View the Powerpoint here:

Poster for Post Closure Analysis in Power Point

Michael Sequino (msequino@directionaltech.com) (Directional Technologies, Inc., Wallingford, Connecticut, USA), George Losonsky, Ph.D., P.G. (glosonsky@cox.net) (Losonsky & Associates, Inc., Baton Rouge, Louisiana, USA), James R. Oppenheim (jroppenheim@sunocoinc.com) (Sunoco, Inc. Lester, PA, USA) and Kevin W. Martin, CHMM (km@acquaterra-tech.com) (Aquaterra Technologies, Inc., West Chester, PA USA)

Introduction

A major oil storage facility (MOSF) in New York was successfully remediated in the late 1990’s using a horizontal well air sparge system. The site is closed and no remedial efforts have been required since 2003. The subject property is known as Griffith Terminal, (“site”) located on the south side of Route 352 in Big Flats, Chemung County, NY. Elevated concentrations of dissolved phase hydrocarbons were detected in site groundwater monitoring wells in 1995. Non-aqueous phase liquids (NAPL) were subsequently discovered in the vicinity of a six-inch diameter petroleum product pipeline that passed under Route 352. Product inventory records indicated that approximately 50,000 gallons of product were released between 1994 and 1995, including mixed oils, gasoline, No.2 fuel oil and kerosene. The plume related to the product release encompassed approximately 15 acres. Sensitive receptors include residences with private wells, the Chemung River, and a quarry with a pond. The objectives were to prevent exposure to sensitive receptors, control the plume by preventing migration beyond the existing plume boundaries, and remediate the plume using a logical, phased implementation of the remediation program.

Geographic Setting

Site Location and Description – The subject property is known as Griffith Terminal (“site”) encompassing approximately 20 acres, located on the south side of Route 352 between the intersections of South Corning Road and Route 17 in Big Flats, Chemung County, NY. The site’s elevation is approximately 900 feet above mean sea level. The surrounding area has low topographic relief, with a shallow slope southeast toward the Chemung River, which flows adjacent to the site along the south and southeastern property boundaries. Surrounding land use includes residential properties, light commercial properties and an industrial property (Figure 1).

Description of Facility

The site is a major oil storage facility (MOSF), containing several slab-on-grade buildings, a product loading rack and six aboveground bulk petroleum storage tanks (AST) located in the approximate center of the site. The site also contains two 30,000-gallon bulk propane ASTs located in the southern portion of the site on what used to be known as the Stueben property. Approximately 15% of the property is covered by impermeable surfaces while the remainder is covered with gravel or turf (Figure 1).

Site Hydrogeology

The Corning Elmira Primary Aquifers consisting of highly permeable gravel deposits exist beneath the site. A constant rate 24-hour pumping test performed at a constant rate of 14 gallons per minute revealed calculated hydraulic parameters including transmissivity of 13,000 gallons per day per foot, storativity of 0.02 and hydraulic conductivity of 210 gallons per day per square foot, which is equal to 28.1 feet per day.

Groundwater depth varies between 9 and 17 feet across the site. Historical groundwater monitoring well data indicate that the site water table is subject to seasonal fluctuations of approximately four feet. Groundwater flow is predominantly southeast toward the Chemung River (Figure 2).

The site bedrock consists of the West Falls Group comprised of Devonian shale and siltstone. Overburden soils consist of approximately 65 feet unconsolidated glacial and fluvial deposits. Logs for groundwater monitoring, recovery, soil vapor extraction and sparge wells installed post-release indicate that silt and fine-grained sands are present at the surface, grading to well-rounded gravel and sand at 10-15 feet below ground surface (bgs). The sand and gravel layer is approximately 15-feet thick and is underlain by coarse to medium sand. A confining layer of silt and gravel is present at approximately 45 feet bgs and a second confining layer of glacial till overlies the bedrock (see Figure 3).

The recharge and product yield obtained from monitoring wells is highly variable due to the presence of zones of highly transmissive stringers or buried erosional channels which may or may not intersect the boring depending on location. It is impossible to predict the location of such zones when advancing vertical borings.

Site History

Elevated concentrations of dissolved phase hydrocarbons were detected in site groundwater monitoring wells in May 1995. This discovery triggered an investigation and non-aqueous phase liquids (NAPL) were subsequently discovered in the vicinity of a six-inch diameter product delivery pipeline that crossed under Route 352 during July 1995 (see Figure 1). The product delivery pipeline reportedly exhibited corrosion when inspected. A NAPL sample was collected and analyzed and found to be less than one year old. The greatest NAPL thickness of 2.3 feet was detected in monitoring well MW-21 in 1995. It was estimated that at least 50,000 gallons of petroleum products were released to the subsurface between 1994 and 1995. Product lost during this period included mixed oils, gasoline, No. 2 fuel oil and kerosene. An eight-inch diameter pipeline was subsequently installed to replace the six-inch pipeline.

Remedial Actions

A series of remedial actions have been implemented, including excavation, soil vapor extraction, air sparging, oil skimming and total fluids extraction using vertical wells, air sparging with horizontal wells, and wellhead treatment of water supply wells. Only the wellhead treatment is ongoing.
A combination of floating skimmers, an automatic product-only pump, an auto-skimmer, hand bailing and vacuum enhanced pumping were employed to recover NAPL from July 1995 to August 1995. Sixty tons of soil were excavated in 1996. A total phase extraction system was also installed, connected to monitoring wells MW-10, 15, 16, 17, 33 and 42 and operated from 1999 and 2000 during low water table conditions (late summer through early fall). Approximately 19,000 gallons of SPH was recovered using these combined technologies.

Conceptual Site Model

The primary means of groundwater and contaminant mobility is through buried channels of transmissive sand and gravel. Boring logs show sand and gravel lenses interbedded with less permeable silty and clayey soils. A poorly cemented bed of gravel, sand and silt occurs at 10 to 15 feet bgs and appears to act as a barrier to vertical movement of soil vapor.
The groundwater gradient is eastward toward the quarry. Groundwater fluctuates about 4 feet, creating a substantial smear zone. Free product has collected in the smear zone. This makes the subsurface amenable to air sparging and soil vapor extraction. In addition the site has a dissolved groundwater plume.

Vapor phase contamination is present throughout the area defined as underlain by free product. Lateral migration of soil vapor contamination is limited. Soil vapor contamination does not appear to extend beyond the dissolved contaminant plume.

Remedial Objectives

The plume related to the product release initially encompassed approximately 15 acres (see Figure 2). The most significant sensitive receptor is the Chemung River, adjacent to and hydraulically downgradient of the site.
Private commercial and residential wells have documented petroleum impact, including toluene, MTBE, and ethylbenzene, and are operating with point of entry treatment systems.
The site remediation objectives were to: 1) prevent impacts beyond known plume boundaries; and 2) aggressively remediate the plume in a step-wise phased approach.
• Limit the migration of product and dissolved plume to the potential receptors including potable drinking water wells in the vicinity, the quarry to the east, and the Chemung River.
• Remove free product
• Remediate soil and groundwater to established site-specific cleanup levels required by NYSDEC

Remediation Strategy

The horizontal wells were installed as part of a large field-scale program to address the groundwater plume by removing source area contaminant mass, decreasing the area of the plume that exceeds regulatory standards, promoting natural attenuation. Plume stability is the expected result of attainment of these remediation goals. Engineering analysis revealed that a total of 129 vertical wells coupled with 23 blowers would be required to remediate the entire affected area via bio-sparge technology. Analysis further revealed that only 12 horizontal bio-sparge wells coupled with 6 blowers would be required to remediate the same area. As a gross approximation the time and cost of a remedial project such as this is inversely proportional to the amount of well screen that can be placed in contact with contaminated media. For this project, the vertical well system would provide only 272 feet of well screen versus 2,650 feet of well screen for the horizontal well system. On a generalized basis it can be stated that horizontal well installation costs are greater than vertical wells; however, when assessing total life cycle costs, horizontal well systems are substantively less expensive than their vertical counterparts. In summary, the horizontal bio-sparge well option was selected on the basis of (1) large plume area, (2) expected rapid remediation versus vertical well configuration and (3) life cycle cost.

Pilot testing included conventional pump-and-treat, total fluids extraction, vacuum enhanced pump and treat, product only recovery, and soil vapor extraction. Soil vapor extraction pilot testing indicated up to 60 feet of radial influence at 10 inches of mercury, which is half the vacuum required for total fluids recovery.

Air sparge pilot testing of vertical wells indicated that dissolved oxygen (D.O.) increased significantly above ambient levels up to 44 feet away from the sparge wells. Pressure influence was detected up to 72 feet away.

Interim product recovery was achieved with direct pumping of pure product, and dual diaphragm pump recovery. Pilot testing for product recovery included conventional pump-and-treat, total fluids extraction, soil vapor extraction, and air sparging.

Horizontal Air Sparge Well System

Two networks of horizontal bio-sparge wells were installed during site remediation; the first in 1996 and the second in 1997. A perimeter network of six horizontal bio-sparge wells designed to protect the Chemung River and an adjoining quarry was installed in 1996. A second goal of installing the perimeter horizontal bio-sparge wells was to determine the constructability and effectiveness of the horizontal remediation well technology in advance of designing and installing the “Core” plume area horizontal bio-sparge well network of 13 horizontal wells in 1997. The six-well perimeter system included horizontal wells HSW-2, HSW-3, HSR-1, HSR-2, HSQ-1 and HSQ-2 (Figure 4). The core area horizontal bio-sparge wells installed in 1997 included HSW-1, HSF-1, HSF-2, HSF-3, HST-2 and HST-3. The system operated from 1996 to 2003.

Performance Evaluation

NAPL decreased from a thickness of 2.3’ in 1995 to essentially a sheen by 2003. Measured by volume, bio-sparging eliminated at least 30,000 gallons of NAPL. Soil samples collected in the release area circa 1996 uniformly and significantly exceeded regulatory criteria. Soil samples collected from the same approximate locations in the release area circa 2005 were either “non-detect” or below actionable clean-up criteria.

The estimated volume of spilled free product was 50,000 gallons. The original remediation design expected 37, 500 gallons to be recoverable. Vertical recovery wells extracted over 18,000 gallons by 1995, using various product recovery methods and by soil vapor extraction. Product recovery methods included direct product recovery, floating skimmers, and hand bailing.

The area with separate-phase hydrocarbons before the installation of the horizontal well system covered an area of 7 acres, shown in Figure 5 with a dashed outline. Product thicknesses ranged from a sheen to 0.9 feet. After the horizontal well air sparge system was shut off, no separate-phase hydrocarbons remained.

The dissolved hydrocarbon plume experienced a dramatic size reduction from 1995 to 2003 (see Figure 6). Analysis of changes in concentrations of constituents of concern over time was used to assess the effectiveness of the remediation system, and to evaluate plume stability. The slope of the natural log as a function of time describes attenuation characteristics. Decreasing concentration trends occurring both in the source area and in the downgradient plume would suggest that active remediation and natural attenuation is occurring. Increasing trends in both source and downgradient areas would indicate that the rate of contaminant loading into the
aquifer exceeded the rate of natural attenuation. Plume stability is typically indicated by unchanging concentrations.

Groundwater contaminant concentration trend analysis demonstrated steady improvement from 2000-2005. The trend analysis reveals decreasing trends in most wells at the site, suggesting that active remediation and natural attenuation has reduced the source mass and decreased the footprint area of the dissolved plume (Figure 7). In one exception, concentrations of ethylbenzene and xylene are increasing, however, this trend does not account for high concentrations that would be in a well that was excluded from the sampling and analysis because it contains separate-phase hydrocarbons.

Conclusions

Conventional wisdom says that bio-sparging is ineffective for remediating NAPL because it is not feasible to introduce sufficient quantities of oxygen across the source mass to sustain bacteria populations that use the hydrocarbons as a nutrient source. This project proves that conventional wisdom is not always reliable. At this site, bio-sparging with horizontal wells eliminated at least 30,000 gallons of NAPL at this site. NAPL bio-sparging can be successfully implemented when applied using properly designed and installed horizontal remediation wells. A horizontal well system provide significantly greater source mass contact area and oxygen quantity compared to an equivalent system of vertical wells. Horizontal remediation well systems can be very effective and less expensive at remediating large expanse plumes than vertical well systems. In addition, this project demonstrates that horizontal sparge wells can be used as very effective barriers to plume migration.

KSU Conference April 2011 – contains updated data from previous conference

Old Version:
Battelle Remediation of Chlorinated and Recalcitrant Compounds May 2010

Paper E-037, in: Bruce M. Sass (Conference Chair), Remediation of Chlorinated and Recalcitrant Compounds—2008. Proceedings of the Sixth International Conference on Remediation of Chlorinated and Recalcitrant Compounds (Monterey,CA; May 2008). ISBN 1-57477-163-9, published by Battelle, Columbus, OH, www.battelle.org/chlorcon.

George Losonsky, Ph.D., P.G. (Losonsky & Associates, Inc., Baton Rouge, Louisiana)
Michael J. Sequino (Directional Technologies, Inc., North Haven, Connecticut)

ABSTRACT:

Environmental injection wells are used to deliver a variety of fluids into
porous media for the purpose of in situ soil and groundwater remediation. Injection wells
are used to deliver nutrients for enhanced bioremediation; various amendments for bioaugmentation; and potassium or sodium permanganate, or mixtures of ozone and peroxide for chemical oxidation. Remediation agents are injected in both liquid and gas states, and they have a wide range of physical properties, such as temperature, density and viscosity. Long screens are often specified in order to deliver efficiently into large soil or aquifer volumes. The screens may be vertical, horizontal, or inclined, depending on hydrogeology and plume geometry. Some screens have variable inclinations, such as horizontal screens that continue into the riser section. Complete and efficient sweep of the targeted volume of porous material is ensured if long screens are engineered to provide
uniform injection of the fluid.

Mechanical requirements for achieving uniform fluid injection center around three
design elements: pipe flow, orifice flow, and porous media flow, which are used to define
the three stages of the injection process: generation of positive pressure by the treatment
system at the ground surface; establishment of a pressure gradient through the conveyance
pipe and well screen; flow through the pipe; flow through the screen slots; and flow
into the formation. The flow process depends on physical specifications of the pipe and
screen, and on formation properties. Pipe and screen specifications include pipe material
and roughness factor; inner and outer pipe diameter; pipe inlet and outlet elevations;
length of conveyance pipe, well riser, and screen; screen slot dimensions and orifice coefficient; and open area distribution along the screen. Formation characteristics include
pneumatic or hydraulic conductivity of the formation; and hydraulic head elevation and
gradients. Sensitivity analyses provide an understanding of the relative influence of these
parameters, and guidance in focusing the design effort.

Design of an injection system accommodates boundary conditions set by the treatment
system capacity, well material properties, drilling conditions, and anticipated response
of the aquifer or soil zone to the injection process. Monitoring of wellbore and
aquifer conditions during well drilling and development provides data about subsurface
conditions that can be used to adjust and optimize the operation parameters for the injection
system. A series of examples illustrates typical and end-member cases of injection of
liquids and gases with different fluid properties, into wells made of commonly used pipe
materials, and into various geologic media.

INTRODUCTION

Efficient delivery or extraction wells are the lifeblood of effective in situ remediation
systems. A successful screen design achieves the necessary balance of screen dimensions
and operational parameters, and is appropriately tuned to material properties of the various
media involved, to achieve uniform delivery or extraction of sufficient volumes of
fluid or gas to effectively remediate the target zone. Fluids and gases used in such systems
include:

• Air
• Water
• Bioamendment solutions
• Chemical oxidant solutions
• Soil gas
• Groundwater and non-aqueous fluids

This paper examines the effect of key design parameters on the efficiency of well
screens. Although a well screen of any length should be engineered to be effective, the
design parameters discussed in this paper are especially relevant to long screens. The
need for proper screen design applies equally to vertical wells, horizontal wells, and inclined
wells.

DESIGN PARAMETERS

The size of the zone of influence of a fluid injection or soil vapor well is directly
proportional to the flow rate. For a fluid injection well, the zone of influence is proportional
to the rate of delivery of fluid into the soil formation. For a soil vapor extraction
well, the zone of influence is proportional to the extraction rate of air out of the formation.
As the flow rate increases, so does the hydraulic head gradient from an injection
well into the formation, or the negative pressure gradient between a soil vapor extraction
well and the formation. The efficiency of the fluid delivery or vapor extraction system
depends on the balance of pressures and flow rates applied at the surface, and resistance
to flow along the entire flow path, which is influenced by a series of material characteristics
encountered by the fluid or vapor, including:

• Roughness coefficient of the pipe
• Slot apertures
• Hydraulic radius of slots or orifices
• Depths of screen slots (pipe wall thickness)
• Distribution of slots along the screen (percent open area of screen)
• Hydraulic or pneumatic conductivity of the formation
• Skin factor of the well

The skin factor of the well is a measure of formation damage that develops adjacent
to the well screen, which reduces the well’s effectiveness in delivering or extracting fluid
or gas. By contrast, operation of a well can cause beneficial changes in formation adjacent
to the well screen. This process is causes the formation of a natural filter pack. This
self-development of a well depends on several key formation parameters:

• Sphericity of the grains comprising the formation
• Average diameter of the grains
• Grain size distribution, or grain sorting

During the operation of a fluid injection well or a soil vapor extraction well, the pressure
gradient within the formation depends on a series of physical parameters, many of which are
interrelated. These parameters include the material characteristics listed above, and:

• Viscosity of the injected fluid or extracted gas
• Pressure gradient along the well screen
• Flow rate of air through the well screen
• Reynolds number
• Diameter of the well
• Length of the well screen
• Length of the riser pipe
• Length and diameter of conveyance piping
• Depth of the well

The delivery of fluid into the formation or extraction of vapor out of the formation
sets up a pressure gradient from the well to the edge of the well’s zone of influence. This
pressure gradient drives fluid or vapor transfer between the well and the formation. The
steeper the pressure gradient, the faster groundwater or soil vapor will be replaced within
the pore spaces of the treatment zone. Pore fluid around an injection well becomes enriched
with the constituents of the injected fluid. Soil gas around a soil vapor extraction
well is removed from the formation by flowing into the extraction well. Various degrees
of mixing occur in both the fluid injection and soil vapor extraction processes. Treatment
of the target zone is expedited when the fluid or vapor phase is driven to move quickly
through the well’s zone of influence. Pressure gradients around the well directly impact
the rate of treatment of the target zone.

PRESSURE GRADIENT AND INJECTION RATE

Four plots of the pressure gradient as a function of the unit injection rate along the
well screen illustrate the effect of injection rate on the pressure gradient that drives the
rate of treatment of a target zone. Two plots are shown in Figure 1, representing different
open area values.

The pressure gradient is normalized with respect to the gradient. The plots show a direct
proportional increase of pressure gradient with increasing injection rate. A three-fold increase
in injection rate produces a three-fold increase in pressure gradient. The effect of screen open
area is to increase the rate at which pressure increases with increasing flow rate.
Figure 1 represents flow at low Reynolds Number values. Flow under these conditions
is considered to be laminar. Liquid injection can be designed to maintain laminar
flow through the well screen, preventing encrustation of the well. By contrast, soil vapor
extraction wells commonly develop turbulent flow conditions. The two plots in Figure 2
represent the increase in pressure gradient in response to increasing flow rates under flow
conditions with relatively high Reynolds Number values. The plots show a non-linear
increase of the normalized pressure gradient with increasing flow rate. Under these flow
conditions, the pressure gradient increases by several orders of magnitude while the flow
rate increases by integer amounts. The two plots in Figure 2 represent screen open areas
that differ by a factor of two, and illustrate the increased sensitivity of flow rate to the
pressure gradient with decreasing open area of the well screen.

SCREEN OPEN AREA AND FLOW DISTRIBUTION

Besides showing the effect of injection rate on the pressure gradient that drives the
rate of treatment, the plots above also illustrate the sensitivity of fluid injection well effectiveness to the open area along the well screen. The following example illustrates how
the efficiency of a liquid injection well is influenced by small changes in the percent open
area along the well screen. Figure 3 shows the pressure distribution and cumulative flow
rate along a properly designed, 230-foot long well screen. The left hand side of Figure 3
shows an even pressure distribution along the screen, while the right hand side shows that
the cumulative injection rate reaches 100 percent at the end of the well screen.
Small changes in open area can be detrimental to the effectiveness of the well. Figure 4 shows the impact along the well screen of a ten percent increase in open area over the optimal amount. The cumulative injection rate reaches 100 percent about 20 feet before the end of the well screen. This means that 20 percent of the screen is ineffective.

Pressure drop along a well screen may become excessive, as shown in Figure 5, if the
well specifications, such as well diameter, length, depth, and screen design do not match
the operational requirements, such as injection rate and injection pressure, and do not
take into consideration material properties of the well pipe and target formation. The red
plot on the left hand side of Figure 5 represents pressure drop, measured in feet of water.
The lower plot represents background pressure in the formation. Excessive pressure drop
can lead to a declining injection rate along the well screen, as shown on the right hand
side of Figure 6. The same engineering principles apply to the design of extraction well
screens. For example, an imbalance of vacuum and flow rate against the well specifications
and formation characteristics can lead to a large pressure drop and unacceptably
declining extraction rates along a soil vapor extraction well screen. Problems with high
pressure drop are more common under typical operating conditions for soil vapor extraction
wells than they are for fluid injection wells.
ZONED SCREENS
Problems associated with high pressure drop in an injection or extraction well can be
resolved by adjusting the well diameter or by progressively changing the open area along
the well screen. Typically, the open area is increased in 100-foot intervals toward the distal
end of a well screen, which is the point furthest away from the source or blower. This
is referred to as zoning the screen. Open area zones are designed so that the average injection
or extraction rate is constant among the zones, as illustrated in Figure 6 for a soil
vapor extraction well.

INCLINED SCREENS

Changes in screen elevation also affect flow distribution along a horizontal well. The
example in Figure 7 shows changes in fluid injection rate along the screen of a liquid injection
well with slots extending partially into the curved riser section of the well. The
screen represented on the left hand side of Figure 7 has a constant open area along the
entire screen, including the shallower portion of the screen on the left. The screen represented
on the right hand side of Figure 7 is zoned to adjust to the changing screen
elevation. As in the previous example of a soil vapor extraction well, the liquid injection
well can be designed to achieve even distribution of flow along the well by dividing the
screen into several zones with progressively changing open areas. The liquid injection
screen in the example below is divided into three sections, or open area zones.

PIPE DIAMETER

The sensitivity of pressure and air flow distribution is enhanced at high vacuum, high
air flow rate, and small well diameter. The latter is illustrated by the comparison of extraction
rates along two wells with different diameters, in Figure 8. The distribution of the
extraction rate along a 2-inch diameter screen is shown on the left hand side, and the extraction
rate along a 4-inch diameter screen is shown on the right hand side of Figure 8.
The flow rate is essentially constant in the 4-inch well, but decreases away from the
blower in the 2-inch well.

CONVEYANCE PIPING

Finally, the importance of taking into account the length or diameter of conveyance
piping between the blower or fluid source and the wellhead is illustrated by the Figure 9,
which shows the cumulative extraction rate for a soil vapor extraction well which was
designed without taking into account several hundred feet of conveyance piping. As a
result, the well produces only 70 percent of the air flow required by the design. The low
air flow rate will result in diminished performance of the remediation system. The design
must also account for changes in pipe diameter or pipe materials. Compared to the effect
of pipe diameter, the effect of pipe materials is small.
2″ SVE Well with 350′ Conveyance

CONCLUSION

Proper sizing and design of wells and well screens is essential to successful delivery
or extraction of agents used for in situ soil and groundwater remediation. The open area
of a wells screen determines its ability to distribute flow evenly along the entire length of
the screen. Screen zonation is one mechanism for equalizing flow, but equal flow can
also be achieved by adjusting well diameter or operational parameters, such as pressure
or total flow rate. Problems of variable well elevation can also be addressed with screen
zonation. Proper engineering of well screens must also account for conveyance lines.
Most important among the design considerations is that the design assumes operational
parameters that will cause sufficient volumes of fluid or gas to be delivered or extracted
for the system to effectively remediate the target soil zone, water-bearing unit, or aquifer.