1. Development of the Site Conceptual Model

The conceptual model for the site is based upon information about the source areas and hy-drogeology at the site presented in the final remedial investigation report. The regional geologic setting for the site is the Quaternary Gulf Coastal Plain of Texas. This region comprises a series of sedimentary depositional plains, the youngest of which is of recent, postglacial deposition (Holocene deposits). Sediments of the Holocene are deposited along the coast and in the alluvial flood plains of existing river systems. The site is located specifically in the surface sediments of the Beaumont Formation.

The average depth of the Beaumont Formation is 0-20 ft below grade. The Lissie Formation lies below the Beaumont and extends to approximately 200 ft below the surface. The majority of the historical site soil borings were drilled to a depth of 65 ft, although deeper borings

Significant Water Bearing Zone
Figure 7 Generalized cross section of the site.

were drilled to depths of 200 ft below the surface. Sediments making up these units consist of a top stratum of cohesive soils (sandy clays and silty clays) and a substratum of cohesionless soils (silty sands, clayey sands, and poorly graded sands). Primary water-bearing zones and their corresponding thickness encountered at the site are organized as follows: shallow zone (10-21 ft), intermediate zone (115-127 ft), and deep zone (174-200 ft). A generalized cross section of these zones is presented in Figure 7.

The 1986 RI report states that the shallow water-bearing units consist predominantly of silty and clayey sands with occasional gradations to sand and clayey silts. The base of this unit is irregular and slopes gently to the east. Shallow zone groundwater trends at the site generally slope to the west at a gradient averaging about 20 ft/mi. Site surveys indicated that the shallow zone extends continuously offsite to the west, toward a bayou. This drainage feature may act as a groundwater discharge area, thereby influencing westward-trending gradients. A surface impoundment located offsite along the east property line may cause localized groundwater mounding, which also would produce westward gradients in the shallow zone.

The hydraulic conductivity of the shallow zone was evaluated by conducting falling-head field permeability tests at selected well locations. Table 26 presents the field hydraulic conductivity test results for wells screened in the shallow water-bearing zone. Soils of the waterbearing unit are also typed for future consideration. Measured field horizontal hydraulic conductivities for the shallow zone ranged from 2.4 x 10~4 to 2.2 X 10~3 cm/sec and averaged 8.3 x 10~4 cm/sec.

The next significant water-bearing unit is the intermediate water-bearing zone, which is overlain by a clay layer and is not considered to be affected by historical site activities. Hydraulic conductivities for this zone are not currently available. The continuous nature of the intermediate aquitard excludes this zone from the modeling effort.

The deep zone is also considered to be a confined water-bearing unit. However, historical data indicate that a deep well was screened in this unit, and records of the well's abandonment are not available. Physical data are not available for this particular zone. It is believed that the deep zone gradient slopes to the south-southwest at approximately 6 ft/mi.

Vertical seepage rates between shallow and intermediate water-bearing units are extremely low because of the nature of the confining strata. However, a potential conduit between the shallow zone and the deep may exist in the form of a leaking improperly abandoned well in the southern portion of the site. For this reason, the deep water-bearing unit is a concern of EPA V and is considered in the modeling effort.

Given the nature of the site and the characteristics of the underlying strata, contamination of the shallow zone by infiltration of precipitation is the most obvious groundwater contami-

Table 26 Field Hydraulic Conductivity Test Results—Shallow Groundwater Zone

Field horizontal








interval (ft)





2.4 x 10~4


Silty sand


4.5 X 10~4




4.5 x 10~4




2.2 x 10~3

Avg. 8.3 x UT4

nation mechanism. Contamination reaching the shallow zone would in all likelihood move in the westward direction of shallow zone groundwater migration. Vertical and horizontal dispersion would then occur, with vertical movement bounded below by the intermediate aquitard. If the improperly abandoned well exists, contaminants might also enter the deep water-bearing zone in which the well was screened. Groundwater in the deep water-bearing zone in which the well was screened. Groundwater in the deep water-bearing zone trends toward the southern property boundary.

The site conceptual model consists of three elements: (1) leaching of constituents of interest (COIs) from affected soils, (2) migration and subsequent dispersion of the COIs with respect to shallow groundwater flow direction, and (3) migration from the shallow zone to the deep water-bearing unit through the open conduit and subsequent dispersion. The site conceptual model is presented in Figure 8. An important aspect of this conceptual model is that it is chemically conservative, assuming that the cPAHs are not biodegraded or adsorbed to soils before, during, or after migration.

2. Constituent Mobility

Polynuclear Aromatic Hydrocarbons (PAHs), in general, are immobile constituents in environmental media. Recent research shows that PAHs bind to soil surfaces as a result of their van der Waals forces [15]. Van der Waals forces act solely between molecules within close proximity of each other. As a rule, the larger the molecular size, the greater the van der Waals forces. cPAHs, being generally larger than noncarcinogenic nPAHs, are even less mobile in environmental media. Similarly, relatively low water solubilities and vapor pressures add to the inherent environmental immobility of cPAHs.

A semiquantitative assessment of theoretical mobility developed by Laskowski et al. [16] can be used to describe the immobility of cPAHs in the environment based on known physical and chemical constants. The basis for the assessment is an algorithm that utilizes water solubility (5), vapor pressure (VP), and the organic carbon partition coefficient (K<*.) of the constituents to determine a relative mobility index (MI). The MI is defined as

A relative scale is then used to evaluate the MI derived for each cPAH [17]. The scale is a descriptive one, comparing a numerical MI to the categories extremely mobile, very mobile, slightly mobile, immobile, and very immobile.

Mobility index



Extremely mobile

0 to 5

Very mobile

-5 to 0

Slightly mobile

-10 to -5



Very immobile

Mobility index values and the physical-chemical constants for the cPAH are presented in Table 27. The Mis for the cPAHs range from values of -8.9 (immobile) for benzo[fe]fluoran-thene to -11.7 (very immobile) for chrysene.

PAH mobility may be enhanced if these constituents bind onto dissolved organic macro-molecules such as humic acids or suspended particulate organic matter in groundwaters. It is important to note that in most soil-water systems, these macromolecules are, themselves, not


Figure 8 Conceptual model of potential containment migration. I, Organic Leachate Model (OLM); II, Vertical and Horizontal Spreading (VHS) model; III, two-dimensional horizontal line source model (continuous solute line source).


Figure 8 Conceptual model of potential containment migration. I, Organic Leachate Model (OLM); II, Vertical and Horizontal Spreading (VHS) model; III, two-dimensional horizontal line source model (continuous solute line source).

Table 27 Physical-Chemical Constants and Relative Mobility Indices for cPAHs




Mol wt.






(ppm @ 25°C)

(mm Hg)

'«g Koc





5.6 x 10~9 (25°C)





0.0057 (20°C)

2.2 x 10~8 (20°C)






5 x HT7 (20°C)






5 x 10~7 (20°C)






6.3 x 10"9 (25°C)



Dibenzo[a ,/i]anthracene



7 x 10~8



Indeno[l ,2,3-a/]pyrene



7 x KT8



Source: U.S. EPA Aquatic Fate Process Data for Organic Priority Pollutants [11],

Source: U.S. EPA Aquatic Fate Process Data for Organic Priority Pollutants [11], mobile. Nevertheless, this site conceptual model considers that the cPAHs could potentially migrate advectively, bound to mobile dissolved organic matter in the underlying shallow aquifer and the deeper zones. Retardation will not be considered in order to account for the maximum cPAH mobility in groundwater at the site.

3. Governing Equations

Element 1. Contaminant Leaching. The potential concentration of an organic constituent that will leach from source area soils [C(Z>] is estimated using the 95% confidence version of EPA's OLM. The OLM is a multiple regression equation that is derived from a database of measured leachate concentrations. The database is composed of TCLP data, EP-Tox data, and field lysimeter results for many of EPA's Target Compound List (TCL) compounds. The log form of the equation used to calculate the 95% percent CI is

95% CI = io.o5.^-2 • MSE • [X'h lX'X)~1 Xh] (11)

where /0 05 is the factor for the 95% CI, r) — 2 is the number of degrees of freedom, MSE is mean square error, and X, X', Xh, Xh' are various versions of the waste and leachate concentration data matrix.

The OLM equation is

Cw is the estimated concentration of an organic constituent in source area soils, and Sw is the water solubility of the constituent. Since benzo[fc]fluoranthene is the most relatively mobile cPAH as a result of its water solubility, the modeling effort will focus on this constituent.

Element ¡1. Dispersion in the Shallow Aquifer. The transport of a constituent by shallow groundwater to the APC results in dilution by vertical and horizontal dispersion processes. The VHS model [13] provides a conservative estimate of the dilution provided by these dispersive transport phenomena and is used as the final element in the site conceptual model. The VHS model equation is

CAPC = C0[erf [<r/4>f 5] erf {X/[4(a,y)° 5]}] (13)

where cAPC is the estimated constituent concentration at the APC, c„ is equal to C(/) developed by OLM since dilution prior to transport is not a consideration, Y' is the width of the source area parallel to the direction of groundwater flow, Y is the distance from the source area to the APC, a, is the transverse dispersivity, and x is the length of the source area perpendicular to groundwater flow.

The error function of some number z, erf(z), can be estimated using an approximation method. The equation is an integral part of the VHS algorithm and can be expressed as erf(z) =


The calculated error function can be checked against published error function values such as those presented in EPA's Water Quality Assessment Document [18] to ensure its legitimacy. The VHS model is based on the dispersion-convection equation

(d2C d2C

In essence, a continuous source-contaminated parcel moving at a steady one-dimensional velocity is subject to a transverse spreading process. D, is the transverse dispersion coefficient, y is the spatial coordinate collinear with the velocity of the contaminant V , and x and z represent the horizontal and vertical spatial coordinates perpendicular to groundwater flow direction. The problem is viewed as a two-dimensional semiinfinite medium bounded at the top by z = 0, the flux boundary bC/bZ = 0.

where Z and X represent the penetration depth at the waste area boundary and the width of the waste unit perpendicular to groundwater flow direction, respectively. The solution to Equation ( 14) given the boundary conditions is

This is a two-dimensional version of a solution presented by Morgenau and Murphy [20], Maximum contaminant concentrations occur at x - 0 and z — 0. Solving at jc = 0 and z = 0 gives

erf I

This solution was first provided by Domenico and Palciauskas [19] and was used in obtaining exclusions for solid wastes on a generator-specific basis. Applying for an exclusion involves a public comment period during which modification of the model is suggested. Modifications to the model were due to a change in the manner in which penetration depth (Z) is calculated. EPA agreed with the comments regarding the modification of the Z term. Since Z is calculated differently, notation in the equation changes slightly. Modification in the calculation of the Z term and notation changes account for the differences between Equations ( 13) and (18). The Z term and its importance in this modeling effort are discussed in subsequent sections of this chapter.

Element ///. Dispersion in the Deep Aquifer. A direct conduit to the deep aquifer in the form of an improperly abandoned well could exist in the south-central portion of the site. Historical records indicate the presence and approximate location of a deep well, and although U.S. Geological Services (USGS) boring logs exist, records of the well's abandonment were not located.

A number of studies were conducted in 1986 to locate the improperly abandoned well. Geophysical testing and excavation were undertaken in the area of the site described in the company records. These investigations failed to locate the deep well conduit. However, the study did reveal several magnetic anomalies that could correspond to the presence of the deep well casing. Excavation was undertaken as a result of the magnetometric survey, but the exact location of the well was never found. It is unlikely that this is a major pathway of cPAH migration at the site, and the uncertainty in the existence of the open conduit makes modeling this particular pathway difficult. If the well does exist, it could be a conduit to the deep aquifer.

The approach used in estimating the potential impact on the deep aquifer considers two-dimensional horizontal flow with continuous solute line sources [18]. The aquifer is considered to be homogeneous and isotropic. Benzo[f>]fluoranthene is assumed to discharge continuously and uniformly within the deep aquifer. Steady-state conditions are assumed, to maintain consistency with other portions of the site conceptual model. The equation is where C„ is the acceptable concentration of COIs emanating from a source area that is protective of the shallow aquifer (p.g/L); Q is the flow emanating from the shallow aquifer as a result of the hypothetical conduit (m3/day); b is the saturated thickness of the deep aquifers (m); p is the porosity of the deeper aquifers (dimensionless); Dx, Dy are the dispersion coefficients in the x and y directions, respectively (m2/day); B = 2DJVd; Vd is the Darcy velocity (m/day); P is an estimate of retardation (dimensionless); Vx is the seepage velocity (Darcy velocity) of regional flow in the x direction (m/day); x the distance from the source area to the APC (m); y the distance between the centerline of the plume and the APC (m); and

4. Site-Specific Modeling Inputs

The site conceptual model comprises three elements described by analytical solutions predictive of an approximate groundwater concentration at an APC resulting from a soil concentration of cPAHs at a source area.

Given the performance standard (CAPC), the VHS model is solved for each source area capable of producing an impact on the shallow zone groundwater quality. The VHS reduces to where CAPC is the performance standard in the shallow aquifer (10 (Xg/L); and DF„ and DFW are the dilution factors derived from vertical and horizontal dispersion, respectively.

The equation is then rearranged and solved for C0, which is derived for each source area of interest at the site. Then C0 is entered into the rearranged OLM equation, which becomes

Cw, the concentration of the constituent in the soil, becomes the soil cleanup goal protective of the shallow aquifer at the APC and meets cAPC.

The performance standard associated with the established soil leaching criteria is then assessed with respect to the deep aquifer. If the groundwater quality in the deep unit is unaffected at the APC, the performance standard established for the shallow zone is also protective of the deep aquifers and no modification to the soil leaching criteria is necessary. If the groundwater quality at the deeper aquifers' APC is affected, the no leaching potential criteria for the shallow zone may be modified to also be protective of the deep aquifer. The performance standard for the deep aquifer is assumed to be the practical quantitation limit (PQL) of any individual cPAH.

The two source areas located in the southern portion of the study area are the most influential source areas (with respect to the shallow aquifer) by virtue of their size and their proximity to the APC. The western property boundary was chosen as the APC for the shallow aquifer because it is hydrogeologically downgradient of both potential source areas. The southern property boundary was chosen as the APC for the deep aquifer for the same reason. The source areas in the northern portion of the site seem relatively small and too far from either APC to produce a significant impact on groundwater quality. Since it is possible that cPAHs emanating from the northern source areas are additive, the two small areas are combined and considered to be a single source of cPAHs to simplify the modeling effort. Combined, the northern source areas may have some impact at the APC.

As discussed previously, the most conservative approach to modeling a mixture of cPAHs is to assume that the total cPAH soil value is attributable to the occurrence of the single most water-soluble cPAH and not a mixture of the seven constituents (e.g., the concentration of benzo[6]fluoranthene (Cw) is 700 mg/kg, and no other cPAHs are detected). This is an extreme worst-case scenario, since the concentration of cPAHs in groundwater at the APC is directly proportional to the constituent soil concentration at the source area of interest and cPAHs typically occur as a mixture at the site as a result of past practices, not just as individual components.

Figure 9 presents the source areas, the location of the hypothetical abandoned well, and the locations of the shallow zone and deep zone APCs used in the conceptual modeling effort. The source areas were delineated during the 1986 site remedial investigation as those soils having visible staining between the 0.5-ft and 6.0-ft depths, an indication of cPAH contamination. The 0.5-6.0-ft interval was selected because of its proximity to the shallow zone and its potential to affect groundwater quality in that zone. Furthermore, the location of the hypothetical improperly abandoned deep well is approximate, as excavation failed to identify its exact location.

The remainder of this section discusses the numerical inputs to the algorithms and the subsequent groundwater results at the APC as a function of constituent concentrations in soils at the source areas.

5. Shallow Aquifer

Three general areas were evaluated to determine soil cPAH concentrations that are protective of the shallow aquifer at the APC. These are the southeastern, southwestern, and northern areas. Each area its corresponding modeling inputs are discussed below.

Regional shallow groundwater trends indicate that groundwater beneath the site flows from east to west, perpendicular to the southwestern source area. Estimated source width (Y'), source length (X), and distance to the APC (10 are 25, 275, and 230 m, respectively. Only unpaved areas were considered in calculating the source area dimensions, because pavement would prevent infiltration of precipitation and subsequent leaching of cPAHs.

Figure 9 Approximate location of source areas and points of compliance. Hatching indicates areas in which visual staining was observed in I subsurface soils. _

The source located in the southwestern portion of the site is closer to the APC than those in the southeastern areas and should therefore have a greater effect on groundwater quality at the APC. Input estimates for the southwestern source area include width (i" = 45.7 m), length (X = 69 m), and distance to the APC (Y = 47 m). The central area of the source is paved and is not considered part of the overall source area since migration of the cPAHs by leaching from soils beneath the paved area will be negligible without the infiltration of precipitation.

Groundwater in the northern portion of the site also flows from east to west perpendicular to the two source areas. The two source areas are combined to account for the potential co-contribution of cPAHs to the shallow zone and simplify the estimation of a soil cleanup goal. The source inputs for width (K'), length (X), and distance to the point of compliance {Y) are 111, 78, and 91 m, respectively.

6. Deep Aquifer

To determine the potential impact of shallow zone groundwater concentrations on the deeper aquifer, a number of assumptions must be made. For the sake of conservatism, it is assumed that the improperly abandoned well exists as a conduit to the deep water-bearing zone. Impact on the deep aquifer is influenced by the integrity of the abandoned well and the physical characteristics of the aquifers. Little site-specific information pertaining to the deep well conduit or the deep water-bearing zones is available; therefore, hydraulic conductivity, horizontal gradient, porosity, and aquifer thickness must be estimated from the available literature.

The 530-ft water-bearing unit corresponds to the base of the Chicot Aquifer in the Alta Loma Sands. This layer is approximately 200-400 ft thick in the vicinity of the site and is fairly homogeneous, being 60-80% sands. Hydraulic conductivities of 10~5 cm/sec and a porosity of 35% are considered to be typical of the formation [21].

The estimated Darcy velocity for both the deep zones used in the modeling effort is therefore 1 x 10~5 cm/day.

Steady-state flow from the hypothetical conduit is a function of the integrity of the well and the availability of groundwater present in the shallow aquifer. A characteristic of the shallow aquifer at the site is its relatively low yield. Yields of 1-1.5 gal/min (gpm) are common. Pump tests conducted in 1988 used a pumping rate of 1.25 gpm, and significant drawdown was observed in each of the observed wells. If shallow zone groundwater is flowing to the deeper zones through the conduit, a trend toward the conduit would be apparent in the shallow zone groundwater contours. Shallow zone groundwater contours presented in the RI report do not indicate the presence of a significant trend associated with an open conduit losing 1.25 gpm to deeper aquifers. Therefore, a conservative estimate of flow rate (Q) equal to one-tenth of the 1.25 gpm rate (0.125 gpm) is used as the maximum potential flow from the shallow zone through the hypothetical conduit.

The exact location of the hypothetical conduit is also unknown. U.S. Geological Survey records and magnetometric surveys indicate that the general location is believed to be in the south-central portion of the site, approximately 55 m north of the southern property boundary.

0 0

Post a comment