrafiltration, potassium nickel hexacyanoferrate(II), selective ion exchange, adsorption, complex compounds Introduction Approximately 40000 m3 diluted liquid radioactive wastewater is generated yearly in the Hungarian Pressurized Water Reactor (PWR) at Paks. At this time there is no selective waste collection. The collected liquid waste is the mixture of low or medium activity and/or salt content wastewater and often decontamination solution. The collected solutions are regularly concentrated by evaporation using NaOH for stabilization of the sodium borates in the evaporation bottom and are stored in common stainless steel tanks. The major component of the concentrate is sodium borate, and the solution is alkaline (pH=l4). The final disposal of this waste is not solved. Sodium borates are incompatible with the embedding matrix material (concrete). In addition the direct cementation of the radioactive waste results a final volume two-three times larger then the volume of the evaporator bottom. Consequently it causes increasing burial cost. The major ingredients of the evaporation concentrates are sodium borate. sodium nitrate and sodium hydroxide. They contain long-lived radioactive isotopes ct34Cs, 137Cs and 60Co) in ultra-low concentrations. Because 75-90% of the radioactivity is due to cesium isotopes its selective separation before solidification is the most important step. The selective removal and accumulation of these long-lived radionuclides in a small volume ion exchanger reduces the volume of the buried radioactive waste and the cost of the final disposal. [11 It is well known that the heavy metal hexacyanoferrates {nickel-KNiFC. copper and cobalt- KCoFC) are selective cesium ion exchangers. The review of their properties and preparations can be found in [2]. Some methods are known to stabilize and improve their capacity e.g. KNiFC loaded zeolite [3]. Si02-· KCoFC composite [4] etc. The aim of the present work is the investigation of the possible improvement of the selective cesium removal using granulated potassium nickel hexacyanoferrate(II} ion exchangers and development of a suitable granular ion exchanger. In this process we investigated various preparation methods, the effect of other factors, such as temperature, age of the exchanger, and other chemical components present in the solution. The other important long·lived radionudide in the evaporation concentrate is 00Co. It is present as Contact information: E-mail: szoke-sz.ktt@chem.bme.hu; Tel.:463~1945; Fax.:463-19l.3 128 Table 1 Average chemical composition and radioisotope content of the evaporation concentrates of PWR Paks pH Na+ K" N03 H3BO, evaporated residue 134Cs 137Cs 60Co Concentration or specific activit g/dm' g/dm3 gldm' gldm3 gldm3 MBq/dm3 MBq/dm3 MBq/dm3 14 80 24 45 250 534 7.2 38.4 0.7 suspended matter, colloidal particles and in complex form. The dissolved cobalt can be removed by adsorption using for example charcoal [5] or montmorillonit [6]. In our research we investigated the removal of the suspended particulates of cobalt from the evaporation concentrate via filtration and ultrafiltration and the other forms via adsorption. Laboratory Experiments 1J RB I!!!Di!lliJ18.6 g. ] RF !iimlll!!l!lili!iBl 40.5 i;l g ND •••Bwml. I·"I·IIIIB·· 59.2 ., i NB jl!la 12.8 0:: NF 23.0 0 20 40 60 Removed(%) 80 Fig.] Effect of preparation techniques for Cs-137 removal (R: reverse order of portioning of reagents, N: normal order of portioning of reagents; D: dehydration method, B: boiling-gel method, F: freeze-thaw method) 60% 50% 40% ., " " ~ 30% >4 20% 10% 0% 70 80 90 100 110 120 Treatment temperature ( °C ) A mixture of the original evaporation concentrates of Fig .l Cs-137 removal capacity for treatment temperature the PWR Paks was used in the laboratory experiments (Table 1). Three types of processes were examined, filtration (5 and 1 Jlm pore size filter), adsorption with The effects of other factors active carbon and ultrafiltration (molecular weigh cut off 15000 Dalton) for its removaL Development and testing of a Cs-selective ion- exchanger Preparation of the ion-exchanger Six different techniques were investigated for the preparation of a granular sorbent: freeze-thawing, boiling-gel and dehydration method using normal and reverse order of portioning of reagents (potassium hexacyanoferrate(U) and nickel sulfate solution). The capacity of the produced sorbents was tested by static method. .Ion exchange measurement technique in static method 100 mg of the granular ion-exchanger was added to 200 cm3 of thermostated evaporation concentrate and stirred for an hour. then filtered with micropore filter paper. A Nai (Ti) detector and an 8000 channel EMG-8ll0 gamma-spectrum analyzer before and after the handling measured the Cs-content of a 100 cm3 sample. The measurement time was 2000 seconds. The results are summarized in Fig.J. We found that the best Cs sorbent is the "ND" granular product. Temperature Because the best Cs-sorbent was prepared by drying (dehydration) we tried to optimize the treatment temperature of ND granular product. Fig.2 shows the Cs removal capacity as a function of the dehydration temperature. The optimum is found between 80 and 100 oc. Storage time (age of the exchanger) The long-term stability of the product was depending on the temperature of dehydration. The sorbent that was handled at 90 oc was more stable than that at 80 oc. (see Fig.3) Complexing agents The evaporation concentrates contain complex forming compounds (oxalic acid, citric acid and EDTA) which decrease the sorbent ion exchange capacity. To minimize this effect by fixing these free complexing agents, in our next ion exchange experiments we applied as inactive reagents eight different metal salt {0.01 M salt concentration in every experiment). Two of them caused a significant increase in the cesium ion exchange capacity; namely the nickel sulfate and the cobalt nitrate (Fig.4). A 52.4 ~ 8 53. I ~ c 43.7 D 32.5 0 20 40 60 Removed(%) Fig.3 Effect of storage time on Cs-137 removal (A: handling at 90 oc two weeks later; B: handling at 90 °C; C: handling at 80 oc two weeks later; D: handling at 80 oc ) 70'/o 60% 50% 140'/o 0 B30'/o ~ 20% 10% 0% z g c 0 (;;" z u.. 0 :::1 "' "' c: Q Q Q ::0 ::0 (;;" u.. Tw.atrrenfl! Fig .4 Effect of metal ions on the Cs-137 capacity Based on the static experiment results we concluded that the Cs-removal capacity depended on the preparation method of the sorbent, the temperature of pretreatment, the storage time and the amonnt of complexing agents (EDTA, citric acid etc.) and the free complexing agents could be :'deactivated" by reacting with added nickel and cobalt compounds. Column tests (dynamic tests) In the column experiments we used a glass ion- exchange column (ID = 5 nnn) filled with 2 cm3 granulated (0.125-0.2 nnn) potassium nickel hexacyanoferrate(Il) prepared by dehydration method (ND). The initial flow rate was 50 cm3 (25 BV) per hour. The effluent concentrate was collected in 100 cm3 plastic flasks. The measurement technique was similar to the static test. 100 cm3 of solution was measured during 2000 seconds by Nai(TI) detector and an 8000 channel EMG-8110 gannna-spectrum analyzer. In our dynamic experiments we compared the efficiency of the cesium removal from the concentrate and from the pretreated (with cobalt-nitrate) concentrate. We found that the pretreatment increased the Cs-capacity of the exchanger. The breakthrough curves are shown in Fig.S. Fig.5 shows that the Cs + breakthrough is smaller in case of the pretreated concentrate. This result means that the dosed cobalt compound reacted with the aggressive complex compounds and the decomposition of the exchanger was smaller. 129 Table 2 Summary of filtration processes Microfiltration Microfiltration Isotope (5 J.Ull C1 J.Ull pore size) pore size) removed removed 0.18% 0.21% 0% 6.3% 7.2% 6.2% Activated carbon filter (with Co(NO,)z) removed 30.2% 30.4% 78.4% Ultrafiltration (I J.Ull prefi1tered solution) removed 1.9% 2.4% 5.0% % Removed = 100"[1-(solution activity after the treatment/ solution activity before the treatment) J E 45 ~ "' ~~ : ~ 25 ] 20 l 15 Fig.S Cesium ion exchange breakthrough curves of un- and pretreated concentrates Fig.6 Filtration and ultrafiltration setup Filtration, adsorption and ultrafiltration experiments The effective removal of the various forms of the long- lived radionuclides from evaporation concentrate was also investigated using filtration, ultrafiltration and adsorption processes. In the filtration and ultrafitration experiments we used a WATTENTECHNIK microfiltration module and a MICRO CARBOSEP 40 M5 ultrafiltration membrane (the operating pressure was 2 bar). The experimental setup is shown in Fig.6. According to our previous experiments the cesium and cobalt radionuclides may be present in the concentrate as ions, suspended and colloidal particles and in complex forms. The results of our filtration and ultrafiltration experiments showed that a small pru1 of the cobalt content could be removed by filtration (6%) and the ultrafiltration had also low efficiency (5%). Although the adsorption of cobalt by an active carbon filter ~med effective means of separation, unfortunately the capacity of the carbon filter is very low. Our filtration experiments showed that the long-lived radionuclides definitely occur in non- filterable (ionic or complex) form (Table 2). 130 Table 3 Summary of filtration processes from the mother lye Isotope Microfiltration (1 f.IID) removed 0.3% 0% 0% Ultrafiltration (1 f.IID prefiltered solution) removed 1.6% 3.5% 0% On-site experiments at the PWR Paks The results of the dynamic experiments obtained in our laboratories were controlled and tested at Pak:s NPP on site. In the experiment we used a freshly ''produced" and an "older" evaporation concentrate solutions. One was the mother lye (partially neutralized evaporator concentrate) of an old stored concentrate (pH=9.5, H3B03 16.1 g/cm 3 ) the other a freshly evaporated co11centrate (pH=14, H3B03 109 g/cm 3 ). The fresh concentrate contained medium-lived radionuclides such as 54Mn, 58Co, 57Co and the long-lived isotopes ct34Cs, 137Cs and 60Co): All the measurements were performed at the isotope laboratory of the PWR using a calibrated HPGe semiconductor detector (1000 seconds measurement time; 100 cm3 sample in plastic flasks). Selective ion exchange In this method we used wider.column (ID=8 rom) filled with 2 cm3 granulated (0.2-0.3 rom) potassium nickel hexacyanoferrate(II) prepared by our preparation technique. The flow rate was 100 cm3 (50 BV) per hour. The ion-exchange separation efficiency from the "fresh" concentrate was very low, because manganese dioxide precipitated continuously from the concentrate onto the surface of the exchanger. In the case of the mother lye the efficiency was excellent, the average decontamination factor was 250 and the volume reduction factor more than 1700. We could not determine the breakthrough point because of lack of the limited experimental time. The breakthrough curve is shown in the Fig.7. Filtration and ultrafiltration The same laboratory filtration and UF equipment were used in these experiments. Only 20% of the cobalt could be removed from the fresh concentrate by a I p.m. pol'e size filter and consecutive ultrafiltration. In the case of the mother lye the filtration processes were unsuccessful (Table 3 and 4). Conclusion According to the laboratory and on-site experiments we conclude that the cobalt content of the evaporation concentrates at the PWR Paks is mostly not in suspended and colloidal forms but is rather present in Table 4 Summary of filtration processes from the fresh concentrate Isotope Mn 57 Co ssco t3•cs 137Cs 60Co 0,5 0,1 Microfiltration(1 f.IID Ultrafiltration pore size filter) (1 f.1ID prefiltered solution) removed removed 1.1% 9.0% 8.7% 3.7% 4.4% 15.4% 6.5% 11.7% 1.3% 10.6% 0.2% 15.8% . .I ...... / ,/~\ \I ~ 1500 2000 2500 Effluent VOlume {em~ 3000 3500 Fig. 7 Cesium ion exchange breakthrough curve of the mother lye complexes. The only applied technique that could reduce the quantity of complex form radiocobalt is adsorption on activated carbon. We prepared a granular Cs-selective ion exchanger,· determined its optimum parameter of preparation, drying and storage. We successfully tested its selective cesium separation efficiency in laboratory and on-site at the PWR Paks. Acknowledgment The authors thank Peter Tilky and Ferenc Feil for his kind help in the Nuclear Power Plant at Pak:s. REFERENCES I. PATZAY G., WEISER L., T61H B., PALMA! G. and FEIL F.: Per. Polytech. Ser. Chern. Eng., 1995, 39(2), 147-154 2. HAAs P. A. : Sep. Sci. Tech., 1993, 28, 2479 3. MlMuRA H., KIMURA M. and .AKmA K.: Sep. Sci. Techol., 1999, 34(1), 17-28 4. MARDAN A., AlAZ R., MEHMOOD A., RAZA S. M. and GHAFFAR A.: Sep. Pur. Tech., 1999, 16, 147- 158 5. FLoRES R. M., OLGUIN M. T., SOLACHE-RIOS M., LoNGORIA S.C. and BULBULIAN S.: J. Radioanal. Nucl. Chern., 1998, 238(1-2) ,199-201 6. PACHERO G., NAVA-GALVE G., BOSCH P. and BULBULIAN S.: J. Radioanal. Nucl. Chern. Letters, 1995, 200(3), 259-264 Page 126 Page 127 Page 128 Page 129