New England Interstate
Water Pollution Control
Commission
Boott Mills South
1OO Foot of John Street
Lowell, Massachusetts
01852-1124
Bulletin 43
March
2OO3
Report On Federal & State Programs To Conirol Leaking Underground Storage Tanks
THE TORTOISE AND TOE HARE REVISITED
Reaching the Goal Effectively and Efficiently:
The Path Not Yet Taken
W! all know Aesop's fable of the boastful hare and
"he slow but persistent tortoise. We all kni^w
•oho wins the race and. why. For the purposes of
this article, I've recast the story and given the character^ a
more challenging and important goal than simply winning a
race. I've morphed the hare into the petroleum marketing
industry and the tortoise into the national UST regulatory
program. Their goal is to protect human health and the envi-
ronment, which is defined as reducing the number ofpetfo-
leum releases from USTs to the absolute minimum. And I've
made the story more politically correct by having the tortoise
and the hare be part of the same team, struggling to reach ihe
same goal. At issue in this story is how our players go about
reaching their goal—how to keep the team members evenly
matched so that they can more effectively and efficiently riin
the race. >
• continued on page 2
Inside
^(Jg People and UST Systems _
j j| / Tpt^*l. ' - "~T ." ~~ • ." " ' ------
;^Overpressurization Problem Baffles Tank Owner
t An Overview of MTBE and Other Oxygenates in Fuel
Lbo We Need to Worry About IWTBE Acid Hydrolysis?
i RBCA is not RBCA
I & A: Pay for Performance
Tank Systems in a Jam
uST-Related Explosions in Kentucky and Pennsylvania
J Building a Better Internet Presence
J^Should LUSTUne (Jp Electronic?
-------
LUSTLine Bulletin 43 • March 2003
• Tortoise and Hare from page 1
The Hare
One of my favorite stories regarding
service station construction is in a
1981 American Petroleum Institute
publication titled The Origin and Evo-
lution of Gasoline Marketing. The story
tells how Shell Oil expanded in Cali-
fornia in the early 1920s:
E.H. Sanders, who was assistant
division manager of the Central
division at the time, loves to tell
how they invaded the San Jose-to-
Santa Barbara territory, building
eight depots (bulk plants) and a
hundred service stations all in a
matter of six weeks. Sanders and
Claude Donaldson, the Central
division's traffic manager, first
rode the 260-mile stretch of road
between the two towns and picked
likely sites for depots and stations.
They were followed closely by real
estate men who, to save time and
expense, leased rather than bought
most of the desired sites. Then
came the construction crews in
Jjonald Poitak, NEIWPCC Executive Director ^
\ i £*ynn DePont, EJFA Project "Officer
l I JJ I U h ML 1 IHi
::'UISTjfme is a product of the New England
~" "crslifs Water Pollution Control Commis-
'-''uce'd through a
!„,: IJgfigsn §E|WPCC" and the U.S.
IngnylronniienEn Protection Agency.
UP LUSTLme is issued as a communication
Jj;JJ!MFMK5Wale SuBtitle I RCRA
'l"'" ........ iiiS m iSinir-i—mminin ..... M, npr,..ip.niipiiiip.|iiBmiiiniiiinn ........ •IJ ------- —
je opinioi
Jt$ ffjgseof the authors and do not neces-
iiiiiii iarfly"re||ecf"the opinions of NEIWPCC.
i !'l"'Hli" ........ biilKife&liriBliirSi Tajm liSifnir-i—mminin ..... M, npr,..ip.niip|jiiiip.|iiBmiiiniiiinn ........ •IJ ------- J— —
,i;:|iLazardous & Solid Waste 'Amendments
-ir-r - «wisf"j omulgaHprTprocess.
JTL»«e is produced to promote
'" " ' '"""' Se on UST/LUST issues.
j-iSliis'pubHcanbn may be" copied.
!**"~r * jcreditto NEIWPCC.
llENllmBIITIIIIITllllillllillllHlfc^
fCCwas esfabTished by an Act of
..'tfieNprtheast United States
jiffi_cogrdjnation of the multi-
~™ ntal activities
.153 y?, v.~>.aect!culj Maine,
HnsK.ilMsetts, Nfew Hampshire,
fprE"Khode liTand, and Vermont.
LOSTUne t« printed on Recycled Paper
I ; ; :
relays. First, crews to dig holes in
the proposed service station lots
and bury the underground storage
tanks the station would need. Next,
crews to pour the concrete founda-
tion for the station, leaving bolt
studs sticking out of the concrete.
Meanwhile the crates containing
the complete "A" station would be
delivered. As soon as the concrete
had hardened, a crew arrived to
install the pumps and bolt the sta-
tion in place, anchoring it to the
foundation by means of the stud
bolts protruding from the concrete.
Then came the paint crew to apply
a coat of red and yellow, and the
station, once its storage tanks had
been filled from the new depot,
was in business "about ten days
from start to finish."
This orgy of station construction
would continue until World War II.
In fact, 1939 was the peak year for the
number of retail gasoline outlets in
the country, when some 400,000 retail
outlets were active. The downside of
this huge number of facilities was
that the average facility was pumping
about 40,000 gallons of gasoline per
year. By comparison, National Petro-
leum News recently estimated that
there were 170,678 retail outlets
active in the first quarter of 2002, with
the average convenience store selling
1.3 million gallons per year. (Conve-
nience stores represent 73 percent of
the retail outlets in the U.S.)
Gasoline marketing continues to
be a dynamic industry marked by
rapid and continual change. The
1950s saw another construction
boom that lasted through the 1960s,
only to come to an abrupt halt with
the oil embargoes and gasoline allo-
cation and price controls of the 1970s.
The 1970s also saw the introduction
of unleaded fuel and the rise of self-
service fueling. The 1980s saw the
transition from the traditional service
station, which offered vehicle main-
tenance and repair, to the conve-
nience store model, which features
cigarettes, soda, beer, and snacks
along with the fuel.
The 1980s also saw the introduc-
tion of federal regulation of under-
ground storage systems, although
most of the tank upgrade and
replacement activity took place in the
1990s, except for the major oil com-
panies, which got started earlier. The
1990s also saw increasingly high-vol-
ume gasoline retailing facilities, cul-
minating in the addition of gasoline
to the products offered by "Big Box"
retailers, such as Wai Mart. The late
1990s also saw the consolidation of
major oil companies and a continued
decline in the total number of retail
gasoline outlets.
While the pace of change in
petroleum marketing is generally
rapid, underground storage technol-
ogy evolves much more slowly. This
is easy to understand because petro-
leum marketing has always been an
extremely cost- and customer-con-
scious industry, and not many cus-
tomers buy fuel based on the type of
storage system a facility has installed.
There have been three major gen-
erations of UST systems in the U.S.:
• First generation: 1910-1985
These storage systems consist of steel
tanks with galvanized steel pipe and
a suction pump (first manual, then
electric) to deliver fuel. The bare-steel
tank reigned at petroleum retailing
facilities virtually unchanged for 70
years, except for size and the change
from riveted to welded construction.
Galvanized piping had a similarly
long life.
• Second generation: 1965-Present
These storage systems consist of
fiberglass and corrosion-protected
steel tanks with fiberglass pipe and
submersible pumps. Though fiber-
glass and factory-engineered corro-
sion-protected steel tanks have been
available since 1965 and 1969, respec-
tively, their market penetration was
relatively small until the federal
interim prohibition, in effect, forbade
the installation of bare-steel tanks
after May 1985.
• Third generation: 1985-Prosent
The latest system feature is double-
walled, corrosion-resistant tanks and
double-walled piping (especially
flexible pipe after 1990) with contain-
ment sumps and submersible pumps.
Though double-walled tanks were
developed in Europe in the mid
1960s, the technology was slow to
cross the Atlantic because of cost.
The dates I cite are approximate,
and both second- and third-genera-
tion systems are in common use
today. Second-generation storage
systems, which represent technology
-------
March 2003 • LUSTLine Bulletin 43
tkat is over 30 years old, are accepted
by the federal rules, even for today's
installations.
The Tortoise
Storage system regulation is the tor-
toise in our fable. Environmentally
based rules regarding underground
storage began at local levels in the
late '70s but rose rapidly to the fed-
eral level with the signing of the
Subtitle IRCRA amendments, estab-
lishing a federal underground stor-
age system regulatory program in
November 1984. This resulted, after a
four-year gestation period, in the
promulgation of the federal tank reg-
ulations, which in turn established a
10-year program for upgrading the
nation's storage systems.
While the introduction and sign-
ing of the 1984 RCRA amendment
into law happened with lightning
speed, the development of the regu-
lations was deliberately slow in an
attempt to create a defensible and
workable rule. Since publication on
September 23, 1988, the rules have
remained remarkably stable. Except
for a small change to the overfill pre-
vention portion of the regulations in
1991, the federal UST rules have not
moved a whisker. This came home to
me as I reached for my taped-
together, dog-eared copy of the Fed-
eral Register and realized that this
14-year old document is still perfectly
viable as a reference for technical
standards for UST systems today. Is
this a good thing?
The Revisit
Consistency has many proponents.
"Slow and steady wins the race" is
our tortoise's motto. Consistency is
touted as a fundamental quality of
good parenting. Consistency is
important in the workplace so that
workers and customers know what
to expect. There is something to be
said for consistency in regulations,
for much the same reason as in par-
enting and in business. But consis-
tency has its limits. The parenting
techniques appropriate for a toddler
will not work on a teenager. A com-
pany that has consistent policies but
is consistently losing money would
be wise to assess whether consistency
with outmoded policies was con-
tributing to the red ink. A carved-in-
stone regulatory program that is
operating in a fluid marketplace may
find that its effectiveness is eroding.
When j the original tank rules
were written, the focus was on
updating pasting UST technology.
This was clearly appropriate because
the dominant technologies for tanks
and piping in use at the time were
hopelessly* outmoded. In recent
years, however, the light has dawned
on many in the UST world that tech-
nological issues must now take a
back seat to operational issues.
It has slowly been recognized that
the best of hardware will not function
well unless, as Ray Powers of the
Pennsylvania Department of Environ-
mental Protection recently stated in an
e-mail, it is installed properly, pro-
grammed correctly, well maintained,
and effectively responded to when an
alarm is triggered.
We have upgraded our hard-
ware, but this has not entirely solved
the problem because the rules did not
look comprehensively at -what would
happen aftjar the new hardware was
installed. By and large, the rules
placed thej burden of defining the
maintenance needs of leak-detection
equipmentjon ^e ecluipinent manu-
facturers themselves by requiring that
equipment be maintained according
to the manufacturers' specifications.
Equipment manufacturers have
tended to shy away from this respon-
sibility, preferring to imply, if not
overtly state, that their equipment is
"maintenance free." With few excep-
tions (e.g., {he annual testing of line-
leak detectors and the triennial
monitoring; of CP systems), the origi-
nal rules adopted the "bury it and for-
get it" attittide that was a substantial
contributor jto the UST problem in the
first place. !
While jsome technology-based
issues are clearly still with us today,
the weak link in today's strategy for
protecting j human health and the
environment from leaking under-
ground storage systems seems to be a
human one (e.g., ignoring equipment
maintenance and the burying of
heads in the sand when leak-detec-
tion equipment alarms sound). But
while the ijssues have evolved, the
regulatory strategy for dealing with
the issues has remained solidly chis-
eled in stone.
A Strategy for the Tortoise
As a resulj: of the different paths
taken by our hare and our tortoise, it
is not at all clear to me that our tor-
toise is doing its share to protect
human health and the environment.
Here is my short list of things that
could be addressed in rule revisions
to help our tortoise meet the chal-
lenge:
• Address maintenance of installed
hardware. A few states have ad-
dressed this issue by requiring
annual inspections of leak-detection,
and in some cases, overfill-preven-
tion hardware. A study that I con-
ducted for the State of Maine a few
years ago documented that checking
operational status of hardware is a
worthwhile endeavor—it was con-
servatively estimated that some 29
percent of facility inspections discov-
ered significant problems with corro-
sion-protection, leak-detection, spill-
containment, or overfill-prevention
hardware.
The Maine study also revealed
that merely inspecting the hardware
is not enough. It found that 39 per-
cent of tank owners failed to address
deficiencies that were pointed out to
them in the inspections. The inspec-
tion requirement must be accompa-
nied by enforcement measures with
sufficient teeth to get the problems
fixed. (Maine subsequently enacted
regulations to deny product deliver-
ies to tanks with problems that are
not corrected.)
• Increase scrutiny of some UST
technologies. The Iowa tank lining
study some years ago indicated that
interior tank lining may have a sub-
stantially greater failure rate than
U.S. EPA believed in 1988. (See
"Iowa's Tank Lining Study," LIZ'S T-
Hne Bulletin #30.) Perhaps the initial
lining inspection should be con-
ducted after five years rather than the
ten specified in the current rule.
Anecdotal evidence indicates
that many retrofitted cathodic-pro-
tection systems are also not function-
ing as they should. Perhaps an
annual monitoring frequency, rather
than every three years, would be
more appropriate. Given the large
number of older steel storage systems
still in service, should increased
attention be focussed on internal cor-
rosion?
California is learning that a very
high percentage of installed tank and
dispenser sumps, elements that are
• continued on page 4
-------
LUSTLine Bulletin 43 • March 2003
• Tortoise and Hare from page 3
critical to the functioning of sec-
ondary containment systems, are not
tight. It is not clear whether this is
due to initial installation problems or
the subsequent "aging" of the sys-
tem. Either way, some periodic eval-
uation of the integrity of these
components would seem to be critical
to the effectiveness of secondary con-
tainment.
The long-term performance of
flexible piping has also recently been
called into question. (See "Flexible-
Pipe Concerns Drive Home the Need
for Tank-Owner Vigilance," LUST-
Line #42.) It may be time to consider
new measures, such as improved
performance standards for flexible-
piping technology or increased vigi-
lance over existing installations to
address these concerns.
• Plug leak loopholes. The existing
rules do not include leak detection
for submersible-pump manifolds or
for dispensers. Yet field experience
indicates that these components can
be frequent sources of release. (See
"Field Notes," LUSTLine #41.) With
the exception of secondary contain-
ment, and in some cases statistical
inventory reconciliation (SIR), no
other commonly used method of leak
detection is effective on these compo-
nents. Another loophole is that the
rules never considered the possible
impact of vapor leaks from UST sys-
tems. California has recently docu-
mented that this type of leak is
commonplace and will soon be
requiring that UST systems be vapor
tight. Is this an issue that should be
considered nationally?
• Improve regulatory compliance.
Whether such an effort consists of
restructuring state cleanup funds (see
"Square Operators, Round Tanks,
and Regulatory Hammers: A Pe-
troleum Marketer's Perspective,"
LUSTLine #42), developing some
requirements for operator certifica-
tion, or some other technique, market
forces that encourage compliance
should be brought to bear on the tank
owner. Compliance with tank regula-
tions must be strongly and directly
linked to the tank owner's bottom
line if the regulations are ever to be
effective. It might also be worthwhile
to review the rules from a "human
engineering" perspective to find
ways to simplify compliance so that
reliance on fallible humans can be
minimized. (See "People and UST
Systems" on page 8.)
• Tighten leak-detection standards.
The present-day rules have petrified
leak-detection technology by specify-
ing generous leak-detection stan-
dards of 0.2 gallons per hour (gph)
(1,752 gallons per year). While the
official interpretation of the rule is
that no leak is acceptable, the de facto
use of the rule is that anything less
than 0.2 gph is not a leak. Most volu-
metric leak-detection methods avail-
able today are capable of reaching
accuracies of 0.1 gph. While this is
still equivalent to 876 gallons per
year, it is half the 0.2 gph leak rate
that is the standard today.
Test reliability could be im-
proved by calculating the probability
of detection (Pd) and probability of
false alarm (Pfa) for each volumetric
test conducted, which would give a
much better indication of the reliabil-
ity of a specific test result. The limita-
tions of various technologies should
also be evaluated and spelled out by
manufacturers or regulations. For
example, inventory control, auto-
matic tank gauges, statistical inven-
tory reconciliation, and line-leak
detectors decline in effectiveness as
throughput increases, but the limits
of their effectiveness have not been
clearly specified. With the prolifera-
tion of very high-volume retail facili-
ties, it is important that tank owners
know ahead of time which leak-
detection options will work for them.
• Revisit the spill-containment and
overfill-prevention issue. As I have
described several times in LUSTLine
(Bulletins #31, #21, #18), I do not
believe that overfill prevention as it is
currently practiced is safe, effective,
or efficient. (See also "Field Notes"
on page 11.) Surely something better
can be promoted by redefining the
regulatory criteria.
• Update the rule to reflect knowl-
edge and experience gained in the
last 14 years. Do we still really need
groundwater and soil-vapor moni-
toring as leak-detection options?
Should operating guidelines for SIR
be specified? Should language
regarding the 1998 deadline be
changed to reflect that this date has
passed? Should issues concerning the
appropriate leak rate for testing line-
leak detectors be addressed? Can a
sensor in a secondary-containment
sump take the place of a line-leak
detector? Should specific require-
ments be set for electronic line-leak
detectors? These questions must be
answered in a process that leads to an
improved rule.
Rule revision is not a chore that
many relish, but some states find that
it is worth the effort. These states
have evaluated the effectiveness of
their regulations and have made
adjustments accordingly numerous
times. Florida's rules, originally
enacted in 1984, were revised in 1992,
1994,1996, and 1998. California's first
rules went into effect in 1985. Since
then, the technical rules have been
modified in 1991, 1994, 1998, 2001,
and 2002. Maine's first UST rules
were adopted in 1986 and modified
in 1987, 1990, 1991, 1996, and 2002.
Why have these states made changes
to their rules? (See "Keeping the Tor-
toise in Shape" on page 6 for their
stories.)
But If If Ain't Broke...
"But wait!" you say. "The data show
that the number of new releases is
dropping fast nationwide. The cur-
rent rules are working just fine." (For
the latest U.S. EPA data, see "Memo
from Cliff Rothenstein to UST
Regional Division Directors," Decem-
ber 23, 2002, FY 2002 End-of-Year
Activity Report.)
OK, let's look at those statistics.
(If you want to take a look, go to
www.epa.gov/oust.) Yes, the data
show a significant decline in the
absolute number of new releases
being reported to U.S. EPA. But you
would expect that to be the case
because the number of active UST
systems has also decreased dramati-
cally since 1988—from 2 million
tanks in 1988 to 700,000 active USTs
today. If we have truly been making
progress in protecting human health
and the environment, then there
should be a measurable decline in the
number of leak incidents as a percent-
age of the remaining active UST popula-
tion.
Figure 1 is a plot of data taken
from U.S. EPA semiannual reports
from 1991 through 2002. Though the
data are somewhat erratic, there does
appear to have been a declining trend
(based on the linear regression line)
in the number of reported releases as
-------
March 2003 • LUSTLine Bulletin 43
a percentage of the active facility
population of from approximately 5
percent to around 2 percent. In other
words, in 1991, the number of newly
reported releases equaled about 5
percent of the active UST-facility
population, while in 2002, the num-
ber of newly reported releases
equaled about 2 percent of the active
UST-facility population.
However, these data don't tell
the whole story for two reasons. First,
the data have a lot of scatter, and the
trend is not a strong one. Second,
because historically most leaks have
been detected by storage system clo-
sure. Because the number of closures
has been declining since 1993, the
corresponding decline in reported
releases may well be an artifact of the
way releases are discovered rather
than a true estimate of the actual rate
of releases at active facilities today.
(See Figure 2.) Of course, the ultimate
answer would be to go out and test a
random sample of the active UST
population.
Though not exactly a random test
because a large number of tank own-
ers refused to participate in the study,
182 storage systems in California were
tested using enhanced leak detection
(a version of a Tracer test). (See the
report at: http://www.swrcb.ca.gov/
cwphome/ust/docs/fbr/FBR_Final
_Report.pdf.) The study found only
one small liquid leak in a piping run
but found that 61 percent of the sys-
tems tested had released detectable
amounts of hydrocarbons along with
the tracer compound via vapor
releases, primarily from the top of the
underground storage tank. But Cali-
fornia is a state that requires annual
inspections, and only 10 percent of the
systems tested were entirely single-
walled, so the leak status of California
storage systems may not be represen-
tative of the national norm.
A Petroleum Equipment Institute
(PEI) estimate of UST system perfor-
mance, though clearly not quantita-
tive, is considerably more pessimistic
about UST-system releases. (See
"Field Notes," LUSTLine #41.). The
PEI members estimated that they
would find a total of 47 leaking com-
ponents inside 100 dispensers and 44
leaking components associated with
100 submersible pumps.
The point is we really don't
know with any degree of certainty
how today's UST systems are per-
Figure 1
Tfcnd.-;of Newly Reported Releases Over Time
The "a" and- "b" suffixes after the year indicate data points for the first and second halves of the federal
fiscal year, fhe original EPA data indicate the number of active tanks. The active tank data were con-
verted to a facility count by dividing by 2.86, the tank/facility ratio used in the preamble to the federal
rule. The number of newly reported releases for each six-month interval was divided by the estimated
number of active facilities for the same time interval and multiplied by 100 to arrive at each data point.
R9ure 2 | i G^mftarison of the Trends of Newly Closed
I, Fabiljtfes and Newlyi Reported Releases Over Time
Newly Reported Releases
Newly Closed Facilities
The "a" and "b" suffixes after the year indicate data points for the first and second halves of the federal
fiscal year. The original EPA data indicate the number of newly closed tanks. The newly closed tank data
were converted to a newly closed facility count by dividing by 2.86, the tank/facility ratio used in the pre-
amble to the federal rule. Because most new releases are discovered via UST facility closure, the parallel
trends in newly closed facilities and newly discovered releases over time may indicate that the reduction
in releases discovered is more closely related to the reduction in the number of facility closures than the
effectiveness of the UST program in reducing leaks.
forming. Without reliable data on the
number of actively leaking UST sys-
tems and the nature of the leaks, we
cannot know how far from the finish
line our hare and tortoise presently
are. Perhaps another area for rule
revision is to put in place a mecha-
nism that will generate the data we
need to chart the team's progress.
Consistency has much to recom-
mend it. Biit in a world where the
only constant is change, consistency
must not be! confused with paralysis.
Businesses must evolve to survive. Is
this not true, for regulations as well? I
believe that! regulations should con-
sistently adapt to the changing cir-
cumstances, knowledge, and experi-
ences encountered on the path to
achieving the regulatory goal. The
tortoise, after all, did not win the race
by standing still. •
Marcel Moreau is a nationally recog-
nized petroleum storage specialist
whose column, "Tank-nically Speak-
ing," is a regular feature of
LUSTLine. Marcel would welcome
some interactive dialogue on "The Tor-
toise and the Hare Revisited." He can
be reached at
marcel.moreau@juno.com.
-------
LUSTLinc Bulletin 43 • March 2003
Keeping; the Tortoise in Shape
by Ellen Frye
Several states have revised their UST regulations from time to time since they were
originally promulgated. (See "The Tortoise and the Hare Revisited" on page 1.)
I asked UST program managers from three such states—Maine, California, and
Florida—ivhy they did this, why they feel it is important to continue to do this, and how
they evaluate the effectiveness of their rules. For the states that can be no more stringent
than the federal rules, this is not meant to rub salt into the wound. Clearly, these three
states had the authority to move ahead on what they deemed were necessary changes.
Maine
"The bottom line in this program has
got to be preventing discharges and
eliminating public exposure," says
George Seel, Maine Department of
Environmental Protection (MDEP)
Director of Technical Services,
Bureau of Remediaton and Waste
Management. "Unless you've got
your head in the sand, you've got to
adapt your program to eliminate the
problems. To make any necessary
revisions to our regulations, we eval-
uate our program on an ongoing
basis, as data becomes available. We
try to answer three basic questions:
How many new discharges have we
had? What were the causes? What
were the environmental, public
health, and fiscal (insurance pro-
gram) impacts? If s all about the
lessons we learn answering these
questions."
Over the years, MDEP has con-
ducted a number of studies instigated
by what they have found during rou-
tine compliance inspections and
release investigations. For example,
compliance inspections indicated that
many required annual inspections
were not being conducted and that
defective equipment was not being
repaired. MDEP commissioned a
study to review the annual inspection
reports of a random sample of UST
facilities. The study confirmed that a
large percentage of owners/operators
were not doing what they were
supposed to do with respect to
annual inspection requirements (e.g.,
cathodic protection monitoring, leak-
detection equipment inspections).
Furthermore, when problems were
found during annual inspections,
about 40 percent of the time they
were not corrected.
Based on this information, MDEP
got approval from the legislature to
institute a third-party inspection pro-
gram to ensure that annual equip-
ment inspections are conducted,
deficiencies are corrected, and a certi-
fied report of the inspection is sub-
mitted to MDEP.
A review of expenses revealed
that about about 40 percent of the
agency's annual remediation budget
was spent on 1 percent of the dis-
charge sites in the state. This was pri-
marily because the sites located in
sensitive geological areas require a
high level of long-term groundwater
and drinking water remediation.
Seel says they have learned that
the primary causes of poor UST facil-
ity performance are equipment fail-
ure and human error or negligence
by owner/operators. "We've found
that even the best-designed facility
can still have discharges," says Seel.
"Within the first five years of a facil-
ity's existence, we have found that it
is likely to have at least one dis-
charge—and many have multiple
discharges. The problems are often
the result of human error, and we've
concluded that you can't really cor-
rect human error problems." (See
"People and UST Systems" on page 8.)
Because some level of UST fail-
ure is predictable and seemingly
inevitable, MDEP concluded that
there are places in Maine that
shouldn't have gas stations. The leg-
islature was provided with docu-
mentation to this effect, and facility
siting standards were adopted. (See
"There Ought to Be a Law: Maine's
New UST Siting and Inspection
Lavfs,"LUSTLine#38.)
California
"I believe that in any type of work
you do, in order to be successful, you
have to evaluate what you are doing
to determine if it is effective," says
Shahla Farahnak, Chief of Leak Pre-
vention at the California State Water
Resources Control Board (SWRCB).
"If you need to make the changes, be
willing to make the changes—and
that is not easy. It is very time-con-
suming. You have the challenge of
coming up with an ideal of what
needs to be done and ultimately bal-
ancing that with what you want to
achieve in terms of environmental
protection."
The SWRCB has undertaken var-
ious research projects to determine
how effective a specific regulatory
standard or requirement is when
applied in the field. (See these reports
at http://www.swrcb.ca.gov.cwphome/
ust/docs/fbr/index.) Based on the
findings of these projects, the SWRCB
has moved forward to make changes.
"We need numbers and documenta-
tion to justify our recommendations,"
says Farahnak.
For example, in 1998 and 1999, an
Advisory Panel evaluated the effec-
tiveness of the state's UST system
upgrade requirements in terms of
protecting the environment. (See
"California's Field-Based Research
Project Looks for Answers," LUST-
Line Bulletin #38.) That effort led to
the state's new UST Leak Prevention
and Enforcement Provisions, which
went into effect on January 1, 2003.
(See "News from California," LUST-
Line #42.)
Florida
"Technology changes. We have to
keep up. Reference standards change.
We have to keep up," says Marshall
Mott-Smith, Administrator of the
-------
March 2003 • LUSTLine Bulletin 43
Florida Department of Environmen-
tal Protection (FLDEP) Bureau of
Petroleum Storage. "Plus, state agen-
cies need to be honest with them-
selves about what they see in the
field and make changes that will pre-
vent incidents and discharges."
FLDEP is substantially revising
its 1998 UST rules to not only keep up
with technology and reference-stan-
dard issues but also to make the rules
easier to understand. In addition,
FLDEP wants to resolve several prob-
lem areas in the rules, eliminate tables
and deadlines that have passed, and
address problems encountered with
flexible piping and piping sumps.
Many of the proposed changes
were based on the preliminary
results from a Cause of Leak Study
that FLDEP is conducting in conjunc-
tion with the U.S. EPA Office of
Underground Storage Tanks. In this
study, reports from district and
county agency leak investigations
have been reviewed with an eye
toward gathering data to improve the
rules. Early results from the study
indicate that in today's UST systems,
piping is the major cause of leaks, fol-
lowed by dispensers. (A future issue
of LUSTLine will carry the results of
the study.)
Mott-Smith says that in many
cases, piping sumps are becoming
primary containment. These sumps
are real headaches for a variety of
reasons, including inadequate de-
sign, poor installation, bad entry
boots, and problems with some types
of flexible polyethylene piping. He
says that the state's secondary con-
tainment requirement (a rule revision
instituted in 1992) was a critical
departure from the federal rule,
which doesn't require UST systems
to have that second barrier.
"If you see free product between
the walls of a double-walled UST sys-
tem," says Mott-Smith, "you know
that second wall was the only thing
that prevented the product from get-
ting into the environment. You have
a leak, but you don't have a release,
and that makes all the difference in
the world." •
For more information on FLDEP's
= proposed rule revisions, go to
www.dep.state.fl.us/waste/
categories/tanks.
1
J
UL Proposes Changes to
Standard on Nonmetallic
Underground Piping
On December 6, 2002, a stan-
dairds technical panel, made
up'of some 30 state regulators,
industry experts, and representatives
of nonmetellic piping manufacturers,
met in Norfhbrook, Illinois, to discuss
proposed changes to Underwriters
Laboratories' (UL) Nonmetallic Un-
derground1 Piping for Flammable Liq-
uids Standard (UL 971).
The proposed standard ad-
dresses two key aspects of the piping:
mechanical failure and permeation.
UL released the mechanical portion
of the proposed standard for com-
ments on ^ebruary 4 and organized
an ad hoc committee to further
address permeation. Under discus-
sion is the proposed permeation
value of 0.1 g/m2/day, the methodol-
ogy used to analyze this value, and
the proposed new test method for
measuring permeability. The com-
mittee is considering these issues in
conjunction with the California State
Water Resources Control Board.
UL expects to publish the stan-
dard in early spring (the comment
period ended on March 4, 2003). The
standard then becomes effective 18
months after publication, giving
makers of flexible piping the oppor-
tunity to have their products tested
to meet the standard. For more infor-
mation, contact Maggie Carroll at:
marguerite.e.carroll@us.ul.com. •
Mississippi and Florida Notify
UST Owners and Contractors
about Flex-Pipe Concerns
Both the Mississippi Department
of Environmental Quality and
the Florida Department of Envi-
ronmental I Protection have notified
UST owner/operators and installa-
tion contractors about concerns with
thermoplastic flexible-piping sys-
tems. (Seje "Flexible-Pipe Con-
cerns..." h} LUSTLine #42.) You can
view Mississippi's letter and related
photographs at: www.deq.state.ms.
us/newweb/homepages.nsf. If this
link does not work you can go to
www.deq.state.ms.us and click on
the "Underground Storage Tanks"
link.
To view the "Special Notice to
Storage Tank System Owners and
Operators about the Use of Thermo-
plastic Flexible Piping in Florida," go
to: www.dep.state.fl.us/waste/cate-
gories/tanks.
U.S. EPA has posted comprehensive data concerning reformulated gasoline
(RFG) fuel properties and emissions performance averages by year/season
and RFG area. The tables and graphs at the site are based on EPA's analysis of
data generated from surveys conducted by the RFG Survey Association, an
association of refiners, importers, and blenders, as a requirement of EPA reg-
ulations. The site provides detailed data on all gasoline oxygenates.
To access "Information on Reformulated Gasoline Properties and Emissions
Performahce by Area and Season," go to:
www.epa.gov/otaq/regs/fuels/rfg/properf/rfgperf.htm.
-------
LUSTLine Bulletin 43 • March 2003
People and UST Systems
By Richard S. Bradley, Jr.
Over drinks at business dinners,
we discuss industry horror sto-
ries with associates. We shake
our heads in disbelief at the ignorant,
lazy—or worse, unlawful—behavior of
people. Together we count the cost to
individuals and communities, and occa-
sionally the damage done to the environ-
ment, knowing there were opportunities
to prevent the incident beforehand.
This is the first installment of a
multipart series examining the role of
human behavior in the management of
UST systems. Human behavior plays a
dominant role in our ability to manage
those systems for the prevention of conta-
mination and the protection of human
health and the environment. One of my
favorite dinner and wine stories is ger-
mane to this topic. It concerns a major
gasoline release from the piping of an
UST system caused by the tank owner's
failure to search for the root cause behind
an ongoing alarm on his automatic tank
gauge (ATG) console.
A Story of System Failure via
Human Error
This UST owner had a retail location
that pumped a high volume of gaso-
line every month. He attributed years
of success to an unwavering focus on
his customers, resulting in a very
loyal clientele. He understood that
interruptions to his business were an
inconvenience to his customers. Con-
sequently, interruptions had a clear
impact on his bottom line.
So, when it came time to replace
his UST system, he was determined
to purchase state-of-the-art equip-
ment with the most up-to-date tech-
nology. He wanted an UST system
that would allow him to operate his
facility without the interruptions
resulting from product releases and
associated contamination.
He installed redundant backups
for preventing and detecting releases
of product. He put in double-walled
tanks, double-walled piping, sec-
ondary-containment sumps for the
submersibles, sensors in the sumps,
and high-tech electronic line-leak
detectors capable of detecting leak
rates smaller than the regulatory
requirement.
8
Unfortunately, this sys-
tem had one flaw—a flaw
that would later reveal how
people behave when con-
tinually confronted with
problems they cannot, or
will not, correct.1
For all of its high-tech
wizardry, this UST system
suffered from a common
problem—the ubiquitous
leaking sump. After invest-
ing a significant sum for the
"best" money could buy,
the owner was surprised
and frustrated to discover
that these "high-tech"
sumps were not watertight.
The sensors in the sumps
went into alarm shortly
after a heavy rain. Rains
were frequent at that time
of the year, and after several
attempts at removing the
water from the surnps and resetting
the alarm console, the owner and
employees became indifferent to the
sensor alarms. After all, it was only
water.
Trouble
As the early morning rush hour
began to wind down during one
blustery cold winter's day, several
customers entered the store to tell the
cashiers on duty that they were
unable to pump regular gasoline. The
owner looked at his ATG console and
discovered that it had shut down the
product submersible because the
product line failed a gross line test.
He called the contractor. Shortly after
arriving at the site, the technician
tried resetting the ATG system and
rerunning the line test. Again, the
product line immediately failed the
test and the ATG system shutdown
title submersible.
1. The topic of human error is fascinat-
ing. Two of the more useful and interest-
ing works include: The Logic of Failure:
Recognizing and Avoiding Error in Com-
plex Situations by Dietrich Dorner, pub-
lished by Perseus Books, 1996, and;
Inviting Disaster: Lessons From the Edge of
Technology by James R. Chiles, published
by HarperCollins, 2001.
He decided to isolate the product
line and rerun the line test to deter-
mine if the problem was with the
submersible. At this time, the techni-
cian noticed a substantial amount of
water in the submersible sump. The
water covered the piping, entry
boots, and the submersible. The
owner responded that water was
always getting into the sumps, but it
was just water. Unable to reach the
submersible because of the water and
seeing that there was no gasoline on
the surface, the technician isolated
the product line from the dispensers
and ran the line test. The product line
again failed the gross line test.
The leak was probably some-
where in the product line. However,
the absence of gasoline in the sub-
mersible sump was bothersome
because the UST system had double-
walled piping. Had there been a leak
in the primary piping it should have
drained back to the submersible
sump. Fearing the worst, the contrac-
tor contacted the tank- and line-test-
ing company, requesting that a tester
be sent to the facility to determine
whether the secondary piping was
tight, and if the testing came back
negative, to locate the source of the
leak. The testing service supervisor
informed the manager that they
-------
March 2003 • LUSTLine Bulletin 43
•would, be unable to perform the test-
ing or locate any leak that might be
present until the owner removed the
water from the submersible sump.
Pumping the Sump
The technician informed the owner of
the need to pump the water out of the
sump. The owner said that was no
problem; he had a sump pump in the
back office that he used to pump
water out of the car wash during the
last heavy rain. He could use that to
pump the water out of the sump onto
the driveway. The technician politely
explained to the owner that that was
illegal. The government classified
water removed from the sump of an
UST system used for storing gasoline
as contaminated waste. A licensed
waste hauler with the necessary
equipment for pumping water from
the sump and transporting it to a
licensed treatment facility was
required.
The owner reluctantly agreed
and called the waste contractor rec-
ommended by the testing contractor.
Shortly afterward, a large tanker
truck arrived at the facility and began
pumping the water from the sump
into the tanker. Water continued
entering the sump almost as fast as
the tanker could pump it out. A large
volume of water had collected in the
tank pit during the previous months.
After several hours, the tanker truck
was filled to its 6,000-gallon capacity.
Water no longer continued to enter
the submersible sump and the testing
technician was able to begin running
the necessary tests.
The Helium Test
Performing a helium test at this par-
ticular location required injecting
helium into the interstitial space of
the double-walled piping system
through a fitting on the entry boot
and waiting an hour to allow move-
ment of the helium through the back-
fill. Since the helium is lighter than
air, it would immediately rise toward
the surface upon entering the backfill
wherever a hole existed in the sec-
ondary piping, causing the detector
to alarm.
Using the owner's "as built"
drawings as a guide, the technician
chose several locations along the pip-
ing pathway to drill small-diameter
holes through the concrete to get to
tke backfill material. The technician
then placed a sensing device
designed for "sniffing" helium over
the holes, j
The dfetector alarmed at the hole
closest to the dispenser that was near-
est the building but well outside the
tank pit, confirming that there was a
leak in the piping system. Excavation
revealed that water from the sumps
had filled the interstitial space
between the primary and secondary
piping. An exceptionally cold winter
had causejd both the ground and the
water in the interstitial space to freeze.
Since ; freezing water expands,
there was; nowhere for the water to
go; it burst the secondary pipe and
crushed the primary pipe. Several
hundred gallons of gasoline entered
the groufid because the line-leak
detector had been unable to run a test
for more than two hours due to the
constant running of the submersible
pump, which was pumping product
to meet customer demand during
rush hour.
The Cost
The UST owner's significant invest-
ment in technology resulted in an
invoice of thousands of dollars to
locate a leak, and thousands more to
clean up the underground contami-
nation. It .Would be years before he
would recover those costs—all
because he had not taken decisive
action to properly resolve the prob-
lem of the jeaking sumps.
This story, almost apocryphal in
nature, sHows that people play an
important Irole in the management of
UST systems. Human behavior,
where it concerns the management of
technology and equipment to achieve
a desired outcome, requires the use
of processes in order to assure suc-
cess—in this case, the prevention of
gasoline releases to the underground
environment from the UST system.
Those processes include daily visual
inspection^ alarm response and noti-
fication, maintenance and repair, and
periodic testing. In addition, the
management of an UST system
requires tKe investigation of incidents
to identify root causes and develop
solutions to prevent future problems.
Human Error and Processes
When people use processes, there is
always thd potential for making mis-
takes. Regardless of the reasons
behind their actions, the conse-
quences can be severe. The actions of
the UST owner in this story pushed
technology, in this case, fiberglass-
reinforced plastic, beyond its
performance limits. The resultant
equipment failure and malfunction of
the UST system had catastrophic con-
sequences.
Numerous studies have revealed
that the dominant source of mistakes
is human error.2 There are three com-
mon reasons for that error. Mistakes
occur when people do the following:
• fail to perform required actions
(e.g., removal of the water from
the sumps)
• perform prohibited actions (e.g.,
ignoring the submersible-sump
sensor alarms)
• misinterpret information (e.g., per-
formance characteristics of con-
tainment sumps) critical for the
performance of actions3
Reducing and eliminating human
error is an age-old problem.
Human error can occur during
the design and engineering of the
equipment used in UST systems or
during the manufacture of that
equipment. It can occur during the
construction and installation of the
UST equipment or during the opera-
tion and maintenance of that equip-
ment. It can occur because of human
actions during recordkeeping of leak-
detection-testing results, inventory
reconciliation, or corrosion-protec-
tion testing. It can ultimately occur
while responding to emergencies
associated with the UST equipment
or with the remediation and cleanup
of releases from UST systems.
Clearly, human error is something
we cannot afford to ignore, especially
when it can lead to the types of
defects that create catastrophic
results.
Human Error and Complexity
Only in the last 20 years has it
become apparent that there is a single
common underlying factor linking
• continued on page 10
2. C. Martin Hinckley, Make No Mistake!
An Outcome-Based Approach to Mistake-
Proofing, Productivity Press, Portland,
OR, 2001, pp 10-12
3. Ibid., pg. 10.
-------
LUSTLine Bulletin 43 • March 2003
• People and ISSTsjrom page 9
the frequency of human error to
defects (in the case of managing
petroleum UST systems, the defect
would be the release of gasoline to
the environment). This discovery was
the direct result of attempts by both
the Japanese automobile and U.S.
electronic industries to improve man-
ufacturing processes in order to
reduce defects in their products dur-
ing the 1980s and 1990s.4 That single
factor was complexity. According to
the research, three components com-
prise complexity:
• objects (i,e., material, equipment,
hardware, tools)
• information (i,e., data, communi-
cation, training)
• human activity (i.e., the difficulty
and number of steps to perform
those activities)
When we consider the complex-
ity of UST systems in light of the
need to manage human error, it is
important to keep in mind the key
function of those systems—to store
hazardous substances and petroleum
products safely, in a way that will
prevent the release of product to the
environment. Furthermore, if re-
leases do occur, UST systems should
be capable of detecting those releases
rapidly and effectively, thus enhanc-
ing any opportunity to minimize the
volume of released product. Soundly
engineered, constructed, installed,
operated, and maintained UST sys-
tems should continue to perform
these functions effectively through-
out their life cycle.
It should be apparent from my
story and the research findings that
UST systems are not simply com-
prised of equipment and technolo-
gies. Owners and operators—
humans—interact with UST systems
at some or all points in their life
cycles, developing processes and pro-
cedures to manage those systems.
Therefore, the core elements of all
UST systems are: technology, equip-
ment, people, and processes.
4. C. Martin Hinddey, Quality By Design -
Eliminating Defects Through the Control of
Variation, Mistakes, and Complexity,
Assured Quality, 50th Annual ASQC
Quality Congress, Chicago, IL, 1996.
10
The System Can't Do It Alone
Given the complexity of UST systems
and the potential for human error, it
stands to reason that we must have
clearly defined, properly engineered,
well-documented, consistently exe-
cuted, and periodically evaluated
processes. If such processes are not
present, we should not expect UST
systems to deliver years of reliable
service, regardless of the capabilities
of the technologies or the quality of
the equipment.
error^ it
' J ' '* •>
-stands to reason that we must have
.,, .#; .-;,-'• r,i=ss ft..^T:i:. .' cti
•• clearly defined, properly
;;" ;,Trn'7',™i;'EEiist:t::»fenif pfessr77?r 1
; ; consistently executed, and
K::^;-:^-''^^^^1^*^:^^**^1'^
. periodically evaluated processes.
" ' ' '
\
Ongoing discoveries of contami-
nation of public drinking water wells
and land surrounding UST facilities
are causing people to question con-
cepts previously considered inviolate
regarding UST technology and
equipment. It is not clear whether the
sources of contamination are from
the upgraded equipment, the result
of releases that occurred before the
UST systems were upgraded, the
mismanagement of UST systems,
other factors such as compatibility, or
a combination of several of these fac-
tors. Unfortunately, it may be several
years before we can positively iden-
tify the actual root causes or sources
of these problems.
Recent court settlements address-
ing the contamination of entire com-
munity drinking water systems with
petroleum-based compounds such as
MTBE and other chemicals from UST
systems illustrate that these problems
are immediate. The potential eco-
nomic and health consequences of
these releases to communities can be
significant. There are several hun-
dred thousand UST systems in opera-
tion throughout North America.
Consequently, the immediate need
must be to address how to manage
the operation and maintenance activ-
ities of those UST systems.
How well owners, operators, and
the regulatory community under-
stand the factors affecting human
error will influence how they
approach solving the problems of
managing UST systems. This knowl-
edge can guide owners and operators
in their selection of technology and
equipment for use in new UST sys-
tems. More importantly, it can pro-
vide a framework for developing the
management processes needed to
ensure maximum performance from
existing UST systems, particularly
during their operation and mainte-
nance.
Likewise, this knowledge can
guide the regulatory community in
determining how best to regulate
new and existing UST systems and in
their development of any future leg-
islation to address the operation and
maintenance of UST systems. Ulti-
mately, understanding the factors
affecting human error will determine
whether the decisions made by legis-
lators and regulators contribute to
the prevention of, or are among the
root causes of, the contamination of
our water resources.
In future issues of LUSTLine, I'll
take a closer look at this story and
other stories like this and examine
what went wrong, what should have
been done to prevent the problem,
and what could be done differently in
the future. We will explore such ques-
tions as: What happens to people
when confronted with something
they do not understand? What role
does training have with the manage-
ment of UST systems? How does
boredom or inattention affect a per-
son's ability to effectively manage
UST systems? Finally, we will review
the tools available to UST owners and
operators to help them select UST sys-
tem equipment and design features
that will deliver the necessaiy perfor-
mance to prevent releases of product
at competitive costs throughout the
total life cycle of the system. H
Richard S. Bradley, Jr., LTSA,
AS ACT, is an Environmental Coordi-
nator and Advanced Safety Auditing
(ASA) Trainer for BP Products North
America Inc., Marketing Environmen-
tal Management Systems, East Coast
Retail Business Unit. Richard can be
reached at bradlers@bp.com,
or at (770) 576-3080.
-------
March 2003 • LUSTLine Bulletin 43
from Robert N. Renkes, Executive Vice President, Petroleum Equipment Institute
Overpressurization Problem Baffles Tank Owner
What Do You Think?
A gasoline distributor in the
Northeast installed under-
ground petroleum storage
systems at one of his stations in Sep-
tember of 1992. The 10,000-gallon-
tanks are double-walled fiberglass
with a brine solution in the intersti-
tial space that is continuously moni-
tored by an automatic tank gauge
(ATG). The station has a two-point
Stage I vapor-recovery system and a
balance Stage II system. The prod-
uct piping is a double-walled flexi-
ble system.
The unleaded (UL) tank was
filled on January 10, 2003, without
apparent incident. On January 12, a
driver had difficulty dropping his
entire load, although the ATG indi-
cated that there was plenty of room
in the tank. The driver later
remarked that he heard a high-
pitched air noise coming from the
18-inch diameter manhole cover
over the ATG probe for the tank.
During this delivery, a customer
also complained that vapors were
coming from the nozzle while he
was fueling his car. It was decided to
carefully observe the next delivery.
Delivery Difficulties
On January 15, a driver arrived at
the station at 1:00 a.m. to make a
scheduled gasoline delivery. Before
dropping his load, he took an ATG
reading for the tanks, confirming
the load would fit. He dropped
2,800 gallons of UL into the 10,000-
gallon tank. He also noticed that
vapors were emanating from the
same 18-inch manhole cover that
had made noises on January 12.
Then approximately 40 gallons of
gasoline came out from beneath the
18-inch manhole cover. An investi-
gation revealed that the ATG riser
cap, which was located under the
18-inch manhole cover, had blown
off. The cap was very corroded, and
it appeared that there had been suf-
ficient pressure in
the tank;to force
the cap open.
Also,; the
ATG probe itself
was blown up out of
the tank. It was being
held up;out of the tank 26
inches higher than normal by the
plastic alignment tabs on the probe
that were resting on top of the 4-
inch riser pipe. Because the probe
was now ^positioned more than two
feet higher than its normal position,
it was dramatically underestimating
the volume of gasoline in the tank.
Because the driver relied upon the
ATG reading and did not manually
stick the tank before the delivery,
the tank had been filled beyond its
capacity. I The next day the ATG
riser caps were replaced on all the
tanks and the tank owner thought
the problejrn was solved.
i
Two Days Later
Another driver, using a different
tank trailer, arrived two days later
at 1:00 a.m. in 4° weather to drop
another load into the UL tank. The
driver stutk the tank before making
the delivery and confirmed that
there was plenty of room in the
tank. Halfway through dropping
the 2,800-gallon compartment of his
tank trailejr, the flow seemed to stop,
and a small amount of gasoline
leaked frpm the delivery elbow
attached tp the tank fill pipe.
The driver shut the valve on the
trailer and disconnected the deliv-
ery elbow. When he did, gasoline
shot out from the fill pipe approxi-
mately 7 feet in the air, completely
covering the driver. He tried unsuc-
cessfully jto reattach the delivery
elbow to the fill pipe. By the time
the flow from the fill pipe stopped,
500 gallons had spilled some 300
feet down the street and into two
telephone-fcompany manholes. The
fire department responded by
spraying the spill with foam and
water, which immediately froze as
it hit the pavement. The local Fire
Marshal closed the station as well as
the main road for 12 hours until the
spill was cleaned up.
The System Test
The Fire Marshal insisted that a
tank-testing company check the
tank system before putting it back
into service. The initial theory was
that the ball of the float-vent valve
must have stuck with the ball in the
"up," or closed position. To inspect
the float vent valve, the testing com-
pany had to remove the Stage I
vapor fitting.
Unfortunately, they had to beat
on the vapor fitting in order to get it
off. When they looked down the
riser pipe, they could see that the
ball of the float-vent valve was in
the bottom of its wire basket. There
was some suspicion, however, that
the blows to the vapor adapter
might have dislodged the ball.
When removed for examination, the
ball of the float-vent valve showed
some rust staining, apparently from
the steel pipe into which the ball
seated when the tank overfill level
was reached. The testing company
noted that the wire basket for the
float-vent valve was not damaged
and reinstalled the assembly.
The testing company next
checked the Stage I vapor adapter
and determined that it was working
perfectly. The company did a pres-
sure-decay test on the tank ullage
and found a couple of drain valves
I continued on page 12
11
-------
LUSTI.«ieBH/teKn43 • March 2003
m Overpressurization Problem Baffles Tank Owner from page 11
leaking on the spill containment fill
boxes and one leaking dispenser
hose breakaway. Eventually, they
were able to get a passing pressure-
decay test on the storage system
vapor space as well as a liquid-
blockage test on the vapor return
lines from the dispensers to the
tanks.
The testing company assumed
that a stuck ball float caused the
Overpressurization of the tank. The
next 7,000-gallon delivery was
made with no trouble at all.
What Do You Think?
The owner's delivery policy is to fill
his USTs as close to 90 percent of
capacity as possible. The ball-float
assembly was measured and found
to be the correct length. The tank
owner is concerned because if the
ball seats at 90 percent and doesn't
release before the next fill, the same
problem could occur again. The
owner wrote me at the Petroleum
Equipment Institute and asked if
he's the only one to have such an
experience and whether there could
be other things besides a stuck ball
float that could cause the Overpres-
surization. He also wondered if
there aren't any other theories or
reasons, then what caused the float
ball to hang up? A week of really
cold weather? Corrosion? Or both?
LUSTLine readers can compare
their answers to those of the PEI
members who responded to the
question by going to the "mystery
tank" link at www.pei.org/frd.
Your comments can be sent to
rrenkes@pei.org. •
An Overview of MTBE and Other
Oxygenates in Fuel
by Patricia Ellis
Conventional wisdom holds that
methyl tertiary-lwfyZ ether
(MTBE) first came into wide-
spread use in the United States in 1979;
first as a replacement for alkyl lead com-
pounds, and later (in the 1990s) as a
"required" ingredient in oxygenated
fuels (oxyfuel) and reformulated gasoline
(RFC). But MTBE has a much longer
association loith automotive fuels, as do
many of the other oxygenate fuel addi-
tives, whether it's because they were
used to increase the oxygen content of
the fuel, to enhance the octane content,
or because they were part of a propri-
etary package of additives. Although
there are still significant gaps in our
knowledge base, let's begin our overview
of these oxygenates by piecing together
some of the highlights in the history of
oxygenates.1 We'll start with the ethers,
with special emphasis on MTBE.
The Ethers
An English chemist first synthesized
MTBE in 1842 (Faulk et al., 2000);
however, a U.S. patent wasn't
granted until a century later. Litera-
1. Much of the information in the history
discussion is adapted from. Morrison and
Associates unless otherwise indicated.
(http://www.rmomson.com/mtbe.htai)
12
ture published by the American
Petroleum Institute (API) in the 1950s
discusses the applicability of using
MTBE in gasoline (Drogos, 2000). In
the late 1960s, Chevron field-tested
the use of MTBE and tertiary-amyl
methyl ether (TAME) in taxicab
fleets. MTBE was used commercially
in Italian gasoline blends in the early
1970s (Hart/IPJ Fuels, 2000).
Then in the late 1970s, U.S. EPA
issued a waiver under the "substan-
tially similar" rule (see sidebar on
page 15) for 7 percent MTBE by vol-
ume in motor fuel. MTBE was soon
blended into East Coast gasoline for
octane enhancement, corresponding
with the phaseout of alkyl-lead addi-
tives (McKinnon et al., 1984; Garrett
et al., 1986; Davidson, 2000).
To achieve the same octane levels
as leaded gasoline, MTBE was used
at concentrations of from 4 percent to
7 percent by volume. As tetraethyl
lead in gasoline was phased out in
the early 1980s, the use of MTBE
rapidly increased. From 1980 to 1986,
the commercial production of MTBE
increased at a rate of about 40 percent
per year (Suflita and Mormile, 1993).
Through the late 1980s, EPA granted
waivers for several oxygenated-gaso-
line blends, some of which used
MTBE at increasingly higher levels.
In 1987, the "wintertime oxyfuel
program" was first implemented in
Denver, Colorado, initially with
MTBE, then later with ethanol, due to
consumer complaints about MTBE
(Harvey, 1998; Drogos, 2000). Report-
-------
March 2003 • LUSTLine Bulletin 43
edly in tKe same year, U.S. EPA
issued a waiver for a maximum of 15
percent MTBE by volume in gasoline.
At about this time, it is also reported
that MTBE was among the top 50
chemicals manufactured in the U.S.
(Uhler et al., 2000). By 1989, winter-
time-oxyfuel programs were imple-
mented in Phoenix, Las Vegas, Reno,
and Albuquerque, initially with
MTBE; they later switched to ethanol.
In 1990, the Clean Air Act
Amendments (CAAA) were enacted,
and the maximum oxygen content of
gasoline was raised to 2.7 percent by
weight (15 percent by volume for
MTBE). Production of MTBE contin-
ued to increase, and since 1993, it has
become the second most produced
organic chemical in the United States
(USEPA, 1998)~-350 billion liters (92
billion gallons) of MTBE were pro-
duced in 1997 (Zogorski et al., 1998).
By 1998, more than 10.5 million gal-
lons per day of MTBE were used in
the U.S.; 4.2 million gallons per day
were being used in California alone
(Johnson, et al., 2000). U.S. produc-
tion of MTBE can meet about two
thirds of the country's demand for
oxygenates. The remaining demand
is met primarily with MTBE from the
Middle East and Canada (California,
CARB, 1998).
TAME was first produced in
1907, but its first use as an automo-
tive fuel additive wasn't documented
until the late 1960s, and then appar-
ently only on a somewhat limited
basis. It wasn't until after the Clean
Air Act Amendments (CAAA) of
1990 that TAME and ethyl tertiary-
butyl ether (ETBE) came into general
use as oxygenates (Peterson, 2000).
Alcohols
Alcohols have been manufactured for
many centuries, and ethanol has been
used as an automotive fuel since the
1930s. In 1975, Nebraska began its
gasohol (ethanol blend) program.
U.S. EPA issued a waiver for 10 per-
cent (by volume) ethanol blends in
1976 (Gibbs, 1998), and in 1978 gaso-
hol (10 percent ethanol by volume to
90 percent gasoline) blends were first
used in Nebraska.
ARCO reportedly blended ter-
tiary-butyl alcohol (TEA) into gaso-
line for the first time in 1969 (Harvey,
1998; Peterson, 2000; Drogos, 2000).
In 1979 U.S. EPA issued a waiver for
methanol and TBA at 2.5 percent by
volume, i From 1979
through 1988, U.S. EPA
granted | waivers for
methanol/TBA blends at
substantially higher lev-
els. !
In th6 1980s, M85
(methanol iat 85 percent to
15 percent gasoline by
volume) vfas introduced
arid ethanpl was tested as
an octane! booster (Har-
vey, 1998). Refiners in
California began to pro-
duce ethariol blends in
response to widespread
MTBE contamination of groundwater
resources because they are consid-
ered alternative fuels.
I
Industry and the CAAA
The CAAA has two basic require-
ments for automotive fuels, neither of
which requires the use of MTBE or
any other' specific oxygenate. The
first requirement is that RFC contain-
ing a minimum of 2 percent oxygen
by weight be used year-round in
severe or extreme ozone nonattain-
ment areas to help reduce vehicle
emissions. The second requirement is
that gasolines used during the win-
tertime in 'carbon monoxide nonat-
tainment areas contain 2.7 percent
oxygen by weight.
The choice of which oxygenate to
use was, an|d still is, entirely an indus-
try decisioh. Primarily for economic
and logistical reasons, a majority of
the gasolines produced in the U.S.
contain MTBE rather than any other
oxygenate !(e.g., ethanol). The Clean
Air Act legislation of 1990 included a
number o| provisions that would
have led to! the introduction of alter-
native (noripetroleum) fuels, such as
methanol. [However, the petroleum
and oxygenate industries responded
to these provisions by offering the
REG program as a substitute.
In fact, j industry documents con-
firm that itjwas the refining industry
that initiated and promoted the use
of MTBE in reformulated gasoline
(see STPUE? references). Their repre-
sentatives -^vere largely successful in
expanding! the use of MTBE over
ethanol, thereby boosting the prof-
itability of the industry.
The two feedstocks of the pri-
mary process from which MTBE is
manufactured are methanol and
isobutylene. Although the methanol
must be purchased, isobutylene is a
refinery waste product and its dis-
posal is expensive. The introduction
of MTBE into fuel in large quantities
turned a liability into a significant
asset. MTBE, and ethers in general,
have blending properties that are
more favorable than alcohols. Ethers
can be blended with the bulk fuel at
the refinery and then transported
through pipelines. Alcohols, on the
other hand, cannot be blended with
fuel at the refinery and alcohol-
blended fuel cannot be transported
through pipelines. Alcohols must be
blended at the transportation or dis-
tribution terminal.
Geographic Distribution
of MTBE
Compliance with the CAAA in
nonattainment areas requires the use
of oxygenated fuel, either RFC or
oxyfuel. Beginning in 1992, 36 areas
of the country were required to use
oxyfuel in the winter. Other areas of
the country were allowed to volun-
tarily opt-in to the RFG program to
improve air quality. By the 1998-99
season, however, only 17 areas were
still implementing the program, as 19
areas were redesignated as carbon
monoxide attainment. This was due,
in part, to the implementation of an
oxyfuel program along with other
control measures. But this is only a
small part of the story.
There are estimated to be more
than 700,000 regulated gasoline USTs
at approximately 400,000 facilities.
"Major" oil companies own about 20
percent of the nation's approximately
182,000 retail outlets (Blue Ribbon
Panel Report, 1999). In California
alone, more than 10,000 sites cur-
• continued on page 14
13
-------
LUSTLine Bulletin 43 • March 2003
• An Overview of Oxygenates
from page 13
rently regulated as leaking tank sites
are suspected to have released MTBE
to groundwater (Happel, Beckenbach
and Halden, 1998). In addition, there
are approximately 3 million under-
ground fuel storage tanks nationwide
that are exempt from federal regula-
tions (e.g., certain farm and residen-
tial gasoline storage tanks and home
heating oil tanks), although some
states do regulate heating oil tanks.
MTBE and a variety of other
ether and alcohol oxygenates have
been used for many years as octane
enhancers, and therefore, fuel blends
containing these oxygenates could
potentially have been stored in any
UST system anywhere in the U.S. In
fact, due to the fungible nature of the
fuel supply in the U.S., MTBE has
been found in gasoline (and the envi-
ronment!) in every state at one time
or another whether or not an area is
or has been subject to RFC or oxyfuel
requirements (CRS report for Con-
gress, February 25,2000).
Industry documents (see STPUD
references) reveal that industry was
aware in the early 1980s that
"exchange agreements" would result
in MTBE-laden fuel being delivered
to areas where it was not needed.
Further complicating the distribution
picture is that MTBE has been docu-
mented to be present in other fuels
(heating oil, diesel fuel, aviation
gasoline) for which it was never
intended. Again, industry documents
(see STPUD references) reveal that
industry was also aware of this prob-
lem in the early 1980s. More recent
studies (Robbins, 1999; Robbins and
Henebry, 1999; and Robbins and
Zack, 2000) of oxygenates in other
fuels have confirmed that this is a
widespread phenomenon in the New
England states.
Sources of MTBE in the
Environment
Given MTBE's widespread use as a
gasoline additive and the large vol-
umes of gasoline that are stored,
transported, and used in all areas of
the country, releases of MTBE to the
nation's ground and surface waters
occur in a number of ways. But, the
largest source of gasoline releases to
the environment is leaking under-
ground storage tank (LUST) systems.
Gasoline storage and distribution
facilities can release relatively large
volumes of gasoline (hundreds to
thousands of gallons) that can affect
ground and surface water.
MTBE releases can also occur
from automobile-related accidents,
homeowner spills or misuse, marine
engine emissions into lakes and
reservoirs, and atmospheric washout.
Releases from LUST systems con-
tinue to occur, even at upgraded
facilities, and releases of MTBE
vapors, without a release of liquid,
are especially problematic (SCVWD,
2000). Stormwater runoff can also be
a source of MTBE contamination.
Runoff can be contaminated with
MTBE from dissolution of residual
from parking lots (such as at service
stations and retail businesses) and
roadways. MTBE can also enter the
environment through atmospheric
deposition.
Unprecedented History of
Contamination
Perhaps the most remarkable aspect
of the history of MTBE contamination
is the sheer speed at which MTBE
contamination has spread nation-
wide. Documented cases of ground-
water contamination from LUST sites
date back to 1980 on the East Coast
(see STPUD references). MTBE and
other gasoline constituents were
detected in private wells in a subdivi-
sion in the town of Jacksonville,
Maryland, in 1980. Also around this
time, several oil companies had doc-
umented that MTBE could migrate
farther and faster than other fuel con-
stituents.
Furthermore, industry was
aware of the low taste and odor
threshold of MTBE in water and that
carbon adsorption was not effective
for removing MTBE from water. For
example, an MTBE plume in Rock-
away, New Jersey, that contaminated
the municipal well was known in
1983 to extend for 1,500 feet, while
the hydrocarbon portion of the
plume was much shorter. Consumers
served by the well had complained of
taste and odor problems, which were
attributable to MTBE contamination.
Through the 1980s and 1990s the
number of MTBE-contamination
cases rapidly increased—and they
were not limited to RFG or oxyfuel
states. For example, Kansas, which
may be described as the "buckle of
the cornbelt" (where ethanol would
presumably be the oxygenate of
choice) and where the use of neither
RFG nor oxyfuel is or has been
required, reported that nearly 90 per-
cent of its LUST cases had docu-
mented MTBE contamination
(NEIWPCC, 2000).
Today, thousands of private
wells from coast to coast have been
impacted, as well as numerous public
water supply wells. Public wells have
been shut down in many states,
including California, Rhode Island,
Maine, Indiana, and New York.
Recent studies of ambient groundwa-
ter in California indicate that MTBE
is a ubiquitous contaminant (though
generally at low concentrations) in
Table i : i ! ^qlirreiice ('%) of oxygenates in cinvenfion^l Versus 1 1
i RFG and fegflar versus premiujti gasjjljnes m Neiy Hampshire samples,; | j
MTBE
TAME
ETBE
DIPE
TBA
Conventional Gasoline
Regular
100%
100%
45%
10%
0%
Premium
100%
100%
85%
3%
0%
Reformulated Gasoline
Regular
100%-
76%
18%
18%
0%
Premium
100%
100%
52%
0%
14%
Data from Fred McGarry, NH DES.
14
-------
March 2003 • LUSTLine Bulletin 43
Oxygenate
Ethyl tert-Butyl Ethsr (ETBE)
Ethyl tert-Butyl Alcohol (ETBA)
Tert-amyl Methyl Ether (TAME)
Diisopropyl Ether (DIPE)
Tert-Butyl Formate (TBF)
Tert-Butyl Alcohol (TBA)
Tert-amyl Alcohol (TAA)
Ethanol
Percent
Detected
5/1/01
6%
14%
28%
36%
0%
22%
33%
14%
Percent
Detected
5/1/02
5%
27%
35%
38%.
3%
24%
26%
12%
Percent
Detected
11/1/02
4%
31%
38%
43%
7%
28%
47%
11%
Analytical Method 8260B used to sample for oxygenates and ethanol.
Highest
Concentration
(ng/L)
60
7,940
1,700
8,700
20,800
39,400
76,000
9,800,000
virtually all groundwater that is less
than 20 years old (Small, 2002).
The Other Oxygenates
Little data are available on the use of
the other oxygenates in gasoline or
groundwater, but where states are
analyzing for them, they are finding
them. In Delaware, we already have
domestic wells with TAME and TBA
contamination, and ETBE has been
detected in at least one public well. I
offer some data from two states—
New Hampshire and South Car-
olina—where studies have been done
to determine the extent of the use of
other oxygenates and related conta-
mination.
In 2000, New Hampshire con-
ducted a study in which samples of
regular and premium grades of both
RFG and conventional gasoline were
collected and analyzed for five of the
oxygenates. (See Table 1.) MTBE was
found in all samples, while TAME
was found in all premium gasoline
samples, all regular grade conven-
tional samples, and 76 percent of the
regular-grade RFG samples. ETBE
was found in all types of samples—
the percentage of the samples con-
taining ETBE varied by grade and
conventional versus RFG. DIPE was
found to a lesser extent, but was
identified in at least some of the sam-
ples of all types of gasolines except
premium-grade RFG. TBA was
found in only the premium RFG sam-
ples (14 percent of the samples).
Four counties in New Hampshire
are RFG counties and six counties use
conventional gasoline. Keep in mind
that oxygenates can be used to meet
octane requirements, not just oxygen
requirements for gasoline.
Southj Carolina, as part of all
LUST assessment activities, has been
sampling; groundwater for MTBE
since June 1995. MTBE has been
detected at 72 percent of all con-
firmed releases. Since June 2000, the
state has! been split sampling all
active corrective actions where MTBE
is known to be present to see if other
oxygenates were present. Sixty-eight
samples Were collected at 40 different
LUST sites since June 2000. In addi-
tion to the percentage MTBE
detected, Table 2 shows the percent-
ages of other oxygenates that were
found.
One thing to keep in mind in con-
sidering these figures is that South
Carolina is not an RFG state! Yet the
data in Table 2 support the fact that
not only MTBE but other oxygenates
are being found. The figures pro-
vided by South Carolina also show
that these oxygenates aren't just pre-
sent in trace amounts.
Art Shrader, Director of the
Assessment and Corrective Action
Division of the South Carolina UST
program, says that the oxygenates
aren't just an environmental liability,
they are also expensive to remediate.
"The bottom line to MTBE from a
LUST cleanup perspective, is because
of MTBE's higher solubility, its
plumes tend to be 20 percent larger
than petroleum plumes without
MTBE," Shrader says. "Therefore,
we've found that the assessment and
cleanup costs tend to cost 20 percent
more than those of sites without
• continued on page 16
Substantially Similar?
All additives for motor-vehicle gasoline are required to be registered by U.S.
EPA, pursuant to 40 CFR 79 ("Registration of Fuels and Fuel Additives")
before their introduction into commerce. To obtain registration, the additive
manufacturer must: (1) provide a chemical description of the additive, (2)
provide certain information on the combustion and evaporative emissions of
gasoline containing the additive at its maximum treat rate, (3) provide cer-
tain technical and marketing information, and (4) comply with the "substan-
tially similar" interpretive rule for gasoline and gasoline additives (56 FR
5352, February 11,1991) or have received a waiver under section 211 (f)(4)
of the Clean Air Act.
The basic registration regulations have been in effect since 1970, under
authority of the 1967 Air Quality Act/1970 Clean Air Act. They were designed
to provide U.S. EPA with information on the health and welfare effects of
motor-vehicle emissions. In 1977 the Clean Air Act was amended to require
that motor-vehicle fuels and additives be "substantially similar" to the fuels
and additives used by the auto makers in emissions certification. This
requirement was to prevent the use of additives that might damage emission
controls, particularly the catalytic converter. The manufacturer of an additive
that was not "substantially similar" could obtain a waiver if it was demon-
strated that the additive would not cause vehicles to fail emission standards.
What this means is that when a waiver application is made (e.g., for an
ether), U.S. EPA has no choice but to grant the waiver (since, as in our
example, ethers have a long history of usage in gasoline). And, because
these waivers are issued under regulations that govern the air program,
other U.S. EPA programs (e.g., water and USTs) have no legal basis on
which to challenge the waivers. •
15
-------
LUSTLinc Bulletin 43 • March 2003
m An Overview of Oxygenates
from page 15
MTBE."
It seems clear to me that sam-
pling and analysis for all of the fuel
oxygenates needs to become routine,
at least as part of an initial site assess-
ment or as a one-time event for exist-
ing LUST sites. A remediation system
that is designed to address MTBE
may be ineffective for TEA. Remedia-
tion technologies designed for one
specific oxygenate may be inappro-
priate or ineffective for other oxy-
genates. Just as we need to fully
investigate the extent of contamina-
tion at LUST sites, both laterally and
vertically, we need to fully investi-
gate the range of chemicals present,
including the complete suite of oxy-
genates, or we are not being protec-
tive of human health, safety, or the
environment. •
Pat Ellis is a hydrologist with the
Delaware Department of Natural
Resources and Environmental Control,
Tank Management Branch, and served
as a member ofEPA's Blue Ribbon
Panel on MTBE. She is a technical
advisor and regular contributor to
LllSTLine and can be reached at
Patricia.Ellis@state.de.us.
References
Blue Ribbon Panel. 1999. Achieving Clean Air and
Clean Water: The Report of the Blue Ribbon Panel on
Oxygenates in Gasoline. EPA420-R-99-021,
September 15. http-Mttnvw.epa.gov/otaij/consumer/fueIs/
oxypaiicUr99021.pdf.
CSS Report for Congress. MTBE in gasoline: clean air
and drinking water issues. Congressional Research
Service, Library of Congress. Updated May 15,2001.
hllpJIwww.cnie.orglnlclairJl6.html. The January 7,
2003, update of this report is available at
Iillp:/ttvtcw.cnic/nle!cr$reporls/03]anl98290.pdf.
Davidson. J. 2000. Fate, transport and remediation of
MTBE. Presented at Mealey's MTBE Conference,
May 11-12, Marina del Rey, CA.
Drogos, D. 2000. Fate, transport and remediation of
MTBE. Presented at Mealey's MTBE Conference,
May 11-12, Marina del Rey, CA.
Faulk, R., and J. Gray. 2000. An evaluation and update
of the emerging underground storage tank litigation.
Presented at Mealey's MTBE Conference, May 11-12,
Marina del Rey, CA.
Garrctt> P, Moreau, M., and J. Lowry. 1986. MTBE as a
ground water contaminant. In Proceedings of the
Petroleum Hydrocarbons and Organic Chemicals in
Ground Water: Prevention, Detection and Restoration.
National Water Well Association. Dublin, OH,
pp.227-238.
Gibbs, L. 1998. Oxygenate use in gasoline: when, what,
and why. In: Proceedings of the Southwest Focused
Ground Water Conference: Discussing the issue of
MTBE and perchlorate in ground water. Supple-
ment. National Ground Water Association (NGWA).
June 2-3, Anaheim, CA, p.17.
Happel, A., E. Beckenbach, and R. Halden. 1998. An
Evaluation of MTBE Impacts to California Ground-
water Resources. Lawrence Livermore National Lab-
oratory. litlp'Jheww_erd.llnl.gov/mtbe/new_ mtbe.html.
16
RFG vs. Oxyfuel
T") eformulated gasoline (RFG) was initially developed and promoted
JL vby the petroleum industry to forestall the development and use of
alternative (nonpetroleum) fuels. In touting the benefits of RFG, the
industry claimed that it would serve as a replacement for leaded fuels
and help reduce vehicle emissions in severe or extreme ozone nonattain-
ment-areas. The final Clean Air Act Amendments of 1990 signed by
President George H.W. Bush not only set emission performance require-
ments for RFG, but also included a mandate for RFG to contain oxy-
genates (http://www.epa.gov/otaq/rfgorig.htm).
By changing the fuel formulation, RFG reduces the formation of
ozone precursors with its effectiveness measured relative to a 1990 fuel
baseline of NOX toxics and volatile organic compound (VOC) emissions.
RFG also limits the maximum benzene level and requires a minimum 2
percent oxygen by weight (11 percent by volume MTBE, or 5.7 percent
by volume ethanol).
In addition to the RFG program, the CAAA of 1990 required the
establishment of a wintertime oxyfuel program. Under this program,
oxyfuel containing a minimum of 2.7 percent oxygen by weight must be
used during the winter in areas that are not in attainment for the
National Ambient Air Quality Standards for carbon monoxide The
required oxygen content can be achieved by using approximately 15 per-
cent MTBE by volume, or 7.7 percent ethanol by volume. •
Hart/IRI Fuels Information Services. 2000. MTBE vs.
other oxygenates. IPresented at Mealey's MTBE
Conference, May 11-12, Marina del Rey, CA.
Harvey, E. 1998. Changes in the composition of gaso-
line and blended products. In: Environmental Foren-
sics: Determining Liability through Applied Science.
International Business Communications (IBC). IBC,
Southborough, MA.
Johnson, R., J. Pankow, D. Bender, C. Price, and J.
Zogorski. 2000. MTBE: To what extent will past
releases contaminate community water xupply
wells? Environmental Science and Technology, May.
http://pubs.acs.org/hotartcl/est/2000iresearch/0666_00moy
_pankow.pdf
McKinnon, R., and J. Dyksen. 1984. Removing organics
from groundwater through aeration plus GAC. Jour-
nal American Water Works Association. May, pp 42-47.
NEIWPCC. 2000. A Survey of wtate experiences with
MTBE contamination at LUST sites. http://www.
epa.gov/swerustl/mtbe/mtbestat.htm
Peterson, N. 2000. Interpretative considerations for fin-
gerprinting gasoline. Environmental Litigation:
Advanced Forensics and Legal Strategies. Department of
Engineering Professional Development. University
of Wisconsin at Madison. April 13-14, 2000. San
Francisco, CA.
Robbins, G., et al. 1999. Evidence for MTBE in heating
oil, Ground Water Monitoring and Remediation, Spring
1999,pp.65-69.
Robbins, G.A., arid B.J. Henebry. 1999. Evidence for
contamination of heating oil and diesel fuel with
MTBE. LUSTLine 32,: 22-23.
Robbins, G.A. and P. Zack. 2000. Connecticut Survey
finds MTBE-contaminated Heating oil and diesel
fuel statewide. LUSTLine 35:24-25.
Santa Clara Valley Water District (SCVWD). 2000. An
evaluation of MtBE occurrence at fuel leak sites with
operating gasoline USTs. May. ftp://www.scvwd.
dst.ca.us/users/pdfs/USTMtBEStudyFinal.pdf.
Small, M. 2002. Personal communication.
State of California, California Air Resources Board
(CARB). 1998. Ah overview of the use of oxygenates
inbasoline. September.
South Tahoe Public Utility District (STPUD). 2002. Var-
ious legal documents pertaining to their successful
lawsuit against Arco, et al. http://www.ewg.org/
reports/withknowledge/pdf/:
Plaintiffs exhibit #13; Testimony of William J. Piel
(013_001.pdf)
Plaintiffs exhibit #31; Testimony of Curtis Stanley,
Equilon (
Plaintiffs exhibit #32; Testimony of Curtis Stanley,
Equilon (032_001.pdf)
Deposition exhibit #16; Presentation prepared by
Curtis Stanley, Equilon (032_004.pdf)
Exhibit #2; Letter from T.G.Kirkpatrick, Shell, to
Ms. Carmen Carlson, API, June 10, 1983
(036_002.pdfl
Plaintiff's exhibit #56; Testimony of Mr. Anderson,
Exxon (056_002.pdf)
Exhibit #19; "Underground Leak Study," Exxon,
Nov. 1973 (160_002.pdfl
Exhibit #15; "Arco Chemical Co. Testimony to the
Colorado Air Quality Control Commission on
Proposed Regulation No. 13 (Oxygenate Man-
date Program)," 1987 (Oll_002.pdf)
Suflita, J.M. and M.R. Mormile. 1993. Anaerobic
biodegradation of known and potential gasoline
oxygenates in the terrestrial subsurface. Environmen-
tal Science and Technology, vol. 27, no. 5,976-978.
Uhler, A., S. Stout, R. Uhler, and K. McCarthy. 2000.
Considerations for the accurate chemical analysis of
MTBE and other gasoline oxygenates. Soil Sediments
& Groundwater. MTBE Special Issue: 70-72.
U.S. Environmental Protection Agency (USEPA). 1998.
MTBE Fact Sheet #3: Use and Distribution of MTBE
and Ethanol, USEPA Office of Underground Storage
Tanks. Washington, D.C. EPA/625/K-98/001. Sep-
tember. http://wzovj.epa.gov/swerustl/mtbe/mtbefs3.pdf.
Zogorski, J.S., G.C. Delzer, D.A. Bender, PJ. Squillace,
T.J. Lopes, A.L. Baehr, P.E. Stackelberg, J.E. Land-
meyer, C.J. Boughton, M.S. Lico, J.F. Pankow, R.L.
Johnson, and N.R. Thomson. 1998. MTBE: Summary
of Findings and research by the U.S. Geological Sur-
vey, In Proceedings of 1998 Annual Conference of
Water Quality, Atlanta, Ga. American Water Works
Association.
-------
March 2003 • LUSTLine Bulletin 43
Oxygenate Measurements in Groundwater:
Do We Need to Worry About MTBE
Acid Hydrolysis?
by Bruce Bauman
Is there one clearly superior proto-
col for analyzing oxygenates in
water samples? Of course not.
Like many aspects of UST release
prevention, detection, or corrective
action, there is no single "best" way
to get the analytical job done. Rather,
decision makers must have a good
understanding of the general princi-
ples and site-specific variables that
can influence the performance of the
various approaches commonly used
to generate data characterizing
groundwater quality.
In this article, I'd like to throw in
my two cents on the subject of oxy-
genate analysis—especially with
regard to questions surrounding
methyl tertiary-butyl ether (MTBE)
transformations to tertiary-butyl alco-
hol (TEA) via acid hydrolysis. Several
articles and conference presentations
in the past year, including one in
LUSTLine #42 (LL#42), "Analytical
Methods for Fuel Oxygenates," dis-
cuss approaches to sample handling
and analysis and how such ap-
proaches may influence the accuracy
of results for MTBE, TBA (for more
about TBA, see LUSTLine #36), and
other oxygenates in groundwater.
You also may have seen a draft copy
of the U.S. EPA Office of Under-
ground Storage Tank's widely dis-
tributed "Environmental Fact Sheet:
Analytical Methods for Fuel Oxy-
genates."
These discussions have focused
on the most popular protocols for
oxygenate analysis, EPA SW-846
methods 5030B (purge and trap) and
8260B (gas chromatograph/mass
spectrometer [GC-MS]). To some-
what oversimplify those discussions,
questions have been raised regarding
the ability of these protocols to pro-
vide accurate measurements of
MTBE, TBA, and other oxygenates in
groundwater.
In particular, critics of the proto-
cols question the following: (a) the
effectiveness of purge and trap (P&T)
techniques for extracting
all oxygenates (especially
alcohols jsuch as TBA)
from water samples, and
(b) the I potential for
transforming MTBE to
TBA by acid hydrolysis.
As these issues can have profound
implications on data quality and
interpretations for corrective action
decisions at any site, they deserve the
considerable discussion surrounding
them. !
"A tremendous amount of
oxygenate data from leaking UST
sites have been generated over
the past several years, yet there
is understandable concern as to
whether these data are valid. In
general, these concerns are
related to two issues: analytical
obstacles, and ether hydrolysis
(particularly of MTBE to TBA)."
LUSTLine #42
"Analytical Methods for Fuel Oxygenates"
Points of Debate
This discussion will focus on the
LL#42 article, as many LUSTLine
readers are familiar with it. That arti-
cle did a vfery good job of addressing
a broad variety of complex issues
associated, with chemical analysis of
oxygenates, not just purging effi-
ciency and MTBE hydrolysis. The
authors provided a good deal of key
information about sampling and ana-
lyzing gasoline oxygenates in
groundwater. They highlighted some
obvious problems with certain ana-
lytical methods and addressed the
important Steps associated with sam-
ple preservation before analysis.
The authors also provided a "rec-
ommended protocol" that "will
greatly improve the quality of the
data" by avoiding the some of the
perceived'problems with existing
methods. VJ/hile the overall guidance
provided in the LL#42 protocol is
excellent, I would like to address two
of the recommendations that deserve
a little more discussion, and, per-
haps, debate:
• To ensure good-quality TBA data,
P&T extraction procedures used
for oxygenates should preferably
be heated to 80°C
• To prevent hydrolysis of oxy-
genates, field preservation of
groundwater samples should use
a base (trisodium phosphate
dodecahydrate [TSP]) instead of
hydrochloric acid.
Let me begin by summarizing a
few key "take-home messages" that
readers would have extracted from
the LL#42 article and that will serve
as the basis for my discussion.
• It is important to analyze for all
oxygenates known to be present in
gasoline using methods that pro-
vide accurate results.
• Standard P&T at ambient temper-
ature does not adequately extract
TBA from water samples, so sam-
ples should be heated to 80°C.
• Existing conventional approaches
to groundwater sample preserva-
tion, extraction, and analysis may
be prone to the acid hydrolysis of
MTBE, which could lead to under-
estimates of concentrations of
MTBE and overestimates of the
TBA concentrations.
Do You Need to Analyze for
All Oxygenates?
Yes, indeed. Only a few states cur-
rently include TBA as a routine target
• continued on page 18
17
-------
LUSTti«cB«/W/K43 • March 2003
m MTBE Acid Hydrolysis
from pnge 17
analyte for UST sites—this despite
numerous recommendations by a
wide variety of authorities over the
last five years. (See page 21 for a list
of reviews and guidance.) Analyzing
for all alkyl ethers (e.g., MTBE,
TAME, DEPE, ETBE) and alcohols has
been widely recommended ever
since the first comprehensive evalua-
tion of oxygenate impacts to water
quality was published in the National
Science and Technology Council's
1997 report Interagency Assessment of
Oxygenated Fuels. On an annual basis
this recommendation has been reiter-
ated by a series of other EPA reports
or pronouncements.
So, obviously, every site should
be assessed for the potential presence
of all oxygenates at some point,
preferably early on, as LL#42 sug-
gests. Does this mean you have to
analyze every site, for every oxy-
genate, every time using GC-MS? Of
course not, and LL#42 makes that
same point, indicating that there will
be situations where simple
approaches like Method 8015 will
work just fine.
Certainly it makes sense to evalu-
ate every site for all oxygenates at
some time to make sure you know
whaf s there. Your data needs may be
different for any given sampling
event during title Hfe cycle of any site,
and a careful decision on the appro-
priate analytical method to use
should made for each round of data
collection.
Do You Need to Heat
Samples to 80°C to
Effectively Extract TBA and
Ethers from Water Samples
During P&T?
LL#42 states that "if purge and trap is
used...it must be modified to
increase method sensitivity. One
straightforward approach is to heat
the sample to 80°C." While it is true
that P&T techniques are most effec-
tive for nonpolar chemicals like
BTEX, there is also good documenta-
tion that more polar compounds,
such as MTBE and other ethers, can
be purged successfully at either
ambient room temperature (~25°C)
or slightly heated (e.g., <45°C).
EPA's oxygenates method evalu-
-_
ation study (U.S. EPA, 2002) showed
good recoveries (>85%) of MTBE and
other ethers at ambient temperatures.
That same study did show poorer
TBA recoveries at ambient tempera-
ture and also showed slightly better
recoveries of ethers using the 80°C
heated purge. It did not evaluate
purge recoveries at temperatures
between ambient and 80°C.
However, P&T can be an effec-
tive TBA extraction method even at
ambient temperatures as shown by a
Lawrence Livermore National Labo-
ratory study (Halden et al., 2001).
These same authors provided similar
information several years earlier in a
report for California EPA (Happel et
al., 1998). Their results showed that
gasoline ether oxygenates (i.e.,
MTBE, TAME, ETBE, and DIPE) and
TBA could be effectively measured
using P&T/GC-MS at ambient tem-
peratures. They observed that a
method detection limit of 35 jWg/L
could be obtained for TBA but noted
that at 40°C, and using a larger sam-
ple volume (10 ml versus 5 ml), they
were able to reach a detection limit of
4.6 ^fg/L. So if a heated purge is
desirable, a temperature of 45°C
should probably be high enough. It
should provide acceptable recoveries,
and minimize some of the potential
problems noted later in this article.
In a very recent paper discussed
in more detail later in this article,
Evans and Colsman (2003) demon-
strate another example of effective
MTBE and TBA analysis using ambi-
ent temperature P&T. In that study,
three California labs using Methods
5030/8260 were able to get good
MTBE and TBA recoveries, as well as
accurate results at low detection lev-
els (-12 pg/L). While their results
don't mean that every lab performs
as well, they do document that those
methods are capable of delivering
satisfactory performance.
Purging effectiveness is clearly
more important for TBA than for the
ethers, but keep in mind that if
proper calibration procedures are
performed, good results can be
obtained, even with lower recoveries.
The relative recovery of TBA from a
lab's calibration sample should be
similar to that from field samples,
and so a correction factor can be
applied to determine the "correct"
concentration. Higher recoveries (i.e.,
75 percent and more) are certainly
What's Purge
and Trap?
Think of the purge and trap tech-
nique as a very small-scale air-
sparging application in a closed
system in which you are captur-
ing the volatilized gases in a car-
bon canister. In practice, a small
volume of water (usually 5 ml,
but sometimes larger volumes
are used to improve method per-
formance) is purged with a gas
(usually helium) to remove the
soluble VOCs from that sample.
The sample may be heated (e.g.,
in the 20° to 45°C range) to
enhance the removal process,
and the purging time may vary,
although 11 minutes is very
common. The volatilized chemi-
cals are "trapped" on a sorbent
column that is then heated
slowly to facilitate the sequential
separation of chemicals with
varying boiling points in the gas
chromatograph (GC). The con-
stituents are then identified in
the mass spectrometer (MS).
desirable, but may not always be nec-
essary to document good method
performance if recoveries are consis-
tent among all samples.
Finally, there is some anecdotal
evidence from lab technicians that
when using an 80°C P&T extraction,
complicating issues with water man-
agement may result in worse recov-
ery of TBA than would be the case at
45°C or lower. The potential problem
here is that at this high temperature,
some water may also be evaporated,
ending up in the instrument's water
trap. Some TBA may be in that water
and not find its way into the GC,
leading to poor recovery. So perhaps
45°C is a good compromise tempera-
ture when TBA is a target analyte.
Is Hydrolysis of MTBE to TBA
a Concern under Normal
Conditions?
The rate of hydrolysis of MTBE to
TBA is affected primarily by pH,
temperature, concentration, and
time. The impact of pH and tempera-
ture on the rate of MTBE hydrolysis
-------
March 2003 • LUSTLine Bulletin 43
"Under normal environmental
conditions ethers do not undergo
hydrolysis at significant rates
without enzyme catalysis; even in
acidified (pH < 2) groundwater
samples, ethers are generally
stable (Church, etal., 1999)."
"Therefore, if water samples are
preserved with acid, there is an
understandable concern as to
whether or not any of these data
are valid. ... If [acidified]
groundwater samples are
refrigerated before analysis and
all the sample preparative
methods are carried out at
ambient temperature (as
opposed to an elevated
temperature of80°C), there is
minimal opportunity for hydrolysis
of the ether oxygenates."
LUSTLine #42
"Analytical Metods for Fuel Oxygenates "
has been well documented (O'Reilly
et al., 2001). At very low pH (e.g.,
below 1.5) and at high temperatures
(e.g., 80°C) hydrolysis rates will be
relatively fast, but those conditions
are unusual except perhaps during
the heated P&T or heated headspace
extractions.
As shown in LL#42, at 4°C,
MTBE hydrolysis is negligible under
almost any condition. Even at ambi-
ent temperature there is insignificant
hydrolysis over the standard holding
times of VOC samples (14 days) if the
sample pH is 2. (Of course, samples
should never be stored this way!)
For example, in a recent paper by
Douthit et al. (2002), the calculated
pseudo-first-order-rate constant for
MTBE hydrolysis for a solution con-
taining 900,000 ppb of MTBE at 26°C
and pH 2 is 0.0022 per day—that is,
0.2 percent of the MTBE will be trans-
formed into TEA each day. At
100,000 jWg/L MTBE and the same
temperature and pH, the calculated
rate constant is 50 percent slower
(0.0011/day). That same study
showed no measurable hydrolysis of
MTBE to TEA above detection limits
as low as 5 ^fg/L in samples stored at
4°C for 7 to 31 days and analyzed by
P&T/GC-MS. Unpreserved control
samples were also analyzed. No TEA
was detected at 5 to 20 ppb.
The | authors state that their
results show that "acid hydrolysis of
MTBE, in properly handled ground-
water samples, does not compromise
the integrity of dissolved MTBE and
TEA analyses." This is very good evi-
dence that hydrolysis should not be a
MTBE at pH 1.5 at temperatures
ranging from 20 to 80°C. Inspecting
Figure 1 at the typical purge time of
11 minutes, reportable TEA (> 5
jHg/L) can occur at temperatures
above 65°C. In practice, samples with
concentrations above 500 to 1,000
Figure 1
30
theoretical TBA Production from a 500 ugjl
i ' SolutibnofMTBEatpH1.5
Heating Time (imin)
12
»S 10
CO
to 8
CJ
CD
•§—»
CO
w .
•i 4-
Figure 2
problem' when
samples are han-
dled and ana-
lyzed according
to commonly
used standard
protocolsj
Usin^; the
method outlined
in O'Reilly et al.,
it is useful to cal-
culate some pos-
sible effects of
hydrolysis under
a variety of typi-
cal analytical
conditions.
Using anj upper
range of j heated purge conditions
(45°C), a sample containing 1,000
jWg/L of | MTBE theoretically will
yield 0.7,4g/L of TBA at pH 1. Under
the same conditions, a sample con-
taining 10,000 pg/L MTBE will yield
7 ^g/L of TEA, and a sample contain-
ing 100 ;*g/L of MTBE will yield 0.1
jig/L of TBA. So even when pH is
extreme, if moderate heating is used,
hydrolysis should be insignificant.
To illustrate this point more rig-
orously, Figure 1 shows the calcu-
lated amount of TBA that can
theoretically result from the hydroly-
sis of a sample containing 500
Changes in Hydrolysis Rate
Constant with Temperature
20 40 60 80
Temperature, degrees C
100
are diluted prior to analysis.
This practice both dilutes the acid,
raising the pH, and reduces the
MTBE concentration to a point where
any resulting TBA concentration is
below reporting limits.
To see that temperature effects
are of greatest concern above 40°C,
look at Figure 2. As temperature
increases, the rate constant for
hydrolysis increases by nearly an
order of magnitude between 45°C
and 80°C. Keeping purge tempera-
tures below this range should greatly
reduce potential rates of hydrolysis.
• continued on page 20
19
-------
LUSTLineBulletin43 • March2003
m MTBE Acid Hydrolysis
from page 19
A Multi-Lab Performance
Study Validation of Methods
5030/8260
As mentioned earlier, the recent arti-
cle by Evans and Colsman (2003) pro-
vides an excellent summary of a
study that addresses many of the
issues raised about the data quality
provided by P&T/GC-MS analysis.
The authors note that the U.S. EPA
Office of Solid Waste is a strong
advocate of performance-based ana-
lytical methodology, which is a
requirement for several RCRA pro-
grams. Briefly defined, "the perfor-
mance-based approach requires that
methods be selected and approved
based upon their ability to meet the
data quality goals of a given project
in the actual matrix to be sampled."
(Readers looking for more informa-
tion this topic are referred to EPA's
site at: http://www.epa.gov/epaoswer/
hazwaste/test/pbms.htm.
The authors compared the per-
formance of three different California
commercial labs for MTBE and TEA
that use routine P&T/GC-MS analy-
sis. They wanted to make sure that
the data they were collecting for their
site (the world-famous Port Huen-
eme) would be adequate to quantify
MTBE at 5 jig/L and TEA at 12 pg/L
(2-propanol and acetone were two
other target analytes).
The authors do not indicate that
they had concerns regarding possible
hydrolysis of MTBE to TEA, but they
do note that some critics question the
purging efficiency of these methods
for nonpolar compounds like MTBE
and especially TBA.
They submitted a variety of Port
Hueneme groundwater samples to
the three labs (concentrations ranged
from 1 ^g/L to about 3,000 pg/L in
the samples submitted to each lab)
and asked them to run their standard
method. Two labs used EPA 5030 /
8260 and one used EPA 5242, a simi-
lar P&T/GC-MS method. All three
labs used a purge cycle of 11 minutes.
Two labs purged at ambient tempera-
ture, and the other used 40°C. A vari-
ety of quality assurance and quality
control procedures were also
employed to confirm method perfor-
mance.
20
To very briefly summarize their
results, using the standards of U.S.
EPA's performance evaluation proto-
col criteria, all three labs performed
within acceptable limits of accuracy
and precision for all four compounds
for the entire range of concentrations
studies (i.e., 1 to 3,000 Kg/L)- MTBE
and TBA recoveries were almost
always within the acceptable range
(i.e., between 75 and 125 percent
recovery). In all cases, the desired
low ppb detection limits were
reached.
The authors conclude that the
results reflect a successful demon-
stration of method applicability
(P&T/GC-MS) for the compounds
studied. This single study of three
labs does not, of course, mean that
each and every one of the almost
1,000 environmental labs nationwide
routinely performs as well. However,
the authors conclude, "The re-
sults... appear to be a successful
demonstration.. .that.. .purge-and-trap
GC-MS is an acceptable technology
for the analysis of MTBE and its oxy-
genated breakdown products."
There is a lot of other good infor-
mation and discussion in that well-
written article, and I strongly
encourage you to take a closer look
for yourself. It provides a good tem-
plate for the kind of information you
should be able to obtain from your
laboratory regarding the quality of its
analytical performance.
With Good Data and
Attention to Protocol...
We know a lot more today than we
did a few years back about the proper
approaches for oxygenate analysis.
We also know a lot about MTBE acid
hydrolysis and its potential effects on
groundwater analysis, but that's not
to say we know it all. I suspect that
this topic will continue to be of high
interest and that we will learn a lot
more in the coming year.
The studies cited in this article
and in LL#42 collectively provide a
lot of interesting information, but
more information will be required to
conclusively document the perfor-
mance of the analytical methods cur-
rently in common use. To be sure,
following the recommended protocol
provided in LL#42 would help you
become more aware of the quality of
the data for the sites you manage,
and any potential limitations to that
data. That information should serve
you well, and your corrective action
decision-making should improve.
However, at most sites you prob-
ably do not need to use a heated P&T
extraction to generate good data for
MTBE and TBA. If heating is shown
to improve method performance,
then temperatures of 40° to 45°C are
likely to be sufficient, and at these
temperatures, MTBE hydrolysis
should not compromise data quality.
Purge-and-trap extraction of water
samples at ambient or slightly ele-
vated temperatures (combined with
good calibration procedures) should
adequately recover any TBA present
in samples.
Furthermore, it is probably
unnecessary in most situations to use
TSP as your field preservative,
instead of acid. While both low acid-
ity (pH) and high temperature are
known to enhance the rates of MTBE
hydrolysis, an evaluation of these fac-
tors and a review of the standard
analytical procedures employed at
many commercial laboratories indi-
cate that it is unlikely that significant
MTBE hydrolysis occurs. For condi-
tions used at most UST release sites—
using standard acid preservation of
samples, followed by P&T/GC-MS
extraction and analysis—hydrolysis
of MTBE to TBA will be very limited
and will not significantly influence
data quality. Hydrolysis of MTBE to
TBA should not be a concern during
storage and analysis of acid-pre-
served samples.
Data users should have a good
understanding of the basic perfor-
mance of the sampling and analytical
procedures they use to generate oxy-
genate data. If they have concerns
about data quality for ongoing evalu-
ations of corrective action at UST
sites where gasoline oxygenates are
present in groundwater, they should
review the protocols in use, the list of
target oxygenate analytes, and their
performance criteria (e.g., recoveries
and detection limits) to determine if
any changes might improve perfor-
mance. I would also recommend that
you talk about these issues with some
of the commercial labs in your area to
determine what they think are the
best approaches to maintaining or
improving data quality. •
-------
March 2003 • LUSTLine Bulletin 43
Bruce Bauman is the Soil and Ground-
water Research Program Coordinator
at the American Petroleum Institute.
He has been involved with UST and
groundwater issues at API since 1985,
and is an occasional LUSTLine
contributor.
[Note from Bruce: Everything I've
needed to know about oxygenate analy-
sis I've learned from: Ileana Rhodes
(Shell Global Solutions); Mike Miller
(ExxonMobil); and Kirk O'Reilly,
Michael Moir, and Al Verstuyft
(ChevronTexaco). Their substantial
contributions to this article are most
gratefully acknowledged. I assume sole
responsibility for any errors and
misstatements.]
References:
Douthit, T. L. et al. 2002. The importance of acid
hydrolysis of MTBE to TBA in properly handled
groundwater samples. API-NGWA Petroleum
Hydrocarbons in Groundwater conference pro-
ceedings.
Evans, J.D. and MR. Colsman. 2003. A GC-MS purge-
and-trap method comparison study for MTBE
analysis in groundwater. LCGC Magazine Online.
Feb 1 2003 http://www.lcgcmag.com/ Icgc/data/article-
standard/lcgc/052003/45064/artide.pdf.
Halden, R. U., A. M. Happel, and S. R. Schoen. 2001.
Evaluation of standard methods for the analysis of
Methyl tert-Butyl Ether and related oxygenates in
gasoline-dontaminated groundwater. Environmen-
tal Science & Technology 35 (7): 1469-1474. (Addi-
tions and Corrections, 35 (7): 1560)
Happel, A.M., E.H. Beckenbach, and R.U. Halden.
1998. Evaluation of EPA and ASTM methods for
analysis of oxygenates in gasoline-contaminated
groundwater. Chapter 2in.An evaluation of MTBE
impacts to California groundwater resources, Lawrence
Livermore National Laboratory UCRL-AR-130897.
June 11,1998.68 p. http://www-erd.Unl.gov/mtbe/new-
mtbe.html.
National Science and Technology Council. 1997. Inter-
agency Assessment of Oxygenated Fuels Online at:
http://wvno.epa.gov/ncea/oxygenates/Ji4lasmnt.htm.
O'Reilly, K.M., et.al. 2001 "Hydrolysis of tert-Butyl
Methyl Ether (MTBE) in Dilute Aqueous Solu-
tions." ES&T 35:3954.
Rhodes, I.A.L. and A.W. Verstuyft 2001 "Selecting
Analytical Methods for the Determination of Oxy-
genates in Environmental Samples and Gasolines"
ET&Avol.lO:24.
U.S. EPA. 2000. Data quality objectives process for
hazardous waste site investigations. EPA/600/
R-00/007. http://www.epa.gov/rlOearth/offices/oea/
epaqag4h.pdf.
U.S. EPA. 2002. Development and evaluation of
methods for the analysis of MTBE. EPA contract
No. 68-WO-0122 WA No. 0-08.
U.S. EPA. SW-846 Method 5030B: Purge-and-Trap for
Aqueous Samples, http://www.epa.gov/epaoswer/
hazwasteltestlpdfsl5030b.pdf.
U.S. EPA. SW-846 Method 8260B: Volatile Organic
Compounds by Gas Chromatography/Mass Spec-
trometry (GC/MS). http://www.epa.goV/epaoswer/h
azwasteltestlpdfil8260b.pdf.
White, H. et al. 2002 "Analytical Methods for Fuel
Oxygenates." LUSTLine Bulletin 42. Posted on the
OUST website at: http://www.epa.gov/oust/mtbe/
LL42Analytical.pdf.
Analyze for All Oxygenates, Got It?
As shown by the following quotes from authoritative oxygenate
reviews and guidance during the last six years, there is no excuse for
not knowing what other oxygenates besides MTBE may be present at
your UST sites.
• 1997 National Science and Technology Council report, Interagency
Assessment of Oxygenated Fuels:
"Add the alkyl ether oxygenates MTBE, ETBE, TAME, and DIPE to
existing VOC analytical schedules and as U.S. EPA routine target ana-
lytes for drinking water, wastewater, surface water, groundwater, and
remediation studies." This report noted that at that time, there were no
validated methods for TBA analysis in groundwater, and so it did not
recommend TBA analysis.
• 1998 EPA Office of Research & Development report Oxygenates in
Water: Critical Information and Research Needs (EPA/600/R-
98/048): Section 5.2 (p. 18):
"Oxygenates should be added to existing VOC analyte schedules and
included as routine target analytes for VOCs in drinking water, waste
water, surface water, groundwater, and remediation sites.... Given the
existence of TBA as a primary oxygenate, as a contaminant of MTBE,
and as a degradation product of MTBE, the inclusion of TBA in ambi-
ent groundwater quality monitoring programs is advisable. It also
would be useful to monitor for TBA at specific sites where MTBE cont-
amination is known or suspected to have occurred."
• August 1999 EPA Blue Ribbon Panel Report Recommendations:
"Establish routine systems to collect and publish, at least annually, all
available monitoring data on: use of MTBE, other ethers, and Ethanol;
levels of MTBE, Ethanol, and petroleum hydrocarbons found in
ground; surface, and drinking water; trends in detections and levels of
MTBE, Ethanol, and petroleum hydrocarbons in ground and drinking
water. Identify and begin to collect additional data necessary to ade-
quately assist the current and potential future state of contamination."
B EPA OUST letter, January 2000, "Monitoring and Reporting of MTBE
and Other Oxygenates at UST Release Sites":
"While MTBE has received most of the publicity recently, it is by no
means the only chemical of concern for which you should be monitor-
ing and reporting. For example, the oxygenate TBA is both a degrada-
tion product of MTBE and a fuel additive in its own right; it is also
potentially more toxic than MTBE. You should also carefully consider
assessing for other oxygenates (that include, but are not limited to,
TAME, DIPE, ETBE, ethanol, and methanol)."
H Halden et al.: 2001, "Evaluation of Standard Methods for the Analy-
sis of Methyl tert-Butyl Ether and Related Oxygenates in Gasoline-
Contaminated Groundwater":
"The results of this study indicate that project managers of LUST sites
and chemists working with gasoline-containing groundwater samples
may turn to EPA Method 8260B and ASTM Method D4815 in their
search for a robust, accurate, precise, and widely applicable monitor-
ing tool for ether oxygenate and TBA analysis."
• NEIWPCC 2001: "A Survey of State Experiences with MTBE Contam-
ination at LUST Sites":
"States were asked if they analyze for any of the following oxygenates:
ethanol, TBA, TAME, ETBE, DIPE, or any others. The overwhelming
majority of states indicated that they never analyze for any of these
substances."
21
-------
Ll/STLi»«>BH//eli«43 • March 2003
The Other Fuel Oxygenates
A Summary of the ASTSWMO Fuel Oxygenate
Symposium
by JeffKtthn and Bob Haslam
T A ~Fmi tjte toPic tffad oxygenates comes up,
\/\ I most people probably think of methyl ter-
F V tiary-fcufyZ ether (MTBE). MTBE has
played so prominently in the news the past few years
that most people probably use "MTBE" and "oxy-
genate" interchangeably. But those of us who are
more involved rvitii fuels, fuel storage, and fuel cont-
amination know that MTBE is but one of several
oxygenates that are commonly used in fuel. Other
oxygenates include tertiary-amyZ methyl ether
(TAME), ethyl tertiary-fcwfyZ ether (ETBE), di-iso-
propyl ether (DIPE), tertiary-amyZ ethyl ether (TAEE),
* *. - . - . . * *»•« *>.-. t i t 1 srn ,
tertiary-&Hfi/Z alcohol (TBA), tertiary-amyZ alcohol (TAA),
ethanol, and methanol.
Despite our best efforts at keeping fuel in fuel storage tanks where it belongs, fuel components
(including MTBE and other oxygenates) frequently escape the confines of tank systems and find their
way into the environment. MTBE has also been found in many types of fuels, such as home heating oil,
in which its presence ivas never intended. Because of the nature of the nation's fuel production and dis-
tribution system, it is likely that other oxygenates are also present in fuels other than gasoline. It is
becoming increasingly evident that MTBE as well as other oxygenates are quite often discovered in fuel
releases, if one dares to look for them; all that's needed is the appropriate analytical method.
Over the years, we've become relatively familiar (and hence comfortable) with the behavior, cleanup
approaclies, and toxic properties of the hydrocarbon constituents in fuel, and we're now getting some-
what familiar with MTBE. But what about the other oxygenates? What do we know about them? Are
these oxygenates going to be as widespread and difficult to address as MTBE?
TJiese questions (and undoubtedly many others) prompted the Association of State and Territorial
Solid Waste Management Officials (ASTSWMO) to hold a one-and-a-half-day symposium on fuel oxy-
genates other than MTBE. The symposium, entitled: "State Symposium on Fuel Oxygenates: Analysis, Assessment, and Remedia-
tion" was held October 22 and 23,2002, in Arlington, Virginia. Representatives from 25 states, two territories, and the District of
Columbia were represented, as well as six U.S. EPA regional offices and U.S. EPA's Office of Underground Storage Tanks (OUST),
Office of Research and Development (ORD), Office of Solid Waste (OSW), and Technology Innovation Office. Also represented were
the U.S. Geological Survey (USGS), the American Petroleum Institute, two consulting firms, and NEIWPCC. This article briefly
summarizes the main issues identified by participants of the symposium.
Symposium Highlights
The symposium offered all partici-
pants the opportunity to provide their
perspectives on the issue of fuel oxy-
genates as contaminants in soil and
groundwater. The first day of the
event featured presentations on a vari-
ety of issues, including physical and
chemical properties of oxygenates and
methods for detecting them in soil and
groundwater, occurrence of oxy-
genates in fuel and at LUST sites, site
characterization, remediation, regula-
tory trends, and industry perspec-
tives. The second day was devoted
largely to breakout sessions for small-
group discussions of state experiences
with oxygenate occurrence, remedia-
tion strategies, the effect of ethanol on
22
hydrocarbon plumes, and the effect of
state and/or federal regulatory levels
for various oxygenates. The outcome
of all of this information exchange is
summarized as follows:
• TBA is an emerging problem and
is anticipated to be a large problem.
Little is known about the health
effects of TBA and whether current
treatment/remediation technologies
can adequately address soil and
groundwater contamination. Also,
common analytical practice is yield-
ing inaccurate data on the occurrence
and magnitude of TBA in environ-
mental samples. This complicates
decision making on treatment and
remediation and can lead to unneces-
sary expenditure of scarce resources.
• States generally feel that there is
a need for U.S. EPA and others,
such as USGS, to be proactive
(rather than reactive) in regard to
the other oxygenates in fuel. States
expressed a desire to be notified of
changes in fuel composition so that
they can adjust their site assessments
to look for any new fuel additives. A
key component of this issue is the
need for a comprehensive, multime-
dia life-cycle analysis of the various
oxygenates being considered as a
replacement for MTBE. States are
concerned about what other com-
pounds may be added to fuel that can
end up in the groundwater and are
fearful of another 60 Minutes-style
debacle, this time regarding (most
immediately) TBA.
-------
March 2003 • LUSTLine Bulletin 43
• There is uncertainty and some
anxiety in the states regarding ana-
lytical methods for fuel oxygenates.
While the recent U.S. EPA study on
analytical methods and sample
preservation answered some ques-
tions regarding these issues (see
"Analytical Methods for Fuel Oxy-
genates," LUSTLine Bulletin #42),
states expressed a strong desire for
U.S. EPA to release formal guidance
on these and other issues.
• A major issue states face with
regard to oxygenates, including
MTBE, is the lack of federal drinking
water standards for these contami-
nants. While some states are testing
for a broad range of oxygenates, oth-
ers are only looking for MTBE. Sev-
eral states that have started to focus
their attention on TEA are concerned
about the frequency of discovery of
the compound in drinking water.
Many are frustrated by the absence of
any federal mandate to consider the
impacts of oxygenates other than
MTBE.
• Recent product analysis studies
performed in nine states show that
the presence of other oxygenates is
widespread—oxygenates were de-
tected in virtually all the gasoline
sampled. These data suggest that if
MTBE is present in gasoline, then
other oxygenates will also be present.
• States need training and guidance
on how to put together planning
documents to generate useable
data for site characterization and
remediation options (e.g., site-spe-
cific sampling and analysis plans
[SAPs]). This need for guidance also
extends to the issue of whether sites
should be resampled for TEA (and in
some states for MTBE where resam-
pling has not yet been initiated).
Some states suggested that OUST
develop guidance with specific crite-
ria for resampling sites contaminated
with MTBE and / or TEA.
• Nationally, there is an increasing
concern over vapor migration and
indoor air issues.
• To date, there have been no
advances in release detection,
specifically in the lowering of the
leak detection threshold. This neces-
sitates greater reliance on contain-
ment and cleanup strategies, which
are much snore expensive than pre-
vention. States suggested that gas sta-
tions could be better designed to pre-
vent releases of oxygenates into the
environment.
j
• States'would like to have U.S. EPA
better coordinate the development
of strategies to more effectively
deal with oxygenate issues as a
whole. This coordination should be
in the fojrm of both intra- and inter-
agency Workgroups that hold regular
meetings, and report findings.
The Three Most Significant
Issues !
I
ASTSWtvjO representatives compiled
and evaluated the issues raised dur-
ing the l_ symposium and selected
three thajt were of such significance
that they were detailed in a letter to
OUST Director Cliff Rothenstein
requesting "further discussion and
support ifrom OUST." The three
issues identified and discussed in the
letter are as follows:
• States ^eed to look for TBA and
the other common fuel oxygenates.
Sampling for TBA and other oxy-
genates should be completed rou-
tinely as: part of an initial scan.
ASTSWMO urged OUST to provide
guidance !on TBA and the other oxy-
genates so that states can have a
better understanding of these com-
pounds (e.g., toxicology, appropriate
remediation alternatives) when they
find them in groundwater and drink-
ing water. Some remediation tech-
nologies that are effective for MTBE
may be inappropriate, ineffective, or
may exacerbate contamination from
some of the other oxygenates.
ASTSJWMO suggested that an
OUST memo issued by Sammy Ng in
2000 (http;://www.epa.gov/swerustl/
mtbe/janlSOO.pdf), which recom-
mended ttiat states test LUST sites for
MTBE ancl the other oxygenates, be
updated to more strongly emphasize
the concern for TBA and the impor-
tance of using appropriate sampling
methodologies during sample preser-
vation and analysis.
• Finalize and release the "Analyti-
cal Methods for Fuel Oxygenates"
fact sheet, prepared by OUST, OSW,
and ORD. iThis is timely and critical
information. In fact, some states are
already using the draft fact sheet
methodologies. This work comple-
ments the efforts of many researchers
and the recommendations of the EPA
Blue Ribbon Panel on MTBE—efforts
that occurred in response to the dis-
covery of a national MTBE problem.
States need this guidance to help ver-
ify that they are using the most
appropriate analytical methodologies
to detect groundwater impacts from
the various oxygenates. As many
states are already finding these com-
pounds in groundwater and drinking
water, the research under way by
OUST will provide a welcome level
of assurance that state analytical
approaches are either valid or in need
of change.
The issue of testing for oxy-
genates other than MTBE was previ-
ously highlighted in technical
presentations made at the last two
National UST/LUST Conferences,
presentations made in ASTSWMO
MTBE and Fuel Oxygenates Work-
group meetings, and discussions
between federal and state managers.
While state concerns about testing for
the list of common oxygenates are
not new, the level of knowledge
regarding these contaminants has
greatly advanced. TBA is now recog-
nized as a significant contaminant of
concern by a number of states. There-
fore, knowledge of analytical meth-
ods that have the potential to provide
a more accurate analysis is more
important than ever.
• Form an interagency work group
consisting of USGS and the U.S.
EPA Offices of Solid Waste and
Emergency Response, Research
and Development, Air and Radia-
tion, and Ground Water and Drink-
ing Water that meets quarterly to
discuss fuel oxygenates and addi-
tives that have the potential to
impact groundwater and drinking
water. This work group would also
address life cycle considerations for
these compounds.
Symposium participants agreed
that a federal interagency workgroup
would be an important component of
future discussions on fuel oxygenates
and additives. Interagency coordina-
tion and communication between
U.S. EPA, USGS, the Departments of
Agriculture and Energy, industry,
and water suppliers was also put
forth in the final recommendations of
the Blue Ribbon Panel on MTBE
D continued on page 24
23
-------
LUSTLine Bulletin 43 • March 2003
m Symposium Summary from page 23
(final recommendations dated July
27,1999).
ASTSWMO envisions that the
focus of this new work group would
be to enhance interagency communi-
cation to prevent future incidents
similar to that caused by the unantici-
pated impacts of the federal Refor-
mulated Gasoline (RFC) program.
With this in mind, the Blue Ribbon
Panel had called for a multimedia life
cycle analysis of any proposed fuel
additives. In light of the anticipated
increase in the use of ethanol nation-
wide, ASTSWMO recommends that
life cycle analyses be completed on
new formulations of gasoline.
Future Outlook
The ASTWMO Fuel Oxygenates
Symposium underscored the contin-
ued high level of state concern over
the long-term implications of moni-
toring for, and detecting, fuel oxy-
genates in groundwater and drinking
water. Oxygenates dearly should not
be ignored; but time and resource
constraints will influence what is
realistically achievable within a rea-
sonable period of time. For example,
the promulgation of federal drinking
water standards is a lengthy process
that could conceivably span more
than a decade. And there are other
factors that may come into play over
which neither U.S. EPA nor state
agencies have any control. In particu-
lar, any of the various bills relating to
fuels and energy that are floating in
Congress could affect federal and
state UST/LUST programs in ways
that we cannot now anticipate.
If you've ever hiked on a talus
slope in the western U.S., you know
the "two steps forward, one step
back" feeling that many of us have
experienced in the battle to keep oxy-
genates out of the environment. The
good news is that we're now above
the timberline—prior to MTBE, we
couldn't see the forest for the trees. •
JeffKuhn is a hydrogeologist and man-
ages the Petroleum Release Section for
the Montana Department of Environ-
mental Quality. He can be reached at
jkuhn@state.mt.us. Jeff chairs the
ASTSWMO MTBE and Fuel Oxy-
genate Workgroup. The workgroup
includes representatives from state
LUST Programs, U.S. EPA, American
Petroleum Institute, U.S. Geological
Survey, Association of California
Water Agencies, Academic Network
for Contaminated Lands Research in
Europe (ANCORE), Montana State
University Center for Biofilm Engi-
neering, and several private consulting
firms. The workgroup is open to all
interested parties.
Bob Haslam is a hydrologist with the
Vermont Department of Environmen-
tal Conservation, Underground
Storage Tank Program.
UST Bill S.195 Set for Action in Senate
Senate Bill S.195, which is nearly identical to last year's S.1850, sponsored
by Senators Chafee, Inhofe, Carper, Warner, and Jeffords, was adopted by
the Environment and Public Works Committee in February and is ready for
action in front of the full chamber. (The full Senate never voted on S.1850.)
Under S.195, states would be required to ensure that UST facilities are
inspected every two years and that they implement a strategy for training UST
operators. It provides states with the authority to prohibit gasoline deliveries
to noncompliant tanks.
The bill also requires that EPA distribute to the states at least 80 percent
of the corrective action funds appropriated each year from the Leaking Under-
ground Storage Tank Trust Fund. Substantial funding is authorized for MTBE
cleanup, enforcement, inspection, and corrective action. While much of the
intent of the bill has the support of state UST regulators, program managers
caution that the added workload associated with certain requirements in the
bill could only be accomplished with additional federal resources. The House is
expected to introduce a similar bill soon. •
Northeast States
Powwow on
Compliance with
Marketers
On January 22, EPA Region 1
and New England Interstate
Water Pollution Control Com-
mission (NEIWPCC) hosted a
meeting with UST program man-
agers from the New England
states and New York and repre-
sentatives from six major petro-
leum marketers in the region.
Discussion focused on the pos-
sibility of building partnerships
to improve compliance with tank
regulations.
Petroleum marketing
trends of greater consolidation,
centralized data collection, use
of outside contracted mainte-
nance services, and electroni-
cally monitored systems have
fostered low significant opera-
tional compliance rates and a
move away from on-site facility
management. UST inspectors
are frustrated by the lack of
required on-site documentation;
disconnects between corporate
headquarters, contract service
personnel, and facility operators;
the lack of on-site system
knowledge; and confusion about
compliance responsibility.
At the meeting, regulators
asked industry representatives
how they could work together to
improve compliance. Regulators
offered to provide inspection
forms to the companies to let
them know what they are look-
ing for in the field, to let corpo-
rate headquarters know about
problems inspectors are finding,
and to work together to find
solutions. The corporate partici-
pants were interested in doing
this and in continuing the dia-
logue. NEIWPCC has mailed out
state inspection checklists and
plans to hold another meeting
this spring. •
24
-------
March 2003 • LUSTLim Bulletin 43
Pay for Performance: California's Cleanup Goal Metric
When RBCA is not HBCA
by Robert S. Cohen, David Charter, George Cook, and Kevin Graves
California's Pay for Performance
(PFP)1 implementation process
yielded a substantial tool for state
trust funds, in general. In California, as
in many states, the regulators have a
strict standard of cleanup for site reme-
diation. This standard is not always
obtainable, yet much effort and money
can be spent in the chase. The U.S. EPA
and the California Underground Storage
Tank Cleanup Fund (The Fund) worked
together to find a way to satisfy the regu-
lators' environmental agenda and simul-
taneously achieve fiscal control. The
solution was found in implementing a
Preliminary Active Remediation Goal
as a means for concluding PFP con-
tracts—a cleanup target goal that is
something short of full closure. The
PARC tool combines elements of RBCA
with PFP—bringing peace and satisfac-
tion to regulators and trust fund admin-
istrators alike.
The Problem
Cleanup authority in California lies
with nine Regional Water Quality
Control Boards and 20 local agencies
that are under contract to the state.
Each regional board determines the
cleanup standards required for clo-
sure of UST sites in its jurisdiction,
based on the present and future bene-
ficial use of groundwater in the area.
Standards vary from board to board
and tend to be very protective of
groundwater quality.
Meanwhile, back in Sacramento,
The Fund pays the bills for petroleum
cleanups—up to $1.5 million per inci-
dent. The Fund pays reimbursement
claims, some of which are preap-
proved. It is well financed with an
annual income of over $150 million
and has paid over one $1 billion in
1. Pay for Performance is a contractual
mechanism by which the cleanup con-
sultant is paid upon achieving agreed-
upon environmental milestones. The
cleanups are typically faster and cheaper
than the ordinary time and materials
approach. More information is available
at the EPA Web site: http://www.epa.
gov/swerustl/ pfp/index.htm.
claims. Nevertheless, the lack of
progress pn many sites is straining its
resources.
Remedial projects tend to drag
on for various reasons, with many
cleanups jtaking in excess of 10 years
to complete. Some regional boards
have stringent cleanup standards
that are not practically obtainable at
some sites. Nevertheless, many
cleanups; continue until The Fund
limit is approached, even at sites that
do not warrant a high level of effort.
Without firm and reachable tar-
gets, consultants cannot be held
responsible for their engineering
efforts, and many remediation sys-
tems are riot optimized. In a nutshell,
the regional boards are reluctant to
reduce standards, The Fund is not
agreeable to paying for endless
cleanups,; the consultants are not
accountable, and the owner has a
contaminated site!
An associated problem is that
property owners and their consultants
often don't want to begin remediation
on a site where there are not clearly
defined and achievable goals. As a
result, the! investigation and monitor-
(PARG), a remediation goal that the
consultant, owner, and regulator
agree is obtainable by active remedia-
tion. It is not necessarily the regula-
tory cleanup goal, or MCL, though it
could be. (See Table 1.) The PARC is
simply the concluding goal of the
PFP agreement. Upon reaching and
maintaining the PARC, the PFP
agreement is concluded and one of
three events may occur:
• The site is issued closure (no fur-
ther action)
• Natural attenuation monitoring
begins
• The site continues with active
remediation, and a new PARC is
negotiated ,
PFP is impossible without realis-
tic goals. With a defined and obtain-
able interim goal, a remediation
system can be efficiently designed
and operated. Without the PARC,
consultants design systems for unob-
tainable goals (i.e., systems that are
doomed to fail). Systems that are
designed without a defined goal such
as a PARC operate with the tenacity
| I ill EXAMPLES ;OFMCLS VERSUS PARGS 1 I
Constituent
Benzene j
MTBE
Cleanup goal
1.0 ppb
13.0ppb
PARG
100 ppb
200 ppb
ing phases5 are drawn out at hundreds
of sites. Because they have little confi-
dence that they can achieve the regu-
latory goals, owners and consultants
adopt the delaying tactic of proposing
additional monitoring wells, long-
term monitoring, endless pilot tests,
and feasibility studies. In looking for a
means of introducing a PFP approach
to California's cleanup conundrum,
our challenge was to find a solution
that would implement site cleanup
without the dawdle factor.
!
The Solution
The solution was found in the Pre-
liminary Active Remediation Goal
of the EverReady Bunny, yet the sites
never dose. For the less-than-scrupu-
lous consultant, this creates a long-
term annuity; for the scrupulous
consultant, this creates the frustration
of not satisfying the clients' desire for
a no-further-action status.
The LUST Paradox
The goal of LUST regulators and
fund administrators, nationwide, is
to use available resources to protect
the environment from the impact of
leaking underground storage tanks.
Yet our enthusiasm to achieve the
• continued on page 26
25
-------
LUSTLine Bulletin 43 • March 2003
• When RBCA is not RBCA
from page 25
fullest protection may result in less
protection! Let us explain this para-
dox in terms of programs in which
the regulator is separate from the
fund administrator; although the
argument also applies to instances
when the regulator and fund admin-
istrator are not separate.
The regulator is primarily con-
cerned with protecting water
resources and restoring contami-
nated resources to a legislatively
defined standard of acceptability.
State fund administrators are univer-
sally concerned with using their lim-
ited financial resources to achieve the
regulatory goals for as many sites as
possible. The separation of a fund
from a regulatory authority is
intended to assure that the financial
resources do not determine regula-
tory standards. In this arrangement
the regulator may demand restora-
tion to the strictest standard and only
allow less stringency when "best
efforts" have failed to achieve the
results.
The operative words here are
"best efforts." If the consultant is not
working toward an obtainable goal,
the engineering design cannot be
optimized and "best efforts" will
remain elusive. Best efforts become
the longest and most expensive
efforts. In the PARC scenario, the
goal is one that the stakeholders have
agreed is obtainable and, therefore,
metrics can be established to track
progress. These metrics become the
payment milestones of PEP.
Contract Goal versus
Cleanup Goal
Some stakeholders in California have
argued that the PARC reduces
cleanup standards and, by doing so,
is simply a back door to RBCA, over-
riding what the local regulations
have specified. This is not true; the
PARC does not obligate the regulator
to close the site. Rather it directs
resources toward achieving obtain-
able goals while leaving full regula-
tory structure and authority in place.
The PARC is a contract goal, not a
cleanup goal. This approach leads to
faster, more efficient cleanups and
provides the regulator with a greater
level of satisfaction.
26
Consider a hypothetical example
of how a PARC might work. Lef s say
Joe's U-Pump suffers a loss of 100
gallons from a pipe-connection fail-
ure. The released gasoline migrates
50 feet through the vadose zone to
the surficial aquifer (water table).
There are no sensitive receptors, such
as potable wells or surface water,
within five miles of the discharge.
The subsurface is composed of
interfingered sands and days.
The MCLs for benzene and
MTBE are 1.0 ppb and 13 ppb,
respectively. Due to the clays it is
quite difficult to achieve these levels.
With clays under the building having
absorbed product, the only way to
fully achieve the MCLs would be to
remove the building and excavate the
soil or install vapor laterals under the
building. The structural engineer is
opposed to laterals. Similar facilities
in the same town have had vapor
extraction and air-sparging systems
operating for years, without appre-
ciable success.
Using the traditional approach, a
consultant would continue "best
efforts" until reaching the funding
limit, and then the regulatory agency
would consider a petition for no-fur-
ther-action with contamination still
present, though at reduced levels. In
the PARC scenario, the regulator,
consultant, owner/operator, and
fund administrator meet upon com-
pletion of the assessment. They agree
upon obtainable and practical
interim cleanup goals that will pro-
tect receptors and minimize off-site
migration. These goals are the
PARGs.
From Process to Results
A PFP agreement motivates the con-
sultant to install effective and effi-
cient technology and motivates the
consultant to operate diligently. PFP
encourages and allows the regulator
to refocus resources from process to
results. The fund administrator
preapproves the PFP agreement pay-
ment terms and no longer needs to
review complex time and materials
invoices. Payments are based upon
milestone contamination reductions
within agreed-upon time parameters.
As in all PFP agreements, there
are termination clauses for unfore-
seen events, such as a new discharge.
Upon reaching the PARC, the PFP
agreement is terminated. At this
point, the regulator will most likely
either issue a site closure or require
additional active remediation, or nat-
ural attenuation monitoring.
In the example above, the regula-
tory authority was not reduced under
a PARC scenario, yet effective conta-
mination reduction was achieved.
Most importantly, the reduction was
greater and faster than the neighbor-
ing facilities' attempts to reach the
MCL without the PARC and PFP.
Elements for Success
Many sites in California are proceed-
ing according to the PFP method and
satisfying all concerned. Through the
simple combination of the PARC and
PFP, cleanups can proceed without
the tension between the fund admin-
istrator and the regulators. The key
elements for success are:
• Accurate and complete assess-
ment, at least at the source area
• Early post-assessment agreement
of practical interim goals
• A PFP agreement to reach the
PARC
• Project review once the PARC is
achieved
The PARC may be a site-specific
goal or a regional goal. The Santa
Clara Valley Water District (SCVWD)
promulgated a regional PARC for
PFP sites that do not have an impact
on sensitive receptors. This is notable
since the district is considered strin-
gent. It took up the PARC to break
the gridlock of endless and inefficient
remediations. Let's look at two
SCVWD cases studies.
Santa Clara County
There are currently two PFP sites in
Santa Clara County that have active
remediation systems and two sites
that are in the process of installing
their remediation system. The first
site began remediation in December
2001 and has reduced contamination
levels by 70 percent. The second site
began remediation in May 2002 and
has achieved a 75 percent reduction
in contamination levels. Both of these
cases were in the program for over 10
years with little or no progress made
towards achieving closure prior to
entry into the PFP program.
-------
March 2003 • LUSTLine Bulletin 43
The SCVWD has developed a
standard set of PARGs for total petro-
leum hydrocarbons as gas (1,000
ppb), benzene (100 ppb), toluene (200
ppb), ethylbenzene (500 ppb),
xylenes (300 ppb), and MTBE (200
ppb). These PARGs are applicable for
all PFP sites in Santa Clara County.
The district believes that they are
achievable and will likely allow the
majority of sites to be closed, if
achieved. The SCVWD does not
guarantee closure upon reaching the
PARC but does guarantee that the
site will be reviewed for closure
when the PARGs are achieved.
• Case Example #1
Fuel leak site #1 was first opened as
an active case in January 1990, fol-
lowing the discovery of petroleum
hydrocarbons in soil during the
removal of two 8,000-gallon fuel
USTs and a 550-gallon waste oil UST.
Soil and groundwater investigations
and quarterly monitoring events con-
ducted between 1993 and 2000
detected total petroleum hydrocar-
bons as gasoline (TPHG) and
benzene in groundwater at concen-
trations up to 140,000 and 34,000
parts per billion (ppb), respectively.
Additionally, free product was
detected at thicknesses up to 1.65
feet. No site remediation was com-
pleted in the 12 years following the
initial report of a release at the site.
The reasons for this delay in site
remediation are not clear, but the PFP
initiative gave the cleanup its jump-
start.
A PFP cleanup agreement for the
site was reached in June 2001 and
was modified in April 2002 to take
into consideration additional conta-
mination discovered in December
2001. The consultant proposed the
use of groundwater extraction (GWE)
in combination with soil vapor
extraction (SVE) and air sparging
(AS).
The SVE and AS systems are being
conducted with the use of a mobile
unit that allows considerable flexibil-
ity in selecting the wells to be used
for extraction. After six months of
GWE and two months of SVE and
AS, total BTEX concentrations in key
monitoring wells at the site have
been reduced by 77 percent. Current
maximum concentrations at the site
are 19,430 ppb benzene (a 43 percent
reduction) and 38,000 ppb TPHG (a
73 percent reduction).
The site consultant feels the PFP
approach resulted in faster payment
from thej cleanup fund, eliminated
much of ithe administrative hassles
that were: involved with the standard
cleanup fund process, and provided
him an opportunity to increase prof-
its. We believe this is a win-win situa-
tion for all parties.
1 i
i
• Case Example #2
Fuel leak jsite #2 was first opened as a
leak site in February 1992 following
the discovery of petroleum hydrocar-
bons in soil during an UST removal.
Maximuni concentrations at the site
were 31,OpO ppb benzene, 10,000 ppb
MTBE, and 200,000 ppb TPHG. Reme-
diation activities performed at the site
between 1991 and 2001 included addi-
tional excavation of contaminated soil
in 1991 a4d 1993 and the installation
and operation of SVE and GWE sys-
tems from: 1996 to 2001.
Following the initial period of
remediation, benzene, MTBE, and
TPHG concentrations were reduced
by 72 percent, 97 percent, and 73 per-
cent, respectively. Although the total
reduction! of contaminants was posi-
tive, most, of this progress was made
between 1996 and 1998, and the con-
centratiorls remained stable from
1998 to 2001.
The responsible party and the con-
sultant entered into a PFP contract in
November 2001. The consultant pro-
posed to continue to use the existing
systems but to upgrade the equip-
ment to provide increased flow rates
and to increase the number of extrac-
tion wellsJ The upgraded system was
started up in December 2001. After
eight months of operation, total
BTEX concentrations were reduced
by 70 percent. Benzene, MTBE, and
TPHG concentrations were reduced
by 78 percent, 46 percent, and 37 per-
cent, respectively.
These case studies show that the PFP
program is effective in moving stag-
nant sites into successful remediation.
Setting realistic PARGs—goals that
are achievable and will provide pro-
tection to Ijhe groundwater basin—is
a critical component of the PFP pro-
gram. Without realistic PARGs, it
would be 'difficult, at best, for any
consultant to agree on a PFP contract.
Progress Without Angst
Introducing the PARC in combina-
tion with PFP can lead to faster, less
expensive, and, most importantly,
more efficient cleanups. Setting
obtainable goals is a powerful tool for
avoiding regulatory tension with the
trust fund and for getting all sides to
focus on what we all want—maxi-
mum protection of receptors by effi-
cient use of resources. Pay for
Performance has produced remark-
able results in many states by pro-
moting faster and less expensive
cleanups. (See LUSTLine #42, "Pay for
Performance: Does It Work? The
Data.") The PARC combines ele-
ments of RBCA with PFP to acceler-
ate cleanups through efficient
operations. •
Robert S. Cohen is a professional geolo-
gist specializing in LUST cost contain-
ment and risk management issues. He
is a consultant to both the public and
private sectors. He has conducted over
30 PFP workshops and studies on
behalf of the EPA and various states.
For more information, contact Bob at
bobcohen@ivs.edu or
(352) 337-2600
David F. Charter is a Senior Engineer-
ing Geologist in the California Under-
ground Storage Tank Cleanup Fund.
He has been involved in UST removal,
assessment, and remediation since
1990. He can be reached at
dcharter@cwp.swrcb.ca.gov.
George E. Cook Jr., R.G., is an Assis-
tant Engineering Geologist with the
Santa Clara Valley Water District. He
is responsible for oversight of 100
active fuel leak sites in Santa Clara
County, California. He can be reached
at (408) 265-2607 ext. 2665 or
gcook@valleywater.org.
Kevin Graves is Chief of the UST
Cleanup Unit at the State Water
Resources Control Board in California.
He has worked in the environmental
remediation field for the past 10 years
as a regulator, researcher, and consul-
tant. He can be reached at
gravesk@swrcb.ca.gov.
27
-------
LUSTLine Bulletin 43 • March 2003
a
At the llth Annual State Fund Administrators Conference, June 2003, in Boise, Idaho,
there was much discussion concerning LUST cleanup reimbursement, Pay for Perfor-
mance (PFP) style. We've selected some of the questions asked and some of the answers
provided that we felt might be of interest to state fund administrators considering imple-
menting PFP.
PAY FOR PERFORMANCE
C2. Wlien you negotiate a PFP
agreement, how do you know when
you get a good price?
Experience, experience,
experience! The more experience
state officials and contractors have in
PFP, the more confident all the par-
ties will be with the negotiated price.
For example, Florida program per-
sonnel, with over 250 agreements,
have gained considerable experience
at knowing what things cost. For
each project, three state project man-
agers independently review the
cleanup plan developed by the nego-
tiating contractor and then develop
an estimate of the price of the
cleanup. These three estimates are
then compared to the price proposed
by the contractor. "If the contractor
had a very high price and then sud-
denly comes down to meet your esti-
mate, you know there was a lot of fat
in the proposal to begin with," said
Florida DEP's Brian Dougherty.
Oklahoma, on the other hand, uses
the Tank Racer costing program as a
basis for negotiations. "Tank Racer
removes any question of arbitrariness
and has proven useful in court," said
Dave Kelley, Oklahoma Corporate
Commission, PST Division.
C<2. What were some of the most
unexpected benefits of PFP?
One state representative
noted that contractors make their sys-
tems work better without contacting
agency staff. He said the back and
forth calls between state staff and
contractor, the change orders, and the
cost overruns all go away under PFP.
Art Shrader, South Carolina
DHEC, recalled one occasion when a
hurricane was approaching. The PFP
contractor notified his office that he
was shutting down the systems. Six
systems were under water, and the
state didn't have to pay to replace
any of them.
"PFP helped whittle out all the
bad consultants. Now, the technical
people that we work with are more
28
knowledgeable. The good consul-
tants rise to the top and stand behind
their work," said Dave Kelley.
O. How has PFP changed the
way you approach site characteriza-
tion?
. "You need more thorough
site characterization for consultants
to feel comfortable about entering a
PFP agreement," said Chuck Schwer,
Vermont DEC. "Our fund pays a bit
more for a better characterization, but
if s worth it."
Art Shrader noted that if you don't
have a good site assessment, you're
gonna find out about it in the price
tag.
d. Do you allow for real-time
sampling after the contract is
awarded?
. In South Carolina, baseline
levels are established before entering
a PFP agreement. If site conditions
change significantly between the time
of the baseline sampling and the
implementation of toe agreement
(e.g., free product levels double), the
contractor can be released from the
agreement and the project can be re-
bid.
&,. Does state lead on a cleanup
seem to work more smoothly?
. Florida experience says get
as much state lead as you can. You
have more problems when owners
choose the contractor. Vermont, on
the other hand, has not done any
state-lead PFP work; however,
they've found that if they work very
closely with the owner /operator, a
better agreement can be developed.
VTDEC staff have helped owners/
operators conduct bidding and rou-
tinely meet with contractors during
the development of the agreement.
"Working as a team, owners /opera-
tors and the state can be each other's
best ally," said Chuck Schwer.
O. How do you justify which
contractor is chosen?
-. To have an objective way to
evaluate bids, Utah uses a Proposal
Evaluation Form that consists of a list
of the criteria required for each bid-
der's proposal and a scoring system
for evaluating the proposals accord-
ing to the criteria. Six people score
the proposals using the same rating
scale. Higher ratings are assigned to
bidders that clearly define the work
to be performed. Higher ratings are
also assigned to realistic assumptions
of equipment performance, flexible
cleanup plans that cover a range of
potentially unknown conditions, and
offers that provide adequate equip-
ment and backup in the event of
breakdown, that adequately address
all environmental and health and
safety concerns, and that minimize
the risk of not completing work on
schedule. (For a copy of Utah's
Request for Bid Form, contact Randy
Taylor at rtaylor@deq.state.ut.us.)
O. What techniques do states
use to control contractors?
. Since the stakes are high
under PFP, you need to ensure that
your agreements have some stiff con-
sequences if the contractor walks
away from the project. For example,
in Vermont, if the contractor doesn't
fulfill his obligations, he risks losing
work under the fund for three years.
In Oklahoma, if a contractor fails to
fulfill his contract, he cannot perform
any remediation work for two years,
by legislation. "If you don't have
such legislation, you should put it
into the contract," remarked Kelley.
O. Does the ability to encumber
funds through PFP help prevent raids
on state funds?
In 2002, the Vermont fund
was raided by the legislature. If not
for PFP and monies encumbered for
the cleanups, the amount taken by
the legislature would have been
much larger. In Oklahoma, because
of PFP, they were able to encumber
funds. Florida has found that if you
encumber funds, money is saved. Hi
-------
March 2003 • LUSTLine Bulletin 43
Tank Systems in a Jam
Contaminated Gasoline from a Kentucky
Refinery Spurs a Flurry of Tank
Cleanups and Lingering Concerns
by Ellen Frye
In summer 2002, incidents at sev-
eral retail gasoline facilities in
West Virginia caught the atten-
tion of inspectors from the state's
Department of Environmental Pro-
tection (WVDEP) and Division of
Labor Weights and Measures Section
(W&M). "In June, we were hearing
from tank owners who were doing
inventory control and statistical
inventory reconciliation (SIR) and
getting inconclusives and faulty
meter readings/' says Gil Sattler,
WVDEP's UST program manager.
In August, Dennis Harrison, a
supervisor with W&M, was con-
tacted by a major jobber who said he
was being overcharged for his 87-
blend gasoline and suspected that
there was a calibration problem at the
bulk storage loading racks. "We
quickly discovered that the turbine
meters at the loading rack were out of
calibration," says Harrison. "We sus-
pected there was something amiss at
the marine terminals.
"The first physical evidence I saw
was at an Exxon station, where a sub-
mersible pump was plugged with par-
ticulates. One of the pumps had
already been replaced," says Harrison.
Harrison contacted Marathon
Ashland Petroleum LCC (MAP), the
refiner/distributor, and told them of
his finding. A MAP representative
instructed him to remove the filters
and empty them into a pan. Upon
doing this, Harrison found that the
filters were filled with particulates
that had been sucked up by the
pumps. Marathon immediately sent
teams of investigators to its refineries
and barges.
About two weeks later, Harrison
started hearing from other station
owners, complaining of meters
breaking, filters clogging prema-
turely, and some pumps failing. "Job-
bers aren't big on talking to each and
comparing these kinds of notes," says
Harrison. "So unfortunately it can
sometimes take them awhile to real-
ize that it is not just their problem."
MAP inspected more than 2,100
retail stations that often had as many
as 20 mefers per site. The company
replaced 26 broken meters, less than
one tenth of one percent of the meters
inspected;,
j
The Problem
Fuel contamination was soon traced
to MAP's refinery in Catlettsburg,
Kentucky^ ^ source of about 70 per-
cent of West Virginia's fuel. The
refinery, according to MAP, had been
"distributing gasoline containing for-
eign material." The company took
immediate steps to stop the contami-
nation at the source and.clean all
affected points along the distribution
system. Ultimately, the problem was
identified! at facilities receiving prod-
uct from this refinery in West Vir-
ginia, Kentucky, Ohio, and Indiana.
According to a brochure, Fuel
Quality: Everybody's Business, pre-
pared by MAP and distributed to the
jobber and service station network in
the affected area, the company per-
formed analytical sampling on more
than 2001 terminal and bulk plant
tanks in tijie Catlettsburg distribution
system. ;
MAP j identified the following
three types of contaminants that
could be (traced to the Catlettsburg
fuel qualify issue.
i
• Excessive spent sodium hydrox-
ide (NlaOH) water (caustic, high
pH) that is normally recovered
from gasoline before it leaves the
refinery. NaOH is used in conjunc-
tion with air and a catalyst to react
with mercaptan and ultimately
remove sulfur compounds. Mer-
captan is a sulfur-containing
organic compound that is a
byproduct of tike gasoline refining
process;
Excessive rust participate gener-
ated from within the refinery and
the barges that was carried down-
stream;' According to MAP, the
participate itself was not gener-
ated from UST corrosion. MAP
says that other scavengers used to
remove mercaptan kept particu-
late from bulk tanks and barges in
suspension and carried it down-
stream to retail facilities.
• Thick sludge resulting from a
reaction of chemical additives
with finished product. The use of
these additives has now been dis-
continued.
More than 70 of the tanks at the
terminal and bulk plants were
cleaned. All Catlettsburg-sourced
barges were inspected—of 124
barges, 43 required cleaning. More
than 220 company-owned, commer-
cial, and private transports delivering
product to the four-state area were
inspected, and 90 were cleaned. Of
the more than 2,100 retail sites in the
four-state area that were inspected,
approximately 1,300 of those sites
underwent tank cleanings. MAP has
also addressed claims by car owners
concerning problems with fuel sys-
tems, primarily fuel injectors.
State inspectors also found signs
of microbial blooms at the
product/tank system interface dur-
ing and after cleaning. "The truth is, I
don't understand it all," says Dennis
Harrison. "We routinely check the
water bottoms. In some of these tanks
we've seen a noticeable amount of jet
black water... after they'd been
cleaned."
"MAP's efforts to look at under-
ground tank bottoms have been
unprecedented," says MAP spokes-
man Shawn Lyon. "The dark liquid
found at the bottom of tanks is a nor-
mal occurrence. Using various sam-
pling methods, field tests conducted
outside of the impacted Catlettsburg
area further prove this point. With
that understanding, MAP's cleaning
efforts focused on removing particu-
lates."
Long-Term Concerns
Based on what inspectors were see-
ing in affected tanks, WVDEP's UST
Advisory Committee discussed some
long-term concerns they had about
the potential for corrosion or other
adverse impacts to the linings of
USTs.
• continued on page 30
29
-------
LUSTLine Bulletin 43 • March 2003
m Tanks in a Jam from page 29
"During the tank-cleaning pro-
cess, we noticed that the water in the
tank bottoms looked different at dif-
ferent locations," says WVDEP UST
inspector Michael Young. "In steel
tanks (those that are bare steel on the
inside) the water was a rusty brown
color. In lined and fiberglass tanks
we noticed that it was a dark
red/purple/ black color. It was also
more viscous and some places it was
sticky.
"The results from some tank-bot-
tom sampling that we conducted
indicated that there were significant
metals and some odd organics in the
watery tank bottoms," says Young.
"Some of these constituents are the
same as the binder material that is
used in the tank linings; however,
they can also be found in parking lot
runoff and soil. We realized that
these samples were of little use
because they were not definitive.
Even if we looked into some of the
lined tanks and measured the thick-
ness of the liners, we have no base-
line data indicating what the lining
thickness was when it was first
installed. Without this data it would
impossible to determine if any liner
loss had occurred."
Responding to WVDEP's con-
cerns about long-term tank integrity
and microbial growth, MAP'S Lyon
explains that of all the samples taken
at the retail station level, die pH read-
ings were normal (in the 6 to 8 range)
with only a nominal number of sam-
ples slightly elevated to 10. Further
analysis by a third-party testing lab
noted that NaOH water found in the
system is no more corrosive than nor-
mal water.
"Essentially, what we have
occurring is normal corrosion, and, as
a result, we implemented a plan to
remove the NaOH water and the
solids in the system," says Lyon. "In
addition, industry standards for sta-
tion equipment such as UL-listed
storage tanks and piping include test-
ing to withstand exposure to high pH
solutions, which should further vali-
date tank integrity."
With regard to microbial growth,
Lyon contends that this is not a
Catlettsburg issue but that "it is
becoming more common in today's
environment of [EPA-regulated]
cleaner burning fuels. Therefore,
30
there is an increasing need for educa-
tion and awareness of this topic."
Lyon notes that "microbial-induced
corrosion is typically much more
aggressive than normal corrosion
associated with the presence of
water." (See "Does Your Tank Sys-
tem Have Bugs?" below.)
In the meantime, WVDEP inspec-
tors remain somewhat puzzled about
the goings on in so many tank sys-
tems. "While Marathon has been
very responsive to this problem,"
says Gil Sattler, "we still have some
concerns about the long-term effects
of all of this."
What Should Tank Owners Do
to Ensure Product Quality?
In its brochure to affected jobbers and
station owners, MAP describes what
it is doing to ensure product quality
throughout its distribution system.
The company also notes that jobbers
and dealers have a "responsibility to
continue efforts to assure product
quality after product leaves the ter-
minal." Jobbers and dealers can help
achieve this by
• regularly monitoring and mini-
mizing water levels in under-
ground and aboveground storage
tanks,
• scheduling regular changing of
dispenser filters, and
• developing a plan to proactively
monitor for microbial growth at all
tank locations. •
Does Your Tank System Have Bugs?
We've heard from several UST inspectors who are concerned that they are
seeing internal corrosion associated with microbial contamination in USTs. In
LUSTLine Bulletins #39 and #40, industrial microbial ecologist Fred Passman
provided our readers with extensive information on this subject.
If you are wondering whether microbes are likely to be causing problems
with your UST or dispenser systems, here are a few quick checks that Fred has
provided us that will help put you on the right track.
• What does your bottom sample look like? Is there a third layer in the
sample that looks like mousse and tends to cling to the sides of a gently tilted
glass jar? If you see this in your bottom sample, you can be better than 90
percent certain that you have enough microbial contamination to be causing
problems.
• DO you have a flow-rate restriction? Note the totalizer reading when you
change dispenser filters. If your flow rate falls below 7 gpm in less than
50,000 gallons through the filter, there's a better than 75 percent chance that
microbes are a major, if not the major, cause.
• Have you replaced dispenser flow-control valves more than once in
the past year? If so, there's a better than 75 percent chance that microbes are
causing the fuel to become corrosive.
• Have you experienced more than one of these problems at a retail
site? If so, there is a 95 percent chance that you have a microbe problem in
your UST.
Also, ASTM Manual 47: Microbial Contamination of Fuels and Fuel Systems
is scheduled for publication in spring 2003. This manual will contain three
original chapters plus all of the nonvolume 5 ASTM and IP standards to which
ASTM's D6469 Standard Guide to Microbial Contamination in Fuels and Fuel
Systems refers (See LL#38). The three original chapters in Manual 47 are:
• Fuel Microbiology Basics
• Fuel Sampling for Microbiological Examination
• Fuel System Contamination Control
The manual also includes a glossary of microbiological and filtration terms
with which petroleum industry people might not be familiar.
For more information, contact Dr. Fred Passman, President, BCA, Inc.
bca-fjp@ix.netcom.com or (609) 716-0200. •
-------
March 2003 • LUSTLine Bulletin 43
UST-Related Explosions in Kentucky and
Pennsylvania Are Unfortunate Reminders of
What Not to Do
by Ellen Frye
"A RIVER OF GAS" AND
BOOM
On the morning of December 12,
2002, Virgil Burgan, the operator of
the Mount Eden Country Store in
Shelby County, Kentucky, was on his
way out of the store to fill a cus-
tomer's kerosene can. He'd noticed
that a gasoline delivery truck was
making a drop but that it was parked
in a different direction than usual. As
Burgan walked to the kerosene
pump, he saw what he said looked
like a "river of gas" flowing along the
left side of the store to the back of the
building.
Burgan yelled to the driver and
went back inside, where he found the
smell of gas was already strong. He
asked everyone inside to put out any
smoking materials and turn off the
stove burners. But as folks started to
get up and leave, they heard—and
felt—an explosion. They ran from the
building just before the store caught
fire and burned to the ground. No
one was injured.
Several factors caused this deliv-
ery incident, according to Dale Man-
cuso, Senior Deputy State Fire
Marshal. "First and foremost, a liq-
uid-tight connection was not estab-
lished before filling the tank," says
Mancuso. A bulk delivery nozzle
used to fill aboveground tanks had
been placed inside the opening of the
fill riser pipe, and the product was
pumped into the tank.
Underground tanks are equipped
with an adapter on the fill riser pipe
that allows the delivery truck to hook
up without the possibility of the hose
disengaging. Once this connection is
established, the valve to the storage
compartment of the truck is opened
and the product flows by gravity into
the tank. "Even if a liquid-tight con-
nection had been used, the. delivery
would still have gone against code
requirements," says Mancuso, "since
overfill-prevention devices used in
underground tanks are not listed for
pump-fed deliveries."
According to Mancuso, another
factor that contributed to the release
was that the nozzle being used to fill
the tank did not have an automatic
shut-off device. Although state codes
don't require this safety feature in
bulk-fuel nozzles, Mancuso says that
such a device could have prevented
the releas^ and the subsequent fire.
The Driver had also parked his
truck so that the rear of the truck was
in the front of the store. The UST fill
connectioin was on the side of the
store, so the driver could not have
seen it from the rear of the truck since
the store (blocked the line of sight.
The driver explained to investigators
that he had been going back and
forth between watching the meter at
the back of the truck and his nozzle at
the fill connection.
"The amount of product ordered
played a role too, in that the Burgans
only ordered 400 gallons of gasoline.
The proper way to make this delivery
into a UST, would have been to meter
400 gallons of gasoline into a com-
partment !of the tank truck and then
deliver it. Instead, the driver chose to
fill up the compartment," explains
Mancuso.
Mancuso says the driver had to
be at the rear of the truck to make
sure that only 400 gallons went into
the tank. \ The tank didn't have a
meter with a preset cutoff, a device
that shuts off the flow of product
when a preset amount is reached.
As for the gasoline spill, at this
point investigators do not know what
caused it. Mancuso suggests a few
possibilities: The nozzle could have
been kicked out when the driver was
going back and forth from the meter
to the fill. The restriction in pipe size
from the UST overfill-prevention
device could have caused excessive
turbulence in the tube and created
backpressure. A restriction in the
vent line to the UST could have
caused the problem.
"One fact still remains,' says
Mancuso, "had a tight fill connection
been used along with a gravity fill,
this release would never have hap-
pened."
TANKS BLAST OUT OF
DORMANCY AT MTBE
CONTAMINATION SITE
On January 7, 2003, two under-
ground tanks at a former Mobil sta-
tion at the Pool's Corner area of
Doylestown, Pennsylvania, exploded,
injuring two workers, rocking the
neighborhood, and sending a shower
of rocks and tank debris onto passing
cars and residential properties. The
station had been closed for over two
years because of a gasoline release
that contaminated a dozen nearby
wells with MTBE. Both the Mobil sta-
tion and a nearby Exxon station had
voluntarily closed to clean up the
Sltes' • continued on page 32
Recommendations for Preventing a Similar
Fuel Delivery Situation
In the wake of the Mount Eden fire, Senior Dputy State Fire Marshall Dale Man-
cuso made the following recommendations.
• Provide better training for drivers/operators.
• Use tight fill connections.
• Fill USTs by gravity, not by pump.
• Follow the manufacturer's recommendations on product usage. Specifi-
cally, do not use a pump in conjunction with an overfill-prevention device
listed for USTs.
• Make deliveries to USTs with premetered amounts of fuel.
• Make sure that the driver/operator can always see the tank-fill connection
as well as the truck.
31
-------
LUSTLineBiiHefm43 • March2003
m UST-Related
Explosions from page 31
On. the day of the
explosions, contractors
were in the process of
upgrading the tops of the
four 10,000-gallon tanks
at the site, replacing the
lines and dispensers, and
installing a remediation
system. New owner
Conoco Inc. was doing
this in preparation for a
reopening of the station.
When the station was
dosed, the tanks had been emptied of
product but not cleaned—about an
inch of sludge remained at the bot-
tom. At the time, the tanks were filled
with water, as ballast.
"There seemed to be an assump-
tion that the water would negate the
need to monitor the tanks inside and
out for vapors," says Kathy Nagle,
Pennsylvania Department of Environ-
mental Protection Water Quality Spe-
cialist Supervisor. "We haven't
identified the ignition source, but the
vapors from the tanks were clearly at
levels high enough to trigger the
Left: Exploded tank closest to the road at
Pool's Corner Mobil Station. The foam in the
tank area is fire-fighting foam. At upper left
are the two containment sumps for the mid-
dle two tanks In the field.
Bottom: The tank at the opposite end of the
tank field blew out enough of the gravel back
fill to cause the macadam to collapse.
explosions, first in one tank
and then in the other."
Well users in the area
have relied on bottled
water since the MTBE was
discovered. Some of them
have sued Tosco Refining
Co., former owner of the
Mobil station, and Exxon-
Mobil Corporation, owner of the
Exxon station. The explosions have
raised new concerns about well conta-
mination. ConocoPhillips and Exxon-
Mobil have stepped up well sampling
schedules to assess possible impacts
on contaminant levels. Ongoing sam-
pling is under review by PA DEP. The
MTBE-impacted homes (i.e., MTBE
levels above the cleanup standard of
20 ppb) have had point-of-entry niters
installed. H
Building a Better Internet Presence
Ten Things Yon Can Do to Improve the Performance of
Your UST Web Site
by Ben Tiiomas
Technology can be a blessing and
a curse when it comes to deliv-
ering technical assistance on
underground storage tanks via the
Internet. If s a blessing when you can
provide UST operators with perpetual
access to a plethora of rules, forms,
lists, and documents to help them
achieve operational compliance. But
it's a curse when operators can't find
what they want on your Web site.
Chances are the stuff they seek is
there, but it's buried somewhere on
your Web site. The good news is that
you can fix this. All it takes is some
analysis, design work, some coding,
the nod of the boss/and viola! You
improve your Web site, increase
accessibility, and have a happier,
more educated operator.
32
So, at that juncture of your next
Web page update, consider the mak-
ing following improvements:
JL« Know your audience.
For a Web page to meet its goal of
increased access to useful informa-
tion, it must be tailored to its
intended audiences. Who uses your
Web page the most? Owners? Con-
tractors? Consultants? Other govern-
ment employees? Knowing your
audience will help you organize your
information so that it is as eye-catch-
ing and intuitive as possible. The
most effective UST Web pages out
there today are the ones intentionally
designed to greet the target audi-
ences the moment they show up at
the front door.
Idea: Analyze your audience by
percentage of user types, and see if
your Web page is designed for those
audiences. Redevelop the page as
necessary, based on user type.
2* Realize that users don't read,
they surf.
Internet users don't navigate through
your Web page like you might think
they do. Studies show that users
don't really read introductory state-
ments. They don't study any particu-
lar Web page at length. They don't
read it like a book. Basically, they
surf. They surf and surf and surf until
they find what they want, and then
they bail out. Users roam a Web page
like a driver in a bustling city scan-
ning for street signs. If a user wants
-------
March 2003 • LUSTLine Bulletin 43
an UST installation form, then the
glowing "form" button is all they see;
all else is blurrily irrelevant. If you
understand better how a user typi-
cally navigates through your page,
you can arrange the layout for opti-
mal performance.
Idea: Study your Web page, and
look for wordy distractions that
could be culled. As Strunk and White
suggest in their little book, The
Elements of Style: "Omit needless
words!" I would add to this:
• Relocate lengthy introductory
statements with links or buttons
that say something intuitive like
"Mission" or "About Us."
• Remove or relocate esoteric topics.
• Use bullets to organize topics
whenever possible by deleting
accompanying text that doesn't
help navigation.
Think of the sum of your Web
pages as a series of traffic signs that
are designed to get drivers to their
destination as simply and clearly as
possible.
«»• Prioritize information.
If a user comes to your Web page
looking for something basic like a
form or a list of service providers,
they shouldn't have to spend any
time looking for the link. It should be
obviously and immediately located
right there at the top of the home
page. Organize the links to these
types of common items by making
them abundantly apparent the sec-
ond a user hits your page. Steve
Krug's excellent book, Don't Make Me
Think, says users who are scanning
your Web page should need take
only a few milliseconds to decide
where to go next in search of their
objective.
Idea: Write down what you think
is the most popular information on
your Web page, and see if you can
find it at first glance on entering your
Web page. Think of the top half of the
first page as waterfront property—
prime real estate. Use that space for
all your mission-critical elements.
The less common items should be rel-
egated to the lower half of your page.
4» Keep your home page short.
Your home page, where your user
presumably starts, should be short.
Users generally avoid scrolling
downward (and never, ever across)
and with a well-designed home page,
you have ;all the basics laid out in an
intuitive fkshion.
Idea: If you have a lengthy intro-
ductory page, chop it down to a one-
pager. For example, shrink all text
fonts (8 point is fine) except for topic
headers. (Shrink the banner (your
logo). Surnmarize and hyperlink the
common things people look for (e.g.,
news, forrns, regs, documents, lists).
Consolidate banners, navigation
bars, and links to reduce the amount
of empty space.
51
• Avoid jargon.
While jargon helps streamline the
language lamong regulators, it is a
primary barrier to communication
with the general public. Jargon only
increases b| arriers to user comprehen-
sion. Nearly all UST Web pages out
there could do with a little bit of jar-
gon reduction. Even if you think your
audience knows your particular fla-
vor of acronyms, make your Web
page supejr-intuitive by saying, for
example, "Installation Form" instead
of "State Form 27B-6."
Idea: ; Review your pages for
acronyms,! ]'argon, and shop talk.
Remove this language that is not
common or useful to the audience.
The acronym "UST" is probably okay
to keep. Use common language to
highlight concepts.
fj» Keep a uniform look to all
pages. |
This may s^em obvious, but the more
your Webj pages all look the same,
the more cbmfortable your user will
be navigating through your site on
the way to jtechnical nirvana. Naviga-
tional bars, banners, fonts, and gen-
eral layoutlshould all be as similar as
possible. I
Idea: Scan your pages and look
for thematic similarities among your
pages. Correct the odd-duck sites and
strive for a |Uniform look.
T i
4 • Use/fix navigation bars.
Navigation! bars are common tools on
Web pageslthese days, and every sin-
gle one of your Web pages should
have them1. This lends itself to the
uniform look mentioned above. Nav-
igation bar's should contain founda-
tion-type links, such as "Contact
Informatioiji," "Search," and "Index."
Today's W^b user has come to expect
this navigational system ancf will
look for these features on your pages.
Idea: Survey your pages and
evaluate whether you have effective
navigational bars, or whether you
have them at all. Strive for simple,
compact, consistent bars.
O« Avoid gadgets.
An ambitious and creative Web
designer can easily get carried away
with the arsenal of tools available to
enhance a Web page — features such
as animated graphics, scrolling
banners, and snazzy background
designs. Often, these gimmicks, if not
tightly controlled, distract rather than
enhance the content. We all know
that the poor content of a speaker's
message cannot hide behind the
appeal of a nice-looking PowerPoint
presentation. The same is true for a
Web page. If the enhancement
doesn't enhance, don't bother.
Idea: Review your page to deter-
mine whether your enhancements
help or hinder. Have someone out-
side the program perform the assess-
ment. When in doubt, go basic.
"• Provide a visually pleasing
layout.
Users should get subliminally warm
and fuzzy when they enter the realm
of your Web site. They should feel a
certain amount of ease and comfort
because the information is laid out
intuitively, and certain categories of
information are smartly lumped
together and compartmentalized.
White space allows users to scan the
page and zero in on what they want.
Any white space should have the
subliminal ability to guide users
along. Poorly designed white space
can confuse users in the matter of a
split second.
Idea: Study highly successful
Web pages such as www.ebay.com
or www.amazon.com. See how they
lay out a page. Study the use of white
space, the lumping together of topics,
the hierarchy of information, and the
overall appeal, including font, colors,
and background design. Then evalu-
ate your page.
Have searchable databases.
A number of government Web pages
have searchable UST and LUST data-
bases. In these lucky states, opera-
tors, prospective property buyers,
• continued on page 34
! 33
-------
LUSTLine Bulletin 43 • March 2003
• Web Pages from page 33
tank workers, and environmental
consultants all have access to either
thumbnail or exhaustive UST data-
base files. Even mildly savvy Web
users are coming to expect this sort of
service. Plus, it will save you tons of
time and money rooting though your
database files.
Idea: Convince your database
staff and your program manager that
your Web page will be infinitely more
effective if it has a searchable UST
and LUST database. While your state
may have database platform con-
straints and money issues, keep beat-
ing this drum. Like digitizing other
UST services, it is a new thing and
will only save you money in the end.
So What Makes a Great
Web Page?
Layout, color scheme, organization,
content? Obviously there are lots of
things that make a Web page func-
tional and attractive. Some things are
apparent, but much has to do with
what goes on in the subliminal realm.
If users can surf effortlessly through
your site, grab what they want, and
leave, you've done your deed.
If I had to choose the most
important element of having an effec-
tive Web page, it would probably be
ease of navigation. All of the 10 ele-
ments listed here can be tied back to
how easily a surfer surfs your site.
If you are one of those UST regu-
lators who feels there aren't enough
UST operators online to warrant a
full-blown Web page analysis, don't
despair. Web usage is increasing
daily. By the time you finish your
assessment, the numbers will have
grown—and along with them the
expectation for better, stronger, faster
service. •
Ben Thomas is a former UST regulator
in Alaska. He created one of the first
state UST Web pages to deliver in-
depth technical assistance online. In
1997, he urged program managers at
the national UST/LUST conference to
seriously invest in the Web as a pri-
mary outreach tool. Today as an UST
consultant, Ben is always looking for
ways to improve access to information
about tanks. His Web page is
http://www.bentanks.com.
The United States Sues
New York City Over UST
Violations
The U.S. has filed a civil lawsuit in
Manhattan federal court against the
City of New York (NYC) alleging
that since at least 1997 and continu-
ing until today, the city has been
violating RCRA in connection with
its UST systems. NYC owns at least
1,600 federally regulated USTs in at
least 400 locations throughout the
NYC metropolitan area. The tanks
are operated by at least 16 agencies
or departments of the city.
The complaint charges that
NYC has for many years and con-
tinuing until today committed
numerous violations of RCRA and
the UST regulations, including the
failure to upgrade UST systems in a
timely manner; and properly pro-
vide, operate, maintain, and/or
monitor release-detection methods
in a timely manner; maintain and
furnish records concerning compli-
ance with release-detection require-
ments; report, investigate, and
confirm suspected releases of regu-
lated substances; comply with per-
formance standards for new UST
systems; comply with requirements
for closure of certain UST systems;
and maintain and furnish records
concerning closure of tank systems.
The complaint seeks civil penalties
and injunctive relief from the city.
EPA Issues a Unilateral Order
to Chevron USA for a LUST
Site in Chillum, MD
On November 26, 2002, U.S. EPA
issued a Unilateral Order to Chevron
USA, Inc., requiring it to investigate
the release of petroleum products
from a former Chevron service sta-
tion on Riggs Road in Chillum,
Maryland. The order also requires
Chevron to develop a cleanup plan
to address contamination associated
with the release. The order became
effective December 11.
A release of gasoline of un-
known quantity was discovered in
1989 in response to an investigation
of a traffic accident that damaged a
premium unleaded gasoline dispens-
ing line. The release was initially
addressed by the Maryland Depart-
ment of the Environment (MDE);
however, after many years of pump-
and-treat and monitoring, MDE noti-
fied the District of Columbia
Department of Health in April 2001
that a gasoline plume had migrated
into a residential area in D.C. Cur-
rent data indicate that the plume has
migrated approximately 1,600 feet
into D.C.
In October 2001, due to the juris-
dictional issues of cross-state conta-
mination and at the request of D.C.
elected officials and citizens, U.S.
EPA Region 3 assumed responsibil-
ity for the site. When Chevron has
completed its investigation and rec-
ommends a cleanup plan, EPA will
evaluate the plan and make it avail-
able for public comment.
16 States Pilot Test
Significant Operational
Compliance
Sixteen states, including at least one
from each Region, have volunteered
to take part in pilot testing EPA
OUST's draft revised significant
operational compliance (SOC) per-
formance measures. These revised
measures are the product of an
EPA/state work group that first met
in June 2002. Each state conducted a
minimum of 20 inspections of UST
facilities, checking for compliance
with regulations contained on two
matrices, one for release prevention
and the other for release detection.
The results from these pilot tests
have been analyzed and were pre-
sented at the annual UST/LUST
National Conference in March 2003.
OUST expects that when these mea-
sures become final, the agency will
be able to obtain from the states
more accurate and nationally con-
sistent data regarding compliance
with regulations designed to pre-
vent and quickly detect releases. For
more information, contact Jerry
Parker at (703) 603-7167. •
34
-------
March 2003 • LUSTLine Bulletin 43
. _ _ _^_ Your Vote is Needed!" ==
Should LU.SXUINfi Go Electronic?
Dear Readers, I
These are extraordinarily tough times for state UST/LUST programs. In such times, a publication such as
LUSTLine can play a critical role. Ever since our fjrst issue in 1985, we have worked hard to keep our readers
up-to-date on UST/LUST issues. We promise to Continue to stay on top of these issues and to address your
concerns. This is, after all, your publication. \
Like so many of your state and federal programs, we are not immune to budget concerns. In the interest of
cutting costs, EPA's Office of Underground Storage Tanks, the grantor of this publication, has asked us to look
into having LUSTLine go electronic. This would mean dropping the current print version and instead delivering
LUSTLine to you as a PDF file via e-mail. j
We know that LUSTLine is a valuable source of UST/LUST/state fund information for many readers and
that some of you save your issues for reference purposes. This is why we instituted the index. On one hand,
such archiving of a hard copy isn't so simple with a PDF file. On the other hand, an electronic LUSTLine would
enable you to read it via computer or print it out and read it at your leisure.
Please think about how you use this publication and consider the merits of receiving it electronically versus
by mail. Keep in mind, having it both ways is NOT an option—if s either electronic or in print.
We need to hear from all of you! Do you want to continue to receive this publication in print
by regular mail? Or would you prefer to receive it electronically? Please let us know by filling out the
form in this issue and sending it to us. Or, fill out the online version of the form, available at http://www.neiwpcc.
org/publication.html. Please note that we are also asking you to resubscribe. This will provide us with the infor-
mation needed to make a smooth transition to an electronic version, if that is what our readers prefer.
Thank you for your continued support oif LUSTLine.
Sincerely,
Ellen Frye, Editor
To continue to recievej LUSTLine, you must fully complete this form and
return it by mail OR send! the information to us electronically.
L.U.S.T.LINE
Name
Renewal Form
: Company/Agency,
Please choose one: !
Q I would like to have an electronic PDF version of LUSTLine sent to my e-mail address.
Q I would prefer to have a printed copy of LUSTLine mailed to me.
Please provide both addresses. E-mail Address J __
Mailing Address J
Q One-year subscription. $18.00. j
. i
Q Federal, state, or local government. Exempt from fee. (For home delivery, include request on agency letterhead.)
Please enclose a check or money order (drawn on a U.S. bank) made payable to NEIWPCC.
Send to: New England Interstate Water Pollution Control Commission
Boott Mills South, 100 Foot of John Street, Lowell, MA 01852-1124
Phone: (978) 323-7929 • Fax: (978) 323-7919 • lustline@neiwpcc.org • www.neiwpcc.org
Comments: '
35
-------
Should
g°
, ^ .. I-1 crToni\iir
totally LLCL i ix'-'i iiL
We need your vote.
i
See page 35
LU.ST.UNE
New England Interstate Water
Pollution Control Commission
Boott Mills South
100 Foot of John Street
Lowell, MA 01852-1124
LUSTLine T-Shirts
TWO HBW WACKY designs
BMMtrUKIUM attaint, lint Ma
TWO colon... red and black
TWO vorslons... long and short sleeve
Lento steva $17,00
Snort shave S13,00
Sbes: H,L,X, XXL
TO O»t«»S«idd«cknk»«r»ioioiiU.S.fcmfamIy)lo;
NErvvpcc
Bootl Mills South, 100 Fool of John Street
Lowell. MA OI8S2-I124
Td; (978) 323-7929 • Fax (978) 323-7919
------- |