SECTION
1: INTRODUCTION
The Greater Fallston Association, Inc. extends its thanks to the numerous local and national environmental consultants, hydrogeologists, academicians as well as community experts for donating their time and expertise in analyzing the data presented in the ExxonMobil Site Assessment Reports of October 8, 2004 (SAROct2004) and March 22, 2005 (SARMar2005).
The Maryland Department of the Environment (MDE) is charged
with protecting the health, safety and welfare of the citizens of
In response to the historical elevations of methyl-tertiary-butyl-ether (MTBE) at the ExxonMobil station dating back to 1990 and the occurrence of a significant release of MTBE in 2003 extending into 2004, the MDE charged ExxonMobil to conduct an investigation to determine the extent of contamination from the ExxonMobil station site. This included a comprehensive assessment of the underlying geology; the determination of groundwater, depth, flow rate, direction and storage capacity in the surrounding aquifer; and the measurement of MTBE levels and other volatile organic chemicals (VOCs) in and around the ExxonMobil site, including the surrounding community.
From a scientific standpoint, the conclusions of ExxonMobil’s report are suspect, because the design of the study was built around a single hypothesis: that groundwater in the bedrock aquifer only flows to the southwest at an infinitesimally slow rate. With this limited perspective, ExxonMobil restricted its investigation to substantiate their foregone conclusion.
As will be detailed in this Comprehensive Review, the design of ExxonMobil’s study was flawed from the very beginning, which lead to an inaccurate site conceptual model and, consequently, broad and inaccurate assumptions regarding the transmissivity and storage capacity of the regional aquifer. At best, the conclusions from their investigation are “well specific”, not “area-wide specific”.
Upon extensive review of the data provided by ExxonMobil, as well as additional evidence not presented in ExxonMobil’s report, we present a strong argument demonstrating that the ExxonMobil site at Uppercrossroads, Maryland is the epicenter of a massive MTBE vapor release that has migrated into the regional aquifer and contaminated the private supply wells in the current study area and beyond.
SECTION 2: THE SOIL
1. General
As will be detailed below, review of the boring logs of monitoring wells (MWs) 1 through 27 (SAROct2004, Appendix M) and MWs 28 through 33 (SARMar2005, Appendix F) revealed the following inconsistencies in the soil descriptions of MWs 1 through 4 drilled in 2003 and MWs 5 through 33 drilled in 2004.
First, the boring logs in 2004 do not even indicate the presence of fill that would be present in a developed area. This is a minor notation, but important to establish the accuracy of the soil descriptions to follow.
Second, saprolite, commonly referred to in the field as “rotten rock”, is weathered bedrock in which fine fabrics are retained. At times, the consultant geologists seem to confuse the natural soil for saprolite or fail to recognize the boundary between fill, soil and true saprolite (assumed to be the “schist”) in their field descriptions. These inconsistencies seriously call in to question the accuracy of the soil descriptions and the validity of any conclusions drawn from these descriptions.
Third, the geologist who logged most of the bedrock monitoring wells did not provide the Unified Soil Classification System designations for the unconsolidated materials on the boring logs. This reflects either a lack of attention to detail or, worse, a bias to a particular conclusion.
2. Comparison of 2003 and 2004 Boring Logs
Monitoring wells 2 and 5 are only 25 feet apart. MW-2 was drilled in 2003, and MW-5 was drilled in 2004. The boring log for MW-2 reveals, “clayey sand”, “sand and silt” and “weathered bedrock (saprolite)” beginning at 30 feet. In contrast when MW-5 was drilled a short distance away, the soil becomes “clay”, “clayey sand” and “silty saprolitic clay” at 25 feet with schist appearing at 49 feet. (FIGURES 1-2) (Reference Appendices 1, 2, and 3 may be used throughout this review to view the locations of monitoring wells and pertinent supply wells.)
Monitoring wells 4 and 11, which are only 50 feet apart, show an even more dramatic redefinition. In 2003, the boring log for MW 4 shows “clayey sand” and “clay and silt” with schist bedrock beginning at 40 feet. Suddenly in 2004, however, the soil for MW-11 converts to pure clay and “silty saprolitic clay” beginning at 11 feet with schist bedrock at 51 feet.
In fact, as FIGURE 1 illustrates, the soil in the entire region in and around the Exxon station has
become almost pure clay and “silty saprolitic clay” in the span of a year. Even MW 33 near
Also included for comparison in FIGURE 1 are the “crude” boring logs from 1981,
1987 and 1991 at the Exxon site.
Recognizing that geologists did not record these descriptions, it still
indicates the fact the drillers noted a brown silty loam (a
ExxonMobil would have us believe that the entire property (and its surroundings) is composed of a clay overburden to support their contention that MTBE would not have migrated beyond the confines of the station, because clay retards water flow and its adsorption potential for chemicals such as MTBE is high.
3. Area Percolation
According to the soil descriptions presented by ExxonMobil (SAROct2004, Appendix M), clay is the predominate soil type in the shallow overburden at the ExxonMobil site. Clayey soils demonstrate slow percolation times of 30 minutes or more. Interestingly, the percolation tests performed on the Exxon site in 1975 and 1984 showed percolation times of 4 minutes and 5 minutes respectively, which is characteristic of a silty loam. APPENDIX E-F A fast percolation rate of 2 minutes is characteristic of sandy loams. Indeed, most of the soils in the Fallston area percolate well in the range of 2 to 4 minutes at 5 feet. These percolation times do not comport with the alleged regional presence of a confining clay.
In their report, (SAROct2004), ExxonMobil states that a “locally, … thin perched and/or water table exists just above the weathered bedrock”. To have a perched water table, there must be a confining layer below it. ExxonMobil has not defined the confining layer. The report presents no figures or cross-sections demonstrating the lateral extent or thickness of the confining layer or the perched water table. ExxonMobil’s report (SAROct2004) only presents ill-constructed cross sections that illustrate “fill” versus “silty saprolitic clay”. Stating that the perched water is just above the weathered bedrock confuses matters even further. (SARMar2005, p. 11) By definition, perched water is separated from the water table aquifer by a confining layer. If that is so, then the “silty saprolitic clay” described by ExxonMobil must be the same as “weathered bedrock.” However, ExxonMobil seems to differentiate the two as separate lithologic units.
Simply put, the report fails to establish that the overburden acts as a confining layer given the inconsistent terminology; lack of geotechnical confirmation of the percentage of clay in the overburden; and the lack of cross-sections showing the extent of the so-called confining layer.
4. Soil Vacuum Extraction
Exxon claims that the soil vacuum extraction (SVE) system has been effective in reducing the levels of MTBE as evidenced by the low levels of MTBE in subsequent soil analysis. (SARMar2005, Table 2) The report states that MTBE levels dropped from 4940 ppb in June of 2004 to non-detect (ND) in January 2005. While we are happy to see reductions of MTBE vapors in site soils, we must point out that SVE is not highly effective in clay soils, because of clay’s low permeability and chemically adsorptive properties. Thus, if the overburden at the Exxon site truly consists of clay, the SVE system would not have recovered MTBE vapor completely in less than eight months. Exxon’s conclusion about the effectiveness of SVE at the site thus contradicts its own soil descriptions.
5.
Summary
1. ExxonMobil’s describes the soil characteristics of the site as almost entirely clay and “silty saprolitic clay”. This description does not match the descriptions provided in previous borings on the site and is not confirmed by reports on well completion logs of adjacent properties.
2. Soil classifications have not been confirmed by geotechnical analysis.
3. Soil percolations on the ExxonMobil site show fast percolation not characteristic of clayey soils.
4. The documented effectiveness of SVE in removing MTBE from the soils on site is inconsistent with the alleged high clay content in the soil.
SECTION 3: MONITORING WELL DEPTH
1. General
One of the first steps in any hydrogeological investigation of a contaminated aquifer is to determine the depth of the wells in the area, particularly the contaminated ones. Boring logs from wells already drilled provide a wealth of information about the aquifer, and allow geologists to best position monitoring wells and determine potential water-bearing zones in the subsurface. As a general rule, well depths should be converted to elevation above mean sea level (amsl) rather than distance from ground surface, especially in areas with moderate to high topographic relief. Exxon, however, solely uses depth from ground surface. While ExxonMobil’s blanket drilling of bedrock monitoring wells to the same depth of 200 feet “below ground surface” (bgs) may be acceptable in certain sedimentary geologic regimes where the regional dip is negligible and topographic relief is low, it is not appropriate in an area such as Uppercrossroads where the subsurface is composed of folded and fractured metamorphic rock under a variable surface topography. Exxon’s installation of bedrock wells, all to the same depth from surface in this complex geologic formation, does not allow the adequate characterization of the aquifer.
2. Depth Determination
ExxonMobil bored 12 bedrock
monitoring wells as part of their site assessment in August and September 2004
and three additional bedrock monitoring wells in late 2004. With exception of MW 10, all were bored to
200 feet bgs. MW 10 was bored to 250
feet bgs (337 feet amsl). FIGURE 3 ExxonMobil acknowledged that the community’s
supply well data was not reviewed prior to the construction of the wells in a
meeting on
3. Community Supply Well Depth Investigation
The GFA obtained all available boring logs for wells in a one-half mile radius of the Exxon site as well as properties greater than a half-mile south of the site. For wells with no boring logs available, the GFA made a direct request to residents for the pertinent information. One hundred and forty eight (148) well boring logs were obtained. Seventy three percent (73%) of the community “well base” elevations were substantially below the “well base” elevations of the ExxonMobil monitoring wells. FIGURE 4 One hundred eight (108) of 148 community wells bottom out at elevations significantly less than 337 feet amsl (the base elevation of ExxonMobil’s deepest well). In addition to the static graphic representation, a 3-D rotational computer generated graphic compares the depth of the community “well bases” with those of the ExxonMobil monitoring wells. One static image from this rotational graphic is presented in FIGURE 5. As the graphic is rotated, it is clear that the ExxonMobil bedrock monitoring wells are shallow compared to the community supply wells. (3-D rotational graphics are available upon request.) As such, they cannot adequately assess the water-bearing zones at depth that supply a majority of the community’s wells. Moreover, these bedrock monitoring wells can provide few meaningful data regarding the relationship between shallow and deep water-bearing zones in the bedrock.
4. Well Yields and Information from Local Well
Drillers
Another good source of information
about a local aquifer are the drillers that drilled the community’s supply
wells. Consultation with different well drilling companies in the Fallston area
(Barlow 410-838-6910, Bel Air, Caswell,
Inc 410-557-9355, Jarrettsville, Jones 410-692-6981, Jarrettsville) provided
important information on the depth of water bearing zones in the
Uppercrossroads region. According to the
local well drillers, the average well depth was 280-350 feet. In Haddon Hurst,
water was not reached until 700 feet bgs; in Delmar Farms, 350-500 feet bgs;
and in
This information strongly counters the position taken by ExxonMobil that the probability of hitting significant water bearing fractures below 100 feet diminishes dramatically. The fact that private supply wells may be receiving water from upper water bearing zones in the bedrock, which the Exxon monitoring wells do intersect, does not mean that there are no significant water-bearing zones below the 100 foot bgs depth mark. If that were so, only a few wells would be dug so deep.
5. Well Reservoir Depth
ExxonMobil contends that the deep
wells in the community were only bored to provide a reservoir of water. If that were true, many community wells would
only extend to about 233 feet bgs on average.
However, the following analysis demonstrates otherwise. The state of
Required depth = (500 gal – [(yield gal/min) x 120 min])/1.5 gal/feet) + static water level
The formula is based on the fact that each foot of well
depth provides 1-1/2 gallons of water storage.
Yields in the area range from 1 gallon/minute to as high as 8
gallons/minute. Static water level
ranges from 40-80 feet. TABLE 1
below provides the range of required well depths calculated with this formula.
TABLE 1 Required Well Depths as a Function of Yield
and Static Water Level
|
Yield (gal/min) |
|
1 |
2 |
3 |
4 |
5 |
6 |
|
Static water level (Feet) |
|
|
|
|
|
|
|
|
40 (min) |
|
293 |
213 |
133 |
53 |
0 |
0 |
|
80 (max) |
|
333 |
253 |
173 |
93 |
13 |
0 |
Assuming and average yield of 2 gallons per minute and a static water level of 60 feet, the required well depth would be 233 feet. From the data collected, 90 of the 148 wells or 61% of community supply wells had well depths that exceeded 233 feet. Even assuming a yield of 1 gallon per minute and 60 feet of static depth, 34 percent, or more than a third, of the community’s supply wells have depths greater than the required well depth.
In conclusion, ExxonMobil’s cursory comment that
deep supply wells are bored solely for purposes of establishing a reservoir is
unfounded and misleading. Wells are
drilled with both purposes in mind: (1) establishing a reservoir for storage
and (2) reaching deeper water bearing fracture zones. The fact that deep wells were bored in the
Uppercrossroads area supports our contention that water-bearing zones exist as deep as 655 feet bgs in the area.
6. Geophysical
Data Confirming Fractures Below 100 feet
ExxonMobil asserts that there are few “fractures capable of transmitting water” below the 100 ft depth. (SARMar2005, p.32) Nevertheless, the geophysical data do not support this conclusion. To begin, the P278 (Morgan well) is only 77 feet deep; therefore, data from this well can imply nothing about the bedrock beyond the 100-foot depth. As for the remaining wells tested, there were data in each well indicating zones of secondary porosity. For instance, the D.R. Horton Lot 33 well exhibited potential water bearing zones at the 156-170 foot bgs interval and at 206 feet bgs. The JD (formerly Wawa) P244 supply well exhibited signs of water bearing zones at the 180-187 foot bgs and the 276-277 foot bgs intervals. There were also indications of water bearing zones at the 120-125 foot bgs, 135-140 foot bgs, 150-165 foot bgs, and 175-190 foot bgs intervals in the monitoring wells tested. These correspond with “soft zones” identified in the boring logs for the monitoring wells. Such soft zones can indicate fractures. It should also be noted that MW-22 was meant to be 200 feet deep, like the rest of the monitoring wells, but it collapsed at 180 feet and was finally set at 150 feet. The most likely cause of this collapse was a major fracture zone. The presence of an extraordinary high level of MTBE in this well and the fact that there was deep collapse is strongly suggestive that this well contained a water-bearing fracture zone near 180 feet.
All of this data provides strong evidence that the folded metamorphic rock in the Uppercrossroads region contains water-bearing fractures extending below 100 feet. ExxonMobil’s conclusion is simply incorrect.
7. Cross-Contamination of Deep Fracture
Zones via the Well Boreholes
While most of groundwater may flow
in fractures closer to the top of bedrock and these intervals may be the most
heavily contaminated, Exxon ignores the risk of contamination of deeper zones
via the wells in the area and on the Exxon site. The possibility of cross-contamination via
the boreholes of wells drilled through a contaminated zone and into deeper
bedrock is well accepted. This is also a
significant element of the hydrogeological framework. The presence of so many supply wells
extending to deep water-bearing zones in the bedrock creates a leaky aquifer
scenario, which affects the analysis of any aquifer test data. This will be discussed further in the section
on groundwater and aquifer testing.
8. Summary
1. Monitoring wells were bored without the consideration of available boring logs from local supply wells in the contaminated region.
2. The depth of the monitoring wells was based on total depth below ground surface rather than elevation. Well design based on a comparison of the elevation of “well bases” to contaminated supply wells is the only way to ensure proper assessment of the fracture zones in the bedrock aquifer system.
3. Seventy three percent (73%) of the community’s supply wells are deeper than ExxonMobil’s deepest monitoring well.
4. The depths of the community’s wells are not solely reservoir based.
5. Water bearing fractures in the Uppercrossroads area occur between 100 to 200 feet bgs on average with water bearing fracture zones as deep as 800 feet bgs according to the well drillers for the developments of Haddon Hurst, Franklins Chance and Delmar Farms. Water was not hit until 500 feet for the Dipasquales property. A yield of 50 gallons per minute at 685 feet bgs at a well adjacent to the study area is conclusive evidence of deep aquifers in the study region.
6. ExxonMobil’s geophysical data supports the fact that water bearing fractures occur below 100 feet bgs and even as low 276 feet bgs.
The failure to consider the depths of the existing supply wells in establishing the depths of the monitoring wells and to construct the monitoring wells so as to intersect all the water-bearing zones contributing water to community’s wells is a major flaw in study design. This seriously calls into question the validity of any conclusion regarding groundwater flow in the bedrock aquifer.
SECTION 4: SURFACE
ELECTRICAL RESISTIVITY TESTING
In general, ExxonMobil failed to perform a thorough investigation by including areas northwest of the site and to use the surface electrical resistivity profiles to their greatest advantage. The result is a cursory discussion that does not contribute to the greater understanding of the hydrogeological framework of the area around the site. What might have been more useful was a detailed soil investigation on the ExxonMobil site itself that included sampling along the product lines, near the dispensers, in the tank field and near the garage. This was not accomplished.
1. The Northern Transect
ExxonMobil failed to perform any
Surface Electrical Resistivity (SER) Testing in the sector northwest of the
site. The only test performed north of
the station was a transect oriented east to west and located 100 feet north of
the
2. The Transect behind the ExxonMobil
Station building
The results of the SER test for Line E offer striking evidence of a permeable zone in the overburden of the Exxon site. Line E runs southeast to northwest along the back of the ExxonMobil station site, paralleling the “B line” drainage system. FIGURE 6 provides a composite of information from SARMar2005 and illustrates transect lines from the SER tests. There is a markedly less resistive zone extending to the lower limits of depth detection toward the weathered bedrock and competent bedrock that extends 50 to 60 feet below surface. This area courses through the “epicenter” of the contaminant zone. While a low resistivity area can indicate the presence of clay, it also suggests increasing porosity with water saturation and/or dissolved ions in the groundwater. This zone is evidence of a plume on the ExxonMobil site itself, and at the very minimum, lends support to the premise that the overburden is permeable and a probable route for MTBE migration, especially as this area falls under the southerly drainage ditch discussed later in SECTION 6 “STORM WATER DRAINAGE”. This is consistent with the high level of contaminations found in MW 22 and MW 9.
3. Summary
1. A single Surface Electrical Resistivity Test northward above the shopping mall indicated the presence of two areas of low resistivity that were not investigated. The casings for MW 25 and 26 were set too deep to yield meaningful information on the transport of contaminants in the upper unit north of the site.
2. An SER profile located south of the ExxonMobil site revealed an area of low resistivity in close proximity to the contaminant “epicenter” on the ExxonMobil site. These findings support the conclusion that the Exxon station is the source of contamination.
SECTION 5: NORTHERN PROPERTIES AND THE NORTHWEST SECTOR
1. General
ExxonMobil’s investigation and data interpretation is based on the singular hypothesis that groundwater flow in both the overburden and the bedrock is solely towards the southwest. Consequently, ExxonMobil has refused to perform any meaningful investigation of flow to the north of the site.
2. The Strike
[CU1]While
the predominant strike in the area is N30oE to N50oE, the
predominant dip is to the northwest at N25oW to N35oW,
with a less prominent dip to the southeast at S35oE. This means that groundwater can flow to the
northwest and the southeast. Furthermore, there are a substantial number of
northwest trending lineaments and features.
Given these data, the northwest sector is a critical area of study. ExxonMobil has refused to even consider
studying these sectors in spite of its own data. In line with the ExxonMobil hypothesis, two
bedrock wells and zero overburden wells were installed northeast of the site,
and no bedrock or overburden monitoring wells were drilled in the quadrant
northwest of the site. FIGURE 2
First, the strikes represented in a rose diagram only suggest the probability of the strike direction. Both MW-25 and the Wawa well (P244) are situated on the outer margin of ExxonMobil’s own strike axis as depicted on their rose diagram. (SARMar2005, Figure 7) FIGURE 7 A contaminant will follow the path least resistance, and in a fracture-controlled aquifer system, that path will be a fracture. Locating that fracture line becomes more difficult the further one moves from the “epicenter” as the arc of the potential contaminant paths widens with increasing distance from the epicenter. FIGURE 8 Although MW 26 falls within the axis range, it must be emphasized that the strike axis only represents a probability that the strike will fall at any point in the path of the axis. A relatively short distance away, still within the strike axis zone, might be the predominant water-bearing fracture path.
At the ExxonMobil site there are differentials in MTBE contamination of over 1000 ppb in a 200-foot distance, e.g., MW-22 with 3770 ppb MTBE versus MW-18 with 2 ppb MTBE. FIGURE 9[CU3]. Thus, it is just as probable that there will be similar MTBE level differentials to the east or west of the P244 supply well or the other monitoring wells and supply wells north of the site.
There are also other potential pathways in the overburden that could have directed groundwater to the northwest. For example, the construction of MD 165 and MD 152 resulted in significant disturbance of the overburden. MTBE could have migrated through the overburden under the roadway. Exxon failed to address this possibility as well as the potential migration of MTBE through the storm water drainage system both to the southeast and the northwest. The storm water drainage systems on the site will be discussed in greater detail in SECTION 6 “STORM WATER DRAINAGE” below.
3. Levels of MTBE at Northern Supply Wells
Comparing the P244 well to an ExxonMobil monitoring well in terms of MTBE level is comparing apples to oranges. Note for example, MW 10 with an MTBE level of 7.5 (250 bgs, 337 amsl, 43) versus P244 with a MTBE level of 8.7 (504 bgs, 71 amsl, 44). The well at P244 is 254 feet deeper than MW 10. The reservoir (assuming a standard 6” pipe) holds 381 more gallons of water than MW 10. An adjusted level of MTBE for the P244 well accounting for the dilutional effect of the reservoir is 15.2 ppb MTBE as a baseline comparison. The Packer Test (SARMar2005, Figure 9) shows the expected result for MW 10 with the shallow zone demonstrating a level equal to the pretest equilibrated value. For P244, all zones shows significantly less MTBE than the pre-equilibration value. This can only be explained by the fact the Packer itself obstructed a feeding fracture zone.
4. Water and Contaminant Flow Northward
Another major concern is that there is significant horizontal flow of groundwater in the overburden in addition to vertical flow through the overburden and into the bedrock. The point at which horizontal flow descends into the bedrock in all four quadrants is not known, but this is especially true to the north. A major flaw in the study design is that there were no overburden wells bored northward.
The fact that groundwater flow in
the overburden does not solely follow a preferential path southeastward will be
discussed under the SECTION 7, “GROUNDWATER”, but it is worth reemphasizing
that the ExxonMobil site sits at a topographic high for the study region. It is on top of the “mountain” on
ExxonMobil’s selective use of surface electrical resistivity profiling (SAR Mar2005, Subsection 6.1) reflects the predeterminative approach of the entire investigation. Exxon refuses to consider that any water could have flowed northwest of the site. Thus, it did not perform a surface electrical resistivity profile for the area northwest of the site, and has not installed any monitoring wells in that area, not even to confirm its staunch presupposition. As long as Exxon refuses to collect any data from this quadrant, it cannot state conclusively that no water flows to the north of the site.
Through its consultant, Roux
Associates, ExxonMobil argued at the
As will be discussed below, there
are three significant locations in line with the strike northeastward that have
MTBE elevations. The property at 2716
Farmview (15-20 degrees NE, 0 .17 mile) showed a maximum MTBE level of 11.6 ppb
in the summer of 2004 and 7.5 ppb in January 2005. The property at 2719 Fallston (shopping
center) (20 degrees NE, 0.1 mile) showed a maximum MTBE level of 7.6 ppb in
January 2005. The property at
Second, these wells are private supply wells, and their casings (2716 Farmview-63 feet bgs, 2719 Fallston- 54 feet bgs, 2801 Fallston- 40 feet bgs) extend into bedrock. Consequently, these wells do not directly receive any groundwater from the overburden.
Third the depths of these wells are not comparable to the depths of the ExxonMobil monitoring wells (2716 Farmview 350 feet bgs, 2719 Fallston 400 feet bgs and 2801 Fallston 505 feet bgs). Hence the measured levels in these wells represent a diluted concentration of MTBE compared to the ExxonMobil monitoring wells as was so readily observed in the P244 site (Wawa supply well).
Fourth, there is a significant
lineament extending northwestward right up to 2716 Farmview and ending near the
entrance to Cross Country Estates.
Lineaments again are only indicators of the underlying geologic
morphology. The magnitude of this NW-SE
lineament is equally impressive as the NE-SW lineament. It is highly probable that there is a
juncture between the NE-SW and NW-SE lineaments in the bedrock driving
contaminated water in the direction of Cross Country Estates and
5. Summary
- Geophysical, lineaments and bedrock outcropping features have clearly demonstrated the presence of a major NE-SW strike, a major Northwest dip as well as a minor but significant NW-SE strike. Despite these data, ExxonMobil has not undertaken any investigation to determine the geomorphology of the area northwest of the site.
- To support their only hypothesis, ExxonMobil only placed two bedrock wells to the north of the site. One of these wells was placed at outer boundary of their described strike axis of N30oE and N50oE. Hence 50% (1 of the only 2) wells were placed in a location with a low probability of intersecting the main axis of contaminant flow.
- A major rationale for not exploring the northwest sector is ExxonMobil’s contention that groundwater only flows southeastward. They base this conclusion on static water isocontour maps, topography, surface electrical resistivity, hydraulic conductivity and the transmissivity of the aquifer. These factors will be addressed in subsequent sections. Furthermore, they also contend that since the MTBE level at P244 is low in comparison to an equidistant point southwest, that contamination from the site has not migrated northward. However, both the P244 well and MW 25 are beyond the limits of ExxonMobil’s own defined axis of the strike. (Note, that even though the P244 supply well is not within the strike axis, it is still contaminated). The decline in MTBE levels over short distances is amply demonstrated on the monitoring wells on ExxonMobil’s own site. Hence, even though the level of MTBE at P244 is only 8.7, it is possible that a short distance away the level could be significantly higher. The lower level may reflect the fact that P244 is not in the direct path of contaminant flux. Four properties as far as 0.17 miles northward show evidence of significant MTBE contamination. Clearly, there is a geomorphologic pathway for contaminant flux northward. The rational behind ExxonMobil’s theory cannot be supported.
- ExxonMobil sits on a topographic high with a relief of 100 feet within one-half mile of the Exxon site in all four quadrants. The steepest and closest downward slope is to the northwest.
- A
major lineament comparable to the NE-SW lineament extends southeast to
northwest ending at the base of
SECTION
6: STORM WATER DRAINAGE
1. General
On of the most disturbing elements regarding the layout of the ExxonMobil site is the location of the storm drain inlets, pathways and outlets in relationship to potential point sources of contamination.
The storm drain is deeply trenched as seen in the photograph of FIGURE 10. There are two main drainage culverts on the station. (FIGURE 11)
The first system, labeled “A” runs in an open ditch along the front peripheral edge of the station. The open ditch eventually enters a culvert and discharges northwestward into a rip- rap channel just alongside the MD 152 easement. One storm water inlet, one storm drain inlet and a storm water infiltration device empty into the same conduit leading to the rip-rap channel.
3. Southeastern
Storm Drainage Culvert
The second drainage system labeled “B” in FIGURE 11 is a drainage system of major concern. Water collecting to the rear of the property flows behind the station building directly over MW 22 and MW 9. Bedrock MW 22 is the “epicenter” well on site with an MTBE level of 3770 ppb. MW 9 is an overburden well with an MTBE level of 1150 ppb MTBE. The drainage courses behind the station building along the northeast edge of P278 (Morgan Property) before turning northwestward and emptying into the same rip-rap channel as drainage system “A”. What is most disconcerting is that drainage channel "B" is less than 10 feet from the septic system on the property. On site Sewage Disposal Records and HCHD records as noted previously have demonstrated that the soil exhibited a fast percolation rate (4 minutes and 5 minutes) (at 5 feet). Hence any superficial contamination would be quickly moved by water sieving through the drainage ditch. Contamination could migrate in two directions a) vertically into the soil of the drainage ditch by percolation or b) carried by the drainage waters to the rip-rap channel adjacent to MD 152. Because MTBE is extremely hydrophilic, water flowing over the highly contaminated area near MW 22 and MW 9 could draw the contamination with the flow, spreading the contamination along the course of the ditch an into the rip-rap channel. The fact that drainage ditch course along the northeast property line of P278 and the fact that the well on P278 is shallow at 75 feet suggests that contaminant migration via the drainage ditch could have been an alternative overburden route that contributed to the contamination of this well. ExxonMobil failed to adequately investigate and address this potential pathway of contaminant flow.
4. Storm
Drain Egress Topography
Of greater concern is the fate of
any contaminant that reached the rip-rap channel. Given the topography of the rip-rap site,
the water has no alternative but to percolate into the overburden. The only question remaining is what direction
the water will take once it penetrates into the overburden. A detailed topographic map of the elevation
isocontours in the northwest region of the ExxonMobil site near the riprap
channel illustrates a major route of potential contaminant spread to be northwestward. (FIGURE
12)
The water in the rip-rap channel is
initially trapped from flowing southward by the 565 feet and 570 feet
isocontour lines that broadly encompass the area. As the water perks into the ground, the path
of least resistance under gravity flow would pull the water northwestward
directly toward the subdivision of
The failure of ExxonMobil to provide a detailed topographic map at the discharge site of its storm drainage system in relationship to detailed topography of the surrounding region is inexplicable. The March report only shows a USGS topographic map (SAROct2004, Figure 1) It is submitted in this report as FIGURE 13. The scale of the USGS topographic map is inappropriate for illustrating the topography and surface features of the site and the surrounding region. Furthermore, Subsection 4.4 (SARMar2005) is totally inadequate in describing the relationship between the storm water runoff and potential contaminant routes. Nowhere in the report is there a detailed analysis of the soil or water VOC analysis in the rip-rap channel or a discussion about the fate of the drainage water than courses over the point with the highest contamination on the ExxonMobil property. This cursory treatment of a potentially major contaminating route by ExxonMobil, Groundwater Environmental Services, Inc. (GES) and Roux Associates, Inc. is inexcusable.
5. Summary
- A
major storm drain pathway traverses the contamination epicenter on the
ExxonMobil Site, potentially distributing contamination along the path of
the drainage ditch into the overburden and transporting the contamination
to a rip rap channel that eventually flows downhill towards
- ExxonMobil never investigated or discussed this major potential route of contamination to the area northwest of the site.
SECTION 7:
GROUNDWATER AND AQUIFER TESTING
1. Direction
In the preceding section, documentary evidence has been presented to demonstrate that the ExxonMobil site is at a topographic high for the region under study. Although surface water flow may be diverted or redirected by developmental structures such as highways and easements, the direction of water flow in the overburden tends to follow the surface topography and the topography of the top of the bedrock. In this regard, the ExxonMobil site is at a groundwater divide, with the potential for water to flow in all four directional quadrants north, east, south, and west. Indeed, in their recent report (SARMar2005, p. 63) ExxonMobil states, “The surface topography generally mimics the underlying bedrock topography”.
ExxonMobil from the outset predicted that groundwater flowed only to the southwest parallel to the major NE-SW axis. As mentioned, previously but worth restating, ExxonMobil suggests that not only does the overburden water flow to the southwest, but there is a topographic bedrock high northeast (elevation 530-540 amsl) on the site sloping downward to the southwest (approximate elevations of 500-510 feet amsl near Scarff Road). The report fails to mention that the same gradation also occurs northwestward. As there were no measurements of groundwater flow in this northwest sector, ExxonMobil has concluded that water only flows in the direction that it prefers the water to flow. This conclusion by ExxonMobil again demonstrates their single-minded approach to a complex situation that suits their forgone conclusions.
In order to obtain accurate information about the static level of water in the various wells in the proximity of the ExxonMobil Site, manual water levels at the east and west supply wells of 2719 Fallston Road (shopping center) and P244 (Wawa) were recorded. In the pre-pump test “Middle Unit Potentiometric Surface Map” (SARMar2005, Figure 12), the ground water elevations (feet amsl) for the following middle units were noted; Wawa (541.78), Shopping Center East (528.94), Shopping Center West (546.14), MW 26 (546.57) and MW 25 (547.92). This is the only map for which groundwater elevations were recorded for properties north of the ExxonMobil site. The fact that the water levels at P244 and the East shopping Center supply wells were 5 and 18 feet below the highest middle unit potentiometric surface contour at the ExxonMobil site (547 feet) clearly suggests that there is a potentiometic gradient northward. Regardless of this finding, there has been no further measurement of static water levels in the northern/ northeastern well. Reviewing the general topography map of the Uppercrossroads region (FIGURE 14), one can clearly make the inference that, if studied, the potentiometric isocontour lines would form a circle around the ExxonMobil site. (FIGURE 15) It is unequivocal that groundwater flow is not restricted to a southeastern path.
2. Depth
In the SARMar2005, ExxonMobil presents a site conceptual hydrogeological model of the ExxonMobil site and surrounding areas. Exxon Mobil defines the first 40 feet bgs as “saprolite”. Although a misnomer, its composition is silty clay with a relic structure. Weathered bedrock (the true saprolite) extends from 40-65 feet bgs. Below the weathered bedrock is competent fractured bedrock between 65 to 100 feet with a decreasing number of fractures (according to ExxonMobil) from 75 to 100 feet.
This definition of the site conceptual geology model is flawed. Developing a uniform conceptual geologic model for highly fractured and folded metamorphic bedrock involves the making of many major assumptions. The location, direction and depth of fractures may follow a general trend, but isolated point borings, even considered in their aggregate, are not predictive of the entire study area. As clearly demonstrated in SECTION 3, “WELL DEPTHS”, the average depth of water bearing fractures in the bedrock in the Uppercrossroads area is between 100 and 200 feet. In addition, although less frequent, high yield water bearing zones exist at depths of 655 feet. Only in the ExxonMobil monitoring wells were there lower yields of groundwater at depths greater than 75-100 feet.
The only conclusion that can be
made regarding a conceptual model is that once the bedrock is reached, the
model represents the characteristics of the individual monitoring wells
tested. Clearly, the aquifer in the
Uppercrossroads area is extensive and deep.
Furthermore, it is well know than MTBE dives deep into the aquifer. If, for instance, a monitoring well was
extended another 100 feet deep it is possible that the FLUTe test (which
revealed no water bearing fractures below one hundred feet) would provide
completely different results. For example,
the well at
ExxonMobil’s conceptual model
consists of an upper unit (weathered bedrock 40-65 feet bgs) and a middle unit,
which is less weathered bedrock from 65-100 feet. Where the overlying “silty saprolitic clay”
is saturated and in communication with the weathered bedrock, the “silty
saprolitic clay” is considered part of the upper unit. There is no confining layer between the upper
and middle units. We believe that there
is no evidence of a confining layer, but even if there were, ExxonMobil has not
provided any indication where the upper and middle unit aquifers are confined
and where they are unconfined. According
to ExxonMobil, the following figure demonstrates the definition of a “confined upper and middle unit” and an
“unconfined upper and middle unit”. (FIGURE
17)
Figure 17 ExxonMobil's Definition of a Confined and Unconfined Upper and Middle Unit
This is an extremely important point as the presence of a confining layer in the overburden suggests an impedance to water flow and hence a retarding of the movement of contaminants. As noted previously, if the “silty saprolitic clay” were a confining layer, water would be perched above the level of this silty clay relic. There is no evidence to support this.
Although, ExxonMobil is quick to conclude that the aquifer system is confined or semi-confined in some regions (SARMar2005, p. 51), it should be noted that the so-called upper and middle hydrostratigraphic units were affected by barometric pressure. This would seem to imply unconfined conditions. The liberal and interpretative application of “confined” is not supported either geologically or by ExxonMobil’s own data.
In summary, ExxonMobil’s Geologic Hydrostratigraphic Column is the basis of their derived site conceptual model. There are three fundamental characteristics that define their proposed column:
1. There is limited fracturing of the competent bedrock below 100 feet.
2. There is no hydraulic interconnectivity between the upper/middle and lower units.
3. There is minimal groundwater flow in the lower unit from either the primary lower zone fractures or from any secondary vertical interconnections.
The lower unit is starved even further of groundwater by the claim that the fast percolating “silty saprolitic clay” becomes impermeable under the conditions described in Figure 17.
We find all of these characterizations untenable.
ExxonMobil bored 15 bedrock-monitoring wells to the depths they felt were representative of the typical depths of the zonal aquifer layers of the Uppercrossroads region. Averaging the extensive geophysical data derived from the “shallow” bedrock wells, ExxonMobil formulated their geological hydrostratigraphic column. Given the unpredictable, complex geomorphology of the bedrock system, one must ask if the true descriptive nature of the aquifer can be based on the average descriptions from all of these wells. If the underlying strata were uniform, the answer might be yes. However as demonstrated by the finding of 2 dry wells on the ExxonMobil Site (MW8 and MW 15), productive wells can be found in very close proximity to dry wells. The productivity of a well and hence its intersection with the zonal fractures is a function of both depth and location of the well.
A total site conceptual model can therefore not be based on an average of descriptions from all the wells. An analogy is akin to the example of a pathologist who takes 100 slices from a biopsy specimen and finds cancer in one slice. Does the patient have 1/100 of a cancer? Or does the patient have cancer? The answer is obvious. In other words, as the site is investigated, and as wells are discovered that exhibit characteristics of groundwater flow in the deeper zones that are not apparent in other wells, those characteristics become the defining model for the aquifer. The discovery of these characteristics in a well should lead to further investigations and deeper borings to further define the extent of these properties.
The only monitoring well that even began to approximate the wells typical of the surrounding aquifer was monitoring well 12. Had ExxonMobil investigated the numerous off-site wells drawing from the deeper aquifer, the findings characteristic of MW 12 would have been confirmed. Although the depth of MW12 is insufficient, its geophysical characteristics should have made this well the starting focus of their geophysical investigation. MW 12 demonstrated comparable contaminant levels in both its deep and shallow zones on three separate monthly tests (October, zone 1, 772 ppb MTBE, zone 2 609 ppb MTBE), (December, shallow 627 ppb MTBE, deep 720 ppb MTBE) (January, shallow 482 ppb MTBE, deep 537 ppb MTBE). In addition, with 2 separate packers in place, MW 12 demonstrated hydraulic interconnectivity between upper/middle and lower zones as noted previously.
ExxonMobil states that the similar contaminant levels may have been secondary to poor purging of the lower zones, but the presence of the demonstrated hydraulic conductivity defeats their attempt to minimize the implication of this finding that the Uppercrossroads aquifer does indeed have zonal interconnectivity (to be discussed below).
3. Interconnection
Another important factor that affects groundwater flow is the communication between the hydrostratigraphic units. ExxonMobil acknowledges that the upper and middle units are hydraulically interconnected. In addition, MW 12 demonstrated vertical interconnection with the lower unit. This is consistent with the known frequency of vertical fractures found in this type of bedrock formation. The presence of vertical fractures is difficult to discern, but their presence suggests a rapid route of water and contaminant migration to deeper zones in the aquifer. It is interesting to note that ExxonMobil acknowledges that MTBE levels obtained from Packer test isolation of shallow and deep zones were relatively similar in MW 12. Rather than attribute this to the vertical interconnection that it acknowledged hydraulically or any other alternative explanation, ExxonMobil states, “This may be associated with extremely low yields at depth which do not permit an extensive purging of the lower zone prior to sampling.” (SARMar2005, p.40) Alternative explanations are that the Packer results reflect a low yield water-bearing fracture zone with a high concentration of MTBE or true vertical interconnection.
4. Rate of
Travel
A fourth major issue regarding groundwater flow is its rate of travel. In the SAROct2004, ExxonMobil reported on its investigation of hydraulic conductivity of eight monitoring wells.
A review of the hydraulic conductivities (K value) as a function of location and soil description reveals several inconsistencies. The hydraulic conductivities ranged from values of 0.05 to 230 feet per day. (TABLE 2) Although these hydraulic conductivities do not represent a linear velocity through the site, this would translate to a migration range of 21 feet/year to 84,000 feet/year. Monitoring well 17 is described as penetrating a gravel bed in the soil description log (FIGURE 1) and is only 110 feet away from MW 21 (highest recorded K value of 239 feet /day) and yet it has one of the lowest K values within the test group. It is interesting to observe that only those monitoring wells distant from the contaminant “epicenter” have K values more reflective of the true nature of the subsurface. Again, although one cannot use the K value for linear velocity, a broad comparison of linear velocity can be made using the stated values to understand their “implied “significance. Assuming the K-value is reflective of the entire overburden, a K value of 1 foot/day would cause water to move a distance of one mile in 15 years. More realistically the K value of 2.7 (957 feet/year) would cause water to move a distance of 1 mile in 5.5 years.
TABLE 2 Overburden Aquifer Test Data (K values)
|
Source |
Soil Description |
Estimated K Value feet/day |
Bouwer-Rice K Solution
feet/day |
Feet per year |
|
MW 5 |
CL, SC |
0.1779 |
0.2042 |
75 |
|
MW 6 |
CL, SC, MH |
0.1125 |
0.2245 |
82 |
|
MW 9 |
CL |
0.2464 |
missing |
90* |
|
MW 11 |
ML, CL |
0.04776 |
0.05001 |
18 |
|
MW 17 |
CL |
missing |
0.05681 |
21 |
|
MW 19 |
CL |
2.744 |
2.621 |
957 |
|
MW 21 |
CL |
230.9 |
230.9 |
84,000 |
|
MW 27 |
CL |
missing |
0.2518 |
92 |
* Derived from the estimated value
It is astonishing that MW 19, 21 and 27 with identical labeled soil descriptions (FIGURE 1) demonstrates two orders of magnitude difference in K-values. All three wells have nearly the same proportion of clay or “silty saprolitic clay”. The K-values are inconsistent with the generally known characteristics of the “silty saprolitic clay” ExxonMobil claims is present in the Uppercrossroads area. This material is generally understood to show rapid percolation rates. The values are infinitesimally small and reflect only very local conditions. Even using this data, however, it should be noted that the highest hydraulic conductivity is occurring westward and northward, areas that ExxonMobil has refused to study.
In the SAROct2004, ExxonMobil
disregarded the K value of MW 21 in determining an average K-value. In the
notice from the MDE to ExxonMobil on
5. The
Aquifer (Pump) Test
This is the centerpiece of any hydrogeologic investigation and the discussion of the test seems fairly complete. Nevertheless, the test is deficient in several respects. To begin the test did not include any off-site wells northwest of the site. Second, the use of MW-14 as the pumping well was inappropriate, because MW 14 intersects numerous fracture zones in the bedrock, as well as the middle and lower units of Exxon's designated "hydrostratigraphic units." (Interestingly, the well completion report implies that well MW 14 was screened in the middle and lower hydrostratigraphic units. Because the definition of the hydrostratigraphic units includes weathered bedrock in both the upper and middle units, it cannot be said for sure that MW 14 does or does not intersect a portion of the upper unit.)
As to the results of the test,
Exxon downplays any response in wells north of the site that were included in
the pump test. For example, the report
stated that the supply wells at 2719 Fallston Road (shopping center) which are
0.10 miles northeast of the site “may have had a response to pumping, but data
are inconclusive.” In addition, the
supply wells at the P244 site showed evidence of drawdown, but the report is
quick to dismiss this as due to a high level of commercial water use. “The sustained pumping duration and larger
extraction volumes associated with these commercial supply wells likely has
reversed groundwater flow in a very localized area.” (SARMar2005 p. ix). The reports states that the measurements
of groundwater elevation “in conjunction with the continuous rate pumping test”
had limited effect on the potentiometric head beneath the site. It should be noted that during the pump
test, the supply pump to P244 was disabled or removed (similarly for
True to its conceptual hydrogeologic model, the report discounts any contribution of zones below the 100-foot depth mark to well yield. Saying only that it is based on Packer testing in the pumping well, MW-14, the report assigns 72.5 percent of total flow to be derived from above the 65 ft. bgs mark, 23 percent from the 68-77 ft. bgs interval, and 4.5 percent from below the 80 ft. bgs mark. (SARMar2005, p 52) But ExxonMobil goes further in discounting any flow from the lower regimes by assigning the upper unit (highly weathered bedrock), 72.5 percent of the yield and the middle unit (less weathered bedrock), 27.5 percent of the yield. These percentages were used to calculate the pumping rates for the analysis of the pumping data. Thus, the contribution from fracture zones below the 100 ft. bgs mark were not considered for any of the wells in the test, except MW-12. The problem with this apportionment of the flow rate, besides having a questionable basis in scientific methodology, is that MW 14 may not have intersected the upper unit fully or even at all. It appears that the zone from which the water is coming is temporally variable, based on fractures intersected, sources of water feeding the units, and dewatering of the three units at different rates. The drawdown curve for well MW 14 suggests that multiple sources of water are contributing to the pumping of the well.
An analysis of “drawdown” is complex requiring sophisticated modeling, curve comparison, corrections for distance from the pumping well source etc. and beyond the scope of this discussion. The challenge in this report is not the actual data collected, but its interpretation as well as the failure to collect important data as a result of a flawed study design.
If one reviews the corrected drawdown graphs (SARMar05, Appendix P CD “Corrected Transducer/Manual Files”) and classifies the receptor wells according to their plotted scale, it is quickly apparent that certain units have more pronounced drawdown curves than others. For whatever reason, ExxonMobil plotted the drawdown of the various monitoring and private wells on different “y” scales reflective of the magnitude of the drawdown or the magnitude of associated “noise”. This analysis is presented in TABLE 3.
TABLE 3 Drawdown of Receptors Monitoring Wells and Private Supply Wells According to ‘Y” axis Scale of the Drawdown Plots of Corrected Transducer Manual Files
|
|
|
SCALE |
|
|
Y AXIS SCALE OF CORRECTED DRAWDOWN PLOTS |
|
(in
feet) |
|
|
|
|
|
|
|
> 0 £ 1 |
|
|
>1 £ 3 |
|
>3 £ 5 |
|
>5 £ 7 |
|
>7 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
RECEPTOR |
|
1 |
25 |
|
7 |
|
2 |
|
10 BOT |
|
6 |
|
WELLS |
|
4 |
27 |
|
5 |
|
19 |
|
12TOP |
|
12BOT |
|
|
|
11 |
28 |
|
9 |
|
22BOT |
|
30BOT |
|
31BOT |
|
|
|
13 |
29 |
|
10TOP |
|
18BOT |
|
|
|
SSEAST |
|
|
|
17 |
30 |
|
16 |
|
WAWA
SHALLOW |
|
|
|
SSWEST |
|
|
|
18 |
32 |
|
22 |
|
WAWA MID |
|
|
|
|
|
|
|
20 |
33 |
|
26 |
|
WAWA DEEP |
|
|
|
|
|
|
|
21 |
24TOP |
|
EXXON
SUPPLY SHALLOW |
|
|
|
|
|
|
|
|
|
23 |
31TOP |
|
EXXON
SUPPLY MID |
|
|
|
|
|
|
|
|
|
25 |
MORGAN |
|
EXXON
SUPPLY DEEP |
|
|
|
|
|
|
|
|
|
|
TAFI |
|
|
|
|
|
|
|
|
This is a qualitative rather than quantitative analysis. If we look at the corrected data (adjusted for pre and post trending data of static water levels and barometric pressure), it is clear from this very simplistic approach that drawdown curves from the lower units either had the largest amount of drawdown or fluctuations that required a larger “Y” scale plotting axis. The only conclusion to be pointed out is that clearly the lower units were impacted by the draw of the pumping well.
It is interesting that for the Shopping Center East supply well, the effects of drawdown were inconclusive due to spiking “noise”. There is no mention in the investigation that the cause of these presumed spikes (attributed to continued pumping of the well against a closed cap), was investigated or corrected by simply “pulling the well pump plug”. As the static water level at this well is 18 feet lower than the static water level of the wells on the Exxon Mobil site, it would have been best for ExxonMobil not to implicate this well in its findings.
The Moench solution is the appropriate method for calculating hydraulic conductivity (K) and specific storage in a fractured bedrock aquifer. However, a primary assumption of the Moench solution for hydraulic conductivity is that the aquifer is confined. Exxon has not clearly demonstrated that the sol-called saprolite acts as a confining layer, at least for the wells used to determine the aquifer's hydraulic conductivity by the Moench method. Another assumption of the Moench method is that flow to the wells is entirely through fractures. This does not account for the leakage from the saprolite into the bedrock, or the movement of water from upper water-bearing zones to lower water-bearing zones via the numerous wells in the region. Head data demonstrate a vertical downward gradient between the upper, middle, and lower hydrostratigraphic units in most of the shallow and deep wells, based on pre-pumping data (SARMar2005, Figures 10, 12, and 14) This downward head gradient demonstrates a natural tendency for shallow groundwater to migrate unimpeded into the bedrock aquifer system. Given this downward hydraulic gradient, it is all the more critical to design deep bedrock monitoring wells to minimize vertical cross-contamination of lower water-bearing zones via the borehole.
Exxon has presented the K and S values for only the upper and middle hydrostratigraphic units. As discussed above, MW 14 may only partially receive water from the upper unit. If this is the case, then the K and S values for the upper unit and derived from the data for MW 14 must be discounted or adjusted. The report does not present a groundwater velocity for these units. It should be noted that a fractured bedrock aquifer could have relatively low fracture porosity but have a high velocity. Freeze, R.A. & Cherry, J.A., Groundwater at 409 (1st ed. 1979).
In summary, the aquifer pump test assigned a mean hydraulic conductivity (K) of 2.9 feet/day and a specific storage value (Ss) of 5.3e-05 feet-1. This calculation totally discounts any contribution from the lower unit based on criteria established from the restricted well depths of the original study design. The drawdown extended northeastward despite the fact that there was no commercial use of water in the northern commercial establishments. The fact that there was a detectable drawdown at the P244 supply well, and this was the northernmost well in the test series confirms the original criticism that the monitoring wells used for the pump test were inappropriate for quantifying the characteristics of the aquifer system used by the affected community, especially northward. As ExxonMobil essentially discounted the draw down influence northeastward due to “unexplainable “drawdown spikes” and no receptor wells were included in the northwest sector, the conclusion regarding an anisotropy ratio of 2:1 and 3:1 is invalid. Regardless of these significant criticisms and even assuming these invalid conclusions, the distant and widespread movement of contamination from releases in the past and the present is readily accounted for by this ultra conservative estimate of aquifer transmissivity.
6. Summary
1. The direction of water flow (potentiometric gradient) as determined by surface topography and bedrock topography as well as static potentiomentric levels of north supply wells is not restricted to the southwest. ExxonMobil has refused to investigate the current paths of water flow northward. In addition, the current levels of static water level may not even begin to approach the degree of potentiometric gradients existing in the areas northeast, north, northwest and west of the site.
2. The site conceptual geologic and hydrostratigraphic model is point specific at the level of the bedrock. In a highly fractured metamorphic bedrock system, it is difficult to predict the depth and yield of fractures. There is adequate documentation demonstrating the existence of high yield water-bearing fractures t depth in the study region. Furthermore, although geophysical evaluation suggests a general trend of the strike and dip for a region, there are individual variations in these features for different wells (e.g. P333 (Lin Mar supply well)). While the model may be valid for the specific monitoring wells investigated, the extension of the bedrock portion of the model to the rest of the study area is not valid.
3. The data provided by ExxonMobil support the concept that the different
hydrostratigraphic units are interconnected. Although ExxonMobil acknowledges that the upper and middle units are hydraulically interconnected, and indeed provided evidence of hydraulic interconnectivity with the lower unit (MW12), they consistently make reference to “hydraulic isolation of the lower unit”. (SARMar2005, p. 64) ExxonMobil refuses to acknowledge that there is interconnection with the lower units. The fact that levels of MTBE are comparable in the deep and shallow units (as demonstrated by Packer testing) supports this contention.
4. The hydraulic conductivity measured and reported for the eight monitoring wells in SAROct2004 show a twofold log variation despite close proximity to each other and identical soil descriptions. Furthermore, percolations documented in sanitary construction records in 1975 and 1984 demonstrated a rate of 4 minutes and 5 minutes respectively. The hydraulic conductivities are clearly "well specific" and the values not indicative of the site characteristics.
5. While the methodology for deriving the K and Ss values for the aquifer appears valid, the assumptions for flow rate for the upper, middle and lower units are not valid for the whole aquifer. The first major design flaw pertains to the pumping well. Monitoring well 14, (bored to 200 feet as were all but one of the other monitoring wells) is not representative of the depths of the water bearing fracture zones present throughout the regional aquifer. Moreover, MW 14 may not even receive water from the upper hydrostratigraphic unit. Hence from the start, assigning a low flow rate to the lower unit is merely an assumption based in the restricted study design. Second, the pump test did not include any off-site wells northwest of the site.
6. The fact that P244 and the shopping center wells experienced drawdown (in the absence of any commercial activity) strongly suggests that the ExxonMobil site influences the aquifers. Again it should be noted that the report does not present a groundwater velocity for the hydrostratigraphic units and that a fractured bedrock aquifer could have relatively low fracture porosity with a high velocity
SECTION 8:
MTBE OVERALL CONSIDERATIONS
1. Evidence
supporting the ExxonMobil Site as the wide spread contaminant source
In order to understand the dynamics of contaminant spread in a region with complex geomorphology (metamorphic rock, fractures, dipping planes, complex overburden and a varying topographic terrain), it is necessary to undertake a comprehensive investigation of the hydro-geomorphology of the region. This is especially true when the suspected source of contamination is geographically intermixed in that complex terrain.
Regardless, however, there are certain overwhelming bodies of evidence that overshadow the underlying complexities and give strong indications as to the origins of the contamination.
a. Topography
The ExxonMobil site sits at a topographic high with steep down sloping terrain in the north, east south and west directions. As such, the ExxonMobil station is located on a groundwater divide with flows on the surface and overburden to the north, east south and west.
Since October 2004, the GFA has
challenged the relevance of local topography to groundwater flow. This became increasingly relevant as in the
Executive summary of SAROct2004, where ExxonMobil concludes “groundwater in the overburden flows to the
south and southeast from the Site”. As
noted previously, ExxonMobil never provided a detailed topographical map of the
area. (SAROct2004, Figure 1) (FIGURE 13) The GFA, with assistance from a local
surveyor and an environmental engineer, produced a 3-D elevation topographical
model based on detailed local topographic information. This model demonstrates without question, that
the ExxonMobil station is situated at a
Review of a two dimensional detailed
topographical map of the region also clearly illustrates the four downward
sloping regions from the ExxonMobil station.
(FIGURE14) In
fact, it should be noted that the region with the closest and steepest downward
gradient is northwestward to
The failure of ExxonMobil to undertake a complete investigation of all the quadrants surrounding the station with regards to groundwater flow conforms to its isolated and limited approach of only presenting facts relevant to its own hypothesis. Therefore the study is incomplete and totally inconclusive.
Further, evidence of the isolated and restricted nature of the study is revealed in the SARMar2005, p.63 which states, “The local bedrock topography provides a more detailed assessment of the bedrock conditions beneath and in the immediate vicinity of the site. The local bedrock topographic high is located to the north and east of the site (approximate elevations of 530 to 540 feet amsl) and slopes downward to the southeast (approximate elevations of 500 to 510 feet amsl near Scarff Road).” Figures 20 and 21 of SARMar2005 plot elevations of the bedrock based on GPS determinations of well elevations and a review of well boring logs for the monitoring wells and community supply wells. If one enlarges Figure 20 four times and superimposes it on Figure 21, the 500-foot isocontour line (marked in red and derived from ExxonMobil’s own GPS elevation assessment) is equidistant from almost any point at the station. (FIGURE 19) The first assumption that surface elevation parallels bedrock elevation may be true in an undeveloped terrain, but not at the Exxon Site or at the Highways of MD 152 and 165 and the surrounding commercial establishments. These terrains have essentially been graded flat (as is most of the ExxonMobil site.) Furthermore, even assuming there is some correlation of the underlying bedrock topography with surface topography, there is an equal probability that water will flow northwestward as in fact the 500 bedrock isocontour line is closer to the northwest direction than it is to the southeast direction.
b. Absolute
levels of MTBE
A second overwhelming body of evidence pointing to the ExxonMobil Site as the source of the community wide contamination comes from the absolute levels of MTBE at the site compared to those in the surrounding community. For a significant portion of the community, the level of ExxonMobil's contamination is 4 orders of magnitude higher than the community levels.
Since July of 2004, Cross Country
Estates Community Association, Inc. (CCECA), the GFA and other local community
geologists requested that ExxonMobil reformulate its reporting of MTBE levels
in its Private Supply Well Sampling Map due to the fact that the GES legend
does not make a distinction between any levels of MTBE higher than 20 ppb.
Despite numerous requests to do so, ExxonMobil never complied with this
request. In response, the community generated its own computer database of MTBE
levels from the SARMar2005. From this
data, we produced both a physical model and a computer-generated model of maximum
MTBE levels as of January 2005. (FIGURES
20-21) (The 3-D rotational graphic and
the physical model are available for review upon request). ExxonMobil levels are not included in
these graphic presentations as will be detailed below. All values represented in the model and
graph represent maximum concentrations detected in samples collected between
In the SARMar2005, ExxonMobil shows the level MTBE in MW 22 to be 3770 ppb MTBE (deep) and 4590 ppb MTBE (shallow.) The well was originally bored to 200 feet deep but sustained a 40-foot collapse to 160 feet deep (elevation 421 feet above mean sea level (amsl)). It is the estimation of most consultant geologists that the collapse is most likely explained by the fact that an extensive fracture bed was penetrated. Monitoring well 22 is bedrock well. Fifteen percent of community’s supply wells for which boring logs were available had well base depths as shallow as 160 feet. (FIGURE 22) With this information, it is absolutely fair and reasonable to conclude that MW 22 is a bedrock well that can be compared to other bedrock wells in the community. Given this fact, the ExxonMobil level of MTBE shown in the physical model (and computer model) of community MTBE levels would be 10 times higher than the highest level seen in FIGURE 20 and the physical model.
In essence, the ExxonMobil station has a monumentally high level of MTBE as a point source and sits at a topographical high. Logic would defy any other explanation other than the fact that contamination would spread in all directions, far and wide in the community aquifer.
.
c. Absence of a BTEX plume
A third major factor supporting the contention that ExxonMobil is the site responsible for MTBE contamination in the study area and beyond is the fact that there have been no documented contaminant plumes for benzene, toluene, ethybenzene or toluene (BTEX) products lagging behind the MTBE plume. As demonstrated in Table 11 of (SAROct2004) the overburden and bedrock monitoring wells showed no evidence of BTEX products. Likewise, soil analysis for BTEX, TPH-DRO and TPH-GRO were non-detect. The only petroleum contaminant was MTBE in both the monitoring wells and in soil samples from MW 5 and 6 (352 ppb MTBE and 795 ppb MTBE respectively). Likewise, the private supply wells in the community show only evidence of MTBE contamination without an associated BTEX component. ExxonMobil has proposed the idea that the community’s wells were contaminated by the homeowners via heating oil or gasoline spillage or previous petroleum spillage by unidentified sources. Clearly, if the community were responsible for contamination via these routes, they would have occurred in a random fashion with variable BTEX plumes following the MTBE plume. The question begs to be answered. How can a community be contaminated with MTBE alone if petroleum spills are the most likely source of community caused contamination? Is there any situation in the community where MTBE vapor releases can occur in the absence of a gasoline spill? The answer is unequivocally no.
2. Dating the contamination
A fourth major issue that affects the interpretation of all the contamination in the Uppercrossroads region is the date and amount of current and past releases. There is an assumption by ExxonMobil that the MTBE contamination currently being investigated is the result of a recent contamination. It is indisputable that there have been releases in the past (1990, 45 ppb MTBE in the supply well at ExxonMobil, and in 1998 when 126 ppb MTBE was detected at the 2404 Baldwin Mill supply well (adjoining property). In October of 2003, ExxonMobil submitted to MDE the results of a subsurface Environmental investigation that demonstrated 3730 ppb MTBE in MW 4. In fact, given the massive release currently documented at MW 22 (max 4500 ppb) and the current level of 12.9 in the supply well (one-third of the 1990 level), it is not only possible but likely that a release of even greater magnitude occurred as far back as 1990 if one assumes a proportional change based on similar overburden and underlying geological structure. This historical trending makes it indisputable that past releases of significant quantities of MTBE have occurred. As is currently evidenced, these releases were likely vapor in nature. As vapor releases have only been recently recognized, and past monitoring of USTs were designed to detect liquid releases, it is highly plausible that unrecognized massive vapor releases occurred throughout the station site history.
There is some evidence in the literature that MTBE, under aerobic conditions, can degrade to tertiary butyl alcohol (TBA).
Role
of Natural Attenuation in the Life Cycle of MTBE Plumes. Wilson, John T. and
Natural
Attenuation of MTBE in the Subsurface under Methanogenic Conditions. 2000.
Use
of Compound-Specific Stable Carbon Isotope Analyses to Demonstrate Anaerobic
Biodegradation of MTBE in Groundwater at Gasoline Release Site.
Environmental
Behavior and Fate of Methyl tert-Butyl Ether (MTBE). Squillace, Paul J.,
Pankow, James F., Korte, Nic E., and Zogorski, John S.
MDE requested that TBA be added to the list of VOCs as part of the routine private supply well sampling regime.
TBA is present in gasoline, but in very small quantities. An argument has been raised that since there is no TBA in either the ExxonMobil MWs or in the community’s supply wells, that there has been little time for degradation and hence the contamination is recent. This argument is strongly countered by the fact that there is little to no TBA in the community’s supply wells. If the MTBE contamination in the community occurred through petroleum spillage by households or other commercial sources, these contaminations would have been expected to occur randomly over the past 2 decades. There should be varying levels of TBA in the community’s supply wells.
The assumption that the absence of TBA is due to recent spillage is simply wrongheaded, since its absence is best explained by the slow degradation due presumably to anaerobic conditions community wide. Hence the presence or absence of TBA cannot be used as an argument for or against a recent release.
It is the contention of most consultants that in addition to the massive current release in 2003-2004, there have been significant releases in the past that have been migrating through the community for up to 18 years.
As noted above, the date(s) of the release(s) are unknown but clearly date back to 1990 and 1998. Using the well specific K values measured by ExxonMobil and discussed above, the “implied” water and contaminant migration in the overburden alone, (purely an implication as the K value can not be used as a linear velocity) would be 1 mile in 5.5 years, clearly ample time for transmission to these areas.
.
3. Summary
- The contamination of MTBE at the ExxonMobil site is four orders of magnitude greater than the levels of contamination for many of the supply wells in the community. The massive gradient imposed by this differential makes he ExxonMobil site the most likely source of the community wide contamination in all directions in the current study area and beyond
- ExxonMobil site is situated at a topographic high. Steep 100 feet declines in just over a one-half mile distance occur in all directions, north, east, south and west. ExxonMobil’s assertion that the majority of water flows in the upper and middle zones above the competent bedrock (the overburden) makes their position at a topographic high all the more vulnerable to downward potentiometric gradient flow in all directional quadrants.
- The absence of BTEX products in the community’s supply wells makes community-based point source of petroleum spillage extremely unlikely.
- As the industry focus in the past was on liquid releases from USTs, it is highly likely that there were significant prior vapor releases at this ExxonMobil site that went undetected over the past two decades. This is supported by the finding of MTBE alone in ExxonMobil’s supply wells and monitoring well dating back to 1990, 1998 and 2003. It is unclear but highly probable that the current release is the same or a different release than the one that occurred in 2003. These releases would account for the now amorphous and widespread MTBE contamination of the Uppercrossroads aquifer.
SECTION 9: CONTAMINATION IN THE COMMUNITY
1. Failure to assign a probability or level of contribution to the MTBE Contamination
Exxon claims that all MTBE contamination north of the site is not the result of operations at the site. To make this claim, however, Exxon must give evidence showing that there are other potentially significant sources of MTBE in the area. Subsection 4.5, Other Potential Sources in Vicinity of the Site (SarMar2005), should accomplish this, but again fails by giving vague references to a report prepared by a Geologic Services Corporation (GSC) last year and presented in SAROct2004. The first major problem with the subcontractor’s report is that it is merely a list of all potential sources, but it does not evaluate the probability that each potential source contributed to the MTBE contamination in the area or the contribution volume of each likely source. For example, the report states that there have been 71 accidents involving 141 vehicles at the intersection of Baldwin Mill and Fallston Roads between 1980 and 2003. However, the report does not list any other facts important to the determination of this potential source’s contribution. Such facts would include how many accidents occurred before 1990 when MTBE was required to be added to gasoline or how many accidents involved ruptured fuel tanks or some gasoline spillage. A fender-bender accident obviously would not necessarily involve a leak or spill of gasoline on to the roadway. Regardless of the number of site identified by ExxonMobil, not a single site has revealed elevations in MTBE of a magnitude that could contaminate the aquifer locally or regionally. In fact, most of these site show only low levels of MTBE especially in comparison to the 3-4 log increases in MTBE levels present at the ExxonMobil site.
2. Inaccurate
and misleading information
A second major problem with the GSC
area report is that it is not field verified. An example is Figure AR 2 in
SarOct2004, which gives a graphic representation of homes in the study area
with heating oil tanks. The data was
not verified, and therefore is misleading and grossly inaccurate. It does not distinguish between USTs, ASTs
and it grossly mislabels properties as having heating oil tanks when they do
not exist. As a demonstration in point,
residents of
3. Case By Case Analysis
It is important to focus on several critical properties where MTBE levels are unusually high to ascertain if there is a possible contributing cause to the elevated MTBE level at those properties.
1.
On
TABLE 4 Summary of Table 24.1 from SAROct2004
|
Analyte |
TPH-GRO |
TPH-DRO |
Benzene |
Toluene |
Ethylbenzene |
Xylenes |
MTBE |
|
|
(mg/kg) |
(mg/kg) |
(mg/kg) |
(mg/kg) |
(mg/kg) |
(mg/kg) |
(mg/kg) |
|
Stockpile |
11 |
1930 |
16 |
203 |
199 |
816 |
305 |
|
2
feet below tank |
ND |
89 |
ND |
ND |
ND |
ND |
ND |
|
3
feet below tank |
ND |
24 |
ND |
ND |
ND |
ND |
ND |
Note: (mg/kg) =parts per billion (ppb)
ExxonMobil has not provided a
detailed evaluation of
The owners of this property have been maligned through inaccurate press releases that have mischaracterized the nature of the soil that was removed. Witnesses at the tank removal made clear statements that this was not a source of contamination. The supportive facts and data noted above are totally consistent with the assertion that the heating oil tank was not the cause of the elevated MTBE level at this property or any of the adjoining properties.
2.
Ten (10) salvaged ASTs and 40- 5 gallon buckets of waste oil were present on site. MDE inspector on site made no notation of spillage. The owner was given a 30-day notice to remove and dispose of the tanks and containerized oil.
The property at 2808 has been
non-detect since the summer of 2004. The
adjoining property northwest at a lower elevation likewise continues to be
non-detect. Southeastward the adjoining
properties are 2817 and
3. 2414 Haddon
The levels of MTBE at this site have been 11.7 ppb (Summer 04), 13.9 ppb (October 2004) and 10.3 (January 2005). There are no USTs on site.
There is no associated elevation of BTEX components in the supply well.
There is no open MDE investigation at this site.
4. 2808
The property at this site reveals an MTBE level of 31.5 ppb MTBE in January 2005. There is no onsite UST. There are no associated BTEX products.
4. Major
Sites with Historical Past Contaminations
a. Walkers Garage
The historical elevations of levels of MTBE and other petroleum products in the supply wells at Walkers Garage are well documented.
1990 Removal of three USTs (reported in the MD State registry as LUST (leaking underground storage tanks). The soil was grossly contaminated and allowed to remain in situ per MDE. Benzene 250 ppb Toluene 2 ppb, Ethyl benzene 2 ppb and total xylenes 40 ppb.
1994 MTBE in the monitoring well at 156.8 ppb and benzene at 382.7 ppb
1997 MTBE 12 ppb and monitoring well abandoned
2005 Investigation reopened. 3 monitoring wells at 25 feet. No detection of MTBE or other VOCs.
It can be assumed that the Walkers Garage site
sustained a liquid petroleum release at or prior to 1990. Unfortunately, the failure of the MDE to
require additional monitoring wells and off site domestic well sampling in 1990
has blurred the investigation at this site, essentially closing the Walker
Garage case. The contribution of this site to the recent high elevations of
MTBE on
b. Scarff’s Garage
1998 Removal
of two USTs. The soil was grossly
contaminated but allowed to remain in situ per MDE. Photovac microtip showed 400-3400 ppm of
total volatiles. Potable wells at 2419 (Dispasquales), 2800 (Exxon) and 2404
(Mamma Liberas) Baldwin Mill Road showed 11 ppb
5. Summary
1. Other than providing a compendium of potential contaminant sources, ExxonMobil has not investigation the probability or contribution of these alternative sites to the widespread contamination in the Uppercrossroads area.
2. The report of Other Potential Sources in Vicinity of the Site contains gross misinformation as evidenced by the inaccuracies in the document “Identified Heating Oil Properties”.
3.
The commonly
referred to case of
4.
The absence of
USTs and BTEX products in any of the supply wells in Haddon Hurst (2414) and
5.
The
contribution of Walkers Garage to the MTBE contamination on
6. Scarff's Garage had tanks that were out of service for 28 years prior to their removal. As this predates MTBE and there are no BTEX products in the affected wells, it can be concluded that Scarff’s is not a contributing site.
SECTION
10: FATE AND TRANSPORT
MODELING
1. ExxonMobil’s Fate and Transport Model
A major component of ExxonMobil’s assessment of the MTBE contamination in the Uppercrossroads study region was quarterly sampling of MTBE levels in private supply wells. Although initial sample procurement methods and testing regimes underestimated MTBE levels, revised methodologies resulted in the generation of large volumes of sensitive and specific data essential to the understanding of contaminant origin and migration.
In a meeting with the Maryland Department of the Environment n December of 2004, the GFA stressed the importance of fate and transport modeling to assist in the interpretation of the sampling data. Despite this request, the SARMar2005 failed to provide this vital information.
In the response of the MDE to
ExxonMobil on
Producing a fate and transport model is complex, requiring sophisticated analysis and incorporating many factors including groundwater flow, inorganic chemical analysis of well water along with an analysis of contaminant levels. There is no question that the contaminant spread is complex, especially in light of the fact that the MDE did not require an investigation of the community’s wells when the original releases at the ExxonMobil station were detected in 1990 and 1998. Secondary point sources originating from the primary (original) contamination makes interpretation extremely difficult especially as more than a decade has passed.
In the absence of these technologies and supplemental data, the GFA has undertaken the enormous task of analyzing and grouping MTBE levels with time and according to location.
2.
ExxonMobil Approach to Fate and Transport
Analysis
On the first page of the Executive Summary and page 20 of the SARMar2005, ExxonMobil states that only 205 of the 360 wells sampled from May 2004 through January 2005 showed no detection or estimated MTBE concentrations below the laboratory-reporting limit and were qualified with a “J”. A “J” value is estimated, as this level is less than the minimum accuracy standard prescribed by method and quality control protocol used. In other words, the gradations of the ruler are too big to make a fine measurement. Nonetheless, if the ruler was graded finer, it could be measured.
This statement is wholly irresponsible and reflects the cavalier approach of ExxonMobil to understanding the dynamics of the MTBE contamination and migration.
A contamination is a contamination. A “J” level is in no way equated with or comparable to a non-detectable level of MTBE. The occurrence, location timing and magnitude (whether a “ J” value or a measurable quantity), provides vital information to understating plume migration and potential contaminant sources. The presence of a “J” value indicates that a supply well is in communication with a contaminant path.
To ExxonMobil, the focus is only on the high levels of MTBE (greater than 20 ppb MTBE), rather than assessing all the contaminant levels to develop a clear picture of an area conceptual model.
3.
Data analysis
The community’s supply well MTBE levels were tabulated using varying geographic criteria in an effort to determine correlations and trends. While some of the comparisons are purely academic, others provide convincing arguments against the occurrence of primary community point source contaminations. Many of the comparisons clearly suggest a central “epicenter” contamination source.
As an initial example, a summary of private supply well contamination levels for January 2005 is presented in TABLE 5. (Data from SARMar05, Figure 4),
TABLE
Total Number of Wells in each concentration range with associated percentages.
|
|
MTBE |
LEVEL |
(ppb) |
|
|
|
|
Location |
ND |
J
(>ND
to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to <
20 |
³ 20 |
|
Within
½ mile radius |
44 |
32 |
43 |
7 |
8 |
10 |
Percent
|
30.6% |
22.2% |
29.9% |
4.9% |
5.5% |
6.9% |
|
Outside
½ mile radius |
84 |
36 |
50 |
4 |
1 |
1 |
Percent |
47.7% |
20.5% |
28.4% |
2.3% |
0.6% |
0.6% |
|
Total |
128 |
68 |
93 |
11 |
9 |
11 |
|
Percent |
40% |
21.3% |
29.1% |
3.4% |
2.8% |
3.4% |
Accordingly, as of January 2005, 60% of all wells in the study area showed evidence of MTBE contamination. Nearly 70 % of all wells within the one-half mile radius showed evidence of MTBE contamination.
This analysis provides some suggestion of a central focus of contamination. ExxonMobil dismisses any contribution to the contamination north of the ExxonMobil site. Indeed, the ½ mile sector above the station should provide a “dilutional” effect on the percent contamination within the ½ mile study radius if this contamination were from isolated spills versus a major leakage from a major commercial petroleum site. Just the opposite was observed.
An even more interesting comparison is between wells north of the ExxonMobil site and those South of the ExxonMobil site. (TABLE 6)
TABLE 6 A Comparison of MTBE levels in Supply wells North of the Exxon Site and South of the Exxon Site
Total Number of Wells in each concentration range with associated percentages
|
|
MTBE |
LEVEL |
(ppb) |
|
|
|
|
Location |
ND |
J
(>ND
to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to <
20 |
³ 20 |
|
North |
66 |
34 |
39 |
7 |
1 |
3 |
Percent
|
44.0% |
22.7% |
26.0% |
4.7% |
0.7% |
2.0% |
|
South |
63 |
34 |
54 |
4 |
7 |
8 |
Percent
|
37.1% |
20.0% |
31.8% |
2.4% |
4.1% |
4.8% |
It is remarkable that the distributions of contamination in the north and south are comparable. Given this similarity, it defies logic that in the area northward, the community caused the contamination while in the area southward, ExxonMobil is the major potential contaminant source. This would imply that communities northward produced a petroleum release of comparable magnitude to a release from a major commercial gasoline distribution site. If indeed the contamination northward is community based, then more than 50% of all properties in the county and state should show a similar distribution of MTBE contamination (based on the non detect rate in the northern sector). Of note, ExxonMobil refers to the HCHD map of reported MTBE levels outside the study area to justify their claim that MTBE contamination is widespread. The levels and distribution of MTBE as shown in the HCHD map do not even approach those seen in the northern sectors of the study area. Furthermore, the spatial distribution and levels of reported MTBE contaminations in the total one-mile radius surrounding the ExxonMobil site (outside the study area) are totally consistent with the ExxonMobil site being the source of these contaminations as well. To use a HCHD data as a comparative model, ExxonMobil would need to prove that an area, comparable to the current study area, showed the same sector type of MTBE distribution, especially in the sectors that ExxonMobil claims it did not contaminate. ExxonMobil has not conducted such a comparable study.
ExxonMobil has never conceded the fact that there are discreet areas of no contamination surrounding large areas of contamination. This would contradict their presumption of random community based contaminations. FIGURES 25, 26 and 27 provide shaded maps of MTBE distribution for August 2004, October 2004 and January 2005 in the defined study area. Areas shaded in grey reflect properties that are non-detect. Areas shaded yellow indicate contaminated properties with “J” values. In all three maps, north, east south and west, there are large areas without contamination around the periphery of the study area. In particular, these areas are;
1.
the northwestern properties on Franklin’s
2.
the southwestern properties bordering the highly
contaminated
3. the northeastern properties of Delmar Farms, and
4.
the southeastern properties along
The spatial distribution of these non-detect properties is a strong argument against the random community based contamination theory of ExxonMobil.
It is interesting to note the
presence of a northeast to southwest contamination corridor as shown in a
redraw of the October shaded map. (FIGURE 28)
Although there are areas of non-detect within the corridor, contaminant
migration is dependent on many factors (casing depth, topographic elevation,
bedrock fracture depth, dip angle etc) that would account for these
interdispersed non-detect supply wells.
Contamination in a bedrock system does not follow a totally predictable
route and characteristically shows skip areas with random appearing spikes of
contamination due to preferential routes of transport in the fracture
fabric. It is interesting that the
“barrier” property between Cross Country Estates and the corridor (
Using ExxonMobil’s hypothesis that there is no “boundary” to the contamination, i.e. contaminations are random, it can only be assumed that individuals in these peripheral non-detect areas are more adept in the safe handing of petroleum products than individuals in the more central study area, or that ExxonMobil’s hypothesis is misguided.
c. Sector
Analysis
The following tables provide a detailed analysis of MTBE level trending from August 2004 to January 2005. The data was analyzed using three major comparisons
1. Entire Study Area
2. Inside the ½ mile radius versus outside the ½ mile radius
3. Northern sectors and southern sectors.
Each data set is accompanied by a table showing the change in percent distribution for comparative analysis.
1.
Entire Study Area
As seen in TABLES 7 and 8,
there is a 10% decline in non-detect levels in the study area over the two
major time intervals from August 2004 to October 2004 and October 2004 to
January 2005. The decline occurred at a
steady rate of 5% per quarter. This
decline was accompanied by an increase in “J” values to 8% with a more accelerated
increase in “J” values in the first quarter.
There was a corresponding small increase in the “green” range levels (³ 0.5 to < 5
ppb MTBE) referred
to hereon as “green” levels. The blue
levels, (³ 5 to < 10
ppb MTBE), orange levels (³ 10 to < 20 ppb MTBE)
and red levels (³ 20 ppb MTBE) showed
evidence of inter-group shifting. TABLE
9 presents a graphic color
representation of the MTBE level changes among the blue, orange and red level
groups along with their community locations to gain a better understanding of
this shifting.
TABLE 7 Percent Distribution of MTBE levels in the Entire Study Area
|
|
MTBE % |
|
|
|
|
|
MONTH |
ND |
J
(>ND to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to <
20 |
³ 20 |
|
August |
50.0 |
13.3 |
26.8 |
3.4 |
3.4 |
3.1 |
|
October |
44.8 |
18.3 |
26.3 |
4.1 |
3.5 |
2.9 |
|
January |
40.0 |
21.3 |
29.1 |
3.4 |
2.8 |
3.4 |
TABLE 8 Change in Percent Distribution of MTBE in the Entire Study Area
PERIOD |
ND |
J
(>ND to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to <
20 |
³ 20 |
|
D August to October |
-5.2 |
5.0 |
-0.5 |
0.7 |
0.1 |
-0.2 |
|
D October to January |
-4.8 |
3.0 |
2.8 |
-0.7 |
-0.7 |
0.5 |
|
TOTAL |
-10.0 |
8.0 |
2.3 |
0 |
-0.6 |
0.3 |
TABLE 9 Conversion of MTBE Levels in “Higher”
Contaminant Levels
|
MTBE
(ppb) |
³ 5 to < 10 |
³ 10 to < 20 |
³ 20 |
|||
|
|
|
|
|
|||
|
RED
CONVERSION |
|
|
|
|
||
|
|
|
MONTH |
|
|
|
|
|
LOCATION |
August |
October |
January |
|
||
|
|
|
|
|
|
|
|
|
Scarff/165 |
R |
O |
R |
|
||
|
ORANGE
CONVERSION |
|
|
||
|
|
|
MONTH |
|
|
|
LOCATION |
August |
October |
January |
|
|
|
|
|
|
|
|
Cross
Country |
O |
B |
|
|
|
Delmar |
|
O |
B |
|
|
|
|
B |
O |
|
|
|
|
O |
R |
|
|
|
|
|
O |
R |
|
|
|
|
O |
R |
|
BLUE
CONVERSION |
|
|
||
|
|
|
MONTH |
|
|
|
LOCATION |
August |
October |
January |
|
|
|
|
|
|
|
|
Cross
Country |
O |
B |
|
|
|
Delmar |
|
O |
B |
|
|
|
|
B |
O |
|
|
|
G |
B |
G |
|
|
|
G |
B |
G |
|
|
|
G |
B |
J |
|
|
|
|
B |
G |
|
|
Haddon
Hurst |
B |
G |
|
|
|
Hunting
Ridge |
B |
G |
B |
|
|
Hunting
Ridge |
G |
B |
|
|
|
Hunting
Ridge |
|
G |
B |
|
|
152 East |
|
|
B |
G |
|
Scarff |
|
|
B |
G |
From this table it can be see that
there was up shifting of levels mostly around
In order to get a more
comprehensive understanding of the changes in levels with time, FIGURE 29 and FIGURE 30 present
single perspective views from a 3-D rotational computer generated model of
change in MTBE Levels over the two time quarters from Summer 2004 to October
2004 and October 2004 to January 2005. (3-D rotational graphics are
available upon request.)
As an example, the properties along
This approach to data analysis is the only way to clearly visualize plume migration. ExxonMobil’s approach to focus solely on the few properties with high MTBE levels is cursory and uninformative.
Over the past year, the GFA has insisted that the majority of new contaminations were occurring in the periphery of the study area. Again this would not be consistent with ExxonMobil’s hypothesis of random contamination. To assess new contaminations, FIGURES 31 and 32 provide locational plots of new contaminations as a function of the quarterly time periods. As can been seen from these graphs, it is unequivocal that the contamination is spreading to the periphery in all quadrants in the now defined study area. Occurrence of contamination is of course dependent on many factors including distance from the contaminant source, secondary point sources of contamination from primarily affected wells, and the location and elevation of the water bearing fractures. This accounts for the presence of “J” (and non-detect) levels of MTBE occurring within areas that have already exhibited MTBE contamination. However, the fact that new contaminations are occurring in all quadrants makes the “epicenter” model of contamination the most plausible explanation. Consistent with the outward migration from a central focus is the fact that there is clustering of J-values in the peripheral areas. This is most pronounced in the following areas;
- The east side Cross Country estates and along Baldwin Mill road northward
- The end of MD 152 eastward and northward toward Choate road
- Franklin’s
In conclusion, MTBE contamination in the study area is bounded by a periphery of non-detect wells and the contamination is spreading outward in all quadrants of the study area. This makes a central source of contamination with distant migration the most likely site conceptual model. Furthermore, the fact there are a large number of new contaminations appearing with the same low values of MTBE, would suggest that every one of the new point sources are leaking the same low amount of MTBE at the same time. This proposition is simply untenable.
2.
Inside versus Outside the ½ mile radius
As can be seen in Tables 10, 11, 12 and 13 the most important fact to note is that the levels of non-detect MTBE inside the ½ mile radius are significantly less than the non-detect levels outside the ½ mile radius. Again, this argues against the random contamination theory of ExxonMobil. This fall in non-detect levels, accompanied by a predominate rise in “J” values northward and “J” and green levels southward, is a strong indicator that new site contamination is due to a progression of contamination rather than new random primary point source contaminations. Levels in the blue, orange and red zones showed minimal group inter-shifting.
TABLE 10 Percent Distribution of MTBE levels Inside the ½ Mile Radius
|
|
MTBE % |
|
|
|
|
|
MONTH |
ND |
J
(>ND to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to <
20 |
³ 20 |
|
August |
39.4 |
14.2 |
29.0 |
3.9 |
7.1 |
6.5 |
|
October |
34.2 |
20.6 |
27.1 |
5.2 |
7.1 |
5.8 |
|
January |
30.6 |
22.2 |
29.9 |
4.9 |
5.6 |
6.9 |
TABLE 11 Change in Percent Distribution of MTBE Inside the ½ Mile Radius
PERIOD |
ND |
J
(>ND to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to < 20 |
³ 20 |
|
D August to October |
-5.2 |
6.4 |
-1.9 |
1.3 |
0.0 |
-0.7 |
|
D October to January |
-3.6 |
1.6 |
2.8 |
-0.3 |
-1.5 |
1.1 |
|
TOTAL |
-8.8 |
8.0 |
0.9 |
1.0 |
-1.5 |
0.4 |
TABLE 12 Percent Distribution of MTBE Outside the ½ Mile Radius
|
|
MTBE % |
|
|
|
|
|
MONTH |
ND |
J
(>ND to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to <
20 |
³ 20 |
|
August |
58.3 |
12.6 |
25.1 |
3.0 |
0.5 |
0.5 |
|
October |
53.8 |
16.3 |
25.5 |
3.3 |
0.5 |
0.5 |
|
January |
47.7 |
20.5 |
28.4 |
2.3 |
0.6 |
0.6 |
TABLE 13 Change in Percent Distribution of MTBE Outside the ½ Mile Radius
PERIOD |
ND |
J
(>ND to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to <
20 |
³ 20 |
|
D August to October |
-4.5 |
3.7 |
0.4 |
0.3 |
0.0 |
0.0 |
|
D October to January |
-6.1 |
4.2 |
2.9 |
-1.0 |
0.1 |
0.1 |
|
TOTAL |
-10.6 |
7.9 |
3.3 |
-0.7 |
0.1 |
0.1 |
3.
Northern and Southern Sectors
The data comparing MTBE levels over the quarterly assessment period for the northern and southern sectors is presented in TABLES 14, 15, 16 and 17. The most remarkable feature comparing the northern and southern sectors is that the distribution of MTBE contamination is comparable over the two quarter time periods. Again this argues strongly against the random contamination model of ExxonMobil. The effect of a major release with contamination migrating south to southwest is mirrored in the distribution of MTBE contamination spreading northward.
TABLE 14 Percent Distribution of MTBE North of the ExxonMobil Site
|
|
MTBE % |
|
|
|
|
|
MONTH |
ND |
J
(>ND to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to <
20 |
³ 20 |
|
August |
55.5 |
14.0 |
22.0 |
4.3 |
2.4 |
1.8 |
|
October |
47.3 |
19.4 |
23.6 |
6.7 |
1.2 |
1.8 |
|
January |
44.0 |
22.7 |
26.0 |
4.7 |
0.7 |
2.0 |
TABLE 15 Change in Percent Distribution of MTBE North of the ExxonMobil Site
PERIOD |
ND |
J
(>ND to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to <
20 |
³ 20 |
|
D August to October |
-8.2 |
5.4 |
1.6 |
2.4 |
-1.2 |
0.0 |
|
D October to January |
-3.3 |
3.3 |
2.4 |
-2.0 |
-0.5 |
0.2 |
|
TOTAL |
-11.5 |
8.7 |
4.0 |
0.4 |
-1.7 |
0.2 |
TABLE 16 Percent Distribution of MTBE South of the ExxonMobil Site
|
|
MTBE % |
|
|
|
|
|
MONTH |
ND |
J
(>ND to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to <
20 |
³ 20 |
|
August |
45.3 |
12.6 |
31.1 |
2.6 |
4.2 |
4.2 |
|
October |
42.5 |
17.2 |
28.7 |
1.7 |
5.7 |
4.0 |
|
January |
37.1 |
20.0 |
31.8 |
2.4 |
4.1 |
4.7 |
TABLE 17 Change in Percent Distribution of MTBE South of the ExxonMobil Site
|
PERIOD |
ND |
J
(>ND to <0.5) |
³ 0.5 to <
5 |
³ 5 to < 10 |
³ 10 to <
20 |
³ 20 |
|
D August to October |
-2.8 |
4.6 |
-2.4 |
-0.9 |
1.5 |
-0.2 |
|
D October to January |
-5.4 |
2.8 |
3.1 |
0.7 |
-1.6 |
0.7 |
|
TOTAL |
-8.2 |
7.4 |
0.7 |
-0.2 |
-0.1 |
0.5 |
4. Summary
ExxonMobil provided a single page in their March 22nd report devoted to an analysis of fate and transport to help construct a conceptual site model. Although the discussion presented above did not approach the true level of a fate and transport analysis, it provided the best available interpretation of large volume of MTBE level data collected. The cursory analysis provided by Exxon Mobil and the lack of emphasis it placed on this valuable analytic tool reflects their simplistic and biased approach to viewing only one hypothesis.
- The presence of a low level distribution of MTBE uniformly in all four quadrants of the study area compared to the 4 log level of MTBE at the ExxonMobil site is convincing evidence of a central epicenter source of contamination affecting the entire aquifer. The expected contamination pattern of a highly fractured and deep bedrock aquifer over time is a widespread amorphous distribution of the contaminant (aquifer dilution) with spiking levels at random points from the epicenter (preferential migration in fracture paths). This is exactly what is being observed in the Uppercrossroads aquifer.
- ExxonMobil views “J” values as insignificant rather than as a contamination that it truly represents. For this reason, ExxonMobil has underestimated the extent of the contamination in the Uppercrossroads area.
- ExxonMobil’s claim that MTBE contamination is random and widespread as supported by the HCHD contamination map is flawed as the percent distributions found in the study area far exceed those found in the HCHD map. (HCHD map is not a regional study of MTBE level reporting but a random reporting of MTBE levels from owners of private supply wells outside the circumference of the study area. As the reported data are random and sporadic, they cannot be scientifically compared to the ExxonMobil study area. Regardless of this point however, if one looks at the level of contamination in the one mile radius surrounding the study area, the levels are consistent with ExxonMobil being the source of their contamination.)
- Graphic illustrations support the contention that there is a “boundary” of non-detect levels in all four quadrants surrounding the study area. The presence of a “boundary” does not support ExxonMobil’s contention that contaminations are random.
- Graphic illustrations demonstrate that “J” values are a prominent interface between non-detect levels. This implies that contamination is progressive and not random.
- Fate and Transport modeling shows a migration of MTBE contamination toward the periphery of the study area. No conclusions can be drawn until full fate and transport modeling is complete.