36-35

Frederick M. Mann and Raymond J. Puigh II
Fluor Daniel Northwest
Richland, Washington

The Hanford Immobilized Low-Activity Tank Waste Performance Assessment (1) examines the long-term environmental and human health effects associated with the disposal of the vitrified low-level fraction of waste presently contained in Hanford tanks. The tank waste is the by-product of separating special nuclear materials from irradiated nuclear fuels over the past 50 years. This waste has been stored in underground single- and double-shell tanks. The tank waste is to be retrieved, separated into low- and high-activity fractions, and then immobilized by private vendors. The U.S. Department of Energy (DOE) will receive the vitrified waste from private vendors and dispose of it in the Hanford Site 200 East Area.

The performance assessment activity will continue beyond this first assessment. The activity will collect data on the geotechnical features of the disposal sites, on the disposal facility design and construction, and on the long-term performance of the waste form. This activity will also perform analyses on the impact of such new data or information collected from other programs. Better estimates of long-term performance will be produced on a regular basis.

The performance assessment analyzes the long-term performance of the disposal system in order to (1) set requirements on the waste form and on the facility design which will protect the public’s health and safety and protect the environment into the future and (2) demonstrate that such requirements can be met.

The calculations in this performance assessment show that there exists a "reasonable expectation" that the disposal of the immobilized low-activity fraction of Hanford tank waste can meet environmental and health performance objectives.

The Hanford Site in south-central Washington State has been used extensively as a location for defense materials production by DOE and its predecessor agencies. Over the last 50 years, radioactive and mixed waste from materials production and related activities have been stored on the Hanford Site, primarily in underground single- and double-shell tanks.

As part of the Hanford Site's environmental restoration and waste management mission, DOE is proceeding with plans to retrieve the waste from the tanks (some of which have already leaked some of their contents), to separate the waste into a small quantity of high-level waste and a much larger quantity of low-activity waste, to immobilize both waste streams, to store the immobilized high-level waste until it can be sent to a federal geologic repository, and to dispose of the immobilized low-activity waste on-site in near-surface low-level waste disposal facilities. This plan is based on Revision 6 of the Hanford Federal Facility Agreement and Consent Order (Tri-Party Agreement)(2) and on the Record of Decision for the Tank Remediation Systems, Hanford Site, Richland, Washington (3). More than 200,000 m³ (7,000,000 ft³) of immobilized low-activity waste will be disposed of under this plan. This quantity is among the largest amounts in the DOE Complex and contains one of the largest inventories of long-lived radionuclides at a low-level waste facility.

The waste in the Hanford tanks is high-level radioactive waste. The staff of the U.S. Nuclear Regulatory Commission (NRC) has indicated (4) that the low-activity waste would be considered "incidental waste" or low-level waste (i.e., not high-level waste) if DOE follows its program plan for separating and immobilizing the waste and if the performance assessments continue to indicate that public health and safety would be protected to standards comparable to those established by the NRC.

The current program plan is to utilize existing disposal vaults and to construct additional facilities for the disposal of the immobilized low-activity tank waste (ILAW). An earlier program to dispose of the tank waste built four large concrete subsurface vaults (having a total usable volume of about 15,000 m³). These vaults will be modified and will accept the first waste to be immobilized in the second half of the year 2002. Based on planned ILAW production schedules, additional disposal facilities (of a different design) will be needed in 2005. ILAW production is scheduled to continue until 2024, with closure being later in the decade. The closing of the tanks themselves is a separate program and will occur between 2010 and 2030, depending upon the tank farm.

DOE Order 435.1, Radioactive Waste Management (5), will be the primary regulation currently governing management and disposal of radioactive waste at DOE facilities. Prior to the disposal of low-level radioactive waste, a Waste Disposal Authorization Statement (WDAS) must be issued from DOE-headquarters to the field office. The issuance of this WDAS is predicated on many analyses, one of which is the performance assessment which investigates the ability of the disposal system to provide long-term environmental, public health, and safety protection.

Because of the long time frame of the program, because of the variability of the ILAW produced over those many years, and because of the likelihood of different disposal facility designs, the approach taken in this performance assessment is to analyze for likely conditions in such a way as to provide requirements on the disposal facility design and the ILAW product quality. The requirement on the performance assessment to show "reasonable expectation" that public health and safety will be protected is met by showing that the requirements set in the performance assessment can be met. Future performance assessments will show how the requirements were met.

As more data are collected (through performance assessment activity data collection, through tank retrieval sampling, through ILAW production experience, through disposal facility operation history, and through other research), this performance assessment will be modified. Because of the requirements of the DOE Order and from good business practices, this performance assessment will be revised to reflect our better knowledge and understanding.

This commitment to iterative analysis is demonstrated by noting that this performance assessment is actually the third set of environmental analyses performed for the program. The first set of analyses (6) provided the background for disposal facility conceptual design and for waste form quality. The second set of documents, Hanford Low-Level Tank Waste Interim Performance Assessment(7), which provided a set of analyses based on the current DOE Order on radioactive waste management (5820.2A)(8) showed the disposal of ILAW was likely to be successful in meeting its performance objectives based on DOE’s current plans and on current knowledge. This document builds upon the analyses presented in the interim performance assessment. However, instead of only presenting the results of the calculations, this document proceeds further by setting requirements on the disposal facility design and waste form quality and by showing that such requirements could be met.

Based on applicable regulations and earlier performance assessments, performance objectives were established (9) to protect:

The data used in this performance assessment are documented in Data Packages for the Hanford Low-Level Tank Waste Interim Performance Assessment (10). The base analysis and sensitivity cases are provided in Definition of the Base Analysis Case of the Interim Performance Assessment (11).

Disposal will occur at two facilities approximately 2 kilometers (1.5 miles) apart. The first facility to be used consists of four existing concrete vaults located just east of the Hanford Site 200 East Area. The vaults which have an outer layer of asphalt ~1 meter thick were constructed around 1990 as the first of 34 vaults for the disposal of double-shell tank waste in a grouted waste form. The location of the remaining disposal facility is to the southwest in a previously unused area. This disposal facility is assumed to also consist of concrete vaults, but without the asphalt layer. Both disposal facilities are planned to have a surface cover (to minimize the flow of water or other potential intrusions into the facility) as well as a sand-gravel capillary barrier (to divert water around the waste form).

Geologic, hydraulic, geochemical, and water infiltration data obtained for the 200 Area plateau were used in this analysis and are considered to be representative of the disposal areas. Additional site-specific data are being collected.

The inventory of contaminants in the waste form is based on estimates for the tank waste inventory and using a conservative estimate to project the low-activity fraction of radionuclides immobilized in the waste form after the separation and immobilization processes. The tank waste inventory estimate is based on computer simulations of the production reactor history and the known reprocessing histories.

The release rate of contaminants from the waste form (4.4 parts per million per year) used in the base analysis case is based on the Request for Proposal (12) issued by the Richland Operations Office for the separation and immobilization of tank waste. Sensitivity cases were also performed for a well-studied low-level waste glass using a computer simulation code to estimate the rate at which this glass would release the contaminants.

A three-dimensional computer code was used to simulate moisture flow and the transport of contaminants from the waste form through the vadose zone to the groundwater. Another three-dimensional computer code simulated the flow and transport in the groundwater. The results from these two codes were combined with inventory and dosimetry data to provide radionuclide concentrations in groundwater and dose rates. Explicit calculations were conducted to 100,000 years after disposal with extrapolations used to extend the results to 65 million years. For inadvertent intruder analyses, a spreadsheet was used with calculations extending from 100 to 1,000 years.

Because of the very slow predicted release of contaminants from the waste form (hundreds of thousands of years), the estimated concentration of radionuclides in the groundwater does not show a peak, but rather a broad plateau (see, for example, the beta/photon drinking water dose rate shown in Figure 1). This contrasts with most other environmental assessments, in which the contaminant release time is short compared to the contaminant travel time resulting in a peaked response.

The estimated all-pathways doses are significantly lower than the performance objectives. The sensitivity cases show the all-pathways performance objective would be exceeded if one or more of the following conditions exist for the actual waste disposal action:

During the first 10,000 years (the time of compliance), the estimated doses are at most 1/3 of the performance objective (25 mrem in a year as stated in the DOE order). ⁹⁹Tc is estimated to contribute 58 percent of this dose. The peak all-pathways dose (23 mrem in a year) is estimated to occur at about 50,000 years. At the peak, uranium and its daughters are the main contributors.

The other two performance measures (all-pathways including other Hanford actions and a design that produces doses as low as reasonably achievable [ALARA]) are not expected to exceed 100 mrem in a year or 500 persons-rem per year at any time.

A one-time dose (an acute exposure) scenario as well as a continuous exposure scenario (a homesteader) are defined. Both performance objectives (500 mrem and 100 mrem in a year) are met.

The acute dose (estimated to be 5.5 mrem by assuming a person drills a well through the disposal facility) is much less than the performance objective. The continuous dose (which includes the ingestion of contaminated food and water, the inhalation of air, and direct radiation exposure) is over a factor of 3 lower than the performance objective. At the time of compliance (500 years) ¹²⁶Sn contributes more than 95 percent of the dose.

The performance objectives (beta/photon drinking water dose of 4 mrem in a year, alpha emitter concentration of 15 pCi/l, and a radium concentration of 3 pCi/l) are based on the federal drinking water standards. The time of compliance is 10,000 years and the point of compliance is at a well 100 meters down gradient of the disposal facility. The estimated impact from beta emitters is a factor of 2 less than the performance objective and the estimated impact from alpha emitters is a factor of 5 less than the performance objective. The concentration of radium is insignificant.

The most important drivers are the inventory of technetium (for beta/photon emitters) and uranium (for alpha emitters), the release rate from the waste form, the amount of mixing in the aquifer, and the geometry of the disposal facility. For the impact for alpha emitters, the amount of retardation experienced by the uranium isotopes in the vadose zone is also important.

For the most part, other geotechnical data (water infiltration rate, hydraulic parameters, and geochemical factors) are less important, because they mainly affect the time at which the plateau is reached. However, there are two exceptions. If the water infiltration rate is a factor of 5 lower than assumed (which is 0.5 mm/y for the first 1,000 years [the period during which the surface barrier is assumed to function] and 3.0 mm/y thereafter), then the most mobile radionuclides do not reach the groundwater in significant quantities during the compliance period. Alternatively, if the infiltration rate is a factor of 30 higher than assumed and if no capillary barrier is in place to divert the infiltration, then the uranium group arrives in significant amounts at the water table during the compliance period.

The beta/gamma drinking water dose rate is not estimated to exceed 4 mrem in a year for 750,000 years, reaching a maximum value of 14 mrem in a year at the end of the simulation period (65 million years). The concentration of alpha emitters is estimated never to exceed 15.0 pCi/

, reaching a maximum of 8.2 pCi/

at 50,000 years.

The time of compliance is 10,000 years and the point of compliance is at a well intersecting the groundwater just before the groundwater mixes with the Columbia River. The estimated impacts are over an order of magnitude lower than the performance objectives (which are comparable to those established for groundwater protection). The calculations indicate that the impacts never reach the values given as performance objectives. Because of the large flow of the Columbia River, tremendous mixing occurs in the river and the predicted impacts in the river itself would be far lower.

The time of compliance is 10,000 years and the point of compliance is just above the disposal facility. The estimated impacts are significantly lower than the values prescribed in the federal clean air regulations.

Based on the computer simulations, relatively simple restrictions on disposal facility design and waste form quality can be set. The restrictions are more complex than those normally set (inventory concentrations based on inadvertent intruder protection and total inventory limits based on groundwater protection), but they are quite similar.

The restriction set by protecting the inadvertent intruder is more comprehensive than the concentration limits established by the Nuclear Regulatory Commission (NRC). For the protection of the inadvertent homesteader, the following equations can be used

where the first sum is over contaminants i, the second sum is over containers j in a vertical column emplaced within the disposal facility, and where

I_ij =	the inventory of contaminant i in container j
V_j =	the volume of container j
d_i^h =	the dosimetry factor relating response to concentration of contaminant i in the inadvertent homesteader scenario
k_i^h =	the factor which accounts for the fraction of waste exhumed during drilling, the mixing of the waste in the soil and then transport to point of exposure
H_j =	the height of container j
D^h =	the maximum dose allowable in the homesteader scenario (100 mrem in a year)
Y_i =	[D^h / (d_i^h k_i^h)].

The parameters d_i^h and D^h can be specified and the parameter k_i^h can be calculated from data presented in this performance assessment. The TWRS Immobilized Waste Program will place restrictions on the quotient of inventory (I_ij) and container volume (V_j). Although the height of an individual container is known, it is not known how many will be stacked on top of each other. Therefore, the program will also place restrictions on the summation over the vertical column.

The TWRS Immobilized Waste Program has also decided to place additional restrictions on waste concentrations. In order to satisfy the Nuclear Regulatory Commission (4) in their determination that the Immobilized Low-Activity Waste is not high-level waste, the concentration of all radionuclides will be below the Class C limits set in 10 CFR 61 (13).

Also, the Department of Energy has mandated (12) concentration limits for ⁹⁰Sr, ⁹⁹Tc, and ¹³⁷Cs for the first phase of waste form production. All of the waste slated to be placed in the Existing Disposal Vaults will be produced under this contract. Therefore, these contract requirements will also be imposed on the waste to be placed in the Existing Disposal Vaults. Although it is also expected that most of the waste in the first set of units in the New Disposal Facilities will also be produced under this contract, most of the waste that will be contained in the New Disposal Facilities will be produced under a different contract. Therefore, in order to provide maximum flexibility in future decisions, these contract limitations will not be placed on waste in the New Disposal Facilities.

For expected stacking heights (~7.2 meters), the NRC Class C limits will be more restrictive than the ones found in this analysis for most of the isotopes. The cause of this is that the glass waste form makes ingestion or inhalation of the radioisotopes very difficult even after being brought to the surface. A few isotopes (mainly actinides) may be more restricted by this analysis than by the NRC restriction:

Note that the radioisotope of greatest concern for intruder protection (¹²⁶Sn) is not one of the radioisotopes restricted by the NRC.

The computer analysis shows that for groundwater protection the main requirement is the contaminant flux leaving the disposal facility and the amount of water into which the flux evidently flows. Unlike most environmental analyses where the rate of release is relatively a minor concern, this analysis shows that it is a driving concern. The groundwater scenario places the restriction that

I_i =	the inventory of contaminant i
R_i =	the release rate of contaminant i from the waste form
d_i^gw =	the dosimetry factor relating response to concentration of contaminant i in the groundwater scenario
k_i^gw =	the factor which accounts for vadose zone and aquifer transport for contaminant i
L =	the effective length of the disposal facility and accounts for the direction of groundwater flow
D^gw =	the maximum dose allowable in the groundwater scenario
X_i =	[D^gw / (d_i^gw k_i^gw )]

The parameters d_i^gw and D^gw can be specified and the parameters k_i^gw can be calculated from data presented in this performance assessment. The drinking water scenario and the all-pathways scenario are considered in establishing the restrictions. Also, the plume overlap caused by the upgradient facility is taken into account. The TWRS Immobilized Waste Program will place restrictions on the inventory (I_i) and the release rate (R_i). The effective disposal facility length (L) is a special case. For the Existing Disposal Vaults, L can be calculated. For the New Disposal Facilities, since they have not been designed, the program will place restrictions on L.

The isotopes facing the greatest restrictions relative to the expected performance are ⁹⁹Tc and ⁷⁹Se. This is not surprising since these are the most mobile and because most of the uranium and transuranic elements have been separated from the low-activity waste form. The long-term release limits found here are less restrictive than the short-term release limits found in the Privatization Request for Proposal (RFP) (12).

Most of the requirements imposed by the performance assessment analysis are on the waste form. However, there are a few that are imposed on the disposal facilities. The major requirements deal with subsidence, recharge rate, layout, interactions with the waste form, and intruder protection.

The performance assessment assumes that subsidence is small based on the slow degradation of the waste form and the lack of voids in the disposal facility. Thus, the facility must be constructed without significant void space. In addition, after waste is placed inside the facility, the spaces between the waste containers must be filled with a dry material.

Because the waste form releases contaminants so slowly, the time dependence of the exposures show more of a plateau structure than of having a peaked shaped. Therefore, the major effect of the recharge rate is to delay the arrival of contaminants to the groundwater. If the slightly retarded group of contaminants (such as uranium) would arrive before 10,000 years, the all-pathways dose performance objective would be violated and hence restrictions would have to be placed on the recharge rate. Based on the sensitivity analyses, the recharge rate must be limited to ~3.0 mm/year (i.e., the natural rate) if no hydraulic diverter is included in the design. If, however, a hydraulic diverter is included then a recharge rate of 100 mm/year would not violate performance objectives. Therefore, if no hydraulic diverter is included in the design, then the surface barrier must limit the recharge rate to no more than the natural rate of 3 mm/year. That is, gravel surface barriers such as used in the Hanford Site Tank Farms would not be suitable. Note that the surface barrier must perform the function of inadvertent intruder diversion and does minimize the already low air releases.

The requirement for groundwater protection [å (I_i R_i /L ) / X_i < 1] is actually on the disposal system. The designers of the disposal structures must insure that materials are not used that would accelerate waste form degradation and that the vault layout in relationship to groundwater flow is such to have a sufficient effective length (L). Alternatively, the designers can add components (for example, hydraulic diverters, getters) which could minimize the requirements on the waste form.

Designers of the engineered system may wish to add components to provide greater defense-in-depth. The major components would be a surface barrier which reduces recharge, a hydraulic barrier which diverts moisture away from the waste, concrete pads which would trap uranium, and other getter materials which would trap important radionuclides such as technetium. The recharge rate is the main driving function for the system. Thus, by having a surface barrier which could reduce this rate, the contaminants would take even longer to reach the groundwater. The diversion of water away from the waste would likely reduce the contaminant release rate from the waste form and would also create a greater moisture shadow under the disposal system which would also delay contaminant travel. Concrete is known to highly retard uranium isotopes, thus reducing its impact during the time of compliance. If an inexpensive getter could be found for technetium, such a material could also have important impacts.

Not only must the performance assessment establish restrictions to provide a reasonable expectation that the environment and the public health and safety will be protected, but the document must also show that these restrictions can be expected to be met. The major restrictions deal with inventory concentrations, long-term waste form release rates, and disposal facility design.

For almost all the radionuclides, a stack of containers each having the maximal concentration would be of no concern, the concentration values being very far below the acceptance limits. For most of the remaining radionuclides (⁹⁰Sr, ⁹⁹Tc, ¹³⁷Cs, ²³⁹Pu, ²⁴⁰Pu, and ²⁴¹Am) which are near or over the limit, the producers of the ILAW are required to have the waste form meet NRC Class C limits (13) which are for these radionuclides the basis of the waste acceptance.

The only other radionuclide of concern in meeting the acceptance requirements based on inadvertent intruder protection is ¹²⁶Sn. This radionuclide does not have a Class C limit, so its waste acceptance limit is based on this performance assessment. If the ILAW containers having only wastes from tanks A-105, A-106, or AX-104 were stacked on top of each other, then the intruder dose would exceed the 100 mrem in a year limit. However, a number of alternatives exist. This performance assessment conservatively assumes that all of the tin would go to the ILAW product. However, a significant fraction may be diverted to the high-level waste stream during separations and treatment. The three tanks of concern have small volumes of waste (19,000 gallons, 125,000 gallons, and 7,000 gallons, respectively). Thus during retrieval the tank contents are likely to be mixed with other tanks which have significantly lower ¹²⁶Sn concentrations. Additionally, the operators of the disposal facility have the option of placing containers having low concentrations of ¹²⁶Sn are on top of a container having a high concentration which would make the stack compliant with the disposal requirements. Finally, as these tanks are likely to be process during the second phase of immobilization, the Department of Energy could by contract have the ILAW producers separate the ¹²⁶Sn from the low-activity waste and hence insure that the ¹²⁶Sn is below the acceptance limits.

Even if all of the ILAW inventory is placed in each set of disposal facilities, for each radionuclide the (I_i R_i / L) product is less than the requirement. Moreover, using the mandated reduced inventory for ⁹⁹Tc in the Existing Disposal Vaults, then the sum of (I_i R_i / L) is less than unity (as required) for each facility, being 0.54 for the Existing Disposal Vaults and 0.34 for the New Disposal Facility.

Thus, given the conservative assumptions it is reasonable to expect that groundwater will be protected. In particular, the maximum release rate was used, but this maximum release rate is calculated not to occur until 10,000 years after disposal (i.e., at the time of compliance), However, it takes many thousands of years for the contaminants to reach groundwater, so in fact the contamination level in the first 10,000 years will be lower.

The information in this performance assessment can also be used to back out the maximum contaminant release rate from the facility. For the Existing Disposal vaults, the maximum release rate is 2.4 ppm/year assuming that all the inventory of ⁹⁹Tc is placed in the New Disposal Facility. For the Existing Disposal Vaults, the maximum contaminant release rate is higher, being 3.9 ppm/year assuming that the maximum amount of ⁹⁹Tc is placed in this facility.

Because most of the performance requirements are on the waste form, the restrictions on the disposal facility design are relatively few and can be easily met. Whether a sand-gravel hydraulic moisture diverter is actually used will depend on engineering and cost tradeoffs.

Because of the early stage of this project, conservative assumptions have been used. Given such assumptions, it is gratifying that all the estimated impacts meet the performance objectives. Restrictions placed on the waste product and the disposal facility design will not require heroic efforts.

The numerous sensitivity cases that were run show that the results presented in this assessment are quite robust. The computer simulations of long-term dissolution rates for low-level glass (LD6-5412) show that the 4.4 parts per million per year rate can be met. Concerning radionuclide inventory, the calculations are most sensitive to the amount of technetium and to the peak concentration of ¹²⁶Sn. For the base analysis case no credit is taken for enhanced chemical separation or separation occurring during immobilization. Computer simulations of flow and transport under a wide variety of conditions show that slightly increased impacts may occur, but that most changes would result in larger decreases.

Future performance assessments (required by DOE Order) will benefit from increased knowledge of the waste form and the disposal facility design as well as from an extensive data collection activity for the generation of site-specific estimates for geochemical data, hydraulic parameters, and water infiltration rates and waste form release rates. These performance assessments are expected to confirm this analysis that the on-site disposal of the low-level Hanford tank waste can meet the performance objectives with a high degree of assurance.