United States
Environmental Protection
Agency
Risk Reduction
Engineering Laboratory
Cincinnati, OH 45268
Research and Development
EPA/600/SR-92/057 May 1992
^ EPA Project Summary
Technical Aspects of
Underground Storage Tank
Closure
The U.S. Environmental Protection
Agency (EPA) Is currently evaluating
several technical and regulatory as-
pects of UST closures, such as appro-
priate tank cleaning upon its removal
from service. This study has developed
a deeper understanding of Underground
Storage Tank (UST) residuals at clo-
sure: their quantities, origins, physical/
chemical properties, ease of removal
by various cleaning methods, and their
environmental mobility and persistence.
The investigation covered only un-
derground storage tanks that held the
following products: gasoline, diesel oil,
and fuel oil. It obtained information in
two phases. Phase I elicited data via
telephone contacts with knowledgeable
Individuals Including tank cleaning
companies, from literature cited by
these experts, in site visits, and from
questionnaires completed by state rep-
resentatives.
Phase II monitored selected tank
cleaning cases and made quantitative
measurements of the amounts of re-
siduals left in USTs before and after
cleaning, characterizing the nature of
the residuals and any rinses generated
during the cleaning process. To sup-
port the objectives of the study, the
following Information was collected for
each UST site Included In the study:
estimates of volumes of tank residuals
and secondary wastes, hazardous char-
acteristics and chemical composition
of the residuals and secondary wastes,
detailed descriptions of the cleaning
methods used, and background Infor-
mation on the UST site that relates to
the nature of the residuals.
This report documents the study find-
Ings In order to aid regulators and to
assist those implementing/overseeing
closure activities. This report covers a
period from August 1988 to May 1990,
and work was completed as of May
1990. '
This Project Summary was devel-
oped by EPA's Risk Reduction Engi-
neering Laboratory, Cincinnati, OH
45268, to announce key findings of the
research project that Is fully docu-
mented In a separate report of the same
title (see Project Report ordering Infor-
mation at back).
Introduction
The overall objective of this study was
to develop a deeper understanding of UST
residuals at closure: their quantities, ori-
gins, physical/chemical properties, ease
of removal by various cleaning methods,
and their environmental mobility and per-
sistence. The investigation covered un-
derground storage tanks containing: gaso-
line, diesel oil, and fuel oil. The work pro-
gressed in two phases.
Phase I: Preliminary Investigation of
UST Residuals and UST Cleaning/
Closure Methods
To obtain preliminary information oh
UST residuals, researchers employed the
following sources: telephone interviews,
review of literature cited by expert tele-
phone contacts, observation during site
Visits to four tank cleaning/removal opera-
tions, a survey of various,state represen-
tatives at a National UST seminar, and
engineering calculations on residual vol-
umes and costs of cleaning/closure.
The telephone surveys of experts elic-
ited citations of published and unpublished
data that were subsequently reviewed. Site
visits provided an opportunity to observe
tank cleaning and removal operations by
two companies at three different sites.
To supplement data from the telephone
survey, a focused survey was conducted
at the November 1988 "Workshop for State
Tank Program Managers Conference," in
Santa Fe, New Mexico, sponsored by the
U.S. Environmental Protection Agency's
Office of Underground Storage Tanks
(OUST). The data provided by the tar-
-------
geted respondents illuminated some com-
mon, jurisdfctfonal, closure practices and
Indicate which practices are prevalent.
Engineering calculations, detailed in the
full report, provided estimates on the vol-
ume of residuals likely to be found in
USTs, the amount of water and rust or
scale that might be expected in a UST,
and the costs of UST cleaning and clo-
sure.
Phase II: Hold Sampling and Analyses
of Residuals at Sites Undergoing UST
Cleaning/Closure
Under an agreement with a UST clean-
ing/removal contractor and with the per-
mission of UST owners, Phase II moni-
tored selected tank cleaning cases and
made quantitative measurements of the
amounts of residuals left in USTs before
and after cleaning, characterizing the na-
ture of the residuals and any rinses gen-
erated during the cleaning process. To
support the objectives of the study, the
following information was collected for each"
UST site included in the study: estimates
of volumes of tank residuals and second-
ary wastes, hazardous characteristics and
chemical composition of the residuals and
secondary wastes, detailed descriptions
of the cleaning methods used, and back-
ground information on the UST site that
relates to the nature of the residuals.
UST Residuals
Gasoline and diesel oil USTs have been
found to contain significant quantities of
residuals at closure, typically tens to hun-
dreds of gallons. The tanks can usually
be emptied by .ne owner/operator to within
4-6 in. of the tank bottom. This dimension,
which determines residual quantity in an
"empty" tank before cleaning, translates
into about 100-200 gal for a 10,000^al
tank. Both the Phase I and Phase II find-
ings indicated that the median volume of
residuals found in gasoline and diesel oil
USTs before cleaning was slightly below
100 gal. Some USTs, however, are found
to contain several thousand gal, consist-
ing of abandoned product and/or water
which has leaked into them.
Quantity
Field personnel often describe the vol-
ume of a tank's contents in terms of inches
of residuals on the bottom of the tank.
Figure 1 illustrates the residuals in a UST
and the formula for calculating the volume
of residuals.
By design, the submersible pump sys-
tems used to supply product, drop down
no farther than 4 in. above the tank bot-
tom in steel tanks. This provides 4 in. of
dead tank space, used to trap sediments
and water in the tank to ensure that they
will not be pumped out to the customer.
For fiberglass-reinforced plastic (FRP)
tanks, the tube usually ends 6 in. above
the tank bottom to allow for any settling
and deformation of the FRP tank. This
design feature leaves at least 4 in. of
residuals in steel tanks and 4-6 in. of
residuals in FRP tanks after the tank has
been "pumped dry" by the tank owner.
This represents residuals from 95 to 264
gal for a 10,000-gal tank a mid-sized
UST.
The volume of residuals found in gaso-
line tanks at any one site can vary signifi-
cantly. The majority of the reporting par-
ticipants estimated residual quantities up
to 1,000 gal. The mean of the values
reported was 160 gal; the median, 75 gal.
Most respondents agreed that diesel oil
Cross-Sectional
tanks contained more residuals than gaso-
line tanks, with a range of up to 200 gal
and a mean value of 58 gal. The median
estimate was approximately 75 gal. They
also concurred that fuel oil tanks produced
a greater amount of residuals than oil
tanks. The two respondents that provided
numbers for this product reported 500 and
1,000 gal, averaging to 750 gal signifi-
cantly higher than gasoline and diesel oil.
The volume of used rinse solutions gen-
erated during the cleaning procedures can
vary widely with the type of cleaning pro-
cedure used. Estimates ranged from 100
to 3,300 gal, with an average of 1,200 gal.
The American Petroleum Institute (Rec-
ommended Practice 1604) calls for filling
the tank nearly to the top for cleaning
and/or vapor removal purposes. This prac-
tice would generate much greater volumes
Draw Pipe
View
D = Tank Diameter
L = Tank Length
d = Depth of Residuals
V = Volume of Residuals in UST
View
t
s Draw Pipe
Vent
Pipe
"'/'
T
u
I JL
"KT
2 arccosl
ffl
- d
sin
2 arccosl
Figure 1, Schematic of UST tank for estimate of residuals volume.
-------
of used rinse solutions than actual prod-
uct.
Origin and Composition of
Residuals
The basic components of tank residuals
usually include residual product, water,
product-related residuals, tank rust and
scale, soil, dirt and other foreign objects,
and microorganisms. Residual product
probably constitutes 70-90% .of total re-
siduals in an aged tank. The other com-
ponents make up the remaining 10-30%,
with microorganisms represented in large
numbers but a very small percentage of
the total weight.
Product
Residual product would represent ap-
proximately 100 gal in a 10,000-gal tank.
The purity of the product must be deter-
mined in each case. Resale of gasoline,
for example/might require filtration, dewa-
tering, or further treatment. If the tank
being removed was abandoned for a long
time (months to years), significant changes
in the nature or composition of the residu-
als might take place due to volatilization,
water infiltration, rust formation or biologi-
cal action. Some tank cleaners do not
attempt a separate recovery of this re-
sidual product to facilitate reuse; they pump
it into the same tank truck used for rinses
and/or residuals from other tanks. They
then send the mixed fuel to a treatment
facility which separates, treats, and/or dis-
poses of the components.
Product-Related Residuals
Survey respondents discussed the pres-
ence of some product-related residuals
(e.g., gums, sediment) but estimated their'
total amounts to be relatively small. These
include gums and tars (high molecular
weight organics left by heavy fuels), poly-
mers formed in situ from reactive compo-
nents of the fuel (e.g., unsaturated hydro-
carbons) that can sink to the bottom of
the tank, sediment present in product on
delivery that sinks to the tank bottom, and
certain fuel components that attach to tank
walls or other solid residuals through sorp-
tion.
Water
Significant amounts of water lie at the
bottom of many, if not most, USTs. The
sources of such water include accumu-
lated water delivered in the product, con-
densation in the tank from infiltrating mois-
ture-laden air, surface runoff entering fill
pipe, and groundwater leaking into tank or
fill pipe.
Water residuals in USTs may play a
significant role in the internal corrosion of
steel tanks. Water present in a UST can
exist partly as a separate phase and partly
in solution with the fuel. It is' a common
practice for owners of USTs in service to
periodically check for the presence of wa-
ter (and sediment), pumping out any ex-
cess over 1 in. prior to refilling the tank
(about 12-18 gal in a 10,000-gal tank).
Water found in USTs prior to cleaning
generally would contain a significant
amount of dissolved hydrocarbons (-100-
300 mg/L), dissolved salts (e.g., Na*, Cr,
Fe*2, HCO3-, Pb*2) and other soluble com-
ponents or additives in the fuels (e.g.,
ethanol, MTBE, detergents).
Tank Rust or Scale
The survey and information cited by
respondents indicated that steel tanks are
likely, over time, to shed rust particles
(Fe^O3), and iron scale. This internal cor-
rosion may be caused by galvanic action
or bacterial action. Concentrated internal
corrosion often occurs directly under the
fill tube where the gauge stick strikes the
bottom of the UST. Several measures can
prevent tank failure in this location. Never-
theless, surveys of UST removals have
clearly demonstrated the importance of
internal corrosion to UST failures. Study
calculations estimate about 10 Ib of rust
generation in a 10,000 gal tank.
Soil, Dirt and Other Foreign
Objects
The Phase I survey and field trips pro-
vided evidence of the following foreign
objects in USTs: soil, dirt, rubber hoses,
soft drink cans, and similar trash. Although
this material probably entered via the fill
tube, some may have been discarded in
the tank prior to its initial use. There is
also potential for the entry of foreign ob-
jects at other times (e.g., repairs).
Microorganisms
Like water, microorganisms appear to
be fairly ubiquitous in petroleum storage
and distribution systems. They can reside
in the tank before it is used, and enter
from the outer environment via an open
fill tube or cracks. While they may appear
to be present in large numbers (102 to 103
organisms/L), their combined mass is
small. At times, however, large floes can
form, clogging fuel lines and filters. The
microorganisms in USTs include several
varieties of bacteria and fungi. One espe-
cially important class (sulfate-reducing bac-
teria) can cause significant iron and steel
corrosion.
Microorganisms need water to thrive
and, in storage tanks, are usually found at
the fuel-water interface. The mix of hydro-
carbons, water, oxygen (low for anaer-
obes), nutrients, and a compatible pH all
contribute to their growth. They appar-
ently thrive better in fuel oil than in gaso-
line.
Tank/Site Factors Affecting
Residual Quantity and
Composition
A number of tank and site factors con-
trol the nature, quantity, and composition
of UST residuals, such as tank design,
use, cleaning procedures, repair practices,
age, total volume throughput, site factors,
hydrogeology, meteorology, product type,
and product composition. These factors
also suggest ways to reduce the volume
and/or control the composition of
UST residuals. For example, lowering the
suction tube deeper into the tank increases
the maximum pumpable by the owner/
operator, and therefore lowers the volume
of remaining product. The origins of the
various components of the residuals are
fairly discernible. This knowledge and in-
formation on relevant site/tank factors can
help to control the future quantity and
quality of residuals. For example, the
growth of microorganisms can be con-
trolled by the use of biocides and/or the
elimination of water; this would reduce the
microbiological mass as well as the amount
of internal corrosion and rust generation.
Cleaning and Closure
Cleaning Procedures
A variety of tank cleaning and removal
procedures appear to be in use; many are
variations on a simple, logical theme. Many
steps are dictated by safety considerations
and state and local regulations, rather than
concern for tank cleanliness. The guiding
set of objectives in emptying/cleaning
USTs (whether they are actively in use or
set for closure) should entail minimizing
the environmental/health hazards pre-
sented by the tank and its residuals, the
explosion hazard of removing the UST,
the volume of secondary waste gener-
ated, and the cost of UST closure.
Rinses
In one way or another, most proce-
dures begin by pumping residuals with a'
suction line, then rinsing the tank with
water, and finally removing the used rinse
solution. For USTs with especially viscous
residuals, a light fuel oil (e.g., No. 2),
sprayed into the tank, may assist in the
cleaning.
Manholes
Several tank cleaning companies, after
the initial pumping of liquid residuals, cut
a manhole into the UST so that a work-
-------
man can enter, and then manually re-
move bottom grit and, with a "squeegee,"
wipe liquids adhering to the side walls.
The risk of explosion is significant, par-
ticularly for tanks which have not been
properly purged. However, benefits gained
from the increased cleaning efficiencies
and closer inspection of the tank may
sometimes outweigh the hazards.
Disposal of Residuals
Some companies put both initially-
pumped residuals and used aqueous
rinses in the same tank truck for off-site
treatment and disposal. Other companies
segregate the residuals from the rinses,
thus facilitating subsequent treatment.
Disposal of Tanks
For tanks that will be crushed/cut and
remelted, a modest amount of retained
residuals may be environmentally accept-
able. For tanks that are filled in place or
landfilled, the retained residuals are likely
to pose only a smali-to-negiigibie risk of
adverse environmental impact due to the
small volume of retained residuals, limited
environmental mobility for most constitu-
ents, and limited lexicological significance
for the bulk of the constituents.
American Petroleum Institute
Recommendations
The basis of most UST cleaning meth-
ods identified through the survey is API's
Publication 1604, "Removal and Disposal
of Used Underground Petroleum Storage
Tanks," and API's Publication 2015,
"Cleaning Petroleum Storage Tanks." Pub-
lication 1604 does not address cleaning
methods explicitly, but it does describe
the removal process. Publication 2015 de-
scribes a recommended cleaning process,
using the following format:
1. Completing preliminary preparations
2. Determining that the dike area is
free of flammable or toxic materials
before personnel are permitted to
enter the tank
3. Controlling sources of ignition in,
around, and on the tank
4. Emptying the tank by pumping out
residual liquid and floating it with
water
5. Blinding off the tank and de-ener-
gizing electrical circuits after as
much of the contents as possible
have been removed
6. Vapor-freeing the tank (mechani-
cal, steam, and natural ventilation
are alternatives)
7. Testing the tank for oxygen, hydro-
carbon vapors, and toxic gases
8. Opening the tank for entry
9. Removing bottom residuals and
sending them for appropriate dis-
posal
The UST is then transported to a licensed
UST disposal facility for ultimate disposal.
Additional Practices Reported
The Phase I survey of tank cleaning
and tank removal contractors provided a
variety of cleaning procedures in addition
to that described above. The full report
lists some of the more interesting varia-
tions, such as cleaning residuals from the
tank while it is still in the ground by spray-
ing rinse through fill or vent pipes and
then pumping the rinse out (an alternative
to a manhole). Such variations may de-
pend on many factors, e.g., residual type,
future tank fate, tank size/design, and the
availability of water.
Secondary Wastes
Secondary waste streams consist of the
tank material and rinse solutions. Spent
rinse solutions are generated in the clean-
ing procedure when water, steam, deter-
gent, or some other agent is used to clean
the tank. The rinse volumes may vary
depending on the nature and volume of
residuals found in the USTs. As noted
above, survey respondents reported rinse
volumes ranging from 100 galAank to one
third of the tank volume. Little information
was found on methods used to treat and
dispose of the secondary wastes gener-
ated. However, the treatment and disposal
of oil/water wastes is successfully accom-
plished by numerous demonstrated and
commercially available processes, such as
phase separation followed by incineration
of the organic phase and a two-step (e.g.,
physical/chemical and biological) treatment
of the aqueous phase.
Effectiveness of Cleaning
Procedures
The Phase I survey revealed no con-
tractor contacted knew just how clean a
tank their procedure(s) could achieve. Most
contractors believe that if they follow the
company's standard cleaning procedures,
then the tank will be "clean." Visual in-
spections of "clean" are also common.
When UST closure procedures preclude
the use of a manhole in the UST, visual
inspection of "clean" is quite- difficult. At
present, no standard measure of the clean-
ing effectiveness seems to have been set.
Phase II attempted to resolve this ques-
tion by actually visiting tank cleaning/re-
moval operations and characterizing the
residuals before and after cleaning.
Field Studies of UST Closures
This phase collected information at ac-
tual UST cleaning/removal sites, includ-
ing:
Background information on the UST
site that relates to the nature of the
residuals
Detailed descriptions of the cleaning
methods used
Estimates of volumes of tank residu-
als and secondary wastes
Hazardous characteristics and chemi-
cal composition of the residuals and
secondary wastes
Costs of cleaning and closure
Field case studies were conducted in
concert with a company that offered a
range of environmental services, including
UST cleaning and removal. The company
provided a list of six representative UST
closure jobs that met study objectives.
Permission was obtained from the site
owner/operator to monitor the job and per-
form sampling during the normal course
of the closure; cleaning techniques were
not modified for the study. The study fo-
cused on tanks containing gasoline and
No. 2 fuel oil. Table 1 summarizes back-
ground information on these tanks.
UST Removal and Cleaning
Procedures
Observers noted common steps in
cleaning procedures, e.g., vacuuming re-
sidual product (No. 2 fuel oil or gasoline)
from the UST into a tank truck; adding dry
ice to gasoline to displace oxygen with
carbon dioxide; excavating overlying soil
(at this point, pulling some tanks from the
pit); cutting a manhole in the top or side of
the tank to allow worker entry; scraping
tank interior (manually) to remove residu-
als; rinsing the tank interior with tap water
and vacuuming it into the tank truck dur-
ing rinsing; pulling the UST from the exca-
vation pit; and scraping tank exterior clean
before transport to a tank yard.
Characterizing Residuals
In general, 3 types of samples were
collected from each UST for laboratory
analysis: original fuel product (if present),
bottom residuals, and aqueous rinse solu-
tions. These samples were analyzed for a
series of chemical parameters and haz-
ardous characteristics, including specific
RCRA metals and VOCs.
-------
Table 1. Specifications of Underground Storage Tanks (USTs) Sampled
Site
^ No.
1
2
3
4
5
6
Size
(gai)
ą4,000
1,000
10,000
ą1,000
ą500
ą2,000
Fuel
Type
No. 2
Nb:2
No. 2
Gasoline
Gasoline
Gasoline
Material
Type
Steel
Steel
Steel
Steel
Steel
Steel
Condition
Very good, no rust
Fairly rusted
Good, some rust at ends
Rusty, but intact
Rusty, but intact
Very good, no rust
Age
(yr)
15
15
20
11+
20+
11+
Depth to
Groundwater
(ft)
Unknown
Unknown
Unknown
ą20
4
4
Depth to
Tank (ft)
20
4
ą3
ą4
ą2
ą3-4
Product
Volume in
Tank (gal)
4,400
800
94
ą90
ą2
ą55
Field Study Results
Volume of Residuals After
Cleaning
The first direct indication of effective
UST cleaning is the visual examination of
residual organic liquid, sludge, or aque-
ous rinse remaining in the UST after it
was cleaned. The residual volume esti-
mates of either organic liquid, sludge or
rinse varied between negligible amounts
and 3 gal of residual. These residual vol-
umes were less than 1% of the total tank
volumes. In addition, the volume of re-
siduals appeared to be independent of
the tank volumes. Any variation in vol-
umes of residuals is probably dependent
upon the daily variations in field condi-
tions and operating procedures followed
at a given site.
Analyses of UST Residuals
The second measure of effective UST
cleaning is the concentration of chemical
constituents found in the residuals remain-
ing in the USTs after cleaning.
Product
Laboratory analyses of the two types of
fuel products removed from the USTs in
this survey (i.e., gasoline and No. 2 fuel
oil) did not yield any unusual results (Table
2). VOCs, metals, TPH, and flash point
measurements were all within ranges that
are consistent with those for No. 2 fuel or
gasoline. As expected, the BTEX concen-
trations for gasoline were higher than those
for No. 2 fuel. Metal concentrations were
either below the detection limit or exhibit
some lead. The fact that the reported TPH
measurements on the fuel products did
not match 100% TPH (1,000,000 ppm)
does not necessarily reflect non-TPH con-
tamination in the fuel, since the specified
analytical procedure used a synthetic non-
fuel standard for instrument calibration.
The flash point measurements indicate that
the gasoline would be considered a haz-
ardous waste because of its ignitabil'rty
characteristics (flash point below 140"F).
The fuel oil would not be considered haz-
ardous by this characteristic.
Bottom Residuals
These materials were probably a com-
bination of settled petroleum products, tank
scale, and accidentally introduced soil. The
results of laboratory analyses performed
on this material (Table 3) were consistent
with its sources. TPH and VOC concen-
trations were slightly lower than the fuel
products, flash points were roughly similar
to fuel products, and metals concentra-
tions were higher than the fuel product.
The origin of the metals could either be
from settled impurities or additives in the
gasoline (such as tetraethyl lead), impuri-
ties in the tank steel, or constituents of
soil that was accidentally introduced into
the tank. (Laboratory personnel indicated
that high barium concentrations are often
seen in analyses of petroleum products.)
In addition to the routine TPH, metals,
and VOC measurements, the bottom re-
siduals were also subjected to a TCLP
extraction to assess what concentration of
metals, VOCs, and ABNs (Acid/Base
Neutrals) could potentially become mobile
in the presence of an acidic leachate.
TCLP results (Table 4) indicated that only
a fraction of the metals and VOCs present
were potentially mobile as aqueous, sol-
utes. Based upon these TCLP results and
the recently revised TCLP criteria, bottom
residuals from two of the gasoline tanks
would be considered hazardous waste.
sludges by the EPA. The regulatory levels
and exceedances are as shown in Table
5. The only unexplained TCLP result is
Table 2 Summary of Typical Analytical Results for Fuel Product in USTs Before Cleaning
Site
No.
1
2
Fuel
Type
No. 2
No. 2
TPH
(ppm)
788,000
702,000"
Flash
Point
fF)
>200
185
Metals
Detected
(Ppm)
BDL'
BDL
VOCs Detected
(ppm)
Toluene 743
Ethylbenzene 222
Totalxylenes 2,810
Benzene 37
Toluene 220
Ethylbenzene 150 .
Totalxylenes 977
3 No.2c
4 Gasoline
5 Gasoline
6 Gasoline
518,000" 25
485,000" 21
634,000" 23
Lead 5.3 Benzene
Toluene
Ethylbenzene
Lead 1,370 Benzene
Toluene
Ethylbenzene :
Total xylenes
BDL Benzene
Toluene
Ethylbenzene
Totalxylenes
12,000
30,800
53,700
17,700"
39,400"
* 13,900"
78,600"
13,000"
37,000"
14,500"
75,500"
BDL - below detection limit.
" Average of two values.
c No analyses performed
-------
Tabla 3. Summary of Analytical Results for Bottom Residuals In USTs During Cleaning
Flash
Site Fuel
No. Type
TP.H
(ppm)
Point
fF)
Metals Detected
(ppm)
VOCs Detected
(ppm)
No. 2
2 No. 2°
237,000 181 Arsenic 0.83' Toluene 110
Barium 5.7' Ethylbenzene 196
Lead 20.9' Total xylenes 993
3 No. 2 355,000 205 Arsenic
Barium
Cadmium
Chromium
Lead
4 Gasoline 114,000 45 Arsenic
Barium
Cadmium
Chromium
Lead
Silver
5 Gasoline Arsenic
Barium
Cadmium
Chromium
Lead
Silver
6 Gasoline
2.7
157'
2.3
12.7
' Benzene
Toluene
' Ethylbenzene
' Total xylenes
17
133
138
640
59.2'
25.8
23.9
19.8
51.3
2,230
2.2
8.4
22.8
13.5
50.4
232'
264'
Avoraga of two values. " No bottom residuals in tank. "
Benzene
Toluene
Ethylbenzene
Total xylenes
' Benzene
Toluene
' Ethylbenzene
' Total xylenes
No analyses performed
5.2
370
774
334
624
639
284
765
tal Organic Carbon (TOG), and pH. The
oil and grease surements reflect the pres-
ence of high molecular weight organics in
the fuel. BOD is used as a measure of the
amount of degradable organic material
present in the waste, and TOO is a surro-
gate measure of organic carbon present.
The pH range of the samples collected,
4.7-6.6, is consistent with the range in
natural waters. The tanks at sites 5 and 6
were washed with non-municipal ground-
water, which may account for the lower
pH measurements (4.7 and 5.4, respec-
tively).
Organic Vapor Concentrations
The concentration of organic vapors
found inside the tanks after cleaning indi-
cates the effectiveness of the cleaning as
well as the potential explosion hazard that
the tank may present. The concentration
of organic vapors was measured in three
of the tanks following the cleaning proce-
dures using an HNu organics analyzer
equipped with a photo-ionization detector.
The organic vapor concentrations in the
tanks ranged from 26 ppm to 250 ppm.
These concentrations are well below the
the presence of methylene chloride at site
No. 1; the chemical may have been intro-
duced during the laboratory analysis.
Aqueous Rinse
The rinse solutions analyzed in this sur-
vey were intended to simulate the rinse
water used during the final rinse of the
fuel tanks. As indicated in Table 6, the
TPH concentrations ranged from 4 to 379
ppm, and metals concentrations were ei-
ther below the detection limit or a fraction
of the concentrations found in the bottom
residuals. For example, at Site No. 4 the
concentration of lead in the rinse was
12.6 ppm whereas the concentration of
lead in the bottom residuals was 2230
ppm. VOC concentrations in the aqueous
rinse reflected the VOC concentrations in
the fuel product stored in the tank. Tanks
that stored gasoline had higher VOC con-
centrations than those that contained No.
2 fuel. The presence of low levels of
trihatomethanes, such as chloroform and
bromodichloromethane, in some of the
aqueous rinse samples probably reflects
the presence of trihalomethanes in the
public drinking water used to clean the
tanks in Sites 1, 2, and 3.
Additional tests of the aqueous rinse
solutions compared its quality with the
guidelines for discharge of industrial wa-
ters containing the following materials to
sewers serving Publicly Owned Treatment
Works (POTWs): oil and grease, 5-day
Biochemical Oxygen Demand (BOD), To-
Tablo4. Summary of TCLP Analyses on UST Bottom Residuals
Site Fuel Metals Detected VOCs Detected
No. Type (ppm)
Semi-VOCs Detected
(ppm)
1 No. 2
2 No. 2"
3 No. 2
4 Gasoline
5 Gasoline
6 * Gasoline
' Average of two
Barium
Cadmium
Chromium
Lead
Barium
Lead
Arsenic
Barium
Cadmium
Lead
Barium
Lead
values.
3.23
0.019
0.005
0.047
10.5
0.83
0.031
3.58
0.19
23.2
2.6'
0.34'
Methylene
chloride
Acetone
Benzene
Tetrachloro-
ethane
Toluene
Ethylbenzene
Total xylenes
Benzene
Toluene
Ethylbenzene
Total xylenes
Benzene
Toluene
Ethylbenzene
Total xylenes
Benzene
Toluene
Ethylbenzene
Total xylenes
" No bottom residuals in
0.24
20
0.23
0.49
0.69
0.15
0.82
0.15'
0.40'
0.158'
0.87'
29.7
23.6
2.3
14.3
23.1
32.1
4.8
23.2
tank. °
Naphthalene
2-Methyl-
naphthalene
Acenaphthylene
Diethylphthalate
Di-n-butyl-
phthalate
Bis(2-ethylhexyl)
phthalate
Naphthalene
Phenol
2-Methylphenol
2,4-Dimethyl-
phenol
Naphthalene
2-Methyl-
naphthalene
Phenol
Benzyl alcohol
2-Methylphenol
4-Methylphenol
2,4-Dimethyl-
phenol
Naphthalene
2-Methyl-
naphthalene
0.10
0.41
0.0002
0.033
0.044
0.044
0.170
0.14
1.12
0.39
0.22
0.83
0.51
0.024
0.63
0.81
0.26
0.20
0.028
No analyses performed
-------
Table S. TCLP Regulatory Levels and Exceedances
-EPA/TCLP Exceedances
Chemical
Arsenic
Barium
Cadmium
Lead
Benzene
.Benzene
Criterion (ppm)
5
100
1.0
5.0
0.5
0.5
Tank
4
4
5
Cone, (ppm)
None
None
None
23.2
29.7
23.1
lower flammable limits for gasoline (>1.2%
by volume).
Hazardous Composition of
Residuals
The Phase II field studies indicated that
residuals from gasoline tanks would typi-
cally be classed as hazardous waste be-
cause of their ignitability characteristic
{flash point below 140°F) and Toxicity
Characteristic Leaching Procedure (TCLP)
values for lead and benzene. In addition,
USTs containing gasoline residuals typi-
cally present vapors in concentrations
above the lower explosive limit and above
levels that would impair human health af-
ter even short-term exposures. Removal
of these vapors is absolutely essential to
eliminate rjsk from fires, explosions, and
the inhalation of toxic vapors. By contrast,
No. 2 fuel oil residuals were not found to
be hazardous based on ignitability (flash
points above 180°F) or TCLP criteria.
Bottom residuals from both gasoline and
No. 2 fuel oil USTs contained significant
concentrations of lead, barium, chromium,
cadmium, and arsenic. As expected, prod-
uct residuals from both also contained sig-
nificant concentrations of benzene, tolu-
ene, ethylbenzene and xylene (BTEX). The
BTEX fraction comprised 10-15 percent of
the gasoline residuals and 0.1-0.4 percent
of the No. 2 fuel oil residuals. Used aque-
ous rinses from tank cleaning operations
contained levels of total petroleum hydro-
carbons (up to 480 ppm) and BTEX (up to
70 ppm) that would likely bar their direct
discharge to sanitary sewers.
UST Cleaning and Closure
Costs
The costs of cleaning and closing (by
removal) USTs are highly variable, rang-
ing from under $1,000 to over $10,000
(1988) for individual tanks in the 1,000^
10,000 gal range. The range of per unit
tank size is a little narrower, 0.3-1.0/gal of
tank capacity in most cases. Extreme val-
ues of up to $36,700/tank and $8/gal of
capacity have been reported. The cost
variability relates to the site-specific time
and equipment requirements for tank
cleaning and closure, as well as the na-
ture and depth of covering, proximity to
structures and utilities, residuals volume,
required level of worker protection, equip-
ment availability, inspection logistics, sam-
pling needs, etc.
The cleaning method does seem to play
a significant role in the total cost. Costs of
labor, equipment and materials for 3 ma-
jor steps in tank cleaning and removal
have been estimated as high as $10,920
Table 6. Summary of Analytical Results for Aqueous Rinse Samples
5-day Oil and
Site Fuel TPH BOD TOC Grease Metals Detected
No. Type (ppm) (ppm) (ppm) (ppm) pH (ppm)
1" No. 2
2 No. 2 156' 210 109 6.6 BDLe
3 No. 2 379' 330 646 405' 6.1 BDL
4 Gasoline 20.3* 240 150 23.9' 6.0 Arsenic 0.047
Chromium 0.27
Lead 12.6
5 Gasoline 4.4' 2,165 1,168 12.5- 5.4 Cadmium 0.17
Chromium 0.33
Lead 4.2
6 Gasoline 74.3 35.0 33.7 83.1 4.7 BDL
VOCs Detected
(ppm)
Chloroform
Bromodichloromethane
Benzene
Toluene
Ethylbenzene
Total xylenes
Chloroform
Benzene
Toluene
Ethylbenzene
Total xylenes
Benzene
Toluene
Ethylbenzene
Total xylenes
Benzene
Toluene
Ethylbenzene
Total xylenes
Benzene
Toluene
Ethylbenzene
Total xylenes
0.016
0.009
0.009
0.082
0.087
0.332
0.009'
0.015'
0.54'
0.039'
0.395'
4.98
12.0
3.57
14.0
11.5
28.1
7.32
24.0
0.848
31.4
1.28
3.89
' Average of two values. " No sample collected.
Below detection limit.
ŁU.S. GOVERNMENT PRINTING OFFICE: 1993 - 7SO-07I/SOOM
-------
the high end of the range of reported
1988 costs indicating either a more
complete coverage of all costs (e.g., 1988
costs may have excluded backfill, tank
disposal, etc.) or an overly conservative
approach. In this hypothetical forecast, la-
bor accounts for 33% of the costs; equip-
ment, 61%; and materials, 6%.
Conclusions
This study has developed greater un-
derstanding of the technical aspects of
UST cleaning and closure. The informa-
tion collected from the Phase I interviews
indicated that current tank cleaning meth-
ods appear to satisfactorily clean most
gasoline and light oil .tanks, even though
little to no formal guidance is available on
UST cleaning. The sampling, analysis, and.
physical measurements from the Phase If
study verified the effectiveness of the
cleaning procedures and documented the
steps used to meet 3 criteria:
Avoiding hazards from explosions and
toxic vapors;
Complying with the requirements of
the Department of Transportation
(DOT); and
This Project Summary was prepared by staff of Camp Dresser & McKee, Inc., Cambridge,
MA 02192-1401.
Anthony N. Tafurlis the EPA Project Officer (see below).
The complete report, entitled "Technical Aspects of Underground Storage Tank Closure,"
(Order No. PB92-161199/AS; Cost: $19.00, subject to change) will be available only
from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Edison, NJ 08837
Meeting the cleanliness and explo-
sion hazard requirements of the dis-
posal facility.
The Phase II study also quantified the
volumes of tank residuals and secondary
waste,' in addition to characterizing the
contents. After cleaning, only one to three
gallons of unrecovered rinse remained in
USTs, posing a relatively low level of haz-
ard. Post-cleaning vapor levels were safe
in terms of explosion potential and acute
toxicity.
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati, OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Official Business
Penalty for Private Use $300
EPA/600/SR-92/057
------- |