Publications

Absence of mineral colloids in high ionic strength solutions associated with salt formations: Experimental determination and applications for nuclear waste management

Xiong, Yongliang X.; Kirkes, Leslie D.; Kim, Sungtae K.; Marrs, Cassandra M.; Knox, Jandi L.; Dean, Justin; Deng, Haoran; Nemer, Martin N.

Radionuclides and heavy metals easily sorb onto colloids. This phenomenon can have a beneficial impact on environmental clean-up activities if one is trying to scavenge hazardous elements from soil for example. On the other hand, it can have a negative impact in cases where one is trying to immobilize these hazardous elements and keep them isolated from the public. Such is the case in the field of radioactive waste disposal. Colloids in the aqueous phase in a radioactive waste repository could facilitate transport of contaminants including radioactive nuclides. Salt formations have been recommended for nuclear waste isolation since the 1950's by the U.S. National Academy of Science. In this capacity, salt formations are ideal for isolation of radioactive waste. However, salt formations contain brine (the aqueous phase), and colloids could possibly be present. If present in the brines associated with salt formations, colloids are highly relevant to the isolation safety concept for radioactive waste. The Waste Isolation Pilot Plant (WIPP) in southeast New Mexico is a premier example where a salt formation is being used as the primary isolation barrier for radioactive waste. WIPP is a U.S. Department of Energy geological repository for the permanent disposal of defenserelated transuranic (TRU) waste. In addition to the geological barrier that the bedded salt formation provides, an engineered barrier of MgO added to the disposal rooms is used in WIPP. Industrial-grade MgO, consisting mainly of the mineral periclase, is in fact the only engineered barrier certified by the U.S. Environmental Protection Agency (EPA) for emplacement in the WIPP. Of interest, an Mg(OH)2-based engineered barrier consisting mainly of the mineral brucite is to be employed in the Asse repository in Germany. The Asse repository is located in a domal salt formation and is another example of using salt formations for disposal of radioactive waste. Should colloids be present in salt formations, they would facilitate transport of contaminants including actinides. In the case of colloids derived from emplaced MgO, it is the hydration and carbonation products that are of interest. These colloids could possibly form under conditions relevant in particular to the WIPP. In this chapter, we report a systematic experimental study performed at Sandia National Laboratories in Carlsbad, New Mexico, related to the WIPP engineered barrier, MgO. The aim of this work is to confirm the presence or absence of mineral fragment colloids related to MgO in high ionic strength solutions (brines). The results from such a study provides information about the stability of colloids in high ionic strength solutions in general, not just for the WIPP. We evaluated the possible formation of mineral fragment colloids using two approaches. The first approach is an analysis of long-term MgO hydration and carbonation experiments performed at Sandia National Laboratories (SNL) as a function of equivalent pore sizes. The MgO hydration products include Mg(OH)2 (brucite) and Mg3 Cl(OH)5•4H2O (phase 5), and the carbonation product includes Mg5(CO3)4(OH)2•4H2O (hydromagnesite). All these phases contain magnesium. Therefore, if mineral fragment colloids of these hydration and carbonation products were formed in the SNL experiments mentioned above, magnesium concentrations in the filtrate from the experiments would show a dependence on ultrafiltration. In other words, there would be a decrease in magnesium concentrations as a function of ultrafiltration with decreasing molecular weight (MW) cut-offs for the filtration. Therefore, we performed ultrafiltration on solution samples from the SNL hydration and carbonation experiments as a function of equivalent pore size. We filtered solutions using a series of MW cut-off filters at 100 kD, 50 kD, 30 kD and 10 kD. Our results demonstrate that the magnesium concentrations remain constant with decreasing MW cutoffs, implying the absence of mineral fragment colloids. The second approach uses spiked Cs+ to indicate the possible presence of mineral fragment colloids. Cs+ is easily absorbed by colloids. Therefore, we added Cs+ to a subset of SNL MgO hydration and carbonation experiments. Again, we filtered the solutions with a series of MW cut-off filters at 100 kD, 50 kD, 30 kD and 10 kD. This time we measured the concentrations of Cs. The concentrations of Cs do not change as a function of MW cut-offs, indicating the absence of colloids from MgO hydration and carbonation products. Therefore, both approaches demonstrate the absence of mineral fragment colloids from MgO hydration and carbonation products. Based on our experimental results, we acknowledge that mineral fragment colloids were not formed in the SNL MgO hydration and carbonation experiments, and we further conclude that high ionic strength solutions associated with salt formations prevent the formation of mineral fragment colloids. This is due to the fact that the high ionic strength solutions associated with salt formations have high concentrations of both monovalent and divalent metal ions that are orders of magnitude higher than the critical coagulation concentrations for mineral fragment colloids. The absence of mineral fragment colloids in high ionic strength solutions implies that contributions from mineral fragment colloids to the total mobile source term of radionuclides in a salt repository are minimal.