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 failing—bare 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
          numbers—judgments based upon experience; permit database;
          release notification database; corrective action database—can
          these databases be put together to confirm these numbers.


RELEASE DETECTION TECHNIQUES


     o    What types of leak detection are in'use—both 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 technique—at installation
          or after;

     o    What sizes of releases are being found with each technique.


PROGRAM IMPLEMENTATION QUESTIONS


     o    What are the installation requirements—how does the regulator
          insure proper installation;

     o    What is the phase-in schedule for upgrading—is it being met;

     o    What repairs and retrofits are allowed—how does the regulator
          insure proper repair;

     o    What are the spill and overfill control requirements—what is
          the phase-in schedule;

     o    What is the Leak Detection phase-in schedule—is 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 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 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*



                                          ;•••ป••ซ••    • ปป •••'
                                          •••••••••    • •• •<•
                                          *•••ซ••ซ•    • •• •*•
                                          •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

-------

-------