Edited by: Rizlan Bernier-Latmani, École Polytechnique Fédérale de Lausanne, Switzerland
Reviewed by: Eric D. van Hullebusch, UMR7154 Institut de Physique du Globe de Paris (IPGP), France; Naji Milad Bassil, The University of Manchester, United Kingdom
This article was submitted to Microbiological Chemistry and Geomicrobiology, a section of the journal Frontiers in Microbiology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Deep geologic repositories (DGR) in Canada are designed to contain and isolate low- and intermediate-level radioactive waste. Microbial degradation of the waste potentially produces methane, carbon dioxide and hydrogen gas. The generation of these gases increase rock cavity pressure and limit water ingress which delays the mobility of water soluble radionuclides. The objective of this study was to measure gas pressure and composition over 7 years in experiments containing cellulosic material with various starting conditions relevant to a DGR and to identify micro-organisms generating gas. For this purpose, we conducted experiments in glass bottles containing (1) wet cellulosic material, (2) wet cellulosic material with compost Maker, and (3) wet cellulosic material with compost Accelerator. Results demonstrated that compost accelerated the pressure build-up in the containers and that methane gas was produced in one experiment with compost and one experiment without compost because the pH remained neutral for the duration of the 464 days experiment. Methane was not formed in the other experiment because the pH became acidic. Once the pressure became similar in all containers after 464 days, we then monitored gas pressure and composition in glass bottle containing wet cellulosic material in (1) acidic conditions, (2) neutral conditions, and (3) with an enzyme that accelerated degradation of cellulose over 1965 days. In these experiments, acetogenic bacteria degraded cellulose and produced acetic acid, which acidity suppressed methane production. Microbial community analyses suggested a diverse community of archaea, bacteria and fungi actively degrading cellulose. DNA analyses also confirmed the presence of methanogens and acetogens in our experiments. This study suggests that methane gas will be generated in DGRs if pH remains neutral. However, our results showed that microbial degradation of cellulose not only generated gas, but also generated acidity. This finding is important as acids can limit bentonite swelling and potentially degrade cement and rock barriers, thus this requires consideration in the safety case as appropriate.
Nuclear power plant operations produce low- and intermediate-level radioactive waste (LILW). Low-level waste is radioactive waste in which the concentration of radionuclides is above 1 Becquerel per gram and contains primarily radionuclides with half-lives shorter than or equal to 30 years. Intermediate-level waste is radioactive non-fuel waste, containing significant quantities of radionuclides with half-lives greater than 30 years (
Low- and intermediate-level radioactive waste typically consist of ion-exchange resins, filters, carbon steel and ample cellulosic material (
Reaction involved in the biodegradation of cellulose and resulting gas formation.
Microbial process | Reaction equation | Reaction # |
---|---|---|
Hydrolysis of solid cellulose into β- |
Cellulose → β- |
Reaction 1 |
Hydrolysis of β- |
CH2O → organic acids + CO2 + H2 | Reaction 2 |
Hydrogenotrophic methanogenesis | 4 H2 + CO2 → CH4 + 2 H2O | Reaction 3 |
Acetogenesis | 4 H2 + 2CO2 CO2 → CH3COOH + 2 H2O | Reaction 4 |
Aceticlastic methanogenesis | CH3COOH → CH4 + CO2 | Reaction 5 |
Degradation of cellulose-based waste by fungal, bacterial and archaeal communities involves several carbohydrate active enzymes, such as the glycoside hydrolases (
It is challenging to estimate the length of time that cellulosic material will be available to microbial degradation because of the limited number of microbial organisms able to degrade solid cellulose. As a result, microbial degradation of solid cellulose can often be the rate limiting step. Some studies suggest cellulose and associated breakdown product β-
The rate of cellulose degradation is important to predict how a DGR will isolate and contain radionuclides in the LILW because it will influence the timing and onset of gas generation and subsequent pressure build-up over time. The sooner the gas pressure rises to equilibrium with preconstruction water pressure, the sooner water stops entering a DGR, which will dry up and remain as such as long as its gas pressure is maintained above the water pressure and as long as water is formed through methanogenesis and acetogenesis. However, if water vapor becomes limited, dry conditions could limit microbial activity, the gas pressure would slowly decrease and water will eventually resume migrating into the DGR. Then, microbial gas generation would be stimulated as long as a cellulosic substrate is available. This wet and dry cycle is likely to happen as long as cellulosic material remains available for microbial activity. As a result, during dry periods, water soluble radionuclides such as Cl-36 and I-129 would be effectively immobilized. In contrast, gaseous radionuclides such as tritium and C-14 could migrate if a pathway exists within a fault or within the excavated damage zone between the host rock and the shaft seal material. The extent of C-14 and tritium gas migration will depend on the pressure build-up in the DGR cavity. As carbon dioxide and hydrogen gas are formed, the pressure could increase and favor gaseous migration. In such situation, it should be noted that the potential migration of C-14 and tritium gases with relatively short half-lives is less of a concern because DGRs are designed to contain and isolate radionuclides with much longer half-lives. In addition, methanogenesis is assumed to decrease gas pressure by consuming both carbon dioxide and hydrogen, thereby decreasing the likely migration of C-14 and tritium gas. Hence, ideal conditions to contain water soluble and gaseous radionuclides are at a pressure that limits water ingress but also limits gaseous radionuclide migration.
Therefore, it is important to determine how microbial degradation of cellulosic material will affect the gas composition and pressure in a DGR. As there is currently no underground laboratory or DGR in Canada, we initiated laboratory experiments over a 7 years timeframe. The objective of these laboratory studies was to measure gas pressure and composition, as well as to identify microbial groups and functions associated with the production of gas.
To test whether methane can be the terminal gas produced within a repository isolating and containing LILW, we conducted 6 un-replicated experiments using surrogate cellulosic material. Initial starting conditions were either without or with commercial compost additives to accelerate the evolution of gases. We first monitored gas pressure and composition in two experiments with commercial compost additives and in an additional experiment without commercial compost additives (control) to investigate the effect of additives on gas pressure and composition. Once pressure build-up stabilized after 464 days, we initiated new experiments in which we monitored gas pressure and composition in one experiment with added cellulase enzymes from a fungal source, in another experiment in acidic conditions (pH 5) and a third experiment at neutral pH for an additional 1965 days.
We used mop heads as surrogate cellulosic material because these are used for routine cleaning of nuclear facilities located within radiological zones. Approximately 100 g of mop material was used for each experiment. For the first three experiments, mop samples were wetted with sterile reverse osmosis deionized water. Mop samples were wetted to achieve at least 60% of the moisture holding capacity. The wetted samples were then inserted into separate sterile 1 L glass bottles. Before sealing the bottles, we added 0.031g of Green Cross Original Fertosan compost maker which contained 11% total nitrogen, 5% phosphoric acid, 5% soluble potash, 2% magnesium, 1% calcium, and 55% organic matter into one bottle (Compost Maker treatment). In the second bottle, we added 0.36 g of Green Earth 100% natural compost accelerator containing bone meal and char, feather meal, sulfate, potash, granular calp, and soil enzymes (Compost accelerator treatment). These two mop samples wetted with sterile water were amended to test the effects of additives on the rate of gas generation.
Following the 464 days experiments, three new experiments were initiated with new cellulosic material to test the effect of an additive and pH on gas pressure and composition. One mop sample was wetted with sterile neutral phosphate buffer (BioUltra, >0.044 M Na2HPO4 from Sigma-Aldrich) and another mop sample was wetted with citrate buffer (pH 5). The other container received 1 unit of purified cellulase enzyme from
Sample names, starting condition and the elapsed time in days when the headspace gases were sampled.
Experiment name | Starting condition | Headspace sampling: elapsed time in days |
---|---|---|
Control (un-amended) | No commercial compost additive added | 464 |
Compost accelerator | Commercial blend: nutrients and enzymes | 464 |
Compost maker | Commercial blend: nutrients and microorganisms | 464 |
Neutral | pH 7 | 150, 730, and 1965 |
Acidic | pH 5 | 150, 730, and 1965 |
Cellulase | pH 7 with Cellulase from |
150, 730, and 1965 |
All six containers were sealed using media bottle lids that had been modified to allow temperature and pressure gauges to be situated inside of the bottle. These penetrations were sealed with stainless steel tubing. These penetrations also provided the means for periodic sampling of the container headspace for gases. The temperature and pressure inside the containers were monitored continuously using thermocouples and differential pressure cells (Cole Parmer, 14.7–15 psig). The pressure measurements were corrected for temperature. Data from the thermocouples and pressure cells were logged automatically using a Keithley data logger. To prevent over pressurizing the containers as the mop heads degraded, each container was fitted with a 68.95 kPa (10 psi) pressure relief valve.
Headspace gas analysis was performed using a Agilent 6890N gas chromatograph (GC) by connecting the pressurized sample containers (vessels) to a vacuum line, and then evacuating the line to a pressure <1 torr absolute, as measured by an MKS Baratron Capacitance Manometer. The vacuum pump was then isolated, and a sample of the headspace gas within the container was bled into the vacuum system by a low flow metering valve to a pressure of roughly 760 torr. The sample inlet was then isolated, and time was given for the pressure to stabilize in the vacuum system. This process was repeated three times on each sample to ensure sample consistency. A 100 μL sample of the analyte gas was then directed to the GC using a 6-port 2-position sampling valve. The GC was configured with an Agilent HP-PLOT Molsieve 15 m column for separation of permanent gases (Ar, O2, N2, H2, and CH4), and an Agilent GASPRO 30 m column for separation of corrosive gases (CO2 and H2S). A valve was installed to allow for column selection midway through the GC run. The analysis was performed independently using both He and Ar carrier gases for each sample to ensure detection of hydrogen gas. Detection of the gases was performed using a Thermal Conductivity Detector (TCD) and a reference gas matching the carrier (both He and Ar, respectively). Calibration of the GC was performed using certified standard mixes of the analyte gases using the same sample introduction scheme at sample pressures ranging from 10 to 760 torr. Examples of the analysis output are provided in
The volume of the leachate obtained from each experiment, although sufficient for DNA analysis, was not sufficient to allow for pH measurement with conventional pH meter electrodes. Therefore, the pH was measured using pH-indicator paper. The indicator paper provided reasonable precision to compare the pH in our leachate from one experiment to the other.
Following the final headspace gas sampling, the sample bottles were opened and two pieces of material approximately 1 cm in size were cut out and processed for DNA extraction using the protocol from the PowerSoil® DNA Isolation kit (MoBio, 12888). DNA extracts were also obtained from liquid leachate collected from the samples using the Rapid Water DNA extraction kit. Leachate (5 ml) was first filtered through a 0.2 μm Supor polyether sulfone membrane. The extracted DNA was quantified by fluorescence using the intercalating dye, PicoGreen, from the Quant-iT PicoGreen dsDNA assay kit (Life Sciences, P7589). The detected concentration ranged from 0.08 to 0.39 ng/μL, with two samples below our detection limit (
Microbial communities are challenging to characterize due to the phylogenetic and functional diversity required to degrade solid wastes. To gauge the cellulolytic, fermentative, sulfate-reducing, acetogenic and methanogenic microbes, a literature search was performed to find primers that targeted the genes of interest. The cellulase enzymes are involved in the early steps of cellulose degradation by catalyzing the hydrolysis of glycosidic bonds. There are several different types of cellulases based on the reaction to be catalyzed. The glycoside hydrolase family of cellulose enzymes consist of 131 protein families based on their sequence and structure information (
Details of the primer sequences, targets, expected amplicon size and thermal cycler conditions are provided in
All ddPCR reactions were set up inside of a laminar flow hood wiped with 70% ethanol and exposed to ultraviolet light for 3 min prior to entering the hood. Individual reactions for ddPCR contained a final primer concentration of 150 nM with 2x QX200 ddPCR EvaGreen Supermix (Bio-Rad #186–4033) following manufacturer’s instructions. Two microliters of DNA were loaded into a total reaction volume of 21 μL. All ddPCR runs included a no template negative control and a positive control on some occasion.
To determine the optimal concentration of DNA for ddPCR, a dilution series was constructed using two samples: one sample that displayed high expression of the target and one that displayed low expression of the target from the PCR.
To determine the optimal annealing temperature (
The gene abundance (concentration) was reported as copy number per μL. The positive droplets and the total accepted droplets were also reported. To fulfill Poisson’s distribution rules, samples must have more than 13,000 droplets to be analyzed. If the samples did not reach this threshold, individual samples were removed. The threshold to determine quantification of the target was set manually using samples with a well-discriminated fluorescence difference between positive and negative droplets.
Cellulase activity within the leachate was determined after opening the sample bottles. The mop materials were removed from the bottles, and the liquid was extracted from the top, middle, and bottom portions of each sample by inserting each portion into a sterile 30 mL syringe and squeezing liquid into sterile plastic tubes. The leachates were stored at -20°C until analyzed for cellulase activity. Cellulase activity was measured in the surrogate wastes using the fluorometric-based Cellulase Activity Assay Kit (Abcam, ab189817) that detects the release of a fluorescent compound, resorufin, from a cellulase substrate, resorufin-β-
The development of gas pressure inside each of the test containers varied in terms of the timing of gas pressure onset, the rate of gas pressure onset and the rate of gas pressure change over the monitoring period. The rate of change of pressure within the un-amended test container displayed an initial decline in gas pressure that lasted 80 days, followed by a slight increase starting at about day 130, which then leveled off at 115 kPa for the next 120 days before displaying another period of increasing pressure to 130 kPa at day 464 (
Evolution of gas pressure and headspace gas compositions without adding compost (control) and after adding Compost Accelerator and Compost Maker over 464 days
Community differences for cellulose and leachate by differential relative abundances of gene copies detected within the surrogate waste for: Archaeal rRNA (A16S), fungal rRNA (18s), and bacterial glutamine synthetase (glnA), genes.
Addition of the compost additives, Compost Maker (with microorganisms within its composition) and Compost Accelerator (with enzymes within its composition), increased the rate of gas pressure onset (
The cellulosic material amended with the additive that included microorganisms within its composition (Compost Maker,
The cellulosic material with added cellulase displayed an initial rapid rate of gas pressure rise, followed by a rapid leveling off within the 1st week. Gas pressure then remained constant at 126 kPa for over 300 days. The gas pressure continued to evolve with a second rapid rise to 185 kPa followed by a period of variable pressure changes before displaying a continuous decline to similar pressure measured in the neutral experiment at day 1965. The gas composition was 66, 21, and 69% carbon dioxide, 30, 71, and 30% nitrogen, 0, 0.01, and 0% methane, and 3, 7, and 0.1% hydrogen at days 150, 730, and 1965, respectively. The pH in the leachate at the end of the experiment was at pH 3. It should be noted that hydrogen gas had depleted considerably by day 1965.
The test containers for the acidic starting condition displayed an initially rapid rate of gas pressure rise, followed by a rapid leveling off within the 1st week of starting the gas pressure monitoring, the duration of which lasted about 50 days. Afterwards, gas pressure rapidly declined to 110 kPa and remained at this pressure until the end of monitoring at day 1965 (
The test container with cellulosic material exposed to neutral pH conditions at the start of gas pressure monitoring displayed a gas pressure evolution that first declined then rose, followed by a rate of change that slowly leveled off over approximately 1500 days and remained unchanged at 132 kPa for the remainder of the monitoring period. The gas composition was 18, 14, and 25% carbon dioxide, 70, 79, and 72% nitrogen, 0, 0.01, and 0% methane, and 11, 6, and 0.8% hydrogen at days 150, 730, and 1965, respectively. The pH in the leachate was at pH 6. Similar to the cellulase and pH 7 experiments, hydrogen gas had depleted by day 1965.
The results of this assay are shown in
Cellulase activity based on the glycoside hydrolase family 1 and 3 activities detected in the sample leachates.
Sample | Portion of the sampled extracted | Emission 595 nm | Concentration of resorufin | Cellulase enzyme activity | Fraction of enzyme activity∗ |
---|---|---|---|---|---|
RLU | nM | μmole/mL/min | % | ||
Cellulase, 100× dilution | Top | 58.6 | 1834 (40) | 3.00 × 10-3 | 446 |
Middle | 60.4 | 1878 (23) | 3.13 × 10-3 | 465 | |
Bottom | 59.2 | 1824 (162) | 3.04 × 10-3 | 452 | |
Compost accelerator | Top | 20.7 | Not detected | – | – |
Middle | 20.7 | Not detected | – | – | |
Bottom | 20.5 | Not detected | – | – | |
Compost maker | Top | 35.0 | 15.2 (0.7) | 2.54 × 10-6 | 0.4 |
Middle | 50.4 | 28.8 | 4.79 × 10-5 | 7.1 | |
Bottom | 39.1 | 14.0 | 2.34 × 10-5 | 3.5 | |
Acidic un-amended, pH5 | Top | 21.4 | Not detected | – | – |
Middle | 21.7 | Not detected | – | – | |
Bottom | 21.7 | Not detected | – | – | |
Neutral un-amended, pH7 | Top | 60.9 | 33.1 (0.2) | 5.52 × 10-5 | 8 |
Middle | 105.4 | 77.1 (0.2) | 1.29 × 10-4 | 19 | |
Bottom | 152.0 | 118.1 (0.8) | 1.97 × 10-4 | 29 | |
Un-amended | Top | 25.0 | 1.7 (0.3) | 2.87 × 10-6 | 0.4 |
Middle | 24.6 | 1.4 (0.4) | 2.28 × 10-6 | 0.3 | |
Bottom | 23.8 | 0.7 (0.2) | 1.10 × 10-6 | 0.2 | |
The two test containers whose pH was acidic [the sample with an acidic starting condition and the sample with the added Compost Accelerator (
The results for diversity of bacteria, Archaea and fungi are shown in
Community differences for cellulose and leachate by differential relative abundances of gene copies detected within the surrogate waste for dissimilatory sulfate reductase (dsr1), glycoside hydrolase family 61 (gh61.3.2 and gh61.5.4), methyl co-enzyme M (mcrA), and acyl co-enzyme A synthase (acas) genes.
The Archaea were numerically dominant on the solid material across the samples except the sample amended with added cellulase (A16s,
The community compositions within the leachate were different from the community compositions from the solid cellulosic material (
In this study, we monitored gas pressure onset, gas pressure evolution and headspace gas compositions as a result of microbial degradation of cellulose. We also tested the microbiology of the cellulosic material indirectly by measuring cellulase activity and directly by quantifying specific marker gene targets. Here, we will discuss our results, and their implications for the design of DGRs containing and isolating LILW.
The microbial degradation of cellulose is a complex process (
In our tests, the only cellulolytic gene associated with the solid cellulosic material was from the fungal glycoside hydrolase family 61 (now called Auxiliary Activity Family 9). Copies of these genes were associated with both the solid cellulose (
Separately, we also detected glycoside hydrolase activity, from families 1 and 3, at various levels in the leachate (
The microbial degradation of cellulosic material generates carbon dioxide, methane, and hydrogen gas, which build up pressure in waste containers, and eventually in DGR caverns. In our tests, cellulosic material was subjected to various treatments establishing initial conditions for added microbes, enzymes and initial pH. The resulting gas evolution of pressure and composition was monitored throughout. In one set of tests, lasting 464 days, hydrogen gas was a minor component of the gas composition making up to 0.2% of the gas volume. However, it was above the ∼0.01% level of the measured threshold for methanogenesis (
In contrast, there was little methane formed in the Compost Accelerator experiment. Correspondingly, carbon dioxide and hydrogen levels were higher in proportion than in the other two experiments that lasted 464 days. In the Compost Accelerator experiment, the low pH likely suppressed hydrogenotrophic and acetoclastic methanogenesis (Reactions 3 and 5;
In the experiments that lasted 1965 days, carbon dioxide and hydrogen were the predominant gases. Hydrogen gas made up to 11% of the gas volume in the first 730 days of these experiments only (
The gas pressures were relatively similar between the different experiments with the exception of the acidic experiment (
Evidence from our laboratory experiments suggests that if pH is maintained at neutrality in waste containers, there will be active microbial degradation of cellulose, which will produce carbon dioxide and hydrogen gas, which can then transform into methane. These conditions are predicted by typical gas models (
However, some experiments in our study suggest microbial degradation of cellulosic material within waste containers will generate acid and limit gas pressure build-up. At start, the fermentation of cellulosic material contained in LILW would produce organic acids (Reaction 2;
It is possible that acidic leachate accumulate in several waste containers within a repository. As the acidic leachate build-up inside the containers, it can slowly deteriorate the waste containers until it is punctured and the acidic leachate is released into the repository cavity. Low- and intermediate-level waste are often encased in cement within deep geological repositories (
The length of the advection channels and whether they could reach the host rock or shaft seal would depend on the amount of acidic leachate accumulated inside each waste container but also on the gas pressure build-up inside the DGR cavity. Our experiment suggests that there is very low gas pressure when acidic conditions are observed inside waste containers. Therefore, it is possible that gas pressure remain low in DGR cavities. Low pressure would allow resaturation with meteoric water of the DGR cavities depending on hydraulic conductivity in shaft seal and host rock and the presence of fractures. If the channels extend outside of the cement barriers, the acidic leachate could be buffered by the accumulating water that bathes the cement material at pH > 12 and the host rock. Further channeling of the host rock and shaft seal material by acidic leachate would therefore be limited.
Based on this reasonable scenario, the extent of barrier degradation, in combination with mobility of water soluble radionuclides, would require additional consideration when designing a DGR. For instance, sufficient quantities of Portland Cement may be required not only in the shaft seal and as support in the DGR cavity chambers and tunnels, but as backfill into repository chambers to buffer the production of acidity from degradation of cellulose in the waste container. The importance of buffering material for the management of LILW is acknowledged by the recent results from the long-term gas generation experiment in Finland.
One uncertainty of this study is if our results could be reproducible at high salinity as encountered in deep groundwater in sedimentary rock. Although this is less of an issue in granitic rock, it is possible that cellulosic degradation at high salt concentration be limited.
Our experiments were conducted in surrogate cellulosic material that did not contain any radioactivity. Radiations will be present in geological repositories but total absorbed doses and dose rates will cover a large range because of the different radionuclide inventories of waste packages, their spatial distribution and their evolution with time. Although ionizing radiation is detrimental to the survival of living organisms, a number of different mechanisms are used by bacteria to avoid lethal damage (e.g.,
Long-term pressure monitoring of sealed containers and periodic analysis of the headspace gas composition showed that methane may not be the terminal gas if the acidity generated from the microbial degradation of cellulose is not sufficiently buffered to a pH above 6. Further replication with other waste types is warranted to substantiate our observations. In the meantime, predicting gas generation within a DGR safety case first requires quantification of cellulose in low-level waste. Then, a prediction of acid generation from cellulose degradation is needed along with an estimate of the localized buffering capacity of the cement and host rock. The acid generation prediction is complicated by the biodegradation of cellulosic material having multiple steps, which is rate limited by the initial breakdown of solid cellulose. The cellulose breakdown rate is also reduced by its own production of acidity because the cellulase enzyme activity is inhibited. If the cellulose in low-level waste is present in sufficient quantity to produce enough acidity to puncture the containers, the released acidic plume may dissolve the cement used as backfill and create channels. The extent of channels need to be estimated to determine if modification to DGR designs are required to prevent the migration of water soluble radionuclides.
The extent of deterioration of natural and engineered barriers caused by the generation of acidity would remain localized by excavating geological repositories several hundred meters below ground level. Siting repositories at great depth provides isolation, and additional margin of safety to counter balance uncertainties associated with the impact of acidity on natural and engineered barriers. In addition, international guidance requires demonstration that even if the shaft seal material of DGRs is severely degraded by any agent including acidity, dose to the public and the environment would remain acceptable for hundred thousands of years until the radioactive inventory decay to background levels.
Nonetheless, future research should investigate if there is enough cellulosic material in low-level waste to generate sufficient acidity to affect natural and engineered barriers. This is important to reduce uncertainty in DGR safety cases and optimize the depth of DGR hosting LILW so that people and the environment are adequately protected.
DB designed and conducted the experiments. PP was responsible for DNA analysis and primers and reviewed sections pertaining to primers and ddPCR. RG participated in the design of the experiments and provided oversight of the project and wrote the manuscript with DB.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We would like to thank Grace Chen, Mandy Serran, Jeff Keyes, and Randy Perron for laboratory assistance. We would also like to thank Joe Small, Minna Vikman, Kristel Mijnendonckx, and Natalie Leys for constructive comments during the 2018 Microbiology in Nuclear Waste Disposal workshop, Lausanne, Switzerland. We also thank Grant Ferris, Simsha Stroes-Gascoyne, Ian Clark, Joe Small and Philip Bennett for their discussion about acidity production in DGR at a workshop organized by the Canadian Nuclear Safety Commission in 2013.
The Supplementary Material for this article can be found online at: