In Situ Enhanced Source Removal

2.0    Site Characterization

2.1    Historical Data

2.1.1    Location

Hill AFB is located in northern Utah, approximately 25 miles north of Salt Lake City and five miles south of Ogden as shown in Figure 2.1.1.1.


Figure 2.1.1.1 Location map

The Base occupies approximately 6,700 acres in Davis and Weber Counties. The western boundary of Hill AFB is Interstate 15 and the southern boundary is State Route 193. The northern and eastern perimeters are bounded by the privately-owned Davis-Weber irrigation canal.

2.1.2    Description and History

As shown in Figure 2.1.1.1, Operable Unit 1 (OU 1) shown in Figure 2.1.2.1 is located near the northeastern boundary of Hill AFB and consists of the following disposal sites:


Figure 2.1.2.1 disposal sites at OU1
 

Chemical Disposal Pits (CDPs) 1 and 2: An industrial liquid waste disposal site in operation from 1952 through 1973 (Montgomery, 1992). Wastes (principally petroleum hydrocarbons and spent solvents) were burned periodically at these sites. These are the two yellow regions inside the magenta region of Figure 2.1.2.1.

Landfill 3 (LF 3): An industrial (dump and burn) liquid and solid waste disposal site in operation from 1940 through 1978. Materials disposed at LF 3 included unidentified chemicals, industrial sludge, waste solvents, and residues from solvent cleaning operations.

Landfill 4 (LF 4): A trench-and-fill sanitary refuse landfill in operation from 1967 through 1973. LF 4 received domestic and construction refuse, and industrial waste consisting of drying-bed sludge, sulfuric acid, chromic acid, and methyl ethyl ketone.

Fire Training Area 1 (FTA 1): Used by Hill AFB until 1973 as a practice area to extinguish simulated aircraft fires. Fuels burned at FTA 1 include jet fuel, oil, and combustible waste chemicals.

Fire Training Area 2 (FTA 2): Used by Hill AFB from 1973 until the present as a practice area to extinguish simulated aircraft fires. Only on-specification jet fuel or, more recently, propane, have been used in the fire training exercises at FTA 2.

Waste Phenol/Oil Pit (WPOP): A brick-lined pit used periodically from 1954 through 1965 to dispose and burn waste oil and phenol. During excavation of Pond 10 at OU 1, soil at the WPOP that appeared contaminated was excavated and placed under the cap that covers LF 3.

Waste Oil Storage Tank (WOST) Site: Four 25,000-gallon, above-ground storage tanks used to store fuel oil, jet fuel, and hydraulic oil; removed in 1985.

2.2    Geologic and Hydrogeologic Setting

2.2.1    Topography

Hill AFB is on a terrace of the Weber Delta approximately 300 feet above the Weber River Valley floor. The valley runs east-west immediately north of Hill AFB. The delta surface slopes to the west and has slight to moderate relief with elevations varying from approximately 5,045 feet relative to the national geodetic vertical datum of 1929 (NGVD) along the eastern boundary of Hill AFB to 4,600 feet along the western boundary. The elevation at OU 1 is approximately 4,800 feet. The escarpment at OU 1 has moderate to steep slopes toward the Weber River Valley. About four miles to the east of Hill AFB, the Wasatch Mountains rise abruptly from the valley floor to an elevation of over 9,500 feet. The Great Salt Lake lies approximately 12 miles west of Hill AFB and its current (1995) elevation is 4,200 feet.

2.2.2    Geomorphology and Geology

The subsurface stratigraphy at OU 1 can be characterized as follows: (1) Upper Sand and Gravel Unit: An upper unit consisting of fine to coarse, clean to silty sand interbedded with gravel that ranges in thickness from zero to 62 feet and averages 30 feet. The lower 3 feet of this unit generally are water bearing, although the saturated thickness ranges from zero to 10 feet. (2) Silty Clay Unit: An intermediate unit primarily consisting of silty clay interbedded with fat clay and silt, and containing thin stringers (tenths of inches to a few inches thick) of very fine sand. The unit is approximately 200 feet thick and appears to be saturated with intermittent saturated sand stringers from its top to as deep as it has been penetrated by drilling (approximately 150 to 200 feet into the clay). (3) Deeper Sand Unit: A possible deeper unit of unknown thickness consisting of clean sand with occasional stringers and interbeds of silty to fat clay.

2.2.3    Hydraulic Characteristics of the Shallow Aquifer

The potentiometric surface of the shallow, unconfined on-Base aquifer slopes to the north-northwest toward the Weber River Valley, and appears to have remained relatively constant during several years of monitoring. The apparent horizontal gradient is relatively flat on the OU 1 bench and in the Weber River Valley, and relatively steep on the escarpment.

The saturated portion of the upper sand and gravel unit on Base has a horizontal hydraulic conductivity ranging from 10-1 to 10-2 cm/sec based on aquifer test data, and from 10-2 to 10-5 cm/sec based on slug test data. The horizontal hydraulic conductivity of the underlying clay unit ranges from 10-4 to 10-5 cm/sec based on slug test data. The calculated average linear horizontal velocity of on-Base ground-water ranges from 310 to 4,656 feet per year (average of approximately 1950 linear feet per year) in the upper sand and gravel unit, and from 0.4 to 50 feet per year (average of approximately 10 feet per year) in the clay unit. The vertical hydraulic gradient of the shallow on-Base aquifer in November/December 1994 ranged from 0.05 foot/foot (ft/ft) upward to 0.9 ft/ft downward with an average vertical hydraulic gradient of 0.25 ft/ft downward.

The vertical hydraulic conductivity of the clay unit ranges from 10-5 to 10-8 cm/sec and averages less than 10-7 cm/sec based on constant-head permeability testing of core samples collected from the unit. Average linear vertical velocities in the clay range from 0.014 to 0.240 feet per year based on the vertical hydraulic conductivities, vertical gradients, and an assumed porosity for the clay of 50 percent (Freeze and Cherry, 1979). These results imply that, although the shallow on-Base aquifer has a downward vertical hydraulic gradient into the clay unit, the primary component of shallow ground-water flow is horizontal because it is easier for the ground water to flow horizontally in the sand and gravel unit than vertically into the silty clay unit.

2.3    Contaminant Composition and Distribution

The following paragraphs summarize the current understanding of the distribution of contaminants at OU 1. Based on the available data, the current principal contaminant sources at OU 1 are the contaminated soils in the vadose zone underlying the disposal sites, the residual- and free-phase LNAPL in the capillary fringe, and the contaminated aquifer matrix soils at and downgradient of CDPs 1 and 2, FTA 1, and the western part of LF 3. The contaminants include VOCs, BNAEs, PCBs, dioxins, and furans. The following paragraphs discuss contaminant types and distribution in the various media at OU 1, and focus on the area adjacent to CDPs 1 and 2 where this study is performed.

2.3.1    LNAPL and Soil Contamination

Based on the history of the OU 1 disposal sites (Montgomery, 1992, 1993), bulk quantities of liquid wastes were disposed and periodically burned at the CDPs.  The waste mixture combined with the burning activities created a very complex NAPL.  Some of the components in this mixture appear to have coated the soil particles changing the soil from a hydrophilic material to a hydrophobic material. This apparent transformation makes it more difficult to extract the contaminants of interest.  During previous drilling operations, a distinct "black or oily" layer with a strong hydrocarbon odor was observed in most boreholes associated with the CDPs. The layer was encountered just above and/or below the water table, at elevations ranging from approximately 4,768 to 4,780 feet NGVD (i.e., 9 to 31 feet below ground surface (bgs). The thickness of the layer ranged from approximately 1 to 10 feet, and the layer generally ended at the contact between the sand and gravel unit and the silty clay unit. Stained soil was observed in the silty clay unit in several boreholes, but the staining extended less than two feet into this unit. The variation in thickness of the stained layer probably results from variability in soil types, proximity to source, frequency with which contaminants were dumped into the soil, and smearing caused by seasonal fluctuations of the water table. The ground-water elevation at the CDPs fluctuates from approximately 19 to 24 ft bgs. Hydrocarbon staining was observed in sand stringers in the silty clay unit down to an elevation of approximately 4,762 feet (25 to 37 feet bgs) in boreholes associated with the CDPs (U1-129 through U1-133A, and U1-732).

In several soil borings associated with the CDPs, a distinct fuel layer was observed above the black layer. The layer consisted of soil visibly coated with a substance that had a strong fuel odor. This substance caused a sheen on the soil particles, but did not discolor or stain the soil. The fuel was observed in the sand and gravel unit at elevations ranging from approximately 4,775 to 4,785 feet (12 to 27 feet bgs). The apparent thickness of the fuel layer varied from approximately 3 to 10 feet. This fuel layer probably represents LNAPL saturation in the soil above the water table.

Based on laboratory analyses (Montgomery, 1992, 1993), the LNAPL is composed primarily of jet fuel; however, other contaminants that have been disposed at OU 1 have solubilized and accumulated in the LNAPL due to co-solvency. VOCs, BNAEs, pesticides, PCBs, dioxins, and furans have been detected in the LNAPL associated with CDPs 1 and 2. Compounds detected in the LNAPL and their concentrations are summarized in Table 2.3.1.1.

Recently (HAFB, 1998), Hill Air Force Base has concluded, based on risk factors, areas where it considers the soils to be contaminated and areas where it considers the soils to present an environmental risk.  These areas are delineated in the  soil contamination map .  The southern point of the area indicated by WP002 is centered between the test cells used for surfactant mobilization and macromolecular solubilization.  The center of the northern point of the area indicated by WP002 is slightly south of the test cell used for cosolvent solubilization.  Contamination levels in surfactant mobilization and macromolecular solubilization were significantly higher than the remainder of the test cells in study presented in this document.

In the current study, core samples were analyzed for decane as well as other compounds frequently found at OU1.  Geostatistics, discussed later, were used to map the distribution of decane in the study area. The results of this analysis are shown based only on the samples collected from within the test cells.  The decane distribution shows the site contamination is heterogeneous for this portion of OU1.

2.3.2    Ground Water

During Phases I and II of the Remedial Investigations (Montgomery 1992, 1993) and the quarterly monitoring program, 27 wells in the CDP area on Base and on the hillside downgradient of the CDPs were sampled. Table 2.3.2.1 summarizes the concentration ranges for analytes detected in samples from the wells. Generally, the types of compounds identified in the ground water at the CDPs are consistent with the history of disposal activities in the area. The majority of the wastes disposed in this area were spent chemical solvents and petroleum hydrocarbons.

Chlorinated hydrocarbons are the most common contaminants detected in the ground water at the CDPs. The VOC detected at the highest concentration and with the widest distribution is 1,2-DCE. Chlorinated benzene compounds were detected over a wide area and in relatively high concentrations, although their distribution is much more limited than that of 1,2-DCE. The chlorinated benzene compounds detected include chlorobenzene, 1,4-dichlorobenzene, 1,2-dichlorobenzene, and 1,2,4-trichlorobenzene.  Hill AFB has prepared a map, in their Environmental Restoration Management Action Plan May 1998, of the dichloroethane contamination  OU1.  The areas labeled WP002 are the same as the Chemical Disposal Pits shown in yellow surrounded in magenta in Figure 2.1.2.1.

2.4 Characterization Methods

2.4.1    CPT

Cone Penetrometer Technology (CPT) was used extensively at Hill AFB to install sampling and monitoring equipment in the pilot treatment systems.  The CPT has advantages in being able to place instruments in the ground with minimal disturbance and without producing waste cuttings.  At Hill AFB, a 1.75" O.D. rod with a 1" I.D. was used to install multilevel samplers, oxygen sensors, thermocouples, and TPH sensors.  The formation at Hill was not conducive to collecting core samples with the CPT due to the large cobbles which commonly plugged the core barrel making the core recovery unacceptable.  In addition to instrument installation,  the CPT was used to collect more traditional geotechnical measurements, including: tip and sleeve resistance, pore pressure, and resistivity.  These measurements were used to locate the clay interface, define the stratigraphy, and map the NAPL distribution.

2.4.2    LIF

Laser Induced Fluorescence (LIF) uses a light source on the surface connected to a fiber optic cable to transmit light down into the profile through a sapphire window into the soil.  The light excites organic compounds having a ring structure, causing them to emit light at a higher wavelength (fluoresce).   Thus, a fluorescence response is an indicator of the presence of these compounds and can be used to map the vertical and areal extent of contamination.

2.4.3    Cores

Soil cores are effective methods for characterizing contaminants in a particular soil sample. A soil sample typically represents only a very small portion of the total potential volume of the subsurface under investigation. Therefore, a large number of soil cores are necessary to develop statistical confidence in the total mass or the spatial distribution of a particular contaminant in the system. In addition, for sites where the contamination is a complex mixture of chemicals like the NAPL at OU 1, it is extremely difficult, if not impossible, to identify all of the individual constituents.  Quantifying the total mass of the contaminants in the mixture is even more difficult.

Transects of soil cores were taken in each of the test cells before and after the remedial activities. These cores were extracted in the field and then analyzed in the laboratory for selected target analytes. Results were reported on a mass/mass soil concentration. Target analytes shown in Table 2.4.3.1 were selected as major components of different classes of chemicals which could be analyzed. Approximately eight cores were taken from each test cell. Soil samples were taken at approximately 30 to 60 cm vertical intervals in the treatment zone or when there was an obvious change in the stratigraphy or level of contamination. Samples were taken throughout the contaminated portion of the profile based on either visual detection of the contamination or PID readings above background. Core samples were collected with a hollow stem auger (HSA).


Figure 2.4.3.1  NAPL thickness estimated from core and LIF data.

Combining the LIF data with the core data, it is possible to generate a map of the approximate thickness of the NAPL (Fig 2.4.3.1).   In general, the thickness of the NAPL-contaminated zone increases from north to south over the study site.  Additional data to better define the southern limits of the NAPL smear zone would be needed to generate a defensible estimate of the total NAPL volume.  During our investigation, we did not have permission to collect samples outside the locations presented here. Combining core and LIF data to estimate contaminant mass for this site is problematic.  For core samples, NAPL mass was inferred from the selected target analytes; whereas, NAPL mass is based on the presence of  fluorescing compounds for the LIF data.  Data presented here suggest inconsistencies between the core and LIF data.

2.4.4    Ground-water Sampling

Cleanup standards are often based on potential exposure by ingestion of drinking water. Thus, knowing the activity of regulated compounds in the ground water before and after remediation gives a good indication as to the immediate hazard through the ingestion route but gives no indication as to the mass of water which might be impacted at this level by dissolution from the source. Each of the test cells had twelve multilevel samplers and each of the samplers had five or more sampling points. The multilevel samplers were vertically positioned within the zone targeted for remediation. In addition to the multilevel samplers, samples were collected from seven injection/extraction wells which were screened throughout the target zone. These sampling points were analyzed for target constituents before and after remediation to evaluate where and how much improvement in the solution concentration was observed.

2.4.5    Ground-penetrating Radar

Two surveys were performed using ground-penetrating radar (GPR) to assist in determining the depth to clay.  The primary purpose was to assist in the design of the test cells which were constructed of sheet piles driven into the clay to hydraulically isolate them from the surrounding aquifer.  The first survey was performed around cell f  (cosolvent solubilization) which was the first cell constructed.  This survey was very successful in delineating the depth to clay(Young and Sun, 1996).  The survey was calibrated with several CPT determined clay interface depths.  It was decided to move the remainder of the study south of the first study.  Thus, a second survey (Young and Sun, 1998) was performed in a fenced area at the site of a former lined evaporation pond and the site of one of the chemical waste pits see Figure 2.1.2.1 .  There were numerous interferences with man-made objects in the second survey.  The results of the second survey which were performed prior to the installation of the test cells are shown in Figure 2.4.5.1.  The actual cell installation was completed prior to the completion of the radar data analysis.  For a more complete description of the GPR survey, see Young and Sun, 1996, 1998.