|
By John T. Leethem,
C.P.G.
Background
for MtBE Chemical Oxidation
Chemical
oxidation has been used in situ for remediation of
contaminants such as chlorinated compounds, benzene, and mixes
of organic compounds. Although MtBE has probably be treated
by in situ oxidation inadvertently during these treatments for
gasoline, recently MtBE has become a driver for cleanup and
now in situ chemical oxidations are beginning to be done to
remediate the MtBE as well as gasoline. Although chemical
oxidation processes tested for MtBE in drinking water have
generated undesirable intermediates (Liang, et al.), this is
less of an issue with in situ chemical oxidations.
Groundwater chemical oxidation treatments generally have
longer contact times and oxidant concentrations are high,
residual intermediates of oxidation are of less concern and
excess oxidant is not a problem since it will be consumed by
natural organic matter. In addition, the groundwater has
days, years or decades until it reaches potential human
consumption while drinking water may be at the faucet in
minutes to hours. Generation of intermediates in groundwater
is less of an issue because large doses of oxidizers are added
and can be repeated, and intermediates are potentiated for
subsequent natural biological degradation.
In situ
oxidation of MtBE is just beginning to be used. Several
different oxidants have been tried (see article on
permanganate oxidation in this issue). The oldest, most
understood and cost effective in situ chemical oxidant system
is the peroxide/iron Fenton’s reagent system. The cost versus
efficiency of the various systems for MtBE are not yet
established, are impacted by the presence and concentrations
of other contaminants at the site as well as any naturally
occurring readily oxidizable organic materials.
The Texas
Site Setting
This case
study site is an active north Texas gas station situated in an
urban setting on a busy street corner. The site has low
permeability soils, which are dry most of the year. However,
the UST tankhold was excavated and backfilled with sand and
gravel so that only the surrounding soils were native. The
tankhold retains water that infiltrates in during heavy
rainfall events. This causes the water level in the tankhold
to fluctuate from approximately three to six feet below
grade. During construction activities associated with a UST
upgrade at the gas station, the product lines were left
unconnected to the submersible pumps on each of the three USTs.
The connections between the product lines and submersible
pumps were located inside plastic sumps on top of the USTs.
During a subsequent heavy rain event, the sumps filled with
water and water flowed into each of the USTs. As water filled
the USTs, gasoline was displaced out of the top of each UST
through the fill ports. The gasoline pooled on top of the
tank pit and in a larger area covered with pea gravel, where
concrete had been removed during construction. During and
following the rainfall event, gasoline and water infiltrated
into the pore space among the sand and pea gravel within the
tank pit. After discovery of the release on the morning of
May 18, 1999, approximately 7,052 gallons of gasoline were
determined to have been released. The initial response was to
recover as much of the free-phase product and dissolved phase
as practical. To accomplish this, water was recovered from the
UST tankhold and surrounding area, based on the volume of
water recovered. This was done initially to recover the
released free-phase hydrocarbons and continued in an attempt
to restore the site. That same day, 18,184 gallons of water
and product were pumped out of the tank pit and from the four
in-place observation wells surrounding the tankhold. In
addition, 5,040, 2,946, and 6,037 gallons of water/product
were pumped out of the tank pit on May 19, 20, and 24, 1999,
for a total of 32,207 gallons.
URS
Corporation (URS) then conducted a subsurface investigation in
the vicinity of the tankhold. Contamination of native soils
was not identified in the vicinity of the tankhold, and the
Texas Natural Resource Conservation Commission (TNRCC) granted
closure of the Leaking Petroleum Storage Tank (LPST)
facility. At the owner’s request, URS collected water samples
from the tank pit observation wells. Laboratory analyses
reported gasoline constituents present that had not been
addressed by the previous groundwater recovery efforts.
After several
more extraction events and recovering a total of more than
50,000 gallons of water from the tankhold, the hydrocarbon
concentrations stabilized in the groundwater. This indicated
that little, if any, additional efficient hydrocarbon removal
was likely with continued groundwater removal. Groundwater
recovery was stopped and it was concluded that a pump and
treat technology was probably no longer appropriate as a
speedy, site remedy.
TECHNICAL
APPROACH
The Remedy
Evaluation
Although the
free-phase gasoline was recovered using the initial tankhold
pumping, dissolved-phase hydrocarbons remained in the tankhold
water. To evaluate the several available technologies for
site restoration, several remedial options were compared for
schedule and cost. Additional factors considered in the
selection of the remedy included the need to continue
operations at this active gas station and minimize the impact
of remediation on the business. Consequently, the remedy
needed to employ operations conducted at night and all
remediation equipment needed to be setup after business hours
and removed before business hours to minimize business
impacts.
The
comparison included pump and treat as a baseline as well as
several other more innovative technologies. The previous
recovery activities, including the site wells had already
provided a good evaluation of how effective pump and treat
might be at the site. One technology that has been proven
effective on MtBE and other hydrocarbons in saturated zones at
other sites is enhanced aerobic bioremediation using Regensis’
Oxygen Release Compound (Koenigsberg & Norris, 1999). However
this remedial technique is typically only effective on lower
concentrations of hydrocarbons where free product is not
present. Due to the relatively high concentrations of MtBE,
BTEX and other hydrocarbons, combined with the need for a
quick solution, this remedy was not selected as a final
remedy. Since in situ chemical oxidation is among the
technologies proven effective on higher concentrations of MtBE,
such as those present at this site, in situ chemical oxidation
was selected as the remedy. In November 2000, URS initiated
in-situ chemical oxidation of the UST tankhold sand and gravel
backfill, and tankhold water.
The In
Situ Chemical Oxidation Process
The specific
technical approach selected was the in situ oxidation of MtBE
and gasoline residuals using Fenton’s reagent. The Fenton’s
reagent process uses Fenton’s chemistry to generate primarily
hydroxyl radicals. The petroleum hydrocarbon destruction is
the result of a strong chemical oxidation by the reactive
chemical species generated by the Fenton’s Reaction. Fenton’s
Reaction is named for H.J.H. Fenton, who reported in the
1890’s that hydrogen peroxide, in the presence of iron salts
or iron oxides, formed hydroxyl radicals (OH-). These
hydroxyl radicals are among the strongest oxidants for
organics and react rapidly with organic compounds. The
hydroxyl radicals cleave the petroleum hydrocarbon bonds,
eventually reducing the hydrocarbons to water and CO2 as
shown below, if adequate peroxide is provided.
Chemical oxidation
of MTBE (C5H120)
C5H120
+ 15H2O2
(Fe)
5CO2 + 21H2O
→
On a
stochiometric or theoretical basis, 5.78 pounds of hydrogen
peroxide (H2O2) are required for every pound of MtBE oxidized.
This is the theoretical amount required, but due to other
factors such as co-contaminants, naturally occurring organic
compounds and other complicating factors, the chemical oxidant
needed is often five to seven times greater. This illustrates
one of the drawbacks to chemical oxidation processes.
Oxidation processes can not typically target only the
contaminant(s) of concern. Instead, they oxidize all the
susceptible compounds; more rapidly those compounds that are
most susceptible to oxidation and more slowly those compounds
that are less susceptible. However, application of high but
still economical concentrations of oxidants at sites usually
overcomes this competitive oxidation limitation. One of the
additional advantages of the Fenton’s reagent reaction is that
hydrogen peroxide will rapidly degrade and dissipate leaving
no residual except water, oxygen, carbon dioxide, and possibly
salts from non-hydrocarbon components. That assures the
treatment is not introducing metals other than iron.
REMEDIATION
Remedial
Operations
At this site,
URS worked with Hill Liebert, Inc. (HLI), a company with a
patented, delivery technique. The delivery system is capable
of injecting liquid amendments into soil and groundwater using
a high pressure, low volume injection system. Although not
critical at this site because of the gravel and sand matrix,
amendments can be injected at pressures as high as 5,000 psi.
When appropriate to the site matrix, HLI claims that injecting
low volumes of liquid amendments at high pressure creates
micro-fractures in soils with low permeability. Once the
micro-fractures are opened, the amendments can disperse into
the soil and groundwater. The lance is dual purpose, both
boring the access holes and injecting the amendments. For
contamination in both the unsaturated and saturated zones, the
lance injects amendments throughout the depth range that is
contaminated.
This delivery
technique provides the intimate chemical contact between the
amendment and the contaminant necessary to achieve contaminant
oxidation. The peroxide and iron catalyst may be
simultaneously injected into the subsurface through separate
ports in the injection. The sand around the hole is to capture
foam slurry that can emerge from the hole during the vigorous
reaction of the hydrogen peroxide with the ferrous sulfate and
subsequent reactions of the Fenton’s reagent with the
organics. The lance is designed to allow the delivery of the
amendments where they are needed, without the need for
injection wells. The lance can be easily inserted using its
injection blast to bore a hole. For these operations, the
work area was walled off with plastic sheeting to minimize any
potential spatter during injection. The work was carried out
at night to minimize the disruption to the business during
peak business hours.
RESULTS
Results To Date
Based on the
sampling done by URS, the pretreatment concentrations of the
contaminants of concern were:
-
MtBE -
411 to 475 milligrams per liter (mg/l);
-
Benzene -
14.4 to 15.8 mg/l;
-
Toluene -
27.9 to 28.0 mg/l;
-
Ethylbenzene
- 1.45 to 2.25 mg/l; and
-
Xylenes -
1.45 to 2.25 mg/l.
URS
calculated the mass of hydrocarbons in the backfill sand and
water, based on measured concentrations from sampling. Using
this information, the masses of hydrogen peroxide and other
amendments required for site remediation were estimated. In
general, for highly contaminated sites, several injections may
be necessary to assure achievement of the low target
concentrations for BTEX, benzene in particular, and MtBE.
A total of 37
drums of 33% hydrogen peroxide were injected into the tankhold
for a total of about 2,000 gallons. Two hundred and twenty
gallons of ferrous sulfate (12% concentration) was injected
prior to the peroxide addition. The injector pressure at the
nozzle was estimated to be 3,500 psi. After the first
treatment, the in situ chemical oxidation achieved average
MtBE and BTEX reduction of 83.4% and 69.3%, respectfully. The
reduction of TPH was lower, at about 60.4%. The four existing
UST tankhold observation wells were used to monitor the
progress of the treatments by periodic sampling and analyses
for the major contaminants of concern, BTEX and MtBE.
The data is
for the southwest monitoring well (MW-SW) where the highest
concentrations of gasoline components were detected. Similar
results were seen at the other wells although they were not as
highly contaminated. The changes in MtBE and to a lesser
extent BTEX concentrations are quite good. The Total
Petroleum Hydrocarbons (TPH) were less impacted by the in situ
oxidation process. This may be due to the incomplete
oxidation of some hydrocarbons or an artifact of using the
groundwater, in which the bulk of TPH has a poor solubility,
as the primary indicator of treatment. The results suggest
that MtBE and BTEX are “better” chemical targets for Fenton’s
reagent oxidation than the moderate-sized linear alkanes and
isoalkanes and other components measured as TPH in gasoline.
No monitoring of MtBE or BTEX intermediates was done to
determine if they were produced or completely oxidized.
Experience with drinking water suggests that the intermediates
would be present at much lower concentrations than the MtBE (Liang,
et al). In any case, in this type of setting, with the large
treatment dosage, it would seem that these intermediates would
be rapidly oxidized, or subsequently naturally attenuated
because they are partially oxidized, making them more
susceptible to biological metabolism.
Conclusions
In situ
chemical oxidation of MtBE gasoline releases was an effective
mechanism for this localized, highly concentrated release.
The process was cost-effective for this small groundwater
release and was done with minimal site disturbance and minimal
business disruption. An understanding of the possible
underground structures is essential and consideration should
be given to the effect of the oxidant system on underground
structures. The initial results for this site suggest that
MtBE and BTEX compounds might be preferentially oxidized.
Evaluation of a rebound effect has not yet been tested at this
site.
There are a
variety of processes for introducing in situ oxidants,
including injection of liquids under pressure and through
installed wells. Existing wells can also be used but may not
provide the coverage for efficient oxidation of the
contaminated area. The use of lance injections, while
somewhat messy, allows cost-effective injection of the
oxidants at numerous locations, if needed. This process
demonstrated another efficient remediation technique in the
arsenal for MtBE and gasoline releases. The effectiveness of
this process for a compact area emphasizes the need to monitor
and detect leaks early so that the area to be treated is
small, maximizing the range of economical techniques to select
from for remediation. In addition this site is representative
of the size and type of sites most commonly remediated by the
gasoline marketing industry.
REFERENCES
Koenigsberg,
S.S. and R.D. Norris. 1999. Accelerated Bioremediation
Using Slow Release Compounds: Selected Battelle Conference
Papers: 1993-1999.
Liang, S.,
Yates, R.S., Palencia, L.S., and Bruno, J.M., 1999,
Oxidation of methyl tertiary-butyl ether (MTBE) by Ozone and
Peroxone and Identification of By-products, American Water
Works Association Annual Conference Proceedings, Chicago,
Illinois, June 20-24, 1999.
John
Leethem is a senior project manager with URS Corporation in
Houston, Texas specializing in advanced in situ remediation
technologies.
Top |
