2005 Closure Assessments for WMA-C Tank Farms: Numerical Simulations

Freedman, Vicky L. ; Zhang, Z. F. ; Waichler, Scott R. ; Wurstner, Signe K.

Pacific Northwest National Laboratory (U.S.)

关键词: Technetium 99;    Leaks Tank Farms;    Storage Facilities;    Ground Water;    Computerized Simulation;   

DOI  :  10.2172/882106
RP-ID  :  PNNL-15377
RP-ID  :  AC05-76RL01830
RP-ID  :  882106

美国|英语

来源: UNT Digital Library

【 摘 要 】

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.