4.4.7.3   Long-Term Effectiveness and Permanence -- Alternative 7

The long-term effectiveness of solvent extraction was assessed in
terms of the risk remaining after treatment.  Risks may be posed
by chemicals not removed from the contaminated material during
treatment.  Performance data, factors that influence performance,
and system limitations are discussed below.

Solvent Extraction Performance

Full- and pilot-scale performance data are currently available
from RCC CF Systems, and Terra-Kleen.  Brief site descriptions of
successful treatment of PCB-contaminated wastes are presented
below for each vendor technology.  Performance data for the waste
treated at these sites is presented in Table 4-14.

Pilot-scale tests were conducted in Massena, New York, to
determine the RCC's B.E.S.T.  system ability to treat six wastes
from an aluminum manufacturing facility.  These waste types
originated from a soluble oil lagoon, a sanitary lagoon, a 60-
acre lagoon, a waste lubricating oil landfill, an oily waste
landfill, and the Dennison crossroads.  PCBs were the target
contaminants in all six wastes. The objective of these tests was
to determine whether the B.E.S.T.  process was capable of
reducing the PCB concentrations in the soils and sludges to less
than 2.0 mg/kg.

A pilot-scale SITE demonstration of RCC's B.E.S.T  process was
conducted in July 1992 near the Grand Calumet River in Gary,
Indiana.  The demonstration was conducted in cooperation with the
Great Lakes National Program Office and the U.S. Army Corps of
Engineers.  River bottom sediment collected from two separate
locations along the Grand Calumet River was treated during the
demonstration.  Two different sampling locations were selected so
that a variation in contaminant concentration (including oil
percentage) could be evaluated in regard to treatment
applicability.

The B.E.S.T.  pilot-scale solvent extraction unit was also used
to conduct a treatability study in Greenville, Ohio.  The soil at
a machining operations site in Greenville was contaminated with
PCBs as a result of the disposal of lubricants from machining
operations.  The objective of the treatability study was to
determine whether the B.E.S.T.  process was capable of reducing
the PCB concentrations in the soils and sludges to less than 10
mg/kg.  The treatment process achieved a 98.1 percent reduction
in PCB contamination.
                                 
RCC's full-scale B.E.S.T.  unit was used to treat oily sludge at
the General Refining site in Garden City, Georgia.  This oily
sludge was generated during oil waste reclamation and re-refining
activities and was disposed of in unlined lagoons.  The system
was tested and modified, and then treated approximately 3,700
tons of oily sludge.  Contaminants present in the feed included
PCBs, metals, VOCs, and SVOCs.  The organics from the feed were
concentrated in the oil fraction.  Metals remained in the treated
solid.  Low concentrations of metals were also detected in the
water stream discharged from the unit.  

Terra-Kleen's Solvent Extraction Treatment System was used to
remediate soils at NASNI, which is located at the northern end of
the peninsula that forms the San Diego Bay.  NASNI was officially
commissioned in 1917 as a support facility to provide services
and material for aviation activities and naval operations.  Large
quantities of hazardous waste were generated and disposed of at
NASNI after the United States entered World War II.  As part of
the Naval Environmental Leadership Program (NELP), NASNI
contracted Terra-Kleen to treat about 5 tons of PCB-contaminated
soil.  The demonstration was conducted at Site 4, which was used
to store miscellaneous materials including electrical
transformers containing PCBs.  Soils at Site 4 were contaminated
with heavy metals, VOCs, SVOCs, PCBs (Aroclor 1260), dioxins, and
furans.  The soils were homogenized before being loaded into the
five extraction tanks, so that the PCB distribution would be
similar in each tank.  The system had a 98.5 percent PCB removal 
rate and the treated solids contained some residue solvent.

CF Systems LG-SX technology was used to remediate sediment at the
New Bedford Harbor Superfund site located on the Acushnet River
Estuary north of Buzzard's Bay in New Bedford, Massachusetts. 
The site is listed on the NPL.  Harbor sediments at the site
contained pollutants discharged to the harbor from various
industrial sources.  The pollutants included PCBs, PAHs, and
heavy metals.  PCBs presented the greatest toxic threat, and
concentrations as high as 30,000 ppm were observed.  An estimated
878,000 yd3 of sediment was contaminated with 0 to 50 ppm PCBs;
236,000 yd3 was contaminated with 50 to 500 ppm range; 91,000 yd3
was contaminated with 500 to 5,000 ppm range; and 16,000 yd3 was
contaminated with greater than 5,000 ppm PCB contamination. 
Samples of harbor sediments were dredged by the U.S. Army Corps
of Engineers and stored in 55-gallon drums for use in a pilot-
scale SITE demonstration.  The pilot-scale unit used during the
demonstration achieved greater than 90 percent reduction in the
PCB-contaminant level conducted by CF Systems.

Applicable performance data demonstrate that solvent extraction
is potentially capable of reducing PCB concentrations in the
materials from the six CD sites to below 2 ppm.  Therefore, risk
associated with PCBs in the treated solids should be minimal. 
Trace residual solvent may remain in the treated solids.  For the
B.E.S.T  system, the treated solids typically retain 150 ppm of
TEA.  The typical extraction solvents used by the currently
available systems either volatilize quickly from the treated
solids or biodegrade easily.  Therefore, little risk associated
with extraction solvent would remain in the treated solids.

Other treatment residuals include concentrated oily waste and
product water.  Long-term effectiveness will also depend on the
adequate treatment of these residuals.  Concentrated oily waste
may include PCBs, oil and grease, naturally occurring organic
substances present in the untreated waste, and extraction
solvent.  Concentrated oily waste could be destroyed by
incineration or dechlorination.  Treatment of product water is
dependent on the concentration of contaminant present in the
water and the flow rate and volume of residual water.  In
general, the water should be sampled and will often be suitable
for discharge to a local POTW.

Factors that Influence Performance

Particle size distribution can affect system performance because
PCBs tend to sorb preferentially to finer-sized particles, and
fine solids are more difficult to extract (ATSDR 1994).  For
example, PCB-contaminated soils with greater than 15 percent clay
are difficult to treat by solvent extraction.  The clayey soils
present at the CD sites might not be treated effectively without
mixing them with coarser waste types.  Larger particles may not
pass through process equipment and may interfere the pumping of
the sediment slurry when required.  Particle size depends on the
specific system design selected and the scale of the processing
equipment.  Diameter ranges of 0.2- to 1-inch have been reported
as maximum values for particle size.

Moisture content can also affect system performance.  Moisture
content requirements depend on the specific system design.  If
the system is designed to treat pumpable sludge or slurries, it
may be necessary to add water to the solids to form a pumpable
slurry.  Sludge from the Winston-Thomas Sewage Treatment Plant
site may need to be dewatered depending on the system used;
however, most systems require reduction of the moisture content
in order to treat contaminated media effectively.  For example,
for the Terra-Kleen system, soil containing more than 20 percent
moisture must be dewatered prior to treatment.  Excess water
dilutes the solvent, reducing contaminant solubilization and
transport efficiency.

The contaminant concentration in the waste affects the number of
wash cycles required to achieve the cleanup concentration. 
Theoretically, solvent extraction has no upper limit in the
contaminant concentration that can be treated.  Even wastes with
concentrations as high as 330,000 ppm of PCBs can be treated, but
they would require more wash cycles.  For the B.E.S.T.  process,
approximately 80 percent of the contamination is removed during
each wash cycle; therefore, about eight wash cycles would be
required to remove PCBs from a concentration of 330,000 ppm in
the waste feed to a residual concentration of 2 ppm.

Organic carbon content can also influence system performance. 
High concentrations of organic materials in the waste matrix can
reduce the extraction efficiency and process rate (ATSDR 1994). 
Excavated material containing high levels of organic waste needs
to be blended before being fed into the extraction system.  This
blending process may need to be implemented for the landfill
wastes from the six sites.

The presence of detergents and emulsifiers can also unfavorably
influence extraction performance and material throughput (EPA
1994b).  Water soluble detergents in some wastes especially
municipal waste, will dissolve and retain organic pollutants. 
This factor can impede a system's ability to achieve low
concentration treatment levels.  Detergents and emulsifiers can
promote the formation of foam, which hinders separation and
settling characteristics and generally decreases material
throughput.

System Limitations

Solvent extraction is not effective for treating heavy metals
(ATSDR 1994).  Metals are insoluble in solvents normally used by
the solvent extraction process and typically remain in the soil. 
The remaining metals are a potential hazard because the metals
can leach out of the soil.  In order to effectively immobilize
the metals, chemical fixation or stabilization may be required. 
Organically bound metals can be extracted and become a component
of an organic waste stream, creating additional restrictions on
disposal.  Metals present in the wastes at the six sites may
require further treatment after the solvent extraction process. 
Solvent extraction is generally effective for treating materials
contaminated with VOCs and SVOCs.  If the VOCs are in high
concentrations however, solvent extraction is unsuitable.  It may
be possible to modify the treatment system to use additional
solvents.  These solvents will extract the remaining organic
contaminants not affected by the common process solvents. 
Dioxins have been treated on a limited basis by solvent
extraction, and furans may potentially be treated by this
technology.

The solid material input to the system would need to undergo
screening to meet the particle size limitation of 0.2- to 1-inch
nominal diameter.  Also, the material may need to be limited in
organic and clay content, with greater than 15 percent fines. 
Additionally, solvent extraction is generally not effective for
treating municipal solid waste and therefore would only be
effective in treating the municipal solid wastes at the   CD
sites that meet the screening requirements.  The remaining waste
would be treated as discussed in Section 3.3.2.3.  Moisture
content limitations also apply depending on which option is
implemented.  Adjustment of pH may be necessary for some systems
to ensure solvent stability or to protect process equipment from
corrosion.  For example, pH adjustment of greater than 10 is
required for TEA extraction.

Most organic solvents are relatively volatile, requiring
emissions control.  Some solvents can be toxic to some organisms,
requiring very efficient separation of the solvent from the
solids prior to disposal.  The typical extraction solvents used
in currently available systems either volatilize quickly from the
treated soils or biodegrade easily.  For example, the vendor
claims that TEA, the solvent used in the B.E.S.T.  process, is
biodegradable, does not accumulate in the environment, and occurs
naturally in the food chain (EPA 1993d).