DSHW-2011-006328 - 0901a06880233751

Launch Systems Group HAND DELIVERED
P.O. Box 707
Brigham City.UT 84302 MAY 2 3 2011
www atl<.com
UTAH DIVISION OF
16 May 2011 SOLID & HAZARDOUS WASTE
8200-FY12-012 ^tll.QI^^^
Scott T Anderson
Executive Secretary,
c/o UDEQ Division of Solid and Hazardous Waste
PO Box 144880
SALT LAKE CITY UT 84114-4880
Subject: Response to Division Comments on Corrective Measure Pilot Test Reports Ex-Situ
Soil and In-Situ Groundwater ATK Launch Systems Promontory Facility, EPA ID
#UTD00908]357
Dear Mr. Anderson
Enclosed, please find responses to the Division's comments on two ATK Launch Systems
Promontory reports, "Evaluation of the Enhanced Bioremediation Pilot Test" and
"Corrective Measure Pilot Test to Evaluate Ex-Situ Remediation of Perchlorate
Contaminated Soil.
Please contact Paul Hancock at [phone] if you have any comments or questions.
Sincerely,
David P. Gosen, P.E.,
Director, Environmental Services
ATK Launch Systems Promontory Response to January 20, 2011 Division of Solid
and Hazardous Waste Comments on Ex Situ Soil Remediation Evaluation of the In-Situ
Bioremediation Pilot Test for
Contaminated Groundwater
Comment: 1 The text states that the effluent is primarily water with residual volatile organic
acids. However, the exact nature and concentration ofthe volatile organics is not
discussed in the text. Please elaborate and Indicate whether there is reason to believe that
the volatile organics could present a risk to down-gradient receptors, if this test is to be
scaled up to full production in the future.
Response: Detailed speciation of organic acids In bioreactor effluent has not been
performed because of the general variability and utility of such an analysis. However,
acetic acid is the primary organic acid produced during microbial fermentation of sucrose.
An acetic acid odor is common to the tiioreactor effluent. Others include butyric acid, lactic
acid and propionic acid. Tests for total organic carbon (TOC) and chemical oxygen
demand (COD) have been used to estimate organic acid concentration. TOC results have
shown between 3,000 and 3,500 mg/L. COD has shown between 3,000 and 5,000 mg/L.
At this concentration, the volatile organics could present a risk to natural waterways if
introduced directly. Aware ofthis risk, ATK has developed the ex situ process to greatly
limit the probability of a release of raw bioreactor effluent to natural waterways.
Comment 2: The text states that saturation was only achieved in the center of the mounds,
but that the outer areas of the mounds did not display any discernible reduction in
perchlorate concentrations, because these areas did not receive enough effluent. It is
unclear why this geometry was selected at the beginning of the test. Furthermore, it is
unclear why the mound configuration was not changed once it became apparent that the
design of the mounds was inadequate to provide for treatment of all soils. Was untreated
soil from both the manure test and the bioreactor effluent test used in the final procedure
using roll-off dumpsters? Please elaborate.
Response: The mound configuration was a holdover from the previous test using manure.
The initial testing was to prove the concept of using bioreactor effluent as an electron
donor. Certainly, it was hoped that transition to the liquid donor could be implemented with
minimal changes to the existing test plan that used manure. Although the initial test mound
revealed the operational challenges of using a liquid donor, it validated that perchlorate
reduction could be achieved in soil saturated with bioreactor effluent under field conditions.
A few options were considered to optimize the mound configuration so that all of the soil
would stay saturated with bioreactor effluent. Most of these options involved spreading the
soil over a larger area or using heavy equipment to turn the soil once the center of the
mound had reached treatment goals. Due to the large quantity of soil in the test mound,
significantly more space would have been needed to create a thinner layer. Additionally,
maintaining a cover over the large area would have been problematic. Periodically turning
the soil - similar to the test plan using manure - was considered to be too inefficient in
terms of meeting the treatment goal within a reasonable time frame. Since, the options for
changing the mound configuration provided their own risks, ATK decided to abandon
treatment in mounds in favor of treatment in roll-off dumpsters. All soil in the test mound
was subsequently treated in the bioreactor effluent/roll-off dumpster process.
Comment 3: Does ATK plan on conducting any further work on evaluating this remediation
technique?
Response: Any further work will focus on optimizing the type of container used for the
treatment process.
Comment: 4) Page 2. The text states that, in the vicinity ofthe M-136 burning grounds, the
groundwater potentiometric surface is substantially reduced, and caused, in part, by locally
elevated hydraulic conductivities. Please explain how a reduced potentiometric slope can be
caused by elevated conductivities beneath the burning grounds. A review ofthe 2010 site-wide
potentiometric surface map reveals that a reduced slope is perhaps discernible to the south of
the burning grounds, but not at the burning grounds proper.
Response: As described in the 2005 Groundwater Flow and Contaminant Transport Model
Report, "The slope of the potentiometric surface in the vicinity of the Burning Grounds (between
wells C-6/C-8 and B-4/F-1) is significantly less than that in adjacent areas immediately up- and
downgradient. This decrease in slope is likely caused by a combination of a local increase in
the hydraulic conductivity (due to an extremely fractured fault zone or solution cavities) and the
presence of a downgradient groundwater flow barrier. Chen-Northern (1992a, 1992b, and
1992c) and Bolke and Price (1972) suggest that an up-faulted ridge of limestone on the
downgradient edge of this area, trending northeast to southwest across Blue Creek Valley, acts
as a low-conductivity barrier. However, Plate 2-2 indicates that the hydraulic gradient does not
increase until groundwater flows back into the unconsolidated materials at and downgradient
from wells F-1 and E-9. A review of hydraulic conductivity data obtained from site monitoring
wells (discussed in Section 3.1 of the report) indicates that this steepening of the potentiometric
surface occurs in an area of lower conductivity alluvial deposits."
Basically high hydraulic conductivities allow the groundwater to move at a higher flow rate and
the potentiometric surface flattens out where the water moves through the subsurface with less
to impede flow whereas as stated above, hydraulic gradient steepens when the groundwater
flows into unconsolidated material.
Comment: 5) Page 7. The text states that Figures 2 and 3 display the point decay rates for
perchlorate at wells C-2 and A-5. While we agree that the decay rates at well C-2 appear to
have different slopes, we are not sure if the decay rate at well A-5 is indeed significantly
different for pre- and post-treatment times. Did ATK perform any statistical tests regarding
significant differences ofthe respective slopes?
Response: The decay rate at C-2 and A-5 are represented by straight line segments and
shown in standard slope-intercept form in the equation y=mx+b; where m is the slope of the line
and b is the y-intercept, which is the y-coordinate of the point where the line crosses the y axis.
Visually both graphs show increased slopes and therefore increased degradation or decay. The
slope on both C-2 and A-5 increases post injection. At C-2, the slope increases 4 times from
0.0004 to 0.0016 and at A-5, 2.3 times from 0.0003 to 0,0007. ATK did not perform statistical
testing on the slope differential because it was concluded from the slope-intercept equation that
the decay rate had accelerated following injection of oil.
Comment: 6) Page 8. The text states that the decay rate of well T-2 is larger than that of well D-
2, which, in turn, is smaller than that of well C-2, which proves that this test demonstrates the
efficient degradation of perchlorate in groundwater. However, it appears that the slope of decay
in well D-2 is positive, which contradicts the above line of reasoning. Please elaborate. We also
question whether the above statement is perhaps not well founded, as it appears to be based
on three wells only, and a determination of statistical significance of the respective decay slopes
is missing here as well.
Response: Wells D-2, T-2, and C-2 are situated in a southeast to northwest line. Well D-2 is
topographically higher in elevation than T-2 which is higher than C-2. Well D-2 appears to be
cross-gradient both topographically and potentiometrically from the injection well A-1 and is
therefore less likely to be impacted by the vegetable oil injection. It was not clear in the text that
this well represents conditions not impacted by vegetable oil injection and therefore shows no
response to anaerobic degradation. The reviewer is correct in stating that the text does not
clearly state the intended purpose of citing these three wells.
Additionally, it is understood that statistical significance of the respective decay slopes may not
be well founded based on only three wells. This pilot test is limited in scope and extent and to
increase the number of wells that may be immediately impacted would require a greater volume
of vegetable oil and a longer time over which to inject and monitor the impact.
Comment: 7) Page 9. The text states that vinyl chloride (VC) was observed one year after the
test began, at a concentration of 4.1 ug/l in injection well A-1. While VC, classified as a
mutagen, and as a known human carcinogen, is indeed short-lived, once it has the opportunity
to volatilize, please be aware that, for a carcinogenic target risk level of 1E-6, the EPA regional
screening level for ingestion is 0.017 ug/l, well below the concentration displayed in well A-1.
We are concerned that this in-situ test, if scaled up to full production, would not account
adequately for the generation and dispersion of VC into the groundwater system. Please
comment.
Response: ATK understands why the Division is concerned about progressing to a full scale
implementation of the vegetable oil bioremediation at the M-136 site. However, many sites
across the nation use a similar process to remediate chlorinated VOCs through reductive
dechlorination.
The EPA's Fact Sheet on Vinyl Chloride states: If released to water, vinyl chloride will rapidly
evaporate. Using a reported Henry's Law constant of 0.0560 atm/cu m-mole, a half-life of 0.805
hr was calculated for evaporation from a model river 1 m deep with a current of 3 m/sec and
with a wind velocity of 3 m/sec. In waters containing photosensitizers such as humic acid,
photodegradation will occur fairly rapidly. Limited existing data indicate that vinyl chloride is
resistant to biodegradation in aerobic systems and therefore, it may not be subject to
biodegradation in aerobic soils and natural waters. It will not be expected to hydrolyze in soils or
natural waters under normal environmental conditions.
Some data indicate that vinyl chloride is too readily volatilized to undergo bioaccumulation,
except perhaps in the most extreme exposure conditions. Based on a reported water solubility
of 2,700 mg/l, a BCF of 7 was estimated, indicating that vinyl chloride will not be expected to
significantly bioconcentrate in aquatic organisms.
"If VC is released into water, it is not expected to adsorb to suspended solids and sediment in
the water. The biodegradation half-life of vinyl chloride in aerobic and anaerobic waters was
reported as 28 and 110 days, respectively. Volatilization from water surfaces is expected to be
an important fate process. The estimated volatilization half-lives for a model river and model
lake are 1 hour and 3 days, respectively.
Vinyl chloride is transitory or relatively short lived especially in an enhanced biodegradation
process. It is recognized that VC is a necessary step to achieve complete degradation. There
have been no significant VC detections in the past and therefore any VC is most likely an
indication of bacterial degradation of the solvents in groundwater. If full scale implementation of
vegetable oil remediation were to occur at the site, ATK would monitor downgradient wells
frequently for the presence of VC,
It is likely that anaerobic dechlorination has been happening for years at this site. Until recently,
wells at the M-136 area have shown measureable amounts of mineral oil on the water surface.
Concentrations began to climb at this site when the mineral oil was no longer detectable. No
significant concentration of VC has been detected downgradient of the site over the last several
years during a period of time when mineral oils most likely encouraged reductive dechlorination.
ATK believes that enhanced bioremediation can be safely applied to the M-136 site with
strategic downgradient groundwater monitoring.
8) Figure 1. The North arrow appears to point to a wrong direction. Please correct in any future
submittals that utilize this figure.
Response: The north arrow on Figure IA has been changed to point in the appropriate
direction.
9) Page 10. Does ATK plan on conducting any further work or monitoring to evaluate this
remediation technique?
Response: ATK believes there is value to continue monitoring the wells within the treatment
area of the burn grounds. At this time, there are no specific plans to inject additional oil, ATK
is focused on the obtaining results from the groundwater risk assessment before determining
appropriate remedial actions.