SUMMARY OF THE REMEDIATION TECHNOLOGIES DEVELOPMENT FORUM
PHYTOREMEDIATION ACTION TEAM
ALTERNATIVE COVER ASSESSMENT PROGRAM MEETING

U.S. Environmental Protection Agency
75 Hawthorne Street
San Francisco, California
March 22 and 23, 1999

WELCOME AND OVERVIEW OF REGION 9 ACTIVITIES
Steve Wall, U.S. Environmental Protection Agency (EPA), Region 9
Jeff Scott, EPA, Region 9

Steve Wall opened the meeting by welcoming participants (see Attachment A) to the second annual Alternative Cover Assessment Program (ACAP) meeting. He explained that EPA's Region 9 Solid Waste program was hosting the meeting and introduced the deputy director of that program, Jeff Scott.

Scott provided a brief overview of waste containment regulation history. There was little interest in long-term waste containment issues, he said, prior to the late 1980s/early 1990s. This changed with the passing of Subtitle D, which established nationwide waste containment standards. These regulations, Scott explained, were written by a small group of people and under a tight schedule. Writing them proved to be a daunting task because they had to impose protective nationwide standards without stripping individual states flexibility in choosing site-specific designs. This state-level flexibility is important, Scott noted, because regulatory-approved prescriptive covers often perform suboptimally if site-specific climatic issues are not addressed. (For example, Scott noted, clay covers often crack in arid and desert-like environments.)

Scott said that state regulators are responsible for enforcing waste containment regulations and approving waste containment designs. At times, he said, state regulators are not allowed to implement containment designs that are more stringent than Subtitle D's specifications. As a result, the most protective measures are not always adopted. Scott said that EPA's Region 9 would like to research alternative cover (i.e., nonprescriptive covers) performance in dry and remote sites. He thanked EPA's Office of Research and Development, the Desert Research Institute (DRI), and other participating organizations for coming together to discuss alternative waste containment designs. The main goal, Scott concluded, is to identify designs that are environmentally protective, economically feasible, and in compliance with waste management regulations.

PROGRESS REPORT - ACAP PHASE I

Introduction to ACAP
Steve Rock, EPA

Steve Rock provided an overview of waste containment issues and ACAP's goals (see Attachment B). He noted that landfills and other waste units pose human health threats by contaminating underlying ground water, causing adverse health effects via dermal contact, and producing gases. Effective waste containment systems, he said, must address all three threat components. Ground-water contamination can be prevented, he said, by installing a cover so that precipitation cannot percolate through wastes and into underlying aquifers. When precipitation falls on a cover, Rock said, it can either run off, sink in, evaporate, or transpire (via plants). Rock said that prescriptive covers minimize water infiltration by increasing the degree to which precipitation runs off. He said that the ACAP team wants to determine whether the same end goal can be obtained by designing covers that increase transpiration and/or manipulate water storage capacity. The ACAP's goal, Rock said, is to determine whether less costly and less complex alternative covers can offer the same degree of protectiveness as prescriptive covers.

Rock said that ACAP will evaluate a number of alternative cover designs, some of which will be evapotranspiration (ET) covers. He said that ET covers are designed to achieve a water balance, meaning that water entering a system exits before it percolates downward through wastes. (Rock noted that it can take up to 3 or 4 years for water balance to be achieved and an ET cap to become protective.) Rock said that ET covers are being investigated mostly in arid and semiarid environments, where shallow-rooted plants (e.g., local prairie grasses with 0 to 3 feet rooting depth) can be used to increase soil storage capacity and ET. In addition, some investigators are experimenting with ET covers in more humid environments, using poplar and willow trees. By using trees with deep root systems, Rock noted, investigators may be creating a situation where roots contact and remediate wastes (via microbial degradation). Allowing contact is a novel idea, Rock commented, and one that EPA has shied away from for years. Rock said that the University of Oklahoma's John Fletcher hopes to show that an ET cover can change over time, from a treatment cover to a containment cover. (Fletcher claims that waste will degrade when penetrated by plant roots. Remediation will continue through the layers until the roots reach their natural root length limit. At this point, remediation will cease, water balance will be achieved, and the cap will act as a containment cover.) Rock expressed enthusiasm about the potential for phyto-based covers to remediate waste, but stressed that the ACAP team will focus only on evaluating how well alternative covers operate as containment units.

Rock said that the ACAP is part of the Remediation Technologies Development Forum (RTDF), an organization that consists of seven different Action Teams that are working toward developing and improving environmental remediation technologies. One of these teams, the Phytoremediation Action Team, met in 1996 and decided to focus on: (1) assessing the status of current phytoremediation research, (2) validating phytoremediation system designs through monitoring, (3) determining appropriate applications for phytoremediation, and (4) providing data for guidance and regulatory writers. The ACAP is a subgroup of the Phytoremediation Action Team. Although ACAP evolved out of a phyto-based research group, Rock said, the ACAP team will also evaluate alternative cover designs that do not use plants.

Rock said that ACAP is being conducted to monitor the performance of alternative covers over a wide variety of geographical, geological, hydrological, and climatological regions. The project will provide information on alternative covers for landfills, treatment units, and other areas where containment is the environmental goal. Rock said that the ACAP involves establishing a network of sites with similar instrumentation, then monitoring these sites over a 5-year study period. Data will be collected in a central location, he said, and distributed to project participants before being released to the public. This data collection effort, Rock noted, will provide long-term performance information that will help regulators, owners, and operators decide if alternative cover designs are viable for specific sites. Rock said that several site owners have already expressed interest in participating in ACAP's network of sites, but he encouraged additional interested parties to contact him or other ACAP members (e.g., DRI's William Albright and Glenn Wilson, Battelle's Glendon Gee, the University of Wisconsin's Craig Benson).

Phase I Task A Report (Facilities Review)
William Albright, Desert Research Institute (DRI)

Albright opened his talk by explaining that ACAP will be conducted in several phases. He said that Phase I is nearly completed and that summary reports will be released in the near future. During Phase I, the ACAP team (1) identified facilities that already have performance-monitoring equipment in place and (2) reviewed available computer models. Albright summarized the results of the facility identification effort, noting that DRI identified several facilities with water-balance lysimeters during their investigation (see Attachment C). Albright said that lysimeters are critical for measuring landfill cover performance because they provide direct drainage measurements. (Other types of instrumentation merely provide estimates.)

Albright briefly described 20 of the identified sites. (Other sites were identified as well, but Albright said that he would not discuss agriculturally based lysimeter sites [e.g., the Coshocton, Ohio site] during his talk.) Albright stressed that much more detailed site-specific information on existing cover designs, soil data (e.g., engineering and hydraulic properties), climate data (e.g., annual precipitation, temperature, and snowfall information), and plant data will be released in ACAP's Phase I report. Albright provided the following site information:

Albright concluded his discussion by noting that lysimeter-based facilities are not evenly distributed across regions. He broke down the number of sites by annual precipitation: areas with less than 20 centimeters (4 sites), 20 to 40 centimeters (6 sites), 40 to 60 centimeters (2 sites), 60 to 80 centimeters (1 site), 80 to 100 centimeters (4 sites) and more than 100 centimeters (2 sites). Albright said that many areas of the country receive about 60 to 80 centimeters of precipitation a year, and expressed a strong desire to include more sites with this precipitation profile. He noted that this meeting was intended to serve, in part, as a sales meeting, with the purpose of enticing more site owners to become involved.

Phase 1 Task B Report (Models Review) Glenn Wilson, DRI

Wilson noted that one of ACAP's main objectives is to develop, improve, review, and evaluate numerical models so that investigators have better tools for predicting alternative cover performance.

Wilson reminded participants that numerical models are inherently inaccurate, but said that the ACAP hopes to make improvements. He noted that there are three types of models: conceptual models (which are created based on observations), experimental models, and numerical models. All of these are dependent on each other—a numerical model is a mathematical representation of the conceptual model, and is only as good as the experimental model it is founded upon.

Wilson noted that ACAP's model development program is split into four phases: Phase I (model description), Phase II (model modification and evaluations), Phase III (performance predictions), and Phase IV (guidance documentation). Phase I has been completed and a report, Phase I: Task B Report: Review of Models for Designing and Monitoring Alternative Landfill Covers, will be released soon. During Phase I, Wilson said, ACAP team members identified models that are currently being used, identified the attributes of those models, evaluated their robustness by performing sensitivity analyses, and performed verification and validation testing. Wilson defined the following terms for participants:

• Sensitivity analysis. In conducting this type of analysis, investigators systematically change one parameter at a time to see how the changes will impact a model's overall output. Sensitivity analyses are important, Wilson explained, because they can be used to (1) reveal numerical errors that are overlooked during verification testing, (2) compare codes, (3) guide site characterization and monitoring, and (4) evaluate the significance of processes. Wilson provided the following equation:
  1. 

• Verification testing. Verifying a code, Wilson explained, involves testing it to ensure that the numerics are accurate. It includes an evaluation of the code's executable statement and numerics and a comparison of modeled outputs to analytical solutions, hand calculations, and/or other "verified" codes.

• Validation testing. Wilson said that validating a code involves comparing modeled outputs to real-world field data. He stressed that lysimeter-generated drainage data and site-specific meteorologic data will play an important role in validating numerical models.

As a kick-off to Phase I, Wilson explained, ACAP team members distributed a country-wide survey to researchers, regulators, and consultants to determine which numerical models are being used to design or evaluate alternative covers. Wilson said that 10 codes (HELP, EPIC, HYDRUS2D, UNSAT-H, SHAW, CREAMS, LEACHM, TOUGH2, MULTIMED, and POLLUTE) were identified, some of which were not originally drafted to predict landfill cover performance. Of the 10 codes identified, Wilson noted, five were selected for more detailed evaluations and are discussed in the Phase I report. Wilson's presentation focused on four of these codes. He provided a brief description of their attributes and the types of testing they underwent prior to ACAP's Phase I activities:

• HELP. Wilson said that this code, which was generated by Paul Schroeder, has become the industry standard for landfill design and evaluation. The code is a water-balance model, Wilson explained, and computes a "pseudo" 2-dimensional drainage. Wilson said that this code is process-rigorous and accounts for runoff, snow melt, plant growth, and geosynthetic properties. Prior to ACAP's Phase I activities, Wilson noted, all verification testing and sensitivity analyses had been conducted by Paul Schroeder. Validation tests, which were conducted by several investigators, yielded varying results. (While modeled predictions matched field measurements in some studies—Peyton and Schroeder [1988] and Stephens and Coons [1994])—drainage was significantly overpredicted in other studies—Nichols [1991], Fayer and Gee [1997], Thompson and Tyler [1994], and Benson and Pliska [1996]. In other studies—Ascough et al [1997] and Berger et al [1996]—some of the code's components proved valid, while others did not.)

• EPIC. This code, Wilson said, was generated for agricultural applications by David Knisel. Wilson said that the code is a 1-dimensional water-balance code that calculates vertical drainage and accounts for runoff, erosion, snow melt, and plant growth. Wilson said that Knisel has performed verification tests and several investigators have conducted validation tests. Most of the validation tests have focused on agricultural processes, however, with very limited work related to landfill applications. Wilson did cite one landfill application validation test that was conducted in 1994 by Hauser and Shaw; this test showed predictions for water storage to be valid, but no direct field drainage measurements were available for comparison.

• HYDRUS2D. Wilson said that Rien van Genuchten and Jirka Simunek developed the HYDRUS2D code to evaluate general vadose zone hydrology. In recent years, however, several people have started using this code to evaluate capillary barrier designs. Wilson said that the code is fully 2-dimensional, solves Richards' equation, and accounts for solute transport and heterogeneities (e.g., spatial variability). He noted that the code's developers conducted verification testing and that several validation tests have been performed. The majority of validation testing has not focused, however, on landfill cover applications. (Wilson said that validation testing has also been conducted on HYDRUS2D's parent code, SWMS.)

• UNSAT-H. Like HYDRUS2D, this code solves Richards' equation. This code is not, however, 2-dimensional.

As part of Phase I, Wilson said, the ACAP team performed sensitivity analyses on the HELP, EPIC, and HYDRUS2D codes. (Sensitivity analyses are currently being performed on UNSAT-H, but results are not yet available.) The team performed the sensitivity analyses, Wilson explained, using data for two different environmental scenarios: Columbus, Ohio (a humid environment with substantial snowfall) and Laramie, Wyoming (an arid environment). (Climate-specific data for these sites were generated using the HELP model.) The results for both environmental conditions are summarized in the Phase I report, but Wilson only provided results for the arid scenario. Wilson said that the team performed sensitivity analyses on water-balance parameters only. (He acknowledged that other types of parameters [e.g., plant growth parameters and snow melt] are also important in influencing modeled outputs, but said that the team tried to focus their efforts narrowly during this phase of ACAP.) For each parameter evaluated, the ACAP team evaluated the model's sensitivity and whether it provided a realistic response. The following parameters were evaluated to determine how they impact modeled drainage outputs:

• Hydraulic conductivity (Ks). The sensitivity of EPIC could not be determined, Wilson explained, because no drainage output was predicted with this model. HELP (Sr of 60%) and HYDRUS2D (Sr of 93%) were both sensitive to Ks and both models exhibited realistic responses (i.e., drainage decreased as Ks decreased). (See Attachment D for additional details.)

• Parameters that impact water storage capacity. Wilson stressed that effective plant-based alternative covers must store water in their soils long enough for plant roots to tap into the water. Water storage capacity, he said, is a function of:

• Available water capacity (AWC). Wilson said that the AWC is a function of the water-retention curve, which describes the relationship between water content and matric head. (With water-balance models like HELP and EPIC, the water-retention curve can be generated using field capacity [FC] and wilting point [WP] data. Above the field capacity, water drains out before plants can utilize it. Below the wilting point, retention forces are too strong to allow plants to utilize water. In a Richards' equation-based model like HYDRUS2D, the water-retention curve is calculated using the van Genuchten model, a model which employs alpha and "n" parameter data.) Wilson said that the ACAP team varied the AWS parameter by manipulating water-retention curves. The team found that all three models were sensitive to changes in the AWS parameter (Sr values ranged from 126% to 184%—see Attachment D), but that they differed in their response patterns. While HYDRUS2D exhibited a realistic response pattern, HELP exhibited an unrealistic pattern and EPIC's pattern was uncertain. Wilson said that HELP and EPIC's response patterns have raised some concern among ACAP members. Wilson said that ACAP also performed sensitivity analyses on alpha and "n" parameters. (Results are available in Attachment D).

• Depth. Wilson noted that a cover's water storage capacity is influenced by the depth of the cover profile. He said that the ACAP team ran sensitivity analyses, expecting to see that drainage decreased as cover depths increased. EPIC's sensitivity to changing depth could not be determined because no drainage was predicted. HELP exhibited low sensitivity (Sr of 24%) and unrealistic response patterns, but HYDRUS2D was highly sensitive (Sr of 308%) and realistic in its response. Wilson said that HELP's unrealistic response resulted because this model only accounts for the evaporative depth of a cover's profile rather than the entire profile (see Attachment D). (HYDRUS2D accounts for the entire profile because it accounts for an upward hydraulic gradient that replenishes water as it is removed by plant roots and evaporative forces.)

Also as part of Phase I, ACAP performed some validation testing on EPIC, HYDRUS2D, UNSAT-H, and several different versions of HELP (See Attachment D for results). Wilson said that the data used for the test were obtained from the Hanford site. (Drainage data was collected from a lysimeter and site-specific weather data was obtained from an onsite meteorological station.) Using these data, EPIC and HYDRUS2D predicted no drainage, but HELP and UNSAT-H predicted some. The predictions generated with HELP were version-dependent, with improved accuracy in more recent versions. Even the best predictions generated with HELP, however, were still an order of magnitude too high. (HELP estimated drainage to be 50 millimeters per year, but the actual field data indicated that the drainage was 5 millimeters per year.) The UNSAT-H model matched the results closely (predicting a drainage of 2.5 millimeters per year) when hysteresis was taken into account, but predicted no drainage when this component was ignored. The ACAP team concluded that including hysteresis is important when making drainage predictions with Richards' equation-based models. As a result, strong consideration is being given to incorporating hysteresis into the HYDRUS2D model. Wilson said that a simple 1-dimensional comparison of UNSAT-H and HYDRUS2D was conducted. The models exhibited the same response, he said, but a lag was exhibited with HYDRUS2D. Wilson said that ACAP is concerned about this, but is working with model developers to solve this problem.

Based on the Phase I findings, Wilson recommended (1) continuing to validate models against common databases across hydrogeologic regions and (2) improving Richards' equation-based models so that they are more specific to landfill cover designs (i.e., develop a process-rigorous 1-dimensional code, and develop a 2-dimensional code for capillary barrier evaluation.) Before concluding, Wilson listed the modeling development activities that have already been proposed for future ACAP phases. During Phase II, he said, team members will (1) help facilitate communication between model developers and practitioners, (2) integrate data from the ACAP site network into numerical models, (3) serve as an independent third party during validation testing and evaluation, (2) provide oversight on model modifications/developments, and (5) develop and test models for performance-assessment monitoring. During Phase III, Wilson continued, the ACAP team will combine field results with improved numerical models to predict long-term performance of alternative cover systems over a wide range of climatic, geologic, and vegetative changes. During Phase IV, the ACAP team hopes to construct a decision tree that will help regulators and consultants select appropriate codes for site-specific situations. Ultimately, Wilson said, the ACAP hopes to provide guidance on how to utilize different codes.

THINKING AHEAD—ACAP PHASE II

History of Alternatives and ACAP Data Objectives
Glendon Gee, Battelle

Gee said that ACAP's Phase II field test program goals are to (1) evaluate alternative cover designs, (2) conduct field-scale tests at selected landfills, and (3) synthesize data to develop performance criteria. He explained that alternative covers have sparked strong interest among several organizations because they may offer a cost-effective approach to landfill closure. Gee provided cost estimates for four different types of landfill closures:

Gee noted that Mark Ankeny has generated cost estimates for alternative cover systems in Medford, Oregon; Los Lunas, New Mexico; Roswell, New Mexico; and Cheyenne, Wyoming. According to Gee, Ankeny's predictions indicate that these covers will cost between $0.15 million and $0.3 million per acre. Gee did note, however, that some have wondered whether these estimates are low.

Gee said that there are several regulatory requirements that currently exist for landfill covers. For example, EPA's minimum national requirements state that (1) cover materials must not be more permeable than underlying liners, (2) covers must have at least one layer of material with < 10^-7 centimeter per second permeability, and (3) 30 years of post-closure monitoring must be conducted. In addition to these minimum requirements, Gee noted, a number of other regulations exist. As an example, Gee cited NRC's requirements, which call for minimizing moisture infiltration and designing a low-maintenance cover that meets lowest reasonable achievable specifications, is at least 5 meters thick, and is within dose requirements. At some radioactive-waste sites, Gee said, NRC is requiring designers to create covers that will prevent human intrusion for up to 500 years. When designing covers for long-term periods, Gee said, researchers must account for long-term climatic change (e.g., flash floods, droughts, precipitation/snowfall distributions), extreme events (e.g., fires), cover durability (e.g., stability of the layered system and durability of the cover materials), erosion processes, and soil-buildup processes. In addition, they must install institutional measures to ensure that maintenance and repairs are conducted over the long term. Gee indicated that many cover design choices have implications on several levels. Thus, designers must balance issues carefully to ensure that they design a protective, practical, and durable system. For example, he said, when deciding on a cover's side slope, designers have to consider costs (e.g., steeper slopes are less stable, but less expensive than gentle slopes) and balance the risk of erosion against the risk of water infiltration (e.g., erosion is minimal when rock side slopes are used, but water infiltrates this substrate easily).

Gee said that ACAP's goal is to test a variety of alternative covers to determine whether they adequately prevent leachate production and meet the design objectives and requirements discussed above. He said that it is crucial to collect drainage measurements because current models often give false predictions. (He noted that inaccurate drainage rate predictions were generated for the covers installed at Hill Air Force Base [AFB].) Before the ACAP team is able to fully assess the performance of alternative covers, Gee said, several pieces of data need to be collected. He said that sites participating in the ACAP will be asked to collect the following:

• Drainage measurements. Gee said that drainage must be measured directly with a lysimeter rather than relying on estimates calculated with other monitoring systems. Indirect estimates are prone to major errors, he explained. For example, errors of about 2000% are possible if drainage is calculated in an arid environment using only data from a micromet station. Gee said that the ACAP team is drafting guidelines and specifications for lysimeter installation. Ideally, he said, drainage measurements should be collected on a daily basis for several years.

• Surface-water runoff measurements. Gee said that it will be important to collect information on event-driven runoff.

• Climatic data. Gee said that investigators need to gather site-specific information on rainfall, snowfall, solar radiation, wind speed, temperature, and humidity so that investigators can determine precipitation and ET rates. Gee said that weather data should be collected on an hourly basis and that daily potential evapotransipiration (PET) summaries should be prepared on a daily basis. He said that ACAP has already purchased weather station equipment and is currently testing it.

• Vegetative data. Investigators will need to collect data on plant type, plant density, and rooting depth.

• Soil variables. Gee said that investigators will need to collect data on soil depth, texture, and hydraulic properties. This information can be used, he said, to estimate FC, WP, van Genuchten parameters, and AWC profiles. (These soil data are essential for the further development of ACAP's model development and evaluation program.) Gee recommended collecting daily water storage information. He noted that measuring the AWC at a site is important because this variable can impact the amount of drainage that flows through a system. As was shown during side-by-side demonstrations at Hill AFB, he commented, drainage increased as AWC decreased and vice versa. (Five cover designs were tested at Hill AFB: an ET cover, a modified RCRA cover, two capillary barriers, and a Hanford-type cover. Lysimeter-generated drainage measurements were collected over 6 years. The cover that exhibited the highest AWC measurements, the Hanford-type cover, also generated the lowest drainage rates. Gee said that the Hanford-type cover had the highest AWC because it was composed of fine soils, which hold water better than courser [sandy] soils.)

Engineering and Monitoring of ACAP Sites
Craig Benson, University of Wisconsin

Benson noted that a variety of alternative covers will be evaluated under ACAP, but that each will have similar monitoring equipment. The data obtained will help investigators learn more about the way water runs off and percolates through alternative covers. In addition, he said, the data will provide valuable information on water storage and a valuable database to be used for model validation.

Benson said that ACAP plans to release a Test Section Installation Manual to provide guidance and specifications on how to construct and install lysimeters, weather stations, probes, and surface-water drainage collection systems. Benson said that facility owners will not be required to use the exact materials that are recommended in the document, but that selected materials should be of equivalent quality. Benson provided a brief overview of the ACAP test section design. The test system has been designed, Benson said, so that (1) full field-scale construction conditions will be simulated (i.e., large machinery will be used rather than picks and shovels), (2) installation will be simple and straightforward, (3) accurate drainage measurements will be obtained, (4) all water-balance components will be monitored, and (5) the lysimeter will have minimal influence on water-balance components. Benson stressed that the test section has been designed to be virtually maintenance-free. To ensure that everything is working properly, however, an ACAP team member will conduct semiannual site visits.

Benson presented a series of construction documents (See Attachment E) indicating how the ACAP's lysimeter should be constructed. He said that the lysimeter will rest on a compacted base and will be about 10 meters by 20 meters in size and surrounded by a 5-meter buffer zone on all sides. A geomembrane, made of a low density polyethylene, will serve as the bottom layer of the lysimeter and will be tested for leaks after it has been anchored in place. (The geomembrane will be textured on both sides to prevent slippage.) A geocomposite drainage layer will be laid over the geomembrane. This material, Benson said, will consist of a geonet sandwiched between two geotextiles and will be highly puncture- resistant. On top of this, an interim soil cover will be laid down with great care to ensure that no damage occurs to underlying materials. To ensure that roots do not penetrate the lysimeter, Benson noted, a geosynthetic root barrier (containing an herbicide) will be placed on top of the interim soil cover. Finally, he concluded, a test cover design will be built on top. Benson said that the cover's slope will be site-specific, ranging from 0 to 3:1.

Benson provided a brief description of the ACAP test section's water-collection system, noting that both percolation and surface-water runoff will be collected. Percolation traveling through the cover, Benson said, will be routed through the lysimeter's drainage layer to a sump. From there, water will collect in a factory-manufactured boot that is joined to a PVC pipe. The connections between the boot and pipe are critical, Benson noted, because any leakage could grossly skew results. (A flex connector should be installed to ensure that the pipe does not break if some movement occurs.) From the pipe, Benson continued, water will flow to a 100-gallon percolation collection tank. Surface-water runoff, Benson explained, will be captured by diversion berms that surround the lysimeter. The runoff will be diverted to a downward sloping PVC pipe. At the top, the pipe will be 4 inches wide, but will narrow to 2 inches as it reaches a 300-gallon water-collection tank. Flow rates at the percolation and surface water runoff tank will be measured with a tipping bucket (during low-flow events) and a dosing siphon (during high-flow events).

Benson provided a brief description of some of the probes that the ACAP team plans to use, including water-content reflectometers (for gathering information on the apparent soil dielectric constant) and type-T thermocouples (for measuring soil temperature). At some sites, Benson noted, the ACAP team may recommend installing heat-dissipation probes and minirhizotron tubes to measure matric suction and rooting characteristics, respectively. Benson and Gee provided a brief description of ACAP's weather stations and data loggers (see Attachment E). Using these systems, they explained, data will be funneled to a central location and then posted on a password-protected area on the Internet. This will allow ACAP members to gather up-to-date precipitation, runoff, water storage, drainage, and PET data at all times. (Rock said that the password will be made available to those who contribute sites to the ACAP and to other ACAP team members.)

Benson said that he will include cost estimates in the Test Section Installation Manual. He said that it would cost roughly $50,000 to install the ACAP test section. He did note, however, that costs are highly site-specific, and would likely be lower if site owners have their own construction equipment and labor force. Rock said that some of the monitoring costs will be shared between facility owners and the ACAP team. He said that ACAP funds have already been used to purchase weather stations, data loggers, and telemetry instruments.

ACAP Partnership: Roles and Commitment
Steve Rock, EPA

Rock reiterated that the ACAP hopes to test alternative caps at a network of sites. He said that demonstration sites can be set up at three types of sites, those which:

• Have site owners who are willing to install ACAP's recommended monitoring equipment over a waste-free area of their site.

• Already have lined and fully instrument landfill covers in place. (At this type of site, Rock said, the instrumented liner system will serve as a multi-acre lysimeter system.)

• Do not have a liner or lysimeter in place. (Rock said that some abandoned sites fall into this category.)

Rock outlined the steps that site owners (or principal investigators) must take to enroll a site in the program. First, he said, a site owner (or principal investigator) must express interest in participating and discuss cover designs and monitoring systems with ACAP team members. If it appears that the site fits well into ACAP, the site owner must sign a research agreement with EPA, agreeing to participate in the program for 5 years and to allow site access to ACAP members. (Rock said that EPA forms Cooperative Research and Development Agreements [CRADAs] with private parties and Memorandums of Understanding [MOU] with public entities.) Before installation can take place, Rock noted, state approval and EPA regional concurrence must be granted. (In some cases, he warned, regulators ask for a contingency plan before granting approval.) Assuming approval is granted, Rock said, the site owners will install a lysimeter, and ACAP members will install monitoring systems, collect data, make semiannual visits, and compile annual reports. Rock said that the ACAP will provide site owners with performance data, but will not interpret these data for them or make a determination about a cover's protectiveness to the environment. One participant asked what ACAP will do if performance is not deemed adequate. Rock said that the ACAP cannot guarantee that site owners will be happy with results, but that the team will make suggestions about ways to "tweak" the design (e.g., adding additional cover depth or trying different plants) if the results are unsatisfactory.

To date, Rock said, 11 site owners (or principal investigators) have expressed interest in enrolling a test site in the ACAP program. The following table summarizes their progress:

Site	Expressed Interest?	Discussed cover design?	Discussed monitoring equipment?	Research agreement with EPA?	Gained state approval?	Gained EPA regional concurrence?	Installation phase started?
Phelan
Milliken
Bluestem			1/2		1/2	1/2
Kiefer
Lancaster
Lebanon					1/2
Marina					1/2	1/2
Lake County
Lewis & Clark
Green II
Center Hill				1/2

Rock encouraged other site owners and/or principal investigators to come forth if they have a site to include in the program. He said that ACAP members would like to test at least 10 to 12 sites, but would be willing to work on as many as 24 if additional funding becomes available.

CASE STUDIES

Overview of Roles in ACAP Demonstration Project
Craig Benson, University of Wisconsin

Benson said that each ACAP demonstration project will consist of four components:

• Regulatory component. Benson said that regulators should be included in early discussions about cover and monitoring designs, pointing out that success is most likely when site owners, consultants, and regulators work together as a team. He said that ACAP members could help establish communication lines between regulators and site owners.

• Design component. Benson reminded participants that ACAP will provide specific guidance on what monitoring equipment to use at a site. As for landfill covers, however, site owners and consultants will need to work together to design a system. Consultants will be responsible, Benson said, for drafting the landfill cover plans, drawings, and construction specifications. ACAP's role will be limited to that of an advisory board, that offers suggestions for improvement. Benson told meeting participants to contact him if they need help identifying a contractor. If a site owner desires, Benson said, ACAP members can be hired independently of the ACAP to provide consulting services.

• Construction component. Benson said that the site owner, ACAP members, and consultants will all be involved in the construction phase. While site owners and consultants will be responsible for actual construction activities, the ACAP will provide oversight and guidance.

• Monitoring component. Benson said that ACAP members will collect, analyze, and summarize data. Albright noted that the ACAP team's data collection software is set up to look for erroneous data points. He said that the software will send ACAP members e-mails if any data points fall outside of expected boundaries. This kind of feedback, Benson said, allows ACAP members to detect equipment failures and to make corrections immediately. Before releasing annual summary reports, Benson said, the ACAP team will perform the arduous task of quality assurance.

Kiefer Landfill Site in California

Site Description and Owner-Operator's Perspective
Chris Richgels, County of Sacramento Waste Management and Recycling

Chris Richgels provided a brief description of Kiefer Landfill. He explained that the landfill consists of a lined portion (about 67 acres) and an unlined portion (about 165 acres). While he expected that regulators would consider an alternative cover for the lined portion, he was not sure whether they would for the unlined portion. Richgels noted that regulators require performance data to decide whether to approve alternative covers. To gather this information, he said, a demonstration project will be initiated under ACAP in the spring or summer of 1999.

Richgels said that CH₂M Hill did the majority of the modeling and design work for the alternative cover demonstration project, but noted that Albright performed some of the hydraulic characterization of soil. (Albright rendered these services independently of ACAP.) During the demonstration project, Richgels explained, investigators will evaluate the performance of a 4-foot-thick cover that is vegetated with grasses and shrubs. Richgels emphasized the importance of using locally derived materials in the cover, noting that the cost of a cover increases when imported materials are used. (Richgels said the local soils at Kiefer have been well characterized and consist of [1] silty sandy clay with cobbles, [2] red sandy clay with a permeability between 10^-6 and 10^-7, and [3] soil types that are a mix of silt, clay, and sand. This latter soil type, Richgels said, will likely support a high AWC because it holds water well. Richgels said that a geological soil investigation is ongoing to determine how much of each soil type is present for utilization.) Richgels said that he may also establish a second test cover, which would consist of an 8-foot-thick pad covered with eucalyptus trees. Richgels said that eucalyptus grow well at Kiefer Landfill because they are fairly resistant to methane. (Richgels noted that gas extraction systems are in place to reduce methane concentrations. Prior to installation, gas levels were so high that nearly all of the vegetation died off. Lou Licht pointed out that methane does not directly kill plants. Rather, it is the lack of oxygen that does so.)

Richgels explained that he is interested in alternative covers because he suspects they will be cheaper and less difficult to maintain than prescriptive covers. Richgels said that geosynthetic liners are commonly employed in his region and that conventional covers typically include 2 feet of foundation soil and 1 foot of low-permeability clay. In many cases, he said, the low-permeability clay has to be imported from other areas, at costs reaching $12 per cubic yard (yd³). Richgels said that he is reluctant to use clay in his cover because this material is prone to desiccation and cracking. In addition, he also expressed a reluctance to use geosynthetics because he questioned their long-term stability and feared that they could require extensive maintenance. (Richgels said that a base liner has been installed at Kiefer Landfill. During a heavy rain event, a large bubble accumulated under the plastic, sloughing sands off and forcing them to the landfill's toe. Richgels said that he does not have the manpower to push the sands back up the side slope after each rain or seismic event.)

Richgels expressed a respect for natural systems, noting that they are often more self-sustaining than engineered systems. For this reason, he said, he is interested in ET covers. He acknowledged that ET covers will require some maintenance (e.g., replanting), but suspected that these efforts would be less extensive than those required with other systems. Richgels said that cost will be an important factor in deciding whether to use alternative covers over prescriptive designs. He said that he is unsure how the two will compare, but reminded meeting participants to incorporate long-term maintenance costs when calculating expenses.

Regulatory Perspective in California
Glenn Young, California Integrated Waste Management Board (CIWMB)
Elizabeth Haven, California Water Resources Control Board (CAWRCB)

Glenn Young and Elizabeth Haven noted that Kiefer Landfill falls under the jurisdiction of several state regulatory agencies. Young said that he has not been involved with the site yet, but that he will in the near future. Young and Haven briefly described the roles of some of the regulatory agencies in California. They noted that state regulations are written and revised by CAWRCB, but that the task of approving final closure plans and landfill cover designs falls upon regional water quality control boards (RWQCBs), CIWMB, and local enforcement agencies. These same agencies are also responsible for granting approval on final closure plans and approving closure certification reports. Once approval has been granted, Young explained, CIWMB has the authority to release owner/operators from their financial assurance and return money to them for closure activities.

Haven said that the state regulations require landfill covers to be no more permeable than their liners and to contain a layer with a permeability of no more than 10^-6 centimeters per second. Given these requirements, she said, the majority of landfill covers approved for composite-lined units are composite covers that use plastics and a compacted clay layer. Haven admitted that the lifetime of these composite covers is unknown. She noted that California has stringent post-closure requirements, requiring post-closure activities for as long as wastes pose a threat to water quality. Although most of the approved covers in California fit a specific profile, Haven said, provisions are in place for regulators to approve alternative covers. Before approval can be granted for such a cover, however, researchers must prove that it performs as well as prescriptive designs in its ability to minimize infiltration. Haven said that some alternative covers have already been approved in California and Young said that several proposals have recently been submitted. For example, Young said, several landfills in southern California (e.g., Milliken Landfill, Coyote Canyon, and Coachella Landfills) are proposing to use alternative earthen covers.

Young said that his agency has several concerns that need to be addressed regarding alternative earthen covers. For starters, he said, there is some concern that they will create a "bathtub effect" (i.e., floating garbage) because they do not meet permeability requirements specified in the 27 CCR regulations. Also, he said, some of the proposals rely too heavily on modeled information. (Modeled predictions of hydrologic performance must be validated with field data, he explained.) Lastly, Young noted, the construction quality assurance for alternative cover designs needs to be better defined. Young said that CIWMB recently contributed $15,000 to ACAP because it thinks that ACAP's activities can help provide answers to their regulatory concerns.

Haven said that she commends the ACAP for focusing some of their efforts on model development and validation and she encouraged them to take a close look at storm events. She said that the CAWRCB evaluate models carefully to determine whether a given system is protective. She said information on modeling can be found at www.swrcb.ca.gov/~cwphome/chap15. If significant assumptions are made when running models, she explained, the CAWRCB may require that these assumptions are met before granting a permit.

One participant asked whether California's regulatory agencies allow waste materials to be incorporated into final covers. More specifically, the participant asked whether leachate recycling is allowed. Haven said that leachate can be applied, but only over composite-lined areas. Young said that a leachate-recycling demonstration project has been initiated at a lined landfill in Yellow County. Another participant noted that California lags behind other states in bioreactor research. The participant said that the biocell philosophy differs greatly from that advocated by containment approaches. Therefore, it would probably be best if a different group pursued this kind of technology.

Center Hill Landfill
Steve Rock, EPA
Dennis Murphey, City of Cincinnati

Rock and Dennis Murphey provided a description of the Center Hill Landfill. Murphey said that incinerator ash comprises the bulk of the landfill, but that municipal solid waste (MSW) has also been disposed. Murphey said that dumping activities ceased about 25 years ago and that the site has been leveled off, abandoned, and left to the City of Cincinnati to manage. Ideally, he said, the city would like to reuse this property in the future. Rock noted that no landfill liner has been installed, and only a shallow soil layer separates wastes from the atmosphere. (In some areas of the landfill, tires and rebar poke through the soil layer, but the average soil thickness is estimated at about 4 feet.) Rock said that the wastes produce methane, but gas extraction systems are in place to reduce levels. (Rock said that the proliferation of vegetative growth on the cover indicates that methane levels are not phytotoxic.)

Rock said that the hydrology of the site is complicated and not well characterized. He suspects that a certain amount of ground water, coming from upgradient sources, moves laterally though the site below the waste layers. In addition, he said, leachate is introduced into the system when rainfall percolates downward through waste layers. Rock said that the laterally moving ground water and the leachate are collected in a leachate-collection trench. From there, he said, water is directed to a sewage treatment facility. At present, the amount of water collected in the trenches is not measured. Murphey said that onsite ground-water quality has been assessed during monitoring activities. Although several contaminants have been detected, he reported, most have been below Maximum Contaminant Levels. About 6 years ago, however, PCB was detected at concentrations that raised regulatory concern and prompted investigators to score the site under the Hazard Ranking system. Murphey said that he doubted that the site will be placed on the National Priority List.

Rock said that the Center Hill Landfill is 45 acres in size and located near an inactive incinerator and a privately owned landfill. The latter, Rock said, has been closed since February 1998 and is in the process of formal closure. Mill Creek, a stream that drains to the Ohio River, abuts the Center Hill Landfill. (Several parties are interested in performing ecological restoration and flood hazard reduction plans.) Rock described Mill Creek as a very small creek that drains a very large valley. Murphey said that the creek has been abused for years and has serving as a discharge point for industrial waste water since the 1800s. Additionally, Murphey noted, the creek receives combined sewer overflows during large rainfalls. (The metropolitan sewer system is currently drafting a plan to prevent future overflows.) Murphey said that the Center Hill Landfill has impacted the creek via leachate releases. Although the leachate may be "cleaner" than the creek, such releases are unacceptable to Ohio's EPA and the following will be done in an attempt to:

• Reduce leachate production in the vadose zone. Rock said that an engineered phytoremediation cover will be installed at this site. He said that existing plants will be removed and replaced with willows and poplars. (One participant asked why the existing vegetation is being torn up to accommodate the trees. Rock said the trees will utilize more water and grow deeper than existing vegetation. Another participant asked why coniferous trees are not being considered. Rock said that junipers operated effectively at a Maryland site, but that it is difficult to match the efficacy of poplars and willows.) Rock said that investigators hope that the tree cover will stop leachate formation within 3 or 4 years and will dry up the site. Investigators are also optimistic, he said, that the tree roots will promote remedial activities by converting ash and garbage to soil. In addition, Rock thinks that the plant cover might reduce methane levels by supporting methane-eating microbes and creating more aerobic underground environments.

• Reduce leachate releases to Mill Creek. Murphey noted that erosional forces have removed about 15 feet of the landfill's bank over the last 2 years. As a result, the leachate-collection trench is close to Mill Creek's shoreline and regulators are concerned that the system could be washed into the creek during a storm event. Murphey said that an enhanced leachate-collection system will be constructed. Additionally, a soil bioengineering stabilization project has been started to prevent additional erosion. Murphey said that the stabilization project involves using plants rather than concrete or riprap for slope stabilization. Using a more natural system, he said, is in line with efforts initiated under the Green Waste Master plan.

Rock provided a brief overview of how ACAP plans to monitor the performance of the Center Hill Landfill phytoremediation cap. Unlike other ACAP facilities, he said, a lysimeter will not be installed because installation would disrupt the wastes. Instead, meters will be installed at the leachate collection trenches to measure the amount of drainage from the site. David McMillan noted that it will be difficult to determine whether the cap prevents leachate generation if the deep, laterally moving ground-water aquifer continues to release to the collection system. McMillan said that some people will say that the cover failed if any water is detected in the trenches. He and another participant recommended gathering additional hydrologic information on upgradient water. Rock agreed that obtaining these data would be useful, and said that it might be collected over the next several years. In general, however, Rock was not very concerned with the deeper ground water because it is probably "clean." Murphey said that the Ohio EPA has already indicated that the results from the Center Hill Landfill will not serve as precedent for other sites. For this reason, he noted, it may not be important to study all aspects of the hydrology. He said that the City does not require a definite quantifiable measure of water-control performance because using the phytoremediation approach carries enough secondary benefits (i.e., bank stabilization, flood-hazard reduction,) to make it a worthwhile project.

PANEL DISCUSSION ABOUT ALTERNATIVE COVERS

During this section of the meeting, attendees were allowed to question a panel of ACAP members. Albright, Benson, Gee, and Rock served on the panel, and Wilson acted as a moderator. Discussion revolved around the following topics:

• Using landfill gas as an energy source. Wilson said that approximately 130 facilities capture landfill gas for use as an energy source. Rock said that he does not want to focus on this issue under ACAP.

• Relationship between landfill cover design and gas transfer. Panel members noted that cover types influence gas-venting rates. Impermeable RCRA-C type covers, for example, allow little venting and cause landfill gases to build up. Benson said that blanket statements cannot be made regarding whether alternative covers vent gas more readily than prescriptive covers. While gases vent readily through some types of alternative covers, they are more easily trapped in others (e.g., earthen covers). Rock said that not enough information is available to use cover design as a means of controlling gas venting. Thus, sites must have a gas-management plan established before entering ACAP.

• Relationship between gas emissions and air quality. Rock said that he will conduct a study to determine how gas moves through an uncompacted vegetated soil layer. Benson said that high emissions from a landfill can violate air quality standards. Gary Lass said that the South Coast Air Quality Management District lists requirements for monitoring surface emissions.

• Generating a protocol for measuring gas transfer. Licht asked whether ACAP members will provide a protocol for measuring gas transfer through landfill caps. He strongly encouraged ACAP members to do this, noting that several European agencies are evaluating gas transfer in detail. Wilson said that there are no plans to develop such a protocol at this point. He did acknowledge, however, that ACAP is in its infancy and may be willing to address other issues like this in the future. Benson said that ACAP's goal is to provide a baseline system for evaluating water flow through a cover. If site owners wish to evaluate other processes simultaneously, he said, they should feel free to pursue additional lines of testing.

• Relationship between landfill gas and water movement and production. Meeting participants noted that landfill gas can push moisture through a cover. Additionally, methane combustion processes generate water (see Mark Ankeny's presentation below).

• Altering boundary conditions. One participant asked whether placing a lysimeter under a cover will create a capillary break and alter boundary conditions. Gee said this should not be a problem if the lysimeter is placed deep enough. Benson admitted that the lysimeter will create a capillary break, but noted that this same type of break is also created when covers are placed over solid waste. This occurs any time, he continued, when a finer medium is placed on top of a coarser medium.

• Retrofitting existing covers for the ACAP. Earlier in the meeting, McMillan pointed out, Rock said that some lined sites can be retrofitted to produce the data that ACAP requires. McMillan asked whether ACAP will provide guidance on how to retrofit these sites. Rock said that ACAP will provide guidance as the sites are brought into the program.

• Alternatives to lysimeters. At this point, Gee said, lysimeters are the only commercially available product that measure water flux directly. Several other instruments can be used to estimate flux, but the accuracy of their estimates is uncertain. Gee said that his laboratory is developing a meter that will measure water flux in unsaturated soils. He plans to test the instrument over the next couple of years. Assuming that it performs well, it will serve as a cost-effective alternative to lysimeters.

• Relationship between water-content and water-flux measurements. Gee stressed that water-content measurements cannot always be used to infer water movement through a system. At a site in France, for example, researchers incorrectly assumed that no water movement was occurring because no changes in water content were detected. (Gee said that water-content changes of 0.1% in coarse sand can cause fluxes up to 10 to 30 inches per year.)

• Preferential flow. Several participants asked whether ACAP members have considered potential impacts from channeling and preferential flow. Wilson acknowledged that plant roots, burrowing rodents, worms, and ants can create channels, but he stressed that only a small percentage of macropores operate as water conduits. Wilson said that the ACAP team has not made efforts to study preferential flow in numerical models yet, but that this process may be evaluated later in the program. Wilson noted that two currently available models—an Australian code and the Agriculture Research Service's model of root zone water quality—do evaluate preferential flow.

• Impact of plant die-offs. One participant asked whether studies are being conducted to determine how plant death impacts covers. Rock reminded participants that vegetated covers will require maintenance; the impact of dying plants should be offset by replanting activities.

• ACAP's goals. Licht asked what ACAP hopes to accomplish during their 5-year evaluation period and whether their data will be used to change existing problematic laws. Benson said that ACAP does not intend to eliminate the use of prescriptive covers. Rather than preempting existing practices, ACAP hopes to provide additional options for waste containment. (Gee reminded participants that no cover designs are expected to work efficiently in all climatic conditions. For example, ET covers have proved ineffective at Hill AFB and at a particular site in Idaho, where snowmelts during certain times of the year caused plutonium to leach.) Rock said that the ACAP team will not write new regulations, but that the data generated could help regulation writers make modifications to existing laws. As was indicated earlier, Rock said, ACAP has two main goals: (1) give regulators enough information to make informed decisions about alternative covers and (2) improve numerical models so that they provide accurate drainage predictions. Benson said that developing tools to better predict and manipulate drainage may prove useful for a variety of researchers, including containment-cover designers and bioreactor designers. Perhaps one day, Benson said, bioreactors and containment covers will both be used at a given waste site—the former could be used to remediate wastes and the latter could be installed afterwards to contain remaining wastes. Gee stressed that researchers cannot expect to answer all questions about alternative covers over just 5 years. Rock reminded participants that the project does not have to end after 5 years and that data can continue to be collected.

• Data required by regulators. Albright asked regulators to list the information that they still need before they can feel confident in their ability to evaluate alternative covers. Dixie Lass listed the following:
  • Site-specific soil and climate data. Lass said that she will not accept modeled predictions unless site-specific soil and climate data are incorporated. She stressed the importance of accounting for different microclimates. Panel members assured her that this type of data will be collected at ACAP sites. She recommended installing meteorological equipment and other monitoring systems one year before starting the design phase.

  • Detailed information about plant community dynamics. Lass said that models should incorporate information about specific types of plants. During a presentation, she said, she wants investigators to indicate which plants are incorporated into their models and to explain why those specific plants were chosen. She said she will ask whether botanists participated in plant selection. Wilson said there are some agriculture-based codes that model plant community dynamics. Unfortunately, few of these models evaluate the plant types that are being considered for use in landfill covers.

  • Gas extraction system. Lass said that she will not approve covers unless gas extraction systems are in place. She also mentioned that gas removal systems remove a substantial amount of moisture and she recommended accounting for this removal in numeric models. Wilson said that some available codes (e.g., UNSAT-H) can already account for vapor transport processes.

• Comparing performance of alternative to prescriptive covers. Benson noted that it will be difficult to compare performance between prescriptive and alternative covers because so little data have been collected on RCRA-C and RCRA-D covers. He said straightforward comparisons can be made if site owners are willing to conduct side-by-side comparisons at their sites. In fact, he said, the ACAP originally proposed setting up three test pads at each site to test one prescriptive cover and two alternative covers. This proposal was revised, however, due to the expense associated with installing three lysimeters.

• Long-term changes in performance. Jody Waugh said that several uranium tailing sites have been covered under the Uranium Mill Tailings Remedial Action (UMTRA) project. Efforts are currently underway, he said, to determine whether these covers will behave as designed over long period of time and whether changes in performance translate to changes in risk.

• Plant monitoring and modeling activities. Albright said that ACAP hopes to draw local universities into their projects. He said that evaluating a small component of a cover, such as analyzing plant communities, would serve as an excellent Master's thesis.

• Greenhouse tests. One participant asked whether it makes more sense to include fewer sites in ACAP and focus intensive efforts on simulating different environmental events. Panel members said that artificial weather events fail to mimic real-life conditions. For example, Gee noted, it is difficult to simulate the kinetic energy and humidity of rainfalls. Gee said that it might be better to look at analog sites (i.e., sites that have a similar design) to see how they perform during natural weather events.

• Statistical dependability of data sets. McMillan asked whether ACAP members are worried that they will receive criticism for using just one lysimeter per site. Panel members said that they would prefer to install multiple lysimeters per site, but that this is cost-prohibitive.

• Integrating state regulatory agencies. One participant noted that state regulators across the country are being asked to evaluate alternative cover designs. He asked whether attempts have been made to consolidate the states' efforts within one organization. Albright said that no organization has been formed, but that the Association of State and Territorial Solid Waste Management Officials (ASTSWMO) has been contacted. One participant noted that ASTSWMO is holding a conference in June or July 1999 in San Diego, California. Rock recommended sending Albright to the meeting to make a "sales pitch" for ACAP. Rock reminded meeting participants that ACAP welcomes new members. He encouraged participants to contact him if they know of people who should be invited to future meetings.

• RTDF Web site. Rock encouraged meeting participants to visit the RTDF web site (http:\\www.rtdf.org) for a summary of previous ACAP meetings and conference calls. He noted that the Web site also contains a bibliography that lists more than 1,400 papers on phytoremediation. (The list is updated twice a year.)

CAPPING RESEARCH ACTIVITIES IN OTHER AGENCIES

Alternative Landfill Cover Activities at DOE Sites
Scott McMullin, DOE

Scott McMullin said that he works out of DOE's Office of Technology Development and focuses on subsurface-contamination problems. He said that DOE has several waste units located around the country. While some of these units contain sanitary, hazardous, or mixed waste, others contain radionuclide materials. Since some of these contents have half-lives in the 10,000-year range, many of the radionuclide sites require containment systems that will perform over long time periods. McMullin said that DOE has experimented with a wide variety of subsurface-management options. While some have been "low-energy" approaches that require minimal disruption, others are "high-energy" approaches (e.g., excavation) and are quite disruptive. McMullin said that alternative covers fall into the "low-energy" category and are being tested at several DOE sites, including:

• Hill AFB. McMullin said that a "Hanford-type" cover was installed at this site with the purpose of investigating long-term performance. (A climatic change study was performed, extrapolating the cover's performance under a different climatic scenario.) McMullin said that activities at this site laid the benchmark for long-term performance evaluations.

• Los Alamos. McMullin said that alternative cover performance studies are ongoing at this site, despite limited funding. At this site, he said, studies were conducted to evaluate the impacts of plant intrusion.

• Sandia Laboratories. McMullin said that studies performed at this site are being conducted to provide incentives for deploying alternative covers. He said that a variety of cover designs have been evaluated based on cost and performance and that a summary report will be submitted to EPA in the future. For this site, McMullin explained, performance criteria were established during an early meeting that was attended by New Mexico's regulators. He strongly encouraged meeting participants to involve regulators early on in their demonstration projects.

• Savannah River. McMullin said that subsidence studies are being performed at this site (see Mike Serrato=s talk below). (McMullin said that other subsidence studies are ongoing at a site in Nevada, but he did not provide any details.)

McMullin said that DOE has performed work at a variety of other sites as well, and that a CD-ROM has been created to compile 15 years of data. McMullin said that DOE will produce additional copies of the CD-ROM and he encouraged meeting participants to contact him if they want a copy. He stressed that it is important to get all available data into one searchable format so that researchers do not waste time "reinventing the wheel." McMullin said that it is a challenge to collect a broad range of performance data. He said that researchers try to research topics that are generic to a number of sites without completely sacrificing site-specific considerations. He commended ACAP's approach of researching a network of sites.

McMullin said that DOE is struggling to determine how to dispose of materials in a manner that minimizes risk and causes minimal intrusion. About 4 years ago, McMullin said, regulators, end-users, and researchers met in Utah to talk about long-term performance. Today, the Long Term Capping/Cover System Strategy is in place to address this issue. This programmatic strategy is intended to help researchers meet the DOE 2006 Accelerated Cleanup Plan goals: (1) coupling long-term performance to risk, (2) designing guidance that is focused on varied life cycles, and (3) installing performance verification.

McMullin provided examples of ways that risk analyses can be used to drive design decisions. One example involved a waste site in Idaho. The site, he explained, did not have a hydraulic component and only posed a health risk via dermal exposures. For this reason, big boulders were placed on the waste. McMullin said that the presence or absence of institutional controls also can affect design decisions. If institutional controls will be in place forever, he said, covers do not have to be designed to perform for thousands of years because site owners will continually repair the site to mitigate risks. If the site has been abandoned, however, a cover designed like a lower-maintenance natural system should be considered. McMullin noted that covers will not need to be designed for long-term performance if they are going to cover a waste site that has rapidly degrading radionuclide wastes. One participant asked McMullin which models are being used to assess risk. McMullin said that using a risk approach to justify design is a relatively new approach. As a result, DOE has not yet identified which risk models they prefer.

McMullin said that DOE hopes to produce user-friendly guidance documents that include decision matrices for designers. He said that end-users will be asked to review the document to ensure that it is useful. McMullin said that DOE is evaluating several approaches to minimize maintenance and troubleshoot problems. For example, he said, DOE has developed ways to detect flaws that occur during construction activities. (For example, holes in the centimeter range have been detected and reported.) Also, he said, DOE is researching the possibility of using airborne surveillance to evaluate covers in remote areas. Additionally, DOE is experimenting with "smart materials" (i.e., those that provide feedback on their own performance).

In closing, McMullin encouraged interagency collaboration on alternative cover performance evaluations. He said DOE would like to work with ACAP, EPA, the Department of Defense (DOD), state regulators and other organizations. (DOE has already established a working relationship with the Southern States Energy Board and the state of California.)

Activities at Mining Sites
Mark Ankeny, Daniel B. Stephens & Associates, Inc.

Ankeny said that he has worked on several mining sites, noting that final closure at these sites can be very costly due to their size. (For example, one of Ankeny's clients has an area of 10 square miles that requires closure.) Several complicated issues are involved with reclaiming and recovering mining sites, he said, because (1) reduced materials are located at the bottom of a system and "good" materials are located on top, (2) uniformity is usually lacking, and (3) coarse, acid-generating materials are often present. Complicating the problem, he said, is the fact that mining sites have become financially fragile due to increased copper mining in Chile and decreasing gold prices. Although several U.S. mines lose money by staying open, many continue to mine gold and copper to avoid initiating expensive closure activities. (The regulations do not require active mining sites to initiate final closure activities.)

In addition to his work at mining sites, Ankeny said, he has also been involved with containment projects at MSW sites, Subtitle-C sites, and radionuclide-waste sites. He encouraged the use of containment systems that will perform as long as their encased wastes remain hazardous. He pointed out that plastic liners will degrade far earlier than radionuclide wastes.

Ankeny brought up four issues that he wanted meeting participants to consider:

• "Gas tubbing" can result when trying to avoid "bathtubbing." Ankeny said that permeability requirements in federal waste containment regulations were written to prevent "bathtubbing" from occurring. These regulations, he said, were written with humid climates in mind. Using the same permeability requirements in more arid climates can cause "gas tubbing" (i.e., the movement of gas through waste layers). Some studies indicate that carrier gases (e.g., methane and carbon dioxide) cause problems by reducing metals and dragging chlorinated solvents. Reduced metals, Ankeny explained, can be construed as contaminants if they reach underlying ground-water layers. Often times he said, subsurface processes lead to methane combustion, a process that generates significant amounts of water. This water, Ankeny said, can act as the vehicle that carries the reduced metals to the ground water. Ankeny recommended including water created in waste layers in numeric models.

• Ground water can flow uphill. Ankeny said it is not uncommon to find ground water flowing uphill in the western part of the country. He said that numerical water-balance models should account for this flow pattern.

• Plant root functions need to be parameterized. Ankeny encouraged the ACAP team to develop models that predict plant behavior.

• Water-balance data need to be tied to overall facility performance. Ankeny said that most water- balance models provide simple 1-dimensional water flow estimates. He said this information needs to be tied to overall facility performance to make risk estimations.

In closing, Ankeny presented two decision trees, the Cover Design Event Tree and the Cover Design Fault Tree. He recommend providing them to site owners (See Attachment F).

OTHER TECHNICAL ISSUES FOR ALTERNATIVE COVERS

Landfill Gas Issues for Design of Monofill Alternative Covers
Horacio Ferriz, GeoLogic Associates

Ferriz said that there are two schools of thought regarding the need to address landfill gas issues in the design of monofill alternative covers:

• The cover is not intended to control gas. Under this assumption, Ferriz said, all gas management is controlled by an extraction system and gas migration through the cover does not need to be addressed in comparing the performance of one type of cover with that of another.

• The cover is part of the gas control system. Gas extraction systems are not always present in old and abandoned landfills. Gas exits the landfill, therefore, solely through diffusion and convective flow. Ferriz said that uncontrolled releases are of concern to toxicologists, air-pollution regulators, and landfill landscapers.

Horacio Ferriz said that he is actively participating in efforts to estimate landfill gas flow. He said that the mathematics are complicated by the heterogeneity of a porous medium, barometric pumping in response to diurnal variations in barometric pressure, departures from ideal gas behavior, and spatial variations in landfill gas pressure. He said that gas flows through the subsurface via convective flow (movement of gases from high-pressure areas to low-pressure areas) and diffusive flow (movement of gas from areas of high partial pressure to low partial pressure). He said that modeling gas flow through porous materials requires a set of equations describing mass transport for each gas, including terms for convective and diffusive flow. Ferriz provided several equations that he thinks can be used to estimate flow. These equations, which include Darcy's law, the mechanical energy balance equation, Leva's correlation, and Fick's law, are presented in Attachment G.

Ferriz stressed that he is at the beginning stages of his research. He acknowledged that the ACAP focuses on water-infiltration issues rather than gas-transfer issues. He encouraged participants who are interested gas migration, however, to instrument their sites to collect information on gas pressures (using a flux box) and gas concentrations (using a device that measures partial pressure) at various levels in the cover. He expressed strong enthusiasm for creating numerical models that predict gas migration through porous media. In closing, he encouraged meeting participants to review available literature. (A partial reference list is included in Attachment G.)

Dynamic Stability of Alternative Final Covers
Robert Anderson, California Energy Commission

Robert Anderson briefly described how seismic activity affects alternative covers. He said that he has evaluated monofill, GCL, GCL/monofill, and FML/monofill covers. Several alternative covers of this type, he explained, are located in California, including monofill covers at Baker Landfill, Mission Canyon Landfills, Puente Hills Landfill (side slopes only), and Milliken Landfill (east mound only); and GCL covers at Annapolis Landfill, Laytonville Landfill, Penrose Landfill, Cold Canyon Landfill, McCourtney Road Landfill, and Morongo Valley Landfill. Anderson said that Earthquake SPECTRA (Volume 14, Issue 2, May 1998, pages 319-334) provides information on how some of these landfills fared during the 1987 Whittier Narrows earthquake, the 1989 Loma Prieta earthquake, the 1992 Landers/Big Bear earthquakes, and the 1994 Northridge earthquake. (This article is presented in Attachment H.)

Anderson said that some damage (e.g., cracking or tearing) is likely to happen during an earthquake. Investigators must decide how much damage is acceptable on a site-by-site basis, he said. He reminded participants that cover damage is not the only problem that occurs during earthquakes; in many cases, earthquakes shut down landfill gas control systems by causing power outages. Anderson said that a schedule has been compiled to rate damage to landfills.

Anderson talked briefly about two landfills that experienced damage: the Chiquita Canyon Landfill and the Mission Canyon Landfill. At the former, he said, the geomembrane tore and the interim cover was sheared and broken during an earthquake. Significant damage (i.e., cracks about 40 to 50 meter long and about 3 meters deep) has also been reported at the Mission Canyon Landfill, but investigators concluded that the damage did not result from an earthquake. (Neglect and improper maintenance of drainage systems were identified as causative factors.).

Anderson said that several factors are important in determining how stable a landfill will be during an earthquake. These include (1) side slope inclination, (2) peak and residual friction of GCL and GCL interfaces, (3) peak and residual friction interfaces for FML (rough and smooth), (4) peak and residual strength of soils (cohesive versus noncohesive) and interfaces, (5) drainage elements, (6) anticipated peak horizontal ground acceleration, (7) site maintenance history, and (8) dynamic compaction of waste. Additionally, he said, neighboring features can pose risks to landfills. For example, he noted, landfills that are located next to unstable hills may be impacted during an earthquake.

Anderson said that the literature offers a limited number of observations regarding landfill performance during earthquakes. Most of the observations have been on monofill covers and indicate that landfills perform reasonably well during strong ground shaking; it is hard to hurt a landfill. Anderson encouraged people to contact him if they have additional questions (see Attachment A for contact information).

Subsidence Impacts at DOE's Savannah River Site
Michael Serrato, Westinghouse Savannah River Company

Michael Serrato provided a brief overview of the Savannah River site, noting that this 310-square-mile site has five inactive reactors. Serrato said that regulators have already given DOE (the site owners) permission to use waste containment covers consisting of 3 feet of clay. Installing such covers, however, proves to be a challenge. Therefore, DOE is investigating whether alternative covers, consisting of composite systems with GCL geomembranes, can be used at this site. Serrato said that attention is being directed to foundation layers, noting that these are important both hydrologically and structurally. Serrato said it will be a challenge to convince regulators that thin alternative covers will provide the same protection as thick clay covers.

Serrato said that DOE plans to install covers over MSW areas. Because these materials degrade, he explained, there is some concern that subsidence will impact cover performance. To address these concerns, Serrato said, investigators are evaluating how subsidence causes landfill collapse and impacts the in situ hydraulic conductivity of a system. Their experiments involve using an industrial vacuum truck to suck sand out of test systems. About 1 yd³ is removed once every 2 weeks. Collapse is considered to occur when surface deflections reach 1 foot. Experiments are being carried out in two phases, Serrato explained. During Phase I, test pads were configured with the following layers: (1) 1 to 2 feet of cover soil; (2) 2 feet of compacted sandy clay; and (3) 4 feet of loose sand at the bottom. Four test pads with this profile were constructed to mimic the following conditions: (1) high-stress conditions in saturated soil, (2) high-stress conditions in unsaturated soil, (3) low-stress conditions in saturated soil, and (4) low-stress conditions in unsaturated soils. (High and low stress are defined by the intensity with which experimenters removed sand.) Serrato said that each test pad has a footprint of about 50 feet by 125 feet and is instrumented so that surface deflections can be measured and moisture fronts can be monitored.

With the low-stress pads, Serrato said, no major deflections or changes in hydraulic conductivity have been observed even though about 8 yd³ of materials have already been removed. The test pads subjected to high stress, however, did collapse after about 8 yd³ of sand removal. (The cavity span reached 5 feet in the saturated condition and 6 feet in the unsaturated condition before collapsing.) Investigators found that hydraulic conductivity (measured at about 2 x 10⁴) did not change substantially until about 2 weeks before collapse occurred. This indicates, Serrato said, that hydraulic conductivity did not degrade substantially in response to a growing subsurface cavity.

During Phase II, alternative composite systems with GCL geomembranes are being tested over three test pads. Sand removal is currently ongoing. Serrato said that metallic tape has been placed on the liner's seams so that movement can be detected. Also, investigators have placed heavy loads on the surface to determine whether this would induce collapse or damage the system. To date, about 8 yd³ of sand have been removed and only small deflections have been observed. The study is ongoing.

Side Slope Configurations
Glendon Gee, Battelle
W. Jody Waugh, Roy F. Weston, Inc.

Gee noted that site designers must think carefully about stabilizing side slopes without sacrificing resistance to water infiltration. If stability is not considered, he said, extreme storm events may cause erosion and slope failure. If water-infiltration resistance is sacrificed, however, environmental problems can result. (At one radioactive site at which Gee worked, the most dangerous and soluble wastes were located in the corner of a landfill, which is typically the area that is covered with side slope.)

Gee described a study that was performed to compare two side slope configurations. The study involved one cover, which had a clean-fill dike configuration on one side and a rock slope configuration on the other. (The clean-fill dike side slope was composed of pit-run gravel and was relatively flat [10:1 slope] and had a large footprint [55 meters from the soil edge]. The rock slope was composed of fractured basalt and was relatively steep [2:1 slope] and had a small footprint [12 meters from the soil edge]). The side slopes were evaluated for their ability to protect fine soil layers and to minimize space requirements, infiltration, underflow, and recharge. After 4 years of study, no significant subsidence or deflection was observed on either side slope. Investigators concluded, therefore, that both slopes are stable. To evaluate their protectiveness against water infiltration, investigators simulated extreme weather conditions by applying three times the annual average precipitation through the year and then simulating a 1,000-year extreme storm event once a year. Investigators measured water-balance components (e.g., water inputs, water storage, drainage, runoff, and ET) and found that the timing of precipitation events greatly impacts the side slope's performance. Gee said that less water drained out of the bottom of the rock slope. The difference in drainage between the two types of side slopes was hypothesized to be due to advective drying processes. In fact, early modeling results (obtained through STOMP) confirm this hypothesis. Modeling results also indicate that unprotected side slopes allow about 20% to 30% of precipitation into the waste system.

Waugh emphasized that there are many issues to consider when designing stable side slopes and that different designs are required for different climates. For example, he said, grassy side slopes perform well in eastern humid areas, but this type of vegetation cannot always be supported in western semiarid and arid sites. Likewise, while boulder side slopes may be effective in western climates, these are often overtaken by deep-rooted woody plants in more humid eastern areas. He said that a rock slope was applied at a site in Arizona, but it has caused a water-harvesting problem.

Waugh recommended studying naturally stable side slopes to get a better idea of how to design engineered systems. In Colorado, Waugh continued, he researched the Beaver Gulch side slope. He said that geological studies indicate that this naturally-occurring slope has been stable for a long time. (This conclusion was made by evaluating desert varnish [i.e., magnesium oxide coatings on rocks], lichen-growth patterns, and soil profiles.) Also, Waugh said, evidence suggests that the slope exhibits a favorable water balance. (Information about water movement was gathered by evaluating soil-profile structures and collecting data on soil-moisture profiles. Without direct flux measurements, Waugh said, further investigation would be needed before making a more conclusive statement about water balance.)

Waugh said that the Beaver Gulch side slope received 20 centimeters of precipitation a year. It rolls on for about 1,000 to 1,500 feet and has an average slope of 5:1, with some parts at the top that are steeper (3:1 slope) and other parts that are less steep (10:1 slope). The slope's surface is covered with rocks (e.g., boulder-size basalt), soils (e.g., silts, clays, and loams), litter (shales and sandstones), and plants. The rocks cover about 37% of the slope and the plants cover about 46%. The plant mix consists of cheat grass (65%), galleta (32%), Indian rice grass and wheat grass (1%), and shrubs (2%).

Waugh said that the information gathered from the Beaver Gulch side slope may help designers engineer slopes at nearby landfills. He said that two landfills are located nearby. One is an UMTRA site and already has a rock side slope. The other, however, is a MSW landfill and the cover and side slope design have not yet been determined.

CLOSING COMMENTS
Steve Rock, EPA

Rock thanked EPA Region 9 and DRI for their roles in planning the meeting. He also thanked meeting attendees for their participation. Rock noted that there is often conflict between regulators and the regulated community. Although this conflict creates creative tension that spurs innovation, it is not always desirable. Rock said that ACAP can serve as the vehicle to get regulators, site owners, researchers, and consultants together to work in unison. He recommended thinking of ACAP as a roundtable, with several chairs that still remain empty for future participants.

Attachments A-H

Attachment A: Final Attendee List

Attachment B: Introduction to ACAP (Steve Rock)

Attachment C: Phase I Task A Report (Facilities Review) (William Albright)

Attachment D: Phase I Task B Report (Models Review) (Glenn Wilson)

Attachment E: Engineering and Monitoring of ACAP Sites (Craig Benson)

Attachment F: Cover Design Event and Fault Trees (Mark Ankeny)

Attachment G: Landfill Gas Issues for Design of Monofill Alternative Covers (Horacio Ferriz)

Attachment H: Journal article from Earthquake SPECTRA