EPA-600/2-81-212
September 1981
A HAZARDOUS MATERIALS SPILL WARNING SYSTEM
by
Milton Kirsch
Robert Melvold
John Vrolyk
Rockwell International
Newbury Park, California 91320
Contract No. 68-03-2080
Project Officer
Joseph P. LaFornara
Oil & Hazardous Materials Spills Branch
Municipal Environmental Research Laboratory-Ci.
Edison, New Jersey 08817
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory-Cincinnati, U.S. Environmental Protection Agency, and approved for
publication. Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or rec-
ommendation for use.
ii
-------
FOREWORD
The U.S. Environmental Protection Agency was created because of increas-
ing public and government concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled land
are tragic testimonies to the deterioration of our natural environment. The
complexity of that environment and the interplay of its components require a
concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem solution;
it involves defining the problem, measuring its impact, and searching for solu-
tions. The Municipal Environmental Research Laboratory develops new and im-
proved technology and systems to prevent, treat, and manage wastewater and
solid and hazardous waste pollutant discharges from municipal and community
sources, to preserve and treat public drinking water supplies, and to minimize
the adverse economic, social, health, and aesthetic effects of pollution. This
publication is one of the products of that research and provides a most vital
communications link between the researcher and the user community.
One of the requirements to control pollution of our waterways by hazard-
ous materials is a device to detect their presence. The program reported
herein was designed to fulfill this need and is part of a continuing program
of the Oil and Hazardous Materials Spills Branch, MERL-Ci, to detect and miti-
gate pollution from hazardous material spills.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
iii
-------
ABSTRACT
The Environmental Protection Agency has developed a list of materials
defined as hazardous substances based on their aquatic toxicity. In addition,
certain materials have been designated as "priority pollutants." Often, a
spill of toxic materials into a moving water stream can occur without the
spiller being aware, or without the spiller notifying authorities. According-
ly, a system was needed to detect the presence of hazardous toxic materials in
streams and rivers. This need has been filled by providing a spill alarm sys-
tem, which was designed, fabricated, and tested prior to delivery. It con-
sists of nonselective detection components which together serve to detect all
the materials on the designated hazardous materials list, and the priority pol-
lutants. The system was mounted in an automotive trailer and delivered to the
Oil and Hazardous Materials Spills Branch in Edison, New Jersey.
This report was submitted in fulfillment of Contract 68-03-2080 by
Rockwell International under the sponsorship of the U.S. Environmental Protec-
tion Agency. This report covers the period from June 28, 1974, to December 1,
1978. Rockwell work was completed as of October 1, 1977.
iv
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Tables vii
List of Abbreviations and Symbols viii
1. Introduction and Summary 1
Background 1
Summary 2
2. Conclusions and Recommendations 6
Conclusions 6
Recommendations 6
3. Operation of the Hazardous Materials Spill Warning System .... 8
Overview 8
Detail 9
4. System Testing 28
Laboratory Evaluation of Individual Detection Components. . 28
Laboratory Evaluation of Integrated System 34
Trailer Mounting 41
Field Testing of the Hazardous Material
Spill Warning System at the Rocketdyne
Santa Susana Field Laboratory 41
Field Testing on the Los Angeles River 45
References 47
Appendices
A. Instrument Manufacturers 48
B. Operating Instructions 51
C. Spill Alarm System Interim Report, Environmental Emergency
Response Unit 57
-------
FIGURES
Number Page
1 Hazardous Material Spill Warning System, block diagram 8
2 Hazardous Material Spill Warning System, pictorial view 10
3 Hazardous Material Spill Warning System 11
4 Warning system flow diagram showing nominal flowrates 12
5 Rotary injection valve Model DC-60 TOCA 15
6 Alarm control system, schematic 17
7 Alarm control system, pictorial view 18
8 Relation between total dissolved solids and electrical
conductivity 22
9 UV absorptimeter rezero fluid control system 25
vi
-------
TABLES
Number Page
1 Characteristics of Prosser Pumps for Water Quality Monitoring . . 13
2 Hazardous Materials Detectable at the 0.1 ppm Level Above
Buffer Capacity by Measurement of pH 20
3 Hazardous Materials Detectable by Oxidation-Reduction Potential. . 21
4 Specific Conductivity of Some Hazardous Materials in Water .... 22
5 Hazardous Materials that are Electrically Conducting in
Aqueous Solution 23
6 Hazardous Materials Detectable by Ultraviolet Absorption 26
7 Sensitivity of Ultraviolet Absorption Sensor for Some
Hazardous Materials 29
8 Carbon Contents of Aqueous Solutions of Selected Organic
Compounds as Measured by the Dohrmann DC-60 Carbon Analyzer ... 32
9 Effect of Some Inorganic Compounds on pH, ORP, Conductivity
and UV Absorption 35
10 Response of the Warning System to the Presence of Some Organic
and Metal-Organic Compounds 39
11 Change in Response Due to Stream Spills of Hazardous Materials . . 44
vii
-------
LIST OF ABBREVIATIONS AND SYMBOLS
amps — amperes
C — carbon
°C — degrees Centigrade
cc — cubic centimeters
C/L — carbon per liter
cm — centimeters
C02 — carbon dioxide
cps — cycles per second
EDTA — ethylenediamine tetraacetic acid
EERU — Environmental Emergency Response Unit
FS — full scale
g — gram
gph — gallons per hour
gpm — gallons per minute
HAHC — hydroxylamine hydrochloride
HC1 — hydrochloric acid
ir — infrared
KC1 — potassium chloride
KHP — potassium acid phthalate
L — liter
M — molar (moles per liter)
m — meter
mg — milligrams
ml — milliliter
mm — millimeter
mmho/cm — millimhos per centimeter
mV — millivolt
um — micrometer
Umho — micromho
N — normal (equivalents per liter)
N2 — nitrogen
NaOH — Sodium hydroxide
nm — nanometers
ORP — oxidation-reduction potential
ppm — parts per million
psig — pounds per square inch gauge
PVC — polyvinylchloride
TC — total carbon
TOG — total organic carbon
TOCA — total organic carbon analyzer
uv — ultraviolet
VDC — volts, direct current
viii
-------
SECTION 1
INTRODUCTION AND SUMMARY
BACKGROUND
The legislative base upon which the development reported herein rests is
the Federal Water Pollution Control Act Amendments of 1972, Public Law 92-500.
In consecutive paragraphs of Section 311 of the law, it is stated that:
"(b)(l) The Congress hereby declares that it is the policy of
the United States that there should be no discharges of oil or
hazardous substances into or upon the navigable waters of the
United States, adjoining shorelines, or into or upon the waters
of the contiguous zone."
"(2)(A) The Administrator (of the Environmental Protection
Agency) shall develop, promulgate, and revise as may be appro-
priate, regulations designating as hazardous substances, other
than oil as defined in this section, such elements and compo-
nents which, when discharged in any quantity into or upon the
navigable waters of the United States or adjoining shorelines
or the waters of the contiguous zone, present an imminent and
substantial danger to the public health or welfare, including,
but not limited to, fish, shellfish, wildlife, shorelines, and
beaches."
Under this authority the Environmental Protection Agency concurrently
began the process of designating 271 materials as hazardous, and this program
to develop a warning system to detect spills of hazardous materials in natural
waterways. The advance notice of this intent to designate hazardous materials
was published in the August 22, 1974 issue of the Federal Register, and the
proposed rules appeared in the December 30, 1975 issue with some modifications,
and inclusion of harmful quantities and penalty rates. Final notice of the
rules was published March 13, 1978, and amendments to the Clean Water Act of
1977 changing enforcement for chemical spills were passed on October 14, 1978.
This was followed on February 16, 1979, with a series of amendments to the
final rules and the designation of 28 additional hazardous materials
(40CFR116). In June 1978, a court settlement involving the EPA and certain
concerned plaintiffs required the EPA to identify those materials which were
considered toxic. This is the "EPA Consent Decree" and as a result EPA iden-
tified 65 toxic chemicals, which formed the Toxic Pollutant List; this list was
eventually expanded to 129; these chemicals are commonly referred to as Prior-
ity Pollutants, based upon frequency of occurrence ia worldwide water analysis
(1).
-------
The Federal program to combat spills is dominated by three goals:
1. To prevent spills.
2. To detect spills.
3. To contain, remove, and clean up spills.
The spill alarm system was designed to advance the state of the art in
detecting that a spill has occurred. Quick notification is essential to miti-
gating the spill effects.
The detailed list of materials designated as hazardous has undergone some
changes since the initiation of the program, but such changes were expected.
Detection components were therefore chosen from among nonselective candidates,
each of which could respond to many materials on the list. The rationale
underlying this approach was that it was more important to detect all spills
of hazardous materials than to avoid false alarms caused by spills of nonhaz-
ardous materials to which the detection components would also respond. It was
recognized at the outset that the sensitivity for some of the most toxic haz-
ardous substances could be less than desirable, but this limitation was accepted
as necessary to build a system from commercially available major components.
SUMMARY
A hazardous material spill warning system has been designed, fabricated,
and tested. It meets the following criteria laid down when the program was
started in July 1974.
1. Detection components, where possible, shall be off-the-shelf items
ready for use with a minimum of modifications.
2. Detection components shall be "rugged" enough for use in an untended
remote station and capable of functioning around the clock.
3. The final package shall contain components which are resistant to
fouling, scaling, etc., to a degree where cleaning is necessary only
once every 14 days or longer.
4. Detection components shall produce an electrical response which can
be transmitted by wire or radio to a distant receiver.
5. Detection components shall remain sensitive under adverse conditions
such as high flow or rapid current in a watercourse.
6. Detection components shall be capable of distinguishing between a
slug of a given hazardous material(s) and concentration levels of the
substances normally present in a watercourse.
7. Detection components shall not degrade any of the beneficial uses of
a watercourse.
-------
8. The final package, consisting of one or more detection components,
shall be capable of sensing both organic and inorganic hazardous
materials. The final package need not contain components which
selectively detect the presence of each substance on the list.
Rather, several detectors capable of sensing wide classes of these
substances would be desirable both in terms of low initial cost and
of ease of maintenance. The detectors, as stated previously, must
be capable of distinguishing a slug of a hazardous substance from
the materials normally present in fresh, brackish, or saline waters.
The entire instrument package is housed in an air-conditioned, 8.2-m (27-
foot) automotive trailer. The warning system can, therefore, be moved conven-
iently from one site to another. A submersible pump in the waterway feeds the
instrument package through 5-cm- (2-inch)-diameter plastic hose. The only
materials added to the natural water as it passes through the trailer are small
quantities of hydrochloric acid (200 ml [7 fluid ounces] of 0.5 N acid per day)
as part of the determination of total organic carbon content, and a minute
quantity of a biocide to reduce proliferation of bacteria and algae in a res-
ervoir of rezero fluid.
Total water flow is about 38 L (10 gallons) per minute. Quantities of
materials added are too small to affect any of the desirable uses of the water.
The detection components that have been incorporated into the integrated
system are:
1. pH probe - Leeds & Northrup.
2. Oxidation-reduction potential, ORP sensor - Leeds & Northrup.
3. Electrical conductivity sensor - Leeds & Northrup.
4. Ultraviolet absorptimeter - Teledyne.
5. Total Organic Carbon Analyzer, TOGA - Dohrmann.
The strip chart recorder channel for each detection component has a
built-in alarm circuit, the response level of which can be preset so that an
electrical signal is transmitted when the level is exceeded. (The pH and ORP
circuits have a settable alarm level below a certain limit as well as above an
upper limit to the "normal" range.) When an alarm condition is reached, the
following events occur:
1. A signal light on the control panel indicates which detection compo-
nent produced the alarm.
2, A solenoid valve is opened for a selected time to permit collection
of a 3.8-L (1-gallon) grab sample of the water for later chemical
analysis.
3. A telephone dialer is actuated and transmits a prerecorded message to
any chosen telephone station. The dedicated telephone receiver in use
-------
has been outfitted with an automatic telephone answerer so that the
prerecorded alarm message is recorded at the receiving station.
4. The recorder chart speeds (for all but the TOC analyzer) are increased
from the normal 1.3 cm (0.5 inch) per hour to 15 cm (6 inches) per
hour to show more detail. After the spilled material has passed com-
pletely through the detection components, the alarm lights are auto-
matically turned off and after a preset time, a second relay is ener-
gized to prepare to collect a second grab sample when a second spill
occurs unless the TOGA was alarmed by the first spill. After a. sec-
ond spill alarm, the telephone dialer sends a second prerecorded
message.
Sufficient stability of.the ORP, pH, and electrical conductivity detec-
tion components for a two-week period unattended operation has been available
in a number of commercial instruments for some time. Stability of the ultra-
violet analyzer stems from a combination of an automatic electronic rezero
circuit and provision for simultaneous flushing of the sample cell with fluid
to maintain a constant baseline. The total organic carbon analyzer is also
equipped with an automatic rezero circuit which allows adequate zero drift
control over the two-week period of unattended operation.
The integrated warning system was first tested in the laboratory. Water
was recirculated from a 208-L (55-gallon) drum. Tests of individual hazardous
materials were made by addition of known quantities to water in the drum to
determine sensitivity levels. When no hazardous materials were added, the
system ran continuously for two weeks without attention.
The entire system, mounted in the automotive trailer, was successfully
moved to the Rockwell Santa Susana Field Laboratory (SSFL), where a simulated
natural waterway was available for further testing the warning system. The
simulated waterway is part of a wastewater control system within which are sev-
eral ponds where effluents from rocket firings and other operations can be held
for cleanup before discharge to the Ventura County sewage system. At low flow-
rates, the simulated waterway is a possible carrier for added hazardous mate-
rials which can be transported by the submersible pump to the warning system.
Occasionally, as when rocket fuel tanks are purged or cleaned, the waterway
carries high concentrations of organic compounds. Under both circumstances
the warning system was tested at SSFL, and it responded both to hazardous ma-
terials placed in the waterway specifically for testing it, and to foreign
material introduced by other operations without necessarily prior notification
of the warning system personnel.
A test at the Los Angeles River was conducted successfully. Passage of a
uv-absorbing material was detected, indicating introduction of some foreign
substance upstream.
A field test by Mason & Hanger, the Environmental Emergency Response Unit
contractor, showed that unattended operation of two weeks was not achieved.
The alarms received from the system were of a continuous nature and distinction
between equipment failure, zero drift, and actual spill was not possible. The
results are partially attributed to the parameters of the groundwater runoff
-------
and the soils saturated with various organic and inorganic materials present
at the test site.
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
CONCLUSIONS
The warning system has been shown to be capable of detecting a wide vari-
ety of hazardous materials, and of unattended operation for a two-week period.
It performed better than expected on a simulated waterway near a rocket engine
test site where it was subjected to high vibration levels and extremes in
water flowrates. On the Los Angeles River, where it was located 82 m (270
feet) from and 9 m (30 feet) above the stream bed, it detected a highly uv-
absorbing material that took more than 40 minutes to pass the sampling point.
Its usefulness in detecting spills has therefore been amply demonstrated. Sug-
gestions for further testing and improvements will increase its usefulness and
capability.
RECOMMENDATIONS
The warning system should undergo extensive field testing under a wide
variety of conditions to clarify what improvements would permit longer periods
of unattended operation. Such testing will need to include provision for re-
ducing the possibility of vandalism by displaying warning signs and audible
onsite as well as remote alarms in case of tampering. Tests should be con-
ducted on waterways that differ in dissolved solids content, temperatures and
frequency of pollution by hazardous materials. The grab sample collected when
an alarm is produced should be analyzed qualitatively and quantitatively to
determine what material(s) caused the alarm.
Improvements of several types may be made for the next generation of warn-
ing systems, and improved techniques for using the present one will increase
its usefulness. Judicious placement of the submersible pump may increase the
sensitivity of the system to water-insoluble floaters or sinkers. A multi-
channel telephone dialer may be used to identify in the transmitted message
which of the individual sensors triggered the alarm signal. Improved solenoid
valves should be installed to reduce the need for servicing.
At present, the TOCA alarm does not reset automatically once the material
which produced the alarm is no longer present. The second grab sample circuit-
ry will therefore not be actuated if the first alarm is, in part, the result of
high TOC content in the stream. A relay arrangement to reset the TOCA alarm
automatically should be installed to overcome this difficulty.
A temporary power failure will not produce any serious problem in the
warning system, but it will extinguish the deuterium lamp in the uv analyzer.
-------
Additional circuitry is required to restart the lamp automatically when, power
is restored. (At present, a spring-loaded switch must be momentarily engaged
manually to start the deuterium lamp.)
-------
SECTION 3
OPERATION OF THE HAZARDOUS MATERIALS SPILL WARNING SYSTEM
OVERVIEW
A block diagram (Figure 1) of some of the important components of the
warning system shows the relationship among the detectors, the general flow-
pattern of the water sample, the grab samplers, and the alarm circuitry. A
submersible pump (Prosser Model 9-04011) in the watercourse supplies a water
sample continuously to the instrument package in the trailer. Most of the
water flows through the redox potential (ORP), pH, and conductivity cells and
back to the waterway. A small isokinetically split sidestream goes on to the
homogenizer and thence to both the differential ultraviolet absorptimeter and
WATERCOURSE
•DENOTES ISOKINETIC f LOW SPLITTING DEVICE
Figure 1. Hazardous Materials Spill Warning System, block diagram.
8
-------
the total organic carbon analyzer. When an alarm condition is detected by any
of the major detection components, a grab sample is collected in the 3.8-L
(1-gallon) sample bottle and the telephone dialer sends a prerecorded message
to the automatic telephone answerer.
The 8.2-m (27-foot) automotive trailer which houses the instrument pack-
age is shown in Figure 2. The system as a whole, depicted in Figure 3, con-
sists of three main consoles. The central unit contains the Teledyne General-
Purpose Analyzer, Model 611DS2, across the top. Upon its frame are mounted
the Leeds & Northrup pH, ORP, and conductivity cells across the bottom front.
At the back of this central unit are housed much of the auxiliary plumbing
that integrates the individual major components. The right-hand console is
the Dohrmann Model DC-60 Total Organic Carbon Analyzer with its built-in
recorder.
To the left is the control console with the two Leeds & Northrup record-
ers across the top. The left-hand recorder is a three-pen type and also con-
tains the necessary circuits to power the L&N detectors. The blue pen of the
right-hand two-channel recorder is connected to the output of the Teledyne
analyzer, the differential ultraviolet absorptimeter, and contains a spare
channel for some device to be added later. The charts on both these recorders
normally run at 1.3 cm (0.5 inch) per hour. When an alarm condition is
reached on either recorder, both charts automatically speed up to 15 cm (6
inches) per hour to show more detail.
The panel below the recorders in the control console houses the timer-
relays that control the solenoid valves on the grab samplers, the alarm
lights, and the connections to the automatic telephone dialer.
The lower panel contains the timer and counter relays to control the flow
of fluid to and from the uv rezeroing reservoir and the addition of faiocide to
it. The counter is actuated every hour by the rezeroing cam on the uv absorp-
tion instrument which flips a three-way solenoid valve to the position that
allows fluid from the rezero reservoir, instead of sample, to flow through the
uv cell. The counter resets to 24 after it reaches zero. At that time a
timer-relay is engaged that energizes a solenoid valve to empty the reservoir.
A second timer operates a solenoid valve to refill the reservoir with fresh
fluid and a third timer controls a valve that permits addition of biocide to
the reservoir for a predetermined time. The 1-liter bottle in which the bio-
cide is stored is large enough to contain a two-week supply.
DETAIL
Water Flow
A detailed flow diagram is shown in Figure 4. The main submersible pump,
placed in the waterway, and not depicted in the diagram, has manufacturer's
performance characteristics listed in Table 1. The pump is surrounded by a
coarse cylindrical stainless steel screen filter containing 3-mm (1/8-inch)-
diameter holes. Particles smaller than this size can be easily handled by the
ORP, pH, and conductivity detectors. The Raytheon homogenizer breaks down par-
ticles in this size range to 0.1 mm or less so that they can be handled by the
small lines of the uv instrument and TOG analyzer.
-------
Figure 2. Hazardous Materials Spill Warning System, pictorial view.
-------
Figure 3. Hazardous Materials Spill Warning System
-------
U><^1 >T *
-r 60UCTOH fil T
•&3 <-I I ' <
© 0 Ul FLOW
v-/ ^-^ x-^J AP /O\
> 'I _j 4
D 0^—1»»
DRAIN TO W /—\
(<)
DflAIN
Figure 4. Warning system flow diagram showing nominal flowrates
(main flow in heavy lines)
12
-------
TABLE 1. CHARACTERISTICS OF PROSSER PUMPS FOR WATER QUALITY MONITORING
Maximum
*
TDH(m)
3.05
6.10
9.15
12.2
15.2
18.3
21.3
24.4
27.4
Maximum
TDK* (feet)
10
20
30
40
50
60
70
80
90
performance
L/minute.
144
140
132
121
106
95
79
61
38
performance
US GPM
38
37
35
32
28
25
21
16
10
Head loss (friction) per
30.5 m of 2.5-cm-diameter oiue
@ L'/minute
19
38
57
76
95
114
132
Valve =
Fitting
Loss in m
0.6
2.4
5.2
8.5
12.8
18.3
24.1
6.1 m of pipe
= 1.5 m of pipe
Head loss (friction) per
100 feet of 1- inch-diameter pipe
-------
The heavy lines in Figure 4 show the direction of flow during normal
operation of the warning system when no alarm condition exists. The numbers
indicate the ideal flowrates in the various parts of the system. Of the 38 L
(10 gallons) per minute supplied by the submersible pump to the trailer, 1.1 L
(0.3 gallon) per minute is isokinetically split at Ij; 22 L (5.8 gallons) per
minute runs through the ORP, pH, conductivity cell, and out to waste. The re-
maining 15 L (4 gallons) per minute serves as the heat exchange fluid to cool
the recirculating liquid that acts as coolant around the insulation for the
furnace on the TOG analyzer. It also acts as a constant supply for the eductor
and produces the pressure drop required to .flush the various small streams to
waste.
The 1100 ml/minute (17.4 gallons/hour [gph]) is split just above the
homogenizer, into an 800 ml/minute (12.7 gph) stream and a 300 ml/minute (4.8
gph) stream. The larger stream usually goes to waste, but it also supplies the
grab sample bottles when an alarm condition is detected and the three-way sole-
noid valve is opened in the appropriate direction to allow filling of the bot-
tle. The 300 ml/minute output of the homogenizer normally feeds the ultra-
violet absorptimeter and TOG analyzer, 70 ml/minute (1.1 gph) going to the uv
cell and 230 ml/minute (3.6 gpm) to the TOG analyzer overflow.
The TOG analyzer requires only 25 ml/minute (0.4 gph). The remainder of
the 230 ml/minute flow usually overflows and goes directly to waste. Once
each day, however, the excess is used to fill the uv rezero reservoir auto-
matically. When that operation occurs, a small quantity of biocide is also
added from the disinfectant reservoir to produce a concentration of about 5
ppm of quaternary ammonium chloride to keep down algae growth.
Once per hour the rezero fluid automatically flushes out the uv cell at
the same time as the electronic rezero circuit is engaged. This feature main-
tains a stable baseline for the uv output signal over the two-week period of
unattended operation. The rezero fluid reservoir is conical and Teflon-lined
so that particulate matter will not accumulate at the bottom, but will be
swept out when it is emptied. The stirring motor, with which it is provided,
maintains a water quality representative of that in the watercourse.
The TOG analyzer operates semicontinuously. The water sample is pumped
continuously by a Masterflex pump at the rate of 25 ml/min (0.4 gph), and hydro-
chloric acid is added just before entry of the mixture to an air sparger where
the carbon dioxide produced by the inorganic carbon compounds is stripped off.
The sample containing only organic carbon compounds is then fed to the in-
jector valve which incorporates several unique features. As shown in Figure
5, the injector face is continuously flushed with fresh fluid delivered by the
sparger. The opening through which a constant volume of sample is injected
into the furnace is flushed once to waste just before the entire injection
port is turned to position it over the combustion tube and the fresh sample is
forced into the furnace. Combustion of the sample at 900° C in a stream of
oxygen converts all the carbon remaining in the water sample to carbon diox-
ide. The carbon dioxide content in the gas is sensed by a nondispersive ir-
absorption unit. The strip-chart recorder can be made direct-reading by cal-
ibration with both a known concentration of carbon dioxide in a gas and a
standard solution of potassium hydrogen phthalate. The injection valve
14
-------
a.
a.
O
CC
O
LU
a
at
>
<
5
LU
£ %
s-
§ o
=• ai
53
is
LU
85
23
«S
5 w
u. <
Sj*8
LU
•xl
Cfl
M
LU
cc cc
LU LU
3
33
S
LU
W)
U)
o
IU
a
LU
3
o
LU
O
x 3 =
z S
cc i2
< Q. a. o a.
LU
<1
O O
32
ii
E CC
a. X
M H
LU
r- I- 2 «
3 3 ± 3
Z Z < I-
H ni
"I
°- -d
o <
y
o s
M I.I •-
V)
cc a
U. i" Z X
xJ££
(/] 3 » «
CM CM CM CM CM CM
0)
•o
o
a
c
o
O
01
-------
operation frequency is determined by a selector switch that allows injection
at 5-, 10-, 15-, 30-, 60-, or 180-minute intervals.
The entire warning system operates with the addition of insignificant
quantities of extraneous materials to the waterway. The TOG analyzer consumes
about 200 ml of 0.5 N hydrochloric acid per day to remove carbonate and bi-
carbonate from the water sample. With the normal total flow through the warn-
ing system of 38 liters per minute (10 gpm), this amount of hydrochloric acid
corresponds to the introduction of 0.07 ppm (1.8 x 10"^ N) HC1 to the efflu-
ent from the warning system if no inorganic carbon is present in the influent
water. This concentration of acid is not enough to affect any beneficial uses
of the waterway.
The rezero system for the ultraviolet absorptimeter requires a small
(5 ppm) concentration of a quaternary ammonium chloride as a biocide to reduce
algae growth. Each day, about 3.8 L (1 gallon) of the rezero fluid (pri-
marily natural water from the watercourse being monitored) flows through the
uv cell. The biocide present in the effluent will have a concentration of
3.4 x 10"^ ppm, assuming normal flow through the entire system, much too small
to affect the beneficial uses of the waterway.
Alarm Control System
The alarm control system wiring diagram appears in Figure 6 and the cor-
responding front panel on the control console is shown pictorially in Figure
7. From the wiring diagram, it is seen that closure of any one of the single-
pole, single-throw alarm switches, which occurs when the predetermined "normal"
range of any one of the variables is exceeded, results in turning on the ap-
propriate signal light and transmission of a signal to the time-delay relay
TR]_. This relay fulfills several functions, the most obvious of which is to
open the solenoid valve, SVj_ (18 of Figure 4), for a selected time to allow
filling of the first grab sample bottle. When the alarm condition that trig-
gered TRj_ no longer exists, the transfer relay TR.^ serves as a monitor to per-
mit actuation of TR2 if sufficient (preset) time elapses before the next alarm
signal occurs. Then the second grab sample bottle is filled as a result of
the solenoid valve SV (20 of Figure 4) being opened for a selected time.
The occurrence of each alarm condition with sufficient "normal" opera-
tion between,them also actuates one channel of a dual-channel automatic tele-
phone dialer. Each channel is connected through a relay in parallel with the
alarm signal light of each spill circuit. The device dials any selected tele-
phone number and transmits a prerecorded message that announces the occurrence
of the spill. The announcement can be made to an ordinary telephone if it is
answered when the number is dialed. However, to avoid the need for continuous
human monitoring of the receiving telephone, an automatic telephone answerer
has been connected to the dedicated telephone receiving line. The Phonemate
Remote 9000 records the message received from the automatic telephone dialer.
This model includes a useful feature for this application. The answerer can
be interrogated from any telephone by use of a pocket-size, battery-powered
tone key that is coded to the answerer. Whether an alarm has occurred can be
determined by calling the telephone answerer periodically without requiring
constant surveillance of the telephone line.
16
-------
WHITE „
~~——,.., _p -*ai]-*u ir
[L
I CHANC1NO *—(D
T WHING OIACHAM I
GACENl WHITE
CHASSIS
GROUND
SOLENOID
SAMPLE l\
Figure 6. Alarm control system schematic.
17
-------
Figure 7. Alarm control system, pictorial view.
-------
Detector Probes
The warning system was designed to detect as many as possible of the
materials designated as hazardous by the EPA that appeared in 40CFR116 and
the Consent Decree.
pH Probe—
The substances detected by the pH electrode at a concentration of about
0.1 ppm above the buffer capacity of the watercourse are listed in Table 2.
This list assumes an equivalent weight of about 100 and a detection limit of
1 pH unit above or below that of the natural waterway. If the alarm condi-
tions are to be preset for below pH 6.5 or above pH 8.5, strong acids or
bases would have to be present in the water sample at a concentration level of
more than 3 x 10~^ equivalents per liter (i.e., at about 0.3 ppm) above the
buffer capacity of the natural water. The substances listed are either strong
acids or bases or form strong acids or bases rapidly in contact with water.
Weak acids and bases or substances that react with water to form them will
also be detected by the pH electrode, but at higher concentration levels that
depend on the strength of the acid or base.
Oxidation-Reduction Potential Sensor—
Hazardous substances detectable by the oxidation-reduction potential
(ORP) electrode are listed in Table 3. Because redox equilibria are often es-
tablished slowly and because of the variable oxygen content of the water, it
is difficult to calculate precisely the concentration level at which each of
these substances is detectable. The complexity of processes that normally
occur in natural water leads to the measurement of a mixed potential. Despite
the difficulty of interpreting the ORP precisely, it is claimed to be useful
in detecting spills of chemicals (2) and it is for this purpose that the
ORP electrode was included in the warning system.
Conductivity Sensor—
The electrical conductivity detector responds to water-soluble electro-
lytes. Hazardous substances that fall into this category are most salts
(organic and inorganic), weak and strong acids and bases and substances that
react with water to form them. Specific electrical conductivities of some
typical hazardous materials are listed in Table 4 at a concentration level of
one weight percent (3). For many salts the specific conductivity falls within a
fairly narrow range at the one weight percent concentration level. The elec-
trical conductivity is approximately directly proportional to concentration at
lower concentrations. Hence the sensitivity of the electrical conductivity
detector is roughly independent of the particular soluble salt that has been
spilled. Strong acids and bases are more highly conducting than salts, but
are generally detectable at lower concentration levels by the pH detector than
by electrical conductivity.
The sensitivity of the conductivity sensor to detect a spill depends on
the background conductivity of the natural water which is in turn related
directly to the total dissolved (inorganic) solids in the water. This rela-
tion is shown graphically in Figure 8 (3).
19
-------
TABLE 2. HAZARDOUS MATERIALS DETECTABLE AT THE 0.1-PPM LEVEL
ABOVE BUTTER CAPACITY BY MEASUREMENT OF pH
Strong Acids or Acid Forming
Strong Bases or Base Forming
Acetyl bromide
Acetyl chloride
Antimony tribromide
Antimony pentachloride
Arsenic trichloride
Benzoyl chloride
Chlorine
Chlorosulfonic acid
Chlorophenol
Chromic acid
Dichlorophenol
Hydrochloric acid
Nitric acid
Pentachlorophenol
Phenol
Phosphoric acid
Phosphorus oxychloride
Phosphorus pentasulfide
Phosphorus trichloride
Sulfuric acid
Calcium carbide
Calcium hydroxide
Calcium oxide
Diphenylhydrazine
Potassium hydroxide
Sodium
Sodium hydroxide
Sodium methylate
Sodium sulfide
If it is assumed that a sudden 20-percent increase in conductivity can be
reliably ascribed to a spill of an electrolyte, this would correspond to an
increase of about 20 ppm in the dissolved solids content of a river with about
100 ppm natural dissolved solids, such as the Columbia River (3). However, in
20
-------
TABLE 3. HAZARDOUS MATERIALS DETECTABLE BY OXIDATION-REDUCTION POTENTIAL
Ammonium bisulfite Potassium chromate
Ammonium hypophosphite Sodium bichromate
Ammonium oxalate Sodium chromate
Ammonium persulfate Iron (II) compounds (3)
Ammonium sulfide Hydroxylamine
Ammonium sulfite Lead compounds
Ammonium thiosulfate Mercury compounds
Arsenic compounds Phosgene
Cadmium compounds Nickel compounds
Calcium arsenite Nitrosomines
Calcium hypochlorite Potassium permanganate
Chlorine Resorcinol
Ammonium bichromate Selenium compounds
Chromic acid Silver compounds
Chromous chloride Sodium bisulfite
Calcium chromate Sodium arsenite
Copper compounds Sodium hypochlorite
Potassium arsenite Thallium sulfate
Potassium bichromate Zinc compounds
the Rio Grande River with its 791 ppm dissolved solids (3), the same criterion
would be met only if the spill produced a 160-ppm increase in the dissolved
solids content.
The more than 150 hazardous substances detectable by the electrical con-
ductivity increase produced when they are spilled in natural water are listed
in Table 5 (3, 4). Many of them are not detected by the other sensors in the
warning system package. Although more sensitive instruments are available for
detecting certain species, e.g., specific ion electrodes for particular metal-
lic ions or anions, their selectivity would complicate the instrument package
by requiring that many specific electrodes be incorporated into the device.
Ultraviolet Absorptimeter—
The differential ultraviolet absorption instrument records the ratio of
the sample absorption at two wavelengths determined by the choice of optical
filters inserted in a rapidly rotating filter wheel. As outfitted for the
warning system, the absorptimeter is equipped with a 390-nm filter as the ref-
erence wavelength and a 230-nm filter for the measuring wavelength. The
21
-------
TABLE 4. SPECIFIC CONDUCTIVITY OF SOME HAZARDOUS MATERIALS IN WATER
(Concentration = 1 weight percent)
Substance
Specific
Conductivity
(mmno/cm)
Substance
Specific
Conductivity
(mmho/cm)
Acetic acid
Ammonia
Cadmium chloride
Cupric sulfate
pentahydrate
Formic acid
Hydrochloric acid
Lead nitrate
Nickel sulfate
0.7
0.7
5.3
5.4
2.4
92.9
5.4
5.8
Nitric acid
Phosphoric acid
Potassium chromate
Potassium dichromate
Potassium hydroxide
Potassium permanganate
Sodium dichromate
Sodium hydroxide
Sulfuric acid
56.1
10.1
10.8
7.3
38.3
6.9
6.9
43.6
47.8
Q-
O.
o
I—
O
CJ
00
Q
O WELL WATERS
— APPRQX. AVERAGE LINE
8' 100 2 468 1000 2 4
CONDUCTIVITY - MICROMHOS PER CM AT 25°C
Figure 8. Relation between total dissolved solids and
electrical conductivity (3).
22
-------
TABLE 5. HAZARDOUS MATERIALS THAT ARE ELECTRICALLY
CONDUCTING IN AQUEOUS SOLUTION
Substance
No. of
Compounds*
Substance
No. of
Compounds*
Acetyl bromide
Acetyl chloride
Aluminum sulfate
Ammonia
Ammonium compounds
Antimony trichloride
Arsenic compounds
Benzoyl chloride
Beryllium compounds
Cadmium compounds
Calcium compounds
Chlorine
Chlorosulfonic acid
Chromium compounds
Cobalt compounds
Copper compounds
Cyanide compounds
Dodecylbenzenesulfonic
acid salts
Hydrochloric acid
Iron compounds
Lead fluoborate
Lead nitrate
Mercury compounds
Nickel compounds
26
3
3
4
14
4
8
4
4
Nitric acid
Phosphoric acid
Phosphorus trichloride
Potassium hydroxide
Potassium permanganate
Selenium dioxide
Sodium selenite
Sodium
Sodium bisulfite
Sodium bifluoride
Sodium fluoride
Sodium hydrosulfide
Sodium hydroxide
Sodium hypochlorite
Sodium methylate
Sodium nitrite
Sodium phosphate
Sodium sulfide
Sulfuric acid
Uranium compounds
Vanadium compounds
Zinc compounds
Zirconium compounds
4
2
13
6
*Number of compounds in class on Hazardous Substances List.
23
-------
measuring wavelength selected was chosen because the absorption of many inor-
ganic complexes increases rapidly with decreasing wavelength below 240 nm.
The sensitivity for detection of many aromatic organic compounds is still high
at the 230-nm measuring wavelength. The choice will therefore provide better
sensitivity for many inorganic complexes than the more usual choice of 254 nm.
This higher measuring wavelength is also frequently chosen because of the high
intensity of a mercury arc lamp at 254 nm. On the other hand, the continuous
emission spectrum produced by the deuterium lamp in the model used for the
warning system allows more flexibility in the choice of reference and measur-
ing wavelengths than can be available from a light source such as a mercury arc
lamp. For some purposes it may be desirable to choose another combination of
wavelengths. This change requires only replacement of the filters in the fil-
ter wheel of the Teledyne General Purpose Analyzer.
The Teledyne instrument has a built-in electronic rezeroing system which
is engaged every hour and lends baseline stability. Associated with this sys-
tem is provision for an auxiliary system to flush out the sample cell with
fluid stored in a reservoir. The electronic control for this rezero fluid
system is shown in Figure 9. The rezero fluid is replaced once each day and
biocide is added to reduce proliferation of living matter. This hourly flush-
ing of the cell simultaneous with electronic rezero aids in the maintenance of
a stable zero reading over an extended period.
A partial list of hazardous substances detectable by the ultraviolet ab-
sorptimeter appears in Table 6.
Total Organic Carbon Analyzer—
All organic compounds are detectable by the Dohrmann Model DC-60 Total
Organic Carbon (TOC) Analyzer. This instrument operates semicontinuously.
Samples can be injected at intervals no less than 5 minutes apart. After re-
moval of the inorganic carbon compounds, a controlled sample volume is in-
jected into a high-temperature furnace where oxygen converts all the organic
carbon into carbon dioxide. The quantity of carbon dioxide in the effluent
gas is measured in a nondispersive infrared unit and is related to the organic
content of the original water sample. The entire process takes place auto-
matically and periodic electronic rezeroing and a. stable infrared detection
system combine to make zero drift small over the two-week unattended period.
When the first alarm includes triggering by the TOGA, the second alarm
will not occur, since the TOCA alarm is not reset automatically when the
alarming material passes the sampling point.
The water sample passes through a homogenizer before it arrives at the
small (1.5-mm- [0.059-inch-] -diameter lines in the ultraviolet and TOC detec-
tors. The Raytheon homogenizer consists of a conical ceramic rotor with ad-
justable clearance between it and the ceramic stator allowing variable extent
of comminution of particles. Suspended solid material is reduced in size to
»100 um.
Detector Calibration
Calibration of the major detection components is accomplished by
24
-------
TABLE 6. HAZARDOUS MATERIALS DETECTABLE BY ULTRAVIOLET ABSORPTION
Aromatic Organic
Acenaphthene
Aldrin/Dieldrin
Aniline
Benzene
Benzidine
Benzoic acid
Benzonitrile
Benzoyl chloride
Benzyl chloride
Chlordane
Chlorinated benzenes
Chlorinated naphthalene
Chlorinated phenols
Cresol
Dichlorobenzedine
DDT
2,4-D (acid)
2,4-D (esters)
2,4,5,T-(acid)
2,4,5,T-(esters)
Dinitrobenzene
Dinitrophenol
Diphenylhydrazine
Endrin
Ethylbenzene
Fluoranthene
Guthion
Heptachlor
Kelthane
Aromatic Organic
Methoxychlor
Naphthalene
Nitrobenzene
Nitrophenols
Pentachlorophenol
Phenol
Phthalate esters
Polychlorinated biphenyls
Polynuclear aromatic hydrocarbons
Quinoline
Resorcinol
Styrene
Toluene
TCDD
Trichlorophenol
Xylene
Xylenol
Inorganic
Cobalt compounds
Copper compounds
Iron compounds
Nickel compounds
Sodium nitrite
Uranium compounds
No. of
Compounds*
4
8
9
5
* Number of compounds in class on Hazardous Substances List.
26
-------
following instructions in the individual operating manuals. To perform these
calibrations, the following solutions are required:
1. Commercially available buffer solutions at pH 4 and 7 to calibrate
the pH electrode.
2. Solutions of potassium chloride of concentrations that exhibit about
the same electrical conductivity as that of the waterway.
3. Potassium hydrogen phthalate, 10 ppm, to calibrate the uv absorpti-
meter. [The material(s) causing high uv absorption at 230 nm may
not be identified; this calibration therefore does not suffice to
assign a concentration to the spilled material.]
4. Potassium hydrogen phthalate, 375 ppm, to calibrate the TOG analyzer.
5. Zobell solution to calibrate the redox potential electrode. The com-
position of this solution is:
0.003 M potassium ferricyanide
0.003 M potassium ferrocyanide
0.1 M potassium chloride
The redox potential of this solution as measured by a platinum elec-
trode versus a saturated calomel electrode is given as:
E (observed) *• 0.185 + 0.00164 (25-t°C) (5).
Calibration of the detection components should be performed at least once
per month, and more frequently if a malfunction is suspected.
Other Materials Required
In addition to the calibration solutions, other materials required for the
operation of the warning system for the two-week period are:
1. Biocide for the uv rezero fluid - one liter, 300 ppm aliphatic
quaternary ammonium chloride.
2. TOG analyzer acid - three liters 0.5 N hydrochloric acid.
3. TOG analyzer compressed gas - 240 cubic feet (STP) oxygen to oxidize
the organic compounds in the water sample to carbon dioxide and
water.
27
-------
SECTION 4
SYSTEM TESTING
LABORATORY EVALUATION OF INDIVIDUAL DETECTION COMPONENTS
The major detection components were first placed into operation as they
were received and tested briefly to become familiar with their individual
characteristics. The Leeds & Northrup pH, ORP, and conductivity detection
components were treated as a single unit since their power supplies and re-
corder circuits were housed in a single cabinet.
Ultraviolet Absorptimeter
The Teledyne photometric analyzer was the first detection component to
be placed in operation. Its sensitivity to detect some of the substances on
the hazardous materials list in distilled water was determined using the 230-
nm measuring wavelength and the 390-nm reference wavelength with which the
instrument was delivered, and the 2.54-mm (0.1-inch) -thick optical cell.
Solutions of known concentration of various materials were prepared by dis-
solving weighed quantities of various solids in distilled water and making up
to volume in volumetric flasks.
Lower concentrations were prepared from these stock solutions by succes-
sive dilution. Flow of sample through the 2.54-mm (0.1-inch) absorption cell
was alternated with flow of distilled water, using a manual valving arrange-
ment. The instrument output was recorded on a Sargent Type SR strip-chart
recorder initially, and later on a Hewlett-Packard Model 7100B recorder. The
difference in signal between sample and distilled.water was plotted against
the sample concentration, and at low concentrations the plots were generally
linear. The results were expressed as a concentration at which the hazardous
material would produce an alarm signal of 0.5 mV.
For benzene, a solution was prepared in distilled water by injecting a
volume of benzene measured with a microliter syringe into water and making up
to a known volume. Various lower concentrations were prepared by dilution.
The benzene-in-water solutions were fed by gravity from a reservoir through
rubber and polypropylene tubing to the absorption cell. The benzene may have
been strongly sorbed by the tubing, so that the actual concentration flowing
through the uv absorption cell may have been less than in the reservoir.
The Aldrin tests were done by starting with a saturated solution of the
Technical grade insecticide in water, filtering the excess undissolved solid
and diluting the solution tenfold. The actual concentration of Aldrin, or in
28
-------
fact, the identity of the uv absorbing species in the test solution, is not
determined.
The results are listed in Table 7. Both the concentration required to
produce a signal of 0.5 mV and the sensitivity in mV/ppm are tabulated. The
results for benzene are subject to error because of likely loss in the tubing
connections between the sample reservoir and the uv absorption cell.
The results listed in Table 7 were obtained using distilled water as the
reference liquid. Natural water may very well exhibit some uv absorption it-
self at the 230-nm measuring wavelength. This natural absorption would act as
a background above which the spilled hazardous material would have to absorb
to provide an alarm signal. Incidentally, sodium nitrate, which is not a
hazardous material, apparently absorbs less strongly than sodium nitrite,
which is on the hazardous materials list.
The electronic stability was measured and found to be adequate. How-
ever, zero drift was observed as a result of condensation of moisture on the
windows. This problem was overcome by installing a muffin fan in the instru-
ment and providing a vent in the light source housing. Circulation of air
through the previous stagnant areas around the flow cell essentially prevented
moisture condensation.
TABLE 7. SENSITIVITY OF ULTRAVIOLET ABSORPTION SENSOR
FOR SOME HAZARDOUS MATERIALS
Concentration, ppm,
to produce 0.5 mV Sensitivity,
Material alarm signal mV/ppm
Potassium 0.4 1.25
hydrogen phthalate
Resorcinol
Benzene*
Sodium nitrate
Sodium nitrite
Copper sulfate
Aldrin
0.5
23
0.9
0.5
4.9
1/10 saturation
1.0
0.022
0.56
1.0
0.10
~—
* See text for possible explanation of benzene loss during tests.
29
-------
Malfunctions in the uv absorptimeter observed during the nearly two
years of intermittent operation included the following:
1. An operational amplifier (the Extended Voltage Amplifier) failed.
Satisfactory operation resulted when the component was replaced
with a standard off-the-shelf unit.
2. A solid-state voltage regulator that supplies the rest of the elec-
tronics with a stable +15 VDC was replaced.
3. The deuterium lamps required periodic replacement.
4. The entire instrument was serviced by a factory representative in
September 1977, before testing at the Los Angeles River. Several
components were upgraded during this procedure.
ORP, pH, Electrical Conductivity
The Leeds & Northrup module, consisting of pH, electrical conductivity,
and oxidation-reduction potential (ORP) sensors, together with the power
supply-recorder unit, was installed in a flow loop to permit evaluation of its
performance.
The flow loop employed the Prosser submersible pump immersed in a 208-
liter (55-gallon) drum to circulate about 180 liters of water at a rate of 30
liters per minute through the three housings which contained the individual
sensors. A means was also provided to isolate the sensing points from the
flow loop and to replace the fluid surrounding the sensors with a calibration
fluid.
The recorder was first calibrated utilizing electrical potentials simu-
lating outputs from the sensors. The sensor and measuring instrument-recorder
combinations were then further calibrated by comparing indicated instrument
output with the known values of standard solutions that had been checked with
laboratory instruments. The above three parameter sensing systems appeared to
be functioning properly.
The Leeds & Northrup recorder and the alarm level set points associated
with the ORP, pH, and conductivity sensors were checked and found functioning
properly. Measurements were made of the dead-band inherent in the mechanical
alarm level set points. A dead-band of between 1/3 and 1/2 of 1 percent of
full scale was measured, which is entirely acceptable.
A preliminary experiment simulating a spill of calcium hypochlorite was
made by suddenly adding one gram of Ca(C10)2 to 135 liters of tap water which
was being circulated through the sensing flow loop. This caused a change of
240 mV in the oxidation-reduction potential which was more than enough change
than required to give an alarm indication. Similarly, 25 cc of acetic acid
added to 135 liters of tap water changed the pH from 8.1 to 4.8, again enough
to activate the alarm relay for pH on the low side. The addition of 19 grains
of NaOH then caused a similar upper alarm signal to be given since the pH was
raised to about 9.2.
30
-------
The electrical conductivity was driven off scale on the high side after
addition of the sodium hydroxide, giving an additional alarm signal for high
conductivity.
Samples of the circulating fluid were taken at intervals and measured
with calibrated laboratory instruments. Good agreement was obtained between
all three process sensors and their respective laboratory counterparts.
The stability of the above three measurement systems with respect to
line voltage variations was also determined. The pH and ORP outputs were
found to be virtually independent of line voltage variations between 90 and
120 volts. The conductance was found to change about 2 percent of span in the
same direction as the line voltage over the above range of line voltages.
This stability was considered to be adequate when compared to conductivity
variations to be found in a natural watercourse.
The Leeds & Northrup instruments have required no repair since they were
placed in operation.
TOG Analyzer
The Dohrmann Model DC-60 Process Total Organic Carbon Analyzer (TOCA)
selected was a high-sensitivity unit equipped with two operating ranges, 0-10
and 0-500 mg carbon (C) per liter. Its performance was checked following the
operating instructions supplied with the instrument and found to be within
specifications.
To gain experience in the operation of the TOCA and to provide information
on the carbon content of in-house distilled and tap waters, distilled and tap
water streams with and without the addition of acid and subsequent sparging were
monitored by the TOCA. Results showed no inorganic carbon and very little or-
ganic carbon (considerably less than 1 mg C per liter sample) present in the
distilled water, confirming the belief that standards for calibration of the
TOCA could be formulated using the distilled water as received. Tap water ex-
hibited a much greater total carbon (TC) level than the distilled water and a
tenfold decrease to TOC when the sample stream was acidified and sparged.
The instrument was calibrated using a standard solution of potassium
hydrogen phthalate containing 7.5 mg C per liter. Solutions of several organic
compounds in distilled water were prepared containing 9.0 mg C per liter. The
results obtained in both the TC and TOC modes of operation of the instrument
are shown in Table 8. The results show that the aqueous solutions of chloro-
form and chlorobenzene produced extremely low TOC values. This effect appears
to be principally due Co the sparging (in the TOC mode) of the aqueous solu-
tions of these volatile, hydrophobic organic compounds. However, even without
the acid sparging, their respective TC values were somewhat lower than nominal.
Another interesting phenomenon was observed in the case of nitrobenzene,
i.e., both the TC and TOC values appeared to approach a steady-state value
asymptotically. This behavior is probably due to absorption of the nitro-
benzene by the Tygon tubing used for fluid transfer lines in the DC-60. Sup-
port of this belief was obtained subsequent to the nitrobenzene runs; about
31
-------
TABLE 8. CARBON CONTENTS OF AQUEOUS SOLUTIONS* OF SELECTED ORGANIC COMPOUNDS
AS MEASURED BY THE DOHRMANN DC-60 CARBON ANALYZER
Organic
Compounds
Acetic acid**
Chlorobenzene**
Chloroform**
Methanol
Nitrobenzene**
Pyrogallol***
Tannic acid***
TC
Average
8.9
3.0
4.3
9.4
7.1
8.9
7.8
Range
8.8-8.95
2.7-3.2
4.2-4.4
9.2-9.6
6.8-7.5
8.8-9.1
7.7-8.2
TOG
Average
8.6
0.1
0.3
8.7
5.3
8.6
7.7
Range
8.6-8.7
0.1-0.1
0.1-0.5
8.45-9.15
4.75-5.8
8.4-8.8
7.4-8.8
* All monitored solutions were formulated to a TOG of 9 mg C/l, and
all data are listed in units of mg C/l.
** On December 30, 1975 Hazardous Substances List.
*** On August 22, 1974 Hazardous Substances List.
a dozen distilled water injections were required before the TOC readout fell
to the nominal value of 0.1 mg C per liter for distilled water. Replacement
of the manufacturer-supplied Tygon tubing with Teflon tubing may eliminate
this absorption problem. The remaining organics produced TOC values close to
their nominal values.
When the TOGA was operated on an around-the-clock basis, with injection
of a water sample at 5-minute intervals, several minor problems were elimin-
ated one at a time by slight modifications in procedure and/or in the instru-
ment. Chloride deposition in the ir-sample cell was reduced/eliminated by
decreasing the concentration of hydrochloric acid used in the sparging process
from 3N to 0.5N. The plunger of the rotary injection valve tended to bind
over a period of time; its operation was greatly improved by replacing the
bare stainless steel plunger with a Teflon-coated one supplied by the manufac-
turer. The operation of the gas-liquid separator was improved and simplified
and pressure buildup prevented by changing the position of one of the lines
and leaving it open to the air.
Other malfunctions were observed during extended operation of the TOGA
and were corrected. Introduction of water samples containing sufficiently high
carbon contents to exceed the alarm levels did not always produce an alarm.
This malfunction was the result of defective or failed components on the clock/
32
-------
overrange circuit board. This board was upgraded, on being returned to the
, manufacturer, and placed back into service.
, Another difficulty manifested itself in the following symptom: a 375 mg
C per liter potassium hydrogen phthalate (KHP) solution registered a response
in the high ir range, "HI OPER" function position (although lower than usual),
yet a 7.5 mg C per liter KHP solution did not register a response in the low ir
range, "LO OPER" function position, with all other aspects of the operation
appearing normal. While tests were being conducted which showed that the ir
module was responsive to C02/N2 span gas in both range settings, it was ob-
served that the recorder pen responded (in the detector output position) by
first moving downscale before returning and moving upscale, the normal proce-
dure being to move directly upscale. It was this peculiarity which essentially
led to a definition of the problem and its ultimate resolution. The problem was
caused by an unbalance between the reference and sample cells in their module,
a situation that (according to Dohrmann engineers) occurs periodically and
which can be and was adjusted electronically.
The Homogenizer
Although it is not a detection component, one of the important parts of
the warning system is the Raytheon Model 2650 homogenizer. This device is
capable of reducing the size of particles from 6 mm (1/4 inch) to 100 to 1200
urn, depending on adjustments with the small throughputs necessary to feed the
uv and TOG sensors. (As shown in Figure 4, the nominal flowrate through the
homogenizer is 300 ml/minute, or about 4.8 gallons/hour — much smaller than
the minimum flowrate for which most on-stream homogenizers are designed.)
Two major problems were encountered with the homogenizer as received,
aside from relatively minor leaks that were eliminated by judicious use of
Teflon tape. The homogenizer was first installed and used in February 1976,
and the first major difficulty traceable to it was plugging of all downstream
lines in May 1976 after several continuous runs in the laboratory, the last
one of six days' duration. Despite efforts to continue the test by repeatedly
cleaning valves and lines to the uv and TOG, and reassembling the system, flow
below the homogenizer remained far less than nominal. When the homogenizer
was disassembled, the aluminum housing was found to be badly corroded. A sub-
stantial quantity of gelatinous material, apparently aluminum hydroxide, cov-
ered the inside surface of the housing. The presence of this material would
explain the observed impediment to flow.
After removing the gelatinous deposit on the aluminum housing, an attempt
was made to reduce the rate of corrosion by applying a water-repellent silicone
coating, Union Carbide L'-31. This coating was not sufficiently impervious and/
or adherent to prevent formation of more aluminum hydroxide during the next
continuity test. Accordingly, the aluminum housing was replaced by one fabri-
cated from stainless steel. This improvement entirely eliminated the line-
plugging due to corrosion products.
After a 19-day continuity run made in July 1976 with the stainless steel
housing, the homogenizer was again the cause for terminating the test. On this
occasion, the ceramic conical stator separated from its mount and came into
33
-------
contact with the rotor. Both rotor and stator were considerably abraded and
produced a quantity of fines that plugged the downstream lines. The homogen-
izer was then returned to the manufacturer for repair. Since its reinstalla-
tion in the warning system, the homogenizer has operated satisfactorily in
both laboratory and field tests.
LABORATORY EVALUATION OF INTEGRATED SYSTEM
The integrated system was evaluated in the laboratory by continuously
pumping a tap water sample from a 208-liter (55-gallon) drum through the en-
tire system using the Prosser pump immersed in water in the drum. Recirculat-
ing the same water resulted in a gradual but considerable temperature rise.
To prevent undue overheating, a portion of the water was withdrawn and re-
placed with fresh tap water. Only when a hazardous material was added to the
water in the drum was the partial replacement temporarily stopped to assess
the effect of the addition. Between additions of different hazardous mater-
ials, the rate of water replacement was increased to accelerate flushing one
material out of the system before testing the next. Usually, whatever alarm
condition was produced by the hazardous material was eliminated by flushing
before the next hazardous material was added.
During a period when the TOG Analyzer was not in operation, a number of
inorganic materials were added incrementally to about 180 liters (50 gallons)
of water in the drum. The response of the other four detection components is
shown in Table 9. For many of the inorganic materials tested, the increase of
electrical conductivity that they produce will serve as the means for their detec-
tion. However, several of the colored materials would be detected at lower concen-
tration levels by the differential uv absorption. This observation applies also
to colorless nitrates, such as lead nitrate and ammonium nitrate, which will be
detected more reliably by the uv device than the other three sensors.
The effect of ammonium dichromate on the uv device results from its
stronger absorption at the reference wavelength of 390 nm than at the measur-
ing wavelength of 230 nm. The measured differential absorption of the water,
therefore, falls in the presence of ammonium dichromate. Only 11 ppm of the
salt.(2 grams in 180 liters) is sufficient to produce a differential uv absorp-
tion that is negative under the stated conditions.
During these tests the alarm levels were set as follows: pH ±0.5, ORP
±10% of full scale, conductivity +10 ymho/cm, uv -HD.5 mV. Under these alarm
settings, all the inorganic materials tested would have produced an alarm from
one or more of the sensors at the 16 ppm level or less except zirconium oxy-
chloride. It should be noted that although these alarm settings are adequate
for the tap water in use, it was recognized that they would probably need to
be reset for a natural waterway. In particular, the electrical conductivity
alarm level setting would depend on the baseline conductivity of the water.
Organic and metal-organic compounds were tested, usually when the TOG
analyzer was operating, by the same incremental addition of weighed quantities
to 180 liters of recirculated tap water in the drum. The results are dis-
played in Table 10. For several common nonvolatile water-soluble organic com-
pounds such as acetic anhydride, benzoic acid, phenol, diethylamine, and
34
-------
TABLE 9. EFFECT OF SOME INORGANIC COMPOUNDS ON pll, ORP, CONDUCTIVITY AND UV ABSORPTION
u>
ui
Material
Tap water (180 liters)
Potassium permanganate
1 8
Tap water
Potassium permanganate
1 g
+ 1 g
+ 1 g
+ 1 g
Tap water
Cupric chloride
(anhydrous)
1 g
+ 1 g
+ 2 g
Tap water
Ammonium nitrate
1 g
+ 1 g
+ 1 g
+ 1 g
+ 1 g
pH
7.8
7.8
8.6
8.5
8.5
8.6
8.7
8.5
8.3
8.2
7.9
8.6
8.6
8.5
8.5
8.5
8.4
ORP
% Full Scale
64
64
64
62
61
60
60
65
66
66
67
67
60
59
58
57
55
Conductivity
umhos/cm
319
328
319
328
333
338
343
324
328
333
338
324
333
345
348
357
367
UV Absorption
% Full Scale
5.0
6.5
5.0
6.5
9.5
11.0
14.5
5.0
5.0
8.0
12.5
5.5
9.0
11.8
14.5
17.5
20.0
(continued)
-------
TABLE 9. (continued)
co
ON
Material
Tap water
Ammonium carbonate
1 g
+ 1 g
Tap water
Zirconium oxychloride
1 g
+ 1 g
Tap water
Ammonium chloride
1 g
+ 1 g
Tap water
Ammonium hydroxide
(concentrated) 1.75 ml
+ 1.75 ml
Ammonium dichromate
1 g
+ 1 g
Ammonium bifluoride
i e
+ i g
PH
8.4
8.5
8.6
8.2
8.3
8.3
8.6
8.5
8.5
8.6
9.0
9.2
8.4
8.2
8.2
8.0
ORP
% Full Scale
65
55
51
68
71
71
72
64
57
69
53
50
58
57
58
57
Conductivity
|jmhos/cm
324
328
343
324
328
328
328
343
357
324
338
346
333
343
333
348
UV Absorption
% Full Scale
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.1
5.1
5.0
5.0
5.0
4.0
< 0.0
5.0
5.0 (continued]
-------
TABLE 9. (continued)
Material
Ammonium persulfate
1 8
+ 1 g
Tap water
Aluminum fluoride
1 g
+ 1 g
Cupric nitrate
1 g
+ 1 g
Lead nitrate
1 g
+ 1 g
Sodium hydroxide
1 g
+ 1 g
Nitric acid (70%)
1.4 g
+ 1.4 g
Potassium hydroxide
1 g
+ 1 g
PH
8.4
8.4
8.5
8.5
8.5
8.4
8.1
8.4
8.3
9.1
9.3
8.3
8.0
9.2
9.5
ORP
% Full Scale
57
55
71
70
70
70
58
68
67
69
64
74
77
74
77
Conductivity
pmhos/cm
328
338
318
328
328
328
333
323
326
343
348
328
333
333
348
UV Absorption
% Full Scale
5.0
5.0
5.0
5.0
5.0
6.0
7.0
7.2
8.2
5.4
5.4
5.9
6.4
5.0
5.0 (continued
-------
TABLE 9. (continued)
ORP
Material pn % Full Scale
Nickel nitrate
1 g 8.4 71
+ 1 g 8.4 72
Phosphoric acid (85%)
1.2 g 8.0 75
+ 1.2 g 7.8 76
Conductivity
umhos/cm
338
348
333
338
UV Absorption
% Full Scale
5.4
5.4
5.0
5.0
u> UV absorption - full scale = 5.0 mV output.
CO
Alarm settings - pll +0.5, ORP +10% full scale, conductivity +10 innho/cin, UV +10% full scale.
-------
TABLE 10. (continued)
Concentration
Material
Benzole acid
Chlorobenzene
•
Phenol
Chloroform
Diethylamine
Para formaldehyde
Resorcinol
ppra
(Total)
5.5
11
6
12
270
6.6
12.1
8.2
16.4
400
3.8
7.6
5
10
65
5.5
11
ppm (C)
3.8
7.6
3.8
7.7
173
5.1
9.3
0.83
1.66
40.4
2.5
5.0
2
4
26
3.6
7.2
ApH
0
0
0
0
0
0
0
0
0
0
0.2
0.3
0
0
0
0
0
AORP
% Full Scale
0
0
2
2
2
0
-2
0
0
0
-9
-13
-2
-14
-22
-18
-36
AConductivity
|imhos/cm
0
0
5
5
5
0
0
0
0
0
5
10
5
5
5
0
0
AUV Absorption
% Full Scale
27
60
0
0
17
10
20
0
0
0
3
7
0
0
3
60
>100
TOG
% Full Scale
30
45
0
0
*
30
55
0
0
30
30
*
10
20
>100
20
50
Responses listed are changes from tap water baseline.
TOC full scale = 10 ppm carbon.
UV absorption full scale = 5 mV output.
* Sufficient time was not allowed for full response to develop before next addition.
** Benzene added in the form of one liter of saturated solution.
-------
resorcinol, the response of the TOG analyzer (calibrated to measure 10 ppm carbon
at full scale) was nearly quantitative. For volatile, water-insoluble com-
pounds, the response is noticeably less because the sparging treatment to which
the sample is subjected removes a considerable fraction of the material before
it is injected into the furnace. Even with this drawback, however, benzene
and chloroform do not escape detection when present in large concentration.
One facet of the laboratory evaluation consisted of running the entire
system continuously for as long as possible. A 19-day period of continuous
operation was achieved with only minor interruptions, which it should now be
possible to improve upon since various parts of the system have been upgraded.
This continuous run was terminated because of a serious failure of the homo-
genizer. The ceramic stator separated from its mount, fell against the rotor,
and a large quantity of abrasive fines was produced. The fines plugged the
downstream lines, necessitating dismantling and cleanout. The homogenizer was
returned to the manufacturer for repair and since its return it has operated
satisfactorily.
TRAILER MOUNTING
Following the laboratory evaluation phase, the components of the warning
system were mounted in an 8.2-m (27-foot) automotive trailer for increased
mobility. The custom-built trailer was provided with inlet and outlet for the
water sample, and large integral eye bolts to which tiedowns could be attached
for holding the major components in place during transport. The central (uv)
console is permanently in position in the trailer attached to the plastic in-
let and outlet lines. The control console is placed in the aft corner of the
trailer with Ethafoam padding protecting it from damage, and several tiedowns
keep the console in position. The Dohrmann DC-60 TOG Analyzer is placed on a
custom-built pad and also tied down for transport. Connections between the
TOG Analyzer and the rest of the system are undone to simplify moving from one
site to another. The mounting arrangements have been demonstrated to be ade-
quate for several short moves of the warning system from one site to another,
and for transcontinental shipment as well.
FIELD TESTING OF THE HAZARDOUS MATERIALS SPILL WARNING
SYSTEM AT THE ROCKETDYNE SANTA SUSANA FIELD LABORATORY
The hazardous material spill warning system trailer was moved to the
Rocketdyne Santa Susana Field Laboratory (SSFL) in March 1977. It was located
at the Alfa I Test Stand flame deflector channel, about 15 m (50 feet) from a
creek bed. The submersible pump was placed in the creek aftar first rechan-
neling the creek to accommodate it. Thirty in (100 feet) of 5-cm (2-inch) re-
inforced PVC tubing connected the pump to the trailer and an additional 30 m
(100 feet) of tubing returned the effluent to the creek downstream of the pump.
A right-angle, sharp-crested weir was installed 100 m (390 feet) upstream of
the pump to permit determination of the stream flowrate.
Water to the creek bed was provided by "natural runoff" from a set of
sprinklers hooked up to the SSFL water reclamation system. The water from the
sprinklers flowed approximately 100 m (390 feet) down a hillside containing
41
-------
vegetation indigenous to the area and then entered the stream 100 m (390 feet)
upstream of the weir.
The warning system equipment consoles and associated gear, which had been
anchored securely for the trip, were untied and reconnected in the usual oper-
ational configuration prior to engaging the system. In a logical, stepwise
fashion, the warning system modules and various subsystems were activated,
checked out, and placed on stream. With the SSFL water reclamation sprinklers
running and the stream flowing at flowrates up to 240 L (63 gallons) per min-
ute (measured at the weir), the submersible pump was engaged and system main-
stream flow was obtained. Mainstream flow, which is monitored by ORP, pH, and
conductivity sensors, was held at 20 to 22 L (5.3 to 5.8 gallons) per minute.
Flow to the uv absorptimeter and the TOGA was held at 1.1 L (.29 gallon) per
minute and was obtained by opening the valves to the small-bore tubing after
the homogenizer was engaged.
With flow in these small-bore lines and with the TOGA furnace cooling
water pump operating, the uv absorptimeter and the TOCA were turned on. The
respective readouts appeared normal, except for the conductivity, which was
off scale, apparently because of the high specific conductivity of the stream.
The original conductivity cell, therefore, was replaced with a higher cell-
constant conductivity cell.
All of the sensor alarm cams and alarm pots required readjustment from
the laboratory tap water settings to the new "natural" waterway conditions.
The only sensors that were calibrated in situ with standard solutions during
•field trials at SSFL were the conductivity, TOC, and uv absorptimeter. The
reason for this was that only these sensors had been modified after all the
sensors had been originally calibrated in the laboratory using standard solu-
tions. (The conductivity system was modified by the exchange of conductivity
cells, the TOCA by its being used in the "HI OPER" range position rather than
the original "LO OPER" mode, and the uv absorptimeter by the replacement of a
deuterium lamp, cleaning of cell windows, and the adjusting of circuit board
pots.)
The conductivity system was calibrated using KC1 solutions of selected
specific conductivity. Solutions of 0.01 and 0.02N KC1 were prepared and,
with specific conductivities at 25°C of 1413 and 2765 umhos/cm, respectively,
produced a calibration constant of 57.5 ymhos/cm/percent full scale with the
new, higher cell-constant conductivity cell in the system.
The TOCA was calibrated using a 375-ppm C:KHP solution. The usual proce-
dure involves adjusting a calibration circuit pot so that the TOCA readout for
this solution reads 75'percent of full scale. Thus, a full-scale deflection
would read 500 ppm. As a result of the calibration, the background reading
fell from 30 to 20 ppm carbon.
The uv absorptimeter was calibrated using KHP solutions which were passed
through the uv cell and the response measured on the uv recorder. A 50-ppm
KHP solution was higher than the full-scale reading, while a 25-ppm solution
was close to 50 percent full scale.
42
-------
With the entire warning system in operation, spills of selected hazard-
ous chemicals were conducted. Spills of resorcinol, hydroxylamine hydrochlor-
ide (HAHC), and aluminum sulfate were performed by adding usually 28.4 g of a
chemical sample to 500 ml of stream water, stirring the mixture, and then
dumping it into the stream approximately 50 m (165 feet) upstream of the sub-
mersible pump. The measured changes in response of the various sensors to the
spills of these hazardous materials in the stream are listed in Table 11.
(Baseline values for each sensor just before the spill are indicated by the
value after the slash [/].)
With a weir-measured flowrate of approximately 260 L (70 gallons) per
minute, it took 10 minutes for the warning system to first sense the resor-
cinol spills. Both the uv absorptimeter and the TOGA alarmed with resorcinol.
With a weir-measured flowrate of 180 L (48 gallons) per minute, the first re-
sponse of the warning system to the HAHC and the EDTA occurred after 15 min-
utes, although the EDTA did not alarm for another 15 minutes. The HAHC
alarmed the ORP and affected the pH, while the EDTA alarmed the TOCA. Alumi-
num sulfate alarmed the pH after 10 minutes at a weir flowrate of 220 L (58
gallons) per minute. In general, the data show that the hazardous materials
spill warning system has the capability of responding to a wide variety of
compounds within a reasonable period of time. This timely behavior allows the
grabbing of a sample of stream water at the time of the alarm so that a repre-
sentative sample will be available for later chemanalysis.
During the time that the warning system was being field tested at SSFL,
a number of improvements were made to the system. The telephone dialer was
received and installed in the trailer. It was hooked up to the warning system
electronics through two relays connected individually to the first and the
second spill warning circuits. The telephone dialer tape was then programmed
so that channel A described the first spill and channel B the second.
A Phone-Mate telephone recording device was received and connected to a
dedicated telephone line. This device provides the warning system with the
capability for taping any incoming telephone-dialer messages so that the oper-
ator does not have to be present at a specific telephone station to receive a
warning system phone call. In addition, the Phone-Mate has a remote playback
feature which permits the operator to call the unit from any telephone and
interrogate the device to determine if a spill has occurred. The telephone
dialer, coupled with the telephone recorder, provides the hazardous spills
warning system with an unparalleled ability for unattended operation.
Further, a gel-cell battery was connected to the telephone dialer as
part of a fail-safe capability built into the system. If the 110-volt power
fails after an alarm is- tripped or if power interruptions occur, the dialer
can still operate up to 6 hours on the battery charge.
To prevent the TOCA furnace cooling water pump from cavitating, the
plumbing line at the base of the original 1-inch standpipe, which is the water
reservoir for the cooling subsystem, was rerouted. The water line now is
shorter with fewer bends and does not cross directly over the warm pump motor.
Also, the 1-inch-diameter standpipe was replaced with a 1%-inch one, and this
greatly improved the cooling capacity of the subsystem.
43
-------
TABLE 11. CHANGE IN RESPONSE DUE TO STREAM SPILLS OF HAZARDOUS MATERIALS
SPILL
MATERIAL
Resorcinol
Resorcinol
Resorcinol
HAIIC
EDTA
Resorcinol
Resorcinol
Resorcinol
Resorcinol
Al Sulfate
EDTA
EOTA
Al Sulfatet
Resorcinol
WEIR
FLOW-
RATE
(L/min)
260
260
260
180
180
180
60
60
60
260
260
220
220
220
TIME FOR
SYSTEM
TO ALARM
(min)
_
10
10
15
30
15
25
25
25
-
-
25
10
10
ORP
(* F.S.)t
14/14
13/13
12/13
6/21
9/9
17/17
8/8
8/8
8/8
27/28
28/29
29/28
26/28
25/27
pll
(pll units)
8.7/8.6
8.8/8.7
8.8/fl.7
7.5/7.9
7.9/8.0
8.1/8.1
7.9/7.9
8.0/8.0
8.0/8.0
7.0/7.4
7.2/7.3
7.4/7.5
6.6/7.6
8.1/8.0
CONDUCTIVITY
(« F.S.)t
12/12
12/12
12/12
12/12
12/12
12/12
12/12
12/12
12/12
10/10
11/11
11/11
11/11
11/11
U.V.
(% F.S.)t
100/6
100/10
100/15
4/4
6/6
100/28
>8Q/14
100/14
100/10
0/0
1/0
0/0
23/0**
100/1
TOC
(% F.S.)t
5.5/3
5.5/3
8.5/3.5
2.5/2.5
4/2.5
2.5/2
5.5/3.5
6/3.5
6/3.5
3/3
4/3*
4/3*
4/3
6/3
All spill samples were prefwred in a concentration of 28.4 g per 500 nil of stream water (except t which was
prepared in a concentration of 142 g per 500 ml of stream water) and then the mixtures were dumped into the
stream 50 m upstream of the submersible puinp.
•These changes in TOC value were small because of extremely low TOCA sample flow due to plugging of the
first isokinetic flow splitter.
"This response (s believed to have been due to some foreign nutter which had entered the stream; this
can happen since the stream was uncontrolled.
t t F.S. = % Full Scale.
-------
Although a full two-week unattended test run was not actually achieved
at the SSFL test site, much valuable information on the practical aspects of
operating the system was obtained. This knowledge was collected from testing
under conditions considerably more severe and demanding than the system would
normally encounter in practice. Since the trailer was located about 100 m
(330 feet) from an installation on which rocket engines were periodically test-
fired, the warning system was exposed to much higher levels of mechanical shock
and vibration than it would normally have to withstand. The system was essen-
tially unaffected by these high vibration levels.
The test firings were preceded and followed by water purges, which in-
troduced much larger volumes of water than usual, upstream of the sampling
pump. This large flow loosened debris and silt and piled it around the pump,
necessitating temporary removal of the pump and cleaning of the immediate area
to maintain proper flow through the warning system. More important, these
purges also carried quantities of organic solvent or hydrocarbon fuel. The
petroleum fuel was detected by the uv absorptimeter when it was in operation
immediately after a purge. This observation constituted the first detection
of a spill of a hazardous material by the warning system.
Since each engine firing included several test stand water purges, the
stream background readings were continually and dramatically changing on those
days when firings were scheduled. For this reason, either an operator would
have to be continuously on hand at the trailer to reset the warning system so
that it could continue operating uninterruptedly, or the system would have to
be placed on standby until the stream perturbation had abated. A number of
times the latter course was selected as the more desirable and it was for this
reason, especially, that a series of short test periods were obtained rather
than the desired uninterrupted two-week test. Of course, hazardous chemical
spills probably would not occur in actual practice with such a frequency as
they did during the SSFL test trials and, thus, the degree of attention demand-
ed by the warning system in actual field use would be reduced considerably.
FIELD TESTING ON THE LOS ANGELES RIVER
With the cooperation and assistance of the Los Angeles County Flood Con-
trol District, Water Quality Section, possible sites for testing the warning
system on the Los Angeles River were investigated. Of several possible loca-
tions, the one selected (Dominguez pumphouse) provided the best security and a
convenient source of power. The trailer was inside a doubly fenced area out
of sight of most individuals who might use the flood control channel for rec-
reational purposes, such as bicycling. It was about 80 m (270 feet) from, and
10 m (30 feet) above the watercourse in the middle of the concrete-lined
channel.
The depth of the water, about 12 cm (5 inches), was insufficient to cover
the submersible pump completely when it lay horizontally in the river. A first
attempt to run the warning system under these conditions produced too little
flow and pressure to fill the uv cell. Only the Leeds & Northrup module could
provide any data. The water characteristics were: pH=r9.0, conductivity» 1490
to 1610 umho/cm, ORP =• 4 percent of full scale. The conductivity was higher
than that required to produce an alarm at the preset value of 20 percent full
scale.
45
-------
The following week, the area of the watercourse was sandbagged just
downstream of the pump so that the effective water depth in the region of the
pump was increased to 18 cm (7 inches). This change produced a. nominal water
flowrate and pressure at the inlet to the warning system in the trailer,
despite the horizontal and vertical distance between the stream and the in-
struments. The uv absorptimeter was started up after adequate flow was es-
tablished. The conductivity alarm was reset to 35 percent of full scale (ap-
proximately 2000 ymho/cm) and the system was allowed to monitor the stream.
After about 1% hours of stable operation with pH at 9.0, ORP at 2 to 4 percent
full scale, conductivity at 1670 umho/cm, and uv absorption at 0, uv absorption
increased rapidly and climbed off scale. This high value of uv absorption was
exhibited for more than 40 minutes. During this period, the rezero function
operated automatically with water from the river placed in the rezero reser-
voir before the highly absorbing material passed the sampling site. When the
high uv absorption condition persisted, the system was also rezeroed manually
to check that the instrument was operating properly. After staying off scale
for 45 minutes, the uv trace came back on scale and returned to zero within 5
minutes, showing definitely that a highly absorbing material had passed the
sampling point.
The high uv absorption was not accompanied by any significant change in
pH, ORP, or conductivity. The TOG content of the grab sample collected was
not significantly different from the baseline. However, the uv absorptimeter
is more sensitive to certain organic compounds than the TOGA. Therefore, it
is certainly possible that some organic compounds, to which the uv analyzer
responds strongly, may not affect the TOG noticeably.
Additional field tests were performed by Mason & Hanger, contractor for
the EEE.U. These tests are discussed in Appendix C.
46
-------
REFERENCES
1. Shackelford, J., and L. Keith. Frequency of Organic Compounds Identified
in Water. EPA-600/476-062, U.S. Environmental Protection Agency, Cincin-
nati, Ohio, 1976.
2. Ciaccio, L.L., et al. Water and Water Pollution Handbook, Vol. 4. Marcel
Dekker, Inc., New York, New York, 1973. p. 1488.
3. Lange, N.A. Handbook of Chemistry. McGraw-Hill Book Co., Inc., New York,
New York, 1961. 1969 pp.
4. Camp, T.R. Water and Its Impurities. Reinhold Publishing Co., New York,
New York, 1963. p. 132.
5. Langmuir, D. E^-pH Determination. Procedures in Sedimentary Petrology,
Robert E. Carver, ed. Wiley-Interscience, New York, New York, 1971. p. 617.
47
-------
APPENDIX A
INSTRUMENT MANUFACTURERS
At the beginning of the program, when selection of particular instru-
ments had to be started, there was a variety of process-type pH, ORP, and con-
ductivity instruments commercially available. The selection of the Leeds &
Northrup instruments was somewhat arbitrary, but conditioned in part by their
time-honored reputation in the field of process instruments and the availabil-
ity of recorders with chart speed that could be increased when a set point was
exceeded.
The possible selection of ultraviolet absorptimeter was limited to the
two commercially available instruments made by duPont or Teledyne. Of these,
the built-in electronic rezero and the possibility of hourly flushing of the
uv cell, along with the single beam feature, argued strongly for the Teledyne
General Purpose Analyzer. The particular model selected, 611DS2, incorporates
a deuterium light source (which makes possible selection of a wider variety of
measuring and reference wavelengths than the more common mercury arc source).
Reference and measuring wavelengths actually in use are 390 nm and 230 nm
respectively.
The most difficult detection component to select was the Total Organic
Carbon Analyzer. No completely satisfactory instrument was available. On-
line instruments were claimed to be available by Astro Ecology, Delta, Dohr-
mann and Raytheon. The Astro and Raytheon instruments were continuous instru-
ments. However, both were subject to substantial zero drift that appeared to
make them unsuitable for operation for an unattended period of two weeks. The
Delta instrument appeared not to be quite ready for delivery within the re-
quired time frame. Hence, the Dohrmann Model DC-60 was chosen for the instru-
ment package. Its major shortcoming is that it operates semicontinuously; a
sample is injected into the furnace no more frequently than once every 5 min-
utes. This discontinuous operation, however, leads to certain advantages.
Fouling of the furnace may be slower than with the continuously operating in-
struments. In addition,, the Dohrmann instrument employs a more stable nondis-
persive IR unit for the measurement of the carbon dioxide produced than the
other available furnace-type TOG analyzers.
A partial list of manufacturers contacted appears in Table A-l.
48
-------
TABLE A-l. PARTIAL LIST OF INSTRUMENT MANUFACTURERS CONSULTED
Aquatronics Inc.
4th 5 Cumberland Streets
Philadelphia, PA 19133
(215)739-0695
Astro Ecology Corporation
P.O. Box 53159
Houston, TX 77053
(713)332-2484
Balsbaugh Laboratories Inc.
25 Industrial Park Road
So. Hingham, MA 02043
(617)749-1360
Beckman Instruments Inc.
2500 Harbor Boulevard
Fullerton, CA 92634
(714)871-4843
Bendix Corporation
Environmental Science Division
1400 Taylor Avenue
Baltimore, MD 21204
(301)825-5200
Buchler Instruments Division
1327 Sixteenth Street
Fort Lee, NJ 07024
(201)224-3333
Chondra Tech Inc.
1053 Shary Circle
Concord, CA 94518
(415)825-4533
Delta Scientific Corp.
120 E. Hoffman Ave.
Lindenhurst, NY 11757
(516)884-4422
Dohrmann Division-Envirotech
3240 Scott Boulevard
Santa Clara, CA 95050
(408)249-6000
E. I. DuPont de Nemours § Co.
Instrument Products Division
Monrovia, CA 91016
(213)357-2111
Ecologic Instrument Corp.
132 Wilbur Place
Bohemia, NY 11716
(516)567-9000
Enviro Control Inc.
960 Thompson Avenue
Rockville, MD 20852
(301)881-2660
Gil son Medical Electronics
3000 West Beltline
Middleton, WI 53562
(608)836-1551
Great Lakes Instruments Inc.
7552 North Teutonia Avenue
Milwaukee, WI 53209
(414)351-1250
Hach Chemical Company
P.O. Box 907
Ames, IA 50010
(515)232-2533
Honeywell Inc.
6620 Telegraph Road
Los Angeles, CA 90040
(213)723-6611
(continued)
49
-------
TABLE A-l (continued)
Inter Ocean Systems Inc.
3510 Kurtz St.
San Diego, CA 92110
(714)299-4500
Ionics Instrument Division
65 Grove Street
Watertown, MA 02172
(617)926-2500
Instrumentation Specialties Company
P.O. Box 5547
Lincoln, NE 68505
(402)464-0231
Kernco Instruments Inc.
19 Walt Whitman Road
Huntington Station, NY 11746
(516)427-4354
Leeds and Northrup
1360 S. Anaheim Blvd., Ste 225
Anaheim, CA 92805
(213)924-1647
Martek Instruments Inc.
879 West 16th Street
Newport Beach, CA 92660
(714)645-1170
Montedoro-Whitney Corporation
P.O. Box 1401
San Luis Obispo, CA 93401
(805)543-7337
Raytheon Company
Submarine Signal Division
West Main Street
Portsmouth, RI 02871
(401)847-8000
Technicon Industrial Systems
840 Tenth Street
Manhattan Beach, CA 90266
(213)374-5528
Teledyne Analytical Instruments
333 W. Mission Dr.
San Gabriel, CA 91776
(213)283-7181
Uniloc, Inc.
17401 Armstrong Ave.
Santa Ana, CA 92705
(714)546-8700
Wallace and Tieraan Division
25 Main Street
Belleville, NJ 07109
(201)759-8000
Wesmar Industrial Systems Division
905 Dexter Avenue North
Seattle, WA 98109
(206)285-2420
Weston and Stack Inc.
446 Lancaster Ave.
Malvern, PA 19355
(215)647-2400
50
-------
APPENDIX B
OPEEATING INSTRUCTIONS
INSTALLATION AND INITIAL STARTUP
Check all systems for damage. Lower the trailer jacks and block chem so
the trailer will be essentially level. Care should be taken when untying the
instrument consoles. Remove the tie-downs which hold the consoles in place
and put them out of the way in a drawer. Arrange both the control console
and the total organic carbon analyzer (TOGA) so they sit side by side with,
and face the same direction as, the uv absorptimeter/flow loop console. The
uv console remains in place facing toward the trailer doors. (Be careful dur-
ing arrangement of the consoles not to put any tension on electrical wiring
connecting the control console and the uv console.) Anchor the TOGA by lower-
ing the front support bolts.
Additional Equipment Required (to be supplied by user)
1. A 110-volt, 60-cps line capable of delivering 60 amps should be
connected to the trailer if there is sufficient power available
at the site. As an alternative system, the portable generator
installed in the trailer would provide up to 12 hours of opera-
tion at nominal power levels with one filling of the gasoline tank
(50 L).
2. A phone line will have to be installed for the telephone dialer.
(It is important that the operator reprogram the telephone dialer
tape to correspond to the new circumstances under which the warning
system will be used.)
3. A source of potable water should be available for the personnel
using the trailer.
4. The trailer's toilet facilities should be hooked up to onsite dis-
posal facilities. An alternative would be to provide a chemical or
other toilet separate from the trailer.
5. A supply of chemicals and reagents should be on hand, including dis-
tilled water, potassium acid phthalate (KHP), Zobell solution, hydro-
chloric acid (HC1), and a biocide.
6. Compressed gases, including oxygen, and 400 ppm C02 in N2, and
appropriate pressure regulators are also required.
51
-------
7. Chart paper for the two L&N recorders and the Dohrmann TOGA recorder
should be provided.
8. A dedicated telephone jack for the automatic telephone answerer must
be installed at the receiving station.
Service Connections
All instrument console main power switches should be turned off before
any power cords are connected.
Submersible Pump/Main Flow
1. Place the submersible pump in the stream. As long as it is com-
pletely submerged, it does not matter if it is positioned in a ver-
tical or horizontal attitude.
2. Connect a suitable length of 5-cm (2-inch) PVC tubing to the sub-
mersible pump and attach it to the main flow inlet to the trailer.
Likewise, attach a similar length of tubing to the trailer main flow
outlet and run it to a point downstream of the pump.
3. Uncoil suitable lengths of electrical extension cord to reach from
the trailer to the pump and attach to the trailer through outside
demand-pump hatch.
4. After checking the main flow valve (A)* and opening it a little,
open valve (2)* half way and plug in the power cord to the submers-
ible pump. Observe main pressure gauge (B) while opening valve (A)
slowly. Bring pressure up to 20 psig and adjust valve (3) so that
main flow measured by flowmeter (9) is 21 L (5.5 gallons) per min-
ute. Check to see that flow is properly exiting return line.
Control Console
Place the first and second spill switches in the "not ready" position,
the chart speed switch in the "low" position, and the counter switch "off."
(The main power switch to the uv module located on the front panel of the uv
control box should be off.) Unplug homogenizer (22) cord from electrical
plug-in strip in the rear of the control console.
1. Attach TOGA alarm cable to labelled terminal strip in rear of con-
trol console. Connect the wire lugs to terminal positions 4 and 18,
numbered from" left to right.
2. Plug main control console power cord into nearest wall outlet. The
"not-ready" lights should come on.
3. Remove polyethylene sheet pen protectors which have been inserted
between pens and chart paper in both L&N recorders.
* Numbers and letters in parentheses refer to items so identified in Fig. 4.
52
-------
Turn on each L&N recorder. The switch can be reached by opening
recorder door and depressing latch on right front of recorder while
carefully pulling chassis out. Do not pull chassis completely out'.
The power switch is located on the right side toward rear.
TOGA
All external connections are located at the rear of the instrument (41).
Make sure master power switch and pump module switch are off.
1. Hook up the TOGA cooling water system located in the rear of the uv
console by connecting to the TOCA the two copper lines attached to
the two diagonal angle-iron support struts of the uv console. The
strut nearest the TOCA contains the cooling water return line and
should be connected to the "cooling water out" fitting; the strut
farthest from the TOCA contains the cooling water supply line and
should be connected to the "cooling water in" fitting. Top off
water in TOCA cooling water stand pipe (15) located behind uv ab-
sorptimeter front panel.
2. Plug the TOCA main power cord and the TOCA cooling water pump power
cord to their nearest respective wall outlet. Check for water flow
and leaks in the TOCA cooling water system.
CAUTION
With TOCA main power cord plugged in and
TOCA master power switch off, some terminals
within the TOCA are HOT! Use care.
3. Turn on TOCA master power switch. Red light should come on. Set
cycle interval for 3 hours. (Since the TOCA furnace will be warm-
ing up, the operator can periodically monitor furnace temperature
by observing meter on TOCA front panel while alternating readout
between upper and lower furnace zones by means of front panel
switch.)
4. Attach pressure regulator (45) to oxygen K-bottle (44) and then at-
tach copper line from reactant gas fitting at rear of TOCA to the
regulator. With regulator set for zero delivery pressure, open main
cylinder valve. Set regulator pressure for 40 psig. Observe pres-
sure of 1 to 2 psi on "reaction zone pressure" gauge on pump module
front panel. Reactant flowmeter should read approximately 40 mm.
Check for proper gas flow through the system by removing right side
panel from TOCA and inserting the free end of Tygon tubing from ir
analyzer gas outlet into a container with a few centimeters of water.
A steady stream of gas bubbles should emerge from the tube.
5. Prepare a 0.5 N HC1 solution and place in the 3.8-L (1-gallon) plas-
tic container (43) behind the TOCA. Hook up 3-mm (1/8-inch) Tygon
line with special connector to acid inlet at the rear of the TOCA
for delivery of acid.
53
-------
6. Remove air compressor (48) from cabinet and attach air line to rear
of TOGA at fitting labelled "compressed air." Plug compressor
power cord into appropriate outlet at rear of TOCA. Observe air
pressure gauge on pump module front panel to see that pressure holds
between 30-40 psig. A pressure relief valve on compressor will re-
lieve pressure above that setting. Verify a setting of approximately
20mm on the sparge air flowmeter on pump module front panel.
7. Attach copper sample flow line from TOCA overflow device (40) to
rear TOCA "sample inlet" connection. The return line consisting of
Tygon tubing attaches to a fitting so labelled in the nearest diag-
onal support strut of the uv console.
UV Absorptimeter Console (Flow Loop)
1. Prop end of absorptimeter ozone transfer tube (exiting right-hand
box on front panel) between trailer window and outside screen.
2. Turn on uv absorptimeter master power switch. (This will probably
cause a uv alarm signal on control console. Ignore for the time
being.) Continue turning switch to start position and release.
Being spring-loaded, it will return to "on" position. Check to see
if deuterium arc lamp is lit by briefly looking (while wearing
plastic safety glasses) through the ozone transfer tube at the top
of the right-hand box.
3. Open valve (39) to the TOCA overflow device (40).
4. Plug in homogenizer power cord in rear of main power console.
5. Observe flow to the TOCA overflow device and, thence, to funnel (38)
and sump (49). Check to see that float valve (50) is operating
properly and water is leaving sump to the eductor (51).
6. Open valve (24) and observe flow to the uv absorption cell (29) by
watching outflow to funnel (38).
7. Plug power cord of stirrer (36) for uv rezero fluid reservoir (33)
into wall outlet. Leave stirrer switch in off position.
8. Open valve (31) about one and one-half turns.
OPERATION
It is expected that all sensors have been calibrated in accordance with
instructions in their individual operating manuals. All alarm settings have
been selected based on information developed from field trials conducted by
Rocketdyne. However, due to the likelihood of variations in background from
one location to another, adjustment of the alarm settings from nominal may be
necessary. ORP, pH, and conductivity alarms are adjustable at the cams lo-
cated within the three-pen L&N recorder. The uv absorption alarm level is
controlled by an adjustment pot on the front panel of the uv control box,
54
-------
while the TOGA alarm is set by an alarm level pot located within the TOCA
programmer/integrator module.
1. Switch on stirrer (36). Flip counter switch on front panel of con-
trol console to "on" position. Push and release "start" button and
then push and release "advance" button until counter advances from
24 to 0. At this point the uv rezero fluid reservoir will star-
emptying. At appropriate times preset on the respective timers,
the reservoir will stop emptying, start filling, add disinfectant,
and stop filling. Upon completion of this cycle, the uv absorpti-
meter will be ready for operation. It will now automatically re-
zero its baseline once an hour against the reservoir fluid back-
ground.
2. Remove right side panel from TOCA. Attach jumper cable between the
two wires on top of pump module housing near front of console. This
step bypasses the pump module flow interruption, shutdown circuit.
Flip pump switch (located on front panel) to "on" position. If
flowmeter lights and pumps do not go on, momentarily depress red
reset button located just below jumper connections on pump module
housing. Observe movement of bubbles through pump tubing and ad-
vance of acid solution through tubing to air sparger in rear of
pump module. When no further flow discontinuities are observed
(usually about 5 minutes after starting pumps), remove jumper cable
and replace right side panel to TOCA.
3. Turn TOCA cycle interval switch to 5-minute injection cycle. Set
TOCA function switch (pull out programmer/integrator module) to "HI
OPER" position and push "program start" button momentarily. Ob-
serve that the TOCA rotary injector valve is now following a pro-
grammed sequence of commands involving injection of a sample from
the flow stream into the TOCA furnace. The TOCA will now operate
automatically with injections and readouts occurring once every 5
minutes.
4. Check grab sample bottles (19) and (21) to see that they are empty.
Open valve (17) approximately six turns. Check control console to
ensure that no alarm condition exists. Push the reset buttons on
the alarm panel of the control console and throw the first and sec-
ond spill switches from "not ready" to "ready." The warning system
will now automatically record the responses of the sensors, close
alarm circuits if preset alarm levels are exceeded, and grab a sam-
ple of the stream fluid if an alarm occurs. (Up to two samples can
be grabbed, if there is a discontinuity in the alarm signal after a
preset period of time. Such a condition could occur in the event
of two separate hazardous material spills, unless the first spill is
that of an organic material.)
5. Plug telephone dialer transformer into wall outlet. Open lid to
telephone dialer and turn switch to "on" position. The warning
system is now ready to automatically call any number(s) and mes-
sage (s) that have been preprogrammed on the cartridge tape.
55
-------
6. Connect the telephone recording device to a phone pickup. The de-
vice can now receive and record automatically any incoming phone
messages, such as those from the telephone dialer. Also, the re-
mote playback feature allows the operator to interrogate the device
from another phone.
SYSTEM SHUTDOWN
1. Turn off telephone dialer.
2. Set TOGA cycle interval switch for 3 hours. Turn off TOGA pump
module and TOGA master power switches. (Again, the operator can
periodically monitor the furnace temperature by observing meter on
TOGA front panel.) Unplug air compressor power cord at rear of
TOGA. Open the air compressor pressure relief valve by pulling out
the spring-loaded red knob for a few seconds until the pressure
bleeds down completely.
3. Turn off uv master power switch on front panel of uv control box.
Check to see that deuterium lamp is no longer lit.
4. Close valves (39) and (31).
5. Unplug main control console power cord. This will turn off the
homogenizer and L&N recorders.
6. Close valve (24). Observe that water flows into funnel (38) have
ceased.
7. Insert pen protectors between pens and chart paper in both L&N
recorders.
8. Switch off stirrer (36).
9. When the TOGA furnace temperature falls below 400° C the TOGA cool-
ing water pump can be unplugged.
10. Turn off oxygen regulator valve.
11. Unplug electrical extension cord to the submersible pump and then
close main flow valve (A).
12. Unplug TOGA power cord. This completes system shutdown.
56
-------
APPENDIX C
SPILL ALASM SYSTEM REPORT
ENVIRONMENTAL EMERGENCY RESPONSE UNIT
Mason & Hanger - Silas Mason Co., Inc.
Edison, New Jersey 08837
Contract No. 68-03-2647
ABSTRACT
The Spill Alarm System was transferred to the Environmental Emergency
Response Unit in November 1977. The initial shakedown consisted of minor re-
pairs and the checkout of individual components. Short-term tests were con-
ducted using specific materials to produce response and alarm from the detection
components. A long-term test was set up on Mill Brook at a site adjacent to the
Kin-Buc landfill in Edison, New Jersey. The Spill Alarm System was successfully
used for emergency field response at Pittston, Pennsylvania, from October 19,
1979, to November 16, 1979.
This report covers a period from January 1978 to and including March 1981.
BACKGROUND
The Spill Alarm System was developed for the U.S. Environmental Protection
Agency, Oil and Hazardous Materials Spills Branch, Edison, New Jersey, by the
Rocketdyne Division, Rockwell International, Canoga Park, California, under
Contract No. 68-03-2080. The system was designed to detect spills of hazardous
materials in natural waterways.
The instrument package is installed in a modified air-conditioned 8.2-
meter (27-foot) camping trailer. The detection components that have been in-
corporated into the instrument package are a pH meter, oxidation-reduction
potential cell, electrical conductivity cell, ultraviolet adsorptimeter, and
total organic carbon analyzer. A strip chart recorder channel with an adjust-
able alarm sensor is associated with each detection component. Upon detection
of a concentration of hazardous material exceeding the alarm level of any of
the detection components, an alarm indication is sent via automatic telephone
equipment to a designated telephone where the alarm indication is recorded.
SHAKEDOWN
The Spill Alarm Trailer was first tested by the Environmental Emergency
Response Unit (EERU) at the U.S. Environmental Protection Agency facility in
57
-------
Edison, New Jersey. The system was set up for short-term testing by recirculat-
ing water from a 208-liter (55-gallon) container. Tests of the individual de-
tection components were made by the addition of known amounts of hazardous mate-
rials into the water stream. This method enabled the operator to establish a
background zero and note response and alarm of each detection component.
Several problems were encountered during the checkout of the detection
components. The peristaltic pumps in the Total Organic Carbon Analyzer malfunc-
tioned and were replaced. An examination of the pumps revealed the lubricant
in the gear box had hardened and overheating was suspected of causing the prob-
lem. A cooling fan was installed in the unit to improve reliability and pre-
vent overheating of the pumps. Plugging of small lines and tubing caused inter-
mittent flow problems, especially in the sample feed to the ultraviolet absorp-
timeter cell. Air pressure was used to blow out the plugged line and small
pieces of plastic were found to be the cause. After replacing most of the small
tubing, the problem was traced to a section of vinyl-lined stainless steel tub-
ing. Erratic operation of the Total Organic Carbon Analyzer was traced to a de-
fective pyrolysis tube requiring replacement of the heating unit and thermo-
couples. On June 2 the initial checkout was completed and the system was oper-
ating satisfactorily. The system was subsequently operated for approximately
2 months with a minimum of attendance.
FIELD TEST
On August 18, 1978, the Spill Alarm System was moved to a site on Mill
Brook adjacent to the Kin-Buc landfill in Edison, New Jersey. The system was
set up approximately 60 meters (195 feet) from the stream bed. A submersible
pump with 2.5-cm (1-inch) vinyl tubing was used to pump the sample water to the
Spill Alarm System and return to the stream. Numerous problems were encountered
during the 4 months the system was located at the landfill site. The majority
of the problems were associated with the Total Organic Carbon Analyzer and the
ultraviolet absorptimeter. The water-cooling system on the Total Organic Car-
bon Analyzer malfunctioned and required replacement. The cooling system had no
indicators for the pressure of the water being circulated in the system. A
flow meter and pressure gauge were added to the cooling system for visual mon-
itoring of flow and for adjustment of proper operating temperature. The small
tubing connected to the ultraviolet cell continued to plug and stop sample flow,
causing the system to alarm. This plugging was attributed to solids passing
through the homogenizing unit. Subsequent investigation revealed that the
abrasive wheel in the homogenizer had fractured and particles of the abrasive
wheel were plugging the system tubing. The homogenizer was replaced with a
500-micron cartridge filter to completely remove any solids present in the
water sample.
The time frame of 2 weeks unattended operation was never achieved. The
alarms received from the system were of a continuous nature and distinction
between equipment failure, zero drift, and actual spill .were not possible. The
results are partially attributed to the parameters of the groundwater runoff
and the soils saturated with various organic and inorganic materials present
at the test site.
58
-------
The Spill Alarm System was returned to the EERU facility for winteriza-
tion and maintenance on December 29, 1978.
FIELD RESPONSE
During the month of October 1979 the services of the Environmental Emer-
gency Response Unit (EERU) were requested in Pittston, Pennsylvania, by EPA
Region III, to monitor mine discharge water that was contaminated with a multi-
tude of unknown chemicals and oil.
Oil spills on the Susquehanna River were traced to an abandoned coal
mine, locally known as The Old Butler Mine, that discharged from 4 to 12 mil-
lion gallons of water per day into the Susquehanna. The source of the contam-
ination was suspected to be a. chemical salvage company that had been illegally
using the old mine as a chemical waste disposal site.
EERU responded on October 19, 1979, with the Spill Alarm trailer and aux-
iliary equipment. On October 22, 1979, the On-Scene Coordinator (OSC) for EPA
Region III requested additional EERU equipment for support operations. The
Spill Assessment van and the Mobile Office/Decontamination Trailer were dis-
patched from Edison and were on site at Butler Mine the following morning. The
mine drainage was continually monitored for total organic carbon, uv adsorp-
tion, oxidation/reduction potential, pH, and conductivity. The EERU crew and
equipment were released by EPA Region III in November 1979.
59
-------
TECHNICAL REPORT DATA
/Please read Iiixzructio'is on the reverse before completing)
1. REPORT NO.
EPA-600/2-81-^ JZ.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
A Hazardous Materials Spill Warning System
5. REPORT DATE
September 1981
5. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Milton Kirsch, Robert Mel void, John Vrolyk
8. PERFORMING ORGANIZATION REPORT NO.
RI/RD77-263
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Rockwell International Corporation
Environmental Monitoring & Services Center
2421 W. Hi 11 crest Drive
Newbury Park, California 91320
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-03-2080
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory-Gin.,OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final Report (June 74-Dec 78
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Contact: Ira Wilder
(201) 321-6635
16. ABSTRACT
The Environmental Protection Agency has, developed a list of materials defined
as hazardous substances based.un their aquati,c toxicity. In addition, certain
materials have been designated as "priority pollutants." Often, a spill of toxic
materials into a moving water stream can occur without the spiller being aware,
or without the spiller notifying authorities. Accordingly, a system was needed
to detect the presence of hazardous toxic materials in streams and rivers. This
need has been filled by providing a spill alarm system, which was designed, fabri-
cated, and tested prior to delivery. It consists of nonselective detection compo-
nents which together serve to detect all the materials on the designated hazardous
materials list, and the priority pollutants. The system was mounted on an auto-
motive trailer and delivered to the Oil and Hazardous Materials Spills Branch in
Edison, New Jersey.
This report was submitted in fulfillment of Contract 68-03-2080 by Rockwell
International under the sponsorship of the U.S. Environmental Protection Agency.
This report covers the period from June 28, 1974, to December. 1, 1978. Rockwell
work was completed as of October 1, 1977.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Hazardous Materials
Warning Systems
Monitors
Recording Instruments
Design
Water
PH
Conductivity
UV Absorption
Total Organic Carbon
Analysis
Unattended, Operation
13B
3. DISTRIBUTION STATEMENT
Release to public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
68
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
60
------- |