General Recipe and Properties of a Four Inch Hydroceramic Waste Form

Sodium bearing waste (SBW) having a high loading of sodium salts (typical of the SBW stored at Hanford) must be pretreated before it can be incorporated in a hydroceramic waste form. One such method, using thermal decomposition of the nitrate/nitrite ions in the SBW is described. Calcination of the SBW with sucrose and various aluminosilicate sources produced a granular solid that was solidified by mixing it with additional metakaolinite and a minimal amount of NaOH solution. The caustic and metakaolinite combine and form tectosilicates and in doing so bind the sample together forming a strong and highly stable solid we are calling a hydroceramic. A summary of the systematic study that allowed us to optimize both recipe and associated processing is reported. Scoping experiments use gram sized samples, thus a 4 inch cube was also produced using the optimized recipe and procedure in order to test for the homogeneity of a larger sample. The scale-up sample was evenly divided into 27 sub-samples. The properties of all the sub-samples indicate that the larger sample is homogeneous and thus it might very well be feasible to produce a massive hydroceramic nuclear waste form for the large-scale solidification/stabilization of SBW stored at Hanford. INTRODUCTION A monolithic hydroceramic is a new nuclear waste form that has been developed to solidify liquid waste commonly called low activity or sodium bearing waste (SBW). 1-4 This waste is currently stored in underground tanks at the Department of Energy’s (DOE’s) Savannah River and Hanford sites. The DOE has been mandated to clean up this legacy waste. In order to do this they are required to produce a solid from the waste that is measurably strong and insoluble enough to qualify as a waste form suitable for storage either on site or off site at a national repository. Generally the DOE favors making a glass from the more easily vitrified insoluble precipitates (sludge) at the bottom of the tanks. 5 The remaining soluble fraction is what we call SBW consisting mainly of NaOH, NaNO3 and NaNO2 with much smaller amounts of soluble carbonates, sulfates and a host of organic compounds. The difficulty of making a glass from a SBW is that many sodium silicates are soluble, which is not an entirely desirable property of a waste form. The problem can be resolved however by diluting the sodium content of the glass by adding a large amount of glass making (silicate) frit and produce, as an example, a borosilicate glass like Pyrex that contains 2.8 wt% Na. The concern here is that the cost of a glass log (the name of the solidified waste in a 10 foot high stainless steel canister) is approximately one million dollars. Waste loading will be minimal and the cost enormous. Therefore, alternates are always under consideration. Portland cement grout in one of many forms has been used and continues to be used to dispose of SBW. 6 Oak Ridge National Lab was among the first to promote Portland cement based waste forms that contained Indian River clay, some blast furnace slag and fly ash residue (from burning coal) to form zeolite-like phases. Savannah River has an ongoing program in which properly pretreated SBW (i.e. waste that has had its cesium and strontium removed) is solidified using a compositionally related Portland cement grout that they call Saltcrete. The major problem with any Portland cement waste form, even ones proposed to be used to fill empty tanks at Savannah River or Hanford is the fact that they are calcium silicate based. The hydrates that form during setting are calcium silicate hydrates and smaller amounts of zeolitic phases. Zeolites are able to host sodium ions and cancrinite can host nitrate and possibly nitrite, but calcium silicate hydrate itself can host neither. 2 For this reason the majority of the salts in the waste are not hosted by one of the hydrates that form and thus remain part of the interstitial fluid that is highly leachable. Let us focus for a moment on zeolites. Zeolites are presently forming in the World’s oceans, in saline lakes and other alkaline environments that have accumulated volcanic ash or fine silty clays (smectites). Natural zeolites are good scavengers for nuclear waste. 7 Over the years we have developed this concept into one in which we are now able to use the SBW itself as one of the ingredients to make a zeolite-containing monolith without using any Portland cement at all. Taking a recipe for Saltcrete and making it without Portland cement is a starting point for thinking about what we are doing. However, one must also remove all calcium containing ingredients such as blast furnace slag as well. This leaves the SBW and clay. We have found that the SBW can be reacted with a thermally treated clay (much more reactive) and, if the SBW is limited and the paste that forms is very thick and extrudable, the mixture will harden at ambient in a month (faster at 90°C) to form a strong low leachable solid. 4 We have been calling this solid a hydroceramic, because it is tan in color, vitreous in appearance and rings like a ceramic when hit with a steel rod. The phases that form in the hydroceramic during curing, mimic those that form in nature. They consist of tectosilicates, namely zeolites and feldspathoids. In both cases these phases are able to host sodium and a variety of anions in their structures. The zeolites we have formed generally are Zeolite A, Na-P1 and hydroxysodalite. The feldspathoids are generally cancrinite-like. Although Na is exchangeable, ions such as Cs and Sr are held preferentially. It has been demonstrated that the leachability of hydroceramics approach those of DOE’s best glass waste forms at considerably lower cost. We are proposing that hydroceramics be used to solidify DOE’s remaining SBW. One of the advantages of doing so vis à vis Saltcrete for example is the ability of a hydroceramic to host Cs and Sr. Because they are encapsulated in a zeolite they are not easy to leach. 7 This fact could help Savannah River and Hanford convince stakeholders that it is safe to dispose of their slightly higher activity wastes on site rather than shipping them offsite as high level waste. Finally, filling empty tanks with a pumpable hydroceramic (in this case a mixture of Class F fly ash and concentrated NaOH) will harden and form zeolites. The hydroceramic will support the overburden for millennia and as an added benefit the zeolites that form will (it is so presumed) interact favorable with in tank residuals. These residuals are possibly silica rich phases and traces of hardened sludge that could slowly dissolve and leak from the tank in the far future. If the zeolites were able to adsorb these ions and contain them and release sodium instead this would provide another protective mechanism not provided by Portland cement based tank fill materials. One can make hydroceramics in a number of ways more or less dictated by the composition of the SBW. The proposed reaction works best if the SBW consists entirely of concentrated NaOH solution (let us say 4-15 molar). The caustic dissolves some of the aluminum silicate (metakaolinite) and once in solution the anions begin to rearrange themselves around the solvated sodium ions. The structures that form have very short range order and are therefore not detectable by X-ray diffraction, but they are nevertheless tectosilicate precursors (NMR tells us so). Silicon and aluminum ions change their coordination from 3 and 6 to 4 and the newly formed tetrahedra link together. If one waits for a few weeks or a month the samples will begin to develop long range order and one will begin to see crystalline zeolites and feldspathoid peaks develop as long range order becomes more widely established. As man is loath to wait, the rate of reaction can be increased by raising the temperature to as high as 180°C (e.g. in a steam heated autoclave). For the work reported here we have focused on 90°C because one could achieve this environment in a well insulated room with conventional but modified heating ventilation air conditioning (HVAC) systems. Further more mixing can be carried out using pug mills and the like which is available off the shelf because these same mixers are used to prepare clay based ceramics. Which brings us back one again the SBW composition. We have found that if the waste contains a small amount of NaNO3 and NaNO2 salts the process will work as well as if the waste were pure NaOH. The only difference being the fact that cancrinite phases form. This is important because cancrinite is able to host nitrate and possibly nitrite in its crystal structure. If however the waste contains an excess of these salts some will remain unreacted and the waste form will have a very large and unacceptable leach rate. We have found that the upper limit of salts in the waste is 25 mol% NOx calculated as the ratio of total moles [(NOx/total moles Na) x 100]. SBW falling below this value can be solidified directly, in one step. 3 SBW that has a higher salt content (typical of the SBW at Hanford) must first be pretreated in some way to get its salt content into an acceptable range after which it can be incorporated into a monolithic hydroceramic. We can pretreat the waste in a number of ways. Calcination has been used here. What is given below is a summary of the calcination work we have accomplished to date. It provides the reader with an optimized recipe for dealing with the calcination process and the best way to solidify the calcine that is produced with a hydroceramic binder. The making of a hydroceramic has two parts and each part has several steps. The first part consists of denitration/denitrition by heating (calcination), which has the following steps: (1) The composition of the SBW is analyzed. The total moles of NO3 and NO2 and the total moles of Na are calculated. (2) Sucrose is added to the liquid waste at a ratio of 38 g sugar/mole of NO3 and NO2 (NOx) in the SBW. (A little excess of sugar is acceptable.) (3) Metakaolin is added to the SBW at a mole ratio of metakaolin/Na = 0.7. (A little excess of metakaolin is also acceptable.) After the addition of the sugar and metakaolin to the SBW the mixture becomes a slurry. (4) The slurry is dried at 90 o C and then calcined at 525 o C for 10 h allowing an additional 3 hours to raise the temperature to 525 o C. This process denitrates/denitrites the sample and at the same time encourages the formation of sodium aluminosilicate compounds that are collectively known as calcine. The second part is the preparation of a monolithic hydroceramic, from the calcine. This second process includes the steps of (5) Calcine is mixed with additional metakaolin at a weight ratio of calcine/metakaolin = 3/2 and then mixed with 4 M NaOH at the mole ratio of additional Na/additional metakaolin=0.5 to form a paste. The paste can be placed in a container or in a mold. It is thick but yet thin enough to self compact with vibration. (6) The sample is precured in the mold at 40 o C and 100% humidity until the sample becomes hard enough to demold. At this point the sample has minimal strength, but had it been allowed to cure longer it would get progressively stronger with time. (7) The sample is now hydrothermally cured at 90 o C or higher for 24 hours or longer during which the precursor phases crystallize and bind the mass together to form the desired hydroceramic monolith. In reality, every SBW now stored in tanks at the DOE sites are different from one another. Therefore, it is important to study the general applicability of this recipe and procedure to dispose of SBW having different compositions. The data given below summarizes our findings for a Savannah River and a Hanford simulated SBW, which have different compositions and different concentrations of NOx, Na and Cs. These were tested in order to show that the method works and the data are summarized here as an example of what one may expect from a hydroceramic produced from different types of waste. The Savannah River samples were small 2.54 by 5.1 cm long ( 1” by 2” cylinders , whereas the Hanford simulant was incorporated in a 4inch cube of hydroceramic. Results show that both hydroceramic samples had low leachability and adequate strength. The data for 27 subsamples taken from the large cube indicate that the larger sample is homogeneous and thus suggests that larger samples could be made without concerns for separation during setting and curing. EXPERIMENTAL Materials Metakaolin (mainly, metakaolinite, Al2O3·2SiO2) as mentioned before 2 was used as the aluminosilicate source. Two nuclear waste simulants, i.e., a Savannah River Tank 44 simulant (SRS) 3 and a Hanford Tank AN-107 simulant, 4,8 as showed in Table I and Table II, respectively, were prepared from reagent grade chemicals. Table I. The composition of SRS a Compound Concentration (g/l) CsNO3 97.4545 KNO2 8.5107 NaNO2 27.6000 NaAlO2 16.3958 NaOH 179.9910 a The total concentration of nitrate and nitrite anions is 1.0 mol/L. The total concentration of Na is 5.1 mol/L. The concentration of Cs is 0.5 mol/L. Table II. The Hanford simulated SBW Compound Concentration (g/l) Compound Concentration (g/l) Al(NO3)3·9H2O 121.1723 PbO 0.3172 Ca(NO3)2·4H2O 2.5171 NaCl 2.2999 Na2Cr2O7·2H2O 0.6532 NaF 5.5596 CsNO3 0.0157 Na2HPO4 4.0852 Fe(NO3)3·9H2O 15.0138 Na2SO4 11.7463 KOH 2.0483 NaNO2 69.0892 La2O3 0.0307 NaNO3 155.6883 NaOH 105.9035 Na2CO3 123.0559 NiO 0.4637 b The total concentration of nitrate and nitrite anions is 3.93 mol/L. The total concentration of Na is 8.2 mol/L. The concentration of Cs is 8.055 x 10 -5 mol/L. The Hanford simulated SBW has a higher nitrate/nitrite and Na concentration than SRS SBW. However, SRS has a much higher Cs concentration than the Hanford simulated SBW.