New England Interstate
Water Pollution Control
Commission
                           Boott Mills South
                           10O Foot of John Street
                           Lowell, Massachusetts
                           01852-1124
Bulletin 45
October
2003
LUST.
A Report On Federal & State Programs To Control Leaking Underground Storage Tanks
Oh What a Taiioled Web!
Gasoline Oxygenates, Petroleum Distribution Networks, and
Detections in Ground water at LUST Sites

by Michael Martin
          inson

 o


          the past several years, gnnindwQter
        tinalifses at leaking underground storage
        tank (LUST) sites across th? U.S. have
detected \u> ether oxygenates, including MtBE,
from reletises of gasoline, fuel ail, itnti other petroleum
products. Interestingly, ninny of Ou&e oxygenate:
detections occur in state locations ivtierc the use of
ether-oxygenated or reformulated gasoline (RFC) lias
never been mandated for clean air requirements. South
Carolina, for example, has iwver required the use of
oxygenated gasoline, yt'f Ml HE has been found at 72
percent of till 1 LIST release sites and at 85 percent of
all corrective action releases (Shrader 2002).
   So what's jfomg OH here? Certainly the use of
oxygenates far all purposes, including octane boost,
RFC, or oxyfuei, all contribute to the "cross-
contamination" issue. However, the U.S. pipeline
distribution system and its operations also offer
plausible explanations for the widespread detections
and occurrences of gasoline oxygenates in LUST site
groundwatcr. Let's take a look at this tangled nnd
perplexing system.

Petroleum Shipments in the U.S.
Data for 2001 indicate that approximately 19.5
million barrels (819 million gallons) per day of
petroleum products are consumed in the U.S.
{Trench, 2001), Approximately two-thirds of the
petroleum shipped in  the U.S. travels via oil
pipelines. The balance of the distribution methods
includes barge trucking, railroad, and waterborne
shipments.
                     • continued on page 2
                                   U.S. REFINED PRODUCT PIPELINE NETWORK
                                                     Inside
                                         NEIWPCC Survey on Oxygenates at LUST Sites: Part 2
                                         To Drinking Water, with Love
                                         TRIAD Approach to Site Investigation
                                         EPA MtBE Treatment Profile Web Site
                                         Garbage In, Garbage Out
                                         Leak Detection in Europe
                                         One American's View of European Leak Detection
                                         Changes in Petroleum Marketing Scene
                                         Will the Real Quantitative SIR Method Stand Up?
                                         Ftex Piping Class Action Suits
                                         Legal Update
                                         Brownfields Beat
                                         Exploding Cell Phones: Fact or Fiction?
                                         UST and Energy Legislation Ongoing and Likely to Be Merged

-------
                •
• Tangled Web front ;>«
    The interregional flows of crude
and refined  petroleum  are  built
upon a  national  infrastructure of
pipelines designed to move oil and
refined products from the producing
regions to the consuming regions
(Pennwell, 2002).  According  to the
Association of Oil Pipe Lines' data
(AOPL,  March 2000),  in 1998, the
total network for crude and refined
petroleum products constituted over
200,500 miles of pipeline. Crude oil
and gathering  lines  account  for
114,000 miles; product lines make up
another 86,500 miles.

PADDS
Five regions,  referred to  as "Petro-
leum Administration for Defense Dis-
tricts"  (PADDs),  were delineated
during World War II to facilitate the
pipeline transmission of refined prod-
ucts. Up until World War II, domestic
distribution   relied   primarily  on
tanker shipments,  but the disruption
           LUSTLine
           Ellen Frye, Editor
          Ricki Pappo, Layout
     Marcel Moreau, Technical Advisor
    Patrscia Ellis, Ph.D., Technical Advisor
 Ronald Poltak, NEIWPCC Executive Director
     Lynn DePont, EPA Project Officer

 LUSTLine is a product of the New England
 Interstate Water Pollution Control Commis-
 sion (NEIWPCC). It is produced through a
  cooperative agreement (#CT825782-01-0)
     between NEIWPCC and the U.S.
    Environmental Protection Agency.
   LUSTLine is issued as a communication
      service for the Subtitle I RCRA
   Hazardous & Solid Waste Amendments
       rule promulgation process.
     LUSTLine is produced to promote
 information exchange on UST/LUST issues.
 The opinions and information stated herein
  are those of the authors and do not neces-
   sarily reflect the opinions of NEIWPCC.
     This publication may be copied.
     Please give credit to NEIWPCC.
   NEIWPCC was established by an Act of
   Congress in 1947 and remains the oldest
   agency in the Northeast United States
  concerned with coordination of the multi-
      media environmental activities
    of the states of Connecticut, Maine,
     Massachusetts, New Hampshire,
   New York, Rhode Island, and Vermont.

             NEIWPCC
  Boott Mills South, 100 Foot of John Street
        Lowell, MA 01852-IT24
       Telephone: (978) 323-7929
          Fax: (978) 323-7919
         Iustline@nciwpcc.org

    $,9 LUSTLine Is printed on Recycled Paper
of wartime shipments created a need
for the development of long-distance,
large-diameter pipelines. The logisti-
cal hubs of the PADDs serve as gate-
ways  for  regional  supplies  of
petroleum products.
    Allegro Energy Group's Decem-
ber 2001 document, Hint1 Pipeline*
Make the Oil Market Work-Their Net-
works,   Operation and  Regulation,
explains how the transmission of
petroleum products through the oil
market's   pipeline   infrastructure
helps balance the oil market—mov-
ing oil from  producing regions to
consuming regions. The regional
PADDs are summarized as follows.

• East Coast (PADD 1) has little or no
  crude oil production,  some refin-
  ing capacity, and  one of the high-
  est PADD  demands  for refined
  products.

• Midwest (PADD 2) is the source of
  approximately 10 percent  of the
  region's crude oil needs. The bal-
  ance of crude oil is obtained from
  outside  the region  (i.e., Canada,
  the  Gulf Coast  region, foreign
  crude imports). While the  Mid-
  west refining capacity meets most
  of the needs of the region, refined
  supplies are  also supplemented
                      with   shipments   from   other
                      regions,  most  notably the  Gulf
                      Coast.

                      Gulf Coast (PADD 3) provides the
                      U.S. with the largest regional pro-
                      duction  of  crude oil  (55%) and
                      refined products (47%). In terms of
                      interregional  PADD  trading  of
                      crude and refined petroleum, this
                      region provides 90 percent of the
                      crude and 80 percent of the refined
                      petroleum that is shared among all
                      PADDs.  Refined products are pri-
                      marily shipped to the East Coast; a
                      smaller portion is shipped to the
                      Midwest,

                      Rocky Mountain  (PADD 4) obtains
                      crude  oil from  local production
                      and supplements refinery inputs
                      with Canadian crude. Geographic
                      limitations   of   distances  and
                      topography create an inadequate
                      infrastructure that relies on interre-
                      gional trading to maintain supply
                      and demand balances, despite its
                      being the lowest petroleum-con-
                      suming PADD region in the U.S.

                      West Coast (PADD 5) is a region that
                      is mostly separate from the other
                      PADDs in the U.S. Alaska supplies
                      approximately 55 percent of the
                      crude oil inputs to the refineries;
 figure 1
 U.S. PRODUCERS OF MtBE
               (1999)
                        Motiva
                         2%
              Lyondell
                2%
                 I Coastal
          8P Amoco   2%
   MarathonX  2%
     2%.
Chevron
                 Cilgo
                 3%
      : ClicmExpii 2IKHI Chemical Prafilr MtKL

-------
                                                                                         •
   the balance of the oil production
   takes place in California, Almost
   all of the refining capacity is met
   from California  state  refineries
   that produce unique product spec-
   ifications.

    The manufacturers of MtBE and
other ether oxygenates are numer-
ous. (See Figure 1.) In 1999, MtBE
oxygenate supplies were produced
from at least 28  U.S.  suppliers
(ChemExpo Chemical Profile, 2000).
In 1998, approximately 25 percent of
the MtBE used in the "U.S. was from
imports (Oak Ridge National Labo-
ratories, 2000).
    The produced oxygenated gaso-
line products vary in  their oxy-
genates content, depending on clean
air mandates, use as octane boosters
in the gasoline, and other factors
associated with supply and demand.
MtBE is combined with refined gaso-
line per shipping  specifications for
shipment  to  ultimate  distribution
points.
    As oxygenated  gasoline enters
the refined products distribution net-
work of pipelines, it enters a system
encompassing various geographies,
numerous manufacturers, and gaso-
line products with varying MtBE and
other oxygenate contents. These all
contribute to  a national complex of
widespread MtBE distribution, both
intended and unintended. In addi-
tion, the refined petroleum product
passes through many of the more
than 2,500 pipeline terminals, start-
ing at the point of production, dur-
ing the pipeline transmission process
to its final distribution point (Penn
Well, 2002).
    In  Maine, a state  where oxy-
genate-containing  RFC use is not
required, there is considerable vari-
ability in MtBE content in gasoline.
The Maine Department of Environ-
mental Protection  (DEP) monitors
and  reports annually on levels  of
MtBE in shipments of gasoline  to
storage  terminals  that  have   a
throughput of more than 20,000 gal-
lons of gasoline per day in the state.
Terminals in  Maine reporting data
were owned by Gulf, Irving, Mobil,
Motiva, and Webber. Although the
goal for Maine has been to eliminate
MtBE from gasoline this has, so far,
been next to impossible. The average
level of MtBE in gasoline for 2002
was 2.44 percent  (by  volume)  in
  Figure 2
             TYPICAL SEQUENCE OF PETROLEUM
           PRODUCTS FLOW THROUGH A PIPELINE
                                 TRAHSMIX
                    COMPATIBLE
                    INTERFACES
                                 REGULAR VPREHHMl REGULAR
                                 GASOLINE I GASOLINE (GASOLINE
                   TRANSMIX (interface material
                   which must be reporcessed)
 gasoline, ranging from 0 to 14.53 per-
 cent MtBE (Maine DEP, 2003). Ship-
 ments with 11 percent or higher are
 most likely loads of RFC that have
 ended  up  in  Maine, one  way  or
 another, when they shouldn't have.

 Residual Refined Product
 Mixing During Distribution
 Another contribution to the nation-
 wide distribution of MtBE and other
 oxygenates, even to locations that do
 not use or need them, is the mixing
 of  residual  petroleum  products
 within the pipelines, terminal stor-
 age tanks, bulk shipments in barges,
 and final distribution to retail sites
 via tanker trucks. Even fuel oil sup-
 plies have been found to have signif-
 icant  MtBE and  other oxygenate
 concentrations due to residual vol-
 umes of oxygenated gasoline mixing
             with fuel oil shipments in pipelines,
             barges, and tanker trucks.
                 In pipeline transmission opera-
             tions, it is common practice to ship
             different  petroleum  products  or
             grades  of the  same  product  in
             sequence  through a pipeline, with
             each product or "batch" distinct from
             the preceding or following (Allegro,
             2001). Transmix interface materials
             are used to separate refined petro-
             leum products (e.g., fuel oil or diesel
             fuels separated with a transmix from
             gasoline shipments). (See Figure 2.)
             However,  the various grades of gaso-
             line products are not typically sepa-
             rated during pipeline transmission.
             The mixing of gasoline grades and
             their respective varying oxygenate
             concentrations can result in the inad-
             vertent  distribution  of   residual
                               I continued on page 4
Figure 3
STATE MtBE BANS
       March 2003
   | Ban Currently in Place
      (or date ban takes eflecl) -16 states
   | Legislative MtBE Bans Considered as of
      March 2Q03 - 3 states
   | Stales with Goal of Banning - Z states
                                                  Source; Britct" Bauman, API.

-------
              i  •
• Tangled Web from page 3

refined products containing ether
oxygenates such as MtBE.

Can We Predict Detections of
Ether Oxygenates at LUST
Sites?
Unintended  MtBE and oxygenate
distribution to states is plausible, if
not likely, given the complexities in
MtBE and other oxygenates produc-
tion, refined product composition,
pipeline and terminal network oper-
ations, and final distribution of the
refined products to the retail distrib-
utor. The refined product pipeline
network  works with the supply and
demand from the PADDs network to
cause an unpredictable distribution
of oxygetvatc-containing petroleum
products.
    When pipeline networks move
refined products from refining cen-
ters through states that do not require
the use of oxygenated gasoline, the
supplied refined products tapped by
states along the transmission process
are not necessarily formulated to
meet individual state bans on the use
of MtBE as a gasoline oxygenate. (See
map of state MtBE bans, Figure 3.)
Thus, LUST  releases  of gasoline
products anywhere in the U.S. have a
good chance of containing some level
of MtBE or other ether oxygenates
due to the complex nature of refined
petroleum distribution networks and
their operations. •



                  :,'-•  lilt. lilt,! llh>-
        '1'innfiila:
     il nl mnwvrriiiMin'lvli'lt.
    •    • • ''.I'd n' ( "i/ ri/
                 WWW j

          Illji i\ "Mtipl I
  •
  It'tiin li/.s/i.
  •
                          ..com
  •
References
Chi-mExpLi Chemical Profile: MTBE , 2000 December
    re\-JMon found at:
Djvis, Stacy C., 201X1. rraiisivrtatian liimfy Ootii Rank.
 tdition 20, preparvd for Office "( Transportation
 TecHnotu>;tE'v L S Department of Mnergy under
 Contract  No. LH--AC05-OOOR22725, Oak Ridge
 National l-aboralorios, October 2000
     Department ol  Environmental Protection
 Bureau ol Air Quality. February 21X11 2002 Maine
 l-uels Report. ttttptfftMBf&&HUne,gpDf4epfsirftttObft£l
 futtstpCUX&.piff
IVnnWell MAI'Sc-iirch, , 2MB. Petroleum diMnl'ulion
 iiiir,isirurlure  maps  and  data found  at:
    T, Arl , South Caroiina-DHEC, 2002. Personal
 communication.
Trench, Cheryl I, 2tNH.Hmc /"^v/iru-s Matt th? Oil
 Market W>irk - Thi'ir Nrfjrorts Operation ami Ursula
 tittii, A Memurandum Prepared for the Association
 ot Oil Pipe Lines and the American Petroleum Insti-
 tute's Pipeline Committee, Allegrit llnergy C.roup,
 December 2001.
 NEIWPCC  Survey Shows  How States Are

 Dealing with Oxygenates at LUST Sites:  Part  2

 by Lllcu Fri/e

    Jn the Insf issue wfLUSTLine (#44), I provided a }>nrthii report on the Nc~w England Interstate Water Pollu-
    tion Control Commission's (NEIWPCC's) 2003 "Stinvy of State Exfwrieuccs ii'itli MtBE ami Other O.ry-
    ^t'ndte Contamination at LUST Sites," conducted through a grant from the EPA Office of Underground
 Storage Tanks. The survey focused on the fa!!onith oxygenates and what they are finding, A full report along with a compilation of the state
 answers to each question if posted on NEIWPCC's Web site at www.neiwpcc.org. States can correct information
 at any time.
Oxygenate Analysis
Most of the states are sampling for,
analyzing for, and undertaking reme-
diation of MtBE  associated  with
petroleum  releases at  LUST sites,
even without standards. Far fewer,
however, are addressing the poten-
tial presence of other oxygenates at
these sites. All  but two states indi-
cated  they  require  sampling and
analysis for at least MtBE in ground-
water.  Only 34 states  responded that
they require sampling and analysis
for MtBE in soil.
    Forty-one states  indicated that
they request analysis for  MtBE in
ground water 80 to 100 percent of the
time; 28 states request this analysis in
soil 80 to 100 percent of the time. Sev-
eral states indicated they request this
analysis for the various other  oxy-
genates at an 80 to 100 percent fre-
quency. Another small  group  of
states indicated they request  this
analysis for the other oxygenates up
to 20 percent of the time.
    For both groundwater and  soil,
U.S. EPA SW-846 Method 8240/8260
(GC/MS) is the most commonly  used
analytical  method  for all  the  oxy-
genates, U.S. EPA SW-846 Method
8020/8021 (GC/P1D), a combination
of   8020/21  and
8240/60, and U.S.
EPA  Drinking
Water Method
524 (GC/MS) are
the  second,  third,
and fourth  most com-
monly used methods.
Site Assessment
Only four states said they investigate
MtBE or other oxygenate plumes dif-
ferently from BTEX plumes because
of the potential for "diving" plumes.
Seventeen  states said  they  do  so
rarely. Twenty-five states said they

-------
                                                                                         •
do not investigate oxygenate plumes
differently. It is clear, however, that
many states are aware of the diving
plume issue but that many deal with
this situation on a case-by-case, risk-
related basis.
    Only  11 states require three-
dimensional  characterization   of
plumes, and less than  half of the
states are taking extra steps to make
sure oxygenates are not migrating
beyond  standard monitoring para-
meters.  The typical types of steps
cited include:
• Multilevel sampling (nested wells,
  shorter-screened wells)

• Deeper wells

• Monitoring of  drinking water
  wells downgradient or in the area
  of an MtBE release
    In instances where  there is no
state standard for an oxygenate com-
pound,  factors states  use to deter-
mine  when to  test for them  are
primarily  proximity  to  drinking
water receptor and general vulnera-
bility analysis.
    Thirty-five   states  allow  for
dynamic work plans (i.e., field-deter-
mined based on site conditions) with
respect to well placement and screen
positions.
    Based on the responses  to this
survey, most states do not intend to
reopen closed sites to look for MtBE
or TBA  unless they have reason to
suspect a problem. Yet 32 states said
MtBE plumes are often or sometimes
longer than typical BTEX plumes.
The survey indicates  that MtBE in
groundwater is detected in gasoline
releases (averaged among the states)
60 percent of the time.
    Sixteen  states responded  that
oxygenate  levels exceed ground-
water  action  levels  somewhere
between 0 and 20 percent of the time;
five said 20 to 40 percent of the time,
six said 40 to 60 percent, six said 60 to
80, and nine said 80 to 100.
    Reported highest concentrations
of MtBE in the hot spot/core of a
plume ranged from 200 to 9,131,994
ppb; receptor concentrations ranged
from 6 to 28,000 ppb. (These concen-
trations are considerably higher than
those reported in 2000.)  Concentra-
tions in TBA hot spots ranged from
215 to 250,000 ppb; receptor concen-
trations ranged from 12 to 1,000 ppb.
Concentrations in TAME hot spots
ranged from 41 to 170,000 ppb; recep-
tor concentrations ranged from <5 to
1,000 ppb.

MtBE Plume Lengths
Eighteen states said they track MtBE
plume lengths from gasoline releases;
M
Average MtBE plume lengths
PLUME LENGTH (feet)
10-50
51-100
101-250
251-500
>50Q
Don't know
# OF STATES
0
3
12
10
2
8
Table z Maximum length any
MtBE plume observed in a state
PLUME LENGTH (feel)
50-250
250-500
500-1000
1000-5000
>9,000
Don't know
# OF STATES
1
3
2
26
1
16
 Table a Maximum length any MtBE
  plume observed in bedrock in a state
PLUME LENGTH (feet)
50-250
250-500
500-1000
1000-5000
Don't know
# OF STATES
0
3
3
11
30
15 said they sometimes do. Sixteen
states reported that these plumes are
often longer  than  typical  BTEX
plumes; 16 said they are sometimes
longer, three said rarely, and 12 did
not know. Tables 1 and 2 summarize
the average and maximum lengths of
MtBE plumes observed in individual
states. The MtBE  plume  in  East
Patchogue (Long Island), New York
was more than 9,000-feet long. Table
3 summarizes  the maximum length
of any  MtBE  plume observed in
bedrock. Only  17 states were able to
provide estimates for this question.

MtBE Impacts in
Drinking Water
Twenty-four  states  reported  that
their  drinking water program re-
quires routine analysis for MtBE in
drinking water. {This number was
the same in 2000.) Seven states did
not know the answer to this question.
The failure of  26 states to routinely
analyze for MtBE reflects back to the
lack of a federal MCL.
    Drinking  water  programs  in
Minnesota, New  Hampshire, New
Mexico, and Rhode Island began ana-
lyzing for MtBE in the mid to  late
1980s. Most of  the other states began
analysis  in the late  1990s to early
2000s.
    Sixteen states reported that their
LUST program  routinely  reviews
MtBE data from the drinking water
program. Based on these responses, it
appears that there  is a disconnect
between  more  than half of the state
LUST programs and drinking water
programs.
    Table 4 shows state estimates of
numbers of public and private drink-
ing water wells that have been conta-
minated  by  MtBE  at  any  level.
                • continued on ptift! 6
Table 4 State estimates of numbers of public and private
drinking water wells that have been contaminated by MtBE at any level.
f OF WELLS
1-10
11-50
51-100
101-500
> 500 (provide an estimate)
# OF STATES (PRIVATE)
9
3
6
g
NH: 30,000 -40,000
NY: 866
t OF STATES (PUBLIC)
12
7
2
5

                                                                                                        5

-------
 I Survey from pa$? 5
Thirteen states did not know or did
not have access to information as to
how many public and private drink-
ing water wells in their state  have
been contaminated by MtBE at any
level. (The survey  defined public
wells as groundwater supply systems
that serve more than 25 households.)

Oxygenate Remediation
Thirty-thivr stales soy that MtBE dri-
ves cleanup/investigative activities
less than 20 percent of the time or
never. Most of the states say there are
very few cases to none where MtBE is
the only concern. In most states, less
than 10 percent of the sites have situ-
ations where BTEX has been success-
fully remediated but MtBE remains.
Thirty-four states say  they   have
remediated sites with MtBE to clo-
sure. When asked  approximately
how many such sites have  been
closed,  however,  state  responses
indicate that there have been  rela-
tively few. More than half of the
states are not particularly aggressive
in non-aqueous phase liquid (NAPL)
recovery.
    States were asked to  rate  tech-
nologies they have used to remediate
oxygenates in soil and groundwater.
It was interesting to see the variety of
MtBE ratings for just about every
technology. The most widely  used
technologies for remediation of MtBE
in soil were soil vapor extraction
(SVE) and  biodcgradation, and prob-
ably excavation (added by the states
under  the "other"  category). As
would  be expected,  the ratings for
SVE leaned toward the "good" end of
the  spectrum,  while  those  for
biodegradation leaned toward the
"poor" end.
    The variety of ratings raises some
interesting and important questions.
While it is  no surprise that pump and
treat and monitori-d natural .itlunua-
tion got overall poor marks and that
point-of-use treatment and soil exca-
vation get overall high marks,  even
these technologies have some more
positive and more negative ratings.
Ratings for most of the other ground-
water technologies  seemed to go
every which way. Ratings for the
other oxygenates were too few  to be
significant.
    Does  this variation have  any-
thing to do  with geography and
soils? Does it have to do with limited
experience with a technology such
that whatever the degree of success
or failure, that is where current state
opinion rests? Does it have to do with
the manner in which the technology
has been applied—-competence ver-
sus incompetence in  the field? The
heart of  the  answer probably lies
with the  fact  that ail remediation is
site-specific. The state responses pro-
vide  us with  a snapshot of experi-
ences with technologies, but not  a
compelling sense of success trends.


   Twenty-three states say MtBE has
  had a noticeable impact on the cost
  of remediation in their state. Could
   this be an additional incentive to
   leave oxygenates that are not a
  direct threat to a receptor in place?
Remediation Cost Impacts
Twenty-three states say MtBE  has
had a noticeable impact on the cost of
remediation in their state. Could  this
be an additional incentive to leave
oxygenates that are not a direct threat
to a receptor in place? When we find
the relatively few states that seem to
be struggling hard with analyzing
for, discovering,  and cleaning up
oxygenates, we are left to wonder,
why them?
   States  provided  the  following
reasons why MtBE has driven up
remediation costs:

   •  Longer plumes

   •  Difficult to air strip

   •  Inefficiency of carbon
   •  Depends on environmental sen-
     sitivity of area
   •  More mobile, less biodegrad-
     able than BTEX

   •  More wells, more investigation

   »  Need for 3-D site characteriza-
     tion
   *  Higher installation and  O&M
     costs
   •  MtBE-only sites, so all costs are
     MtBE-related
   •  Additional monitoring

   •  Cost  of   corrective  action
     directly  related to size, length,
     width, depth of plume

   •  Receptor impacts

   •  High MtBE concentrations

   •  Carbon break-through

Your Thoughts?
Given the state responses to this sur-
vey,  at first glance it would appear
that the remediation of MtBE is not
something  states are  overly  con-
cerned with  right now. At second
glance, it would appear that we are
figuring out how to live with oxy-
genates in the environment.
   Most of the states use some form
of  risk-based   corrective  action
process and most account for MtBE
in this process. Most states have some
kind of no-further-action provision
that allows for inactivation of a file
with the possibility for a reopen in
the future if circumstances warrant.
Most states have some kind of long-
term  management  provision  to
ensure that sites with residual conta-
mination will  not be a threat to recep-
tors.  Most states do not intend to
reopen sites that were closed before
they had an MtBE standard.
   Although we have learned much
through this effort, there is still much
more we need to
know. If you
get a chance
to   examine
this survey on
NEFWPCCs
Web  site,   we
would  be  very
interested in know-
ing your thoughts
and observations. •
 Check Out NEIWPCC's
 Online Transformation
 Major improvements to the
 design, navigation, and content
 of NEIWPCC's Web cite should
 be readily apparent to visitors.
 The site's debut is scheduled for
 the end of November.
 Be our guest!
 www.neiwpcc.org

-------
                                                                                         L(J.*i
To  Drinking Water, with Love
         While we had several very
         positive responses to Patri-
         cia Ellis'  article, "A  Hot
Dog by Any Other Name Could by
Your Drinking Water,"  in the last
issue of  LUSTLiiie (#44), we  did
receive a somewhat outraged  re-
sponse from a reader on the drinking
water  side of  the equation, who
found three statements in the article
to be "false or misleading":
• "But most water suppliers analyze
  for a couple dozen contaminants
  at most." (page 2, second column,
  second line)
• "Generally  an  accounting gim-
  mick, such as 30-day average con-
  centration, is employed so that it
  can  be  claimed  that although
  detected above the limit, the con-
  centration did not exceed "permis-
  sible" levels and the water is safe
  to drink." (page 2, second column,
  second paragraph)

• "...we have no idea what contami-
  nants are really  in the water  we
  drink.  ..." (page 4, third column,
  third paragraph)

    Taken out of context,  these state-
ments appear quite brazen. As editor,
I  could have  tempered these state-
ments  to  avoid controversy.  But as
editor of a publication produced by
the  New  England Interstate  Water
Pollution Control Commission and
as an environmental professional, I,
too, make water (the "blue gold" of
our time)  my primary concern.
    This article points out, in  one
LUST  regulator's  voice, concerns
shared by many regulators. This is
why I ran the article on the cover; I
felt the message was important. After
all, as Ms. Ellis mentions in the arti-
cle:  "Although not limited to organic
compounds, the Chemical Abstract
Service (CAS) assigns unique regis-
tration numbers (known as CAS or
CASRN) to new chemicals at a rate of
about 4,000 per day!!!"
    This article is not meant to be an
attack on state and  federal drinking
water programs and water suppliers.
Though not stated  explicitly  in the
introduction,  the  article  refers
throughout to organic contaminants
and more specifically to the several
hundred that  occur in petroleum
products (especially gasoline). For
any federal or state drinking water
program to think it could stay on top
of this chemical assault would be like
tilting at windmills. The real solution,
reducing our use of these chemicals
and preventing them from getting
into  our environment,  is  equally
quixotic.
   Nevertheless, 1  will attempt  to
shed some light on the reader's con-
cerns and Ms. Ellis' responses,

A Couple Dozen
Contaminants?
In response to the reader's concerns,
Ms. Ellis  states that "under  Sate
Drinking Water Act (SDWA) require-
ments, analysis is required for less
than 100 contaminants. About one-
fifth  of these are organics, and  of
these, only four (BTEX) are found  in
petroleum products. Standard drink-
ing water analysis methods  (e.g.,
Method 502.2) have only a "couple
dozen" organic  contaminants  on
their respective lists of  target ana-
lytes, and of those that are on the list,
typically the only petroleum con-
stituents are BTEX!"
   As reported in NEIWPCCs 2003
oxygenate survey of the 50 states,
only 24 states indicated that  their
drinking water  program  requires
routine analysis for MtBE in drinking
water.  There is no federal standard
for MtBE.

Accounting Gimmick?
With regard to the accounting gim-
mick issue, the reader writes  that
"EPA establishes standards and the
means  for which compliance is deter-
mined. For  several contaminants,
averaging is used to determine com-
pliance. The guidelines for  doing  so
are set forth in state statutory  rule
requirements and are legally binding.
They are not 'accounting gimmicks'
by which water suppliers can manip-
ulate analytical results to their advan-
tage. If an operator does so, he/she
will   be   held
legally liable for
their actions."
    In  her  re-
sponse, Ms. Ellis
says  that  if  a
contaminant  is
detected  at any
concentration, it
is in fact present
in  the drinking water. "Whether or
not the concentration exceeds some
regulatory level (e.g., an MCL) is rele-
vant from a legal perspective," Ellis
says, "and although a water  pur-
veyor may not be legally required to
disclose the presence of the contami-
nant to the consumers, this does not
necessarily mean that the water is in
fact safe to drink."
    The author explains that all of
these regulatory limits  were estab-
lished  with the assumption that no
other contaminants were also pre-
sent. As discussed in the article, there
is essentially no information on the
effects of dilute mixtures (as would
be found  in drinking water). At best
the presumption is that effects are
additive, if they're even considered at
all. This  is not necessarily a good
assumption, and  it's certainly  not
erring on  the side of caution.

We Have NO idea?
The  drinking water reader writes:
"The word 'no' is an extreme exag-
geration especially in reference to the
writer's first sentence on page two—
'The drinking water supply systems
in  the United States are unquestion-
ably the  best  in the world.'" The
reader goes on  to say that  EPA
undergoes an extensive process to set
standards and ensure the safety of
our  drinking water.  "The   1996
amendments to the SDWA require
EPA to go through several steps to
determine, first,  whether setting a
standard is appropriate for a particu-
lar contaminant, and if so, what the
standard should be," says the reader.
    "Peer-reviewed science and data
support an intensive technological
evaluation, which includes many fac-

                • continued on page 8

-------
/ IK/'/ inr i'.'j//. .'<•'( 4'- » (',(,•
• Drinking Water from page 7

tors: occurrence in the environment;
human exposure and risks of adverse
health effects in the general popula-
tion and sensitive subpopulations;
analytical methods of detection; tech-
nical feasibility; and impacts of regu-
lation   on  water  systems,   the
economy, and public  health,"  the
reader explains,  "While unknown
contaminants  in  gasoline certainly
cannot be regulated, particularly if
analytical methods  are  not  estab-
lished, it is inappropriate to claim we
have 'no idea' what we are drinking."
    Ms. Ellis says the fact is we only
know whether a very few contami-
nants are actually in our drinking
water  at any given  time,  "Just
because a contaminant isn't analyzed
for does not mean that it is positively
not present in the sample. If there
isn't an  analytical result (with suffi-
cient  QA/QC documentation   to
demonstrate   the  accuracy of  the
analysis) that shows that the analyte
is not present in the sample, then it
could very well be present," she says.
    "But we're obviously not con-
cerned   with  every  conceivable
organic contaminant," says  Ellis,
"only those that might reasonably be
expected to be present, and this num-
ber is probably a couple orders of
magnitude  greater  than  the  21
organic chemicals on the SDWA list.
Granted, a lot of them may be at very
low concentrations, but we  have
absolutely  no idea what the health
effects may be of a dilute  solution
that contains a random subset of the
entire suite of organic chemicals pre-
sent in gasoline."
    Ellis explains that methods for
analysis of petroleum hydrocarbons
and other fuel constituents are well
established (for the most part). She
says that adequate calibration stan-
dards that contain the several hun-
dred  constituents of gasoline  (for
example) are lacking. Taking Method
502.2 as an example, she says there
are approximately 70 compounds on
the target analyte list.
    "Of these," says Ellis, "nearly 50
are chlorinated compounds not pre-
sent in gasoline. BTEX and about five
to six other hydrocarbons are on the
list (plus a couple of PAHs); most of
these compounds are typically pre-
sent in gasoline, but an analysis that
shows that these compounds are  not
8
present,  tells  us  nothing  about
whether or not any of the other cou-
ple hundred petroleum constituents
are present."
   The author's final point is that
the focus of the article is on doing a
better job analyzing for petroleum
constituents in soil and ground water
at LUST sites, which are typically
located a considerable distance from
drinking water sources.
   "The point is," says  Ellis, "if we
don't even know what contaminants
are present near their source (because
we aren't looking for them), how can
we claim to know that they aren't in
our  drinking water  (especially if
we're not  looking  for  them  there
either)? Our  goal is  to spur  more
comprehensive characterization  of
contaminants near  the  source—if
they're not there, then it's likely that
they're  not in the drinking water
either.
    "When there has been a known
release of gasoline at a LUST site,"
explains  Ellis,  "how  can  it  be
assumed that by analyzing for only
four organic compounds (i.e., BTEX)
found in gasoline (and a bunch  of
other organic compounds that are not
in gasoline, such as chlorinated sol-
vents) that, if those four compounds
are not present, we can be comfort-
able that other constituents in gaso-
line are also not present?" As noted
in the article,  "absence of proof is not
proof of absence," •
  NFPA Adopts Amendment on Static Discharge
  Warning Signs at Dispensers
  The National Fire Protection Association (NFPA) has adopted a Tentative Interim
  Amendment (TIA) to its Code for Motor Fuel Dispensing Facilities and Repair
  Garages (NFPA 3QA) that adds language to be carried on warning signs in the
  dispensing area of fueling facilities. The language can be found in the 2003 edi-
  tion of NFPA 30A. The amended code adds language that warns motorists of
  the danger of static discharge, explains what to do if a fire should occur, and
  prohibits children under licensed age from using the pump.
    Now that the 2003 addition of NFPA 30A is out, it can be adopted by the
  appropriate jurisdictional authorities. The parts of the country that follow the
  International Fire Code developed by the International Fire Code Institute (IFCI)
  should be aware that language identical to what is now included in NFPA 30A is
  being considered by IFCI in its code development hearings.
    Two separate sections were amended by this TIA. The first part (9.2.5.4)  is in
  the code itself and requires identical or equivalent wording. The second section
  (A.9.2.5.4) is in the Annex to NFPA 30A and is not obligatory. It was included to
  offer suggested text for signage that permits the owner/operator of the station
  to choose the best method of educating the public about how to refuel their
  vehicles.
    The warning signs are to be conspicuously posted in the dispensing area
  and must incorporate the following or equivalent wording  {new language is  in
  italics):

  Warning
  * It is unlawful and dangerous to dispense gasoline into tmapproved
    containers.
  • No smoking.
  * Stop motor,
  • No filling of portable containers in or on a motor vehicle.
  • Place container on ground before filling.
  • Discharge your static electricity before fueling by touching a metal
    surface away from the nozzle.
  • Do not re-enter your vehicle while gasoline is pumping.
  •Ifa fire starts, do not remove nozzle—back a way immediately.
  • Do not allow children under licensed age to use the pump.

-------
LUST Innovations, TRIAD,  and Computer
Imaging  Move LUST  Site Investigation into
the  21st  Century
/'!/1 tine Taylor


      Good site investigation is
      the most critical factor in
      successful  remediation.
Having enough data from a large
portion of the site reduces uncer-
tainty and allows targeting  of
cleanup  activities. The  ability
to adapt plans as new informa-
tion becomes available further
improves cleanups  and saves
money. These ideas are the dri-
ving force behind U.S. EPA's
push for the TRIAD approach to
site characterization.
   The TRIAD approach to manag-
ing the  inherent  uncertainties of
investigating  contaminated  sites
incorporates three key elements:

• Systematic  planning  (knowing
  what you are looking for and what
  you will do when you find it)

• Dynamic work plan  strategics
  (being able to change course/let
  new data guide actions)

• The use of real-time measure-
  ments to accelerate and improve
  the cleanup process (using in-field
  technologies to get  more data
  more rapidly)
   Most of the real-time/in-the-field
measurement techniques were first
given  widespread application at
LUST sites. With the vast number of
sites, it was simply too expensive to
do things the old fashioned way—to
"poke and hope," drilling a few mon-
itoring wells and hoping to hit the
"hot spot" of contamination.
   "EPA's Office of Underground
Storage Tanks (OUST) was very open
to trying new technologies and giv-
ing people flexibility—that helped
innovation," says Deana Crumbling,
an environmental  scientist  with
EPA's Technology Innovation Office
(T1O)  which  is promoting  TRIAD.
Also, the liability problems of getting
it  wrong—missing a  plume that
would continue to pollute drinking
water—pushed industry and regula-
tors to find better methods.
           The TRIAD
Systematic
  Project
 Planning
Uncertainty
Management
 Dynamic
Work Plan
Strategies
  Real-time Measurement Technologies
   The Bigger Picture
   Advantage
   Unlike numerous hazardous waste
   sites that may have a wide range of
   contaminants, most LUST sites bene-
   fit from having only a limited num-
   ber of very detcctible volatile organic
   compounds (VOCs).
       "With  LUST  sites, you know
   what the constituents are  you're
   looking for; you're looking for VOCs,
   for benzene, for MtBE," notes Tom
   Schruben, an environmental engineer
   with long experience in LUST issues.
       Since the costs of using direct
     sensing devices are so much lower,
      you can get that vertical data at
       multiple points across a site,
     yielding 3 much better "picture" of
       subsurface contaminants. By
        contrast, drilling individual
      monitoring wells is expensive: so
      fewer spots are tested, leaving
           greater uncertainty.
       A range of field-based and hand-
   held devices allows for quick detec-
   tion of such compounds throughout
    the site without the need to send
    samples to the lab. Examples of
    these direct/in situ sensing tools
    are membrane interface probes
    (MIPs) and laser induced fluo-
    rescence (LIF). These techniques
    are much faster than pulling up
    water samples from a well and
    shipping them to a laboratory for
    analysis. Instead of drilling mon-
    itoring wrells that pull up sam-
    ples every few feet, these devices
    can  take measurements every
    inch down to 100 feet, giving a
continuous vertical profile of contam-
ination,
    "This is important because we're
finding that the subsurface is very
heterogeneous," notes TIO's Crum-
bling, "You can have a sand layer, for
example, and within  8 inches what
you'll find will be very different. It's
easy to miss the contamination."
    Since the costs of using direct
sensing devices are so much lower,
you can get that vertical data at mul-
tiple points across a site, yielding a
much better "picture" of subsurface
contaminants. By contrast, drilling
individual monitoring wells is expen-
sive—so fewer spots are tested, leav-
ing greater uncertainty.
    "The differences in the amount of
data you can  get for similar costs is
amazing," says Schruben, adding,
"This really makes a big difference in
characterizing a site; you're likely to
find not just hot spots, but also the
source of the contamination and the
extent of contamination  across a
whole site, as well as how the conta-
mination relates to soil strata and
pathways."
    The TRIAD approach encourages
site  managers to adjust their investi-
gation as data  reveal contaminant
hotspots and pathways. This is now
much easier to do quickly, given in-
field techniques that provide instant
or real-time readings of contamina-
tion levels. (See a list of sources for
information describing MIPs,  LIF,
              • continued on f>agc If)

-------
             . .  .    .fart
• TRIAD from page 9
and  other  rapid  in-field analytical
techniques in the box below),

The Key Is Sufficient Data
One criticism of in-field technologies
is that they  may not be as sensitive as
traditional  laboratory analysis.  In
response to this criticism, Bruce Bau-
rnan, a geologist and member of the
American Petroleum Institute's (API)
Soil and Groundwater  Task Force,
points out that, for site characteriza-
tion, sensitivity  is not key. "For
VOCs, a lot of data points  showing
parts per million is much better than
a few showing parts per billion; they
will give you a much better chance of
finding the trouble spots." He also
says that recent tests of membrane
interface probes show much greater
sensitivity  in detecting MtBE, thus
offering the petroleum  industry an
important lower-cost tool for dealing
with this problem.
   Dennis   Rounds,   Director   of
South Dakota's Petroleum Release
Compensation  Fund   (PRCF),   is
embracing  TRIAD and  the in-field
techniques. Like most states, South
Dakota  is striving to get more work
done with  the available funds and
sees these emerging technologies as a
means to that end. Rounds believes
the TRIAD approach and  the field
techniques on which it is based—
push probes, L1F, and  MIPs—will
fundamentally  change his  state's
cleanup program.
   In addition to providing more
data, these techniques are able to dis-
tinguish between gasoline, diesel, or
other contaminants, greatly  aiding
remediation  planning. "I  don't
understand why this approach and
these technologies aren't  the stan-
dard; you can do it in one day and it
costs less," says Rounds.

The Beauty of
Computer Imaging
Rounds and his staff are also taking
advantage of one of the most cutting-
edge innovations in the field: com-
puter  imaging  of the rich data
provided by in-field probes. Using
state-of-the-art software, a 3-D image
of a site is created showing the sub-
surface contaminants. "It's like med-
ical imaging, an MRI for UST sites,"
says Rounds, "you can  actually  see

10
where the hot spots are."
    South  Dakota began  working
with Dakota Technologies, innova-
tors in using LIF to get more accurate


  "It's like medical imaging, an MRI
  for UST sites, you can actually see
      where the hot spots are."
         Dennis Rounds, South Dakota PRCF
information on  the extent of light
non-aqueous phase liquid (LNAPL)
at sites with petroleum contaminants.
Dakota partnered  with Columbia
Technologies, a Maryland-based firm
that was an early leader in membrane
interface probes and vertical profil-
ing and has more recently developed
imaging software to present subsur-
face data as pictures rather than com-
plex data  graphs.  (See Figure 1 on
page 11.)  The saying "a picture is
worth a thousand words" holds true
here. No one needs a technical back-
ground to  understand the data.
    "We like that it is a representa-
tion of what's actually there, not a
'guesstimate'/' says Rounds, adding,
"Another advantage of this comput-
erized data presentation is that it can
be shared on a Web site, so decision-
makers in different parts of the state
can all see it on screen at the same
time. You can even rotate the 3-D
image, making it immediately obvi-
ous as to what you need to focus on."
    Deana  Crumbling  agrees  that
sharing of data over the Web is a big
advantage. "You can do a joint call
and everyone, no matter where in the
country, can see it. There's no transfer
of paperwork that can take weeks."
She notes that EPA Region V is devel-
oping software that will incorporate
some Graphic Information  System
(GIS) features that will also be helpful
to state UST/ LUST programs. API's
Bauman likes to think of all  these
innovations   as  "Better   Living
Through Computerization."

A No-Brainer
The  TRIAD approach, in combina-
tion with the innovations  in-field
analysis that make it possible and
cutting-edge  computer  imaging,
offers much to UST/ LUST program
managers and their colleagues  in
industry. As Bruce Bauman puts it,
"To me it's a no-brainer that this is
the trend of  the future.  More mea-
surements mean a better picture of
what's going on;  it's a huge improve-
ment." •

  /id/i' l'tn/1,11' i> tin ,-iirii'c
        'V worked as .   •   ltan\ far
 E/V<       i '  Underground St.-
       'it'll tit kaylorjune@aol.com.
   More Information about TRIAD, Direct
   Sensing Technologies, and Computer
   Visualization
       http://www.epa .gav/tia/triad/index.Mm Explains the concept of the TRIAD approach.
       http://www.clu-in.org/ERA'S Technology Innovation Program supports this site on
   hazardous waste Cleanup INformation (CLU-IN), which includes TRIAD (hnpj/www.clu-
   in.orgfiriad/) and has links to many related remediation topics.
       http://www.calumbiadata.com/mip/mip.cfni This site gives Information on mem-
   brane interface probes (MIPs)), which provide detailed mapping of volatile organic com-
   pounds, including petroleum, for vertical profiling of subsurface contaminants.
       ltttp://www.dakotatechnologies.com/LIFanimatioi.him Describes the laser-induced
   fluorescence (LIF) technology used to provide data points for the graphic accompanying
   this article.
       http://smartdata.columbiadata.com To see computerized graphics of a petroleum ter-
   minal and a manufacturing facility, enter the username and password as "demo," then click
   on the PID Images and the Movies sections to see data points converted to a highly accurate
   color 3-D representation of the contaminants. Such images can be put on the Web in near
   real time, aiding investigation and remediation decisions.

-------
                                                                                    -
 Figure 1
       Leaks and Spills Clearly Defined and Pictured: Diesel and Gasoline
                 from Aboveground Storage Tanks (ASTs)
         ASTs
   Pump
   Islam!
                        Computer graphic by Columbia Technologies/SmartData Solutions

  In this depiction ot LWAPL at a Soutli Dakota bulk fueling and dispensing service sta-
  tion, trie lighter gray areas are the more intense hot spots. (In full color these would
  be yellow, orange and red.) A dozen vertical probes using laser-induced fluorescence
  (LIP) accurately measured the source areas and the depth and extent of the LNAPL
  plume, Previously, the state had sunk 35 monitoring wells covering a two-block long
  area, but not gotten a clear picture of the vertical and horizontal distribution of the
  LNAPL at the site. In fact, the state has concerns that improperly screened wells
  drilled through a perched watertable may be conduits for LNAPL to move deeper into
  the sandstone formation beneath the site. •
Alaska Implementing TRIAD
Approach at LUST Sites in
Fairbanks
On October 9, representatives of the
Alaska Department of Environmental
Conservation (ADEC), the Argonne
National Laboratory (ANL), EPA's Tech-
nology Innovation Program (TIP),
Region 10, and the EPA Office of
Underground Storage Tanks conducted
a kick-off conference call to discuss the
use of the TIP "TRIAD Approach" on
contaminated UST sites in Fairbanks.
(See related article on page 9.)
  ADEC is currently implementing an
areawide approach to address clusters
of LUSTs and other contaminated sites
(consistent with OSWER's Land Revi-
talization and One Cleanup initiatives)
in an area of Fairbanks designated as
the Fairbanks Areawide Industrial
Reclamation (FAIR) project. Given
Alaska's limited  resources to continue
with this project and the pending sun-
set of its state cleanup fund, TIP will
work with ANL and other public/private
partners to assist the state in develop-
ing the best process for collecting
information to meet project and pro-
gram goals in a timely and cost-effec-
tive manner. As the FAIR project
moves forward, Alaska intends to apply
this streamlined areawide management
and risk-based approach to other clus-
ters of sites in the state. •
EPA's  Treatment  Profile  Web  Site Provides
Real-World Data  about the Treatment of MtBE
and Other Fuel  Oxygenates
by John Qitnnder and Michael Bcrman


      Data  from  the U.S.  EPA's
      Office of Underground Stor-
      age Tanks indicate that out of
436,500 confirmed releases of gaso-
line into the environment, 139,500
still require  cleanup (EPA, 2003).
Many of these sites are contaminated
with fuel hydrocarbons, most often
benzene, ethylbenzene, toluene, and
xylene (BTEX) compounds, as well
as the common  fuel  oxygenate
methyl tertiary-butyl ether (MtBE)
and other fuel oxygenates, such as
ethyl tertiary-butyl ether (EtBE), ter-
tiary-amyl methyl ether (TAME),
diisopropyl ether (D1PE), tertiary-
butyl  alcohol (TBA), ethanol, and
methanol.
   There are many challenges asso-
ciated with the characterization and
remediation  of sites contaminated
with MtBE and other oxygenates. For
example, fuel oxygenates are gener-
ally more soluble, less likely to parti-
tion to organic matter  in soil, and
slower to biodegrade than other cont-
aminants in fuel, such as BTEX con-
stituents. These properties  lead to
larger and more widespread ground-
 water plumes and challenges with
 employing certain treatment tech-
 nologies. These factors also have an
 impact on our ability to characterize
 the nature and extent of contamina-
 tion involving fuel oxygenates.
    Technologies    available    tu
 cleanup MtBE and other oxygenates
 in soil, ground water, and drinking
 water include: air sparging, bioreme-
 diation (in situ and ex situ), in situ
 chemical  oxidation,  groundwater
 pump and treat, multiphase extrac-
 tion  (MPE), soil  vapor  extraction
 (SVE), phytoremediation, and ther-
               • continued on page 12

                             11

-------
                •
• Treatment Profile Web Site
 from page 11


mal treatment. Until recently, limited
information  was  available  about
technologies used to address sites
contaminated  with MtBE and other
oxygenates. To  address this need,
EPA has worked to make available a
great deal of information  about the
characterization and treatment  of
sites contaminated with MtBE and
other oxygenates, including the pub-
lication  of  fact  sheets,  technical
reports, and other documents.

Real-World Data
Since 2000,  EPA's Office of Solid
Waste and Emergency Response has
compiled information about cleanup
sites where  the  treatment of MtBE
and other oxygenates has taken place
to provide information to regulators,
remediation consultants, technology
vendors, and other  interested parties.
    In April 2002, EPA published an
online database of this information  as
MtBE Treatment Profiles, at the Web
si te   http:/icltt in, orgtproduci$/m the/.
This Web site is intended to be used
as a starting point for identifying
technologies that have been used for
the treatment of MtBE and other oxy-
genates, as well as for identifying
other environmental professionals,
technology providers, or remediation
consultants   that  may  serve   as
resources.
    The  MtBE Treatment Profiles
Web site is a searchable database  of
project profiles for  sites treating oxy-
genates in drinking water, ground-
water, or soil. The Web site contains
treatment profiles that include back-
ground  information, cost and perfor-
mance   information,  point(s)   of
contact,  and references. As of Septem-
ber 2003, the site contained 340 treat-
ment profiles.  In  addition to the
search capability, the site allows users
lo submit new prolik-s  and  updolr
existing  profiles. New  or updated
profiles are submitted frequently.
    The site provides a search engine
that allows a user to search by conta-
minant,  media, technology, scale, sta-
tus, state, site name, or keyword.
Alternately, a user  may browse a list
of all the profiles in the database. The
site is visited more than 600 times per
month.
    In addition to  serving as a tool
for identifying  existing and  com-
pleted cleanup projects, the Web site
provides a portal to other environ-
mental professionals and technology
providers. Each profile   provides
information on  point(s) of contact,
allowing more detailed information
to be acquired directly from individ-
uals involved  with the site.  EPA
encourages regulators, remediation
consultants, and technology vendors
to add new treatment profiles to the
site. In summer 2003, EPA updated
much of the information in the data-
base, including cost information for
57 projects.
   The site provides a search engine
    that allows a user to search by
   contaminant, media, technology,
   scale, status, state, site name, or
   keyword. Alternately, a user may
    browse a list of all the profiles
          in the database.
    To prepare site information, EPA
obtained data from  site managers,
regulatory officials, and technology
providers, as well as  from published
reports, conference proceedings, and
other available reference materials.
Consequently,  each profile  has a
varying level of detail, depending on
the data and information  that are
available. In addition, some  of the
profiles include active links to more
detailed case studies, which present
further, in-depth information.
    The data are provided by project
managers and technology  vendors,
and no additional testing of technolo-
gies or independent review is per-
formed. The profiles contained in the
database do  not represent all projects
treating fuel oxygenates. The perfor-
mance and cost data included for the
projects  are, provided  as general
information and should not be used
as a sole basis to select future MtBE
remediation projects or to  compare
technologies. EPA does not guaran-
tee the accuracy or completeness of
this data.

Data from 340  Projects
The following discussion provides a
summary of the MtBE  Treatment
Profiles, including a brief overview of
the treatment technologies used, con-
taminants treated, and available per-
formance  and  cost  data.   More
information for each of the projects is
available at the MtBE Treatment Pro-
files Web site.

• Treatment Technologies
The 340 projects employed all of the
technologies commonly used to treat
MtBE and other oxygenates. The pro-
jects include those using active treat-
ment  technologies.  The database
does not include projects that primar-
ily  employ  non-active treatment
remedies, such as natural attenuation
or institutional controls. In addition,
the database generally does  not
include projects where the only reme-
dial  technology employed  was a
removal technology, such as excava-
tion or product recovery.
   Table I summarizes technologies
in the database employed at 104 com-
pleted and 233 ongoing projects. The
technologies are: air sparging, pump
and treat, and in situ bioremediation
(more frequently used to remediate
groundwater and soil contaminated
with MtBE); in situ chemical oxida-
tion, MPE, SVE, drinking water treat-
ment, phytoremediation, and thermal
desorption. Most (83%)  of the  340
projects are full-scale, 14 percent are
pilot-scale, and 3 percent are bench-
scale.
Tame 1 Breakdown of
Technologies lor 340 Projects
TREATMENT
TECHNOLGIES
Air Sparging
Pump and Treat
Bioremediation
Soil Vapor Extraction
In Situ Chemical Oxidation
Drinking Water Treatment
Multiphase Extraction
Phytoremediatian
Thermal Desorption
# OF PROJECTS
124 (36%)
88 (26%)
79 (23%)
43 (13%)
25 (7%)
15 (5%)
14(4%)
8 (2%)
1 (<1%)
Note: May be more than one technology per project
• Contaminants Treated
While the profiles primarily focus on
projects treating MtBE, a number of
these projects also provide informa-
tion  about  other  contaminants.
Where this information is provided.

-------


it is included in the treatment pro-
files. As shown on Table 2, data were
available for several contaminants
including MtBE, TEA, TAME, DIPE,
ethanol,  and  BTEX.  All  projects
reported MtBE as a contaminant with
many  (71%) also reporting BTEX.
TBA  (11%), TAME  (2%),  ethanol
(1%), and DIPE (<1%) were reported
as being present for only a small per-
centage of the sites. For the 340 pro-
jects,  275  (81%)  provided  MtBE
concentrations (either initial or final
concentrations, or both) and 84 (25%)
provided an MtBE  cleanup goal,
Reported  MtBE  treatment  goals
ranged from 5 /'g/L to more than 10
mg/L, with a median treatment goal
• Treatment  Technology Perfor-
mance
Of the 340 projects, 104 (30%) were
complete as of the July 2003 update.
Technology performance information
is included for these projects in the
treatment profiles, primarily in terms
of changes in concentration of MtBE
in the ground water. Soil data are sel-
dom reported and not included in the
database. MtBE concentrations before
and /or after treatment are available
for 275 projects in the database. In
general,  the  highest concentration
reported prior to treatment and the
highest concentration after treatment
was completed (shown as "final con-
centration") is provided.

• Treatment Technology Cost
Of the 340 projects, 162 provide some
form of cost data. Most of these pro-
jects (151 of the 162 projects with cost
data)  employed   bioremediation,
pump and  treat, SVE, or air sparging
alone, or one of three combinations of
technologies:  air sparging and SVE;
air sparging,  SVE, and pump and
treat; or SVE  and pump and  treat.
Both total costs and unit costs based
on  area treated  for the  projects
employing pump  and  treat  either
alone or in combination with other
technologies were upwards of twice
that of projects employing only in situ
technologies.
    It should be noted that most of
the costs included in this analysis are
for ongoing projects (129 of 162 pro-
jects). Therefore, the total costs for
many of  the projects may eventually
be greater than what  is currently
reported in the treatment profiles. In
Contaminant Distribution for 340 Projects
CONTAMINANT
TYPE
PROJECTS REPORTING
CONTAMINANT
PROJECTS PROVIDING
CONCENTRATION DATA
PROJECTS PROVIDING
CLEANUP GOALS
OXYGENATES
MtBE
TBA
TAME
Ethanol
DIPE
340 (100%)
36 (11%)
6 (2%)
3 (1%)
1 (<1%)
275 (81%)
29 (8%)
0
0
0
84 (25%)
4 (1%)
0
0
0
OTHER CONTAMINANTS
BTEX
243 (71%)
190 (56%)
13 (4%)
general,  there is limited cost data
available. However, an analysis of
this  data  suggests  that  systems
designed to treat  other fuel oxy-
genates in addition to MtBE may be
more costly than those that treat only
MtBE.

Other Resources for
MtBE Treatment
Additional  EPA  resources  about
MtBE and other oxygenates are listed
below.

• EPA's MtBE Web Page - A list of
Frequently  Asked  Questions pro-
vides basic background information
on MtBE, as well as links  to other
Web sites.  Available at http:llwww.
epn.goi'/mtbe

• EPA's Office of Underground Stor-
age Tanks MtBE Web Page - Gen-
eral  information about MtBE and
USTs. Available at http:ilwww.epa.gavl
* Clu-ln - Information about innova-
tive treatment and site characteriza-
tion technologies. Serves as a forum
for waste remediation stakeholders,
Available on line at http:l Iwuip.duin,
Bf%

• TechDirect -  Hosted  by EPA's
Office of Superfund Remediation and
Technology Innovation,  a monthly e-
mail that highlights new publications
and events of interest to site remedia-
tion and site assessment profession-
als. Sign  up online  at http://itww.
• Technology News and Trends - A
newsletter about soil, sediment, and
groundwater  characterization  and
remediation technologies. Available
on line at http://www.chtin.org/prod-
itcts/neivsltrs/tnaiicit/mU











               ider.john












              ' bt-rrn,)i.

References
Blue Ribbon Panel on Oxygenates in Gasoline, 1999,
 Achieving Clean Air and Clean Water: The Report
 of the Blue Ribbon Panel on Oxygenates in Gaso-
 line; EPA 42Q-R-99-021. ;if
F.PA, 2003, UST Program Facts: Office of Solid Waste
 and Emergency  Response,  hflp-.tlww.cpll.gov!
 oust/f'Xhstusipragrtimfacte.piif,
EPA, 2003a, MlBE Treatment Profiles: hll^/lcluin.^rif/
 prottiictsfmltf.
EPA, 2001,  Remediation Technology Cost Com-
 pendium - Year 2000 (EPA-542-R-01-009): Office of
 Solid Waste and Emergency Response. September.
 htty://cl»-in.orx/dcni'nloail/rcmal/542rai009.ptlf.
U.S. General Accounting Office (GAO), 2002, Testi-
 mony Before the Subcommittee on Environment
 and Hazardous Materials, Committee on Energy
 
-------
 !  i o o o o c c   |x|   ! 't o o o o c c    ! M T
 UI  11\ UIIU»-  III,  UI  11\ UI  IU'-  U •_» I

 The Trouble with Modeling the Behavior of
 MtBE in the Environment
       Often abbreviated as GIGO, this is a famous computer axiom meaning that if invalid data is entered into a system, the
        resulting output will also be invalid—garbage in, garbage out. Although originally applied to computer software, the
        axiom holds true for all systems, including, for example, decision-making systems. As a state regulator, I have become an
 anti-modeler. Wliile some of the data input into groundwnter-modeling programs is not garbage per se, the choice of data will, in
 many cases, result in output that may not be protective of human health, safety, or the environment. Allow me to unleash my pent-
 up feelings on this theme, emphasizing some of the problems I've seen in modeling the behavior of MtBE in the environment, pri-
 marily with various fate and transport models.
Plume Characterization
One of the first tasks that needs doing
before you try to run any kind of
modeling program for a site is col-
lecting sufficient data to characterize
the existing plume. This means that
you must collect data that defines the
extent of the plume in three dimen-
sions, the contaminant concentration
at various locations in the plume, and
the hydraulic characteristics of the
aquifer.
    Many of the models that I have
reviewed have had poorly character-
ized plumes. Areas of maximum con-
centration have not been located, and
the data from  the "hottest" area
observed might not be the "hottest"
area that is  input into the model.
Likewise, just because concentrations
decrease in wells as you move down-
gradient does not mean that  you
have reached the end of the BTEX or
MtBE plume. If you  haven't looked
deeper, you may have missed a div-
ing plume.
    In East Patchogue, Long Island,
the beginning of  the MtBE  plume
was beyond the farthest extent of the
toluene, ethylbenzene,  and xylene
plumes, and beyond the "apparent"
benzene plume as defined by moni-
toring wells screened at the top of the
water table. Because investigation of
the East Patchogue plume began near
the middle of the plume, with an
impacted domestic well, the investi-
gation proceeded both  up-  and
downgradient and included multi-
level sampling. The MtBE plume and
distal portion of the benzene plume
would both have been missed if mul-
tilevel sampling had not occurred
(Weaver and Wilson, 2002).

Concentration Values
Because of spatial and temporal vari-
ability, true current concentration val-
ues may not be well defined. If you
input concentrations from monitoring
wells into a model, you will likely
underestimate the concentration pre-
sent in the aquifer.  Measured concen-
trations in monitoring wells with
relatively long screens are always less
than the  true maximum, because
groundwater samples  withdrawn
from each well are a composite of the
concentrations  over   the   entire
screened interval (White, 2002a).
    To further explore the effects of
well-screen length and placement on
average borehole concentrations, visit
ORD's OnSite  Calculator  at  http:/
Iwww.ept~t.gov/athensllearn2modellpart-
two/onsite/abc.htm. Table 1 shows the
effects of varying the screen place-
ment and length with data from East
Patchogue, New  York, using  the
Average Borehole Calculator.
    This underestimation of maxi-
mum  concentrations also results in
an underestimation of the length of
the plume. Modeling by Weaver and
Small  (2002) showed that decreasing
the initial concentration of MtBE in a
release would proportionally reduce
peak and average concentrations in
the groundwater  plume but  that
plume lengths would not be propor-
tionally reduced.

Mass and Duration
of a Release
Initial contaminant concentrations,
duration of a release, and total vol-
ume of a  release are seldom well
known, unless the release occurred as
a catastrophic failure. Contamination
is often discovered years  after a

-------
                                                                                            •
Borehole concentrations in ppb for various combinations of screen elevation and length.
Data from East Patchogue New York, calculated using the Average Borehole Calculator.
SCREEN
LENGTH
5 feet
10 feet
20 feet
TOP OF
SCREEN
AT 20 FT.
0
0
119.5
TOP OF
SCREEN
AT 30 FT.
279.7
209.8
760.7
TOP OF
SCREEN
AT 40 FT,
1054.0
1805.0
3524.0
TOP OF
SCREEN
AT 50 FT.
5677.0
4849.0
3028.0
TOP OF
SCREEN
AT 60 FT.
877.1
1054.0
1130.0
TOP OF
SCREEN
AT 70 FT.
1155.0
994.7
743.5
TOP OF
SCREEN
AT 80 FT,
560.7
475.5
316.3
TOP OF
SCREEN
AT 90 FT.
79.99
61.83
36.48
release   occurs.  Certain   models
require  you  to choose whether a
release  was  instantaneous  or had
occurred over a length of time. (See
the section below on Definition of
"Source."} This can  be especially
important in the case of MtBE, as the
lack of sorption to soils will  allow a
one-time release  of MtBE to move
more as a slug through groundwater
than the BTEX constituents, some-
times resulting in a detached plume,
whereas a one-time release of BTEX
compounds results in a plume that
continues to  be  connected to the
source area due to sorption and retar-
dation.  (See the sections below on
Darcy Velocity and Sorption  and
Retardation.)

Definition of "Source"
The definition of "source" varies from
model to model and person to person.
One report that I reviewed, using Bio-
plume II, showed that all contami-
nants would be reduced  to below
MCLs within six months. I thought
this slightly odd, since  the site had
already  been  around for  about 10
years, and, while showing decreasing
levels in a few of the wells, it looked
like it might take another 10 or 20
years to achieve MCLs at the rate at
which it was proceeding.
    The consultant was obviously
proud of his modeling  effort, but  I
had my doubts. As a matter of fact,
over the next few quarters, concen-
trations increased in most wells as
water levels fluctuated. We  went
back and reviewed  some  of the
model inputs and assumptions.
    The  consultant  said  that the
model assumed that the source had
been removed. I asked him what the
"source" was for the site. He said that
the old tanks had been removed, and
the new tanks were located in a dif-
ferent area onsite and weren't leak-
ing. It didn't seem to occur to him
that the area of the  old tank  field,
where some of the wells still periodi-
cally had free product, could be con-
sidered to be a continuing source, so
he really had  a  plume that should
more correctly have been modeled as
a continuing source, rather than as an
instantaneous release.
    Many of the models that I have
      reviewed have had poorly
    characterized plumes. Areas of
   maximum concentration have not
  been located, and the data from the
  "hottest" area observed might not tie
   the "hottest" area that is out there
       for input into the model.
Sorption and Retardation
The transport of organic chemicals in
groundwater is typically slower (i.e.,
"retarded") relative to the velocity of
the groundwater. Retardation (R),
which is both chemical specific and
organic-content  specific,  is due  to
sorption and is defined as:

        R=l +(pb/n)-Kd
    where:
    pb is  the bulk density of the
aquifer material (soil)
    n is the porosity
    Kj is the chemical-specific soil-
adsorption coefficient (or distribution
coefficient), which is defined as:
              = focK
    where:
    Foc is fraction organic carbon
    Kuc is the organic carbon parti-
tion coefficient (chemical specific)

    The higher the Kd, the greater the
retardation. The retardation factor for
benzene is different than that of MtBE
because the organic carbon partition
coefficients are different. Retardation
also varies with soil types, based on
the amount of organic carbon present.
If you are going to model fate and
transport of chemicals in the environ-
ment, you need to have a model that
will account for these differences. You
wouldn't expect Benzo(a)pyrene  to
behave the same as BTEX, so why
should you expect MtBE to behave
the same as BTEX? Fraction organic
carbon is sometimes measured at spe-
cific sites, but many modelers are con-
tent to use default values. Particularly
when trying to model MtBE,  you
need to use a model that accounts for
the extremely low retardation.

Dilution and Dispersion
Dispersion coefficients (i.e., longitu-
dinal, transverse, and vertical) are
seldom measured directly. The role
of dilution and dispersion at a partic-
ular site are usually  extrapolated
from  available  literature, or deter-
mined by calibrating a fate and trans-
port model to field data by adjusting
the modeled value for dispersion to
best fit the existing monitoring data.
    Most transport models assume a
uniform flow direction and velocity.
As a result, spreading of the plume
due to variations in flow direction
and velocity are attributed to disper-
sion and not to the uncertainty  in
monitoring data describing the direc-
tion of groundwater flow.
    If there is little variation in flow
direction, modeled  MtBE plumes
usually appear long and skinny. If
flow direction is variable, apparent
dispersion may be deceiving. What
will appear to be lateral dispersion
may actually be longitudinal disper-
sion occurring in different directions.
Dispersivity is almost  never mea-
sured on a site-specific basis, despite
its importance in determining model
outcomes.
                • continued on pa$f 16

                             ~\5

-------
                • •
• GlGO/iwi page 75

Hydraulic
Conductivity,
Homogeneity,
and Isotropy
Hydraulic conductiv-
ity is often measured at a
specific site; although spa-
tial variability can give a wide
range in the values of the mea-
surement.  If the  measured
values are different at differ-
ent places in the aquifer, the
aquifer is heterogeneous (it
would be homogeneous  if
the measured values are the
same regardless of where they are
measured in the aquifer).
    If  the measured  values  differ,
depending on the direction in which
they are measured, the aquifer  is
anisotropic (isotropic means the val-
ues are the same regardless  of the
direction  in which  they  are  mea-
sured). Despite the fact that very few
aquifers  are actually homogeneous
and isotropic, most models require
the assumption that they are and that
there are no preferential flowpaths.
    Aquifer tests provide essential
information on aquifer  response  to
pumping stress; however, if  wells
with long screens are used, the tests
may yield erroneous information. It is
essential that the pumping well and
the monitoring wells tap the same
hydrostratigraphic unit. Aquifer tests
provide only a gross estimate of aver-
age aquifer permeability and yield.
While  this may  be the  objective  of
water  supply investigations, it  is
essentially useless for determining
contaminant travel time. Because con-
taminants migrate along preferential
flow paths that generally have higher
than average permeability, the "true"
transport velocity of contaminants
may be significantly underestimated
and contaminants  may  arrive  at
potential receptors much earlier than
expected (White, 2002a),
    In  aquifers with significant varia-
tion in hydraulic conductivity with
depth, it is important to ensure that
the  regions  with  the  highest
hydraulic conductivity are sampled.
If significant variation is suspected,
all of the monitoring wells should  be
tested  for their specific capacity. The
wells with the highest specific capac-
ity should have the greatest weight in

16
the interpretation of the dimensions
of the plume. Data from wells with
low specific capacity should be inter-
preted with caution (Wilson, 2003),

Darcy Velocity
We use Darcy's law to estimate the
average linear velocity of flowing
groundwater.  We must  remember
that the velocity calculated represents
the average linear velocity of the cen-
ter of mass  of a contaminant slug.
Due to the paths traveled by ground-
water through any geologic medium,
some  water will  be moving faster
than this average and some slower.
From  a  practical standpoint,  this
means that some molecules are mov-
ing faster than others, but as they
advance in the aquifer, previously
uncontaminated water becomes cont-
aminated {although  the concentra-
tions are reduced by dilution).
    Such contaminants may appear
long before the arrival time estimated
based  on  Darcy's  law,  because
Darcy's law is a strictly advective cal-
culation that does not account  for
macrodispersive  effects.  In  sandy
soil, the front of the plume may be
traveling at a velocity 50 percent or
more higher than  the Darcy velocity.
The tail of the plume, in the same
geology, may be traveling at less than
50 percent of the velocity of the cen-
ter of mass (Cleary, 1998). Contrary
to the idea that, due to dispersion,
contaminants  move  faster  than
groundwater, what actually happens
is that some contaminants may move
faster than the average linear velocity
of the groundwater (White, 2002).
Biodegradation Constants,
Half-Lives, and Electron
Acceptors
There is an increasing body of litera-
ture documenting the biodegradation
of MtBE.  However,  MtBE will not
biodegrade in every environment at a
rate sufficient to achieve remedial
objectives within a reasonable period
of time.  In some environments, it
may degrade almost as  quickly as
benzene,  while  in others, the rate
may be  so slow  as to  be almost
nonexistent. Only real field data will
give a reliable estimate of degrada-
tion rates. The  most  conservative
approach is to run the model with no
decay.
    The first-order decay coefficient
(lambda) is a rate coefficient describ-
ing the first-order decay process for
dissolved  constituents.  In  BIO-
SCREENII,  the  first-order  decay
process assumes  that the  rate of
biodegradation depends only on the
concentration of the contaminant and
the rate coefficient.
    Typical methods for selection of
decay coefficients include the follow-
ing:
•  Literature values - Various pub-
   lished references are  available list-
   ing  decay half-life values  for
   hydrolysis  and  biodegradation
   (e.g., see Howard et al., 1991).
   Note that many references report
   the half-lives.  These values can be
   converted to the first-order decay
   coefficients using k = 0.693 / tl /2.

*  Calibrate to Existing  Plume Data -
   If the plume  is in a steady state
   or  diminishing condition,  BIO-
   SCREEN can be used to determine
   first-order decay coefficients that
   best match the observed site con-
   centrations. One may adopt a trial-
   and-error procedure to derive a

-------

  best-fit decay coefficient for each
  contaminant. For still-expanding
  plumes, this steady-state calibra-
  tion method  may  overestimate
  actual  decay-rate coefficients and
  contribute to an underestimation
  of predicted  concentration levels,

Calibration and
Sensitivity Analysis
Model  calibration  is   extremely
important. However, in calibrating
the mode! there may be ten different
inputs that can be varied to get the
results to look  like the real-site data,
therefore the solution is nonunique!
How  do you  know that you are
tweaking the  right ones  to get the
model to look  like the real-life data?
(Or worse, what if the inputs art-
being tweaked to  give a favorable
outcome  and not necessarily  a rea-
sonably accurate one?)
   Part of the  solution is to conduct
a sensitivity analysis. When a given
parameter is changed by a specific
amount,  how  much difference is
made to the final answer? A model
will be more sensitive to certain para-
meters. In  developing Delaware's
RBCA   program   (DERBCAP—
Delaware's Risk-Based  Corrective
Program),  a  sensitivity  analysis
showed that ihe modeling was most
sensitive to  the inputs for fraction
organic carbon, depth to ground wa-
ter, source area width, and ground-
water  velocity. We recommended
that consultants stick with the default
parameters  for  the  other inputs,
because the above-1 is ted inputs were
the ones that made the largest differ-
ence in the calculated cleanup levels.

Lab-Determined Values v.
Field-Determined Values
Biodegradation rates approximately
double with every IO°F  change in
temperature. Why should you use a
value from an experiment conducted
in the lab at 70'JF, when the tempera-
ture in the aquifer may be closer to
50°F?
   Not enough is known yet about
conditions amenable for biodegrada-
tion  of MtBE  in  aquifers.  Is the
aquifer aerobic or anaerobic? Are
there  sufficient electron acceptors?
Are there sufficient microorganisms
of the correct kind? The conservative
approach would  be  to  turn off
First-order decay rate constant lor degradation (per year).

Number ol reported rates
Mean
Median
Benzene (Suarez
and Rifai, 1999)
28
3.7
NA
Benzene (Aronson
and Howard, 1997)
16
3.9
1.5
MtBE
(Wilson)
10
2.3
0.44/0.56
Sen rcc: VVV/.-M "I. J'HM,
biodegradation in your model. Occa-
sionally, I see a consultant usf instan-
taneous decay,  but  more  usually,
first-order  decay is  chosen as the
modeling option.
    The degradation  of MtBE under
aerobic or denitrifying conditions has
been documented at a wide variety of
locations. In general, the rates in lab-
oratory  experiments, or pilot-scale
demonstrations where oxygen is not
   A good report of a modeling effort
   wilt provide a sufficient amount of
   text to explain what has been done
   and why, a$ well as presenting the
    results of the modeling effort.
limiting, are rapid. The median rate is
five per year, corresponding to a half-
life of two months. Rates  that  have
been documented in the  field are
much slower, on the order of 0.4 per
year or a half-life of two years. 11  is
likely that the field-scale rates reflect
the rate of reaeration of the plume, as
well as the rate of degradation when
oxygen is available.
    "It is extremely important to real-
ize that laboratory-derived rales of
biodegradation  are  almost  never
comparable to rates observed in I he
field," says EPA Microbiologist |ohn
Wilson. Almost  without exception,
laboratory rates are much higher, and
estimations (or simulations) of the
time required  to reach remediation
goals should never be based on labo-
ratory-derived rates (Wilson, 2003).
    There is one we 11-documented
study of natural attenuation under
iron-reducing  conditions  (Land-
meyer et al., 1998). To date, no one
has shown MtBE biodegradation in
aquifer  sediments under sulfate-
reducing  conditions, and data  of
biodegradation  of  MtBE  under
methanogenic conditions are mixed
(Wilson, 2003). The geochemistry of a
site must be well characterized before
you should even hazard a guess as to
the half-life of MtBE at a site, because
there may be  little  to  no  natural
biodegradation occurring.
    It is generally considered  that the
rate  of natural bioattenuation of
MtBE is much slower than the rate of
benzene bioattenuation. Table 2 com-
pares rates of natural bioattenuation
of benzene in the field as reviewed by
Suarez and Rifai (1999), or extracted
from  the  review of  Aronson  and
Howard (1997),  to the rates of natural
attenuation of MtBE (Wilson, 2003,
Tables 3-3 and 3-4). There  is  little
practical difference in the mean rate
of natural attenuation of benzene and
MtBE. However, the median rate of
MtBE degradation is one-third that of
benzene.
    In BIOSCREEN,  if individual
constituents are being modeled with
the instantaneous reaction assump-
tion, note that other constituents pre-
sent in the plume must reduce the
total  biodegradation  capacity to
account for electron acceptor utiliza-
tion. For example, in  order to model
benzene as an individual constituent
using  the instantaneous reaction
model in a BTEX plume containing
equal source concentrations  of ben-
zene, toluene,  ethylbenzene,   and
xylene,  the  amount  of oxygen,
nitrate, sulfate, iron, and methane
should be reduced by 75 percent to
account for utilization to toluene, eth-
ylbenzene, and xylene (EPA, I99h,
'1997).
    In a recently  submitted report
using  BIOSCREEN,  a  consultant
modeled each of the constituents sep-
arately (i.e., benzene, toluene, ethyl-
benzene, xylenes, and MtBE), trying
to account for the differing physical
and chemical characteristics of the
contaminants.  However, you can't
run the model  separately for  each
constituent using the electron-accep-
tor approach because more than one
constituent uses the electron accep-
               • continued tut page IS

                             17

-------

GIGO fmti
                1 ',
tors. That's like promising the same
piece of candy tt> five different kids!
    There will also be other chemi-
cals that weren't taken into consider-
ation that would also be utilizing the
same electron receptors. In addition,
the  consultant chose  to  use  the
default concentrations for the elec-
tron acceptors, rather than field-mea-
sured data, thereby introducing yet
another source of error. For several
recent sites I've had, where the con-
sultant  actually collected  electron
acceptor  data,  the  values were
extremely low. Default values would
have given erroneously high results
for biadegradation potential. Don't
assume that  the electron acceptors
are there!
    In addition to data on  the geo-
chemistry of the site, it is wise to col-
lect field data on bacterial consortia
actually present. Bacterial counts can
be done for hydrocarbon degraders,
BTEX degraders, MtBE degraders,
etc. The  site geochemistry may be
ideal, but  if the  correct  kinds of
degraders are nol present, degrada-
tion  may not occur unless bioaug-
mentation is done.

Variability and  Uncertainty
Subsurface transport models require
various  input  parameters.  These
parameters  are both variable  and
uncertain. The variability originates
with   subsurface   heterogeneity.
Uncertainty results from inadequacies
in  measurements  (U,S.F.PA,2003;
Weaver et al., 2002b).
    The EPA Web site on modeling
subsurface petroleum hydrocarbon
transport includes a Concentration
Uncertainty model thai can be used
to assess uncertainty in  parameter
estimate?, bv end-ring ranges of  val-
ues for the transport parameters  and
source definition. The model then
runs for all combinations of the vari-
able quantities. Tables 3a and 3b list
each parameter of the one-dimen-
sional  model and  indicate which
inputs can be treated as variable.
    With  nine  potentially  variable
parameters, the model  is  run  512
times. The applet tabulates the results
and  finds the minimum and maxi-
mum of three quantities:
  Contaminant first arrival time
Table 3a Inputs to
Concentration Uncertainty Model
PARAMETER
Hydraulic Conductivity
Porosity
Gradient
Dispersivity
Fraction Organic Carbon
Organic Carbon Partition
Coefficient (Koc)
Half-Life
VARIABLE?
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Table 3ti Inputs tO
Concentration Uncertainty Model
SCENARIO DEFINITION
Hydraulic Conductivity
Scenario Definition
Source Concentration
Source Duration
Distance to Receptor
Chemical Contaminant
Threshold Concentration
VARIABLE?
Yes
Variable?
Yes
Yes
No
No
No
                                   •  Maximum contaminant concentra-
                                     tion

                                   •  Duration of contamination
                                      The model also draws the break-
                                   through curve for the mean of the
                                   parameter ranges. The range for the
                                   site is wide. The answer for a site is
                                   probably somewhere  in the mean
                                   values, but if worst case values are
                                   assumed in all cases {conservative,
                                   protective assumptions), the plume
                                   would reach  a receptor in a short
                                   time, with high values.
                                      Although there are many uncer-
                                   tainties in the study of groundwater
                                   movement, we can be certain of one
                                   thing: the geology will always be
                                   non-homogeneous. The degree  to
                                   which it is non-homogeneous and
                                   also,  perhaps, anisotropic,  depends
                                   on the geologic history of the  area
                                   (Cleary, 1998). In contaminant stud-
                                   ies, the variation in true velocities,
                                   especially as they relate to variations
                                   in hydraulic conductivity and effec-
                                   tive porosity for flow, will have a crit-
                                   ical effect on the accuracy of our
                                   results.

                                   Regulatory Review
                                   It's difficult to review a model you're
                                   not familiar with and to which you
                                   don't have access. Most of the models
                                   I have seen submitted bv consultants
provide a minimal amount of infor-
mation. A good report of a modeling
effort  will  provide  a   sufficient
amount of text to explain what has
been done and why, as well as pre-
senting the results of the  modeling
effort.
    Consultants need to  provide
more explanation than is often given
of the source of the various inputs.
Which values  were field measured,
and which came from literature val-
ues? Are the results presented from
various runs that vary the inputs? If
you know where in the mathematics
of a model a certain input is, you can
have an idea how much the results
may vary and in what direction by
tweaking the input a certain amount.
(Is  the value in the numerator, the
denominator, or a squared term?)
    Regulators must use conserva-
tive values when running models to
be certain that the results are protec-
tive of human  health, safety, and the
environment. As I have shown, sev-
eral of the inputs are extremely sensi-
tive as to how and where the data are
collected, and the characteristics of
MtBE make these inputs more critical
than for some of the less mobile con-
stituents.

Summary
Models can provide useful informa-
tion about specific sites, but we need
to make  sure that the data that we
put into them is the best that we can
collect. Site-specific data can signifi-
cantly  improve the results, and we
need to be sure we  are measuring
what we  think we are measuring. As
regulators, we need to err on the side
of conservatism to be protective, but
to assume a worst-case input in all
cases would   drive  most cases to
remediation.
    We also need to be sure that the
model used for a particular site is
appropriate. If the model requires a
host of assumptions that are clearly
contrary  to actual conditions at the
site, then no  matter how good the
input data are, the output will  be
meaningless.
    Some of  the modeling efforts
that have been submitted to me have
definitely fit the acronym GIGO, but
with more thought and effort given
to choosing assumptions and inputs,
the output can be greatly improved.

       • References con tinned on page 26
18

-------
Leak  Detection in  the European Union
by Jamie Thompson

        When a station owner or operator installs a leak detection system, he or she
        immediate]]/ feels much more comfortable with the facility.  The same can be
        said for the regulator, who is concerned with whether the leak detection si/s-
tem has been installed and who checks to see if it is operating properly. Botli could be
considered to be doing all that if iiecessnrif to comply u>ifh regulations, as well as doing
their bit with regard to environmental responsibility...or are they?
    As I see it, one of the biggest problems in the petroleum marketing industry is our
understanding of the words "leak detection." I developed mi/ understanding of leak
detection after 40 years in this  industry. I used  to think that London had one of the best
leak detection systems in Europe—the London  Underground Railway. If a gas station
leaked, product soon found its way into one of the 600 miles of tunnel.  "Dangerous and
not very satisfactory," I can  hear you say. Following are some of my present dai/
though ts o11 leak detectioii.
The Need for Reexamination
Leak detection for underground stor-
age tanks and pipes developed fol-
lowing many years of problems, in the
United States, Europe, and through-
out the world. One of the problems of
technology development is that  the
technologies in  the forefront at one
moment in time can quickly become
the accepted practise and standard. It
is essential, though, that we challenge
those standards and reexamine our
methods as time goes on and better
systems are developed,
   A prime example is the detection
of leaks and  what is considered
acceptable. Take, for example,  the
American standard requiring that
760 mL/hr (0.2 gal/hr) be detected
95 percent of the time. Applying the
commonly used threshold  for declar-
ing a leak of 378 mL/hr (0.1 gal/hr),
this standard allows some 3,316 liters
(876 gallons) of fuel a year to enter
the ground  before  a  release  is
detected. The regulatory  standard
for inventory reconciliation  in  the
U.S. is even worse. This rule allows
monthly losses of up to 1  percent of
the sales plus 492 liters (130 gallons)
before failing the tank. For a modest
throughput   station,  this  could
amount to losses of more than 2,000
liters  {528 gallons)  per month  for
monthly  sales  of  200,000  liters
(52,840 gallons).
   While these amounts may well
have been accepted in the past when
our technology was not as sophisti-
cated and  we used  single-walled
storage systems, a  conscientious
operator or regulator would not  be
happy to have this amount of prod-
uct enter the environment today. Yet
these are the standards that many of
us still use.
   The European Union (EU) has
been drafting common standards for
many years, and  one area of those
standards  that has  been worked
upon and published  this year is BS
EN 13160-1 to 7 Leak Detection Sys-
tems. These standards cover the vari-
ous methods of leak detection that
can be used on  tanks and pipes. They
are performance standards and have
detailed test methods for evaluating
equipment to show that it will work
(Though they  are sometimes more
prescriptive than U.S. standards—
See Ken Wilcox Article on page 21.).

Positive Leak Detection
One  solution to underground stor-
age system releases  that has been
widely adopted in Europe is what I
prefer to call leak prevention or positive
leak detection. This involves the con-
tinuous monitoring of the integrity
of the two skins of a double-walled
tank  or pipe. This system has been
used in some Exiropean countries for
many years, and in Germany  for
underground tanks since 1968. For
comparison, the American standard
only   requires   monitoring   the
integrity of the inner wall of a dou-
ble-walled tank or pipe.
   The basic  principle  of positive
leak prevention is quite simple: you
provide a pressure or a vacuum in
the tank and piping interstices, and
the two skins are always under test
for the whole life of the installation.
As soon as one of the walls  is
breached, an alarm sounds, the tank
is emptied of product, and no prod-
uct enters the environment. It is an
attractive alternative for both regula-
tor and operator.

Pressure v. Vacuum v.
Tradition
Pressure  systems  pressurize  the
interstitial  space to 440 millibars
(6.38 psi) in such a way that a leak
anywhere in the interstitial  space
will cause  a  detectable  pressure
drop.  Minor  long-term  pressure
losses  in  the system are restored
(within very tight limits laid down in
the standard) by a small pressure
pump—a low-capacity pump that is
provided  as part of the  system to
maintain  pressure  despite  these
small pressure losses, thus avoiding
periodic false alarms due to  "nor-
mal" pressure loss.
    If a small leak occurs in the inter-
nal wall,  the pressure pump will
attempt to maintain the pressure in
the interstice. A major advantage of
the positive-pressure system is that
the pressure in  the  interstice pre-
vents any gasoline from entering the
interstitial space. When the pressure
drops to around 330 millibars (4.79
psi), an alarm will sound, alerting
the operator that there is a problem
before any stored product can escape
to the environment.
    Vacuum systems work in a simi-
lar way to the pressure systems, but
in reverse. A small vacuum pump is
used to maintain the vacuum  in the
interstice. If the vacuum decreases to
below  a  set point, or if liquid is
drawn into the tubing leading  to the
vacuum pump,  an alarm sounds. A
disadvantage of the vacuum system
is that a leak in  the inner wall will
allow product to enter the interstitial
space. The presence of product in the
interstice may complicate the process
of  repairing the tank, if a repair is
                • continued on page 20

-------

• Leak Detection in Europe
fnini fagc 19
feasible.  Again, the alarm sounds
before any product can enter the
environment
    The real value of the pressure
and vacuum  systems is that  they
have no detrimental affect whatso-
ever on the environment. The pres-
sure or vacuum is maintained using
air or inert gas. This is a different cat-
egorv ol leak detection than the older
technology of using a  food-grade
glycol  and water solution (brine
solution in the U.S.) to monitor the
integrity of thy walls  of  double-
walled  tanks.  The  move toward
these systems is being encouraged
by many European authorities  who
prefer  this over having the small
amounts of glycol released into the
ground, as would happen when the
outer wall of  a glycol-based leak-
detection system developed a leak.

Tech Transfer Issues
The European Union parameters for
pressurized leak-detection systems
must be closely examined if applied
to UL tanks, as distortion or failure
of the inner tank could occur.  Vac-
uum leak-detection systems may be
more suitable for tanks built to U.S.
standards. With pipes, the use of
plastic is widespread in Europe  with
polyethylene the more popular (note
that European polyethylene piping is
more rigid than U.S. flex pipe and
has thermowelded  joints). Class  1
and 3  leak-detection systems  (see
table below) are normally used on
pipes in Europe.
    In Europe,  most pumping sys-
tems are still  suction, although the
move to pressure is being made. The
application of leak-detection Classes
1 and 2 to pressurized pumping sys-
tems that are  common  in the  U.S.
could be problematic because of the
significant pressures that the outer
wall of the piping must withstand.
Class I pressure could be a problem
with flexible pipes, as the standard
recommends a pressure in the inter-
stice of 1 bar (14.5 psi) over the nor-
mal working pressure in the primary
pipe. Vacuum systems should work
well on pipes, assuming the system
can be made tight enough to hold a
vacuum.

The Class System
With the variety of European coun-
tries, types of installations, and dif-
fering regulations in the past, there
was a need to have the leak-detection
systems  divided  into  classes  to
enable countries to adopt the stan-
dards according to their needs. The
classification system applies to  both
tanks and piping.
    You will, of course, recognize
Class 3 (interstitial monitoring at
atmospheric pressure), Class 4 (mon-
itoring tank contents), and Class 5
(monitoring wells), as the slandards
commonly used in the U.S. In addi-
tion, the U.S. has electronic line-leak
detection on  pressure systems; in
Europe we are also at present draft-
ing such a standard, which  J hope
will be added to this standard in the
future as 13160-8.  Even so, line-leak
detectors are only a Class 4 system.
    With regard  to  the European
standards, the most important point
in using any system is to  understand
the class, the method, and the limita-
tions of the systems before applying
it to a site. Guidance on when to use
the classes is  given  to  regulators,
operators,  and  engineers  in an
APE A/IP book,  Guidance for the
Design, Construct ion, Modification and
Maintenance of Petrol Filling Stations
ISBN 0 85293217 0. The chapter on
leak detection gives excellent advice
on how and when to apply the sys-
tems according to risk.
    The world is a smaller place, and
you therefore probably would not be
surprised to learn that a number of
U.S. experts and organizations had
input  into  these  standards—for
which we are grateful. Through this
effort,  1 believe we have provided
industry a way to improve the con-
struction of gas stations in Europe.

For More Information
The following EU standards can be
ordered through iiwu\bsi-%lobal.cow:
• BS EN  13160-1 2003 Leak detec-
  tion systems general principles
• BS EN 13160-2 2003 Pressure and
  vacuum systems
• BS EN  13160-3 2003 Liquid sys-
  tems for tanks
• BS EN  13160-4 2003 Liquid and
  vapour sensor systems
• BS EN  13160-5  2003 Contents
  gauge systems (being published
  2004)
• BS EN 13160-6  2003 Sensors in
  monitoring wells
• BS EN 13160-7 2003 Requirements
  and  test methods for interstitial
  spaces, tank linings and jackets
    For  information  on  how  to
obtain  Guidance for the Design, Con-
struction,  Modification and Mainte-
nance of Petrol Filling Stations ISBN 0
85293  217  0,  write  to  admin@
apea.org.uk. •

          ''if/'M'ii ;




                   !




  jamiethomp1                uk.
CLASS
1
2
3
4
5
METHOD OF OPERATION
• Monitors air pressure or vacuum
between the skins.
• Alarms if pressure changes
Monitors pressure of a liquid filling the
"gap between skins
Detects presence of liquid or vapour
within the Iwo skins.
Analyses rates of charge in tank contents.
Monitoring welts around installation
DETECTION CAPABILITY
Detects teaks anywhere in double-skin equipment
irrespective of product level.
Detects leaks anywhere in double-skin equipment.
irrespective of product level
Detects leaks in double-skin systems with large
interstitial space
Detects leaks below liquid level in tanks and pipes
Detects leaks below the liquid surface in tanks and
pipes.
EFFECTIVENESS
Very secure system alarms betore product can reach
environment
Very secure system alarms on leak. Leak detection fluid
enters environment on leak.
Sensors at low pomis unable to detect failure o) outer
skin or inner skin above liquid level.
Product is released to the environment before a leak is
detected.
Product is released to the environment before a leak is
detected.
20

-------
One  American's  View  of the  European
Approach to  Leak Detection
by Ken Wilcox
      As most people in the U.S. leak
      detection industry are aware,
      U.S.EPA requires that leak-
detection equipment meet specific
performance standards before it can
be used legally to monitor under-
ground  storage  tanks containing
motor fuels. This performance must
be verified, usually by an indepen-
dent third-party evaluation, using
specific  testing procedures defined
in a series of documents produced by
EPA in  1990, More recently, other
test protocols have been developed
by nongovernmental organizations
and approved by the National Work
Group on Leak Detection  Evalua-
tions (NWGLDE).
    A parallel effort to develop stan-
dard procedures has been occurring
over the past  ten years or so  in
Europe.  (See "Leak Detection in the
European Union" on page 19.) While
these procedures are of interest  to
U.S. UST technology developers and
others with  an interest  in selling
equipment in Europe, they have not
yet had  much direct impact on U.S.
evaluation and  testing of leak-detec-
tion  equipment. This may  change,
however, because the NWGLDE has
officially adopted one of the Euro-
pean leak detection evaluation proce-
dures,   making it  an acceptable
evaluation procedure here in the U.S.
    The  document adopted is EN
13160-2, "Leak Detection Systems -
Part 2: Pressure and  Vacuum Sys-
tems."  (This document is copy-
righted  and  can only be obtained
from one of the European standards
organizations for a nominal  fee. One
such organization's Web site is con-
tained in the  article on page 19.) The
protocol applies to interstitial moni-
toring systems in both double-walled
tanks and pipelines and contains pro-
cedures  for the evaluation of moni-
toring   methods  that  rely  on
maintaining  either pressure or vac-
uum in the interstice.
    Although the document covers a
number  of types of leak-detection
systems, I will focus my  comments
only on pressure/vacuum systems
for interstitial monitoring, because
this is the only part of the document
officially adopted by the NWGLDE.

Contrasting Approaches to
Environmental Protection
There are a number of features of the
European approach to protecting the
environment that differ significantly
from that  of the  EPA. Space pre-
cludes a detailed evaluation of all the
differences, but here are three that
merit comment.
• The emphasis in Europe is on the
  prevention of loss of product to
  the  environment  rather  than
  detecting a leak of a specific size
  (e.g., 0.1 gallon/hour).

• The  evaluation  protocols from
  Europe are somewhat prescriptive
  in nature. Not only is the  testing
  procedure described, but descrip-
  tions of how the leak detection
  equipment should be designed are
  also provided.

• The testing is focused on the hard-
  ware and its capability to continue
  to operate under a  very wide
  range of temperature conditions
  ranging  from -25°C (-13°F)  to
  +70°C (168°F). The test procedures
  do not, for the most part, address
  all the other  variables that might
  be present in a  field installation
  {e.g., size of tank, length of  piping,
  operating pressures of the piping).

Allowable Leaks
The first difference listed above is
very fundamental. The Europeans
have divided leak detection into five
classes. (See article on page 19.) It is
interesting to note that most of the
methods used in the U.S. are  in cate-
gories 3,4, and 5, which do not report
leaks until fuel is already  in the
ground. Most   of  these methods
would not even be allowed as a pri-
mary leak-detection method in some
countries in Europe.
   This fundamental difference has
already had an impact in California.
With  the  adoption  of  SB 2481
(u'ivu>.5ivrcb.cn.gov/ust/docs/eld/index.
html), the regulatory emphasis  has
shifted to the European focus on loss
prevention rather than leak detection.
The requirements described in SB
2481 state that pressure, vacuum, or
hydrostatic   monitoring  be  main-
tained in the interstitial space of dou-
ble-walled pipe for loss prevention
purposes.
   These rules apply  to  all new
installations after July  2004. They
apply to all newly installed storage
systems as  of July  1, 2003, but the
implementation  date  has  been
delayed by  one year because there
was no approved technology avail-
able that met the rules that could be
applied to American-style  pressur-
ized piping systems.
   This change in  emphasis is ini-
tially painful, particularly for station
owners, because a whole new type of
leak-detection system must now be
developed and installed to comply
with the new regulations. Although
most thoughtful people would agree
that the overall objective of loss pre-
vention is a good one, there are some
cautions along the way.
   It is important for the regulatory
community  to  recognize that some
things simply cannot be done as well
and as quickly as they would like until
the leak-detection industry has  an
opportunity to develop equipment to
meet the new standards. In addition,
we need an  opportunity to work out
the bugs  in  the new systems before
deadlines kick in. Murphy's  Law will
likely be in effect, and these new sys-
tems should not be expected to work
perfectly during their initial use.

Approach to Standards
The second  difference is a result of
two significantly different  ways to
approach standards for leak detec-
tion / prevention: performance-based
standards v. design-based standards.
               • continued on page 22

                             21

-------
                .
• American's View twin page 21

EPA's rules on leak  detection are
largely  performance based  in that
they define an end result (e.g., must
detect a leak of 0.2 gal/h) and leave
the method to obtain the results up to
the leak-detection industry to work
out.  Almost nothing is provided on
how the equipment should work or
what physical principles should be
used to obtain  the   results. The
emphasis is primarily on how the
system should be evaluated to deter-
mine if it works.
   The European protocols tend to
include design parameters as part of
the evaluation process. Part  2 of the
European protocols describes in con-
siderable detail  how the pressure/
vacuum leak detector should be con-
structed and how it should  work. For
example, the flow rates provided by
the pressure and vacuum pumps are
defined. Specific  types  of check
valves are to be included. The size
jnd color of tubing used to construct
the equipment are provided. In short,
till leak-detection systems  that meet
the strict guidelines of the European
protocol will be similar.
    Is one of these approaches better?
It is partly a matter of the philosophy
of [he regulating organization. The
performance-based standards allow
for much more creativity and  innova-
tion  of  new methods. This  can  be
both good  and bad. Better  mouse-
traps have  been a basic tenet of
American ingenuity. This can, how-
ever, result in equipment that may
pass an evaluation but stil! not work
well  in an actual field application.
   A more prescriptive  approach
assures that fewer of these poor per-
formers will be  developed because
the  basic engineering  behind  all
methods that meet the requirements
is known to be reliable. Having all the
equipment basically look and work
the same also simplifies the task of
conducting regulatory inspections.

The Variables
The third difference should be of con-
cern, particularly where the technol-
ogy  is  new  and  unfamiliar. The
California requirements specify that
virtually the entire fuel system  be
doubly contained right up to  the dis-
penser.  This  includes  sumps, lids,
penetration  fittings,  piping, and


22~
virtually any other piece of hardware
that is in contact with fuel.
    Exactly how  all this will work
together is not known at  this time.
What is certain is that it will not work
very well at first. The problem is that
the evaluation required by the Euro-
pean protocol does not address the
system as a whole, a basic feature of
the U.S. approach.
    The European testing addresses
the  leak-detector function  when
attached to a one-liter test cell over a
wide range of temperatures, A  leak
detector passing these requirements
will continue to function equally  well
in Alaska in the winter or Phoenix in
the summer.
     // does seem to us that some
  consideration should be given to Hie
   complete fuel system to which the
  leak detector is to be applied.  We do
  not want to be accused of certifying
   equipment in our laboratories that
  later turns out to be a dismal tailure
           in the field.
    What   is  not  determined   is
whether the system will detect leaks
reliably without false alarms when
connected to a complex fuel system.
In fact, a leak detector passing the
Part 2 requirements can be hooked to
virtually any system, no matter how
complex, without any further consid-
erations. A detailed look at how this
would be evaluated is beyond the
scope of this article, but suffice it to
say the complexity of the fueling sys-
tem must be addressed  sooner or
later if the leak detection is to be reli-
able.
    Since my company, KWA Inc., is
primarily a testing laboratory, we are
not directly concerned with the per-
formance of equipment in the field.
But it does seem to us that some con-
sideration  should  be given to the
complete fuel system to which the
leak detector is to be applied. We do
not want to be accused of certifying
equipment in our  laboratories that
later turns out to be a dismal failure
in the field.
    This  concern  is  particularly
important when you realize that the
fuel systems in Europe are somewhat
different than in the U.S. Many of the
European pumping systems utilize
suction technology (although this
seems to be changing). Some of their
systems  incorporate  welded-stecl
pipe that  is much thicker than the
metal pipe used  in the U.S. Other
significant differences may also be
present.

A Final Note
The process for developing new eval-
uation protocols needs to be recon-
sidered. The rate of development of
new technology  far  outstrips the
capability of the regulatory commu-
nity to provide reliable evaluation
procedures. For example, it took ten
years for the final documents of the
European protocols to be issued, and
Part 5, dealing with tank gauges,  is
still not complete.
    The process to amend the testing
protocols or create new ones needs to
be streamlined so that when new
equipment is  developed, the  leak-
detection industry can expect that the
approval   process will  function
smoothly and efficiently.
    There are other differences in the
U.S. and  European leak-detection
approaches, but I have limited my
discussion to those that seem to have
the most impact in the U.S. My com-
ments are not meant to be a criticism
of our many friends in Europe, but to
point out that our procedures are dif-
ferent. They are ahead of us in their
emphasis on loss prevention.
    When  I  checked   with  Mr,
Thompson, author of "Leak Detec-
tion in the European Union," on how
the Europeans would view devia-
tions from the protocol, he said that
the European Standards committee
would need to be convened to review
the requested modifications. This is
not different from what happens in
the U.S. •


 M\ '

-------
                                                                                      .
from Robert N. Renkes, Executive Vice President, Petroleum Equipment Institute

CHANGES IN THE PETROLEUM MARKETING SCENE MAY PRESENT NEW
WORRIES FOR UST REGULATORS
       Consider the  following an-
       nouncements released  by
       major oil companies during
 the last several months:
 •  ChevronTexaco said it would
 sell 550 gasoline service stations in
 the United States and 900 in Asia
 and Africa as it restructures its
 downstream refining and market-
 ing business,
 •  Alimentation Couche-Tard Inc.
 reached an agreement in October
 with ConocoPhillips to acquire the
 Circle K  Corp.,  which operates
 1,663 corporate stores in 16 states
 and has a franchising or licensing
 relationship with more than 350
 additional stores. Approximately 86
 percent of those stores sell gasoline.
 The Circle K sale was expected, as
 ConocoPhillips told investors last
 December it would sell 3,200 mar-
 keting sites this year.
 •  Shell is planning to shed stations
 across a broad section of the coun-
 try, particularly in markets where
 margins have produced  subpar
 results. The company is also in the
 process of turning over 1,750 com-
 pany-owned stores to "multi-site
 operators," or MSOs, but will retain
 responsibility for the fuel opera-
 tions.
    [t seems clear that the major oil
 companies have decided to reduce
 their capital investments in refining
 and marketing activities to boost
 overall returns. They have deter-
 mined, one by one, that they are
 better at and can make more money
 by producing oil, rather than run-
 ning  retail gasoline  and conve-
 nience store operations. And there
 are other compelling reasons for the
 majors to  shift their  focus away
 from retail marketing:
 •  All  traditional gasoline  mar-
 keters, including the major  oil com-
 panies, have faced stiff competition
 in the form of heavily subsidized
 gasoline marketed  by hypermar-
kets.  Gasoline  marketers  have
acknowledged that hypermarkets
are not going away.
•  The oil companies' experience
with hypermarkets in Europe, most
notably in France and Great Britain,
has shown that double-digit gaso-
line  margins  are  not likely  to
bounce back. As one major oil com-
pany vice president of European
retail operations told us privately
four years ago: "We have given up
trying to make money at  retail. It
just isn't there."
•  Price-competitive and very effi-
cient convenience store operators
have expanded rapidly into high-
growth  markets and  now have a
dominant position in them.
   This is not to say that the major
oil companies are abandoning retail
operations. They will still own tens
of thousands of gasoline stations
after they sell off the thousands that
are on the block. They will still con-
tinue to market gasoline  in areas
where they  have strong brands,
well-established  supply  agree-
ments,  and  competitive market
shares. But  the days of an ever-
growing retail presence appear to
be over.
    What will this mean to under-
ground storage  tank regulators?
Here are several possibilities.
•  In the 1980s and 1990s, when the
majors were expanding their retail
locations, they employed a full staff
of construction and service engi-
neers. One major oil company engi-
neer  told us  that  he  had  150
engineers and trainees on his retail
marketing staff at one time. Now
six engineers  service the  entire
country for the company. Oil com-
pany personnel  are spread thin,
and UST compliance issues may fall
through the cracks. That's a natural
result when you have to do more
with less.
•  Chances are the new owners of
stations sold by the majors will not
be as sophisticated from an engi-
neering standpoint.  Some  pur-
chasers may be first-time owners
and unfamiliar with what is hap-
pening underground. For example,
Acme Petroleum & Fuel Co. sold 33
convenience stores at auction  on
October 2. One company purchased
10 units and another firm bought
nine. Seven buyers purchased one
unit  each. The 33 stores and  14
dealer accounts went to a total of 14
buyers. The majority of these buy-
ers are not oil companies with expe-
rience in managing tank programs.
They are more likely to be recent
immigrants  to the United States
and speak English as a second lan-
guage.
•  The  "Mom  and Pops" {one- or
two-station owners) are back.  They
contributed significantly to the UST
problem in the 1970s and 1980s and
now account for the ownership of
over 70,000 convenience stores in
the United States. Outreach to  these
UST system owners may be expen-
sive and time-consuming.
•  Look for more station closures in
the future.  In general, the retail
petroleum   marketing   industry
today  is barely  making enough
money to stay afloat. Unprofitable
sites that can't be sold will  be
closed. Stations involved in bank-
ruptcy  proceedings may stay  shut-
tered for quite some  time. And if
station  owners   can't  generate
enough money to stay in business,
it's difficult to assume that these
same people  will have  enough
money to spend on their tank sys-
tems.
    The  petroleum   marketing
industry is currently undergoing
huge changes. Those changes will
require UST programs to change as
well if the  public  is going to  be
served  in the future as well as  it has
in the past. •
                                                                                                   23

-------
               •
                       - InicafJy Speaking
                           by Marcel Moreau
                                       Marcel Moreau if a
                                       •;>uzi'd fwtrvleum storage specialist
                                       'column, Tank-nically Speaking,
                                     is a regular feature of LUSTLine. A*
                                       i/*. in- welcome your continents ami
                                           S. |f there are technical ;>sins
                                         yew would like to ham- Marcel
                                          i/i'sniss, /['/ /H';H tHinr (if
                                        marcel.moreau@juno.com
Will the Real Quantitative  SIR  Method
Please Stand Up?
      S
-------


    While the presence or absence of
a numeric leak or gain rate contained
in the SIR report is the most obvious
distinction between quantitative and
qualitative SIR methods to the casual
observer, the distinction between the
methods goes much deeper than this.
The statistical techniques used in the
evaluation protocol for qualitative
and  quantitative SIR  methods are
very different.
    The quantitative evaluation is
much more rigorous because it com-
pares the leak rate calculated by the
SIR method to the 0,0, 0.1, or 0.2 gal-
lon per hour leak rate introduced into
the inventory data  set by the SIR
evatuator. In a qualitative SIR evalua-
tion,  there  are  only  two possible
answers that the method  can pro-
duce—pass and fail. As in a true/
false test question, the SIR method
has a 50 percent chance of getting the
correct answer without making any
calculations. Because the evaluation
is based solely on the pass/fail con-
clusion, no evaluation is made of the
ability of the qualitative SIR method
to accurately estimate leak rates.

Performance Standard or
Reporting Requirement?
Because of Florida DEP's determina-
tion that SELA did not meet the new
regulatory requirements, compliance
inspectors were authorized  to cite
SELA customers for failing to con-
duct leak detection because their SIR
reports did not include values for cal-
culated leak rate, threshold, or mini-
mum detectable leak. To remedy this
deficiency, SELA started reporting
these   numbers   in   subsequent
monthly reports. SELA did this with-
out the benefit of a third-party evalu-
ation to show that  these  numbers
were accurate.
    When  confronted  with  this,
SELA argued that  the Florida rule
had  not changed  in 1998, that the
new regulatory language consisted
merely  of a  requirement to report
additional information and that the
performance standard for  SIR  of
detecting a 0.2 gallon per hour leak
had not changed. Because the perfor-
mance standard had not  changed,
SELA argued, no  new certification
was required.
    Florida DEP pointed out that the
requirement to report these numbers
was contained in a section of the rule
entitled  "Performance Standards"
and that the new requirements were
both reporting requirements and per-
formance standards. The DEP argued
that in addition to requiring SIR ven-
dors to include leak rates, thresholds,
and minimum detectable leak values
on their results forms, they must also
have some certification  indicating
that these values are accurately calcu-
lated.
    Attempts to settle  the  matter
were not successful, and Florida DEP
eventually sent a letter of intent to
revoke SELA's approval to sell leak-
detection  services  for  regulatory
compliance. SELA then filed a peti-
tion for an  administrative hearing to
decide the matter before a judge. The
hearing  to argue the matter took
place this past August. A final deci-
sion from the judge is not expected
before next spring. Until a decision is
rendered and the  legal process has
run  its course,  SELA continues to
operate in Florida.

Your Bottom Line
So why is any of this important out-
side of Florida? Because the bottom-
line  message is that leak-detection
equipment   vendors  and   service
providers still need to be kept honest
by a careful comparison of the regu-
latory requirements to the equipment
certification. Specifically with SIR, a
SIR method holding only a qualita-
tive  certification does not,  in  my
opinion, meet a regulatory require-
ment to report leak  rates merely by
adding a leak rate to the output of the
software.
    The  regulatory  requirement to
report numeric leak rates is a  perfor-
mance standard, because leak rates
should not be numbers pulled out of
a hat. There has to be a level of confi-
dence associated with these numbers;
a level of  confidence provided by
putting the SIR software through the
quantitative EPA evaluation protocol
and applying statistical measures to
determine  the accuracy of the leak-
rate calculation.
    If your state UST  requirements
specify that SIR results include a leak
rate, you have effectively specified
that only  SIR methods  holding  a
valid  quantitative  SIR  evaluation
may be used in your state. •
SIR  Terms
When used in this article,
SIR terms mean the following:

•  Performance standard  This
refers to the regulatory require-
ments that  a  SIR  method must
meet to  be  deemed a valid leak-
detection method. Such require-
ments may include (but  not be
limited to) the ability to  reliably
detect whether a storage system is
leaking above a specified leak rate,
the ability to accurately calculate
leak rates in a data set, or  specifi-
cations for how inventory data to
be used  in a SIR analysis must be
gathered.

• Threshold  For quantitative leak-
detection methods, this means the
measured (or calculated) leak rate
that defines the boundary between
a "pass"  and a "fail" test result. To
achieve a 95 percent probability of
detection, the threshold leak rate is
typically  about one-half the perfor-
mance standard.

• Minimun detectable leak Fora
given  inventory data  set,  this
means the smallest leak rate that
can be reliably detected in a SIR
analysis.

• Pass  If the calculated leak rate
produced by a  SIR analysis is less
than the leak threshold, and the
minimum detectable leak rate is
less than or equal to the certified
performance standard (0,2 gallons
per hour), the test result is "pass."

• Fail   If the calculated leak rate
produced by  a SIR analysis  is
greater than  the leak threshold, the
test result is "fail."

• Inconclusive  If the calculated
leak rate  produced by a SIR analy-
sis is less than the leak threshold,
and the  minimum detectable leak
rate exceeds the certified  perfor-
mance standard (0.2 gallons per
hour), the test result is "inconclu-
sive." If  for any other reason, the
SIR test result is not conclusive
(i.e., "pass"  or "fail"), the result is
"inconclusive."

-------
                 .
^
  esponse to Comment on
California's Field-Based
Research Study
Carol Eig limey's  "Thoughts on the
Tortoise and the Hare Revisited" arti-
cle (LUSTLiw #44) raised ww.t'[>a.gfn>/adaf
  (siHefimixlek/hpsfrti html
USEPA, 2003. Average Borehole Concentration cal-
  culator, htttKHutiWfpit.gpriathens/lrarttliHoiltHpart-
  la.ta!amileliAf.Mm Site visittd 9/25/03.
USEPA, 2003. Modeling Subsurface Transport of
  Petroleum Hydrocarbons. Uncertainty in Model
  Outcomes  http:Viwu',epa.giivtiithfn*th'arii2m(tdelt
  parttwti.'imsiMuiicertainty.hrm; Uncertainly Calcula-
  tions httpi/ln'WU'.fpa.gavlalhens/ ltar»2modellpart-
  itm/onsilehiiKerlatnti/()2.hlm.
  Sites visited 9/25/0}
Weaver, James W. and Matthew C. Small. 2002.
  MlBE: Is a Link- Bit OK?, Presented at the National
  Ground Water Association's  2002  Petroleum
  \ hydrocarbons and Organic Chemicals in Ground
  Water: Prevention, Assessment, and Remediation
  Conference. November 6-8, 21X12 Atlanta, GA..
  pgs 206-219.
Weaver, James W,, and [ohn T, Wilson. 2002. Diving
  Plumes and  Vertical Migration at  Petroleum
  Hydrocarbon Release Sites, LUSTl.int' Bulletin 36.
  November 2(XX). pg",. 12-15.
Weaver, fames W., Caroline L. Tebes-Slevens, and
  Kurt Wolfe. 2002. Uncertainty in  Model Predictions
  - Plausible Outcomes from Estimates of Input
  Ranges. Presented at Brownfields 2002, Charlotte,
  N.C., November 13-15, 2002. httfKl/u'ti'U>.epa.gai'/
  it!lieiis/!carn2miiiM/i>art ttpo
                                                                                   While, Hal, 2001  MlBE Taste and Odor Thresholds
                                                                                    ... The Myth of Proler riveness. LUSTI.ine Bulletin
                                                                                    ,19. November 21X11. pgs. 6-7.
                                                                                   White, Hal. 2002a. Do Monitoring Wells Monitor
                                                                                    Well? Part I, LUSTLine Bulletin 40. March 2002,
                                                                                    pR5. 13-16.
                                                                                   White, Hal. 2002b. The Eleven Myths about MtBE.
                                                                                    tUSTUne Bulletin, #42. pgs. 9-16.
                                                                                   Wilson, John T.. 2003 Fnfr and Transfer! vfMtRe and
                                                                                    Other Gasalitti,' Ctntifh>utut^. Chapter 3 in Remedia-
                                                                                    lum Handbook, edited by Ellen F.. Mover and Paul
                                                                                    T,  Kostecki.  Amherst  Scientific  Publishers,
                                                                                    Amherst, MA pgs 19-61.
26

-------

Flex  Pipe  Class Action  Suits Filed
by Thomas J, Schruben

     Flexible pipe litigation seems to
     be on the upswing. LIST owners
     have filed suits against several
flexible-piping installers, manufac-
turers, and  component suppliers.
Some of the suits have been brought
by individual retail operators, one
was  brought by a former installer
and  distributor, and  two national
class action suits have been filed on
behalf of UST owners. The two class
action suits, Russell  Petroleum  v.
Environ and May's Distributing v.
Total Containment, are similar  in
many respects:
• They are both asking for National
  Class Action status

• They were filed in Alabama

• They are each backed by consor-
  tiums of plaintiffs attorneys
• They allege similar defects and
  damages

• They name the installer, the manu-
  facturers, and the suppliers of var-
  ious components as defendants

Russell Petroleum v. Environ
This suit was filed in August, 2003, in
Montgomery County Alabama. The
plaintiff, Russell Petroleum Corpora-
tion, of Montgomery, Alabama, dis-
tributes oil and gasoline products
and operates convenience stores and
motor-fuel dispensing facilities, Rus-
sell Petroleum brought the suit indi-
vidually and  as a class representative
for all those motor-fuel dispensing
facility owners similarly situated in
the United States.  In  1998, Russell
Petroleum contracted  with   Ken's
Sales  and Service Company,  Inc. of
Ashford, Alabama (Ken's), to install
underground thermoplastic flexible
pipes and sump systems distributed
by Environ Products, Inc. (Environ)
and designed and manufactured by
one or more of the  following defen-
dants:
• Dayco Products, Inc.,
• Mark IV Industries, Ltd.

• Parker Hannifin Corporation

• Atofina Chemicals, Inc.
    The suit also names "fictitious
defendants" A through FF to allow
for the later addition of other parties
that are found to have manufactured
or supplied components to Environ.
    The suit alleges that on or about
March 4, 2002, Russell Petroleum dis-
covered that the Environ flexible pipe
at its location  was defective. Upon
inspection and  testing,  it was discov-
ered that the very properties of the
Environ flexible pipe had changed, in
that fuel was permitted to permeate
outside the barriers of the flexible
pipe and that the pipe failed to retain
its shape and rigidity,  experiencing
elongation,  swelling,  and other
defects.  Russell Petroleum  claims
that it suffered damages as a result of
the defective Environ  flexible pipe.
The suit seeks injunctive relief requir-
ing Environ to pay for the removal
and replacement of its  defective flex
pipe. The suit states that the principal
common issues for the class are:
•  Whether the Environ flexible pipe
   is defective
•  Whether the components of  the
   flexible pipe  are defective

•  Whether the defendants knew or
   should have known  that  the flexi-
   ble pipe was defective

•  Whether the defendants  know-
   ingly sold a defective product

•  Whether the conduct of the defen-
   dants was fraudulent

»  Whether the conduct of the defen-
   dants constituted negligence, reck-
   lessness and/or wantonness  in
   regard to the class

•  Whether defendants negligently,
   recklessly and/or wantonly  de-
   signed,  manufactured,   and/or
   marketed Environ flexible pipe

•  Whether defendants failed to ade-
   quately inspect or test the flexible
   pipe

•  Whether defendants failed to give
   warnings or  to  give  adequate
   warnings regarding the limitations
   of the Environ flexible pipe

•  Whether  defendants made  an
  express warranty(ies) concerning
  Environ flexible pipe
• Whether the defendants breached
  an express warranty(ies) regard-
  ing the flexible pipe
• Whether the defendants breached
  an implied warranty(ies) regard-
  ing the Environ flexible pipe

    Because of the wide use of flexi-
ble-pipe and sump systems at motor-
fuel dispensing  facilities  over a
substantial period of  time  in  the
United  States,  Russell Petroleum
believes there are thousands of mem-
bers of the class.
    The lead law firm of the consor-
tium of plaintiff's attorneys is DeHay
& Elliston,  LLP of Houston, Texas.
The lead attorney for DeHay & Ellis-
ton on flexible-piping failure cases is
Jay  Cawley.

May's Distributing v. Total
Containment
This suit  was  originally filed  in
March  2003,  in Bullock  County
Alabama and was amended in July.
The plaintiff.  May's   Distributing
Company,  Inc. of Union Springs,
Alabama, distributes oil  and gasoline
products and operates convenience
stores  and motor-fuel dispensing
facilities. May's brought  the suit indi-
vidually and on behalf of all those
motor-fuel dispensing facility owners
similarly situated in  the   United
States, in 1997, May's  Distributing
contracted with Oil Equipment Com-
pany, Inc.  (OEC) to  install  under-
ground  thermoplastic flexible pipes
and sump  systems distributed by
Total Containment, Inc. (TCI) and
designed and manufactured by one
or more of the following defendants:
• Dayco Products, Inc.
• Mark IV Industries, Ltd.

• Parker Hannifin Corporation

• Ticona Polymers, Inc.
• Shell Chemical, LP,

• Cleveland Tubing, Inc.

                • continued on page 28

                              27

-------
               •
• Flex-Pipe Suit/rum ;«X'' 27

   The suit also names "fictitious
defendants" A through Z to allow lor
the later addition of other parties that
are found to have manufactured or
supplied components to Total Con-
tainment.
   May's alleges  that on or about
September 5, 2002, May's discovered
that the TCI flexible pipe at their loca-
tion was defective. OEC inspected
and tested the TCI flexible pipe and
discovered that its properties had
changed in that it permitted perme-
ation of fuel outside the barriers of
fhe TCI  flexible pipe, and that  it
failed to retain its shape and rigidity,
by elongation, swelling,  and  other
defects. The suit  seeks  injunctive
relief requiring TCI to pay for the
removal and replacement of its defec-
tive flex pipe.
   The principal common issues for
the class listed in the suit  are the
same as listed in Russell Petroleum v.
Environ, except that TCI is named
instead  of Environ, The May's Dis-
tributing v. Total Containment class
action is limited. The class is all own-
ers of retail petroleum service sta-
tions that have TCI tlex pipe. Because
of the wide use of flexible pipe and
sump systems at motor-fuel dispens-
ing facilities over a substantial period
of time in the United States, May's
believes there are thousands of mem-
bers of the class.
   The lead law firm of the consor-
tium of plaintiff's attorneys is What-
ley Drake,  LLC  of   Birmingham,
Alabama. Whatlev Drake is known
for class action involving mass torts,
including lead exposure and  poison-
ing of children, MtBE, asbestos and
silicosis exposure of workers, injuries
arising from medical and pharma-
ceutical products, and  problems
related  to  artificial  stucco  (EIFS)
housing.

Remediation Expense
Recovery Not Sought
Neither of  the class action  suits  is
attempting to  recover remediation
expenses. According to an attorney
close to the suits, the class actions do
not involve sites where a release has
occurred above state action levels
and remediation is required because
remediation costs are site specific and
may not meet the Rule 23 require-
ments for a class. It is anticipated that
additional individual suits  will be
brought as additional UST  owners
attempt  to join one of the class action
suits and do not qualify because their
piping has caused a major release,
   The  defendants in  both of the
Russell  Petroleum v. Environ  and
May's Distributing v. Total Contain-
ment class action suits have been
served and have filed general denials
with the court. The court certification
of the class will probably be opposed
by some or all of the defendants. •
    :
                      llltllllt. Hi'
LEGAL  UPDATE
New Hampshire Sues 22
Oil Companies Over MtBE
New Hampshire has become the first
state to sue oil companies over the
gasoline additive  methyl  tertian/
butyl ether (MtBE). On October 6,
2003, the state filed a lawsuit in state
court against 22 major oil companies,
including ExxonMobil Corporation
and Lyondell Chemical Corporation,
claiming the companies have added
increasing amounts of MtBE to New
Hampshire's gasoline even though
they knew years ago that it would
contaminate water supplies,
    The state alleges that the manu-
facturers and refiners produced  a
defective product, created a public
nuisance, and violated state environ-
mental  and consumer  protection
laws. The state is asking the court to
hold the companies responsible for
all costs associated  with addressing
the problem, including investigative
and  cleanup  costs, and to assess
monetary penalties.
    Calling MtBE "the Houdini of
pollutants" because the chemical is
water soluble and seems to be able to
escape from underground storage
system tanks and pipes, state Attor-
ney General Peter Heed said at a
news  conference  with Governor
Craig Benson  that "New  Hamp-
shire's groimdwater  and surface
waters are under attack."
    MtBE has been associated with
adverse health effects and can  render
water unpalatable, even at very low
levels. Because it dissolves easily in
water, it travels  faster and  farther
than other gasoline constituents and
is more difficult to find and remove,
making cleanup more expensive. To
protect public health,  the state has
established a health-based standard
of 13 ppb for MtBE, which triggers a
regulatory  cleanup response. State
funding mechanisms cover many of
those  cleanup expenses, but  the
funds are not available for cleanup if
the health-based  standard   is  not
exceeded.
    Approximately 60 percent of the
state's population relies on ground-
water wells for drinking water. As of
2002, MtBE was detected in more than
15 percent of the state's public water
supplies tested statewide. A prelimi-
nary analysis of recent data generated
by a joint NHDES/ U.S.G.S. study of
Rockingham County's public water
systems,  which used relatively low
detection  limits, shows that 41 percent
of the systems tested contain some
level of MtBE,  The  state also esti-
mates, based on studies from other
states, that about 40,000 private wells
in New Hampshire contain some level
of MtBE.
    New  Hampshire has chosen to
remove itself from the federal oxy-
genated fuels program in the absence
of Congressional action on MtBE but
is still awaiting federal approval. For
a copy of the state's lawsuit, go to the
New Hampshire Department of Jus-
tice Web site at uni'w.state.nh.us/nluioj.

-------
Sacramento County and 10
Water Utilities Sue Over
MtBE in California
As  if  the  New  Hampshire  suit
weren't enough, the MtBE pot is
being stirred big-time in California.
The  Sacramento  County  District
Attorney's Office and 10 water utili-
ties are suing the nation's major gaso-
line producers to pay for the cleanup
of MtBE-polluted groundwater. The
lawsuit, filed on October 2, 2003, in
Sacramento  Superior   Court,  is
believed to be the first of its kind to
claim damages for threatening drink-
ing water wells.
   The claimants allege in the law-
suit that the oil companies knew that
MtBE was a risk to drinking water
years before they doubled its volume
in gasoline to meet clean-air require-
ments, yet they went ahead and mar-
keted the product as being better for
the environment. The suit seeks to
obtain  from the oil companies "all
necessary funds to investigate, moni-
tor, prevent, abate, treat, and con-
tain" MtBE pollution of groundwater
in the county and of any wells that
may become contaminated in the
future.  In addition, it seeks to be able
to pursue civil penalties for violation
of the state's laws for protection of
and  safe  disposal  of  hazardous
waste.
   Both this and the New Hamp-
shire legal action come as Congress
debates a "safe harbor," provision in
the federal energy bill, which would
protect the makers, distributors, and
users of MtBE from liability for envi-
ronmental cleanup.  (See related arti-
cle on page 32.)
ARCO/BP to Pay San Diego
S4 Million for UST
Violations
BP West Coast Products, operator of
more than 100 ARCO stations in San
Diego County, agreed on September
29 to pay the county and city of San
Diego $4 million for violations of Cal-
ifornia law regarding the monitoring
and maintenance of USTs. After four
years of  inspections, the city  and
county discovered more than 1,300
violations at  ARCO station, many
involving sensors installed to detect
leaks that "had been tampered with
so that they wouldn't go off."
Father and Son Go to
Prison for UST Violations
in Connecticut
Russell W. Mahler, 77, and his son,
Russell Mahler 11, 43, pleaded guilty
in Connecticut Superior Court to
multiple charges, including the dis-
charge of gasoline into water and fail-
ure to properly close an underground
storage  tank. The father, who has
gone to  prison twice before for ille-
gally dumping toxic waste into land-
fills in New  York  and abandoned
coa! mines in Pennsylvania, pleaded
guilty to five counts and received a
sentence of six years of incarceration
with three years suspended. The son
pleaded guilty  to four counts and
was sentenced to three years of incar-
ceration with two years suspended.
The father was  also ordered  to pay
nearly $498,000 in restitution.
   According to state prosecutors,
the Mahlers  failed  to update and
monitor the  tanks as required by
state law and in some  cases aban-
doned some gas station/convenience
 store sites without proper tank clo-
 sure. Failure to follow the LIST regu-
 lations at one location resulted in the
 discharge of 2,500 gallons of gasoline
 into  the  groundwater  and  put  a
 nearby theatre at risk of exploding.
    After serving a year in federal
 prison for dumping waste, including
 cyanide,  into an  abandoned coal
 mine in Pennsylvania that ultimately
 left a 30-mile slick on the  Susque-
 hanna River, Russell Mahler came to
 Connecticut and  opened a number of
 gas  stations  with  his  son. They
 owned and operated these businesses
 under a  variety of corporate and
 entity names.
    For at least 10 years the Connecti-
 cut Department of Environmental
 Protection  had  initiated   several
 enforcement  actions  against  both
 Mahlers. The  elder Mahler had the
 dubious distinction of being the sub-
 ject for about a page and  a half in the
 book Poisoning for Profit: The Mafia
 and Toxic Waste  in America  by Alan
 Block and Frank Scarpitti. •
                     EM HO UPDATE
  States and EPA Implement
  New UST Program
  Performance Measures
  for FY 2004
  U.S. EPA's Office of Underground
  Storage Tanks (OUST), in coopera-
  tion with the Association of State
  and Territorial Solid Waste Man-
  agement  Officials,  revised  its
  approach for measuring signifi-
  cant aspects of operational compli-
  ance.  The revised measures will
  be used to evaluate the UST pro-
  gram's success in promoting  the
  environmentally safe operation of
  USTs.  A memorandum describing
  the new reporting requirements
  and the criteria for determining
  "significant operational  compli-
  ance" (SOC) was sent to the states
  and U.S. EPA Regions. The memo-
  randum and related documents,
  including matrices, can be seen at
  iiranv.epa.gov/oust/cmplns tc/soc, h t m.
USTfields Pilot Results in
Four New Homes in
Oakland, California
On October 11, 2003, the East Bay
Habitat for Humanity and its  part-
ners welcomed  four  families  to
their new homes at 2662 Fruitvale
Avenue in Oakland. The property
had been a contaminated gas sta-
tion site.  A partnership consisting
of the East Bay Habitat for Human-
ity, the State of California, the City
of Oakland, the Alameda County
Department   of  Environmental
Health, and   U.S. EPA  worked
together to investigate and clean up
petroleum contamination to make
the property ready for reuse as a
site for affordable housing. The
remediation work had been funded
in part by an USTfields Pilot grant.
For   further   information,   see
ivivw.epa.govi'region09.  To find out
more about the U.S.EPA/Habitat
for Humanity partnership,  see
ivwiv. epa, goi'/b rowi i fields/pdf/ha bitat,
pdf.
                                                                                                    29

-------
               •
                me  Browntields  Bear
 2003 Brownfieids Awards
 Include Petroleum-
 Contaminated Sites
 On June 20, 2003, as part of it's on-
 going efforts to promote economic
 revirali/ation while safeguarding
 the environment and public health,
 U.S. EPA announced $73.1 million
 in Brownfieids funds for a variety
 of different grants made available
 from the Small Business Liability
 Relief and Brownfieids Revitaliza-
 tion Act. The 2002 law expanded
 the definition of what's considered
 a Brownfield, so communities may
 now focus on sites contaminated
 with petroleum,  as well as  lands
 scarred by mining. The Act autho-
 rizes up to $250  million in funds
 annually for Brownfieids grants,
 including up to $50 million for the
 assessment and cleanup of low-risk
 petroleum contaminated sites. This
 year, a  total  of  $22,500,000 was
 awarded for Brownfield petroleum
 sites    and   petroleum-related
 research, the first such set of awards
 under the Act. (See Table 1.)
    Brownfieids  are  abandoned,
 idled, or  under-used  industrial
 and  commercial facilities where
 expansion  or  redevelopment  is
 complicated by real or perceived
environmental contamination. The
program provides funding incen-
tives, feasibility tools, and individual
grants up to $1,000,000 to help states,
tribes, communities, and other orga-
nizations prevent assess, safely clean
up, and reuse Brownfieids, For more
information, go  to http://ivww.epa.
gov/brownfieids/,

Proposals for FY 2004
Petroleum Brownfieids
Grants Now Being Accepted
U.S. EPA's Proposal Guidelines for
Brownfieids Assessment, Revolving
Loan Fund, and Cleanup Grants is
now  available   on  the  Web  at
http://umnv,epa,gav/brou'nfie1ds/appli-
cat.htm.  The  guidance will  assist
applicants in preparing  proposals for
assessment,  revolving  loan funds,
and cleanup grants to address Brown-
fields sites, including those impacted
by petroleum. The proposal guide-
lines include specific eligibility and
proposal  ranking criteria. Grant
applications for fiscal year 2004 are
due December 4, 2003. As stipulated
in the 2002 Brownfieids law, 25 per-
cent of the total grant funds must be
made available  for the assessment
and cleanup of petroleum-contami-
nated sites.
       -NEW!
LUST	INDEX
August IBttrVBulletin «1 - March i9«!l>rfcuLL*tJii Mfi
 The NEW LUSTUne
 Index—the  long and
 action-packed story of
 USTs and LUSTs is ONLY
 available on-line.
 To download the
 LUSTLine Index, go to
 www.neiwpcc.org, click
 on Publications, and then
 click on  LUSTLine.
  LUSTUne T-Shirts
    TWO new WACK» designs
    anud tf LUSniat aitomttt. nut > »
     TWO colors... nA and black
  TWO versions... long and short sleeve

          Lang sleeve $17.00
          Short sleeve S1 3.00
          Slass: M. L, X. XXL
   lEI . l^nJitxiki.i rm.r*-v older rrft.prj'.i i.jr [/ S iSj',1.-., •.'.'*: to

           NE1WFCC
    BOOH Mills South. 10(1 Foot al John Strtvl,
        Lowell, MA QJSK-I1Z4
    Tut «78) 323- 7929 • Hj.l: [978) .123-7919
Table 1
2003 PETROLEUM BROWNFIELD SITE AWARDS
REGIONS
1
2
3
4
5
6
7
8
9
10
TOTAL
RESEARCH GRANTS

TOTAL
ASSESSMENT
9
7
4
3
18
3
4
1
6
2
57



CLEANUP
3
3

REVOLVING LOAN FUND
1
1
1

9
3
2

5
5
30



4


1
2
2
12



TOTAL GRANTS
13
11
5
3
31
6
6
2
13
9
99
2


TOTAL FUNDING ($)
2,994,700
2,380,000
835,000
600,000
5,505,925
875,000
1,100,000
950,000
2,925,000
4,115,442
$22,281,067
$ 148,933
$ 70,000
$ 218,933
TOTAL $22,500,000
30

-------
                                                                                    ...
The Exploding  Cell Phone
at  Gas  Stations:  Fact or  Fiction?
by Ben Thomas

     You've  probably  heard  the
     story about the  cell phone
     somewhere that sparked dur-
ing a vehicle fueling and caused a
fire or explosion at  a gas station.
Allegedly, the  cell phone was the
source of a spark that, combining
with a cloud of gas vapors, accumu-
lated  around  the dispenser, and
rocketed the cell user off his/her feet
and into the hospital.
    The scenario sounds plausible,
and most of the folks who tell me the
tale swear they heard it from some
reliable source. But  truth be  told,
there has not been one officially doc-
umented  case  by the  petroleum
industry of a cell-phone induced fire
at a gas station. Both the American
Petroleum Institute (API) and the
Petroleum Equipment Institute (PEI)
have been tracking this issue for sev-
eral years and confirm: There are no
cases of this incident that have been
verified. A spokesman for API told
me rather bluntly, "You would think
that with 11 billion refueling stops
each year and probably at least that
many  cell phones, we would know
about this by now."
   Sorry to burst your bubble but it
seems like the exploding cell phone is
like a  number of excellent but mis-
leading urban legends. Like the Poo-
dle in  the Microwave, the Kentucky
Fried Rat, the Vanishing Hitchhiker,
the Exploding Cell Phone is just one
of those things made popular and
accessible by the Internet. In fact,
when  I typed in "cell phone gas sta-
tion"  into a  search engine on the
Internet, guess what was the first hit?
www.urbanlcgends.com For the full
story,  see http^/itrbank'gends.about.
com/librari/fu'cckli/ faa062399.htm.
   For those who say "yeah, but it
could  happen," I leave you with this
fact: Last year,  the  U.S. Minerals
Management Service investigated an
alleged exploding cell phone by run-
ning a series of complex tests in a lab-
oratory,  trying  to  recreate  the
conditions that caused the accident,
The result? They could not get the
      THE SIGN AT THE DISPENSER:
       To keep you fire-free
        or simply focused?
phone to blow. No matter how hard
they tried and tried, no boom. For
those who want to see the report, go
to http://ivww.mms.goV/safetyalerti/6.
htm.
   So what's up with the signs at the
pumps that say DON'T USE CELL
PHONE? My personal theory is that
the  industry wants people paying
attention  during  fuel up  and not
yakking to the spouse about what's
for dinner. Which is actually pretty
good advice. •

  /vij 1 iit'Hitt? I id* ii Ll^T M')/-.II//I;/X
  iuWHt'S.s lliirtll t't'^itlith'. Hi1 .
  rea> lit\i .it bthurrtas@whidbey.com.
 /1i^  r        www.bentanks.com.
  Q.L.UST.LINE  subscript
  Name.
            ion 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	_^^_

  Mailing Address.	
  G  One-year subscription. $18.00.

  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.
  Si-mi 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 • Justline@neiwpcc.org I
  Comments:	  	 	
                               I www.neiwpcc.org
                                                                                                31

-------
UST and Energy Legislation Ongoing and Likely to Be Merged
        Wii waited until the last pos-
        sible moment to give you
        the scoop on  the  fate of
Congressional UST and comprehen-
sive energy legislation. As of LUST-
Line press time, a few contentious
provisions have held up passage of
the energy bill,  and  we  can't  say
whether they will be resolved in the
near future. Congress  is working
hard to finalize the energy bill before
this session ends (probably in mid-
November). If they do not finish the
bill this session,  they will likely take
it up again when they return after
the new year.
   On October 17, Representative
Paul Gilmore introduced H,R. 3335,
the "Underground  Storage Tank
Compliance Act  of 2003." Simultane-
ously, Congress has  been nearing
completion of an omnibus  energy
bill. Both the Senate and the House
have passed energy bills, and a con-
ference committee is currently craft-
ing a compromise bill to send to each
body for a final vote.
   While both the Senate and House
versions of the energy bill contained a
few provisions directly modifying the
UST  program, the conference com-
mittee plans to  incorporate  a more
comprehensive  UST  bill  into  the
omnibus energy bill. Sources say the
committee intends to include H.R.
3335, or something very close to it, in
the final energy bill. H.R. 3335 con-
tains a number of similarities with S.
195, which the Senate passed in May.
   The issues being debated on the
energy bill that are of concern to the
UST-related community are a renew-
able  fuels mandate,  a repeal of the
oxygenate mandate,  an MtBE ban,
and a liability protection, or "safe har-
bor," provision that gives oil, ethanol,
and  chemical industries immunity
from claims that the fuel additives are
"defective in design or manufacture."
Most state and local  UST regulators
oppose   the   inclusion  of  any
MtBE/ethanol immunity provision in
the final version of the energy bill or
any other pending legislation dealing
with  liability waivers.

Major Provisions  in H.R. 3335
Since Congress  may include  H.R.
3335 in the energy bill, let's look at
some of the major provisions. The bill:
• Authorizes $605 million per year
  from FY04 - FY08.
• Expands eligible uses of the LUST
  Trust Fund to include various com-
  pliance and enforcement activities.
• Requires   on-site   inspections
  within two  years for all USTs that
  have not  been inspected since
  12/22/98.  After  completion  of
  these inspections, the bill requires
  the inspection of  all USTs every
  three years. EPA may grant a one-
  year extension of the first three-
  year period.
• Allows state assurance funds to
  use LUST funds to pay for cleanup
  in  financial hardship cases. Cost
recovery is  prohibited in these
instances. However, any state that
diverts funds from its state fund
for non-UST purposes is no longer
eligible to use LUST funds in this
way. EPA also has the authority to
withdraw approval of a state fund
to be used as a financial responsi-
bility mechanism in State Program
Approval states without having to
withdraw state program approval.
Requires all  UST operators to be
trained, EPA to develop training
guidance, and states to develop
training requirements based on
EPA's guidance. Operators are to
be trained in accordance with the
state requirements, both initially
and again if their facility is deter-
mined to be out of compliance.
Provides specific authority and
authorization for appropriation to
clean up releases of fuel containing
oxygenates.
Makes it illegatto deliver products
into a UST at  a  facility that has
been identified by EPA or a state
as being ineligible for delivery.
EPA and states must develop both
a delivery prohibition roster listing
the ineligible  facilities  and  a
process for installing tamper-proof
"boots" that block the fill pipes for
USTs determined to be ineligible
for delivery. EPA and states would
have one year to develop guidance
documenting the processes and
procedures to implement these
provisions. •
LUST.UNE
New England Interstate Water
Pollution Control Commission
Boott Mills South
100 Foot of John Street
Lowell, MA 01852-1124

-------