Evaluation of New Tank Fill Materials for Radioactive Waste Management at Hanford and Savannah River

Transcription

1 Evaluation of New Tank Fill Materials for Radioactive Waste Management at Hanford and Savannah River May 31, 2006 Prepared with the support of DOE Grant No. DE-FG02-05ER63966 Authors: Paul Brenner, Yun Bao, Maria Dicola, Michael W. Grutzeck The Pennsylvania State University Materials Research Institute Materials Research Laboratory Building University Park, PA 16802

2 Introduction The U.S. Department of Energy is working on methods to reduce/eliminate its 80 million gallon inventory of radioactive waste currently stored in underground tanks at its Hanford and Savannah River sites. The DOE is planning to remove the waste (sludge, salt cake, liquid) from the tanks and convert the waste into a solid that can then be disposed of either on or off site. The majority of the waste has been in storage for close to 50 years. It has undergone a considerable number of decay cycles, now making it relatively low activity waste. The waste consists of a highly alkaline liquid (supernate), precipitated salts at the top edges of the tank (salt cake) and a small amount of insoluble sludge (precipitate) which has settled to the bottom of the tank. Because of the potential hazards associated with leaks, the DOE is actively converting the liquid waste and the sludge into solid waste forms. Vitrification of the sludge is underway at Savannah River and a glass melter is being built at Hanford. See for example a review article by Grutzeck et al [1]. The supernate is extremely voluminous and it would be impossible to convert it to glass in a timely fashion using the existing/proposed melters. Thus alternate solidification schemes have been adopted by Savannah River (solidification of a treated Cs/Sr-free supernate with a blended pozzolanic Portland cement to form Saltcrete) and are being evaluated by Hanford (bulk vitrification in a disposable melter) in order to speed up the clean up process. Bulk Vit, as the process is known, is an alternate to melting in a glass plant where the liquid waste is mixed with native soil in a disposable Dempsey Dumpster-sized melter. After vitrification the melter is fitted with a lid and buried on site in an engineered land fill. As the clean up process moves forward, the DOE will find itself with an inventory of empty underground tanks. Washington and South Carolina both require that the tanks be filled in order to prevent subsidence of overburden once the tanks rust away. The current choice of fill material is a variant of a basic Portland cement concrete [2]. Concrete is one of the World s most ubiquitous materials, having been used to fill almost every underground opening including oil wells, gasoline tanks, oil tanks, subway tunnels, water wells, and abandoned mines. But as is the case with most materials some applications are better suited for the use of Portland cement concrete than others. DOE s waste tanks will contain traces of radioactive substances that are tightly cemented to the bottom and edges of the tanks. These probably include insoluble sodium silicates, sodium salts and sludge that can not be removed using tank washing techniques. Legislation has been passed by Washington State that will allow contractors to leave this waste in the tank and thus dispose of it in situ. This waste is called residual waste. It is known that Portland cement is unable to host alkalis and therefore Portland cement concrete would not react with and chemically bind the alkalies that may remain in the tank [3]. It is proposed that an alternate type of concrete based upon an alkali aluminosilicate hydrate (zeolite-containing matrix) might do a better job at protecting the environment from migrating residual ions after the tank has undergone closure and the tank has rusted away [3]. Current work on such an alternate fill material has progressed to the point where it is now possible to formulate and cure mixtures that have adequate strength and also cation exchange properties that are selective to cesium adsorption [4]. It is suggested that these materials can be made into suitable concretes that will provide both long-term stability and at the same time have the potential to act as a sponge for the residual waste ions. The fill materials described below consist of a mixture of two fly ashes mixed with alkali solutions. The chemistry is such that the mixture will harden and develop a calcium silicate hydrate (cementitious phase) mixed with a zeolite (adsorptive phase). In this instance it is proposed that such an alternate concrete, one based on a so called zeolite binder would be a better choice for decommissioning. The adsorptive properties of the zeolites in the concrete should attract ions such as Cs and Sr once the

3 tank has rusted away and adsorb them and thus prevent them from migrating into the surrounding soil. The proposed zeolite-concrete can be mixed and placed much the same as Portland cement concrete, but in addition to filling space the zeolite-concrete also has the ability to absorb/chemically attract the remaining sludge and the soluble alkaline/alkaline earth ions present in the empty tanks. The work reported below, provides a summary of preliminary work undertaken in order to explore the potential of formulating equivalent concretes that set up at room temperature and at the same time contain zeolites that able to adsorb and react with residual waste ions. Two tasks are summarized: formulation of concretes able to set up at room temperature, and the ability of these materials to cation exchange with dilute CsCl solutions. The work is a logical extension of previous work by Palomo et al. [5], Krishnamurthy et al. [6] and Bao et al. [7-8]. Previous Work The best of a set of preliminary formulations previously studied [4] were chosen for further work. Two fly ashes [a Class C fly ash/dry FGD (flue gas desulfurization) material from Xcel s Sherco plant] and a Class F fly ash from the Ft. Martin power plant (part of Allegheny Energy) were mixed with various molarities of NaOH either alone or preblended with sodium silicate (PQ Type D). On occasion (in order to improve on the zeolite content) a small amount of metakaolin produced from Troy clay mined near Troy Idaho was substituted for some of the ashes. Metakaolinite tends to accelerate rates of reaction and more rapid setting. The nominal composition of metakaolin clay and Class F fly ash is: Al 2 Si 2 O 7. See Table 1. The metakaolin Table 1. Chemical Composition of Starting Materials (wt%) Oxide Ft. Martin Sherco Troy metakaolin Al 2 O B 2 O nd BaO CaO Fe 2 O K 2 O MgO MnO Na 2 O SiO SrO TiO Sulfite as SO 2 < nd Sulfate as SO nd LOI (90 C) nd nd 2.07 Total Nd=not determined was made from kaolin clay by firing it in a rotary kiln at 700 C for a few hours. The Class F fly ash consists of very fine spherical particles of partially devitrified glass derived from the

4 combustion of bituminous coal. Both contain a high level of silica which is ultimately responsible for stronger chemical bonds. Moreover, working with fly ash and/or metakaolin produces less carbon dioxide vis à vis Portland cement which is ultimately beneficial for the environment. The Sherco FGD material is a Class C ash that has been used to adsorb SO 2 from the power plant s flue gas. It is a Ca-rich glass with a coating of gypsum and hannebachite. It is self cementing and for our purposes represents a substitute for ordinary Portland cement in our mixtures. Chemical analyses of the starting materials are given in Table 1. Minimum Required Strength, Setting Time, and Cation Exchange for Tank Fill Materials One design parameter used to evaluate the minimum strength of the fill material was derived as follows.. The minimum strength required for the fill at the very bottom of the tank to support its own weight was calculated. The tanks at Hanford and Savannah River have a volume of a million gallons which means that the solid fill material at the very bottom of the tank would have to support a substantial height of solid. It was proposed that the fill material should, at a minimum, be able to support twice its overburden. It was also proposed that the fill material harden within two to three days, in as much as there is no need to have it harden in a few hours. The fact that it is slow to harden allows one to fill the tank completely without the formation of stratified layers with discrete joints between them. Finally it was proposed that the fill material provide not only strength and durability but it also provide some protection from migration of residual elements remaining in the tanks. It was required that the tank fill have the ability to selectively adsorb Cs and Sr and in this way prevent their loss to the environment after the tank has rusted away. Tanks are typically 75 feet wide and hold 1 million gallons = ft 3 of radioactive waste 1 Pi*r 2 *h = volume of cylinder. Which calculates out to 3.14* (75ft/2) 2 * h = ft 3. This suggests that the tank is ~30.3 feet high. A unit weight for fill materials under development is based on the assumption that the final fill material will have the following characteristics (these are based upon preliminary data for samples evaluated to date): molds for cubes are 2in x 2in x 2in = 8 in ft 3 an average weight of a 2 cube is 255g lb lb ft 3 = lb/ft lb = lb/in 3 3 8in To find the pressure of the fill at a depth of 30.3 ft = in the following calculation was carried out: lb/ft 3 * ft = lb/ft lb/in 2 * in = lb/in 2. It was proposed earlier that a safety factor of two would be enough. This sets a 50 psi lower limit. From a construction point of view however, this seems very low, so it has been arbitrarily decided to adopt 100 psi as the required lower limit. Because this requirement is relatively low, it allowed us to favor Class F fly ash-based mixtures (in this case Ft. Martin fly ash) and their propensity to form zeolites when mixed with NaOH. The Class C ash from Sherco is more Portland cement like and less likely to form zeolites and thus less likely to have the required cation exchange capability. 1

5 Exploratory Mixtures and Results Mixtures containing Class F and/or Class C FGD fly ash and on occasion metakaolin were mixed with sodium hydroxide and/or a sodium silicate solution (PQ Type D) to form thick pastes that were then placed in 2 inch cube molds and allowed to cure. Tables 2a and b represent the first attempts at narrowing selected formulations identified in our earlier report [4] to a more workable few. Five samples were produced in which the percentage of Sherco versus Ft. Martin ash was gradually varied. Earlier work had determined that samples made with 4M versus 8M NaOH solution performed adequately, and in some cases better than the equivalent 8M sample. It was determined that the 4M samples tended to achieve adequate strength at half the cost of 8M NaOH. Thus 4M NaOH was chosen for this portion of the work (Table 2a). As a check the 50:50 mixture highlighted in red was reformulated with 8M NaOH instead of 4M, all other variables remained the same, and indeed the results did in fact confirm the earlier data. (Table 2b). Table 2a. Mixing Solution was 4 M NaOH and all Samples were Cured at RT for 7 days Mixture (% Sherco/%Ft. Martin) 100/0 75/25 50/50 25/75 0 Sherco fly ash (g)* /530 0 Ft Martin fly ash (g)** NaOH (g) & molarity 405/4 422/4 392/4 340/4 303/4 Solution/Solid Ratio Curing conditions 7 days RT 7 days RT 7 days RT 7 days RT 7 days RT Compressive strength (psi) *** 426 still soft The recipe highlighted in red is the one we chose for further work. *Sherco ash is from Northern States Power s Sherburne Co. Unit 3 generating station. **Ft Martin Class F ash is from Allegheny Energy s Ft Martin generating station. ***An equivalent sample made with 8M NaOH tested out at 189 psi. Table 2b. Mixing Solution was 8M NaOH, Samples were Cured at RT for 7 days: A Comparison of 2 Separate Mixtures Mixture (% Sherco/%Ft. Martin) 50/50 (earlier data) 50/50 (current data) Sherco fly ash (g)* Ft Martin fly ash (g)** NaOH (g) & molarity 722/8 676/8 Solution/Solid Ratio Curing conditions 7 days RT 7 days RT Compressive strength (psi) It was observed that the strength of a given mixture was directly proportional to the amount of Sherco Class C FGD in the mixture. Even though the 100% Sherco sample was the strongest, it was removed from consideration because it was found that it contained little or no zeolite phases and thus would not be able to adsorb and hold residual ions in its structure. The Sherco sample is more Portland cement-like than the other samples and thus lacks the chemistry necessary to produce zeolites. In addition, the 100% Sherco mixture required so much more sodium hydroxide solution to make the mixture liquid enough to pour (solution/solid ratio = 0.62) that that the effect of using lower molarity solutions to save on cost was somewhat negated. The 100% Ft. Martin sample was also eliminated because it had not yet hardened at 7 days. It did however form copious amounts of zeolites when cured at slightly higher temperatures or for longer times. Ft. Martin samples would harden in a matter of weeks, but even then their strength was normally in the vicinity of 100 psi leaving little room for error. Having eliminated these two mixtures, three mixtures remained for further study. These could be

6 cast in place and would harden at room temperature. They developed compressive strengths ranging from 1125 to 426 psi. The mixtures containing the most Sherco Class F FGD ash (75/25) were the strongest and fastest setting. At this point it was clear that the Sherco ash was going to act as a hardening agent while the Ft. Martin ash was going to act as a raw material for zeolite production. In as much as this seemed logical because in fact phases that develop are composition sensitive, it was decided to chose the 50/50 mixture for further study 2, it was strong and lay half way between the extremes. This mixture was reformulated to make a series of three mortar samples containing progressively larger amounts of sand and lesser amounts of the matrix. Recipes and results of testing are given in Table 3. The samples were cast in 2 inch cube molds. After curing for 7 days at room temperature their compressive strengths were tested using 2 samples each. Sample mass was recorded in order to monitor the rate at which water would evaporate from the samples during curing. Samples were covered or cured in sealed plastic bags, thus weight loss was minimal. Samples were examined using X-ray diffraction to evaluate the quantity of crystalline zeolites that formed in each sample. Table 3- Mortar Formulations: Mixing Solution was 4 M NaOH - Cured at RT for 7 days* Mixture 75% Sand Mortar 50% Sand Mortar 25% Sand Mortar Sherco fly ash (g)* Ft Martin fly ash (g)** Sand C Graded quartz river sand NaOH (g) & molarity 114 / / / 4 Curing conditions 7 days RT 7 days RT 7 days RT Compressive strength (psi) *Samples were covered with a glass plate and cured in covered plastic containers. It is assumed that the RH ~ 100. As one might expect, as the amount of sand in the mortar was increased from 25 to 75%, the strength of the mortar decreased. What was unexpected however was the observation that the 25 and 50% mortars were stronger than their paste counterparts (1118 and 975 psi versus 762 psi). All mortar strengths exceeded the 100psi design minimum adopted earlier. X-ray diffraction examination of these mixtures suggested that crystalline zeolites were not produced to any great extent. See later. A few very small peaks appeared after 4 months. There was however a large X- ray amorphous hump present in the plot that suggested that nano-crystallites existed and perhaps these might be zeolite-related. It seems as if any or all of these formulations could be used to fill the empty tanks at Hanford. A final question that was addressed was how would this 50:50 mix of Sherco and Ft. Martin fly ash perform as a concrete? Recent reports by Rangan and his coworkers at Curtin University of Technology in Perth Australia provided guidance as to how to scale up a paste to a concrete [9,10]. They have been making geopolymer concretes using Class F ash and sodium hydroxide/sodium silicate exclusively. Our work is related because we used the same amounts of sand and aggregate but instead of 100% Class F ash we used combinations of Class C and Class 2 After the fact it was discovered that the 50:50 mixture did not develop crystalline zeolites even after ~4 months of RT curing. But at the same time it was also observed that the 75% Ft. Martin/25% Sherco sample described in relation to Table 2a (compressive strength 426 psi) did develop zeolites. It has been suggested that these zeolites would act as ionic sponges to attract and hold Cs and Sr and other residuals left in the tank. Thus a few concrete samples have been made and remain to be tested. X-ray data and SEM images are presented as a companion to the 50:50 mixture so examined later in the text.

7 F in order to produce concretes that cure at 21 C rather than 60 C required by Rangan s concrete. Five variations on the 50/50 mix were prepared and cast as either 3 by 6 inch or 4 by 8 inch cylinders. Samples were cast in plastic disposable molds and allowed to cure at room temperature. A series of limestone aggregates were prepared by sieving. These were prepared using a locally available construction approved limestone from a quarry in Oak Hall, PA operated by Hanson. Sizes are given in mm. Details of formulations are given in Table 4a and b. In terms of strength, the concretes turned out to be quite good. The addition of zeolite A seeds to encourage the development of zeolites in the mixture seemed to have an adverse effect on strength. These samples tested out at significantly lower compressive strengths. What is not clear, but seems to be the more important issue was why the use of 8M NaOH in these same samples would lead to lower strengths. As was the case earlier, the 8M samples tended to be less strong. The samples highlighted in red confirm this observation once again. The 8M samples are ½ to ⅓ as strong as the 4M samples (550 psi for 4M samples versus 185 psi for the 8M samples). Therefore it is now recommended that tank fill materials be formulated without zeolite seeds and with 4M NaOH solution as mixing agent. A lower solution/solid ratio seemed to increase the strength of the full concrete that is listed last on the table (3/21/06 mixture). This mixture contained large as well as small aggregate and tested out at 830 psi. This is a very respectable strength. Once again, the X-ray diffraction analyses failed to identify the presence of abundant zeolites. For this reason, cation exchange experiments are currently being run to determine their characteristic properties for this parameter. Table 4a. Formulation of Concretes: Solids Used (continued Table 4b). Mix ID (date) 19mm 12.5mm 6.3mm 4mm Sand Zeolite A 50/50 Fly ash* Total solids 3/23/ /23/ /04/ /30/ /21/ *The fly ash listed as 50/50 was a 50:50 blend of the Sherco and Ft. Martin ashes that had been preblended and stored in a bucket. The mixtures highlighted in red are duplicate samples made on different days. Table 4b. Formulation of Concretes: Liquids Used, Cure Time, Size and Strength NaOH g/m Solid/liquid ratio Cure time (days) Diameter (inches) Strength (psi) 953/ / / /8 0.19* / *Made with 1 gram superplasticizer. Naphtalene type. X-ray Data A X-ray diffraction pattern for the 50:50 paste mix cured at room temperature for 7 and 112 days is given below (Figure 1-upper pair). It is clear that zeolite production if it occurs at all must be at the nanoscale. An X-ray diffraction pattern tends to identify crystalline materials. X- ray amorphous materials may be crystalline, but they have particle sizes that are too small and too disorganized to diffract X-rays, i.e. they do not have the required long range order needed to diffract X-rays. Apparently not very much happens to this mixture s internal structure after

8 prolonged curing. The pattern for the 112 day sample is essentially the same as the pattern taken after 7 days. What both have in common however is a pronounced amorphous hump extending from 15 to 35 2θ. This seems to signify that the sample contains two types of silicates: one has a network structure whose disorganized peaks are centered around 21, a spacing usually associated with quartz, the other phase is more layer structure-like with disorganized peaks centered around 29, a position usually associated with a calcium silicate hydrate, the main phase in hydrated Portland cement. In fact the presence of a low angle jennite peak at ~8 2θ would also account for the amorphous hump at 29. Jennite is a cement hydrate. The major peaks in the plots are due to quartz and mullite in the starting material fly ashes. Thaumasite is an expansive phase often associated with ettringite. These phases are usually found in super sulfated cements. It also contains carbonate that it would get from the air over a four month period. Because the Sherco is a mix of Class C ash and CaSO x, this identification makes sense. The two X-ray patterns directly below the 50:50 mix represent the patterns for the 25% Sherco:75% Ft. Martin mix. The 7 day sample closely resembles the 50:50 samples suggesting that much of the early reactivity is due to the reaction of the Sherco fly ash. Over the longer term, the Class F ash also reacts and seems to account for the development of the zeolite phases found in the 4 month sample. Q J T M C Q M J Z M Z P,Z M,Z M M M Q Q DEGREES 2θ Figure 1. A comparison of X-ray diffraction data for the 50:50 and 25:75 wt. % mixtures of Sherco:Ft. Martin fly ash mixtures cured at room temperature for 7 and 112 days. The 50:50 mix does not change with time; the 25:75 mix develops zeolite peaks. M=mullite and Q=quartz are present in the fly ash, J=jennite, C=Ca(OH) 2 and T=thaumasite are calcium rich phases, and Z=unidentified Na-zeolite and P=phillipsite are zeolites.

9 SEM Data Two rather typical SEM images (1,000x and 3,000x original magnification) of a 50:50 and 25:75 Sherco:Ft. Martin pastes cured at room temperature for ~4 months are given in Figures 2 and 3, respectively. These are the same samples that were tested and reported upon in Table 2a, i.e. the 762 and 426 psi samples tested after 7 days. The samples were stored in plastic bags in the lab (~21 C) for an additional 4 months prior to this examination. Small pieces were mounted on conductive tape and gold coated. The samples were interesting in as much as the microstructures of the two samples was distinctly different. Figure 2. 50:50 mixture cured for 112 days. The 1000x image portrays the sample as being very dense, i.e. it lacks a large amount of porosity. This is a broken sample. During breaking some overgrowths on fly ash grains have pulled away leaving very smooth fly ash cores behind. The 3000x image provides a closer look at the matrix. It seems to consist of intergrown platelets which could be jennite-like calcium silicate hydrate. Figure 3: 50:50 mixture cured for 112 days. Compared to the 50:50 mixture this sample appears to have a greater amount of porosity and platy crystals. These crystals are less massive and do not fill space as well. The 3000x image confirms this more platy/needle like appearance. The crystals appear fuzzy-they seem to have small powdery crystals as overgrowths. This could be carbonation associated with thaumasite?

10 Considering the fact that the starting materials were Sherco Class C FGD ash and Class F Ft. Martin ash mixed with 4M NaOH solution one can image the following progression of reactions. These pastes consist of a mixture of varying sizes of glassy fly ash spheres (1-30 µm) that are surrounded by films of 4M NaOH solution. Packing of the various particles occurs during mixing as small particles fill spaces between close packed larger particles. With time the NaOH dissolves the most reactive glasses first. The Sherco is Ca-rich, which, from a glass technologist viewpoint, is a crumby glass. It is very soluble, easily etched even with plain water. It probably goes into solution and as concentrations of Ca, Al and Si in solution build, calcium aluminosilicate hydrates begin to nucleate and grow. Precipitates form on the surfaces of the Class F ash and on the dissolving Sherco ash. The Ft. Martin is also dissolving but at a slower rate. It is a low calcium glass that is more resistant to attack. Precipitates eventually bridge open spaces, connect particles and begin to fill in the previously water filled spaces. The SEM images suggest that both the 50:50 and 25:50 do a good job at space filling, although on first appearance, the 50:50 mixture seems to be better in regard to its general lack of open porosity. Cation Exchange Three of the better performing mixtures (the ones chosen from Table 2a earlier and some made with metakaolin instead of the Sherco or Ft. Martin fly ashes) were remixed using 8M NaOH and cured at RT and 40 C for 7 days. These were then ground to a powder and 100 mg of each powder was placed in 20 ml of a NaCl/CsCl solution on a shaking table for 7 days. Then they were analyzed for Na and Cs and Kd values were calculated. The formulations and cation exchange data are given in Table 5. The best performing RT cured mixture was the Ft Martinmetakaolin formulation [Kd (Cs)=26.0%], followed by the Sherco-Metakaolin formulation [Kd (Cs)=17.6]%. The least adsorptive were the 50:50 Sherco:Ft. Martin fly ash samples (Sherco-Ft Martin with a Kd (Cs)=7.8%). These numbers were reported earlier in [4]. They are indicative of trends. Samples were made with 8M NaOH, but it is suggested that trends with 4M should be similar. Table5. Cation Exchange behavior of Three Mixtures Cured at RT and 40 C for 7 days Sample Tested Cs % adsorption = Kd (Cs) Na mg/l (A0-Af)/A0 ml/mg mg/l P24-0 Stock solution [ N CsCl dissolved in 0.02N NaCl] :50 Ft. Martin and Sherco fly ashes mixed with 8 M NaOH cured at 40 C 7 days % 5.4% 640 P24-1 P :50 Ft. Martin and Sherco fly ashes mixed with 8 M NaOH cured at RT 7 days % 7.8% 650 P :50 Ft. Martin fly ash and metakaolin* mixed with 8 M NaOH cured at 40 C 7 days % 29.9% 490 P :50 Ft. Martin fly ash and metakaolin mixed with 8 M NaOH cured at RT 7 days % 26.0% 490 P :50 Sherco fly ash and metakaolin mixed with 8 M NaOH cured at 40 C 7 days % 13.2% 640 P :50 Sherco fly ash and metakaolin mixed with 8 M NaOH cured at RT 7 days % 17.6% 620 P :50 Sherco fly ash and metakaolin mixed with 15 M NaOH cured at 40 C 7 days % 7.1% 670 *Adding metakaolin to any mixture as a substitute for either fly ash dramatically increases Cs selectivity.

11 It is hypothesized that the Sherco-metakaolin and Ft. Martin-Kaolin samples reacted faster than the two fly ashes themselves. Thus their Kd values were higher (i.e. better for Cs adsorption). With longer curing the RT samples will develop more zeolites and the Kd values should go up. An ordinary Portland cement paste will be similarly evaluated in the future. Metakaolin is a very reactive aluminosilicate derived from kaolinite clay by heating at 700 C. The heat treatment drives off the water and in doing so leaves behind a collection of highly disorganized platelets that no longer have long range order. Crystallite sizes may be 1 or 2 layers only rather than massive kaolinite books present in natural clay. It is suggested that the 4 month samples discussed above (25/75 Ft. Martin/Sherco will perform better because they now contain zeolites). These tests are underway. Conclusions The study that was conducted showed that it was possible to make concrete like equivalents from Class C and Class F fly ashes activated with 4M NaOH. These mixtures were mixed in a conventional fashion to a consistency that could be pumped into an empty waste tank. The concrete would fill space with our without vibration and it tended to harden to the touch overnight. The long working time would allow one to fill the tank using a continuous pour. This will allow the contractor a bit more flexibility in making his pours without the fear of having dry joints between layers. The heat given off by the hydration process should be lower vis à vis Portland cement concrete, an issue that could be of concern when pouring mass concrete. This was an important consideration when one is pouring a million cubic feet of concrete. The 50:50 mixture was stronger than the 25:75 mixture, but it did not develop more than a trace of crystalline zeolites over a 4 month period. It is now recommended that the tanks be filled with the 25% Sherco:75% Ft. Martin ash. Tests are currently underway to evaluate two things: long term compressive strength and the ability to attract and hold Cs and Sr. The ability to attract and hold ions is called cation exchange. This test is performed by suspending sieved/sized cured tank fill material powder in dilute NaCl solutions containing CsCl and then monitoring the change in concentration of Cs and Na in the solution after 24 or more hours in the solution. Future work Follow up work is underway. A series of the 25:75 wt% Sherco:Ft. Martin ashes mixed with 4M NaOH have been formulated as concretes and molded in 3x6inch cylinders. These are being cured at C (RT) for 7, 28, 56, 90 days. Samples from the compression tests will be sieved to isolate the matrix phase for cation exchange and X-ray diffraction analysis. The inclusion of metakaolinite as a partial replacement of Class F fly ash (same composition) may provide an alternate route to increasing the amount of zeolite in a given concrete and in turn increasing the cation exchange properties of the tank fill material. Once a formulation has been chosen it will be evaluated for shrinkage, freeze thaw resistance, and wet dry cycling stability. These data will be published as a third report on tank fill materials for Hanford.

12 References 1. Bao, Y., M.W. Grutzeck, C.M. Jantzen, Preparation and Properties of Hydroceramic Waste Forms Made with Simulated Hanford Low-Activity Waste, Journal American Ceramic Society 88(12), (2005) 2. Lorier, T.H., D.H. Miller, J.R. Harbour, C.A. Langton, W. L. Mhyre, Grout Formulations for Closing Hanford High-Level Waste Tanks-Bench-Scale Study, pp in Ceramic Transactions 168 (Environmental Issues and Waste Management Technologies in the Ceramic and Nuclear Industries X), American Ceramic Society, Westerville OH (2005) 3. Bao, Y., M.W. Grutzeck, Solidification of Sodium Bearing Waste using Hydroceramic and Portland Cement Binders, pp in Ceramic Transactions 168 (Environmental Issues and Waste Management Technologies in the Ceramic and Nuclear Industries X), American Ceramic Society, Westerville OH (2005) 4. Belot, G., Y. Bao, M.W.Grutzeck, Preliminary Studies of Tank Fill Materials for Radioactive Waste Management at Hanford and Savannah River, January 30, Submitted to DOE. Available on line or from the authors. 5. Palomo, A., M.W. Grutzeck, M.T. Blanco, Alkali-Activated Fly Ashes: A Cement for the Future, Cement Concrete Research 29(8), (1999). 6. Krishnamurthy, N., M.W. Grutzeck, S. Kwan, D.D. Siemer, Hydroceramics for Savannah River Laboratory's Sodium Bearing Waste, pp in Ceramic Transactions 119 (Environmental Issues and Waste Management Technologies in the Ceramic and Nuclear Industries VI), American Ceramic Society, Westerville OH (2001) 7. Bao, Y., S. Kwan, D.D. Siemer, M.W. Grutzeck, Binders for Radioactive Waste Forms Made from Pretreated Calcined Sodium Bearing Waste (SBW), Journal Materials Science 39(2), (2004) 8. Bao, Y. M.W. Grutzeck, General Recipe and Properties of a Four Inch Hydroceramic Waste Form, pp in Ceramic Transactions 176 (Environmental Issues and Waste Management Technologies in the Ceramic and Nuclear Industries XI) American Ceramic Society, Westerville OH (2006) 9. Sumajouw, M.D.J. and B.V. Rangan, Low-Calcium Fly Ash-Based Geopolymer Concrete: Reinforced Beams and Columns, Research Report GC 3, Faculty of Engineering, Curtin University of Technology, Perth, Australia (2006). Available on line from the Geopolymer Institute Hardjito, D. and B.V. Rangan, Development and Properties of Low-Calcium Fly Ash-Based Geopolymer Concrete, Research Report GC 1, Faculty of Engineering, Curtin University of Technology, Perth, Australia (2005). Available on line from the Geopolymer Institute.