Corroded Magnox sludge and plutonium waste cementation Stephen A. Parry and Francis R. Livens

1. Corroded Magnox sludge and plutonium waste cementationStephen A. Parry and Francis R. Livens Thank Chair � greet room. Good afternoon everyone � today I�m going to present the data from our study of corroded Magnox sludge formed in spent fuel storage pond, its chemical form, its interaction with plutonium, and the results from initial studies of the structure of cement for immobilisation of this waste.Thank Chair � greet room. Good afternoon everyone � today I�m going to present the data from our study of corroded Magnox sludge formed in spent fuel storage pond, its chemical form, its interaction with plutonium, and the results from initial studies of the structure of cement for immobilisation of this waste.

2. Outline of talk Introduction (context) Magnox fuel storage ponds Study of corroded Magnox sludge (CMS) Interaction of CMS with Pu Characterisation of CMS Carbonation experiment results (real-time synchrotron study) Cementation Zr doped cement structure Current / further work So an outline of the talk � I�ll first give some background to this study with a brief description of Magnox fuel storage ponds where this CMS waste we�re interested in cementing is generated. And then the I�ve split the talk into 2. First I�ll go over our characterisation of CMS and its interaction with Pu in the storage pond environment. And then I�ll present our initial data on the structure of cements, similar to those that are intended for immobilisation of this CMS waste. Finally I�ll end on the intended extension of this work.So an outline of the talk � I�ll first give some background to this study with a brief description of Magnox fuel storage ponds where this CMS waste we�re interested in cementing is generated. And then the I�ve split the talk into 2. First I�ll go over our characterisation of CMS and its interaction with Pu in the storage pond environment. And then I�ll present our initial data on the structure of cements, similar to those that are intended for immobilisation of this CMS waste. Finally I�ll end on the intended extension of this work.

3. Magnox fuel storage ponds Spent Magnox fuel rods stored in large water-filled storage ponds Pond maintained at high pH to minimise corrosion of the fuel rods magnesium-alloy cladding However, cladding easily corroded in aqueous environment � and large quantity of corroded sludge has accumulated (primarily at Sellafield site) Include: ponds constantly purged to limit organic and carbonate concentrationsInclude: ponds constantly purged to limit organic and carbonate concentrations

4. Microfiltration experiments So from our previous studies on this system � we looked at how plutonium interacts with CMS in solution at a number of pH�s and with the interaction of a number of possible reagents with the storage pond / effluent treatment plant environment. After performing filtration experiments we found that plutonium shows a strong affinity to CMS in solution even at low concentrations. In our experiments were Pu was present in 10 micromolar concentrations and CMS was present at 3 ppm, So from our previous studies on this system � we looked at how plutonium interacts with CMS in solution at a number of pH�s and with the interaction of a number of possible reagents with the storage pond / effluent treatment plant environment. After performing filtration experiments we found that plutonium shows a strong affinity to CMS in solution even at low concentrations. In our experiments were Pu was present in 10 micromolar concentrations and CMS was present at 3 ppm,

5. Pu association with CMS Pu found to associate with CMS particulate / colloid (Kd = 1.5 x 106 ml/g) CMS within storage pond contaminated with particles of spent fuel and Pu sorbed from the pond liquor. So from our previous studies on this system � we looked at how plutonium interacts with CMS in solution at a number of pH�s and with the interaction of a number of possible reagents with the storage pond / effluent treatment plant environment. After performing filtration experiments we found that plutonium shows a strong affinity to CMS in solution even at low concentrations. In our experiments were Pu was present in 10 micromolar concentrations and CMS was present at 3 ppm, So from our previous studies on this system � we looked at how plutonium interacts with CMS in solution at a number of pH�s and with the interaction of a number of possible reagents with the storage pond / effluent treatment plant environment. After performing filtration experiments we found that plutonium shows a strong affinity to CMS in solution even at low concentrations. In our experiments were Pu was present in 10 micromolar concentrations and CMS was present at 3 ppm,

6. Project Overall goal � to achieve a better understanding of cementitious wasteforms for encapsulation of Magnox sludge (well characterised, predictable) Requirements The waste: Corroded Magnox sludge good understanding of the composition of CMS and its evolution with time. Cementation Have a knowledge of microstructure of cementitious wasteform investigate actinide stability within wasteform relate these two properties Evaluate cement evolution with time.

7. Corroded Magnox sludge (CMS) X-Ray diffraction and IR analysis - CMS composed of: Brucite - Mg(OH)2 Artinite - Mg2CO3(OH).3H2O

8. CMS characterisation - ESEM

9. Synchrotron time-resolved XRD experiment We�ve been successful in obtaining synchrotron beamtime to study the reaction of brucite with the atmosphere and carbonate saturated solution.We�ve been successful in obtaining synchrotron beamtime to study the reaction of brucite with the atmosphere and carbonate saturated solution.

10. Time-resolved Energy-dispersive XRD results

11. Time-resolved study continued Mention that more data recently been collected over greater temperature range with pH controlled with hydroxide. More time coming up.Mention that more data recently been collected over greater temperature range with pH controlled with hydroxide. More time coming up.

12. Recent / Further work Recently more synchrotron EDXRD data collected pH conditioned controlled (pH 9-11) Range of temperatures studied (20-60�C) More synchrotron time secured and scheduled A further set of time-resolved EDXRD were collected last month though with solution pH controlled by sodium hydroxide addition during the CO2 bubbling. Experiments were also made at elevated temperatures. At the minute this data is just being analysed but already we have observed a faster rate of reaction, as we anticipated. We have some further beamtime next month to collect more data which should allow us to derive some arrhenius related kinetics data. It will also show if the reaction pathways are the same with variations in temperature and pH.A further set of time-resolved EDXRD were collected last month though with solution pH controlled by sodium hydroxide addition during the CO2 bubbling. Experiments were also made at elevated temperatures. At the minute this data is just being analysed but already we have observed a faster rate of reaction, as we anticipated. We have some further beamtime next month to collect more data which should allow us to derive some arrhenius related kinetics data. It will also show if the reaction pathways are the same with variations in temperature and pH.

13. Cementation of legacy waste So on to the second and last part of my talk are aim to produce a well-characterised cement for the immobilisation of CMS waste. I should mentioned that much of this work has been helped along by Neil Milestone from Sheffield University, who has been a provided much of the background behind the current plans for immobilisation and also supplied our cement that is to the same specification to that currently used for real watse immobilisation. So CMS waste poses a challenge for these cements as it�s a Mg-rich, high pH sludge. So factors such as pH compatibility, and rheology need to be considered. Along with determination of the fate of actinides incorporated into the cement from the CMS. So factors such as to which material do actinides associate in the cement may significantly influence the stability and performance of the cement to retain actinides.So on to the second and last part of my talk are aim to produce a well-characterised cement for the immobilisation of CMS waste. I should mentioned that much of this work has been helped along by Neil Milestone from Sheffield University, who has been a provided much of the background behind the current plans for immobilisation and also supplied our cement that is to the same specification to that currently used for real watse immobilisation. So CMS waste poses a challenge for these cements as it�s a Mg-rich, high pH sludge. So factors such as pH compatibility, and rheology need to be considered. Along with determination of the fate of actinides incorporated into the cement from the CMS. So factors such as to which material do actinides associate in the cement may significantly influence the stability and performance of the cement to retain actinides.

14. Cementation Zr & U Ordinary Portland Cement (OPC) samples Microstructure images using: ESEM / SEM Elemental mapping using: EDX EMPA So we�ve began looking at trial cements doped with Zr and U � not quite on this scale�. Here�s a cast block and here are a few sectioned and polished samples. A number of sample have been examined using ESEM (say long), SEM (say long), and EMPA (say long) and election backscattered images, and elemental data have been collected. We�ve used Zr as an analogue for +4 actinides, but its intended to examine a U containing cement next, for which a sample has already been prepared. Before So we�ve began looking at trial cements doped with Zr and U � not quite on this scale�. Here�s a cast block and here are a few sectioned and polished samples. A number of sample have been examined using ESEM (say long), SEM (say long), and EMPA (say long) and election backscattered images, and elemental data have been collected. We�ve used Zr as an analogue for +4 actinides, but its intended to examine a U containing cement next, for which a sample has already been prepared. Before

15. ESEM image of Zr doped cement Cement after ~30 days curing: Anhydrous cement grains encapsulated in C-S-H gel phase and Ca(OH)2. Phases determined from greyscale GSE image. This is an example of one of cements under ESEM � we see large anhydrous cement grains, these are actually larger than in typical cements as larger cement grain are chosen for the cement as this improved mixing. Between the grains we have CSH (calcium silicate hydrate) phases, with brighter spot corresponding to Ca(OH)2. The very bright spots (point) are due to charging on the surface.This is an example of one of cements under ESEM � we see large anhydrous cement grains, these are actually larger than in typical cements as larger cement grain are chosen for the cement as this improved mixing. Between the grains we have CSH (calcium silicate hydrate) phases, with brighter spot corresponding to Ca(OH)2. The very bright spots (point) are due to charging on the surface.

16. ESEM � elemental analysis If we look at the energy dispersive elemental map of the sample we see most of the sample consists of calcium silicate phases with a high Mg, Al grain incorporated in the matrix.If we look at the energy dispersive elemental map of the sample we see most of the sample consists of calcium silicate phases with a high Mg, Al grain incorporated in the matrix.

17. EMPA: Zr doped cement If we look at a similar sample using EMPA we have a similar result � I�m sorry I don�t have an image of the sample but here we have the Zr results. No Zr in the cement grains but a significant amount in the intergranular spaces, and some small concentrated regions at the edge of grains. We see as before the grains are composed of Ca and Si. And there is some Fe present � thou the Fe and Zr are not associated with each other. EMPA offers better detection limits than SEM so we can work with smaller actinide concentrations � which is why we�re developing this technique.If we look at a similar sample using EMPA we have a similar result � I�m sorry I don�t have an image of the sample but here we have the Zr results. No Zr in the cement grains but a significant amount in the intergranular spaces, and some small concentrated regions at the edge of grains. We see as before the grains are composed of Ca and Si. And there is some Fe present � thou the Fe and Zr are not associated with each other. EMPA offers better detection limits than SEM so we can work with smaller actinide concentrations � which is why we�re developing this technique.

18. Further work Continue cement structure investigation with U containing cements Examine the effect of curing temperatures up to 60�C on the cement (accelerated aging) Prepare and examine a Pu containing cement Apply for synchrotron EXAFS time Determine Pu oxidation state & Pu co-ordination number Actinide leaching experiments (groundwater)

19. Neil Milestone & Claire Utton University of Sheffield Luke O�Brien Nexia Solutions Thank you for your attention Acknowledgements