INSTALLATION OF REACTIVE METAL GROUNDWATER
COLLECTION AND TREATMENT SYSTEMS AT THE ROCKY
FLATS ENVIRONMENTAL TECHNOLOGY SITE GOLDEN, CO
John K. Hopkins, Ph.D. and Annette L. Primrose
Rocky Mountain Remediation Services, L.L.C.
Norma Castenada
U.S. Department of Energy
ABSTRACT
Three groundwater plumes contaminated with volatile organic compounds (VOC's) and radionuclides at the Rocky Flats Environmental Technology Site are scheduled for remediation by 1999 based on the Rocky Flats Cleanup Agreement (RFCA). These three plumes are among the top 20 environmental cleanup sites at Rocky Flats. One of these plumes, the Mound Site Plume, is derived from a previous drum storage area, and daylights as seeps near the South Walnut Creek drainage. Final design for remediation of the Mound Site Plume has been completed based on use of reactive metals to treat the contaminated groundwater, and construction is scheduled for early 1998.
The two other plumes, the 903 Pad/Ryan's Pit and the East Trenches plumes are derived from VOCs associated with drums that either leaked or were disposed in trenches. These two plumes are undergoing characterization and conceptual design in 1998 and construction is scheduled in 1999. The contaminants of concern in these plumes are tetrachloroethene, trichloroethene, carbon tetrachloride and low levels of plutonium and americium.
The RFCA includes a two phase groundwater strategy: 1) to remove the source of contamination, and 2) to prevent contamination of surface water by applying action levels to groundwater which are protective of surface water. Collection and treatment of the contaminated groundwater plume derived from the Mound Site is consistent with this strategy.
The objective for Rocky Mountain Remediation Services, L.L.C. was to design and install a simple intercept and treatment system for the Mound Site Plume to protect surface water to the extent practicable, treat contaminated groundwater to levels consistent with the RFCA for unrestricted discharge, to minimize the generation of low level mixed waste, and to demonstrate a low cost and effective technology for other contaminated groundwater plumes at RFETS and at other DOE sites. The selected treatment system will utilize an innovative, primarily passive, reactive metals treatment system for VOC and radionuclide removal. The Mound Site Plume collection system is designed to be effective for a low flow, low permeability claystone regime.
EPA's SITE Program will monitor the operation of the Mound Site Plume. The data gathered from the monitoring program will be used by RMRS and ETI to optimize the design of the collection and treatment systems for the 903 Pad/Ryan's Pit and East Trenches Plumes.
INTRODUCTION
The Mound Site Plume contains chlorinated organic compounds, americium and uranium in excess of action levels defined in the RFCA (DOE 1996). The proposed action will consist of constructing a subsurface groundwater collection system coupled with a passive reactive metals treatment system to treat contaminated groundwater from the Mound Site Plume and asurface water seep to the surface water action levels specified in the Rocky Flats Cleanup Agreement (RFCA) (DOE 1996). The project will be conducted in accordance with RFCA, DOE Orders and RFETS policies and procedures. The project will remediate one of the top ten Individual Hazardous Substance Sites (IHSSs) at RFETS. Funding for this project was provided by DOE EM50 - Subsurface Contaminant Focus Area.
Background
The Mound Site Plume is located on the east side of the RFETS Industrial Area, and east of the protected area fence (see Figure 1). This plume of primarily volatile organic compound (VOC) contaminated groundwater is believed to originate from the Mound Site, and extend northward to where the plume discharges as seeps and subsurface flow into the South Walnut Creek drainage. Low levels of uranium and metals below background levels have been detected at a seep.
Most of the groundwater contamination is believed to be derived from the Mound Site where approximately 1,405 intact drums were stored on the ground surface, covered with soil between April 1954 and September 1958. The drums contained uranium and beryllium-contaminated lathe coolant (a mixture of approximately 70 percent hydraulic oil and 30 percent carbon tetrachloride). Historical information also indicates that some of the coolant contained low levels of plutonium. In 1970, all drums along with some radiologically-contaminated soil were removed from the Mound Site. Approximately 10 percent of the drums were thought to be leaking at the time of removal. However, there are no records of the volume of contaminants released to the soils at the Mound Site (DOE 1992).
An accelerated removal action was completed in Spring 1997 to excavate the soil contaminated with VOCs above Tier I action levels from the Mound Site (DOE 1997a). Low temperature thermal desorption technology was used in the Summer of 1997 to remove the VOC contaminants of concern from the excavated soils. The treated soil below specified action levels was returned to the Mound Site excavation and the area was revegetated.
Hydrogeologic Setting
At the source area for the Mound Site Plume, bedrock unconformably underlies approximately 12 feet of surficial deposits and consists of weathered claystone and minor sandstones of the Cretaceous Arapahoe and Laramie Formations (DOE 1995, DOE 1997a). The Arapahoe No. 1 Sandstone subcrops under the northwest corner of the Mound Site, and is truncated to the north by the South Walnut Creek drainage. Groundwater within the Arapahoe No. 1 Sandstone exits into the colluvium, causing a higher water table, and an increase in vegetation in an area of intermittent seeps (DOE 1995, RMRS 1996a). The Arapahoe No. 1 Sandstone is absent under the eastern portion of the Mound Site. Figure 1 illustrates the hydrogeologic setting for the plume.
Near the distal end of the plume, clay-rich colluvium partially derived from the Rocky Flats Alluvium unconformably overlies Laramie Formation claystone (DOE 1995, EG&G 1995a, EG&G 1995b, RMRS 1996c, and RMRS 1996a). The elevation to bedrock is variable as this area has been extensively disturbed by landslides and/or slumps. Aerial photographs showed that the area was extensively regraded in 1962.. The bedrock surface forms a shallow trough plunging to the north, which probably directs groundwater flow. Depth to the bedrock surface varies from 5 to 15 feet over much of the area. At the eastern extent, bedrock is 25 feet below ground surface due to fill material brought in for the eastern perimeter road.
The groundwater occurs in the alluvium, colluvium, in the weathered bedrock, and in the underlying Arapahoe No. 1 SandstoneNumber One Sandstone. Groundwater flow in the alluvium and colluvium is primarily to the north and follows along the bedrock surface. Recharge occurs primarily through local infiltration of precipitation or local runoff. Geometric mean hydraulic conductivities are 6 x 10-04 cm/sec for the Rocky Flats Alluvium and 8 x 10-08 cm/sec for the weathered claystone (DOE 1995). Geometric mean hydraulic conductivity for the colluvium is 9 x 10-05 cm/sec (EG&G 1995b).
As shown on Figure 2, the water table elevation contours parallel the stream, with the major flow direction to the north. The water table elevation is very close to the elevation of the stream, necessitating care in placing the collection system to ensure that stream water is not captured.
The groundwater discharges through subsurface seeps from the bedrock into the colluvium along the hillside, seeps on the south bank of South Walnut Creek, and through evapotranspiration. Infiltration into the underlying unweathered claystone is limited (DOE 1995, EG&G 1995b). Depending on the season, unsaturated areas may occur within the plume (DOE 1996b, EG&G 1995b, RMRS 1996a). At seep SW059, groundwater containing low levels of VOCs with trace amounts of radionuclides discharges at a rate averaging less than 0.5 gallons per minute. The seep water is currently collected, transported and treated onsite (DOE 1995, RMRS 1996b). The approximate quantities of water collected from seep SW059 are listed in Table I. The Spring of 1995 was exceptionally wet.
Table I. Quantity of Water Collected from Seep SW059 by Quarter
Quarter | Gallons/quarter |
July-Sept. 1995 | 32,526 |
Oct.-Dec. 1995 | 16,930 |
Jan.-Mar. 1996 | 17,285 |
April-June 1996 | 18,775 |
July-Sept. 1996 | 13,095 |
Oct.-Dec. 1996 | 11,605 |
Jan.-Mar. 1997 | 7,268 |
Based on historical flow rates from seep SW059, available hydrogeologic data, and typical rates for other groundwater drains at the Site, the groundwater flow for the Mound Site Plume was calculated to be 0.1 to 2 gallon per minute. This flow rate assumes that water will not be depleted from South Walnut Creek, recharge to the hill side is not significantly altered from current conditions and field data is representative of the actual conditions. During extended periods of below average precipitation, it is possible that the flow will effectively approach zero.
Mound Site Plume Contamination Data Summary
Tetrachloroethene is the predominant contaminant found in soil and groundwater at the Mound Site, with the highest historic groundwater concentration of 528,000 ug/l in Well 0174 (Figure 1). Concentrations decrease towards South Walnut Creek, which supports the Mound Site as the source area for the contaminants seen in this plume.
From the source area to the distal end of the groundwater plume, the most commonly detected groundwater contaminants in the Mound Site Plume are tetrachloroethene and trichloroethene. Carbon tetrachloride is detected only on the weastern side of the plume; at seep SW059 and at location 10397 (Figure 1). This may indicate that there is a separate source of contamination in the road fill. Both dichloroethene and vinyl chloride are present in the distal portion of the plume, and are degradation products of trichloroethene and tetrachloroethene (RMRS 1996a, DOE 1995, DOE 1996b).
Soil and groundwater samples were collected near the distal end of the plume during Spring 1997 to support the design of the collection and treatment system. The analyses indicate that the highest groundwater concentrations in the distal end of the plume are trichloroethene (TCE) at 844.5 ug/l, tetrachloroethene (PCE) at 260.8 ug/l, and cis-1,2 dichloroethene at 808 ug/l seen at location 10797, directly downgradient of the Mound Site source area (Figure 1). Table II summarizes the groundwater results of this investigation.
Table II. Groundwater Contaminants of Concern from Recent
Investigation Results (in ug/l).
|
|
|
|
Number of Detects |
Tier
I |
Tier
II |
Vinyl Chloride | Nd |
55.0 |
13.0 |
5 |
200 |
2 |
1,1 Dichloroethene | Nd |
94.2 |
18.0 |
8 |
700 |
7 |
cis 1,2 Dichloroethene | Nd |
808.0 |
169.0 |
9 |
7000 |
70 |
Carbon Tetrachloride | Nd |
6.6 |
0.8 |
1 |
500 |
5 |
Chloroform | Nd |
177 |
17 |
6 |
10000 |
100 |
Trichloroethene | Nd |
844 |
195 |
9 |
500 |
5 |
Tetrachloroethene | Nd |
261 |
66 |
8 |
500 |
5 |
Methylene Chloride | Nd |
18 |
4 |
3 |
500 |
5 |
Americium 241* | 0.25 |
0.25 |
0.25 |
1 |
15 |
0.15 |
Gross Beta* | 3.1 |
28 |
8 |
14 |
19** |
|
Uranium 234* | 3.4 |
3.4 |
3.4 |
1 |
298 |
3 |
* Insufficient water to obtain radiological analyses, data is from seep
SW059
** Surface water action level for Walnut Creek
PROJECT APPROACH
A downgradient capture system will be installed in early 1998 near South Walnut Creek to intercept contaminated groundwater and to minimize impacts to surface water. A subsurface groundwater collection system will be coupled with a passive reactive metals treatment system to treat contaminated groundwater from the Mound Site Plume to the appropriate surface water action level specified in the RFCA (DOE 1996)Rocky Flats Cleanup Agreement (RFCA). The downgradient capture system was chosen as the best remediation method following an evaluation of other more traditional options in the Groundwater Conceptual Plan (RMRS 1996a). The passive treatment system was chosen as it effectively treats VOCs and radionuclides to below action levels at a lower operations and maintenance cost than other treatment options. The treated water will then be discharged to surface water.
Proposed Action
A funnel (impermeable barrier groundwater collection system) and gate (treatment system) will be keyed into the underlying claystone for flow cut-off and treatment of the collected groundwater (Figure 3). Based on the available data, to capture the contaminant plume, a groundwater collection system will be installed that extends from the western road approximately 250 feet to the east (Figure 1).
The variable elevation of the bedrock surface and the similarity between the clay-rich colluvium and bedrock makes it difficult to install a collection system keyed a certain depth into bedrock. The clay-rich colluvium and bedrock have similar properties, effective collection of the contaminated groundwater is not dependent on being keyed into bedrock. Therefore, the collection system will be installed at a variable depth of approximately 8 to 15 feet across the site, at least 6 inches, but up to several feet, into claystone, without regard to whether this is colluvium or bedrock. The contaminated groundwater will be treated in a series of cells containing reactive iron filings to remove VOCs and radionuclides. Under normal operations, the treated water will be discharged to groundwater using an infiltration galley located adjacent to South Walnut Creek. However, the system is designed to allow discharge directly to surface water in South Walnut Creek.
After installation of the funnel and gate system, reclamation of the collection/treatment area will be performed. The existing seep collection system will continue to operate to the extent practical, until the new system is operational. However, it is likely that installation of the funnel and gate system will require decommissioning of the existing system.
Installation of Funnel and Gate System
Conventional excavation/trenching techniques will be used to install the funnel and gate system. Silt fences will be installed downgradient of the excavation to control potential release of sediment to the drainages. During trench construction, the material removed from the trench will be stockpiled adjacent to the trench. A horizontal groundwater-collection line will be installed on the upgradient side of the impermeable barrier. Filter pack or pea gravel will be installed from the top of the claystone to the level of the horizontal collection line. The trench will then be backfilled and excess fill will be spread over the top of the collection system. Figure 4 shows the details of the trench construction.
Treatment and Discharge
A reactive metals treatment system will be used to degrade dissolved VOCs and remove radionuclides from groundwater. The reactive metal media works by inducing conditions that cause substitution of hydrogen for chlorine in the chlorinated VOCs. The end-products of the process are completely dehalogenated hydrocarbons and non-toxic salts. Examples of end-products of chlorinated VOCs degraded by this process are ethene, ethane, and chloride ions. Radionuclides are removed by undergoing a reduction and/or absorption process.
The treatment system was designed based on the results of laboratory treatability studies conducted by Envirometal Technologies, Inc. (ETI), the patent holder for the reactive iron filings technology, and by Sandia National Laboratories (Sandia) for radionuclide removal. ETI's and Sandia's recommendations on the volume of reactive media and retention times required to meet the surface water action levels were incorporated into the final design of the treatment system.
Sandia tested the ability of media to remove the metals and radionuclides found in Mound Site seep water by performing column test using a surrogate water sample. Their results show removal of metals and radionuclides in approximately 12 minutes.
For their laboratory treatability study, ETI used uncontaminated groundwater from RFETS and spiked it to the maximum contaminants levels expected for the three RFETS VOC plumes. Initial concentrations used in the column testing and concentrations in the treated effluent are shown in Table III.
All VOCs, with the exception of methylene chloride (dichloromethane) were removed to below surface water action levels. The concentrations of methylene chloride in the Mound Site Plume (Table III) are already not detectable or low level, and surface water action levels would be met.
Table III. Results of ETI Bench Scale Testing - Connelly Iron
Compound | Influent Conc. |
Effluent Conc. |
Surface Water Action Level (ug/L) |
Carbon Tetrachloride | 1,004 |
nd |
5 |
Trichloromethane | 110 |
nd |
8 |
Dichloromethane | 111 |
105 |
5 |
Tetrachloroethene | 5,496 |
nd |
5 |
Trichloroethene | 5,250 |
nd |
5 |
Cis-1,2-dichloroethene | 64 |
nd |
70 |
1,1-Dichloroethene | 318 |
nd |
7 |
Vinyl Chloride | 102 |
nd |
2 |
1,1,1-Trichloroethane | 37 |
nd |
200 |
nd = non detect
CONCLUSIONS
Treatability studies have shown that reactive (zero valent) iron will treat contaminated groundwater in the three RFETS VOC plumes to regulatory action levels. After the Mound Site Plume's containment and treatment system is installed, the EPA's SITE Program will monitor the system for up to one year to obtain operating data. The operating data will be used by the Site to refine the design and installation of similar containment and treatment systems for two other VOC plumes.
REFERENCES
Please contact author for copies of the figures.
Mr. John Hopkins
V-(303)966-4974
F-(303)966-5198
Fig. 1. Mound Site Plume Area
Fig. 2. SW059 Site Area Water Tape Elevation April 30,-May 8,
1997 with Water Annotations
Fig. 3. Mound Groundwater Plume
Fig. 4. Mound Groundwater Plume