United States Solid Waste And EPA510-R-92-703
Environmental Protection Emergency Response September 1987
Agency 5403W
vv-EPA Causes of Release
From UST Systems
Attachments
^Printed on Recycled Paper
-------
so
-------
FINAL REPORT TO U.S. EPA/OUST
ATTACHMENTS
CAUSES OF RELEASE
FROM UST SYSTEMS
EPA CONTRACT: 68-01-7053
SUBCONTRACT: 939-5
WORK ASSIGNMENT: 24
September 30,1987
-------
-------
ATTACHMENT A
QUESTIONS FOR MAJOR COMPANIES
WITH UST SYSTEM MANAGEMENT PROGRAMS
1. UST SYSTEM PERFORMANCE/CAUSES OF RELEASE
o Types, numbers, typical ages;
o Most common source of release (%):
- Overfills and spill;
Tanks;
- Piping. (suction and pressurized)
o Most common causes of failure for each source of release:
Corrosion (external and internal);
- Installation; and
Other.
o What are the sources of the worst (largest or most costly)
releases?
o What procedures are used to assure proper installation?
2. RELEASE DETECTION PROGRAM
o What types of leak detection are in use - both manual and
automatic?
o How many locations are using each technique?
o How did the company decide to use each technique?
o How has each technique been performing? (i.e., operating
problems, false alarms, defects, etc.)?
o How are releases being discovered?
o What fraction of the releases are discovered by each technique?
o What fraction of the releases are discovered by other means?
o How often are releases discovered upon installation of the
device?
o What sizes of releases are discovered .with each technique?
35
-------
ATTACHMENT A
PAGE 2
3. RELEASE CONFIRMATION AND INVESTIGATION
- *
o What "suspected releases" does your company investigate?
o What techniques or procedures are following to confirm release?
o What are typical points of contact and activity with local and
state governments concerning release confirmation?
o How often do "false alarms" occur and how are they resolved?
4. REPAIR OR RETROFITTING OF UST SYSTEMS
o Are repairs allowed and, if so, what types and how is the
quality controlled?
o What types of retrofitting of existing tank systems are
performed?
o What oversight and performance checks are performed on upgraded
systems?
o Do you allow re-use of tanks removed from the ground?
5.
CLOSURE
o What procedures are typically used at closure?
o Is some form of site assessment/testing performed to assure
clean closure?
o What are the result of closure testing?
6.
SPILLS AND OVERFILLS
What are the correct practices and equipment used to prevent.-
spills and overfills?
What is the estimated contribution- of spills and overfills to
releases identified in response program (e.g., frequency,
sizes)?
36
-------
ATTACHMENT B
QUESTIONS FOR LOCAL REGULATORS I
EPA OUST has questions in three general areas of underground storage
tank regulation and enforcement:
o Causes of release;
o Release detection techniques; and
o Program implementation.
Questions in each of these topic areas are identified below.
CAUSES OF RELEASE
Of the total population of tanks systems in the jurisdiction:
o What is the incidence of releases:
o Has the incidence gone down since starting the program;
o How many of the releases are due to overspills;
o How many of each type of tank systems are failingbare steel,
coated steel, CP steel, clad steel, FRP galvanized pipe, CP
pipe, FRP pipe, suction piping, pressurized piping;
o Where are the leaks occurring (i.e., at pipe joints, tank
bottoms, underfill tubes);
o What is the age distribution of the failed tank and piping
systems;
o Can the failures be attributed to improper installation or
repair;
o What basis does the local jurisdiction have for the above
numbersjudgments based upon experience; permit database;
release notification database; corrective action databasecan
these databases be put together to confirm these numbers.
RELEASE DETECTION TECHNIQUES
o What types of leak detection are in'useboth manual and
automatic;
37
-------
QUESTIONS FOR LOCAL REGULATORS
PAGE 2
o How many of each;
o How did the regulator choose limitations on use;
. o
o How did the regulator decide to allow each technique;
o How has each technique been performing; " *
o Are release being discovered with the techniqueat installation
or after;
o What sizes of releases are being found with each technique.
PROGRAM IMPLEMENTATION QUESTIONS
o What are the installation requirementshow does the regulator
insure proper installation;
o What is the phase-in schedule for upgradingis it being met;
o What repairs and retrofits are allowedhow does the regulator
insure proper repair;
o What are the spill and overfill control requirementswhat is
the phase-in schedule;
o What is the Leak Detection phase-in scheduleis it being met;
o What are the Leak Detection reporting and confirmation
procedures;
o What are the closure requirements;
o Are many releases being discovered through the closure process;
o How many people on staff;
o How many inspections are performed;
o How many permits are filed.
38
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Arkansas Best Freight Systems
CITY: Fort Smith, AR
VISIT DATE: June 4, 1987
INTERVIEWER/S: Robin Parker (JEG)
CONTACT: Jim Halladay
DATA COLLECTION METHOD: Telephone contact
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS 3%
TANK FAILURES 25%
PIPE FAILURES 72%
UNDETERMINED
BASIS FOR DATA:
39
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIEWER/S:
CONTACT:
Amoco Oil Co.
Chicago, IL
June 18, 1987
Robin Parker (JEG)
Randall Farmer - Staff Director, Environmental &
Safety
Jay Scrivener - Environmental Specialist
Dennis Strock - Engineering
DATA COLLECTION METHOD: On-site interview
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
PERCENT OF TOTAL
10%
35%
55%
BASTS FOR DATA:
Corporate data
40
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Ashland Oil
CITY: Ashland, Kentucky
VISIT DATE: June 16, 1987
INTERVIEWER/S: Robin Parker (JEG)
CONTACT: Dick Hale
DATA COLLECTION METHOD: Telephone conversation
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS Infrequent
TANK FAILURES 10%
PIPE FAILURES 90% .
UNDETERMINED
BASIS FOR DATA:
Corporate records
41
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIEWER/S:
CONTACT:
Ashland Oil
Ash1and, Kentucky
July 25, 1987
Ramesh Maraj (JEG)
Dick Hale
DATA COLLECTION METHOD: Telephone Conversation
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
PERCENT OF TOTAL
9 releases (8%)
34 releases (92%)
BASIS FOR DATA:
FRP tankage and piping - corporate records.
42
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Austin Underground Storage Program
CITY: Austin, TX
VISIT DATE: March 9 and April 27 - Both visits combined in
report
INTERVIEWER/S: Tom Willard (VERSAR)
Dave O'Brien (USEPA)
CONTACT:
DATA COLLECTION METHOD: On-site interview
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS
TANK FAILURES 24%
PIPE FAILURES 44%
UNDETERMINED
BASIS FOR DATA:
1. Review of local tank testing reports.
2. Analysis of tank test-records conducted by a local vendor.
15% of UST systems failed integrity test.
43
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIEWER/S:
CONTACT:
Boeing
Seattle, WA
June 11, 1987
Robin Parker (JEG)
David Smukowski
DATA COLLECTION METHOD: Telephone interview
CAUSES OF RELEASE
PERCENT OF TOTAL
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
12%
6%
12%
BASIS FOR DATA:
Corporate records (internal study)
44
-------
SUMMARY OF CAUSES OF RELEASE
a,
ORGANIZATION: Environmental Quality Control Board
ป CITY: Broward County, FL
VISIT DATE: April 2, 1987
INTER VIE WER/S: Kim Green (EPA OUST)
Bill Meyers (VERSAR)
Tom Schruben (JEG)
CONTACT: Jim Andrews (BCEQCB)
Greg Balicki (BCEQCB)
Fran Henderson (BCEQCB)
DATA COLLECTION METHOD: On-site interview
SAMPLE SIZE: 14 incident reports
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERSPILLS 46%
TANK FAILURES 23%
PIPE FAILURES 31%
UNDETERMINED
BASIS FOR DATA:
Examination of incident release reports.
2% of UST systems have reported releases.
45
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Department of Environmental Resource Management
CITY: Miami, FL DADE COUNTY
VISIT DATE: April 3, 1987 TRIP I
INTERVIEWER/S: Kim Green (EPA OUST)
Bill Meyers (VERSAR)
Tom Schruben (JEG)
CONTACT: Amando Villanueva (DERM)
DATA COLLECTION METHOD: :On-site interview
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS #1 major cause
TANK FAILURES #3 cause
PIPE FAILURES #2 cause
UNDETERMINED
BASIS FOR DATA:
Impressions of Mr. Amando Villanueva.
46
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Department of Environmental Resource Management
CITY: Miami, FL BADE COUNTY FILES
VISIT DATE: May 13/14, 1987 TRIP II
INTERVIEWER/S: Frank Hicks
Tom Abel
CONTACT: Amando Villanueva (DERM)
DATA COLLECTION METHOD: On-site interview
SAMPLE SIZE: 148 release files (out of possible 215)
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS 45.1% .
TANK FAILURES 25%
PIPE FAILURES 21.5%
UNDETERMINED 8.4%
BASIS FOR DATA:
1. Monthly status report on contamination incidents.
2. Information contained in inspection reports.
47
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIEWER/S:
CONTACT:
Fuel Quality Services
Buford, GA
May 18, 1987
Al Nugent (Hart)
Howard Chesneau, President
DATA COLLECTION METHOD: Telephone data collection
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
PERCENT OF TOTAL
Unknown
20%
80%
BASTS FOR DATA:
Personal recollections
48
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Circle K Stores
CITY: Phoenix, AZ
VISIT DATE: July 7, 1987
INTERVIEWER/S: Robin Parker (JEG)
CONTACT: Don Esperson
DATA COLLECTION METHOD: Telephone contact
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS 0%
TANK FAILURES 1%
PIPE FAILURES 99%
UNDETERMINED
BASIS FOR DATA:
Impressions of Don Esperson who has approximately 20 years of
experience in the UST industry.
49
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIE>VER/S:
CONTACT:
Consolidated Freightways
Menlo Park, CA
August 1, 1987
Mary Ann Parker (JEG)
Sue Black
DATA COLLECTION METHOD: Telephone contact
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
PERCENT OF TOTAL
5%
40%
55%
BASIS FOR DATA:
Corporate files
50
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Corrosion Research Centre & Ministry of the
Environment
CITY: Denmark
VISIT DATE: April 2nd through April 3rd, 1987
INTERVIEWER/S: Phil Stapleton (D&M)
CONTACT: Mr. Finn Yding
DATA COLLECTION METHOD: On-site interview
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS . 65%
TANK FAILURES 14%
PIPE FAILURES 21%
UNDETERMINED
BASTS FOR DATA:
Regulatory records
51
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIEWER/S:
CONTACT:
Fiberglass Petroleum Tank and Pipe Institute
McLean, VA
July 14, 1987
Robin Parker (JEG)
Ed Nieshoff
DATA COLLECTION METHOD: Meeting for comment presentation -. /:
' ' *" -'
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
' '.-^ '^:..vJ*fJ"S-' -;4*:" ' '' .J./'* '""
PERCENT OF TOTAL
Not addressed
0.25% of Population
No Data
BASIS FOR DATA:
FRP Tankage data from Owens Corning files, 1978 & 1984.
52
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Murphy Oil
CITY: El Dorado, AR
VISIT DATE: May 27, 1987
INTERVIEWER/S: Robin Parker (JEG)
CONTACT: Dave Petty
DATA COLLECTION METHOD: On-site interview
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS < 1%
TANK FAILURES 50 - 60%
PIPE FAILURES 39-44%
UNDETERMINED
BASIS FOR DATA:
Impressions of Dave Petty.
53
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIEWER/S:
CONTACT:
Ministry of Consumer & Customer Relations
Fuel Safety Branch
Ontario, Canada
May 20th through May 22nd, 1987
Robin Parker (JEG)
Bill Drake (FSB)
Paul Beaumont (FSB)
DATA COLLECTION METHOD: On-site interview
SAMPLE SIZE: 50 documented releases in 1986 (.16%)*
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
PERCENT OF TOTAL
44%
56%
BASTS FOR DATA;
Examination of incident release reports.
* Probably represents a steady-state condition; incidence has
decreased with program phase-in.
54
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Ministry of the Environment
CITY: Ontario, Canada
VISIT DATE: July 28, 1987
INTER VIE WER/S: Ramesh Maraj (JEG)
CONTACT: Ray Valentine
DATA COLLECTION METHOD: Telephone conversation
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS Unknown
TANK FAILURES 9 releases
PIPE FAILURES 2 releases
UNDETERMINED
BASIS FOR DATA:
1983 Records
55
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Petroleum Equipment Institute
CITY: Dallas/Ft. Worth Airport
VISIT DATE: July 8, 1987
INTERVIEWER/S: Robin Parker (JEG)
David O'Brien (EPA)
CONTACT: Bob Renkes
Sandra Parkison
DATA COLLECTION METHOD: Expert Panel
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS 1-4%
TANK FAILURES 20-30%
PIPE FAILURES 70 - 80%
UNDETERMINED
BASIS FOR DATA:
Consensus opinion of expert group.
56
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIEWER/S:
PIECO (Petroleum & Industrial Equipment)
Hialeah, FL
May 20, 1987
Frank Hicks
Tom Abel
CONTACT: William F. Wynne
DATA COLLECTION METHOD: On-site interview
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
PERCENT OF TOTAL
Majority of releases
' 3rd major cause of release
2nd major cause of release
BASIS FOR DATA:
Impressions of William Wynne, President of PIECO.
57
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTER VIEWER/S:
Ryder Truck.Rental, Inc.
Miami, FL
May 18, 1987
Frank Hicks (JEG)
Tom Abel (JEG)
CONTACT: James R. Barr, Manager of Environmental Services
DATA COLLECTION METHOD: On-site interview
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
PERCENT OF TOTAL
50%
5%
45%
BASIS FOR DATA;
Impressions of James Barr, Manage of Environmental Services, Properties
and Construction.
58
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIEWER/S:
CONTACT:
DATA COLLECTION METHOD: On-site interview
Department of Health Services
San Diego, CA
June 1, 1987
VERSAR Personnel
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
PERCENT OF TOTAL
50%
BASIS FOR DATA:
Field experience of inspectors.
59
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Regional Water Quality Control Board
CITY: San Francisco, CA
VISIT DATE: March 27, 1987
INTERVIEWER/S: Tom Schruben (JEG)
CONTACT: Peter Johnson (SFBRWQCB)
Tom Callaghan (SFBRWQCB)
Steve Morse (SFBRWQCB)
DATA COLLECTION METHOD: On-site interview
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
* No information obtained during interview concerning causes of release.
Referred to files at Sunnyvale and San Jose.
BASIS FOR DATA:
60
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: San Jose City Fire Prevention Bureau
CITY: San Jose, CA
VISIT DATE: April 3, 1987
INTER VIE WER/S: Tom Schruben (JEG)
CONTACT: . Joe Afong (HMD)
DATA COLLECTION METHOD: On-site visit and interview
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS 25 - 33%
TANK FAILURES 67 - 75%
PIPE FAILURES 67 - 75%
UNDETERMINED
BASIS FOR DATA:
Personal recollections
61
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: The Southland Corporation
CITY: Dallas, TX
VISIT DATE: July 9, 1987
INTERVIEWER/S: Robin Parker (JEG)
CONTACT: Hal Miller
DATA COLLECTION METHOD: Telephone conversation and on-site interview
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS Unknown
TANK FAILURES 23%
PIPE FAILURES 39%
UNDETERMINED
BASIS FOR DATA:
Corporate records (for FRP)
62
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIEWER/S:
CONTACT:
Suffolk County Health Department
Long Island, NY
March 4, 1987
David O'Brien (EPA)
Jim Pirn
DATA COLLECTION METHOD: On-site interview
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
PERCENT OF TOTAL
Not addressed
10%
90%
BASIS FOR DATA:
Suffolk County data
63
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIEWER/S:
Department of Public Safety
Sunnyvale, CA TRIP I
March 27, 1987
Karen Reed (EPA OUST)
Eric Yunker (EPA Region 9)
Bill Meyers (VERSAR)
Tom Schruben (JEG)
CONTACT: Ruben Grijalva (DPS)
DATA COLLECTION METHOD: On-site interview
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
PERCENT OF TOTAL
Majority of incidents
Substantial cause of release
Substantial cause of release
BASIS FOR DATA:
Impressions of Ruben Grijalva.
18% of UST systems have reported releases since 1983.
64
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: Fire Department
CITY: San Jose, CA TRIP II
VISIT DATE: April 26, 1987
INTERVIEWER/S: Bill Meyers (VERSAR)
CONTACT: Ruben Grijalva
DATA COLLECTION METHOD: On-site interview
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS Less than 2%
TANK FAILURES 60%
PIPE FAILURES 38%
UNDETERMINED 52 cases
BASIS FOR DATA:
124 incident release reports, 41 with reported cause of release.
65
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION:
CITY:
VISIT DATE:
INTERVIEWER/S:
CONTACT:
City o'f Sunnyvale - Department of Public Safety
Sunnyvale, CA TRIP II
April 26, 1987
Bill Meyers (VERSAR)
Ruben Grijalva
DATA COLLECTION METHOD: On-site interview
CAUSES OF RELEASE
SPILLS/OVERFILLS
TANK FAILURES
PIPE FAILURES
UNDETERMINED
PERCENT OF TOTAL
N.A.
44%
56%
27 cases
BASIS FOR DATA:
45 incident release reports, 18 had reported cause of release.
66
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: VARIOUS
CITY: Sunnyvale, CA
VISIT DATE: May 22, 1987
INTERVIEWER/S: Bill Meyers (VERSAR)
CONTACT: . Vendors of various organizations
DATA COLLECTION METHOD: Telephone interviews, TRIP II
SAMPLE SIZE: 7 vendors
CAUSES OF RELEASE - VENDOR INFORMATION
Vendor 1: Estimated 40% of releases due to pipes, 20 - 30% of releases due
to tanks.
Vendor 2: Majority of releases due to pipes'(if steel piping).
If fiberglass piping, more tank leaks observed.
Some releases caused by overfills.
Vendor 3: Majority of releases occur from tanks.
Vendor 4: 2/3 of releases are associated with piping, especially the
fittings.
Vendor 5: Majority of releases along the piping at unions and fittings.
Vendor 6: Most leaks occur in the piping and secondarily at the vent pipe.
Vendor 7: Releases evenly distributed between piping and tanks.
tr
BASIS FOR DATA:
* Impressions of vendors who manufacture, install and repair tanks.
67
-------
SUMMARY OF CAUSES OF RELEASE
ORGANIZATION: State of Virginia (Pollution Response Office)
CITY: Richmond, VA
VISIT DATE: August 5, 1987
INTERVIEWER/S: Elaine Strass (JEG)
CONTACT: Mary Ann Sykes
DATA COLLECTION METHOD: On-site visit
CAUSES OF RELEASE PERCENT OF TOTAL
SPILLS/OVERFILLS 12%
TANK FAILURES 75%
PIPE FAILURES 13%
UNDETERMINED
BASIS FOR DATA;
State of Virginia files
68
-------
TRIP REPORT SUMMARY
PIECO
May 20, 1987
Pieco supplies and installs equipment for service stations.
Overfills and spills are felt to be the major cause of releases.
Piping leaks are more prevalent than tank leakage. An estimated 85
percent of tank leakage is related to pipe fittings or connectors.
Corrosion is a major problem and is thought to be a result of bad backfill
or locating tanks near power plants. The worst releases are overfilling
and long term, undetected leaks, which produce a contamination problem.
Proper installation is felt to be a major factor in preventing future
releases. It was suggested that from the time the tank is designed to the
time it is put in service it should be inspected by certified and
professional people. Electronic gauging systems are preferred over
automatic release detection equipment, which has little value due to lack
of maintenance and operator attention.
Monitor wells are the most prevalent means of release detection.
Additional wells or excavation are used to confirm releases. Repair of
pumps and lines is accomplished but fiberglass coating of existing tanks
was frowned upon. Tanks are removed in 99 percent of closures; otherwise
they are filled with sand.
The protection system for spills and overfills was a "double overfill
protection system" consisting of a ball float that stopped filling at 95-
98% full and a containment system at the manway.
69
-------
TRIP REPORT SUMMARY
SAN DIEGO FIRE MARSHALL
SAN DIEGO, CALIFORNIA
MAY 19, 1987
Modifications in the original state regulations have eliminated an
exemption from secondary containment for new installations of pressurized *
product lines. The regulatory personnel in the county are in agreement
that pressurized product lines require a means of both detecting and
containing release. They pointed out that a number of severe incidents
were caused by a pressurized product line that failed.
There is a evidence that the San Diego program has succeeded in
drastically reducing the number of overfill incidents associated with
tanker re-fueling operations. The county imposes a $1000 fine for either
owners or delivery personnel that fail to report an overfill incident.
The air quality regulatory personnel imposed these fines through their
field monitoring of vapor recovery operations. As a result of the
enforcement effort, vendors now fire any employee that fails to report an
overfill incident, and will also dismiss any employee if they are involved
in more than one overfill incident. This regulatory enforcement approach
coupled with the cooperation of the bulk storage and delivery facilities
have led to only 6 overfill incidents reported in 1986 out of over 25,000
deliveries.
o
Ground water and vadose zone monitoring systems have not been
permitted for use in San Diego County. In order for a new installation to
70
-------
be granted an operation permit that employs either of these methods the
county requires that the owner provide evidence that supports the claim
that they will be able to effectively detect unauthorized release on-site.
This evidence must include a determination of the soil characteristics,
background contamination, and ground water flow. The regulated community
has come to the conclusion that the cost of designing and installing such
systems exceeds the cost of installing a double-walled system. Ground
water monitoring may, however, be approved for some chemical tank systems
in the future that are not currently permitted to use inventory
monitoring. The use of ground water monitoring will then only be
permitted in non-beneficial use areas.
The county has proposed a modification to the California Legislature
that would establish a state-wide clearinghouse to evaluate and approve
for use equipment testing procedures and devices associated with the
storage of hazardous substances underground. The regulators in the county
stated that such a clearing house would remove the burden, form the local
enforcement agency, of evaluating a steady stream of new devices and
enable qualified experts in the area of leak detection technology to
assess the reliability and effectiveness of the devices.
When the program began in April 1984, emphasis was placed on the
identification and investigation of suspected releases. As of June 1987,
the office had an active file of 520 cases where a release is suspected.
Of those active cases, the initial means of identification were classified
as: soil contamination, failed precision test, or complaint. The soil
71
-------
contamination incidents, which accounts for 276 or 53% of the active
cases, were commonly identified through tank removal or replacement.
Failed precision test accounted for 139 or 27% of the active cases. The
investigators indicated that most of the suspected releases were
determined to involved small quantities of lost product that can usually
be corrected by removing the contaminated soil.
The degree of remediation required is determined on site by site
basis. The general scheme is to allow no more than 1000 ppm total
petroleum hydrocarbons in the soil in non-beneficial use areas, and no
greater than 100 ppm in beneficial use areas.
72
-------
TRIP REPORT SUMMARY
SAN FRANCISCO BAY j
March 27, 1987
In California, the regional water quality control boards are
responsible for groundwater corrective actions. Santa Clara County is
thought to have the best fuel release detection and corrective action
program. The cause of release is categorized into fuel releases or
hazardous substance release.
The San Francisco Bay Regional Water Quality Control Board (SFBRWQCB)
tracks fuel releases but not causes. Approximately 40 percent of the 1249
reported underground fuel leak cases are in Santa Clara County.
Considering the region's population distribution, this percentage is
higher than expected.
Guidelines have been established for addressing fuel leaks and the
SFBRWQCB is working with the localities to implement the program. A
ranking system for the severity of the releases helps the localities put
their limited resources into the cases that pose the greatest threat to
the environment.
Hazardous substance releases are being tracked separately from fuel
releases and the corrective action activities are not being delegated to
the local water districts as are the fuel leak groups.
Two databases-are used for cause of release analysis. The first
contains materials of construction, corrosion protection, release
detection techniques, and type of release. The second contains the site
description and status report.
The progress of the corrective action programs is being limited due
to current staffing levels. However, they have developed corrective
action tracking systems for hazardous substance releases and fuel
releases. The group with hazardous substance has initiated a number of
corrective actions and has closed some cases. The group with fuel
releases is still developing its corrective action program in cooperation
with local agencies.
73
-------
TRIP REPORT SUMMARY
SUNNYVALE, CALIFORNIA TRIP I
MARCH 27, 1987
Overall management of the Sunnyvale UST Program is placed in the
Department of Public Safety, with the State Board of Water Resources
assuming the lead for guaranteeing the proper installation of monitoring
devices and any necessary remedial actions that may be required for
releases of hazardous waste and/or chemicals. All sites are required to
have an operating license issued by The Department of Public Safety. The
department reviews and approves both the location of the monitoring wells
and any device used for detection and/or containment.
Sunnyvale has a total of approximately 575 tanks. Since 1983, 102
releases have been reported, with the majority due to overfills and
spills. Wrapped-steel tanks make up 70% of the tank population and 90% of
tank failures come from this group. Most of the failures are in systems
that are over 20 years old. Only one double-walled tank has failed. Most
of the piping releases occur in the pressurized FRP piping, due to piping
glue joint failure.
All new UST's are required to have secondary containment on piping
and the tank. Existing tanks are required to have monitoring wells or
automatic inventory control (with periodic precision testing), 518 tanks ,
have monitoring wells.
74
-------
Most of the releases are discovered through precision testing and a
number of previously unknown releases are being discovered.
Approximately 70% of the facilities are in compliance with the
monitoring regulation. Repairs are allowed only if the tank can retain
its U.L. listing. Existing tanks are allowed to remain in use
indefinitely (with monitoring). Tank closures are inspected and
approximately four closures per month are performed. Abandoning in place
is not allowed in Sunnyvale.
Whenever a potential release is discovered, the facility is required
to either perform a precision test or pull the tank system. If site
contamination is observed, a form letter is sent to the facility requiring
them to initiate a remedial action'.
75
-------
TRIP REPORT SUMMARY
BROVVARD COUNTY, FLORIDA
April 2, 1987
Broward County UST Program is governed by state and county
regulations. The bulk of investigation responsibilities is directed to
the County Environment Quality Control Board (EQCB).
A rather prolific sand formation, the Biscayne Aquifer, lies beneath
the county at depths ranging from one to thirty feet. In order to protect
the Aquifer, the county has calculated a 270-day flow radius about their
in place well fields.
There are 1,050 permitted facilities that have a combined total of
3,500 to 4,000 underground storage tanks (UST). An additional 500 tanks
are estimated to be at unpermitted facilities. The tank failures seen by
the EQCB are only at older installations. Only one new piping system
failed due to poor installation. Broward County investigated 46 release
cases; in 14 cases, cause of release was determined. The greater percent
of releases was due to overfills and spills.
All facilities, regardless of location, require groundwater
monitoring wells. Secondary containment and a continuous monitoring
system is required for the sites within the zone of influence.
76
-------
Piping system leak detectors (PSLD) are required of all UST and have
proved to be effective. The majority of the releases are discovered at
the time of monitoring well installation and when a tank system is closed.
a
Besides the groundwater monitoring and PSLD, Broward County is also
using overflow protection. All three of these systems have been installed
at all facilities within three to six months after license issuance.
Pipe repairs are allowed but lining the tank is not. Abandonment of
tank in place is permitted after properly following regulations. Fifteen
to twenty releases have been found out of 300 closures.
The county realizes that their program is not perfect but they also
recognize its good features. The county UST program is staffed by three
groups: 1) the industrial waste section; 2) the technical support
section; and 3) the enforcement section. These groups meet with the state
staff on a monthly basis to divide release cases. Another feature about
the UST program is the preparation of as-built drawings by professional
engineers which helps to assure improved installations and has eased the
permit evaluation burden.
77
-------
TRIP REPORT SUMMARY
BADE COUNTY, FLORIDA
April 3, 1987
Dade County Underground Storage Tank (UST) Program is governed by two
sets of regulations those by the state and those by the county. Unlike-
Broward, Dade County does not have a cooperative relationship with the
state to share release investigations.
Also, in contrast to Broward County, Dade County's Biscayne Aquifer
(a prolific sand formation that lies beneath the county) lies at very
shallow depths, usually five to seven feet and often more shallow.
There are 1,405 permitted facilities in Dade County and an estimated
4,000 to 6,000 tanks at these facilities. Since 1983, 600 of the 1,405
facilities have been inspected and 215 releases have been discovered.
They expect to find 200 to 300 more cases as they inspect more facilities.
The majority of the releases are due to overfills and spills. The second
major.cause of release is loose piping and the third is corrosion of bare
steel tanks. Sometimes FRP tanks fail at installation or from gauging
stick punctures.
Dade County has a fairly complex approach to protecting its water
supply. They divide the Aquifer into four zones according to distance
from their currently in-place pumping well. The. requirements become more
stringent as the zone gets closer to the well field. All zones require
78
-------
continuous monitoring devices while only some require secondary
containment. >
Secondary containment provision allows systems ranging from synthetic
"skirts" to double walled tanks. Cathodic protection for metal tanks is
required but cannot be substituted for secondary containment and release
detection systems. A Florida PE must approve all systems and should be on
site when the system is constructed. Both the containment tank system and
the monitoring wells .must have as-built drawings in the final submission.
For sensitive areas, in addition to the above requirements, it is
required that existing wells be bailed and checked for sheen.
The variety of release detecting devices that are used include: 1)
monitoring devices, which have detected approximately 90% of the releases
discovered; 2) manual groundwater monitoring; 3) automatic groundwater
monitoring (pollutant); 4) soluble cable-type detectors; and 5) inventory
79
-------
controls which has been more useful in determining the volume lost than in
discovering releases.
The County's program implementation includes: 1) modifying the
secondary containment and monitoring retirements to a four-tiered system;
2) repair of tank systems, including lining; 3) development of overfill
and spill control requirements; 4) monitoring wells; and 5) requirement to
report releases within four hours.
In-place abandonment of tanks is allowed but it is required that
monitoring wells be placed around the tank and checked for sheen. If the
tank is to be pulled, it is checked for free floating product in the tank
hole.
It is felt that the best feature of Dade's UST program is the
extensive communication with UST owners and operators and that the best
way to implement a program is on an local level.
80
-------
ONTARIO PROVINCE, CANADA
MINISTRY OF CONSUMER AND CUSTOMER RELATIONS
FUEL SAFETY BRANCH
SUMMARY
Ontario Province has a successful regulatory program for underground
fuel storage tanks which was initiated in 1974 and presently regulates
30,000 tanks at some 8,000 locations monitored by 33 inspectors.
Voluntary compliance is working extremely well and those subject to
regulation are seeking more, not less regulation. While the number of
locations has decreased from 13,5000 to 8,000, the decline appears to be
based on economic reasons other than regulatory costs. Maximum
environmental protection is obtained with standards less than proposed by
USEPA. Cost of implementation is not oppressive to small businesses.
A protracted 15 year phase-in of regulations of pre-regulation tanks
has not produced any appreciable environmental impact. Simple upgrading
by addition of cathodic protection and interior lining appears to be
working successfully. FSB estimates 72% of pre-regulation tanks have been
retrofitted.
Utilization of existing association standards for equipment and
installation functiOons as well as federally instituted standards.
Release incidents on tanks and piping covered only 0.16% of the tanks
under regulation. Post regulation release incidents are virtually all
piping failures; post regulation tank failures are attributable to faulty
and improper installation rather than corrosion. Of the reported
releases, 34% resulted in off-site migration of liquid or vapor. The
majority of releases were site confined without secondary containment and
required disposal of an average of 10 cubic yards of soil, 90% of releases
were owner reported and tank releases are 95% from pre-regulation tanks
and detected by manual DIPS. FSB data indicates that reported release
incidents have peaked and appear to be on a downward curve.
81
-------
TRIP REPORT SUMMARY
CITY OF AUSTIN
AUSTIN, TEXAS
MARCH 9, 1987
Austin estimates that close to 90% of the registered facilities are
in compliance with their regulations. The apparent success of the
Austin UST program is due to a large degree on their inspection
program that involves witnessing each tank test, closure and repair
activity. The tracking systems used, which relies on well developed
concise forms, supplies documentation on the case history or each
facility.
Two independent research efforts, that used local tank testing and
repairs records as the data sources, found that most of the problems
identified were not related to the tank itself. One study, conducted
by the Austin Fire Department, examined 197 tank system test reports.
Only 30, or 15%, did not pass the integrity test. From a review of
inspectors' field notes, repair records and tank testers comments it
was found that 5 of the 30 failed systems were attributed to
corrosion of the tank surface or structural failure of the tank.
Structural failure of the tank, was interpreted for this study, as a
loss of structural integrity, for example breaks in weld seams. The
problem area identified for 25 of the 30 failed systems were
attributed to : vent line leaks, leaking pipe unions, loose fittings
on top of the tank, or corrosion of the piping. The results of this
study were confirmed through the field experience of the Austin
inspectors. Versar interviewed the principal investigator of the
study and confirmed the completeness of the information sources used
and the procedures followed to obtain these findings.
A local vendor, that services the Texas area who has performed more
than 5000 tank tests since 1981, conducted a similar analysis of 483
tank test records. He found that of 217 problems identified by tank
testing, only 13, or 8%, had a problem with the tank (other than
loose fittings). According to this study, 37% of the repairs that
were identified through tank testing were related to the product
line. The specific leak catagories included under this heading were:
pipe, union, and flex hose leaks. The next most common problem area
was identified as tank fittings that represent 33% of the repairs
required. These findings were corroborated by interviews with other
vendors that service the Austin area.
Through interview with the Austin regulatory personnel, it was
determined that tank system closures generally occur for one of three-
reasons: 1) the tank system has been abandoned or taken out of
service,, 2) the tank system failed a tightness test and cannot
identify the problem, or 3) the parent company removes the existing
tank system as part of a tank replacement program.
82
-------
A review of the program file that included both tank testing and
closure reports indicate that most of the problems identified were
not with the tank, but rather with the tank fittings, vents, and
product lines. For example of 41 closure records that had comments
related to the condition of the tank, only 7 indicated the presence
of holes or perforations on the surface.
83
-------
TRIP REPORT
THE SOUTHLAND CORPORATION
July 9, 1987
The visit was a follow-up on telephone conversations between Messrs.
Lynch and Miller of Southland and Dave O'Brien of USEPA documented in Dave
O'Brien's memo to the file on April 7, 1987. The main area of concern was
in FRP tankage. Mr. Miller indicated that out of a population of some
3,000 tanks, installed by Southland, only 30 had failed. Some FRP tanks
obtained by Southland in a buy-out in Florida had a high rate of failure
which is traceable to one installation contractor who had failed to follow
proper installation procedures. FRP performance is 95%+ dependent on
proper installation.
84
-------
TRIP REPORT SUMMARY
RYDER TRUCK RENTAL, INC.
May 18, 1987
Ryder has 1,200 facilities with an estimated 6,000 tanks.
Approximately 50% of the releases are due to overfills_and spills, Jok
pipes account for 45% of releases with the other 5% being attributed to
leaking tanks. The principle cause of line failures is corrosion and it
is felt to be the most severe problem.
All piping systems throughout the company have "Red Jacket" leak
detectors^ In the preferred installations, tanks have double walls with a
space between the tank walls that is partially filled with brine. A level
detector will inform the owner if either tank wall is leaking. Inventory
control is also used to locate releases but monitor wells were not thought
of favorably.
After a suspected release is reported, an initial investigation is
made Precision tests of tanks and lines are made and if the release is
confirmed, the state government, the local fire department and the local
environmental program, as applicable, are notified.
Tanks are never reused. The only retrofits that are used are to add
cathodic protection or overfill protection. Tanks are removed from the
ground only if in an environmentally sensitive area, or in the event that
the tank has evidenced a past problem. If a tank is removed, soil or
groundwater samples are taken.
85
-------
TRIP REPORT SUMMARY
AMOCO OIL COMPANY
JUNE 18, 1987
Amoco oil company operates 4,675 outlets with a tank census of 21,500
tanks, 20,000 which are steel. .An internal tank retrofit program, Amoco
Leak Assessment and Risk Management (ALARM), was initiated in 1984 for
tanks 10 years of age or older to reduce risk and liability exposure.
Since program inception, reported releases have been in the 40-50 range
annually (0.2% of tank population).
Major causes of release are 90% from tanks and piping and 10% from
overfills. Response to the major causes is accommodate by the
retrofit/replacement aspects of the ALARM program.
Release detection is accomplished through inventory reconciliations
and monitoring wells. Suspect releases are approached through API's
recommendations.
Amoco believes precision test methods cannot be accurate and
repeatable below 1.0 gph. Further, the Rogers age and soil correlation is
not accurate; internal formula have twice the accuracy.
Amoco supports the UST regulation by EPA but fears the specifitivity
of the pending state regulations.
86
-------
TRIP REPORT SUMMARY
MURPHY OIL COMPANY
MAY 27, 1987
Murphy Oil operates 250 outlets with a tank census approaching 800
tanks. An internal program is in place to replace all tankage with new
FRP units; the program has slowed to determine what EPA's final regulation
will require.
Causes of release are estimated as: Tanks 50 - 60%, Piping 39 - 44%,
and Overfills 1% or less. Only 3 releases have been reported in the last
year. Release detection is from manual inventory reconciliations.
87
-------
TRIP REPORT SUMMARY
SAN JOSE CITY
April 3, 1987
Responsibility for implementing the underground storage tank (UST)
regulation in San Jose is with the fire department but is shared with the
state water board, department of health services and the Santa Clara water
district.
San Jose considers their entire aquifer to be equally valuable,
regardless of the distance from existing wells. Implementation of the UST-
regulation is going slow due to: (1) lack of resource; (2) policy of
requiring the installation of monitoring devices only after city
inspection; and (3) responsibility for implementation of the program is
too divided.
San Jose has an estimated total of 1,300 to 1,500 tanks in
approximately 800 locations. Since 1983, there have been 250 reported
releases. Release incidents are increasing and it is felt that the
increases are partially due to increasing city enforcement and a large
number of old UST. Approximately one fourth to one third of releases were
due to overfills. Other releases are attributed to 10-15 piping failures
per year. These are largely bare steel, galvanized steel and FRP pipe.
Approximately 30 steel tanks have been repaired. New tankage installed
are double wall tanks made of FRP, clad steel or coated and cathodically
protected steel. Three of the new FRP tanks have failed.
Existing tanks are required to have either monitoring wells or
automatic inventory control (with periodic precision testing). Vadose
zone monitoring is required unless the site can demonstrate that their
tanks are in the groundwater at all times. Vadose zone systems consist of
at least one well in the backfill near each gasoline tank, two for each
diesel tank, one for each island, and one or more for the piping. The
installation must be certified by a registered geologist. Clayey-
material, San Jose'.s primary soil composition, could have a detrimental
effect on the functioning of the Vadose system.
Abandonment of a tank in place is allowed. When a tank is removed, a
soil sample must be taken from the bottom of the hole and tested for
hydrocarbon content.
San Jose has been unable to develop and implement a data management
system because of the lack of resources. The shortage has also prevented
the program from being fully operational.
88
-------
TRIP REPORT SUMMARY
PEI INSTALLERS
DALLAS, TEXAS
JULY 7, 1987
The meeting with the PEI installers resulted in several important
findings:
o The experienced and knowledgeable installers believe state
licensure would be a giant step in the right direction. Most
installation problems must be solved by the installers
themselves,
o Corrosion is the major release cause with the tanks; installer
errors or accidents with piping.
o A build-up of poor attitudes and the survival of bad practices
must be overcome to solve the installation problem.
o Great progress has been made through the use of interim
prohibition tanks. (They confirmed the 0.5% failure rate of FRP
tanks.) However, piping is more complicated and less forgiving
of mistakes that are human and unavoidable to a large extent.
o Pressurized piping is more of a threat and they recommend
pressure leak detectors be installed as quickly as possible on
existing tanks.
o The influence of corrosion will vary in different parts of the
country. The use of clean, sandy backfills appears to eliminate
the localized (or pitting) corrosion phenomenon, but not
generalized corrosion. All corrosion holes do not leakmany
get plugged up by the corrosion by-products or surrounding dirt.
o Failures in delivery piping are inevitably going to continue
because of human control problems. Annual testing and pressure
detectors are needed. More than 70% of volume releases will be
thereby avoided in the future.
o Overfill shut-off is the best approach, but not yet widely
available in this country.
o Double-walled tanks and piping are good ideas, but there is not
much experience with them and most people will not voluntarily
buy it.
89
-------
TRIP REPORT SUMMARY
STATE OF VIRGINIA
August 5, 1987
The Pollution Response Office of the Virginia Water Control Board
(VWCB) maintains three data sources: (1) a telephone log book of
pollution complaints received by phone at the VWCB; (2) monthly
investigative reports of pollution complaints received by the VWCB; and
(3) a computerized database of oil spill incidents from June 1985 to the
present.
From July 1986 to June 1987, VWCB's database has 160 UST-related
pollution reports on record. Half of the reported cases were concerning
LUST and another twenty percent involved line leaks. Unlike the telephone
log or monthly report, all of the incidents on database were investigated.
There is no computer database set up for data collected earlier than July
1986. Information for that time period would have to be derived from the
telephone logs and monthly reports.
90
-------
TELEPHONE REPORTS SUMMARY
FRED C. HART j
RESULTS OF TELEPHONE DATA COLLECTION ON CAUSES OF RELEASE
Telephone calls were made to ascertain whether documentation existed
on the causes of releases from underground storage tanks containing
petroleum products, and to determine if the documentation could be made
available to EPA. The calls also attempted to qualitatively prioritize
the major causes of releases in terms of their frequency of occurrence.
The following 7 persons were called, 6 representing the transportation
industry and one representing a regulatory agency.
o Richard Cross
Commercial Carrier Journal
o John Java
Yellow Freight System
o Don Dawson
Roadway Express
o Bill Scherrer
Consolidated Freightways
o Captain John Hall
Los Angeles City Fire Department
o Bill Tracy
American Trucking Association
Maintenance Council
o Howard Chesneau
Fuel Quality Services
The calls identified piping problems as the most common cause of
releases, with installation being the second most identified cause. Most
respondents admitted their information might not be able to accurately
identify whether the piping failed due to faulty material or design or
failed due to installation errors such as improper backfill or improper
protection from overhead loads.
Most respondents were hesitant to say that they actually had well-
documented records which could be used to identify causes of release.
However, the following persons agreed to discuss the matter further with
EPA and agreed to potentially give EPA access to their records (all
emphasized the importance of EPA specifying exactly what information was
required):
o John Java
Yellow Freight Systems
91
-------
o Bill Scherrer
Consolidated Freightways
o Captain Hohn Hall
Los Angeles City Fire Department
Additional persons were also identified who might have records on the*
causes of releases. Most of these additional references were in the
regulatory agencies who receive reports or releases as they occur. Calls
to these persons will be made when we are authorized by MRI and EPA. The *
additional contacts which we recommend be made are:
o Harry Fink
ABF
o Karl Sjoberg
Los Angeles County Engineer
o Hank Yacomb or Hosh Workman
Los Angeles Regional Water Quality Control Board
o Bob Flesher
AGA Gas
92
-------
93
-------
-------
IDENTIFIED LERKS (STEEL TflNKS)
LBflKINQ TflfKS
LBKINQ FITTINGS
LEfiKINS MEWTS
Percent
to the
INSTflLLED TflNK flGE
of total tanks tested by installed age bracket, 100% corresponds
total number of tanks tested in that catagory. ( See graph below )
TOTflL STEEL TflNKS TESTED
0ป ^ 0
-J
-------
-------
SERVICE STATION TESTING INC.
L.U. S.T. SURVEY
"COMPARISON OF FIBERGLASS TANK TIGHTNESS RESULTS BY AGE OF TANK"
INSTALLED
TANK
AGE
NEW
1
2
3
4
5
6
7
8
9
0
t
2
3
4
UNKNOWN
TIGHT
t!
118
15
13
12
8
10
3
1
5
1
1
1
6
3
6
16
SYSTEMS * I
%
92.2%
93.8%
92.9%
92.3%
72.7%
100.0%
1 00. O%
100.O%
83.3%
100.0%
160.0%
160.6%
6.6%
160.6%
160.6%
160.6%
-EAKIN
tt
6
6
6
6
0
0
O
0
0
0
6
6
6
6
6
6
G TANKS
*
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
LEAKING
tt
9
0
0
1
0
0
O
0
0
0
0
6
6
0
0
6
FITTINGS
*
7.0%
0.0%
0.0%
7.7%
0.0%
6.0%
6.6%
0.0%
0.0%
0.0%
0.6%
0.0%
0.0%
0.0%
0.0%
0.0%
LEAK I
tt
1
1
1
0
3
0
0
0
1
6
6
0
0
0
0
0
NG VENTS
*
0.8%
6.3%
7. 1%
0-. 0%
27.3%
0.0%
0.0%
0.0%
16.7%
0.0%
0.0%
0.0%
0. 0%
0.0%
0.0%
0.0%
TOTAL
TANK SYSTEMS
TESTED
128
16
13
1 1
10
3
1
6
1
1
1
0
3
6
16
TOTALS
267
0.6%
16
4.5*
3.1%
* Systems as used here include the tank, fittings attached directly to' the tank, and the tank
vents, product linos are excluded except for the fitting attaching the line to the tank.
L.U.S.T.. Leaking underground storage tanks . ,
No available data for fiberglass HI th installed ages older than 14
' \
-------
-------
IDENTIFIED LEflKS CFIBERSLflSS TflNKS)
LEfiKING FITTINฎ
5-i
r
?
f
I
K
?
?
r
I
E
&-
?
1..
s_
ป
?
1
L_
K
)
N
EH
i$
^3
1
B
3
J 1
IM
s
/
tn
^^
^
)
^
^
x^
^
f
3
^
M
KM
H
-
3
^
=
i
C
5 i
1
1
> 7 ฃ
__
IM
=
^
=
~
1
,
1 { t 1 t 1 1 1 1 1
i I i I i 1 1 I 1 I
9 10 11 12 13 14 ? TTL
INSTALLED TflNK flGE
Percent of total tanks tested by installed age bracket, 100% corresponds to
the total number of tanks tested in that catagory. { See graph below-)
TOTRL FIBERGLRSS TRNKS TESTED
r " ri>r
6 7 8 9 18 U 12 15 14 ?
INSTPLLED TfN< AGE
TTL
-------
-------
Notes of the
July 8th, 1987 Meeting
with PEI Installers
at Dallas/Fort Worth Airport
July 15, 1987
U.S. Environmental Protection Agency
Office of Underground Storage Tanks
-------
-------
I
PE1 MEETING OUTLINE
SECTION TITLE PAGE
PREFACE 1
EXECUTIVE SUMMARY * 2
I I. BACKGROUND INFORMATION AND MAJOR 3
RECOMMENDATIONS
III. TECHNICAL DISCUSSIONS 9
o Tanks ฐ
o Piping ฃ
o Spills/Overfills lb
IV. FAILURE ANALYSIS REACTION
18
-1-
-------
* T
I *
-------
PREFACE
On July 8, 1987, Robin Parker (Jacobs Engineering) and David
O'Brien (ERA-OUST) met with Bob Renkes (Executive Vice President) and
Sandra Parkison, (Administrative Director) of the Petroleum Equipment
Institute) and eight PEI member/insta1lers from different parts of the
country to discuss underground storage tanks. This meeting was arranged
to take place at the Dallas/Fort Worth Airport by PEI. EPA reimbursed
the PEI installers their costs of travel for getting to the meeting.
Letters were sent out to each meeting attendee by PEI prior to the
meeting and included a list of questions prepared by EPA'and several
trip reports summarizing results of local visits by EPA contractors
ซ
concerning causes of release. Several of the installers had obviously
discussed those questions with several of their work crews in
preparation for the meeting. The average experience of these installers
was between 20-30 years. EPA's purpose of this meeting was to record
(as viewed "from the trenches"), the response of several experienced
installers concerning some important technical questions pertaining to
underground storage tank systems.
-1-
-------
I. EXECUTIVE SUMMARY
The meeting with the PE I installers resulted in several important
findings:
o The experienced and knowledgeable installers believe state
licensure would be a giant step in the right direction. Most
installation problems must be solved by the installers
themselves.
o Corrosion is the major release cause with the tanks; installer
errors or accidents with the piping.
i
o A build-up of poor attitudes and the survival of bad practices
must be over-come to solve the installation problem.
o Great progress has been made through the use of interim
prohibition tanks. (They confirmed the 0.5% failure rate of
FRP tanks.) However, piping is more complicated and less
forgiving of mistakes that are human and unavoidable to a
large extent.
o Pressurized piping is more of a threat and they recommend
pressure leak detectors be installed as quickly as possible on
existing tanks.
-2-
-------
o The influence of corrosion will vary in different parts of the
country. The use of clean, sandy backfills appears to
eliminate the localized (or pitting) corrosion phenomenon, but
not generalized corrosion. All corrosion holes do not leak --
many get plugged up by the corrosion by-products or
surrounding dirt.
o Failures in delivery piping are inevitably going to continue
because of human control problems. Annual testing and pressure
detectors are needed. More than JQ% of volume releases will
be thereby avoided"in the future.
o Overfill shut-off is the best approach, but not yet widely
available in this country.
o Double-walled tanks and piping are good Ideas, but there is
not much experience with them and most people will not
voluntar ily buy it.
II. IMPORTANT BACKGROUND INFORMATION AND MAJOR RECOMMENDATIONS
(Provided by PEI Representatives)
A. The Profile of PEI Installers
It was noted at the meeting that the installers in attendance were
unique in several respects. First, they were from all parts of the
country, including California, Colorado, Iowa, Texas, Wisconsin,
-3-
-------
Georgia, Delaware, and Pennsylvania. (PEI made sure a wide cross-
section was in attendance.)
Second, they were very experienced with servicing underground
storage tank systems and reported 20 years and 38 years as the shortest
and longest experience in the group. (Twenty-five years was most
common).
Third, Mr. Renkes reported that this was a very successful group of
businessmen each knew the technical aspects of UST
installation/servicing and repair business quite thoroughly.
Fourth, all of these installers provided both steel or fiberglass
tanks and several tested both tanks and lines.
Fifth, most of these gentlemen only got 10 to 20% of their business
from major oil companies (one got approximately 50* from that sector of
the industry). In sum, Mr. Cooper succeeded in gathering a unique group
of men who were independent-minded, who had not gotten together before
the meeting; who clearly had wide-ranging and long-standing experience
with the underground storage tank system technology; and who clearly
felt free to provide technical information that could be helpful to EPA
in finalizing the UST system regulations. It became obvious in the
meeting that these esteemed gentlemen knew what they were talking about.
-------
C. Important Changes within the Industry
During the meeting, these installers reported that a significant
transformation has taken place over the last 15 years in the petroleum
marketing industry. Fifteen years ago, the major oil companies
dominated the retail motor fuel market place. Each had very experienced
teams of engineers who developed their own firm's detailed company
specifications for designing retail stations (including blueprints and
drawings). The on-the-job oversight by the companies was extensive
then, and resulted in the actual training (in effect) of many of the
present PEI membership. This was also significant because, prior to the
oil crisis of the early 1970's, major oil also controlled many of the
gasoline stations. Thus, although major oil company engineers would
often look for economical tank systems, they generally insisted upon
adequate technical designs as part of the "bottom line."
Since the early 1970's, there has been a major shift in the retail
motor fuel marketing industry. Major oil companies no longer dominate
and have progressively withdrawn as builders of gasoline stations.
Major independents and jobbers have become the major purchasers of new
tanks and have prompted significant changes in customer attitudes faced
by PEI members. The independents are still very concerned with costs as
a bottom line, but often do not have the technical knowledge to insist
on effective equipment selection and installation. Thus, the low bidder
often gets the new installation job or the independent gasoline station
owner arranges several deals which cover separate contracts within the
-5-
-------
new tank installation. Cut-rate backfilling of piping or construction
of substandard concrete pads is not uncommon in an effort to reduce
costs. Also, there are many new non-PEI installers in the business
(many simply plumbers with spare time), who do not know proper system
equipment choice and installation. Their inadequate understanding and
quality control enables them to low bid new installations.
The PEI members reported they are often called by these new guys
concerning various technical matters, and it is obvious that a whole new
round of learning is necessary for them to become successful installers
in the future but, there is no one around to train them now. Some of
these new firms are highly capitalized construction companies, but with
little actual experience in this area. The major oil companies are also
generally running "mean and lean" right now and their reduced
engineering teams are often recent college graduates with budgets of
between 2 to 8 million dollars to spend, but rio field experiences.
(Apparently, the same major oil company representatives now call PEI
members for technical advice; not vice versa.)
It was suggested, at the meeting that all these major changes
portend the difficult times ahead in trying to improve installation
performance through regulation. The basic infra-structure for
installation is still undergoing major changes. The misinformed or
under-informed customer is now causing technical problems to be glossed
over by this unsuspecting "low-bidders" type of decisionmaking;
according to these experienced installers.
-6-
-------
C. Implementation Is the Key
These PEI members suggested that problems in the industry have been
buried for years and covered up by an "out-of-sight, out-of-mind"
mentality. Implementation of good practices is the key issue, and
therefore, they strongly recommend to EPA that installer certification
or licensure be a requirement in the final rules. They cited the
Iicensure of electricians as an example of the kind of internalization
of standards and professionalization of the industry that is needed to
make proper installation an industry-wide practice. This other industry
was noted "as having self-directed technicians, primarily because
licensure assures their livelihood is threatened if they do not do their
job right.
These PEI members believed that self-directed installers are the
only way to assure proper implementation. They noted that, at present,
there are many inexperienced third party inspectors of installations
making the rounds. In addition, most installations are not inspected at
all. Thus, most decisions about equipment choice and proper
installation are completely upon the shoulder of the installer "in the
trenches." State licensure programs would be a giant step forward in
assuring sound installation practices are followed.
-7-
-------
D. Enforcement Is Important
The PEI members at the meeting also emphasized the need for a
regulator's presence in the field. They were quick to point out the past
bad practices and short-cuts that resulted in environmental releases had
no adverse reprecussions to these irresponsible installers. Apparently,
the cost of going back and repairing mistakes was not a sufficient
enough incentive to assure sound initial installations. While growing
liability problems will change this attitude somewhat in the future, new
standards will be ignored if the regulated community does not believe
that EPA and-the states will enforce requirements. These experienced
PEI members would welcome an independent third party inspection
requiring minimum standards are applied consistently at each site.
These members were emphatic to point out that there are presently
real problems in the field because special interests and connections
within the industry are well established. The business goals they
support are in direct conflict with societal goals of environmental
protection. The example of this conflict they described at the meeting
was the case of a tank tester who serves the regulated community and
finds a leak. Must he report it? If he is good at finding leaks, will
his service be used? They believe that an absence of enforcement will
assure these established patterns, conflict of interest, and ties remain
undisturbed, and current behavior and attitude problems will not change.
-8-
-------
E. The Major Causes of Release
The PE I members cited two major causes of release that had to be
controlled/regulated by EPA.
1. There is a need to have better control over who installs the
tanks, what is hooked up to them, and did they do it right.
As previously mentioned, they believe state licensure and
enforcement are the key controls needed. The "system" is an
installer's problem.
2. They believe that corrosion (the primary failure mode) must be
prevented. They also believe that the new "interim
prohibition tanks" will largely solve this problem for the
tanks.
111. TECHNICAL DISCUSSIONS
Attached to this trip report is a copy of the questions that were
mailed out to the PEI members prior to the meeting (Attachment I). Each
of the four technical topics (piping, fittings, tanks, and
spills/overfills), were discussed at the meeting in consideration of
these basic questions. The installers emphasized that corrosion and
improper installation are the major causes of release.
-9-
-------
They provided the general estimate that k-S years ago, 50 percent
of the systems would not have passed a tank tightness test and 10
percent were significant problems. They estimate that this earlier
level has decreased somewhat due to the on-set of new state and local
UST programs and on-going risk reduction programs initiated by several
major companies, including most major oil companies. They confirmed
that their collective experience has shown many systems with
perforations that do not leak. It was estimated that perhaps 15 to 20
percent of existing systems are leaking (not counting spills and
overfilIs).
A. Tanks
Concerning tankage, the installers reiterated that corrosion and
improper installation were the primary causes of tank failure. Improper
installation in their minds includes the use of improper soils (loam or
clay soils that foster corrosion, or poorly compacted soils) in the
backfill. They estimated tank failures as contributing 20 to 30% of all
equipment-related releases.
Bare Steel Tanks
They expressed the opinion that external corrosion of bare steel
tanks was primarily induced by contaminated backfill (including the use
of clays or other native loams). They confirmed that the use of sand or
gravel backfills has been seen to produce generalized, but not localized
(or point) corrosion. External corrosion was often seen to occur at
-10-
-------
the bottom of the tank. End cap seam corrosion was another culprit.
With internal corrosion, the bottom foot to eighteen inches of the tank
is the first to go; with a focus on the shell to head seam.
An informal polling revealed a wide disparity in the reported on-
set of corrosion-induced tank system failures with age and location:
Age
of Tank
0-10 years
10-20 years
21-30 years
South
California
30*
60%
Few
Mid-Atlantic
Coast
30*
60*
Few
Atlanta
0*
40*
Few
Wisconsin*
0*
10*
25*
Denver
0*
10*
25*
* Depends on groundwater level.
Thus, while the consensus was that corrosion-induced failures of
bare steel tanks was significant, it is apparently a very site-specific
phenomenon. One of the mid-west installers reported that less than 100
of the 300 tanks he closed and his dealer cut-up last year were leaking;
75* of the leakers were corrosion-induced (one was internal corrosion).
Protected Steel Tanks
The discussion of coated and protected or clad tanks related that
experience is not yet extensive, but no one had seen any failures yet.
One installer noted he had closed several tanks with anodic protection
and they were in fine shape. There was some experience with wrapped
tanks, which appear to improve upon bare steel performance, but had some
-11-
-------
problems with installation and water getting in between the tank and the
wrap. The general sense was that the premanufactured protected steel
tanks work and will result in at least an 80 to 90? improvement over the
current bare steel tank corrosion situation.
Fiberglass Tanks
Initial discussion of FRP tanks revealed some reservations about
the technology and a swapping of "war stories" An informal polling
revealed, however, that these installers had installed an estimated
total of 17500 to 2,000 FRP tanks over the years with approximately 7-8
failures. These installers reported only one failure over the last 10
years. They all complimented the industry training work that has been
provided towards the installation of FRP tanks and expressed their
awareness of the need to carefully install FRP.
Double-walled Tanks
Many of the installers expressed a preference for double-walled
tanks as a way of further reducing leaks. However, at least one was
concerned that it may introduce new problems that are as yet unknown.
Few appeared to have extensive experience with this technology and
expressed reservations as to how well it would be received by cost-
conscious buyers, if required by EPA in their state.
-12-
-------
B. P i p i ng
The installers expressed their opinions/experiences that piping
failures are the main problem by a 5 to 10 to 1 ratio compared to
problems with the tank. They reported that it is not uncommon to re-
pipe tanks for reasons other than they are leaking, e.g., site
modifications. In general, they expressed the concern that piping and
equipment fitted to the tanks are less forgiving (than the tanks) of
installation errors and are generally more complex to install there
are more conditions prevailing in assuring proper installation of piping
versus tanks (e.g., a 1Q% breakage rate by other trades during
construction was cited as typical with the piping).
Steel Piping
They witnessed that more than a few "pipe-benders" are still in
operation and these practices weaken piping and induce corrosion. Steel
or galvanized piping has corrosion problems similar to bare steel tanks.
They also see numerous fitting failures/stress breaks; accidents at the
site that jog the unanchored impact values at the dispenser and break
the piping; and corner-cutting by the owners at installation that can
weaken piping. One example cited was numerous sub-contractual
arrangements that had someone else placing backfill around completed
pipework, final grading, and doing inappropriate pad construction that
is inadequate in protecting the piping from normal stress forces from
above.
-13-
-------
FRP Piping
Some expressed work crew preferences to work with FRP piping as
intrinsically easier and cleaner to use. Some FRP product piping are
damaged during shipping but, in general, the materials are improving.
they emphasized their collective experience is that you cannot mix the
different types of FRP piping and proper installation must closely
follow manufacturer's instructions.
It can get tricky to install FRP piping in cold weather, and they
fear other construction site workers damaging it once it has been
completed and left in place. The same fears were highlighted about
possible adverse affects of frost heaves upon FRP piping failures.
Also, not enough is yet known about how much future work on-site (e.g.,
digging up asphalt) will disturb FRP piping. They all insist it be
integrity tested after the full job at the site is completed.
Suction versus Pressure Lines
Suction systems were cited as generally having only minor "leakage
problems." The installers pointed out that suction pumping has
operational problems in high altitude areas and some hot places during
the summer months. Besides certain technical/operational advantages,
pressure systems can be less expensive:
-------
o one suction pump and line $1100
o pressure pump lines $400-$500
o one pressure pump that can service $800-$900
several dispenser lines
In general, failure modes for suction versus pressure lines are the
same. But, pressure systems are acknowledged to pose more of a release
threat. The buyers only focus on costs and ignore the additional risks
posed by pressure piping. Significant losses can occur with pressure
lines, particularly when owners do not carefully track their product.
Each installer said he could easily cite at least 10 cases of
significant damages posed by releases from pressurized piping failures.
One example cited was where a faulty pressurized pipe resulted in an
owner refilling a tank twice before noticing the breakage. The group
further estimated that at a minimum, 70 percent of line volume losses
could be eliminated by retrofitting continuous pressure leak detectors
on all product lines within twenty four months. Such devices would only
increase the typical $10,000 piping installation costs by $500 at new
faci1i ties.
Because of disruptions caused by future work at the site,
accidental damage during system operations, and installation errors, the
installers may only be able to reduce the volume loss about 20 percent
if automatic pressure in line detectors are not required.
Costs of retrofitting continuous pressure sensors are about $165
per unit plus 2 hours labor cost in 80 percent of the cases. In
-15-
-------
some situations, retrofitting Is not straight forward and an additional
extractable section of the pump must be added (and an additional cost of
$650), taking 12 hours more labor. They further estimated that
retrofitting on existing systems would result in 20-25? of the lines to
be identified as leaking.
Automated line detectors are widely available in Southern
California and cost $6,000 for 3 tanks and lines plus $2,000
installation costs. Pipe trench liners or double-walled pipe was
estimated to cost within $500 of each other. Annual line testing costs
about $200 pefstation and one member cTted that at 100 Texaco or SheH
stations, the following results from annual testing could be estimated:
10 - 12 faulty impact valves
1 leak (every 2 years)
2-3 bad Red Jackets
The installers did not believe manifolded pipes or fill tubes
presented a leaking problem, except in overfill situations. Ninety
percent of all commercial accounts (e.g. truck fleets) are suction
pumping and 95 percent of all -new retail/motor fuel accounts are
pressure systems.
C. Fittings/Overfill and Spills
Three basic kinds of overfill/spills control equipment were
discussed:
-16-
-------
o An overfill shut-off device located in the tank or vent/fill
lines (several types), most not widely available in the United
States.
o A small containment catch basin at the fill pipe.
o A large volume catch basin at the fill pipe or man way catch
basins available from tank manufacturers.
The large catchment basins/manway equipment are believed to pose
safety threats due to explosions however, no examples were available
of this happening yet.
They estimated that they see 75 out of 100 closed sites as having
been contaminated by releases (at least somewhat), and 50 of them appear
to be caused by overfill/spills. Two former transporters in the group
provided the following estimated spill/overfill rates:
1 out of 25 deliveries have "minute" spills of 2 to 3 gallons
1 out of 100 deliveries have several tens of gallons of
spillage.
They reminded us that most deliveries take place at the evening and
drivers get paid by the load which encourages fast deliveries. They
introduced a new meaning to the term "midnight dumping."
-17-
-------
Although tank fittings were generally believed to be a small source
of releases, they do pose leakage problems during overfill and spill
situations which do not result in visible losses at the fill pipe
connection (as discussed above).
A number of contractors were cited as taking installation short-
cuts believing that vent lines and bung holes only handle "air." Some
noted short-cuts were use of PVC pipe for vent lines, failures to
tighten vent line joints, failure to glue joints, failure to tighten
bung holes or use proper fixtures.
IV. REVIEW OF EPA FAILURE ANALYSIS
The installers were asked at the meeting's end to comment on some
of the probabilities of failure that EPA is presently using in its
analysis. The mark-up sheet follows. In general, they did not believe
non-corrosion induced seam leaks or tank/fitting ruptures were real
failure events. They also believed FRP tank failures estimates were
much too high. More failures due to pipe corrosion or loose fittings
was predicted. They emphasized that all tanks with holes did not
necessarily leak.
-18-
-------
PROJECTION OF EQUIPMENT FAILURE ON 1,000 TANKS
IN A 30-YEAR PERIOD
CAUSE OF FAILURE ปCF P_งi
External Tank Corrosion 770 770
Internal Tank Corrosion 150 10
Seam Leaks ^3
Tank Breakage 1* NIL
Pipe Corrosion 280 400
Pipe Rupture 34 -0-
Loose Fittings -*- 50 100
Tank Rupture (Steel, Single Wall) 60 10
FRP Tank Rupture 230 25
Double Wall Tank Rupture 84 10
Steel Pipe Rupture 96 ~0~
Double Wall Pipe Rupture 120 -0-
Screw Fitting Rupture 150 10
Steel Tank Installation Damage 30 30
FRP Tank Installation Damage 40 40
Pipe Tank Installation Damage 10 10
Fitting Installation Damage
-19-
-------
Sign-up Sheet
Name
David O'Brien
Robin Parker
Sandra Parkison
Edward Ettl (WI)
Leland M. Freeman (TX)
William Thompson (DE)
Bob Renkes
Orville Kubat (CO)
George Watkins (GA)
Jerry Thomas (CA)
B.D. Cooper (IA)
M.W. Farmer (PA)
Company
USEPA-OUST
Jacobs Engineering
PEI
Badger Oil Equipment
SE 6 M, Inc.
Cal1is-Thompson, Inc.
PEI
Kubat Equipment & Ser. Co.
Watkins Service Co.
C.E. Thomas Co.
P.E.S., Inc.
M. W. Farmer & Co.
Telephone
(202)
(202)
(918)
(414)
(512)
(302)
(303)
(404)
(213)
(319)
(717)
382-7815
783-1560
743-9941
258-2777
573-7449
398-3253
743-99^1
777-2044
768-7867
321-8420
377-6357
326-1541
-20-
-------
QUESTIONS
Piping (used for product transport between the tanks and
from the tanks to the dispensers) .
o Why does piping leak? What are the most common modes
of failure you have seen (e.g. loosejoints^ corrosion,
breakage, other)? ,*;'
o What is your experience concerning the success/
failure of FRP piping?
ป _ ป - " ' - * ^ 4 e -.
o What is the relative frequency at piping failures as
compared to tank failure? Is there any experience to
how often it will fail (at 100 stations over 20 years,
for example).
~~" _ - - . ..-*
o What are the typical sizes of releases you have seen
from piping?
Are there larger releases from pressure
versus suction lines (even with systems
that have red jacket type devices on them?)
o Are the impacts from piping release often significant?
Less or more significant than from tanks? Suction
piping less significant?
o What will the world be like 20 years from now if the
federal proposal was implemented? Does FRP piping
solve the problem?
Fittings (vents, bung holes, and fill pipes on top of the
tank)
o Why and to what extent are fittings a source of leakage
from the tank system?
o What is the relative frequency of fittings failures as
compared to tank or piping failures? (How often will
fittings fail at 100 stations over 20 years?)
o What typical sizes and impacts are the leaks from loose
or failed fittings? Have you seen significant impacts
that were based on failed fittings?
o Are fittings problems controllable and avoidable? What
will be the situation concerning fittings 20 years from
now?
-------
Tanks (actual storage vessel excluding the fittings)
>
o What are the primary causes of tank failures? What are
the most common failures of steel versus FRP tanks you
have seen?
o What is the relative failure of tanks versus piping or
fittings failures? How often do unprotected tanks fail
(e.g. at 100 stations over 20 years, for example)?
Have you seen protected tank failures; for
what reasons; how often do you estimate they
have failed/will fail?
o What are the impacts/sizes of leaks from failed tanks?
Does this differ from steel versus FRP?
o Will present problems be the same or different 20 years
from now if the federal proposal was implemented?
Spills/Overfills
o What is their experience concerning spills or
overfills?
Frequency
Sizes
Impacts
o What equipment works or doesn't work?
Overfill shut-off devices?
Spills containment basins?
o Are shut-off devices and containment devices needed?
o Should EPA allow 10 years for equipment to be applied
to existing tanks?
o How difficult and costly is it to apply various
equipment?
-------
TEST IMG COM !=>ซ=ป MY I IMC
P.O. BOX 79io3,f;,;.-J
San Antonio, Texas 7QS7$j-1032
July 21, 1987
Ms. Cecily Beal 1 ;
MIDWEST RESEARCH INSTITUTE
52O1 Leesburg Pike
Suite 209
Falls Church, VA 22041
Dear Ms. Beal 1 :
Thank you -For the opportunity to participate in your study.
I hope the i n-formati on herein will facilitate the E.P.A. in
their current quest to determine underground storage tank
policies. Please keep in mind that the data we have
provided is totally unbiased, as we .do not sell nor install
tank equipment.
In gathering this information, we thoroughly read over 6OO
test -files each containing eight to twenty pages of
documentation covering two to five tank tests each. We
determined the number of leaks and in what part of the tank
or tank system they were found and categorized them on a
data sheet. After all the files were reviewed, columns were
then totaled giving us the actual number of tight and non
tight components. Using a computer to compile this data, we
converted them into percentage form and generated pertinant
graphs to relay the gist of the findings. By using the
categories listed in the data sheets, there are more analyze
that can be evaluated beyond our summarization.
A basic tank system consists of the parts illistrated in
Diagram i (see attached). In our study we categorized a
leaking system in two categories-- "Tank or "Li"6"/ Under
the tank category we included 1). the tank, 7). the vent,
8). riser pipes and 9). bung plugs. Under the line category
we included 2). the product line piping, 3). unions 4> .
swing joints, 5). f 1 ex hose and 6) . submerged pump.
In a nut shell, leaks occuring in the line category
resulted in operational leaks; ^hereby, hydrocarbons escaped
from the system into the ground causing varying degrees o+
contamination. Leaks occuring in the tank category caused
contamination only if the leak was in the tank itself or if
-------
the tank was filled slightly above its full capacity (not
knowing how often the tanks are overfilled during their
lifetime, it is important to test the entire tank system to
insure its tightness).
After gathering this data and reviewing it, one observation
I was able to make was the unusally low number of actual
bad tanks we found as compared to those found in other
studies. Could this possibly be due to a terminology
problem?
-------
k/'
1.
2.
3.
4.
5.
6.
7,
8,
9
10
Tank
Product Line Piping
Unions
Swing Joint
Flex Hose
Submerged Pump
Vent Line
Fill Pip Riser
4" Plug Into Tank Bung
Dispensing Unit
DIAGRAM 1
-------
-------
TANK SYSTEM TEST RESULTS
Geographical
Region
C&lorado
Texas
Aflstin (TX)
San Antonio (TX)
TOTAL
TANK TEST RESULTS
TOTAL
TANKS
205
331
544
S41
1921
TANK
6
17
18
51
92
LEAKS
PITTING/VENT
87
35
53
97
272
TOTAL
93
52
71
148
364
LIKE TEST RESULTS
TOTAL
LINES
189
325
510
801
1825
LEAKS
10
40
50
76
17C
GBXSO TOTALS
TOTAL
LEAKS
103
92
121
224
540
TIGHT
TESTS .
291
564'
933
1411
3206
TOTAL
TESTS
394
656
1054
1642
3746
LINE TEST RESULTS
LINE TYPE
SUCTION
474
REMOTE
1351
TOTAL
LINES
1825
LEAKS
PIPING
96
UNION
80
TOTAL
'176
FAILED
LIMB TEST
156
TIGHT
LINS TESTS
1669
fisrtcuss
228
XATIRIXL '
HBraoLxsj
srtei.
57
sTtrt,
1636
1921
: ^Tx Ttsr aisut/rs
| ' (-FIKS ,.. .. ^
92
FITTING _
140
124
TOTAt
364
325
TANK TEST
1596
-------
-------
SERVICE STRTICN TESTING INC. L.U.S.T. SURVO 81-87 (COMBINED TEST DfiTfl)
92 Tank Leaks
272 F/V Leaks
176 Line Leaks
* ปป ป ป
ซ * mm -^mmmm m mm *t ..
*^^H ** *
*i ^V ป
mW mm m ป
. ซ
" *
: : :
: : ::,
ป j
f : -: :J
I : ; :J
*
:
L i i
^ " '
*- i
^- 11
ซa >ซ
*
3206 Tight Tes
.F/V (refers to fittings and vents)
PERCENTAGE
Tank Leaks
F/V Leaks
Line Leaks
Tight Tests
2.45%
7.26%
4.69%
85.58%
-------
SERVICE STflTION TESTING INC
18 Tank Leaks
53 F/V Leaks
Line Leaks
L.U.S.T. SURVEY 81-87 (flUSTIN)
'^^ปMM^MxSMs^^M&MKซ&m^
933 Tight Tests
PERCENTAGES
Tank Leaks
F/V Leaks
Line Leaks
Tight Tests
1.1%
5.0296
4.74%
88.51%
-------
SERVICE STATION TESTING INC. L.U.S.T. SURVEY 81-87 CflUSTIN)
18 Tank Leaks
53 P/V Leaks
473 Tight Tanks
PERCENTAGES
Tank Leaks
F/V Leaks
Tight Tanks
3.3%
9.74%
86.94%
-------
SERVICE STflTION TESTING INC. L.U.S.T. SURVEY 81-87 CflUSTIN)
58 Line Leaks
460 Tight Line
PERCENTAGES
Line Leaks
Tight- Lines
9.8%
90.19%
-------
SERVICE STRTION TESTING INC. L.U.S.T. SURVEV 81-87 CfEXftS)
Line Leaks
285 Tight Lines
PERCENTAGES
Line Leaks 12.3%
Tight Lines 87.69%
-------
SERVICE STflTION TESTING INC. L.U.S.T. SURVEY 81-87 CTEXflS)
17 Tank Leaks
35 F/V Leaks
279 Tight Tanks
PERCENTAGES
Tank Leaks
F/V Leaks
Tight Tanks
5.13*
10.57*
84.29*
-------
SERVICE STflTION TESTING INC. L.U.S.T. SURVEY 81-87 CTEXflS)
35 F/V Leaks
17 Tank Leaks
40 Line Leaks
564 Tight Tests
PERCENTAGES
Tank Leaks
F/V Leaks
Line Leaks
Tight Lines
2,59%
5.33%
6.09%
85.97%
-------
SERVICE STRTICN TESTING INC. L.U.S.T. SURVEY 81-87 (San fintonio)
51 Tank Leaks
97 F/V Leaks
693 Tight Tanks
PERCENTAGES
Tank Tests
F/V Leaks
Tight Tanks
, 6.06%
11.53%
82.4%
-------
SERVICE STflTiaTTESTM INC. L.U.S.T. SLEVEY 81-87 (San flntonio)
51 Tank Leaks
97 F/V Leaks
76 Line Leaks
1418 Tight Test
PEPCENTAGES
Tank Leaks
F/V Leaks
Line Leaks
Tight Tests-
3.1%
5.9%
4.62%
86.35%
-------
SERVICE STfiTICN TESTING INC. L.U.S.T. SURVEY 81-87 (San fintonio)
725 Tight Lines
76 Line Leaks
PERCENTAGES
Line Leaks 9.48%
Tight Lines 90.51%
-------
SERVICE STflTION TESTING INC. L.U.S.T. SURVEV 81-87 CCOLORRDO)
6 Tank Leaks
87 F/V Leaks
10 Line Leaks
_
!!ป *ปปป*
;ปซ ปป '
<
*ซซ *
ง*ซ
(ป ป ป
* *
:;;:::;;: i ;
r" ~ ! !!
,: J " ซ ;;::
::::
::::
, *
L
'r
tซJ
J
I if
_A
1
'" S
291 Tight Tests
PERCENTAGES
Tank Leaks
F/V Leaks
Line Leaks
Tight Tests
1.52%
22.08%
2.53%
73.85%
-------
SERVICE STflTION TESTING INC. L.U.S.T. SURVEY 81-87 (COLGRflBO
6 Tank Leaks
112 Tight Tanks
87 F/V Leaks
PERCENTAGES
Tank Leaks
F/V Leaks
Tight Tanks
2.29%
42.43%
54.63%
-------
SERVICE STRTION TESTING INC. L.U.S.T. SURVEY 81-87 (COLORADO)
10 Line Leaks
179 Tight Lines
PERCENTAGES
Line Leaks
Tight Lines
5.29%
94.7%
-------
-------
SERVICE STATION TESTING, INC.
P.O. BOX 791032 SAN ANTONIO, TEXAS 78279-1032
Phone (512) 344-3537
DON W. MAY, JR.
President
September 15, 1997
Ms. Cecily Beall
MIDWEST RESEARCH INSTITUTE
425 Volker Boulevard
Kansas City, MG 64110
REF: PURCHASE ORDER'- 92343
Dear Cecily:
Please find enclosed the survey, as requested, in compliance
with Purchase Order #92343. If you have any questions, please
feel free to contact me.
Si ncerelv
Donald W. May, Jr
President
DWM/tlm
enclosure
-------
-------
ST<=*TIOIM TESTING co.
L.U.S.T- SURVEY
L.I1NIE T I OHTlvJESS TEST RESULTS
REMOTE LINES SUCTION LINES
*
UEAKIK6
133! REMOTE LINES
124 (9.11)
474 SUCTION LINES
32 (6.81)
TOTAL LINES
*
182! TOTAL LINES TESTED
134 (8.5X)
TIGHT
1227 (90.81)
442 (93.21)
1469 (91.31)
Lฃf*XN6 PttOUCT LUES AS PSCSCT OF TOTAL TESTS) P3 CAFMORV
Y////7A REMOTE LIhES
SUCTION LITCS
REMOTE + SUCTION
NUMBER CF PRODUCT LINES TESTED (PER CflTflGORY)
-------
-------
SERVICE STATION TESTING INC.
L.U. S.T. SURVEY
"COMPARISON OF STEEL TANK TIGHTNESS TEST RESULTS BY AGE OF TANK"
1ST ALL-ED
TANK
AGE
TEST RESULTS
TIGHT SYSTEMS * LEAKING TANKS LEAKING FITTINGS
It % tt % H %
LEAKING VENTS
ซ %
NEW
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26+
JNKNOHN
67
34
18
16
7
15
22
10
25
7
44
16
30
27
41
70
78
11
29
15
68
32
16
16
21
34
40
523
90.5%
91.9%
66.7%
80.0%
53.8%
75.0%
81.5%
62.5%
86.2%
53.8%
84.6%
80.0%
76.9%
79.4%
85.4%
81.4%
87.6%
57.9%
82. 9%
55.6%
77.3%
71 . 1%
76.2%
61.5%
84.0%
85.0%
1 00. 0%
83.3%
0
0
2
0
3
0
1
0
0
0
2
2
6
1
3
7
4
1
2
1
4
1
1
2
0
2
3
44
0.0%
0.0%
7.4%
0.0%
23. 1%
0.0%
3.7%
0.0%
0.0%
0.0%
3.8%
10.0%
15.4%
2. 9%
6.3%
8. 1%
4.5%
5.3%
5.7%
3.7%
4.5%
2.2%
4.8%
7.7%
0.0%
5.0%
0.0%
7.0%
7
2
6
2
1
2
2
4
3
1
6
2
0
3
0
5
5
2
2
3
8
6
4
5
3
1
0
33
9.5%
5.4%
22. 2%
10.0%
7.7%
10. 0%
7.4%
25.0%
16.3%
7.7%
11.5%
10.0%
0.0%
8.8%
0.0%
5.8%
5.6%
16.5%
5. 7% ,
11.1% ^
9. 1%
13. 3%
19. 0%
19.2%
12.6%
2.5%
0. 0%
5.3%
0
1
1
2
2
3
2
2
1
5
0
0
3
3
4
4
2
5
2
8
8
6
e
3
1
3
0
28
0.0%
2.7%
3.7%
10.0%
15.4%
1 5. 0%
7.4%
12.5%
3.4%
38.5%
0.0%
0.0%
7.7%
8.8%
8.3%
4.7%
2.2%
26. 3%
5.7%
29.6%
9. 1%
13.3%
0.0%
11.5%
4.0%
7.5%
0.0%
4.5%
TOTAL
TANK SYSTEMS *
TESTED
74
37
27
20
13
20
. 27
16
29
13
52
20
39
34
48
86
89
19
35
27
88
45
21
26
25
46
43
628
OTALS
1332
81. 1%
92
5.6%
1 18
7.2%
99
6.0%
1641
Sy*tปrป* *s used here include the tank, fittings attached directly to the tank, and the tank
vents, product lines are excluded except for the fitting attaching the line to the lank.
I . U>,S.T., Leaki ng* underground storage tanks
-------
------- |