[en] Model verification and validation are both important steps in the development of reactive transport models. In this paper, a distinction is made between verification and validation, and the focus is on codifying the issues of verification for a numerical, reactive transport flow model. First, the conceptual basis of model verification is reviewed, which shows that verification should be understood as a first step in model development, and be followed by a protocol that assures that the model accurately represents system behavior. Second, commonly used procedures and methods of model verification are presented. In the third part of this paper, an intercomparison of models is used to demonstrate that model verification can be performed despite differences in hydrogeochemical transport code formulations. Results of an example simulation of transport are presented in which the numerical model is tested against other hydrogeochemical codes. Different kinetic formulations between solid and aqueous phases used among numerical models complicates model verification. This test problem involves uranium transport under conditions of varying pH and oxidation potential, with reversible precipitation of calcium uranate and coffinite. Results between the different hydrogeochemical transport codes show differences in oxidation potentials, but similarities in mineral assemblages and aqueous transport patterns. Because model verification can be further complicated by differences in the approach for solving redox problems, a comparison of a fugacity approach to both the external approach (based on hypothetical electron activity) and effective internal approach (based on conservation of electrons) is performed. The comparison demonstrates that the oxygen fugacity approach produces different redox potentials and mineral assemblages than both the effective internal and external approaches

[en] CH2MHill Hanford Group, Inc. (CHG) is designing and assessing the performance of an Integrated Disposal Facility (IDF) to receive immobilized low-activity waste (ILAW), Low-Level and Mixed Low-Level Wastes (LLW/MLLW), and the Waste Treatment Plant (WTP) melters used to vitrify the ILAW. The IDF Performance Assessment (PA) assesses the performance of the disposal facility to provide a reasonable expectation that the disposal of the waste is protective of the general public, groundwater resources, air resources, surface water resources, and inadvertent intruders. The PA requires prediction of contaminant migration from the facilities, which is expected to occur primarily via the movement of water through the facilities and the consequent transport of dissolved contaminants in the pore water of the vadose zone. Pacific Northwest National Laboratory (PNNL) assists CHG in its performance assessment activities. One of PNNL's tasks is to provide estimates of the physical, hydraulic, and transport properties of the materials comprising the disposal facilities and the disturbed region around them. These materials are referred to as the near-field materials. Their properties are expressed as parameters of constitutive models used in simulations of subsurface flow and transport. In addition to the best-estimate parameter values, information on uncertainty in the parameter values and estimates of the changes in parameter values over time are required to complete the PA. These parameter estimates and information were previously presented in a report prepared for the 2001 ILAW PA. This report updates the parameter estimates for the 2005 IDF PA using additional information and data collected since publication of the earlier report

[en] In support of CH2M HILL Hanford Group, Inc.'s (CHG) preparation of a Field Investigative Report (FIR) for the Hanford Site Single-Shell Tank Waste Management Area (WMA) T and TX-TY, a suite of numerical simulations of flow and solute transport was executed using the STOMP code to predict the performance of surface barriers for reducing long-term risks from potential groundwater contamination at the T and TX-TY WMA. The scope and parametric data for these simulations were defined by a modeling data package provided by CHG. This report documents the simulation involving 2-D cross sections through the T Tank and the TX-TY Tank Farm. Eight cases were carried out for the cross sections to simulate the effects of interim barrier, water line leak, inventory distribution, and surface recharge on water flow and the transport of long-lived radionuclides (i.e., technecium-99 and uranium) and chemicals (i.e., nitrate and chromium). For simulations with barriers, it is assumed that an interim barrier is in place by the year 2010. It was also assumed that, for all simulations, as part of tank farm closure, a closure barrier was in place by the year 2040. The modeling considers the estimated inventories of contaminants within the vadose zone and calculates the associated risk. It assumes that no tanks will leak in the future. Initial conditions for contaminant concentration are provided as part of inventory estimates for uranium, technetium-99, nitrate, and chromium. For moisture flow modeling, Neumann boundary conditions are prescribed at the surface with the flux equal to the recharge rate estimate. For transport modeling, a zero flux boundary is prescribed at the surface for uranium, technetium-99, nitrate, and chromium. The western and eastern boundaries are assigned no-flux boundaries for both flow and transport. The water table boundary is prescribed by water table elevations and the unconfined aquifer hydraulic gradient. No-flux boundaries are used for the lower boundary. Numerical results were obtained for compliance at the WMA boundary, 200 Areas boundary, exclusion boundary beyond the 200 Areas, and the Columbia River (DOE-RL 2000). Streamtube/analytical models were used to route computed contaminant concentrations at the water table to the downstream compliance points. When the interim barrier was applied at 2010, the soil was desaturated gradually. The difference in saturation of the soil with and without the interim barrier was the largest at 2040, the time the closure barrier was applied. After this, the difference in saturation in the two cases became smaller with time. Generally, the solutes broke though faster if there was a water line leak. A relative small five-day leak (Case 4) had little effect on the peak concentration, while a large 20-yr leak (Case 3) increased the peak concentration significantly and reduced the solute travel in the vadose zone. The distribution of the inventory, either uniform or nonuniform, has little effect on peak arrival time; the peak concentrations of the conservative solutes varied by -6.9 to 0.2% for the T tank farm and by 11 to 49.4% for the TX tank farm. The reduction of the meteoric recharge before the barrier was applied led to less soil saturation, as expected, and thus longer solute travel time in the vadose zone and smaller peak fence line concentration. The effect on soil saturation lasted for about another 50 years after the barrier was applied at 2050. However, the reduced recharge rate affected the breakthrough curve till the end of the simulation. The fence line concentrations at the year 3000 were always higher for cases with reduced natural recharge than for those of the base case, which indicates that the fundamental impact of the reduced natural recharge is a smoothing of the breakthrough concentrations at the compliance points

[en] In support of CH2M HILL Hanford Group, Inc.'s (CHG) preparation of a Field Investigative Report (FIR) for the closure of the Hanford Site Single-Shell Tank (SST) Waste Management Area (WMA) tank farms, a set of numerical simulations of flow and solute transport was executed to predict the performance of surface barriers for reducing long-term risks from potential groundwater contamination at the C Farm WMA. This report documents the simulation of 14 cases (and two verification cases) involving two-dimensional cross sections through the C Farm WMA tanks C-103 - C-112. Utilizing a unit release scenario at Tank C-112, four different types of leaks were simulated. These simulations assessed the impact of leakage during retrieval, past leaks, and tank residual wastes and tank ancillary equipment following closure activities. . Two transported solutes were considered: uranium-238 (U-238) and technetium-99 (Tc-99). To evaluate the impact of sorption to the subsurface materials, six different retardation coefficients were simulated for U-238. Overall, simulations results for the C Farm WMA showed that only a small fraction of the U-238 with retardation factors greater than 0.6 migrated from the vadose zone in all of the cases. For the conservative solute, Tc-99, results showed that the simulations investigating leakages during retrieval demonstrated the highest WMA peak concentrations and the earliest arrival times due to the high infiltration rate before the use of surface barriers and the addition of water into the system. Simulations investigating past leaks showed similar peaks and arrival times as the retrieval leak cases. Several different release rates were used to investigate contaminant transport from residual tank wastes. All showed similar peak concentrations and arrival times, except for the lowest initial release rate, which was 1,000 times slower than the highest release rate. Past leaks were also investigated with different release rate models, including advection dominated, diffusion dominated, and solubility controlled releases. Of the three models, peak concentrations were lowest and arrival times were later for the solubility controlled model. For the tank ancillary equipment leak case, the diffusion dominated release rate model yielded peak concentrations and late arrival times that were similar to the majority of the past leak cases for residual tank wastes

[en] This report documents numerical flow and transport simulations performed that establish initial waste acceptance criteria for the potential waste streams that may be safely sequestered in the 221-U Building and similar canyon structures. Specifically, simulations were executed to identify the maximum loading of contaminant mass (without respect to volume) that can be emplaced within the 221-U Building with no more than 1 pCi/m2 of contaminant migrating outside the structure within a 1,000 year time period. The initial scoping simulations were executed in one dimension to assess important processes, and then two dimensions to establish waste acceptance criteria. Two monolithic conditions were assessed: (1) a grouted canyon monolith; and (2) a canyon monolith filled with sand, both assuming no cracks or fissures were present to cause preferential transport. A three-staged approach was taken to account for different processes that may impact the amount of contaminant that can be safely sequestered in canyon structure. In the first stage, flow and transport simulations established waste acceptance criteria based on a linear (Kd) isotherm approach. In the second stage, impacts on thermal loading were examined and the differences in waste acceptance criteria quantified. In the third stage of modeling, precipitation/dissolution reactions were considered on the release and transport of the contaminants, and the subsequent impact on the maximum contaminant loading. The reactive transport modeling is considered a demonstration of the reactive transport capability, and shows the importance of its use for future performance predictions once site-specific data have been obtained

[en] In support of CH2M HILL Hanford Group, Inc.'s (CHG) preparation of a Field Investigative Report (FIR) for the closure of the Hanford Site Single-Shell Tank (SST) Waste Management Area (WMA) tank farms, a set of numerical simulations of flow and solute transport was executed to investigate different potential contaminant source scenarios that may pose long-term risks to groundwater from the closure of the S-SX Tank Farm. This report documents the simulation of 7 cases (plus two verification) involving two-dimensional cross sections through the S Tank Farm (Tanks S-101, S102, and S-103) and the simulation of one case involving three-dimensional domain of the S Tank Farm. Using a unit release scenario at Tank S-103, three different types of leaks were simulated. These simulations assessed the effect of leaks during retrieval as well as residual wastes and ancillary equipment after closure. Two transported solutes were considered: uranium-238 (U-238) and technetium-99 (Tc 99). To evaluate the effect of sorption on contaminant transport, six different sorption coefficients were simulated for U 238. Overall, simulations results for the S Tank Farm showed that only a small fraction (< 0.4%) of the U-238 with sorption coefficients 0.6 mL/g migrated from the vadose zone in all of the cases. For the conservative solute, Tc-99, results showed that the simulations investigating leaks during retrieval demonstrated the highest peak concentrations and the earliest arrival times due to the high infiltration rate before water was added and surface barriers installed. Residual leaks were investigated with different release rate models, including uniform release, advection-dominated, diffusion-dominated, and saltcake (solubility-controlled) release models. Of the four models, peak concentrations were lowest and arrival times later for the uniform release model due to the lower release rate of the residual tank waste solids; similar high peak concentrations occurred for the advection-dominated and the salt cake models due to the higher release rate. For the tank ancillary equipment leak case, the diffusion-dominated release rate model yielded peak concentrations and arrival times that were similar to the majority of the past leak cases for residual tank wastes. Comparison between the results of the two-dimensional and those of the three-dimensional simulations show that the two-dimensional simulation significantly overestimated the peak concentrations of the contaminants by a factor of about 41 for Tc-99 and 37 for U-238 with sorption coefficient of 0.03 mL/g

[en] In support of CH2M HILL Hanford Group, Inc.'s (CHG) closure of the Hanford Site Single-Shell Tank (SST) Waste Management Area (WMA) tank farms, numerical simulations of flow and solute transport were executed to investigate different potential contaminant source scenarios that may pose long-term risks to groundwater from the closure of the C Tank Farm. These simulations were based on the initial assessment effort (Zhang et al., 2003), but implemented a revised approach that examined a range of key parameters and multiple base cases. Four different potential source types were identified to represent the four base cases, and included past leaks, diffusion releases from residual wastes, leaks during retrieval, and ancillary equipment sources. Using a two-dimensional cross section through the C Tank Farm (Tanks C-103?C-112) and a unit release from Tank C-112, two solutes (uranium-238 (U-238) and technetium-99 (Tc 99)) were transported through the problem domain. To evaluate the effect of sorption on contaminant transport, seven different sorption coefficients were simulated for U 238. Apart from differences in source releases, all four base cases utilized the same median parameter values to describe flow and contaminant transport at the WMA C. Forty-six additional cases were also run that examined individual transport responses to the upper and lower limits of the median parameter values implemented in the base case systems. For the conservative solute, Tc-99, results amongst the base cases showed that the simulations investigating past leaks demonstrated the highest peak concentrations and the earliest arrival times (48 years) due to the proximity of the plume to the water table and the high recharge rate before surface barriers were installed. Simulations investigating leaks during retrieval predicted peak concentrations ∼60 times smaller than the past leak cases, and corresponding arrival times that occurred ∼70 years later. The diffusion release base case predicted the lowest peak concentrations and arrival times for all solutes. Even after 10,000 years of simulation, only 11.2% of the Tc-99 mass migrated past the fence line compliance point in the groundwater. Although ancillary equipment cases released the contaminant at a similar depth as the diffusion cases, nearly all of the Tc-99 (99.0%) exited the groundwater domain by the end of the simulation due to differences in release rates. These differences were also reflected in the peak arrival times, which were ∼8,500 years for the diffusion base case, and ∼3,700 years for the base ancillary equipment release. In the diffusion cases, peak concentration predictions were sensitive to the rate of diffusion, but had no impact on the peak concentration arrival times. The average peak concentration was ∼3.2 times higher than the base case value for the upper estimate of diffusion, and 3.2 ? 10-3 lower for the lower bounding estimate. The past leak, ancillary equipment and retrieval leak cases were sensitive to the estimate of the pre-barrier installment recharge rate. For example, on average for the past leaks, relative concentrations increased by ∼2.2 times for the upper recharge estimate, and decreased by ∼0.14 times for the lower bound. Faster arrival times were associated with the upper recharge estimate, and slower arrival times with the lower estimate. Similar trends in both predicted peaks and arrival times occurred for the ancillary equipment and retrieval leaks scenarios that investigated the uncertainty in the pre-barrier installment recharge rate. Uncertainty in the plume depth also impacted predicted peak concentrations and arrival times for the past leak scenario. Trends similar to the pre-barrier installment recharge rate resulted, with higher concentrations and earlier breakthroughs associated with a lower plume depth, and lower concentrations and later breakthroughs with a higher plume depth

[en] From the mid-1940s through the 1980s, large volumes of wastewater were discharged at the Hanford Site in southeastern Washington State, causing a large-scale rise (in excess of 20 m) in the water table. When wastewater discharges ceased in 1988, groundwater mounds began to dissipate. This caused a large number of wells to go dry and has made it difficult to monitor contaminant plume migration. To identify the wells that could potentially go dry, a first order uncertainty analysis was performed using a three-dimensional, finite element code (CFEST) coupled with UCODE, a nonlinear parameter estimation code. The analysis was conducted in four steps. First, key parameter values were identified by calibrating to historical hydraulic head data. Second, the model was tested for linearity, a strict requirement for representing output uncertainty. Third, results from the calibration period were used to verify model predictions by comparing monitoring wells? wet/dry status with field data. In the final step, predictions on the number and locations of dry wells were made through the year 2048. A non-physically based model that extrapolated trends at each individual well was also tested as a predictor of a well?s wet/dry status. Results demonstrated that when uncertainty in both parameter estimates and measurement error was considered, the CFEST-based model successfully predicted the majority of dry wells, outperforming the trend model. Predictions made through the year 2048 identified approximately 50% of the wells in the monitoring well network are likely to go dry, which can aid in decisions for their replacement

[en] A critical component is determining the suitability of disposing glassified, low activity waste is the identification of key mineral assemblages affecting the porosity and permeability of both the glass and near- and far- field materials. In this study, two different classes of geochemical models are used to identify minerals precipitation and dissolution potentials for an immobilized low-activity waste (ILAW) disposal facility in Hanford, Washington. The first is a static geochemical model that does not consider the effects of transport. The second model is dynamic, and combines geochemical reactions with hydrogeological processes such as advection, diffusion and dispersion. This reactive transport model also includes an innovative application of a depositional film model for determining changes in permeability due to mineral precipitation and dissolution reactions. Although both models describe solid-aqueous phase reactions kinetically, the two models identify two different sets of mineral assemblages affecting the porosity and permeability of the media

[en] The effects of secondary mineral precipitates on radionuclide sequestration at the Hanford Site were investigated by reacting quartz and Hanford sediment (Warden Soil) with caustic solution of high ionic strength (2 M NaNO3), high pH (∼13), high temperature (∼90 C), and dissolved Al(0.01 M Al(NO3)3). Continuous Si dissolution and concomitant secondary mineral precipitation were the principal reactions observed. Nitrate-cancrinite was the dominant secondary precipitate on mineral surfaces after 3-10 days reaction time. The presence of dissolved Al in the simulated tank fluid was found to depress the net Si dissolved concentration. Based on batch equilibrium sorption results, secondary precipitates (cancrinite) at the primary mineral surfaces enhanced the sorption capacity of typical Hanford sediment for radionuclides of major concern at the Hanford Site such as 129I, 79Se, 99Tc, and 90Sr