FINAL REPORT
ASSESSMENT OF THE TECHNICAL ,
ENVIRONMENTAL AND SAFETY ASPECTS
OF STORAGE OF HAZARDOUS WASTE
IN UNDERGROUND TANKS
Prepared for:
U.S. Environmental Protection Agency
Office of Solid Waste
401 M Street, SW
Washington, D.C. 20460
Prepared by:
SCS Engineers
11260 Roger Bacon Drive
Reston, Virginia 22090
February, 1984

-------
DISCLAIMER
The information is this document has been funded wholly
or in part by the United States Environmental Protection
Agency under Contract 68-02-3179 to SCS Engineers. It has
not been subjected to the Agency's peer and administrative
review and has not been approved for publication. Mention of
trade names or commercial products does not constitute endor-
sement or recommendation for use.

-------
n
(file 0
113*f-
U.S. Environmental Pratection Agency
Region ill Information Resource
Center (5PM52)
841 Chestnut Street
CONTENTS	Philadelphia, PA 19107
Page
F1 gures	.... ;	i v
Tables		 . v
Summary		 . . vi
1	Introduction	1-1
2	Underground Tank Use for Hazardous Waste
Storage	2-1
Introduction			2-1
Tank Use Characteristics	2-1
3	Damage Case and Spill Event Review	3-1
Introduction 		3-1
Information Sources	3-1
Conclusions	3-26
Exhibit 3-1 . . . 		3-29
Exhibit 3-2	3-37
4	Relative Release Probability and Magnitude	4-1
Introduction 		4-1
Release Events and Variables 		4-1
"Typical" Facility 			4-14
"Typical" Facility Release Magnitudes
and Probabilities	4-17
Data Limitations	4-35
Summary		4-36
5 Analysis of Selected Management Alternatives	5-1
Introduction 	 5-1
Model Facilities	5-4
Leak and Rupture Release Migration/
Prevention	5-8
Conclusions	5-40
PRO"^
Regional Center for Environmental Information
US EPA Region III
1650 Arch St.
Philadelphia. PA 19103
V_
i i i

-------
TABLE OF CONTENTS, (CONTINUED)
Page
Appendices
A. Excerpts from "Technology for the Storage
of Hazardous Wastes, A State-of-the-Art Review"
New York State Department of Environmental
Conservation, January 1983	A-l
B.	"Corrosion Guide" for Fiberglass Reinforced
Plastic Tanks, from Raven Industries Inc.,
Sioux Falls, South Dakota	B-l
C.	Summary of Design Standards for Underground
Storage Tanks prepared by SCS Engineers	C-l
D.	Memoranda from SCS Engineers to Bill Kline
regarding Tank Shell Thickness	D-l
E. Chapter 8, "Inspections", from "Permit Writers
Guidance Manual for Hazardous Waste Tanks", Draft
Prepared by Ba-ttel le-Col umbus Di vision for Region
II, U.S. Environmental Protection Agency,
July 1983	E-l
F.	Cost Estimates for Construction of Storage
Tank Systems	F-l
G.	Summaries of Selected References	G-l
H.	Leak Testing Methods		H-l
I.	Procedures Used to Estimate the Number of
Underground Tanks in New York State	1-1
J. Underground Protected Steel Tank Study
Statistical Analysis of Corrosion Failures	J-l

-------
FIGURES
Number	Page
2-1 Frequency distribution of tank age	2-4
2-2 Frequency distribution of internal inspections
of underground tanks 	 2-5
2-3 Distribution of all underground tanks per
facility with underground tanks	2-9
2-4 Distribution of underground storage tanks per
facility with underground tanks	2-10
2-5	Distribution of underground treatment tanks per
facility with underground tanks	2-11
3-1	Location of ground water basins in the San
Francisco Bay area	3-11
4-1	Events leading to releases from underground
hazardous waste storage facilities . . . . . . . 4-3
4-2 Variables effecting tank overflow at an
underground hazardous waste storage facility . . 4-4
4-3 Variables effecting tank leaks at an underground
hazardous storage facility 	 4-5
4-4 Variables effecting containment system failures
at an underground hazardous waste storage
facility	4-6
4-5 Variables effecting tank rupture at an
underground hazardous waste storage facility . . 4-7
4-6 Variables effecting ancillary equipment leaks' at
an underground hazardous waste storage
facility	4-8
4-7 Variables effecting ancillary equipment
rupture at an underground hazardous waste
storage facility 	 4-9
4-8 Variables effecting fire of explosion at an
underground hazardous waste storage facility . . 4-10
v

-------
FIGURES (Continued)
4-9 Variables effecting other accidents at an
underground hazardous waste storage facility . .4-
4-10 Diagram of the "typical" underground storage
faci 1 i ty . . . . 		.... 4-
4-11 Release events and probabilities associated with
a "typical" underground hazardous waste storage
facility	4-
4-12 Variables and release probabilities associated
with tank overflow at a "typical" underground
hazardous waste storage facility 	 4-
4-13 Variables and release probabilities associated
with tank leaks at a "typical" underground
hazardous waste storage facility ........ 4-
4-14 Variables and release probabilities associated
with tank rupture at a "typical" underground
hazardous waste storage facility 	 4-
4-15 Variables and release probabilities associated
with ancillary equipment leaks at a "typical"
underground hazardous waste storage facility . . 4-
4-16 Variables and release probabilities associated
with ancillary equipment rupture at a "typical"
underground hazardous waste storage facility . . 4-
4-17 Variables and release probabilities associated
with fire or explosion at a "typical" under-
ground hazardous waste storage facility	4-
4-18 Variables and release probabilities associated
with other accidents at a "typical" underground
waste storage facility 	 4-
11
16
18
19
20
21
22
23
24
25
vi

-------
TABLES
Number Page
2-1 Underground Tanks Characteristics 	 2-3
2-2	Percent of Underground Tanks by Type of Waste
Stored and Construction Material of Tank	2-7
3-1	Petroleum Product Leaks by Category From the
API "Tank and Piping Leak Survey". (1977-1980). . . 3-3
3-2 Steel Tank Leaks Identified During the API
"Tank and Piping Leak Survey". (1977-1980)	 3-4
3-3 Fiberglass Tank Leaks Identified During the API
"Tank and Pipe Leak Survey". (1977-1980)	 3-8
3-4 Piping Leaks Identified During the API "Tank
and Pipe Leak Survey". (1977-1980). . 		 3-9
3-5 San Francisco Bay Region Questionnaire Data for
Selected Facilities*	3-14
3-6 Contamination Found at 20 Sites in San
Francisco Bay Area With Tank System Failures* . . . 3-17
3-7 Prince Geogre's County, Maryland Tank Testing
Program Results for Underground Gasoline Tanks
as of January 1978	 3-22
3-8 Sources of Well Contamination by Gasoline as
Reported in a Survey of Local Health Units 1n
New York in 1978*	 3-24
3-9 Results of Suffolk County, New York Tank Testing
Program as of December 1982 	 3-25
3-10	Summary of Reported Tank System Leaks 	 3-27
4-1	"Typical" Facility Release Magnitudes 	 4-30
4-2	Relative Release Probabilities and Magnitudes
Associated With the "Typical" Facility	4-37
5-1	Secondary Containment*. . . 		5-11
v i i

-------
TABLES (Continued)
5-2 Tank System Testing Summary*	5-16
5-3 Environmental Monitoring Summary* 		5-22
5-4 Inventory Monitoring Summary* 		5-29
5-5 Inspection Summary* 		5-34
5-6 Corrosion Protection Summary* 		5-37
5-7 "Solution" Comparison Summary*	5-41
5-8 Model Facility Release Probabilities* 		5-43
5-9 Incremental Cost and Effectiveness Summary -
Existing Facilities*	5-45
5-10 Incremental Cost and Effectiveness Summary - New
Facilities*	5-46
v i i i

-------
SUMMARY
INTRODUCTION
The objectives of this project were to define current
practices for hazardous waste storage in underground tanks;
evaluate these practices in relation to spill and damage event
data and best engineering judgement; estimate the relative
probability' and magnitude of waste release from underground
tanks; and examine appropriate alternatives for prevention and/or
mitigation of releases. The results of activities performed in
pursuit of these objectives are summarized below.
UNDERGROUND TANK USE
Based on the results of the II. S. Environmental Protection
Agency (EPA) mail survey of 1981 hazardous waste management
practices, hazardous wastes are stored in underground tanks which
range up to 50,000 gallons in capacity and 35 years in age. The
median tank capacity is 3,000 gallons while 90 percent of the
tanks have a capacity of 10,400 gallons or less. The median tank
age is 8 years and 90 percent of the tanks are less than 25 years
old.
A majority of the tanks are constructed of carbon steel,
although concrete, stainless steel, fiberglass reinforced plastic
(FRP) and other materials are also used. Ignitable wastes are
the most commonly stored waste type, followed by corrosive,
toxic, E. P. toxic wastes. Underground tanks are used to store
other types of hazardous waste significantly relatively infre-
quently.
Facilities with underground tanks which are used for hazard-
ous waste storage have up to 15 such tanks, with a majority of
facilities (55 percent) having only one underground tank. Under-
ground tank capacity ranges up to 95,000 gallons per facility
with a median capacity of 10,000 gallons. A majority of these
facilities (63 percent) store ignitable waste, with the next most
common waste types being toxic (34 percent) and corrosive (28
percent) .
DAMAGE CASES AND SPILL EVENT REVIEW
Damage cases and spill events were reviewed as part of the
effort to assess the adequacy of current practices for storage of
hazardous waste in underground tanks. Available information
which was reviewed came primarily from state and local government
agencies and trade associations. A majority of this information
is derived from petroleum product storage facilities since very
limited information is available for hazardous waste storage
facilities.
Data from an American Petroleum Institute (API) survey of
gasoline storage tanks which were found to be leaking indicate
i x

-------
that corrosion is the primary cause of steel tank leaks. The
ages of the leaking tanks covered by the survey ranged from 1 to
more than 31 years, with 86 percent of the responses for tanks in
the 6 to 25 year range. For FRP tanks, breakage or tank
separation (i.e., a physical separation of tank wall material)
accounted for all of the leaks. For piping, which was also
frequently cited as a leak source, corrosion was again reported
to be the primary cause of the leakage. Additional conclusions
which can be derived from the information are that poor installa-
tion can contribute to leaks, primarily through corrosion or
loose fittings, and that leaks can occur from tank systems
provided with corrosion protection if design, installation or
maintenance is inadequate.
Information collected from local government organizations
such as the Cape Cod Plannning and Economic Development Commi-
sion; Suffolk County, New York; and Prince George's County,
Maryland led to conclusions similar to those presented above for
the API survey (e.g., corrosion of existing steel tanks is
resulting in a significant number of releases). As a result,
local ordinances have been or are being developed to more closely
monitor the integrity of underground storage tanks. Similar-
efforts have also occurred at the state level in Michigan and New
York.
A'survey conducted by the California Regional Water Quality
Control Board, San Francisco Bay Region, of facilities storing
hazardous materials identified more than'80 facilities which used
underground tanks (primarily for product and waste solvent
storage) and were judged to have a high potential for leaking
hazardous materials. As of May 1983, tank system failures had
been found to be the cause of releases to soil and/or ground
water at 72 percent of the 57 facilities for which investigations
had been completed. Additional leaking tanks are expected to be
found as lower priorty groups of tanks are investigated
Prior to the conduct of this survey by the San Francisco Bay
Region, 21 facilities were found to have leaking underground
hazardous materials storage tanks. In order to incorporate
information from these facilities (which pre-date the question-
naire survey) into this report, two case studies were prepared.
In combination, leaks from the two facilities resulted in the
closing of more than a dozen water supply wells serving about
3,000 people and clean-up costs which were estimated to have
reached $20 million by May 1983. In addition, numerous law suits
have been filed in an attempt to establish responsi-bi 1 ity ¦ for the
leaks and to require payment of compensatory damages.
RELATIVE RELEASE PROBABILITY AND MAGNITUDE
To provide a basis for comparing the effectiveness of
alternative approaches to prevention and/or mitigation of re-
leases from underground tank storage systems, relative release
probabilities were estimated for a /typical" underground tank
x

-------
facility. This was accomplished through categorization of re-'
lease events, development of "typical" tank system character-
istics, and estimation of relative release magnitudes and proba-
bilities.
Six types of release events were identified,- including: tank
overflow, tank leak, tank rupture, ancillary equipment leak,
ancillary equipment rupture, fire/explosion and other incidents
(e.g., vandalism, earthquakes, etc.). For each type of release,
the causes of release were also categorized. The tank and
ancillary equipment leak and rupture release event categories
were found to share the same release cause categories (design
deficiency, installation practices, equipment failure and opera-
tional error).
The features of the "typical" tank system used for comparison
purposes were defined by the median of the EPA mailed question-
naire survey responses to the extent that data were available.
Thus, the "typical" tank system consists of a single 3,000 gallon
carbon steel tank which is filled through gravity cast iron
piping. The tank has been in service for 8 years and is used to
store ignitable waste. The tank was installed in accordance with
specifications commonly used at the time of installation in soils
which contribute to corrosion (resistivity less than 10,000 ohm-
cent i meters ) .
The release magnitude associated with each of these six
release ev'ent categories depends on the release rate and dura-
tion. Release rates are based on assumptions judged to be
conservative and release durations are based on assumptions
concerning the frequency of testing and tank level measurement.
Rased on the assumptions made, tank leak is the largest volume
release event for the small model facility, followed by tank
rupture and ancillary equipment leak. For the medium sized model
facility, tank rupture is the largest volume release event,
followed by tank leak. Tank and ancillary equipment leak were
judged to have the highest relative release probability.
PREVENTION MITIGATION MEASURES
Six measures intended to prevent or mitigate releases from
underground tank systems due to tank or ancillary equipment
failure were examined, including: secondary containment, tank
system testing, environmental monitoring, inventory monitoring,
internal inspection and corrosion protection. Each measure is
examined in terms of advantages and effectiveness, disadvantages
and limitations and equivalent uniform annual costs (FIJAC) for
two model facility sizes.
Secondary containment is shown to be the most expensive
(based on equivalent unifrom annual cost (EUAC)) of the control
methods examined for both the small and medium sized model
facilities under both new and retrofit conditions. ' Internal in-
spection is the second most expensive method, with corrosion
protection the least expensive method.
xi

-------
Although secondary containment is the most expensive of the
control methods examined, it is clearly the most effective means
of preventing both leak and rupture events since it is the only
method which reduces both the estimated probability and magnitude
of release. Corrosion protection also serves to reduce the
estimated release probability, and as shown, it also can control
all four release events. Other measures, such as tank system
testing and environmental monitoring, serve to mitigate the
effects of releases by decreasing the release magnitude but are
judged to have- relatively little impact on the estimated release
probabi1i ty.
From a release probability perspective, secondary containment
is the most cost effective method analyzed (under the conditions
assumed). This statement is made since secondary containment for
tank and ancillary equipment provides a three order of magnitude
greater decrease in release probability than corrosion protection
at a cost which is less than two orders of magnitude greater.
Secondary containment for both tank and ancillary equipment also
provides a 99 percent decrease in the estimated release magni-
tude. Although the cost associated with this approach is among
the highest shown, the cost per unit of release reduction is
approximately the same as for tank containment alone. Thus,
containment for the entire tank system is indicated to be a
better investment in light of the very significant reduction in
release probability provided.
Mitigation measures such as tank system testing and environ-
mental monitoring are shown to provide significant reductions in
the estimated release magnitude at costs per unit of reduction
which are about half those asssociated with secondary contain-
ment. However, they provide no reduction in the estimated
release probability.
Inspections are also shown to result in reductions in release
magnitude without impacting the release probability. While tank
inspection can result in the identification of developing prob-
lems before a leak or rupture occurs, measurements are taken on a
relatively small percentage of the tank surface area. Thus, it
was judged that while some reduction in the estimated relative
release probability will occur with tank inspection, the reduc-
tion will be less than one order of magnitude.
A prevention measure which has no impact on the estimated
release magnitude but which results in an estimated release
probability reduction of one order of magnitude is corrosion
protection. Based on the assumption that corrosion protection is
provided by an external coating and sacrificial anode(s), corro-
sion protection is shown to be the least expensive method of
achieving a reduction in estimated release probability.
xi i

-------
SECTION 1
INTRODUCTION
Hazardous wastes are stored using a variety of methods in-
cluding surface impoundments, tanks and containers. Since hazar-
dous waste mismanagement has been shown to have costly and dam-
aging consequences, there is a continuing interest in ensuring
that the management practices utilized will protect human health
and the environment. Recently, numerous cases of leaking under-
ground storage tanks have been discovered. As a result, a study
of underground hazardous waste storage facilities (defined as
tanks and appurtenances which are completely buried and are used
for storage of hazardous waste for more than 90 days) was ini-
tiated.
The objectives of this project were to define current under-
ground tank storage practices and to evaluate them in relation to
spill and damage event data and best engineering practice. Once
evaluated, this information was used to identify management al-
ternatives. Five management alternatives for mitigation/
prevention of waste release were then selected for evaluation,
which included examination of applicability, availability, com-
plexity, cost and effectiveness (expressed as the estimated rela-
tive probabilities and magnitudes of release).
The results of this investigation are presented in the fo1 - •
lowing four sections. In Section 2, -data derived from the EPA
Hazardous Waste Tank Questionnaire (0MB no. 2000-0424) are pre-
sented and discussed. These data provide a characterization of
underground tanks used for hazardous waste storage in terms of:
types of wastes typically stored, tank sizes, tank age, materials
of construction, methods of leak detection and frequency of use,
prevalence of tank linings, type of tank liners as a function of
waste type and type of tank liner as a function of waste type and
type of tank liner as a function of tank material.
Section 3 presents information regarding release events
associated with hazardous materials storage (most frequently
petroleum products). The, sources of this information were State
and Local agencies, trade associations and industry. In addi-
tion, two case studies associated with hazardous waste and mater-
ials storage are included. The implications of these data with
respect to the prevalence of tank systems failures are also dis-
cussed.
Section 4 presents an analysis of estimated relative release
probabilities and magnitudes associated with a "typical" under-
ground tank storage facility for seven types of release events
(i.e., tank leak, ancillary equipment rupture, fire or explosion,
etc.). The "typical" facility used for reference is based on the
most common current practice as determined from the data pre-
sented in Section 2 in conjunction with other relevant sources.
1-1

-------
Section 5 is a discussion of five waste release mitigation/
prevention measures selected to represent the range of possibili-
ties for reducing the relative probability and magnitude of re-
leases for underground hazardous waste storage tanks. Each mea-
sure is discussed with respect to both existing and new tanks.
The discussion provides a description of each option and the
associated costs, change in probabilities and magnitudes of re-
lease and advantages and disadvantages.
1-2

-------
SECTION 2
UNDERGROUND TANK USE FOR HAZARDOUS WASTE STORAGE
INTRODUCTION
To put the discussion of hazardous waste storage which
appears in the following three sections in perspective and to
provide input to the determination of representative facility
characteristics, a profile of underground tank used for hazardous
waste storage is presented here. The presentation is based on
responses to selected portions of a mail survey of the 1981
hazardous waste management practices regulated under Subtitle C
of the Resource Conservation and Recovery Act of 1976 (RCRA) con-
ducted for the U.S. Environmental Protection Agency's Office of
Solid Waste (EPA OSW) [1]. A description of the design of the
survey and how the responses may be used to generalize about all
hazardous waste storage tanks is currently being prepared [2].
The discussion presented is based on the questionnaire
responses with a focus on percentages of tanks with specific
characteristics. Data are also presented at the facility level
for selected characteristics such as overall underground tank
storage capacity.
TANK USE CHARACTERISTICS
Statistics on selected variables from, the mail survey were
found to be of interest for this report. One part of the survey
asked for a detailed description of all hazardous waste tanks at
a facility. Data were obtained concerning the tank descriptions
of the underground hazardous waste tanks. Variables selected
from the mail survey for inclusion in this report were as fol-
lows:
•	Capacity and age of underground tanks;
•	Interval of time between underground tank inspections;
•	Integrity testing of underground tanks;
•	Safety equipment on underground tanks;
•	Liners of underground tanks;
•	Construction material of underground tanks;
•	Wastes stored in underground tanks and at facilities with
underground tanks;
•	Number of underground tanks per facility at facilities
with underground tanks; and
9 Capacity of underground tanks per facility at facilities
with underground tanks.
2-1

-------
Following is a summary of the mail survey results concerning
underground tanks. The responses include a total of 169 under-
ground tanks, of which none were used for wastewater treatment.
Most of the tables and figures are based on data for less than
169 underground tanks. Many of the mail survey responses were
reported either as not ascertained, unknown, or with a blank. In
addition, some questions on the mail survey relate to only a sub-
set of the 169 tanks.
Table 2-1 lists the cumulative percent of underground tanks
by design capacity, volume contained and age. The median design
capacity and median "average volume contained" of underground
tanks are 3,000 gallons and 1,260 gallons, respectively. The
median age that an underground tank has been in use is eight
years, with the oldest tank in use for 35 years. Figure 2-1 re-
veals that the most frequent response to the number of years that
an underground tank has been in use is 10 years.
A total of 111 out of 168 underground tanks (66 percent of
168 responses) can be entered for internal inspection. The
median interval between internal inspections was reported to be
12 months based on 70 responses. Figure 2-2 also shows that the
most frequent response to the average number of months between
internal inspections is one year.
Many methods are used to check the integrity of underground
tanks; The percentages of underground tanks using different
types of integrity testing methods are as follows (based on 118
responses) :
Testing Method
U1trasoni c
Percent Using
0
Air
9
Penetrant dye
0
Vacuum box
0
Water/hydrostati c
13
Kent-Moore/Petro-tite
24
Other
37
Various types of safety equipment are employed for under-
ground tanks. The percentages of underground tanks using the
different types of safety equipment are as follows:

-------
TABLE 2-1. UNDERGROUND TANKS CHARACTERISTICS
Cumulati ve
Percent
Des i gn
Capaci ty
(gallons)
Average
Volume
Contai ned
(gallons)
Tank Age
(.years)
10
1,000
140
2
25
1,500
700
4
50
3,000
1,260
8
75
8,000
3,000
14
90
10,400
6,000
24
100
50,000
27 ,000
35
Total Number
of Responses
155
151
165
2-3

-------
1 2 3 4 5 6 7 8 9 10 11 13 14 15 16 17 18 20 22 23 24 25 29 30 31 34 35
Tank age (years)
FIGURE 2-1. Frequency distribution of tank age.

-------
3	6	12	15 24 36 48 60
Average number of months between internal inspections
Figure 2-2. Frequency distribution of internal inspections
of underground tanks.

-------
Safety Equipment
Percent Using
Lightning arrestors	13
Sparkless motors and wiring	37
Flame arrestors	27
Nitrogen blanketing	2
Other	25
The results of the type of safety equipment used are based on 167
responses.
The vast majority of underground tanks are constructed of
carbon steel. The percentages of underground tanks by type of
construction material are broken down as follows:
Construction Material	Percent Using
Carbon Steel	60
Stainless Steel	9
Concrete	17
Fiberglass	9
Other	5
The results of the type of tank construction material are based
on all the 169 underground tanks reported in the mail survey.
Data were also collected on the use of tank liners. Of the
39 underground tanks reported to have linings, most have plastic
liners (54 percent) or a liner made of a material other than rub-
ber, fiberglass, or steel (36 percent). Carbon steel tanks with
plastic liners make up the vast majority of lined tanks (43.5
percent of the 39 lined underground tanks). Of the 39 lined
underground tanks, most store corrosive wastes (72 percent) or
ignitable wastes (54 percent). The majority of lined underground
tanks store corrosive wastes in plastic-lined tanks (46 percent).
Many of these tanks store ignitable wastes in plastic-lined tanks
(28 percent).
Table 2-2 presents statistics on the types of waste stored
in underground tanks and the construction materials of the tanks.
Based on responses for all 169 underground tanks reported in the
mail survey, ignitable waste is the most common (46 percent)
waste type. Carbon steel tanks which store ignitable wastes
2-6

-------
TABLE 2-2. PERCENT OF UNDERGROUND TANKS BY TYPE
OF WASTE STORED AND CONSTRUCTION MATERIAL OF TANK


Percent
Used By
Construction
Material of
Tanks*
Waste
Stored
Percent
of Tanks*
Carbon
Steel
Stai n
less
Steel
Concrete
Fiber-
glass
Other
Ignitable
46
34
7
2
3
0
Corrosive
32
17
2
6.5
6.5
1
Reactive
11
1
2
4
3.5
•5
E.P. Toxi c
24
5
0
7
7
5
Toxic
28
14
3
6
5
0
Acutely
Hazardous
6
6
0
0
0
0
Other
10
10
0
0
0
0
* Percent of all underground tanks (169 tanks) with the specified type
of waste stored and construction material of tank indicated; i.e.,
46 percent of the tanks (78 tanks) store ignitahle waste, of which
34 percent of the tanks (57 tanks) store ignitable waste and are
constructed of carbon steel. Total of this column exceeds 100
percent since some tanks store waste which is classified as being in
more than one category.

-------
comprise 34 percent of all underground tanks. Acutely hazardous
wastes and other types of wastes besides ignitables, corrosives,
reactives, E.P. toxics, and toxics are stored only in carbon
steel tanks.
Selected cumulative percentages for the number of under-
ground tanks per facility with underground tanks were found to
be:
Number of Storage and/or
Cumulative	Treatment Tanks per
Percent		Facility	
10	1
25	1 •
50	1
75	3
90	6
100	15
As indicated, the median number of underground tanks per
facility with underground tanks is one since 55.4 percent of the
"65 facilities with underground tanks have a single underground,
tank. Figure 2-3 also indicates that most facilities with under-
ground tanks have one underground tank.- Figures 2-4 and'2-5 in-
dicate the distribution of underground storage and treatment
tanks respectively. As shown, the data for these two subsets of
underground tanks follows the same trend displayed by Figure 2-3.
In addition, these data indicate that most underground tanks are
storage tanks.
The majority of facilities with underground tanks (63 per-
cent of 65 facilities) store ignitable wastes in underground
tanks. The percent of facilities with underground tanks for each
type of waste stored in under ground tanks is as follows:
Waste Stored	Percent With Waste
Ignitable	63
Corrosive	28
Reactive	9
E.P. Toxic	18
Toxic	34
Acutely hazardous	3
Other	0
2-8

-------
Figure 2-3. Distribution of all underground tanks per facility
with underground tanks.

-------
2	3	4	5	6	7	9	10 13 15
Number of underground storage tanks/facility
Figure 2-4. Distribution of underground storage tanks per
facility with underground tanks.

-------
12	3	4
Number of underground treatment tanks/facility
Figure 2-5. Distribution of underground treatment tanks per facility
with underground tanks.

-------
The percentages of wastes stored are based on 65 responses for
each type of waste.
Selected cumulative percentages of underground tank capacity
per facility with underground tanks are listed below:
Cumulati ve
Percent
10
25
50
75
90
100
Capacity of Storage and/or
Treatment Tanks per
Faci1i ty
	(Gal 1ons )	
1,000
3,300
10,000
20,000
31,200
95,000
For all underground tanks, the median total capacity per facility
is 10,000 gallons, based on responses for 59 facilities.
In summary, results of th& mail.survey on underground hazar-
dous waste tanks reveal that a typical (median) underground tank
has the following characteristics:
Design capacity of 3,000 gallons;
Average volume contained of 1,500 gallons;
Installed for eight years;
Checked for integrity by method other than ultrasonic,
air, penetrant dye, vacuum box, hydrostatic, or Kent-
Moore/Petro-tite methods;
Constructed of carbon steel ;
Unlined; and
Stores i gnitable wastes.
Most facilities with underground tanks generally have three or
less underground tanks which typically store ignitable wastes.
The median capacity of underground tanks per facility with under-
ground tanks is 10,000 gallons.
2-12

-------
REFERENCES
1.	U.S. Environmental Protection Agency. Hazardous Waste
Treatment Storage and Disposal Questionnaire-Tank
Questionnaire. 0MB# 2000-0424. Washington, O.C. December
31, 1982 expiration date.
2.	Westat, Inc. Documentation of Mail Survey Questionnaire for
U.S. Environmental Protection Agency. Washington, D.C. (in
preparation).
2-13

-------
SECTION 3
DAMAGE CASE AND SPILL EVENT REVIEW
INTRODUCTION
One objective of this study was to assess the adequacy of
current practices for storage of hazardous waste in underground
tanks with regard to the protection of human health and the
environment. As part of this assessment effort, available data
on damage cases and spill events (releases) were reviewed in
order to determine the extent to which releases occurred and the
causes of these releases. For the purpose of this review effort,
"available data" were defined as readily available reports and
papers which contained compilations of individual incidents. In
addition, organizations with information regarding tank investi-
gation programs and detailed investigation of two specific sites
wereincluded.
A listing of the data sources included in this section is
presented below:
« American Petroleum Institute "Tank and Piping Leak
Survey
•	California Water Quality Control Board, San Fransisco Bay
Region
•	Cape Cod Planning and Economic Development Commission
•	Maryland Petroleum Association, "Prince George's County,
Maryland, Tank Testing Program
•	Michigan Department of National Resources
•	New York Department of Environmental Conservation
•	Suffolk County, New York
A summary of the overall findings resulting from the review
of these sources is also presented at the end of this section.
INFORMATION SOURCES
The relevant source of information evaluated during this
effort are individually discussed below. Each review was
prepared to provide an overview of the programs or incidents
responsible for initiating each study; to present the data
compiled during each effort; to point out some of the limitations
associated with the data presented; and to present conclusions
which can be drawn from the data.
American Petroleum Institute "Tank and Piping Leak Survey"
The American Petroleum Institute (API) "Tank and Piping Leak
3-1

-------
Survey" was conducted from the Fall of 1977 to the Summer of 1980
to "identify the location of perforations in leaking tanks to
support the effectiveness of tank testing procedures which mea-
sure liquid level loss in a tank". [1] As the study progressed
API requested additional information such as tank age, cause of
leak, leak detection method, piping system information, etc.. in
order to better understand the circumstances surrounding tank
leaks. This information was collected using a general question-
naire form distributed to the chief engineer (or the appropriate
individual (s) who handle reported leaks) at each of the major oil
companies (i.e., Exxon, Mobil, Shell, Gulf, ARCO, Chevron,
etc...) and to representatives of the Petroleum- Equipment Insti-
tute (PEI) (Note: Many of PEI's member organizations are in-
volved in supplying or installing replacement tanks.) These
questionnaires were then distributed to service station owners
(or managers) who had reported leaks from underground storage
systems and who volunteered to complete the survey form. Because
of this process, only leaking systems were reported (i.e., sta-
tions without leaks did not respond to the survey) and survey
forms were not completed for every leak occurring during the data
collection effort (i.e., since the survey was voluntary not all
stations with leaks completed survey forms). In addition, the
majority of responses w.ere from service stations owned by the
major oil companies.
The data from the survey were compiled by API and are pre-
sented below. As noted in Table 3-1, a total of-1953 leaks were
reported; 204 of these leaks could not be categorized. Some' 64
percent of the categorized leaks were attributed to steel tanks
without cathodic protection and 33 percent were attributed to
piping. (Note: Piping leak information was not requested on the
survey form until a year after the survey was started; by that
time, 400-500 questionnaires had already been collected. As a
result, more of the reported leaks may be attributable to piping
leaks.) The remaining categories only accounted for 3 percent of
the reported leaks.
A discussion of the results for each of the three categories
of leaks (from steel tanks, fiberglass tanks, and piping) is pre-
sented below. Questionnaires were not completed consistently,
which resulted in different numbers of responses for the various
question; many questions going unanswered; and a need to intepret
some of the answers. Although not statistically valid (i.e., the
total universe of stations was unknown, the survey was voluntary
and as a result not all stations with leaking tanks responded;
only data from stations with leaking underground storage systems
were surveyed; and the methods used to distribute the survey
forms tended to biase the results to represent conditions at
facilities owned by the major oil companies) the data shows rela-
tive frequencies and trends regarding leaks.
Steel Tank Leaks (see Table 3-2)
Corrosion was the primary cause of steel tank leaks
3-2

-------
TABLE 3-1. PETROLEUM PRODUCT LEAKS BY CATEGORY FROM
THE API "TANK AND PIPING LEAK SURVEY". (1977-1980)


Total
Percentage
Steel Tanks

1,112
63.6
Fiberglass Tanks

28
1.6
Steel Tanks with
Anodes
Sacri f i c i a1
2
0.1
Steel Tanks with
rent Cathodic
Impressed Cur-
Protect i on
19
1.1
Interior Coated
Steel
5
0.3
Piping

583
33.3
Subtotal

1,749
100.0
Unspecified Tanks

204

•TOTAL

1,953

3-3

-------
TABLE 3-2. STEEL TANK LEAKS IDENTIFIED DURING THE API
"TANK AND PIPING LEAK SURVEY". (1977-1980)
NUMBER OF RESPONSES
CAUSE OF LEAK
Corrosion Hole
Loose Fitting
Breakage
Other
Subtotal
Unanswered
AGE OF TANK
0-1 Year
2-5 Years
6-10 Years
11-15 Years
16-20 Years
21-25 Years
26-30 Years
31 - + Years
Subtotal
Unanswered
HOW LEAK WAS DETECTED
Inventory Shortage
Water in Tank
Leak Detector
Tank Test
Product in Sewers, Wells, Etc.
Other
Subtotal.
Unanswered
DID TANK HAVE A FILL TUBE?
Yes
No
Subtotal
Unanswered
1,112
Total	Percentage
970	92.3
9	0.9
17	1.6
55	5.2
1,051	100.0
61
2	0.2
14	1.4
117	11.8
262	26.5
296	30.0
176	17.8
80	"8.1
•41	4.2
988	100.0
124
134	17.5
584	55.4
3	0.3
122	11.5
45	4.3
116	11.0
1,054	100.0
58
84 5	30.5
205	19.5
1,050	100.0
62
3-4

-------
TABLE 3-2. (CONTINUED)
WAS LEAK BENEATH FILL TUBE?
Yes
No
Subtotal
Unanswered
WAS PART OF TANK IN GROUND WATER?
Yes
No
Subtotal
Unanswered
Total	Percentage
180	2 5.4
528	74.6
708	100.0
137
713	68.4
329	31,. 6
1,042	100.0
70
3-5

-------
accounting for 92 percent of the responses. Addi-
tional data pointed to external (63 percent of report-
ed cases of leaks due to corrosion) corrosion as the
primary type of corrosion.
Ages of leaking tanks ranged from less than 1 to more
than 31 years. Some 98 percent of the responses were
for tanks more than 6 years old and 86 percent were
for tanks in the 6 to 25 age range.
Water in the tank was the primary means of leak dis-
covery (this is usually found using a water-finder
paste on the bottom of the tank level gauging dip
stick), accounting for 55 percent of the finds. This
method of detection was followed by inventory shortage
and tank testing (primarily Petro-Tite) with 18 per-
cent and 12 percent of the responses, respectively.
(Note: .68 percent of the tanks were located in ground
water.) ; and
Of the 845 tanks reporting to have fill tubes, (i.e.,
a pipe extending from the surface down into the inter-
ior of the tank which is used for filling purposes) 21
percent reportedly had leaks beneath the fill tube.
The actual number may have been greater since the
question pertaining to this type of leak was not an -
swered for 137 of the tanks with fill tubes.
The results of the survey indicate that corrosion is a
major cause of steel tank leaks, with a notable percent-
age of .these leaks at the base of fill tubes. Additional
data showed that 22 percent of steel tank leaks reported
had some type of "point anode" (i.e., a point from which
electric current leaves the surface of the tank, resul-
ting in a destructive alteration or eating away of the
metal) at the leak point. These data, along with other
information presented, indicate that corrosion is influ-
enced or even enhanced by a number of factors such as:
The resistivity, pH, moisture content and sulphide
content of the soils surrounding the tank;
The existence of "point anodes" which may result from
foreign particles (i.e., cinders, clay etc...) on the
tank surface or physical damage of the tank coating
such as a scrape which may occur during installation;
and
Tank age. [2]
One of the most common causes considered is tank age, but
due to the broad range of ages over which leaks were re-
ported, age appears to be only one of possible variables
which influence the occurence of leaks due to corrosion.
3-6

-------
•	Fiberglass Tank Leaks (see Table 3-3)
With only 28 responses, one can conclude that it may
be less commonplace for fiberglass tanks to leak.
This conclusion is supported by the virtual elimina-
tion of the major source of leaks in steel tanks -
corrosion. On the other hand, this conclusion may be
biased by the smaller number (a total of 28 responses)
and the shorter duration of use of fiberglass tanks
(approximately 15 years) compared to steel tanks.
Breakage or tank separation (i.e., a physical separa-
tion of tank wall material) accounted for all leaks.
One third of these were caused by dip stick punctures
[l];
Tank age ranged from less than 1 year to 15 years with
96 percent of the responses falling between 0 and 10
years ; and
Inventory shortage was the primary means of leak de-
tection followed by water in the tanks.
Due to the limited number of responses from facilities
with fiberglass tanks which leak, few conclusions can be
drawn from the data. The principal point to be made is
that fiberglass tanks require careful handling during in-
stallation and operation (i.e., dip stick level measure-
ments) of the facility.
•	Piping Leaks (see Table 3-4)
Corrosion was the primary cause of pipe leak, accoun-
ting for 64 percent of the responses;
Pipe age ranges from less than 1 year to over 31 years
with 84 percent of the responses falling between 6 and
20 years.
Inventory shortage was the primary means of leak de-
tection, accounting for 45 percent of the responses.
The results of the survey indicate that with steel or cast
iron piping corrosion was a primary cause of release. This
is partially influenced by pipe age, but due to the wide
range of responses, pipe age is not the only factor that
should be considered (i.e., factors such as soil character-
istics and installation "practices may also influence leak
events).
The API "Tank and Piping Leak Survey" served its purpose in
identifying the locations of the leaks in tanks. [1] The addi-
tional data obtained during the survey, though not consistently
collected or statistically based, provides insight into tank and
pipe leak occurrences.. Additional conclusions can he derived
from these data:
3-7

-------
TABLE 3-3. FIBERGLASS TANK LEAKS IDENTIFIED DURING THE
API "TANK AND PIPE LEAK SURVEY" (1977-1980)
NUMBER OF RESPONSES
28

CAUSE OF LEAK
Total
Percentage
Breakage
17
60.7
Other (tank separation)
11
39.3
Subtotal
28
100.0
AGE OF TANK


0-1 Year
8
29.6
2-5 Years
7
26.0
6-10 Years
11
40.7
11-15 Years
1
3.7
Subtotal
27
100.0
Unanswered
1

HOW LEAK WAS DETECTED
.

Inventory Shortage
15
53.6
Water in Tank
9
32.1
Tank Test
1
3.6
Product in Sewers, Wells, etc.
1
3.6
Ot he r
2
7.1
Subtotal
28
100.0
3-8

-------
TABLE 3-4. PIPING LEAKS IDENTIFIED DURING THE
API "TANK AND PIPE LEAK SURVEY" (1977-1980)
NUMBER OF RESPONSES
CAUSE OF LEAK
Corrosi on
Loose Fitting
Flex Connector Failure
Breakage
Other
583
Total	Percentage
353	63.9
64	11.6
38	6.9
43	7 .8
54	9.8
Subtotal	552	100.0
Unanswered	31
AGE OF LEAKING PIPING
0-1 Year	10	2.1
2-5 Years	31	6.4
5-10 Years	158	32.8
11-15 Years	159	. 33.1
16-20 Years	87	18.1
21-25 Years	24	5.0
26-30' Years	11	2.3
31- + Years 		1_	0.2
Subtotal	481	100.0
Unanswered	102
HOW LEAK WAS DETECTED
Inventory Shortage	261	45.2
Water in Tank	19	3.3
Leak Detector	76	13.2
Line Test	82	14.2
Product in Sewers, Wells, etc.	60	10.4
Other	79	13.7
Subtotal	577	100.0
Unanswered	6
3-9

-------
•	Poor installation can contribute to leaks either by in-
ducing corrosion or resulting in loose fittings or tank
plugs (in many instances, loose fittings were tightened
and not reported on survey forms [1]); and
t Corrosion protection systems can fail as indicated by the
21 leaks reported for steel tanks with sacrificial anodes
or impressed current cathodic protect. These failures
may be result of a number of factors such as inadequate
sizing of the sacrificial anode, improper installation,
inadequate maintenance, equipment failure or other rea-
sons.
California Water Quality Control Board, San Francisco Bay Region
In September 1980, a case of ground water contamination
associated with underground storage of chemicals was discovered
at an electronic components manufacturing plant in Santa Clara
County. Subsequently, other plants in the region began to exa-
mine their hazardous materials storage practices. As a result of
these voluntary materials storage surveys, the California Re-
gional Water Quality Control Board, San Francisco Bay Region, be-
came aware of 21 facilities with leaks of hazardous materials
(mostly solvents) from underground tanks and sumps by the end of
1981.
In March 1982, the Regional Board initiated a 1eak. detection
program. The purpose of the leak detection program was to deter-
mine the overall magnitude of hazardous materials leakage (both
product and waste) from underground storage and handling facili-
ties in selected parts of the San Francisco Bay area. This leak
detection program was divided into three phases:
•	Detecti on - To determine all sources of hazardous mater-
i a 1s leaks to usable ground waters;
•	Remedi al Act ion - To identify the extent of leak contami-
nation, take remedial action to prevent further migra-
tion, and clean up contaminated ground water; and
•	Prevention - To develop construction and monitoring stan-
dards for underground storage and handling of hazardous
materials.
Activities performed by the Regional Board in each of these
phases are as fol1ows :
•	Detection - The Santa Clara Valley, Niles Cone, and
Livermore-Amador Valley ground water basins are important
supplies of potable water within the San Francisco Bay
region.. Figure 3-1 shows the location of the ground
water basins of concern. In April 1982, the Regional
Board sent a mandatory questionnaire to approximately
1,400 facilities within the three ground water basins.
3-10

-------
Figure 3-1.
Location of ground water
San Francisco Bay area.4
basins in the
3-11

-------
cities had enacted major portions of the model ordinance.
These eight cities require underground storage facili-
ties, including gasoline stations, to test their tanks
for leaks.. New or replacement tanks must have concrete
vaults or comparable forms of double containment. The
Regional Board has also worked to obtain statewide pas-
sage of the model ordinance.
The Regional Board is currently nearing completion of Phase
I, the Underground Leak Detection Program. A status report sum-
marizing the efforts of the Regional Board from April 1982 to
April 1983 is available. [3] Results of corrective actions
undertaken at sites with documented subsurface contamination
(Phase II) are not yet available. The effectiveness of preven-
tive measures (Phase III) has yet to be determined, since adop-
tion of the model ordinance occurred only recently. Because work
on Phases II and III is still in the early stages, results of the
program are presented only for Phase I.
As part of the Leak Detection Program, 1 ,294 out of 1,950
facilities responded to the mandatory questionnaire as of May
1983. Questionnaires for the remaining. 656 facilities were
either undel iverable (i.e., returned by the Post Office), receiv-
ed by the facility but not completed and returned, due at a later
date, or mailed to facilities outside of the study area. A total
of 429 facilities indicated that . they use or have used under-
ground tanks and/or sumps.
Of these, as	mentioned above, 87 facilities with under-
ground tanks and/or	sumps were judged to have the highest po-
tential for leaking	hazardous materials. As of May 1983, leak
monitoring and data	interpretation had been completed for 36 of
these 87 facilities,	with the following results:-
% of 36
No. of facilities	Comp1eted	Status
20	56	Contamination due to tank
system fai1ure
5	14	Contamination detected;
not due to tank system.
11	30	No contamination detected
Table 3-5 presents a comparison of facility character-
istics for the 36 facilities with known monitoring results and
the subset of 20 facilities with tank system failures. As shown,
the facility characteristics of the sites with tank system fail-
ures are comparable to the characteristies of all the sites with
known monitoring results. The typical facility has two under-
ground solvent storage tanks and one underground waste solvent
tank. Over 80 percent of the tanks are not vaulted and more than
one-half are steel. Corrosion protection consists mainly of
3-13

-------
Another 550 facilities not included in the original mail-
ing list were later sent questionnaires for a total of
1,950 questionnaires. Based on the responses to the sur-
vey, the Regional Board selected 87 facilities judged to
have a high potential for leaking hazardous materials,
especially solvents, and required these facilities to in-
ititate soil and/or ground water monitoring for under-
ground contamination. The 87 priority facilities are in
addition to the 21 ongoing cases discussed above. The
Regional Board placed the 87 facilities required to in-
stitute subsurface monitoring on either' a Priority 1 or
Priority 2 list. Priority 1 facilities have or have had
either a:
Non-vaulted buried waste solvent tank(s) without cor-
rosion pratection which was placed in operation before
January 1, 1975; or
Concrete sump(s) used for the storage, treatment,
separation, or disposal of solvents.
All other facilities which have or have had any product
or waste solvent tanks, regardless of installation date
or corrosion protection, were included in the Priority 2
list.
•	Remedial Action - Currently the Regional Board staff is
working with the faci1ities which reported a detectable
level of contamination in the soil and/or ground water.
This effort includes the 21 original cases (identified
prior to March 1982) and 36 out of the 50 subsurface in-
vestigations submitted to the Regional Board as of May
1983. (Fourteen of the 50 facilities were found to have
contamination, but the sources were not determined as of
May 1983.) The Regional Board is still waiting for re-
sults from 37 of the 87 facilities ordered to perform
subsurface monitoring.
Corrective measures undertaken by the industries in-
clude:
Identification of the lateral and vertical extent of
contaminant migration;
Actions to preclude further migration of contaminants;
and
Remedial action to cleanup contaminated ground waters
and soils.
•	Prevention - The Regional 3oard staff was actively in-
volved fn~ developing a model ordinance for underground
storage and handling of hazardous materials in Santa
Clara County. In March 1982, a task force established by
the Santa Clara County Fire Chiefs began meeting to
develop a model ordinance. Sixteen months later, eight
3-12

-------
TABLE 3-5. SAN FRANCISCO BAY REGION QUESTIONNAIRE
DATA FOR SELECTED FACILITIES*
Item
No.
36
Faci11ty
20 Facility
Subset
of
Tanks Percent
No. of Tanks
Percent
Unit type




product storage tank
68
50
48
57
waste storage tank
31
23
22
26
waste treatment tank
17
12
7
8
concrete sump
13
13
8
9
other
3
2
0
0
Vaulted unit




yes
24
18
12
14
no
113
82
73
86
Unit material




.steel
68
50
49
57
stainless steel
'3
2
1
. 1
concrete
23
17
12
14
fiberglass
9
7
0
0
alumi n um
11
8
11
13
other
2
1
2
2
unknown
21
15
11
13
Material contained in unit




sol vents
87
64
59
69
corrosives
2
1
2
2
wastewaters
.19
14
6
7
not in use
11
8
9
10.5
unknown
18
I3
9
10.5
Unit coating/wrapping


53
62
yes
68
50
no
34
25
11
13
unknown
35
25
21
25
Unit corrosion protection+


64
75
no
97
71
sacraficial anodes
1
1
0
0
impressed current
4
3
0
0
unknown
35
25
21
25
3-14

-------
TABLE 3-5. (CONTINUED)
Item
No.
36
Faci.l i ty
20 Facility
Subset
of
Tanks Percent
No. of Tanks
Percent
Corrosion protection




maintenance




yes
3
2
0
0
no
0
0
0
0
u nknown
134
98
85
100
Integrity checking of unit**




yes
57
42
39
46
n o
36
26
23
27
unknown
44
32
23
27
Internal inspection of unit




yes
20
15
12
14
no
117
85
73
86
Ground water monitoring




no
137
100
85
100
Tank testing




yes
26
19
24
28
n o
111
81
61
72
Inventory monitoring


14

yes
24
18
16
no
113
82
71
84
* Data for 36 facilities with known monitoring results (as of May-.
1983) and a subset of 20 facilities with tank system failures. <3
It should be noted that the data presented include all underground
tanks reported at the facilities. Information on which tank
systems have failed was unavailable.
+ Other than coating or wrapping.
** Integrity checking of unit indicates that one or more of the
following practices was performed prior to receipt of the
questionnaire: internal inspection of unit, ground water
monitoring, tank testing, inventory control, or another type of
integrity checking.
3-15

-------
coating and/or wrapping of the tanks. Most facilities report
that they do not provide internal inspection of tanks, ground
water monitoring, tank testing, or inventory control.
Table 3-6 reports the levels of contamination in the soils
and ground water at the 20 sites with tank system failures. The
chemicals are grouped by ranges of concentrations as follows:
•	Greater than 1 ,000 parts per billion (p p b);
•	Between 500 and 1,000 ppb;
•	Between 100 and 500 ppb; and
•	Less than 100 ppb.
In most cases the chemicals detected at the 20 sites were various
mixtures of a variety of industrial solvents including:
•	Acetone;
•	Benzene;
•	Dichlorobenzene;
•	Dichloroethane (DCA);
•	Dichloroethy1ene (DCE);
•	Ethyl benzene;
•	Freon;
•	Isopropyl alcohol (IPA)
•	Methyl ethyl ketone (MEK);
•	1,1,1-Trichloroethane.(TCA);
•	Trichloroethylene (TCE);
•	Tetrachloroethylene (PCE);
•	Toluene; and
•	. Xylene.
Although both ground water and soil data were not available for
all facilities, these data indicate that both ground water and
soil contamination were detected more frequently than either type
of contamination alone. At 10 facilities, soil contamination
levels exceeded 1,000 ppb for at least one chemical, while 11
facilities had ground water levels over 1,000 ppb. The chemicals
detected generally were distributed over a variety of concenr
tration ranges in both media for the 20 sites as a group.
As noted above, questionnaire data on the 21 facilities with
leak problems reported prior to initiation of the 3-phased leak
control program were not available. In order to incorporate in-
formation from these 21 facilities, which pre-date the question-
naire into this report, case studies at two of these facilities
were prepared. These case studies, which are presented as Exhi-
bits 3-1 and 3-2, describe the facility characteristics, the en -
vironmental setting, the release events and the associated conse-
quences. In combination, leaks from the two facilities resulted
in the closing of more than a dozen water supply wells serving
about 3,000 people and clean-up costs which are currently esti-
mated at about $20 million and are continuing. Numerous law
suits have been filed in an attempt to establish responsibility
for the leak and to require payment of compensatory damages.
3-16

-------
TABLE 3-6. CONTAMINATION FOUND AT 20 SITES IN SAN FRANCISCO BAY AREA WITH TANK SYSTEM FAILURES*

SOIL (PPB)
GROUNDWATER (PPB)
Facility
>1000
500-1000
100-500
<100
>1000
500-1000
100-500
<100
1
Bls-2-ethyIhexy1-
phthalate
Ethy1benzene
Naphthalene
8enze(a)pyrene
ND*
ND
ND
ND
2
NO
ND
MEK
ND
MEK
IPA
Cellosolve
Cyclohexanone
NO
ND
ND
3
NT*
NT
NT
NT
IPA
ND
NO
ND
4
ND
ND
CyanIde
NO
NT
NT
NT
NT
5
IPA
Olchlororoethone
ND
Acetone
ND
ND
ND
ND
6
• •



TCA
TCE
IPA
ND
ND
ND
7
ND
ND
ND
TCA
TCE
DCE
Toluene
Methylene Chloride
Benzene
Chlorobenzene
1,2-Trans-DCE
Ethy1 benzene
Methylene Chloride
Toluene
TCE
1,1,2-Trlchloro-
1,2,2-Trl-
flouroethane
MethyIcylohexane
TCA
ND
ND
8
Methylene Chloride
MEK
Acetone
ND
ND
NO
Methylene Chloride
MEK
Acetone
DCE
Ethy1 benzene
Methylene Chloride
ND
ND
ND
9









-------
TABLE 3-6. (CONTINUED)


SOIL
(PPB>


GROUNDWATER (PPB)
FaclIIty
>1000
500-1000
100-500
1000
500-1000
100-500
<100
10
ND
ND
ND
TCA
Xy lene
Trans-1,2-DCE
1,2-Dlchloro-l,2,2-
Trl-fluoroethane
1,1,2-Trlchloro-
fluoroethane
TCE
Vinyl Chloride
1,1-DCE
II
to
ND
ND
ND
HtK

ND
ND
12
HEK
Cyclohexanone
IPA
ND
ND
M)
MEK
Cyclohexanone
IPA
Acetone '
Xylene
ND
Toluene
Freon
TCA
TCE
13
M)
ND
Toluene
ND
Ethy1benzene
Xylene
TCE
Chlorobenzene
Toluene
Dlchlorobenzene
ND
PCE
Benzene
DCE
OCA
Freon
M
Tr 1 ch 1orobenzene
Dlchlorobenzene
ND
TCE
Freon
PCE
TCA
ND
TCE
Freon
ND
PCE
HeMane
Acetone
Ethy1 Benzene
Toluene
Benzene
15
Phenol
Methanol
TCA
IPA
Xylene
n-Butyl Acetate
ND
Acetone
Trlchloro-
Huoroethane
Methylene Chloride




16
Stoddard Solvent
ND
ND
ND




17
TCE
Xylene
Freon
ND
ND
ND
ND
ND
ND
18
NO
ND
ND
ND
Diesel
Naptha
Xylene
Toluene
Cellosoive
Acetate
MEK
M)
ND

-------
TABLE 3-6. (CONTINUED)


SOIL
(PPB)


GROUNDWATER (PPB)
FaclfIty
>1000
500-1000
100-500
<100
>1000
500-1000
100-500
<100
19
o
Dlchlorobenzene
Freon
TCE
PCE
DCE
NO
Chloroform
TCA
NO
Xylenes
Ethy 1 benzene
TCE
DCE
PCE
20
Methylene Chloride
Oxyblsethanol
Heptanone
NO
NO
Xy fene
Cyclopetnane
Methoxyethanof
Methoxypropanol
Methylene Chloride
Caprolactum
Hexanolc Acid
Hexanol
Heptanot
Octanol
Acetone
Other Heptanols
*	See Table 3-5 for additional Information on these facilities
*	Mot detected
' Not tested
¦" Blank Indicates that data were not available as of May 1903
Key to abbreviations;
DCA 3 Dlchloroethane
DCS ¦ Olchloroethylene
IPA * IsopropyI alcohol
MEK 3 Methyl ethyl Ketone
PCE =¦ Perchloroethylene 
-------
In summary, the Regional Board has discovered numerous leaks
of solvents from underground storage systems in the San Francisco
Bay Region and more are expected to be found. The Regional Board
has nearly completed the first phase of its program to identify,
correct, and prevent chemical leaks from underground storage sys-
tems. As of May 1983, tank system failures had released solvents
into the soil and/or ground water at 41 sites (21 sites were
identified before the questionnaire was developed and the 20
sites shown in Table 3-5). This represents 72 percent of the 57
facilities with know monitoring results as of May 1983, and near-
ly 10 percent of the 429 facilities found by the survey to use
underground tanks or. sumps in the Santa Clara Valley Region.
Remedial measures at these facilities and at other sites
where contamination has been detected but not linked with tank
system failures are continuing. The Regional Board judged the
potential hazards associated with leaks to be high enough to sup-
port efforts to develop and enact ordinances at the local level
and to assist in the development of legislation statewide to
regulate underground tanks storing hazardous materials.
Cape Cod Planning and Economic Development Commission
After completing their 208 Water Ouality Management Plan in
1978, the Cape Cod Planning and Economic Development Commission
(CCPEDC) developed a number of model groun-d water protection by-
laws and regulations as guidelines for a ground water protection
program. Since May, 1980, 14 of the 15 Cape Cod communities have
adopted one or more of these ordinances which have resulted in
various levels of requirements such as tank registration, tank
inspection and zoning restrictions in ground water recharge
areas. Since enactment of these ordinances, eight of the more
than 159 underground tanks tested were found to be leaking (in-
formation on the details of these leak events was unavailable).
A telephone conversation with the local health official in Barn-
stable, Massachusetts revealed that many of the larger oil com-
panies replaced the steel tanks at their service stations as soon
as the ordinances were passed. [4]
Maryland Petroleum Association, "Prince George's County,
Maryland, Tank Testing Program"
In 1977 The Prince George's County (P.G. County) Government
passed legislation requiring tank and piping system testing for
tank storage facilities in response to a number of gasoline leak
incidents at service stations in the County. Although P.G.
County does not maintain statistics on its tank testing program,
the Maryland Petroleum Association compiled the results of tests
conducted on underground tank systems as of January, 1978.
These data are presented in Table 3-7 and represent the re-
sults of Petro-Tite (Kent-Moore) testing of service station tank
systems. (Note: Even though piping system testing was included
in these tests, no distinction between tank and pipe leaks was
made in the available statistics.) It is important to note that
3-20

-------
only petroleum products and waste oils associated with service
stations are stored underground in P.G. County and that all of
the tanks tested as of January, 1978 were more than 10 years
old. [6]
As shown in Table 3-7, 18 percent of the tank systems tested
by the Petro-Tite method were indicated to be leaking. Further
investigation of these 108 tanks revealed 61 "verified" leaks (10
percent versus 18 percent of tanks tested). (Note: Information
on how leaks were verified could not be obtained.) These data
indicate that the tank testing method used (under circumstances
of application and verification about which little is known)' was,
at best, about 50 percent accurate. However, the testing ap-
proach used did identify a significant number of leaking tanks
which were subsequently removed from service or reconditioned.
Michigan Department of Natural Resources
As a result of a growing number of reports of spills and
leaks from underground storage of petroleum fuels and an in-
creasing awareness of the potential for ground water contamina-
tion, the Michigan Department of Natural Resources (DNR) under-
took an investigation to evaluate the problems associated with
the underground storage of petroleum fuels. This evaluation re-
sulted in a report which was released in September 1981, and con-
tained information pertaining to spill and leak events in the
State. [7] •
Approximately 25,,000 underground commercial fuel tanks are
in use in the State. These do not include abandoned, private or
underground bulk storage tanks. As reported in Michigan's Pollu-
tion Emergency Alerting System (PEAS) files, a total of 396 re-
ports of pollution of soils and/or groundwater by petroleum fuels
from underground tanks were submitted from 1977 to 1978. A
breakdown of these reports showed 30 percent were due to overfil-
ling, 26 percent were leaks from underground tanks, 9 percent
were pipe leaks, and 36 percent from unknown sources. Another
study completed in 1978 that assessed ground water contamination
in Michigan, showed that 21 percent of the 268 known ground water
contamination sites "involved petroleum contamination either
known or suspected to be from underground tanks". [7]
The data presented above only represents releases reported
over a two year period. On-going work by the DNR is finding that
more releases are reported from the discovery of gasoline in
drinking water wells, subsurface construction sites and buried
cable systems than from reports of spills or product loss from
tanks. [7] This leads one to believe that the study conducted
may have only identified a small portion of the total number of
leaking tanks. In addition, it should also be noted that while
30 percent of the reported cases are due to tank leaks and 9 per-
cent of the reports are due to pipe leaks, it is plausible that
many of the reported cases from unknown sources are also likely
to be due to one of these two events.
3-21

-------
TABLE 3-7. PRINCE GEORGE'S COUNTY, MARYLAND TANK TESTING
PROGRAM RESULTS FOR UNDERGROUND GASOLINE TANKS
AS OF JANUARY 1978.
Number	Percent
Tank systems tested	604	40*
Tank systems that failed the test	108	-18
Number of verified leaking tank systems	61	10
(56 percent accuracy of test results)
* All tanks tested as of January 1978 , were more than 10 years
old. The tested tanks represent approximately 40 percent of
all underground commercial gasoline tanks in the County, based
on extrapolation from an average of 3.72 tanks per station for
310 (with known numbers of tanks) of the total 406 stations in
the County.
Participating: Amoco, BP, Cities, Crown Central, Exxon, Gulf,
Mobil, Phillips, Shell, Sun, Texaco and Tenneco
Source: Maryland Petroleum Association. [4],
3-22

-------
New York Department of Environmental Conservation
The New York Department of Environmental Conservation (DEC)
undertook a two-year bulk storage study program in an effort to
reduce petroleum and hazardous liquid leaks and spills into the
environment. As part of their program, information was compiled
on the number of underground tanks in the State and the inci-
dences of well contamination by gasoline. Additional data per-
taining to oil spills reported in 1979 were also included, but no
distinction was made between underground and aboveground tank in-
cidences.
The DEC estimates that there are 83,000 functioning under-
ground tanks in the State and that 20 percent of these currently
leak. The methods used to derive these estimates are presented
in Appendix I. In addition, the State expects that many of the
estimated 28,000 underground tanks that have been abandoned over
the past 10 years contain materials (primarily gasoline) which,
if the tanks are steel, will leak once the tanks corrode. [8]
In a 1979 survey of local health units in New York, 187
wells were reportedly contaminated by gasoline. The information
obtained from this survey is presented in Table 3-8. [8]
The work that New York has conducted to date shows that a
significant • number of wells have been contaminated as a result of
leaks from underground storage facilities (primarily petroleum).
These figures, which are four years old and do not 'cover the
whole state, combined with the estimated number of current leaks,
indicate that more well contamination incidents may have already
occurred or will occur in the future.
Suffolk County, New York
In September, 1979, Article 12 of the Suffolk County San i -
tary Code was enacted to control ground water contamination re-
sulting from the storage of hazardous materials in underground
and aboveground tanks. As a result of' the permitting, inspection
and testing program subsequently conducted by the Suffolk County
Department of Health Services, the information presented in Table
3-9 was obtained for underground tanks. [9]
These data represent the results of the first phase of
regulation implementation, conducted from 1980 to 1982 for all
tanks 20-years old or older. Additional phases of implementation
are currently underway which will eventually result in the per-
mitting, inspection and testing of all tanks in the county.
As shown above, of the 4554 underground tanks registered
(primarily petroleum product storage tanks) as of December 1982,
1024 privately-owned and 82 county-owned tanks over 20 years old
had been tested. The test results showed that approximately 10
percent (98 privately-owned and 15 county-owned tanks) of the 20-
years old or older underground storage tanks were leaking. If
piping system leaks were included in these statistics the number
3-23

-------
TABLE 3-8. SOURCES OF WELL CONTAMINATION BY GASOLINE AS REPORTED
IN A SURVEY OF LOCAL HEALTH UNITS IN NEW YORK IN 1978*

Sources of Contamination
Number of
Incidences
Percent of
Total
Gasoline Stations
94
50
Buried Gasoline Tanks at Sites
Other Than Gasoline Stations
16
8
Other**
24
14
Unknown
53
28
TOTAL
187
100
* Data from 13 counties in the state were not included in this
survey.
* * Includes contamination from other sources such as transfer
spills, tank truck accidents, etc...
3-24

-------
TABLE 3-9. RESULTS OF SUFFOLK COUNTY, NEW YORK
Tank Testing Program as of December 1982.
Privately owned tanks tested*	1,024
Privately owned tanks leaking	98
County owned tanks tested*	^ 92
County owned tanks leaking	15
Percent of total tanks leaking	10
Tanks registered as of December 1982
(includes all tanks all ages)	4,5 54
* All tanks tested were >20 years old and almost all were steel.
Tanks were tested using the Petro-Tite (Kent-Moore) test under
the supervision of Suffolk County Department of Health Services
Personnel .
(Source: Article 12, Suffolk • County Sanitary Code Statistics
(December 1982))	¦	>
3-25

-------
of leaking systems would be closer to 30 percent. [10] (Note:
Statistics on piping system leaks were not available.) Specific
information as to the causes, volumes, durations or impacts of
these leaks was not available.
As noted above, the implementation of Article 12 has re-
sulted in the discovery of a number of underground storage system
leaks, the majority of which can be attributed to piping system
leaks. As a result of the efforts of Suffolk County Department
of Health Services, these leaks have been remedied and a number
of tanks (911) have been removed or abandonded. However, it can
be assumed that additional leaks will be discovered at facilities
not yet tested, even though the remaining universe of tanks is
less than 20 years old. (This assumption is supported by work
conducted by Warren Rogers Associates [2] (see Appendix J) which
indicates that tank age is not the principal factor controlling
when a tank will begin leaking.)
CONCLUSIONS
The reports, studies and papers presented in this section,
though not all-inclusive, document a number of cases of leaks
from underground storage facilities as shown. (Note: A majority
of these cases are product related (primarily petroleum) due to
the historical awareness of the costs associated with product
loss.) The following conclusions can be drawn from the informa-
tion reviewed.
•	A large number of leaking underground storage tank sys-
tems have been discovered (primarily petroleum product
storage) over the past 6 years, and indications are
that many more will be discovered in the future as in-
vestigations continue and awareness increases. This is
supported by the efforts in the San Francisco Bay
Region where investigations showed that 72 percent of
the facilities tested (with test results available as
of May 1983) had one or more leaking tank systems (see
Table 3-10) .
•	Due to the range of percentages of leaking tank systems
to tank systems tested (see Table 3-10), it is diffi-
cult to draw quantitative conclusions as to the extent
of the problem of underground storage.
•	Once a leak occurs it may go undetected for years and
may result in clean-up costs totalling in the 10's of
millions of dollars, as evidenced by the case studies
presented in this section.
« The impact from underground storage releases may be
effected significantly by the geologic conditions of
the area.
3-26

-------
TABLE 3-10. SUMMARY OF REPORTED TANK SYSTEM LEAKS
Source
Number Tank
System Leaks
Reported
Number of
Tank Systems
Tested
Tank System
Leaking
Tank System
Tested %
Uni verse
of Tank
System
API "Tank and Pipe Leak Survey
(Petroleum Products)
1,953
Unknown
NA
Unknown
California Regional Water Quality
Board, San Francisco Bay Region
Statistics^
41
57
72
429
Cape Cod Area Statistics
(Petroleum Products)
8
159
5
Unknown
Michigan DNR "Underground Gasoline
Storage Study"
452
Unknown
NA
25,000
New York DEC Statistics
(estimates)
Prince George's County, Maryland
1977 Statistics (Petroleum Products)
16 ,000
(estimated)
61
NA
604
20
(esti mated)
10
83,000
(esti mated)
Unknown
Suffolk County, New York
Statistics (Petroleum Products)
103
1,116
9
Unknown
1)	Tank systems include both tanks and pipe leaks for a facility
with one or more underground storage tanks.
2)	These values represent the number of facilities with one or more
tank systems. The number of tank systems per facility is unknown
(these range from 1 to >100 underground tanks per facility) and the
number of leaking systems per facility is unknown.

-------
In addition to the figures presented in Table 3-10, Warren
Rogers Associates, which has collected data from approximately
10,000 gasoline tank storage sites in the United States and
Canada, estimates that there are currently 75,000 leaking gaso-
line storage tanks in the U.S. [11]
If one assumes that many of the existing underground haz-
ardous waste storage facilities employ . simi1ar storage practices
(i.e., unprotected steel tanks and piping), an assumption which
appears to be confirmed by available data, the potential for sim-
ilar problems occurring is probably significantly higher than for
gasoline unless installation methods and designs are improved.
This is based on the assumption that hazardous waste storage
facilities store a variety of wastes, some of which may be corro-
sive or incompatible with tank materials used at the facility.
This increases the chances of operator error (e.g., storing waste
in the wrong tank or not testing a waste to determine which of
the several tanks to store it in) and, as a result, increases the
possibility of tank failure.
3-28

-------
EXHIBIT 3-1
SITE A STORAGE SYSTEM FAILURE DAMAGE CASE SUMMARY
FACILITY INFORMATION
Site A manufactures electronic components (SIC 367) and uses
a variety of solvents or solvent-based chemicals in the manufac-
turing process including:
•	Acetone;
•	1,1-Dichloroethylene (DCE);
•	Freon 113; o Hexelmethyldisi 1ane (HMOS);
•	Isopropyl alcohol (IPA);
•	Methyl alcohol ;
•	1,1,1-Trichloroethane (TCA); and
•	Xylene
Waste solvents, generated by the manufacturing operations are
stored	in one of three ways, as follows:
•	Containers -- A variety of strippers with propriety for-
mulations supplied by outside vendors are used in the
manufacturing process. Since the specific chemical for-
mulations of these materials are not known by operating
personnel at Site A, waste strippers are seg- regated for
storage by using containers to avoid potential compata-
bility problems'.
e Small waste tank -- Most mixing of chemicals for use in
.the production process occurs in one area of the plant.
In this area, containers in which chemicals are re-
ceived are cleaned so that they can he disposed in a san-
itary landfill. Waste from the container washing process
are stored in a 550 gallon underground tank.
•	Large waste tank -- Formerly, waste solvents generated
throughout the plant were collected from sink drains with
a gravity piping system and stored in a 6,000 gallon
underground fiberglass tank. Following failure of this
tank, which is the subject of this damage case discus-
sion, waste solvents have been stored in a temporary
1 ,000 gal 1 on tank.
ENVIRONMENTAL SETTING
Site A is located in a suburban area adjacent to a major
city and is surrounded by residential neighborhoods and small
farms. Shopping centers, park land, and other industrial facili-
ties are also situated in the i mmedi ate vi ci n i ty. The site lies
in a valley approximately 210 feet above sea level between a
ground water recharge area and several water supply wells.
The geology of the broad alluvial valley surrounding Site A
is the result of active stream erosion and deposition. Streams
3-29

-------
flowing out of the highlands and into the valley have deposited
large quantities of debris as alluvial fans and outwash plains.
The alluvial sediments range in thickness from zero along the
hills bordering the valley to 400 feet in the center of the
valley. These alluvial fan deposits are very permeable, and the
discontinuous clay beds in the area are poor barriers to vertical
ground water migration. The large number of high production
(over 1,000 gpm) water wells reflects the permeability of the
alluvial sediments. Within a one-mile radius of Site A, there
are 25 active or potentially active water supply wells.
The alluvial deposits at Site A vary in thickness from 330
to 360 feet and contain four aquifers. The aquifers are desig-
nated "A", "B", "C", and "D" with increasing depth from the
ground .surface, as follows:
Approximate Depth Below
Aquifer	Ground Surface (Feet)
"A"	50
"B"	60 - 100
"C"	150 - 190
"0"	220 - 270
All four aquifers average approximately 40 percent sand and gra-
vel over their total depth. These deposits are separated by silt
and silty clay layers ranging .from a few feet to 60 feet in
thickness. Aquifers "A", "B", and "C" have percentages of silt
and clay varying between 3 percent and 19 percent. Aquifer "D"
has a slightly higher silt and clay content. Although the silt
and clay layers separating the aquifers at the site are discon-
tinuous, they cause ground water to flow primarily in a horizon-
tal di recti on.
Primary recharge to the aquifers under Site A comes from in-
filtration ponds along a creek situated approximately 4,000 feet
to the east. Ground water elevations indicate a local flow to
the west, except when irrigation wells north of Site A cause the
flow to be in a more northwesterly direction. A well owned by a
local water company that is part of a drinking water supply sys-
tem for about 700 residents is located approximately 2,000 feet
northwest of Site A. Thus, ground water flow from the primary
aquifer recharge area passes through Site A toward drinking water
sources.
Recharge to the aquifers in the region also occurs by infil-
tration of irrigation water applied to lawns and agricultural
lands and by percolation of percipitation. The estimated direct
recharge from irrigation water is small relative to recharge from
the percolation ponds. Average annual precipitation for the site
area is moderate. (Seasonal rainfall at a nearby weather station
averages 14.2 inches (360 millimeters) per year.)
3-30

-------
RELEASE FACTS
As a result of construction activities unrelated to Site A,
waste solvent storage tanks, solvent contamination of soils and
subsequently ground water was discovered. Follow-up investiga-
tions identified a 6,000 gallon fiberglass waste solvent storage
tank as the source of the solvent contamination. Visual inspec-
tion of the tank following excavation revealed that the tank
walls had deteriorated to the extent that in some areas only the
reinforcing ribs remained. The cause of this tank wall failure
has not been determined and is currently under litigation.
The duration and maximum magnitude of the leak has been es-
timated at 1-1/2 years and 58,000 gal 1ons , respectively, based on
a mass balance analysis of solvent purchase and waste removal
records. Ideally, a mass balance analysis can be performed by
matching the total mass of materials entering a fixed system with
the total mass of all material's leaving the system plus any
accumulation. However, correlation of solvent purchased with
waste solvent removed is difficult under real industrial condi-
tions for several reasons. First, solvent is usually purchased
well in advance of the time it is actually used. Second, waste
solvent is not removed until the holding tanks or drums accumu-
late a specified volume. Moreover, some solvent remains either
in the original container or on the surfaces of the material
cleaned. Thus, mass balance variances of 5 to 10 percent can be
expected due to these factors. In addition, approximately 11
percent of all solvent is lost through evaporation.
Table 1 summarizes the results of the mass balance analysis
conducted for Site A. As shown, essentially ¦ al1 of the solvent
purchased at Site A is accounted for during years 1 to 4-1/2 and
year 6. However, an imbalance between "solvents used" and "total
out" began in the middle of year 4 and continued until the leak
was detected near the end of year 5. During this 1-1/2 year per-
iod, the facility can account for the removal of only 43 percent
of the solvents used. Thus, the maximum amount of solvent lost
appears to equal 57 percent of the total solvent used or approxi-
mately 58,000 gallons. The amount of 1,1,1-trichloroethane (TCA)
used is included in the table since high concentrations of this
solvent were found in the soil and ground water at Site A, as
described later in this report.
During the year following the discovery of the leaking tank,
Site A had approximately 76 wells drilled on and off their prop-
erty to determine the areal extent and severity of contamination
resulting from the tank leak. In addition to 28 on-site and 40
off-site observation wells, there were 5 on-site and 3 off-site
pumping wells installed. The facility also performed sampling at
9 nearby irrigation and drinking water wells. Since ground
water flow at Site A moved in a westerly to northwesterly direc-
tion, the welis radiated out from the facility in this direction
(downgradi ent) for a distance of approximately 1 mile from the
site.
3-31

-------
TABLE 1. SOLVENTS USED v.' SOLVENTS REMOVED AT SITE A
(IN GALLONS!


Solvents Used
*
Solvents 1
Removed**

Sol vents

Year
Months#
Total Solvents Used
(TCA)##
Drums
Bulk
Evaporation+
Total Out++
Lost %
1
1-12
1,587
62
0
0
175
175

2
13-24
9,755
79
2,875
5,040
1,073
8,988

3
25-36
18,081
327
4,260
8,758
1,989
15,004

4
37-42
16,952
671
14,814
. 5,000
1,865
21,680

SUBTOTAL
46,375
1,139
21,949
1 8,798
5,102
45,849
1
4
43-48
35,726
386
19,032
1,400
3,930
24,360

5
49-60
66,720
3,570
9,465
2,904
7,339
19,708

SUBTOTAL
102,446
3,956
28,497
4,304
11,269
44,068
57
6
61-72
19,068
3,614
6,750
9,425
2,097
18,272
4
* Based on three-month moving average of solvents purchased. ++ Quantity accounted for by removal or evaporation
** Based on date solvent accumulated for removal.	H Consecutive from facility startup.
+ Estimated	## 1,1,1-Tri-chloroethane

-------
A variety of well drilling and construction methods were
employed at Site A. Borings were drilled by either a continuous
flight, hollow stem auger; a mud rotary rig; the reverse cir-
culation drilling method; or the caisson auger technique. Var-
ious diameters of steel and PVC casing were used in the wells.
Although different types of well casings and other construction
methods may affect the accuracy of sampling results, no informa-
tion was available on how these factors may have affected the
analytical results.
Ground water level measurements were taken using a Soil Test
M-scope. Most ground water samples were obtained using a sub-
mersible bladder, and in a few cases a teflon bailer was used.
Analysis of the ground water was performed in accordance with EPA
Standards. Soil samples were taken with splitspoon and auger re-
turn samplers.- Physical testing of the soils included moisture
content, dry density, liquid limits, plastic limits, grain size
distribution and permeability. Chemical testing of the soils
involved several methods. Most soil samples were analyzed by
purge and trap/gas chromatography/f1ame ionization detection
(PAT/GC/FIQ) . Some were analyzed by purge and trap/gas
chromatography/mass spectrometry (PAT/GC/MS) for quality control
or improved quantification.
Results of this ground water and soil chemical testing
program revealed a solvent (especially TCA) plume in aquifers
"A", "B", and "C" .which extended approximately 4 ,500 feet west-
northwest of the site with a maximum width of about 2,000 feet.
The highest concentrations of solvents in the soil and in the
ground water were obtained from auger caisson borings 32 to 38
feet below the ground surface and from aquifer "A" monitoring
wells located within 50 feet of the former waste solvent tank,
respectively. Table 2 reports the mean solvent concentrations
found at Site A within 50 feet of the tank that failed.
REMEDIAL MEASURES
Remedial measures undertaken at Site A included on-site and
off-site work. The on-site remedial effort included removal of
soils in the area of the former waste solvent storage tank since
these contaminated soils had the potential to act as a continuing
source of solvent to the ground water system. In addition, a
series of ground water purge wells were installed to hydraulical-
ly contain solvents on-site. The off-site remedial plan involved
the placement of a four-tiered ground water purge well system to
reduce the width and length of the solvent plume. This, series of
redundant recovery wells was designed to lower the concentration
of TCA which had contaminated and resulted in the closure of the
drinking water well located 2,000 feet from the faulty tank. The
well system also was installed to prevent TCA from reaching an-
other drinking water we 11 situated approximately 6,000 feet in a
hydraulically downgradient direction.
3-33

-------
TABLE 2. MEAN SOLVENT CONCENTRATIONS FOUND IN THE SOIL
AND GROUNDWATER AT SITE A (PPM)*
Soil	Ground Water
(Dry Weight)	(Wet Weight)
1,1,1-Trichloroethane (TCA) 1,000	540
Xylenes and ethyl benzene 600	80
Acetone 1,000	25,000
Isopropanol (IPA) 3,000	43,000
Freon 113	.18
* Solvent concentrations found within 50 feet of the tank that failed.

-------
The augered caisson method was considered to be the only
cost-effective way to remove the soils surrounding the former
leaky tank. Open excavation would have undermined the building
footings at Site A and the use of tie-back pilings was considered
too risky and costly. The soil removal area extended approxi-
mately 50 feet wide by 65 feet long and 52 feet deep. The es-
timated 3,400 cubic yards of soil removed and transported to a
Class I disposal site contained an estimated 38,000 -pounds of
solvent.
Water pumped from the ground water purge wells located on
the periphery of the soil removal area was loaded into tank
trucks and hauled to a licensed off-site disposal facility.
Pumping the other on-site recovery wells which are downgradient
of Site A near the property line lowered the water levels in ob-
servation wells beyond the solvent plume. Water from these wells
is treated by carbon absorption and discharged to a nearby creek
via storm sewers. TCA concentrations, at one of these wells de-
creased from 6.8 ppm to 0.55 ppm after the first three months of
pumping. However, the length of time required to reduce the sol-
vent concentrations in the on-site ground water purge system to
stable and acceptable levels is unknown.
The offsite drinking water supply well that was closed be-
cause of solvent contamination was returned to service as a
ground water purge well. Water from this well was treated by a
carbon absorption system at Site A and discharged to storm
sewers. Treatment of ground water from this wel-1 was stopped
after one year when TCA concentrations fell sharply and met dis-
charge permit discharge limits. The state has yet to determine
what residual level of TCA is acceptable for drinking water.
Presently the state's action level for TCA in the ground water is
0.3 ppm.
Data from the tiers of other off-site ground water recovery/
observation wells indicated a reduction of solvent concentrations
within the plume and a reversal of downgradient migration. The
one aquifer "A" well located near the second tier of the ground
water purge system rarely showed any detectable levels of chemi-
cals. Solvent concentrations in the "8" aquifer decreased by ap-
proximately one order of magnitude from the first tier to the
third tier of off-site observation wells about 3,000 feet apart
where TCA levels dropped to less than 0.005 ppm after one year of
pumping. The highest TCA concentration measured in the off-site
"B" aquifer decreased from 11 ppm to 0.12 ppm after less than 12
months of pumping. The greatest TCA level in the off-site "C"
aquifer was found approximately 3,400 feet from the faulty tank
and dropped from 0.23 ppm to 0.15 ppm in less than a month of
ground water purging. Concentrations of TCA recorded for aquifer
"D" off-site wells never exceeded the permit limit and most were
not detectable. Freon and DCE were the only solvents other than
TCA detected off-site. The maximum levels of freon and DCE were
recorded in aquifer "B" about 1,000 feet from the waste solvent
tank at 0.026 ppm and 0.047 ppm, respectively.
3-35

-------
As of May 1983, remedial measures involving purge well pump-
ing and treatment of the extracted ground water were continuing.
Completion of these activities is dependent on state agency
acceptance of aquifer water quality, but may be accomplished by
the end of 1983.
RELEASE CONSEQUENCES
As of May 1983 , Site A had spent an estimated $12 million
over 1-1/2 years on cleanup of the contamination, and costs con-
tinue to be incurred. Although pollution levels have been re-
duced significantly as a result of remedial measures, engineers
agree that pumping will probably never completely remove the con-
taminants from the aquifers.
One drinking water supply well located about 2,000 feet from
the leak site, which served 700 residents, was closed because of
high TCA concentrations. Several individual water supply wells
were also closed, and the leak may also have resulted in minor
contamination of another major drinking water source about 6,000
feet from the tank.
The spill also spawned a multimillion-dollar lawsuit by
nearby residents who have charged the site with negligent con-
tamination and with being the cause of numerous birth defects in
the neighborhood. Site A maintains that no scientific link has
been established between tts leak and the alleged high number of
birth defects in a nearby neighborhood. TCA, the solvent found
in the drinking water well at concentrations far exceeding the
state's recommended level, is an organic that can cause damage to
the central nervous system, the liver and the cardiovascular sys-
tem if ingested in large doses. In addition, it can cause loss
of coordination, eye irritation and dizziness. The National Tox-
icology Program concluded in a recent draft report that TCA is a
liver carcinogen in mice but not in rats.
SUMMARY
Lack of inventory and/or environmental monitoring, tank in-
spection or tank testing programs at Site A allowed a waste sol-
vent storage tank leak to go undetected for approximately 1-1/2
years. The leaked material contaminated soil and ground water.
As a result of the duration and size of the leak and the hydro-
geology of the site, transport of the contamination into three
aquifers and over an area of about 1/3 square mile occurred.
One drinking water well serving a total of about 700 people,
several private wells, and possibly another public drinking water
well were closed because of contaminants found in the wells.
Cleanup costs have exceeded $12 million and are continuing. These
efforts have been effective in reducing levels of ground water
contamination. In addition, law suits concerning damages resul-
ting from consumption of contaminated ground water have been
filed.
3-36

-------
EXHIBIT 3-2
SITE B STORAGE SYSTEM FAILURE DAMAGE CASE SUMMARY
FACILITY INFORMATION
Site B manufacture's electronic computing equipment (SIC
3573), semi-conductors and related devices (SIC 3674) and uses
numerous underground tanks for the storage and treatment of pro-
cess chemicals and industrial wastes. Site B has over 100 under-
ground tanks or concrete sumps and about 190,000 feet of under-
ground piping. The tanks are constructed of a variety of mater-
ials, including carbon steel, stainless steel, fiberglass, and
polypropylene. Information on piping materials in use was un-
available.
Of the 32 existing underground product storage tanks at
Site B, 26 (81 percent) are vaulted (i.e., located in a concrete
vault). Nearly all of the vaulted product storage tanks are
6,000 or 7,000 gallon capacity, are constructed of stainless
steel, and are less than six years old. Chemicals stored in
these vaulted tanks include:
•	Acetone;
•	Ethyl amy.l ketone ( EAK) ;
•	Freon 113;
•	Isophorone;
•	Isopropyl alcohol (IPA) ;
¦ • Kerosene; and
•	Nitrogen.
The remaining eight product storage tanks (19 percent) are
non-vaulted. The typical non-vaulted tank at Site B has a capac-
ity less than 3,000 gallons, is made of carbon steel, stores gas-
oline, and was installed more than 10 years ago.
Of the 31 existing underground waste storage tanks at Site
B, 28 are vaulted. Most of the vaulted tanks are made of steel,
are less than 5 years old, have a capacity of several thousand
gallons, and contain waste solvents such as acetone; EAK; freon;
IPA; isophorone; 1,1,1-trich1oroethane (TCA); 1,1,1-trichloroe-
thylene (TCE); or xylene. The three non-vaulted waste storage
tanks are older than the vaulted tanks and have smaller capaci-
ties.
Site B has 49 existing treatment tanks or concrete sumps.
About one-half of the treatment units are vaulted tanks and one-
half are concrete sumps. In addition, there are two nonvaulted
treatment tanks. A typical vaulted treatment unit has a 1,000
gallon capacity, is fiberglass, and is less than six years old.
The concrete sumps tend to be older than the treatment tanks and
range in capacity from 150 to 10,000 gallons.
The' facility has removed, abandoned, or relocated about 64
additional tanks or sumps, and more than 3 ,000 feet of piping.
3-37

-------
Some of these units were removed from service as a result of con-
tamination detected near the units. Reasons for discontinuing
use of other units are not known. Of the six areas where sources
of contamination have been detected at Site B, five have resulted
in the removal of underground storage units and the excavation of
surrounding soils. These areas include:
•	Tank Farm A;
•	Tank Farm B;
t	Building 14 chemical waste transfer sump;
•	Building 2 5 ink waste tank; and
•	Building 100 chromic acid tank.
The three cases of documented ground water contamination at Site
B are Tank Farms A and B and the area surrounding Wells A-30 and
A-31. Remedial actions at the other three areas appear to have
prevented migration of chemicals into the ground water.
Following is a presentation of the known facts concerning
tanks removed because of associated contamination. (Note: The
excavated tanks and related underground equipment were not all
leaking. The corporate practice manual for Site B concerning
containment of industrial liquids requires that underground sys-
tems with actual or potential leaks be replaced in accordance
with the most stringent government regulation, safety and fire
protection requirements, or other corporate standards and prac-
tices. The Site B corporate practice states that all newly con-
structed or replaced facilities storing solvents underground
shall have secondary containment. The definition of secondary
containment is one layer each of chemical/physical resistant
coating and liner or two single layers of liner which are applied
to or supported by an appropriate structure.)
•	At Tank Farm A, 17 non-vaulted solvent tanks and one 2000
gallon non-vaulted waste solvent tank were removed after
11 years of operation. One 24,000 gallon concrete vault
containing mixed solvent waste was also removed after
five years of use. All units were monitored by level
gauges and, except for the concrete tank, were con-
structed of asphalt-coated carbon steel with capacities
between 2,000 and 10,000 gallons, with a median capacity
of 2,000 gallons. Solvents stored included acetone, EAK,
IPA, isophorone, kerosene, sodium hydroxide, petroleum
naptha 365 , and 1,1,1-trich1oroethane (TCA). The speci-
fic cause of the leaking tanks at Tank Farm A is unknown,
but possible sources may be attributed to improper dis-
posal of the chemicals or past operational problems. It
is unknown how the leaks were discovered.
9 At Tank Farm B, nine solvent and six waste solvent under-
ground storage tanks were removed after less than eight
3-38

-------
years of operation. Six of the tanks were excavated
after about three years of use. These six tanks (five
product and one waste) each had a capacity of 2,700 gal-
lons and were double-walled with an inner wall of stain-
less steel and an outer wall of carbon steel. The re-
maining tanks ranged in size from 1,000 to 5,000 gallons,
were constructed of stainless steel or carbon steel with
cathodic protection, and were provided with vapor detec-
tors. Chemicals stored included acetone, IPA, freon,
methylene, chloride, N-methyl-2-pyrrolidone, and mixed
solvents. During excavation of the tanks, a drainline
from one of the mixed solvent waste tanks was found to be
severely corroded. Just prior to excavation, however,
the tank and drainline were tested and revealed no pro-
fa 1 ems .
o At Building 14, the 440 gallon capacity concrete waste
transfer sump and its liner were replaced after approx-
imately nine years of operation. Elevated levels of
chromium had been found in soil samples taken from out-
side the building. However, these levels were thought to
be due to the mineral content of backfill material
brought on-site during construction of Building 14,
rather than tank or piping leaks.
o At Building 25, the 4,000 gallon capacity ink waste tank
and the surrounding soils were excavated six years after
installation. The excavation appears to have stopped
inks from migrating to the ground water. Information on
the cause of the leak and how it was discovered was un-
available.
o At Building 100, the 1,000 gallon capacity concrete tank
with a plastic liner to hold chromic acid waste was aban-
doned six years after installation and removed four years
after abandonment. During removal a total of about 340
cubic yards of material (tank and associated soils) were
disposed at a Class I Site. Specifics on why the tank
failed and how the leak was discovered were not avail-
able.
ENVIRONMENTAL SETTING
Site B is located in a suburban area adjacent to a major
city and is surrounded by residential neighborhoods, small farms,
a hospital, freeway, and golf course. Other industrial facil-
ities are situated in the immediate vicinity. The site occupies
an area of approximately one square mile with more than 30 build-
ings built in a valley between a ground water recharge basin and
numerous wel1s.
The geology of the broad alluvial valley surrounding Site B
resulted from stream erosion and deposition. Streams flowing out
3-39

-------
of the highlands and into the valley deposited large quantities
of debris as alluvial fans and outwash plains. The alluvial sed-
iments range in thickness from zero along the hills bordering the
valley to 400 feet in the center of the valley. These deposits
are very permeable and the discontinuous clay beds in the area
are poor barriers to vertical ground water migration. The high
permeability of the alluvial sediments is reflected by a large
number of water supply wells. In the area of ground water flow
downgradient of and near Site B, there are 18 active and 7 inac-
tive wel1s.
Site B is underlain by a four aquifer system, designated as
aquifers "A", "B", "C", and "D" with increasing depth from the
ground surface. Aquifers "B" and "C" appear to be interconnected
at numerous random locations and, thus, they do not act as inde-
pendent permeable formations. The shallow "A" aquifer is gen-
erally between 20 and 50 feet below the ground surface. The
underlying suite of aquifers begin at about 60 feet and extend to
approximately 300 feet below the ground surface.
A generalized description of the subsoil conditions at Site
B is as follows:
o Surface to 20 feet - Moist dense brown clayey silt and
sti f f si 1 ty cl ay ;
d 20 feet to 30 feet - Saturated brown sandy silt and silty
sand with lenses of silty clay;
o 30 feet to 60 feet - Stiff brown and blue-grey silty clay
and clay; and
o 60 feet and deeper - Interbedded sands, silts, clays, and
gravels.
A creek which is the primary recharge source for the ground
water aqui.fers is located about 4 ,000 feet northeast from the
center of Site B. Infiltration ponds along the creek are a few
miles downstream. Recharge also occurs to a lesser extent by in-
filtration of irrigation water and through percolation of rain-,
fall which averages 14.2 inches per year. The flow of ground
water at Site B is in a west-northwesterly direction.
RELEASE FACTS
As discussed above, six areas of documented contamination
have occurred at Site B. Three of these areas have polluted the
ground water as far away as 7,000 feet or more. The plume of
contaminated ground water emanating from the site has been linked
with the contamination of two public drinking water wells located
3,000 feet west and 1 mile northwest of the facility. The water
company voluntarily closed the two drinking water sources when
trace amounts of TCA and Freon 113 were detected in the wells.
Contamination from the site has also been linked with the
3-40

-------
contamination of 10 private wells. The County environmental
health sanitarian recommended that four of the 10 contaminated
private wells be closed, including one well serving a mobile home
park with 400 residences and three owned by individuals.
The ground water investigation for Site B involves three
areas which have been documented as sources of contamination.
These areas are termed "Tank Farm A", "Tank Farm B", and "Wells
A-30/A-31". At each location the facility has conducted field
explorations of the extent of soil and ground water contamination
in an effort to determine the most effective remedial strategy.
The facility has drilled nearly 250 wells, including more than
200 on-site and about 40 off-site wells, as part of the investi-
gation.
Comparative analyses of well installation and sampling
methods were performed at Site B. Most well borings were drilled
using a mud rotary rig; some were drilled using a hollow-stem
auger. The analyses for organics in the soils and groundwater
showed that the drilling procedure did not effect the results.
Tests were also performed to examine the difference in results
due to collecting samples with a teflon bailer as opposed to a
polyethylene, disposable bottle; the sorption potential of PVC;
and the use of PVC glue for joining well casings. It was found
that the use of polyethylene sample bottles and PVC well casing
during normal sampling times did not produce any significant
analytical differences. However, the practice of using PVC glue
for joining well casings at Site B' was stopped after testing in-
dicated the potential for the organics found in PVC glue to ad-
sorb or desorb organic materials.
The analyses of the soils and ground water were performed in
accordance with EPA standards. Soil samples were taken with a
drill rig equipped with hollow-stem augers and a split-barrel
sampler. Ground water sampling in boreholes involved lowering a
fresh polyethylene bottle in the bore of the auger and allowing
it to fill when submerged. Ground water sampling from wells was
performed using a submersible electric pump and PVC pipe. All
water samples were taken from the pump discharge line in plastic
and glass bottles, except for the volatile organics samples which
were taken by lowering a clean polyethylene bottle into the well.
Quality control was maintained by taking duplicate and blank sam-
ples.'
The distribution of Freon 113 and TCA in aquifer "A" two
years after the contamination at Site B was detected was largely
concentrated in the on-site area. Both chemicals had spread
4,000 feet by 2,500 feet in a west northwest direction and ap-
peared to originate from the same dual sources, Tank Farms A and
B, as shown in Figures 1 and 2. The similarity of the two con-
taminant pi umes suggested that the chemicals migrated with the
general ground water flow.
The horizontal movement, however, was very slow in aquifer
"A", considering that the chemicals had probably been in the
3-41

-------
-p»
ro
contaminated
DRINKING WATER
WELL
?nE B PROPERTY
V-IMt
CONTAMINATED
DRINKING WATER
WELL
KEY:
» "A"AQUIFER MOHITORiNGi WELL
Figure 1. Concentrations of Freon 113 (ppb) in the "A" Aquifer
(Average of data for a 2\ month period)

-------
Figure 2.
Concentrations of TCA (ppb) in the "A" Aquifer
(Average of data for a 2\ month period)

-------
ground for several years. Since aquifer "A" has limited trans-
mi ss i vi ty and is only partially saturated, lateral migration of
the chemicals was limited. Freon 113 and TCA, being mo re dense
than water, appear to have migrated downward into the underlying
aquifers. Once in the hydraulically interconnected aquifers "8"
and "C", the contaminants moved horizontally with the ground
water due to the high transmissivity and saturation of the forma-
tion.
In aquifer "B" the distribution of freon and TCA two years
after the detection of contamination at Site B revealed that
their plumes extended northwesterly for 7,000 feet or more from
the source, as shown in Figures 3 and 4. The alignment of the
chemical plume with the contaminated public drinking water well
located one mile northwest of the site suggests that pumping of
the well influenced ground water flow rate and direction. A spur
from the main plume toward another polluted public drinking water
well located approximately 3,000 feet west of Site B also sug-
gests that pumping attracted contaminants toward this well. The
estimated ambient ground water velocity of five feet per day or
1 ,800 feet per year implies that a plume extending over 7,000
feet required the chemicals to reside in the aquifer at least
four years. Given the length of time that the Tank Farms had
occupied the site, a four-year interval is plausible.
At the time the contour maps shown in Figures 1 to. 4 were
prepared, two years after the contamination was detected and
remedial work was started, the maximum concentration of Freon 113
in the "A" aquifer was found near Tank Farm "B" at a level of 11
ppm. Tank Farm "A" revealed the greatest concentrations of TCA
in the "A" aquifer at 50 ppm. In aquifer "B", the highest con-
centration of ' freon (1.6 ppm) was found less than 1 ,000 feat
downgradient of Tank Farm "B", and the greatest concent ratioo of
TCA was about 0.1 ppm found approximately one mile off-site.
High concentrations of another chemical, 1,1,1-trich 1 oroe-
thylene (TCE), were found in the soil and ground water near Wells
A-30 and A-31. Possible contamination sources may have been im-
proper disposal of the chemical in the area of the wells or past
operational problems at the abandoned Tank Farm A. High levels
of TCE appear to have been confined to Wells A-30 and A - 31.
Nearly all concentrations of TCE reported for aquifers "A" and
"B" monitoring wells in the vicinity of Wells A-30 and A-31 were
less than 1 ppb, whereas Well A-30 showed as much as 410 ppb in
the "A" aquifer.
REMEDIAL MEASURES
Remedial measures undertaken at Site B included localized
ground water extraction at the three sources of contamination and
soils removal. The on-site cleanup system involved a series of
removal wells placed in the "A" and "B" aquifers near the site
boundary in a west-northwest orientation. The system also in-
cluded an extensive array of monitoring wells to evaluate the
3-44

-------
CO
I
-£>
cn
SITE e> PROPERTY"
line
TANK
FARM VW
• /I
"TANK
FARM x"B"

Figure 3.
Concentrations of Freon 113 (ppb) in the "B" Aquifer
(Average of data for a 2^ month period)

-------
Figure 4. Concentrations of TCA (ppb) in the "B" Aquifer
(Average of data for a 2h month period)

-------
efficiency of the cleanup. The plan for off-site remedial work
was not available at the time that this case study was conducted.
Remedial measures began in 1980 and are continuing.
Remedial measures in the immediate vicinity of Tank Farm "A"
included the removal of 17 solvent tanks, a waste solvent tank,
and a waste solvent concrete vault. The specific cause(s) of
contamination and the volume of solvent released in this area are
unknown. Improper disposal of the chemicals or past operational
problems, however, are possible sources of contamination at the
abandoned Tank Farm. In other words leaking tanks or piping have
not been reported as the cause of the contamination.
The excavated soil (about 7,000 cubic yards) and ground
water from the Tank Farm "A" area were disposed of at a Class I
disposal facility. A biological oxidation and activated carbon
adsorption system was constructed for ground water treatment at
the abandoned facility. During excavation, soil and water sam-
ples were taken and analyzed by gas chromatography. The highest
concentrations found were:
Aceton.e
EAK
I PA
Isophorone
Kerosene
Petroleum naptha
TCA
Xylene
Soi 1
( PPm)
5,000
5,000
150
12,000
25,000
3,300
3
Ground Water
(PPm)
220,000*
70
23,000
. 45
3,500
3,300
2,200
290
Water sample extracted from soil
Remedial actions at Tank Farm "8" were started about one-
year after the remedial actions were started at Tank Farm A. The
cleanup effort included the removal of nine solvent and six waste
solvent underground tanks. The cause of contamination appeared
to be a severely corroded three-inch drainline from one of the
waste solvent tanks. The spilled material surrounding the tank
consisted mostly of Freon 113, though acetone, IPA, TCA, and
methylene chloride were also detected. Concentrations of the
material that was stored in the waste solvent tank were as
foilows:
Freon 113	93%
TCA	0.9 ppm
Methylene	chloride 4.3	ppm
Acetone	1.0 ppm
IPA	280 ppm
The excavated soil and ground water from the Tank Farm "B" area
were hauled to a Class I disposal site.
3-47

-------
As of this writing cleanup measures for the third area of
contamination at Site B, Wells A-3 and A-31 were not known. Soil
boring work has been done to identify the possible sources of TCE
found in the wells. Remedial efforts are continuing on-site and
off-site at Site B, and the date for completion of these activi-
ties is unknown.
RELEASE CONSEQUENCES
Site B has not publicly disclosed the cost of remedial mea-
sures for cleanup of underground storage system failures. How-
ever, it has been estimated that more than $10 million has been
spent over three years. Two public drinking water supply wells
located about 3,000 feet and one mile from Site B and serving at
least 2,000 people were taken out of service because of Freon 113
and TCA contamination. The plume of chemicals from Site 8 also
contaminated 10 private wells. The county health department
recommended that four out of the 10 polluted private wells be
closed as a result of the most recent data linking TCA to cancer.
The public wells were taken out of use even though levels of TCA
were 10 times less than the State's recommended limit of .3 ppm.
There is no recommended limit for freon.
SUMMARY
Lack of inventory and/or environmental monitoring, tank in-
spection or tank testirvg programs at Site B allowed many leaks to
go undetected for as long as 11 years before detection. The
source of pollution has been determined for only one of the three
areas found to have soil and ground water contamination. The
duration and size of the leaks and the hydrogeology of the area
allowed the released chemicals to enter three aquifers and to
travel for a distance of more than a mile.
Two public drinking water supply wells serving at least
2,000 people and 4 out of 10 contaminated private wells have been
closed as a result of underground tank system leaks at Site B.
Remedial measures are in progress and have been estimated to have
cost approximately S10 million through Hay 1983 (including on-
site excavation of tanks, piping and soils). The cleanup efforts
have prevented contamination levels from increasing or spreading
to a larger area.
3-48

-------
REFERENCES
1.	"Tank and Piping Leak Survey Results". Memorandum, F. 8.
Killmar to Members of the Operations and Engineering Commit-
tee and Underground Leakage Task Force, American Petroleum
Institute, February 5, 1981.
2.	Warren Rogers Associates (W. Rogers), Newport, Rhode Island.
Personal communications with SCS Engineers, May through July
1983.
3.	California Regional Water Quality Control Board. "Under-
ground Tank Leak Detection Program - Status Report, April
1982 to April 1983", (Internal Memorandum). San Francisco
Bay Region, Oakland, California, May 2, 1983.
4.	"Status Report on Local Water Resources Protection Measures",
(Memorandum). S. W. Horsley to D. A. Hall, Cape Code Plan-
ning and Economic Development Commission, September 10, 1982.
5.	New York State Department of Environmental Conservation.
"Bulk Storage of Hazardous Liquids Study Program, Paper No.
6", (Draft). Albany, New York, April 1981.
6.	Prince George's County Fire Department (Sgt. Bradburn and
Capt. Becker), ,Prinee George's County, Maryland. Personal
communications with SCS Engineers, March 1983.
7.	Michigan Department of Natural Resources, Water Quality
Division. "Study of the Underground Storage of Gasoline".
September 1981.
8.	New York State Department of Environmental Conservation.
"Bulk Storage of Hazardous Liquids Study Program, Paper No.
5, Problem Assessment Report", (Draft). Albany, New York,
April 1981.
9.	Suffolk County Department of Health Services. "Suffolk
County Sanitary Code, Article 12, Toxic and Toxic and
Hazardous Materials Storage and Handling Control (and
Associated Data)". Farmi n g vi 11 e, New York, July 28-, 1982.
10.	Suffolk County Department of Health Services (J. Pim),
Suffolk County, New York. Personal communications with SCS
Engineers, May 1983.
3-49

-------
SECTION 4
RELATIVE RELEASE PROBABILITY AND MAGNITUDE
INTRODUCTION
Hazardous waste releases to the environment from underground
storage facilities can result from a number of events such as
spills during filling or emptying operations, tank overflow or
tank and piping system leaks and failures. The purpose of this
section is to evaluate the potential for release from underground
storage systems both in terms of magnitude and probability. In
order to organize the presentation of the methods and results of
the release potential assessment, this .section was subdivided
into five parts as follows:
•	A general overview of release events and variables
effecting them for all types of underground hazardous
waste storage facilities is provided. Release events and
variables associated with transfer are excluded.
•	A "typical" underground hazardous waste storage facility
which will be used to evaluate the relative importance of
release events in terms of release probability and
magnitude is described.
•	The specific release events and variables as well as the
relative release magnitudes and probabilities associated
with the "typical" underground storage facility are
revi ewed.
•	Brief discussions of additional factors such as envi-
ronmental setting and waste type which effect release
variables and release magnitude, are provided; and
•	Data limitations are noted.
Specific underground storage management options and their
relative impact on reducing hazardous waste releases as compared
to the "typical" facility are discussed in Section 5. The health
and environmental concerns, such as environmental pathways and
human exposure that arise once the stored material leaves a tank
system, are not addressed in this report
RELEASE EVENTS AND VARIABLES
The purpose of this subsection is to provide a "shopping
list" of release events and variables which can be used to
construct fault-trees for the many different types of underground
hazardous waste storage facilities. The release events
4-1

-------
considered in this analysis include tank overflow, tank leak,
tank rupture, ancillary equipment leak, ancillary equipment
rupture, fire or explosion and other incidents. The occurrence
of these events and their magnitudes may be influenced by one or
a number of different variables which are presented in Figures 4-
1 through 4-9.
To show the relationship between the release events and
variables resulting in hazardous waste releases to the environ-
ment a fault-tree was developed. The basic components of this
fault tree were derived from the fault-tree analyses for
aboveground facilities conducted by F.G. Bercha and Associates
Limited [1, 2], JRB Associates [3] and the information obtained
from telephone conversations and documents presented in Appendix
G •
The components of the fault-tree presented in Figures 4-1
through 4-9 are connected byeither "and" gates or "or" gates
which determine the specific relationship between the
probabilities of events and/or variables occurring. An "and"
gate represents a situation where both of the components must
exist or occur for the next step up on the tree to be affected.
This situation will result in a multiplication of probabilities
of occurrence. For example, in Figures 4-3 and 4-4, the
probability of tank leak will be reduced significantly if, when a
tank leak occurs, the facility has secondary containment with a
leak detection system. Similarly, in Figure 4-8 both an ignition
and fuel source (i.e., tank overflow, tank leak, etc.) must be
present before a fire or explosion will occur.
An "or" gate represents a situation where the component will
occur regardless of the existence or occurrence of the other
factors. The "or" gate situation will result in addition of the
probabilities of occurrence. This is evident in Figure 4-6 where
variables such as corrosion, seal failure and operator error can
occur with or without the occurrence of the others.
The events and the variables presented in Figures 4-1 through
4-9 are in most cases directly related to the waste stored and/or
the complexity of the storage facility. For example, a fire or
explosion will not occur unless an ignitable or reactive waste is
stored, and release cannot be influenced by the failure of a
corrosion protection system or an overflow prevention system when
they do not exist. As a result, the variables' presented in these
tables may or may not influence the release from specific under-
ground storage facility.
Each of the release events, with the exception of fire or
explosion and other incidents, are divided into four general
categories: design/installation deficiency, operator error,
4-2

-------
lank
Mupturs
4^
I
CO
fdult Irw Lagtmdi
Major Evonts
( )
Variables oINctlny
Major Evimti
~
"And" lidtosi (probabl I It I**
mul I lf»l led)
a
"Or" Uatas (probabilities
atfilud)
Figure 4-1.
Events leading to releases from underground hazardous waste
storage facilities.

-------
Tdflfc
Overflow
Figure 4-2. Variables effecting tank overflow at an underground hazardous
waste storage facility.

-------
IdMk
I uok
I
CJ1
IH*$lyn/l»s1dl Ia-
tlon Deficiency
f\Inadequate corroslon\
I protection system
(Tank material lncum-\
pdilbla with Hdila J
Figure 4-3.
Variables effecting tank
waste storage facility.
leak
tqulpoeftt
f al lure
Release Containment
Systm Fdllure
(See Figure 4-4)
I link
tull
«*ll\
ui«i	y

-------
Figure 4-4. Variables effecting containment system failures at an underground
hazardous waste storage facility.

-------
F i gure
4-5.
Variables effecting tank
storage facility.
rupture at an underground hazardous waste

-------
ure 4-6. Variables effecting ancillary equipment leaks at an underground
hazardous waste storage facility.

-------
Figure 4-7. Variables effecting ancillary equipment rupture at an underground
hazardous waste storage facility.

-------
Figure 4-8.
Variables effecting
hazardous waste sto
fire or explosion
rage facility.
at an underground

-------
-p»
I
Figure 4-9.
Variables effecting other accidents at an underground
hazardous waste storage facility.

-------
equipment failure and control system failure. These categories
are further divided into the variables which have the greatest
influence on release potential and magnitude. Brief descriptions
of each of these events and the variables associated with release
potential and magnitude are presented below.
Tank Overflow
Tank overflow occurs while filling. As shown in Figure 4-1,
this event is influenced by variables such as inadequate design
of the overflow control system, overestimation of the available
capacity of the tank, failure of the tank level gauge and failure
of the automatic shutdown system. The major cause of tank
overflow is operator error which depends on factors such as tank
filling method, operator competence and whether or not the
facility has an automatic shutdown system. In most cases release
magnitude is influenced by the operator's ability to identify the
problem and the time it subsequently takes to stop the filling
operation.
Tank Leak
Tank leak occurs at relatively low rates over an extended
period of time (i.e., weeks, months, years). One of the primary
causes of leaks in steel tanks is external corrosion which is
influenced by installation procedures, soil characteristics
(especially soil resistivity and moisture content), tank age and
corrosion protection system effectiveness. Since steel tanks are'
not the only tanks used in underground storage, tank material,
installation procedures and leak prevention systems (i.e.,
secondary containment, corrosion protect ion systems, etc.) must
be addressed when assessing leak potential (see Figures 4-3 and
4-4).
Once a leak occurs, the magnitude of release is dependent
upon the size of the leak (i.e., gallons per day) as well as the
time it takes for the leak to be detected and stopped. As a
result, the existence of a leak detection system or secondary
containment, the operators' ability to note level discrepancies
in the tank and/or the frequency of tank testing will be the
primary factors effecting release magnitude.
Tank Rupture
Tank rupture is defined as the release of large quantities of
stored material over a relatively short period of time (i.e.,
minutes, hours, days). Tank ruptures result from the failure of
the tank material due to factors such as failure of the venting
system, puncture or cracking o-f a Fiberglass Reinforced Plastic
(FRP) tank, uneven settling from improper installation, and tank
4-12

-------
material failure resulting from the introduction of incompatible
wastes. As shown in Figures 4-4 and 4-5, additional variables
also contribute to release potential.
The magnitude of release from tank ruptures is dependent upon
the size of the tank, the volume of waste in tne tank when the
rupture occurs, the existence of a release detection system or
secondary containment and the operator's ability to note the loss
of stored material. Tank rupture will be identified sooner than
tank leaks due to the drastic change in volume of the tank
contents or other evidence of system failure.
Ancillary Equipment Leak
Ancillary equipment leaks occur from pipes, pumps, valves,
etc. at relatively low rates (i.e., a few gallons per day), over
an extended period of time (i.e., weeks, months, years). The
major factors, as shown in Figures 4-4 and 4-6, that influence
ancillary equipment leaks are external corrosion of steel piping
and loose fittings or joints which may result from improper
installation or time-induced stresses (i.e., vibration, settling,
etc.) .
The magnitude of releases are affected by the same factors as
tank leaks--the size of the leak (i.e., gallons per day) and
detection by the operator. However, due to the lower rate of
release, these leaks may go undetected for longer periods.of
time. As with tank leaks, a leak detection system or secondary
containment and/or frequent system testing decreases release
magnitudes from ancillary equipment.
Ancillary Equipment Rupture
Ancillary equipment ruptures are defined as the release of
large quantities (i.e., greater than 10 percent of the material
being transported) of material from pipes, pumps, valves, etc.,
over a relatively short period of time (i.e., minutes, hours,
days). These ruptures are similar to tank rupture in that
release usually results from equipment failure due to
overpressurization, piping system fracture from induced stresses
or improper installation, and piping system failure. (See Figures
4-4 and 4-7).
The magnitude of release from ancillary equipment rupture is
dependent upon the volume of waste being transported through the
system and the operator's ability to identify the problem.
Controls such as secondary containment or release detection
systems and tank level monitoring will reduce the magnitude of
release.
4-13

-------
Fire or Explosion
Fire or explosion is defined as a sudden release of a
portion or all of the stored material in a tank as a result of
the ignition and/or sudden expansion of a flammable or reactive
waste. These events are slightly different from the events
described above in that they may occur as the result of an
overflow, leak or rupture, or they may arise from conditions
within the system such as chemical reactions and thermal
expansion (see Figure 4-8).
As mentioned above, these events can result in the release of
a portion or all of the stored material depending on the
circumstances leading to the events and the control systems
available (i.e., foam system, sprinkler systems, etc.).
Other Incidents
Other incidents are defined as events that occur due to
natural phenomena, vandalism, etc., which have not been discussed
under the other headings. These incidents are dependent on the
facility location in the case of variables such as earthquakes
and flood, and on the uncontrolled or unpredictable nature of
people. Steps such as proper designs and security systems can
reduce the probability of their occurrence.
The magnitudes of these release events are variable depending
on the extent of the damage incurred. For example, an earthquake
may result in either a leak or a rupture depending on the system
design.
"TYPICAL" FACILITY
In this subsection, a "typical" underground storage facility
is defined to provide a baseline for developing release pro-
babilities and magnitudes. The characteristics of this facility
were selected by evaluating current practices to determine the
most common features of underground storage facilities. Current
practices were defined using data from the "Hazardous Waste Tank
questionnaire" (0MB # : 2000-0424), information collected from
equipment manufacturers, trade associations and "standards"
organizations (i.e., American Society for Testing and Materials,
National Fire Protection Association, etc.), information
presented in Section 3 and Appendix G.
Current practices were found to include a wide variety of
equipment types (i.e., tanks, ancillary equipment and control
systems), facility ages, management/mai ntenance programs,
installation practices and environmental settings. In fact, the
features of the alternative management systems discussed in
Section 5 are currently used to varying degrees. In this section,
4-14

-------
a single, "typical" facility (see Figure 4-10) which represents
the most common underground storage facility (as defined by the
information sources noted above) was evaluated. The
characteristics of this facility are presented below.
Alternative systems designed to represent a range of options for
reducing the probability and magnitude of release as compared to
the "typical" facility are discussed in Section 5.
"Typical" Facility Characteristics
Equipment Type--
The "typical" underground storage facility was assumed to
consist of:
One 3,000 gallon carbon steel tank conforming to UL58
with a black asphaltum coating;
Waste being stored is ignitable;
Unprotected cast iron piping;
A trap to prevent vapor migration to the point production
faci1i ty;
Steel vent pipes; and
Gravity piping (so no pumps are included).
Faci1ity Age--
Facility age was assumed to be 8 years.
Management/Mai ntenance Programs--
Until recently, the concern for management/maintenance
programs for underground hazardous waste storage facilities has
been minimal. This statement is supported by the large number of
releases from underground product storage tanks noted in Section
3 that had gone undetected for relatively long periods of time
and by the assumption that the loss of stored product would be of
more concern (due to cost considerations) than would waste
release. Consequently, the management/maintenance program for
the "typical" facility under consideration is assumed to be
limited to a simple tank level checking program (i.e., once a
week) with waste removal when the tank is three quarters full and
tank testing (Petro-tite or equivalent) every 5 years.
Installation Procedures-
Installation of underground tanks is
with specifications of the NFPA (NFPA
Petroleum Institute (API 1615) [5] and/or
Installation usually involves excavating
normally in accordance
30) [4] the American
the tank manufacturer,
an appropriately sized
hole and trench for tanks and piping, placing tank and piping in
4-15

-------
(e) NON-CORROSIVE . INERT BACKFILL
(j) TRAFFIC BOLLARDS
Figure 4-10. Diagram of the "typical" Underground Storage Facility.

-------
the excavated areas on sand or gravel bedding, anchoring the
tank, if it is located in an area subject to a high groundwater
table or flooding, attaching ancillary equipment such as vents,
pumps and valves, backfilling with sand or gravel, and installing
a concrete pad if the tank is to bear traffic or barriers if
traffic is to be prohibited. These procedures were assumed to be
used to install the "typical" storage facility with the exception
of anchoring since it is assumed that the facility is not subject
to high groundwater or flooding.
Environmental Setting--
The environmental setting selected- for this analysis
consists of poorly drained acidic soils with a resistivity of
less than 10,000 OHM-centimeters [4]. (An assumption was made
that no corrosion protection system was installed even though
NFPA specifies corrosion protection for soils with resistivities
at this level. This assumption was made because the results of
the API "Leak Survey" showed a large number of tanks without
corrosion protection.) In addition, the site is not located in a
flood plain and depth to groundwater is at least 20 feet.
"TYPICAL" FACILITY RELEASE MAGNITUDES AND PROBABILITIES
The purpose of this subsection is to present the release
events, magnitudes and probabilities associated with the
"typical." facility. Initially, the assumptions associated, with
each event are discussed along with the resulting release
magnitudes. This is then followed by a discussion of the
relative release probabilities of each event.
To provide a breakdown of the variables effecting the
"typical" facility release events, the fault-tree shown in
Figures 4-11 through 4-18 was created from the "shopping list" of
events and variables presented earlier in Figures 4-1 through 4-
9. As shown, parts of the general fault-tree such as the
"Release Containment System Failure" shown in Figures 4-3 and 4-4
were not included (see Figure 4-13) since they did not apply to
ttie "typical" facility.
The procedures and assumptions used to develop the release
magnitudes and relative probabilities are described below.
Re 1 ease Magnitudes
Magnitudes of the release events associated with the
"typical" facility were estimated by assuming an average volume
of waste stored in the tank and the time which might elapse
before the release was detected. Release magnitudes for the
"typical" facility are discussed below in terms of the events
leading to the release, the variables which influence their
4-17

-------




Tank

Tank

Over 11 cm

Leak

tI0"2)

(10


H»Ims« to the
Environment
<	V





















Tank

Ancl 1 lary

Ancl1lary

Fire end
Rupture

Equlpnent

Equipment

Explosion



Leak

Rupture





(10
-4.

(I0~6>
Other
Incidents
(IO-«>
I—»
00	Fault Tree Legend:
Major Events
( )
~
A
Variables Effecting
Major Events
"And" bates (probabilities
inul t Ipl led)
"Or" (jates (probabilities
added)
Figure 4-11
Release events and probabilities associated with a "typical"
underground hazardous waste storage facility.

-------
I
Figure 4-12.
Variables and release probabilities associated with tank
overflow at a "typical" underground hazardous waste
storaae facility.

-------
Figure 4-13.
Variables and release probabilities associated with tank
leaks at a "typical" underground hazardous waste storage
f aci 1 i ty..

-------
-P*
I
ro
Ueslgn/lnstalla-
t fOft Oetlclency
IIO"5>
Inadequate consider-
ation ol Induced
Vstresses
[ Codes not met
'improper Installa^
material	—
^patlbte with waste J



Pressure k&1 let

System tallure

<10

Figure 4-14.
Variables and release probabilities associated with tank rupture
at a "typical" underground hazardous waste storage facility.

-------
•fa.
I
l\i
ro
F i gure
4-15.
Variables arid release probabilities associated with ancillary
equipment leaks at a "typical" underground hazardous waste
storage facility.

-------
Figure 4-16.
Variables and release probabilities associated with ancillary
equipment rupture at a "typical" underground hazardous waste
storage faci1ity.

-------
ro
4^
Figure 4-17.
Variables and release probabilities associated with fire or
explosion on a "typical" underground hazardous waste storage
facility.

-------
ro
<_n
Figure 4-18.
Variables
acci dents
and release probabilities associated with other
at a "typical" underground waste storage facility.

-------
magnitude arid the assumptions made to calculate release
magnitudes. The values derived from this exercise are discussed
below.
•	Tank Overflow: The volume of waste released due to tank
overflow is influenced primarily by the operator's
failure to have the tank emptied (it is assumed that the
tank under consideration is emptied when it reaches 75
percent capacity or every 15 days, whichever comes
first), or overestimating the available capacity of the
tank. The assumptions made to estimate the "typical"
facility release magnitude include:
-	overflow would only occur during filling operations;
-	a single batch load of 150 gallons is drained to the
tank daily (assume 75 gallon capacity available in tank
at time of filling resultin-g in overflow of 50 percent
(75 gallons) of the batch discharge); and
-	the overflow would be noticed by the facility operator
the same day it occurred.
A release of approximately 75 gallons would occur as a
result of this event. [50 percent of batch lost x'150
.gallons/batch=75gallons].
•	Tank Leak: The volume of waste released due to tank leak
is dependent on the number, size and location of
perforations in the tank wall, the existence of secondary
containment and/or leak monitoring systems, and the time
it takes the operator to detect the leak. The
assumptions made to estimate the "typical" facility
release magnitude include:
-	the number and size of tank wall perforations are su-ch
that 2 gallons of stored material leak from the tank
each day;
-	all of the perforations are in the lower third of the
tank;
-	tank testing is conducted once every 5 years and the
leak occurred 6 months after the last test (i.e.,. the
leak would go undetected for 4.5 years).
A release of approximately 3,300 gallons would occur as a
result of this event. [1,643 days x 2 gallons/day =
3,285 gallons].
4-26

-------
t Tank Rupture: The volume of waste released due to tank
rupture is influenced primarily by the location of the
opening, the volume of waste in the tank at the time of
the event, the existence of secondary containment and/or
a relese detection system, and the time it takes the
operator to detect the loss of material. The assumptions
made to estimate the "typical" facility release magnitude
i nclude:
-	the opening is located at the bottom of the tank;
-	the tank contains 1,500 gallons of waste material at
the time of rupture;
-	the entire contents 1,(500 gallons) of the tank are
lost over a period of a few days; and
-	the rupture is detected after 1 week when the operator
makes his weekly tank level reading. (Assume a loss of
150 gal 1ons per day).
A release of approximately 2,550 gallons would occur as a
result of this event. [1,500 gallons (tank content) + (7
days x 150 gallons/day) = 2,550 gallons].
•Ancillary Equipment Leak: The volume of waste released
due to ancillary equipment leaks is primarily controlled
by the size of the leak, the existence of leak-monitoring
systems and/or secondary containment and the frequency of
tank and pipe testing. The assumptions made to estimate
the "typical" facility release magnitude include:
-	the size and location of the leak are such that one
percent (1.5 gallons per day) of the daily batch
discharge to the tank is released;
-	the facility does not have a leak monitoring system or
secondary containment;
-	tank and pipe testing are conducted once every 5 years;
and
-	the leak occurred 6 months after the last test (i.e.,
the leak would go undetected until the next test is
performed (4.5 years)).
A release of approximately 2,470 gallons would occur as a
result of this event. [1,643 days x 1.5 gallons/day =
2,465 gal 1ons].
4-27

-------
•	Ancillary Equipment Rupture: The volume of waste
released due to ancillary equipment rupture is primarily
controlled by the volume of waste transported through the
system, the existence of a leak monitoring system and/or
secondary containment and the ability of the operator to
detect level discrepancies in the tank. -The assumptions
made to estimate the "typical" facility release magnitude
i nclude:
the release is due to a pipe break and 90 percent (135
gallons per day) of the daily batch discharge to the
tank is released;
-	the facility does not have a leak monitoring system or
secondary containment; and
-	the release would go undetected for 2 weeks. (The leak
would be detected as a result of level discrepancies
noted during tank level reading.)
A release of approximately 1,890 gallons would occur as a
result of this event. [14 days x 135 gal Tons/day = 1,880
gallons].
•	Fire or Explosion: The volume of waste released due to
fire or explosion is primarily controlled by storage
facility safety practices and control measures such as
spark arrestors on vent pipes, safety training programs
for employees and fire suppression systems. The
assumptions made to estimate the "typical" facility
release magnitude include:
-	the release is the result of a fire followed by an
explosion;
-	the storage facility does not have a safety training
program, fire suppression system or any other fire or
explosion prevention equipment;
-	the tank contains 1,500 gallons at the time of the
event; and
-	the entire contents are released as a result of the
event.
Approximately 500 gallons would be lost with a portion
being combusted and the balance being released to the
environment (i.e., land and air).
4-28

-------
• Other Incidents: The volume of waste released due to
other incidents is dependent on the type of event. The
primary factors which influence this event are facility
location (i.e., whether the facility is located in a
fault zone, flood plain, etc.) and facility security. The
assumptions made to estimate the "typical" facility
release magnitude include:
-	the release is a result of arson;
-	the storage facility does not have afire suppression
system;
-	the tank contains 1,500 gallons at the time of the
event; and
-	the entire contents of the tank are released as a
result of the event.
Approximately 1,500 gallons would be lost with a portion
being combusted and the balance being released to the
environment (i.e., land and air).
A summary of the release volumes from these events is
presented in Table 4-1.
Relative Release'Probabil ities
Release probabilities used in this analysis were derived
using judgment supported by values from studies done by F.G.
Bercha [1] and JR8 [3]. The principal reference source for
estimating release probabilities was an F.G. Bercha report [1]
since these values were relative rather than absolute, and thus
were more closely appropriate for the analysis conducted in this
section.
These values from F.G. Bercha [1] were then compared to those
used in the JRB report to check the relative relationship between
fault-tree components. Probability values from the F.G. Bercha
study [1] were based primarily on the "Reactor Safety Study"
prepared for the U.S. Nuclear Regulatory Commission (NRC) in
October 1975 [6], correspondence with equipment manufacturers and
facility operators and judgment [3]. As a result, these values
represent estimates of bulk plant storage relative release
probabilities. Probability values in the JRB report [3] were
based on the NRC data mentioned above and additional sources.
These values represented actual vs. relative values and were
considered inconsistent with the fault-tree developed for this
analysis which considers relative- rather than absolute
probabilities. In both cases, NRC data cannot be considered to
4-29

-------
•	Tank Rupture: As shown in Figure 4-13, the primary
causes of tank rupture are design/inst|l1ation deficiency
(10~5) and/or equipment failure (10 ~ 5 ). The principal
variable influencing design/instal1ation deficiency is
improper installation which may result in excessive
stress due to uneven settling, etc. In the case of
equipment failure, the primary cause is tank wall failure
which is caused by corrosion. Since the "typical"
facility has a carbon steel tank that is not pressurized
(i.e., pump fed), there is less likelihood of rupture
than for a facility with FRP tank or a pump fed system.
• Ancillary Equipment Leak: As shown in Figure 4-14, the
primary causes of ancillary equipment leaks are
design/insta 1 1 ation deficiency (10~M and/or equipment
failure (10 ~ * ) ~ Design/installation deficiency is the
most significant factor and occurs primarily due to
improper installation procedures such as inadequate
tightening and sealing of fittings and inadequate care
taken to prevent conditions (i.e., point anodes) which
induce corrosion. Equipment failure occurs to a lesser
degree, but is still a significant cause of ancillary
equipment leaks. The primary causes of equipment failure
are corrosion and seal failure.
•	Ancillary Equipment Rupture: As shown in Figure 4-15,
'the primary causes of ancillary equipment rupture are
design/insta 1 1 ation deficiency (10~4) and/or equipment
failure (10"^). The principal variables influencing
design/instal 1ation deficiency are improper installation
and subsequent induced stresses, both of which may result
in excessive strain on the system due to differential
settlement or vehicular traffic. Equipment failure is
somewhat related to design/insta11 ation deficiency since
pipe wall or equipment failure may result from induced
stresses combined with corrosion induced weaknesses. In
addition, seal failure may result in equipment failure.
•	Fire or Explosion: As shown in Figure 4-16, fire or
explosion is directly attributable to the probabilities
of the previously discussed events occurring as well as
the existence of an ignition source. Since the waste
stored at the "typical" facility is ignitable, this event
may occur, but its probability will be low since both
ignition and material sources must be available for this
event to occur. As a result, the probabilities of each
must be multiplied to obtain the probability of this
event occurring. As presented earlier, the events which
will most likely result in providing a source for
combustion are tank leaks (10"*) and ancillary equipment
4-32

-------
leaks (10~1). This, combined with a probability of
ignition of 10"^, results in a probability of fire or
explosion of 10"°.
•	Other Incidents: As shown in Figure 4-17, the primary
cause of other incidents is vandalism (10~°) Since the
"typical" facility is underground; is located outside of
the flood plain; and is not located in a region of high
seismic activity, there is little likelihood of this
event occurring.
From the information presented above, the most likely events
leading to releases of hazardous waste to the environment are
tank and ancillary equipment leaks. Release probabilities for
each event are controlled by the principal variables mentioned
above since, once a release occurs, there are no control systems
to prevent the material from entering the environment. Release
probabilities estimated for facilities with alternative
characteristics including overfill prevention, tank inspection,
more frequent testing, etc. are presented in the next section.
Other Factors Influencing Release Probability
The release probabilities and magnitudes discussed above were
estimated based on a number of assumptions regarding the
characteristics (i.e., management practices, environmental,
setting, tank material, waste type, etc.) of the "typical"
facility. Different faci1ities, types of waste and environmental
settings will'cause probabilities and magnitudes to be different.
Some of these differences are discussed below.
•	Environmental Setting: In this analysis, environmental
setting considerations consist of geographic location,
soil characteristics and groundwater levels, each of
which influence the variables effecting release. These
factors are all interrelated, but each plays a slightly
different role in this analysis.
-	Geographic location considerations are based primarily
on whether or not the facility is located in a fault
zone or a flood plain. Since the "typical" facility
was assumed not to be located in these types of areas,
the probability of release due to natural phenomena in
the category of "Other Incidents" is approaching zero.
The actual change in relative probability value would
vary by site-specific consideration such as the
frequency of floods or earthquakes;
-	Soil characteristics, particularly resistivity, are
measures of the corrosion potential of steel tank and
piping systems. Corrosion is a major consideration in
steel tank, ancillary equipment leak and rupture
4-33

-------
events. If conditions were less conducive to corrosion
than those assumed for the "typical" facility, the
release probabilities associated with these events
might be lower. The actual change in relative
probability would depend on specific site conditions
and facility configuration.
- Ground water levels are important when considering
corrosion potential and installation procedures, both
of which may influence tank and ancillary equipment
leak and rupture events. Soil moisture content effects
corrosion potential. As a result, tanks and ancillary
equipment situated in groundwater will be more prone to
corrosion (Note: the extent of change in corrosion
potential is unknown). Fluctuating groundwater tables
may cause a partially filled tank to "float" if it is
not properly anchored. This "floating" problem may
result in tank and/or ancillary equipment leaks or
ruptures. Since the "typical" facility was not
influenced by groundwater, the relative probabilities
of release in the example, may be lower than in a
situation where groundwater is of concern. The actual
change in relative probabilities would be site
speci fi c .
Tank and Ancillary Equipment Material: Tank and
ancillary equipment material is a majorconsideration
when assessing the system's susceptibility to corrosion
and structural durability. For example, concrete and
steel storage systems are more susceptible to corrosion
than FRP storage systems and as a result, have higher
probabilities of release associated with events
influenced by corrosion. On the other hand, structural
durability is of less concern with steel storage systems
than with FRP systems. FRP tanks and piping systems have
a higher probability of release as a result of puncture
and/or fracture due to installation error, puncture due
to operator error (i.e., dip stick punctures) [7], and
fracture due to induced stresses. The actual change in
relative release probabilities varies by site.
• Tank Age: The age of the equipment is one indicator as
to how much longer the facility can be expected to be
serviceable. However, other factors such as corrosion,
puncture due to operator error, and installation
deficiencies have a larger effect than age. For example,
work by Warren Rogers and Associates [8] has shown that
factors such as soil resistivity, pH, sulfide content and
moisture content affect corrosion far more than tank age.
In a given soil environment a steel tank may last for
more than 20 years, whereas in a corrosive soil, the same
4-34

-------
tank may fail within 2 years. Thus, tank age cannot be
the only factor considered when determining release
probabi1i ty.
•	Waste Type: Waste type is of concern when selecting
compatible material for the underground storage facility
and later when considering facility operation. As the
number of different wastes handled or the number of tanks
at a storage facility increases, the probability that a
waste will be accidentally or intentionally emptied into
a storage system constructed of an incompatible material
increases along with the probability that two chemically
incompatible waste types will be mixed. If this
situation exists, the relative probabilities of release
due to tanks and ancillary equipment rupture and fire or
explosion may increase in relationship to those presented
for the "typical" facility which handles only^one waste
type.
DATA LIMITATIONS
As noted previously, a number of assumptions based on a
variety of data sources and judgments were used in this section
to define current practices and release events and to develop
relative release probabilities and magnitudes. Due to their
importance, the major assumptions are reiterated below.
•	Release events and variables associated with the fault-
tree analyses were developed from aboveground bulk plant
storage facility studies and information obtained from
telephone conversations and documents reviewed for this
report.
•	Current practices for storing hazardous wastes
underground were defined based on information gathered
from equipment manufacturers, trade associations and
11 standards" organizations, in-house knowledge about
storage facilities, and literature sources. Information
from the tank and general hazardous waste storage
questionnaires was originally intended as the primary
source of this information, but was unavailable.
•	Release probabilities presented in this section are
relative values and are not intended to represent actual
release probabilities. These values represent judgment
based on previous studies of bulk petroleum product
storage facilities [1] [3] and aboveground hazardous
waste storage tanks.
4-35

-------
•	Release magnitudes were based on judgment using the
assumptions presented in this section.
SUMMARY
Examination of information from previous studies of petroleum
product storage facilities and contact with equipment
manufacturers, trade associations, and "standards" organizations
lead to the identification of seven events that cause releases
from underground storage tanks, as follows:
•	Tank overflow
•	Tank leak
•	Tank rupture
•	Ancillary equipment leak
•	Ancillary equipment rupture
•	Fire and/or explosion
•	Other incidents (e.g., earthquakes, floods, vandalism)
Relative release probabilities a.nd magnitudes are affected
significantly by the specific facility features such as. tank and
ancillary equipment materials, type of waste stored, management
practices, method of waste delivery to the storage tank, etc.
Thus, a "typical" facility was identified which was believed to
represent the most common practice. This facility also serves as
a baseline for comparison with alternative practices discussed in
Section 5.
Based on the characteristics of this "typical" facility,
estimates of relative release probabilities and magnitudes were
developed (see Table 4-2). As shown, two of the events with the
highest relative probability of occurrence, tank and ancillary
equipment leak, also have two of the highest estimated magnitudes
of release. The principal assumption affecting the magnitude
associated with these events is the duration of the leak (in this
case 4.5 years), which is based on a testing frequency of 5
years.
Duration is also a principal factor in determining the
magnitudes of release due to tank and ancillary equipment
ruptures. These events have a lower relative probability of
occurrence, but if they occur and go undetected for longer than
the time periods assumed (1 week for tank rupture and 2 weeks for
ancillary equipment rupture) their magnitudes could be much
higher. For example, if a tank rupture went undetected for 1
4-35

-------
TABLE 4-2. RELATIVE RELEASE PROBABILITIES AND
MAGNITUDES ASSOCIATED WITH THE "TYPICAL"
FACILITY.
Event Relative Release	Probability* Release Magnitude
		:	(Gallons) 	
Tank Overflow 10~2	75
Tank Leak 10"^	3,300
Tank Rupture 10"^	2,550
Anci11ary Equi pment
Leak 10"1	2,470
Ancillary Equipment
Rupture 10"4	1,890
Fire or Explosion 10"6	1,500
Other Incidents 10~®	1,500
* Release probabilities presented in this table are relative
versus absolute values and represent the probability of
release over the life of the facility.
4-37

-------
month and an ancillary equipment rupture went undetected for 2
months, the resulting magnitudes would be approximately 6,000
gallons and 8,100 gallons for tank and ancillary equipment
ruptures, respectively.
4-38

-------
REFERENCES
1.	F.G. Bercha and Associates Limited., "Bulk Plant Risk
Organization." Department of the Environment, Environment
Protection Service, Hull, Quebec, Canada, December 1982, 232
pp.
2.	F.G. Bercha and Associates Limited. (Mr. Rockne), Calgary,
Alberta, Canada. Personal Communications with SCS Engineers,
June 1983.
3.	JRB Associates. "Failure Incident Analyses: Evaluation of
Storage Failure Points," (Draft). U.S. Environmental
Protection Agency, Washington, D.C., March 1982.
4.	National Fire Protection Association, National Fire Code.
Flammable and Combustible Liquids Code. NFPA 30-1931.
Quincy, Massachusetts.
5.	American Petroleum Institute. Installation of Underground
Petroleum Storage Systems. API 1615-1979. Washington, D.C.
6.	U.S. Nuclear Regulatory Commission. "Reactor Safety Study",
Washington,'D.C., October 1975.
7.	American Petroleum Institute. "Results of API Tank and
Piping Leak Survey, February 5, 1981 and Updated Statistical
Data." Washington, D.C., 1981.
8.	Warren Rogers Associates (W. Rogers), Newport, Rhode Island.
Personal communications with SCS Engineers, May through July
1983.
4-39

-------
SECTION 5
ANALYSIS OF SELECTED MANAGEMENT ALTERNATIVES
INTRODUCTION
Review of Sections 3 and 4 of this report indicates that
current underground storage tank system practices can result in
environmental release of hazardous waste. These releases have
been described by seven categories of release events, as
fol1ows:
•	tank leak, defined as release at relatively low rates over
an extended period of time (i.e., weeks, months, years);
•	tank rupture, defined as	release of large quantities
(relative to tank volume)	of stored material over a
relatively short period of	time (i.e., minutes, hours,
days );
•	ancillary equipment leak,	defined as release from pipes,
pumps, valves, etc. at relatively low rates over an
extended period of time;
•	ancillary equipment rupture, defined as release of large
quantities (relative to the quantity of material	h,andled)
of material from pipes, pumps, valves, etc.	over a
relatively short period of time;
•	tank overflow, defined as release associated with over-
filling of the storage tank;
•	fire/exp1osion, defined as sudden release of a portion or
all of the stored material from a tank system as a result
of the ignition and/or sudden expansion of a flammable or
reactive waste; and
•	other, defined as other miscellaneous events which occur
due to natural phenomena, vandalism, etc.
With respect to storage of hazardous waste in underground
tanks, 40 CFR Parts 264 and 265 include requirements related to
prevention of tank overfilling (264.192 and 265.192), fire/explo-
sion (264.198, 264.199, 265.198 and 265.199) and vandalism
(264.14 and 265.14). The impacts of natural phenomena on release
from tank storage facilities have been the subject of other-
investigations [1], Thus, hazardous waste releases from the
first four categories of release events listed above are the
subject of this analysis of management alternatives.
Environmental releases resulting from leaks and ruptures of
underground tanks and associated ancillary equipment are of
particular concern since the occurence of these events often goes
5-1

-------
undetected for long periods of time (i.e., a year or more). In
order to prevent and/or minimize the impact of these release
events, it is necessary to understand the causes. Four types of
release causes were identified in Section 4 and are used here to
facilitate the analysis, as follows:
•design deficiency;
•	installation practices;
•	equipment failure; and
•	operational error.
Desi gn
Although design deficiencies may occur at any stage in the
development of a storage tank facility, they are thought to be
the least common of the four types of release causes. The
specific type of error which occurs will be determined on a case-
by-case basis, but generally the errors are caused by the same
type of factors which contribute to other types of engineeering
errors, and include:
•	inaccurate information;
•< incomplete information;
•	inexperience (on the part of the engineer, equipment
manufacturer and/or facility operator); and
•	errors in judgement.
Solutions to deficiencies associated with facility design ob-
viously involve correcting these deficiencies (i.e., through
improved availability and accuracy of baseline design informa-
tion, etc.). Thus, improvements can be expected if the effort is
made. However, errors will still occur even with improved
practices, although with a lower frequency.
Installation
Installation practices are indicated to be an important
source of problems at existing facilities, with the type of
problem often depending on the the tank system materials. For
all types of tank systems, improper joining of piping and
appurtenances are a significant source of leaks. For steel
systems, the primary concern is the increased rate of corrosion
(expecially non-uniform corrosion) caused by events such as:
•	damage to a cathodic protection system (i.e., sacraficial
anode or impressed current equipment attached to the
tank);

-------
•	lack of homogeneous and inert backfill material;
•	damage to protective coatings; and
t attachment of mud clods to the tank or similar contribu-
tors to point corrosion.
Other concerns associated with steel tank installation in-
clude inadequate fill compaction leading to differential settling
and damage to piping connections and improper anchoring. For FRP
systems, problems are generally related to puncture or breakage
of the tank due to foreign objects in the excavation or fill
material; damage due to floating of inadequately anchored tanks;
and breakage of the tank or piping due to differential settle-
ment. For concrete systems, concerns include stress cracks and
cracks resulting from settlement, both of which may lead to
leakage. Minimizing these problems generally involves confor-
mance with manufacturers recommendations, applicable codes and
standards and guidance available from organization such as ASTM,
API, UL, etc.
Equi pment
Equipment failure has also been indicated to be an important
cause of release, although equipment failures are also frequently
linked to the three other types of release event causes. The
equipment failures which occur are of many different types.
Probably the most significant from a release perspective are
related to corrosion and/or failure of ancillary equipment.
Corrosion-induced failures cover the range of types of
corrosion (i.e., uniform, erosion, stray current, pitting, gal-
vanic; etc.) and may be aggravated by improper installation,
incompatible waste and/or design deficiencies. Ancillary equip-
ment failures may involve pump diaphragms and packing, valve
seals, piping connections, etc. and be caused by excessive
pressures, design deficiencies, improper installation, incompa-
table materials and many other factors. Thus, equipment failures
are minimized primarily by utilizing improved practices associ-
ated with equipment selection, installation and operation.
Operati on
Operation is the fourth type of release event cause identi-
fied above and is an important factor in some types of release
events. Operational errors may result from a variety of factors
including lack of training, lack of maintenance, lack of secur-
ity, human shortcomings (i.e., carelessness) or a lack of
contingency planning and preparedness. Such operational errors
may result in direct releases or may trigger other causes of
release, as in the case of an accidental addition of an
incompatible waste into a tank which results in equipment failure
due to accelerated corrosion or explosion. As reflected hy
current regulations, improved practices can lead to decreases in
5-3

-------
release from operational causes. For example, 40 CFR Part
264.194 requires daily inspection of overfill control equipment
to insure proper operation. In spite of such measures, however,
operational releases are likely to remain an important, although
reduced, source of release.
For the four release event causes common to the four release
events categories discussed here, reductions in the frequency and
size of waste releases can be accomplished through application of
improved knowledge and practices. However, these causes cannot
be completely eliminated. Thus, additional measures can be taken
to reduce the frequency and size of tank system releases.
In this section of the report, five types of measures
designed to provide for reduced levels of environmental release
from underground tanks are discussed , as follows (see Section 2
for a discussion of how these measures were selected for
consideration):
•	secondary containment;
•	corrosion protect ion;
•	system testing;
•	system monitoring (inventory and/or environmental); and
•	inspection.
In order to provide a basis for comparison of these five types of
approaches to reducing tank system releases, model facilities
were developed. A discussion of the two model facility sizes
used, including the relative importance of the four categories of
release events at these facilities, is also provided.
MODEL FACILITIES
In order to provide a common point of reference for compari-
son of the various management alternatives, two sizes of model
facilities were selected to represent small and medium sized
facilities. The specific sizes of the model facilities selected
were based on data from three sources. One source of data used
was the preliminary data from the U.S. EPA Hazardous Waste Tank
Questionnaire (0MB No. 2000-0424) [2]. A second data source was
the San Francisco Bay Region of the California Regional Water
Quality Control Board [3],- and a third source was a profile of
hazardous waste tank and container storage facilities which
relied primarily on the Hazardous Waste Data Management System
(HWDMS) for input data [4].
From these data sources, facility sizes of one 1,000 gallon
tank and two 5,000 gallon tanks were selected to represent small
and medium sized facilities respectively. Data from the Hazard-
ous Waste Tank Questionnaire indicate that the median facility
5-4

-------
has an underground hazardous waste storage tank capacity of
10,000 gallons provided by three or less tanks, and that 14
percent of the underground hazardous waste tanks have a capacity
of 1,000 gallons or less.
The physical and operational characteristics of the facili-
ties assumed in the subsequent analysis of prevention and
mitigation options are the same as those of the "typical"
facility presented in Sections 4 unless otherwise noted. Parti-
cularly noteworthy characteristics are as follows:
•	equipment and operation;
carbon steel tanks conforming to UL 58;
stored waste is ignitable;
- waste enters tanks through gravity feed piping;
waste supply piping is underground and 20 feet in
length;
tank vent piping runs parallel with the supply piping
to the building and'then up the side of the building;
waste is transferred to the small tank in 50 gallon
batches once each week;
waste is transferred to each tank at .the medium
facility in 150 gallon batches twice each week;
tank level measured daily;
•	installation was conducted in accordance with appropriate
specifications available at the time of installation;
•	located in poorly drained acidic soils with a resistivity
considered to be conducive to corrosion; and
•tank age is 8 years.
The specific features of these facilities are assumed to be
the same as for the "typical" facility discussed in Section 4
(see Section 4 for details) with the exceptions , noted in this
Section. In addition, hazardous waste storage facility charac-
teristics associated with compliance with 40 CFR Part 264
Subparts B through G and J are assumed.
Subpart B addresses waste analysis, security, general inspec-
tion requirements and personnel training. As applied to under-
ground storage facilities, waste analysis requirements include
chemical and physical analysis of the waste prior to storage,
repeat analysis as necessary to insure that it is accurate and
up-to-date and a waste analysis plan. Facilities receiving waste
5-5

-------
from off-site sources must specify the procedures to be used to
insure that the characteristics of waste received match the
accompanying manifest.
For underground tank storage facilities, compliance with the
security requirements could take several forms depending on site
specific conditions. For example, facilities that also perform
treatment and/or disposal functions would presumably integrate
compliance with security requirements for all of their hazardous
waste functions. In most instances, compliance would likely
involve surveillance and/or fencing with gates to control access.
The general inspection requirements of Subpart B require that
a facility must conduct (and record) inspections with a frequency
sufficient to identify problems in time to correct them before
harm to human health or the environment occurs. The type of
inspection which is feasible for underground tanks and associated
equipment varies with the installed configuration. For the model
facilities, it was assumed that the piping is not accessible for
inspection while the tank is accessible for inspection. For
tanks which were not provided with manways at the time of
construction, a manway may be retrofitted (see below for more
details). The type -and frequency of tank inspection are dis-
cussed below.
Preparedness and Prevention (Subpart C) and Contingency Plan
and Emergency Procedures (Subpart D) require that design, con-
struction, maintenance and operation minimize the possibility of
unplanned waste releases. In addition, specific equipment (espe-
cially for fire control) and procedures (especially for ingitable
or reactive wastes) are required unless specifically waived by
the Regional Administrator.
Supart E defines requirements for the manifest system,
recordkeeping and reporting which apply to hazardous waste
storage facilities. Subpart G defines. closure and post-closure
requirements, which are also mentioned in Subpart J. Subpart J,
which specifically addresses hazardous waste storage tanks (ex-
cepting underground tanks which cannot be entered for internal
inspection) requires sufficient shell strength to prevent col-
lapse or rupture (see also Appendix D) and that tank materials
(or liners) are compatible with the waste stored (see also
Appendix A).
In addition, requirements which expand on those in other
Subparts regarding inspection, closure, reactive/ignitab!e waste
and incompatible waste are also included in Subpart J. Require-
ments for internal inspection of tanks which can be entered for
inspection are specifically excluded from the model facility
since they are discussed below as one of the five approaches for
preventing and/or mitigating releases.
Review of the release probabilities presented in Section d
for tank and ancillary equipment leak and rupture release events
5-6

-------
indicates that compliance with these regulatory requirements does
not change the release probabilities of these events (see Tables
4-13 through 4-16). This occurs primarily due to the assumption
that the model facilities are also existing (rather than new)
facilities and to the exclusion of compliance with the internal
inspection requirement for the model facilities (since inspection-
is discussed below as one of five mitigation/prevention mea-
sures).
The model facility release magnitudes, on the other hand, are
not the same as for the "typical" facility discussed in Section 4
due to changes in tanks size and operating assumptions. The
values derived for the model facilities are as follows:
Volume of Waste Released (gallons)
Event	per Event (gal Ions) by F"aci 1 ity Size
Smal1	Med i um*
Tank Leak	1600	1600
Tank Rupture	500	2500
Ancillary Equipment	Leak 120	700
Ancillary Equipment	Rupture 90	540
* Values are rounded to two significant figures.
The volume of waste released due to tank and ancillary equipment
leak and rupture depend 6n the duration and rate of the .event.
Since empirical data for use in deriving estimated release rates
are extremely limited, the values above are based primarily on
assumptions. The key assumptions are presented below.
Tank Leak: The volume of waste released due to tank leak
depends primarily on the number, size and locations of
perforations in the tank wall with respect to the liquid
level in the tank, the type of waste stored and the time
it takes the.operator to detect the leak. The assumptions
used with respect to leak rate and duration are:
the leak rate averages 1 gallon per day, with the
initial rate lower and the final rate higher than the
average. Thus, the leak rate at the time of detection
is slightly above .the rate which is detectable with
most tank testing procedures and is the same as the
rate assumed for the "typical" facility. The assumed
leak rate (which is thought to be conservative) is
based on judgement since empirical data were unavail-
able; and
tank system testing is conducted once every 5 years and
the leak occurred 6 months after the last test (i.e.,
the leak would go undetected for 4.5 years. This
assumed testing frequency is based on judgement since
no empirical data were available, the range of testing
frequencies actually used is thought to be large, with

-------
some facilities testing as often as every six months
and others not at all.
•	Tank Rupture: The volume of waste released due to tank
rupture is determined primarily by location of the open-
ing, the volume of waste in the tank at the time of the
event and the time it takes the operator to detect the
loss. The assumptions used with respect to these vari-
ables are:
- the rupture occurs in the bottom of the tank;
the tank contains 50 percent of capacity when the
rupture occurs and the tank contents are released over
a period of 1 to 2 days; and
the rupture is detected after one day when the operator
makes a daily tank level reading.
•	Ancillary Equipment Leak: The volume of waste released
due to ancillary equipment leak is determined primarily by
the size of the leak, waste transfer characteristics and
the time it takes for the operator to detect the leak.
Assumptions used with respect to these variables are:
one percent of each batch discharge to the tank is
1eaked; and
tank system testing is conducted once every 5 years and
the leak occured 6 months after the last test (i.e.,
the leak would go undetected for 4.5 years.
•	Ancillary Equipment Rupture: The volume of waste released
due to ancillary equpment rupture is primarily controlled
by the waste transfer characteristics and the time it
takes for the operator to detect the leak. The assump-
tions used for these variables are:
the release is due to a pipe break and 90 percent of
each waste transfer is released; and
the release would go undetected for.2 weeks, at which
time the operator would notice that the tank level had
increased only nominally.
LEAK AND RUPTURE RELEASE MITIGATION/PREVENTION
Five types of prevention/mitigation measures are discussed
here. For each approach, the following are provided:
•	a brief description;
•	a general discussion of the types of release causes and
events which the measure mitigates/prevents;
5-3

-------
•	identification of the specific choices available for
implementation;
•	presentation of selected implementation options, including
costs, advantages and disadvantages; and
•	a brief summary.
In addition, summary tabulations of release probabilities, costs
and effectiveness are included.
Secondary Contai nment
Secondary containment as discussed here includes both the
provision of a containment structure in addition to the tank and
interstitial leak detection equipment for identifying the failure
of either the primary or secondary containment structure. It can
be applied to both tanks and ancillary equipment to prevent
environmental release of the stored waste in the event of a leak
or rupture, and has the following features:
0 provides a second line of defense against tank and
ancillary equipment design deficiencies;
•	removes concern for problems associated with undetected
leakage due to installation errors, except for damage to
the monitoring system which may occur during installation;
and
•	provides protection against equipment failures, except for
failure of the monitoring equipment.
For both existing and new facilities, containment can be
provided a number of different ways. For tanks, the secondary
containment options include double-walled tanks, concrete vaults,
and liners of various types, such as clay or synthetic membranes.
For piping, containment options include covered trenches (i.e., a
concrete utility trench), double-walled piping and tunnels.
Depending on the type of containment used for the tank and
piping, interstitial (between the primary and secondary contain-
ment units) monitoring can be accomplished using vacuum, pres-
sure, sensors or visual inspection.
Selection of one of the above methods for use at a storage
facility will depend on a variety of factors, such as number,
size and location of tanks; waste type; and environmental
setting, including soil and groundwater characteristics. These
factors vary such that most if not all of the secondary
containment methods identified above will see some use. Thus,
most are discussed below. Clay liners are not discussed due to
the substantial variations in cost as a function of clay
availability and the similarity of applicability to synthetic
liners. Tunnels for piping are also not discussed due to the
substantially higher cost than the other options.
5-9

-------
Double-walled tanks are available in steel, stainless steel,
fiberglass or combinations of these materials., although fiber-
glass tanks are not available in sizes larger than 4,000 gallons.
As discussed above, use of steel tanks for the model facilities
.is assumed since this is thought to be the material most commonly
used (see Sect ion 2). Alternatives are also available for the
extent of secondary containment (i.e., complete containment,
double walls only on the bottom half of the tank, etc.) and the
type of monitoring system used (i.e., measurement of vacuum or
resistivity to detect water and/or waste in the interstitial
space).
The advantages, disadvantages and costs associated with the
use of double-walled tanks for both existing and new facilities
are presented in Table 5-1. As shown, the primary disadvantage
is the lack of availability in some materials and sizes and the
primary advantages are greater ease of cleanup if primary
containment does fail, and lower cost.
The initial costs for existing facilities assume cleaning and
removal of the existing tank, replacement (in the same excava-
tion) with a double wall tank and reuse of the existing ancillary
equipment. The initial costs for new- facilities represent the
difference between the cost of the facility with a double wall
tank and the cost with a single wall tank. Annual costs for both
existing and new facilities are based on the assumption that the
interstitial monitors must be checked each operating day to
comply with 40 CFR 264.194. This daily checking of the monitor-
ing equipment is estimated to require 5 minutes per day, 260 days
per year at a cost of $16 per hour. Thus, the annual cost is
$350 per year.
The costs and advantages and disadvantages associated with
the concrete vault approach to secondary containment for under-
ground tanks are also shown in Table 5-1. The primary advantage
of the concrete vault approach to secondary containment is that
the containment structure will not need replacement in the event
of tank failure. The principal disadvantages are the generally
higher cost than for double wall tanks; the increased risk of
fire or exposion in the event of release of ignitable or reactive
waste from the tank (as compared to a directly buried tank); and
the requirement of some local codes that the vault be backfilled
ifthetankcontains ignitablewaste.
Since concrete is porous and susceptible to cracking, it is
assumed that the containment structure is lined with an epoxy or
similar material which is compatable with the waste to be
contained. In addition, it is assumed that the exterior of the
vault is water proofed to help prevent water from entering the
secondary containment area. Use of a liner material on the
concrete vault adds relatively little cost to the system and will
also facilitate clean-up (if waste is released from the tank) and
closure (since concrete will not be contaminated).
5-10

-------
TABLE 5-1. SECONDARY CONTAINMENT*
Model
Facility Type
Method
Disadvantages/limitat ions
Advantages
Incremental Cost ft}"*"
Initial Annual EUAC**
Existiny-smalI
Concrete vault for tank(s) with
continuous monitoring
cn
l
Some local codes require backfilling
If tank contains ignitables. This
prevents periodic visual inspection
and complicates clean-up if a release
occurs. Maintenance of sensors for
monitoring is also more difficult of
the tank exterior and secondary con-
tainment .
May require lining, depending primar-
ily on waste type.
-	Available
-	Containment will rarely need re-
placement following a tank release
-	If not backfilled, clean-up should
should be relatively fast and in-
expensive.
-	Provides for containment and de-
tention of tank releases and moni-
toring of containment integrity.
16,000 350
1,400
- Cracking may impare Integrity
Synthetic liner for tank
excavation
Double walled tanks
Concrete trench for
ancillary equipment
containment
-	Clean-up of releases relatively
expensive as compared to other
methods of secondary containment.
-	Liner incompatible with some wastes.
Not available in all materials and
tank sizes.
May require lining, depending on
waste type.
Some local codes may require back-
fill. See concrete vault for tanks
above.
-	Expensive to install relative to 38,000 350 2,900
other secondary containment
methods.
-	Provides for containment and de-
tention of tank releases and moni-
toring of containment Integrity.
-	Available
-	Least expensive clean-up follow- 16,000 350 1,400
ing tank release.
-	Provides for containment and de-
tention of tank releases and moni-
toring of containment integrity.
-	Available	6,000 350 750
-	Provides for containment and de-
tention of ancillary equipment and
monitoring of containment integrity.

-------
TABLE 5-1 (Continued)
	Ho3e1	
Facility Type
Method"
Disadvantages/1 imitations
Advantages"
Existing -
ined i um
New - small
New - medium
Double walled
piping
Incremental Cost~TTr,r
Initial Annual EUAC**
Replacement relatively expensive
unless pipe walls are independent.
Does not control releases from pumps
valves, and other ancillary equip-
ment.
Available
Provides for containment and de-
tention of pipe releases and moni-
toring of containment integrity.
May be more practical than
trenches for retrofit installa-
tion in many situations.
1,500 350
450
Tank concrete vault
Same as above
Same as above
44,000
350
3,300
Synthetic liner for
tank excavation
M
"
67,000
350
4,900
Double walled tank
-
H
46,000
350
3,400
Piping trench
»
H
6,000
350
750
Double walled piping
II
It
2,500
350
520
Tank concrete vault
II
II
9,200
350
970
Synthetic liner for
tank excavation
"
II
33,000
350
2,600
Double walled tank
M
II
9,200
350
970
Piping trench
II
II
6,000
350
750
Double walled piping
-
II
1,300
350
440
Tank concrete vault
»

18,000
350
1,600
Synthetic liner for
tank excavation
•1

" 50,000
350
3,700
Double walled tank
-

31,000
350
2,400
Piping trench
»

6,000
350
750
Doub1e wa11ed p i p i ng
'•

2,100
350
490
~ Increase in cost from the baseTine facility." Costs "for tanks assume pfpfngTs Teft unchanged arid pfpe costs assume taiiks are feft unchanged. ff secondary
containment of tanks and concrete trench for piping are combined, the initial cost will be $2,600 less and the annual cost will be $250 less than the sum of the
two costs presented here since one monitoring system control unit can be eliminated. If tank secondary containment is provided by a vault and a concrete trench
is used for piping, an additional $2,500 initial cost savings will result from elimination of the piping trench sump.
t+All methods presented assume continuous monitoring.
*See accompanying text for additional information on assumptions used in developing this table. Costs are rounded to two significant figures.
"Equivalent Uniform Annual Cost.

-------
It is also assumed that the vault is provided with a manway
to permit inspection of the vault liner material, the tank and
the sensors which are assumed to be used to continuously monitor
for leakage in the secondary containment area. The initial costs
for exisiting facilities also assume removal of the existing
tank, construction of a concrete vault in the same excavation,
reuse of the exisiting tank in the vault and reuse of the
ancillary equipment. The initial costs for a new facility
represent the incremental cost for inclusion of the concrete
vault and associated monitoring equipment. Annual costs for both
existing and new facilities assume daily checking of the monitor-
ing equipment at a cost of S350 per year. In addition, one
inspection per year of the vault lining, tank exterior and
monitoring sensor at a cost of $24 (one and one-half hours at $16
per hour) is assumed. Thus, annual costs are $374.
Use of a synthetic liner below the tank is the third method
of tank secondary containment presented in Table 5-1. As shown,
it is a more expensive method of containment than either of the
other two methods discussed under the assumptions used here. The
key construction assumptions effecting cost are the slope at
which the liner is installed on the sides of the tank excavation
and the number of tanks placed within a single liner. For the
costs presented in Table 5-1, it was assumed that a slope of 2 to
1 (2 feet horizontal to 1 foot vertical) was used. Installation
on steeper slopes may be possible, but such applications are not
warranteed by the liner manufacturers. Use of a 1 to 1 slope,
however, would result in initial costs which are less than
instead of greater than those for the other two containment
methods.
For the medium sized model facility, the wastes contained in
the two tanks are assumed to be sufficiently compatable to permit
both tanks to be installed within one liner. If separate liners
are required, the costs would be significantly higher.
The initial costs presented for an existing facility also
assume removal of the existing tank, additonal excavation, liner
installation, and reuse of the existing tank and ancillary
equipment. The initial costs for new facilities represent the
incremental cost for inclusion of the liner and associated
monitoring equipment (resistivity sensor and control unit). The
annual costs for both exisiting and new facilities assume daily
checking of the monitoring equipment at a cost of $350 per year.
Use of a concrete utility trench with resistivity sensors to
detect leakage of either the ancillary equipment or the trench
itself is one of two methods of ancillary equipment secondary
containment presented in Table 5-1. This method has several
advantages, including the ability to use a containment structure
for ancillary equipment associated with several tanks, to replace
failed ancillary equipment without replacing the containment
structure and to integrate leak sensing with tank secondary
5-13

-------
containment. The principal disadvantage is that some local codes
may require backfilling for ignitable wastes.
The initial costs for exisiting facilities assume installa-
tion of a pre-cast concrete trench with new piping and abandon-
ment of exisiting piping in place. The initial costs for a new
facility represent the cost for inclusion of the trench and
associated leak monitoring equipment (resistivity sensor and
control unit). If the concrete utility trench approach to
secondary containment for ancillary equipment is used in conjunc-
tion with tank secondary containment (which is assumed to include
sensors for containment monitoring), the initial costs for small
and medium sized existing and new facilities will be reduced by
52,200.
The annual costs for both existing and new facilities assume
daily checking of the monitoring equipment at a cost of $350 per
year. If this approach to piping containment is used in conjunc-
tion with tank containment, the annual cost can be assumed to be
eliminated since their will be no separate monitoring devices to
read and record.
Use of double walled piping is the second ancillary equipment
containment method presented in Table 5-1. As shown, it has the
advantage of being easier and less expensive to install than a
concrete trench in many situations. The principal disadvantage
is the lack of economies of scale which are possible with a
concrete trench both .in terms of containment and leak detection
monitoring.	•
The initial costs for existing facilities assume installation
of double wall piping with pressurization of the interstitial
space and abandonment of the existing piping in place. The
initial costs for- a new facility represent the differential
between the installed costs of single and double wall piping.
The annual costs for both existing and new facilities assume
daily checking of a pressure gauge at a cost of $350 per year.
All of the above methods of tank and ancillary equipment
secondary containment have the advantage of significantly reduc-
ing the magnitude and probably of release from both leak and
rupture events. Magnitudes are reduced because event duration is
reduced due to the use of continuous monitoring equipment, as
fol1ows :
Reduction i-n Waste Released per
Event by Model Faci1ity Si ze*
Event
Smal 1
gal. percent
Medium
gal. percent
Tank Leak
Tank Rupture
Ancillary Equipment Leak
Ancillary Equipment Rupture
1595	99+
495	99
115	96
85	¦ 94
1595	99+
2495	99+
695	99
535	99
5-14

-------
* Values are based on the following assumptions:
•	for events to result in release, the primary and
secondary containment structures and the leak moni-
toring system must fail simultaneously. It is as-
sumed that failure of the secondary containment
system takes the form of a small leak (1 gallon per
day) in the event of primary containment failure.
•	event duration is five days for both rupture and
leak events since it.is assumed that the leak will
be discovered and the tank drained within this
period due to the daily inspection of the secondary
containment monitoring equipment (the leak detection
system is not discovered to be malfunctioning for
three days and it takes two days to complete pump
out of the tank and secondary containment area.
Use of secondary containment (including continuous monitoring
equipment) for both tanks and ancillary equipment is estimated to
reduce the probability of release due to leak or rupture by four
orders o f magn i tude. THTcTi a "Targe reduction results from the
numerous-""and" gates in the fault tree for the system. Specific-
ally, a release can occur only if the primary containment fails
and the monitoring equipment fails or the operator fails to
respond to an indication of aleak and the secondary containment
structure fails over the same time period. Thus, they provide a
high- level of protection against design, deficiency, installation
error, operator error and equipment failure causes of waste
release.
Use of either tank or ancillary equipment containment alone
fails to provide a reduction in the probability of release from
the facility as a whole since significant events remain uncon-
trolled. Some reductions in estimated release magnitude also
occur, but they are generally small.
Tank System Testi ng
Tank system testing as discussed here includes testing of
both tanks and piping systems to identify the presence, and in
some cases, the rate and/or locations, of leaks. (Other methods
which provide for testing of tanks only are also discussed in
Appendix H.) Thus, tank system testing serves to reduce the
magnitude of tank and ancillary equipment release by reducing the
duration of an undetected leak or rupture. Some leak test
methods only test for tank leaks, but are not considered here
since they offer no particular advantages and have the obvious
disadvantage of failing to detect piping leaks.
As shown in Table 5-2, a variety of methods exist for tank
system testing. (Note: Table 5-2 is not all-inclusive.) Addi-
tional detail on these and other methods is provided in Appendix
5-15

-------
TABLE 5-2 TANK SYSTEM TESTING SUMMARY*
Model
FaciIity Type
Method
Disadvantages/1 imitations
Advantages
FncrementaT Cost (i)*
Initial Annual FDAC**
Existing -
sma 11
Sunmark Leak Lokator
Ul
I
o~>
VacuTect
Smith & Denison
(he Iium)
-	Applicability dependent on waste
type.
-	Cannot detect very small (less
than 0.03 gallons/hour) leaks.
-	Availability limited, but improv-
ing.
-	Tank needs to be full to give
most reliable results.
-	Sophisticated equipment requires
specially trained personnel.
-	Leak rate not measured.
-	Applicability may be limited by
waste type.
-	Tank system must he empty
for testing.
-	Leakage rate not measured.
-	Pressurized testing.
-	Requires specially trained per-
sonnel.
Tests both tanks and pipes.
Reported to be accurate to 0.003
gallons/hour.
Detects leaks throughout tank
depth.
Compensates for temperature
changes.
Relatively short set-up and test-
ing time.
All testing coordinated by one com-
pany, which improves personnel
training and testing reliability.
Tank deficiencies and waste tem-
perature do not affect results.
Short test time.
Tests both tanks and pipes.
Full tank not required.
Generally available.
Tests tank and pipes.
Applicability not dependent on
waste type.
Not affected by temperature
changes or tank deformation.
Relatively short lest duration.
Generally available.
1,500
1,500
500
500
500
500
*See accompanying text for additional information on assumptions used in developing this table. Costs are rounded to" two significant figures.

-------
TABLE 5-2 (CONTINUED)
"MoiTeT"
FaciIity Type
Method'
UFsailvantayes/l iini tat ions
Advantages
Incremental Cost
Initial
Annual ETE
liU'*
Petro-Tite
Ol
Full tank ami extra waste re-
quired.
Relatively long test duration.
Cannot detect very small (less than
0.05 gallons per hour) leaks.
Applicability dependent on waste
type.
- Generally available.
Tests both tanks and pipes.
Reported to be accurate to 0.05
gal Ions per hour.
Temperature effects and tank de-
formations accounted for.
Detects leaks throughout tank
depth.
500
500
Existing
med i urn
New - smal1
New - medium
Sunmark Leak Lokator
Vacutest
Smith & Oenison
Petro-Tite
All 4 methods
Same as above
Same as above
1,500 1,500
800	800
800	U00
800	800
Same as existiny small
Same as existing medium
~ Increase in cost from the baseline facility. Costs for tanks assume piping is left unchanged and pipe costs assume tanks are left unchanged. If secondary
containment of tanks and piping are combined, the cost will be $3,000 less than the sum of the two costs presented here since one monitoring system control unit
can be eliminated.
** Equivalent Uniform Annual Cost.

-------
H. For all of the methods, there are three principal concerns
associated with selection of a testing method: 1) compatibility
of the testing equipment with the waste, 2) the minimum detect-
able size of a leak, and 3) the availability of equipment and
t rai ned personnel.
A potential disadvantage of all of the methods listed is that
there is limited experience with testing tanks used to store
hazardous waste. For tests which require equipment contact with
the waste, waste characteristics are likely to limit the applica-
bility of the testing procedure in some circumstances.
The minimum detectable leak size also varies with the method
used, but is generally in the range of 0.03 to 0.05 gallons per
hour for the more sensitive methods. For some methods, such as
the Smith and Oenison helium testing method, leak rate is not
measured. For other methods, such as hydrostatic testing, the
minimum detectable leak is notably larger. In general, the
sensitivity of tank testing methods is at best aproximately one
gallon per day.
The Sunmark Leak Lokator is reported to have been used to
test commercial, non-petroleum, underground tanks and piping
systems for leaks [5], However, the availability of this method
at a reasonable cost in some areas may be a problem. The Petro-
Tite leak test method has been used primarily on underground
gasoline storage tanks. It appears that the test method could be
used to test . tanks containing hazardous wastes as long as the
stored product was compatible with the testing equipment and
extra product was available to raise the liquid level above the
top of the tank. The requirement of additional product may limit
the extent to which this testing method can he used to test
hazardous waste underground storage tanks for leaks. Availabil-
ity of the other two methods shown in Table 5-2 is more limited.
As shown in Table 5-2, the costs associated with testing tank
and piping systems vary with testing method, but are the same for
existing and new facilities. However, costs for each method may
vary significantly with location. Due to this variation and the
other factors discussed above, selection of a testing method will
generally not be made based on a comparison of the costs
presented here.
For whichever method is used, the benefit derived will be a
reduction in the magnitude of release due to earlier detection.
The magnitude of this reduction will depend primarily on when the
leak occurs in relation to system testing and the leak rate. For
comparison purposes, estimated reductions in release magnitudes
based on an annual testing frequency for the model facilities are
as foilows:	•
5-18

-------
Reduction i n Maste Released per
Event by Model Facility Si ze*
Event	Sma11	Medi urn
gal. percent gal. percent
Tank Leak

1420
89
1420
89
Tank Rupture

0
0
0
0
Anci11ary Equi pment
Leak
110
91
78
89
Anci11 ary Equi pment
Rupture
0
0
0
0
* Values are based on the following assumptions:
t the leak begins at the mid-point of the testing
cycle. Thus, the leak duration for all facilities
is 26 weeks, and leak magnitudes using tank testing
are as follows:
180 gallons (26 weeks x 7 days/week x 1 gal-
lon/day) is the tank leak magnitude for both small
and medium facilities;
13 gallons (26 weeks x 50 gallons/week x 1%) is
the anci11ary¦equipment leak magnitude for small
facilities; and
78 gallons (26 weeks x 300 gallons/week x 1%) is
the ancillary equipment leak magnitude for medium
faci.l ities.
Since developing leaks are not detected, ji£ reduction i n release
probabi1ity is achieved.
Envi'ronmental Monitoring
Another method of reducing release magnitudes without affect-
ing release probabi1i1ities (i.e., detecting releases after they
have occured) involves the monitoring of the environment adjacent
to a hazardous waste storage tank. Such monitoring could be
conducted in the saturated zone and/or unsaturated zone using
observation wells and any of three methods of detection, includ-
ing: 1) thermal conductivity or electrical resistivity sensors,
2) gas detectors; or 3) sample collection and analysis. Regard-
less of the specific method used, the objective of such monitor-
ing would be early detection of tank leakage, thereby minimizing
release volume.
Soil and ground water monitoring of existing hazardous waste
storage tanks has been used for leak detection in various
situations. Perhaps the most concentrated use of this approach
has been the program initiated by the California Regional Water
Quality Control Board, San Francisco Region, to detect potential
leakage from underground tanks. As part of this program, soil
sampling and ground water well installation (where depth to
5-19

-------
ground water was less than 30 feet) have been conducted at
approximately 100 locations since March, 1982, and monitoring at
additional sites is anticipated.
The applicability of environmental monitoring for monitoring
release from underground tanks containing hazardous waste is
dependent on waste type and site conditions. Dependence on waste
type is due to the fact that leak detection after the fact may
not be acceptable for some types of waste (e.g., acutely toxic).
In addition, waste type affects the selection of a specific
environmental monitoring approach. For example, only volatile
wastes can readily be monitored using gas detectors. Dependence
on site conditions is due to soil and ground water characteris-
tics discussed in detail below.
Site specific conditions	and waste type also determine the
practicality of a specific	monitoring approach. As a result,
four alternative approaches	to environmental monitoring are
di scussed:
•	ground water sampling and analysis;
•	ground water wells with continous monitoring sensors;
•	volatile gas monitoring with stationary probes; and
•	soil water sampling and analysis.
i-
In areas where the saturated zone is relatively close to the
surface (i.e. 20-30 feet) ground water wells might be used. In
order to document that contamination, if detected, is originating
from the equipment (e.g., tank or piping) being monitored, both
up-gradient and down-gradient wells are assumed.
The frequency of ground water sampling and analysis has a
significant impact on the cost of implementing this management
alternative. Since the overall objective of the monitoring
program is to detect leakage as quickly as practicable, the
sample collection/analysis interval should be no greater than the
estimated time of migration from the equipment to the well. This
time of migration wil depend primarily on: 1) the distance
between the monitoring well and the equipment, and 2) the rate of
transport in the saturated and unsaturated zones.
The rate of transport is extremely variable due to the
dependence on a wide range of parameters including:
•	soil porosity;
•	soil permeability;
•	waste mass density (which is a function of temperature);
•	waste vi scosi ty;
•	waste saturated hydraulic conductivity; and
•	size/rate of leak.
5-20

-------
Thus, a desirable monitoring frequency based on the rate of waste
transport and the distance between the equipment and the monitor-
ing well could vary over a range of from minutes to many years.
For example, soil transport calculations for benzene which assume
a volumetric loading of 0.4 cubic meters/square meter indicates a
20 meter penetration in about 25 minutes from a surface spill at
20* C (ambient) over coarse sand and a 1 meter penetration in
about 3 years from the same spill over clay till [61. For the
purposes of this analysis, a range of monitoring frequency of A
times per year is assumed.
Assumptions used in estimating the costs associated with
ground water sampling and analysis (see Table 5-3) at existing
facilities are as follows:
•	1 up-gradient and 2 down-gradient wells for both small and
medi um faci1i ti es;
•	well depth of 20 feet;
t	4 inch well diameter with drilling and casing cost of
S16/foot ;
•	sampling equipment cost of $200; and
•	drill rig mobilization cost of S300.
•	sampling costs $50 per well per quarter;
•	sample analysis costs $100 per sample; and
•	samples from down-gradient wells are composited prior to
analysis, so that 2 samples are analyzed each quarter.
Costs associated with new facilities are based on these same
assumptions, with the exception that down-gradient wells will be
replaced with casing installed in the backfill below the tank
such that samples can be collected to monitor for leakage. The
installed cost of this casing is $15/foot and it is assumed that
30 feet of casing are required for each tank.
Based on these assumptions, the advantages, disadvantages and
costs associated with the use of the ground water sampling and
analysis approach to environmental monitoring for both existing
and new facilities are presented in Table 5-3. As shown, the
primary disadvantage is the failure of this approach to provide
for continuous monitoring, while the primary advantage is the
relatively low initial cost.
An alternative to collection and analysis of samples from
wells is the use of monitoring sensors which measure electrical
resistivity or thermal conductivity to detect leaks. As shown in
Table 5-3, a significant advantage of this approach is that it
provides for continuous monitoring at a relatively low EUAC and
5-21

-------
TABLE 5-3. ENVIRONMENTAL MONITORING SUMMARY*
Model
Facility Type
Method
Disadvantages/limitations
Advantages
"incremental""Cost (TF*"	
Initial Annual EUAC**
Existing -
smal 1
Ground water sampling
(quarterly)
Source of any detected contamination
may be difficult to identify.
Duration of leak which can go un-
detected depends on soil permeabil-
ity, waste type, well placement,
sampling frequency, etc.
Does not provide for continuous
monitoring.
-	Can detect both tank and
ancillary equipment leaks.
-	Lower initial cost than other
monitoring methods.
-	Available.
-	Soils can also be sampled during
well installation.
1,500
950
1,100
en
i
ro
no
Ground water wells
with conductivity or
resistivity sensors.
Volatile gas monitor-
ing with stationary
probes.
Source of any detected contami-
nation may be difficult to identify.
Duration of leak which can go un-
detected depends on soil permea-
bility, waste type, well placement,
etc.
Applicability limited to volatile
materials with appropriate sensors
available.
Duration of a leak which can go
undetected depends on soil permeabil-
ity, well placement, waste type, etc.
Source of any detected contamination
may be difficult to identify.
-	Can detect both tank and ancil-	5,000
lary equipment leaks.
-	Available
-	Provides for continuous monitor-
ing.
-	Soils can also be sampled during
well installation.
-	Can,detect both tank and ancil-	3,700
lary equipment leaks and appli-
cability is independent of ground
water depth.
-	Provides for continuous monitor-
ing.
Soils can also be sampled
during probe installation.
Available.
350
690
650
900
Soil water sampling
and analysis.
-	May not be applicable to same
wastes.'
-	Duration of a leak which can go un-
detected depends on soil permeabil-
ity, lysimeter placement, waste
type, etc.
Can detect both tank and ancil-
lary equipment leaks.
Available.
Soils can also be sampled
during Installation.
1,400
950
1,000
*See accompanying text for"aifflTtronaTTnforin.ition on assumptions use3"fh developing this table. Costs are roundeJ to two sTgniTfcant figures.

-------
TABLE 5-3 (CONTINUED)
Model
FaciIity Type
Method
Disadvantages/limitations
Advantages
Initial
Increniental Cost (tp~
Annual EUAC*'
-	Does not provide for continuous
monitoring.
-	Source of any detected contamination.
Existing -
medium
en
i
ro
CO
New - sma11
New - medium
Ground water sampling
Ground water wells
with sensors.
Gas wells with sensors.
Soil water sampling
Ground water sampling
Ground water wells
with sensors.
Gas wells with sensors.
Soil water sampling.
Ground water sampling
Ground water wells
with sensors.
Gas wells with sensors.
Soil water sampling.
Same as ahove.
Same as above.
1,500
5,000
3,700
1,700
1,2/0
3.500
1,000
1,030
1,700
4,700
2,100
1,200
950
350
650
1,000
900
350
350
900
950
350
700
950
1,100
690
900
1,100
990
590
420
970
1,100
670
(140
1,000
t ImVVimso TiT cost Trom Uie Tja se I ine Tac i I ity.
" I (|iiivdIi-nl Uniform Annual Cost.

-------
is equally applicable to both new and existing installations. A
potential disadvantage is that experience with use of sensors in
observation wells for monitoring tank leakage is limited. In
addition, the sensitivity of sensors is less than that for the
sampling and analysis approach.
The costs shown are based on the same assumptions listed
above for the sampling and analysis approach, with the following
changes:
•	the $200 initial cost for sampling equipment is deleted;
t the installed cost for sensor equipment at existing
facilities is $3700;
•	the installed costs for sensor equipment at small and
medium new facilities are $2600 and $3150 respectively;
and
•	annual costs are associated with daily readings of the
sensor control unit, which require 5 minutes per day, 260
days per year at a cost of $16 per hour.
A large percentage of the initial cost is associated with the
sensor control unit which can be used to monitor .multiple
sonsors. Thus, there are significant economies of scale for this
method.
For underground tanks which contain volatile wastes, monitor-
ing for waste vapors is a third method of environmental monitop-
ing. As with monitoring of ground water wells, vapor monitoring
can be accomplished through continuous measurement or sample
collection and analysis. For continuous monitoring, a detection
device is mounted in an observation well, while for the sampl-
ing/analysis approach, samples are periodically taken from the
well for laboratory analysis. Only the continuous approach to to
vapor monitoring is discussed in detail due to the leak detection
advantages of continuous monitoring and the similar costs asso-
ciated with the two approaches.
The advantages, disadvantages and costs associated with the
use of continuous vapor monitoring are shown in Table 5-3. As
shown", the primary advantage of this approach it that applicabil-
ity is independent of ground water depth, hut the primary
disadvantage is that applicability is limited to volatile wastes.
The vapor monitoring approach, has the second highest initial cost
but the second lowest EUAC, based on the following assumptions:
•	both existing facility sizes require 3 monitoring sensors,
with initial costs as follows;
- mobilization cost of $300;
sensor depth of 10 feet;
installed cost of 2 inch casing of $14/foot;
5-24

-------
installed cost of sensors at $980 each;
•	a new small facility requires one sensor installed below
the tank with an installed cost of $1040 complete;
•	a new medium facility requires one sensor under each tank,
with a total installed cost of $2080 for the facility;
•	annual costs for existing facilities and a new medium
facility include daily reading of the monitoring devices
which requires 10 minutes/day, 260 days/year at a cost of
$16/hour; and
•	annual costs for a new small facility include daily
reading of the monitoring device which requires 5 min-
utes/day, 260 days/year at a cost of $16/hour.
The fourth approach to "morn*toring for leak detection involves
the use of suction lysimeters to collect samples for analysis
from unsaturated soils. Suction lysimeters or comparable devices
have been used to collect water samples from unsaturated soils
for a wide variety of applications. Applicability for monitoring
hazardous waste tanks will depend on a variety of factors such as
waste type, soil conditions and climate. Where lysimeters can be
used, soil conditions and tank configuration and size determine
the number and location of samplers required.
The advantages, disadvantages and costs associated with the
use* of lysimeters are shown in Tab.le 5-3.. As shown, the primary .
advantages of this approach are that the cost is relatively low
and it can be used to monitor for leakage of non-volatile wastes
in areas where the saturated zone is relatively deep. The
primary disadvantage is that sample collection and analysis from
lysimeters does not provide for continuous monitoring. In addi-
tion, lysimeters tend to be more susceptable to clogging than
wells.
The estimated costs for monitoring with lysimeters shown in
Table 5-3 indicate that this approach has the lowest initial
cost, but the second highest EUAC based on the following
as s umpt ions:
•	an existing small facility requires 3 lysimeters (includ-
ing one background) with an initial cost as follows;
mobilization cost of $300;
installed lysimeter depth of 10 feet;
drilling cost of $12/foot;
installation cost of $300;
pump, lysimeter and sampling equipment costs of $480;
5-25

-------
•	an existing medium facility requires 4 lysimeters (includ-
ing one background) with an initial cost which is $290
above that of the existing small facility;
•	a new small facility requires 2 lysimeters (one below the
tank and one background) with costs as follows;
mobilization cost of $300;
installation of background lysimeter at 10 depth with a
drilling cost of $12/foot;
installation of 2 lysimeters (with one in fill materi a 1
below the tank) at a cost of $170 each (including the
lysi meter);
pump and sampling equipment costs of $270;
•	a new medium facility requires 3 lysimeters (one back-
ground and one under each tank) at a cost of $170 more
than the new small facility; and
t annual costs include sample collection costs of SBO/lysi-
meter and $800 in analysis costs (one background and one
composite from tank monitoring lysimeters taken quarterly,
yielding 8 samples per year, with analysis costs of $100
each).
All four of the methods of environmental monitoring discussed
above can reduce the estimated magnitude of release from under-
ground tanks by reducing the duration that a leak or rupture goes
undetected. The extent to which . magnitudes are reduced is
extremely dependent on: 1) appropriate selection and placement of
the monitoring devices: 2) the rate of waste migration from the
tank system to the monitor; and 3) waste type (solubility,
viscosity, etc.). The estimated reductions in release magnitude
shown below are based on arbitrary assumptions concerning release
duration and are included only to permit comparison of this
option with the other prevention/mitigation measures discussed:
(The event durations assumed are thought to be reasonable and
sufficient to allow for meaningful comparison with other op-
tions.)
Reduction i n Maste Released per
Event b'y~Rodel Facility Size7
for continuous mom toring
Event
Tank
Tank
Leak
Rupture
ianK Kupture
Ancillary Equipment Leak
Ancillary Equipment Rupture
Smal 1
Med i um
gal, percent gal, percent
1586
0
119
0
99
0
99
0
1586
0
694
0
99
0
99
0
5-26

-------
* These values are based on the assumption that it takes two
weeks for waste released from the tank system to appear at
the monitoring sensors. Thus, ruptures are discovered as a
result of the daily tank level monitoring conducted at the
facility and not as a result of environmental monitoring.
The migration time can expected to be somewhat less for
new installations, but well within the confidence interval
of the these estimates.
Reduction i n Waste Re 1 eased per
Event by Model Faci1ity Size*
for i ntermittent monitoring
Event
Tank Leak
Tank Rupture
Ancillary Equipment Leak
Ancillary Equipment Rupture
Smal 1
Medi um
gal. percent gal. percent
1558
n
117
0
97
n
98
0
1558
0
fi 82
0
97
0
97
0
* These values are based 'on the assumption that it takes two
weeks for waste released from the tank system to appear at
the monitoring well or lysimeter and that the leak occurs
at the midpoint of the monitoring cycle. Thus, ruptures
are discovered as a result of the daily tank level
monitoring conducted at the facility and not as a result
of environmental monitoring.
As discussed above, applicability of the four methods varies
with waste type and environmental setting. Thus, selection of a
particular method will generally be based on site-specific
factors and will not include consideration of relative" effec-
tiveness (in terms of the release probabilities shown above) of
the methods. Since leak detection is after the fact for all of
the monitoring approaches, there is n_£ effect on estimated
release probabilities.
Inventory Monitori ng
Another method of monitoring for tank system leakage involves
monitoring of waste quantities. Delivery of hazardous waste to
the storage tanks at the model facilities is assumed to be
accomplished through a gravity piping system since this appraoch
is generally less costly and more reliable than pressure delivery
and is frequently possible with underground tanks. Methods
available for gauging of gravity flow pipes include liquid level
sensors or Venturi meters. Use of the liquid level measurement
technique requires computation of flow using pipe slope and
roughness coefficients, and .would be inexact in the relatively
small diameter pipe used in underground tank systems. Use of a
venturi meter requires that the pipe be full (since it only
5-27

-------
measures velocity), and this condition is not typical. Thus,
gauging	of the liquid level in the tank over a time period
without	withdrawals or additions of waste (e.g., a weekend) is
the most	probable method of inventory monitoring.
The advantages, disadvantages and costs associated with three
methods of inventory monitoring are shown in Table 5-4. For
storage of product (i.e., gasoline) in underground tanks, the
traditional method of level measurement is the dipstick. Use of
this method for hazardous waste storage tanks has a number of
disadvantages, including:
•	lower accuracy than automated methods;
•	more labor intensive than automated methods;
•	may not be performed as scheduled (i.e., due to inclement
weather);
•	presents the potential for release of the waste stored via
material retained on the dipstick when it is removed from
the tank ;
•	water seepage into a tank in the event of a leak or
rupture mau prevent leak detection; and
•	presents the potential for increased worker exposure to
the waste.
The principal advantage associated with this approach is the lack
of an initial cost, although the EUAC is higher than for some
other methods.
A wide variety of methods exist for level monitoring as
indicated by the some 22 different types of level gauging
equipment discussed in a recent state-of-the-art survey [7].
These range from simple float type level indicators which are
read at the fill port to electronic level sensors with remote
indicator and recorder at a control panel. Selection of a
specific monitoring system for an underground hazardous waste
storage tank by a design engineer would include consideration of
cost, accuracy, reliability, simplicity, time requirements and
possible complications associated with use (e.g., increased
potential for fire associated with bubbler tube measurement of
i g n itable waste).
To represent a range of the equipment which may be used at a
facility, both direct and remote read-out level sensing equipment
are included in Table 5-4. As shown, the remote read-out
approach has the disadvantage of a higher initial cost, but has
the advantages of a lower EUAC and is less susceptible to
operator errors.
5-28

-------
TABLE 5-4
INVENTORY MONITORING SUMMARY*
Model
Facility Type
Method
Disadvantages/!imitations
Advantages
Incremental Cost
Initial Annual
loAC*
Existing -
smal 1
Dip stick
CJl
I
IM
10
Relatively low accuracy.
May not be performed as scheduled.
More labor intensive than automated
methods.
Potential for worker exposure to
waste.
Potential for release of the stored
waste via material retained on the
dipstick where it is removed.
Sensitivity depends on length of
time between measurements and ac-
curate records of previous measure-
ments.
Very low initial cost.
Available.
Detection of large releases.
700
700
Level sensor (pneumatic-
read at tank location).
-	Applicability of specific equipment
affected by waste type.
-	Readings may not be made (i.e., in-
clement weather).
-	More labor intensive than remote
readout systems.
-	Cannot detect small leaks.
-	More accurate than dipstick.
-	Detection of large releases.
-	Available.
850
190
550
Level sensor (electronic-
remote records and readout).
-	Applicability of specific equipment
affected by xaste type.
-	Cannot detect small, leaks.
-	Relatively high initial cost as
compared with the other two methods.
-	Cannot detect small leaks
-	More accurate than dipstick.
-	Detention of large releases.
-	Effectiveness relatively inde-
pendent of operator.
-	Low recurring costs.
-	Available.
2,200
350
500
*See accompanying text for additional information on assumptions used in developing this table. Costs are rounded to two significant figures.

-------
TABLE 5-4 (CONTINUED)
Model
Facility Type
Method
Disadvantages/1 imitations
Advantages
Incremental Cost fj]
Annual
Initial
TJac""
Existing -
med i um
New - sma11
New - medium
Dipstick
Level sensor (at tank)
Level sensor (remote)
Dipstick
Level sensor (at tank)
Level sensor (remote)
Dipstick
Level sensor (at tank)
Level sensor (remote)
Same as above
Same as above
1,700
4,300
U50
2,100
1,700
4,100
1,000
700
350
700
490
350
1,100
700
350
1,100
810
640
700
550
490
1,100
810
630
~Increase in cost from the baseline facility.
** Equivalent Uniform Annua) Cost.

-------
In general, inventory monitoring can help to reduce release
magnitudes but has no effect on release probability since
releases are detected after they occur. However, the ability of
inventory monitoring to reduce release magnitude is limited by
the size of the minimum detectable leak, which is controlled by a
variety of factors, including:
•	the temperature of the tank and waste	contained. The
significance of this factor is determined	by the coeffi-
cient of expansion of the waste and	the degree of
temperature fluctuation;
•	the extent to which waste material is lost through
vaporization, which is affected, by the waste temperature
and vapor pressure;
•	the accuracy of the level measuring technique used, which
is determined by the specific device used, the volume of
the tank and the level of waste in the tank; and
•	the effects of water inflow (in the event of a leak or
rupture) on tank level reading.
Based on these factors and experience with gasoline stations [81,
leaks of less than approximately 15 gallons/day cannot be
reliably detected with inventory monitoring. Thus, inventory
monitoring is helpful in reducing the magnitude of rupture events
but does not reduce the magnitude of leak events for the model
faci1ities , as fol1ows :
R&duct i on i n Wa s t e Released per
Event by Model Fac i1i ty Size*
for conti nuous monitoring
Event	Smal1	Medi um
gal, percent gaTT pe" r c e n t
Tank Leak

0
n
0
0
Tank Rupture

50
10
150
f
Ancillary Equipment
Leak
0
0
0
0
Anci11ary Equipment
Rupture
45
50
. 405
75
* These values are based on the assumptions that leak rates
are below the detection limit of the inventory monitoring
equipment. In addition, -it is assumed that tank rupture
is discovered within one day, but that the tank contents
have been lost by this time (so that a single batch
transfer is the only release reduction). The ancillary
equipment.rupture values were estimated assuming that the
rupture is discovered after one transfer, and that 90
percent of the transfer was released.
5-31

-------
Internal Inspection
Inspection of tanks can be used to detect actual leaks, or to
locate potential leak locations resulting from corrosion or other
damage to the tank, liner or coating material. Since the
hazardous waste storage tanks considered in this project are
completely buried in the ground, inspection of the tanks can only
be accomplished from inside the tank, and then only if the tank
has a manway. Since internal inspection is limited to tanks,
this mitigation/prevention measure does not help to control
ancillary equipment release events.
Before a tank can be inspected, any waste contained must be
pumped into containers, another on-site tank or a tank truck; the
tank atmosphere decontaminated to allow personnel entry with the
minimum of danger to health and safety; and the tank cleaned.
Decontamination of the tank atmosphere may not always be required
prior to cleaning, but is assumed to be a typical part of the
tank inspection process. This is assumed to normally be accom-
plished by creating an inert atmonphere in the tank using dry ice
[9].
Cleaning of the tank can be accomplished via a variety of
methods, including sand blasting, hydro-blasting, steam cleaning
and/or chemical cleaning. Selection of a cleaning method is
somewhat dependent on the tank contents, condition and material
of construction. Data on the relative prevalence of these
methods for cleaning underground tanks are not available.
Sand blasting has the advantage that contractors capable of
providing the service can be assumed to be readily available.
However, it has the disadvantage of creating dust within the tank
which makes monitoring of the cleaning process more difficult.
Hydro-blasting is similar to sand blasting except the abrasives
used in the cleaning process are suspended in water. The
principal advantage of this approach is that the progress of the
cleaning process is more easily monitored (visually) than with
sand blasting. Steam cleaning has characteristics similar to
hydroblasting, and the choice between the two would primarily be
determined by the type of waste in the tank to be cleaned.
Chemical cleaning has the disadvantage of generally being slower
and more costly than the other cleaning options and normally is
used only as a last resort in tank cleaning in preparation for
inspection [9].
Following cleaning, the inspection process is assumed to
proceed with a visual inspection and subsequent use of ultrasonic
equipment. Visual inspection only was considered, but this
approach to inspection was considered to be unacceptable since it
can identify only relatively large defects on the inside of the
tank and cannot detect potential problem areas on the outside of
the tank.
5-32

-------
For aboveground tanks, a ball peen hammer is normally used to
aid in the inspectionn process and is considered to give a
reliable indication of where thinning of the shell has occurred.
On underground tanks, variations in backfill characteristices
such as moisture content, compaction and material make this
approach less reliable. The rationale for selection of ultra-
sonic equipment (rather than other available techniques such as
radiography) to improve the effectiveness of inspection for
detection/prevention of tank leaks is present in Appendix D.
While the visual inspection of the tank interior will help to
locate some potential problem areas (e.g., spot corrosion, etc.),
it obviously will not detect potential problems which exist on
the exterior of the tank shell. Thus, it is assumed that the
entire tank will be tested ultrasonically from the inside.
As with the other alternatives discussed, costs will vary
depending on the specific implementation. For inspection with
ul-trasonic equipment, significant cost variables include: dis-
tance between the tank facility and the location of the inspec-
tion contractor (if a contractor is used); cleaning materials
disposal cost; and method of tank cleaning.
The advantages, disadvantages and costs associated with
internal tank inspection are presented in Table 5-5. As shown,
the primary advantage of internal inspection is that some (but
not all) developing problems may be identified. The primary
disadvantage is that thickness measurements are made on a
relatvely small percentage "of the tank surface. Thus, localized
problems (such as small perforations resulting from point corro-
sion)' may go undetected. Other disadvantages are that ancillary
equipment releases are not effected and that the tank must be
taken our to service to permit inspection. The costs shown for
inspection are based on the following assumptions:
« there will be no initial costs since inspection will be
performed by a contractor. This assumption is made based
on the relatively high cost of the equipment involved as
compared to' contractor rates and the assumed inspection
frequency (annual). Additional information related to
this and other assumptions is provided in Appendix D;
•	the inspection contractor charges for travel time (one-
half hour each way is assumed) between his location and
the tank faci1ity;
•	the operator "empties" the tank prior to the arrival of
the tank cleaning crew using his normal methods and the
cost of this activity is not part of the inspection cost;
•	cleaning a 1,000 gallon tank requires a 2-man crew for 4
hours and cleaning of two 5,000 gallon tanks requires a 2 -
man crew for 8 hours (including travel time) at a rate of
$120/hour for hydroblast cleaning, including materials and
breathing apparatus [9];
5-33

-------
TABLE 5-5. INSPECTION SUMMARY*
Model	Method	Di sadvantages/) imi tat ions	Advantages	IncrementaT Cost fff*	
Faci I i ty Type	Initial Annual EUAC**
Existing
smal I
Visual (internal) and
ultrasonic.
Tank must be enterable for inspec-
tion.
Training and experience required
for proper inspection makes use of
a contracted service desirable.
Not applicable to ancillary equip-
ment .
Both existing and developing prob-
lems may go undetected due to the
point measurement nature of the
equipment.
Since tank wall thickness is
measured, developing as well
as existing leaks can be identi-
fied .
-	No initial cost.
-	Tank cleaning and inspection
contractors are readily avail
able.
730
730
Ex ist ing -
medium
New - small
New - medium
Saine as above
Same as above
Same as above
2,300 2,300
730	730
2,300 2,300
*See accompanying text for additional information on assumptions used in developing this table. Costs are rounded to two significant figures.
+ Increase in cost from the baseline facility.
* Equivalent Uniform Annual Cost.

-------
•	cleaning materials are removed from the tank facility by
the cleaning contractor and disposal of these materials is
included in the cleaning rate charge;
•	an ultrasonic survey of the tank interior with subsequent
additional measurements made in potential problem areas
identified in the survey or thorough visual inspection is
assumed. The survey of the tank interior is assumed to be
performed at a rate of 60 square feet per hour (one
measurement per square foot) while a detailed inspection
rate of 3 square feet per hour was assumed. Detailed
inspection is assumed to be performed on 10 percent of the
tank. The cost for ultrasonic testing is $25/hour.
•	the 1,000 gallon tank has a diameter of 4 feet and length
of 10.6 feet and the 5,000 gallon tanks are each 7 feet in
diameter and 17.4 feet long.
The effectiveness of internal tank inspection in mitigating
tank releases depends primarily on the frequency of inspection
since this controls the release duration. For purposes of
comparison, annual inspection has been assumed, which results in
the following reductions in release magnitude:
Reduct i on in Waste Released pe r
Event by^odel Facility Size*
Event
Tank Leak
Tank Rupture
Ancillary Equipment Leak
Ancillary Equipment Ruptun
* These values are based
Sma 11
Med i um
gal . percent gal. percent
1420
0
0
0
89
0
0
0
1420
0
0
0
R9
0
0
0
on the assumptions that tank leak
begins in the middle of the inspection cycle. Thus, the
leak goes undetected for 180 days and release occurs at a
rate of 1 gallon per day.
Tank inspection may impact release probabilities for two
reasons. First, the methods available for inspection of under-
ground tanks are such that existing small leaks may go undetect-
ed. This may occur if the leaks are not revealed by visual
inspection and the problem is sufficiently localized that it goes
undetected in the ultrasonic survey of the tank. On the other
hand, some developing leaks may be detected before they occur,
thereby tending to reduce the release probability. Within the
context of the order of magnitude estimates of release probabili-
ties developed, the effect of the these two factors are judged to
balance each other such that internal inspection does not effect
release probabi1ity.
5-35

-------
Corrosion Protection
Corrosion protection may be provided for a variety of tank
system materials in a variety of ways. Regardless of the
material being protected or the approach used, the objective of
corrosion protection is to protect the tank system materials
from corrosion during their intended service life. As such,
corrosion protection helps to control tank and ancillary leak and
ruptures by protecting against the most frequent cause of tank
and pipe equipment failure. In addition, it helps to minimize
increased corrosion which may be caused during installation.
Discussion of the mechanisms of corrosion and alternative
approaches to corrosion protection are provided in Appendix A and
are the subject of a large body of literature. As discussed
here, corrosion protection applies to protection of steel tank
system equipment only, although it is recognized that corrosion
protection is occasionally required for other construction mater-
ials as well. A wide variety approaches to corrosion protection
are possible, including anodic protection, cathodic protection,
linings, coatings, compressive strength induction, etc. For
underground tank systems, the three methods which are most
commonly used and which are discussed here are external coatings,
internal linings and cathodic protection (either impressed cur-
rent or sacrificial anodes).
External coatings may be used alone or in conjunction with
cathodic protection (normally sacrificial anode(s)) to protect
underground steel tanks and . .piping from corrosion. A wide
variety of coating materials are commonly used for corrosion
protection, including both generic and trademarked materials.
Selection of a material will depend on site-specific installation
conditions and the equipment supplier (since not every coating
material will be available from a given supplier).
Since coatings may be damaged during shipping and installa-
tion, thereby creating point corrosion problems, coatings are
most effective when used in conjunction with a sacrificial anode.
In fact, sacraficial anodes are normally used only on coated
tanks since the coating significantly reduce the size of the
anode required to provide protection throughout the normal tank
system design life (20 years). Additional discussion of sacrafi-
cial anodes is provided below.
The advantages, disadvantages and costs associated with use
of external coatings for corrosion protection are presented in
Table 5-6. As shown, coatings are not judged to be applicable to
existing tanks, primarily since the cost of retrofit application
makes purchase of a new tank preferable. Use of coatings on new
steel tank systems is common, at least in part due to National
Fire Protection Association Codes 30 and 31 which include'(as of
1981) a responsibility for cathodic protection or corrosion
resistant materials. The costs shown are based on factory
installation of a coal-tar epoxy coating (e.g., Koppers 300M or
equal).
5-36

-------
TABLE 5-6 CORROSION PROTECTION SUMMARY*
Mode)
FaciIity Type
Method
Disadvantages/!imitations
Advantages
Incremental Cost ftp	
Initial Annual EOAC**
Existing -
smal I
Externa) coatings.
Generally not applicable for retro-
fit of existing tanks.
Internal lining for
tanks.
Lining flaws more of a problem
than for factory applied linings
on new tanks.
Tank must be enterable. Small tank
dimensions may make lining Instal-
lation expensive and/or impractical.
-	Available
-	May prolong the service life
of a tank which has developed
minor leaks.
-	May be used to alter tank and
waste compatabi1ity.
2,000
130
CD
I
00
Cathodic protection
impressed current.
Requires partial excavation of
system for installation.
-	Less expensive than replacement.
-	Available.
-	Applicable to both tanks and
piping.
5,000
60
400
Exist ing
medium
Cathodic protection
sacrificial anodes
External coatings.
Internal linings.
Cathodic protection
impressed current
Cathodic protection
sacrificial anodes
-	Requires partial excavation of
system for installation.
-	Applicability limited to tanks
which were coated prior to instal-
1 at ion.
Same as above
-	Less expensive than replacement
-	Available.
-	Applicable to both tanks and
piping.
Same as above.
1,200
00
6,300
5,000
3,400
120
460
460
230
•See accompanying text "for adilTFional information on assumptions" useTe. "TosTs are rounded to two significant fTyifres.

-------
TABLE 5-6 (CONTINUED)
Model
Faci 1 ityType
Method
Disadvantages/1 imitations
New - small	External coatings.
Internal linings.
Drainage during installation may
reduce tank life below that of a
bare tank due to erection of point
anode(s) where accelerated corro-
sion may occur.
Requires that tank be constructed
with a manway.
Cost highly dependent on lining
material.
Generally not applicable to
piping.
cn
i
co
oo
Cathodic protection
impressed current.
-	On-going power consumption
-	More expensive than sacrificial
anodes for small single tank
faci1ities.
Cathodic protection
sacrificial anodes
Anode size depends on site-
specific conditions and design life
New - medium External coating.	Same as above.
Internal lining	"
Cathodic protection-
impressed current	"
Cathodic protection-
sacrificial anodes	"
~Increase in cost from the baseTFne facility.
** Equivalent Uniform Annual Cost.
Advantages
Incremental Cost ($)*
Initial Annual EUAC**
Available
Generally low cost.
No maintenance required.
Available
400
1,200
30
80
May be used to modify tank
and waste compatibility.
-	Available.
-	Applicable to both tanks and
pipes.
5,000
60
400
-	Available.
-	Very little maintenance required.
-	Applicable to both tanks and
piping.
450
30
Same as above.
1,400
4,800
90
320
5,000
120
460
1,800
140

-------
Linings in tanks' and piping may be used to protect them from
the corrosive effects of the material contained. The advantages,
disadvantages and costs associated with tank lining for both
existing and new facilities are presented in Table 5-6. As
shown, the primary limitations are that application to existing
tanks requires access into the tank and that problems with
quality control may be greater than for factory installed linings
on new tanks. On the other hand, linings have the advantages
that they may be used to extend the service life of a tank or
allow a change in the material stored;
The costs shown in Table 5-6 were derived based on the
following assumptions:
•	an existing small facility requires 4 hours to clean at a
cost of $120/ hour prior to lining (158 square feet) with
an epoxy resin which costs $9.50/square foot;
•	an existing medium facility requires 8 hours to clean at a
cost of $120/hour prior to lining (918 square feet total)
with an epoxy resin which costs $6.50/square foot;
•	lining a new tank costs 20 percent less than lining an
existing tank; and
•	tank interiors are prepared in accordance with the Steel
Structures Painting Council Specification No. 6 Commercial
Blast Cleaning.
Cathodic protection can be provided through the use of
impressed current or sacrificial anodes (see Appendix A for more
detail). As shown in Table 5-6, sacrificial anodes have the
disadvantage of having applicability limited for existing tanks
to tanks which were coated prior to installation. The primary
advantage of both approaches is that they can effectively protect
against both internal and external corrosion, excepting corrosion
caused by incompatable waste materials.
As shown, impressed current is substantially more expensive
for use at the model facilities than is the sacraficial anode
method of cathodic protection, and both provide the same type of
protection. As a result, impressed current will normally be used
only at large facilities with a large number of tanks and/or
extensive piping networks to protect. Costs shown in Table 5-6
for cathodic protection are based on the following assumptions:
•	a typical minimum cost for an impressed current corrosion
protection system is 55000;
•	installation of sacraficial anodes at existing facilities
requ i res ;
removal of existing pavement
excavation to the top of the tank
5-39

-------
installation of anodes'
2 at 9 pounds each for the small facility
2 at 24 pounds each for the medium facility
backfill and compaction
replace pavement
haul away old pavement
• for new facilities, the costs are based on factory
installation of a coating and sacraficial anodes (size and
number identical to existing facilities) to provide sti-P3
(TM) type of protection. It should be noted that tanks
porvided with both coatings and sacrificial anodes for
corrosion protection are readily available.
All of the methods of corrosion protection discussed are
judged to reduce the chance of a leak occurring. Use of either-
impressed current or a coating in combination with sacraficial
anodes is estimated to reduce the probabi1ity of release by one
order of magnitude. Use of coatings or linings alone will result
in some reduction in release probability which is estimated to be
less than one order of magnitude. Once a leak occurs, however,
corrosion protection is not thought to have a significant impact
on the leak rate, and therefore does not reduce the estimated
release magnitude associated with tank system leak or rupture.
CONCLUSIONS
The advantages, disadvantages and equivalent uniform annual
costs (EUAC) associated with each of the six release mitiga-
tion/prevention measures discussed above are summarized in Table
5-7. The costs shown are incremental costs, and as a result costs
for new facilities are significantly lower than for existing
facilities for methods which involve significant construction
costs (e.g. secondary containment).
As shown, secondary containment is the most expensive (based
on EUAC) of the control methods examined for both the small and
medium sized model facility under both new and retrofit condi-
tions. Internal inspection is the second most expensive method,
with corrosion protection the least expensive method. The bene-
fit of the greater expense associated with secondary containment
is that this method, unlike all of the others discussed, reduces
both the probability and magnitude of release events.
The effects of the release prevention/mitigation measures
discussed above on the estimated probability of tank leak, tank
rupture, ancillary equipment leak and ancillary equipment rupture
release events are summarized in Table 5-8. As shown, secondary
containment is clearly the most effect means of preventing both
leak and rupture events. Corrosion protection also serves to
reduce the estimated release probability, and as shown, it also
can control all four release events. Other measures, such as
tank system testing and environmental monitoring, serve to
5-40

-------
TABLE 5-7
Type of
"Solution"
Disadvantages/limitations"
Secondary - Double walled tanks have limited
containment	availability in materials other
for tanks	than steel (including stainless).+
-	Concrete vaults may not be applic-
able to ignitable waste due to
local code requirements.
-	Clean-up of a release to a synthe-
tic liner secondary containment area
will be more difficult and expensive
than for other methods, but will still
I	be less expensive than clean-up of an
-P*	environmental release.
Secondary - Concrete trenches may not be ap-
containment plicable to ignitable waste due
for ancil-	to local code requirements,
lary equip-
ment .
Tank sys- - Limited track record for waste
tern testing tank system testing.
-	Applicability of some tests de-
pend on waste type.
-	Will not help to prevent leaks;
rather, it will help to minimize
volume through earlier detection.
-	Does not provide continuous moni-
toring.
"SOLUTION" COMPARISON
SUMMARY*
Advantages
Cleanup of releases to secondary
containment area easier and less
costly than environmental clean-ups,
and especially easy for double
walled tanks.
Concrete vaults applicable to all
types of tank materials.
Provides for containment and de-
tection of tank releases prior to
environmental release.
Provides for detection of secondary
containment failure independent of
primary containment failure.
Effectiveness
TUAC"
Ex
Tf~5y FacilTty
sting
New
small medium sina 11 ineflium
Very effective in pre- 1,400 3,400 970 2,
venting environmental
release from tanks.
Concrete trenches can be integrated
with concrete tank vaults to pro-
vide continued capacity in excess
of tank volume.
Provides for containment and detec-
tion of ancillary equipment (esp.
pipe) releases prior to environ-
mental release.
Clean-up of release to secondary con-
tainment area easier and less costly
than environmental clean-up.
Very effective in pre-
venting environmental
releases from piping and
other ancillary equip-
ment.
1,600 3,800 1,100 2,001
Provides for detection of tank and
piping leaks. Size of detectable
leak varies with test, but Is general-
ly in the range of 0.03 to 0.05 gal-
lons per hour.
Generally effective for 500
tank and pipe leak detec-
tion, although small
leaks may go undetected
and errors in performing
a test may cause enor-
mous results. Also depen-
dent on testing frequency
(assumed here to be annual).
800
500
800

-------
TABLE 5-7 (Continued)
" Type "of-
"Solution"
Environmen-
tal monitor-
ing
0Fsarfvant ages /Tfinf t a t ions "
Advantages
Specific method applicable depends
on site conditions and waste type.
Duration which a leak may go unde-
tected depends on soil permeability,
well/sensor location, waste charac-
teristics, etc.
Source of any detailed contamination
may be difficult to identify.
Can provide continuous monitoring for
tank system leaks.
Soils may also be sampled during in-
stallation. Especially applicable
to existing facilities.
"Effectiveness
Tuac
"IT by Facility Type*
Existing"	
smal 1 medium small" medium
New
Generally effective for
detecting releases.
However, the duration
of leakage prior to de-
tection depends on site
specific conditions.
690
690
590
590
en
i
-Pi
ro
Inventory - Difficult to implement when tanks
monitoring	are filled by trickle gravity flow.
Internal - Tank must be taken out of service,
inspec-	generally for at least one to two
tion	days.
- Does not address possible ancillary
equipment problems.
Corrosion - External coatings required. Damage
protection during installation may increase
corrosion rate.
- Monitoring of performance generally
requires use of one or more of the
other "solutions" discussed.
Hill defect ruptures or large leaks.
Speed of detection depends on fre-
quency of measurement, which may be
continuous or intermittent.
Can detect developing problems before
leaks occur.
Tank cleaning precedes inspection,
so the use of the tank can be changed
relatively easily following Inspection.
Protects against' both tank and piping
leaks resulting from corrosion.
- Little or no maintenance required.
Generally effective in 500 640 490
detecting ruptures or
large leaks.
Can help to prevent	730 2,300 730
leaks, but reliability
is not known; can also
indicate existing
leaks, but is thought
to be less reliable
than tank system
testing.
Generally effective in 80 220 30
preventing tank and
pipe (primarily steel)
releases due to corro-
sion failure, but does
not provide a mechanism
for readily checking the
system performance.
630
2,300
140
»FKP tanks are also available in sizes	up to 4,000 gallons.
* See accompanying text for additional	information on assumptions used in developing this table. Costs are roun.lef,' to two significant figures.
** Equivalent Uniform Annual Costs	which represent least cost methods presented in Tables 5-1 through 5-6. Note that these least costs methods may not be
applicable in some situations.

-------
TABLE 5-8. MODEL FACILITY RELEASE PROBABILITIES*
tn
CO
So 1u 11 on
Me t hod
Order of Magnitude Change In Estimated Release Probability by Event^
Tank Leak
Tank Rapture
Anc11 1 a r y
Equ1proent Leak
Anc111ar y
Equipment Rupture
Tank secondary
Concrete vault for tanks
-4
-4
0
0
conta1nment
w/contlnuous monitoring





Synthetic liner for tank
-4
-4
0
0

excavat1 on





Double walled tanks
-4
-4
0
0
Anc111ary equ1p-
Concrete trench for
0
0
-4
-4
ment secondary
p 1 pes




conta1nment






Double walled piping
0
0
-4
-4
Tank system
Any method Identified
0*
0
0
0
tost 1ng*"
In Table 5-2




Env1ronmenta1
Any method Identified
0
0
0
0
mon1 tor 1ng **
In Table 5-3




1nventory
Any method Identified In
0
0
0
0
son 1 tor 1ng*"
In Table 5-4




1nspect1 on
Visual Inspection with
0
0
0
0

ultrasonlc testing




Corros1 on
Cathod 1c protect Ion
-1,
- 1
-1
-1
protect 1 on





• See accompanying text for additional Information on assumptions used In developing this table. Costs are rounded to
7 significant figures,
t F»f example, a value of -4 Indicates a reduction In the estimated probability of release I0~*
"¦ Will only Identify leaks after they have occurred

-------
mitigate the effects of releases by decreasing the release
magnitude and have no impact on the estimated release probabi1 -
ity.
From a release probability perspective, secondary containment
is the most cost effective method analyzed. This statement is
made since secondary containment for tank and ancillary equipment
provides a three order of magnitude greater decrease in release
probability than corrosion protection at a cost which is less
than two orders of magnitude greater.
It should be noted that the specific costs and release
probabilities which led to the above conclusion are based on a
variety of assumptions presented earlier in this Section. As-
sumptions regarding facility layout, materials of construction,
method of secondary containment, etc., all affect the cost of the
release mitigation/prevention measures discussed. These and
other assumptions also affect the estimated release- probabilities
shown in Table 5-8. However, secondary containment remains the
most cost effective method over a wide range of conditions.
The effects of the release prevention/mitigation measures on
the model facilities as a whole (instead of the effects on
individual release events) are presented in Tables 5-9 and 5-10
for existing and new facilities respectively. Effects on both
estimated release probabilities and magnitudes as well as equiva-
lent uniform annual costs (EUIC) associated with each measure are
also presented. As shown, secondary containment for both tank
and ancillary equipment provides a 99 percent decrease in the
estimated' rel ease magnitude. Although t'he cost associated with
this approach is among the highest shown, the cost per unit of
release reduction is approximately the same as for tank contain-
ment alone. Thus, containment for the entire tank system is
indicated to be a better investment in light of the very
significant reduction in release probability provided.
Mitigation measures such as tank system testing and environ-
mental monitoring are shown to provide significant reductions in
release magnitude at costs per unit of reduction which are about
half those asssociated with secondary containment. However, they
provide no reduction in the estimated release probability.
Inspections are also shown to result in reductions in release
magnitude without impacting the release probability.' While tank
inspection can result in the identification of developing prob-
1 ems before a leak or rupture occurs, measurements are taken on a
relatively small percentage of the tank surface area. Thus, it
was judged that while some reduction in the estimated relative
release probability will occur with tank inspection, the reduc-
tion will be less than one order of magnitude.
A prevention measure which has no impact on the estimated
release magnitude but. which results in an estimated release
probability reduction of one order of magnitude is corrosion
5-44

-------
TABLE 5-9. INCREMENTAL COST AND EFFECTIVENESS SUMMARY - EXISTING FACILITIES*


Reduction In Estimated Release Probability and Volume and Solution
Model Facility Type**
Cost by
So 1u tIon
Method
Sma 1 1
Med 1 urn


Re 1 ease
Proba-
bility
Release
Vo1ume
(gal Ions)
Release
Vo1ume
(percen t)
1ncre-
me nt a 1
E11 AC 
Re 1 ease
Proba-
bl11ty
Release
Vo1ume
(gal 1 ons )
Re 1 ease
Vo1ume
(percent >
1 ncre-
men ta 1
EUAC (S)
lank system
secondary
con t a 1nmen t
-Concrete vault for tank
& concrete trench for
piping with leak detec-
tion sensors and alarm
.0-"
1 595
99
1600
I0-4
3190
99
3500

-Double waited tank and
plplnq w/leak detection
i alarm
1
O
1595
99
1600
10-4
3190
99
37 00

-Synthetic liner contain-
ment for tank w/leak de-
tection sensors & alarm
0
1480 _
93
2900
0
1800
56
4900

-Concrete vault for tank
w/leak detection sensors
and a 1 arm
0
1480
93
1400
0
1800
56
3300

-Double walled tank w/
leak detection and alarm
0
1480
93
1400
0
1800
56
3400
Tank system
test Ing
-Most methods In Table
5-2
0
1420
89
500
0
2840
09
800
E n v1r onmenta 1
Mon1 tor 1ng
-Ground water wells w/
sensors & alarms
0
148 1
93
690
0
18 1 2
57
690
1nven tor y
Mon i t or Inq
-Level sensors w/remote
recorder, readout and
alarm
0
0
0
500
0
0
0
640
Inspect ion
-Internal visual and
ultrasonic Inspection
0
1420
89
730
0
1800
56
2300
Corros1 on
Protection
-Cathodlc Protection
10" '
0
0
80
10" 1
0
0
230
* See accompany Ing text for additional information on assumptions used in developing this table. Costs are rounded to
2 significant figures*
* * Estimated release volumes are presented to the gallon to help document how 1hey were derived. Accuracy Is, at best,
2 significant figures. Reduction volumes tor the medium facility are double those presented in the text in order fo
represent the reduction on per facility basis rather than a per event brisis.

-------
TABLE 5-10. INCREMENTAL COST AND EFFECTIVENESS SUMMARY - NEW FACILITIES*


Reduction In Estimated Release Probability and Volume and Solution
Model Facility Type""
Cost by
So 1u tI on
Method
Smal 1
Med 1um


Re 1 ease
Proba-
bi11ty
Re lease
Vo1ume
(gal Ions)
Re 1 ease
Vo1ume
(percent I
1 ncre-
men ta 1
EUAC 
Release
Proba-
bl II ty
Release
Vo1ume
(gallons)
Release
Vo1ume
(percent)
1ncr e-
mental
EUAC (S>
Tank syst em
secondary
contafnment
•Concrete vault for tank
& concrete trench for
piping with leak detec-
tion sensors and alarm
io-4
1595
99
1 100
IO"4
3190
99
1800

-Double walled tank and
piping w/leak detection
& alarm
10"4
1595
99
1200
I0"4
3190
99
2600

-Synthetic liner contain-
ment for tank w/leak de-
tection sensors & alarm
0
1480
93
2600
0
1800
56
3700

-Concrete vault for tank
w/leak detection sensors
and alarm
0
1480
93
970
0
laoo
56
1600

-Double wa1 led tank w/
leak detection and alarm
0
1480
93
970
0
1800
56
2400
Tank syst em
test 1ng
-Most methods In Table
5-2
0
1420
89
500
0
2840
89
800
Environmental
Hon 1 tor 1ng
-Ground water welts w/
sensors & alarms
0
1 4 B 1
93
590
0
1812
57
670
1n ventor y
Mon1 tor 1ng
-Level sensors w/remote
recorder, readout and
alarm
0
0
0
490
0
0
0
630
Inspection
-Internal visual and
ultrasonic Inspection
0
1420
89
730
0
1600
56
2300
Corros1 on
Protection
-Cathodlc Protect 1 on
10" 1
0
0
30
10" 1
0
0
140
* See accompanying text for additional information on assumptions used In developing this table. Costs are rounded to
2 significant figures.
* * Estimated release volumes are presented to the gallon to help document how they were derived. Accuracy is, at best,
2 significant figures. Reduction volumes for the medium facility are double those presented In the text In order to
represent the reduction on per facility basis rather than a per event basis.

-------
protection. For the costs presented in Tables 5-9 and 5-10,
corrosion protection was assumed to be provided by an external
coating and sacraficial anodes. Based on this assumption, corro-
sion protection is the least expensive method of achieving a
reduction in estimated release probability.
5-47

-------
REFERENCES
1.	Development Planning and Research Associates, Inc. "Work Plan
-- Seismic and Floodplain Regulatory Impact Analysis", Sub-
mitted to Putnam, Hayes & Rartlett, Inc., Cambridge, Mas-
sachusetts and U. S. Environmental Protection Agency, Wash-
ington, D. C. April, 1983.
2.	Development Planning and Research Associates, Inc. "Report
on Underground Hazardous Waste Treatment and Storage Tanks
(Preliminary Mail Survey Data)," U. S. Environmental Protec-
tion Agency, Office of Solid Waste, Washington, D. C. April
28, 1983.
3.	SCS Engineers. "Statistics from California Regional Water
Quality Board, San Francisco Bay Region, 1982, 'Manditory
Facility Ouestionnaire1 ", Reston, Virgina. May, 1983.
4.	JRB Associates. "Macroprofile: Hazardous Waste Tank and
Container Storage Facilities (Draft)", U. S. Environmental
Protection Agency, Office of So-lid Waste, Washington, D. C.
May, 1982.
5.	Hunter Environmental Services (Sunmark Industries), Philadel-
phia, PA. Personnal communication with SCS Engineers,
Reston, Virginia. August, 1983.
6.	Environment Canada. "Technical Information for Problem
Spills: Benzene", (Draft). Ottawa, Canada. December, 1981.
7.	Hall, John. "New Devices, Systems for Level Monitoring",
Instruments and Control Systems. October, 1982.
8.	Warren Rogers Associates (Warren Rogers), Providence, Rhode
Island. Personnal communication with SCS Engineers, Reston,
Virginia. May through July, 1983.
9.	Northwest Tank (Larry Peterson), Seattle, Washington.
Personnal communcation with SCS Engineers, Reston, Virginia.
May, 1983.
5-48

-------
APPENDIX A
Excerpts from "Technology for the Storage of
Hazardous Wastes, A State-of-the-Art Review",
New York State Department of Environmental
Conservation, January 1983

-------
INTRODUCTION
As part of their Bulk Storage Program, the New York State
Department of Environmental Conservation (DEC) prepared a State-
of-the-Art Review Manual applicable to underground and above-
ground storage system. Companion documents prepared as part of
this program included the "Manual on Criteria and Guidance for
Storing Hazardous Substances", the "Model Local Ordinance for
Storage of Hazardous Substances", and the "Siting Manual".
The State-of-the-Art Review Manual compiles much of the
latest information on the equipment available for storing and
handling hazardous liquids. Included are data on tanks, hoses,
overfill prevention devices, piping, valve, and pumps. Important
information is also provided on the field practices and equipment
available for leak detection and spill cleanup. It is a well-
prepared overview. Accordingly, the portion describing under-
ground storage systems and related background information are in-
cluded as an appendix to this report.
This material discusses the technology and practices for
storage of petroleum and other hazardous liquids which could be
accidentally released into the environment. It should be noted
that hazardous liquids vary widely in their characteristics and
in the manner in .which they should be stored.
This manual should serve only as a guide. Each chemical and
each environmental setting requires its own specific storage de-
sign. It is the responsibility of the owner of the storage
facility to seek the assistance of a professional engineer who
has the skills to design a storage system which can be us-ed safe-
ly and which provides the necessary measures for utility and en-
vironmental protection.
The mention of trade names or commercial products in this
manual does not constitute endorsement or recommendation for use
by the DEC, U.S. Environmental Protection Agency, or SCS Engi-
neers .
The following chapters from the State-of-the-Art Review Man-
ual are i ncluded :
•	Title Page, Acknowledgements, Table of Contents and In-
troduction. This material shows what is included in the
entire manual.
#	Part I, Chapter 1: Leaks and Spills of Hazardous
Liquids. Included in this chapter is a discussion re-
garding the generally recognized mechanisms of corrosion.
It should be noted that the primary measure of corrosiv-
ity is the soil resistivity, as evidenced by design stan-
dards and a field test program conducted by the National
Bureau of Standards. [11
A- 1

-------
•	Part I, Chapter 2: Hazardous Substances.
•	Part II, Underground Storage Systems, Chapters 1 to 5
and 7: The design standards which are most commonly fol-
lowed for bare steel tanks are Underwriters Laboratories
(UL) 58 and National Fire Protection Association (NFPA)
30. American Petroleum Institute (API) Publication 1615,
which is generally recognized for installation of under-
ground petroleum storage systems, is also applicable for
systems used to store hazardous wastes. API Publication
1602 is being phased out as a standard for udnerground
gasoline tanks and API Publication 1611 is primarily a
guide for sizing and laying out tankage for service sta-
tions. The primary design and installation standards
applicable for hazardous waste storage systems are brief-
ly described in Appendix C of this report.
•	Appendix B, Compatibility Chart for Fluids, Seals, and
Metals is included as an example for metal tanks, and
should not be interpreted as. a complete presentation.
[1] E. Escalante, "Soils and Underground Corrosion", Chemical
Stability and Corrosion Division, National Bureau of Stan-
dards, Washington, D..C.
A- 2

-------
TECHNOLOGY
FOR THE STORAGE OF
HAZARDOUS LIQUIDS
A State-Of-The-Art Review
NEW YORK STATE
DEPARTMENT OF ENVIRONMENTAL CONSERVATION
DIVISION OF WATER
BUREAU OF WATER RESOURCES
ALBANY, NEW YORK
JANUARY 1983

-------
TABLE OF CONTENTS
Page
Preface 		 i
Acknowledgements	 ii
Table of Contents			iii-vii
List of Figures 	viii-ix
List of Tables 	 x-xi
INTRODUCTION 	 1
A.	Purpose 	 1
B.	Report Overview 		 I
Part I STORAGE OF HAZARDOUS SUBSTANCES 	 2
INTRODUCTION 		 2
CHAPTER 1: LEAKS AND SPILLS OF HAZARDOUS LIQUIDS 4	 2
A.	Behavior of Hazardous Liquids in the Environment 	 2
1.	Background 	 		2
2.	Spill Behavior 		3
3.	The Importance of Spill Prevention 		8
B.	Types and Causes of Spills and Leaks
1.	General		g
2.	Aboveground Storage Systems 		9
3.	Below Ground Storage Systems 		10
C.	Corrosion 				11
.1. Corrosion Mechanisms 				11
2.	Forms of Corrosion 	•.				:	12
3.	Factors Influencing Corrosion 		14
4.	Corrosion Protection 		
References 	i		21-22
CHAPTER 2: HAZARDOUS LIQUIDS 		 23
A.	Listing of Hazardous Liquids 	 23
B.	Properties of Hazardous Liquids 	 23
1.	Chemical Properties 	 23
2.	Relationship Between Temperature, Pressure and Volume Within
a Storage Tank 	 25
C.	Storage and Handling Protocol 	 32
1.	Storage and Handling Systems 			32
2.	Aboveground vs. Underground Storage 		33
3.	Chemical Compatibility 		36
References 		37
Part n UNDERGROUND STORAGE SYSTEMS 		 40
INTRODUCTION 	 40
CHAPTER 1: UNDERGROUND STORAGE TANKS 	 42
A.	Introduction 	 42
B.	Tank Layout 	 42
C.	Types of Underground Storage Systems 	 46
1.	Bare Steel Tanks 		46
2.	Coated Steel Tanks 		50
3.	Cathodically Protected Steel Tanks - Galvanic Protection 		50
4.	Cathodically Protected Steel Tanks - (Impressed Currents) 		52
5.	Fiberglass-Reinforced Plastics 			52
6.	FRP/Steel Bonded Tanks 		54
7.	Tanks of Other Materials 		54
8.	Double Containment Systems 		54
9.	Relined Tanks 		57
iii

-------
TABLE OF CONTENTS
Page
D.	Tank Coatings and Linings 	57
E.	Wrappings 	60
Information on Specifications for Tank Materials and Construction 	60
References 	:	61-62
CHAPTER 2: UNDERGROUND PIPING SYSTEMS 	 63
A.	Introduction 	63
B.	Causes and Methods of Preventing Leaks	63
1.	Proper Design 	63
2.	Piping System Installation 	63
3.	Periodic Testing 	65
4.	Timely Replacement 	65
C.	Types of Piping 	66
D.	Fittings 	66
E.	Expansion Joints and Swing Joints '	66
F.	Underground Pumps 	66
References 	72
CHAPTER 3: UNDERGROUND SPILL CONTAINMENT SYSTEMS 	73
A.	Introduction 		73
1.	Background 	73
2.	Containment Technology 	73
B.	Clay Liners 	74
1.	Chemical and Physical Properties 	74
2.	Design and Installation Requirements 	74
C.	Synthetic Membrane Liners 			74
1.	Chemical and Physical Properties 	74
2.	Design and Installation Requirements 	76
D.	Soil Cement 	'.				76
1.	Soil Cement 			76
2.	Bentonites 	 	79
E.	Concrete Vaults 	80
F.	Double-Walled Tanks 	80
References 	81
CHAPTER 4: TRANSFER SPILL AND OVERFILL PREVENTION SYSTEMS
FOR UNDERGROUND STORAGE TANKS 	82
A.	Introduction 	82
B.	Overfill Prevention Systems for Underground Storage Tanks	82
1.	Elements of an Overfill Prevention System 	82
2.	Specific Level Sensing Devices 	 	83
C.	Transfer Spill Prevention Systems 	86
1.	Check Valves	87
2.	Couplings 	87
D.	Operating Practices for Overfill Protection 	87
References 	90
CHAPTER 5: LEAK AND SPILL MONITORING FOR UNDERGROUND STORAGE ....91
A.	Introduction 	91
B.	Early Warning Leak Detection System 	91
1.	Inventory Control 		91
2.	Interstitial Monitoring in Double-Walled Tanks 	94
3.	Tank Excavation Monitoring Sensors 	94
4.	Tank Excavation Monitoring Systems 	96
C.	Area-Wide Surveillance Methods 	97
1.	Dyes and Tracers	97
2.	Monitoring Wells 	97
D.	Recovery Wells 	103
E.	Examples 	103
References 	107
iv

-------
TABLE OF CONTENTS
Pa
CHAPTER 6: TESTING AND INSPECTION OF
UNDERGROUND STORAGE TANKS 		108
A.	Introduction 		108
B.Tank	Testing Methods 	108
1.	Pneumatic Testing 	Ill
2.	Hydrostatic (Standpipe) Testing 		Ill
3.	Heath Petro-Tite Tank and Line Testing Systems (Kent-Moore Test) 	Ill
4.	The J-Tube Manometer Test	113
5.	The Sunmark Leak Detector 	113
6.	Laser-Beam Leak Detector 	....115
7.	The ARCO Leak Test 	115
C.	Pipeline Testing Procedures 	116
1.	The Suction Piping Test 	 	116
2.	Discharge Line Testing 	116
D.The	Applicability of Tank Tests to the Associated Piping 	117
E.	Physical Inspection 	117
1.	Inspection Prior to Backfilling 	117
2.	Internal Inspection of Installed Tanks 	117
3.	Checking Tank Bottoms 	117
4.	Checking for Water in Tanks 	118
5.	Inspection of Cathodic Protection Systems 	118
References 	119
CHAPTER 7: TEMPORARY CLOSURE, ABANDONMENT AND REMOVAL
OF UNDERGROUND TANKS 	120
A.	Introduction 	120
B.	Temporary Closures 	.	 	120
C.	Permanent Closure 			122
1.	Abandonment in Place 			122
2.	Removal of Tanks 	122
References 	123
PART IH: ABOVEGROUND STORAGE SYSTEMS 	124
INTRODUCTION 			 	124
CHAPTER 1: ABOVEGROUND STORAGE TANKS
A.	Introduction 	126
B.	Tank Layout 	126
C.	Storage Tank Materials 	132
1.	Carbon Steel Tanks 			132
2.	Stainless Steel Tanks 				 134
3.	Fiberglass-Reinforced Plastic 		135
4.	Plastic Tanks 	 	136
5.	Concrete and Aluminum Tanks 	 	136
D.	Coatings, Linings and Cathodic Protection 	138
References 		138-
CHAPTER 2: ABOVEGROUND PIPING SYSTEMS 	 	 .140
A.	Introduction 	140
B.	Causes and Methods of Preventing Leaks 	140
1.	Proper Design and Installaton 	140
2.	Periodic Inspection 	 140
3.	Periodic Testing	 	142
4.	Timely Replacement 	 	142
C.	Piping 	142
1.	Types of Piping Systems 	142
2.	Type's of Piping 	142
3.	Hoses 			145
4.	Pipe Support Elements 	 	145

-------
TABLE OF CONTENTS
Page
D.	Valves 					 145
1.	Functions of Valves 			 145
2.	Types of Valves 	 145
E.	Pumps 			 150
1.	Types of Pumps 	 150
2.	Pump Seals 		 151
3.	Spills and Leaks from Pumps 	 151
F.	Fittings and Joints 	153
References 		153
CHAPTER 3: ABOVEGROUND SPILL CONTAINMENT SYSTEMS 	154
A.	Introduction 				 154
B.	Degrees of Permeability 				154
1.	Levels of Protection 	154
2.	Surface Materials 	159
3.	Choosing a Surface Material 	159
C.	Spill Containment Systems 	160
1.	Dike Systems 	160
2.	Curbs 	'	160
3.	Slurry Trench Cut-Off Walls 	160
4.	Secondary Containment Tanks 	 	166
D.	Spill Collection Systems 	166
E.	On-Site Spill Cleanup 	166
1.	Sorbents	166
2.	Soil Removal 			168
3.	Trenching and Skimming 	 168
4.	Recovery Wells 			170
5.	On-Site Detoxification 	170
6.	Gelling Agents 			170
7.	Biodegradation of Petroleum and Organic Chemical Spills 	171
References 	172
CHAPTER 4: TRANSFER SPILLS AND OVERFILL PREVENTION SYSTEMS
FOR ABOVEGROUND STORAGE 	173
A.	Introduction 	173
B.	Overfill Prevention Systems 	173
1.	Level Sensors and Gauges 	175
2.	High Level Alarms 	181
3.	Automatic Shutdown or Flow Diversion 	181
4.	Emergency Overflow to Adjacent Tank			181
5.	Personal Monitoring of Systems 	Ig2
C.	Dry Disconnect Couplings or Transfer Pipe and Hoses 	183
D.	Redundant Valving and Instrumentation 	183
E.	Use of Established Transfer Stations 	183
F.	Proper Transfer Practices 	183
G.	Regular Inspection and Maintenance 	183
References 		 184
CHAPTER 5: LEAK MONITORING OF ABOVEGROUND TANKS 	185
A.	Introduction 	185
B.	Aboveground Leak Detection Systems 			185
References 	185
CHAPTER 6: INSPECTION AND MAINTENANCE OF ABOVEGROUND TANKS 	186
A.	Introduction 	186
B.	Visual Inspection of Tanks 	186
1.	External Inspection - Tank in Service 	.187
2.	Internal Inspection - Tank Out of Service	188
C.	Visual Inspection of Pipes, Valves, Fittings and Hoses 	188
D.	Inspection of Pumps and Compressors 	 	189
E.	Visual Inspection of Instruments, Control Equipment and Electrical Systems . 189
vi

-------
TABLE OF CONTENTS
Page
F.	Inspection of Vapor Control Systems 		190
G.	Inspection Tools and Electromechanical Equipment 	190
1.	Hammering 			190
2.	Penetrant Dye Method 	191
3.	Vacuum Box 	191
4.	Ultrasonic Instruments 	191
5.	Radiographic Tools 	191
6.	Other Radiation-Type Instruments 	192
7.	Acoustic Emissions Testing 	192
H.	Frequency of Inspections 	192
I.	Maintenance of Aboveground Tanks 	192
References 	193
CHAPTER 7: TEMPORARY CLOSURE, ABANDONMENT AND REMOVAL OF
ABOVEGROUND TANKS 	194
A.	Introduction 			194
B.Tank	Decontamination 	194
C.	Tank Dismantling	196
References 	197
Appendix A: CHEMICAL AND PHYSICAL PROPERTIES OF
VARIOUS HAZARDOUS SUBSTANCES 	198
Various Hazardous Substances 	 198
Definitions 			198
1.	Human Poisons 	 199
Table A-1 	 199
2.	Flammables 	 199
3.	Corrosives 					 199
4.	Volatiles 			:	 199
5.	Floaters 			199
6.	Reactives 	200
7.	Solubles 			200
8.	Solids or Liquids Amenable to Biological Treatment 	 200
9.	Biodegradables 		 200
10.	Compounds Highly Toxic to Aquatic Life 		 200
Gassifications 	200
Table A-2	 201-205
Table A-3.	 205-211
Table A-4	*>12
Appendix B: COMPATIBILITY CHART FOR FLUIDS, SEALS AND METALS 	213-215
Appendix C: PROPERTIES OF PRINCIPAL COATING RESINS 	216-218
References 	216
Appendix D: GLOSSARY 	219-223
vii

-------
LIST OF FIGURES
Number	Title	Page
1.1-1	Product Seepage			4
1.1-2	Trapped Product Droplets	4
1.1-3	Possible Migration of Product to Outcrop Followed by
Second Cycle Contamination	5
1.1-4	Effect of Clay Lens in Soil	5
1.1-5	Typical Behavior in Porous Soil Following a Sudden High Volume Spill	6
1.1-6	Behavior of Product After Spill has Stabilized	7
1.1-7	Electrolytic Corrosion		12
1.1-8	Galvanic Corrosion	13
1.1-9	Corrosion Mechanisms at an Underground Steel Tank	 	15
1.1-10	More Corrosion Mechanisms at an Underground Steel Tank	16
1.1-11	Magnesium Anode Cathodic Protection Typical Configuration		19
1.1-12	Impressed Current Cathodic Protection Typical Configuration	20
1.2-1	Typical Annual Tank Temperature Variation for an Underground
Gasoline Tank	28
1.2-2	Location of Temperature Sensors in the SRI Tank	29
1.2-3	Tank Temperature at Various Heights as a Function
of Time for a 24-Hour Period After Tank Fill-Up		30
1.2-4	Mean Temperature Distribution as a Function of
Depth for Four Different 24-Hour Periods	31
2.1	Elements of an Underground Storage Tank Installation	41
2.1-1	Tank Piping Details - Suction System	45
' 2.1-2	Tank Piping Details - Submerged Systems	45
2.1-3	Anchoring of Tanks Installed in High Groundwater Tables	49
2.1-4	Magnesium Anode Cathodic Protection - Typical Configuration				51
2. i-5	Impressed Current Cathodic Protection - Typical Configuration	 	53
2.1-6	Fiberglass-Reinforced Plastic Double Wall Tank		55
2.1-7	Double Wall Steel Tank with Epoxy Coating and Sacrificial Zinc Anode	56
2.2-1	Diagram of a Universal-Type Expansion Joint		64
2.2-2	Double Walled Pipe	68
2.2-3	Action of the Bellows of an Expansion Joint	69
2.2-4	Swing or Swivel Joints..-	69
2.2-5	Typical Remote Pump Shut-Off Valve	 	70
2.2-6	Typical Remote Pump Shut-Off Valve 	71
2.3-1	Synthetic Liner Installation for Storage of Lighter-Than-Water Liquids
in Area of High Groundwater		79
2.3-2	Synthetic Liner Installation for Storage of Heavier-Than-Water Liquids
in Area of High Groundwater	80
2.4-1	Tape Float Gauge for Underground Storage Tank	86
2.4-2	Float Vent Valves Used for Overfill Prevention	87
2.4-3	Optical Liquid Level Sensing System for Tank Truck	....88
2.4-4	Optical Liquid Level Sensing System for Storage Tank	88
2.4-5	Types of Couplings	89
2.5-1	Typical Applications of a Leak Monitoring System
Based on Thermal Conductivity	95
2.5-2	Examples of Observation Wells	98
2.5-3	Example of a U-Tube Installation	99
2.5-4	Typical Wells for Continuous Gas or Vapor Monitoring	101
2.5-5	Typical Single Monitoring Wet Well	103
2.5-6	Typical Wet Well Cluster	104
2.5-7	Schematic of a Typical Nested Monitoring Well	105
2.5-8	Typical Single-Pump Recovery System	106
viii

-------
LIST OF FIGURES
Number	Title	Page
2.5-9	Typical Two-Pump Recovery System	106
2.6-1	J-Tube Underground Tank Leak Detector	114
3.1	Selected Components of an Aboveground Storage Facility		125
3.1-1 '	Types of Atmospheric and Low-Pressure Tanks	127
3.1-2	Aboveground Tank Connections	130
3.2-1	Types of Valves		 147
3.2-2	Types of Valves	148
3.2-3	Check Valves Used to Prevent Backflow		-..149
3.2-4	Typical Control Valve		 		150
3.2-5	Mechanical Seal Components	151
3.3-1	Schematic Flow Diagram	155
3.3-2	Bulk Plant Layout	161
3.3-3	Dike and Siphon	162
3.3-4	Typical Earth Dikes	163
3.3-5	Typical Curbed Containment Area Drained Through Catch Basin	164
3.3-6	Underground Barrier and Cut-Off Wall	 	165
3.3-7	Imbiber Bead Applications	 	167
3.3-8	Spill Cleanup Using Interceptor Trench	169
3.3-9	Spill Cleanup Using Pumping Well	169
3.4-1	Elements of an Overfill Prevention System	.-	174
3.4-2	Chain and Tape Float Guage Used for Level Control	177
3.4-3	Level and Shaft Float Used for Level Control		.177
3.4-4	Magnetically Coupled Floats Used for Level Control	178
3.4-5	Torque Tube Displacer Used for Level Control	179
3.4-6	'Magnetically Coupled Displacer Used for Level Control	179
3.4-7	Flexure Tube Displacer Used for LeVel Control			 180
3.4-8	Bubble Tube System Used for Leyel Control		180
3.4-9	Loading Arm Equipment with Automatic Shutoff	182
ix

-------
LIST OF TABLES
Number	Title	Page
1.1-1	Types of Leaks from Bulk Storage and Handling Facilities 	8-9
1.1-2	Guide to Discussions of Causes and Mitigative Measures for Spills or Leaks
from Hazardous Liquid Storage Systems	9
1.1-3	Leaks by Source Categories 	10
1.1-4	Causes of Leaks in Steel Tanks 	10
1.1-5	Causes of Piping Leaks 	11
1.1-6	The Galvanic Series of Metals and Alloys 	..14
1.1-7	Common Forms of Localized Corrosion 		14
1.1-8	Soil Corrosivity vs. Soil Resistivity 	17
1.2-1	Listing of Regulated Hazardous Substances 	23
1.2-2	Reports Describing Hazardous Materials 	24
1.2-3	Physical and Chemical Properties of the Twenty-Nine Most Commonly Used
Hazardous Substances in New York State 	26-27
1.2-4	Total Force on Tank Ends Due to Internal Pressure 	 	32
1.2-3	Storage System Components - Methods, Materials and Design Considerations 	33
1.2-6	General Properties of Materials Used for Storage Tanks and Piping 	34-35
1.2-7	Aboveground vs. Underground Storage 	35
1.2-8	Compatiblity Chart: Chemicals vs. Structural Materials 	36
1.2-9	References on Materials and Chemical Compatibility 		36
2.1-1	Characteristics of Underground Storage Tanks 	 43-44
2.1-2	Recommended Thickness of Steel Tanks 	 47
2.1-3	Gallon Capacity per Foot of Length 		48
2.1 -4	Surface Preparation Specifications 	58
2.1-5	Latest Coating Techniques 	59
2.2-1	Important Criteria in the Design of Piping Materials for Underground Service 	 63
2.2-2	Characteristics of Piping Materials for Underground Service 	 67
2.3-1	Comparison of Underground Spill Containment Systems	 75
2.3-2	Comparison of Various Synthetic Polymeric Membranes 	 	 77-78
2.3-3	Chemical Compatibility of Membrane Liners with Hazardous Materials 	 78
2.3-4	Considerations During Liner Placement 		 76
2.3-5	Highlights of Soil Cement Design and Installation 	'.	 79
2.4-1	Elements of a Good Overfill Prevention System 	 82
2.4-2	Characteristics of Pneumatic and Electronic Controls 	 84
2.4-3	Level Detection Devices for Underground Storage Tanks 	 85
2.4-4	Transfer Spill Prevention Systems 		 88
2.5-1	Comparison of Various Leak Monitoring Techniques 	 92-93
2.5-2	Applicability of Types of Leak Sensors in Tank Excavation Areas	 96
2.5-3	Types of Site Data Needed to Design Appropriate Groundwater Monitoring Programs ... 100
2.5-4	Types of Groundwater Monitoring Wells 	 102
2.5-5	Leak Detection System for Manufacturers 	 104
2.6-1	Comparison of Various Leak Detection Tests for Underground Systems 	 109-110
2.6-2	Apparent Loss of Product Volume Due to Tank End Deflection (in Gallons) 	 112
2.7-1	Closure, Abandonment and Removal of Underground Tanks 	 121
3.1-1	Characteristics of Various Types of Aboveground Storage Tanks 	 128-129
3.1-2	Current Publications of the American Petroleum Institute Pertaining to Storage Tanks ... 133
3.1-3	Underwriters Laboratories Recommendations for Metal Thickness of
Horizontal Steel Tanks 	 134
3.1-4 Underwriters Laboratories Recommendations for
Metal Thickness of Vertical Steel Tanks 	 134
3.1-5 Miscellaneous Recommendations for Steel Tank Fabrication 	135
3.1-6 Liquid Chemicals Commonly Stored in Fiberglass Tanks 	135
3.1-7 Performance Factors for Various Types of Fiberglass Tanks 	 	136
3.1-8	Partial Chemical Compatibility Chart for One Type of Polyolefin Tank	137
3.2-1	Piping Installation and Leak Prevention 	 141
x

-------
LIST OF TABLES
Number	Title	Page
3.2-2	Characteristics of Piping Materials 	143-144
3.2-3	The Functions of Valves 	146
3.2-4	Items to Look for During Valve Maintenance	 	150
3.2-5	Comparison of Pump Seals 	152
3.3-1	Spill Containment and Collection Systems 	156
3.3-2	Characteristics of Various Surface Materials 	157-158
3.3-3	Materials Absorbed by Imbiber Beads	168
3.4-1	Transfer Spills Are Prevented by Using the Following Equipment and Practices	173
3.4-2	Level Detection for Overfill Prevention Systems for Aboveground Storage Tanks 	176
3.6-1	Tank Inspection Point Listing 		 186-187
3.7-1	Closure, Abandonment and Removal of Underground Tanks 	195
3.7-2	References for Tank Cleaning Operations	196
xi

-------
INTRODUCTION
A. PURPOSE
Within the past five to ten years there have been
major advances in the technology and practices of stor-
ing and handling hazardous liquids. New tank designs
and tank materials have been applied to solve problems
of corrosion and to prevent leaks. Mechanical and elec-
tronic flow control and level detection devices have
been invented to prevent transfer spills and monitor
storage volumes. Laser technology, capable of measur-
ing the loss of product stored in a tank with a resolution
of 0.000001 inches, has been used experimentally to
test for tank leaks. Secondary containment designs have
been developed and applied by many sectors of the in-
dustry.
It is the purpose of this report to evaluate these and
other aboveground and underground storage practices. It
is a report on the state-of-the-art for the following:
•Tanks for storing hazardous liquids.
•Secondary containment systems.
•Piping and safety valves.
•Overfill prevention systems and practices.
•Inspection, testing and monitoring.
•Closure and abandonment practices.
Hopefully, this report will provide timely informa-
tion for the industry and government officials faced with
problems on the storage of hazardous liquids and will
encourage the use of the best technology and practices
for preventing spills and leaks.
B. REPORT OVERVIEW
This report is divided into three parts. Part I repre-
sents an overview of the general concerns associated
with the underground or aboveground storage of hazard-
ous liquids. This part of die report includes discussion
of:
•The properties and characteristics of various haz-
ardous liquids.
•The compatability of various tank and piping sys-
tem materials with certain hazardous liquids.
•The types of leaks and problems which can occur
in bulk storage facilities.
•The cause of corrosion.
•Other technical factors that must be considered
prior to the storing of hazardous liquids.
Part II of this report addresses the state-of-the-art
for underground storage systems. Part III addresses
these same items for aboveground storage systems.
Some of the material discussed applies to both
aboveground and underground systems, therefore, cross
referencing is employed throughout this report. More
detailed references, such as those prepared by the
American Petroleum Institute, the National Fire Protec-
tion Association, and other institutions, are identified in
the text. The reader is specifically directed to these
references for futher information on the technology and
measures for spill and leak prevention.
Because of the large quantity of gasoline and other
petroleum products handled and stored in the state and
the extensive information on the storage of these mate-
rials available from the American Petroleum Institute
and other organizations, much of the detailed material
presented in this manual is drawn from the experience
of the petroleum industry. Although the basic principles
illustrated are applicable to the storage of all hazardous
liquids, some of the specific details presented may not
be directly applicable in all situations. There are many
ways by which environmentally acceptable storage facil-
ities can be achieved. The manual is intended to serve
as a source of background information and guidance to
aid government officials, designers and users of bulk
liquid storage systems in understanding the many differ-
ent considerations and features which may impinge
upon design and installation of such systems. It is not
intended as a standard or as a substitute for sound en-
gineering practice as applied to the design and installa-
tion of bulk storage systems for specific materials at
specific locations.
1

-------
Part I
STORAGE OF
HAZARDOUS SUBSTANCES
INTRODUCTION
The purpose of Part I of this report is to present
background information describing the characteristics of
hazardous substances, the types and causes of leaks and
spills, and the behavior of these leaks and spills in the
environment. Chapter 1 of this part of the report addres-
ses the types and causes of leaks from both aboveg-
round and underground storage systems, the sources of
leaks and spills and methods available to control them.
This chapter also includes a description of the behavior
of hazardous substances when they are spilled on or in
the ground.
Chapter 2 of this part of the report presents data
on the types of hazardous materials of concern and their
properties. This chapter refers to general listings of haz-
ardous substances, and provides detailed information re-
garding such items as type of hazard, specific gravity,
boiling point, melting point, solubility in water, etc. for
a select group of these substances. This chapter also
provides information regarding the compatibility of vari-
ous types of hazardous substances with the metals or
other materials which may be used to construct storage
system components (tanks, pipes, fittings, etc.).
Part I
CHAPTER Is
LEAKS AND SPILLS OF
HAZARDOUS LIQUIDS
A. BEHAVIOR OF HAZARDOUS LIQUIDS
IN THE ENVIRONMENT
1. Background [1,2,3,4]
Spills (including leaks) of hazardous liquids can
have substantial environmental, public health and social
impacts. Such spills can result in the contamination of
soils, surface water and groundwater supplies and air,
all of which can directly affect crops, wildlife, plants
and ultimately humans. Many hazardous, liquids are
conservative substances (i.e., they do not biodegrade or
decompose), therefore, once the substance has been
spilled it will remain a hazard unless it can be removed
to below the level of harmful concentration.
A hazardous substance which finds its way into the
environment may be a serious threat to the health of
people who come in contact with it. It may contaminate
water supplies, crops and food supplies, fisheries or
wildlife habitat. The non-environmental impacts which
result from spills of hazardous liquids can also be far-
reaching. These impacts can include the following:
•The dislocation of people.
•The loss of valuable product.
•The loss of property resulting from contamination,
fire, explosions, etc.
•The economic and social costs of spill cleanup
[!].
Storage tank leakage problems are more readily
controlled and resolved with aboveground structures be-
cause the leaks are more likely to be visible. Below
ground systems present potentially more serious prob-
lems of contamination because of the likelihood of un-
detected leakage, but minimize the possibility of fires
and explosions when flammable and reactive chemicals
are to be stored. The threat of potential groundwater
pollution must be balanced against other safety consid-
erations. In some parts of New York State, groundwater
is the only source of fresh water, and areas such as
Long Island have already witnessed a large number of
groundwater pollution incidents. A prime example oc-
curred in East Meadow, Long Island where a service
station leaked 50,000 gallons of gasoline from a below
ground storage tank and the hazardous fumes seeped
into the basements of more than a score of the sur-
rounding homes. The gasoline distributor purchased the
homes and they are still uninhabitable [1"].
A contributing factor in the increasing number of
documented incidents like this is the large number of
active and abandoned underground storage tanks. New
York State has more than 100,000 aboveground and
buried bulk storage tanks containing a variety of chemi-
cals (cleaning solvents, pesticides, industrial process
chemicals, etc.). However, most contain petroleum,
primarily gasoline. For these alone, the New york State
Gasoline Retailers Association estimates that at least
68,000 are underground at gasoline service stations.
Roughly 24,000 of these tanks are at abandoned service
stations that went out of business during the recent
period of gasoline shortages. Although in disuse, many
are suspected of containing a residual gasoline supply.
In addition to the gasoline retailers, thousands of stor-
age tanks across the State are used at motor depots,
contracting yards, farms, schools, industrial sites and at
some private homes [2].
Many tanks were installed in the early 1950's when
growth in the chemical industry and highway transporta-
tion was booming. These tanks are now 20-30 years old
- at or beyond their life expectancy. Other contributing
causes to the problem include improper material selec-
tion during the design stage and, just as important, im-
proper installation practices. The percent of failure is
not known but it is estimated that 10 to 20 percent are
leaking [2].
2

-------
The effects of underground leakage and spills can
be both short-and long-term, the short-term effects from
gasoline spills, for example, as well as other hydrocar-
bon type materials, are potentially devastating because
of their volatile nature. Seepage of liquids and fumes,
into underground structures can result in gas and vapor
accumulation and consequent explosion and health haz-
ards. In the long-term, contaminated underground water
supplies (aquifers) are practically impossible to reclaim
once they have been contaminated. To understand the
potential hazard posed by spills and leaks of hazardous
liquids, one must develop some understanding of the
behavior of such spills and leaks.
2. Spill Behavior
Spills and leaks of volatile hazardous liquids may
pose potential air quality and explosion problems. The
level of potential hazard is dependent upoir several fac-
tors including the volatility of the spilled substance and
vapor dispersion characteristics in the vicinity of the
spill or leak. Surface waters may also be contaminated
by spills and leaks traveling across or within soils pos-
ing a potential threat to aquatic life and human health.
However, of primary concern in New York State
is the potential effect that spills or leaks may have upon
soils, surface water and groundwater [1]. Hazardous liq-
uids spilled or leaked into soil typically tend to flow
downward, with some lateral spreading due to gravita-
tional forces, as illustrated in Figure 1.1-1. The rate of
movement in the soil will depend on product properties
such as solubility, miscibility, viscosity, soil permeabil-
ity and compaction, and the rate or volume of the leak
or spill. For example, given the same soil properties,
lighter liquids, such as gasoline, will penetrate the soil
rapidly, while heavier, more viscous liquids will move
more slowly. Alternately, if the soil'has a low permea-
bility, as is characteristic of clays, the product may
have little or no penetration. However, if the soil is
very porous, the product will penetrate it quickly.
Absorption by Soil [3,4). As spilled liquids pass
downward through soils, individual soil particles will be
coated with a thin film of that liquid. In addition, sur-
face tension will act to hold small amounts of that liq-
uid in the voids between soil particles, as shown in fig-
ure 1.1-2. These actions combine to result in the ab-
sorption of the liquid into the soil. Once absorbed, ex-
traction of the liquid is virtually impossible.
A spilled or leaked liquid will move downward
in the soil until:
•It is absorbed by the soil.
•It encounters an impermeable bed or layer.
•It reaches the water table.
•It seeps from groundwater to surface water.
Movement at an Impermeable Layer [3,4]. The
downward movement of a spilled liquid through soil is
affected by variations in permeability of the soil layers
through which it passes. If the flowing liquid encounters
an impermeable layer of soil it will spread laterally until
either it becomes immobile or it comes to the surface
at the outcrop of the impermeable layer. Should the lat-
ter phenomenon occur, a second cycle of soil contami-
nation could begin (Figure 1.1-3).
Note that Figures 1.1-3 through 1.1-6 illustrate the
movement of product which is lighter than water and
immiscible. A spilled chemical with a specific gravity
greater than 1.0 will tend to sink to the bottom of the
aquifer while one with a high solubility will tend to mix
with the groundwater.
Downward movement of spilled material may also
be complicated by the presence of thin lenses of mate-
rial with low permeability (Figure 1.1-4). If such lenses
are present, the fluid path could be substantially altered.
Movement Into Groundwater [3]. The following
excerpt from API Publication 1628 describes the intru-
sion of spilled liquids into groundwater.
"The contact of spilled product with the
water table usually is the most troublesome re-
sult of an on-land spill. This condition greatly
increases the risk of polluting a water supply,
and may increase the chance of movement to
some underground structure, such as a base-
ment, sewer or conduit. The degree of risk de-
fends on the nature of the groundwater system
and the way it is utilized.
Figure 1.15 illustrates a pattern of oil des-
cent to a water table. A sudden, large-volume
spill will depress the water table and spread in
all directions in a layer above the water table.
As the layer becomes thinner, it will begin to
move in the direction of groundwater flow
(Figure 1.1-6).
A slower leak will descend in a narrow cone
and spread in the direction of water movement.
Lateral spreading will usually be slower than
the flow rate of the groundwater. "[3]
More detail discussions of this type can be found in API
Publication 1628 [3], and NFPA 329 [4].
3

-------
Figure 1.1-1
Product Seepage
LAND SURFACE
SEEPAGE INTO
STRATIFIED SOIL
WITH VARYING
PERMEABILITY
SLOW SEEPAGE
INTO
PERMEABLE
SOIL
HIGH VOLUME
SEEPAGE
INTO
PERMEABLE
SOIL
Source: API Publication 1628, Underground Spill Cleanup Manual, 1980.
Figure 1.1-2
Trapped Product Droplets
Source: API Publication 1628, Underground Spill Cleanup Manual, 1980.
4

-------
Figure 1.1-3
Possible Migration of Product to Outcrop
Followed by Second Cvcie Contamination
AREA OF SPILL
PERMEABLE SOIL
IMPERMEABLE CLAY
Source: API Publication 1628, Underground Spill Cleanup Manual, 1980.
Figure 1.1-4
Effect of Clay Lens in Soil
SPILL
SITE
LAND SURFACE
UNSATURATED
ZONE
LAYER OF SATURATION
SEEPAGE PATH
CLAY LENS
CLAY LENS
CLAY LENS
SATURATED ZONE^^
IMPERMEABLE BEDS
Source: API Publication 1628. Underground Spill Cleanup Manual. 1980.
5

-------
Figure 1.1-5
o
Typical liehavior in Porous Soil
Following a Sudden, High Volume Spill
Source: API Publication 1628, Underground Spill Cleanup Manual, 1980.
¦V	. '
K *>v: '•
•;.v -«.''• ,
' POROUS SOIL
SPILL SITE
GROUND SURFACE
SOIL CONTAMINATED
BY RESIDUAL PRODUCT
PRODUCT MIGRATING DOWNWARD
AND ACCUMULATING ON WATER TABLE -
' •r'.i ¦

-------
Figure 1.1-6
Behavior of Product After Spill Has Stabilized
—J
Source: API Publication 1628, Underground Spill Cleanup Manual, 1980.
SPILL SITE
GROUND SURFACE

-------
Movement from Groundwater to Surface Water.
Contaminated groundwater may enter surface water
through springs or by direct influent seepage into a
creek, lake or river. It may now become visible or be
detected by odor. The presence of the contaminant may
also be apparent if there is a fish kill. It will cause im-
mediate concern if the surface water is used for drinking
water or for primary contact recreation (bathing).
3. The Importance of Spill Prevention
The first line of defense against the potential en-
vironmental and public health impacts of hazardous liq-
uids spills and leaks is the implementation of good spill/
leak prevention practices. Such prevention practices are
always more cost-effective and environmentally effec-
tive than attempts to clean up a spill or leak after it has
occurred. For this reason, the importance of adopting
and rigidly following a spill/leak prevention and detec-
tion program cannot be overemphasized.
The practices which can be employed for spill and
leak prevention are discussed in detail throughout the
remainder of this document.
B. TYPES AND CAUSES OF SPILLS
AND LEAKS [1,5,6,7]
1'. General
Spills and leaks of hazardous liquids at bulk stor-
age facilities, either above or below grade, may ema-
nate from any of several sources and may be precipi-
tated by one or more of several causes. The types of
spills which may occur include: (1) large spills such as
those that result from tank or pipe rupture; (2) slow
leaks or drips such as those that result from slow de-
terioration (e.g. corrosion) of a storage system compo-
nent; and (3) small spills such as those that result from
fluid transfer mishaps (e.g. overfills) or other storage
yard spills. Spills or leaks can also occur as a general
result of poor housekeeping practices, or as a result of
vandalism or acts of malicious intent. Spills or leaks
can occur anywhere in the bulk storage facility where
liquids are handled or stored if proper care is not taken.
Examples of leak locations from bulk storage and han-
dling facilities are presented in Table 1.1-1.
An adequate spill/leak prevention and detection
program has a number of key elements, including the
following:
•Tank (material) selection guidelines.
•Tank installation guidelines.
•Piping system (material) selection and installation
guidelines.
•Steps for corrosion and tank failure prevention.
•An overfill prevention system and overfill protec-
tion guidelines.
•Standardized practices for periodic inspection and
preventive maintenance.
•A leak detection system with periodic monitoring.
•Procedures for inventory control.
•Spill containment facilities.
•Emergency response procedures.
•Guidelines and procedures for the closure and
abandonment of storage systems.
•Transfer facility design requirements.
Table 1.1-1
Types of leaks from Bulk Storage and
Handling Facilities
Bulk Storage Facilities - Tank Farms and Tankage
1.	Leaks and overfilling of tanks
2.	Rupture of tanks
3: Leaks in pipe, valves and fittings
4.	Leaks in containment dikes
5.	Inadequate dike volume to hold contents of
leaking tanks
6.	Product flow from dike area through open
dike valve
7.	Leaks from pump seals and maintenance
8.	Level instrument failure allowing tank
overfilling
9.	Piping damage by collision with mobile ^
equipment
10.	Spills from water drawoff from
storage tanks
11.	Spills from tank bottom cleanout and sludge
disposal
12.	Improper disposal of samples
13.	Overflow of wastewater treatment systems
by rainfall flooding
14.	Poor maintenance of pipe, valves
and fittings
15.	Plugging of drainage system by debris
16.	Wastewater treatment systems with insuffi-
cent capacity to remove product
17.	Inadequate secondary containment devices
18.	Spills from line flushing
19.	Spills from pipe and tankage changes
20.	Possible sabotage
21.	I mproper installation
Bulk Handling Facilities - Terminals, Pipelines
1.	Spills from quick-connect coupling
operation
2.	Overfilling tank trucks, tank cars, barges,
tankers, etc.
3.	Lack of curbs, drains and spill collection
system
4.	Improper operation of product/water
separators
8

-------
Tablt 1.1*1 continued
5.	Leaks from loading arms, especially joints
and gaskets
6.	Leaks from underground storage tanks
7.	Improper disposal of sludge from
product filters
8.	Insufficent sump capacity (should be equal
to volume of largest compartment of tank
truck of rail car)
9.	Leaks from damaged loading connections
10.	Operators incorrectly setting loading meters
and tanks overfilling
11.	Level instrument and subsequent sump
pump failure on oil sumps
12.	Leaks from heating coils in heavy fuel tanks
13.	Possible sabotage
14.	Improper installation
Adapted from reference 5.
These causes of storage system leaks and spills and
state-of-the-art methods of controlling them are ad-
dressed briefly in the discussions of aboveground and
underground storage systems that follow this section. A
guide (to the more detailed discussions of these items
throughout Parts II and III of this document is presented
in Table 1.1-2.
use of secondary containment, curbing, pumps, etc.;
and the incorporation of adequate housekeeping and op-
erational procedures.
One of the most important causes of leaks is the
improper installation of storage tanks and related equip-
ment. Many leaks can be traced to problems such as:
Table 1.1-2


Guide to Dbcuskm of Causes ai
nd Miticative Measures for

Spill and Leaks from Hazardous Liquid Storage Systems

Cause of Leak Spill
Equipment Affected
Section oi Report
Equipment deterioration


Corrosion
All components
Pan 1. Chapter 2.

I nderground tanks
Pan II. Chapter 1.

U nderground piping
Part II. Chapter 2

Aboveground tanks
Part {II. Chapter 1

Aboveground piping
Pan III. Chapter 2
Mechanical Failure and crack*
Underground Tanks
Pan U. Chapter 1

Underground piping
Pan il. Chapter 3

Aboveground tank*
Pan til. Chapter 1
Pan III. Chapter 2

Aboveground piping
Transfer spill* and leaks
Underground tanks.
Pan II. Chapters ?

piping and spill
and 4

containment systems


Aboveground tanks.
Pan 111. Chapters 3

piping und spill
and 4

containment systems

Improper installation ol
Underground tanks
Pan II. Chapter 1
system component*



Underground piping
Pan II. Chapter 2

Aboveground piping
Pan III. Chapter 2
Poor Housekeeping
Underground storage systems
Partll. Chapters 5
and 6

Aboveground storage systems
Pan III. Chapter 5
Improper Temporary or
U nderground storage systems
Pan 11. Chapter "
permanent closure



Aboveground storage systems
Pan IH. Chapter?
2. Aboveground Storage Systems [1,5,8,9]
The deterioration of the components of aboveg-
round storage systems can occur for any one of several
reasons. The most common reason for component de-
terioration, particularly the deterioration of metal com-
ponents, is corrosion [1], which is addressed in detail
later in this section. Other reasons include the follow-
ing:
•Mechanical failure, such as failure of valves, gas-
kets or pumps.
•Cracks in tanks, piping or fixtures which could re-
sult from faulty welding, unrelieved stress concen-
trations around fittings, insufficient reinforcement
around openings, settlement or earth movement,
vibration, or poorly designed repairs [8].
Methods of controlling the deterioration of storage
system equipment include: (I) the use of better system
designs; (2) the incorporation of a good preventive
maintenance program; and (3) proper training of em-
ployees.
Leaks and spills due to overflow, overfilling and
other liquid transfer operations are another important
category of product loss from storage systems. These
can be controlled through the use of overflow protection
devices and level sensing devices in storage tanks; the
•Damage to tank coatings.
•Outright structural damage to the tank and other
equipment during transportation and installation.
•The more subtle damage associated with the im-
proper installation of beddings and foundations for
tanks and piping systems.
•The improper connection of system components,
such as the improper installation of valves, flanges
or other fittings.
•Overpressurization caused by overfilling or impro-
per venting of tanks.
Problems such as these can be avoided through
careful adherence to the design and installation require-
ments of storage system components.
Sloppy housekeeping also results in spills and
leaks. Such accidents can be avoided through the im-
plementation of good housekeeping practicies. A clean
and orderly work area reduces the possibility of acci-
dental spills caused by mishandling of equipment and
should reduce safety hazards to plant personnel. Exam-
ples of good housekeeping include neat and orderly
storage of chemicals; prompt removal of small spillage;
regular garbage and rubbish pickup and disposal;
maintenance of dry and clean floors by use of brooms,
vacuum cleaners, or cleaning machines; and provisions
for storage of containers or drums to keep them from
protruding into open walkways or pathways.
9

-------
A good security system is helpful in preventing
hazardous chemical spills or leaks due to vandalism,
theft, sabotage or other improper and illegal use of stor-
age plant facilities. The elements of such a system
could include the following:
•Routine patrols of the plant by security personnel.
•Fencing.
•Good lighting.
•Vehicular traffic control.
•Controlled access to the plant.
•Locked entrances.
•Locks on drain valves and pumps for chemical
storage tanks.
•Television monitoring.
3. Below Ground Storage Systems [1,5,10]
Underground storage systems are susceptible to
leaks and spills from the same types of causes as
aboveground storage systems, and, in general, the same
methods of spill and leak control are applicable. In the
case of thse systems however, data indicates that corro-
sion and poor installation are by far the most common
causes of storage system leaks and spills [1,10]. For ex-
ample, the American Petroleum Institute (API) con-
ducted a survey of 1,717 underground tanks and piping
systems that were known to be leaking. The data was
collected via questionnaire from 1977 through 1980. A
categorization of the reported leaks is displayed in Table
1.1-3. Since no data base exists in this study concerning
the number or age of the various types of tanks in the
ground at the time of this survey or the average ages
of each type of tank, the use of the study for comparing
types of tanks is meaningless. Much Valuable data is
contained in the study but any attempt to compare tank
types would be a misuse of the data.
The life expectancy of any given tank is difficult
to predict. Experience has shown that underground steel
tanks have a finite life, but this life is variable between
5 and 45 years depending on the thickness of the steel,
installation practices, soil resistivity, pH, soil moisture
level, the presence of sulfides, the type of backfill ma-
terial used, and the tank size. The average life expec-
tancy of these tanks is about IS years, but age by itself
is a poor indicator of tank integrity.
The causes of leaks in steel tanks, as determined
by the API Leak Survey, are shown in Table 1.1-4.
Overall, roughly 91 percent of the leaks in steel tanks
were caused by corrosion. Other causes included loose
fittings and physical breakage. Of the 28 leaking
fiberglass tanks included in, the survey, 9 had dip stick
punctures, 4 had breakage from improper handling, 1
had a backhoe puncture, and 14 had experienced phys-
ical breakage or separation due to other causes. For pip-
ing systems, corrosion was also the most common cause
of leaks as shown in Table 1.1-5.
Table 1.1-3


Leaks by Source Categories


Source
Number
Percent
of total
Unprotected Steel Tanks
913
62
Steel tanks with
Impressed Current
13
0.9
Steel tanks with
Sacrificial Anodes
0
0.0
Interior Coated
Steel Tanks
7
0.5
Fiberglass tanks
28
1.9"
Steel piping
454
30.8
Fiberglass piping
50
3.4
Steel piping with
Impressed Current
7
0.5
Sub-Total
1472
100
Unspecified Tanks
216

Unspecified Piping
29

Total
1717

Source Reference 10.


Table 1.1-4


Causes of Leaks in Steel Tanks




Percent
Cause
Total
of Total
Corrosion
775
90.7
Loose Fittings
10
1.2
Physical Breakage
14
1.6
Other
55 ¦
6.4
Sub-Total
854
99.9
Unknown or unanswered
59

10

-------
Table 1.1-5


Causes of Piping Leaks


Cause
Number
Percent
of Total
Corrosion
343
66.6
Flex Connector Failure
31
6
Physical Breakage
34
6.6
Loose Fittings
57
11.1
Other
50
9.7
Sub-Total
515
100
Unknown or Unanswered
25

Note: These tables emphasize the importance of
corrosion as a cause of storage system leaks.
Source: Reference 10.


Underground tanks which are connected together
with siphoning pipes present unique problems. Leak
testing becomes difficult, if not impossible, to accom-
plish. the ususal reason for siphoning between tanks is
tp add capacity to a system. When a small tank does
not provide enough gallonage for increased business
(usually after several years), a second tank is installed
and is connected to the first with a siphon. The new
tank and the new piping become targets for electrolytic
corrosion from the old tank.
C. CORROSION [8,11,12,13,14,15,16,17,18,19,20]
The corrosion of tanks and piping systems is a
complex phenomenon that may take one or more of sev-
eral forms. Corrosion results from interactions betweeen
the tanks and piping and their surroundings, both inter-
nal and external.
The deterioration of plastics and other non-metallic
materials, which are susceptible to softening, cracking,
swelling, etc., is essentially chemical in nature [11],
Non-metallic materials may deteriorate rapidly when ex-
posed to corrosive elements. The corrosion of non-
metallic storage system components can be controlled
and essentially eliminated through proper selection and
careful handling of tank and piping materials.
In metallic materials, corrosion is a chemical or
electrochemical process. Corrosion control in these ma-
terials is therefore more complicated. The remainder of
the discussion in this chapter focuses on the causes of
internal and external corrosion of metals, the factors
which influence this corrosion, and the steps which can
be taken to protect against this form of deterioration.
1. Corrosion Mechanisms
As stated above, the corrosion of metals is primar-
ily an electrical process; it may take the form of either
galvanic or electrolytic corrosion. As shown in Figure
1.1-7, electrolytic corrosion is a result of direct current
from outside sources entering and then leaving a par-
ticular metal structure by way of the electrolyte (sur-
rounding material, such as soil for underground struc-
tures or water for submerged structures). The structure
is usually unaffected or is provided with some degree
of protection at the point the current enters (the cathodic
area). Corrosion occurs where the current leaves the
structure (the anodic area). In underground structures,
this type of corrosion is often referred to as stray cur-
rent corrosion, and is a result, of current entering the
ground from sources of DC current such as street rail-
ways or DC machinery.
The mechanisms of galvanic corrosion are illus-
trated in Figure 1.1-8. Galvanic corrosion is a self-gen-
erated activity resulting from differences in electrical
potential that develop when metal is placed in an elec-
trolyte. These differences in electrical potential can re-
sult from the direct coupling of dissimilar metals, or
they can result from variations in conditions which exist
upon the surface of a single metal. The variations could
include:
•Variations resulting from non-homogeneity of the
metal.
•Differences which exist within the electrolytes.
When two dissimilar metals' are connected electri-
cally and immersed in an electrolyte, as shown in Fig-
ure 1.1-8, current will be generated and galvanic corro-
sion will occur in one of the metals. Current from the
corroding metal will flow into the electrolyte, over to
the non-corroding metal, and then back through the con-
nection between the two metals. The corroding metal
(the one from which current leaves to enter the electro-
lyte) is known as the anode; the metal which receives
current is known as the cathode. Table 1.1-6 shows the
anodic-cathodic (galvanic) series of various metals.
Alternately, as stated previously, the same metal
can develop differences in potential, and, as a result,
portions of the surface of that metal become anodic with
respect to the remainder of the surface. As shown in
figure 1.1-8, corrosion will occur at these anodic loca-
tions.
Electrolytic and galvanic corrosion are similar in
that corrosion always occurs at the anodes. The essen-
tial difference between the two is that in electrolytic
corrosion it is the external current which generates the
corrosion, whereas in galvanic corrosion it is the corro-
sion activity which generates the current.
11

-------
2. Forms of Corrosion [8,11]
The deterioration of tank or piping material may
appear as either general or localized corrosion. General
corrosion appears as a relatively uniform loss of surface
material if viewed without magnification. Localized cor-
rosion results in a non-uniform loss of material from the
corroded structure. Types of localized corrosion are de-
scribed briefly in Table 1.1-7.
Figure 1.1-7
Electrolytic Corrosion
Z' SHi
SRAOE


CURRENT FLOW ALONG PIPELINE
STRAY CttBRENT ON UH0ER6R0UN0 PIPELINE
i+'I
-=- O.C. SOURCE
(-)
CURRENT FLOW ALONG PIPELINE
STRAY CURRENT ON UNOERGROUNO PIPELINE SHOWING SOURCE & LOAD
Source: Harco Corporation
12

-------
Figure 1.1-8
Galvanic Corrosion
CURREHT
THRU CONNECTION
GALVANIC CORROSION • DISSIMILAR METALS
GRADE
h <+)
(-) (-)
(+) N

f .
CATHOOIC
AN001C
CATHOOIC
GALVANIC CORROSION • DISSIMILAR ELECTROLYTES
Source: Harco Corporation
13

-------
Table 1.1-6
The Galvanic Series of Metals and Alloys
Corroded End (Anodic, or Least Noble)
Magnesium
Zinc
Galvanized steel or galvanized wrought iron
Aluminum
Cadmium
Mild Steel
Wrought iron
Cast iron
13 percent Chromium stainless
18-8 stainless type 304
Lead
Tin
Naval brass
Nickel (active)
Inconel (active)
Yellow brass
Aluminum Bronze
Red brass
Copper
Silicon bronze
Nickel (passive)
18-8-3 stainless type 316
Silver
Graphite
Gold
Platinum
Protected end (Cathodic, or Most Noble)
Note: In general, when dissimilar metals are
used in contact with each other in an elec-
trically conductive enviroment, combina-
tions of metals should be chosen that are
as close as possible in the galvanic series.
The coupling of two metals which are far
apart in the series will result in more
rapid deterioration of the more active
metal. However, this table should be used
only as a general guide since exceptions
to this series may be encountered.
Adapted from reference 11.
3. Factors Influencing Corrosion [8,11]
There are innumerable factors that can influence
the presence and rates of internal and external corrosion
in both aboveground and underground storage tanks and
piping systems. The more prominent of these factors in-
clude solution acidity, temperature, moisture levels,
oxygen levels, soil resistivity, and bacterial action. The
following discussion explains the importance of these
and other factors. The more important of these factors
are highlighted in figures 1.1-9 and 1.1-10.
Electrolyte Acidity. The acidity of the electrolyte
(solution, soil, etc.) with which the material is in con-
tact could have a substantial affect on the rate of corro-
sion. Acidic (low pH) electrolytes are, as a general rule,
more corrosive than neutral (pH 7) or alkaline (high pH)
electrolytes in the case of ordinary iron and steel. How-
ever, for the amphoteric metals, such as aluminum and
zinc, highly alkaline electrolytes may be more corrosive
than acidic electrolytes. The effects of electrolytic acid-
ity are highlighted in Figure 1.1-9.
Presence of Oxidizing Agents. The presence of
oxidizing agents, of which oxygen is the most promi-
nent, may accelerate the corrosion of one type of mate-
rial and retard corrosion in another type.
Table 1.1-7

Common Forms of Localized Corrosion
Type
Description
Pitting Corrosion
Formation of shallow de-
pressions or deep pits (cav-
ities of small diameter).
Stess Corrosion ,
Cracking
Corrosion accelerated by
residual stresses resulting
from fabrication opera-
tions or unequal heating
and cooling of structure.
Contact or Crevice
Corrosion
Occurs at the point of con-
tact or crevice between a
metal and non-metal or
two metals.
Intergranular
Corrosion
*
Selective corrosion at the.
grain boundaries (micros-
copic) of a metal or alloy.
14

-------
Figure 1.1-9
Some Corrosion Mechanisms at an Underground Steel Tank
Small differences in electric (ionic) potential can cause serious corrosion of
underground steel tanks and pipes. Such differences can be created when there
is a presence of dissimilar soils or bacterial activity, as shown in the figures
below. The curled arrows *) show the flow direction of electrical current
in these figures.
PAVEMENT
or restivity can lead to corrosion in an underground steel structure.
PAVEMENT
can create corrosive environments around an underground structure.
15

-------
figure 1.1-10
More Corrosion Mechanisms at an Underground Steel Tank
Other items which con promote corrosion ot underground steei tonks include
the presence of dissimilar metals or moisture, as shown in the figures below.
The curled arrows (-*-*¦*) show the flow direction of electrical current in
these figures.

-------
Temperature. The rate of corrosion tends to in-
crease with rising temperature. Temperature also has a
secondary effect through its influence on the solubility
of air (oxygen), which is the most common oxidizing
agent influencing corrosion.
Surface Films. Once corrosion has started, its
progress is often controlled by the nature of the film
that forms on the corroding metal. Some corrosion
products may be insolubale and completely impervious
to the corroding solution and, therefore, completely pro-
tective; or they may be very permeable and thus allow
localized or general corrosion to proceed unhindered. In
addition, discontinuous or non-uniform films may in-
duce localized corrosion in particular areas.
Bacterial Action. The metabolic activity of certain
microorganisms can either directly or indirectly affect
the corrosion of metals. Such activity can:
•Produce a corrosive environment.
•Create electrolytic concentration cells, leading to
crevice corrosion.
•Alter the resistance of surface films.
•Alter the environment composition.
•Influence the rate of anodic or cathodic reaction.
An example of microorganisms that directly influ-
ence corrosion rates are the sulfate-reducing bacteria
found in many soils. These bacteria use hydrogen to re-
duce sulfate contained in the soil. The corrosion of met-
als results in the formation of. hydrogen on the metals
surface. If this hydrogen is not removed corrosion is in-
hibited. Sulfate-reducing bacteria can consume this hy-
drogen, thus speeding up the rate of corrosion. In addi-
tion, the reduction of sulfate results in the formation of
hydrogen sulfide, which, in turn; causes further corro-
sion. This effect is shown in Figure 1.1-9.
Soil Resistivity. Soil resistivity is a measure of the
resistance of soil to the flow of electric current, and is
a very important factor in determining the potential rate
of corrosion of underground pipes and tanks. The lower
the resistivity of the soil, the greater the probability of
corrosion. Soil resistivity is dependent upon several fac-
tors, including soil moisture content. In general, soil re-
sistivity is low where soils are moist and groundwaters
contain high levels of dissolved solids. The relationship
between soil resistivity and corrosivity is demonstrated
in Table 1.1-8.
Moisture Level. The presence of water can also
promote corrosion of metals. The presence of moisture
in soils acts to reduce soil resistivity thereby increasing
•the probability of corrosion (see Figure 1.1-10). Water
accumulation inside tanks is also a major cause of inter-
nal corrosion. Water is often present in tanks due to
condensation, precipitation from tank contents, and be-
cause water is often used as a ballast for underground
tanks.
Soil Variations. Corrosion of underground tanks
and pipes can be influenced by variations in soil condi-
tions along the surfaces of those tanks and pipes. Varia-
tions in soil type, soil resistivity, moisture content, etc.,
can promote galvanic activity in the buried metal, thus
accelerating the rate of corrosion.
Table 1.1-8

Soil Corrosivity vs. Soil Resistivity

The USDA Soil Conservation Service has catego-
rized soil corrosivity levels as follows:

Class of Soil
Resistivity
Corrosivity Type of Soil
(ohm-cm)
Very High Poorly Drained Clay
Below 1.000
High Poorly Drained Clay
1.000 to 2.000
Medium Poorly Drained Clay
2,000 to 5,000
Low Poorly Drained Clay
5,000 to 10.000 ,
Very Low Weil Drained Gravel
Above 10.000
Environmental Elements. Corrosion can also be
influenced by the presence of atmospheric pollutants,
both externally and internally. For example, sulfur
dioxide can form sulfuric acid in the presence of air and
moisture and can thus promote corrosion of certain met-
als.
Adjacent Underground Metal Structures. Corro-
sion of underground tanks and piping can also result
when new structures are installed near existing tanks or
other underground metal structures or when new piping
is installed. Since the older structures have rusted to
some extent, they can become cathodic to the newer
tanks or pipes. The system becomes an electrical cell.
The older tank acts as a cathode. The newer metal
(tanks or pipes) becomes an anode and the moist soil
or fill which separates them becomes an electrolyte. A
current flows through the system, carrying oxide,
chloride, sulfide, etc., ions to the new metal surfaces
and carrying metal ions away from the new surfaces. If
the surface area of the old structure, as for instance a
large tank, is much greater than the new structure (a re-
placed length of pipe), the replacement of the new sur-
face with corrosion products (rust) will proceed at a rel-
atively fast rate. This effect is illustrated in Figure 1.1-
10.
Stray Electrical Currents. Stray underground cur-
rents from nearby electrical facilities using DC current,
such as electrified railway or transit systems, factories,
shops or nearby cathodic protection rectifiers can induce
electrolytic corrosion in underground tanks and pipes.
This effect has been shown in Figure 1.1-7.
Internal corrosion of underground tanks is also
often found directly under the fill pipe. This is fre-
quently caused by repeated impact of the measuring dip
stick. If the stick does not have a soft tip, the impact
breaks down any protective film which may have de-
veloped on the tank surface. The result is selective cor-
rosion.
17

-------
4. Corrosion Protection [8,11,20]
There are a number of methods available to protect
against corrosion. These include the use of soluble in-
hibitors, protective coatings, cathodic protection and the
use of corrosion resistant materials of construction. No
method or material is a universal containment; the con-
taining material or system must be "fitted" to the prod-
uct being contained.
Soluble Corrosion Inhibitors. Soluble inhibitors
are substances which can be added to the contents of a
storage system to inhibit internal corrosion. The choice
of a particular chemical for use as an inhibitor is largely
dependent on the composition of the storage system and
its contents. Typical examples of inhibitors that are used
to minimize the corrosion of iron and steel in aqueous
solutions are the chromates, phosphates and silicates.
These substances act to increase anodic polarization and
are therefore called anodic inhibitors. Substances which
control cathodic polarization, such as certain organic
sulfides or amine materials are effective in minimizing
the corrosion of iron and steel in acid solutions. These
substances are called cathodic inhibitors.
Paints, Coating and Linings. Paints and coatings
are widely used as protective measures against corro-
sion, particularly corrosion due to exposure to atmos-
pheric elements. In these instances the paint helps to ex-
clude water and oxygen from the metal surface, thus
minimizing corrosion. Inhibitive pigments, such as red
lead or chromates, can be used in paints to pit ect m< -
als against corrosion. These pigments can act to inhibit
corrosion through several, mechanisms:
•The pigment may neutralize acids.
•The pigment may promote the formation of pro-
tective ferric oxide films at the iron surface.
•Red lead breaks down sulfur dioxide, which is a
very corrosive constituent of ambient air in urban
and/or industrial areas.
Linings applied to the walls of tanks and piping
can also serve to protect these structures from contact
with their environment, thereby inhibiting corrosion.
Examples of common lining materials are rubber,
epoxies and silicones. A more detailed discussion of
coating and lining properties, and their resistance to
chemical and electrochemical attack, is included in Part
H of this document. It should, however, be noted that
no tank or pipe coating is impervious, no matter how
carefully it is applied. Flaws will eventually develop
and accelerated corrosion will occur at these breaks in
the coatings. Consequently, tanks or pipes that are
coated, without other forms of protection frequently fail
faster than bare structures. Thus, most present-day in-
stallation codes require coating in concert with another
form of corrosion prevention, such as cathodic protec-
tion.
Cathodic Protection. Cathodic protection is a
widely used and highly recommended method of protec-
tion for tanks and pipes. It is particularly effective in
underground applications. The method works by revers-
ing the electrochemical action of corrosion. Instead of
allowing electrons to flow away from the structure
(thereby permitting corrosion to occur) an electron flow
toward the structure is induced, thereby protecting the
structure.
Cathodic protection can be applied to either bare
metal or coated metal, but is more effective and less ex-
pensive on coated structures. On bare tanks, cathodic
protection may be only 90 percent effective, due
primarily to the existence of active pits into which the
protective current cannot penetrate [19]. There are two
basic types of cathodic protection. These are the sacrifi-
cial anode (or galvanic) cathodic protection method and
the impressed current (or electromotive force) method.
The galvanic cathodic protection method employs a
sacrificial anode, such as magnesium.or zinc, in electri-
cal contact with the metal structure to be protected.
These may be anodes buried in the ground for the pro-
tection of underground tanks, or attached to the surface
of materials in electrolytic solutions (i.e., the tank or
pipe). The current required is generated by corrosion of
the sacrificial anode material. A typical galvanic
cathodic protection system for underground tanks and
piping is illustrated in Figure 1.1-11.
The impressed current cathodic protection method
employs direct current provided by an external source.
This current is passed through the system by the use of
non-sacrificial anodes such as carbon, non-corrodible al-
loys, or platinum. These anodes are. buried in the
ground (in case of underground structures) or other-
wise suspended in the electrolyte and connected to the
external power supply. An impressed current system for
underground tanks and piping is illustrated in Figure
1.1-12.
Note that the National Association of Corrosion
Engineers' recommended practice NACE RP-01-69 rec-
ommends a -0.85 volt potential, tank to soil, as mea-
sured by a Cu-CuS04 half cell reference. This will en-
sure continued cathodic protection.
Electrical Isolation. Electrical isolation is another
method of corrosion prevention. As the name implies,
it involves the use of non-conductive dielectric fittings,
bushings, connections, etc. to electrically isolate metal
componer.'s in a storage system: this minimizes the po-
tential for the generation of electrical currents between
dissimilar metals. Electrical isolation is often employed
in concert with other corrosion prevention methods,
such as sacrificial anode cathodic protection, to further
decrease the likelihood of coirosion.
18

-------
Figure 1.M1
Magnesium Anode Cathodic Protection
Typical Configuration
Source: Suggested Ways to Meet Corrosion Protection Codes for Underground Tanks and Piping; The Hinchman
Company, Detroit, MI.
19

-------
Figure 1.1-12
Impressed Current Cathodic Protection
Typical Configuration
Test Box
Anode
Positive Header
Cable
Rectifier
Negative Bond
Source: Suggested Ways to Meet Corrosion Protection Codes for Underground Tanks and Piping; The Hinchman
Company, Detroit, MI.
20

-------
Corrosion Allowance. Often corrosion is antici-
pated, and items are constructed with enough metal to
allow for corrosion to proceed to a point without inter-
fering with the normal function of that item. An exam-
ple of such a corrosion allowance is a tank whose de-
sign thickness is such that appreciable corrosion can be
tolerated before a leak or tank failure will occur.
Corrosion-Resistant Materials of Construction.
Corrosion can also be controlled through the use of cor-
rosion-resistant materials of construction. Examples of
such items include special alloys, fiberglass reinforced
plastic, and fiberglass reinforced plastic coatings. Spe-
cial alloys are most often used when difficult-to-contain
fluids are to be handled. Stainless steel is an example
of such a material. Stainless steel is a family of alloys.
The corrosion resistant properties of the specific mate-
rial chosen for the containment vessel should be appro-
priate for the material being contained.
From the perspective of corrosion resistance,
fiberglass reinforced plastic (FRP) tanks are an effective
means of storing many fluids in underground storage
systems, most notably petroleum products. These tanks
are not subject to corrosion and are strong enough to
withstand most soild and other loading stresses when
they are properly installed. The importance of proper in-
stallation of FRP tanks is discussed in further detail in
Part II. The FRP piping is also applicable in these types
of situations.
Fiberglass reinforced plastic coatings are also avail-
able and generally consist of thick (on the order of '/a
inch) coatings applied .to steel tanks. The concerns ex-
pressed above and elsewhere in this document regarding
the use of coatings apply to these types of coatings as
well.
References
1.	New York State Department of Environmental Con-
servation, Bulk Storage of Hazardous Liquids Study
Program - Problem Assessment Report, Paper #5,
New York State Department of Environmental Con-
servation, 50 Wolf Road, Albany, NY 12233,
April, 1981.
2.	New York State Department of Environmental Con-
servation, New York State Bulk Storage Control -
Study Program, New York State Department of En-
vironmental Conservation, 50 Wolf Road, Albany,
NY 12233, May, 1980.
3.	American Petroleum Institute, Underground Spill
Cleanup Manual, API Publication, American Petro-
leum Institute, 2101 L Street, N.W., Washington,
DC 20037, June, 1980.
4.	National Fire Protection Association, Underground
Leakage of Flammable and Combustible Liquids,
NFPA 329, National Fire Protection Association,
Inc., Batterymarch Park, Quincy, Massachusetts
02269, 1977.
5.	Pace Company Consultants & Engineers, Inc., Spill
Control Manual prepared under EPA Training Grant
No. T-900-175-02-2 to the Department of Environ-
mental Science and Engineering, Rice University,
Houston, TX, February, 1975.
6.	Annual Report of the New York State Oil Preven-
tion, Control and Compensation Program. April I,
1978 to March 30, 1979, New York State Depart-
ment of Environmental Conservation, 50 Wolf
Road, Albany, NY 12233.
7.	Annnal Report of the New York State Oil Preven-
tion, Control and Compensation Program, April 1,
1979 to March 30, 1980, New York State Depart-
ment of Environmental Conservation, 50 Wolf
Road, Albany, NY 12233.
8.	American Petroleum Institute, Guide for Inspection
of Refinery Equipment, Chapter I - Introduction.
Chapter XIII - Inspection of Atmosphere and Low-
Pressure Storage Tanks, American Petroleum Insti-
tute, 2101 L Street, N.W., Washington, DC 20037,
1976.
9.	U.S. Environmental Protection Agency, NPDES
Best Management Practices Guidance Document,
Draft, EPA-600/9-79-045, Industrial Environmental
Research Laboratory, USEPA, Cincinnati, OH
45268, December, 1979.
10.	American Petroleum Institute, Underground Leak
Survey results as reported by F.B. Kiilman to API
Underground Leak Task Force, American Petro-
leum Institute, 2101 L Street, N.W., Washington,
DC 20037, February 5, 1981.
11.	Perry, R.H., Chilton, C.H., Chemical Engineers'
Handbook, Fifth Edition, McGraw-Hill Book Co.,
1221 Avenue of the Americas, New York, NY
10020, 1973.
12.	Husock, B., "Fundamentals of Cathodic Protec-
tion", Paper No. HC-2, Harco Corporation,
Cathodic Protection Division, 1055 West Smith
Road. Medina, Ohio 44256, March 6, 1962.
21

-------
References, continued
13.	Husock, B., "Cathodic Protection - One Way to
Prevent Underground Corrosion", Paper i>.o. HC-4,
Harco Corporation, Cathodic Protection Division,
1055 West Smith Road, Medina, OH 44256.
14.	Husock, B., "Corrosion and Cathod'c Protection of
Underground Tanks at Service Stations", Paper No.
HC-15, Harco Corporation, Cathodic Protection Di-
vision, 1055 West Smith Road, Medina, OH
44256, February, 1965.
15.	Husock, B., "Causes of Underground Corrosion",
Paper No. NC-36, Harco Corporation, Cathodic
Protection Division, 1055 West Smith Road,
Medina, OH 44256, May, 1976.
16.	Rothman, P.S., "Cathodic Protection of Tank and
Underground Structures", Harco Corporation,
Cathodic Protection Division, 244 East County Line
Road, Hatboro, PA 19040, 1978.
17.	U.S. Department of Agriculture, "Control of Un-
derground Corrosion", Design Note No. 12, Soil
Conservation Service, U.S. Department of Agricul-
ture, Soil Conservation Service, Feb. I, 1971.
18.	National Associatiort of Corrosion Engineers, Re-
commended Practice - Control of External Corro-
sion on Underground or Submerged Metallic Piping
Systems, NACE Standard RP-01-69, National As-
sociation of Corrosion Engineers, 1440 South
Creek, Houston, TX 77084, January, 1972.
19.	Fitzgerald, J.H., "Corrosion Control for Buried
Service Station Tanks", Paper No. 52, The Interna-
tional Corrosion Forum Devoted Exclusively to the
Protection and Performance of Materials, April 14-
18, 1975, Toronto, Canada, National Association of
Corrosion Engineers, 1440 South Creek, Houston,
TX 77084.
20.	The Hinchman Company, Suggested Ways to Meet
Corrosion Protection Codes for Underground Tanks
and Piping, Job Number 1079-4542, The Hinchman
Company, Corrosion Engineers, 1605 Mutual
Building, Detroit, Michigan 48226, April 8, 1981.
21.	Husock, B., "Use of Pipe-to-Soil Potential in
Analyzing Underground Corrosion Problems",
Paper No. HC-7, Harco Corporation, Cathodic Pro-
tection Division, 1055 West Smith Road, Medina,
Ohio 44256.
22.	Husock, B., "Pipe-to-Soil Potential Measurements
and Cathodic Protection of Underground Struc-
tures", Paper No. HC-8, Materials and Perfor-
mance, Vol. 10, No. 5, May, 1971.
23.	Rizzo, F.E. "Detection of Active Corrosion", Paper
No. HC-14, Harco Corporation, Cathodic Protec-
tion Division, 1055 West Smith Road, Medina, OH
44256.
24.	Hosford, H.W. "Cathodic Protection of Marine
Structures", Paper No. HC-16, Harco Corporation,
Cathodic Protection Division, 1055 West Smith
Road, Medina, OH 44256.
25.	Rizzo, F.E., Miller, M., "New Technique for CP
Monitoring of Offshore Pipelines", Paper No. HC-
39, Oil and Gas Journal, February 13, 1978.
22

-------
Table 1.2-1
Part I
CHAPTER 2*
HAZARDOUS SUBSTANCES
Table 1.2-1
Listings of Regulated Hazardous
Hazardous Materials Listing
A. LISTINGS OF HAZARDOUS SUBSTANCES
The term hazardous liquids includes a broad range
of chemicals and chemical types. They may be desig-
nated as hazardous because they are flammable, com-
bustible, corrosive, toxic or explosive (reactive). By
their nature, they are of great concern to society, and
to those governmental agencies which are responsible
for public health, environmental protection, transporta-
tion, occupational safety, and fire and emergency re-
sponse.
Several agencies have prepared lists of hazardous
substances and have included these lists in regulations
to control use, transporation and disposal of these mate-
rials. The listings of materials regulated by the U.S. En-
vironmental Protection Agency (EPA) and the U.S. De-
partments of Labor and Transportation are described in
Table 1.2-1. These listings can be obtained from the
Federal Register or the Superintendent of Documents,
U.S. Government Printing Office, Washington, D.C.
20402.
There are also several published reports which pro-
vide information on the physical and chemical proper-
ties of and safe handling practices for various hazardous
materials. Some of the more widely used reports are de-
scribed in Table 1.2-2. A more comprehensive list of
references is provided at the end of this chapter.
B. PROPERTIES OF
HAZARDOUS SUBSTANCES
1. Chemical Properties
Appendix A of this document includes a series of
tables that identifies the chemical and physical charac-
teristics of various solids, liquids and gases that are
classified as toxic or hazardous substances. These tables
identify substances which are poisonous to humans,
flammable, corrosive, reactive and highly toxic to aqua-
tic life. Other properties that are identified are the
biodegradeability of liquids and solids, the amenability
of liquids and solids to biological waste treatment, the
volatility of liquids and the solubility of solids.
Substances
(49 CFR 172.101) — The labeling, packaging and
transportation of these material are regulated by
the U.S. Department of Transportation.
Toxic and Hazardous Substances Listing
(29 CFR 1910 Subpart Z) — Occupational expo-
sure of these substances are controlled by the U.S.
Department of Labor.
Listing of Hazardous Waste
(40 CFR 261) —The disposal of chemical wastes
on this listing are regulated by the Environmental
Protection Agency under the Resource Conserva-
tion and Recovery Act (RCRA).
Designation of Hazardous Substances
(40 CFR 116)—Chemicals which are hazardous
to the environment are identified by the Environ-
mental Protection Agency on this listing.
Available from: Federal Register or
U.S. Goverment Printing Office
Washington, D.C. 20402
A listing of the physical and chemical properties of
twenty-nine commonly used hazardous substances in
New York State is given in Table 1.2-3. This table rep-
resents common chemicals that are stored in bulk.
Usage of these chemicals is from 1 million to 450 mil-
lion pounds per year (not counting petroleum). The
properties which are identified in this table include the
following:
•The physical state at 20°C.
•The melting and boiling points.
•The specific gravity at 20°C or other specified
temperatures.
•The solubility in water.
•The vapor pressure.
•The associated hazard (flammable, corrosive or
toxic).
•The reactivity with common storage tank mate-
rials.
Knowledge of these various physical and chemical
properties is important in determining the proper mode
of storage of these substances. For example:
•The melting and boiling points of substances is
useful in determining the appropriate range of
storage temperatures.
•The solubility of the substance is helpful in deter-
mining whether the substance should be allowed
to come in contact with water.
23

-------
Table 1.2-2
Reports Describing Hazardous Materials
U.S. Dept. of Transportation Hazardous Materials Emergency Response Guidebook [1]:
-	numerical and alphabetical indices of hazardous materials
-	descriptions of health hazards and fire or explosion potential
-	procedure to be followed in the event of fire, spill, leak or personnel exposure.
-	isolation and evacuation distances for selected hazardous materials
U.S. Department of Transportation
Washington, D.C. 20590
National Fire Protection Assn. Publication NFPA 49: Hazardous Chemicals Data [2]:
-	degree of health hazard	- fire explosion hazard
-	potential for reactivity	-	particular life hazards
-	flammability	-	personal protection requirments during handling
-	physical descriptions	-	fire fighting phases
National Fire Protection Association
Batterymarch Park
Quincy, MA. 02269
U.S. EPA Hazardous Materials Spill Monitoring and Safety Handbook and Chemical
Hazard Guide Parts A and B [3]:
Part A	Part B
-	Safety consideration	-	hazard priority number
- - first aid procedures	-	hazards
-	protective equipment	-	safety measures
-	priority listing of hazardous material	-	synonyms
National Technical Information Service
Springfield, VA. 22161
or
U.S. Environmental Protection Agency
Office of Research and development
Environmental Monitoring amd Support Laboratory
Las Vegas, NV. 89114
The Chemical Hazards Response Information System (CHRIS) Manuals [4]:
-	Medical data (exposure hazard)	- physical properties
-	flammability data (fire hazard)	- chemical properties
-	pollution data	- preventative and precautionarey initial
-	biological data	- response information
United States Coast Guard
U.S. Dept. of Transportaion
Washington, D.C. 20590
24

-------
•The vapor pressure of the liquid substances is nec-
essary to determine appropriate storage pressures
(pressures at which significant vapor formation
can be limited).
•The hazard associated with a particular substance
is important in determining handling and storage
protocols.
•How the substances react with various materials of
construction is important in determining the mode
of storage and the materials used in storage.
2. Relationships Between Temperature, Pressure
and Volume Within a Storage Tank
In the handling and storage of hazardous liquids, it
is important to note that most liquids expand and con-
tract with changes in temperature. Variations in the tem-
perature of the stored liquid can lead to changes in the
volume of the stored liquid. In addition, variations in
pressure can lead to changes in the volume of the stor-
age tank itself. These volume variations become ex-
tremely important when one is attempting to detect
small leaks from storage tanks.
The temperature of a liquid stored in a tank either
above or below ground can vary throughout the year.
The reasons for such variations include the following:
•The seasonal variations in ambient temperature.
•Changes in the weather (e.g., hot, sunny days vs.
cold, rainy nights).
•Changes in pressure (compression) of the liquid.
An annual temperature profile for an underground tank
is displayed in Figure 1.2-1. Although this profile was
observed in an underground gasoline tank, it is typical
of the types of variation that can be expected of most
liquids which are stored underground.
Liquid temperatures can also vary throughout the
storage tank itself. The reasons for such variation in-
clude the following:
•Variation in the surface temperature of aboveg-
round tanks due to weather or exposure to the
sun.
•Stratification of temperature in the ground sur-
rounding an underground storage tank.
•The introduction of new liquid into a tank that has
a different temperature than the liquid already
stored in the tank [5.6],
25

-------
Table 1.2-3
Physical and Chemical Properties of the Twenty-Nine Most Commonly Used
Hazardous Substances in New York State
Physical Melting
Substance Stale Point C°
at 20°C
Boiling
Point C°
Specific
Gravity
Solubility2
(mg/1) in H2O
Vapor Pressure1
(mm of Mq)
Hazard4
Petroleum
Tank
Chemical
Tank
Carbi
Stec
1. Petroleum




Insoluble
NA
E, F
OK
	
OK
-Gasoline
Liquid
NA
60-199
0.132
NA
NA
E. F
OK
__
OK
-No. 2 Fuel Oil
Liquid
NA
NA
NA
515
28
E. F

OK
OK
2. Toluene
Liquid
-95
110.4
0.866 20/4
2,860-°
5
E
—
OK
8
3. Tetrachloro-
Liquid
-44
146.5
1.58 25/4






ethane




1.000
74"
E, F
...
—
7,8
4. Methyl Chloride
Gas
-97
-23.7
0.918 20/4






5. Trichloro-
Liquid
-73
87.1
1.45560 25/4
1.000
100"
C
....
—
8
eihylene










6. Tetrachloro-
Liquid
-23.25
121.20
16230 20/4
150
15.8"

--
OK
8
elhylene










7. Methylene
Liquid
-96.7
39.8
1.32
13.200-20,000
362.4
F

—
8
Chloride (Di-










chloromethane)




¦





8. Phenols
Solid
43°
182
1.071 25/4
82.000
0.20"
C. E


7.8
9. Cresols
Liquid
10.9-35.5
191-203
1.048 20/4
25,000
l'«
C

—
8
10. Xylene
Liquid
25.4-75
203-225
NA
NA
NA
NA

OK
NA
or Solid



60
2.600"
C, E

OK
NA
II. Vinyl Chloride
Gas
-160
-13.4
0.908 25/ 25
1,780*'
95.2
E. F

_.
OK
12. Ben/enc
Liquid
5.51
80.093-
0.8794
Very Slightly
5.5
E. F
...
OK
OK



80.094

Insoluble
1-'-
C, E


8
13. Styrene
Liquid
-31
146
0 9074 20 4
NA
5"
NA
....
—
OK
14. Chloroioluene
Liquid
-43
179
1.1026 18 4






15. P-C'hloroben/o-
Liquid
-36
139.3
1.353 15.5 15.5
NA
NA
NA


NA
trilluoride



448"'
lO-'-V
F
...
—
7,8
16. Octyl Phenol
Solid
NA
280-283
0 941 24 4

•




17. Chlorinated
Liquid
-45
131 7
I.I 13 15 5 15.5
480-4,400
96

...
OK
OK
Bcn/encs




7.840
192
....
....

8

-------
Table 1.2-3 continued
18.
Trichlorethane
Liquid
- 4-39
74 .
1.31
NA
NA
F»
—
OK
NA
19.
Chloroform
Liquid
-63.5
61.26
1.49845"
60.000" P-V
C

—
— ¦
OK
20.
Sevin
Solid
142
NA
1.232 20 20
Insoluble
I0M.»
C
—
OK
OK
21.
llydroquinone
Solid
170.5
286.2
1,358 20 4
Insoluble





22.
P-Dichlorobcnzene
Solid
53
173.4
1.4581 20.5 4
10'V
E. F
—
—
OK






60.000
1N."
C
—
—
OK
23.
Pyridine
Liquid
-42
115.3
0 982 20 4
Infinitely
I IIIK B
C
OK
—
OK
24.
Aniline
Liquid
-6.2
184.4
1.20 20 4
36.000'"





25.
Diethylphtha-
late
Liquid
-40.5
296-302
1.110-121 20 20
35.000
71.2
E. F
—
—
OK
26.
2-Butanone
Liquid
-86
79.6
0.805 20,4







(Methyl Ethyl



Insoluble
200".5
E
OK
—
NA

Ketone)




800
100"
—
—
OK
8
27.
Freon 113
Liquid
-35
47.6
1.576 20,4






28.
Carbon
Liquid
-22.6
76.8
1.597
30
0.8715
C. E
OK
—
OK

Tetrachloride









29.
Naphthalene
Solid
80 1
217.9
1,162 20/4






Notes:
1.	Specific gravity at 20° C or as otherwise stated. Where stated, numerator is temperature of substance, denominator
temperature of water.
2.	Solubility at 25° C or as otherwise stated.
3.	Vapor pressure at 20° C or as otherwise slated.
4.	All listed substances are toxic to humans at some concentrations.
E = Explosive
F r Flammable (flashpoint of less than 80° F)
C = Combustable (flashpoint of 80° F or higher)
5.	Compound itself is not flammable but it is usually dissolved in a combustible liquid.
6.	Not recommended.
7.	Corrosive at high temperatures.
8.	Corrosive at high concentrations.
9.	Chemical compatibility may vary from what is shown in this tabic if special resins or other materials are used for
lank construction. Check with the manufacturer for lab analyses of chemical compatibility and for other assurances that
the tank you are using is warrantced for the chemical being stored.
NA - Not available
Sources: References 4. 11. 12. 15. IS). 20. 22

-------
Figure 1.2-1
Typical Annual Tank Temperature Variation
For an Underground Gasoline Tank
Temperature F
90








4^
ir*-y


w



lk,






%
' rfv

IV
fyvy*
ft
J YY
s
Mi
ervice Se
sntrose a
ition No.
: West All
379 House
ibama
on



i r
' I


Te
Dtp. of Fr
oduet In
Jndergrour
id Tank













ao
70
60
50
40
NOV.
1994
DEC.
1994
JAN.
1995
FEB.
1953
MAR.
1959
APR.
1955
MAY
1955
JUN.
1955
JUL.
1955
AUG.
1955
SEP.
1955
OCT.
1955
This graph shows temperature recordings for an entire year by combining the results of 52 weekly graphs. The
vertical lines, either down or up show the immediate effect of the delivery on the tank temperature and the curving
lines show the gradual return to underground temperatures.
The graph also shows a seasonal change of 30°F in underground temperatures occurring in south Texas. Much
greater differences between summer and winter would exist in New York State.
Source: Reference 21.
28

-------
Figures 1.2-2, 1.2-3 and 1.2-4, taken from a Stan-
ford Research Institute study on detection of small
leaks, illustrate temperature variations throughout an un-
derground gasoline tank [5]. Figure 1.2-2 shows the lo-
cation of temperature sensors in the tank. Tank tempera-
tures at these various sensor levels, as a function of
time for a 24-hour period after tank filling, are dis-
played in Figure 1.2-3. As shown in this figure the tank
temperature at each level differs and all these tempera-
tures vary with time. The mean temperature variation,
as a function of depth for this same tank over four dif-
ferent 24-hour periods, is illustrated in Figure 1.2-4.
Again, as shown for each of the 24-hour periods, the
liquid temperature can vary substantially throughout the
tank.
Because liquids expand or contract as their temper-
ature is raised or lowered, seasonal, day-to-day and
tank-wide variations in temperature heavily affect the
detection of small leaks. For example, in an 8-foot di-
ameter, 8,000-gailon storage tank half-full of gasoline,
a 1.2 gallon per day (0.0S gallon per hour) leak would
cause only a 500 micro inch (0.0005 inches) height
change in the gasoline level. A mean gasoline tempera-
ture change of only 0.012°C.(0.022°F) would also result
in a 500 micro inch height change of the gasoline level.
Thus a 1.2 gallon per day leak could be hidden by a
0.012°C rise in mean liquid temperature [5],
Internal tank pressure can also affect tank volume
by leading to increases or decreases in the size of the
tank. For example, the total forces exerted on the ends
of tanks (assuming flat ends) of different diameters by
different pressures are exhibited in Table 1.2-4. This
table shows that a 3 pound per square inch (psi) pres-
sure exerted on a tank's contents results in a force of
over 10 tons on the ends of an 8-foot diameter tank [6].
This is sufficient force to cause the tank ends to bulge
outward some small fraction of an inch, thus increasing
the volume of a tank.
Figure 1.2-2
Location of Temperature Sensors
In The SRI Tank
AMBIENT
TO CHART '

RECORDER
GROUND LEVEL





SRI UNDERGROUND TANK
h


113.5 in




h


78.5 GASOLINE LEVEL H * 85.5 in
h


67.5
h
<
i
44.5
h


21.5
if


10.5
2.5
Source: Stanford Research Institute. "Measurement of Small Leaks
in Underground Gasoline Storage Tanks Using Laser Inter-
ferometry," sponsored by the American Petroleum Institute,
1979.
29

-------
Figure 1.2-3
Tank Temperature at Various Heights
As a Function of Time for a 24-Hour Period
After Tank Fill-up
Source: Stanford Research Institute, "Measurement of Small Leaks in Underground Gasoline Storage Tanks Using
Laser Interferometry," sponsored by the American Petroleum Institute, 1979.
30

-------
Figure 1.2-4
Mean Temperature Distribution as a Function
Of Depth for Four Different 24-Hour Periods
IZ
i
na
<
~—
u.
O
§
g
CO
O
OS
8
7
6 f-
i 2
<
un
-

10-30-78 —


T L j— 10-1-78
9-29-78 —

-
/ \ / /	 10-14-78
. i
~ i
i i
f T P
irr
1 I r
20° 21° 22° 23°
TEMPERATURE - °C
24°
25fl
26°
Source: Stanford Research Institute, "Measurement of Small Leaks in Underground
Gasoline Storage Tanks Using Laser Interferometry," sponsored by the Ameri-
can Petroleum Institute. 1979.
31

-------
Table 1.2-4
Total Force on Tank Ends Due to
Internal Pressure
Tank	Total Force in Tons
Diameter
(inches) 1 psi 2 psi 3 psi 4 psi 5 psi
48
0.9
1.8
2.7
3.6
4.5
64
1.6
3.2
4.8
6.4
8.0
72
2.0
4.0
6.0
8.0
10.0
84
"2.8
5.6
8.4
11.2
14.0
96
3.6
7.2
10.8
14.4
18.0
Source: Reference 6
The extent of tank bulging under pressure is depen-
dent upon several factors, including tank diameter, tank
age, the softness or wetness of the surrounding soil (for
underground tanks)< and past filling practice [6]. How-
ever, an 8-foot diameter tank could bulge enough to in-
crease tank volume by 13 gallons or more [6].
These pressure effects become very important when
one is attempting to detect small leaks using a method
such as standpipe testing which places a pressure on the
tank contents. For example, filling a 4-foot long
standpipe on an 8-foot diameter gasoline tank buried
three feet below grade puts an average pressure of 3.69
psi on the center of the tank [6]. This is sufficient to
put more than 10 tons of force on the ends of the tank,
and will lead to an increase in tank volume and a cor-
responding loss of volume in the standpipe. Thus, the
detection of small leaks using a standpipe testing
method becomes difficult.
C. STORAGE AND HANDLING PROTOCOLS
1. Storage and Handling Systems
Both aboveground and underground bulk storage
systems should be composed of five basic components.
These are:
•The product storage system (storage tanks).
•The product transfer (piping) system.
•An overfill prevention system.
•A spill containment and collection system.
•A leak detection system.
The basic methods and design considerations associated
with these five components are summarized in Table
1.2-5 as a prelude to Parts II and III of this report. The
advantages and disadvantages of the various materials of
construction employed for product storage and transfer
systems are summarized in Table 1.2-6.
2. Aboveground vs. Underground Storage
The choice of aboveground or underground tanks
as an appropriate means of storage for a particular haz-
ardous substance is dependent upon several factors, in-
cluding the following:
•Type and amount of liquid to be stored.
•The availability of space (real estate) for storage.
•The level of product and tank accessibility re-
quired.
•The type of -soil in the area.
•Groundwater levels in the area.
•Fire hazard considerations.
A comparison of the advanatages and disadvantages of
aboveground and underground storage systems is pre-
sented in Table 1.2-7.
The reader should note that the storage of liquefied
or compressed gases, such as liquefied natural gas, re-
quires adherence to special design criteria as described
in the following publications:
•API Standard 2510—The Design, and Construction
of Liquefied Petroleum Gas Installations at Marine
Pipeline Terminals, Natural Gasoline Plants, Re-
fineries and Tank Farms [7],
•NFPA Standard 58—Standard for the Storage and
Handling of Liquefied Petroleum Gas [8].
•NFPA Standard 59—Standard for the Storage and
Handling of Liquefied Petroleum Gas at Utility
Gas Plants [9].
In addition, the American Society of Mechanical
Engineers (ASME) has written the A.S.M.E. Boiler and
Pressure Vessel Code [10], which contains rules for the
design, fabrication and inspection of boilers and pres-
sure vessels. This code consists of eleven sections as
follows:
I	Power Boilers
II	Material Specifications
III	Nuclear Power Plant Components
IV	Heating Boilers
V	Nondestructive Examination
VI	Recommended Rules for Care and Operation
of Heating Boilers
VII	Recommended Rules for Care of Power Boil-
ers
VIII	Pressure Vessels, Division 1
Pressure Vessels, Division 2, Alternative
Rules
IX	Welding Qualifications
X	Fibergalss-Reinforced Plastic Pressure Vessels
32

-------
Table 1.2-5
Storage System Components — Methods, Materials, and Design Considerations
Product Storage
Product Transfer
Piping & Accessories
Overfill Protection
Spill Containment Leak Detection
-aboveground tanks
-underground tanks
-single-walled tanks
-tank linings and coatings
-tank wrappings
-design considerations
corrosion resistance
chemicaJ compatibility
structural strength
pressure relief
foundation require-
ments
safety factors
-tank materials selection
carbon steel
stainless steel
fiberglass-reinforced
plastic (FRP)
fiberglass/ steel bonded
tanks
-coating and lining
materials selection
alkyds
epoxies
phenol ics
-wrappings
vinyl
polyethylene
-surface/ subsurface piping
-hoses
-loading racks
-design considerations
corrosion resistance
chemical compatibility
structural strength
pipe supports
safety factors
-materials selection
carbon steel
stainless steel
fiberglass-reinforced
plastic (FRP)
polyvinyl chloride
(PVC)
polypropylene
-check valves
-emergency shutoff valves
•coupling mates to
prevent mixing of
incompatible chemicals
-level control devices
floats
displacers
gas bubblers
hydrostatic head
devices
capacitance devices
thermal conductivity
devices
ultrasonic devices.
optical devices
nucleonic devices
-automatic shutoff con-
trols and flow
diversion
-high level alarms
-liquid level gages
-check valves
-operating practices for
overfill protection
-dry disconnection hose
valves
-catchment basins
-impervious perimeter
dikes, berms
cutoff walls
curbs
aprons/ slabs
drainage ditches
troughs
-liners
synthetic membranes
asphalt, contrete
-in-situ absorbing/
neutralizing media
for spill containment
-spill collection systems
-secondary containment
tanks (double-walled
tanks)
-clay liners
-inventory con-
trol
-visual inspection
-interstitial
monitoring of
double walled
tanks
-soil/ ground-
water
monitoring
-tightness tests
-structural tests
33

-------
Table 1.2-6
General Properties of Materials Used for Storage Tanks and Piping
Structural Materials
Advantages
Disadvantages
Relative Cost
Carbon Steel
Stainless steel
Fiberglass-reinforced
plastic (FRP)
Polyvinyl chloride
(PVC)
Concrete
Aluminum
FRP/steel bonded
tanks
Linings and Coatings
Alkyds
Compatible with petroleum
products but not compatible
with corrosive chemicals, such
as mineral and oxidizing acids,
without coatings. High struc-
tural strength.
Material has better corrosion
resistance than carbon steel and
higher structural strength. There
are more than 70 standard types
of stainless steel and many
special alloys.
Compatible with petroleum and
several chemical products.
Excellent chemical resistance to
acids, alkalis, and gasoline.
Generally good resistance to
chemical attack when exposed to
dilute organic acids. Epoxy coat-
ings are often applied to concrete
to provide chemical resistance
and decrease permeability.
Excellent resistance to atmos-
pheric conditions and compatible
with mineral and organic acids.
Material has the combined ad-
vantage of the corrosion
resistance of fiberglass and the
structural strength of steel.
Alkyd-phenolics and alkyd-sili-
cones have good weather-ability
and good to excellent resistance
to gasoline, non-halogenated
organic solvents and alchohols.
They may be applied to both the
interior and exterior of tanks
and pipes.
Subject to attack by corrosive
soils and corrosive chemicals
such as mineral and oxidizing
acids.
Lower grade steels (i.e.. marten-
sitic steels) are not suitable for
reducing acids such as HC1.
Lacks the structural strength
and impact resistance of steel
tanks. Not compatible with some
organic solvents.
Plastics have low structural
strength and are less resistant to
mechanical abuse than steels.
They are generally not suited for
the storage or handling of
organic solvents such as benzene,
carbon tetrachloride and acetone.
Concrete is subject to cracking
and spalling with changes in
temperature «uch as during
freeze/ thaw cycles. Generally
poor resistance to chemical attack.
Pure aluminum has relatively low
structural strength and as such is
generally not used in the fabrica-
tion of tanks and pipes. Alumi-
num alloys are available but they
are costly.
The main disadvantage of these
tanks is their cost.
Not compatible with mineral
acids, alkalis, chlorinated solvents,
and organic acids.
Low
Medium to high,
depending on grade
of steel.
Comparable to
coated steel.
Low
High
High
Medium
Low
34

-------
T abb 1.14 continued



Epoxies
These materials include epoxy-
amines. epoxv-esters and epoxy-
phenolics. These materials have
excellent weatherability and excel-
lent chemical resistance to gas-
oline. non-halogenated organic
solvents, alkalis and mineral acids.
Epoxies may be applied to both
interior and exterior of tanks
and pipes.
Generally good to poor resist-
ance to orgranic acids depending
on the acid.
Low
Glass
Used for internal coatings.
Very good chemical resistance.
High cost. Very fragile.
High
Phenolics
Excellent durability and excellent
resistance to gasoline, non-halo-
genated organic solvents, and
alcohols. Phenolic coatings may
be applied to both interior and
exterior of tanks and pipes.
Phenolic coatings generally exhib-
bit poor resistance to alkalis,
mineral acids, chlorinated sol-
vents. and organic acids.
Low
Wrappings
Vinyl
Good resistance to gasoline, non-
halogenated solvents, alkalis and
mineral acids. Vinyl coatings are
usually applied as loose wrap-
pings around tanks and pipes.
Not compatible with chlorinated
solvents, and exhibits excellent to
poor resistance to organic acids
depending on the acid. Wrap-
pings are usually not as effective
as coatings because water often
penetrates the space between the
wrapping material and the tank.
Low
Polyethylene
Very good resistance to oxidizing
acids, and some organic acids
and alkalis. Polyethylene is usual-
ly applied as a loose wrapping
around tanks and pipe.
Not compatible with gasoline and
organic solvents. Wrappings are
usually not as effective as coalings
because water often penetrates the
space between the wrapping mate-
rial and the tank.
Low
Source Reference 3.



Tabic 1.2-7
General Comparison of Aboveground and Underground Storage
Type of Storage
Advantages
Disadvantages
Aboveground
Underground
Accessible for equipment inspection
and surveillance (leak detection).
Allows for storage of much greater
volumes in a single tank.
Tanks accessible for cleanout and
maintenance.
Equipment out of sight: generally
aesthetically pleasing: more efficient
use of limited yard space.
Accidental damage to equipment due
to vehicular traffic (running into
equipment) or similar causes is avoided.
Less subject to lire hazards and
vandalism.
Can be used at existing facilities located
in flood plains if properly designed.
(New facilities, whether aboveground
or underground, may not be located in
a tlood plain under NYS regulation).
Generally not aesthetically pleasing
because equipment is visible.
Requires large spill containment
volume.
Subject to damage in flood area.
Greater potential for vapor loss in
atmospheric tanks due to temperature
fluctuations because tank is exposed.
Equipment is exposed to weathering.
Greater exposure to fire.
Not accessible for inspection, surveil-
lance. or easy maintenance.
One must be concerned with soil
induced corrosion.
Potential for undetected leak and resul-
tant groundwater contamination is
much greater.
35

-------
Table 1.2-8

Capatibility Chart:
Chemicals vs. Structural Materials
Construction
Material
Generally
Incompatible with:
Steel
Mineral acids; nitric, hydroch-
loric, dilute sulfuric acids
Aluminum
Alkalies: potassium hydroxide,
sodium hydroxide,
mineral acids
Magnesium
Mineral acids
Lead
Acetic acid, nitric acid
Copper
Nitric acid, ammonia
Zinc
Hydrochloric acid, nitric acid
Tin
Organic acids, alkalies
Titanium
Sulfuric acid, hydrochloric acid
Fiberglass
Reinforced plastics
Some organic solvents
Lining
Materials
Generally
Incompatible with:
Alkyds
Strong mineral acids, strong
alkalies, alcohols, ketones,
esters, aromatic hydrocarbons
Vinyls (poly-vinyl-
chloride-PVQ
Ketones, esters, aromatic
hydrocarbons
Chlorinated
Rubbers
Organic solvents
Epoxy: (amine-
cured, polyamide
cured, or esters)
Oxidizing acids (nitric acid),
ketones
Coal Tar Epoxy
Strong organic solvents
Latex
Oxidizing acids, ketones, esters
Polyesters
Oxidizing acids, strong alkalies,
mineral acids, ketones, aro-
matic hydrocarbons
Silicones
Strong mineral acids, strong
alkalies, alcohols, ketones,
aromatic hydrocarbons
3. Chemical Compatibility
A primary concern in the handling and storage of
hazardous liquids is the compatibility of these liquids
with the storage system components. If, for example, a
liquid is stored in a tank composed of a material which
is incompatible with that liquid, accelerated and possi-
bly very rapid deterioration of the tank could occur.
This could result in a major leak or spill incident.
General information regarding the compatibility of
various hazardous liquids with different materials of
construction and lining materials is provided in Table
1.2-8. As shown in this table, steel is generally compat-
ible with hydrocarbons, but is incompatible with most
acids. In using Table 1.2-8 as a reference, the reader
should note that when FRP is used as a material of con-
struction, resins that are compatible with the material to
be stored must be used. Other sources of information on
material compatibility are listed in Table 1.2-9.
The compatibility of hazardous liquids with other
types of liquids is also of concern. Liquids may come
in contact with one another if, for example, a tank stor-
ing liquid A is not thoroughly cleaned before it is used
to store liquid B. If these liquids were incompatible, a
violent reaction could ensue with potentially destructive
effects.
Table 1.2-9
References on Material and
Chemical Compatibility
Chemical Engineering Handbook. Perry and
Chilton [I I]
Corrosion Data Survey. National Association of
Corrosion Engineers [12]
"Beat Corrosion With a Rubber Hose," Gal-
lagher. Chemical Engineering, September 8.
1980 [13]
"Guide for Protection of Concrete Against
Chemical Attack by Means of Coatings and
Other Corrosion-Resistant Materials." Ameri-
can Concrete Institute Committee 515 [14]
The Merck Index, Merck and Company [15]
A Method of Determining the Compatibility of
Hazardous Wastes. U.S. EPA-600/1-80-076,
April, 1980 [16]
The Chemical Hazards Response Information
System, Chemical Data Handbook, U.S. Coast
Guard, U.S. Department of
Transportation [ 17]
36

-------
References
1.	U.S. Department of Transportation. Hazardous Ma-
terials Emergency Response Guidebook. USDOT P
5800.2, U.S. Department of Transportation.
Washington, DC. 1980.
2.	National Fire Protection Association. Hazardous
Chemicals Data, NFPA 49, Batterymarch Park.
Quincy, MA 02269, 1975.
3.	U.S. Environmental Protection Agency, Hazardous
Materials Spill Monitoring and Safety Handbook
and Chemical Hazard Guide. Parts A and B, EPA-
600/4-79-008a/b, PB295853 and PB295854, Office
of Research and Development. U.S. Environmental
Protection Agency, P.O. Box 15027, Las Vegas,
Nevada 89114, January, 1979.
4.	U.S. Department of Transportation, Chemical Haz-
ards Response Information System ICHRIS) Manu-
als, U.S. Coast Guard, U.S. Department of Trans-
portation, Washington, DC
5.	Maresca, J.W., Evans, P.C., "Measurement of
Small Leaks in Underground Gasoline Storage
Tanks Using Laser Interferometry," SRI Interna-
tional, Menlo Park, CA 94025, October 31, 1979.
6.	McLean. F.R., "Leak Seeking in Underground
Tanks," Proceedings of the Forty-third Annual Fire
Department Instructors Conference, March 30 -
April 2, 1971, Kansas City, MO.
7.	American Petroleum Institute, The Design and Con-
struction of Liquefied Petroleum Gas Installations
at Marine Pipeline Terminals, Natural Gas Plants,
Refineries and Tank Farms, API Standard 2510,
American Petroleum Institute, 2101 L Street,
N.W., Washington, DC 20037.
8.	National Fire Protection Association, Standard for
the Storage and Handling of Liquefied Petroleum
Gas, NFPA 58, Batterymarch Park. Quincy, MA
02269.
9.	National Fire Protection Association, Standard for
the Storage and Handling of Liquefied Petroleum
Gas at Utility Gas Plants, NFPA 59, Batterymarch
Park, Quincy, MA 02269.
10. American Society of Mechanical Engineers,
A.S.M.E. Boiler and Pressure Vessel Code, Ameri-
can Society of Mechanical Engineers, 345 East 47
Street, New York, NY 10017.
11.	Perry. R.H., Chilton. C.H., Chemical Engineers'
Handbook. Fifth Edition. McGraw-Hill Book Com-
pany. 1221 Avenue of the Americas, New York,
NY 10020. 1973.
12.	Hamner, N.E., Corrosion Data Survey, Fifth Edi-
tion. National Association of Corrosion Engineers,
1440 South Creek. Houston. TX 77084, 1974.
13.	Gallagher, R., "Beat Corrosion With Rubber
Hose". Chemical Engineering. McGraw-Hill Book
Company, 1221 Avenue of the Americas, New
York. NY 10020. September 8. 1980.
14.	American Concrete Institute Committee 515, Guide
for Protection of Concrete Against Chemical Attack
by Means of Coatings and Other Corrosion-Resis-
tant Materials.
15.	Windholy. M.. Budauau. S.. Stroumtsos. L.Y.,
Fertig. M.N., The Merck Index. An Encyclopedia
of Chemicals and Drugs. Ninth Edition. Merck &
Company. Inc., Rahway. NJ. 1976.
16.	U.S. Environmental Protection Agency. A Method
for Determining the Compatibility of Hazardous
Wastes, EPA 600/2-80-076, U.S. Environmental
Protection Agency. Municipal Environmental Re-
search Laboratory. Cincinnati, OH 45268, April,
1980.
17.	New York State Department of Environmental Con-
servation, Properties of Top Ten Hazardous Liquids
Used By Industry in New York State. Division of
Water. New York State Department of Environmen-
tal Conservation. 50 Wolf Road, Albany. NY
12233, October. 1980.
18.	McAnaly, M.A.. Dickerman. J.C., Summary and
Analysis of Data From Gasoline Temperature Sur-
vey Conducted at Service Stations by American Pe-
troleum Institute, API Publication No. 4278. Radian
Corporation. 8500 Shoal Creek. Austin, TX,
November II. 1976.
19.	Sax, Irving N., Dangerous Properties of Industrial
Materials, Van Nostrand Reinhold Company, 135
W. 50 Street, New York, NY 10020, Fifth Edition,
1979.
20.	U.S. Environmental Protection Agency, Innovative
and Alternate Technology Assessment Manual, EPA
430/9-78-009, USEPA MERL. Cincinnati. OH,
February, 1980.
37

-------
References Continued
21.	Heath Consultants, Inc., Procedure Manual for the
Operation of the Petro Tite Tank Tester, Form 582
HPN 5124, Heath Consultants, Inc., 100 Tosca
Drive, Stoughton, MA 02072.
22.	Owens-Coming Fiberglas Corp., Fiberglas Tower,
Toledo, OH 43659.
23. National Association of Corrosion Engineers, Re-
commended Practice RP-01-69, "Control of Exter-
nal Corrosion on Underground or Submerged
Metallic Piping Systems," 2400 West Loop South,
Houston, TX 77027.
Additional Information on Hazardous Chemicals
The following sources may be consulted for addi-
tional information:
A Comprehensive Treatise on Inorganic and Theoretical
Chemistry, J.W. Mellor, John M. Wiley and Sons,
Inc.
Accident Prevention Manual for Industrial Operations,
National Safety Council.
American Insurance Association (Chemical Hazards
Bulletins, Research Reports, Special Interest Bulle-
tins).
Bureau of Mines, U.S. Department of the Interior.
Chemical and Engineering Dictionary, Chemical Pub-
lishing Company.
Condensed Chemical Dictionary, Reinhold Publishing
Company.
Dangerous Properties of Industrial Materials, N. Irving
Sax, Reinhold Publishing Corp.
Encyclopedia of Chemical Technology, Kirk and Oth-
mer, Interscience Publishing Company.
Fire Officer's Guide to Dangerous Chemicals, Charles
W. Bahme, NFPA.
Guide to Safety in the Chemical Laboratory, Manufac-
turing Chemists' Association.
Handbook of Chemistry, N.A. Lange, McGraw-Hill
Book Company, Inc.
Handbook of Chemistry and Physics, The Chemical
Rubber Company.
Handbook of Industrial Loss Prevention, Associated
Factory Mutual Fire Insurance Companies.
Handbook of Laboratory Safety, The Chemical Rubber
Company.
Industrial Hygiene and Toxicology, F.A. Patty, Intersci-
ence Publishers, Inc.
Industrial Safety, National Safety Congress, 1957.
Industrial Toxicology, L.T. Fairhall, William & Wilkins
Company.
International Critical Tables, National Research Coun-
cil, McGraw-Hill Book Company, Inc.
Manufacturing Chemists' Association, Inc. (Chemical'
Safety Data Sheets).
Matheson Gas Data Book, The Matheson Company.
McGraw-Hill Encyclopedia of Science and Technology.
National Fire Codes, National Fire Protection Associa-
tion.
Safety in the Chemical Laboratory, Peters and
Creyghton, Butterworths Scientific Publications.
Standard Methods of Chemical Analysis, F. Furman,
Van Nostrand Company, Inc.
Thorpe's Dictionary of Applied Chemistry, John H.
Wiley and Sons, Inc.
Underwriters' Laboratories, Inc.
U.S. Coast Guard Regulations.
U.S. Department of Transportation, Regulations for
Transportation of Explosives and Other Dangerous
Articles.
38

-------
Part II
UNDERGROUND
STORAGE SYSTEMS
INTRODUCTION
Part II of this report describes, the components and
concerns associated with underground storage facilites
for hazardous liquids.
The basic principles shown in figure 2-1 are appli-
cable to all underground storage systems. However, be-
cause chemicals are highly variable in the characteristics
and the risks they present, underground storage may not
always be an appropriate storage practice, or it may be
appropriate for the design engineer to consider other
practices not illustrated in this generalized drawing.
The items covered in detail in the following chap-
ters include (1) the types of storage tanks available; (2)
piping and pumping system components and their per-
formance; (3) underground spill containment systems;
(4) the types of overfill prevention systems and their
performance; (5) leak monitoring and surveillance; (6)
the testing and inspection of underground storage sys-
tems; and (7) the closure and abandonment of under-
ground storage facilities.
A schematic diagram showing the key components
of an underground storage system is presented in Figure
2-1.
40

-------
Figure 2.1
Elements of an Underground Storage
Tank Installation
\
TANK TRUCK
(a) OVERFILL PREVENTION DEVICE 0
VAPOR RECOVERY LINE
© .
FLOAT VENT VALVE

PROOUCT
DISPENSER
EXCAVATION
CAP
OBSERVATION _
WELL © N
PEA GRAVEL
OR SAND FILL
©AUTOMATIC
SHUTOFF
VALVE
PROOUCT DELIVERY LINE
EXCAVATION WALLS ANO
FLOORS OF IMPERVIOUS
MATERIAL
-LEAK OETECTOR
0 SUBMERGED PUMPASSEMBLY
Weil designed underground storage systems usually contain the following:
I) corrosion resistant tank; 2) striker plate under tank fill line; 3) submerged pump with leak detector on prod-
uct delivery line; 4) float vent valve in tank vent line; 5) excavation walls and floor of impervious material;
6) asphalt or concrete excavation cap; 7) automatic shutoff valve on delivery line at pump island; 8) overfill
prevention device at fill line on tank truck; 9) vapor recovery in tank truck during filling operation; 10) observa-
tion wells located inside excavation boundaries; 11) pea gravel or sand fill for excavation.
These are all important aspects of a good underground storage system.
41

-------
Part II
CHAPTER 1:
UNDERGROUND
STORAGE TANKS
A. INTRODUCTION
Several types of underground storage tanks or stor-
age systems are available for use in today's market.
These include the following:
•Bare steel tanks.
•Steel tanks with coatings.
•Cathodically protected steel tanks (galvanic pro-
tection).
•Cathodically protected steel tanks (impressed cur-
rent protection).
•Fiberglass reinforced plastic (FRP) tanks.
•FRP/steel bonded tanks.
•Double containment systems, such as
•	Double-walled tanks,
•	Vaulted tank storage system, and
•	Impermeable liners.
•Relined tanks.
•Tanks which combine several design features such
as cathodically protected double walled steel tanks
or double walled steel tanks with a fiberglass
bonded-outer shell are available on custom order
from many manufacturers.
Summaries of the characteristics and limitations of
these various types of tanks and storage systems are
presented in Table 2.1-1.
B. TANK LAYOUT
Idealized layouts for underground storage facilities
are illustrated in Figures 2.1-1 and 2.1-2. The diagram
of Figure 2.1-1 depicts a tank equipped with a suction
pump while the illustration in Figure 2.1-2 is that of a
tank equipped with a submerged pump. The discussion
in chapter 2 explains these pumping systems in more
detail.
It is highly desirable for the owner to prepare and
keep at the storage site, a plot plan which shows the
layout of the facility. The plot plan should show age of
tanks, material of construction, depth and location of
pipe galleries, chemical stored in tank and phone
number and address of person to contact in case of an
emergency.
The tank connections shown in Figures 2.1-1 and
2.1-2 include the following:
•A fill and gauge tube.
•A vent line, with a float vent valve installed at the
vent line tank connection for overfill prevention.
•A manhole fitted over the fill and gauge tube to
permit easy access to the tube.
In addition. Figure 2.1-2 shows a manhole over the
pump manifold assembly to permit access for mainte-
nance. and a leak detector mounted on the pump man-
ifold, which detects leaks in the product supply line.
Leak detectors are required by NFPA 30.
The utility and economics of providing a storage
tank with a manhole are subject to debate. They add ap-
proximately $500 to the cost of a tank. A tank can be
cut open and entered for inspection for about 57(30 and
effectively resealed. However, all tanks may not require
inspection in a normal life and only a small fraction
may require re-entry.
Manholes have the advantage of allowing internal
inspection of emptied tanks and are especially useful on
fiberglass tanks where access for measuring diameter
deflections is important. However, Manhole gaskets can
leak, resulting in stormwater entry and can give a false
reading during a leak test. Air pockets formed in the
manhole area can interfere with leak testing, though this
condition can be corrected with a properly placed
bleeder valve.
All tanks, and especially FRP tanks, should be pro-
vided with a striker plate under the fill line. The striker
plate is a heavy metal plate attached to the bottom of
the tank which absorbs the shock of the dip stick when
it is dropped into a tank to measure liquid level. It is
becoming common practice to have striker plates under
all tank openings.
42

-------
Table 2.1-1





Characteristics of Underground Storage Tanks








Remarks



Types of Soil
Major Causes Relative
(advantages and
Type
Description
Applications*
Suitable
of Leaks Costs
disadvantages)
Bare steel
Carbon steel
Compatible with
Not compatible
Corrosion Low
Fifty percent of the bare


fuels and a num-
with corrosive

bare steel tanks leak


ber of other
soil.

after 15 years. Life


chemical prod-


expectancy is dependent


ucts.


on soil corrosivity and


Not compatible


method of installation.


with corrosive





liquids such as





acids.



Coated/lined
Carbon steel
Generally compa-
Generally compa-
Corrosion Low
Coating/ lining must be
steel
with exterior
tible with
tible with cor-
due to de-
properly applied and

coating and/
corrosive chemi-
rosive soil if ex-
fects in coat-
free of defects

or interior
cals such as alkalis
ternal coating is
ing or lining.
(holidays), the effec-

lining.
and organic and
applied.

tiveness of th5 coating/


inorganic acids if


lining will vary with the


internal lining is


type of coating. Internal


applied.


lining can increase life




span of tanks.
Pre-engin-
Steel tank
Compatible with
Can withstand
Internal Medium
Life expectancy of these
eered catho-
with pre-
gasoline, diesel
corrosion in soils
corrosion
tanks is difficult to
dicaily pro-
engineered
fuel, kerosene.
with resistivities

predict but the record
tected steel
corrosion pro-
bunker oil and a
greater than 2000

for the fifteen years that
tanks-galva-
tection con-
number of other
ohm-cm [26].

the tank has been avail-
nic protec-
sisting of sac-
chemical products.


able is impressive.
tion (e.g. Sti-
crificial




P3 and B-IO
anodes, pro-




tanks)
tective coal-





ing and electri-





cal isolation.




Cathodically
Steel tanks
Petroleum pro-
Will withstand
Internal Medium
Good life expectancy if
protected
to which a
ducts and a num-
highly corrosive
corrosion
the cathodic protection
steel tanks-
constant
ber of other
soil if properly

is properly maintained.
impressed
supply of
chemical products.
designed.


current
electric cur-





rent is applied




Vaulted
Tanks are
frequently used
Generally resis-
Low risk of High
Poorly designed con-
tank
installed in
for secondary
tant to soil
leaks
crete vaults are sus- '

concrete
containment of
corrosion.

ceptible to cracking and

valults to
highly hazardous


chemical attack by salts

provide sec-
chemicals.


and acids. Porosity of

condary con-



concrete is a problem.

tainment of





leaks. Vaults





sometimes have




interior coat-





ings and exter-





nal poly-





ethylene wrap-





per to pre-





vent perme-





ation through



,

concrete.




43

-------
Table 2.1-1 continued
Impermeable
liners
Involves the
use of an
impermeable
liner as sec-
ondary con-
tainment in
the tank ex-
cavation.
Examples in-
clude mem-
brane. clay,
and bentonite
liner.
See Part 11.	See Part 11.	Low risk Medium Care must be taken to
Chapter 4	Chapter 4	of leaks	insure that lining
material is compatible
with stored material.
Should include a leak
detection system with-
in the confines of the
liner containment area.
Relined
tanks
Existing steel
tank relined
with corro-
sion resistant
material.
Petroleum pro-
ducts and corro-
sive chemicals.
Used to extend
the lifespan of
underground tanks.
Bare steel tanks
with interior liners
will continue to
corrode in corro-
sive soil.
Defects in Low Condition of the tank is a
linine	key consideration. It is
important that the reiiryng
material be compatible with
the material to be stored
and that workmanship be
according to the API
standards.
Fiberglass
reinforced
plastic
(FRP)
FRP/steel
Double-
walled
Plastic
resins rein-
forced with
glass fiber.
Petroleum and a
number of other
chemical products.
Suitable in highly Tank
corrosive soils. rupture
Outer FRP
Petroleum
Resistance to soil
Low risk
layer fused
products and a
corrosion is com-
of leaks.
to an inner
number of other
parable to that of
but tank is
layer of steel
chemical products.
fiberglass tanks.
susceptible
by a poly-


to internal
ester resin


corrosion.
bond.



Tank within
Applications are
Suitable in highly
Low risk
a tank with
dependent upon
corrosive soils.
of leaks
a vacuum or
the materials of
depending on the

pressurized
construction. See
materials of con-

space be-
descriptions of
struction.

tween the'
FRP and coated.


inner and
galvanically pro-


outer walls.
tected steel tanks


Currently
above.


manufactured



doubled-walled



tanks are com-



posed of either



steel (with



coating and



galvanic catho-



dic protection



on outer wall)



or FRP.



Medium FRP tanks cannot with-
stand loads as does steel
and may easily be dam-
aged if dropped, mis-
handled or subjected to
excessive loads because
of improper installation.
Medium Combines strength of
steel with corrosion
resistance of fiberglass.
High Sonte models only avail-
able in capacities up to
to 4000 gallons. These
tanks usually include a
built-in leak detection
system located between
the inner and outer
walls.
• Refer to Table 1.2-8 and Appendix A lor more tnforamtion on chemical compatibility.
44

-------
Figure 2.1-1
Tank Piping Details - Suction System
Source: API Publication 1615, Installation of Underground Petroleum Storage Systems, 1979.
Figure 2.1-2
Tank Piping Details - Submerged System
UAMuni e VENT LINE TO APPROPRIATE
manhole LOCATION (SLOPE TO TANK)
MANHOLE ~ I, FILLCAP \
jf ^REtNF. CONC. SLAB \
•j». • .• *. *. * '..Vila" *. .*« i/jar*. • . . . r-.¦•-.-••rTs \
LEAK wE TEC fCR	Trt^~ '•) f ...
t>*.J -CLAY TILE \
p.! INi—GRAVEL OR ABSORBENT MATERIAL
[s: -SILL PIPE rP«	
\ "1.
MUPPLr LINE TO'1®
PUMP ISLANDS
pump & mo tor .
k
-t"		 \ ¦' jr
,, II \ PLUG' NIPPLE/7 w
|l T ^BUSHING	'
" 1
M "I
1 ' GASOLINE STORAGE TANK
4; j!	F.LL TUOE
* T 6"MAX
• 1 min !
—-OVERFILL PREVENTION
FLOAT VEN T VALVE
		 SANO OR GRAVEL
18'"** / EXIST SOIL
min'V (UN0ISTUR8E0)
. ^
%
aN?
1 i2" • • ' sr '
"%zrx.^
Source: API Publication 1615, Installation of Underground Petroleum Storage Systems, 1979.
45

-------
C. TYPES OF UNDERGROUND
STORAGE TANKS
1. Bare Steel Tanks
Bare steel tanks constructed of mild carbon steel
may be used in non-corrosive soil environments to store
non-corrosive materials, such as gasoline and other pe-
troleum derivative products. The compatibility of steel
with various petroleum and chemical products is dis-
cussed in Part I and Appendix B. The degree of en-
vironmental protection provided by bare steel tanks is
minimal and consequently their use has declined in re-
cent years.
Please note that many bare steel tanks have a thin
surface coating to protect against rust. This is essen-
tially a cosmetic coating, and should not be confused
with more substantial corrosion protection coatings.
Design Standards. The capacities, dimensions and
construction details for bare steel tanks generally follow
established standards. These include the following:
Underwriter Laboratories (UL) Inc.:
UL 58 - Steel Underground Tanks for Flamma-
ble and combustible Liquids [6].
National Fire Protection Association (NFPA):
NFPA 30 - flammable and combustible Liquids
Code [19].
American Petroleum Institute (API):
API Publication 1602 - Recommended Standard
for Underground Gasoline Tanks [21].
API Publication 1611 - Service Station Tankage
Guide [20].
API Publication 1615 - Installation of Under-
ground Petroleum Storage Systems [1].
American Society of Mechanical Engineers:
ASME/Pressure Vessel Code, Section VIII/[ 14].
Of these standards, the Underwriters Laboratories
standards are the most detailed in that they specify
many of the tank design details. These include steel
thickness, tank head design, bracing requirements for
multi-compartment tanks, the sizes of vent connections,
and tank marking and testing requirements.
The American Society of Mechanical Engineers
Pressure Vessel Code may be used for storage tanks in-
tended for industrial service.
The thicknesses of horizontal, atmospheric-type
steel tanks of various capacities as recommended in UL
58 are shown in Table 2.1-2. This standard also recom-
mends that the length of the tank be no more than six
times its diameter [6]. As a source of reference, the ca-
pacity per foot of length for tanks having diameters of
24 to 144 inches is given in Table 2.1-3.
Installation of Underground Steel Tanks.
Sources of information and recommendations of installa-
tion practices for underground storage tanks include API
Publication 1615 [1], NFPA 30 [19] and the New York
State Department of Environmental Conservation (NYS-
DEC) manual covering standards of practice for bulk
storage of hazardous liquids [30]. Most manufacturers
supply step-by-step procedures for tank installation and
require that these steps be followed to validate the
guarantees and warrantees.
The installation recommendations given in the API
publication identify tank clearance, depth of excavation,
and anchoring and backfilling requirements. Examples
of the recommendations given in API Publication 1615
include the following:
•At least 6 inches and preferably 12 inches of well-
compacted sand or gravel placed underneath the
tank.
•A minimum tank clearance of 12 inches in ail
horizontal directions.
•In areas not subject to traffic, the cover depth
should be a minimum of 24 inches, or not less
than 12 inches plus a reinforced concrete slab not
less than 4 inches in thickness.
•Where tanks are subject to traffic, cover depths
should be a minimum of 36 inches, or not less
than 18 inches of well-tamped material plus at
least 6 inches of reinforced concrete or 8 inches
of asphaltic concrete.
It should be noted that the burial depth of a tank
is dependent upon several factors, including local regu-
lations, the type of finished surface to be applied, soil
conditions, topography, and suction pumping lift re-
quirement [1].
The recommendations of NFPA 30 concur with
those of API Publication 1615, with the addition that
steel underground tanks shall be set on firm foundations
and surrounded with at least 6 inches of noncorrosive
inert material, such as clean sand or gravel, well-
tamped in place [19].
The backfill for steel tanks is typically a clean,
noncorrosive, porous material such as clean washed
sand or gravel. Backfilling operations are very impor-
tant to the life of the installation. It is important that the
backfill be well compacted to avoid undue stresses on
the tank. Application and compaction of the backfill in
layers is often specified to chieve optimum compaction
(eliminate voids in the backfill).
46

-------
Table 2.1-2
Thickness of Steel Tanks
Manufacturers	Nominal Thickness
Maximum*	Standard or
Capacity			Diameter	 Galvanized Sheet	Uncoated ~ Galvanize?
U. S. Gallons
dm3
Inches
m
Gage No.
Inches
mm
Inches
mm
Up to 285
Up to 1078
42
1.07
14
0.075
1.91
0.079
2.01
286 to 560
1082 to 2120
48
1.22
12
0.105
2.67
0.108
2.74
561 to 1100
2124 to 4164
64
1.63
10
0.135
3.43
0.138
3.51
1101 to 4000
4168 to 15142
84
2.13
7
0.179
4.55


4001 to 12,000
15145 to 45425
126
3.20
1/4 inch
0.250
6.35


12,001 to 20,000
45429 to 75708
144
3.66
5/16 inch
0.312
7.9 2


20,001 to 50,000
75712 to 189270
144
3.66
3/8 inch
0.375
9.53


-J
* Length of tank shall be not greater than 6 times the diameter
Source: This material is based on and taken, with permission, from Underwriters Laboratories Inc.
Standard for Safety for Steel Underground Tanks for Flammable and Combustible Liquids, UL 58,
Copyright 1976 (by Underwriters Laboratories Inc.), copies of which may be purchased from Under-
writers Laboratories, Inc., Publication Stock, 333 Pfingsten Road, Northbrook, Illinois 60062.
Note:
UL shall not be responsible to anyone for the use of or reliance upon a UL Standard by anyone.
UL shall not incur any obligation or liability for damages, including consequential damages,
arising out of or in connection with the use, interpretation of, or reliance upon a UL Standard.

-------
Table 2.1-3
Gallon Capacity per Foot of Length
Dia-
U.S.
meter
Gallons
in
1-foot
Inches
lenqth
24
23.50
25
25.50
26
27! 58
27
29.74
28
31.99
29
34.31
30
36.72
31
39.21
32
41.78
33
44.43
34
47.16
35
49.98
36
52.88
37
55.86
38
58.92
39
62.06
40
65.28
41
68.58
42
71.97
43
75.44
44
78.99
45
82.62
46
86.33
47
90.13
48
94.00
49
97.96
50
102.00
51
106.12
52
110.32
53
114.61
54
118.97
55
123.42
56
127.95
57
132.56
58
137.25
59
142.02
60
146.88
61
151.82
62
156.83
63
161.93
64
167.12
Dia-
U.S.
meter
Gallons
in
1-foot
Inches
Lenqth
65
172.38
66
177.72
67
183.15
68
188.66
69
194.25
70
199.92
71
205.67
72
211.51
73
217.42
74
223.42
75
229.50
76
235.56
77
241.90
78
248.23
79
254.63
80
261.12
81
267.69
82
274.34
83
281.07
84
287.88
85
294.78
86
301.76
87
308.81
88
315.95
89
323.18
90
330.48
91
337.86
92
345.33
93
352.88
94
360.51
95
368.22
96
376.01
97
383.89
98
391.84
99
399.88
100
408.00
101
416.00
102
424.48
103
433.10
104
441.80
Dia-
U.S.
meter
Gallons
in
1-foot
Inches
Lenqth
105
449.82
106
458.30
107
467.70
108
475.89
109
485.00
110
493.70
111
502.70
112
511.90
113
521.40
114
530.24
115
540.00
116
549.50
117
558.51
118
568.00
119
577.80
120
-587.52
121
597.70
122
607.27
123
617.26
124
627.00
125 .
638.20
126
647.74
127
658.60
128
668.47
129
678.95
130
690.30
131
700.17
132
710.90
133
721.71
134
732.60
135
743.58
136
754.64
137
765.78
138
776.99
139
788.30
140
799.68
141
811.14
142
822.69
143
834.32
144
846.03
Source: This material is based on and taken, with permission, from Underwriters Laboratories Inc. Standard for Safety for Steel Underground Tanks for
Flammable and combustible Liquids. UL 58. copyright 1976 (by Underwriters Laboratories Inc.), copies of which may be purchased from Underwriter
Laboratories. Inc.. Publication Stock. 333 Pfingsten Road. Nonhbrook. Illinois 60062.
Note: UL shall not be responsible to anyone for the use of or reliance upon a UL Standard by anyone. UL shall not incur any obligation or liability
for damages, including consequential damages, arising out of or in connection with the use. interpretation of. or reliance upon a UL Standard'
48

-------
It is also recommended in API Publication 1615
that tanks be ballasted with the product as soon as pos-
sible after backfilling. Water ballast may be used as an
alternative, but it is necessary to defer installation of
submerged pumping units in the tank until after the
water ballast is removed. If ballasting is necessary in
order to prevent tank flotation (from a high water table
or from rain), the product to be stored should be used
as a first choice [1].
When a high water table is present, anchoring
should be used to prevent tank flotation. A concrete
slab is often used to anchor underground tanks as shown
in Figure 2.1-3. When such a concrete slab is used,
tanks should be separated from the slab by no less than
12 inches of compacted sand. Tanks should not be set
directly on the concrete nor placed on hard or sharp ma-
terial that could cause deformation or damage to the
tank or tank coatings. Anchor straps should be installed
so as not to damage the tank or tank coating. Material
such as asbestos felt or pieces of rubber tire should be
placed between the tank and the anchor straps to pro-
vide electrical isolation [1].
For complete information on the installation of under-
ground steel tanks, the reader is directed to the NYS-
DEC standards of practice document, NFPA 30, API
Publication 1615 and other sources, such as Occupa-
tional Safety and Health Administration regulations (29
CFR, Part 1910, Section 1910.106) [29] and specific
manufacturer recommendations.
Characteristics of Carbon Steel. Carbon steel is
the most common, most versatile and least costly metal
used in industry. It is two-thirds the weight of lead and
three times heavier than aluminum [14]. Carbon steel
may be annealed (i.e. heated and then cooled) to make
it stronger and more flexible, and galvanized (coated
with zinc) to improve its corrosion resistance. The
mechanical properties of carbon steel are strongly influ-
enced by the carbon content.
Over the years, various types of carbon steel have
been developed, for example, structural and pressure
vessel steels. There are only minor metallurgical differ-
ences between these types of steel; the important differ-
ences are in the quality of the steel (resulting from
adherence to tighter specifications).
Figure 2.1-3
Anchoring of Tanks Installed in
High Groundwater Tables
,aURY PER "NFPA'
SPECIFICATIONS
&SAN0 Oft
GRAVEL (TYP.J
GA8LE OR
STRAP (TYP-)
firm soil
SCREW OR EXPANOABlJ
TYPE ANCHOR
,90 LB. ASBESTOS FELT BETWEEN
/ CABLE a TANK ( OR RUBBER TIRE
/ TYf? )		
3RA0E
ANOiOR
CONCRETE SELL-
(OR OEAOMAN )
ANCHOR
¦i- i* <•
//r \enmcjr*TP
T
CONCRETE SLAB
NOTE- SEE MANUFACTURERS RECOMMENOATIONS FOR
ANCHOR ANO INSTALLATION INSTRUCTIONS.
NO SCALE
A PEA GRAVEL FOR
NON - METALLIC TANKS
(OR MFR. APPROVED ALTERNATE}
Source: Reference 2
49

-------
There are a number of standards and specifications
for carbon steel in various forms, such as in the form
of bars, pipe and plate. The American Society for Test-
ing and Materials (ASTM) publishes specifications on
many materials of construction including carbon steels.
For detailed specifications and chemical analyses, refer-
ences should be made to these ASTM standards. The
American Iron and Steel Institute also issues specifica-
tions on a variety of carbon and alloy steels. The Amer-
ican Society of Mechanical Engineers, the American
National Standards Institute, and the American Petro-
leum Institute are also active in the area [14]. For more
information on these steel specifications and how they
may be obtained, please refer to the list of steel specifi-
cation references at the end of this chapter.
2.	Coated Steel Tanks
Organic coatings may be applied to both the inter-
ior and exterior of underground steel tanks. Interior
coatings are often called tank linings. In the case of
shop-assembled tanks, coatings and linings are generally
applied at the factory. The recommendations of the tank
manufacturer should be followed when a coating is re-
quired since improper selection can lead to early failure
and product contamination. When installing the tank,
care must be exercised in order to avoid damage to the
coating. The properties, compatabilities and costs of
common organic coatings are given in Appendix C.
Refer, to Section D of this qhapter for additional infor-
mation on application of coatings and linings.
3.	Cathodically Protected Steel Tanks —
Galvanic Protection
As described in Part I, Chapter 1, Section C of this
report, cathodic protection is used to reduce or eliminate
corrosion of a metallic structure which is in contact with
corrosive soil. This is accomplished by applying an
electric current to the structure which is greater in
strength and opposite in direction to the current that is
causing corrosion.
The galvanic cathodic protection method employs
sacrificial anodes, composed of materials such as mag-
nesium or zinc, in electrical contact with the metal
structure to be protected. These anodes are attached to
the surface of the protected material (tank or pipe) in
the soil or other electrolytic solution, and the required
current is generated by corrosion of the sacrificial
anode. A typical configuration for galvanic protection is
shown in Figure 2.1-4.
The design of an adequate galvanic protection sys-
tem requires making a measurement of the soil resistiv-
ity. If the amount of electric current required to protect
the tank has been determined, the soil resistivity must
be known in order to determine the type and size of
anode(s) required to protect the tank. The life expec-
tancy of the storage system is also important in deter-
mining the number and type of anodes required [22].
Magnesium anodes are the most common type of
sacrificial anode, although zinc anodes may be used in
soils with resistivities less than 1000 ohm-cm. Magne-
sium, because of its higher driving voltage, can be used
quite effectively in soils with resistivities up to 5000
ohm-cm, and on well coated structures can often be
used up to 10,000 ohm-cm or more [10, 12, 27].
In general, because of the low driving voltages of
sacrificial anodes (1.1-1.6 volts) and the low electric
currents generated (usually less than 100 milli-amperes
per anode), it is desirable from both an economic and
an engineering standpoint that galvanically protected
tanks be coated. Bare tanks require a greater electric
current, and hence a larger number of sacrifical anodes
than coated tanks.
Periodic testing of cathodic protection is essen-
tial if the system is to function properly and provide
long-term protection. The current from the anodes may
fail because of anode deterioration or broken lead wires.
Changes in underground conditions (e.g., installation of
a water pipe) or coating deterioration can also change
protective current requirements. Measurements of tank-
to-soil potentials and anode output should be made at
least once a year to ensure proper operation of the sys-
tem [10].
In addition, care must be exercised during the in-
stallation of galvanic protection systems to ensure elec-
trical continuity of the system. This means providing
bonding wires between tanks when several tanks are in-
stalled and across flexible pipe joints, if such joints are
used. Screwed piping should not be relied upon to pro-
vide electrical continuity.
Examples of pre-engineered, galvanically protected
steel tanks include the Sti-P3 tank and the BT-I0 tank.
Both are standard steel tanks provided with three levels
of corrosion protection: cathodic protection, a protective
coating and electrical isolation. High-potential magne-
sium anodes are permanently attached to the heads of
each tank to provide a flow of protective current. In the
case of Sti-P3 tanks, the anodes are packaged in a spe-
cial moisture-holding material which improves conduc-
tivity and current flow from the anodes. The second
protective component in these systems is a coal-tar
epoxy or urethane coating. Electrical isolation is the
third component of these systems; this protects the tanks
against stray currents that could otherwise reach them
via piping connections. In those areas where internal
corrosion may be a problem, optional construction may
include striker plates, internal welding or internal zinc
strips which serve as sacrificial anodes. If the product
to be stored is not compatible with steel, then an inter-
nal lining of compatible material may be applied.
50

-------
Figure 2.1-4
Magnesium Anode Cathodic Protection
Typical Configuration
Test Box
Source: Suggested Ways to Meet Corrosion Protection Codes for Underground Tanks and Piping, The Hinchman
Company, Detroit, MI.
51

-------
4. Cathodically Protected Steel Tanks
(Impressed Currents)
The impressed current cathodic protection method
employs direct current provided by an external source.
This current is passed through the system by the use of
non-sacrifical anodes composed of materials such as
carbon, non-corrodible alloys, or platinum. These
anodes are buried in the ground (in the case of under-
ground structures) or otherwise suspended in the electro-
lyte and connected to the positive terminal of the exter-
nal power supply. The tanks and other structures to be
protected (e.g., pipes) are connected to the negative
side of that power supply [27]. An impressed current
system for underground tanks and piping is illustrated in
Figure 2.1-5.
Impressed current cathodic protection systems are
used extensively at service stations. These types of
cathodic protection systems are particularly applicable
for storage situations in highly corrosive soils. Because
of the large power supply (electric current) provided by
these systems, they can be used to protect bare as well
as coated tanks [10].
A major advantage of impressed current is that
short circuits can be overcome more easily than with
sacrifical anode systems [10]. This facilitates installa-
tion, particularly when electrical continuity must be in-
sured between two or more tanks. Major disadvantages
of these systems are their high power consumption and
the greater possibility of electrical interference on for-
eign structures 110].
As is the case with sacrificial anode systems,
periodic testing of the cathodic protection is necessary
to ensure proper protection. Current may fail because of
rectifier malfunction or interruption of power. The sys-
tem should be tested regularly in accordance with man-
ufacturer's recommendations and adjusted as needed. At
least once a year, tank-to-soil potential measurements
should be made to check the adequacy of protection and
determine if any rectifier adjustments are needed [10].
5. Fiberglass-Reinforced Plastic
Fiberglass-reinforced plastic (FRP) tanks are widely
used for underground storage of flammable and combus-
tible liquids. They are constructed of a plastic resin
which provides chemical resistance, and a fiberglass
material that gives the tank its structural strength. Insur-
ing compatibility of the tank with the stored product is
an important consideration since numerous resins and
glass materials can be used in the fabrication of FRP
tanks. Most fiberglass tanks are designed specifically
for petroleum and its derivatives. However, fiberglass
tanks suitable for storage of other chemicals have been
developed. The tank manufacturer should be consulted
on the selection of a resin which will be compatible
with the product to be contained (see Table 1.2-3).
Two techniques are used to fabricate FRP tanks.
One technique utilizes a centrifugal casting machine
which allows the tank to be made in one continuous
piece. The chopped fiberglass is sprayed on the interior
of a revolving mold which forms the ribs, shell and
hemispheres in one continuous piece. Two identical
pieces can be fabricated by this technique and joined to-
gether in the middle to form the tank. The other fabrica-
tion technique consists of building the tank in alternat-
ing layers of resin and fiberglass. This type of FRP tank
is usually more costly but is also stronger than the
molded tank [13, 15, 29].
Design Standards. As with steel tanks, various
standards have been developed by Underwriters Labora-
tories and the American Society of Testing and Mate-
rials for the design and construction of FRP storage
tanks. For more information on accessing these stan-
dards, please refer to the tank design reference list at
the end of this chapter.
In addition to the UL and ASTM standards. FRP
tanks to be used in underground applications must
adhere to requirements in the following National Fire
Protection Association standards:
•NFPA 30: Flammable and Combustible Liquids
Code [19],
•NFPA 31: Standards for Installation of Oil Burn-
ing Equipment [28].
Installation Requirements. The installation re-
commendations for FRP tanks, as listed in API Publica-
tion 1615, differ somewhat from those recommendations
for steel tanks. Examples of the recommendations made
in that publication for the underground installation of
FRP tanks include the following:
•The tank excavation should provide a minimum
clearance in all horizontal directions of 18 inches.
•The excavation should be deep enough to provide
at least 12 inches of backfill below the tank.
•A uniformly distributed backfill, which conforms
to the tank manufacturer's specifications, must be
used. Proper backfilling is essential to the perfor-
mance of these tanks [1].
As is the case with steel tanks, proper anchoring
and ballasting are important aspects of an FRP tank in-
stallation in an area of high ground water. Strict adher-
ence to manufacturers installation recommendations and
those of API Publication 1615 is important to insure the
integrity of the storage facility.
Extreme care must be exercised in the installation
of FRP tanks because they lack the structural strength
to withstand the high stresses which may be induced
during a difficult or improper installation [8]. For more
complete information of FRP tank installation, refer to
API Publication 1615, NFPA 30, the NYSDEC stan-
dards of practive document [30], and the literature of
specific FRP tank manufacturers.
52

-------
Figure 2.1-5
Impressed Current Cathodic Protection
Typical Configuration
Test Box
Positive Header
Cable
Source: Suggested Ways to Meet Corrosion Protection Codes for Underground Tanks and Piping, The Hinchman
Company, Detroit, MI.
53

-------
6.FRP/Steel	Bonded Tanks
FRP/steel tanks combine the corrosion resistance of
fiberglass-reinforced plastic and the strength of steel.
They are constructed of an outer layer of FRP fused to
an inner layer of carbon steel via a polyester resin bond.
FRP/steel tanks are protected against soil corrosion, but
remain subject to internal decay from corrosive chemi-
cals [7].
The steel inner structure on an FRP/steel tank pro-
vides structural support and serves to keep stresses
evenly spread. FRP/steel tanks may be installed in ac-
cordance with NFPA 30 and API Publication 1615
guidelines (see sections on bare steel and FRP tanks).
Saddles or "chock blocks", which interfere with the
proper distribution of the load, should not be used,
however. In addition, anchoring, as described in the
discussion of steel tanks, should be used to prevent tank
flotation from a high water table.
As is true with steel tanks, FRP/steel tanks are
compatible with petroleum products, such as gasoline
and diesel fuel, and other non-corrosive liquids.
7.	Tanks of Other Materials
This group includes tanks made of such materials
as stainless steel, aluminum and plastic. Although these
materials have a higher resistance to corrosion than car-
bon steel, their use is overshadowed by that of steel,
coated steel and FRP. Plastic tanks, including such ma-
terials as polyvinyl chloride (PVC) and polypropylene,
are not widely used in underground installations because
of their low structural strength, which makes them un-
able to withstand large structural loads. Aluminum tanks
are not widely used because they also lack structural
strength, and stainless steel tanks are not widely used
because of their higher cost.
8. Double Containment Systems
Several methods of double containment for under-
ground tanks are in use. These include the following.
•Double-walled tanks.
•Concrete vaults.
•Impermeable liners.
Double-walled tanks [4.23]. These tanks are es-
sentially a tank within a tank (jacket) with a vacuum or
pressurized space between the inner wall and outer wall.
Leaks due to internal or external corrosion can be de-
tected by loss of pressure or vacuum. Product or water
detecting probes may also be inserted into interstitial
space. Common materials of construction include coated
steel and fiberglass. An inner liner may also be specified
specified for steel tanks. Because double-walled tanks
provide both two wall protection and monitoring of the
interstitial (annular) space, they are well suited for stor-
ing highly toxic chemicals or for storing materials in
sensitive environmental areas.
A double-walled fiberglass tank that is widely used
in Europe is illustrated in Figure 2.1-6. This tank is
constructed with inner and outer fiberglass shells sup-
ported by a concrete bearing wall in between. A built-in
leak detection system monitors a vacuum drawn be-
tween the inner and outer shells. Products that may be
stored in this tank include gasoline, diesel fuel, acid and
caustic solutions, and other hazardous substances [4],
Another type of double-walled tank, manufactured
in Canada, is fabricated of steel and includes a vacuum
leak detection system between the inner and outer walls.
This pre-engineered tank system also includes an exter-
nal epoxy coating and sacrificial anodes to provide cor-
rosion protection. This tank system is shown in Figure
2.1-7 [23].
More and more firms are producing double walled
tanks, some with two walls of steel and an outer layer
of bonded fiberglass, some with a complete outer shell
of steel and epoxy coating, and others with just a dou-
ble bottom to provide protection where corrosion is usu-
ally most severe. The high degree of environmental pro-
tection provided means that their usage will be more
common in the near future.
Concrete vaults. Concrete vaults (also knows as
concrete tanks) are generally used as secondary enclo-
sures intended to contain any spills from the primary
storage tank. Concrete vaults tend to crack with freezing
and thawing, and are also susceptible to chemical attack
by salts and acids. Coatings are often applied to the in-
side of concrete vaults to enhance their resistance to
Chemical attack (see Coating Compatibility Chart Ap-
pendix B).
Impermeable liners: Impermeable liners may also
be used for secondary containment of underground
spills. Examples of such liners include the following:
•Membrane liners.
•Clay liners.
•Bentonite (or similar material) liners.
In instances where impermeable liners of these
types are used, care must be taken to insure that the lin-
ing material is compatible with the material being
stored.
Impermeable liner systems should also include a
leak detection system located within the confines of the
enclosure formed by the liner. These types of systems
are discussed in detail in Chapter 3 of this part of the
report.
54

-------
Figure 2.1-6
Fiberglass-Reinforced Plastic
Double-Walled Tank With Built-in Leak Detection
	\
Leakage warning system < a Suction line
V. 9 Measuring line
10	Mcnway
11	Striker plate
Source: Betco Associates
55

-------
Figure 2.1-7
Double Wall Steel Tank with Epoxy Coating
and Sacriflcal Zinc Anode
Zinc
anode
"Anchor
streps
"Piping
n <**11
Internal
access
t
2
nnertank
wall
5
Fiilspill
collector
Manhole
	
mc>«< fast*
Tank
connection
Epoxy '
coating
Drop-
tube
Vacuum area
Outer jacket
Concrete slab
Model
Capacity
tnalde
Cnaide
No.
Litres
Diameter
Length
1352
5 000
1 600
2 500
1353
10 000
2 000
3 185
1354
15 000
2 500
3 060
1355
25 000
2 500
5 100
1356
35 000
2 500
7 130
1357
50 000
2 500
10 200
1358
SO 000
3 600
U 920
1359
75 000
3 000
7 080
1360
100 000
3 600
9 830
1361
125 000
3 600
12 290
1362
150 000
3 600
1*. 740
Source: Reference 23
56

-------
9. Relined Tanks
D. TANK COATINGS AND LININGS
In-place underground steel tanks may be relined in-
ternally provided the tank is not badly corroded. This
is a widely used practice for extending the useful life
of steel petroleum storage tanks. If consideration is
being given to the relining of a tank which stores a non-
petroleum product, the design engineer should secure
assurances that the lining material is compatible with
the product to be stored and check government regula-
tions to see whether this practice is acceptable.
Interior relining is possible without unearthing the
tank by entering it through a manhole. If the tank is not
. equipped with a manhole, one must be installed prior to
relining the tank [18]. Before entering the tank, how-
ever, it must be completely emptied and freed of toxic
or flammable vapors. Refer to the discussions in Part II,
Chapter 7 and Part III, Chapter 7 for information on the
emptying and degassing of tanks. Prior to relining, the
existing lining must be completely removed and the
tank must be properly prepared. Holes must be plugged
and the surface must be sandblasted to etch a pattern for
good bonding. Refer to the sections on the selection and
application of coatings and linings for specific proce-
dures and considerations. The combination of external
cathodic protection with internal lining provides a
reasonably low cost safeguard system for existing steel
tanks and is useful as a repair technique for a leaking
steel tank in generally good condition. The relining
work should be done only by qualified specialists.
The following inset describes suitability factors for
steel tank relining.
Tank Relining
A steel tank is not normally suitable for interior
lining and should be removed or abandoned if
it has one of the following:
•A split greater than 3 inches.
•A single hole greater than 1 inch diameter.
•More than 10 small perforations (none
larger than 'A inch diameter).
Coatings are those corrosion- and chemical-resis-
tant materials which are sprayed, brushed, or rolled
onto the metal surface of a storage tank. Coatings serve
one of two main purposes: (1) they protect the metal
from attack by a corrosive liquid or environment, and
(2) they protect the product from contamination by cor-
rosion products. Coatings may also be applied to con-
crete vaults. When applied to the interior of a tank,
coatings are often referred to as tank linings. Several
factors affect the effectiveness and durability of tank
linings and coatings. These include the following:
•Proper selection of the coating.
•Preparation of the tank surface.
•Proper application of the coating to the required
thickness.
•Proper treatment (curing) of the coating.
•Testing and inspection of the applied coating.
Selection of coating. To insure the compatibility
of the lining or coating, it is often necessary to consult
with its manufacturer. As a reference tool, the proper-
ties of some commonly used linings and coatings are
given in Appendix B.
Preparation of the Tank Surface. Steel, and for
that matter any surface being coated, should be cleaned
of all dirt, grease, moisture and loose powdery conta-
minants that might interfere with coating adhesion. The
best method of steel surface preparation for most coat-
ing application scenarios consists of sandblasting all sur-
faces to be coated to SSPC-SP6 commercial blast.'
Sandblasting to commercial 6 produces a clean surface
with a good profile for adhesion. This combination gen-
erally provides for maximum effectiveness of the chemi-
cal and physical forces of adhesion between the coating
and the metal surface [16,17].
Surface preparation specifications may differ de-
pending upon the type of application to be made, par-
ticularly in the case of retrofits and field oriented opera-
tions. Various specifications from the Steel Structures
Painting Council are listed and described in Table 2.1-4
[16,32],
Coating Application. The following excerpt de-
scribes conventional coating and lining technology:
"The coating system can be applied a
number of ways including brushing, roilering
and spraying. While brushing, and to a lesser
degree roilering, have the advantage of work-
ing a coating into a rough or irregular surface,
spraying is by far the most common application
method.
With conventional air spraying, air is used to
atomize and propel the paint onto the surface
being coated. The equipment is cheaper than
airless equipment. The principal advantages of
air spraying are the ability to partially trigger
the gun to provide an air blow-down prior to
paint application, and a finer atomization re-
sulting in a smoother finish.
57

-------
Airless spraying utilizes high pressure to
hydraulically push paint through a small
orifice. Upon going from high to low pressure
the paint atomizes in a manner similar to water
from a garden hose. The advantages of airless
spraying are leading to its ever increasing use
in industrial coating work. These include a
thicker Film with less chance of air entrapment,
greater mobility for the painter because there is
no air line, less turbulence in the spray pattern,
and less chance of contamination with moisture
and oils from improperly cleaned field com-
pression equipment." [16]
Newer coating and lining application techniques in-
clude electrostatic spraying, powder coating, force dry-
ing and electron beam curing. Their advantages and dis-
advantages are summarized in Table 2.1-5.
Personnel safety is an important consideration in
coating and lining application. In most instances, re-
spirators should be wom as a minimum during applica-
tion operations.
Table 2.1-4
Surface Peparation Specifications
Specification and sub|ect
SSPC-SP 1,	Solvent Cleaning
SSPC-SP 2.	Hand Tool Cleaning
SSPC-SP 3,	Power Tool Cleaning
SSPC-SP 4,	Flame Cleaning of New Steel
SSPC-SP S,	White Metal Blast Cleaning
SSPC-SP 10,	Near-White Blast Cleaning
SSPC-SP 6,	Commercial Blast Cleaning
SSPC-SP 7,	Brush-Off Blast Cleaning
SSPC-SP 8,	Pickling
Purpose
Removal of oil. grease, dirt. soil, salts, and contaminants by cleaning with
solvent, vapor, alkali, emulsion or steam.
Removal of loose rust, loose mill scale, and loose paint to degree specified,
by hand chipping, scraping, sanding and wire brushing.
Removal of loose rust, loose mill scale, and loose paint to degree specified,
by power tool chipping, descaling, sanding, wire brushing and grinding.
Dehydrating and removal of rust, loose mill scale, and some tight mill scale
by use of flame, followed by wire brushing.
Removal ot all visible rust, mill scale, paint and foreign matter by blast
cleaning by wheel or nozzle (dry or wet) using sand, grit or shot. (For very
corrosive atmosphere where high cost ot cleaning is warranted.)
Blast cleaning nearly to White Metal cleanliness, until at least 95% of each
element of surface area is free of all visible residues. (For high humidity,
chemical atmosphere, marine or other corrosive environment.)
Blast cleaning until at least two-thirds of each element of surface area is
free of all visible residues. (For rather severe conditions of exposure.)
Blast cleaning of all except tightly adhering residues of mill scale, rust and
coatings, exposing numerous evenly distributed flecks of underlying metal.
Complete removal of rust and mill scale by acid pickling, duplex pickling or
electrolytic pickling. May passify surface.
Source: Excerpted by special permissin for Chemical Engineering, December 4, 1972, Copyright (c) 1972, by
McGraw-Hill, Inc., New York, N.Y. 10020.
58

-------
Proper Treatment of the Coating. Proper treat-
ment (curing) of the coating surface is an important step
in the coating application procedure. There are numer-
ous coatings available to protect steel and other metals
with curing requirements varying from baking at tem-
peratures on the order of 400°F to force drying at tem-
peratures slightly above ambient. The proper curing or
drying procedure for the various coating materials is
specified in manufacturers' application instructions [17].
Testing and Inspection of the Applied Lining or
Coating. After the coating is installed, it is the general
practice to inspect its thickness and integrity. There are
a variety of field instruments used to measure coating
thickness and porosity after the coating has been applied
and cured. These include dry gauges, such as the mag-
netic and semi-destructive scratch gauges, and the wet
gauge known as the comb type gauge. Low voltage
pinhole detectors (spark tests) may also be used to de-
tect small imperfections in the coating. Other instru-
ments that may be used include surface temperature
thermometers; sling psychromerers, for calculating de-
wpoint and its relation to the surface being coated; sur-
face profile comparators, for blast-cleaned steel sur-
faces; and moisture meters, for concrete and masonry
surfaces [16].
Table 2.1-5
Latest Coating Techniques
Electrostatic Spraying—An electric cnarge 19 aooiied to
the paint by the spray gun. The cnarged paint particles are
attracted toward the grounded obtect being coated, deposit-
ing at points of maximum electrostatic attraction (thin areas).
Can be combined with air and airless spray metnods.
Advantages: Minimizes overspray. nas " wrap-around"
effect edges and protruding irregularities* receive heavier
coatings.
Disadvantages: Requires electric source, electrostatic at-
traction diminishes as paint thickness increases: water oase
paints, or those using mgniy polar solvents or containing
metallic pigments may be too conductive to oe aooiied by
electrostatic spray. The ooject being coated must oe
grounded and etectncaity conductive.
Powm Coating ¦¦ Coating resins in powder form that are
applied by electrostatic spray, fluidized bed or other
methods. The eoated object is heated, melting and sintenng
the powder to form a continuous coating.
Advantages: insoluble resins can be aooiied such as poly*
ethylene, poiyprooyiene. nylon and ftuorocaroons. as well as
other thermoplastic resins. Either thick or mm coatings can
be apoiied in one application. The obiect can be nanoted
immediately upon cooling.
Disadvantages: Powdered materials present neaith and
explosion Hazard unless prooer orecautions are tanen: ex-
pensive.
Force Ovying—Heating ot coated ootect after aooiieation
to accelerate drying or rate of coating cure, ventilation sys-
tem prevents solvent escape into atmosonere. Dry time to
tooeoat or handle snortened. However, expensive to install
and ooerate.
Electron Beam Curing—Recent innovation in which elec-
trons are accelerated through a vacuum and directed toward
obiect eoated by conventional means with a coating capable
of being crossiinked. The electron beam excites me reacting
molecules, completing crossiinktng and cure within seconds.
Advantages: Raoid handling and cure times, less solvents
*	in paint
Disadvantages: Cost; only a few coatings can oe cros»>
linked at present (polyesters and acrylics); coatings m
excess of seven mils cannot be cured.
Source: Excerpted by special permissin for Chemical Engineering, De-
cember 4, 1972, Copyright (c) 1972, by McGraw-Hill, Inc.,
New York, N.Y. 10020.
59

-------
E. WRAPPINGS
Another method of corrosion protection involves
the application of polyethylene as a loose wrapper
around the tank. Wrappers act as exterior coatings
which minimize the possibility of contact between the
metal and the soil. Wrappers have the advantage that
they are an inexpensive method of corrosion protection;
however, the initial savings must be weighed against
long-term economics and effectiveness. Wrappers are
difficult to install properly and are frequently ineffec-
tive; they may actually trap moisture on the tank surface
and thus lead to accelerated corrosion [10].
The life expectancy and performance of a wrapper
depends upon several factors, including the following:
•Its incompatibility with the surrounding soil.
•Its incompatiblity with the liquid stored if the
wrapper is exposed to that liquid.
•The wrapper thickness.
•The care taken to avoid tearing during its installa-
tion.
Some oil companies have reported extended tank
lives attributable to polyethylene wrappers, but to date
there is insufficient long-term experience to formulate a
solid judgement regarding their effectiveness [10].
Information on Specifications for Tank Materials and Construction
Carbon Steel
American Society for Testing and Materials
1916 Race Street
Philadelphia, PA 19103
American Society of Mechanical Engineers
345 East 47th Street
New York, NY 10017
American National Standards Institute
1430 Broadway
New York, NY 10018
Information on standards and specifications of the
Canadian Standards Association and the Interna-
tional Organization for Standardization may also.be
obtained from ANSI.
American Petroleum Institute
2101 L Street, N.W.
Washington, D.C. 20037
American Iron and Steel Institute
1000 Sixteenth Street, N.W.
Washington, D.C.
American Welding Society
2501 N.W. Seventh Street
Miami, FL 33125
National Association of Corrosion Engineers
1440 South Creek
Houston, TX 77084
Underwriters Laboratories, Inc.
333 Pfingsten Road
Northbrook, IL 60062
National Fire Protection Association
Batterymarch Park
Quincy, MA 02269
Steel Tank Institute
666 Dundee Road
Northbrook, Illinois 60062
Fiberglass-Reinforced
Plastic
American Society for Testing and Materials
1916 Race Street
Philadelphia, PA 19103
Underwriters Laboratories, Inc.
333 Pfingsten Road
Northbrook, IL 60062
National Fire Protection Association
Batterymarch Park
Quincy, MA 02269
Tank Relining -
Surface Preparation
Steel Structures Painting Council
4400 5th Avenue
Pittsburgh, PA 15213
60

-------
References
1.	American Petroleum Institute, Installation of Un-
y derground Petroleum Storage Systems, API Publi-
cation 1615, American Petroleum Institute, 2101 L
Street N.W., Washington, D.C. 20037, November
1979.
2.	Bixby, J.L., Underground Steel Storage Tanks.
The Construction Specifier, The Construction
Specifications Institute, Washington, D.C. 20036,
October, 1973.
3.	Fitzgerald, J.H., Suggested Ways to Meet Protec-
tion Corrosion Codes for Underground Tanks and
Piping, The Hinchman Company, 1605 Mutual
Building, Detroit, MI 48226, August 8, 1981.
4.	Hasse Tank GmbH & Co., KG, The Double Wall
Self-Monitored Tank, Betco Associates, P.O. Box
350, Closter, N.Y. 07624.
5.	Owens-Corning Fiberglas Corp., Fiberglas Tanks
for Fuel Storage, Pub. No. 3-PE-6312-L, Owens-
Coming Fiberglas Corp., Non-Corrosive Products
Division, Fiberglas Tower, Toledo, Ohio 43659,
December 1980.
6.	Underwriters Laboratories, Inc., Steel Underground
Tanks for Flammable and Combustible Liquids', UL
58, Underwriters Laboratories, Inc., 333 Pfingsten
Road, Northbrook, Illinois 60062, October 1976.
7.	Anonymous, "Composite Tanks Fuse Fiberglass to
Steel via Polyester Resin Bond," Petroleum Mar-
keter, McKeand Publications, May-June 1979.
8.	Anonymous, "Proper Installation is Vital to Long-
Life of Fiberglass Tanks," Petroleum Marketer,
McKeand Publications.
9.	Anonymous, "Steel Tank Institute's 'Sti-P3' Tanks
Combine Three-Way Protection," Petroleum Mar-
keter, McKeand Publications, May-June 1973.
10.	Fitzgerald, J.H., "Corrosion Control for Buried
Services Station Tanks," Paper No. 52, The Inter-
national Corrosion Forum Devoted Exclusively to
the Protection and Performance of Materials, April
14-18, 1975, Toronto, Canada, National Associa-
tion of Corrosion Engineers, Publications Depart-
ment, 1440 South Creek, Houston, Texas 77084.
11.	Standards of the Suffolk County Department of
Health Services for the Administration of Article 12
(Toxic and Hazardous Materials Storage and Han-
dling Control), of the Suffolk County Sanitary
Code, .County of Suffolk, Department of Health
Services, Central Islip, N.Y. 11722 dated October
3. 1979.
12.	Husock, B., Corrosion Cathodic Protection and
Common Sense, Paper No. HC-3, Harco Corp.,
Cathodic Protection Division, 1055 West Smith
Road, Medina, Ohio 44256.
13.	Lifetime Fiberglass Tank Co.. Spec-Data Sheet
13P - Liquid and Gas Storage Tanks, Lifetime
Fiberglass Tank Co., 5000 Packing House Road,
Denver, Colorado 80216, February 1981.
14.	Perry, R.H., Chilton C. H., Chemical Engineers;
Handbook, Fifth Ed, McGraw-Hill Book Co., 1221
Avenue of the Americas, New York, N.Y. 10020
15.	Owens-Corning Fiberglas Corp., Fiberglas Tanks
for Liquid Chemicals, Pub. No. 5-PE-6398-G,
Owens-Corning Fiberglas Corp., Non-Corrosive
Products Division, Fiberglas Tower, Toledo, Ohio
43659.
16.	Tator, K.B., Protective Coatings, Chemical En-
gineering Deskbook Issue, McGraw-Hill Book Co.,
1221 Avenue of the Americans, New York, N.Y.,
10020, December 4, 1972.
17.	Falck, S.B., Process Tank Linings, Chemical En-
gineering Deskbook Issue. McGraw-Hill Book Co.,
1221 Avenue of the Americans, New York, N.Y.,
10020, December 4, 1972.
18.	Telephone conversation with Dick Hugo, Ceramic
Coating Corp., Newport, KY., October 14, 1981.
19.	National Fire Protection Association, Flammable
and Combustible Liquids Code, NFPA 30. National
Fire Protection Association, Batteiymarch Park,
Quincy, MA 02269.
20.	American Petroleum Institute Service Station Tan-
kage Guide, API Publication 1611, American Petro-
leum Institute, 2101 L Street N.W., Washington.
D.C. 20037.
61

-------
References Continued
21.	American Petroleum Institute Recommended Stan-
dard for Underground Gasoline Tanks, API Publi-
cation 1602, American Petroleum Institute, 2101 L
Street N.W., Washington, D.C. 20037.
22.	Rothman, P.S., Cathodic Protection of Tanks and
Underground Structures, The Harco Corporation,
Hatboro, PA 19040, 1978.
23.	Clemmer Industries, Ltd., Double-Walled Storage
Tanks, Clemmer Industries (1964) Ltd, 446 Albert
Street, P.O. Box 130, Waterloo, Ontario N2J4A1,
August, 1981.
24.	J&T Ecology Corp., Industrial-Chemical Storage
Tanks, JTEC-979, J&T Ecology Corp., 200 Lam-
bert Avenue, Copiague, New York 11726.
25.	Crown Rotational Molded products. Crown Tanks
for the Industrial-Chemical Industries, Marked
Tree, AR.
26.	Steel Tank Institute, Reliable, Low Cost, Long Life
Design in Underground Steel Storage Tanks, Publi-
cation 15.8 (d)/St, Steel Tank Institute, 666 Dundee
Road, Northbrook, Illinois 60062.
27.	Husock, B., Fundamentals of Corrosion Protec-
tion, Paper No HC-2, Harco Corporation, Cathodic
Protection Division, 1055 West Smith Road,
Medina, Ohio 44256.
28.	National Fire Protection Association, Standards for
Installation of Oil Burning Equipment, NFPA 31,
National Fire Protection Association, Batterymarch
Park, Quincy, MA 02269.
29.	Anonymous, "Denver's Lifetime Co. Constructs
Fiberglass Tanks 'Like Giant Pills'", Petroleum
Marketer, McKeand Publications, March/April
1980.
30.	New York State Department of Environmental Con-
servation, Standards of Practice for the Bulk Stor-
age of Hazardous Liquids, to be written.
31.	Bethlehem Steel Corp., BT-10 Specifications.
Bethlehem Steel Corp., Buffalo Tank Division;
Frankfurst Avenue, Fairfield, Baltimore, MD
21226.
32.	Steel Structure Painting Council, 4400 Fifth Ave.,
Pittsburgh, PA 15213.
62

-------
Part II
CHAPTER 2:
UNDERGROUND
PIPING SYSTEM
A. INTRODUCTION
The components of a piping system for an under-
ground storage facility, as shown in Figure 2.2-1, in-
clude pipe, valves, pumps, and their associated connect-
ing joints and fittings. These components represent a
major potential source of leaks from underground stor-
age facilities. Such leaks may occur due to: (1) corro-
sion; (2) physical breakage; or (3) loose connections at-
tributable to wear or improper installation.
This chapter addresses the causes and methods of
preventing product leaks in the piping system of an un-
derground storage facility. The discussion focuses on
the types of piping, pumps, connecting joints and fit-
tings which are unique to underground (buried) service.
Piping systems described here are typical of those found
in the petroleum industry. Those components which are
usually located aboveground or in accessible locations,
such as valves and large pumps, are discussed in Part
III of this report.
Underground piping leaks can be prevented through
the following:
•Proper design (selection of materials, component
sizing, etc.)
•Proper installation.
•Proper testing.
•Timely replacement.
1.	Proper Design
The design and selection of appropriate compo-
nents for a particular piping system depends upon the
intended use of those components. The items that must
be considered include those listed in Table 2.2-1. In
short, piping system concerns in underground applica-
tions focus on: (1) physical strength of the components:
(2) ability to handle the required volumes (flow rates);
and (3) ability to withstand such phenomena as internal
and external corrosion, thermal loadings due to freezer-
thaw cycles, and the physical loads caused by surges of
liquid flow.
2.	Piping System Installation
Faulty installation of pipe and pipe fittings is a
major cause of leaks and spills at liquid storage facili-
ties. The following is a list of important considerations
during underground pipe installation:
B. CAUSES AND METHODS
OF PREVENTING LEAKS
The major causes of leaks from the piping system
are deterioration of piping system components and im-
proper installation of these components. The deteriora-
tion of piping system components can occur for any one
of several reasons: the most common of these, particu-
larly in the case of metal components or parts, is corro-
sion, which has been discussed in detail in Pan I. Other
reasons included the following:
•Mechanical failure (physical breakage or rupture),
such as the failure of valves or valve seals,
pumps, or the gaskets in fittings.
•Cracks in piping or connecting joints. These could
result from settlement or earth movement, vibra-
tion, or unrelieved stress concentrations.
Many leaks have also been traced to improper han-
dling and installations practices, such as the following:
•The improper connection of system components.
•The improper installation of bedding and founda-
tions for underground piping.
•Structural damage to piping, pumps, etc. during
transportation and installation.
•Corrosion when impressed current cathodic protec-
tion is improperly bonded to the system.
Table 2.2-1
Important Criteria in the Design of
Piping System Components
-	The type of service (transporting liquid, vapor
or slurry).
-	The (corrosive) characteristics of the material to
be transported, and the ability of piping system
components to withstand that corrosion.
-	The volume of material to be transported.
-	The extent to which surges in flow are expected
or likely.
-	The characteristics of the soil or other atmos-
phere to which the piping system components
are exposed.
63

-------
Figure 2.2-1
Piping Systems for Underground Storage
SUCTION LINES TO
PUMP ISLANOS (SLOPE
TO TANKS)
EXTRACTOR ANGLE
CHECK VALVE ASSY.

-NIPPLE
MANHOLE
,FIU. CAP
"8USHINO
-SUCTION TUBE
VENT LINE TO APPROPRIATE
LOCATION (SLOPE TO TANK)
ANGLE CHECK VALVE OR
UNOER PUMP CHIC VALVE. NO
/RISER REQ'O. W/ EARTH COVER.
/OOU8LE SyHM JOINT.
E*." .—Lj—l_1	' x
ft' WIN.
FILL TUBE I l/SUCTION
l| ™e
^^?Tmax L
OVERFILL PREVENTION
FLOAT VENT VALVE
*4
-STORAGE TANK
TT^12^T"
SYSTEM WITH SUCTION PUMPING
A v :;1
SYSTEM WITH SUBMERGED PUMP
Source: API Publication 1615, Installation of Underground Petroleum Storage Systems, 1979.
64

-------
•Product lines should be run in a single trench be-
tween the tank area and their destination. The
same is true for underground vent lines. This will
facilitate access to the piping system for repairs or
replacement of components.
•Before any underground lines are laid, the trench
or ditch for underground piping should receive a
minimum 6 inch-deep bed of well-compacted non-
corrosive (i.e., provides less corrosive environ-
ment than most native soils) material, such as
clean, washed sand or gravel. All trenches should
be wide enough to permit at least 6 inches of such
protection around all underground lines. This ap-
plies to both metallic and nonmetallic underground
piping. Bedding and the covering backfill should
be of the same material. Providing bedding and
covering in this manner serves two purposes: (1)
it provides proper structural support for the pipe;
and (2) it provides a less corrosive environment
for metallic piping system components than most
native soils.
•Piping should be arranged so that lines do not
cross over underground tanks. This will minimize
the possible creation of electrical connections be-
tween pipes and tanks which could accelerate cor-
rosion. Pipe connections to tanks should be
through insulated bushings.
•Underground product lines should have a mini-
mum cover of 12 inches for adequate protection
from the loads of surface traffic.
•Careful attention must be paid to the tightness of
all joints and pipe fittings. Tightness should be
tested (e.g. all joints and piping should be soap
tested) before covering the pipe.
•The possible breakage of underground piping, or
the loosening of pipe fittings resulting in product
leaks, can be minimized through the use of swing
joints. Swing joints, which are described later in
this chapter, provide for movement at pipe con-
nections without putting stress on the pipe. These
types of joints should be installed in lines at the
points where piping connects with underground
tanks and where the piping ends at pump islands
and vent risers. Fiberglass piping, which is inhe-
rently flexible, does not require swing joints if at
least 4 feet of straight run is provided between
any directional change exceeding 30 degrees.
•The actual location of. pipe should be noted on as-
built drawings, especially if there is a change
from facility design drawings. Photographs of un-
derground piping are also desirable as part of the
permanent record of piping locations. Pipe loca-
tion records of these types minimize the likelihood
of pipe breakage accidents during future excava-
tion at the storage facility.
•Product lines and vent lines should have a uniform
slope toward the tank of not less than '/« inch per
foot. This facilitates pipe drainage and avoids sags
or traps in the line in which liquid can collect.
Sloping is very important in insuring tight check
valve and proper leak detection operation.
More information and direction on pipe installation
practices can be found in API Publication 1615 [2].
3.	Periodic Testing
Periodic testing of underground piping system com-
ponents is also an important aspect of any leak/spill pre-
vention program. Inasmuch as leaks in piping systems
can occur in inaccessible (buried) pipe lengths and
joints because of corrosion, thermal stresses and
mechanical stresses, periodic testing is an important
means of incuring safety and reliability.
Underground piping systems may be tested using
the Kent-Moore Test as well as other types of tests.
These testing techniques and their accuracy are de-
scribed in Chapter 6 of this part of the report.
The required frequency of piping system testing
will vary depending upon the severity of service, avail-
able historical data, and local regulatory requirements:
•Testing will be more frequent when high rates of
internal or external corrosion are expected due to
the nature of the stored products or the soil in
which the piping system is buried.
•If the performance history of underground tanks or
piping in the area indicate the likelihood of rapid
deterioration of buried components, testing will be
more frequent.
•The frequency of testing of underground tanks and
piping may be mandated by law. OSHA regula-
tions require submerged transfer pump piping tests
at five year intervals.
4.	Timely Replacement
Equally important in an adequate leak prevention
program is the repair and/or replacement of deteriorated
or damaged pipe prior to the occurrence of a leak or
spill. Piping should always be of sufficient thickness
and integrity to withstand normal working pressures due
to fluid flow as well as the stresses caused by mechani-
cal loading, hydraulic surge pressures, thermal expan-
sion and contraction, and other conditions which can
impose stresses on piping. When the pipe wall thickness
or structural integrity of the pipe joints, connections,
etc. approach a point at which these stresses cannot be
withstood, that piping should be replaced.
Underground metallic piping at a storage facility
should be replaced when metallic underground storage
tanks are replaced to avoid accelerated corrosion in the
new tanks; such accelerated corrosion could result from
a reaction between the older pipe and the newer tank
(see discussion of corrosion in Part I).
65

-------
C. TYPES OF PIPING
E. EXPANSION JOINTS AND SWING JOINTS
There are a wide variety of types of piping com-
mercially available. Those that are extensively used for
underground applications include the following:
•Carbon steel.
•Stainless steel.
•Plastic.
•Fiberglass-reinforced plastic.
•Galvanized steel.
•Other composites, such as rubber-, plastic-, or
epoxy-lined steel.
These types are typically used in sizes ranging from 1
to 4 inches in diameter, although larger pipe may be
used in certain applications, such as at large bulk stor-
age facilities. Data describing the chemical compatibil-
ity, and the advantages and disadvantages of the types
of pipe listed above are summarized in Table 2.2-2.
In handling very hazardous or toxic liquids there
are double walled pipes available. One type consists of
an outer wall enclosing several pipes of smaller diame-
ter. This whole ensemble is contained by a bulkhead at
the end of each segment. The interstitial space of the
outer wall is slightly pressurized with nitrogen. The
inner pipes are filled with nitrogen at a higher pressure.
With this configuration, pipe runs can be checked for
leakage. A drop in pressure indicates leakage in the
outer wall. A rise in pressure shows a leak in one of
the inner pipes. See figure 2.2-2.
D. FITTINGS
Finings are the connecting links of the piping sys-
tem. This includes pieces of pipe that perform the fol-
lowing functions:
•Join two pieces of pipe, as do couplings and
unions.
•change pipe direction, such as is the case with el-
bows and tees.
•Change pipe diameter, as do reducers.
•Terminate a pipeline, such as is the case with
plugs and caps.
•Join two streams to form a third, as is the case
with tees, wyes and crosses.
•Allow for pipeline directional flexibility, as do
swing joints (or swivel joints) and expansion
joints.
As stated earlier, fittings are frequent locations of
piping system failure due to improper installation,
mechanical stress, or wear. To insure proper operation,
these components should be installed carefully and
tested periodically.
Expansion joints and swing joints are used to add
directional flexibility to pipelines, thereby preventing
the building up of potentially destructive stresses. As
shown in Figure 2.2-3, expansion joints typically con-
sist of a flexible bellows jointed to pipe at each end.
These types of joints can be designed for axial move-
ment, lateral movement, or a combination of axial and
lateral movement (see Figure 2.2-4). The most common
types of expansion joints are rubber reinforced with
steel rings and flexible corrugated metal bellows.
Expansion joints are used in piping systems for the
following purposes:
•Prevent stresses. Piping systems expand and con-
tract with temperature changes: an expansion joint
compensates for this movement.
•Eliminate vibration and noise. Pumps, compres-
sors, engines and pressure surges in pipe lines
create vibration and objectionable noises.
•Compensate for misalignment. Piping and
mechanical equipment often move out of normal
alignment during operation due to wear, load
stresses or settling of buildings and foundations.
•Reduce flange breakage. Undue stress caused by
misalignment, vibration, expansion or contraction
will break metal connecting flanges.
Note that there is a potential for leaks from expan-
sion joints because the repeated flexing of the joint will
eventually cause the joint to fail. To prevent this occur-'
rence, the joint should be periodically tested (piping
system tests) and. where accessible, inspected. They
should not be used unless they are inspectable.
Swing joints or swivel joints are employed to pro-
vide rotational flexibility to a pipeline. As shown in
Fig. 2.2-5, they may be designed to provide one,
two, or three planes of rotation. These types of joints
are used primarily to prevent torsional stresses in
pipelines, thereby reducing the likelihood of flange or
pipeline failure. Swivel joints, however, must be pro-
tected, so that no dirt can enter the area of the bearing
or bearing race.
F. UNDERGROUND PUMPS
The types of pumps used at underground storage
facilities are typically submersible pumps or suction
pumps. Suction pumps are located at grade, either di-
rectly above the storage tank or, as is the case in some
dispensing operations, at some distance from the storage
tank. Suction pumps may be either centrifugal, rotary or
reciprocating pumps. The differences between these
types of pumps and the concerns associated with their
operation are addressed in Part III, Chapter 2, and in
references such as the Chemical Engineers' Handbook
[3].
66

-------
Table 2.2-2
Characteristics of Piping Materials for Underground Service
Type of Pipe Chemical Compatibility
Remarks
Carbon steel
Stainless steel
Cast iron
Plastic tube
and pipe
Plastic lined
piping
Fiberglass and
fiberglass-
reinforced pipe
Compatible with petroleum products but
not compatible with corrosive chemicals
such as acids without coatings.
Compatible with petroleun and corrosive
chemicals, such as acids, depending on
grade.
Resists corrosive attack by natural or
neutral waters and neutral soils as well
as atmospheric corrosion. Resistant to
concentrated acids (nitric, sulfuric and
phosphoric as well as alkaline and caustic
solutions. Dilute acids and acid-salt solu-
tions will attack this material.
Various plastics can be chosen for their
resistance to specific chemicals.
For example:
Polyethylene pipe and tubing have excel-
lent resistance to salts, sodium and
ammonium hydroxides, and sulfuric,
nitric and hydrochloric acids.
Polyvinyl chloride pipe and tubing have
excellent resistance at room temperatures
to salts, alcohol, gasoline, ammonium
hydroxide, and sulfuric, acetic, nitric, and
hydrochloric acids: may be damaged by
ketones, aromatics and some chlorinated
hydrocarbons.
Polypropylene pipe and tubing having
excellent resistance to most common
organic and mineral acids and their salts,
strong and weak alkalies, and many
organic chemicals.
Same as plastic tube and pipe above.
Compatible with a wide range of
petroleum and chemical products. See
chemical compatibility chart in Part II.
Chapter 1.
Susceptible to corrosion if not coated,
galvanized or cathodically protected.
Relatively inexpensive.
Galvanized steel is used extensively at
service stations and other petroleum
industry applications.
Used when product purity is of great
concern.
High relative costs.
Primarily used for corrosion protection
when coatings will not suffice (e.g. at high
operating temperatures).
Low relative cost.
Provides more metal for less cost than steel
piping systems.
Brittle - has poor resistance to impact or
shock.
More widely used for non-hazardous service
(e.g.. water) than for hazardous chemicals
service.
' Widely used for low pressure service where
corrosion causes extensive loss of metal.
Free from internal and external corrosion.
Do not cause galvanic corrosion when
coupled to metallic material.
Allowable stresses and temperature limits
are low.
Low structrual strength when compared
to steel.
Plastics are suitable for underground
service when U L-approved for the product
being carried by the pipe.
Combines the chemical resistance of the
various plastics and the tensile and struc-
tural strength of steel.
Less structural strength than steel.
High resistance to external and internal
corrosion.
Suitable for underground piping when UL
approved for the product being carried.
67

-------
.Figure 2.2-2
Double Walled Pipe
SERVICE PIPE
INSULATION
CONDUIT
PIPE SUPPORT
SERVICE PIPES
TRAVERSE ARRESTOR
LAGGING SLEEVES
CORRUGATED SPACER
TACK WELDS
SPACER LOAD CARRYING
SUPPORTS
MULTI-PIPE STANOARO WEB
INSULATOR UNOER LAGGING
Source: Perma-Pipe, division of Midwesco, Inc.
68

-------
Figure 2.2-3
Diagram of a Universal-Type
Expansion Joint
Tie rods
23
X.
W\

Swivel or hinged connection
/

X
Source: From Chemical Engineers' Handbook, Perry, R.H. and Chilton, C.H., Copyright (c) 1973, McGraw-
Hill, Inc. Used by special permissionof McGraw-Hill Book Company.
Figure 2.2-4
Action of the Bellows
of an Expansion Joint
rAAAAAA_
"WWW
jwum
wuwjir
Undeflected Position
A» Axial compression
p-A®Axiol extension


Forms of lateral movement
rf^yvia
^^7
Source: From Chemical Engineers' Handbook. Perry, R.H. and Chilton. C.H., Copyright (c) 1973, McGraw-
Hill, Inc. Used by special permissionof McGraw-Hill Book Company.
69

-------
Figure 2.2-5
Swing or Swivel Joints
Cross section. Shows bail bearings
which permit rotational movement.
Cannot be used underground unless
protected to prevent dirt from
entering bearing area.
2 PLANES OF ROTATION
Source: OPW Division/Dover Corp.
70

-------
The submersible pump (submerged transfer pump)
system works on the principle of positive pressure to
push the liquid from a low point to a high point. Unless
there is a leak line detector in the system or the leak
is large, loss of product will not be detected until con-
siderable volume has been lost. A suction pump system
works because of a vacuum at the high end which
draws liquid from the low end. A large leak in the sys-
tem would result in lack of suction and an immediate
indication of trouble. A small leak would result in
drainage of the pipe overnight and a lack of prime in
the morning which would cause the system to be in-
operative. A check valve under the product dispenser
might hold enough product to reprime the pump but
such a system should be discouraged because it would
mask a line leak.
Submersible pumps are mounted inside the tank;
they are centrifugal pumps closely coupled with an elec-
tric motor that can operate when submerged. These
types of pumps may be commonly used in situations
such as gasoline service stations.
When suction pumps are used, leaks in the pump
delivery line that result in significant losses of product
can be detected through a loss of pump suction, result-
ing in inefficient or poorer pump operation. When such
a situation is encountered, operations should be halted
until the source can be identified and corrective action
taken.
In the case of submerged pumping systems, leaks
in the product delivery line can be detected through the
use of product delivery line leak detectors. These de-
vices are mounted immediately above the tank on the
pump delivery line as shown in Figure 2.2-1; they are
designed to detect losses of pressure in the product de-
livery that do not correspond to decreases in the dis-
charge pressure of the submerged pump. Such a loss of
pressure in the product delivery line indicates a loss of
liquid in that line before it reaches the discharge point
of the pipeline.
Another device recommended (actually required by
NFPA 329) for use in submerged pumping systems or
remote pumping systems, is a remote pump shut-off
valve. In service stations these valves are located at the
base of the dispensers. Should the dispenser be over-
turned, due to bumping or impact, the valve automati-
cally closes, preventing extensive product spillage.
These valves also contain a fusible link which closes the
valve upon exposure to excessive heat or fire. A cross-
section of a remote pump shut-off valve is shown in
Figure 2.2-6.
Systems are also available that automatically shut
off pipe flow in case of a drop in pressure or a differ-
ence of input compared to outflow.
STAINLESS STEEL MAIN
SPRING
COPPER-NICKEL-CHROME
PLATED BRASS STEM
TEFLON COATED I.O. OF
PACKING NUT
CORROSION RESISTANT
SEAT RING
STAINLESS STEEL POPPET
RETAINING RING
Source: OPW Division!Dover Corp.
Figure 2.2-6
Typical Remote Pump Shut-Off Valve
71

-------
References
1.	American Petroleum Institute Underground Leak
Survey results as reported by F.B. Killman to API
Underground Leak Task Force, American Petro-
leum Institute, 2101 L Street, N.W. Washington.
DC 20037, February 5, 1981.
2.	American Petroleum Institute, Installation of Un-
derground Petroleum Storage Systems, API Publi-
cation 1615, American Petroleum Institute, 2101 L
Street N.W., Washington, DC 20037, 1979.
3.	Perry, R.H., Chilton, C.H., Chemical Engineers'
Handbook, Fifth Edition, Section 6, McGraw-Hill
Book Company, 1221 Avenue of the Americas,
New York, NY 10020, 1973.
4.	Pace Company Consultants & Engineers, Inc., Spill
Control Manual prepared under EPA Training Grant
No. T-900-175-02-2 to the Department of Environ-
mental Science and Engineering, Rice University,
Houston, TX, February, 1975
5.	American Petroleum Institute, Guide for Inspection
of Refinery Equipment, Chapter XI, Pipes, Valves
and Fittings, American Petroleum Institute, 2101 L
Street, N.W., Washington. DC 20037, 1974
6.	Baumeister. T.. Avallone, E.A., Baumeister III.
T., Marks' Standard Handbook for Mechanical En-
gineers, Eighth Edition. McGraw-Hill Book Com-
pany, 1221 Avenue of the Americas, New York,
NY 10020, 1978.
7.	Dover Corp., OPW Swivel Joints: Ball Bearing
Joints for Flexible Piping Systems, Catalog SJ.
Dover Corporation, OPW Division. P.O. Box
40240, Cincinnati, OH 45240. February. 1980.
8.	Dover Corp., OPW Engineered Service Station
Products, Catalog SSF. Dover Corporation, OPW
Division, P.O. Box 40240. Cincinnati, OH 45240,
July, 1981.
9.	Perma-Pipe., div. of Midwesco, Inc.. 7720 Lehigh
Road, Nile, IL 60648.
72

-------
Part II
CHAPTER 3:
UNDERGROUND SPILL
CONTAINMENT SYSTEMS
A. INTRODUCTION
1. Background
Underground spills or leaks can range in severity
from a minor leak at a pipe joint or tank wall to a catas-
trophic tank rupture that spills the entire tank contents
into the ground. The largest spills come from sizeable
leaks that go undetected for years. Regardless of the
severity of the incident, the resulting impacts in terms
of soil and groundwater contamination are highly unde-
sirable and should be avoided. The steps which can be
taken to prevent long-term and widespread contamina-
tion of the soil and ground water include the following:
•Leak prevention and early detection practices such
as good housekeeping, inventory control, and
monitoring of the immediate storage area.
•Containment of the storage area.
•Removal of contaminated soil or groundwater be-
fore extensive spreading of the contaminant.
Good housekeeping is an important aspect of any
leak prevention program. Many incidents occur due to
carelessness and sloppy housekeeping, practices which
should not be tolerated. Good housekeeping practices
were addressed in Part I of this report.
Early warning leak detection practices such as in-
ventory monitoring and tank excavation monitoring
form the first line of defense against extensive soil and
groundwater damage due to leaks. These techniques are
discussed in detail in Chapter 5 of this part of the re-
port.
U/idergound spill or leak containment systems rep-
resent the second line of defense agaist propagation of
soil or groundwater contamination. They can also act to
enhance the effectiveness of early warning leak monitor-
ing systems by confining the leak or spill until detection
is possible.
The removal of contaminated soil and groundwater
represent the last line of defense against widespread
contamination. Practices such as extensive soil excava-
tion for disposal, or the use of recovery wells to remove
contaminated groundwater fall into this category and are
drastic and relatively expensive steps. In addition, ac-
tions of this type are not always completely successful.
For example, it may be impossible to remove all of a
contaminant from a groundwater if it has spread exten-
sively or become thoroughly mixed in the groundwater
table. Recovery wells are discussed briefly in Part II,
Chapter 5 of this report.
2. Containment Technology
The control technology used to contain under-
ground spills and leaks consists of establishing a barrier
around the storage tank so that any leaked liquid does
not have a free path to escape from the storage area.
The banier materials used for containment include the
following:
•Liners with low soil permeability (clay).
•Synthetic membrane liners.
•Soil sealants, such as soil cement or bentonites.
•Concrete vaults.
•Double-walled tanks.
It is important to include a liquid removal and
monitoring system as part of secondary containment.
The containment floor should be sloped to a sump from
which a sample can be taken for analysis to determine
if product is leaking from the tank. If the secondary
containment does not have an impervious cover, ac-
cumulated rainwater which percolates to the liner should
be removed by siphoning, pumping or via an under-
ground drainage system. In fact some water will proba-
bly collect above the containment liner even with an im-
pervious cover. Such water should be considered as
contaminated and should receive proper treatment after
being drained off.
Selection of the proper containment material for a
particular application depends upon several factors, in-
cluding:
•The type of material being stored.
•Local environmental conditions.
•Legislative requirements.
Type of Material Being Stored. Consideration of
compatibility with the liquid being stored is important:
the liner material must be able to maintain its integrity
and impermeability when exposed to the stored product.
Local Environmental Conditions. The sensitivity
of the environment in the vicinity of the storage facility
can largely affect the level of environmental protection
and hence the type of containment liner required. For
example, in areas where the storage facility is located
near or above an aquifer, greater care may be required
in the selection and installation of the containment liner.
Legislative Requirements. In addition, local gov-
ernments may be highly prescriptive and specific in
terms of the type of containment barrier required. Such
legislation requirements are often based upon local en-
vironmental conditions.
Containment systems will be effective only as long
as they remain intact. Disruption of clay liners or soil
sealants by tree roots, or the ripping of synthetic liners
during handling are examples of incidents that can lead
to ineffective leak or spill containment.
Table 2.3-1 presents a summary comparison of the
various types of underground containment systems.
Further details are provided in the remainder of this
chapter.
73

-------
B. CLAY LINERS
C. SYNTHETIC MEMBRANE LINERS
1. Chemical and Physical Properties [5,8,10]
Due to their general availability in many areas,
clays are often considered the first altemtive for storage
tank containment liners. Clays are relatively inexpensive
liner materials that can be extremely effective for tank
storage. These materials can also be effectively used as
liners for pipe trenches.
Clays are complex minerals that have a wide range
of compositions and properties. They are subject to
changes in composition due to several factors, including
the following:
•Weathering when exposed to air.
•Leaching of components when exposed to ground-
water or other solutions.
•Ion exchange, or the replacement of ions in the
clay with other ions of similar charge, when ex-
posed to substances such as water containing
acids, alkalis, or dissolved salts.
•Destabilization when exposed to some organic sol-
vents.
Other factors that influence the performance of clay
liners include: (1) compatibility with the stored product;
(2) the thickness of the clay liner; (3) the shrink-swell
potential of the clay; (4) the plasticity of the clay; and
(5) the moisture content, density, and degree of com-
paction of the clay. The selection of a clay material for
a particular liner application should be based upon tests
for suitability by a soils engineer or a soils chemist.
2. Design and Installation Requirements [5,8]
Before installing a clay liner, it is necessary to first
drain and stabilize the excavation. A bottom layer is
then laid in place and compacted using a device such
as a steel wheel roller. This bottom layer should be at
least 6 inches deep; depths of 2 to 4 feet are not uncom-
mon. When this bottom layer is more than 6 inches
thick, it is usually the practice to apply it in stages to
ensure proper compaction. The required degree of com-
paction depends upon the composition of the soil itself,
its clay content, density, and moisture content. Once the
bottom layer has been properly installed, the tank
should then be installed in accordance with New York
State standards and guidelines, and the excavation back-
filled with more clay material to provide containment all
around the tank. The installation of clay liners can be
a complex operation requiring a trained contractor to en-
sure high levels of quality control.
1. Chemical and Physical Properties [5,8]
Synthetic membrane liners are polymeric materials,
manufactured in sheet form, that can be spread over the
tank excavation walls or floor to contain a leak or spill.
As a class, these types of liners have several advantages
and disadvantages. The advantages of synthetic mem-
brane liners include the following:
•They can contain a wide variety of liquids with
minimum loss through seepage.
•They have high resistance to bacterial deteriora-
tion.
•They have high resistance to chemical attack.
•They are relatively economical to install and
maintain.
•They are readily installed for many applications.
In general, the disadvantages of synthetic membrane lin-
ers include the following:
•They are vulnerable to attack from ozone and ul-
traviolet light (sunlight) when compared to other
types of liners.
•They have limited ability to withstand heavy
loads.
•They are susceptible to laceration, abrasion, and
puncture.
•They are prone to cracking at low temperatures,
and stretching and distortion at very high tempera-
tures.
The synthetic, polymeric membranes that are most
commonly used to contain chemical and petroleum
products are polyvinyl chloride (PVC), polyethylene,
chlorinated polyethylene (CPE), chlorosulphonated
polyethylene (CSPE or hypalon), oil-resistant polyvinyl
chloride (ORPVC), ethylene proylene diene monomer
(EPDM), butyl rubber and neoprene. In addition. DuP-
ont has developed a proprietary elasticized polyolefin
called 3110. Table 2.3-2 presents a general summation
of the advantages and disadvantages of thse synthetic
materials, and Table 2.3-3 presents a summary of the
compatibility of these substances with various types of
hazardous materials. For more information on the chem-
ical compatibility of synthetic membrane liners, please
refer to references 5 and 8.
74

-------
Table 2.3-1
Comparison of Underground Spill Containment Systems
Type of System
Advantages
Disadvantages
Relative
Cost
Clay Liners
Polymeric Liners
Soil Cement
Bentonite
Concrete Vaults
Double Walled
Tanks
The least expensive liner if clay is available
close to the site.
Use of clay is a well established practice
and standard testing procedures are
available.
Subject to drying and cracking and thus
must be protected with soil cover.
Subject to leaching of components when
exposed to groundwater or other solutions.
Subject to ion exchange when exposed to
water containing acids, alkalis, or dissolved
salts.
Subject to destabilization when exposed to
some organic solvents.
Require subgrade preparation and steril-
ization to reduce risk of puncture.
Must be protected from damage, particu-
larly due to vehicular traffic.
High resistance to bacterial deterioration. Must be protected from sunlight and ozone.
May be attacked by hydrocarbon solvents
particularly those with high aromatic content.
Good oil resistance and good low tempera-
ture properties do not normally go hand
in hand.
Well established solution to problem of con-
taining petroleum products.
Particularly good for temporary storage.
Good durability.
Resistance to aging and weathering
Low permeability.
Does not deteriorate with age.
Self sealing.
Good strength and durability.
Constructed of material (FPR or coated
steel) which is resistant to the stored
product and to external corrosion.
Includes leak detection system in tank
design.
Subject to degradation due to frost heaving
of subgrade.
In place soil usually used; permeability
varies with the type of soil.
Untreated bentonite may deteriorate when
exposed to contaminant.
Requires protective soil cover, typically
18 inches.
Subject to destabilization when exposed to
some organic solvent.
Requires surface coating to insure
impermeability.
Subject to cracking when exposed to
freeze; thaw cycles.
Some models only available in tank sizes
up to 4.000 gallons.
Low
Moderate
to High
Moderate
Moderate
High
High
75

-------
2. Design and Installation Requirements [5,8]
In most instances, installation of the liner is as im-
portant to the overall success of the application as mate-
rial selection. Liner installation is a relatively compli-
cated task that should be performed by a qualified con-
tractor, paying attention to important details such as the
following:
•The base of the excavation should be compacted
to prevent settling under the liner and tank after
they are in use.
•The slope of the excavation should be stable to
avoid collapse after the liner has been installed.
•The base and sidewall areas that will contact the
membrance should be finely finished. All rocks,
rubble and debris which could puncture the lining
should be removed. Sand layers should be placed
above and below the membrane to further prevent
punctures and to facilitate underdrainage. Soil
sterilization with a herbicide may be considered in
instances where vegetation may propagate, but the
herbicide should not be applied indiscriminately.
•The liner should be carefully placed and seamed
(bonded) in accordance with manufaturers' specifi-
cations. Table 2.3-4 summarizes the important
considerations of synthetic liner installation.
Membrane liners are typically used in areas
of high groundwater, although they can be em-
ployed in other instances. When the material to be
stored is lighter than water, the liner is always in-
stalled around the sides of the excavation perime-
ter extending down beneath the groundwater level.
For these types of chemicals, the groundwater acts
as the bottom containment for any leak or spill.
When the liquid to be stored is heavier than
water, the liner is always installed under the tank
along the excavation base as well as along the
sidewalls to prevent the liquid from migrating out-
side the excavation area. With the liner under the
tank, the bottom liner cover area should be
drained before closure. An observation well to a
low point of the membrane could be used to con-
firm liner integrity. The excavation should have
an impervious cover to prevent flooding of the
lined area. Figures 2.3-1 and 2.3-2 illustrate these
types of applications.
Membrane liners can also be used as wrap-
pers around underground storage tanks.
Polyethylene wrappers have been used in such a
manner to enclose steel tanks. Some success in
tank leak prevention has been reported using such
a technique. However, corrosion can occur under
the wrapper if groundwater enters the space be-
tween the tank and wrapper through a tear or
other imperfection. In addition, such a wrapper is
not adequate for use with cathodic protection for
steel tanks [9]. Further discussion of this tank pro-
tection technique has been included in Part II,
Chapter 1.
Table 2 3-4
Considerations During Liner Placement
Use a qualified installation contractor having
experience with membrane liner installation, prer
ferably the generic type of liner being installed.
Plan and implement a quality control pro-
gram which will help insure that the liner meets
specification and the job is installed per specifica-
tions. Inspection should be documented for review
and recordkeeping.
Installation should be done during dry, mod-
erately warm weather if possible.
The excavation base and wall should be firm,
smooth, and free of sharp rocks or debris.
D. SOIL SEALANTS
The types of soil sealants more commonly
used for lining storage tank containment areas are
soil cement and bentonites (clay materials). These
types of sealants are discussed in detail below.
1. Soil Cement [3.8]
Chemical and Physical Characteristics. Soil ce-
ment is a compacted mixture of Portland cement, water,
and selected in-place soils. The result is a low strength
Portland cement concrete with greater stability than nat-
ural soils. The permeability of this mixture varies with
the type of soil used, a more granular soil produces a
more permeable soil cement.
Any soil can be treated with cement. However,
there are some exceptions where cement should not be
used:
•Highly organic soil retards cement hydration be-
cause of absorption of calcium ions.
•Clean well-graded gravels and crushed rock are
sometimes unsuitable because of shrinkage prob-
lem.
•Clays can be unsuitable because of the difficulty
of incorporating a fine cement powder into a wet.
plastic clay and because property changes are not
significantly affected.
•Saline soils are unsuitable, but this can be over-
come by increasing the cement content.
The aging and weathering characteristics of soil ce-
ments are good, especially when exposed to wet-dry
and freeze-thaw cycles. Some degradation has been
noted when this substance is exposed to highly acidic
76

-------
Table 2 J-2



Comparison of Various Synthetic Polymeric Membranes


Liner Type
Advantages
Disadvantages
Relative
Cost
Polyvinyl Chloride
(PVC)
Good resistance to ozone and ultraviolet
ligh when properly stabilized
Good resistance to puncture.
High tensile strength.
Poor hydrocarbon resistance.
May deteriorate in presence of certain
chemicals and in contact with heat.
Low
Oil Resistant PVC
Improved resistance to aromatic hydro-
carbon relative to standard grades of PVC.
Poor low temperature handling properties.
Moderate
to High
Polyethylene1
Great resistance to bacteriological
deterioration.
Good tensile strength.
Few restrictions on chemical exposure.
Good low temperature characteristics.
Poor puncture resistance.
Poor tear strength.
Susceptible to weathering and stress cracking.
Low
Chlorinated Poly-
ethylene (CPE)
Excellent weatherability.
Good tensile and elongation strength.
Good resistance to ultraviolet lignt
and ozone.
Excellent crack and impact resistance at
low temperatures.
Moderate to good hydrocarbon resistance.
Limited range of tolerance for chemicals,
oils and acids.
Low recovery when subject to tensile stress.
Moderate
Chlorosulfonated
Polyethylene
(CSPE or hypalon)
Good puncture resistance.
Good resistance to microbiological attack.
Excellent resistance to low temperature
cracking.
Excellent weather resistance.
Low tensile strength.
Poor resistance to aromatic hydrocarbons.
Moderate

Good resistance to ozone and ultraviolet light.


Flexible and resilient.


Ethylene Propylene
Diene Monomer
Good weathering characteristics.
Good temperature flexibility.
Good heat resistance.
Resistant to mildew, mold, and fungus.
Excellent resistance to water vapor
transmission.
Poor resistance to aromatic hydrocarbons.
Low peel and shear strength.
Moderate
Butyl Rubber
Excellent resistance to water.
Excellent resistance to ultraviolet light and
ozone.
High tolerance for temperature extremes.
Good tensile and shear strength.
Good resistance to puncture.
Ages well in general, but some compounds
will crack on ozone exposure.
Poor resistance to hydrocarbons particu-
larly petroleum solvents, aromatics, and
halogenated solvents.
Poor sealability.
Moderate
to high
77

-------
Tftbfc 1^2 continued



Neoprene
Excellent aging and weathering character-
Not heat or solvent sealable.
High

istics.



Overall good resistance to hydrocarbons.



but shows some swell when exposed to



aromatics and other cyclic hydrocarbons.



Flexible and elastic over a wide range of



temperatures.


Elasticized Poly-
Resistant to ultraviolet light: does not
Relatively untested.
Moderate
olefin
require earth cover.


(DuPont 3110)
Good resistance to weathering and aging.
Vunerable at low temperatures.


Good resistance to ozone attack and soil



microorganisms.



Good resistance to hydrocarbons and will



accomodate a broad range of solvents.


NOTE 1 • Refers to low density polythylene. High density polyethylene is much less susceptible to puncture, tears, weathering and stress cracking.

Table 2 -3-3





Chemical Compatibility of Membrane Liners with Hazardous Materials


Lining Material
Strong
Acids
Strong
Bases
Petroleum
Products
Halogenated
Solvents
Aromatic
Solvents
Polyvinyl chloride (PVQ1
R
R
NR
NR
NR
Chlorinated Polyethylene (CPE)
R
R
NR
NR
NR
Chlorosulfonated Polyethylene
(CSPE or Hypalon)
R
R
NR
NR
NR
Ethylene Propylene Diene Monomer
(EPOM)
R
R
NR
NR
NR
Neoprene
R
R
R
NR
NR
Butyl Rubber
R
R
NR
NR
NR
Oil Resistant Polyvinyl Chloride
(ORPVC)1
R
NR
R
NR
NR
Polyethylene
R
R
R
R
R
NOTES: I. Not recommended if liner is exposed to the atmosphere due to extreme susceptibility to ultraviolet light and lor ozone
R - recommended





NR = not recommended





Source; References 1,2 and 4.





78

-------
environments, but soil cements can resist moderate
amounts of alkali, organic matter, and inorganic salts.
One of the main deficiencies of soil cement as a liner
material is its tendency to crack and shrink on drying.
Severe cracking and detrioration may also result if the
cement content of the mixture is too high.
Design and Installation. The details for construc-
tion, excavation base and wall preparation, and placing
and curing of soil cement liners can be obtained from
documents such as reference 5, 8 and 9, and from con-
sulting engineers in this field. Some of the consider-
ations and procedures are highlighted in Table 2.3-5.
2. Bentonites [8,10]
Chemical and Physical Characteristics. Bento-
nites are naturally occurring inorganic swelling clays
which are typically chemically treated, and are marketed
under various trade names. Mixtures of soil and chemi-
cally treated bentonites may be used to line excavations
for underground tanks and contain spills. When the ben-
tonite is mixed with a sandy soil and saturated with
water, the granular bentonite particles in the soil swell
to fill the voids with a tough leather-like mastic,
thereby forming an impermeable barrier. Bentonite can
swell up to 15 times its dry bulk volume when used in
such a manner. Untreated bentonites are generally not
as effective when used as soil sealants and are more
susceptible to degradation, particularly if the water used
to wet the material during installation contains a high
concentration of dissolved salts (i.e., hard water), acids,
or alkalis. Bentonites are also subject to destabilization
when exposed to some organic solvents.
Design and Installation Requirements. Before in-
stalling a bentonite liner, it is necessary to first drain
and stabilize the excavation. The mixture of soil and
bentonite is then used to line the bottom of the excava-
tion. The mixture is typically wetted to saturation, and
compacted using a wobble wheel or steel wheel roller
[8]. The tank is then installed in accordance with New
York State standards and guidelines and/or the manufac-
turer's recommendations, and the excavation is backfil-
led with more clay/soil mixture. When preparing the
mixture, the manufacturer's recommendations should be
Table 2-3-5
Highlights of Soil Cement Design and Installation
-	Preparation of the base and walls is extremely
important. The base and wall should be prop-
erly finished, and well moistened before plac-
ing the concrete to prevent the liner from dry-
ing too quickly.
-	Concrete mixes should be plastic enough to
consolidate well, but not stiff enough to slip on
side slopes.
-	Proper curing of the liner is important.
79
Figure 2.3-1
Synthetic Liner Installation for Storage of Lighter-Than-Water
Liquids in Area of High Groundwater

-------
Figure 2.3-2
Synthetic Liner Installation for Storage of Heavier-Than-Water
Liquids in Area of High Groundwater
Liner
followed as to the percent of the clay and soil, the
amount and quality of water used for wetting, and- the
degree of compaction required. The mixture varies, but
usually consists of one part bentonite and three parts
clean uncontaminated soil.
E. CONCRETE VAULTS
Concrete vaults are secondary enclosures consisting
of concrete walls and a concrete bottom slab upon
which the tank is fastened. The vault system may in-
clude a cover.
The vaults may contain one, or more than a dozen
tanks. Some vaults have an open interior so that tanks
can be physically inspected while others are filled with
a bedding of sand which provides structural support for
the tanks. When the vaults are of open design, the inter-
ior tanks are supported structurally on cradles. Unusally
the vault contains a sloped floor and a sump installed
with a monitoring probe and a product recovery pump.
Concrete by itself is not an effective liquid barrier.
Leaks through concrete occur in the vapor phase. Con-
crete will pass vapors of many chemicals after only a
few days of exposure.
Coatings to make concrete impermeable are effec-
tive but there is no universal concrete coating for all
chemicals, weather, and moisture conditions. Coatings
will peel, crack or wear (in traffic areas) over time. In
areas of wear, successive layers of coatings are color-
coded, to show wear patterns.
A common practice is to put a vapor barrier around
the outside of the vault. Concrete vaults must be care-
fully designed and constructed; otherwise joints may
leak or the walls and floor may crack when exposed to
freeze-thaw cycles for extended periods or if settling of
the tanks occurs. Concrete vaults are mandatory in New
York City for the underground storage of gasoline and
other fuels.
F. DOUBLE-WALLED TANKS
Spill containment may also be provided with dou-
ble-walled tanks. These tanks are essentially a tank
within a tank and are described in Part II, Chapter 1 of
this report.
80

-------
References
1.	Haxo, H.E., Haxo, R.S., White, R., Liner Mate-
rials Exposed to Hazardous and Toxic Sludges,
First Interim Report, EPA-600/2-77-081, prepared
by MATRECON, INC., Oakland, California, pre-
pared for USEPA Municipal Environmental Re-
search Laboratory, Cincinnati, Ohio 45268, June,
1977.
2.	The Watersaver Company, Membrane Linings,
Catalog No. GB78, The Watersaver Co., P.O. Box
16465, Denver, Colorado 80216.
3.	Suffolk County Department of Health Services, De-
sign of Underground Gasoline and Oil Storage Fa-
cilities, Standards of the Suffolk County Depart-
ment of Health Services for the Administration of
Article 12 (Toxic and Hazardous Materials Storage
and Handling Control), of the Suffolk County
Sanitary Code, Central Islip, New York 11722,
dated October 3, 1979.
4.	Perry, R.H., Chilton, C.H., Chemical Engineer's,
Handbook, Chapter 23, Materials of Construction,
Fifth Ed., McGraw-Hill Inc., 1221 Avenue of the
Americas, New York, New York'10020, 1973.
5.	U.S. Environmental Protection Agency, Lining of
Waste Impoundment and Disposal Facilities,
SW870, USEPA Municipal Environmental Research
Laboratory, Cincinnati, Ohio 45268, September,
1980.
6.	American Colloid Company, Soil Sealants that
Confine Oil and Chemical Leaks or Spills, Bro-
chure No. 290A, American Colloid Company, En-
vironmental Products Division, 5100 Suffield
Court, Skokie, Illinois 60077.
7.	American Colloid Company, Volclay Seepage Con-
trol Systems, Brochure No. 229L, American Col-
loid Company, Environmental Products Division,
5100 Suffield Court, Skokie, Illinois 60077.
8.	Petroleum Association for Conservation of the
Canadian Environment. State of the Art Review -
Petroleum Product Containment Diking, PACE Re-
port No. 792, prepared for PACE by Golder As-
sociates and James F. MacLaren Limited, Suite
400, 130 Albert Street, Ottawa, K1P564, Canada,
March 1979.
9.	Fitzgerald, J.H., "Corrosion Control for Buried
Service Station Tanks," Paper No. 52, The Interna-
tional Corrosion Forum Devoted Exclusively to the
Protection and Performance of Materials, April 14-
18, 1975, Toronto, Canada. National Association of
Corrosion Engineers, Publications Department,
P.O. Box 1499, Houston, Texas 77001.
10.	Jewett, M.A., Buyce, M.R., A Review of Methods
for Determining the Permeability Characteristics of
Colorado Soils Exposed to Hazardous Waste, pre-
pared by Fred C. Hart Associates, Inc., prepared
for U.S. Environmental Protection Agency, Region
VIII, I860 Lincoln Street, Denver, Colorado
80295, June. 1981.
81

-------
Part II
CHAPTER 4:
TRANSFER SPILL
AND OVERFILL PREVENTION
SYSTEMS FOR UNDERGROUND
STORAGE TANKS
A. INTRODUCTION
Spills can occur at underground storage tank facili-
ties because of tank overfilling and drainage from prod-
uct transfer hoses. For example:
•It is common practice to unload products from ve-
hicles into underground storage tanks without au-
tomatic means to prevent overfilling tanks. With-
out such protection, underground tanks can be
overfilled with product which will rise through the
vent lines until it attains a level equal to the prod-
uct level inside the tank truck being unloaded.
•ft is common practice to use quick disconnection
couplings on the ends of discharge hoses rather
than dry-break couplings which are heavier and
more difficult to maneuver. With quick disconnect'
couplings, product remaining in the discharge,
hose is frequently spilled ne2r the tank area. Be-
side the obvious dangers which can result from -
sloppy practices, the daily small spills seeping
into the ground near the tank areas can accumulate
into sizeable and hazardous volumes over a period
of time.
To avoid situations such as these, it is good prac-
tice to equip underground tanks with overfill prevention
systems and to equip transfer hoses with an automatic
shut-off device which prevents backflow when the hose
is disconnected. Adherence to good operating practices
is also a prerequisite.
An ideal underground tank overfill prevention sys-
tem would include the following basic elements: (1) a
level sensing device that monitors and indicates the liq-
uid level in the tank; (2) an alarm to alert the operator
of an impending overfill condition; and (3) an automatic
shut-off device that stops the flow of product when the
tank is full.
Spills from transfer hoses can be prevented by
using' couplings equipped with spring loaded shut-off
valves which stop flow automatically when the hose is
disconnected and by using dry-disconnect couplings.
Emergency shut-off valves may also be used to stop
product flow, such as in the case of fire.
Devices for overfill and spill prevention add to the
system cost, however. Underground tanks are not nor-
mally equipped with overfill prevention devices. Truck
fleets generally are not equipped with dry disconnect
couplings and there is some argument as to the reliabil-
ity of such equipment on trucks.
The following text summarizes overfill and transfer
spill prevention methods for underground storage sys-
tems.
It should be emphasized that the methods described
are far more prevalent with aboveground systems
(where a spill or overfill would be highly visible) than
with underground systems. The evolving awareness of
hazardous conditions that result from loss of product
from an underground system and the financial accounta-
bility for site cleanup are changing the picture. The day
of the deliveryman relying on a calibrated stick to deter-
mine available capacity in a buried tank may be pas-
sing.
B. OVERFILL PREVENTION SYSTEMS
FOR UNDERGROUND
STORAGE TANKS
1. Elements of an Overfill Prevention System.
Overfill protection is accomplished by measuring
and controlling the level of liquid in a tank. A partial
system may include only a gauge which indicates liquid
height in the tank. A sophisticated system could include
automatic flow control system and a backing audible
high level alarm to warn the operator of emergency con-
ditions.
The elements of a complete overfill prevention sys-
tem are highlighted in table 2.4-1. These include the
following:
•Sensors, which detect the level of liquid in the
tank and indicate the liquid level through gauges
or other types of indicators.
•High level alarms, which are activated to warn the
operator of an impending overfill condition.
•Automatic shut-off devices of systems which pre-
vent overfilling from occuring.
Table 2.4-1
Elements of a Good Overfill
Prevention System
-	Level sensing device.
-	Level indicating device.
• High level alarm.
-	Automatic shut-off control system.
-	Interlocking of the unloading process and the
overfill prevention system so that loading cannot
take place if the overfill prevention system is in-
operative.
-	Bypass prevention so that the overfill prevention
system cannot be overridden by the operator.
82

-------
It may be desirable to interlock the unloading pro-
cess with the overfill prevention system so that loading
cannot take place unless the overfill prevention system
is operative. A bypass prevention feature should also be
included so that the overfill prevention system cannot be
overridden by the operator.
Level Sensing Devices and Indicators. There are
a variety of level sensing devices that have been mar-
keted for detecting liquid levels in bulk storage tanks.
These devices generally sense liquid characteristics,
such as capacitance or thermal conductivity, or operate
on such common principles as buoyancy, differential
pressure, and hydrostatic head. Devices which operate
based on these common priniciples are generally inde-
pendent of product flow rate, pressure and temperature
[13].
Indicators for underground tanks are typically re-
motely mounted (e.g., in a control room), although
above-the-tank gauges may be employed in some cases.
The devices are typically gauges, although more sophis-
ticated electronic devices may be used in some overfill
prevention systems.
High Level Alarms. Overfill alarms may be visual
or audible instruments which are remotely mounted.
Audible alarms may be the preferred type of alarm be-
cause they do not require visual monitoring. However,
when several tanks are being monitored in the same
control room, individual warning lights are generally
provided for each tank. Ideally, an audible alarm would
also be included in such systems to alert the operator
that one of the tanks is overfilling.
Automatic Shut-off Controls. Automatic shut-off
control systems interface with level sensing devices to:
(1) prevent tank overfilling by shutting off the tank
loading pump at a preset high level; (2) prevent damage
to the tank unloading pump by shutting it off at low
level; (3) operate various flow valves to control product
flow. These control systems receive a signal from the
level sensing device which is transmitted electrically or
pneumatically to the control system. Pneumatic devices
require a regulated supply of clean and dry instrument
air, generally at 20 pounds per square inch (psi). Elec-
tric (or electronic) devices generally require 115V line
voltage. Table 2.4-2 shows the characteristics of
pneumatic and electronic controls.
Floated activated, capacitance, ultrasonic, optical,
and thermal conductivity sensors can be readily used in
underground tanks. Their applications are summarized
in Table 2.4-3 and discussed below. The other types of
sensors are used in aboveground storage systems , and
are discussed in Part II, Chapter 4 of this report.
Float-Actuated Devices. Float-actuated devices are
characterized by a buoyant member which floats at the
surface of the liquid. Float-actuated devices may be
classified on the basis of the method used to couple the
float motion to the indicating system. Examples of clas-
sifications include tape float guages and float vent
valves.
A simple tape float gauge designed for use in un-
derground gasoline tanks is shown in Figure 2.4-1. The
device provides a local (above the tank) readout of both
gasoline and water levels while prohibiting vapor loss.
Float vent valves are simple, inexpensive devices
that are used to prevent overfilling of underground fuel
tanks. These devices, which are shown in Figure 2.4-2,
are installed in the tank's vent line. The float closes the
vent line when high liquid level is attained, thus block-
ing the escape of air. This action causes the pressures
inside the storage tank to equalize with the discharge
head in the tank truck, thereby interrupting the flow of
liquid.
2. Specific Level Sensing Devices.
The types of level sensing devices available for liq-
uid level detecting in bulk storage tanks can be classi-
fied as follows:
•Float-actuated devices
•Displaced devices
•Hydrostatic head sensors
•Capacitance sensors
•Ultrasonic devices
•Optical devices
•Thermal conductivity sensors
83

-------
Table 2.4-2


Characteristics of Pneumatic and Electronic Controls

Feature
Pneumatic
Electronic
Transmission distance
Limited to few hundred feet
Practically unlimited
Standard transmission signal
3-15 psi practically universal
Varies with manufacturer
Compatibility between
instruments supplied by
different manufacturers
No difficulty
Nonstandard signals require special
consideration and may not be
compatible
Control valve compatibility
Controller output operates
control valve operator
Pneumatic operators with
electropneumatic converters or
electrohydraulic or electric motor
operator required
Compatibility with digital
computer or data logger
Pneumatic-to-electric
converters required for all
inputs
Easily arranged with minimum added
equipment
Reliability
Superior if energized with
clean dry air
Excellent under usual environmental
conditions
Reaction to very low
(freezing) temperatures
Inferior unless air supply is
completely dry
Superior
Reaction to electrical
interference (pickup)
No reaction possible
No reaction with the system if
properly installed
Operation in hazardous
locations (explosive
atmosphere)
Completely safe
Intrinsically safe equipment availbie
equipment must be removed for
most maintenance
Reaction to sudden failure
of energy supply
Superior-capacity of system
provides safety margin- _
backup inexpensive
Inferior-electrical failure may disrupt
plant-backup expensive
Ease and cost of installation
Inferior
Superior
System compatibility
Fair-requires considerable
auxilary equipment
Good-conditioning and auxilary
equipment more compatible to
systems approach
Instrument costs
Lower if installation costs
are not considered
Higher-becomes competitive
when total including installation
is considered
Ease and cost of maintenance
Fair; procedures more
readily mastered by
people with minimim of training
Good-depends upon capability
of personnel
Dynamic response
Slower but adequate for
most situations
Exceilent-frequently valve becomes
limiting factor
Operation in corrosive
atmosphere
Superior-air supply becomes
a purge for most instru-
ments
Inferior, unless special consideration
is given and suitable steps taken
Performance of overall
control systems
Excellent, if transmission
distances are reasonable
Excellent-no restriction on
transmission distance
Politics (the unmentioned
factor that frequently
pops up)
Generally regarded as
acceptable but not the
latest thing
Often regarded as the latest and most
modern approach
Source: Reference 2


84

-------
Table 2.4-3



Level Detection Devices for Underground
Storage Tanks


Type
Monitor
Liquid
Level
Level
Indication
Alarm and
Shutoff
Response [1]
Float Actuated Devices
Tape float gauges
Yes
Gauge
Interfaces with electronic
or pneumatic controls.
Float vent valves
No
None
Automatic Shut-off.
Capacitance devices
Yes
Gauge
Audible alarm and
automatic shutoff elec-
tronic controls.
Thermal conductivity
devices
Yes
Gauge
Audible alarm and
automatic shutoff elec-
tronic controls.
Ultrasonic devices
Yes
Gauge
Audible alarm and
automatic shutoff elec-
tronic controls.
Optical devices
Yes
Gauge
Audible alarm and
automatic shutoff elec-
tronic controls.
The device also includes a pressure build-up relief
bleed hole. Once flow from the tank truck has ceased
due to pressure equalization, the storage tank fill line
can be disconnected. Then, as vapor escapes through
the float vent valve bleed hole, the liquid remaining in
the fill line can drain into the tank. If dry disconnection
couples are used, the liquid will be held in the transfer
line until this draining can occur, thus preventing any
spillage of product.
This device was developed as part of the Vapor
Recovery Stage I system. Its purpose is to prevent prod-
uct spillover into the vapor manifold, which might re-
sult in lead contamination of an unleaded gasoline
grade. Its use as overfill protection has merit, but it is
not used for that purpose generally.
The float vent valve must be installed in an "ex-
tractable tee" connection, which permits removal of the
float valve for tank testing. The Kent-Moore (Heath
Petro-Tite Tank Tightness) Test cannot be run with the
valve in place.
Float-actuated devices are made of a variety of ma-
terials, including aluminum, stainless steel and coated
steel, depending upon the application [10]. They may
be used in conjunction with pneumatic or electronic de-
vices to operate valves, pumps, remote alarms or auto-
matic shut-off systems.
Capacitance Sensors [7], Devices that operate
based on the electrical conductivity of fluids may be
used to monitor liquid level. A typical device consists
of a rod electrode positioned vertically in a vessel, the
other electrode usually being the metallic tank wall. The
electrical capacitance between the electrodes is a meas-
ure of the height of the interface along the rod elec-
trode. The rod is usually electrically insulated from the
liquid in the tank by a coating of plastic.
Capacitance devices are suitable for use with a
wide range of liquids, including the following: petro-
leum products, such as gasoline, diesel fuel, jet fuel and
no. 6 fuel oil: acids; alkalis: solvents: and other hazard-
ous liquids. They may be used in conjunction with elec-
tronic controls to operate pumps, valves, alarms or
other external control systems.
Thermal Conductivity Sensor [18,19]. Devices
which operate on the principle of thermal conductivity
of fluids may be used to monitor liquid level. A typical
device consists of two temperature-sensitive probes con-
nected in a Wheatstone bridge (a type of electrical cir-
cuit configuration). When the probes are in air or gas.
a maximum temperature differential exists between the
active and reference sensors, which results in a great
imbalance in the bridge circuit and a correspondingly
high bridge voltage. When the probes are submerged in
85

-------
Figure 2.4-1
Tape Float Gauge for
Underground Storage Tank
CAP ft 900T
ASStMSLY
AOArroft boot
ASSDMLT
gPPfNMAXfT
iwiimT
OPW 114-DW
The OPW 114-OW
Tank Gage provides
a sealed, accurate
read-out of both gas-
oline and water for
underground storage
tanks, it pronibits va-
por escape, and
makes inventory
control easy. With
this tank gage, the
operator merely lifts
trie cap and reads
the scale through a
viewing glass. Should
condensate form on tne underside oi the gtass. a
turn ot the cap wipes ii clean
These tank gages are lor use m 4" riser pipes,
and are pre-assemoied at the factory One man can
install the 114-OW tank gage by using the fumisned
instruction sneet as a guide.
Materials
Body: hard coat aluminum
Cap: hard coal aluminum
Gaskets: buna-N
Pulley Acetai
Tape: steel with epoxy paint
OPW 114-SW
Similar to OPW 114-OW above, except it indi-
cates product level only.
Source: OPW Division/Dover Corp.
a liquid, the temperature between the sensors is
equalized and the bridge is brought more nearly into
balance. The probes may be installed through the side
wall of a tank or pipe, or assembled together on a self-
supporting mounting and suspended through a top con-
nection on the tank.
Thermal conductivity devices may be used to con-
trol level with great accuracy. They may be used with
any liquid regardless of viscosity or density. They may
also be used with immiscible liquids and slurries and in
conjunction with electronic controls to operate pumps,
valves, alarms or other external control systems.
Ultrasonic Sensors [16,171. Devices which operate
on the principle of sonic-wave propagation in fluids also
may be used to monitor liquid level. These devices use
a piezoelectric transmitter and receiver, separated by a
short gap. When the gap is filled with liquid, ultrasonic
energy is transmitted across the gap to a receiving ele-
ment thereby indicating the liquid level. These devices
may be used in conjuction with electronic controls to
operate pumps, valves, alarms or other external control
systems.
Another sonic technique used for level measure-
ment is a sonar device. A pulsed sound wave, generated
by a transmitting element, is reflected from the interface
between the liquid and the vapor-gas mixture and re-
turned to the receiver element. The level is measured in
terms of the time required for the sound pulse to travel
from the transmitter to the vapor/liquid interface and re-
turn.
Optical Sensor [9]. Devices which operate on the
principle of light beam refraction in fluids may be used
to monitor liquid level. An optical liquid level monitor-
ing system consists of a sensor and an electronic control
device. A specific electronic signal is generated and
aimed at the tank mounted sensors. The sensors convert
the electronic signal to a light pulse. This light pulse is
transmitted into the tank by fiber optics, through a
prism and out again via fiber optics. The light pulse is
then converted to a specific electronic signal to indicate
the liquid level. A distinct advantage of this type of sys-
tem is that it is self-checking. Any interruption will
sound the alarm, so if equipment is damaged or mal-
functions the operator is alerted.
Figures 2.43 and 2.44 show typical applications of
the optical liquid level sensing system for a tank truck
and a bulk storage tank respectively. The sensor detects
the level of liquid in the tank and sends a signal to the
controller device (i.e. control monitor) which in turn ac-
tivates the shut-off valve or the level alarm.
C. TRANSFER SPILL PREVENTION SYSTEMS
Spill prevention during transfer operations can be
accomplished by using couplings equipped with spring
loaded valves which automatically block flow when the
hoses are disconnected. These include quick-disconnect
couplings equipped with ball valves and dry-disconnect
couplings. Emergency shut-off valves may also be pro-
vided in the product transfer line to stop flow in case
of fire. Applications of these spill prevention devices
are summarized in Table 2.4-4 and discussed below.
86

-------
Figure 2.4-2
Float Vent Valves Used
For Overfill Prevention
Source: OPW Division/Dover Corp.
1.	Check Valves
Check valves are commonly used in the discharge
piping of a pump or the fill line of a tank to prevent
reversal of flow. Check valves are available in three
basic designs: (1) swing check valves; (2) lift check
valves; and tilting-disk check valves. They are available
in a wide wariety of sizes and materials of construction
to suit most applications. A more complete description
of these devices can be found in Part II, Chapter 2 of
this report.
2.	Couplings
The use of tight couplings is essential to prevent
spills when transferring hazardous products from one
storage tank to another. Many types of couplings are
available and their selection depends on temperature.
pressure and the chemical properties of the material
conveyed. The higher the temperature and/or pressure,
the more securely the coupling must be attached. Also,
the material being conveyed must not damage the cou-
pling. The factor which determines the amount of pres-
sure a coupling will withstand is generally the strength
of the hose-coupling connection. If correctly applied
(and at moderate working temperatures) bolt clamps will
handle low pressure, bands will take low to medium
pressures, and interlocking clamps and swaged or
crimped ferrules will handle high pressures.
As mentioned earlier in this chapter, quick-discon-
nect couplings are commonly used because they are
generally lighter and easier to handle than other types
of couplings. However, precautions must be taken to
prevent spill or loss of the product remaining in the
transfer lines when these types of couplings are used.
Quick-disconnect couplings equipped with ball valves
and dry-disconnect couplings are used to minimize spills
when the hoses are disconnected. Dry-disconnect cou-
plings are the best type of coupling available in terms
of product spill control. They are equipped with a
spring loaded valve which is normally closed until the
coupling is attached and the valve is manually opened
with a lever. Figure 2.4-5 demonstrates the difference
between the types of couplings.
Another good product transfer practice is the selec-
tive use of couplings and adapters to preclude the mix-
ing of incompatible liquids. By carefully selecting cou-
plings and adapters that are only compatible with each
other, one can prevent undesired mixing of products.
Imbiber beads are useful for soaking up small
spills, for example in the fill box. These beads absorb
hydrocarbons and swell to many times their original
size, but do not absorb water.
D. OPERATING PRACTICES
FOR OVERFILL PROTECTION
Certain operating practices specified in Publication
385 of the National Fire Protection Association may be
used to prevent overfilling of tanks [14]. Practices that
are applicable to the transfer of any hazardous liquid in-
clude the following:
•Loading and unloading of tank vehicles shall be
done in approved locations.
•The driver, operator, or attendant of any tank ve-
hicle shall not remain in the vehicle and shall not
leave the vehicle unattended during the loading or
unloading process. The delivery hose, when at-
tached to a tank vehicle, shall be considered to be
a part of the tank vehicle. Some companies prefer
to have their own trained personnel conduct all
unloading operations so as to minimize the poten-
tial for human error. Whoever does the unloading
must be cognizant of the potential problems and
87

-------
Figure 2.4-3
Optical Liquid Level Sensing System
For Tank Truck
SENSOR
'X I i
DM
c
COILED CABLE
Source: OPW Division/Dover Corp.
Figure 2.4-4
Optical Liquid Level Sensing System
For Bulk Storage System
CONTROL MONITOR
r SENSOR
CONOUIT RUN TYPICAL
Source: OPW Division/Dover Corp.
Table 2.4-4



Tranfer Spill Prevention Systems



System
Function
Spill
Control
Applications
Ordinary quick-disconnect
coupling
Product
transfer
None
Tank vehicles and storage
tanks.
Quick-disconnect coupling
equipped with ball valve
Product
transfer
Built-in valve reduces
spills from disconnect
hoses.
Tank vehicles and storage
tanks.
Dry-disconnect coupling
Product
transfer
No spills from
disconnected hoses.
Tank vehicles and
storage tanks.
Emergency shut-off
valves
Flow
control
A fusible metal link melts
and closes the valve in
case of fire or impact.
For use any place that, in
the event of fire, it is impor-
tant to stop flow.
88

-------
dangers (overfilling, leaks, vapor or liquid explo-
sions, fire, etc.), and must remain alert at all
times. Human error is a major cause of transfer
spill incidents, and in most instances spills could
be avoided through proper personnel training and
alert observation of all operations.
•When transferring Class I (flammable) liquids,
motors of tank vehicles or motors of auxiliary or
portable pumps shall be shut off during making
and breaking hose connections. If loading or un-
loading is done without requiring the use of the
motor of the tank vehicle, the motor shall be shut-
off throughout the transfer operation of the liquid.
These precautions should be taken to minimize the
possibility of fire or explosion.
•No cargo tank (or compartment) containing vol-
atile, flammable or combustible liquid may be
fully loaded. Sufficient space (outage) must be
provided to prevent leakage due to thermal expan-
sion of the liquid transported. One percent is the
minimum outage required.
Please refer to NFPA-385 for more information on
loading and unloading practices [14].
Other precautions which may be taken to prevent
overfills and spills include the following:
•The use of labels, markings or color codes on
hoses and special couplings that can be used only
for transferring product to prevent accidental mix-
ing-of incompatible materials.
•Periodic inspection of hoses for leaks.
•Ensuring that the operator of any loading/unload-
ing operation is properly trained and is aware of
all potential problems. As stated earlier, a high
percentage of transfer spills are caused by human
error.
Figure 2.4-5
Types of Couplings
]. QUICK OltCOHMtCT PLUS SAIL VALVE
3. ORT OISCONNCCT
Source: OPW Division/Dover Corp.
89

-------
References
1.	Perry, R.H., Chilton, C.H., Chemical Engineer's
Handbook, Chapter 22, McGraw-Hill Book Co.,
Fifth Edition, 1221 Avenue of the Americas, New
York, New York 10020, 1973.
2.	Anderson, N.A., Instrumentation for Process Mea-
surement and Control, 2nd Edition, Chilton Book
Co., 5601 Chestnut Street, Philadelphia, PA 19139,
1972.
3.	Andermatten, R., Overviil Protection System for
Underground Storage Tanks of Hazardous Liquid
Products, 16 S. Lewis Place, Rockville Centre,
N.Y. 11570, July 1, 1980.
4.	Annual Report of The New York State Oil Spill
Prevention, Control and Compensation Program,
April 1, 1978 to March 31, 1979. New York State
Department of Environmental Conservation, 50
Wolf Road. Albany, N.Y. 12233.
5.	Annual Report of The New York State Oil Spill
Prevention, Control and Compensation Program,
April 1, 1979 to March 31, 1980, New York State
Department of Environmental Conservation, 50
Wolf Road, Albany, N.Y. 12233.
6.	Petrometer Corp.. Liquid Level Indicating and Con-
trol Systems, Catalog No. 75. Petrometer Corp.,
P.O. Box 245, New Hyde Park, New York 10040.
7.	Amprodux Inc., Solid-State Electronic Bin and
Tank Gauges, Level Alarms, and Level Control In-
struments and Systems. Catalog 15.10/AM. Am-
produc. Inc., 150 West 28th Street, New York,
N.Y. 10001.
8.	Kodata, Inc., Level Measuring System for Bulk Liq-
uids or Solids, Bulletin No. 8090-25, Kodata. Inc.,
3621 McCart Street or P.O. Box 11528. Fort
Worth, Texas 77610.
9.	Dover Corp., Optic Liquid Level Sensing System
for Petroleum Transportation and Storage Applica-
tions, Bulletin OLLS6-80', Dover Corp., OPW Di-
vision, 9393 Princeton-Glendale Road. P.O. Box
40240, Cincinnati, Ohio 45240, June. 1980.
10.	Scully Electronic Systems, Inc., New Scully-Moor-
mann Development for Remote Readout Inventory
control for Existing and new Moormann Liquid
Gauges, Scully Electronic Systems Inc.. 70 Indus-
trial Way, Wilmington, Mass 01887.
11.	Dover Corp., Service Station Vapor Recovery Prod-
ucts and Systems, Catalog SVR, Dover Corp..
OPW Division, 9393 Princeton-Glendale Road,
P.O. Box 40240, Cincinnati, OH 45240, February,
1981.
12.	Perry, R.H., Chilton, C.H., Chemical Engineer's
Handbook, Section 6, McGraw-Hill Book Co.,
Fifth Edition, 1221 Avenue of the Americas, New
York, New York 10020, 1973.
13.	Emco Wheaton, Inc., Fluid Handling Systems,
Catalog 7-8/73, Emco Wheaton, Inc., Chamberlain
Blvd., Conneaut, Ohio 44030, Revised April,
1977.
14.	NFPA 385, Tank Vehciles for Flammable and
Combustible Liquids, National Fire Protection As-
sociation, Batterymarch Park, Quincy, MA. 02269.
1979.
15.	Dover Corp., OPW Kamvalok - Dry Disconnect
Couplings, Dover Corp., OPW Division, 9393
Princeton-Glendale Road, P.O. Box 40240, Cincin-
nati, OH 45240.
16.	Envirotech Corp., Sensall 880 Ultrasonic Non-Con-
tact Continuous Level Transmitter,, Catalogue B-
8800, National Sonics. 250 Marcus Blvd., Haup-
pauge, NY 11787, March, 1981.
17.	Envirotech Corp., Sensall 880 Ultrasonic Non-Con-
tact Continuous Level Transmitter,, Catalogue B-
800. National Sonics, 250 Marcus Blvd., Haup-
pauge, NY 11787, March, 1980.
18.	Fluid Compnents, Inc., Heat Actuated Liquid Level
Controller Model 8-66, Bulletin 8-66/1, Fluid Com-
ponents. Inc., P.O. Box 1165, Canoga Park,
California 91304.
19.	Fluid Compnents. Inc.. Model FR72 Series Liquid
Level Controller, Bulletin FR72-LL/1, Fluid Com-
ponents. Inc.. P.O. Box 1165, Canoga Park.
California 91304.
20.	Dover Corp., OPW Engineered Service Station
Products, Dover Corporation/OPW Division, 9393
Princeton-Glendale Road, P.O. Box 40240, Cincin-
nati. OH 45240.
90

-------
Part II
CHAPTER 5:
LEAK AND SPILL
MONITORING FOR
UNDERGROUND STORAGE
A. INTRODUCTION
Before any extensive discussion begins, a distinc-
tion must be made between system monitoring and sys-
tem testing. Monitoring means either.
a)	Early warning leak detection systems that pro-
vide continuous surveillance for leaks and spills.
b)	Area-wide surveillance methods that may be
used to investigate or pinpoint the source of a
spill or leak.
Testing refers to special equipment and methods
that are not part of normal operations. System testing is
used to determine whether a tank or pipe system is leak-
ing at a particular time. (Chap. II-6 discusses testing at
length). Table 2.5-1 compares the various leak monitor-
ing systems.
B. EARLY WARNING
LEAK DETECTION SYSTEMS
Early warning leak detection systems typically pro-
vide continuous surveillance for the presence of a leak
or spill. The types of early warning monitoring systems
most widely used are the following:
•Inventory monitoring (also called inventory con-
tol).
•Interstitial monitors in double-walled tanks.
•Systems that monitor the storage tank excavation.
These types of systems include observation wells,
U-tubes, and wire grids. The types of leak sensors
used include:
•	thermal conductivity sensors;
•	electric resistance sensors; and
•	gas detectors.
1. Inventory Control
Early detection of leaks may be achieved by proper
product accounting (i.e., recordkeeping), regular inspec-
tions of the visible parts of the product handling system,
and prompt recognition of the conditions that indicate
leaks in underground tanks and piping. Inventory moni-
toring is a technique that is widely applicable to any
stored or transported product.
Evidence of leakage from buried tanks and
pipelines can be gathered from inventory control records
and from abnormal operation of pumping equipment.
The following are some of the more obvious symptoms
of such leaks:
•Loss of product in a tank during periods when
product is not dispensed usually indicates a leak-
ing tank, but might also indicate faulty accounting
or metering of the product, theft, or extreme tem-
perature change.
•An unaccountable increase in water in an under-
ground tank may be caused by a leak in the tank
if the ground surrounding it is saturated. Under
such circumstances, water may leak into the tank
instead of product leaking out. The increase in
water may also be caused by a leaking gauge or
fill cap, and these should be examined and made
watertight, if necessary, before concluding that the
tank is at fault.
•Increasing differences between the amount of
product received and dispensed may indicate a
meter calibration problem, theft, or a leak in tanks
or piping.
•Where fill boxes are located remotely from the
tanks, large differences appearing consistently be-
tween the amounts invoiced and the tank gauges
after deliveries may indicate a leak in the remote
fill line. In such event, the line should be tested.
•A hesitation in the delivery from a standard dis-
pensing pump may indicate a leak in the suction
¦piping, although such hesitation may also be
caused by a leaking foot valve or, in warm wea-
ther, by vaporlock. Should this occur, the inven-
tory control records may indicate whether the
cause is mechanical or whether product is actually
being lost.
•In a remote pumping system, meter spin without
product delivery may indicate a leaking pipe.
•Gasoline odor in spaces below ground adjacent to
the tank may be evidence of underground leaks,
whether in the tank or piping. However, such
odors may also be evidence of underground leaks,
whether in the tank or piping. However, such
odors may also be evidence of product spills dur-
ing product delivery and tank filling.
Should the operator observe any of the foregoing
symptoms, he should immediately notify those responsi-
ble for maintaining the equipment. He should not at-
tempt to correct the condition himself, as the operation
may involve some hazard and may require special
equipment. Furthermore, in some locations, only spe-
cially licensed mechanics can work on storage equip-
ment.
91

-------
Table 2.5-1





Comparison of Various Leak Monitoring Techniques






Substances
Relative

Approach
Description
Applications
Detected
Cost
Advantages; Disadvantages
Inventory
A system based on product
Any storage
Any product
Low
The technique is widely appli-
Control
recordkeeping, regular inspec-
tanks and buried
stored or

cable to any product stored or

tions, and recognition of the
pipelines.
transported.

transported in pipelines. How-

conditions which indicate



ever. it requires good bookkeep-

leaks.



ing, and will not detect small





leaks.
Thermal
Uses a probe that detects the
Can monitor
Any liquid.
Medium
Primary advantage is early
Conductivity
presence of stored product by
groundwater or


detection which makes it possible
Sensors
measuring thermal
normally dry


for leaks and spills to be cor-

conductivity.
areas.


rected before large volumes c?





material are discharged. Typical-





ly requires inch of product on





groundwater to guarantee detec-





tion of product/water interface





in wet (groundwater)





applications [16].
Electric
Consists of one or a series of
Can monitor
Any liquid.
Medium
Primary advantage is the early
Resistivity
sensor cables that deteriorate
groundwater or


detection of spills. Once a leak
Sensors
in the presence of the stored
normally dry


or spill is detected, the sensors

product, thereby indicating a
areas.


must be replaced. Can detect

leak.



small as well as large leaks.
Gas Detectors
Used to monitor the presence
Areas of highly-
Highly vol-
Medium
Once the contaminant is present

of hazardous gases in vapors
permeable, dry
atile liquids.

and detected, gas detectors are

in the soil.
soil, such as ex-
such as

no longer of use until contamin-


cavation backfill
gasoline.

ation has been cleaned up.


or other per-





meable soils.





above ground-





water table.



Sampling
Grabbing soil or water
Universal: pri-
Any
High
Highly accurate intermittent

samples from area for
marily used to
substance

evaluation tool. However, does

analysis.
collect ground-


not provide continuous


water samples, as


monitoring.


would be the case





with tanks stored





in high ground-





water area.



Interstitial
Monitors pressure level or
Double-walled
Pressure
High
Accurate technique which is
Monitoring
vacuum in space between
tanks.
sensors

applicable with any double-
in Doi'jle-
walls of a double-walled tank.

monitor tank

walled tanks.
Walled Tanks


integrity and





are appli-





cable with





any stored





liquid. Fluid





.sensors mon-





itor presence





of any liquid





in a normally





dry area, and





are also





applicable





with any





stored liquid.


92

-------
Tabic 2i*l continued
Groundwater
Monitoring
Weils
(wet weils)
Vapor (sniff)
Wells
Dyes and
Tracers
Wet wells are used to detect
and determine the extent of
contamination in ground-
water tables.
Vapor wells are used to detect
and monitor the presence of
hazardous gases and vapors
in the soil.
Substances with a characteris- Area-wide mon- Dye itself
Area-wide or
Any hazard-
local monitoring
ous liquids
for groundwater
which can
contamination
be detected
from under-
by on-site in-
ground storage
struments or
tanks and pipe-
laboratory
lines. May be
analysis.
used for periodic

sampling or may

employ one of the

sensors described

above to detect

leaks or spills.

Area-wide or
Many dif-
local monitoring
ferent com-
of the soil sur-
bustible and
rounding under-
non-combus-
ground storage
tible gases
tanks and pipe-
and vapors.
lines.

tic color or other characteris-
tics (e.g.. radioactive tracers)
that can be used to trace the
origin of a spill.
itoring of under- is detected
ground tanks and visually or
buried pipelines. with the use
of instru-
ments.
Medium The type, number and location
to High of wet wells depends upon the
site's hydrogeology, the direction
of groundwater flow and the
type of spill containment and
spill collection systems used.
Low The type, number and location
of vapor wells depends upon the
extent of the spill, the volatility
of the product, and the soil
characteristics. Vapor wells are
subject to contamination from
surface spills and cannot be used
at contaminated sites.
Low Dye or tracer could be low in
Medium cost, but the time required to
perform a study could be great.
Also may require the drilling of
observation wells to trace the dye
or other material. Radioactive
tracers require a license and ap-
proval from the Nuclear Regula-
tory Commission or the U.S.
Department of Labor. There-
fore, they are generally
discouraged.
93

-------
There are a number of factors that limit the accu-
racy of inventory control as a leak detection method.
These include the following:
•Product thermal expansion. Fluctuations in tem-
perature can lead to expansion, contraction, evap-
oration and/or condensation of the stored product,
thereby affecting inventory monitoring results.
The relationship between temperature and storage
volume is addressed in Part I of this report.
•Errors associated with faulty reading of dip stick
measurements.
•Errors associated with resolution in meter read-
ings. All meters have aii associated level of error,
typically on the order of 0.5% of the level of re-
solution of the meter.
Given these limitations in accuracy, even a care-
fully conductd inventory monitoring program can only
detect leaks that are an appreciable fraction (typically
0.75%) of the stored volume. It should be emphasized
that inventory control, conscientiously followed, is the
first defense against leaks. Measuring the liquid level,
(and the water level with a water finding paste) twice
a day and comparing these levels with product de-
liveries and sales will indicate trends of product loss or
water gain in a short time. Major oil companies require
inventory records. Any unreported losses become the re-
sponsibility of the operator and reports of consistent
product losses are followed by testing the suspected
tank for leaks .
ij,, Interstitial Monitoring in
Double-Walled Tanks
An early warning monitoring technique characteris-
tic of double-walled tanks involves monitoring the space
betwen the inner and outer walls of the tank, using
either fluid sensors or pressure sensors. Pressure sensors
would be used to monitor tanks that either have a vac-
uum drawn in the space between walls, or have that
space pressurized. Failure of either the inner or outer
wall is detected by loss of vacuum or pressure. Fluid
sensors, on the other hand, would be located between
the tank walls to detect the presence of a liquid due to
failure of the inner wall (detecting stored product) or the
outer wall (detecting water). (These systems may be ap-
plied at atmospheric pressures to vaulted tanks.)
3. Tank Excavation
Monitoring Sensors
Tank excavation monitoring systems are aimed at
detecting a spill or leak before the contamination
spreads beyond the tank excavation or its immediate
surroundings. The leak or spill sensing mechanisms that
may be used in tank excavation monitoring systems in-
clude thermal conductivity sensors, electrical resistivity
sensors, gas detectors and sample analysis.
Thermal Conductivity Sensors [1.2]. Thermal
conductivity sensors detect changes in the thermal con-
ductivity of their surrounding environment to determine
if a leak or spill has occurred. These types of sensors
can be used in wet or dry applications (i.e. areas of
either low or high groundwater), ard are particularly ap-
plicable for the detection of nydrocart>ons such as
gasoline, gasohol, fuel oils, alcohols, and
trichloroethylene.
A system using a thermal conductivity sensor typi-
cally consists of an electronic control device that is con-
nected by cable to a thermal conductivity probe. The
probe is fitted with a sensor that determines if the moni-
tored area is dry, wet with water, or wet with some
other substance. The control device may be located up
to 1,000 feet from the probe and can continuously indi-
cate the site condition through indicator lights. A non-
water liquid presence may also be indicated by an audi-
ble alarm, and recorded using a chart recorder. A relay
contact that can activate external alarms, recovery
pumps or other automatic controls can also be provided.
Figure 2.5-1 shows examples of thermal conductiv-
ity sensors installed in a diked area, a dry well or sump,
and a wet well (in the groundwater table). When used
to monitor a groundwater table, one sensor located in
a monitoring well will only indicate the presence of
contamination but not the extent of it. By using several
. sensors located at various levels the thickness of the
contaminant layer may be ascertained.
Electrical Resistivity Sensors [3]. Systems em-
ploying this leak detection technique rely on the change
in resistance in a wire due to exposure to the stored
product to indicate the presence of leak or spill. The
key to systems of this type is the use of wires or wire
coatings that are highly susceptible to degradation when
exposed to stored product. For example, bare steel
wires may be used in acid storage areas or bare
aluminum wires may be used in areas storing caustics.
Correspondently, if the stored liquid is not corrosive to
metals, the wires must be coated with a degradable ma-
terial, such as rubber coatings in areas storing aromatic
solvents. The wires in turn are connected to an electri-
cal sensing device that passes a current through them to
evaluate their electrical properties. Any degradation
of the wire or its coating will result in a significant
change in the circuit resistivity, thus indicating the exis-
tance of a product leak or spill.
Electrical resistivity sensors are applicable for
either dry or wet (in groundwater) applications. Am-
bient temperature and soil moisture should have mini-
mal effects on sensors of this type, particularly in appli-
cations involving coated wires. The drawbacks of these
types of leak detection devices include the following:
94

-------
Figure 2.5-1
Typical Applications of a Leak Monitoring
System Based on Thermal Conductivity

POLLULfiflT
CSS
siMsons
-t • t •
i i T i
It ft
i a a + -
O
SURFACE WATER

GROUNDWATER
|RY SUMP
Source: Reference 1
95

-------
•Once a leak has been detected, the sensing wire
must be replaced.
•They cannot be used in a previously contaminated
well or soil unless the contamination has been re-
moved. Otherwise they will rapidly deteriorate
and require replacement.
The control units associated with electrical resistiv-
ity sensors can be designed to interface with audible
alarms, visual alarms (e.g., indicator lights), control
equipment such as pumps or valves, and computer con-
trols. Occasional checks of systems of this type would
be required to insure that the power supply and all con-
trols are in working order.
Gas Detectors [4]. Gas detectors are available to
detect a large number of combustible and non-combusti-
ble gases and vapors. These types of devices are gener-
ally applicable in areas of permeable soil or backfill,
where gases and vapors are likely to migrate easily. Gas
detectors are particularly applicable in instances where
the stored product is highly volatile and the storage area
(excavation) is relatively dry (free of groundwater).
There are a wide variety of both portable and per-
manent gas detection devices available that may be op-
erated in conjunction with audible or visual alarm sys-
tems.
Sample Collection. Sample collection typically in-
volves collecting samples from a well in the excavation "
area. Sample collection is an accurate but expensive-
method of leak detection that is particularly applicable
in areas of high groundwater where direct groundwater -
contamination is of concern. Sample analysis can be
performed using any of several techniques such as mass
spectrometry and gas chromatography; therefore sample
collection can be used to detect any stored product.
However, sample collection is an intermittent as op-
posed to a continuous monitoring technique; sample
cannot be collected 24 hours a day, 365 days a year.
Therefore, sample collection may not be as desirable a
monitoring technique as those described above.
4. Tank Excavation
Monitoring Systems
There are several types of leak monitoring systems
which may be employed using the leak sensors or detec-
tion techniques described above to detect leaks in or
around underground tank storage areas. These system
types include the following:
•Wire grids.
•Observation wells.
•U-tubes.
Table 2.5-2 summarizes the applicability of the
types of leak sensors of detection techniques described
above to these types of leak monitoring systems.
Wire Grids. This type of leak detection system
employs electrical resistivity sensors in a wire grid lo-
cated either within or just outside the containment re-
gion (e.g. just inside or outside the containment area
synthetic liner). The wire grid is connected to a mini-
computer that continuously monitors the electrical prop-
erties of each wire in the grid. If a leak occurs, the
mini-computer can determine which wires in the grid
have had their electrical properties altered, thereby iden-
tifying the location and extent of the leak. A drawback
of this type of system is that is is susceptible to disabl-
ing by a spill. The insulation around the grid wire is
dissolved, thereby registering a change in resistivity.
Observation Wells. Observation wells are most
commonly used in areas of high groundwater, where the
underground tank is likely to be anchored in the ground-
water during normal operation. They may employ any
of the types of leak sensors described above to provide
continuous leak surveillance. An example of an obser-
vation well installation is shown in Figure 2.5-2.
Observation wells typically consist of a 4 inch di-
ameter (schedule 40) PVC pipe driven into the tank ex-
cavation. The wells are constructed with a well screen
long enough to provide a length of 5 feet or more above
the water table, or to the well cap, and extending a
minimum of 5 feet into the groundwater or 2 feet below
the tank bottom, whichever is greater. Well screens typ-
ically have a slot size of 0.02 inches, and are extended
to grade and covered with a water proof cap which is
capable of being sealed.
U-tubes. A U-tube typically consists of a 4 inch
diameter (schedule 40) PVC pipe installed as shown in
Figure 2.5-3. The horizontal segment of the pipe is
half-slotted (typical slot size - 0.06 inches), wrapped
with a mesh cloth to prevent backfill infiltration, and
sloped (pitched) toward the sump with a slope on the
order of '/» inch per foot. At the higher end of the pipe
there is a 90 degree sweep to a vertical pipe that is ex-
tended to grade. At the other (lower) end of the hori-
Table 2.5-2
Applicability of Types of Leak Sensors
in Tank Excavation Areas
Sensor Type	Sruveillance Method
Observation	Wire
Wells U-tubes Grids
Thermal
Conductivity	X	X
Electrical
Resistivity	X	XX
Gas Detectors	X	X
Sample Collection
and Analysis	X	X
96

-------
zontai pipe there is a tee connection with a vertical
pipe; this vertical section is extended to grade, and ex-
tended 2 feet below the tee to act as a collection sump.
All vertical pipe sections are imperforated, and the bot-
tom of the sump is sealed so as to be leak proof. All
openings to grade are provided watertight caps capable
of being sealed.
The U-tube is a relatively new design that has not
been extensively tested in the field. It appears to offer
an economical method for monitoring and recovery of
leaks and spills at underground installations. When the
U-tube is installed without an underlying impervious
liner it functions on the assumption that a leak will
trickle downward along the exterior surface of the tank
and drip off the very bottom directly into the U-tube.
When installed with an underlying impervious liner, the
U-tube will collect all liquids moving downward
through the soil in the vicinity of the tank including
rainwater. This provides positive assurance of collecting
a leak from a tank but presents a problem with removal
of rainwater which floods out the system.
U-tube monitoring systems, as shown in Figure
2.5-3, are most effective in areas of low groundwater,
where it is unlikely that the tank will be exposed to
groundwater during normal operation. However, U-tube
installations can be used in conjunction with observation
wells in areas where the groundwater table level is
known to fluctuate to a level above the bottom of the
storage excavation.
U-tubes may employ any of the leak detection de-
vices discussed above to provide continuous surveillance
of the storage installation.
C. AREA-WIDE
SURVEILLANCE METHODS
Area-wide surveillance methods include the use of
monitoring wells and the use of dyes or tracers. These
methods are relied upon to investigate or pinpoint the
source of a known leak or spill.
1. Dyes and Tracers
Dyes and tracers may be used as investiative tools
to track down a source of groundwater contamination.
The technique consists of injecting a strong dye or trac-
ing material into a storage tank suspected of being the
source of the contamination and monitoring the point
where the contamination was first discovered for the ap-
pearance of the dye or tracer. A variety of dyes and
tracers are available and include organic and fluorescent
dyes, metallic tracers, ultraviolet tracers, and radioac-
tive tracers. The use of dyes and tracers is governed by
the prohibitions and limitations of the New York State
DEC Ground Water Effluent Standards (Title 6, Official
Compilation of Codes, Rules and Regulations, Part
703).
Rhodamine B is a fluorescent dye generally recom-
ended for time-of-travel and dispersion measurements.
Fluorescein and Rhodamine dyes are also typically used
for groundwater applications. Pontacyl Pink is a good
tracer dye, but is usually more costly than the others.
In addition, detection of this dye requires a fluorometer
[8],
Techniques that utilize radioactive tracers as detec-
tion elements may also be used to pinpoint the source
of an underground spill. These techniques consist of in-
jecting a small amount of a radioactive material such as
tritium into the underground storage tank or pipeline
and using a detection device to track its movement
through the soil or groundwater [13]. However, there
are a number of problems associated with the use of
radioactive tracers including the following:
•A license or approval from the U.S. Department
of Labor or the Nuclear Regulatory Commission
will be required before such materials can be
used.
•There are poitential ecological and health hazards
associated with the use of radioactive materials in
this manner. Whatever materials are injected into
a leaking tank will enter and remain in the envi-
ronment, possibly generating a problem more seri-
ous than contamination of the environment with
the stored product.
Monitoring techniques using dyes and tracers are
often unsuccessful for several reasons:
•If only vapor is found at the discovery point, the
dye or tracer may be useless [6].
•The dye or tracer may be absorbed by the soil or
bleached by chemicals in the soil before it reaches
the point of discovery [6].
•If underground flow is very slow, the site will
have to be monitored for a long time to detect any
leaked dye or tracer [6],
•The dye or tracer may contaminate underground
water supplies [6]
•The dye or tracer may contaminate the product
[6],
2. Monitoring Weils
Monitoring wells are typically employed as areal
surveillance tools; they are used to investigate the
movement of either a liquid or a gas in the ground.
There are two basic types or categories of monitor-
ing tht can be conducted using monitoring wells. These
are:
•Detective monitoring, which establishes the pres-
ence or absence, of contaminants and the need for
further monitoring.
•Interpretative monitoring, which determines the
extent of contamination.
Detective monitoring can be conducted using contamin-
ant sensing devices such as those described earlier in
this chapter. Interpretative monitoring, on the other
hand, typically requires a sample collection and analysis.
97

-------
Figure 2.5-2
Examples of Observation Wells. Each Well Consists of 4"
Perforated PVC Pipe, Driven at Least 2 Feet Below the Bottom
of the Tank and at Least 5 Feet into the Groundwater
Source: Adapted from Reference 15
98

-------
Figure 2.5-3
Example of a U-Tube Installation
Qb*«rv«t
-------
The attempt has been made here to emphasize the
fact that the design and construction of a monitoring
well program can be a complicated undertaking. For
more detailed information see references 10. 11 and 12.
Vapor Wells. Vapor wells or sniff wells may be
used to detect or monitor the presence of hazardous
gases or vapors in unsaturated soil. That is, these types
of wells may be used in permeable soils in the region
above the groundwater table where vapors or gases are
more likely to migrate.
Vapor wells typically employ gas detectors on per-
manent probes or portable gas sampling devices to mon-
itor for gaseous contaminants. These types of devices
Table 2.5-3
Types of Site Data Needed to Design an
Appropriate Groundwater Monitoring Program
Geologic
-	surface geology (topography and type/ depth of
overburden)
-	lithology of aquifer
-	type of geology formation (local stratigraphy
and structure)
Hydrogeologic
-	depth to water table
-	water-table contours
-	thickness of aquifers)
-	relative hydraulic heads, if more than one
aquifer
-	annual precipitation
-	aquifer permeability and porosity
Geochemical:
-	Background water quality
-	chemistry of geologic formation
-	presence of other sources of chemical or biolog-
ical contamination.
Source: Reference 10
will detect a large number of combustible and non-com-
bustible gases and vapors. It should be noted that many
contaminants have odors that can be manually detected,
even at low concentrations. Gasoline, for example, can
be detected by smell at concentrations as low as 0.1
mg/I (of water).
Vapor wells are an advantageous means of detect-
ing volatile soil contaminants before they are dissolved
into the groundwater. However, their primary use is in
indicating the presence of a contaminant. Once contami-
nation has been detected, the vapors will remain present
until cleanup; another monitoring technique (e.g. sample
collection) will be required for further monitoring until
that cleanup occurs.
A typical vapor well installation is shown in Figure
2.5-4.
Groundwater Monitoring Weils. Groundwater
monitoring wells, or wet wells, may be used to detect
or define the movement of a spilled or leaked substance
in a groundwater table.
These types of wells are typically constructed of a
PVC well casing that is screened or perforated in the re-
gion that is being sampled. The material used to fill the
well borehole must be permeable to allow water to flow
into the screened area of the casing. The well may be
monitored using instruments such as the sensors de-
scribed above, or samples may be collected manually
for laboratory analysis.
There are various types of groundwater monitoring
wells that may be used to detect and define groundwater
contamination. These include the following:
•A well screened or open over a single vertical in-
terval.
•A well cluster.
•A single we|l with multiple sample points.
These advantages and disadvantages of these configura-
tions are summarized in Table 2.5-4 and described
below.
Weils Screened Over a Single Interval. An ex-
ample of a well screened over a single verticle interval
is shown in Figure 2.5-5. This type of monitoring well
configuration is routinely used to monitor groundwater
contamination. However, a single well screened in such
a manner is not effective in providing information on
the vertical distribution of a contaminant.
Well Clusters. Investigators have used well clus-
ters to define the vertical distribution of a contaminant
in the groundwater. As shown in Figure 2.5-6, each
cluster consists of a group of closely-spaced wells com-
pleted at different depths. From these wells, water sam-
ples that are representative of different levels in the
groundwater table can be collected. Thus, careful place-
ment of well clusters will allow delineation of both ver-
tical and areal contaminant distribution. However, re-
gardless of the selected depths of the wells in each clus-
ter, there will remain unsampled regions through which
contaminant may pass undetected.
100

-------
Figure 2.5-4
Typical Wells for Continuous Gas or Vapor Monitoring
101

-------
Table 2.5-4


Types of Groundwater Monitoring Wells

Type of Well
Advantages
Disadvantages
Well Screened or Open
Over a Single Vertical
Interval
Can provide composite groundwater
samples if screen covers saturated thick-
ness of groundwater table.
No information is given on the vertical
spread of the contaminant.
Improper completion depth can cause
error in determining the spread of
contamination.
Screening over much of the aquifer thick-
ness can contribute to vertical movement
of contaminant.
The contaminant may become diluted in
the composite sample, resulting in lower-
than-actual concentrations.
Well Clusters
Excellent vertical sampling made pos-
sible if sufficient number of weils are
constructed.
If only a few wells are installed, large
vertical sections of the aquifer are not
sampled. Artificial constraint on data by
completion depths.

"Tried and true" methodology, ac-
cepted and used in most contamination
studies where vertical sampling is
required.
Small diameter wells can be used only for
monitoring. They cannot be used in
abatement schemes.

Low cost if only a few wells per cluster
are involved and if the drilling con-
tractor has equipment suitable' for in-
stallation of small-diameter wells
(1-4 inches in diameter).
In small-diameter wells, development and
sample collection become tedious and
difficult if water is below suction lift.
Single Well-Multiple
Sample Points
(Nested Well)
Excellent information gained on verti-
cal distribution of the contaminant.
If necessary, well diameter is large
enough for use with pumping equip-
equipment.
Sampling depths are limited only by
the size and lift of the pump.
Rapid installation possible.
Relatively expensive.
Proper well construction and sampling
procedures are critical to successful
application.
It is possible to skip large sections of the
groundwater table and thereby miss the
contamination plume.
Source: Reference 10.


102

-------
Single Well - Multiple Sample Points. Another
method used to provide sampling at multiple levels in
the groundwater table is to use multiple sampling points
in the same well. This type of monitoring well is called
a nested well. An example of such a well is shown in
Figure 2.5-7. This technique requires great care in con-
struction of the well and isolation of the various sampl-
ing depths.
Another method has been recently developed by
the Suffolk County, N.Y. Health Dept. It involves the
use of a hollow stem auger to drill a sampling well
which is first pumped and sampled at the deepest de-
sired level. The screen is then withdrawn to the next
sampling level, pumped and sampled again. The process
is repeated until the top of the water table is reached.
The well screen is usually set there. [17]
D. RECOVERY WELLS
Recovery wells may be used to recover oil or any
other hazardous liquid that has been spilled and is float-
ing on the groundwater table. Such wells are located so
as to take advantage of natural gradients or induced
gradients in the groundwater table in drawing out the
contaminated water. Through judicious placement and
operation of recovery equipment, the spill can be con-
centrated in one-of a few recovery sites.
The factors that must be considered when establish-
ing a recovery well program include:
•The required pumping rate to recover product
from the groundwater.
•The establishment of a well network that insures
adequate coverage of the spill.
•The prevention of soil contamination during re-
covery operations.
•Existing environmental and public health standards
which will be used to determine when spill recov-
ery operations have been completed satisfactorily.
•The required depth of the recovery well.
•The geologic formation.
•The required well diameter.
The information required to adequately address
these factors is generally obtained through separate
pumping tests and through consultation with a hydro-
geologist [7]. Typical single pump and two-pump recov-
ery systems are shown in Figures 2.5-8 and 2.59 re-
spectively.
E. EXAMPLES
Figure 2.5-5
Typical Single Monitoring Wet Well
6* Monhol*-in »rav«l*d
urn.
(4* PVC Scrtwtop in
rtmot* artai
4' PVC Scfitdui* 40
Solid Pip*.
CoupUd or Throadtd *
* Thrtadtd pip* stiouid
b* u*«d «h«n Minptinq
of groundwater it
pMnnod.
^Ground wo»«rv^
4' PVC Sehadut* 40
.020* Ser»*n*d Pip#
app*d 3 Pr«*»ur*
Pl»t*d.
Source: Lawrence Peterec, P.E.. New York State
Department of Transportation, Oil Spill
Prevention and Control Division.
A partial list of manufacturers of leak detection
systems and devices is presented in Table 2.5-5.
103

-------
Figure 2.5-6
Typical Wet Weil Cluster
Source: Reference 10.
Table 2.5-5

Leak Detection System Manufacturers
Poilulert System
Pollullert System
Mailory Components Group
P.O. Box 706
Indianapolis, IN. 46206
(317)261-1130
Leak-X Gas and
Liquid Monitoring
Systems
Leak-X Corporation
560 Sylvan Avenue
Englewood Cliff, NJ. 07632
(201) 569-8989 (212) 822-6767
McTighe Hydro-
carbon Detector
McTighe Industries, Inc.
P.O. Box 370
Huntington, N.Y. 11743
(516)549-0050
104

-------
Figure 2.5-7
Schematic of a Typical
Nested Monitoring Well System
Proteewe Casing A Cevef
		OourtaSuttCT
ObMtvaiien
Gas Onv* Tube witn internal Riser TuOe
3" Qoreftme Caainq Removed
Sand Ml*
BeraomteSeai
Source: Reference 15.
105

-------
Figure 2.5-8
Typical Single-Pump
Recovery System
Source: Reference 7.
Figure 2.5-9
Typical Two-Pump
Recovery System
Source: Reference 7.
106

-------
References
1.	Emhart Industries, Inc., Pollulert System, Form
No. P-100, Emhart Industries, Inc., Mallory Com-
ponents Group, P.O. Box 706, Indianapolis, In-
diana 46206, April, 1980.
2.	Emhart Industries, Inc., Mallory's New Pollulert
System, Form No. P-101, Emhart Industries, Inc.,
Mallory Components Group, P.O. Box 706, In-
dianapolis, Indiana 46206, April, 1980.
3.	Leak-X Corp., Leak Detection Cables, Leak-X
Corp.. 560 Sylvan Avenue, Englewood Cliffs,
N.Y. 07632.
4.	Leak-X Corp.. Hazardous Leak Detector, Leak-X
Corp., 560 Sylvan Avenue, Englewood Cliffs,
N.Y. 07632.
5.	McTighe Industries, Inc., McTighe Hydrocarbon
Detector, McTighe Industries, Inc., P.O. Box 370,
Huntington, N.Y. 11743.
6.	National Fire Protection Association, Underground
Leakage of Flammable and Combustible Liquids,
NFPA 329, National Fire Protection Association,
Batterymarch Park, Quincy, MA 02269, 1977.
7.	Pastrovich, T.L., Baradat, Y., Barthel, R.,
Chiarelli, A., Fussell, D.R., Protection of Ground-
water From Oil Pollution, Prepared by CON-
CAWE's Water Pollution Special Task Force No.
11, CONCAWE, Den Haag, Netherlands, April,
1979.
8.	Wilson, Jr., J.F., Techniques of Water Resources
Investigations of the United States Geological Sur-
vey, Chapter A12, Flourometric Procedures for
Dye Tracing, U.S. Department of the Interior,
available from U.S. Government Printing Office,
Washington, D.C.
9.	American Petroleum Institute, Recommended Prac-
tice for Bulk Liquid Stock Control at Retail Outlets,
API Publication 1621, American Petroleum Insti-
tute, 2101 L Street, N.W., Washington D.C.
20037, 1977.
10.	Fenn, D., Cocozza, E., Isbister, J., Braids. O.,
Yare, B., Roux, P., Procedures Manual for
Ground Water Monitoring at Solid Waste Disposal
Facilities, EPA/530/SW-611, U.S. Environmental
Protection Agency, 401 M Street, S.W.,
Washington D.C. 20460, August, 1977.
11.	Edward E. Johnson, Inc., Ground Water and
Wells, A Reference Book for the Water-Well Indus-
try, Edward E. Johnson, Inc., St. Paul, Minnesota
55104, 1966.
12.	Freeze, R.A., Cherry, J.A., Groundwater, Pre-
ntice-Hall, Inc., Englewood Cliffs, N.J. 07632.
1979.
13.	Hasse Tank GmbH & Co., KG, Hasse: The Double
Wall, Self-Monitored Tank, Betco Associates, P.O.
Box 350, Closter, N.J. 07624.
14.	Clemmer Industries, Ltd.,. Double-Walled Storage
Tanks, Clemmer Industries (1964) Ltd, 446 Albert
Street, P.O. Box 130, Waterloo. Ontario N2J4A1.
August, 1981.
15.	Cadwagan, R., Barvenick, M., "Monitoring Device
Simplifies Sample Collection", Water Well Journal,
National Water Well Association, 550 W. Wilson
Bridge Road. Suite 135, Worthington, Ohio 43085.
Vol XXXIV, No. 11, November. 1980.
16.	Fluid Components, Inc., Heat Actuated Liquid
Level Controller, Model 8-66, Bulletin 8-66/1,
Fluid Components, Inc., P.O. Box 1165, Canoga
Park, California 91304.
17.	Pim, James, Suffolk County (N.Y.) Department of
Health Services, personal correspondence regarding
hollow stem auger well.
107

-------
Part II
B. TEMPORARY CLOSURE
CHAPTER 7:
TEMPORARY CLOSURE,
ABANDONMENT AND REMOVAL
OF UNDERGROUND TANKS
A. INTRODUCTION
The use of proper procedures for the temporary or
permanent closure of underground storage tanks is im-
portant for several reasons:
•Product that is left in the bottom of the tank (e.g.
below the withdrawal line) will eventually leak
out, leading to potential environmental contamina-
tion and health hazard problems.
•Empty tanks left in place underground may be
used for illegal storage or disposal of hazardous
wastes.
•Improperly closed tanks may be accidently filled
with a material that is incompatible with the previ-
ously stored material.
•Accidental intrusion into the abandoned tank site
may occur. A classic example of such is children
playing near an abandoned gasoline storage tank,
• where a casually discarded lit cigarette or match
can lead to catastrophe.
•A tank may be reused in a sensitive application,
such as for food product storage, without being
properly cleaned and decontaminated.
•Tanks left empty may eventually collapse.
•An empty, forgotten tank could pose a long term
threat, such as explosion, if site is excavated.
The options available for the temporary or perma-
nent closure of an underground storage system are:
•Temporary closure: the tank and piping system are
emptied and sealed so as to be "temporarily out
of service".
•Abandonment in place: the tank and piping system
are emptied and sealed and the tank is filled with
an inert material.
•Removal for reuse or disposal: the tank and piping
system are removed from the ground after being
emptied.
The concerns associated with these closure options are
presented in the box below. The steps involved in each
of the closure options are summarized in Table 2.7-1
and discussed in the remainder of this chapter.
Underground storage tanks may be considered tem-
porarily closed or "temporarily out of service" if: (1)
they are idle and in sound condition, and will be re-
turned to service; (2) they are awaiting abandonment in
place; or (3) they are awaiting removal from the
ground. These are typically tanks that are intended to be
returned to service within two years or are scheduled for
abandonment or removal within 90 days.
Temporary closure practices include procedures to:
•Remove product from the tank.
•Cap the lines leading into the tank.
•Secure the tank against tampering.
The product removal requirements can be met in
several ways. The best practice in most instances is to
pump out the residual product and fill the tank with
water containing a corrosion inhibitor. This practice
minimizes the possibility of a leak developing while the
tank lies dormant. In addition, such a practice is neces-
sary in instances where ballasting is required to keep the
tank in place due to a high groundwater table. It should
be noted, however, that when the tank is reactivated for
service a problem exists with the proper disposal of a
large volume of contaminated water. This can be trans-
ported away from the location by a licensed hauler only
and must be disposed of in a manner which takes into
consideration applicable regulations governing air and
water pollution abatement.
In situations where water fill is not used and the
stored product was non-flammable, all product should
be removed from the tank. In the case of flammable liq-
uids, a sufficient quantity (approximately 4 inches) of
product should be left in the tank to ensure a saturated
vapor space. This saturated vapor space is needed to re-
duce the possibility of vapor explosions.
Concerns Associated With the
Closure of Underground Storage Systems
•Monitoring the physical integrity of tanks during
temporary or permanent closure procedures.
•Ensuring that product spills do not occur.
•Ensuring that the possibility of explosions of prod-
uct vapors or fires are minimized or eliminated.
•Ensuring that illegal or accidental access to the
tank is not possible.
•Ensuring to the extent possible that projected fu-
ture uses of the site and surrounding environs are
not adversely affected.
Specific information on residual volume amounts to
ensure saturated vapor space should be available from
the liquid manufacturer.
120

-------
Table 2.7-1
Gosure, Abandonment and Removal of Underground Tanks
Reason for Practice and	Relative
Gosure Practice Circumstances of Application Typical Procedures	Costs
Temporary Gosure
("temporarily out
of service")
Abandoment in
Place
Tank Removal
For sound tanks intended to be
returned to service within two
years or scheduled for aban-
donment or removal within 90
days.
Provides for safeguards against
tampering or accidental use
until ultimate fate of tank is
determined.
Permanent closure technique
which avoids cost of tank
removal.
Application dependent upon
tank age and salvage values
and projected use of site after
closure.
For tanks intended for junk-
ing or reuse.
May be required by local re-
gulation or because of the pro-
jected future use of the site.
Remove product from tank. The best	Relatively
approach is to remove all product and	low costs.
fill the tank with water and a corrosion
inhibitor. In lieu of this: (1) remove all
non-flammable product; or (2> remove
flammable product, leaving sufficient
quantity (approx. 4 inches) to assure
saturated vapor space in tank; or (3)
empty tank and fill with a CO2
atmosphere.
Use concrete cast in place to cap all fill
and draw-off lines, and cut off power
to tank pumps.
Leave any vent lines open.
Remove all liquid possible from tank
and piping systems.
Relative
costs range
from low
to high.
Remove or disconnect and plug all fill, ¦
gauge and product lines and cap.
Purge remaining product by filling
tank with water.
Tank may be opened and filled with an
inert solid like sand, or be pumped full
with a grout mixture.
Remove all liquid from tank and piping Relatively
system.	high costs.
Remove all tank connections and tem-
porarily plug all openings.
Purge tank of flammable vapors. The
sequence should be repeated until
vapors are no longer evident.
Remove tank from ground. Safeguard
against tampering (Vapors must be per-
iodically monitored until final disposi-
tion. Vapor may be released from
sludge and scale and again reach explo-
sive level).
121

-------
In instances where the future use of the tank is to be
different from or incompatible with the cement use,
product removal practices also include procedures to
wash down and rinse the tank.
All fill lines, gauge lines and product lines leading
to the tank should be capped during the temporary clo-
sure to prevent casual or accidental use. For example,
a concrete cap can be poured over the fill line; this cap
can later be tapped out with a hammer. In addition, all
power servicing pumps that are conducted to or
fnounted in the tank should be turned off to prevent ca-
sual or accidental use. However, vent lines sould be left
open in the case of flammable liquids to prevent the ac-
cumulation and pressurization (due to high tempera-
tures) of explosive vapors.
In general, the temporarily closed tank should be
secured against tampering. The use of locked caps or
concrete caps on all plugged lines, and the isolation of
the tank area through the use of locked fence are exam-
ples of precautions which should be taken.
C. PERMANENT CLOSURE
The determination of whether to abandon a tank in
place or remove it for reuse or disposal is dependent
upon several factors, such as the age and condition of
the tank, its salvage value, and its potential for reuse.
Governmental regulations may require tank removal.
Other factors that are important include the following:
•Tank Location. The depth to which the tank is
buried and the type of soil in which it is buried
will affect the ease, and hence, the cost, of tank
removal. The potential for damage to concrete or
asphalt traffic surfaces and nearby utilities should
also be considered.
•Projected Use of the Site After Closure. If site
plans call for development that involves excava-
tion or regrading to the level of the tank, it is very
likely that the tank will have to be removed.
•The Cost and Availability of Labor and Equip-
ment. Tank removal will require the use of heavy
equipment and experienced labor. If the cost of
this labor and equipment are prohibitive, abandon-
ment in place may be the preferred option.
•The Proximity of Disposal Site. The proximity of
the disposal site can also greatly affect the cost of
tank removal. Tank transportation costs could be
prohibitive, making abandonment in place the pre-
ferred option.
•Regulatory Requirements. Local laws or ordi-
nances may require removal of the tank as part of
any permanent closure procedures.
1.	Abandonment in Place
Practices for abandonment in place, or on-site clo-
sure of underground tanks must include procedures for
•Removing all product.
•Disconnecting all plumbing and controls.
•Filling the tank with an inert solid such as sand,
gravel or concrete. This is important to prevent
subsidence of the ground above the tank if and
when the tank corrodes or otherwide deteriorates.
•Capping all fill lines, product lines, vent lines,
etc. to prevent future entry into the tank.
More detailed information on on-site closure of un-
derground tanks is available in NFPA 30 [1] and API
Publication 1604 [2].
2.	Removal of Tanks
Practices for removal of tanks must include proce-
dures for:
•Removing all liquid product.
•Disconnecting and capping all plumbing and con-
trols.
•Temporarily plugging all tank openings except
for a '/s inch hole for venting.
•Removing the tank from the ground.
•Freeing the tank of all flammable or toxic vapors.
•Transporting the tank from the site.
If the tank is to be disposed of, a sufficient number
of holes should .be made in it to render it unfit for
further use. The reason for making holes in the tank is
to discourage possible future use of it as a container for
some edible products that would be contaminated by re-
sidual deposits of the material that was previously
stored in the tank. Sources of more information on the
disposal of storage tanks include NFPA 30 [1] and API
Publication 1604 [2].
If the tank is to be reused, care should be taken
to assure that the tank is properly cleaned and that the
future use of the tank is compatible with the past use.
For example, a tank that stored gasoline should not be
used to store a product destined for human or animal
consumption or a product that reacts adversely with
gasoline. References for the proper cleaning and reuse
of underground tanks include NFPA 327 [6] and API
Publication 2015 [4],
122

-------
References
1.	National Fire Protection Association. Flammable
and Combustible Liquids Code, NFPA 30, National
Fire Protection Association, Batterymarch Park,
Quincy, MA 02269, 1981.
2.	American Petroleum Institute, Recommended Prac-
tice for Abandonment or Removal of Used Under-
ground Service Station Tanks, API Bulletin 1604,
American Petroleum Institute, 2101 L Street,
N.W., Washington D.C. 20037, March 1981.
3.	American Petroleum Institute, Dismantling and Dis-
posing of Steel from Tanks which have Contained
Leaded Gasoline, API Bulletin PSD 2202, Ameri-
can Petroleum Institute, 2101 L Street, N.W.,
Washington D.C. 20037, October 1975.
4.	American Petroleum Institute, Cleaning Petroleum
Storage Tanks, API Publication 2015, Second Ed.,
American Petroleum Institute, 2101 L Street,
N.W., Washington D.C. 20037, November, 1976.
5.	New York State Department of Environmental Con-
servation, Bulk Storage of Hazardous Liquids -
Five Proposed Regulatory Concepts, Wastewater
Bureau, New York State Dept. of Environmental
Conservation, 50 Wolf Road, Albany, N.Y. 12233,
November 1980.
6.	National Fire Protection Association, Cleaning
Small Tanks and Containers, NFPA No. 327, Na-
tional Fire Protection Association. Batterymarch
Park, Quincy. MA 02269, 1982.
7.	National Institute for Occupational Safety and
Health, U.S. Department of Health. Education and
Welfare.
8.	National Fire Protection Association, Underground
Leakage of Ftamable and Combustible Liquids,
NFPA 329, National Fire Protection Association,
Batterymarch Park, Quincy, MA 02269. 1977.
123

-------
Appendix B
COMPATIBILITY CHART
FOR FLUIDS,
SEALS, AND METALS
Appendix B:
CHEMICAL-MATERIALS
COMPATIBILITY
The following chart presents compatibility data for
several common materials of construction, including:
steel, stainless steel, and aluminum. Other excellent
sources of information include the following:
1.	Hamner, Norman, E., Corrosion Data Survey,
Fifth Edition
includes: A comprehensive compilation of
corrosion data for metals and non-
metals.
available from: National Association of
Corrosion Engineers
1440 South Creek
Houston, Texas 77084
(713)492-0535
2.	Gallagher, Raymond, "Beat Corrosion With A
Rubber Hose"
published in: ' Chemical Engineering,
Sept. 8, 1980, pages 105-118,
includes:	Comprehensive information on the
compatibility of various hose mater-
ials with a wide range of chemicals,
available from: McGraw-Hill Book Company
1221 Avenue of the Americas
New York. N.Y. 10020
(212)997-1221
or
Mr. Raymond Gallagher
The Gates Rubber Company
999 South Broadway
Denver, Colorado 80217
(303)744-4041
3.	Perry, R.H., Chilton, C.H.,
Chemical Engineers' Handbook
includes:	A series of tables presenting corro-
sion resistance (chemical compati-
bility) data for several ferrous and
non-ferrous metals, plastics, and
natural and synthetic rubbers,
available from: McGraw-Hill Book Company
1221 Avenue of the Americas
New York, N.Y. 10020
(212)997-1221
213

-------
COMPATABILITY CHART FOR FLUIDS, SEALS ANO METALS
Resistance of Metals & Gasket Materials to Various Compounds
CODE: G = Good; F * Fair; P 3 Poor
FLUID
METAL
SEAL
MATERIAL
FLUID
METAL
SEAL
MATERIAL
ALUMINUM
BRONZE
IRON
STEEL
ST. STEEL
BUNA
BUTYL
NEOPRENF
Z
O
_1
u.
IU
H
Z
0
H
>
ALUMINUM
BRONZE 1
IRON
STEEL. I
ST. STEEL
BUNA
BUTYL
NEOPRENE
Z
o
u.
Ui
VITON
Acetaldehyde
G
P
G
G
G
p |g
P
G

Cyclohexane
G | G
G
G
G
G
p
G | G j G
Acetate Solvents • Crude
G
F
F
F
G
p ip
G
G

Cyciohexanol
G
G
G
G
G
G

G j G i G
Acetate Solvents • Pure
G
G
G
G
G
p

G
G

Cyciohexanone
G
G
P
G
G
P
F
P I G S
Acetic Acid • Pure
G
G
P
P
G
F
F
F
G


1




1
|
Acetic Anyhydride
<3
P
F
P
G
| (J
G
U I
Detergent Oils
|
<3
<3
6
<5
! fl
G 1(3
Acetic • Glacial
G
G
P
P
G
G IG :g
g ;p
Oiacetone Alcohol
G 1 G
G
G
G
p
G
F
G

Acetone
G
G
G
G
G
p I G i F
G
P
Dichtorobenzene
G i G I P
P j G
p
P
P
G
G
Acetyl Acetone
G
G
G
G G
P jP
P
G

Dichloro E thane
Gt Gj
G | G
F

1 G |
Acetyl Chloride
F
F
F
G j G


G 1
Oichloro Ethylene
G j G| G
G
G
P
P IP |G.
Acrylonitrile
G
G i G
G G
P F
G |G j
Diesel Oil
Gi Gj G
G
G
G
1 GIG
Alionatic Hydrocarbons
G
G
G
G G
G |p
G IG |G
Diethyl Ether
G 1 Gi G
G
G
F
; G , G i
Aluminum Chloride
F
F
P
P F
G ! G
G IG |
Diethyl Phtnalate
1 <3
G
G
G
P

? i G 1
Aluminum Nitrate
G

P
P
G
<5 CS
G
G

Oiethylene Glycol
Gi G
G
G
G
G

G | G G
Aluminum Sulohate
G
F
P
P
G
G I G
G
G


1




1 !
Ammonium Hydroxide
F
P
G
G
G
G G
G
G

Ethyl Acetate
G | G
G 1 G
G
P
F | F I G |
Ammonium Liciuors
F
P
G
G G


G

Ethyl Alcohol
G! G
G i G
G
G
G :G (GIG
Ammonium Nitrate
G
P
G
G G
G G
G
G

Ethyl Benzene
GI G
F | F
G
P
P | P 1 G |
Ammonia, Anhydrous
G
P
G
G
G
G
G
G
G

Ethyl Senzoate
Gi G
G j G
G
G
P i P | G | G
Ammonia, Aoua
F
P
G
G
6
f*


ti

Ethyl Chloride
GI G
G | G
G
G
G ! G | G j G
Amyi Alconoi
G
G
G
G | G
G
G
G
£
<5
Ethyl Ether
G
G
G 1 G
G
F
F 1 F
G 1
Anth racene
G
G
G
G | G



G

Ethylene Chloride
G
G
P IG
G

1
G i
Aromatic Myorocarbons
G
G
G
G ! G
G
F *
G
G
Ethylene Glycol
G
G
G 1 G
G
G 1 G ; G
G 1 G
Asohalt
G
G
G
G|G
F
P
F
VJ

ethylene Oxide
j P r G
G
G
p ! ip
G i
Av>ation Gasoline
G
G
G
G 1 G
G

G
G
hG

i 1


i 1 '
1











Patty Acids
G| F| P
P
G
G F
F
G i G.
Seer - Seer Wort
G
G
F
I F
G
P
G
G
G
G
Foods
¦Gi !

G


G 1
Benzene • Benzol
G
G
G
G
G
P
F
P
G
F
Formaldehyde
G| G
F
F
G
G G
G
G j G
Benzyl Alconoi
G
G
P
P
G
P
G
G
G
G
Formic Acid
GI F
P
P
G
P
G
P
G 1
Benzyl Chloride
P
P
P
P
G
P
G
P
G

Freon. Dry
G| Gi G | G
G
G
F
P
G 1 G
Brines

G




G

G

Fuel Oil
Gi GI G
G
G
G
P | G
G | G
Sutadiene
G
G
G
G
G
G


G
G

j







SUtane
G
G
G
G
G
G
P
G
G
G
Gas. Natural • Manufactured
Gt G
GIG
G
G
P
P
G
G
Butyl Acetate
G
G
G
G
G
P
F i F
G

Gasolene. Sour
g| f
F If
G
G 1 p i G
G
G
Butyl Aiconoi - Butanoi
G
G
G
G
G
G
G
G
G
G
Gasolene. Motor
gI g
G i G
G
G P
G
G 1 G
Butyl E ther
G
G


G
F

P
G

Gasolene, Aromatic
G
G
G
G
G
F
P
P
G >
Sutylene
G
G
G
G
G
G


G
G
Gasolene, Aviation
G
G
G
G
G
F
P
P
G | G











Glycerine - Clyccrol
G
G
G
G
G
G
G
G
G
G | G
Calcium Hydroxide
F
G
G
G
G
G
G
G
G
G
Grease
G
G
G
G
G
F
G
G | G
Calcium Nitrate
F
G
G
G
G
G

G
G
G










Carbitol Solvent
G
G



G
G
G
G
G
Hegtane
G
G
G
G
G
G

G 1 G 1 G
Caroolic Acid • Phenol
G
G
P
P
G
P
P
P
G

Hexane
G
G
G
r.
G
G
P
G
G i G
Carbon Oisuiohide
G
F
G
G
G
P
P

G
G
Hexanol - Hcxyl Alcohol
GI Gi G
G
G
G
1 G
G 1 G
Carbon Tetrachloride
F
G
F
G
G
F
P
F
G

Hi-Boilinq Naotha

GI

G

1
G 1 G
Carbonic Ann
G
P
P
P 1 G
r
G IG
G

Hi• Flash Napiha

GI

G


GI G
Castor Oil
G
G
F
F | G
G
G
G
G
G
Hydraulic Oil
| G|

G
G
G F
Gi G
China Wood OH • (Tunq Oil)
G
G
F
F
G
G

G
G
G
Hydrochloric Acid. 150°
p i p! p
P
P
G
G 1 G
G ! G
Chloroocetono
P
G

G



G
G

Hydro*!*.*" Peroxide
G
P
P
P
G
G
G i G 1 GJ G
Chioroethane

G


G
P

P
G

Hydroqeri Sulphide, Wet
G
F
G
G
G
F
G
F
Gi
ChJoro'orm
G
G
G
G
G
P
P ! P
G










1
Chiorohen/ene
G
G
G
G
G
P
P |p
G

isoctane
G
G| G
G
G
G
P
G
GI G
Chlorine Dry
F
F
G
G
G
F
F
F
G

!sot>ent«jne

G


G
G


GI G
Chlorine Gas
F
G
G
G
G



G

isooronvi Acetate
G
G
G
G
G
P
G
P
G '
Chi or ontu thane
G
G
F
F
G
P

P
G

Itopropyl Alcohol
G
G
F
G
G
F
G
F
Gi G
Citric AcmJ
G
F
P
P
G
G
F
G
G
G
Isopronyl E ther
G
G
G
G
G
G
P
F
G| G
Corn Oils
G | G
G
G
G
G
G
F
G
G











Cotton Seed Oil
G j G
G
G
G
G
G
G
G
G
Jet Fuel • JP 4 JP-5
G
G
G
G
G
G
P
F
G
G
C'eosoi
G ! G
G
G
G
P
P
G
G












Creosote. Crude
G | F
G
G
G
G
P
F
G
G
Kerosene
G
G
G
G
G
G
P
G
G
G
Cresyiic Acid
G G
G
G
G
G
P
P
G
G











Cj'i'Oni?
S ¦ r
T

: G
G 1
L• i^r fiolv®nt
r3
F
f
F
G
P
P
P
c.

CufT'"9 0«'s
G | G | G | G
G j G | | | G |G
Lactic Acid
G| Fj P ] P | G
18
F | F | G

Emco Wheaton Inc., Loading Arm Assemblies. Catalog E-12/72, EMCO Wheaton Inc., Chamberlain Blvd., Con-
neaut, OH 44030, Revised Sept. 1974.
214

-------
COMPATABILITY CHART FOR FLUIDS, SEALS AND METALS
FLUID
METAL
SEAL
MATERIAL

METAL
SEAL
MATERIAL
ALUMINUM
BRONZE
z
0
E
STEEL
ST. STEEL
<
2
0
CD
BUTYL
OJ
Z
Ui
E
&.
0
uj
2
TEFLON
V/ITON
FLUID
ALUMINUM
»
| BRONZE |
Z
0
c
J
UI
UJ
1-
ST. STEEL
<
z
D
CD
BUTYL
01
Z
w
£
a.
0
Ui
Z
Z
O
-J
tL
OJ
Z
0
H
>
Lard - Lard Oil
G G
G

G
G
F
G
Gi G
Raceseed Oil
G
Gl

G
1
G
G

Lmseed Oil
G ;G
G
G
G
G
G
G
G I G


i







Lul>e Oil
G G
G
G
G
G
P
G
g! o
Sewage
G
ic

G
G
F
F
G
G








1
Skydroi

Gi

G
G


G
G
Methyl Alcohol ¦ Methanol
G G
G
G
G
G
G
G
G { G
Sodium Bicarbonate
G
g!f
F
G
G
G
G
G
G
Methyl Amyi Alcohol
G G


G
F

P
G 1
Soaium Bisuipnite
F
F jp
P
G
G | G
G 1 G ! G
Methyl Amy' Acetate
G .G


G
P

P i Gi
Soaium Carbonate
F
F !g
G
G
G
G
gIgig
Methyl Acetate
G G
G
G
G
P 1
P ! GI
Sodium Chloride
F
F |G
G
G
G
G
GIGiG
Methyl Chloride
P P
P
P
G
F
F
P
GI
Soaium Cyanide
P
P ;G
G | G
G
G
GiGiG
Methyl g ther
G G


G
F

P
G !
Sodium Hydroxide
P
f !g
G ! G
F
G
F
G !
Methyl Ethyl Key tone
G ; G
G
G
G
P
F
P
g j p
Sodium Hypochlorite
P j P |P
P | G
F
F
F
Gi
Methyl isonu'yi Keytone
G G
G
G
G
P
F
p
g| p
Sodium Metaphosohate
P j P |P
P | G
G
G
Fi
G ! G
Methyl Prooyl Keytone
G i G
G
G
G
P

P
G | P
Sodium Nitrate
G l FIG
G ! G
G | G
F
Gj G
Methylene Chloride
P G 1 G
G
G
P

P
Gi
Sodium Perborate
G j F j F ! G ! G
G j G i Gi Gi G
Milk
G P ! P
P
G
G
G ' G 1 G | G
Sodium Peroxide
G 1 F jF | F
G
F j G | F | Gj
Mineral Oils
G G | G
G
G
G
F j_F | Gi G
Sodium Phosonate, Mono-Basic
G | F ! F i F
G
G { G | G { G i G
Moiasses
G G | G
G
G
G
|G | Gi G
Sodium Phosphate. Oi-Basic
F 1 FIF | F
G
G 1 G G GIG

. |




1 1
Sodium Phosphate, Tri-Basic
P ' P iG sG
G
G|G | G I G i G
Naotna
G G ! G
G
G
G
P
P
GI G
Sodium Silicate
P { F |G | G
G
G |G (GIGiG
Naptnaieno
G |G ! G
G
G
P
P
P
Gi
Sodium Sulphate
G i G|G iG
G
G ; G ! GiGiG
Naptna fjoivents
G !G ! G
G
G
G
P i P
Gi G
Sodium Suiohide
P i P i G | G
G
G G
GiGiG
Natural G.is
r,
a < g
n
G
nip i g
G 1 G
Sodmm Thiosufohate
G j F ! F 1 F • G
G G
GjGiG
Nitric Acid • (Cone)
P
p
o
p
G
p
P
P
G
Soya Sean Oil
G ; GIF i F i G
G
G
G GiG
Nitric Acul. Cfu
-------
APPENDIX B
"Corrosion Guide" for Fiberglass Reinforced
Plastic Tanks, from Raven Industries Inc.,
Sioux Falls, South Dakota

-------
CORROSION GUIDE
E3EMH3
The long useful life of Raven tanks in customer service
illustrates the chemtcal resistance of these tanks. Where
new applications are found, the specific chemical resist-
ance requirements snouid be determined. The following
taoie reoresents a comoosite of testing of the resin sup-
Qjiers. our own chemical testing, and field experience.
Since minor variations m chemical mixture or service
conditions can mane major changes in the chemical re-
sistance of a plastic oart. this table is supplied as a
guide for vour selection and testing and does not
imply a guarantee of the cnemical resistance of any
Droduct.
NOTE The resistance of anv material to chemical attack is a function of several elements—the soecific chemical, the
cr.emtcai concentration, the temperature, and the time of contact. When tne temperature, chemical concentration, ar
::me of contact can oe reduced, the service life of the tank will increase. The life of the tank will oe affected when using
aggressive cnemicais that 'ndicate lower service temperatures.
IEGENO: NR: Not Recommended	Maximum Service
NT: Not tested	Temperature in Decrees F.

%
Fiberglass
Poly-






Concen-
Laminate
ethyl-
Fittings
Grommeta ft O-Rings
Material
tration
Std.
Prem.
ene
PVC
Nylon
Nitriie
Vlton
EPOM
Acetic acid
10
170
210
140*
140
NR
140
104
212
Acetic acid
25
150
210
140
140
NR
100
75
212
Acetic acid
50
NR
180
140
140
NR
NR
NR
212
Acenc acid
75
NR
180
70
140
Nfl
Nfl
NR
140
Acetone
100
NR
NR
Nfl
NR
'25
Nfl
Nfl
212
Aluminum chloride
All
170
210
140
140
NR
212
212
212
Aluminum cntorcrivaroxice
50
170
30
NT
NT
NT
NT
NT
NT
Aluminum ootassium sulfate
All
170
210
140
140
.Nfl
212
212
212
Aluminum sulfate
All
170
210
•40
140
NT
212
212
212
Ammonium Oicarocnate
10
Nfl
150
,\T
140
NT
212
212
212
Amrrcnium oicaroonate
50
NR
150
NT
140
NT
212
212
212
Ammonium caroonate
30
Nfl
•oo
140
'40
?c
212
212
212
Ammonium chloride
All
170
210
140
140
NR
212
212
212
Ammonium hydroxide
5
NR
150
140
140
75
212
104
212
Ammonium Hydroxide
10
NR
150
140
140
75
212
104
212
Ammonium hydroxide
20
NR
150
140
140
75
140
104
140
Ammonium hydroxide
29
Nfl
100
140
140
75
104
75
140
Ammonium nitrate
Up to 50%
170
210
140
140
NT
140
212
140
Ammonium oersuifate
Ud to 25 %
140
150
'40
T40
NT
NT
NT
68
Ammonium sulfate
20 •
170
210
140
T40
NT
212
2*12
212
Amy1 aiconoi
All
170
100
140
"2
NT
140
140
212
Anmne sulfate
AH
NR
210
70
NR
• NT
Nfl
75
NT
Antimony trichloride*
All
170
210
140
140
NR
68
68
212
Sanum caroonate
All
170
210
140
140
NT
!40
140
NT
Sanum chloride
All
770
2T0
140
140
Nfl
140
)40
'40
Barium hydroxide
10
NR
150
140
140
NT
140
140
140
Barium sulfide
All
80
180
140
140
NT
140
140
140
Benzaidehyde
100
NT
NR
NR
NR
NT
NR
NR
68
Benzene
100
30
NR
Nfl
NR
75
NR
68
68
Benzene sulfonic actd
0-75
NR
210
NR
NT
NT
NT
NT
68
Benzoic acid
All
'70
210
140
140
Nfl
212
212
NT
Senzvi aiconoi
All
NR
100
NT
NR.
NT
Nfl
NR
68
Bone aod
All
*70
210
'40
140
75
140
140
140
Bromine
Gas or Vapor
NR
Nfl
NR
140
Nfl
NR
68
NR
9uM atcono/
All
\R
100
140
<40
NT
140
140
140
Butvric acid
25
170
120
NT
72
NR
58
68
'40
Butyric acid
50
80
210
Nr
NR
Nfl
68
NR
140
Caicium chlorate
All
170
210
140
140
NT
140
140
NT
Calcium chloride
All
170
210
140
140
NR
212
212
212
Caicium hydroxide
25
170
210
140
140
NT
140
NT
140
Calcium hypochlorite
All
80
210
140
140
NT
68
140
140
Calcium sulfate
All
170
210
140
140
NT
212
212
NT
Caroon 3io*>ce
All
* "n
:io
-40
140
NT
212
212
212
Carcon disulfide
100
NR
30
NR
NT
NT
Nfl
68
NT
Caroon mono*'de
A^l
: TO
210
:40
;40
NT
140
140
NT
Carocn -etrac^ionae
100
5C
30
NR
NR
125
NR
140
NR
Chlorine, wet
100
3C
">
-------
THIS TABLE IS SUPPLIED AS A GUIDE ANO DOES NOT IMPLY A GUARANTEE
LEGEND: NR: Not Recommended	Maximum Service
NT: Not Tested	Temperature in Degrees F.

%
Fiberglass
Poly-





Material
Concen-
Laminate
ethyl-
Finings
Grommets & O-Rings

tration
Std.
Prem.
ene
PVC
Nylon
Nitrile
Viton
EPOM
Cific ac«2
Ail
170
270
\AQ
140
NR
740
140
210
Copper chloride
All
170
210
140
440
NR
140
140
140
Coooer cvanide
Ail
170
210
140
140
NT
140
140
NT
Capper sulfate
All
170
210
140
^40 •
Nfl
212
212
212
Crude on. sweet ana sour
100
170
210
NR
140
75
212
212
Nfl
DicntoroDenjene
TOO
Nfl
Nfl
NR
NR
NT
NR
68
NT
E'sct-csoi
5
\T
150
70
NT
NT
NT
NT
* NT
Sin-.; aiccnoi
Ail
30
100
140
140
'25
58
68
68
S:rv err,a'
100
\R
NR
Nfl
NP
75
58
NR
NT
5fnvf?n»s c~'Ori'"2e
•00
N R
NR
NR
NP
NT
68
Nfl
^40
E t!-\ier.e -^ciicnae
100
NR
NR
NP
Nfl
NT
NT
NT
140
cnionoe
All
'70
210
140
140
NR
212
212
212
Ferric nitrate
All
170
210
140
140
NT
NT
NT
NT
Ferric sulfate
All
170
210
140
140
NR
NT
NT
212
Ferrous chloride
All
170
210
140
140
NR
NT
NT
212
Ferrous nitrate
All
170
210
140
140
NT
NT
NT
NT
Ferrous sulfate
All
170
210
140
140
Nfl
NT
NT
212
Fluoboric acid
All
30
210
140
140
NT
NT
NT
NT
Fiuosmcic acia
25
Nfl
100
140
140
NR
68
NT
140
Forme acid
All
NR
'00
• 140
140
Nfl
NP
104
140
Gasonne
100
170
100
NR
*40
75
• 140
140
NR
Givcer>n iG'vceroJ*
TOO
• ''u
270
140
140
75
140
140
140
Heorane
too
'70
210
NR
140
NT
140
IV
NR
^varooromtc acia
IE
30
180
'40
140
NT
Nfl
140
212
Hydrobromic acid
50
30
150
140
140
NT
NR
140
140
Hvdrochioric (muriatic) acid
10
170
210
140
140
NR
NR
140
212
Hvcrochioric acid
20
NT
210
140
140
NR
NR
140
212
Hvdrochionc acid
37
NR
210
140
140
Nfl
NR
68
140
Hvdrocvanic acid
10
80
150
140
140
NT
NT
NT
NT
Hvarofiuoric acid
10
NR
150*
70
140
NR
NR
140
NR
-vcrotiuonc acid
20
Nfl
100*
70
72
NR
NR
104
NR
Hvcrogen peroxide
30
MR
150
140
140
Nfl
NR
68
140
^'vcccnicrous acia
10
30
210
.140
140
NT
NT
NT
NT
coc^iorous aod
20
3Q
150
140
14Q
NT
NT
NT
NT
'-vcccmorous add
50
NR
150
140
140
NT
NT
NT
NT
<*fz$ er.e
100
* ?c
100
NR
140
75
140
140
NR
Lactic acid
All
170
210
140
140
75
212
212
68
Lead acerare
All
170
210
140
140
NT
140
212
NT
Linseeooii
100
170
210
NR
140
NT
140
212
140
Magnesium caroonare
All
170
150
140
140
NT
140
140
NT
Magnesium cnionae
All
170
210
140
140
75
212
212
212
Magnesium sulfate
All
170
210
140
140
75
212
212
212
acc
All
•-C
2*0
70
140
.Nfl
212
212
NT
Mercjr-c :r"cr:ae
ai;
'70
210
140
140
75
* 4C
140
140
c-vcce
At!
' 70
210
!40
140
75
14C
140
NT
.VCCCi
All
NP
••x
140
140
NR
68
68
68
xetore
!G0
NR
NR
NR
NR
125
NP
NP
58

*00
vp
Ns
NR
NR
NT
NR
58
Nfl
Napnna
TOO
170
210
NR
140
NT
NT
NT
NR
Naontnaiene
100
170
210
NR
NR
NT
68
68
NR
Nickel cnlonde
All
170
210
140
140
NT
140
140
140
Nickel nitrate
' All
170
210
140
140
NT
140
140
NT
Nickel sulfate
All
170
210
140
140
NT
140
140
140
Nwic aod
5
150
'50
140
140
NR
NR
540
140
3c-c
:c

' ~0
•40
"40
NP
Nfl
•40
*4v
'.::'"ce^:ene
'00
NR
\a
NP
NP
NT
NR
NP
NR
¦: -r-c aca
Ail
¦-TC
VP
NR
•4.:
N*
140
•40
140
" 5u;f'jfC 3C:3'
-•
N P
NP
NP

NP
NR
68
NR

A;;

2'C
:±z
'40
Na
•40
2'2
'40
3cic
'0
\p
150
U0
•40
NP
NP
•40
NT
Oorcmonc acid
30
NR
'00
70
72
NR
Nfl
68
NT
R^ospronc acid
10
¦70
210
140
140
Nfl
140
140
140
®*csonoric acid
25
17C
210
140
140
NR
140
140
140
Phospnonc acid
50
170
210
140
140
NR
140
14C
140
P^csphcnc acid
95
17C
2^0
140
140
Nfl
68
'40
68
'.e-1 -ea-rea

-------
RAVEN CORROSION GUIOE - CON'T.
LEGENO: NR: Not Recommended




Maximum Service


NT: Not Tested



Temperature in Degrees F.



%
Fiberglass
Poly-






Concern
Laminate
ethyl-
Fittings
Grommeta £r O-Rlngs
Material
trati on
Std.
Prem.
ene
PVC
Nylon
Nltrile
Vlton
EPDM
Photographic solutions
All
170
210
140
140
NT
80
104
NT
Phthaiic acid
All
170
210
NT
NT
NT
140
140
NT
Picric (alcoholic) acid
10
80
210
NR
140
NT
68
68
68
Potassium bicarbonate
10
150
180
140
140
75
140
140
140
Potassium carbonate
10
NR
150
140
140
75
140
140
140
Potassium carbonate
25
NR
100
140
140
.75
140
140
140
Potassium carbonate
50
NR
80
140
140
75
140
140
140
Potassium chloride
All
170
210
140
:40
75
212
212
212
Potassium aichromate
All
170
210
140
140
NT
NR
68
140
Potassium ferricvanide
All
170
210
140
140
75
140
212
NT
Potassium hydroxide
10
NR
150
140
140
75
68
NT
212
Potassium nydroxiae
25
NR
150
140
140
NR
68
NT
'40
Potassium nitrate
All
170
210
140
140
NT
140
140
NT
Potassium permanganate
All
80
210
140
72
NR
NR
104
NT
Potassium persuifate
All
80
210
NT
140
NT
NR
212
NT
Potassium sulfate
All
170
210
140
140
75
140
140
140
SeJemous acid
All
NT
210
70
72
NT
NT
NT
NT
Silver nitrate
All
170
210
140
140
NT
176
176
212
Sodium acetate
All
170
210
140
140
NR
NT
NT
NT
Sodium bicarbonate
10
150
180
140
140
75
140
140
140
Sodium bisuifate
All
170
210
140
140
75
NR
140
140
Sodium carbonate
10
NR
150
140
140
NR
140
140
140
Sodium carbonate
25
NR
150
140
140
75
140
140
140
Sodium carbonate
32
NR
150
140
140
75
140
140
140
Sodium chlorate
50
NR
210
140
140
NT
NR
212
140
Sodium chloride
AH
170
210
140
140
75
212
212
212
Sodium cyanide
All
170
210
140
140
NT
140
NT
140
Sodium ferricyanide
All
170
210
140 '
140
NT
NT
NT
NT
Sodium hydroxide
5
NR
210*
140
140
75
140
104
212
Sodium hydroxide
10
NR
180*
140
140
75
140
104
212
Sodium hvaroxide
25
NR
210'
140
140
NT
140
'104
212
Sodium hydroxide
50
NR
210*
140
140
NT
NR
NR
140
Sodium hypochlorite
5'A
NR
150"
140
140
NT
NR
68
140
Sodium hypochlorite
10
NR
ISO"
140
140
NT
NR
68
68
' 68
Sodium hypochlorite
15
NR
180"
140
140
NT
NR
68
Sodium nitrate
All
170
210
140
140
75
140
140
140
Sodium nitrite
All
170
210
NT
140
NT
NR
140
NT
''" Sodium silicate
All
NR
210 •
NT
NT
NT
140
140
140
Sodium sulfate
All
170
210
140
140
75
140
140
140
Sodium sulfide
. All
80
210
140
140
75
140
212
140
Sodium sulfite
All
80
210
140
140
NT
!40
NT
140
Stannic chloride
All
170
210
140
140
NR
212
NT
140
Stannous chionae
All
170
210
140
140
NR
NT
NT
NT
Stearic acid
All
170
210
140
140
75
140
140
68
Sulfonated detergents
100
30
150
NR
NT
75
NT
NT
NT
Sulfuric acid
25
160
210
140
140
NR
104
140
104
Sulfuric acid
50
30
210
140
140
NR
NR
68
NR
Sulfuric acid
70
NR
170
70
140
NR
NR
68
NR
Tannic acid
All
170
210
140
140
NT
68
140
68
Tartaric acid
All
170
210
140
140
NT
68
NT
140
T etrachioroethyiene
100
NR
80
NT
NR
NT
NR
NR
NR
Trichloroacetic acid
50
80
210
NT
NT
NT
140
NR
NT
Tnsodium phosphate
All
NR
210
140
140
75
NT
NT
NT
Toluene
100
80
80
NR
NR
75
68
NR
NR
Urea-ammonium nitrate fertilizer mixture
100
'00
100
•0
140
75
140
140
NT
Water tOistiiied)
All
170
21C
140
*40
212
212
212
212
Water tOemmeraiized)
All
140
210
•40
140
212
212
212
212
Water lOeionizeoi
All
'40
210
'4Q
140
212
212
212
212
Xylene
100
30
50
NR
NR
NT
68
NR
NR
Zinc Chloride
All
170
210
140
140
NR
•40
212
212
Zinc sulfate
All
170
210
140
140
NT
212
212
212
8-8-8 Fertilizer
100
120
100
^0
140
75
104
104
104
THIS TABLE IS SUPPLIED AS A GUIOE AND OOES NOT IMPLY A GUARANTEE.
* Svntnetic veil required.
**Oue to vanaoie service life, factory shouia oe contacted for recommenaauons.
*2CWi-478	~ 37 -
B-3

-------
APPENDIX C
Summary of Design Standards
for Underground Storage Tanks
Prepared by SCS Engineers

-------
UL 58
Title: "Steel Underground Tanks for Flammable and Combustible
Liquid s"""
Scope: UL 58 requirements cover horizontal cylindrical, atmos-
pheric type, welded steel tanks, intended for installa-
tion and use in accordance with NFPA No. 30, the Flam-
mable and Combustible Liquids Code, and NFPA No. 31,
Standard for the Installation of Oil-Burning Equipment.
These tanks are fabricated, inspected, and tested for
leaks before shipment from the factory. Capacities,
dimensions, and metal thicknesses are specified in tables
in the Standard. The stell shall be new, commercial
quality, uncoated or galvanized, and of good welding
q u a 1 i t y .
NFPA 30-1981
Title: "Flammable and Combustible Liquids Code"
Scope: The code applies to all flammable and combustible liquids
except those that are solid at 100 F or above. The code
contains separate chapters on tank storage; piping,
valves, and fittings; container and portable tank stor-
age; industrial plants; bulk plants; se'rvice stations;
processing plants; and refineries, chemical plants, and
distilleries. Rules for storage tanks concern aspects of
tank materials, linings, fabrication, design standards,
installation, site requirements, venting, and control of
spillage.
Note that NFPA 30 i s not a design standard for tanks, but
prescribes which design standards may be used, and pro-
vides certain requirements or recommendations on the as-
pects noted above.
API Publication 1615
Title: Installation of Underground Petroleum Storage System
Scope: This bulletin covers the installation of underground gas-
oline, diesel fuel and waste oil systems and is primarily
applicable at retail and commercial facilities. Emphasis
is on the correct selection of the tank material and
size, location of the tank and ancillary piping and
equipment, correct installation procedures, and testing
both during and after installation to detect leaks.
The material in their bulletin is applicable to hazardous
waste storage systems.
C-l

-------
ASME Section VIII Division I
Title: "Rules for Construction of Pressure Vessels"
Scope: The rules in this Division of Section VIII cover minimum
construction requirements for the design, fabrication,
inspection and certification of pressure vessels other
than those covered in other Sections and other excep-
tions. Subsection A covers the general requirements ap-
plicable to all pressure vessels. Subsection B covers
the specific requirements that are applicable to the var-
ious methods of fabrication: welding, riveting, forging,
and brazing. Subsection C covers specific requirements
applicable to several classes of materials: carbon and
low-alloy steels, non-ferrous metals, high-alloy steels,
non-ferrous metals, cast iron, clad and lined materials,
cast nodular iron, and ferritic steels. The rules have
been formulated on the basis of design principles and
construction practices applicable to vessels designed for
pressures up to 3000 psi. For pressures above 3000 psi,
deviations from and additions to these rules are neces-
sary to meet the requirements of design principles and
construction practices for these higher pressures. The
design temperature shall not be less than the mean metal
temperature (through the thickness) expected under oper-
ating -conditions; in no case shall the surface tempera-
ture ex'ceed the maximum temperature listed in the stress
tables for materials nor exceed temperature limitations
specified elsewhere in Division I or Section VIII.
Corrosion is covered by ASME Section VIII, Division I,
Appendix E, "Suggested Good Practice Regarding Corrosion Allow-
ance", and by ASME Appendix F, "Suggested Good Practice Regarding
Linings". In the former, paragraph IJA-156 states that "when the
rate of corrosion is already predictable, additional wall thick-
ness...shall be provided, which shall be at least equal to the
expected corrosion lpss during the desired life of the vessel".
Paragraph UA-157 states that when the corrosion effects are inde-
terminate prior to design of the vessel, or when corrosion is in-
cidental, localized, and/or variable in rate and extent, the de-
signer must exercise his best judgment in establishing a reason-
able maximum excess shell thickness. Paragraph UA-159 suggests
that, when a vessel goes into corrosive service without previous
service experience, service inspections be made at frequent in-
tervals until the nature and rate of corrosion in service can be
definitely established.
C-2

-------
ASTM D4021-81
Title: "Standard Specification for G1ass-Fiber-Reinforced Poly-
ester Underground Petroleum Storage Tanks"""
Scope: Specification covers fiberglass-reinforced horizontal,
cylindrical, and spherical-type underground tanks for
atmospheric pressure storage of petroleum-based fuels and
oils. The specification covers the materials, the manu-
facture, workmanship, external load requirements, inter-
nal pressure, fitting-moment load and torque load rat-
ings, leakage, internal impact resistance, chemical
resistance, quality control, and test methods.
UL 1316
Title: "Proposed First Edition of the Standard for Glass -
Fiber-Reinforced Plastic Underground Storage Tanks"
(January, 1982 draft!"
Scope: These requirements cover spherical or horizontal cyli n-
drical, atmospheric-type tanks of fiberglass-reinforced
plastic (FRP) that are intended for the underground stor-
age of petroleum-based flammable and combustible liquids.
These tanks are completely assembled and tested for leak-
. age before shipment, and intended for installation and
use in accordance with the Standard Installation of Oi 1-
Burning Equipment, NFPA No. 31, and the Flammable and
Combustible Liquids Code, NFPA No. 30. The standard
allows for the incorporation of manholes; therefore, 40
CFR 264.191(1), which states "The regulations of this
Subpart (Subpart J - Tanks) do not apply to facilities
that treat or store hazardous wastes in covered under-
ground tanks that cannot be entered for inspection", will
remove from consideration those tanks fabricated without
manholes.
AC I 318-77
Title: ACI Standard, Building Code Requirements for Reinforced
Concrete
Scope: This is the basic standard for the proper design and con -
struction of buildings of reinforced concrete. It covers
(1) standards for tests and materials, (2) concrete qual-
ity, (3) mixing and placing concrete, (4) formwork, em-
bedded pipes, and construction joints, (5) details of re-
inforcement, (6) analysis and design, and (7) structural
systems. The code provides minimum requirements for de-
sign and construction of reinforced concrete structural
elements of any structure erected under requirements of
general building codes, but does not specifically cover
C-3

-------
tanks. The code states that for special structures, in-
eluding tanks, provisions of this code shall govern where
applicable.
AC I 350 R-77 (formerly ACI 74-26)
Title: Concrete Sanitary Engineering Structures
Scope: As indicated by "R" in the ACI designation, this is for a
committee report, and as such contains recommendations
for structural design, materials, and construction of
concrete tanks, reservoirs, and other structures commonly
used in water and waste treatment works where dense, im-
permeable concrete with high resistance to chemical
attack is required. Special emphasis is placed on de-
signs which minimize cracking and accomodate vibrating
equipment and other special loads. Chapter 5 - Protec-
tion Against Chemicals - states that concrete made with
the proper type of cement, which has been properly pro-
portioned, mixed, placed, and cured, will be dense,
strong, watertight, and resistant to most chemical deter-
ioration or corrosion. However, in industrial waste
treatment plants, where the pH of acid waste may go as
low as 1.0, the types of protection generally used are
chemical-resistant mortar, acid-proof brick or tile,
thick bituminous coatings, epoxies, and heavy sheets or
liners of rubber or "plastic.
C-4

-------
APPENDIX D
Memoranda from SCS Engineers to Bill Kline
Regarding Tank Shell Thickness

-------
STEARNS. CONRAO ANO SCHMIDT
CONSULTING ENGINEERS. INC.
11260 ROGER BACON ORIVE
RESTON, VIRGINIA 22090-5282
(703)471-6150
SCS ENGINEERS
ROBERT P. STEARNS, PE
E. T. CONRAD. PE
DAVIO H. BAUER
RODERICK A. CARR
LOUIS L. GUY, JR., PE
MILES J. HAVEN
MICHAEL W. MCLAUGHLIN
GARY L. MITCHELL, PE
DAVIO E. ROSS, PE
WILLIAM L. SCHUBERT
JAMES J. WALSH. PE
JOHN P. WOOOYARD, PE
July 22, 1983
File No. 28001-03
MEMO
TO: Mr. William Kline, Environmental Protection Agency
FROM: SCS Engineers
SUBJECT: Tank Shell Thickness as Regards Permit Issuance
A. Shell Thickness Measurement
1. Nondestructive Techniques
The simplest method of making a thickness determination is
to use calipers. Obviously this technique is limited to.areas
of the shell that are within reach of an "opening so that the
calipers can be inserted through the opening and measurement
taken from both inside and outside the shell.(1,2)
b. Ultrasonic Inspection
Ultrasonic instruments can be used to measure tank shell
thickness as well as to determine the location, size and
nature of defects. They can be used while the tank is in
operation as only the outside of the tank needs to be
contacted. They can be used on steel, FRP and concrete
tanks. Two types of ultrasonic instruments, the resonance
and the pulse type, are commonly used for tank thickness
measurement. The pulse type instrument utilizes electric
pulses and transforms them into pulses of ultrasonic waves.
The waves travel through the metal until they reach a
reflecting surface. The waves then are reflected back,
converted into electric pulses, and show up on a time-
baseline of an oscilloscope. The instrument is calibrated
by using a material of known thickness; therefore, the
time interval between the pulses corresponds to a certain
thickness. There are two types of resonance ultrasonic
instruments. In one of these, an electric oscillator
transmits electric energy of constant ultrasonic frequency
to a crystal (transducer) which, in turn, converts this
energy into mechanical pressure waves that travel through
the material being measured in the direction of its thick-
ness. The pressure waves travel at a constant velocity
OFFICES IN RESTON. VIRGINIA: LONG BEACH. CALIFORNIA; BELLEVUE. WASHINGTON; COVINGTON. KENTUCKY; ANO COLUMBIA, SOUTH CAROLINA
a. Calipers
D-l

-------
Mr. William Kline
July 22, 1983
Page 2
and are reflected at the opposite surface back to the crystal.
Because velocity through a given material is constant, the
time required for a wave to circumnavigate is a function of
the distance traversed, in this case, equal to twice the
thickness. Therefore, by measuring the time interval,
the thickness can be determined. In the other type of
resonance device, a crystal (piezoelectric transducer) is
applied to the surface of the wall to be measured, and an
electronic circuit causes it to vibrate over a range of
frequencies. When the vibrating frequency of the crystal
matches the natural frequency of the vibration of the
material being measured, a signal is fed through the circuits
of the instrument and interpreted electronically as an
indicated thickness. This indication is fed through an
oscilloscope and emerges as a series of vertical lines across
the face of the tube. These lines indicate thickness on
a transparent plastic scale mounted directly on the face
(front) of the oscilloscope tube. Ultrasonic instruments
can provide digital readouts, can provide a permanent record
of measurements, and are accurate to within one percent of
the thickness of the tank shell being measured. Ultrasonic
instruments- are the most applicable for shell thickness
measurement. (1,2,3)
A brief synopsis of the cost of ultrasonic inspection is
appropriate at this point. Should a consideration be to buy
ultrasonic inspection equipment the price ranges from $1,500
to $2,600 and provides an accuracy of plus or minus 0.005
inches or better, depending on the price of the equipment.
This type of equipment provides a digital readout and will
give thickness readings of a specific point on the tank sur-
face. Numerous models are available in the marketplace from
different manufacturers. The most exotic ultrasonic inspec-
tion instrument costs in the vicinity of $10,000 and provides
a continuous readout accurate to plus or minus 0.001 inches
with a digital readout. Readings can be made using a sweep-
ing motion of the probe as opposed to a point reading pro-
vided by the equipment discussed above, and therefore will
provide thickness readings along a line or of an area. For
an additional $5,000 approximate cost a strip chart recorder
can be purchased which will record all of the readings taken
in a continuous manner by this continuous readout probe.
Should the option chosen be to hire a professional testing
firm that uses ultrasonic inspection equipment, the prices
are somewhat more reasonable. The current standard hourly
rate for ultrasonic inspection is $25.00 per hour plus mile-
age with a $100 minimum per job.
D-2

-------
Mr. William Kline
July 22, 1983
Page 3
This rate will provide the user with point thickness
measurement and normally the methods involved will cover
a two to four square foot per hour area which allows you
to calculate how much it would cost to do an inspection
of any given size tank by calculating the square area
involved. Should the customer request the strip chart
recorder be utilized as well as the digital readout that
comes with all equipment, the hourly rate rises to $35
per hour and the job minimum rises to $140. The customer
then receives a chart of area thicknesses and can locate
areas of local corrosion and pitting with these. Coverage
using the strip chart recorder is somewhat slower than the
two to four square foot per hour coverage noted above.
Should the customer require a formal report signed by a
professional engineer, the rate rises to $60 per hour.
It becomes obvious that unless the customer intends to make
a multitude of thickness measurements over a short period
of time, the purchase of ultrasonic inspection equipment
is financially disadvantageous. It appears to be much more
appropriate to hire a professional testing firm for each
inspection requirement and pay the hourly rate. (4,5,6,7,8)
c. Radiographic Inspection
Radiography may be used to determine shell thickness, as well
as to detect flaws, such as cracks and voids, and can be
used on steel, fiberglass reinforced plastic, and concrete.
However, access to both sides of the tank shell is required.
The ray source must be on one side and the film on the other.
The rays commonly used in tank shell thickness measurement
are the X-ray and the gamma ray. The X-ray is produced in
a CRT tube within an X-ray machine; the gamma ray is produced
from a radioactive material (source) contained in a small
capsule. The two rays are similar. Each has unique advan-
tages in penetrating power and ease of mobility. Recently,
a gamma ray system has been devised that is portable and,
therefore, much easier to use than the older type fixed
X-ray and gamma ray producing equipments. The rays pass
through the tank shell and are photographed on film. Then
the film is compared with film taken of the same material of
known thickness to determine the thickness of the tank shell
in question. Radiography is better suited to weld inspection
and flaw detection than to thickness measurement. (1,2,3)
Briefly the cost data pertaining to radiographic inspection
equipment is that to buy such equipment would cost approxi-
mately $2,700 at a minimum. The equipment is expensive to use
D-3

-------
Mr. William Kline
July 22, 1983
Page 4
due to the expensive film and the processing of that film
and because the gamma ray source will only last for 90 days
and then must be replaced. In addition, further expense
is incurred because radiographic technicians must be hired
to operate the equipment. Should the customer desire to
hire a professional testing firm that uses radiographic
inspection equipment, the cost currently in the marketplace
is $40 per picture and the process is quite slow and cumber-
some. Apparently ultrasonic inspection techniques are far
superior in both cost and accuracy to radiographic tech-
niques .( 4 , 5 ,8)
2. Destructive Techniques
The one destructive technique that is utilized in measurement
of tank shell thickness is known as the hook gauge. The
technique involves drilling a hole through the empty tank
and inserting the hook gauge through the hole to measure
thickness in the shell at that location.' The hole is then
repaired by tapping threads into the hole and inserting a
plug in accordance with the code under which the tank was
built.(1,2)
B. Minimum Shell Thickness Calculations
Shell thickness calculations are quite different, depending on
the type of material. The discussion below is subdivided by
type of material: steel, fiberglass reinforced plastic (FRP),
and concrete.
1. Carbon Steel and Stainless Steel
Any discussion of steel tank shell thickness must involve
two specific points. The first portion of the tank shell
thickness is determined by calculating the stress that
will be exerted on the steel shell and is a structural
thickness. The remainder of the tank shell thickness is
an allowance to offset the effects of corrosion. The
summation of these two thicknesses, the structural thickness
and the corrosion allowance, as it is known, provides one
with the total shell thickness minimums.(1,3,9)
a. Structural Shell Thickness
All of the codes and standards that deal with steel tank
design and construction discuss structural shell thickness.

-------
Mr. William Kline
July 22, 1983
Page 5
The standards published by the American Petroleum Institute,
specifically API 620 and API 650, discuss in some detail the
calculation of structural shell thickness and provide
formulas for that calculation. (See Attachments A and B)
Additionally, two other standards published by the American
Petroleum Institute, API 12B and API 12D, provide minimum
structural thickness for not only the shell but for the
roof and deck plates in tabular form. These two standards
do not provide formulas for thickness calculation. The
standards published by Underwriters Laboratories, specifically
UL 58 for underground tanks and UL 142 for above ground tanks,
provide the structural thickness in tabular form, but do not
provide formulas for its calculation. In all cases, minimum
shell thicknesses are provided, regardless of what the
structural shell thickness formula dictates. In most cases
the minimum shell thickness for steel tanks is 3/16 in. with
the exception of UL 58 and UL 142, which provide for shell
thicknesses of less than 3/16 in. for smaller tanks.(10,11,
12,13,14,15)
b. Corrosion Allowance
The codes deal with corrosion allowance in a variety of ways,
but, due to an inability to set forth formulas for deter-
mination of corrosion allowance, the codes normally leave
the corrosion allowance thickness up to the purchaser. For
example, API 650 states: "The purchaser shall specify,
when necessary, the corrosion allowance to be provided for
each shell course, the bottom, the roof, and the structurals,
giving consideration to the total effect of liquid stored, the
vapor above the liquid, and the atmospheric environment."
Corrosion is effected by many variables. It is apparent that
those responsible for writing codes for design of steel tanks
are unable to set forth definitive guidance concering corro-
sion allowance. The structural thickness is determined to
the nearest 0.001 in., and then a corrosion allowance to
be determined by the purchaser is added to that structural
thickness to obtain the overall desired minimum thickness
of the tank shell.
Corrosion of steel is affected by a multitude of variables,
as mentioned above. Some of these variables are: the
compatibility of the liquid being stored with the tank
material, the temperature of the liquid being stored, the
pressure inside the tank, the amount of movement of the fluid
D-5

-------
Mr. William Kline
July 22, 1983
Page 6
within the tank, bacterial action that may occur within or
outside the tank, soil resistivity in reference to tanks
that are either underground or partially below the surface,
moisture level either in the soil or in the air in reference
to exterior corrosion, variations in the soil which set up
an electrical current and can cause electrolytic corrosion,
and finally environmental elements such as atmospheric
pollutants. Corrosion and corrosion allowance will be
discussed further in the last major portion of this
memorandum.(1,9,10,11,16,19)
2. Fiberglass Reinforced Plastic (FRP)
FRP is a corrosion resistent, laminated material used in
tank construction. The term corrosion allowance does not
apply when discussing FRP because the failure of FRP due to
the corrosivity of either the fluids being contained or of
the atmosphere and/or soil outside the. tank does not normally
occur by material being corroded away from the shell thick-
ness. Failure occurs when FRP loses its rigidity due to
reaction with a chemical or chemicals either within or outside
the tank. So there is no discussion of corrosion allowance
when addressing FRP, and thickness determination is based on
structural integrity. The specification which guides the
design and construction of FRP tanks is the American Society
of Testing and Materials (ASTM) Standard D3299-81 "Filament-
Wound Glass-Fiber Reinforced Polyester Chemical-Resistent
Tanks". (See Attachment C) FRP tanks constructed in
accordance with this specification are built in layers, as
set forth below: (18,19)
a.	Inner Surface
The inner surface, or surface exposed to the liquid inside
the tank, is a reinforced layer 10 to 20 mils minimum thick-
ness. The reinforcement materials are chemically resistent
glass surface mat, and are present to give form to the
layer rather than add to the overall structural integrity
of the tank.
b.	Interior Layer
The interior layer	is reinforced with noncontinuous glass
strands applied in	a minimum of two plies of chopped strand
mat or alternately	in a minimum of two passes by the spray-
up process. Glass	content is specified as 20 to 30 weight
D-6

-------
Mr. William Kline
July 22, 1983
Page 7
percent. Before filament winding is applied, the interior
layer is allowed to gel completely so the corrosion barrier
will not be squeezed down to a thin layer of glass content
over 30 percent. The combined thickness of the inner
surface and the interior layer is not less than 100 mils.
c.	Exterior Layer - Filament Wound
Subsequent reinforcement is continuous strand roving in
accordance with minimum thickness requirements as set forth
in ASTM D3299-81. The thickness of the filament wound
portion of the tank shell may be varied with tank height
(tapered wall construction), providing minimum thickness
requirements are met at any height level. If additional
longitudinal strength is required, the use of other rein-
forcement such as woven fabric, chopped strand mat, or .
chopped strands, may be interspersed in a winding to provide
additional strength. Glass content of filament winding will
be 50 to 80.weight percent. The minimum thickness of the
summation of the inner surface, the interior layer, and the
exterior layer - filament wound is 180 mils.
d.	Exterior Layer - Contact Molded
The exterior layer or body of the laminate is of chemically
resistent construction suitable for the service intended
and provides additional strength as necessary to meet the
tensile and flexural requirements. Where separate layers
such as mat, cloth, or woven roving are used, all layers
are lapped a minimum of 1 in. Laps are staggered as much
as possible. If woven roving or cloth is used, a layer of
chopped strand glass is placed as alternate layers. The
exterior surface is relatively smooth with no exposed
fibers or sharp projections.
e.	Outer Surface
For added resistence to chemical exposure (spillage), an
exterior surface of chopped glass or surfacing mat, or both,
made from either glass or organic fibers may be employed,
as agreed upon between manufacturer and purchaser. This
layer is used only if contact between the stored liquid and
external surface of the tank is considered likely.
f.	General
As can be seen from the above descriptions of the layers of
an FRP tank, the minimum thickness requirements are quite
D-7

-------
Mr. William Kline
July 22, 1983
Page 8
specific. The main subject of concern with FRP tanks is
the compatibility of the tank material with the liquid to
be stored, as well as with the soil and atmosphere.
Corrosion allowance is not the issue; resistance to loss
of structural strength of FRP caused by chemical reaction
is the issue.
3. Concrete
Tanks constructed of concrete utilize codes just as do the
materials discussed above. The two codes most commonly used
are American Concrete Institute (Ac I) Specification 67-40
entitled Desi qn and Constructi on of Ci rcular Prestressed
Concrete Structures (See Attachment D), and ACI 74-26
entitled Concrete Sani tary Engineering Structures (Copy
provided when recei ved). These specifications provide
formulas for determining the minimum shell thickness required
in concrete tanks and also provide minimum dimensional
limits for the tank shell thickness, as set forth below.
If the tank is cons-tructed of a shot crete - steel diaphragm
type construction, the minimum shell thickness is 3 1/2 in.
If the construction is cast-in-place concrete without vertical
prestress, the minimum shell thickness is 8 in. For cast-
in-place concrete tanks with vertical prestress, . the
minimum shell thickness is 6 in. The minimums sited above
are from ACI 67-40, for prestressed tanks. Similar minimum
thicknesses are provided in ACI 74-26. Minimums are provided
for the overhead dome and the floor of the tank as well. As
with FRP tanks, there is no added thickness provided to
offset corrosion. The problem is dealt with not by adding
thickness to the tank shell, but by providing additives to
the concrete or lining or coating the concrete to prevent
chemical reaction with the stored liquid, or to prevent
reaction with the atmosphere or with the surrounding soil.
Therefore, the thicknesses provided from the formulas in the
ACI specifications are for structural integrity (or ease of
construction), and do not have a corrosion allowance as did
the thickness of steel discussed above. (20,21,22)
C. Visual Inspection versus Actual Shell Thickness Measurement
1. General
The question of when visual inspection is sufficient and when
actual shell thickness measurement utilizing one of the
techniques discussed above is required, is a difficult one.
D-8

-------
Mr. William Kline
July 22, 1983
Page 9
The variables that influence the answer to this question
are numerous. Those variables that have the greatest
impact are discussed below.
2. Vari ables
a.	Leak Impact(l)
The impact that a leak will have on the surrounding
environment must be considered when evaluating visual versus
actual shell thickness measurement. Several relevant ques-
tions are:
•	Will a liquid leak cause a serious threat to the
health of people who come in contact with it?
t Will a leak contaminate water supplies, crops and
food supplies, fisheries or wild life habitat?
•	Will a leak cause dislocation of people?
•• Will a leak cause loss of property resulting from
contamination, fire, explosions, etc?
•	What are the economic and social costs of leak
clean up?
b.	Corrosion(1,9,16,19,22,23) (See Attachments E & F)
The mechanism of corrosion is the primary, concern when dis-
cussing shell thickness in regard t.o both steel and concrete
tanks. Corrosion can take many forms. A common form of
corrosion with steel tanks is electrolytic corrosion. This
form of corrosion is the result of a direct current from
outside sources entering and then leaving a particular metal
structure by way of the electrolyte (surrounding material,
such as soil for underground structures or water for submerged
structures). A similar type of corrosion known as galvanic
corrosion is a self generated activity resulting from dif-
ferences in electrical potential that develop when metal is
placed in an electrolyte. These differences in electrical
potential can result from the direct coupling of dissimilar
metals, or they can result from variations and conditions
which exist upon the surface of a single metal. Electrolytic
and galvanic corrosion are similar in that corrosion occurs
D-9

-------
Mr. William Kline
July 22, 1983
Page 10
at the anodes. The primary difference between the two is
that in electrolytic corrosion the external current generates
the corrosion, whereas in galvanic corrosion the corrosion
activity itself generates the current.
Many factors influence corrosion rates in metals. The
acidity of the electrolyte (solution, soil) with which the
material is in contact can have a substantial effect on
the rate of corrosion. The presence of oxidizing agents,
of which oxygen is the most prominent, may accelerate the
corrosion of one type of material and retard corrosion in
another. The rate of corrosion tends to increase with
rising temperature. Once corrosion has started, its progress
is often controlled by the nature of the film that forms
on the corroding metal. Some corrosion products may be
insoluble and completely protective; or they may be very
permeable and thus allow localized or general corrosion to
proceed unhindered. The metabolic activity of certain
microorganisms can either directly or indirectly affect the
corrosion of metals. The soil resistivity is the largest
single factor controlling the rate of corrosion caused by
either the soil in which the tank is buried or the soil on
which the tank is sited. The lower the resistivity of the
soil, the greater the probability of corrosion. The presence
of water can also promote corrosion of metals. The presence
of moisture in soil acts to reduce soil resistivity, thereby
increasing the probability of corrosion. Water accumulating
inside tanks is also a major cause of internal corrosion.
Corrosion of underground tanks and pipes can be influenced
by variations in soil conditions along the surfaces of those
tanks and pipes. Variations in soil type, soil resistivity,
and moisture content can promote galvanic activity in the
buried metal, thus accelerating the rate of corrosion.
Corrosion can also be influenced by the presence of atmospheri<
pollutants, both externally and internally. For example,
sulfur dioxide can form sulfuric acid in the presence of
air and moisture and can thus promote corrosion of certain
metals.
Concrete tanks also suffer from corrosion; sulfate attack
causes concrete to break down. Sulfate reacts with hydrates,
the resultant compounds expand and rupture the concrete. The
severity of the attack depends on the concentration of the
solution. If the concrete is exposed on one side only,
rather than on both sides, the rate of corrosion will increase
If the concrete is alternately saturated and then allowed to
dry, the corrosion rate caused by sulfate attack will increase
d -1 n

-------
Mr. William Kline
July 22, 1983
Page 11
Concrete is attacked by sea water. This form of corrosion
is more severe on reinforced concrete. The absorption of
salt from the sea water sets up anodic and cathodic areas
within the concrete, causing electrolytic action to take
place. Corrosion products then accumulate on the reinforcing
steel, and this accumulation causes the concrete surrounding
the steel to rupture. Portland cement is highly susceptible
to acid attack. Acid dissolves cement, leaving aggregate
exposed.
Surface treatments have been successful in the prevention or
retardation of the corrosion of concrete. Coal tar pitch,
rubber or bituminous paints, epoxy resins, and magnesium
silico fluoride are some of the surface treatments that have
been used successfully in preventing concrete corrosion or
repairing tanks that have exhibited some corrosion.
c. Compatibi1ity(1,9,16,18,19 ,22)
A major concern in any tank is the compatibility of the
liquid being stored with the material of construction. There
are many questions that arise when discussing compatabi1ity.
t What is the vapor pressure of the liquid?
•	What are the melting and boiling points of the
liquid?
•	Is the liquid flammable, corrosive, toxic or
reacti ve?
•	What will be the allowable pressure inside the
tank?
The main question, of course, is whether the stored liquid
will attack the material of the tank. This compatabi1ity
question is present for not only steel and concrete tanks
but for FRP tanks as well. With steel and concrete, the
liquid normally will corrode or erode the tank wall. With
an FRP tank, a stored liquid that is not compatable with
the components of the laminated shell will cause a loss of
structural integrity which will sometimes be accompanied by
an actual swelling of the shell thickness rather than a
thinning of the shell thickness as takes place during a
corrosive reaction.
D-ll

-------
Mr. Wi11iam K1 ine
July 22, 1983
Page 12
d.	Shell protection(l,2,9,10)
One of the variables that controls how often a tank is
inspected.and whether visual inspection or sophisticated
shell thickness measurement is required involves the
question of whether the tank shell is protected in any way.
In steel tanks, cathodic protection is used to prevent or
retard electrolytic corrosion. In addition, the tank can
be electrically isolated from its surroundings. Tanks
constructed of many different materials beneft from the
protection afforded by various paints, coatings and linings.
This protection may be used internally and/or externally.
e.	Siting(l)
In deciding between visual inspection and tank shell
thickness measurement, the siting of the tank or tanks
must be considered. Underground tanks cannot be as
thoroughly inspected visually as can above ground tanks.
Thus, corrosion and other defects cannot be located as
easily.- This may warrant more frequent, actual shell,
thickness measurement.
f. Historical Data(l)
For a used tank, quite often historical data will provide
an indication of how susceptible the tank is to corrosion
or erosion from either the liquid being stored or the
atmosphere and/or soil outside the tank. If the corrosion
rate indicated by the data shows that the corrosion allowance
is still in tact, actual thickness measurement may not be
required at present. For a new tank, historical data from
like tanks storing similar or the same liquids can be used
to help make the determination.
3. Conclusi on
The question of when visual inspection is sufficient and
when operators or owners should be required to make actual
shell thickness measurements is affected by many variables,
as can be seen from the above discussion. The successful
operation of any tank farm is in large measure dependent
upon a schedule of inspections followed by both preventive
and corrective maintenance predicated on what has been found
during those inspections. Where a visual inspection is
possible, this will always be the first step in shell thick-
ness determination. The purpose of the visual inspection
D-12

-------
Mr. William Kline
July 22, 1983
Page 13
is to seek out any signs of attack on the tank shell
from either the inside due to the liquid being stored
or from the outside by either the atmosphere or the soil.
It should be pointed out that although we have been talking
about shell thickness as the critical measurement, the
measurement of the bottom of the tank for its thickness is
just as critical. The visual inspection should be very
thorough and should be looking for localized corrosion,
pitting, areas where rust is prevalent, blistering, dis-
coloration, stress cracks and any other indications that
a change is taking place in the material of the tank shell.
Previous visual inspections which indicated possible defects
should be used as a basis for increasing the frequency of
inspection. The permit writer must consider the following
factors in deciding whether shell thickness measurement
by instrument 'is required:
•	The impact of the leak on the	surrounding
environment
•	The presence of'factors which	lead to corrosion
or give increased probability	of corrosion
•	The liquid being stored is not compatable with
the tank shell material and no lining or coating
is provided
•	The shell has not been protected using
protection or cathodic or anodic inhibitors or
electrical isolation
•	The tank has not been designed in accordance with
any accepted code or standard
•	The cost of performing actual shell thickness
measurement using ultrasonic techniques.
If there is doubt in the permit writers mind concerning
the sufficiency of visual inspection, actual shell
thickness measurement using ultrasonic techniques should
be considered.
Several scenarios are presented here as examples of appro-
priate actions to be taken by the permit writer. In the
first case, a permit application is received by the permit
writer for three ten thousand gallon above ground FRP tanks.
The tanks have been built in accordance with ASTM specifi-
cations D-3299-81, and the liquid to be stored appears to
D-13

-------
Mr. William Kline
July 22, 1983
Page 14
fit within the guidelines provided by that specification.
A substantial leak of the liquid from the tank would be
very likely to infiltrate the groundwater table and
probably the surface water nearby if allowed to run freely.
An acceptable system of dikes and curbs as secondary
containment have been provided. An inspection schedule
has not been presented by the applicant, but the permit
writer and the applicant have worked out a satisfactory
inspection and maintenance schedule. In this particular
case, the permit writer has no basis for requiring actual
shell thickness measurement as long as the inspection
schedule sets forth conditions under which actual shell
thickness measurement would be required in the future.
The second permit application received by the permit writer
involves a twenty thousand gallon, underground, carbon steel,
storage tank. The applicant has presented somewhat sketchy
historical data which indicates that the tank has been
inspected over its 12 year life, but that no such inspection
has been conducted during the last 5 years. The tank does,
have a man way permitting internal inspection and is lined
with epoxy but there is no information concerning the age
or condition of the epoxy liner. The tank was originally
designed in accordance with the fifth edition of Underwriters
Laboratory Standard for Safety No. 58. In this particular
case, the permit writer does not have sufficient information
concerning the thickness of the tank shell or the condition
and thickness of the epoxy liner. He should require actual
tank thickness measurement using ultrasonic techniques prior
to the issuance of a permit.
The two examples presented above give some indication of the
variables encountered by the permit writer. The decision
to require actual shell thickness measurement is one that
must be made based on the best information available. The
two examples are fairly clear cut, but the permit writer
will not always be able to make his judgement as easily.
Each application must be evaluated on its own merits.
J. D. Wright, P.E.
SCS ENGINEERS
1 a

-------
REFERENCES
1.	New York State Department of Environmental Conservation,
Technology for the Storage of Hazardous Liquids - A State
of the Art Review, New York State Department of Environmental
Conservation, 50 Wolf Road, Albany, NY 12233, January, 1983.
2.	American Petroleum Institute, Guide for Inspection of Refinery
Equipment, Chapter IV - Inspection Tools, American Petroleum
Institute, 2101 L. Street, NW, Washington, DC 20037, July,
1983.
3.	American Society for Metals Handbook, 8th Edition, Volume 11,
"Nondestructive Inspection and Quality Control".
4.	Communication with Richard McGuire, Manager, Technical Services,
American Society for Non-Destructive Testing4 July, 1983.
5.	Communication with Bob Williams, Kraut Kramer Branson, Lewistown,
PA (manufacturer of ultrasonic inspection equipment), July, 1983.
6.	Communication with Joe Monroe, Eastern NDT, Inc., Hopewell, VA
(distributor of Sperry Products Division ultrasonic inspection
equipment), July, 1983.
7.	Communication with Jim Bruner, Law Engineering, McLean, VA.
(tank testing firm), July, 1983.
8.	Communication with Clive Lugmayer, Lugmayer Associates,
Clinton, MD (tank testing firm), July, 1983.
9.	American Society for Metals Handbook, 8th Edition, Volume 1,
"Properties and Selection"
10.	American Petroleum Institute, Recommended Rules for Design	and
Construction of Large, Welded, Low-Pressure Storage Tanks,	API
Standard 620, American Petroleum Institute, 2101 L Street,	NW,
Washington, DC 20037, September, 1982.
11.	American Petroleum Institute, Welded Steel Tanks for Oil Storage,
API Standard 650, American Petroleum Institute, 2101 L Street,
NW, Washington, DC 20037 , November, 1980.
12.	American Petroleum Institute, Specification for Bolted Tanks for
Storage of Production Liquids, API Specification 12B, American
Petroleum Institute, 2101 L Street, NW, Washing ton, DC 20037,
January, 1977.
13.	American Petroleum Institute, Specification for Field Welded
Tanks for Storage of Production Liquids, API Specification 12D,
American Petroleum Institute, 2101 L Street, NW, Washington, DC
20037, January, 1982.
D-15

-------
REFERENCES
Page 2
14.	Underwriters Laboratories, Inc. Standard for Steel Underground
Tanks for Flammable and Combustible Liquids, UL 58, Underwriters
Laboratories, Inc., 333 Pfingsten Road, Northbrood, Illinois
60062, April, 1981.
15.	Underwriters Laboratories, Inc. Standard for Steel Aboveground
Tanks for Flammable and Combustible Liquids, UL 142, Underwriters
Laboratories, .Inc. , 333 Pfingsten Road, Northbrook, Illinois
60062, December, 1982.»
16.	American Petroleum Institute, Guide for Inspection of Refinery
Equipment, Chapter II - Conditions Causing Deterioration or
Failures, American Petroleum Institute, 2101 L Street, NW,
Washington, DC 20037, 1973.
17.	American Petroleum Institute, Guide for Inspection of Refinery
Equipment, Chapter XIII - Atmospheric and Low Pressure Storage
Tanks, American Petroleum Institute, 2101 L Street, NW, Wash-
ington, DC 20037, April, 1981.
18.	American Society for Testing and Materials, Filament-Wound
Glass-Fiber Reinforced Polyester Chemical-Resistant Tanks,
ASTM Standard Specification D-3299-81, American Society for
Testing and Materials, 1916 Race 'Street, Philadelphia, PA
19103, 1981.
19.	Communication with Norman Hamner, National Association of
Corrosion Engineers, July, 1983.
20.	American Concrete Institute, Design and Construction of
Circular Prestressed Concrete Structures, ACI Title Number 67-40,
American Concrete Institute, Box 4754, Redford Station, Detroit,
MI 48219, September, 1970.
21.	American Concrete Institute, Concrete Sanitary Engineering
Structures, ACT Title Number 74-26, American Concrete Institute,
Box 4754, Redford Station, Detroit, MI 48219
22.	A.M. Neville "Properties of Concrete" published by Wiley, 1975.
23.	National Association of Corrosion Engineers, Basic Corrosion
Course, Chapter 8 - "Localized Corrosion" by H.P. Godard, 1969.
n-16

-------
ATTACHMENTS
A.	Excerpt from American Petroleum Institute Standard 620 -
Recommended Rules for Design and Construction of Large, Welded,
Low-Pressure Storage Tanks-concerning shell thickness design.
B.	Excerpt from American Petroleum Institute Standard 650 -
Welded Steel Tanks for Oil Storage - concerning shell thickness
desi gn.
C.	American Society for Testing and Materials Standard Specification
D-3299-81 - Filament-Wound Glass-Fiber Reinforced Polyester
Chemical-Resistant Tanks
D.	American Concrete Institute Title Number 67-40 Design and
Construction of Circular Prestressed Concrete Structures
E.	National Association of Corrosion Engineers, Basic Corrosion
Course, Chapter 8 - Localized Corrosion by H.P. Godard
F.	Excerpt from American Society for Metals Handbook. 8th Edition,
Vol. 1, "Properties and Selection" - concerning carbon steel
and stainless steel corrosion.
D- 17

-------
SCS ENGINEERS
STEARNS. CONRAO ANO SCHMIDT
CONSULTING ENGINEERS. INC.
11260 ROGER BACON DRIVE
RE5T0N. VIRGINIA 22090-5282
(703)471-6150
File No. 28001.03
November 25, 1983
ROBERT P STEARNS. P€
E. T. CONRAO. PE
OAVIO H. BAUER
RODERICK A. CARR
LOUIS L. GUY. JR.. PE
MILES J. HAVEN
MICHAEL W. MCLAUGHLIN
GARY L. MITCHELL. PE
OAVIO E. ROSS. PE
WILLIAM L. SCHUBERT
JAMES J. WALSH. PE
JOHN P. WOOOYARO. PE
MEMORANDUM
TO: Bill Kline
FROM: John Wright
SUBJECT: Addendum to SCS Memorandum dated July 22, 1983 Entitled
"Tank Shell Thickness as Regards Permit Issuance"
A. Shell Thickness Measurement
1. . Nondestructive Techniques
a. Hammering Technique
Hammering, or the physical inspection of a tank using a hammer,
deserves some discussion because it is the technique normally used to
do a routine inspection of a tank prior to using a more sophisticated
technique. The inspection of a tank should always involve a visual
inspection prior to proceeding to one of the other techniques. A
normal follow-on to a visual inspection would be to inspect the tank
using a hammer. The hammer will not tell the inspector what the
thickness of the tank shell is at any location, but will indicate if
the thickness has changed indicating a defect. Other subtle
differences that a trained inspector will look and listen for that may
indicate a defect are vibration, denting, and movement. The key to the
usefulness of this technique is that the operator of the hammer must be
skilled in the art and know what to listen for and what he is feeling.
The only way that he can become skilled is through actual experience.
The accuracy of sounding with the hammer is dependent upon the
operator's ability to distinguish minute differences in sound together
withdifferences in the rebound of the hammer. He must then have the
ability to translate these differences in sound and rebound into
changes in thickness of the shell and/or changes in the structural
integrity of the shell. Once this technique has located a change in
thickness or a difference in structural integrity, another more
accurate technique such as ultrasonics or radiography should be used to
make a determination of thickness and. strength.
D-18
OFFICES IN RESTON. VIRGINIA: LONG BEACH. CALIFORNIA; BELLEVUE. WASHINGTON; COVINGTON. KENTUCKY; ANO COLUMBIA. SOUTH CAROLINA

-------
Care should be exercised when using this technique on tanks that are in
service, so that a failure or rupture is not caused by the hammer
itself. Certain equipment should not be subjected to the hammering
technique to include enameled, ceramic or glass-lined tanks; equipment
in caustic service because stress corrosion cracks may result; brittle
materials such as cast iron, high alloy steels, brass or bronze; and
other locations where hammering might result in damage to the tank and
its appurtenances.
B. Failure of Fiberglass Reinforced Plastic (FRP) Tanks
The failure of the shell of an FRP tank occurs quite differently than the
failure of a steel tank shell. To protect a steel tank from the effects of
corrosion, the tank is lined with a corrosion resistant liner (inside
protection) or cover (outside protection) or the thickness of the shell is
increased with what is called a corrosion allowance and corrosion is
allowed to proceed at a known rate. In the case of FRP tanks, the primary
concerned is still the compatibility of the liquid being stored with the
interior surface of the tank and the compatibility of the outer surface of
the tank with the material around the tank. However, the effects of
incompatibility do not appear as a loss of thickness of the shell as it
does with a steel tank. ^RP tends to lose its structural integrity and may
even swell. This swelling is caused by a reaction between the components
of the tank shell and either the liquid in the tank or the soil and/or
groundwater outside the tank. Another possible indication that a problem
exists with an FRP tank shell is a change in the color of the FRP. This
may also be an indication that a reaction is taking place.
The methods used to combat the incompatibility problem with FRP tanks are
relatively simple. The resins chosen for use in a given tank must be
compatible with both the liquid to be stored and the material on the
outside of the tank (backfill and/ or groundwater). Added thickness of the
tank shell is not particularly effective from a compatibility standpoint
although it may provide some additional structural strength to the tank.
Resins are available thatare resistant to almost any chemical or
combination thereof.
D-19

-------
C. Failure of Conrete tanks
The minimum shell thickness of the concrete walls of a tank are determined
by the structural requirements of the tank, not by any corrosion
considerations. The phenomenon of corrosion is addressed by using
additives to the concrete mix or by using linings or coatings to protect
the concrete from the effects of corrosion. As with FRP, the effects of
corrosion on concrete normally make themselves know by a softening of the
concrete, spalling of the surface of the concrete or by a change in the
color. All of these are indicative that a reaction of some sort has taken
place between the concrete and either the chemical being stored or the
backfill or atmosphere. Sulfate attack is one of the more common types of
problems found with concrete shells. One of several possible sulfates
reacts with a hydration product and the result is an increase in volume.
If the volume increase takes place before the concrete sets, it is not
critical. But if the concrete has already set, the volume increase can
cause explosive forces that can cause the concrete to self-destruct. This
form of attack may be countered by using a high-quality cement paste made
with a sulfate-resistant cement. There is also evidence that suggests that
the substitution of from 15 to 30 percent of an active pozzolanic material
for the cement will help.
i
The effects of seawater on concrete tanks was adequately discussed in the
original memorandum, but the problems resulting from the alternate freezing
and thawing of concrete were not. Normal concrete contains 1 to 2 percent
air entrapped in the mix. If these air voids become filled with water and
the water is then frozen, the water expands when it freezes and forms ice
resulting in hydraulic pressures which can disrupt the concrete and cause
spalling or breaking apart of the surface. Freeze-thaw problems can be
minimized by increasing the percentage of air entrapped in the concrete mix
by using an air-entraining add mixture, thus producing a concrete with from
4 to 8 percent air entrapped in the mix. Freeze-thaw resistance of
concrete is also enhanced by using a high-quality portland cement that
limits the amount of water in proportion to cement used in the concrete
mix. Finally, durable aggregates are absolutely necessary.
D-20

-------
APPENDIX E
Chapter 8/ "Inspections", from "Permit Writers
Guidance Manual for Hazardous Waste Tanks", Draft
Prepared by Battelle-Columbus Division for Region II,
U.S. Environmental Protection Agency, July 1983

-------
APPENDIX E
Chapter 8, "Inspections"/ from "Permit Writers
Guidance Manual for Hazardous Waste Tanks", Draft
Prepared by Battelle-Columbus Division for Region II;
U.S. Environmental Protection Agency, July 1983

-------
DRAFT
on
PERMIT WRITERS' GUIDANCE MANUAL
FOR
HAZARDOUS WASTE TANKS
to
U.S. ENVIRONMENTAL PROTECTION AGENCY
REGION II
from
BATTELLE
Columbus Division
505 King Avenue
Columbus, Ohio 43201
EPA Contract 68-01-6515
WA 02-004

-------
ACKNOWLEDGEMENTS
This document was prepared at Battelle's Columbus Division with
contributions being made by G.C. Stotler (Project Manager), John Hallowell
(Work Assignment Manager), and J.L. Otis, T.S. Reddy, and Dr. D.N. Gideon
(Principal Investigators). Battel!e was acting as a subcontractor to
A.T. Kearney, Inc., with Mr. Joseph Fantasia serving as Regional Liaison
for that organization.
The project was supported and directed by staff of the U.S. EPA
at the Region II office. Early definition of the content of the document
was guided by comments from permit reviewers at many of the EPA Region
offices with special inputs from Heather Ford, Garrett Smith, and D. Fagan;
important contributions to the direction of the project were made by Ernie
Regna and especially the EPA Region II Project Officer, Barbara Kropf.
Comments on the draft report are encouraged. They should be
directed to:
Ms. Barbara Kropf
U.S. EPA Region II
Waste Management Division
26 Federal Plaza
New York City, New York 10278
Telephone: (212) 264-0504.

-------
PREFACE
This Working Draft of the "Permit Writers' Guidance Manual for
Hazardous Waste Tanks" was developed to provide guidance to the permit
reviewer in evaluating the design of hazardous waste tanks, piping, controls,
and ancillaries (containment measures, vents, etc.). The manual is aimed
at presenting supporting material where possible within the manual or,
alternatively, offering summaries or indications of content and coverage
of the extensive technical material contained in existing standards, codes,
handbooks, etc.
Because of the wide spectrum of tank characteristics and applications
anticipated to be encountered by the permit reviewer, the content of this
document contains a diversity of subject matter. Some of the content was
selected for coverage on the basis of comments from Regional Offices contacted
as the first step in the preparation of this manual. The format and arrangement
of the draft manual have been selected with the aim of allowing for addition
»
of supplementary material by the individual user and possibly the incorporation
of uniform changes and additions at some future time.
Similarly, each section contains a detailed listing of content
intended as an aid in assessing material as different requirements arise.

-------
TABLE OF CONTENTS
Pa^e
1.0 USE AND ORGANIZATION OF THE MANUAL		1-2
1.1	How to Use the Manual		1-2
1.2	Manual Organization 		1-3
1.3	References, Codes, and Handbooks 		1-4
REFERENCES		1-6
2.0 GENERAL PROCESS CONSIDERATIONS 		2-2
2.1	Process Flow Diageam (PFD) 		2-2
2.2	Piping and Instrumentation Diagram (P&ID) 		2-4
2.3	Classification of Hazardous Wastes 		2-5
2.4	Installation of Outside Above-Ground Tanks 		2-7
3.0 STANDARD METAL TANK DESIGN 			3-3
3.1	General		3-3
3.2	New Metal Tanks		3-10
4.0 USED, NON-METAL, AND OTHER TANKS . 			4-2
4.1	Used Steel Tanks		4-2
4.2	Fiberglass-Reinforced Plastic Tanks 		4-8
4.3	Rectangular (Polygonal) and Other Non-Standard
Metal Tanks		4-13
4.4	Concrete Tanks		4-14
5.0 TANK ANCILLARIES: PRESSURE AND OTHER CONTROL SYSTEMS ...	5-2
5.1	Internal Pressure and Pressure Controls 		5-2
5.2	Other Controls and Instruments 		5-10
6.0 TANK ANCILLARIES: PUMPS, PIPING AND AGITATOR
STUFFING BOXES 		6-2
6.1	Pumps		6-2
6.2	Piping		6-4
6.3	Agitator Stuffing Boxes 		6-10
7.0 TANK ANCILLARIES: FOUNDATIONS, SUPPORTS, INSULATION,
AND GROUNDING		7-2
7.1	Foundations 			7-2
7.2	Structural Tank Supports			7-5
7.3	Thermal Insulation 		7-6
7.4	Electrical Grounding 		7-6

-------
TABLE OF CONTENTS
(Continued)
Page
8.0 INSPECTIONS	8-2
8.1	Evaluation of Inspection Plan	8-4
8.2	Weekly Above-Ground External Tank Inspection 		8-6
8.3	Detailed Assessment of Tank Condition 		8-6
8.4	Inspection of Auxiliary Equipment 		8-7
8.5	Inspection Tools and Procedures 		8-16
8.6	Frequency of Tank Inspection 		8-18
9.0 SPILLS, LEAKS, AND SECONDARY CONTAINMENT		 9-2
10.0 REACTIONS IN TANKS		10-2
APPENDIX A
TABLE A-l. REGULATORY ANALYSIS OF TANKS	A-2
APPENDIX B
CORROSION AND DETERIORATION OF
MATERIALS OF CONSTRUCTION		B-2
LIST OF TABLES
Table 1-1. References for Use with the Permit Guidance
Manual for Hazardous Waste Tanks 	 1-5
Table 2-1. Buffer Zones for Tanks of Stable Liquids 	 2-8
Table 2-2. Buffer Zones for Tanks Containing Stable
Liquids		2-9
Table 2-3. Buffer Zones for Tanks Containing Boil-Over
Liquids	2-9
Table 2-4. Buffer Zones for Tanks Containing Unstable
Liquids	2-10
Table 2-5. Buffer Zones for Tanks Containing Class III B
Liquids		2-11
Table 2-6. Reference Table for Use With Tables
2-1 to 2-4	2-11

-------
TABLE OF CONTENTS
(Continued)
LIST OF TABLES
(Continued)
Page
Table 3-1. Impact of Selected Properties of the
Waste on Tank Design	3-7
Table 3-2. Summary Checklist for Tank Design 	 3-11
Table 3-3. Materials Groups	3-21
Table 3-4. Maximum Allowable Stress Values for Simple
Tension	3-26
Table 3-5. Maximum Allowable Efficiencies for Arc-Welded
Joints 	3-29
Table 3-6. Permissible Plate Materials and Allowable
Stresses	. . . "	3-33
Table 6-1. Pump Classes and Types	6-3
Table 8-1. Checklist of Tank External - Inspection Points . . . 8-8
Table 8-2. Checklist for Tank Internal Inspection (Tank
Out of Service) 	8-10
Table 8-3. Checklist for Inspection of Piping, Valves, and
Fittings	8-13
Table 8-4. Checklist for Visual Inspection of
Pumps and Compressors	8-14
Table 8-5. Checklist for Inspection of Instruments
and Control Systems 	8-15
Table 8-6. Mandated Inspection Frequencies 	 8-19
Table 9-1. Operational Problems of Tanks 	 9-3
Table 10-1. Oxidation Waste Treatment Applications 	 10-4
Table 10-2. Information for Estimating Corrosion
Against Performance 	 	 10-7
Table 10-3. Reduction Waste Treatment Applications 	 10-8

-------
LIST OF FIGURES
Page
Figure 2-2. A Simplified Piping and Instrumentation
Diagram	2-6
Figure 3-1. Minimum Permissible Design Metal Temperature
for Plates Used in Tank Shells Without Impact
Testing (in degree Fahrenheit) 	 3-20
Figure 3-2. Biaxial Stress Chart for Combined Tension and
Compression, 30,000 to 38,000 psi Yield Strength
Steels	3-39
Figure 3-3. Typical Tank Data Sheet	3-46
Figure 3-4. A Simple Tank Diagram 	 3-47
Figure 5-1. Simplified Process Flow Diagram, Common Vapor
Recovery System	5-9

-------
8-1
CHAPTER CONTENTS
INSPECTIONS
8.1	Evaluation of Inspection Plan
8.2	Weekly Aboveground External Tank Inspection
8.3	Detailed Assessment of Tank Condition, (as scheduled)
8.3.1	External Inspection
8.3.2	Internal Inspection
8.4	Inspection of Auxiliary Equipment
8.4.1	Pipes, Valves, and Fittings
8.4.2	Pumps and Compressors
8.4.3	Instruments, Control Equipment, and Electrical
- Systems
8.5	Inspection Tools and Procedures
8.5.1	Hammering Method
8.5.2	Penetrant-Dye Method
8.5.3	Magnetic-Particle Method
8.5.4	Radiographic Method
8.5.5	Ultrasonic Method
8.5.6	Vacuum-Box Method
8.6	Frequency of Tank Inspection
8.6.1	Regulatory Requirements
8.6.2	Practical Considerations
8.6.2.1	More Frequent Detailed External Inspections
8.6.2.2	Immersed Test Coupons
8.6.2.3	Secondary Containment

-------
8-2
8.0 INSPECTIONS
A tank and its auxiliary equipment must be properly inspected
on a routine basis to ensure that the tank system is in good working order
primarily to prevent uncontrolled discharges of hazardous wastes to the
environment. Inspections may result in the conclusion that the tank should
be derated or no longer used for service if the tank is not economically
repairable. The regulations do not require secondary containment for tanks
and, therefore, any leak or other failure is an extremely hazardous situation
to be avoided. Regular inspections using effective procedures are the only
mechanism available to forecast the possibility of tank failure.
The permit writer is responsible for specifying the minimum
allowable shell thickness and the frequency of inspections. According to
regulation 40 CFR 264.194(b), the applicant is required to develop a pro-
cedure to assess the condition of its tanks. The permit writer should be
concerned that the procedure proposed by the applicant will detect any
defect in the tank before the defect's depth can violate the minimum
shell thickness. The maintenance of a minimum shell thickness
for a tank ma.y be viewed as being similar to the requirement for
secondary containment.
In general, this chapter 8.0 is written with metal tanks in mind.
Fiberglass-reinforced plastic tanks are somewhat different in that they often
fail by different mechanisms of deterioration than metal tanks. Section
4.2 " Fibernlass-Reinforced Plastic Tanks" presents further information
about plastic tanks, including a discussion about minimum shell thickness
equivalent (Section 4.2.3) and frequency of inspection (Section 4.2.4).
Stress corrosion around weld seams, corrosion at the liquid-vapor
interface, oxidative corrosion due to the presence of oxygen (from the air)
in the vapor space of vented atmospheric tanks, caustic embrittlement, and
hydrogen blistering are all types of corrosion which may occur in a non-
uniform way on the surface of the metal. However, careful visual inspection
for these types of corrosion will usually be adequate to detect the possi-
bility of defects which would require more detailed examination. However,
pitting is another form of corrosion that in some cases may not be readily

-------
8-3
detected through visual inpsection. Furthermore, the nature of corrosion by
pitting is such that once the pit has been formed, the rate of corrosion may
be accelerated.
Pitting may occur where the liquid is locally stagnant, and
a concentration gradient of electrolyte may develope that, in turn,
develops a small electrolytic cell, causing localized corrosion in
the form of a pit. Tank bottoms, weld seams and dead pockets are
the tank locations in which pitting often occurs. Liquid streams
containing chlorides are notorious for the possibility of pit corrosion,
as are liquid streams containing sludges which may settle.to the tank
bottom and, thus, form a dead pocket. Pitting has been observed directly
below openings on tanks storing crude petroleum due to rainwater settling
to the bottom and forming electrolytic solutions from the salts cont-
ained in the petroleum. In some cases, the pits formed in the metal may
not appear to be pits upon causual inspection because they have become
filled with corrosion products and sludge.
Thus, the permit writer should
•	Require that the applicant provide information on the
expected corrosion rate of the liquid on the tank material
and the likelihood of pitting and other forms of non-uniform
corrosion.
•	Insist that the applicant provide information supporting
the conclusion that inspections will be performed by quali-
fied personnel using procedures that would detect both
uniform and non-uniform corrosion of all types.
All permit writers unfamiliar with tank inspection procedures
should read the American Petroleum Institute fluide for Inspection of
Refinery Equipment, Chapter XI11,"Atmospheric and Low-Pressure Storage
Tanks". Particular attention should be given to sections 1302 through
1306.03.
A list of tools required for tank inspections is presented in
Section 1304 of the API Inspection Guide as Tables 1 and 2. (Also see
section 8.5 of this manual.) Relatively detailed explanations on how many
of the common tools are used in inspection are presented in the text.

-------
8-4
It should be noted that in the API Guide relatively heavy
reliance is made initially on visual techniques of tank inspection to
detect evidence of non-uniform corrosion; however, upon detection of
potential defects, more sophisticated methods are used to verify and
determine the extent of the defect. For example, pits may be measured by
depth gage; thicknesses determined by calipers or, in some cases, drilling
a hole, which is then measured by hook gage and plugged; cracks measured
by penetrant dye or magnetic particle techniques and leaks verified by a
vacuum-box tester with soapy water. In some cases test specimens may be
removed from some portion of the tank (frequently the bottom )
for detailed examination. Some of the methods mentioned above are destruc-
tive in nature. Ultrasonic thickness detectors are commonly used to
measure for changes in thickness due to uniform corrosion and to detect
other flaws. Ultrasonic inspection has the advantage that measurements
may be made from the exterior of the tank.
8.1 Evaluation of Inspection Plan
The inspection plan proposed in a permit application should
clearly describe all the procedures required to comply with the regula-
tions in 40 CFR 264.194. In brief outline format the required inspections
are:
(1)	Overfilling control equipment, once per day
(2)	Data on tank operating conditions, once per day
(3)	Level in uncovered tanks, once per day
(4)	Above-ground (external) portions of the tank to
detect corrosion and leaks, once per week ¦
(5)	Area around tank to detect signs of leaks, once
per week
(6)	Detailed external and internal assessment of tank
condition adequate to detect cracks, corrosion,
erosion, or wall thinning that may lead to leaks or
inadequate strength according to a predetermined schedule.
The daily inspection of (1) overfilling control equipment is covered in
greater detail in Section 8.4.3.

-------
8-5-
The daily inspection of (2) data on tank operating conditions
such as pressures, temperatures, and liquid levels that should be recorded
on operator's log sheets or on charts from recording instruments, should
be part of the normal operating procedure. Operators and foremen, should
be trained about the range of values that are acceptable practice and to
notify supervision when such values have been violated. Further discussion
of this latter type of inspection is not presented here.
The daily inspection of the (3) level of wastes in uncovered
tanks to assure adequate freeboard to prevent overtopping due to winds
or precipitation is reasonably self-explanatory. Of course, specific
standards should be established to guide operators on the maximum levels
that can be allowed without problems. It would be prudent for the appli-
cant to initially set very conservative maximum levels and then base any
changes on observations made on windy days. The minimum freeboard that
should be allowed is a function of many variables, including maximum
wind velocity, nearby topography and buildings, windscreens, wind
direction, tank diameter, liquid viscosity, and ma'ximum 24-hour (or
longer) rainfall. This inspection is to be made visually and not by
reliance on instruments and other indirect means of data acquisition.
Further discussion on this inspection is not presented herein.
The weekly inspection of (5) the area around tanks to detect
signs of leaks such as wet spots, dried residues, dead vegetation, or
discolored spots does not require further explanation.
The remainder of this section presents information to guide
the permit writer on the inspection of
(1) Over-filling control equipment, daily
(4)	Above-ground (external) tank inspection for
leaks and corrosion, weekly
(5)	Detailed external and internal assessment of tank
condi tion

-------
8-6
8.2 Weekly Above-Ground External Tank Inspection
Regulation 264.194(a)(4) requires inspection of "the construction
materials of the above-ground portions of the tank, at least weekly, to
detect corrosion or erosion and leaking of fixtures and seams". The
intent of this regulation should be viewed more as an attempt,to detect
leaks or the potential for imminent leaks, and less as a detailed assess-
ment of the condition of the tanks. Items to-be assessed during this
inspection include:
•	Erosion around and cracks in the foundation and pads
a Corrosion, leaks, or distortion around nozzles and piping
connected to the tank
•	Evidence of deterioration of protective coatings by the
appearance of corrosion, discoloration, blisters or other
film lifting.
•	. Evidence of corrosion of tank tops or roofs
•	Proper functioning of roof seals ("if any) and roof drains
(if any)
•	Corrosion, discoloration, leaks, cracks, bulges, and
buckles of seams and plates of the tank wall and bottom
(if accessible).
If the external portions of the tank are covered with insulation, then
careful inspection of the insulation for leaks or evidence of leaks such
as discoloration would be the appropriate procedure.
Until potential defects are observed, this inspection is s.trictly
a visual inspection. However, upon detection of a defect, more sophis-
ticated inspection procedures would be appropriate. Of course, if a leak
is detected, further leakage should be stopped and the tank promptly
repaired or replaced.
8.3 Detailed Assessment of Tank Condition,
(as scheduled)
The detailed assessment of tank condition proceeds in two stages,
the external inspection and the internal inspection, as follow:

-------
8-7
3.3.1 External Inspection
Many elements of the external tank inspection may be made while
the tank is in service; for example, ultrasonic examination of the average
shell thickness. (However, the measurement of average shell thickness
is listed as part of the internal inspection procedure in this document.)
A detailed description of the external tank inspection procedure is pre-
sented as section 1304.02 of the API Guide of Inspection of Refinery
Equipment, Chapter XIII, "Atmospheric and Low-Pressure Storaqe Tanks" and
is not repeated herein. However, a checklist of the items to be investi-
gated and what to look for has been presented as Table 8-1 based on the
API Guide. Some external inspection procedures should not be performed
until the tank has been shut down and emptied.
8.3.2 Internal Inspection
The internal inspection described by the applicant should take
place in at least two major phases—emptying the tank and the inspection.
According to 40 CFR 264.184(b), the applicant must establish procedures
for emptying the tank to allow entry and inspection of the interior.
Although the intent of this regulation is not made explicit, the permit
writer should be concerned with safety of personnel, avoidance of spills
to the environment, and other hazardous conditions. A checklist of items
with which to be concerned is presented, in Table 8.2. The checklist
includes consideration of lined and fiberglass-reinforced plastic (FRP)
tanks. (Further information about FRP tanks is presented in Section 4.2.3.)
8.4 Inspection of Auxiliary Equipment
Common auxiliary equipment and system components attached
to tanks used for hazardous waste include pipes, valves, and fittings;
pumps and compressors; and instruments, control equipment, and electrical
systems. Inspection of these are discussed in the following sections.

-------
8-8
TABLE 8-1. CHECKLIST OF TANK EXTERNAL-INSPECTION POINTS
A. Tank In Service
(1)	Ladders, Stairways, Platforms and Walkways
-	worn or broken parts and treads
-	corroding parts
-	cracked or spalled concrete pedestals
-	low spots where water can collect
-	loose rivets and bolts
(2)	Foundations
-	erosion
-	uneven settlement
-	cracks and spalling in concrete pads, base
rings,and piers •
-	deterioration of water seal between tank
bottom and the foundation
-	distortion of anchor bolts
(3)	Pipe Connections
-	external corrosion
-	cracks and distortion
(4)	Electrical Grounds
-	corrosion where enters ground
-	resistance
(5)	Protective Coatings
-	rust spots, blisters, and film lifting
(6)	Tank Walls
-	corrosion (underground and under insulation in Darticular)
-	discoloration of paint surface
-	cracks at nozzle connections, in welded seams,
and at the metal ligament between rivets
-	cracks, buckles, and bulges
-	tightness of bolts or rivets, if applicable
(7)	Tank Roofs
-	malfunctioning of seals
-	blockage or breakage of water drains on roofs
-	corrosion

-------
8-9
TABLE 8.1. (Continued)
(8) Overfilling Control
-	malfunction of controls
-	insufficient freeboard
Tank Out of Service
(1)	Tank Bottoms (only if appropriate)
-	tunneling method
(2)	Pipe Connections
-	hammering
-	at point of entrance at soil line
(3)	Tank roofs, pontoons, double decks, seals, and purlins
-	hammering
-	visual
-	leaks
(4)	Valves and Valve Seats
-	visual
(5)	Auxiliaries
-	vents for plugging, breather valves for seating
-	liquid-level controls for cracks and corrosion
-	pressure gages for plugging and accuracy.

-------
8-10
TABLE 8-2. CHECKLIST FOR TANK INTERNAL INSPECTION
(TANK OUT OF SERVICE)
A. Tank Emptying and Preparation For Inspection
-	avoidance of spills
-	avoidance of hazardous conditions (reaction, ignition,
or toxic exposure)
-	use of appropriate materials of construction for any
temporary storage containers (or tanks) and connecting
systems
-	cleaning and ventilation procedure
-	complete disconnection or blanking off of all connecting
piping
-	air quality check
-	adequate lighting
-	personal safety equipment as appropriate (clothing and
respiratory)
-	Standby equipment and services readily available
B. Interior Inspection of Solid Steel Tanks
(1)	Roof and Structural Supports (visual first for safety)
-	no hazard of falling objects
-	corrosion
(2)	Roof and Structural Supports (more rigorous)
-	loss of metal thickness
-	cracks, leaks at welds
-	cracks at nozzle connections
-	malfunction of floating roof seals
-	water drain system deterioration
-	hammering
(3)	Tank Shell
-	cracks at seams
-	corrosion of vapor space and liquid-level line
-	cracking of plate joints
-	cracking of nozzle connection joints
-	loss of metal thickness

-------
8-n
TABLE 8-2. (CONTINUED)
(4) Tank Bottom
-	hammer testing
-	corrosion pits
-	sprung or cracked seams
-	rivets for tightness and corrosion
-	depressions in the bottom areas around or under
roof supports and pipe supports
-	bottom thickness
-	unevenness of the bottom
-	sample coupons, if appropriate
C.	Interior Inspection of Lined Steel Tanks
NOTE: Some of the procedures and locations to inspect noted
in section B for solid steel .tanks above are equally •
applicable to lined tanks. Tanks may be lined with alloy
steel, lead, rubber, glass, coatings, and concrete.
-	general condition of lining (holes, cracks, gaps, corrosion,
erosion, swelling, hardness, loss of thickness)
-	proper positioning of liner
-	bulges, blistering j or spalling
-	spark testing with rubber, qlass, and orqanic type
coatings
-	ultrasonic examination of steel outer shell thickness
is possible if any deterioration is suspected.
D.	Interior Inspection of Fiberglass Reinforced Plastic Tanks
-	hardness test of any test specimens exposed to liquid
in the tank
-	indentations,cracks, exposed fibers, crazing, checking,
lack of surface resin, and delamination
-	if sufficiently translucent, porosity, air or other
bubbles, other inclusions, and thin areas
-	ultrasonic examination of laminate thickness is possible
if any deterioration is suspected in the polyester matrix.

-------
8-12
8.4.1 Pipes, Valves, and Fittings
Inspections of pipes, valves, and fittings are usually con-
ducted to note any leaks, cracks, corrosion, or losses in metal thickness
owing to external or internal deterioration. The internals of these
equipment parts are subject to erosion or wear because of the effects of
high liquid turbulence or velocity. Areas around pipe bends, elbows, tees,
and other restrictions, such as orifice plates and throttling valves,
are particularly subject to erosion.
Visual inspection techniques include checking for leaks, mis-
alignment, unsound piping supports, vibration or swaying, external cor-
rosion, accumulations of corrosive liquids, and indications of pipe
fouling. Thickness measurements while the pipes are in operation can be
taken utilizing ultrasonic or radiographic techniques.
If the tank is out of service or if a line can be valved off,
with proper safety precautions piping can be opened at various places
by removing a valve, or fitting or by springing the pipe apart at flanged
locations to permit internal visual inspection. A flashlight or exten-
sion light is needed in most cases and a probe-type instrument, such as
a borescope, or a mirror and light will permit a more detailed view. If
corrosion or erosion conditions are noted visually for some parts,
radiographic or ultrasonic techniques can be used to inspect the entire
length of pipe, if inaccessible to visual examination. Replacement may
be more economical than such techniques in some cases if the entire
piping run is suspect. Gaskets should often be replaced if the line is
broken at flanges.
A brief checklist for inspection of piping, valves, and fittings
is presented as Table 8-3.

-------
8-13
TABLE 8-3. CHECKLIST FOR INSPECTION OF PIPING,
VALVES, AND FITTINGS
-	Leaks
-	Cracks or corrosion
-	Metal thickness (by hammering or caliper)
-	Metal thickness (by ultrasonics, radiation, or eddy
current
-	Gasket condition
-	Alignment, distortion, and swaying
-	Valve seats
-	Pipe rack supports or hangers
-	Vibration
-	Erosion
Piping systems that cannot be inspected visually are frequently
pressure tested. They include:
•	Underground and other inaccessible piping
•	Complicated manifold systems
•	Small pipe and tubing systems
•	All systems after a chemical cleaning operation.
The most used media for pressure tests is water. In this type
of test the water is pumped into the pipe such that the quantity of gas
in the pipe is minimal. When the pressure has reached the test pressure,
the system is valved off but with a pressure gage on the closed system.
Small leaks of the incompressible water results in a rapid and significant
drop in pressure and, thus, the probability of a leak is established. Use
of compressible or condensible gases such as steam, air, carbon dioxide,
and so forth is generally less reliable; more reliance must be placed on
hearing the sound of escaping gas or otherwise detecting leaks.

-------
8-14
8.4.2 Pumps and Compressors
Mechanical wear is the predominant cause of deterioration of
pumping and compression equipment, although erosion and corrosion are
also responsible for an appreciable amount of deterioration. Other
deteriorating factors include improper operating conditions, piping
stresses, cavitation, and foundation deterioration causing misalignment
or vibrations.
Since vibration can rapidly deteriorate a pump or compressor,
periodic examination of the vibration level should be made using an
electronic vibration meter. Inspection of all assembly bolts, gaskets,
cover plates, and flanges should be conducted to detect leaks and cracks
as a result of vibration or abnormal operating conditions.
A brief checklist for the visual inspection of pumps and
compressors is presented as Table 8-4.
TABLE 8-4. CHECKLIST FOR VISUAL INSPECTION
OF PUMPS AND COMPRESSORS
-	Misalignment
-	Foundation cracks and uneven settling
-	Missing or broken anchor bolts
-	Leaky piping connections
-	Excessive vibrations and noise
-	Deteriorating insulation
-	Depleted lubrication oil reservoir
-	Missing safety equiDment such as a pump coupling guard
-	Burning odor or smoke
-	Excessive dirt
-	Excessive corrosion
-	Leaks and cracks at assembly bolts, gaskets coverplates,
and flanges
Two pumps are often installed in parallel such that one pump
may be shut down while the other does all the required pumping. Thus,
one pump may undergo a complete internal inspection or replacement
while the system remains in operation.

-------
8-15
8.4.3 Instruments, Control Equipment, and
Electrical Systems
Instruments, control equipment) and electrical systems must be
inspected at the minimum required frequencies given in 40 CFR 265.194
and section 8.5 of this manual to ensure that they are in good working
order. Level controls, emergency shut-off devices, and alarms are among
the most important devices for fail-proof tank operation. Flow rate
controls, temperature gauges, pressure gauges, and analyzers are among
the less important devices.
A brief checklist of what should be inspected regarding the
instruments, control equipment, and electrical systems is presented in
Table 8-5.
TABLE 8-5. CHECKLIST FOR INSPECTION OF INSTRUMENTS
AND CONTROL SYSTEMS
-	Instruments
-	Transmission systems
-	Power supplies
-	Seals
-	Panels and enclosures
-	Electrical Equipment
-	Insulation
-	Operating mechanisms (moving parts)
-	Insulating and lubricating oils
-	Protective relays
-	Bearings
-	Batteries
-	Connectors
-	Rectifiers
The visual inspection should specifically watch out for any
deteriorating effects of the following on electrical systems:
-	Heat
-	Dirt
-	Moisture
-	Chemical attack.

-------
8-16
The instruments and controls must be calibrated by qualified
personnel as per the methodology and frequency recommended by the
vendors.
Inspection of the data gathered by instruments should be included
as an integral part of the overall inspection plan for instruments, control
equipment, and electrical systems. Any unexpected discontinuities or
abnormal peaks in data charts or data logs may indicate that there is some
cause for concern in the control systems.
8.5 Inspection Tools and Procedures
When visual inspection suggests that tools are needed for a more
detailed inspection, simple hand tools may be used as an initial aid.
Tools such as a scraper, digger, flange spreader, knives, paint or crayon,
portable lights, and rules are indispensable for visual inspection.
Additional tools such as hammers, mirrors, magnifiers, magnets, and
internal visual scopes are also helpfyl.
The mechanical measuring tools include calipers, micrometers,
scales and tapes, wire gauges, level and plumb bob and line, depth
gage, hook gage, square, and straightedge.
Approved destructive examination methods include drilling a
hole through the tank wall or bottom, then using a hook gage to measure
thickness, tapping the hole, and inserting a threaded plug. Another
method is to cut large (12 inch by 12 inch) test specimens from the
tank for detailed examinations; this is often performed for tanks where
the bottom cannot be externally inspected. A trepanning saw may be
used to remove a portion of a weld from the tank for examination.
Brief descriptions of other inspection tools and methods follow:
8.5.1 Hammering Method
Full blows of the hammer are used and the sound, feel, and
imprint of the hammer head noted. Where corrosion or erosion is signifi-
cant, the sound will be dull, the feel soft, and a dent or hole likely.
Hammering is frequently performed on tank roofs, bottoms, and on floating
roof components.

-------
8-17
8.5.2	Penetrant-Dye Method
Penetrant dyes are often used to define surface cracks on a tank
that would not be verified by a visual inspection. The penetrant is applied
by either brushing or spraying to a surface carefully cleaned (often by
sandblasting), dried, and then the excess is removed. After a few minutes
of contact to allow penetration into the crack, a chemical developer is
then applied to the surface. The dye stains the developer and exposes the
extent and size (but not the depth) of any defects.
8.5.3	Magnetic-Particle Method
The magnetic-particle method is also used to define surface
cracks on tanks similar to the penetrant-dye method. The surface must
also be carefully cleaned and then iron particles are sprinkled on the
surface. A magnetic field is then imposed near the particles either by
a permanent magnet (especially if flammable materials are stored nearby)
or an electromagnetic device and the particles arrange themselves along
the crack and particularly near the ends of the crack. The magnetic
field should be imposed in two directions to assure there is no crack
or to identify two or more cracks running in different directions. No
indication is given about the depth of the crack. This method may be
used only on tanks constructed of magnetic materials.
8.5.4	Radiographic Method
Welds are often radiographed during tank	fabrication to detect
thickness and flaws of the welds. This method may	also be used to deter-
mine thickness of tank plates. The device may use	either X-rays, or gamma
radiation and must be calibrated prior to use. It	is similar in many
respects to the X-ray machines used for dental and	medical purposes.

-------
8-18
8.5.5	Ultrasonic Method
Ultrasonic instruments can be used to measure the tank's thick-
ness and determine the location, size, and nature of defects. They can
be used while the tank is in operation as only the outside of the tank
needs to be contacted with the device. Two types of ultrasonic instruments,
the resonance and the pulse type, are most commonly used for tanks. The
pulse type utilizes electric pulses and transforms them into pulses of
ultrasonic waves. The waves travel through the metal until they reach a
reflecting surface. The waves then are reflected back, converted to
electrical pulses, and show up on a time-base line of an oscilloscope.
The instrument is calibrated by using a material of known thickness. There-
fore, the time interval between the pulses corresponds to a certain
thickness.
8.5.6	Vacuum-Box Method
The vacuum box is an open box in which the lips of the open
side are covered with a sponge rubber gasket, and the opposite side
is glass. A vacuum gauge and air siphon connection are installed on
the box. The seam of the tank shell where a leak is suspected is first .
wetted with a soap solution, then the vacuum box is pressed tightly
over the seam. The foam-rubber gasket forms a seal, and a vacuum is
achieved inside the box by the air siphon. If a leak exists, bubbles
will form inside the box and can be seen through the glass.
8.6 Frequency of Tank Inspection
There are several requlator.y requirements reqardinq tank inspec-
tions and, in the case of the detailed assessment of tank condition,
other practical considerations.
8.6.1 Regulatory Requirements
The frequency of performing some types of tank inspections is
presented in 40 CFR 264.19 and is summarized in Table 8-6.

-------
8-19
TABLE 8-6. MANDATED INSPECTION FREQUENCIES
At Least Once Per Normal Operating Day
. - Overfilling control equipment
-	Data on tank operating conditions
-	Level in uncovered tanks
At Least Once Per Week
-	Above-ground external portions of tank
-	Area surrounding tank
Although the permit applicant is required to present a schedule
for the detailed assessment of tank condition, the permit writer is
ultimately responsible for specifying the appropriate schedule in the
permit issued to the applicant.
¦ Title 40 CFR 264.15(b) states that the frequency of inspection
for other items should be based on the rate of possible deterioration of
the equipment and the probability of an environmental or human health
incident if the deterioration or malfunction goes undetected. Part 264.194(b)
requirements for periodic comprehensive tank inspections specify the
following additional factors to be used in determining inspection intervals:
•	Material of construction of tank
•	Type of erosion or corrosion protection used
•	Characteristics of waste being stored
•	Rate of corrosion or erosion observed during previous
inspections.
8.6.2 Practical Considerations
The detailed tank assessment is often a costly requirement
for the operator of a hazardous waste storage facility because the tank
must be shut down, blocked off, emptied, cleaned, and undergo detailed
examination by qualified personnel. Unless the operator has spare tanks,
shutdown of the tank may temporarily also necessitate closure of the

-------
8-20
facility. Adequate tank cleaning for personnel safety may also be a
costly step in terms of both elapsed time and other dollar costs. Further-
more, there is always some potential for residual hazardous materials to
remain in nozzles or piping associated with the tank. Thus, tank inspec-
tions must be frequent enough to avoid leaks and spills but should not
become necessarily burdensome to the operator of the facility.
In cases where the corrosion rate data are known at storage
temperature for the specific material of construction of the tank with the
specific liquid to be stored in the tank and only uniform corrosion has
been experienced in prior applications, the expected service life of the
tank can be realistically estimated, which can then be used to establish
a reasonable inspection schedule. During the initial years, scheduled
inspections at 20, 40, and 60 percent of the tank's service life would
be reasonable frequency. For example, a tank with an expected service
life of 25 years might initially be subjected to a comprehensive inspection
every 5 years to establish the actual rate of corrosion or deterioration.
However, after shell thickness measurements were made and the existence
of any nori-uniform corrosion noted, the estimated service life could be
re-estimated and the inspection frequency increased if necessary as the
tank approaches the end of its service life and the probability of leaks
or ruptures increases; for example, the inspection frequency could be
increased to every 1 to 2 years.
If non-uniform corrosion has been experienced by a material
of construction with the liquid to be contained, much more frequent
initial inspections should be scheduled. Pitting and crevice corrosion
are particularly obnoxious because not only does there often seem to be
an induction period with little observable physical damage, but also the
corrosion accelerates once the pit or crevice is formed due to formation
of a larger electrolytic cell. Materials subject to pitting or crevice
corrosion should normally not be selected unless an economic analysis
clearly indicates a preference toward frequent inspections rather than
to a more costly material of construction.

-------
8-21
8.6.2.1	More Frequent Detailed External Inspections. In some
cases where any form of non-uniform corrosion is not expected, the owner
or operator may prefer to conduct more frequent comprehensive external
inspections of the tank to avoid the expense of frequent internal inspec-
tions, providing all portions of the tank are accessible, including the
bottom. In the example cited above, the owner-operator could initially
conduct annual external inspections, which include intensive measurements
of tank shell thickness (i.e, one measurement per square yard of surface
area) and reduce the¦frequency of internal inspections to once every 7
years. As the condition of the tank deteriorates, however, the frequency
of internal inspections should increase to every 1 or 2 years.
8.6.2.2	Immersed Test Coupons. In cases where few corrosion
data are available or proper tank inspection would be very costly, test
specimens (coupons) of material literally from the same heat (or batch) of
the metal used to construct the tank may be immersed in the liquid, with
some coupons allowed to rest on the bottom.of the tank. These test coupons
may be stressed by bending and welding to form crevices to simulate
problem areas in the tank. Samples can be withdrawn annually and measure-
ments made of thickness and observations made about stress, crevice, and
pitting corrosion. The data collected could then be used to suggest an
appropriate inspection schedule; of course, an inspection schedule should
be established that requires greater frequency than that projected by
the data from the coupons.
In the case where a used tank has been installed and no coupons
may be taken from the specific heat (or batch) of material from which
the steel plate was manufactured, and the specific type of material used
is not certain, a small sample of the tank metal may be removed and ana-
lyzed by emission spectroscopy to classify the metal used. Then test
coupons may be made from this type of metal for immersion in the tank.
However, a more conservative inspection schedule should be developed based
upon this circumstantial data than indicated in the prior example.

-------
8-22
8.6.2.3 Secondary Containment. If the tank operates at close
to atmospheric pressure and a leak would not cause undue detriment to
personnel, property, or the environment if the leak were collected in a
secondary containment system, then some reduction could be considered in
the frequency of the internal tank inspection. Obviously, if the material
stored were volatile and toxic upon inhalation or if the waste were highly
reactive with water (rain) or with the material of construction used for the
containment system, then this approach would not be suitable.
Other considerations include the size of the tank or quantity
of material which might-be leaked to the secondary containment system.
Upon detection of a leak it may be possible to rapidly pump the tank's
contents to an alternative tank as a temporary measure and thus avoid
too large a spill. Of course, good housekeeping combined with frequent
inspections would be required to assure that any leaks were detected soon
after failure. The difficulty in cleaning up a spill should also be con-
sidered. Because of the difficulty in inspecting insulated tanks for leaks,
reliance on secondary .containment and early leak detection in this situation
would not be practical.
Another problem emerges with tanks where the bottom rests directly
on a foundation such that the bottom cannot be externally inspected. Ob-
viously, the foundation must be within the secondary containment system.
Furthermore, if pitting or other forms of non-uniform corrosion are experi-
enced in the bottom, a leak may be present for a significant period of
time before it becomes detected. During this period of time, considerable
further deterioration of the tank may continue leading to a major failure.
Particular attention should be given to avoiding ignition of
hazardous wastes if they are combustible . Use of expolosisi on-proof motors and
prohibit of nearby motor vehichles and the like near the secondary
containment system. Also, there should be no possibility of mixing incom-
patible wastes in the same secondary containment area if simultaneous leaks
were to occur.

-------
APPENDIX F
COST ESTIMATES FOR CONSTRUCTION OF STORAGE TANK SYSTEMS
•	Basis and Rationale for Cost Estimates
•	Diagram-Typical Underground' Storage Tanks
•	Cost Estimate for Storage Facility with One
1,000 Gallon Tank
•	Cost Estimate for Storage Facility with Two
5,000 Gallon Tanks

-------
BASIS AND RATIONALE
FOR COST ESTIMATES OF SAMPLE UNDERGROUND
STORAGE TANK INSTALLATIONS
1)	Cost/Price Data: The Means Building Construction Cost
Data, 1983 edition [1] , was the primary source of price data,
with supplementary information obtained from suppliers of tanks,
level indicators and cathodic protection systems. Means cost/
price data is based on average figures nationwide as of January
1983. Future costs will escalate due to inflation and will fluc-
tuate upward and downward due to location and construction acti-
vity; e.g., when there is a lull in construction activity, bid
prices will decrease.
2)	We assumed that the storage facility is installed at an
existing facility and is installed by a contractor who regularly
installs tanks; i.e., the prime contractor does all of the work
with his own forces. Scale factor cost considerations are ac-
counted for by adjusting unit prices from Means, including esti-
mates for mobilization and standby costs, and by varying contin-
gency percentages.
3)	The tanks are installed under a paved area (parking lot
or driveway) to facilitate access by tank trucks that are used to
empty the tanks. Typical industrial plants/facilities try to
minimize unusable space which would be the case if the tank were
installed without pavement protection. Pavement protection is
required by NFPA 30 and API 1615 for underground tanks subjected
to vehicular traffic (and is also recommended by most tank manu-
facturers)'.
4)	The liquids to be stored are ignitable or reactive
wastes. These classes of wastes were chosen because the cost of
storing them underground is less than the added cost of fire pro-
tection required for aboveground storage. Corrosive and toxic
wastes can be stored aboveground at less cost than underground
storage and can be monitored for leaks and easily maintained,
compared to underground installation.
5)	Steel tanks were chosen due to their (1) compatibility
with most flammables and combustibles and (2) lower purchase and
installation cost than FRP tanks. FRP tanks must be evaluated
for compatibility with flammables and combustibles, especially
organic solvents.
6)	The Steel Tank Institute's "sti-P3 System" was chosen
for corrosion protection. It provides corrosion protection
through a protective coating, cathodic protection (sacrificial
anode), and electrical isolation of the tank. NFPA 30 requires
cathodic protection of steel tanks if the soil resistricity is
less than 10,000 ohm-cm or if there are other corrosive condi-
tions. The sti-P3 system carries a 20 year limited warranty for
soil resistivity levels of 2,000 ohm-cm or more (corresponds to
medium corrosive soils). Soils which have resistivity of 10,000
F-l

-------
ohm-cm or more generally have good internal drainage (high perma-
bility and are not saturated) and have low corrosivity.
7)	Level indicators are recommended for hazardous waste
storage systems to minimize the amount of contact with the liquid
being stored. Dip-sticking, the least capital cost method of
determining the liquid level, requires the operator to come into
direct contact with the waste. A direct-reading, float- type,
mechanical level indicator was used for the 1,000 gallon tank in-
stallation. This particular class of level indicator, the first
step above dip-sticking, allows the operator to readily determine
the liquid level at a cost in line with the cost of the facility.
We expect that a facility which has a 1,000 gallon storage tank
does not generate large volumes of waste. The waste is generated
infrequently or at very low rates. Thus, frequent liquid level
monitoring is not required. The operator would inspect the level
in the tank approximately once per week (if there is a regular
flow into the tank) or before and after the waste is transferred
to the tank if discharged infrequently.
An electric level indicator system with remote indicator in
the industrial plant is considered good practice for the two
5,000 gallon tank facility. The cost of the level indicator is
proportional to the cost of the entire storage facility. The
electronic system is composed of a level sensor mounted in the
tank and a level indicator mounted in a control room inside the
industrial plant. This type of level indicating system provides
continuous monitoring of the levels in the tanks without the
operator having to-leave, his normal work place. Continuous moni-
toring is assumed for a facility that has a continuous flow of
wastes into the tank(s) or regular, large volumes of wastes being
transferred to the tank(s). With these types of operations, the
possibility of over filling a tank is higher than the small stor-
age facility. Further, if a tank develops a rupture; more liquid
could drain out before the rupture is detected if a convenient
remote readout device is not available.
8)	With ignitable or reactive wastes being stored, the in-
stallation must conform to NFPA 30. The code requires a firm
foundation at least 6 in. of non-corrosive inert material (well-
drained sand or gravel) surrounding the steel tank and 13 in. of
earth with 6 in. of reiriforced concrete over the tank when sub-
jected to vehicular traffic. API 1615 and most tank manufactur-
ers recommend 12 in. clearance around the steel tank, 6 in.
under the tank and 18 in. over the tank with a 6 in. reinforced
concrete slab (for vehicular traffic). The sti-P3 corrosion pro-
tection system also prescribes 12 in. of well-drained sand or
gravel around and over the tank to isolate the tank from corro-
sive attack.
9)	¦ We assumed that the tank will be installed approximately
20 feet from the industrial plant. The vent is installed against
the building for protection and support with the vent opening 12
feet above grade and directed away from the building or any
building openings. (as per NFPA 30).
F-2

-------
10)	A concrete anchor is provided for the two 5,000 gallon
tanks to prevent the tanks from "floating" from buoyant forces
when empty during periods when the soil is saturated and/or if
the water table is high. With the bottom of the tanks being 11
feet deep (8 feet in diameter plus another 2 feet of cover) it is
often into a zone which is at least partially saturated, or the
ground water might rise to above' the tank bottom. Slab-shaped
anchors were chosen over concrete deadmen for ease of construc-
tion.
On the other hand the 1,000 gallon tank, which is only 4
feet in diameter, will be buried about 7 feet below the surface,
a level which normally should be above the ground water and is in
a zone which is generally fairly well areated. Thus, no anchor
was provided for the 1,000 gallon tank. Furthermore, the weight
of cover soil and the pavement slab over the tank provides resis-
tance against buoyant forces.
11)	The conductors for the electric level indicating system
are installed in conduits to facilitate maintenance and repair
under paved areas and for protection inside the building. The
conductors are assumed to extend 50 feet inside the plant (in-
cluding horizontal and vertical runs) to a control room.
12)	Cast-iron soil pipe is used for gravity flow from the
building to the tank. Cast-iron pipe was chosen due to its long
life, compatibility with the waste being transferred, and similar
installation requirements as the building drain piping. Plumbing
in industrial plants is most commonly cast-iron (for noncorro-
s ives). '
13)	The pump-out port is located directly above the tank.
The top of the port is enclosed in a valve box that is integral
with the concrete pavement slab. The valve box protects the
pump-out port from vehicular traffic.
14)	Manways are good practice; they allow access to inspect
and rehabilitate the tank. In some localities manways are re-
quired for larger tanks. Various manufacturers estimate that 20
to 100 percent of their tanks are furnished with manways. We as-
sumed that one manway is provided for each of the 5,000 gallon
tanks; the 1,000 gallon tank is not large enough to warrant the
additional cost of a manway.
15)	All the apparatus used to remove the waste from the
storage tank is furnished by the hauler. The hauler drives his
truck onto the concrete pavement slab, lifts off the valve box
cover, opens the pump-out port, inserts a reinforced suction hose
and pumps out the contents using a truck-mounted pump. When the
pumping is complete, the hauler shuts off the pump, disconnects
the hose at the truck (using dry-disconnect couplings), replaces
the suction hose on the tank, closes the pump-out port, replaces
the valve box cover, and leaves.
F-3

-------
Robert Snow Means Company/ Inc. Building Construction Cost
Data, 1983. Construction Consultants and Publishers.
Kingston, Massachusetts, 1982, 420p.

-------
DIAGRAM OF TYPICAL UNDERGROUND STORAGE TANK
STEEL TANK (5,000 GALLON)
GRAVITY DRAIN/FILL LINE
TANK VENT
PUMP-OUT PORT
GAGE PORT & MECHANICAL LEVEL
INDICATOR
(6) REMOTE ELECTRIC LEVEL SENSOR.
INDICATOR & ALARM
® CONCRETE ANCHOR SLAB
© NON-CORROSIVE , INERT BACKFILL
® HOLD DOWN STRAP
BO CONCRETE PAVEMENT SLAB
(Jl) MANWAY

-------
COST ESTIMATE FOR STORAGE FACILITY WITH ONE 1,000 GALLON TANK


QUANTITY
MATERIAL OOST
LABOR COST
ENG1NEERING
ESTIMATE

ITEM DESCRIPTION
NUMBER
UNIT
UNIT
COST
TOTAL
UNIT
COST
TOTAL
TOTAL
Tank Installation







1.
1,000 gallon steel tank, with
sti-P3 protection system, set ir
excavation, 49.5" wide x 10'
long
1
EA
715
715
160.0C
160

2.
Excavation, bulk, medium earth,
truck loaded, wheel mounted
backhoe, 3/4 CY capacity, 7*
deep, minimum 12" clearance
around tank, 1' under tank and
2' cover
19
CY
1.4C
27
1.3C
25

3.
Wood sheeting, wales, braces,
salvaged
312
SF
1.35
421
1.0C
312

4.
Haul away spoil, 6 CY dump
truck, 4 mile round trip
18
CY
2.25
41
1 .44
26

5.
Borrow, buy & load at pit, 2
mile haul, place & spread pea
grave 1
4
CY
8.0C
32
2.62
10

6.
Tank bedding, bank sand, 6"
deep, 6' wide x 12* long
2
CY
3.0C
6
2.62
5

7.
Anchor slab 8" thick mesh rein-
forced 6" wide x 12' long, cast-
in place
8
SY
16.34
131
2.12
17

8.
Backfill pea gravel 12" layers,
and compact to top of tank
4
CY
0.46
2
11.47
46

9.
Backfill on-site material and
compact, 12"
3
CY
0.4€
1
11.47
34

10.
Pavement slab, 6" thick, mesh
reinforced, 4,500 psi, 12' long
x 6' wide
8
SY
12.38
99
1.6C
13

11.
Base course for pavement,
crushed stone, 12" deep, 12'
long x 6' wide compacted
8
SY
3.6?
30
0.7 C
6

12.
Pump-out port, 4" diameter,
black steel, schedule 40
2
LF
10.95
22
8.35
17

13.
Valve box, 6" diameter, cast
i ron set i n pavement
Subtotal - Tank Installation
1
EA
65
65
1,592
48
48
719
2,311

-------
COST ESTIMATE FOR STORAGE FACILITY WITH ONE 1,000 GALLON TANK (CONTINUED)


QUANTITY
MATERIAL COST
LABOR COST
ENGINEERING
ESTIMATE

ITEM DESCRIPTION
NUMBER
UNIT
UNIT
COST
TOTAL
UNIT
COST
TOTAL
TOTAL
Level Indicator







1.
Level indicator, float-type,
d i rect read i ng, i n 6" d i ameter
cast iron valve box set in
pavement
Subtotal - Level Indicator
1
LS
200
200
200
70
70
70
270
Pip
1 nq







1.
Steel pipe, 4" diameter with
coup 1i ngs
20
LF
12.05
241
9.19
184

2.
Excavation, 3' deep, 16" wide,
backhoe, wheel mounted, 3/4 CY
capacity
9
CY
1.4C
13
1.3C
12

3.
Pipe trench bedding, 4" deep,
16" wide, in trench, crushed
bank, run gravel
1
SY
1.12
1
0.35
1

4.
Back fill, on-site material,
compacted in 12" layers, by hand
with vibrating plate
2.5
CY
0.46
1
t
11.47
29

5.
Fill tube, 4" 0, schedule 10,
type 6063-T6 aluminum
4
LF
2.8f
11
4.4C
18

7.
Base Course, 12"
3
SY
3.69
11
0.7C
2

8.
Pavement 6" mesh reinforced
concrete
Subtota1 - Piping
6
ST
12.38
74
352
1.60
10
256
608
Vent







1.
Pipe, black steel, 2" diameter
with couplings
35
LF
3.72
130
5.17
181

2.
Excavation, utility trench,
drain trencher, 12" wide, 24"
deep, backfill & compact
20
LF
0.36
7
0.3C
6

3.
Pipe trench bedding, 4" deep,
crushed bank run gravel
Subtotal - Vent
Subtotal - Bare Costs
0.5
SY
1.13
1
138
2,282
0.32
1
188
1,233
326
3,515

-------
COST ESTIMATE FOR STORAGE FACILITY WITH ONE 1,000 GALLON TANK (CONTINUED)
ITEM DESCRIPTION
QUANTITY
MATERIAL COST
LABOR COST
ENGINEERING
ESTIMATE
NUMBER
UNIT
UNIT
COST
TOTAL
UNIT
COST
TOTAL
TOTAL
Mobilization, material tax § 5%,
labor mark-up i 20$, overhead i M%,
profit § 8% and contingency 6 15?
TOTAL





I
2,357
5,872

-------
COST ESTIMATE FOR STORAGE FACILITY WITH TWO 5,000 GALLON TANKS


QUANTITY
MATERIAL
COST
LABOR COST
ENGINEERING
ESTIMATE

ITEM DESCRIPTION
NUMBER
UNIT
UNIT
COST
TOTAL
UNIT
COST
TOTAL
TOTAL
Tank
1nstallation








1.
5,000 gallon steel tanks with
STI-P3 protection system and
manway, set in excavation
2
EA
2,570

5,140
645
1,290

2.
Excavation, bulk, minimum 12"
clear all around, 12" under
tank, 24" cover, 11' deep, 16'
wide x 19' long, 1-1/2 CY capa-
ci ty
124
CY
1.21

150
0.6!
81

3.
Wood sheeting, wales, braces,
sa1vaged
770
SF
1.35

1,040
1.0C
770

4.
Haul-away spoil, 12 CY dump
truck, 4 mile round trip
115
CY
1.77

204
0.81
93

5.
8" thick, mesh reinforced anchor
slab 16' wide x 19' long
34
SY
16.34

556
2.12
72

6.
Borrow, 2 mile haul, place and
spread pea gravel
37
CY
8.0C

296
2.62
97

7.
Backf ill & compact to top of
tank, pea gravel
37
CY
0.4{

17
11.47
424

8.
Backfill on-site material, 12"
and compact
11
CY
o.4e

5
It.47
126

9.
Pavement slab 6" thick mesh
reinforced 16' wide x 19' long
34
SY
12.38

421
1.6C
54

10.
Base course for pavement crushec
stone, 12" deep, 16' wide x 19'
long
34
SY
3.6?

125
0.7 C
24

11.
Pump-out ports, schedule 40
steel
4
LF
10.95

44
8.35
~ 33

12.
Valve boxes, 6" diameter, cast
iron, set in pavement
2
EA
65

130
48
96


Subtotal - Tank 1nstallation




8,128

3,160
11,288
Level Indicator








1.
Electronic level indicator,
sensor and alarm with line
contro1 to bu i1d i ng
2
EA
1,00C

2,000
20C
400


Subtotal - Level' Indicator




2,000

400
2,400

-------
COST ESTIMATE FOR STORAGE FACILITY WITH TWO 5,000 GALLON TANKS (CONTINUED)

QUANTITY
MATERIAL COST
LABOR COST
ENGINEERING
ESTIMATE
ITEM DESCRIPTION
NUMBER
UNIT
UNIT
COST
TOTAL
UNIT
COST
TOTAL
TOTAL
Piping







1. 4" diameter pipe with steel
coup 1i ngs
40
LF
12.05
482
9.19
368

2. Excavation, 31 deep, 16" wide,
3/4 CY capacity
6
CY
1.40
8
1.3C
8

3. Pipe trench bedding, bank run
gravel
1
CY
1.1!
1
0.32
1

4. Backfill, on-site material,
compacted
5
CY
0.46
2
11.47
57

5. Fill tube, 4" diameter, schedule
40 steel, with couplings
16
LF
12.05
193
9.IS
147

6. Pavement, 6" reinforced concrete
9
SY
12.38
111
1.6C
14

Subtotal - Piping



797

595
1,392
Vent







1. Pipe, 2" diameter steel
55
LF
3.72
205
5.17
284

2. Excavate utility trench 12" wide
x 24" deep backfill & compact
40
LF
0.3C
14
0.3C
12

3. Pipe trench bedding, 4" deep
crushed bank run gravel
1.5
SY
1.12
2
0.35
1

4. Base course, 12" thick,
compacted
1.5
SY
3.69
6
0.7C
1

5. Pavement, 6" reinforced concrete
6
SY
12.3E
74
1.6C
10

Subtotal - Vent



301

308
609
Subtotal - Bare Costs



11,226

4,463
15,689
Mobilization, material tax 6 5%,
labor mark-up i 20%, overhead §
12?, profit e 8{ and contingency
e 151






10,303
TOTAL






25,992

-------
APPENDIX G
Summaries of Selected References

-------
RECENT AND ON-GOING FEDERAL, STATE OR LOCAL
AND INDUSTRY OR TRADE ASSOCIATION
STUDIES ON UNDERGROUND TANK STORAGE
FederaI
REFERENCE: A. T. Kearney, Inc. and PEDCO Environmental, Inc. "A Guide for
Preparing RCRA Storage Permit Applications", (Draft). U.S. Envi-
ronmental Protection Agency, Washington, D.C., 1982, 280 pp.
ABSTRACT: This draft guidance is intended for use by owners/operators of
hazardous waste storage facilities in developing Resource Conser-
vation and Recovery Act (RCRA) Part B Permit Applications. Com-
ments from applicants or other persons outside of the U.S. Envi-
ronmental Protection Agency (EPA) have not been incorporated in
this guide. Detailed technical instructions covering the required
contents of the RCRA permit applications and explanations concern-
ing administrative procedures in the permitting process are in-
cluded.
The suggested permit application requires inclusion of the follow-
ing information related to tanks:
•	Tank descrIpt i on;
•	Tank corrosion protection;
•	Tank management practices;
•	Tank i nspectIon;
•	Tank spills and leakage; and
•	Closure of tanks.
REFERENCE: Franklin Associates, Ltd. "Technological Character i zat i on o.f
Waste Oil- Storage", (Draft)'. Prairie Village, Kansas, February
1983, 54 pp.
ABSTRACT: This draft report prepared for the U.S. Environmental Protection
Agency (EPA) discusses waste oil losses in aboveground and under-
ground storage tanks. Estimates of the frequency and magnitude of
leaks in waste oil storage tanks are provided. Underground tanks
have a much greater probability of loss than aboveground tanks.
The probability of leakage in an underground tank is conserv-
atively estimated at 12 to 14 percent, compared to 1.7 percent for
aboveground tanks. Two approaches were used to estimate the prob-
ability of leaks in underground waste oil tanks, while a "fault-
tree" analysis was used for aboveground tanks.
REFERENCE: Fred C. Hart Associates, Inc. "Facilities Storing or Treating
Hazardous Waste in Tanks, a Technical Resource Document for Permit
Writers", (Draft). U.S. Environmental Protection Agency, 1982,
130 pp.
ABSTRACT: Fred C. Hart Associates, Inc. prepared this report for the U.S.
Environmental Protection Agency as part of a series of technical
resource documents on standards for facilities that treat, store
and dispose of hazardous waste. The documents were designed to
assist permit writers in evaluating facilities against standards
(40 Code of Federal Regulations, Part 264) issued under Subtitle C
of the Resource Conservation and Recovery Act (RCRA) of 1976, as
amended. Included in this report is information concerning the
design, inspection, common treatment processes, closure, and costs
of hazardous waste tanks, A checklist of questions and bibliogra-
phies of additional information sources are also included for the
p erm it writer.
G-l

-------
REFERENCE: Fred C. Hart Associates, Inc. "Assessment of Hazardous Waste Mis-
management Damage Case Histories", (Draft). . U.S. Environmental
Protection Agency, December 1982, Approximately 355 pp.
ABSTRACT: This report provides technical support to the U.S. Environmental
Protection Agency (EPA) for promulgating. Implementing, and
revising hazardous waste disposal regulations. it contains a
compilation of damage case histories associated with mismanaged
land and non-land based hazardous waste facilities. These case
histories are the results of the first phase of a two-phase
study.
Damage Incident Summary Forms (DISFs) were completed for 929 sites
across the country as documented In Field Investigation Team
(FIT), Surveillance and Analysis (S&A) files at each of the 10 EPA
Regions. Each completed DISF Identified each site by name, loca-
tion and facility type, media exposed to contamination, the extent
and severity of damage, the event(s) and wastes causing the Inci-
dent, the status of remedial activities and information sources
used. Value judgments were made for many DISF questions.
If one of the 929 sites possessed more than one facility type, a
DISF was completed for each type (e.g., landfill, surface Impound-
ment, land treatment, storage/treatment containers, storage/treat-
ment tanks, and other categories). Of the 1,722 facility types
tabulated, 197 or 11 percent were storage/treatment tanks.
Approximately 70 percent of the tanks recorded in this study were
aboveground facilities. Underground tanks evaluated were presuma-
bly constructed without liners or protective coatings. Tank capa-
cities ranged from 500 to 200,000 gallons on sites typically con-
taining multiple tanks and other facility types.
Associates. "Failure Incident Analyses: Evaluation of Stor-
Fallure Points", (Draft), U.S. Environmental Protection Agen-
Washlngton, D.C., March 1982, 69 pp.
REFERENCE: JRB
age
cy,
ABSTRACT: JRB Associates prepared this report as part of a larger study of
hazardous waste storage and storage-related Issues. This report
attempts to Identify and quantify risks associated with hazardous
waste storage facilities. To overcome the present lack of com-
plete data concerning accidental releases of hazardous waste, JRB
selected two methods for performing the analysis. The first meth-
od Involved the analysis of two data bases set up as requirements
of the Clean Water Act:
*• The Spill Prevention Control and Countermeasure (SPCC) data
base containing all spills affecting Inland waters as re-
ported to the U.S. Environmental Protection Agency (EPA);
and
• The Poflutlon Incident Reporting System (PIRS) with spills
affecting navigable waters as reported to the U.S. Coast
Guard .
The second method chosen for the failure Incident analysis In-
volved the use of. a fault-tree; I.e., a probabilistic logic net-
work that portrays the credible accident sequences by which haz-
ardous wastes could be released. Since both methods had several
I Imitations associated witn their use In this study, certain as-
sumptions had to be made.
The analysis of the SPCC and PIRS data bases revealed two key
points. First, the vast majority of spills occur due to
G-2

-------
operations or failure In ancillary equipment. Second, the size
distribution of spills Is relatively Independent of cause, with
spills over 5,000 gallons accounting for 2 to 13 percent of spills
for a given cause (I.e., failure due to containment device,
operations, ancillary equipment, or other causes of spills).
The fault-tree analysis showed that the most likely cause for re-
lease of stored hazardous waste Is the "loss from tank (and ancil-
lary piping Inside the diked area)" and "dike does not retain
s p I I 111 sequences of events. Tank overflows and manhole leaks were
the chief contributors of losses from the tank and ancillary pip-
ing.
REFERENCE: The MITRE Corporation. "Tanks and Containers for Hazardous
Wastes: Evaluation of Standards", (Draft). McLean, Virginia, No-
vember 1981, 126 pp.
ABSTRACT: This working paper drafted for the U.S. Environmental Protection
Agency (EPA) reviews and evaluates existing design and operational
standards of non-EPA organizations for containers and tanks to
hold hazardous substances. The focus of this study was to
Identify areas not covered In current EPA regulations and Items
suggested for potential Inclusion In EPA tank/ container
regulations.	The main sources of Information were 37
professional, trade, and Industrial standards-setting
organizations, five State environmental agencies, and four non-EPA
Federal agencies.
One of the recommendations pertaining to the EPA regulations for
hazardous waste vessels Is to allow storage of Ignltable wastes In
underground tanks, (OSHA regulations allow storage of flammables
In underground tanks.) However, MITRE recommends that the under-
ground storage of toxics and possibly corrosives be prohibited be-
cause of waste escape hazards and leak detection difficulties.
State or Local
California Department of Health Services. "Criteria for the Sit-
ing of Treatment Technologies for Hazardous Waste Management",
(Draft). Sacramento, California, November 1982, 43 pp.
This report was developed within the California Department of
Health Services as part of the Southern California Hazardous Waste
Management Project. The criteria were Intended for use by facili-
ty planners as recommended guidelines In the waste management or
generating Industries and State or local government agencies. Ty-
pical characteristics of hazardous waste treatment technologies
are provided for five types of facilities other than landfills.
Including waste transfer and storage facilities.
California Department of Health Services. "Siting Criteria for
Hazardous Waste Treatment Facility". Sacramento, California, Oc-
tober 1 981,7pp.
This document was developed within the California Department of
Health Services as part of the Southern California Hazardous Waste
Management Project. It Is a collection of concerns regarding the
siting of hazardous waste treatment facilities. Storage tanks are
not specifically discussed.
REFERENCE:
ABSTRACT:
REFERENCE:
ABSTRACT:
G-3

-------
REFERENCE: California Department of Health Services. "Variances From
Hazardous Waste Facility Permit Requirements" (Memo to Operators
of Hazardous Waste Facilities). Sacramento, California, 1982,
6 pp.
ABSTRACT: This memo from the Hazardous Waste Management Branch of the
California Department of Health Services to operators of hazardous
waste facilities outlines the procedures required for a variance
request. The Department of Health Services has the authority to
grant variances under the California Administrative Code as long
as such action will not result In a hazard to public health and
safety or to the environment. Operators of hazardous waste
facilities must submit a variance application and supporting
documentation (Attachments A and B) to the appropriate regional
office. Facilities with underground tanks must attach Information
(Attachment B) on a proposed ground water monitoring program to be
considered for a variance request. The monitoring program must
meet the requirements specified In the appllcalton.
REFERENCE: California Legislature. "1983-84 Regular Session, Assembly Bill
No. 1362". Introduced by Assemblyman Sher, March_2, 1983, 16 pp.
ABSTRACT: Assemblyman Sher Introduced a bill to regulate the underground
storage of hazardous substances. The bill requires that all tanks
Installed after June 30, 1984 comply with certain design, con-
struction, monitoring system and drainage requirements; tanks In-
stalled prior to this date have a monitoring system Installed, a
means of Inspection, and a permit before January 1, 1985 and be
upgraded to comply with new criteria by January 1', 1994. Each
county Is to be responsible for Implementation of the program to
handle a list of hazardous substances developed by the State De-
partment of Health Services. In addition, each local agency will
be responsible for Inspecting tanks at least once every 3 years.
Permit fees are to be collected to cover the costs of administer-
ing the program,
A list of a few of the key requirements proposed In the regulation
Include:
•	For Tanks Installed After June 30. 1984:
Provision for primary and secondary containment;
Installation of a monitoring system to detect leaks Into
the secondary containment structure;
Provision for overfilling protection either through a
prevention device or an alarm or both; and
Storage of different substances that cause fire, poison-
ous gas, and/or deteriorate primary or secondary contain-
ment when Intermixed In separate primary and secondary
containment structures.
•	For Tanks Installed Prior to June 30, 1984:
Installation of a monitoring system to detect leaks on or
before January 1, 1985;
Maintenance of a monitoring and recordkeeping program as
specified by the local agency;
Provision for visual Inspection of tanks whenever, prac-
tical;
Adherance to spill reporting and clean up as specified in
the bill;
G-4

-------
Adherance to permanent or temporary closure practices as
specified in the bill; and
Provisions for upgrading tanks before January 1, 1994.
REFERENCE: California Regional Water Quality Control Board. "Mandatory Fa-
cility Questionnaire". Oakland, California, 1982, 14 pp.
ABSTRACT: The California Regional Water Quality Control Board decided to im-
plement a program starting in March 1982, to determine the overall
magnitude of numerous subsurface leaks from sumps and subsurface
tanks in the San Francisco Bay Region. This mandatory question-
naire was sent to approximately 1,400 facilities thought to con-
tain potential sources of leaks to the area of concern. The fa-
cilities were asked to answer by May 31, 1982 questions describing
any existing or former sumps, subsurface tanks or subsurface pip-
ing. The responses will be used to determine which facilities
will be required to Implement a leak detection program.
REFERENCE: California Regional Water Quality Control Board. "Water Quality
Control Plan, San Francisco Bay Basin", (Toxic Waste and Hazardous
Waste Section). Oakland, California, July 1982, pp. 4.23-4.24
ABSTRACT: This excerpt from the Water Quality Control Plan for the San Fran-
cisco Bay Basin provides general background information concerning
the program initiated by the California Regional Water Quality
Control Board to deal with the problem of hazardous material sub-
surface leakage In the Santa Clara Vallay, Niles Cone, and Liver-
more-Amador Valley ground water basins. The Regional Board is
currently developing a policy for minimum underground fuel storage
management practices.
REFERENCE: California Regional Water Quality Board. "205(j) Proposal Assess-
ment of Contamination from Leaks of Hazardous Material in the San-
ta Clara Valley Ground Water Basin". Oakland, California, 1982, 6
PP.
ABSTRACT: The California Regional Water Quality Control Board proposed nu-
merous site Investigations in "the Santa Clara Valley ground water
basin to determine the extent of hazardous material contamination
from underground tanks. This proposal outlines the task require-
ments, including time estimates and a tentative budget to perform
the analysis within Fiscal Year 1983-1984, The proposal study
team includes the Regional Board and Santa Clara Valley Water Dis-
trict staffs and a private consultant.
REFERENCE: Cape Cod Planning and Economic Development Commision. (S. Horsley)
Barnstable, Massachusetts. Personal communications with SCS
Engineers, May 1983.
ABSTRACT: The Cape Cod Planning and Economic Development Commission (CCPEDC)
developed two model ordinances which directly address the ground
water contamination problem. These ordinances were designed for
enactment by the area towns since the county government has no
regulatory authority in Massachusetts. As of September, 1982, 14
G-5

-------
of the 15 Cape Cod communities had adopted at least one of the
three types of local protection measures, as follows:
•	Zoning bylaws to protect aquifer recharge areas;
•	Health regulations to prevent leaking of underground fuel
and chemical storage tanks; and
•	Regulations to control the storage, use and disposal of
toxic and hazardous materials.
The two model ground water ordinances are called "Model General
Bylaw/Regulation to Control Toxic and Hazardous Materials"
(December, 1981) and "Model Health Regulation to Prevent Leaking
of Underground Fuel and Chemical Storage Systems" (revised
February, 1982). The latter ordinance has three major provisions
which are registration of underground storage tanks, inventory
control and leak testing, and regulation of new tank
InstaIlatIons .
The first provision of the underground storage tank regulation
requires that all tanks in excess of a certain size must be
registered. Information required for tank registration Includes
size, type, age, location, and material stored. The second
provision requires daily inventory recordkeeping and annual leak
testing of tanks which are 15 years or older. Non-conforming
steel tanks must also be removed and inspected after 20 years.
The third provision of the model ordinance dictates the Instal-
lation of new tanks which must have an approved design and be pro-
tected from Internal and external corrosion. The placement of
tanks Is also regulated with respect to proximity of water
supp I I es .
After two years of Implementing the underground tank regulation,
dozens of leaking underground tanks were discovered. Many of these
tanks were situated within ground water recharge areas. All of
the leaking tanks were removed and replaced. Local officials have
been able to effectively administer the ordinance.
REFERENCE: F.G. Bercha and Associates Limited. "Bulk Plant Risk Opti-
mization". Department of the Environment, Environment Protection
Service, Hull, Quebec, Canada, December 1982, 232 pp.
ABSTRACT: This report analyzes risk-cost optimization of oil bulk storage
plants using conventional fault tree techniques. Seven major
types of accidents are examined, as follows:
•
Tank
overf low;

•
Tank
Leakage;

•
Ta n k
rupture;

•
An c i 1
1ary eq uIpment
leak;
•
Anc I 1
1ary eq u ipment
spill;
•
Fi re
or explosion;
and
•
Other
•

Of the seven major types of accidents, the primary cause of spills
is due to operator error during tank filling. Failures due to
engineering design errors are the least likely to occur. Repair,
service, and inspection are each essential in preventing releases,
f rom tanks.
6-6

-------
Recommendations for optimizing bulk plant operations are as
follows:
•	Workers should take periodic refresher courses to review
operating procedures and Introduce new Industrial
InnovatIons;
•	A high level alarm should be Installed on every tank to
reduce the number of overflow Incidents;
•	Specialized equipment should be Installed If cost effective;
and
•	Though engineering design errors lead to the fewest spills
of petroleum, small Improvements In design may reduce the
primary cause of spills. I.e., operator errors.
REFERENCE: Michigan Department of Natural Resources. "Study on the Under-
ground Storage of Gasoline". Water Quality Division, Lansing,
Michigan, September 1981, 153 pp.
ABSTRACT: This study looks at the problem of ground water contamination
from the underground storage of gasoline and how It relates to the
state of Michigan. The report summarizes known Incidents of pol-
lution from underground storage of petroleum products In Michigan
and provides general Information on the practice of and problems
caused by underground storage.
The study gives recommendations for underground storage In
existing tanks and new tank? and for the testing of tanks and
piping. Highlights of the recommendations Include:
•	State laws should provide the authority for spill coordi-
nators to order tank testing;
•	Work should concentrate on " prevention of ground water con-
tamination rather than depending solely on cleanup measures;
•	Preventive measures should Include education of the public
concerning proper underground storage techniques and the
potential dangers from a leak;
•	Leak detection methods should include the placement of
ground water monitoring wells near all storage tanks;
•	All tanks should be registered and equipped with overfill
p rotectIon;
•	Dally Inventory of tanks should be mandatory and spills and
leaks reported Immediately;
•	All newly Installed tanks should be tested before use and
constructed of either fiberglass reinforced plastic, epoxy
coated steel with cathodlc protection, or a similar system
approved by the state;
•	All existing tanks must be replaced or destroyed If they do
not conform with the approved standards; and
•	Tank testing should be performed with the Kent-Moore Tight-
ness Tester or an equivalent method approved by the state.
G-7

-------
New York State Department of Environmental Conservation (NYDEC).
"Bulk Storage of Hazardous Liquids, Five Proposed Regulatory Con-
cepts". Albany, New York, November 1980, 63 pp.
This document contains summaries of five regulatory concepts for
bulk storage of hazardous liquids as proposed by the NYDEC. These
proposals, if Implemented, would require owners/operators of haz-
ardous liquid storage facilities to:
•	Maintain stock Inventory control records;
•	Post storage plans prepared by a qualified professional en-
g i neer;
•	Register their storage facilities;
•	Submit facility plans, specifications and an application for
a certificate to operate a bulk storage system; or
•	Obtain permits and certificates to operate storage facili-
ties.
New York State Department of Environmental Conservation (NYDEC).
"Bulk Storage of Hazardous Liquids Study Program, Paper No. 5,
Problem Assessment Report", (Draft). Albany, New York, April
1981 , 83.pp.
As part of New York States' Bulk Storage Study Program, NYDEC re-
viewed spill case reports in New York, pertinent literature sourc-
es and Information from NYDEC Regional Oil Spill Engineers and
other personal communications. The results contained in this re-
port describe problems and issues associated with storing, handl-
ing and preventing leaks of hazardous, liquids. The report also
character Izes the type and number of hazardous liquids spills with
any associated environmental damage and gives case histories of
petroleum spills that occurred in each of the nine regions design-
ed by NYDEC.
Findings pertaining specifically to underground tanks include:
•	Approximately 20 percent or 16,600 of the estimated 83,000
functioning underground tanks in New York State leak;
•	To replace or rehabilitate the 16,600 leaky underground
tanks- wou Id cost around $90 million initially and $14 mil-
lion annually thereafter;
•	Underground steel tanks have an extremely variable (between
5 and 45 years) life depending on several factors and an
average life expectancy of 15 years; and
•	About 28,000 underground tanks have been abandoned over the
past 10 years and many were left with 6 inches of product
which will eventually leak.
REFERENCE: New York State Department of Environmental Conservation (NYDEC).
"Bulk Storage of Hazardous Liquids, Paper No. 6, Leak Prevention
Programs of Other States and Localities", (Draft). Albany, New
York, April 1981, 168 pp.
ABSTRACT: As part of New York States' Bulk Storage Study Program, NYDEC re-
searched a number of out-of-state programs to prevent leaks and
REFERENCE :
ABSTRACT:
REFERENCE :
ABSTRACT:
G-8

-------
spills of hazardous liquids. New York officials requested Infor-
mation on manpower Intensity,, overall success, and degree of com-
pliance or cooperation by Industry concerning the prevention pro-
grams. This report contains summaries of leak and spill preven-
tion programs for the States of Massachusetts, Maryland, and Penn-
sylvania; the Province of Manitoba, Canada; and Prince George's
County, Maryland. Names, addresses, and telephone numbers of pro-
gram contacts who can provide additional Information are Included
Immediately following each program discussion.
Highlights of each leak and spill prevention program Investigated
by New York State and related specifically to underground storage
tanks are as follows:
• Massachusetts:
Outlaws unprotected underground steel tanks and pipes ex-
cept where tests prove soils are non-corrosive;
Requires a permit from local officials and possibly an
Inspection for storage of more than 165 gallons of gaso-
line or for Installation, removal or relocation of an un-
derground gasoline storage tank; and
Requires that an accurate dally stock Inventory control
record be maintained by the operator of each underground
storage facility.
Mary land;
Requires a permit for abo.veground and burled oil storage,
of 10,000 gallons or more; and
Requires that gasoline station owners maintain dally
stock Inventory control records.
Prince Seorqe's County, Maryland;
In addition to requirements for dally stock Inventory
control, requires that all tanks, except fiberglass
tanks, which have been burled for 10 or more years be
tested every 5 years.
PennsyI van 1 a:
Requires a permit for all facilities that store hazardous
liquids but has no rigorous requirements for tank test-
ing, leak monitoring or tank replacement.
Province of Manitoba. Canada:
Requires that operators of an underground storage tank
maintain dally stock Inventory control records and report
any losses above normal;
Requires one time tightness tests In critical areas, on
new systems and on rehabilitated systems; and
Requires the removal of an abandoned storage system after
1 year of d Isuse.
REFERENCE: New York State Department of Environmental Conservation (NYDEC).
"Number and Distribution of Sulk Storage Tanks In New York State",
(Draft and Addendum). Albany, New York, August 1980 (Draft) and
April 6, 1981 (Addendum), 22 pp.


G-9

-------
ABSTRACT: This report supports the two-year program conducted by the NYDEC
to Identify and Implement state-of-the-art technology and regula-
tory controls to prevent leaks of hazardous liquids. Ballpark es-
timates of the numbers and locations of hazardous liquid storage
facilities In New York that may be affected by any proposed
regulations are provided. Facilities are broken down Into four
categories:
•	Gasoline stations;
•	Major petroleum facilities;
•	Home heating oil distributors; and
•	Industries storing other hazardous liquids.
Included are the different methodologies used to approximate the
number of each facility type.
REFERENCE: New York State Department of Environmental Conservation (NYDEC).
"Siting Manual for Storing Hazardous Substances: A Practical
Guide for Local Officials". Albany, New York, October 1982, 98
PP .
ABSTRACT: This Is one of a series of manuals to support New York States'
Bulk Storage Program. This manual provides guidance to local of-
ficials who need help In making prudent decisions for the siting
of bulk storage facilities for hazardous substances. Included
topics are:
•	Types of hazards;
•	Causes of leaks and spills;
•	Site evaIuat Ion procedures ;
•	'Risk assessment methods; and
•	Practices for spill prevention and mitigation.
Precautionary designs and practices are provided for different
types of storage facilities. Including both above and underground
tanks. A precautionary storage design (a drawing showing simple
and effective designs to reduce the risk and liability In case of
an accident) Is Illustrated for a pre-engIneered underground stor-
age tank.
O'Brien and Gere Engineers, Inc., staff from NYDEC, Bureau of Wa-
ter Resources, and a review committee comprised of planning repre-
sentatives contributed to this manual. Fred C. Hart Associates,
Inc. provided precautionary storage design drawings.
REFERENCE: New York State Department of Environmental Conservation (NYDEC).
"Technology for the Storage of Hazardous Liquids, A State-of-the-
Art Review". Albany, New York, January 1983, 223 pp.
ABSTRACT: This Is one of a series of manuals to support New York States'
Bulk Storage Program. The report provides timely Information to
Industry and government officials who must face problems concern-
ing the storage of hazardous liquids. This manual also encourages
the use of the best technology and practices for preventing spills
and leaks.
Three parts make up the report. Part I presents background Infor-
mation associated with underground or aboveground storage of haz-
ardous liquids. Parts II and III address the state-of-the-art for
underground and aboveground storage systems, respectively. At the
end of each chapter within each part, references are provided for
further Information.
G-10

-------
Part II describes the components and concerns associated with the
storage of hazardous liquids In underground facilities. Seven
chapters detail underground storage systems, Including:
•	The types of storage tanks available;
•	Piping and pumping system components and their performance;
•	Underground spill containment systems;
•	The types of overfill prevention systems and their perform-
ance;
•	Leak monitoring and surveillance;
•	The testing and Inspection of underground storage systems;
and
•	The closure and abandonment of underground storage facili-
ties.
Fred C. Hart Associates, Inc., staff from NYDEC, Bureau of Water
Resources, and a review committee comprised of Industrial repre-
sentatives, a State engineer and a local health official contri-
buted to this manual.
REFERENCE: Santa Clara County City Managers' Association. "Petroleum Product
Review Committee: Notes from Meeting held In the Sunnyvale City
Council Chamber". Sunnyvale, California, February 17, 1983, 8 pp.
ABSTRACT: Notes from the meeting of the Petroleum Product Review Committee
of the Santa Clara County City Managers' Association held on Feb-
ruary 17, 1983, outline the purpose of the association and topics
of discussion for upcoming meetings. The purpose of the group Is
to review Information on:
•	Monitoring systems;
•	Single- and double-walled containment systems; and
•	Available alternatives for storage of petroleum products.
One of the topics of concern Is scheduled to be discussed at one
of three above meetings In March 1983. The end result will be a
factual presentation of current systems, proposed systems, alter-
natives, costs and any other related Information to the Santa
Clara County Intergovernmental Council.
Attendees at the meeting included representatives from the 24 pe-
troleum and petroleum-related Industries, a municipal water quali-
ty control plant, a citizens activist group, and a local fire de-
partment .
Attached to the meeting notes Is a blank survey form on existing
equipment testing and replacement programs for use by the Commit-
tee.
REFERENCE: Santa Clara County Hazardous Materials Model Code
"Model Hazardous Materials Storage Permit Ordinance".
County, California, February 3, 1983, 45 pp.
Task Force.
Santa Clara
ABSTRACT: The Santa Clara County Hazardous Materials Model Code Task Force
drafted this model permit for local entitles enacting the hazard-
ous materials storage permit ordinance. The permit Is composed of
G—11

-------
14 parts, including containment standards, a hazardous materials
management plan and inventory, inspections and records, etc.
REFERENCE: Suffolk County Department of Health Services. (J. Pim),
Farm i ng v i I I e, New York. Personal communications with SCS Engi-
neers, March 1983.
ABSTRACT: The Suffolk County Board of Health in New York enacted Article 12
of the Suffolk County Sanitary Code, "Toxic and Hazardous
Materials Storage and Handling Control", effective January 1,
1 980. The regulation was developed as a result of the Increasing
number of wells contaminated from leaks and spills of hazardous
materials, especially fuels, in Suffolk County. This region was
one of the first areas In the nation to be designated as a sole
source aquifer by the U.S. Environmental Protection Agency, Suffolk
County was also the first to develop comprehensive regulations
concerning the storage of toxic and hazardous materials in the
nation.
Many countries in Europe (Germany, France, Switzerland and other
Scandinavian countries) have developed regulations to prevent
spills and leaks from tanks, Europeans have been concerned about
the use of hazardous and toxic materials in ground water recharge
areas longer than Americans. They have already designated recharge
protection zones and Installed over 70,000 underground double-
walled tanks.
Article 12 was developed around the concept of double-walled
containment to provide maximum protection of hazardous material
stored underground, aboveground, in portable containers, or at
transfer facilities. By the time the law was passed, however,
there were two exceptions. First, small heating oil tanks were
exempted because of the dificultles In administration. Second,
single-walled 'tanks were allowed for underground storage as long
as they met certain conditions. Some of the key requirements of
Article 12 concerning underground storage include:
•	All new storage facilities constructed on or after Novem-
ber 1 , 1982, must be dou b I e-wa I I ed or some approved
equivalent for use with all non-floatable toxic or hazardous
materials. For use with floatable materials, acceptable
designs are cathodlcally protected steel, glass fibre
reinforced plastic, steel clad with glass fibre reinforced
plastic, doubIe-wa I I ed steel or plastic or some approved
equivalent;
•	All existing storage facilities constructed before Novem-
ber 1, 1982, must comply with all the provisions for new
storage facilities by January 1, 1987, for use with all non-
floatable toxic or hazardous materials and by Jan-
uary ?, 1995, for use with all floatable materials;
•	All existing storage facilities which do not comply with all
the provisions for new storage facilities must be tested and
i nspected; and
•	Accurate records must be kept of all deliveries and
consumption and the figures reconci led da I ly.
After Article 12 had been In effect for three years, testing of
over 1,000 underground tanks resulted in the discovery of 98
leaks. More than 2,000 new or replaced tanks were inspected and
over 900 tanks were removed or abandoned in these first three
years of enactment. Consent orders signed by 158 violators led to
the collection of over $83,000 in fines.
G-12

-------
REFERENCE:
Texas Department of Health Services, Office of Solid Waste (K.
Schoenfelt), Austin, Texas. Personal communication with SCS Engi-
neers, March 1983
ABSTRACT: Texas Department of Health Services has hard copy files of facili-
ties which store hazardous wastes, but no readily accessible in-
ventory on aboveground or underground tanks. The Department has a
checklist for aboveground tank leak assessment, but none for
underground tanks.
Industry or Trade Association
REFERENCE: American Petroleum Institute. "API Industry Recommended Practice
for the Prevention and Detection of Leaks from Underground Tanks
and Piping". Washington, D.C., May 6, 1980, 6 pp.
ABSTRACT: This document prescribes the requirements for the detection and
prevention of leaks of flammable or combustible liquids from un-
derground tank and piping systems. The practice excludes:
•	Storage tanks with capacities under 2,000 gallons which are
located on farms or isolated construction projects; and
•	Fuel oil tanks or containers connected with burning equip-
ment.
The prescribed requirements cover five areas of concern. High-
lights from each area include these requirements:
•	I nventory ControI:
f
An accurate dally Inventory; and
Prompt reporting to the authority having jurisdiction of
abnormal losses,
•	Tank Selection and Installation:
Use of tanks constructed of non-corrosive materials In
corrosive areas or at sites where no corrosion tests have
been conducted;
Placement of at least 12 inches of non-corrosive Inert
material around steel underground tanks; and
Replacement or interior-coating of all underground steel
tanks at a facility which are the same age or older if a
corrosion Induced leak occurs.
•	Piping:
Use of pipes constructed of non-corrosive materials in
corrosive areas or at sites where no corrosion tests have
been conducted; and
Placement of at least 6 Inches of non-corrosive inert ma-
terial around all underground piping.
•	Pumping S ystems :
Installation of a product line leak detector for all new
remote pumping systems;
G-13

-------
Installation of a listed rigidly anchored emergency shut-
off va I ue; and
Placement of at least 6 Inches of non-corrosive Inert ma-
terial around all underground piping.
•	Pump Ing Systems;
Installation of a product line leak detector for all new
remote pumping systems;
Installation of a listed rigidly anchored emergency shut-
off vaIue; and
Discontinuation of a pumping system until corrective ac-
tion for a leak Is completed.
•	TestInq:
Hydrostatic	or pneumatic testing of all piping before be-
ing placed	In use to 150 percent or 100 percent of the
anticipated	pressure of the system, respectively;
Hydrostatic tightness or pneumatic testing of all new un-
derground tanks at not less than 3 pounds per square inch
and not more than 5 pounds per square Inch after Instal-
lation but before being placed In use; and
Use of an on-going preventative maintenance program for
systems cathodlcally protected.
REFERENCE: American Petroleum Institute. "Results of API Tank and Piping
Leak Survey, February 5, 1981 Memorandum and Updated Statistical
Data". Washington, D.C., 1981, 22 pp.
ABSTRACT: The American Petroleum Institute (API) conducted a nationwide
voluntary survey of tank an'd piping leaks from approximately the
fall of 1 977 to the summer of 1980. The February 5, 1981 memo-
randum to the members of the Operations and Engineering Committee
and Underground Leakage Task Force summarizes the results of 1,717
completed survey questionnaire forms. The updated statistical
results complied In about June of 1981 Include an additional 236
reports or a total of 1,953 questionnaires from API member
companies and tank and pump contractors.
The API survey results provide the number of tank and piping leaks
by state and by category of tank construction material (steel or
fiberglass) and tank protection (sacrificial anodes. Impressed
current cathodlc protection, or Interior coated steel). Detailed
statistical data are also provided for three categories of leaks.
I.e., piping, fiberglass tank, and steel tank leaks. For each
category of leaks, a breakdown Is given for the number of
responses, the causes of leaks, the type of backfill material, the
age of the piping or tank, and how the leak was detected. In
addition, Information Is provided on the disposition of a leaking
tank, the product stored, the tank size and location, the type of
corrosion, and the location of leak points.
Of the 1,953 survey responses, the most reported leaks occurred In
California, Pennsylvania, and Virginia. Corrosion caused the most
leaks in underground equipment.	Proper Inventory control
procedures detected the majority of leaks.	Other survey
highlights Include the reporting of no leaks caused by corrosion
or dissolving of fiberglass tanks. Almost one half of all leaking
steel tanks are Interior coated Instead of being replaced or
6-14

-------
abandoned. About 80 percent of the leaking tanks had a capacity of
4,000 gallons or less. The peak age at which leaks occurred in
tanks was 20 years.
REFERENCE: Dames and Moore. "Subsurface Investigation to Evaluate the Inte-
grity of Subsurface Tanks at Van Waters and Rogers' Facility in
San Jose, California", (Letter). San Francisco, California, Feb-
ruary 2, 1983, 13 pp.
ABSTRACT: This letter from Dames and Moore summarizes their investigation to
evaluate the integrity of subsurface tanks containing solvents at
the Van Waters and Rogers' facility in San Jose, California. The
California Regional Water Quality Control Board in Oakland re-
quired Van Waters and Rogers to perform the investigation as part
of the State program to determine the overall magnitude of sub-
surface leaks from sumps and subsurface tanks In the San Francisco
Bay Region. Included are soil sampling results and chemical anal-
yses of soil and ground water at three monitoring wells. Solvents
were detected in the sol I and ground water at the site.
REFERENCE: Exxon Corporation. (G. Gartyser), Houston, Texas. Personal commun-
ications with SCS Engineers, May 1983.
ABSTRACT: In a detailed study of tank corrosion, Exxon found that the most
important factors Influencing the rate of corrosion are tank age
and soil corrosIveness. The soli corrosion index which was
developed during the study is fully described in a proprietary
report entitled "Underground Leak Study" (MERP 7103). This Index
provides the basis of ^xxon's Tankage Upgrading Program started in
January 1980, and expected to be completed in 1984 . The obj'ectives
of the Tankage Upgrading Program are to:
*	Establish the criteria for selecting the appropriate
upgrading action; and
•	Implement a company wide program for protection, repair, or
replacement of all underground steel tanks.
The program encompasses four possible treatment categories at each
facility as follows:
Category	Average Cost
A - Interior lining	$29,000 - 31,000
B - Cathodic protection	$ 3,500 - 5,000
C - Tank tightness testing	$ 2,000 - 3,000
D - Fiberglass replacement	$55,000 - 77,000
Details of the guidelines developed for	the Tankage Upgrading
Program are described in an Exxon document written in December
1981, and revised in July 1981. However,	the revised document Is
proprietary information.
REFERENCE: IBM. "Corporate Facilities Practice 1401 A: Containment of
Industrial Liquids". May 1982, 16 pp.
G-15

-------
ABSTRACT: This document describes the containment criteria which apply to
all newly constructed or replaced IBM facilities which store,
transport, treat or otherwise handle Industrial liquids. The cri-
teria refer to primary containment, i.e., tanks, pipes, and drums,
and secondary containment. The containment criteria is based on
the type of Industrial liquids handled:
• Group I - Liquids which have no environmental hazard
potential require monitoring but not secondary
conta i nment;
•	Group II - Liquids which have a moderate environmental
hazard potential require secondary containment
consisting of a single layer of chemical/ physi-
cal resistant coating, liner or equivalent; and
•	Group III - Liquids which have a high environmental hazard
potential require secondary containment consis-
ting of a double layer for bulk liquid storage
and a single layer for other liquid handling
f acI I 11 i es.
Underground siting of systems is used only when necessitated by
safety or fire protection codes, the liquid properties, or con-
struction constraints. Guidelines are presented for two types of
underground secondary containment systems (concrete and tank jac-
ket).
REFERENCE: IBM Corporation (R. B. Jabblonskl), Tarrytown, New York. Personal
communication with SCS Engineers, March 1983.
ABSTRACT: IBM was going to conduct an Internal survey of all underground
tank storage sites to evaluate current practices, identify prob-
lems and solicits field suggestions for improvements. This survey
has not been undertaken due to resource limitations and shifts in
priorities.
REFERENCE: Motorola. (J. Hinchey and N. Hild), Phoenix, Arizona. Personal
communications with SCS Engineers, May 1983.
ABSTRACT: Motorola has recently had problems with underground storage tanks
leaking solvents. As a result, Motorola developed guidelines for
storage of hazardous materials (products . and wastes) in tanks. The
specifications vary depending on the type of tank and location of
the facility. The primary features of the guidelines developed by
Motoro la include:
•	All tanks shall be located aboveground or underground in
level cement vaults with steel cradles to support the tanks;
•	Where production lines depend on receiving materials from
tanks, double-walled tanks shall be used to prevent an Inter-
ruption in production In case the inner tank wall fails; and
•	Currently the guidelines are in effect for only the semi-
conductor division, but they have been proposed for use by
all Motorola divisions.
G-16

-------
REFERENCE: Petroleum Association for Conservation of the Canadian Environment
(PACE). "Bulk Plant Guidelines for Oil Spill Prevention and Con-
trol". PACE Report No. 80-3, Ottawa, Ontario, Canada, September
1980, 63 pp.
ABSTRACT: This report provides design guidelines for oil spill prevention
and control at bulk plants and terminals. The guidelines detail a
step-by-step procedure for containment, collection, conveyance and
treatment to cover the majority of potential oil spill situations
at bulk plants or terminals. At certain locations, however, some
modifications may be required.
A Task Force of the Product Storage and HandlIng Committee at PACE
prepared the report. Members on the Task Force Included six oil
company representatives.
REFERENCE: Petroleum Association for Conservation of the Canadian Environment
(PACE). "Report on Investigations and Research to Develop a Ser-
vice Station Underground Tank Leak Detector". PACE Report No. 81-
3, Ottawa, Ontario, Canada, October 1981, 36 pp.
ABSTRACT: This report records the Investigation and research Involved by
B. C. Research of Vancouver, British Columbia In developing the
PALD-2 Underground Tank Leak Detector. Findings from numerous
field tests lead to the conclusion that It Is Impossible to design
an apparatus to measure small leak rates (0.05 gallons per hour)
In a 15 to 30 minute test. External factors such as the nature of
soil mechanics, random ground motion and expansion of trapped air
greatly affect the accuracy of leak rate measurement. B. C. Re-
search prepared the report for PACE.
REFERENCE: Petroleum Association for Conservation of the Canadian Environment
(PACE). "Guideline Specification for the Impressed Current Method
of Cathodlc Protection of Underground Service Station Tankage".
PACE Report No. 79-4, Ottawa, Ontario, Canada, June 1979, 23 pp.
ABSTRACT: This report provides a guideline specification for cathodlc pro-
tection of underground service stat Ion tanks. The guideline spe-
cification details requirements for the design, materials, Instal-
lation, Inspection and commissioning, and maintenance of cathodlc
protection and Is Intended for use by any PACE member company.
Corrosion Service Company Limited, corrosion engineering special-
ists In Toronto, Ontario, prepared the report for PACE.
REFERENCE: Petroleum Association for Conservation of the Canadian Environment
(PACE). "Proceedings Underground Tank Testing Symposium". Park
Plaza Hotel, Toronto, Ontario, Canada, May 25-26, 1982, 257 pp.
ABSTRACT: Eight technical papers were presented at the Underground Tank
Testing Symposium sponsored by PACE from May 25-26, 1982 at the
Park Plaza Hotel In Toronto, Ontario, Canada. Three Canadian and
five United States promotors of different tank testing systems
spoke at the two-day symposium.
G-17

-------
Topics covered include:
•	The PALD-2 Underground Tank Leak Detector developed by B. C.
Reserch and the behavior of underground tanks;
•	The accuracy of finding small leaks In underground gasoline
storage tanks using a method tested by SRI International In
CaI If or n I a;
•	A subatmospheric pressure test for detecting leaks In under-
ground hydrocarbon storage tanks developed by Athabasca Re-
search Corporation;
•	A system Invented by Joseph Mooney, PE, to determine the
condition of an underground petroleum product storage tank;
•	The Sun Leak Lokator patented by Sun Refining and Marketing
Company to test underground tanks;
•	Petro-Tlte Tank and Line Testing Equipment (formerly known
as the Kent-Moore system) manufactured by Heath Consultants,
Inc.;
•	"Ethyl" Tank Sentry underground tank leak detector developed
by Texaco and licensed by Ethyl; and
•	On-going developmental work being conducted by Shell Canada
Limited to Innovatively test underground tanks.
REFERENCE: Warren Rogers Associates. Report on the Statistical Analysis of
Corrosion Failures In Unprotected Underground Steel Tanks.
American Petroleum Institute, Washington, D.C. 1982. 76 pp.
excluding Appendix E.
ABSTRACT: This statistical analysis of external and internal corrosion
failures In unprotected underground steel storage tanks was
performed for the American Petroleum Institute. The purpose of
the analysis was to determine if the age at which tank failure
occurs Is related to measurable characteristics of the tank
environment. The study concludes that', although unprotected
underground steel tanks should have a trouble-free lifetime of
over 20 years, unforeseen processes initiated during Installation
can greatly reduce the useful life of a tank. Localized corrosion
may occur in one tank or all tanks Installed at a given site, so
the findings of this study are applicable to a site and not
particular tanks at a site. A small subset of data collected at
approximately 10,000 sites throughout the U.S. and Canada was used
to estimate tank age failure.
About three-quarters (77 percent) of the sites had tanks
experiencing localized external corrosion. The remaining one-
quarter (23 percent) of the sites had tanks which were corroded
uniformly and thus were not corrosion failure problems. Failure
was observed in tanks ranging from as low as 5 to as high as 45
years of age.
REFERENCE: Warren Rogers Associates (W. Rogers), Newport, Rhode Island.
Personal communications with SCS Engineers, May through July 1983.
ABSTRACT: Warren Roger Associates developed a computerized data base on
tanks used in the petroleum industry. Information on tank age Is
G-18

-------
maintained for approximately 46,000 sites. Data concerning tank
leaks is included for about 18,000 facilities. Statistics on leak
cause and volume of release show that 6 percent of the leaks are
due to internal corrosion and that 85 percent of all leaks are
confined on-site, 10 percent migrate off-site but near the
immediate vicinity, and the remaining 5 percent penetrate a large
area. Cleanup costs average $20,000 for on-site leaks, $150,000
for nearby off-site contamination, and over $1 million for
widespread leaks. Some of the data is not available because of
their confidentiality.
Warren Rogers also discussed several leak detection methods. He
feels that both the Kent-Moore and Sunmark Leak Lokator methods
are good, but that the leak locator Is better because It Is
usually administered • by competent crews. Warren Rogers found that
the Kent-Moore method is 95 percent reliable with a good operator
and only 10 to 15 percent reliable otherwise. The laser beam leak
detection method is expensive and difficult to administer,
Cathodic protection of tanks works well In preventing leaks If
Installed and maintained correctly.
Warren Rogers estimates that between 50,000 to 75,000 leaks go
undetected In the United States. The major ol I companies are
concerned about the independent companies who do not often have
the resources or incentive to test or replace tanks. The large
oil companies favor mandatory recordkeeping of inventories.
Another concern is	the proposed methanol additives which react
with the resins used	in FRP tanks currently in use. Warren Rogers
suggests that either	a federal national fund to handle leaks or an
insurance program be	developed.
G-19

-------
APPENDIX H
LEAK TESTING METHODS

-------
APPENDIX H
LEAK TESTING METHODS FOR UNDERGROUND STORAGE TANKS
INTRODUCTION
Approach
The following approach was taken to determine the state-of-
the-art in leak testing methods for underground storage tanks:
t review literature;
•	review manufacturers' and contractors' brochure
i nformati on;
•	telephone conversations with testing contractors; and
•	telephone conversations with test method developers.
Leak test methods are becoming more and more sophisticated.
The early stand pipe test method, which requires only a standpipe
and a measuring tape, is being replaced by complicated measure-
ment devices and microcomputers for analyzing the measurement
data. Test.methods are now being developed which use lasers to
measure small liquid level variations, and hydrophones to detect
the sound of bubbles ingressing through holes in a tank shell.
The state-of-the-art in leak testing methods is changing
rapidly. The major oil companies and other concerned bodies are
striving for more accurate, more reliable, and quicker test
methods. The leak test methods discussed below are those which
have been field tested in the United States and Canada, in
contrast with those which are still in the developmental stage.
Background
Most of the underground storage tank facility leak testing
methods have been developed primarily for detecting and measuring
leaks in underground gasoline storage facilities. Current infor-
mation on testing methods for storage tanks containing non-
petroleum products, such as hazardous wastes, is limited. In
some applications the leak test methods used for petroleum
storage tanks can be and have been applied to tanks storing
liquids other than petroleum products. Material considerations,
the availability of excess stored product, or the stability of
the stored product may preclude the application of certain test
methods. Table H-l gives an overview of the leak testing methods
discussed in this chapter. Emphasis is given to assessing the
applicability of the various leak test methods to underground
storage tanks and piping systems used for hazardous waste
storage.
The American Petroleum Institute (API), the Petroleum Assoc-
iation for Conservation of the Canadian Environment (PACE), and
H-l

-------
Type of Test
Oescr Ipt Ion
ApplIcabl1Ity
Accuracy
Remarks
PneumatIc teak test
Air or other gas Is used
to pressurize the sys-
tem, A drop in pressure
is indicative of a leak
Underground tanks and
piping systems
Pneumatic tests are often
Inconclusive
Air pressure tests are
not recommended for
tanks and piping sys-
tems containing flamma-
ble or combustible
waste* With air test-
ing, there Is also a
serious danger of rup-
turing the tank
HydrostatIc (stand-
pipe) leak test
Water (or another li-
quid) is used to pres-
surize system, A drop
in liquid level Is indi-
cative of a leak
Underground tanks and
piping systems
HydrostatIc tests are
more sensitive than pneu-
mat ic tests
This procedure Is use-
ful where It Is desired
to check the tightness
of any underground
storage tank and its
connected piping for
gross leaks. Does not
compensate for thermal
exports Ion or contrac-
t ion of the stored
waste.
Petro-Tlte (former-
ly Kent*44oore
test)
Accurate type of hydro-
static test
Underground tanks and
piping systems
0.05 gaiIons/hour
Test Is approved by the
National Fire Protec-
tion Association
(NFPA). Requires well-
trained operator. Re-
quires several hours
for completion of ac-
curate test
Ethyl Tank Sentry
Leak Detector
Manometer-type Instru-
ment that detects leaks
by measuring smaJ 1
changes In product level
Underground tanks
Oetects change in liquid
level as small as 0*02
Inches* Accuracy depends
upon the time period over
which the level charge Is
observed (leak of 0.02
Inches over 1 hour Is
larger than leak of 0*02
inches over 10 hours In
same tank)
Easy to transport, as-
semble and operate.
Does not require a con-
tractor crew to oper-
ate* Several tanks can
be tested simultaneous-
ly* Tank, piping and
dispenser openings need
not be sea 1 ed
*Sunmark leak test
System operates on the
principle of hydrostatic
head and uses an analy-
tical balance to measure
small changes In Mould
mass displacement
Underground tanks and
piping systems
0*03 gal Ions/hour
The time for the equip-
ment to be set up and
the test to be complet-
ed Is at least 2 hours*
Compensates for temper-
ature and pressure
Laser-1nterterome-
try
An experimental device
for detecting leaks*
Operates on the princi-
ple of laser Interfero-
metry
Underground tanks
Threshold of detection
has not yet been estab-
1(shed
API has specified that
It wants the device to
be able to detect leaks
as small as 0*05 gal-
lons/hour Instantane-
ously
ARCO HTC teak test
Systems use a float and
light-sensing system to
detect volume changes
Underground storage
tanks and their distri-
bution lines
Less than 0.05 gallons/
hour
System works on tank 75
percent full* It does
not detect leaks in the
upper 25 percent of the
tank or in the fill
1 Irte
TABLE H-l. COMPARISON OF VARIOUS TANK LEAK TEST METHODS
H-2

-------
Type of Test
DescrIpt lor
ApplIcabl1Ity
Accuracy
Remarks
Vacu tect
leak test
Leak test mpthod
based on creating bub-
bles at leak ingress
point rfhlch produce dis-
tinct and detectable
sounds
Underground tanks
Detects presence of leak,
not rate, and relative
location of leak
Limited test data to
date (200 tests by
1982)* Not effected by
pressure temperature or
tank configuration
changes. Subject to
problems similar to
pneumatic test
Smith and Denison
leak test
Helium Is used to pres-
surize tanks and piping
system. Mass spectro-
meter used to monitor
for leaks
Underground tanks and
piping systems
Oetects presence of 'leak,
not.rate, and relative
location of leak
Pressurized testing
system. See preumat Ic
test method remarks.
Tank must be empty for
tank test, can be par-
tlealiy fui1 for pipe
test
Key:
I
• - Method being used in unspecified underground storage tanks for commercial clients (stored material unspecified),
method being used In underground solvent storage tanks
- See text for explanations
^ - See text for references
TAB1E H-I. COMPARISON OF VARIOUS TANK LEAK. TEST METHODS, CONT.
H-
3

-------
the major oil companies have played a major role in the
development of most of the test methods now on the market. These
organizations are continuing their efforts to develop more
reliable, cost effective, and quicker methods for determining
leaks in underground storage tanks and piping systems.
Measurement Techniques and Criteria
There are numerous approaches presently marketed for deter-
mining the presence of leaks in piping systems and underground
tanks. Some of the leak test methods measure volume changes with
time to determine leakage rates, while others test only for the
presence and location of leaks.
The National Fire Protection Association (NFPA) has published
standards and recommended practices for testing piping systems
and underground storage tanks containing flammable and combust-
ible liquids for leaks. NFPA 329 "Recommended Practice for
Handling Underground Leakage of Flammable and Combustible Liq-
uids" is generally accepted as one of the most authoritative
documents on the subject of leak test methods. The following
excerpt is from this publication:
"The Final Test will conclusively determine whether or not an
underground liquid storage and handling system is leaking.
Any testing devices used for the Final Test shall be capable
of detecting leaks as small as 0.05 gal in one hour, adjusted
for variables, a limiting criterion widely accepted by most
authorities."[2]
The Petro-Tite test method (formerly the Kent-Moore Test) is .
believed to be the Final Test referred to in NFPA, although any
test method meeting the NFPA criteria would be acceptable. The
0.05 gal/h criterion has been established based on the most
accurate of the reliable test methods. Some people question
whether any test method can reliably measure to 0.05 gal/h given
all the variables which affect tank testing results. Regardless,
the 0.05 gal/h limit is a reference point.
Other leak test method approaches should also be considered.
For example, two test methods discussed in this chapter, the
Vacutect and Smith & Denison leak test methods, merely detect the
presence and general location of leaks rather than measuring the
leakage rate. These methods are based on a philosophy that no
leak, small or large, is acceptable. These methods, though
mainly used for testing tanks and piping systems containing
petroleum products, may be well suited for application to
hazardous waste underground storage tanks and piping systems,
where the primary concern is identifying the presence of a leak.
This is not to discount the importance of determining the rate of
a leak. Such information, is important in estimating the volume
of leakage and the impact the leak may have on the environment.
Economic considerations will determine whether the test method
developers will invest the monies necessary to adapt their
H-4

-------
systems to non-petroleum storage tank and piping systems (e.g.
hazardous wastes).
API's and PACE'S research efforts have established tentative
performance objectives for new leak test methodologies and
equipment, key elements of which are summarized below [3],[4].
•	sensitivity limits of 0.05 gal/h;
•	fully automatic operation;
t capable of identifying location as well as rate of leak;
•	test duration: 15 to 30 minutes (not including set up);
•	simple to operate;
•	intrinsically safe;
t	easy maintenance through replaceable parts.
Some of these performance objectives may be unrealistic given
the degree of accuracy desired. For example, the equipment and
instrumentation required to measure leaks at 0.05 gal/h within a
15 to 30 minute period will be costly and skilled labor will be
required to insure the reliability of the test results. Also, the
statistical reliability of any measured test resul.t diminishes
with decreasing measurement time. Most of the leak testing
methods currently available take anywhere from 30 minutes to a
day to complete. API's and PACE'S performance specifications
represent goals to achieve, but at the present time no leak
testing method meets them all.
Speci al Considerations i n Testi ng Underground Storage Tanks
According to NFPA 329, the "Final Test" method used to
determine if a tank is leaking must compensate for the affects
of temperature variations on the volume of the stored product and
tank deformation due to pressure surcharges resulting from the
test procedure. Temperature is an important variable to compen-
sate for because of the significant affect it has on the volume
of the liquid stored. Gasoline, for instance, has a coefficient
of expansion of 0.0006 per degree F. A 1 degree F change in
temperature in a 10,000 gal tank will result in a net change in
volume of 6 gal A 0.01 degree F change in temperature will
result in a 0.06 gal change. If the 0.01 degree change occurs
over a one hour testing period, a 0.05 gal/h leak would go
undetected without temperature adjustments to the liquid volume.
In tests conducted by Shell Canada Ltd., it was found that
the maximum temperature stabilization time was 4 hours [5].
Other researchers contend that temperature stabilization may not
occur for hours or even days after the last product delivery [6].
Internal product temperature variations are impacted by the
temperature of the stored and delivered product (and the resul-
tant mixing) and the ground temperature (which varies throughout
the year and leads to temperature stratification in the tank).
This complexity is addressed by most of the newer leak test
methods.

-------
Another important variable that must be considered in volume-
based measurement techniques is tank deformation under surcharge
conditions. Underground storage tanks deform considerably under
surcharge conditions. A surcharge condition occurs when the
pressure applied to a tank is greater than the tank's normal
operating or design pressure. Most underground storage tanks are
designed as low-pressure or atmospheric tanks.
A hydrostatic standpipe type test, which increases the liquid
level above the top of the tank during the test,places signifi-
cant surcharge pressures and resultant forces on the ends of a
tank. Under these surcharge pressures, the ends of a tank will
bulge out, resulting in an apparent loss of product. Figure H-l
and H-2 are presented to demonstrate the significance and
magnitude of this phenomenon.
The deformation may occur immediately, and thus pose no
problem in the volume measurements, or it may take hours or days
to stabilize. Tank deformation can be compensated for by either
of the following methods:
•	slightly surcharging the tank prior to testing and then
relaxing the pressure during the testing; or
•	through tables and graphs relating the extent of bulging
anticipated as a function of the height of storged liquid;
or
> , *
•	testing the tank under normal operating conditions (this
approach assumes that tank deformation has stabilized over
time).
Other variables which impact the accuracy of most test
methods include product density, product expansion coefficient,
trapped air in the tank, and ground vibrations. Each of these
variables impact the testing methods in different ways, depending
on the measurement technique and equipment employed. The vari-
ables of product density and product expansion coefficient can be
easily determined in the field with a hydrometer and other
chemical tables. The presence of air pockets in a full tank test
can render the results of a test totally invalid. Air pockets
can compress or expand in volume significantly, depending on
temperature and the applied pressures. This compressibility
characteristic can set up a spring type action in the fluid which
can be reflected in an increase or decrease in the fluid level in
the tank. This variation could be interpreted as a loss or gain
of product if the frequency of the level variation is greater
than the testing period [8].
LEAK TESTING METHODS
Petro-Tite Test (Formerly Kent-Moore Test)
H-6

-------
Formula: Force Area X Pressure (Ibs./Sq. In.)
Total Force in Tons at:
Tank Dia.
1 Psi.
2 Psi.
3 Psi.
4 Psi.
5 Psi.
48"
0.9
1.8
2.7
3.6
4.5
64"
1.6
3.2
4.8
6.4
8.0
72"
2.0
4.0
6.0
8.0
10.0
84"
2.8
5.6
8.4
11.2
14.0
96"
3.6
7.2
10.8
14.4
18.0
Figure H-l. Total Force on Tank Ends [7]
H-7

-------
FORMULA FOR COMPUTING VOLUME CHANCE DUE TO
TANK HEAD DEFLECTION
VT = [y (fJ ~ -y ) hD(2)
Vy = Totol Volume in Cubic Inches
r s	Rodigs of Tank in Inches
h =	Deflection of Tonk Ends in Inches
Gallon r	231 in.'
Gallons =	(VT) ( j)
Zjl if*.
EXAMPLE COMPUTATION:
Tank Diameter • d 96"
Tonk Radius • r 48"
Hood Deflection • h .125"
1.	VT =-1Q48)j+ (^-)j3(.125) X =904.78 in.J
2.	(904.78	-.) = 3 92 Gallons increase in tonk
231	,	*
copocity due to tonk
head deflection
Figure H-2
Formula for Computing Volume Change Due
to Tank Head Deflection
H-8

-------
The Petro-Tite leak test method is a hydrostatic test capable
of detecting leaks in both storage tanks and connecting piping.
The test adjusts for temperature, pressure, and viscosity varia-
tions. The test was originally developed by F. Ronald McLean of
the Mobil Oil Corporation. The Petro-Tite test method has a
reported accuracy of 0.05 gal/h. The principle application of
the test method has been on underground gasoline storage tanks
and piping systems.
The Petro-Tite test requires specialized equipment, including
a circulation pump, a thermister (a temperature sensing devise
accurate to 1/6Oth degree F), a hydrometer, a standpipe, and a
graduated cylinder (See Figure H-3) [9]. The test can be
conducted by one trained person. Eight hours are generally
required to complete a test on one underground storage tank. The
cost,for testing one tank (including equipment and labor) ranges
between $500 to $600 (assuming no leaks are detected). Discount
rates generally apply for multiple tank installations. [10]. If
a leak is detected during a full system test (piping and tanks),
another test must be run to determine whether the tank or the
piping is the source of the leak. This additional testing will
increase the fee above the $500 to $600 range. Labor costs for a
skilled testing operator can range between $300 and $400 per day.
The Petro-Tite system has the following advantages and
disadvantages:
•	advantages:
available throughout the country
accuracy to 0.05 gal/hr
temperature affects accounted for;
tank deformation accounted for;
tests both tank and piping systems;
detects leaks throughout tank depth.
•	disadvantages:
full tank required and extra product must be available
requires specially trained personnel;
equipment is expensive;
all product transfers must be halted during
testing;
affects of trapped air not accounted for; and
duration of test
The Petro-Tite leak test method has been used primarily on
underground gasoline storage tanks. It appears that the test
method could be used to test tanks containing hazardous wastes as
long as the stored product was compatible with the testing
equipment and extra product was available to raise the liquid
level above the top of the tank. The requirement of additional
product may limit the extent to which this testing method can be
used to test hazardous waste underground storage tanks for leaks.
u _ n

-------
Figure H-3. Petro-Tite Test Schematic Diagram [9]
H-10

-------
Also, handling or disturbing the stored waste may not be
desireable for many types of hazardous wastes stored in under-
ground tanks.
St and Pipe Test Method
The Stand Pipe test method is capable of detecting the
presense of gross leaks in underground storage tanks and connec-
ting piping systems [11]. The detected leaks must be signifi-
cantly greater than the volume variations due to tank deflection
and temperature change or they may not be detected. The hydro-
static Stand Pipe method has an unspecified accuracy.
The Stand Pipe test method does not require any specialized
equipment or personnel. The test procedure can be conducted by
the owner or operator of an underground storage tank system. A
standpipe, a measuring device (i.e. measuring tape), and extra
product are the only items required for the test. Special
precaution should be taken to insure that pumps in siphon sytems
are taken out of service, and that manifolded vent lines (in the
case of multiple tank installations) are disconnected.
The Stand Pipe test method has the following advantages and
disadvantages:
•	advantages:
available throughout the	country;
detects leaks throughout	tank depth;
no specialized equipment	or personnel required; and
inexpensive to run.
•	disadvantages:
full tank required and extra product must be on end;
does not account for volume changes due to product
temperature changes;
does not account for volume changes due to tank
def1ect i on;
only applicable to tanks with gross leaks; and
not recommended as a final test method by NFPA.
As stated above, the Stand Pipe leak test method is only
useful for detecting large leaks. The .accuracy of the test
results are questionable, since temper ature changes and tank
deformation are not accounted for. As with the Petro-Tite test
method, the Stand Pipe test method requires filling the tank
above the top of the tank. As such, sufficient product must be
available to attain this liquid level. The Stand Pipe test
method could be used on underground storage tanks containing
hazardous wastes to detect gross leaks, but it will not meet the
final test performance standards suggested by NFPA.
H-ll

-------
Ai r Test Method (Pneumat i c)
The Air Test method is a pneumatic pressure testing procedure
for testing both tanks and connecting piping systems. In gene-
ral, only leaks above the liquid level in storage tanks can be
detected by this method. NFPA 329 states that pressure tests
with air shall not be' used on tanks storing flammable and
combustible liquids [12]. The inherent risks of tank and pipe
connection failures due to increased pressure during the test are
well documented concerns [13].
Air tests should only be performed by qualified personnel.
Application of the test should be restricted to new tanks which
have not been installed or filled with liquid. According to
NFPA, air testing is more suitable for testing pipelines than
underground storage tanks [12].
The accuracy of air test results is questionable. The
pressure loss in a tank is dependent on the volume of air in the
tank or piping system. For example, the volume of air in a half-
full tank is such that the pressure loss due to a small leak
could go undetected. The air test method also does not compen-
sate for pressure changes due to temperature variations and tank
deflections.
The Air Test method is not recommended for testing of
hazardous waste underground storage tanks due to the inherent
safety problems associated with the test.
Ethyl Tank Sentry Method (J-Tube M anometer Test)
The Ethyl Tank Sentry leak test method was developed by the
Ethyl Corporation and marketed by Texaco Inc. for leak testing
underground gasoline storage tanks. The test procedure is able
to detect small changes in product level with an indicator fluid
which magnifies tank level changes. The devise used to measure
liquid level changes consists of a J-tube manometer attached to a
3-inch pipe reservoir. The basic principals and equipment used
in the testing procedure are shown in Figure H-4. The density of
the indicator fluid must be such that it will not mix with the
product in the tank. This factor limits the extent to which this
method can be used for testing storage tanks containing liquids
other than gasoline.
It is reported that the test method is accurate to 0.02 gal/h
[15]. The test results are valid only if the temperature
differential during the testing period is less than 1 degree F.
Product temperature in underground storage tanks usually stabi-
lizes within 24 hours; therefore, a 24 hour waiting period
following the last product delivery is suggested before commen-
cing the testing procedure. Since the tank is tested at normal
operating conditions and generally 24 hours after the last
H-12

-------
/WATER TRAP
APRON	/FILL PIPE
INSTALLATION SCHEMATIC
FUSL LEVEL
IN TANK —*
OPEN
ai
CLOSED
u
CLOSED
tin
OPEN
•INSTALLATION M TANK
•U
¦ON TEST
OPEN
CLOSED
INITIAL FUEL LEVEL


FINAL FUEL LEVEL


REMOVAL FROM TANK
MAGNIFICATION
Figure H-4. Ethyl Tank Sentry Leak Detector Schematic Diagram [14]
H-13

-------
product deliver, measurement errors due to tank end deflections
and temperature changes are considered insignificant.
The Ethyl Tank Sentry leak testing method should be conducted
by personnel skilled in gasoline equipment maintenance. Two to
three days training is required in order to properly use the test
equipment. Only one man is needed to conduct the test. Set up
requires approximately 30 minutes per tank. A 12 to 24 hour test
period is recommended. During the testing period, the testing
equipment can be left unattended. The method is ideal for over
night testing. This is convenient to many service station
operators who do not want to be shut down during normal operating
hours.
The Ethyl Tank Sentry test equipment costs approximately
$5,000; therefore, it may not be economical for an owner or
operator of a storage tank to own his own testing devise.
Testing costs range between $300 to $400 per tank and $600 for a
three tank installation [16].
The Ethyl Tank Sentry method has the following advantages and
disadvantages:
•	advantages:
-	available throughout the country;
accurate to 0.02 gal/h;
-	easy to .set up;
temperature changes monitored;
test performed under normal operating
conditions;
small level changes easily measured; and
can be performed overnight unattended.
•	di sadvantages:
specially trained operator required;
testing results invalid if temperature
differential greater than 1 degree F;
will not detect leaks above product level;
test duration; and
piping systems cannot be tested along with tanks.
The Ethyl Tank Sentry leak test method is mainly applicable
to testing underground gasoline storage tanks. The Ethyl Corp.
has tried to use the method on other petroleum distillates with
limited success [16]. The indicator fluid is the key to the
test. The specific gravity of the indicator fluid must be
strictly controlled and be slightly greater than the specific
gravity of the stored product to prevent mixing.
It can be expected that the physical characteristics of
hazardous wastes will vary greatly from facility to facility, in
contrast with gasoline which has well defined physical charac-
H-14

-------
teristics. As such, the Ethyl Tank Sentry leak test method has
limited application in the leak testing of underground hazardous
waste storage tanks.
Sunmark Leak Test Method
The Sunmark leak test method, the Leak Lokator, was developed
by the Sunmark Corporation, which is a subsidary of the Sun Oil
Corporation. The test can be used to detect leaks in underground
storage tanks and associated piping. The Leak Lokator measures
mass displacement in a tank via bouyancy changes in a calibrated
apparatus. Any bouyancy change is an indication of a leak.
The principal equipment components used in the test include
an open top-hollow tube sensor filled with the stored product, a
mass balance, a strip chart recorder, a thermister, and a
hydrometer (See Figure H-5) [16],[17]. Test personnel measure
the specific gravity of the stored product prior to beginning the
test. The calculated density is used to relate the mass dis-
placement to volume displacement. The sensor is partially sus-
pended in the stored tank from the analytical balance. As the
volume in the tank changes, the bouyancy of the sensor changes.
The resulting mass displacement is recorded by the balance, and a
permanent record of the test results is produced on the chart
recorder. Temperature is monitored throughout the testing period
at mid-tank level. Product mixing is discouraged. The test
assumes that tank deformation is insignificant, since the test
procedure does not surcharge the tank.
The test can be conducted on any tank which has a fillport 2-
inches or larger. Tank configuration or volume do not affect the
accuracy of the test. The test method was developed for testing
underground gasoline storage tanks; however, the test is present-
ly being applied to underground tanks storing products other than
gasoline (e.g. solvents) [19].
The accuracy of the method depends on the product level in
the tank. The higher the product level, the more accurate the
test results. Sunmark recommends conducting the test under full
tank conditions. At this level, test results are reported as
accurate as 0.03 gal/h [18].
The Sunmark leak test method requires the use of a skilled
operator and specialized equipment. To set up and run the
Sunmark test requires approximately 2 hours per tank. Actual
testing, once all the equipment is in place, takes only 15 to 20
minutes. A typical system test, which might include 4 under-
ground tanks and associated piping, would cost approximately
$500. If a tank is tested alone, the cost is approximately $300
[19]. Sunmark has a different pricing schedule for commercial
clients which they were not free to disclose.
The Sunmark leak test method has the following advantages and
H-15

-------
ANALYTICAL BALANCE





F=H
#•


—=£-



=*=
••
0©


-f—



••


—





RECORDER
PUMP
PIPELINE
r—
"N
-i
7
Figure H-5. Sunmark Leak Test Method Schematic Diagram [17]
H-16

-------
d i sadvantages :
•	advantages:
test results not dependent on tank configuration
or volume;
test both tank and piping systems;
detects leaks throughout tank depth;
accurate to 0.03 gal/h;
compensates for temperature changes;
application to tanks storing various products;
permanent record of test results; and
short set-up and testing time.
•	disadvantages
requires specially trained personnel;
currently only available on the east coast, Texas, and
California (Sunmark is expanding their testing fleet);
tank should be full for most reliable
results;
qualified personnel required to operate equipment; and
equipment is expensive (approximately $52,000 which
includes truck) [21].
Sunmark claims to be using its testing method to test various
underground chemical (e.g. sol vents) .storage tanks for commercial
clients. The sensor in contact with the stored product4 is made
of aluminum, which is resistant to many corrosive liquids.
Sunmark states that they need to know the nature of the stored
liquid prior to testing to insure material compatibility. Alumi-
num is not compatible with alkalies, potassium hydroxide, sodium
hydroxide, and mineral acids [20]. The Leak Lokator appears"
promising with regard to testing underground storage tanks
containing a wide range of hazardous wastes.
Laser Interferometry Test Method
The Laser Interferometry test method was developed by SRI
International under contract with API [22]. It is not so much a
test method as it is a very precise liquid level measuring
devise. The measurement technique was developed to demonstrate
that very small liquid level changes (e.g. 1 micron) could be
detected very quickly through the use of laser interferometry.
The method is. presently not marketed by any firm or used by any
commercial testing contractors.
The measurement technique developed by SRI uses a low-
powered, double-tube laser devise (See Figure H-6). One of the
tubes is closed to the liquid in the tank (simulating a no leak
condition), while the other tube is open to the stored liquid.
The laser beam is reflected off the liquid surface back to a
detecting device, which computes the travel time of the laser
H-17

-------
BEAM SPLITTER
LASER
LASER
BEAM
INTERFEROMETER A
r~f
INTERFEROMETER B
^GROUND
OPTICAL GLASS-
GASOLINE SURFACE
FLOATS
CUBE-CORNER —
REFLECTORS
GASOLINE
NEEDLE VALVE
SB
S

isa
TWO BRASS
TUBES
Figure H-6. Laser Interferometry Test Method Schematic Diagram [22]
H-18

-------
beam. Variations in travel time relate to variations in liquid
levels in the storage tank. Using this device, changes in liquid
levels can be detected almost instantaneously.
The measurement results must be statistically evaluated to
determine whether a detected level change is due to a leak or to
some other background noise. SRI has completed extensive statis-
tical studies on their test results and concluded that the laser
interferometry device could produce results meeting the 0.05
gal/h criteria over a 2 hour test period. The results from a one
hour test period were not considered statistically reliable [22].
The SRI laser measuring device has been primarily used to
detect liquid level changes in underground gasoline storage
tanks. According to the SRI, the device could be used on
virtually any liquid stored in underground tanks, including
hazardous wastes.
ARCO HTC Leak Test Method
The ARCO HTC leak test method was developed by ARCO, Inc. and
has been used exclusively by ARCO in testing their underground
gasoline storage tanks [24]. The test is applicable to tanks
storing liquids of a known density. The reported accuracy of the
method is 0.02 gal/h. Liquid levels must be between 66 to 75
percent of the tank depth to assure reliable test results
[24],[25].
The equipment used in the test method include a specially
fabricated float mechanism, a photo cell, and a voltage meter.
The photo cell and float apparatus are inserted into the tank at
the beginning of the test. The fillport must be at least 3-
inches in diameter. The float is placed at a pre-specified level
in the tank, where temperature changes will not impact the float
level.
A one hour waiting period is recommended to allow for the
temperature of the equipment to stabilize. With changing liquid
levels, the float raises or lowers in the tank. The float
movement forces an ink type solution into or out of the photo
cell. The change in light transmittance in the photo cell
results in a voltage drop across the cell. The voltage change,
which is a function of the liquid level change in the tank, is
measured by a voltage meter. The voltage meter is calibrated
prior to testing with a known quantity of product. The leak test
is performed for one hour. The meter is then calibrated a second
time and another one hour test is run. Testing continues until
two consecutive readings correspond (insures reliability of test
results, i.e. steady state conditions).
The ARCO HTC test method has the following advantages and
disadvantages:
advantages
H-19

-------
accurate to 0.02 gal/h;
compensates for temperature variation and tank deforma-
tion; and
test conducted under normal operationg conditions.
di sadvantages:
equipment is expensive (approximately $4,000);
unable to detect leaks in upper 25 percent of a tank;
density of fluid must be known;
piping systems cannot be tested along with tanks;
requires specially trained personnel; and
used on only ARC0 tanks.
ARC0 has not pursued .other commercial applications for their
test method, although they indicate that the method could be used
for testing tanks containing non-petroleum products [24]. Mater-
ial compatibility would have to be considered when applying the
method to other products. The density of the float, which is
presently designed for gasoline, is also an important factor.
The float density would have to be adjusted for liquid densities
different than gasoline.
VacuTect Leak Test Method
The VacuTect leak test method was developed by the Anthabasca
Research Corp.,.Ltd, Edmonton, Ontario. During the test the tank
is placed under a negative pressure. A special hydrophone probe
installed in the tank monitors for the sound of ingressing air
bubbles. The test procedure is based on the observation that
bubble formation resulting from the ingress of air into a tank
under vacum conditions produces a distinct and detectable sound
(See Figure H-7)[27].
The method does not have.the drawbacks of conventional leak
testing methods such as temperature variation, tank deformation,
or product instability (i.e. vibration). The method does not
measure leak rates, but rather the presence of leaks. In
addition, the method can detect the location of a leak and the
presence of water in the tank. Water is detected with a device
attached to the hydrophone.
The method includes the following equipment:
•	specially fitted vehicle;
•	vacuum pump;
•instrumented probe;
•microprocessor;
•	hoses & fittings; and
•	other misc. special tools.
The equipment is presently designed for use on tanks with 4-
inch diameter fillports. However, Anthabasca is developing a-
H-20

-------
TESTING STEPS
1.	PROBE IS SUBMERGED
2.	VACUUM LINE IS APPLIED.
3.	PRESSURE WITHING THE TANK ULLAGE IS
INCREMENTALLY REDUCED EQUIVALENT TO
FLUID HEAD.
4.	BUBBLE SIGNATURES OF INGRESSING AIR
DETECTED BY HYDROPHONE.
5.	INGRESSING WATER LEVEL IS DETECTED
BY WATER SENSOR.
6.	DATA IS ACCUMULATED AND PRINTED BY
MEANS OF A MINICOMPUTER IN THE TANK
TESTING VEHICLE.
7.	TANK PRESSURE IS RESTORED TO ATMOSPHERIC
WITH THE INJECTION OF NITROGEN.
SUBMERGED PROBE	INGRESSING BUBBLES
(Probe includes hydrophone, pressure sensor, temperature sensor, and water sensor)
Figure H-7. VacuTect Test Method Schematic Diagram [25]
H-21

-------
dapters to fit 3-inch diameter fillports. This adaptation may
allow for testing commercial underground waste and product
storage tanks [28].
The VacuTect test method requires skilled personnel to
operate. Before the the test is conducted, the test personnel
records the operational history of the tank. Inventory records
are evaluated to determine if leaks can be detected from the
data. The vacuum pump and hoses are connected to the tank, and
the probe is lowered into the tank. Pertinent information is
entered into the microprocessor (i.e. tank location, stored
product, name of owner, etc.) and then the test is started. The
tank pressure is lowered step by step to the static head at the
bottom of the tank. In this way the leak can be located with
respect to depth in the tank. Testing costs range between $400
to $500 per tank[24].
The VacuTect test method has the following adavantages and
d i sadv ant ages:
•	advantages:
available throughout the country;
not affected by temperature changes;
tank deformation not a factor;
short testing time;
product transfer from tank can continue during test;
tanks and piping can be tested;
-* detect leaks throughout tank depth; and
full tank not required.
•	disadvantages:
leakage rate not measured;
sophisticated equipment required (costly); and
requires specially trained personnel.
The VacuTect Test Method may be adaptable to use on under-
ground storage tanks containing commercial chemicals and hazard-
ous wastes. At the present time, the VacuTect leak test method
has only been used to test petroleum product storage tanks.
Anthabasca has indicated that if. a substantial market for the
testing underground waste storage tanks emerges, then they may
invest the capital necessary to modify their equipment for use on
non-petroleum storage tanks (e.g. hazardous wastes) [29]. Chem-
ical storage tanks in general are not constructed like gasoline
storage tanks. Fillports are usually 2 inches and less as
compared to the standard 4 inch fillports on gasoline storage
tanks. Changes to their probe would be necessary to be lowered
through small diameter' fillports. Also, the viscosity of some
waste liquids may inhibit the formation of free bubbles. In this
case the VacuTect test method would not be suited for testing for
leaks. The market does not appear to justify such expenditures
H - 2 2

-------
at this time. Anthabasca feels that if a liquid can be stored in
a steiel tank or glass lined tank, their testing system will be
compat i b1e.
Smith & Denison Leak Test Method (Helium Testing)
The Smith & Denison leak test method uses helium to test for
leaks in underground storage tanks and piping systems. This
method does not determine leak rate; it merely indicates the
presence and approximate location of leaks.
The Smith & Denison leak test is usually conducted in in two
steps. The first step involves testing the piping system. The
tank can either be empty or partially full during this testing
stage (preferably partially full in order to isolate the leaks to
the piping system). The piping system is pressureized with
helium. A gas mass spectrometer and gas probe is used to measure
the concentrations of helium in the soild surrounding the piping
system. If the piping system and tank(s) are under pavement, a
grid of 1/2 inch holes are drilled through which a gas probe can
be lowered.
If helium is detected above background levels, a leak is
assumed. The location of the leak(s) can be identified by
further refining the hole grid system until the highest concen-
tration of helium is found. The piping system should be repaired
prior to testin the tank itself.
To test the tank requires placing it under approximately 5
psig pressure and sealing all the opening to the tank. An even
distribution of helium at the surface would indicate a leak near
the bottom of the tank, since the helium concentration will vary
inversely with distance from the leak source. Sharp concentra-
tion peaks would indicate leaks near the surface.
The test can take anywhere from 1 hour to 24 hours to
complete. Equipment needed for the Smith & Denison test include
a mass spectrometer, a gas probe, and a hydraulic jack to drill
through any pavements. Specialized personnel are required to
conduct the test. The mass spectrometer requires care in handl-
ing and set-up. Testing fees are approximately $500 per 10,000
gallon tank. [29], The test is available throughout the United
St ates.
The Smith & Denison system is not approved by NFPA. It is
subject to the same safety constraints as the air test method,
which tests tanks and piping systems under pressure. Smith &
Denison do not feel the pressure tests are unsafe, and as such,
feel their system is suitable for testing underground storage
tanks and piping systems, including hazardous waste storage tanks
and piping systems [29].
The Smith & Denison leak test method has the following
advantages and disadvantages:
H- 23

-------
•	advantages:
short testing duration;
location of leaks can be determined;
not affected by temperature changes;
tank deformation not a factor; and
available throughout the country.
•	disadvantages:
pressurized testing system;
tank must be completely empty for testing;
leakage rate not measured; and
requires specially trained personnel.
Pressurized helium pipe testing has been used for many years
to test natural gas pipelines. Application to storage tanks and
appurtenant piping systems has been limited. Where the Smith &
Denison system has been used on tanks and piping systems, the
tanks have been primarily gasoline storage tanks.
Since helium is an inert gas, compatibility with a hazardous
waste is not a concern. Smith & Denison feel that their system
could be used for testing hazardous waste storage tanks and
piping systems [29].
CONCLUSIONS
Table H-l, in addition to the discussion in the text, gives
an overview of the of the various underground storage tank and
piping system leak test methods. The major concern expressed by
the various test method developers is in reference to test
equipment material compatabi1ity with the stored product and the
tank type (i.e. large enough fillports). The equipment used to
test gasoline storage tanks may not be suitable for certain
aggressive hazardous waste environments.
The most promising methods for testing underground hazardous
waste storage tanks appear to be the following:
•	Sunmark Leak Test Method (detection limit of less than
0.05 gal/h);
« Petro-Tite Test Method (detection limit of 0.05 gal/h);
•	VacuTect Leak Test Method (detects presence of leak not
rate);
•	Smith & Denison Leak Test Method (determines presence of
leak not rate).
The Sunmark leak test method is the only leak test method
currently being used to test commercial chemical storage tanks
for leaks. It is currently available on the East Coast, Califor-
nia, and Texas. The other three are mentioned because of their
application potential. Testing fees for all the tests are
roughly the same, approximately $500 per tank.
H-24

-------
All but the VacuTect test method can test both tank and
piping systems for leaks. The VacuTect test method can only test
for leaks in tanks. The Sunmark and Petro-Tite test methods test
for leakage rates, while the VacuTect and Smith th Denison leak
test methods test for the presence of leaks (one via a vacuum and
the other via pressure). Although only four of the leak test
methods are recommended as most promising, this does not preclude
the possibility of the other test methods being used to test
hazardous waste underground storage tanks.
H-25

-------
REFERENCES
1.	Fred C. Hart Associates. Technology for the Storage of
Hazardous Liquids-State of the Art Review. New York State
Department of Environmental Conservation. January 1983. pp.
109-110.
2.	National Fire Protection Association. Recommended Practice
for Handling Underground Leakage of Flammable and Combust-
ible Liquids. Volume 12, Standard 329, 1980. p. 239-31.
3.	Werner Grundmann. PALD-2 Underground Tank Leak Detector and
Observation of the Behavior of Underground Tanks. In-
Proceeding from the Underground Tank Testing Symposium,
sponsored by the Petroleum Association for the Conservation
of the Canadian Environment. Toronto, Ontario. May 25-26,
1982. p.14.
4.	Communication with a representative from SRI International.
March 1983.
5.	J. Witherspoon. PACE Research Project, Underground Storage
Tank Leak Detector. In- Proceeding from the Underground
Tank Testing Symposium, sponsored by the Petroleum Associa-
tion for the Conservation of the Canadian Environment.
Toronto, Ontario. May 25-26, 192. p. 8.
6.	W. Grundmann, p. 14.
7.	Heath Consultants Inc.. Procedures Manual for the Operation
of the Petro-Tite Tank Tester. 1983.
8.	W. Grundmann, p. 3
9.	Heath Consultants, Inc.. Petro-Tite Tank Tester, Line Test-
er. Informational Brochure. 1983.
10.	Communication with a representative from Calco Inc. (a
Petro-Tite Test Contractor in Maryland). March 1983.
11.	NFPA 329, p. 329-29.
12.	NFPA 329, p. 329-22.
13.	F. Ronald McLean. Leak Seeking in Underground Tanks. Tran-
script from a speech made at the 43rd Annual Fire Department
Instructors Conference. Kansas City, Missouri. March 30,
Apri1 2, 1971.
14.	Ethyl Tank Sentry (Underground Tank Leak Detector). Product
Brochure. March 1983.
H - 2 6

-------
15.	E. P. Watson. Ethyl Tank Sentry (Underground Tank Leak
Detector). In- Proceedings from the Underground Tank Test-
ing Symposium, sponsored by the Petroleum Assoication for
the Conservation of the Canadian Environment. Toronto,
Ontario. May 25-26, 1982. p. 4.
16.	Communications with a representative from the Ethyl Corpora-
tion. Princeton, New Jersey. March 1983.
17.	Leak Lokator. Sun Refining and Marketying Co. Philadel-
phia, PA. Product Brochure. April 1983.
18.	Fred C. Hart Associates, p. 114.
19.	Communications with a representative from Sunmark Indus-
tries March 1983.
20.	Communications with Donna Hymes of Hunter Environmental
(Sunmark Industries) August 9, 1983.
21.	Fred C. Hart Associates, p. 36.
22.	Joseph W. Maresca, Jr.. Method of Determining the Accuracy
of Underground Gasoline Storage Tank Leak Detection Devices.
In- Proceeding from the Underground Tank Testing Symposium,
sponsored by the Petroleum Association for the Conservation
of the Canadian Environment. Toronto, Ontario. May 25-26,
1983. p. 1.
2 3. J. Maresca, p. 36.
24.	Communications with a representative from ARCO. Henry,
Illinois. M-arch 1983.
25.	Fred C. Hart Associates, p. 115.
26.	VacuTect. Tanknology, Edmonton, Alberta. Product Rrochure.
April 1983.
27.	T. Edwin Adams. Development of a Leak Detection System by
Anthabasca Research Corporation, LTD. In- Proceedings from
the Underground Tank Testing Symposium, sponsored by the
Petroleum Association for the Conservation of the Canadian
Environment. Totonto, Ontario. May 25-26, 1982. p. 1.
28.	Communications with a representative from the Anthabascas
Research Corporation, LTD. Edmonton, Alberta. March 1983.
29.	Communications with a representative from Anthabascas Re-
search Corporation, LTD., Edmonton, Alberta. August 9, 1983
H-27

-------
APPENDIX I
Procedures Used to Estimate the Number
Underground Tanks in New York State

-------
APPENDIX I
PROCEDURES USED TO ESTIMATE THE NUMBER OF
UNDERGROUND TANKS IN NEW YORK STATE
The number of underground tanks associated with four types
of hazardous liquid storage facilities (gasoline stations, major
petroleum facilities, home heating oil distributors, and indus-
tries storing other hazardous liquids) were estimated using the
methods presented below. [1]
o Gasoline Stations - Three methods were used to estimate
the total number of operating gasoline stations in the
State by county.
The first method used population distribution data and
the number of operating service stations in the State
(as estimated by the Gasoline Retailers Association of
Northeastern New York) to calculate the number of ser-
vice stations as a function of the percent of State
population represented in each county. For example,
assume that Dutchess County contains 10 percent of the
State's population and assume that the State has a
total of 10,000 service stations. The resulting num-
ber of service stations in Dutchess County would be
1,000 (.10 x 10,000 service stations).
- The second method uses the gasoline consumption rates
by county per year (as estimated by the New York State
Department of Energy (NYS DOE)), and the typical
annual volume (gallons) of sales for a typical station
to calculate the number of stations in each county as
a function of gasoline consumption. For example, if a
typical facility sells 500,000 gallons per year and
the total consumption of gasoline in Tioga County is
4,000,000 gallons per year then the number of service
stations in Tioga County would be 3 (4,000,000 divided
by 500,000) .
The third method was similar to method two except gas-
oline consumption was calculated by determining the
average number of motor vehicles per square mile in
each county.
The results of these methods turned out to be very sim-
ilar so an average of the three values was used to deter-
mine the number of stations per county. These numbers
were then totaled to arrive at the total number of ser-
vice stations in the State (17,475 which includes 10,589
operating and 6,886 closed or abandoned stations) which,
in turn, was multiplied by an assumed number of 4 tanks
per service station to arrive at the total number of
tanks in the State (approximately 68,900 tanks of which
27,544 are closed or abandoned).
1-1

-------
o Major Petroleum Facilities - The number of	major pet ro-
leum facilities in the State totals 390, as	evidenced by
the number of licenses issued by Department	of Transpor-
tation (NYS DOT) .
o Home Heating Oil - The number of home heating oil facili-
ties per county was estimated using NYS DOT's estimates
of the number of heating oil distributors in the State,
subtracting out the major petroleum facilities and dis-
tributing them proportionally by population.
o Industries Storing Other Hazardous Liquids - The number
of industries storing other hazardous liquids were esti-
mated using the DEC's report "An Iventory of Industrial
Hazardous Waste Generation in New York State" and judge-
ment to select the types of industries likely to be using
underground storage for their hazardous wastes.
The number of closed or abandoned facilities were estimated
using the methods described above and assumptions as to the num-
ber of facility closings in the past ten years (as provided by
the Gasoline Retailers Association of Northeastern New York, NYS
DOE and NYS DOT).
The estimate of the number of leaking tanks was determined
using judgement based on results of Prince George's County, Mary-
land, American Petroleum Association and Suffolk County, New York
tank testing programs. [2]
[1]	New York State Department of Environmental Conservation
(NYDEC). "Number and Distribution of Bulk Storage Tanks in
New York State" (Draft and Addendum). Albany, New York,
August 1980 (Draft) and April 6, 1981 (Addendum), 2?. pp.
[2]	New York State Department of Environmental Conservation
(P. Sausville), Albany, New York. Personal Communications
with SCS Engineers, June 1983.
1-2

-------
APPENDIX J
Underground Protected Steel Tank Study
Statistical Analysis of Corrosion Failures

-------
PREFACE
The following is a reprint of a paper prepared by Warren Rogers
Associates which discusses one type of statistical analysis used
to study corrosion failures in underground unprotected steel
tanks.

-------
UNDERGROUND UNPROTECTED STEEL TANK STUDY:
STATISTICAL ANALYSIS OF CORROSION FAILURES
STATISTICAL ANALYSIS
Under contract with the American Petroleum Institute, Warren Rogers
Associates has completed a statistical analysis of data on the occurrence
of corrosion failures in unprotected underground steel storage tanks.
The purpose of the analysis was to establish whether the ages at which
failure occurred were related to measurable characteristics of the tank
environment. The ages at which failures were observed varied from as
low as five years to as high as forty-five. It was determined that
measurements of soil resistivity, moisture, pH, sulphides and tank size
can be incorporated into a mathematical model which explains
approximately 7 5 percent of that variability. In addition, it was found
that the statistical properties of the remaining 25 percent unexplained
variability, are such, that knowledge of the residual error distribution
permits the determination of confidence limits for estimates obtained
from the model.
RESULTS OF ANALYSIS
The immediate practical consequences of the statistical analysis are
as follows:
1.	Tank Site Failure Probabilities -
«
With data, on the variables mentioned above, estimates can be
made of the probability of a failure occurring in a tank at a
specific age, or the probability that a tank may have already
developed a corrosion induced perforation.
2.	Decision Tree Modeling -
In addition to determining a tank failure probability, there are
other factors to be considered when making site-specific
decisions. 3y incorporating relevant costs, environmental risk
assessment, alternative courses of action and such factors as
the economics of continued station operation, the results of
this study provide a management tool to aid in determining a
course of action. For example, a fully automated decision tree
model can be developed to provide the short and long term cost
and benefits of the alternative courses of action.
3.	Leak Prevention Priority Setting -
It has been determined that locations can be prioritized on the
basis of greatest to least probability of corrosion failure
along with other relevant factors mentioned above. This
capability is particularly important in view of the practical
and physical constraints on the time to manage and execute a
tank testing or upgrading program at a great number of sites.
In the absence of site specific data and probability estimates
1

-------
based on Che data, it is difficult and in many cases impossible
to determine the probability of a corosion failure at specific
locations. Therefore, locations having the greatest likelihood
of failure may go untreated while locations having tanks with a
greater life expectancy rceive unwarranted tank testing or
upgrading. It should be emphasized that tank age is, by itself,
a very poor predictor of a corrosion induced failure. As
mentioned above, failures due to external or internal corrosion
have been observed at ages as low as five years and as high as
forty-five.
BACKGROUND
To place this subject in context and to understand the nature of the
statistical analysis and the conclusions that have been reached, it is
important to understand the unexpected nature of underground tank
corrosion failure and the novelty of the technical means required to
solve it. When unprotected underground steel tanks were installed,
engineering state-of-the-art indicated that a trouble-free lifetime of
over 20 years could be anticipated. Data gathered for the statistical
analysis study show that under certain circumstances that belief was
justified. However, the study's conclusions show that unforeseen
processes could be initated during installation, which could greatly
reduce the useful life of an underground tank. The use of improper
backfill material, an impurity in otherwise good backfill (a foreign
substance such as a cinder), physical damage which scraped away a tank's
coating or mill-scale, or any one of several other essentially random
events could serve to create a localized anode on the tank surface.
Whenever a localized anode occurred and either chemical, biological or
other influence created a galvanic cell consisting of the tank and its
surroundings, corrosion was concentrated at one or a few points.
Localized tank corrosion proceeds at a pace determined by the properites
of the backfill in the vicinity of the anode and can ultimately cause a
failure.
Previously no body of theory existed which could serve to predict
the age at which an underground steel tank would develop a corrosion
induced failure. The chemistry of corrosion was well known but a
mathematical model of the particular process which leads to tank
corrosion failure had not been developed.
In an attempt to develop, a predictive tool of that kind, the
petroleum industry launched large scale data collection efforts to gather
information on the factors which the theory of corrosion chemistry
indicated would be relevant. To date, data has been collected at
approximately 10,000 sites throughout the United States and Canada.

-------
STATISTICAL MODELING DEVELOPMENT
The first age-to-failure statistical model developed provides
estimates of the average age at which a tank at a site may fail due to
ext ernal corrosion. In addition, it provides estimates of the
probabilities that a tank has experienced corrosion failure or may fail
at any specific age. These estimates are derived from physical site
measurements of resistivity, pH, moisture, and sulphides in the immediate
backfill and knowledge of tank age and size.
The measurements are combined in the model to provide the mean, or
average age, of a corrosion failure in the specified environment. In
addition, it was found that the pattern of departures from that mean age
conform very well to a well-known mathematical form, the so-called normal
distribution. Because of this, the probabilities mentioned can be
calculated to determine confidence limits for estimates obtained from the
model.
A second model has been generated to predict internal corrosion
induced failures which most often initially develop directly beneath a
tank's fill tube. Such metal failures, the result of the combined
process of erosion and corrosion, have been found to be linearly related
(proportional) to volume of sales and refill rate. Again, as with
external corrosion, both the average age of failure for a given volume
and refill rate and the probabilities of failure at any specified ages
can be estimated.
MODEL VALIDITY
In the initial stages of the research, the approach followed was to
concentrate analytical efforts on one data set and develop a predictive
model. Validation of the model was then sought by reestimating its
parameters from an independently collected data set. With the exception
of a provision which must be made for differences in Canadian and U.S.
tank installations, two essentially identical models have been derived
from the independently collected data sets.
Confidence in recommending the use of these models is based on
several factors:
1.	Successful validation using an independently collected data set.
2.	Semi-empirical structure of the model relates to known
physical process of corrosion.
3.	The model explains 7 5 percent of the variation in the field
data.
4.	Accuracy of predictions derived from the two models were .
verified. Both models were applied to data on non-leaking tanks'
and it was found that they predicted future failures at times
consistent with the reported physical condition.of the tanks.
3

-------
CONCLUSIONS
Statistical analysis of tank site data leads to two basic
conclusions. First, that tank sites fall into two categories:
approximately one quarter (23 percent) of the sites had tanks which were
corroded uniformly and did not present corrosion failure problems; tanks
at the remaining three quarters (77 percent) of the sites experienced
localized external corrosion. Second, the age at which a locally
corroded tank will leak is a normal random variable whose mean and
variance can be estimated from data on the variables noted earlier.
This means that if data is collected on the variables listed and the
calculations for determining the mean age are carried out, the result is
the age at which, on the average, a tank with localized anodes will fail
due to external corrosion. Note that this implies that of a population
of tank sites in similar environments, roughly half will fail before the
mean age and half after.
However, in accordance with recognized statistical formulas, if the
computed mean is subtracted from the actual age and the result is
divided by the standard deviation, then with tables of the standard
normal distribution it is possible to compute the probability that a tank
at a site has developed a perforation, the probability that it will fail
at or before any predetermined age, or the age at which any level of
probability of failure will be attained.
For example, consider	an
characteristics.
age	resistivity	pH
10 yrs. 2000 ohms	6.5
From the model, the estimated	mean
installation with the following
size	moisture sulphides
3000 gal. 1 level 0 constant
age to external leak is 10.5 years.
To compute probabilities of leak at or before specified ages, it is
necessary to first normalize and then use the standard normal tables.
age ^ mean age ¦ 10_ 10.5 = ^5_ = -.2
standard deviation	2.5	2.5
The probability that this tank is now leaking is thus 0.42.
Again using the standard normal distribution tables, the probabilities
of a failure developing at the ages listed are computed.
Age	5.00 6.20 7.70 8.30 10.50
Probability of failure 0.01 0.04 0.13 0.19 00.42
Note that these probabilities are conditioned on the existence of
localized corrosion which was observed in the data to occur in 77 percent
of the cases examined. The unconditional probabilities can thus be
estimated by multiplying by 0.77.
4

-------
ROLE OF INVENTORY MONITORING
It is, of course, impossible to determine directly whether a tank is
uniformly corroding or corroding at a point anode without physical
inspection. However, a statistical procedure has been developed to
assess the probabilities of either condition. Returning to the example
given earlier, it is now assumed that the tank has been closely monitored
by accurate daily inventory control to detect leakage until the tank is
15 years old and has not lost product.
It can be calculated, as before, that if point corrosion existed,
the probability of failure would be:
P(leak / point.corrosion) ¦ 0.96
This suggests that point corrosion is extremely unlikely.
It is possible to be more definitive. With knowledge of the
characteristics of the taethod used to determine tank tightness and the
use of what are called Bayes estimation procedures, an estimate of the
probability that the tank is corroding uniformly can be produced. As the
tank ages past the mean age to leak for a locally corroded condition,
this probability will grow very ranidly.
DECISION TREE MODELING
It must be emphasized that knowlege of the failure probabilities,
while useful, does not provide a complete basis for deciding among the
various courses of action which might be pursued. To form a basis for
rational decison,it is necessary to array all alternatives, estimate all
relevant costs of an undetected failure and compute the expected costs of
each action and outcome combination.
This form of decision tree modeling is, however, well within the
state-of-the-art once failure probabilities can be calculated. Thus with
the ability to calculate probabilities, such a decision making procedure
is feasible.
To implement a program based on this procedure requires four
activities which can proceed concurrently.
1.	Data collection and analysis, probability estimation and
dev-elopment of a prioritized list of installations, e.g.,
highest to lowest probability of leak.
2.	Tank tightness integrity determination at highest priority
locations.
3.	Alternatives evaluation and optimal course of action
determination.
4.	Model revision and update to reflect newly acquired data.
5

-------
1) Data Collection
Much of the difficulty encountered In conducting the statistical
analysis was due to problems in the data. This was to be expected in
that the data was collected in the absence of a model so that the
required accuracy, extent of collection and ultimate sensitivity could
only be estimated. As a result of this research, procedures can be
suggested which should improve future data collectioni
In general, the best data was generated whenever data collection was
supervised by someone technically qualified and familiar with what the
end use would be. It is recommended therefore that whoever is assigned
responsibility for data gathering become extremely familiar with the
statistical procedures developed and with the relevant aspects of soil
chemistry. -This latter should include as a minimum familiarity with the
reasonable ranges of values to be expected in a given geographical
region.
Continuing quality control of field measurements can be maintained
by periodic laboratory analysis. Although it is felt that field
measurements are adequate if they are taken in a reasonably accurate
manner, it is recommended that samples be preserved for further backup
laboratory analysis should anomalous measurements be reported.
Beep cores, down to tank bottom, are preferred since soil from
shallow cores can be reported dry when the tank bottom is in ground
water.
A*carefully monitored pilot program at a few locations can be
productive in ironing out the mechanisms of data collection, lab
analysis, reporting and in developing specifications and procedures for
routine implementation.
2)	Leak. Prevention Priority Setting
In a leak detection and prevention program involving one or more
measures such as inventory control, tightness testing, interior lining,
cathodic protection retrofit, tank replacement, and the like, the site
deserving priority attention can be determined immediately from the
current probabilities of a corrosion induced failure generated by the
appropriate model.
In general, tank testing, replacement or similar measures, when
based on age alone is not recommended. Such actions are arbitrary,
costly and their immediate value questionable at sites which display
extremely low probabilties of failure.
3)	Alternative Evaluation
The conclusions of this project lead to the recommendation of the
development of decision procedures concurrent with data collection and
modeling. There are many leak prevention alternatives which include
taking little action at one extreme and closing the facility at the
6

-------
other. Many complex interactingfactors, economic and otherwise, oust be
considered i't», "choosing among them. Our analysis to date provides one
essential element, the probability of failure. 'What remains to be
determined is the acceptability of any specific probability at- any given
location and the proper and efficient course of action if it is not
acceptable.
The criteria for such decisions are necessarily specific to a
location since they involve several factors such as environmental risks
and the economics of continued station operation. However, once the
probabilities of possible outcomes are known, the other relevant factors
can be.determined. Known procedures can then be implemented to provide
facility owners with the expected relative costs, benefits and risks of
any alternative course of action.
4) Model Revision and Update
As soil data is collected and tank condition results become
available, an ongoing reestimation of parameters and model update is
recommended. This would provide immediate evidence of anomalies. Also,
it would serve as a method to discover the emergence of any previously
undetected phenomena not revealed in the data analyzed to date.
7

-------