EPA-670/2-74-087
November 1974-
ASSESSMENT AND DEVELOPMENT
PLAN FOR MONITORING OF
ORGANICS IN STORM FLOW
-------
EPA-670/2-74-0S7
November 1974
ASSESSMENT AND DEVELOPMENT PLAN FOR
MONITORING OF ORGANICS IN STORM FLOWS
By
Allen Molvar, Ph. I)
Angela Tulumcllo, Ph.D
Raytheon Company
Portsmouth, RI 02871
Contract No. 68-03-0262
Program Element No. 1BBQ34
Project Officer
Hugh Masters
Storm and Combined Sewer Section (Edison, NJ)
Advanced Waste Treatment Research Laboratory
National Environmental Research Center
Cincinnati, Ohio 45268
NATIONAL ONVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
REVIEW NOTICE
The National environmental Research Center--
Cincinnati has reviewed this report and approved
its 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 com-
mercial products constitute endorsement or recom-
mendation for use.
11
-------
FOREWORD
Man and his environment must be protected from the adverse effects of
pesticides, radiation, noise and other forms of pollution., and the
unwise management of solid waste. Efforts to protect the environment
require a focus that recognizes the interplay between the components
of our physical environment--air, water, and land. The National
Environmental Research Centers provide this niulti disciplinary focus
through programs engaged in
o studies on the effects of environmental contaminants
on man and the biosphere, and
o a search for ways to prevent contamination and to
recycle valuable resources.
An evaluation of existing automatic monitoring devices for measuring
organics in sewage was conducted with the objective of developing a
re]iable system for operation under storm and combined sewer condi-
tions. The study described here is a state-of-the-art report based
on the above findings.
A. W. Breidenbach, Ph.D.
Director
National linvironmcntal
Research Center, Cincinnati
-------
ABSTRACT
Sewer line scouring, urban runoff, and combined sewage associated with
storm events represent a substantial organic pollution load. Since storms
usually exhibit high flow rates over a short period of time , the treat-
ment facilities become overloaded and deliver an organic pollution load
to receiving water bodies. Many times a significant amount of the com-
bined sewage bypasses the treatment plant and is discharged untreated.
A method for assessing the organic content of storm related wastewaters
would permit programming discharges, and monitoring and controlling treat-
ment processes. A variety of laboratory techniques have been employed to
measure this organic loading, but only an on-line technique such as con-
tinuous TOC wi11 provide the necessary information on a real or quick-time
basis.
Experience with currently available commercial TOC units has not resulted
in a sense of confidence in the hardware. An evaluation of the instru-
mentation necessary for a reliable TOC in the stormwater environment leads
to the selection of a measurement system based on total combustion of sew-
age and detection of CO2 by infrared methods. Tests are presently under
way to establish sample processing, modifications of the engineering model,
and accumulation of the continuous monitoring data on total organic car-
bon content of storm and combined sewage.
This report was submitted in partial fulfillment of Program Element Mo.
IBBO34, Contract No. 68-03-0262 by Raytheon Company, under the sponsership
of the U.S. Environmental Protection Agency. Work was completed in April
of 1974.
iv
-------
CONTENTS
Page
Foreword
Abstract j v
List of Figures vi
List of Tables viii
Acknowledgement ix
Sections
I Conclusions 1
II Recommendations 2
III Introduction 3
IV Storm and Combined Sewage Characteristics 6
V Technical Literature Related to Automatic
Organic Monitoring Devices 15
VI Evaluation of Available Automatic Total
Organic Carbon Instrumentation 21
VII Performance Specifications 31
VIII Essential Components for a Stormwater
Organic Monitoring System 37
IX A System Design for Sampling and TOC Measure-
ment of the Storm Event 63
X Proposed Operating Mode Test Site Interfaces
and Future Activities 70
XI Bibliography 87
v
-------
No
1
2
3
4
5
6
7
S
9
10
11
12
13
14
15
16
17
18
FIGURES
Page
Rainfall Hyctograph, Calculated and Observed
Runoff Ilydrographs, Oakdale Avenue Basin,
Chicago, Second Storm of July 2, 1960
(Source: Urban Runoff Characteristics, EPA,
Water Quality Office, October 1970, p. 275.) 7
Long, Intense Stoi'm in Combined Sewer District
G-4 (Gpm x 0.0631=l/sec; in. x 2.54=cm) (Ref-
erence: "Journal of Water Control Federation
Vol. 43, p. 2041, October 1971.) 9
Diagrammatic Representation of Available TOC
Instruments 22
Diagrammatic Representation of 11. Wosthoff TOC 24
Diagrammatic Representation of Ionics Model 1224
TOC and TOD 26
Diagrammatic Representation of Enviro-Control TOC
(patent restrains use of device for TOC) 27
Astro-Ecology Corporation Model 1500 TOC and TOD 28
Block Diagram of TOC 39
Raytheon's Blender 42
Discrete Sampling Valve Diagram 44
Temperature Profile of the Continuously Operating
Reactor 49
Field System Diagram 61
Continuous Data on a Flowing Sample 68
Proposed TOC 69
BU Stormwater Detention Facility 71
Application of Sample System and TOC Monitor to
MDC-BU Installation 72
Boston's BU Storm Detention and Chlorination Station 75
Boston's BU Storm Detention and Chlorination Station,
Flow Diagram 76
vi
-------
FIGURES (Cont)
N°- Page
29 Municipal Wastewater Treatment Plant, Cranston,
Rhode Island 78
20 Response Time of Organic Cyclohexanol Solution
to Water and to Sewage 82
VI 1
-------
No
1
7
3
4
5
6
7
8
9
10
11
i&e
10
10
11
19
37
43
46
50
59
62
81
TABLES
Characteristics of Combined Sewer Overflow
Characteristics of Urban Stormwater
Characteristics of Combined Sewage Bulk Sample
Selected Automatic On-Line User Experience
Essential Components for a Stormwater Organic
Monitoring System
Particle Size Reduction
Pumping Systems
Percentage of Organic Carbon Recovered
The Determination of Carbon in Sewage
A1arms
Data Sheet for Characterization of Sewage and
Correlation wi£h TGC Analysis
viii
-------
ACKNOWLEDGEMENT
We wish to acknowledge the support and direction of Messrs. Richard Field,
Chief, and Hugh Masters, Project Officer, of the Storm and Combined Sewer
Section of the U.S. Environmental Protection Agency. We wish to acknowl-
edge the help of Dr. Philip Shelley of Hydrospace-Challenger whose coopera-
tion in describing his Stonnwater Sample Delivery System has made it
possible to define a system of on-line TOC analyses for storm and combined
sewage. We also wish to thank Dr. John Litzkowitz whose description of
his recent contribution to the technology of CO2 detection has helped to
complete this work.
We feel a deep sense of gratitude to Mr. Anthony Ventatuolo and his staff
of the Cranston Municipal Treatment Plant for providing us with the facil-
ities and assistance to obtain sewage samples and conduct the experimental
work necessary to establish the accuracy and repeatability of our data.
-------
SECTION I
CONCLUSIONS
1) High solids, surging flows, and adverse environmental conditions associated
with storm related wastewaters impose difficult service requirements on
automatic rapid organic monitoring systems.
2) Among the various methods of determining organic pollution levels, continu-
ous on-line TOC analysis using the well-known technology of total oxidation
and infrared detection of the CO appears to be a "best" analytical choice.
<6
3) The selection and delivery of representative samples containing substantial
suspended solids arc expected to result from current design and experimental
work in this laboratory and by Dr. Philip E. Shelley of Iiydrospace Research
Corporation.
Because up to 80 percent of the organic content of storm water is caused by
suspended sol ids, any meaningful stormwater organic monitoring system
must take a representative sample from the main flow, suitably transport
and analyze it without the loss of any suspended solids.
4) At this time, no commercially available rapid on-line automatic organic
monitoring system is suitable for stormwater service. However, the authors
believe that most of the components to build such a sysLem are available from
appropriate suppliers. But several essential components such as sample
transporting :uui conditioning equipment and combustion reactors must be
fabricated especially [or stormwater duty.
5) To be successful, the organic monitoring system must be capable of un-
attended operation; moreover, the service requirements should be within the
capability of a typical municipal maintenance staff.
-------
SECTION II
RECOMMENDATIONS
1) ATIcr a decade of laboratory TOC and TOD analyses, the concept and utility
ol' rapid organic monitoring clearly has been demonstrated. The transition
from the laboratory (batch) type to a continuous on-line instrument involves
the fabrication of a rugged, reliable instrument equipped with suitable a!arms
and fail sale modes. The need for mechanical ruggedization cannoL be over-
stressed. Accordingly, an engineering model should be fabricated with com-
mercially available (whenever possible) and custom made components. The
feasibility of representative solids conditioning and tr;msport, along with the
accumulation of accuracy, precision, response time, drift and reliability data
should be determined via stormwater simulations.
2) Published stormwater characterization data indicate that a mixture of raw
sewage ;ind scLLled primary sludge should simulate stormwater, especially
the first flush. During this laboratory study (to be conducted at Cranston, HI
STP), the alarm functions, and startup and shutdown procedure can be
verified in the field. An interim report will describe this phase of the study
and assess the desirability of continuing this project.
?j) After demonstrating the feasibility of the engineering model, the developed
system should be ruggedized, and automated for subsequent field tests under
actual storm conditions. An important aspect of these field tests involves
assessing the ability of a well run municipality Lo operate and maintain the
stormwater organic monitoring system. Arrangements have been made to
test this system at Boston University Stormwater Detection and Chiorination
Station. During the field tests, a series of automated grab ;uxl/or composite
samples will be taken for TOC, and BOD analysis via standard EPA approver!
laboratory methods.
2
-------
SECTION III
INTRODUCTION
The United States Environmental Protection Agency's (EPA) Storm and Combined
Sewer Pollution Control Research and Development Program developed three
storm-event strategies other than separation: control, treatment, ;ind combina-
tions of these two. Briefly, successful treatment and control of stormwater
and combined sewage depends heavily on the ability to measure pollution loads
and subsequent removal efficiencies. Consider, for example, stormwater
storage utilizing the combined sewer's inherent capacity. If the storm is
severe, the storage demand will exceed available capacity and the excess com-
bined sewage must ovcrllow to the receiving water. Based on the strengths of
the incoming combined sewage, the least polluted stormwater can be discharged
into the stream. Without any measurements of organic strength, no rational
method exists for selecting or programming overflows. The organic load {pro-
duct of flow rate and Total Organic Carbon concentration) represents a signifi-
cant pollutant parameter since organic material exerts an oxygen demand on the
receiving system.
Transient flow rate monitoring over large ranges is a difficult problem but pre-
vious efforts and demonstrations clearly indicate that stormwater nnd combined
sewage flow rate monitoring is addressable by well established instruments.
Although several Total Organic Carbon, Total Oxygen Demand or COD analyzers
are commercially available, no existing systemfor automatic monitoring of organics
3
-------
in storm How has been successfully demonstrated to automatically analyze storm
and combined sewage samples. Since storm events lead to rapidly changing-waste-
water concentration profiles with the occurrence of very high suspended solids levels,
a dedicated organic monitoring system must be specifically designed for this service;
satisfactory commercial components, however, should be employed, when available,
so as to keep the project cost to an absolute minimum. Recognizing the need for a suit-
able organic monitoring instrument, EPA awarded Raytheon's Environmental Sys-
tems Center a contract to develop a rapid on-line Automatic Monitoring System
to measure the organic content of storm and combined sewage.
Because laboratory researchers have investigated wastewater total organic carbon
(TOC) determination for the last five years, the technical literature contains a sig-
nificant amount of relevant information. The Assessment and Development Plan pre-
sented herein, evaluates the suitability of reported technology, equipment, and do-
vices with respect to storm and combined sewage TOC monitoring. Moreover, per-
tinent user-experience and comments are presented to give insight into the opera-
tional requirements. From an examination of stormwater and combined sewage
characteristics, a rational set of automatic organic monitoring component specifi-
cations were generated to insure acceptable performance of the developed sys tem
in the storm environment. The commercially unavailable components or unacce pl-
able devices are clearly identified and Raytheon's experiences, with supportingdata,
have been brought to bear in developing the needed components. This report also
discusses the development tasks and laboratory testing program leading to the con-
struction and verification of rapid, reliable, continuous storm and combined sewage
organic monitoring. After successful laboratory demonstration, Phase!, a rugged-
ized, automated analyzer will be thoroughly field tested in MDC's Boston Uni-
versity Stormwater Detention and Chlorination Station. In short, this Assessment
and Development Plan Report summarizes the findings of a comprehensive study
4
-------
of previous TOG investigations, and existing important components and devices;
moreover, it presents a clear cut, action plan with accompanying schedules
Leading to development of a storm and combined sewage automatic organic
monitoring system.
5
-------
SECTION IV
STORM AND COMBINED SEWAGE CHARACTERISTICS
To develop an automatic, on-line organic monitoring instrument, the designers
must be thoroughly acquainted with the characteristics of the intended samples—
storm water and combined sewage. Human factors (i.e., maintenance capabili-
ties), environmental conditions, and operational modes must also be considered
during the instrument design phases. Consequently, a brief description of storm
events and subsequent storm and combined sewage temporal variations follows.
Most of the 1,300 sewered communities in the United States arc served by com-
bined sewers; a much smaller number have segregated facilities where storm
water is transported in a separate sewer. Although storm events only occur 4
to 5 percent of the time, untreated overflows from combined sewers and storm
water are a substantial source of water pollution. For example several stu-
{1 2)
dies ' demonstrated that annual Biochemical Oxygen Demand (BOD) contri-
butions from storm-related wastewaters are approximately equal to the BOD
load of secondary treated sanitary effluent.
Because of the random nature of rainfall events, the quantity of collected run-
off changes continuously; but the integrating effects of the drainage, infiltration,
collections depression storage, gutter-flow and sewer-flow routing processes,
produce relatively smooth hydrographs as illustrated in Figure 1. With respect
to flow-rate properties, storms can be classified as either long or short dura-
tion and high or low intensity.
6
-------
"i ¦"« »"~~Trri" i r 1 1— 1
OAKOAl f AUfNUt 8AHIK
SECOND STOfl* Of *\JlY Zf 1^0
0 10 20 3D JtO to ;u do 50 *00
TIME (MIMJU<0
L P& stohh waiir
hANAQF Hi KT HODiL
80
70
90
Figure 1. Rainfall hyolograph, calculated and observed runoffhydrographs,
Oakdale Avenue Basin, Chicago, second storm of July 2, 1900
(Source: Urban Runoff Characteristics, EPA, Water Quality Office,
October 1970, p. 275.)
7
-------
Stormwater and combined sewage quality parameters vary from storm to storm,
and also changes rapidly during storm events. Frequency of changes in sewage
quality depends upon rain fall intensity, duration, antecedent conditions, land use,
topographyj flush characteristics and other factors. Temporal variation data,
Figure 2, clearly indicate that the organic concentration changes vary rapidly
with respect to time. Notwithstanding that there is no apt description of "typical"
combined sewage, it is informative to examine the general properties of storm
and combined sewage. Because of poor flow characteristics of combined sewers
during dry weather, settled solids build up in the combined sewer. When a
storm occurs, the greatly increased flow scours the collection system and
causes the "first flush" phenomenon with suspended solids concentrations fre-
quently reaching 1, 700 rng/1. In addition as the storm run-off drains from urban
land areas, it picks up accumulated debris. Tables 1 and 2 display concentration
ranges usually encountered in combined sewer overflows and urban stormwater
run-off, respectively. The organic concentration, as measured by either BOD or
Chemical Oxygen Demand (COD), varies over a large range of values; it is esti-
mated that the Total Organic Carbon (TOC) ranges from 10 to 1000 mg/l.
Since stormwater and combined sewage contains an appreciable amount of sus-
pended organic matter, any meaningful, total-organic analyzer must measure
the organic contribution of the suspended solids. The particle-size distributions
and density properties of the suspended matter become important parameters
with regard to sample-taking, transporting and conditioning. One study^ \
related to fixed screening, presents particle-size-distribution data {Tabic 3)
that indicate the bulk of the suspended organics are less than 1. 0 mm in diame-
ter. In general, the further downstream collection points are (i.e., closer to the
end of the delivery system), the smaller the particle sizes are, since transport-
ing under storm conditions usually breaks up the larger organic solids. Moreover,
8
-------
6001 f
tn t>00-
D
Z
<
V)
D
O 4Q0-
300-
2
o
iU
<
* 20Q<
a
100-
CJ
s
w
a
o
v>
j
<
1,000
f
a
w
a
o
>
<
h-
o
500
AUGUST 2, T969 18:05 PM} 85MIN. -2.8
660.000 GPM
40
X
a
2
y.
2
<
EC
<
—J
H
3
U
*0 riG
TIME (MINUTES!
AUGUST 2, 1969 (8:05 PMI 85MIN. -2.8
lOO-
CD
£
O
a
00
50-
0 1
¦/> _J.
Q Q
LU X
Q £
t o
5 V 1.G00
f f) *
UJ *
mS *
p o
1 -
¦J -
o
tf*
D
« C
500
m
UJ
¦<
jr «.
• ¦ * c
b g
>.\
1
' )
1 j
p.
1 «
r i 1 I I if
L—-<
10
ID
rtO 50
TIME (MINUTESI
60
70
HO
90
Figure 2. Long, intense storm in Combined Sewer District G-4 (gpm x 0.0031=
1/sec; in. x 2,54 = cm) (Reference: "Journal of Water Control Fed-
eration, " Vol. 43, p. 2041, October 1971. )
9
-------
Tabic 1, CHARACTERISTICS OF COMBINED SEWER OVERFLOW*
Characteristic
Range of Values
BOD5 (mg/1)
30—600
TSS (mg/1)
20—1, 700
TS (mg/1)
150—2, 300
Volatile TS (mg/1)
15-820
pIT
4. 0—8. 7
Scllleable solids (ml/1)
2—1, 550
Organic N (mg/1)
1.5—3:?. 1
NIIgN (mg/1)
0. 1-12.5
Soluble PO4 (mg/1)
0. 1-6.2
Total col i forms (no./100 ml)
20,000—90 x 106
Fecal coliforms (no./100 ml)
20,000—17 x 10°
Fecal streptococci (no,/100 ml)
20, 000—2 x 10°
*Seleeted data.
Table 2. , CHARACTERISTICS OF URBAN STORMWATER*
Characteristic
Range of Values
BOD5 (mg/1)
I —> 700
COD (mg/1)
5—:j, 100
TSS (mg/1)
2—11,300
TS (mg/1)
450—1-1, GOO
Volatile TS (mg/1)
12-1, (500
SclLleablc solids (mg/1)
0. 5 — 5, 400
Organic N (mg/l)
0. 1— 1G
NII3N (mg/1)
0. 1-2.5
Soluble PO4 (mg/1)
0. 1—10
Total PO4 (mg/1)
0. 1-125
Chlorides (mg/1)
2-25, OGot
Oils (mg/1)
0-110
Phenols (mg/1)
0-0.2
Lead (mg/1)
0—1. 9
Total coliforms (no./100 ml)
200-HO x ,10fi
Fecal coliforins (no. /100 ml)
55-112 x 10(>
Fecal streptococci (no./100 ml)
200-1. 2 x 10(i
^Selected data.
tWith highway cleicing.
10
-------
Tabic 3. CHARACTERISTICS OF COMBINED SEWAGE BULK SAMPLE
5 Nuvcmbur 1969
T i m !.j
05Z0
0620
f) /
OH.10
0. 499
0. (175
TotaI Sol ids, m^/ ]
HH
140
1BU
id 0
4 M
Total Volatile Solids, mg/i
40
44
M
11 Z
Li 4
Total Su bptTicleld SuliriH, eii^/1
14
4 5
M
57
IKS
Tola! Volatile Suspt-ndoci Sol ids, mgjl
IS
U
Z-l
40
5 31
S^ttlcab! v Sol id s T ml/1
5
0, 7
1
t). i'i
0. 7
Fkiatables, mg/l
1, '1
i. 4
^9
L, 3
11. a
Particle Size Distribution, mg/1
>3, i£7 nun
0
t r.ivt1
iract-
1 4, 4
Z'>, 5
K U7-0, W nun
0
t m i v
5 rncc
11, i
14. 1
0, 991-0. Z9 S mm
t race
t r.n c
s, s
4 £.
K>. 5
0, 1, (,
1 3J, 4
6Z6
Huxar.n Extrartablr A1atrri.il, m^/I
4. 1
4
H
7. 2
1 4. S
Total Coliffrms, MPK/1O0 ml
2,
IxlO5
IxH)'1
4. ixia"
3x106
Fucai Coliftjrms, MPN/100 ml
3x1 0
4
5x10
A
1x10
4, is Hj"
9x10''
Fi'Cal Slfi'jitarocri, MPX/10U ml
tjxin3
4. ixllt4
'?x!0 '
4. !xl<) ^
ix 10*'
19 r S
T i ETi t>
104 5
1145
1
4 15
w7 ^
Total Volatile Solids, rng/1
50
f.5
(>0
1 HO
Total Suspended Solids, m^/1
19
4 f>
H
1 1.1
H4
Total Volatile Suaprtidi-d Solids, m^/l
9
51
.'.1
71)
Setll'-abU- Solids, rnl/1
0, J.
1. 5
1
J, H
Fln.-vtablrs, mjj/1
7. 1
J. i<
4. .!
4. 4
i. 9
Particle Si xc Distribution, nij;/l
> 3« Ml mm
0
t)
0
5|i. 5
3
3, A27-0, 99 J mm
0
1. v
4, H
5. 1
4
Q, ?4
BOD, mg/1
14, 7
15
h'j. H
7 7,4
40, B
COL), mr/1
7 6. 8
57, £j
^7. ft
i£l
144
Hcx
h
. it
h
A
Tntcil Gcjliformiij MJ3N/I0f> mi
1. 5x10
4. fix 10
•1, hx 1 0
J. 4s 1 0
¦1, Ox 1 0
Focal Conforms, MPN/100 ml
9. J* 1 0 '
4. ix 10^
j.. 4x1 ah
4, f,il(]J
4, bx ItJ5
Fecal StroptOCncei, MPN/ 100 m I
7. 5xl03
4. 3x10 S
ixlO3
1, i , i :>'
> 1, 1 x 1 0 3
11
-------
Table 3. CHARACTERISTICS OF COMBINED SEWAGE BULK SAMPLE (cont)
£.1 Dti^umbn' 1909
T imr
9 30
1035
1 1 i 5
12,i5
Hill
Flow, cfs
0, 910
0. 17S
0, 169
0, IDS
o, o?a
Total SoJ idis, m^ / 1
I 10
2 HO
210
J^Q
400
Total Voi.ii.ile Solids , 3
so
I&5
115
140
HU
Total Suspended Solids, inju/l
b<3
7
2, "i
7
Klnatablirs, mg/1
^ 3
I.
3, ii
i, ">
4, 2
PoLrliclo Siti' Distribulnm* m^/l
>1, iZl mm
H„ H
0
n
0
3, 3Z7 - 0. 991 mm
14, 1
K. h
r>. V
Iti, A
H, i
(1, 99 1 -0. ^95 mm
9
SA. 4
1 i. 0
lb. 1
M. 7
0. J9S-G. 074 mm
15, i
IB, 1
I 4, 7
H
"5
, -t
lr3, 7
•)U, 7
sixf -At 5(J-prrr*>rPLjli> by v/fitjlit, mm
0. 54
nt n
o, n47
0. 1*)
0. 1 1
BOD, mp/1
U, 6
7^.
4
64, 4
102
CUD, iixu/1
H6, 't
15H
1H2
154
Z^7
i i e xan v K x I r r t ab h ¦ M,iI < - r , a I, i h y / i
4, 4
24 ri
V I
ii. 7
Total Coliforrns, MPN/IGO ml
4, f,x!0S
^ 4x1 nfj
i. -Ixio''
h
i:-;l()
¦1. f.K Hi"
Fccfil Coliforms, MPN/ I 00 ml
4, 3xl04
2, 4xl0f'
-IxlO6
4, (ixlO5
4, «,k10''
Feral Slrupiiu-un.-i, MPN/10Q mi
9. ixlD"
-
-
-
2. -1 s-1 04
ZO January 1970
T ime
IbZb
1725
lKi*>
19 30
20 30
Flow, cfi'
i. 05
(). (,0 5
(), M'M
0, -HI
0, StJ.
Total Solids, m^/1
M
M
1H4
14K
fiO
Tftlal Volatile Solids, mg/1
4 ft
If,
(>H
72
It".
Total Suspended Sol ids , m ^ / I
Z9
30
1-3
5 }
17
Total V^latilr Sut^Ji'iidcd Solids, wg/1
19
2 J
1-i
2H
1 h
Set U c.ibl r Sid ids, ml /1
0, ^
l.
.{
0. 2
(>, (,
Floatahl«*s» m^;/l
4. 2
22. 3
1. ')
4. 7
2. H
I'articlo Si«v Distribution, m^/l
>.*. Ml mm
U
0
D
0
0
i, kiV-0,991 mm
0. 5
1.<¦'
1. J
2. h
». y
0. 991-0. ^9 ^ mm
J, 1
S 7
2, f,
I
1. b
0, i9^-0. 074 mm
<>. 1
1
•1
4,
4. H
<0, 1174 mm
10, 2
17, »
J, 1
2.
a
i?,v n ^0-porc rnlilr by weight, inm
0. 0-GS
(1. 0 ih
0. OfH
0, E)i
0,
BOD, \Tififi
Zl, I
1H. 2
7. 97
U). 4
12. 7
COD, m^/l
Id
127
ill
549
4 '¦
i i exam: ExtractabU- Malarial, ing/1
£
¦14
5. H
10. -1
Tnla 1 Cojjfumis, MPN/ 10-0 ml
3x10"'
i. ijx li)'1
7. 5k1!)*'
i, fix It)'1
>. l,x 1«*'
F^cal C'diforms, Ml3N/ 100 mi
<3xl04
3, 6x I o '
7, ixio'1
!. (,X104
3. '.si!!'®
F e e.i I St r epti>coc c i, MPN/100 ml
2. JxlO4
2. ixlt)*1
¦?. txin5
5. t\K 1 0 ^
7, 3*10 '
12
-------
Table 3, CHARACTERISTICS OF COMBINED SEWAGE BULK SAMPLE (cont)
27 January 1970
T ime
0130
0 210
0145
04 JO
01 30
Flow, cfs
1. 2 J
0, 603
ft. m,
0. 058
0. 0 39
Total SoIkJs, mg/1
95
60
? i
11
90
Total Volatile Solids, mg/1
S5
}2
40
n
4 J
Tuial Suspended Solids, nip/l
8
4
{>
6
Total Volatile Suspended Solids, mg/1
7, 5
4
t>
«»
f.
Si'ttlcablo Solids, ml/1
0. }
0, 1
0, 2
0, OS
0. OS
Floatabl.es, mjj/I
1 I
1
a, »
a, h
0. 9
Particle Size Distribution, rng/l
>J, 327 mm
.1. 991 min
0, -391-0. 295 mm
0, 29^-0. 0?4 mm
<0. 074 mm
al SO-percentih: by weight, mm
0
1. 4
1, 7
9,6
12. 9
0,082
0
0. J
0, 9
2. 4
7
0,0 35
0
!). 2
1. 7
S. •>
9
0, It
0
!), 1
0. ¦>
1, 9
7. 9
0, 021
0
1
1.1
1,9
H
0, 0£H
BOD, rrig/t
3
<1, S
4, 65
1, 95
17. 7
COD, mK/l
fa 4
16, 9
}[
il
S2, 5
Hexane Extraxtable Material, mg/1
S. 2
3. 8
2, f<
I
S, 4
Total Coliforms, MPN/100 ml
2, 3x10b
9xl05
9xlOS
4, 1*1 0f'
TxlO5
Fecal Coliforms, MPN/U10 ml
Fecal Streptococcal, Ml'!.", lri"< ml
4*104
2. JxlO4
<1x103
4
¦1, 3x10
0 9 4 5>
104S
Flow, cfs
1. 0f>9
0. 199
0. 1024
0. 0932
0. 079
Total Solids, mg/1
148
152
432
952
54 H
Total Volatile Solids, mg/1
96
72
lift
24 S
156
Total Suspended Solids, mg/l
49
50
127
95
115
Total Volatile Suspended Solids, mg/1
29, 5
3 J
115
SO
94
Settlcable Solids, mi/1
0, 6
2. 5
11, S
4
14
Floatablcs, mg/1
4.65
I. 83
1, 6H
3, 31,
1.73
Particle Size Distribution, mg/1
>3, 327 mm
0
0
)4. 2
1 26
0
327-0, 991 mm
9, 4
11,6
9. a
56
29
0. 991-0, 295 mm
10. ()
24. 2
2H
47
IB
0. 29^-0. 074 mm
42. 2
?, 4
22, (.
4 )
Ih
<0, 074 mm
53, 4
2, 6
12. 4
7H
hi
size at StJ-perccntile by weight, mm
0, 116
0, 64
0, 64
i, 12
0, 076
BOD, nig /I
21
s*>. «
1 39. 1
168
121. 4
COD, nijj/1
SB. 4
176. H
6
40 2. 2
371. J
Hrxani: Exlraclabli- Material, infi/l
12
7. 7
10. 1
iK
44. 1
21. 5
7
Total Coliforms, MPN/ 100 nil
I, lxlO7
2. 4xl0h
Z, 4x10
1
>1. 1 x JO
1, lx JO
Feral Coliforms, MPN/100 ml
9xl04
4. JxlO5
2. 4x1 0(l
>1. lx !07
9, JxlO5*
y e* r .11 .StrrptiM ri'CI, MPN/ 100 ml
4. 3xl04
4. l.xl0S
2. 4xl0S
4. fix 1 05
9. 5x1ft4
13
-------
most storm and combined sewage exhibit, stratified How properties, where the
grit, sand and other dense particulate material are concentrated near the bottom
of the conduit. It is reasonable to assume that these lower areas contain mostly
inorganic grit and that the bulk of solid organics is collodial, super collodial
or suspended particles less than approximately 1.0—10 mm in diameter.
Although floatable materials such as boards, paper, and other large debris may
contain organic compounds, they can not be considered as part of the analyzer-
sample universe for obvious reasons.
14
-------
SECTION V
TECHNICAL LITERATURE RELATED TO AUTOMATIC
ORGANIC MONITORING DEVICES
In the last five years, an increased awareness of the problems of environmental
pollution motivated research into more effective measuring mid monitoring methods.
One of the problems considered has been Total Organic Carbon (TOC) determin-
ation, because the delivery of unsuitably treated organic wastes, industrial and
municipal, to receiving waters has resulted in eutrifieation of our water resources.
BOD (Biological Oxygen Demand)
In order to prevent the discharge of unsuitably-treated sewage, a number of indices
has evolved which provides information about the organic content of sewage.
Historically, the Biochemical Oxygen Demand (BOD) was the first Lest devised to
measure the organic concentration of waste waters. The BOD test, however, has
two severe shortcomings: poor precision (±18%) and requirement of five days to
complete. Unacelimated seed bacteria, toxicants and human errors are the major
factors which cause this poor precision. Because five days arc required to com-
plete a BOD test, this procedure can only supply historical da La.
RESPIROMETRY
Inasmuch as control techniques require real-time data for implementation of correc-
tive actions, several researchers addressed themselves to developing a rapid BOD
technique/4' ^ commonly referred to as respirometry. Instead of measuring the
15
-------
oxygen depletion after five days, they monitored the oxygen uptake rate during
a 15-minute to 1-hour interval; the uptake rate is used to predict the 5 BOD through
a predetermined estimation coefficient. Although re spirometers wo rk reasonably
well in the laboratory, on-line versions arc usually very unreliable. Moreover,
respirometers exhibit a poorer precision than the BOD test, and are also vulner-
able to the same interferences.
COD {Chemical Oxygen Demand)
Because of the inherent limitations of biological techniques, a more rapid, reliable
lest was necessary to measure organic pollution. Liquid-phase oxidation of
organics by acidulated chromate was developed Lo monitor the organic concentration
of waste waters; this reaction, carried out under reflux conditions, requires two
hours for completion, As the reaction proceeds, the reduction of the chromate to
trivalent chromium can be seen colorimetrically. This measurement, expressed
(7)
as Chemical Oxygen Demand (COD), has been modified by several researchers k
to reduce the reaction time to 15 minutes. Because COD determinations
require consumable reagents, and high temperatures, all COD procedures need
constant attention by skilled technicians. Although use of an Autoanalyzcr ^
can speed the data delivery, it still requires a regular supply of reagents and
inspection by skilled personnel.
Since one of the reaction products of chromate oxidation is carbon dioxide, inst.ru-
(y)
ments have been designed which monitor the carbon dioxide conductome trie ally.
These devices, which display their results as Total Organic Carbon (TOC), have
most of the problems of COD. While inorganic reducing agents do not oxidize to
COg) neither do a number of organics. ^ Despite the improvement in organic
estimation, all automated COD procedures are too slow and not suitable for
unattended operation.
-------
TOC (Total Organic Carbon)
Vupor-phasc oxidation of waste m water (950° C) was developed for the purpose
of providing a complete, rapid determination of the carbon content (TC). Acidify-
ing' the sample and purging with COg free gases, removes the inorganic carbonates;
then this combustion technique measures the total organic concentration. This
procedure has been used to measure Total Organic Carbon in a number of locations,
and has been found to give useful rapid data.
Several investigators attempted to correlate BOD and COD data to corresponding
TOC values. As expected, BOD correlations suffered from the shortcomings
inherent in the BOD test. COD and TOC, however, correlate well^
The earliest forms of the Total Organic Carbon instrumentation used non-
dispersive infrared (NDIR) detection of C09 concentration in the gas stream. In
(14)
one case CO£ was reacted over a platinum catalyst Lo produce CO which was also
monitored by NDIH. While there may be some place where the measurement of
CO offers some advantage, it has not found great acceptance to date; CO2 has
been preferred.
In choosing a means for the measurement of carbon, it is desirable and (for the
purposes of reliable instrumentation) necessary to minimize the complication
of the system. For this reason, ND1R lends itself to considerable economy of
design and has found the most common acceptance.
;Flame ionization detectors (FID) are highly sensitive and linear devices,
requiring the CO^ formed during combustion to be catalytically converted to
(15)
methane. This reaction needs hydrogen. Moreover, aflame is necessary
to provide the ionization and a complex array of electronics is required to
develop output signals. If one neglects the conversion of the carbon to methane,
(1_ G}
the bonding of the hydrocarbon becomes a large factor in the FID signal. In
17
-------
spite of this excessive complication, considerable effort has been spent on this
, (17)
approach. A recent survey of user experiences (Table 4) found FID tech-
niques unreliable in field environments.
Alternative CO^ detection techniques have also been investigated. A differential
conductivity approach has been used by II. Wosthoff. W Because of the need
for calibrated solutions and flow-controlled conditions, differential conductivity
measurements become complicated.
Rather than measure the amount of CO^ formed during combustion, some investi-
gators have measured the oxygen consumed to form CO^; unfortunately, the electro-
chemical oxygen sensors tend to be unreliable and susceptible to contamination.
They arc usually temperature sensitive and require considerable maintenance of
the conducting media which is susceptible to contamination. For wastewater
oxidation,only u small amount of oxygon is consumed out of a high background
concentration; the TOD determinations are calculated by the difference between
input and output concentration with the resultant summation of error. Accordingly,
history of devices of this type is frequently associated with unreliable results.
It has been reported that dissolved organics can be detected by ultraviolet (UV)
absorbcnce but the high organic content of the suspended solids (particulate
matter), would both absorb and disperse UV. Bond breaking occurs under the
influence of UV as does polymerization which is likely to form UV absorbing films
in the system. Consequently for stormwater sample, this approach becomes
inappropriate and cannot be considered as an alternative method for monitoring
the storm-related organics.
Although TOG instrumentation lends itself to automation {because of fast oxidation
and reliable detection), on-line continuous TOO analyzers experienced a number
of difficulties. One of the most serious is that samples containing suspended solids
-------
Table 4. SELECTED AUTOMATIC ON-LINE USER EXPERIENCE
Or^n uizal ion
(User)
Sou rce
Analyzer
Kspn r its rices
Duportl Inc.
Mr, R, Coon
tonics-TOO
Dupont monitors "soluble" influent and effluent *!'C1C* in I heir
Wavncpliorn, Va. activated sludge facility, in spite of their
effort to strain anil filter the sample, the Ionics TOC nnttlyner's
performancc: was unacccplnblu because of frequent sampl inu valve
plugging, and sample flashing difficulties in the combustion
reactor, [t is estimated that 2!!t> man hours of technician time
arc required for proper maintenance! and calibration.
National Air
ami Water
Kcsea re h
Council
Mr. D, liuckley
Astro-Kctilofjy
Tor
In 1072, the National Air and Water Ke.iear -n Council tested an
Astro TOC analyzer in their Tuffs University Laboratory,
liceausc of excessive not sit and drift on Kill" Solutions (stniulard)
and on selected Kraft liquors, field evaluations of this instru-
ment were remcelkKl, Astro corrected the noise problem bv elec-
tron icallv filler inn Use analyzer output Htjjnnls which increased
response time to several hours. Modifications I" Hie renclor
volume and sample flint rates were also implemented, liul no
lest resulls are ava ilable.
C'itY of
Pairs A ltd
Munic ipn 1
Wastewater
Mr. li. Dolv
Mr, R, Niese
Cal ibratcd
Instr,
I In,
A sli'o- K colony
A laboratory evaluation of this TO£' analyzer, which employs
liquid phase ox i rial ion, on raw sewage, primary clarifier, secon-
dary clarifier and final effluents indicated significant errors
(about 2tt'?nl due to rhlorjde interfereilce and loss of volatile orna-
nies.
In nud November 1117:!, an Asl ro Till' instrument started mmii-
toring t'-1' primary effJu<*ut which ronUiined abmiL HO l'lJM oi
suspended solids. As of .January '£, j'-'M no pinning was
observed. However, no correlations with repird to other meth-
ods of measuring the orpanie content, have been performed,
lie suits appear encouraging, hut performance dala are unavail-
able.
International
Paper
Mobile, Ala,
Mr. R, Wise
Ki'A
Astro-Ecology
Initially the Astro TOC analyzer failed when operating mi primurv
effluent due to pliiKUmj; awl corrosion problems. After elia lining
the trouble components to Intamel, tins analyzer perl'ormei.1
without failure. No data is avail utile on accuracy or the sus-
pended solids content of the samples.
Dow
Chemical
Company
Freepnrt,
Texas
Dr. M. Zeitoun
Ionics-'IOC
Dow intermittently monitors TC content of their influent waste-
water ami mixer) liquor from their aeration lank. Th« aeration
tank's mixed liquor contains approximately 2aun mfj/l of MI.VKK;
an elaborate batch homojj,enizatinn system is required to condi-
1 ion the sample prior to analyses. This system has an accuracy
of i-s'i and requires extensive maintenance every three days.
19
-------
could not representatively be introduced into the system. Automatic slider
valves, and injection valves simply plugged up and became inoperative. In addi-
tion, combustion reactors would plug on the refractory components in the sample.
Special consideration should be given to insure adequate heat transfer and
complete oxidation. Also, most sparging systems necessary to eliminate in-
organic carbonates are vulnerable to solids accumulation and plugging. Through
careful system design (utilizing known technology), it is anticipated that these
problems can be solved.
20
-------
SECTION VI
J i V A L U ATION oF A V AILABLK A LITOM A TIC
TOTAL ORGANIC CARBON INSTRUMENTATION
When Instruments are introduced to Individuals who are involved in analyzing
samples, the insl rumcnts are generally treated from the sample point of view.
The sample is processed through a number of operations that must be controlled
in such a way that the instrument will deliver valuable and useful data. One
becomes accustomed to this approach as a student and on exposure to instrument
salesmen. For the purpose of dese ri bins avail able TOC analyzers, this is a
totally acceptable approach.
From the point of view of instrumentation, it is possible to devise a somewhat
more efficient approach than the classical one. This pretty much represents
the point of view of the instrument designer and the usual experimental approach
to engineering' the design. The first and perhaps the only independent choice that
is possible is a transducer. The transducer is a device where the physical or
chemical characteristics of a system are converted to an electrical signal.
When an informed and experienced judgment is made about the transducer effect,
this choice acts as a constraint on all further choices. The sample processing
and the signal processing must be consistent with that initial choice. To demon-
strate this approach we have chosen the format In Figure 3.
21
-------
AUTOMATIONS IN AID OF OPERATION OF INSTRUMENT
SAMPLE PROCESSING
TRANSDUCER
SIGNAL PROCESSING
SIGNAL APPLICATIONS, FEEDBACK CONTROL, FEED FORWARD CONTROL
Figure 3. Diagrammatic representation of available TOC instruments
-------
All example would be an appropriate aid to understanding this approach. A care-
ful examination of the various possible and practical alternatives might lead an
individual to choose to measure TOC by making an infrared measurement of the
CO in the gas stream. It is then necessary that he provide a means for the
efficient oxidation of the carbon compounds in sewage and find means of handling
the sample and delivery system in a way to bring about an efficient quantitative
oxidation. This may be described as material processing.
There are several systems for the measurement of infrared adsorption.
Requirements of sensitivity, control and economy of design restrict the choices
here. A signal is generated for the purpose of information retrieval. This
operation requires signal processing.
To these functions a re generally added a variety of automatic control systems
which protect the instrument and act as a convenience to the operator.
To aid in the evaluation of available automatic TOC instrumentation,
Raytheon devised the following diagrammatic representations (Figures 4 through
7) of the available instruments in order to compare the functions described above.
Efforts to obtain detailed information of the several commercially available TOC
units has proven to be difficult. Based on the more easily available commercial
literature, it is possible to make some comparisons between devices. They are
all quite different devices and are much less automated than would normally be
expected of devices to be operated continuously on sewage samples. Some of
these designs would indicate that they require considerable maintenance.
Figure 4 describes the H. Wostoff device marketed by Calibrated Instruments,
Inc. which is a continuous commercial TOC analyzer employing oxidation of the
organic carbon by chromate to form CO^. This is essentially the process
employed in the estimation of Chemical Oxygen Demand (COD). The CO evolved
-------
H. WOSTHOFF
Calibrated Instruments, Inc.
AUTOMATION OF OPERATIONS, ALARMS
Zero adjusted with distilled water
Calibration solution of KNa tartrate used to adjust instrument.
Several reagents are consumed at the rate of 1 ml/mm and must be supplied frequently.
SAMPLE PROCESSING
Inorganic carbonate and chloride
removed by sulfuric acid. The car-
bon compounds are reacted with
acidulated chrornate to form COj
which is blended with COj free air.
Cooling coils provided for tempera-
ture control of effluent.
TRANSDUCER
Differential conductivity cell using IMaOH.
Monitors the change in conductivity resulting
from the absorption of CQ^ by the dilute
NaOH solution.
TOC operates in the 0 — 500 mg/1 c/range.
SIGNAL APPLICATIONS, FEEDBACK CONTROL
0 20 ma analog signal output for feedback control.
Meter readout.
SIGNAL PROCESSING
Meter readout.
Two electrode conductivity
measurement in which a voltage
is applied and the current monitored
(J. Searle)
Commercially available system for the independent determination of TOC and COD
COD (not considered here) is achieved by chrornate reduction
Modular design makes it possible to stack operations
Figure 4. Diagrammatic representation of H. Wosthoff TOC
-------
from this process is collected in sodium hydroxide solution and determined by
means of differential conductivity.
Figure 5 is a TOC measuring device marketed by Ionics. This device employs
Flame Ionization Detection (FID) as a means of determining carbon. A low tem-
perature combustion furnace is provided for the determination of inorganic car-
bon.
The device described in Figure 6 is produced by Enviro-Control. It measures
TOD which is indirectly associated with the formation of CO^. The fluidizcd bed
reactor is reported to be designed so as to accept solids up to 100 microns.
Astro-Ideology Corporation provides an instrument capable of measuring TOC or
TOD. This device is described in Figure 7. The Astro-Ecology Corporation
Model 1500 uses the MSA Lira as an NDIR to measure CO from a catalytic com-
« Z
bustton process.
Our information indicates that these are the only devices commercially available
at present for the continuous determination of TOC. It is interesting to note that
each of them employs a different transducer effect. A review of user experience
will be of interest to estimate the comparative problems and advantages of each
transducer choice.
TOC USEfi EXPERIENCES
Field experiences were sought from automatic, on-lint: total carbon, and Total
Organic Carbon instrument users, who had applications somewhat similar to
storm and combined sewage. This criteria, in conjunction with user lists
(accumulated from manufacturers and Raytheon's experience in related projects),
lead to the selection of the organizations listed in Table 4, These selected users
utilized most of the commercially available analyzers in a variety of monitoring
tasks. 25
-------
IONICS MODEL 1224 TOCand TOD
AUTOMATION OF OPERATIONS, ALARMS
Continuous operation option possible, single or multiple grab bed or stream samples are possible. Front panel switch selects TO or
TOC option. Pi2+ instrumentation must be supplied. Span set on known solutions. High and Low Level Alarms,
SAMPLE PROCESSING
TRANSDUCER
SIGNAL PROCESSING
Measures organic and inorganic car-
bon. Milton Roy H2 generator sup-
ply. Purifisd air prepared over a
catalyst bed. 40 microliter sample
injection into a catalytic reactor at
800° C gives CO2 converted to
methane and the methane burned
in flame. Inorganic carbon deter-
mined independently by low tem-
perature combustion. Additive
automatically replenishes acid.
FID linear output
High sensitivity 0 —~ 50 rng/l
Flame Ionization Detector with high
sensitivity in the 0 to 50 mg/l range.
Apparently analog signal is con-
verted to digital making convenient
memory storage and subtraction.
Reconversion to graphical output
on command. L and N recorder
included in the instrument Manual
switch sets range.
SIGNAL APPLICATIONS, FEEDBACK CONTROL
Safety Interlocks to overcome problems due to interruption of gas supply.
Figure 5. Diagrammatic representation of Ionics model 1224 TOC and TOD
-------
ENVIRO-CDNTROL-TOD
Patent restrains use of device for TOO
AUTOMATION OF OPERATIONS, ALARMS
Continuous monitor. Pushbutton calibration
Automatic alarm
SAMPLE PROCESSING
TRANSDUCER
SIGNAL PROCESSING
High pressure fluidized bed reactor
accepts solids up to 100 microns.
Controlled air flow. Provision for
O2 sensor.
Switch selectable output 4 ranges.
Meter readout. Remote recorder
option.
calibrated air. Sample delivered by
metering pump. Water condensation
before sensor.
SIGNAL APPLICATIONS, FEEDBACK CONTROL
Remote multiparameter central display.
Automatic range changes.
Figure 6, Diagrammatic representation of Enviro-Controi TOC (patent restrains use of device for TOC)
-------
ASTRO ECOLOGY CORP Model 1500 TOC and TOD
AUTOMATION OF OPERATIONS, ALARMS
Pushbutton selection of TOC, TOD and TC
Multiple sampling, options available
Alarms
SAMPLE PROCESSING
TRANSDUCER
SIGNAL PROCESSING
Continuous sampling or discrete
injected samples delivered to
catalytic combustion tubes. Air
pump. Handles suspended solids.
Condenses water to form liquid
waste. Gas stream delivered to the
Lira.
ND1R Lira
Non-dispersive Infrared Lira
Lira delivers signal to a meter.
SIGNAL APPLICATIONS, FEEDBACK CONTROL
Meter
Figure 7, Astro-Ecology Corporation model 1500 TOC and TOD
-------
Although the Ionics TGC or TC analyzer cannot readily accept samples with sus-
pended solids, intensive conditioning through use of a batch homogenizer per-
mitted Dr. M. Zeitoun of Dow Co. to analyze mixed liquor sample for total car-
bon. lie found that homogenizatton is the most critical process with dilution
ratios and residence time dramatically affecting the results. In Dr. Zeitoun's
opinion, the complex nature and moderate reliability of the batch, draw and fill
sample conditioning system limits the usefulness of this system. Moreover, the
maintenance experiences of Dupont's staff with the Ionics TOC analyzer clearly
indicated the complex nature of flame ionization detection.
User comments with regard to the Astro-Ecology TOC analyzer point up the
potential noisy data or long response time problems. Several users commented
on noisy data during the combustion of solutions which contain organic com-
pounds that are solids at room temperature and during the combustion of sam-
ples containing large particle suspended matter. Almost all of the Astro TOC
analyzers exhibited corrosion problems that could be corrected by replacing all
metallic components with Inconel and by properly insulating (ad i aba tic transport)
the gas delivery tubing. In spite of the hopeful nature of the Astro TOC analy/.cr,
no user data on accuracy, noise, drift, and response time, and on the ability to
analyze samples with high suspended solids are available.
Liquid phase oxidation via acidulated dichromate to determine TOC as practiced
in the Calibrated Instruments device is prone to interferences from chlorides,
partial oxidations and loss of volatile organics. Accordingly, the accuracy of
this device is poor. Because of the slowness of liquid phase oxidation, twenty-
minute-response times are expected.
In closing, the field experiences of several users show that the complicated
flame ionization detection approach of the Ionics instrument is less reliable
29
-------
and requires more maintenance than the simple, straightforward infrared detec-
tion used by Astro-Ecology, Furthermore, continuous analyzers experienced
fewer problems with sample delivery and injection than batch, sampling-valve-
type instruments. Because of the high suspended solids concentration of storm
and combined sewage, it is essential to homogenize the particulate matter to
such a degree that representative samples can be transported through the analy-
tical system; demonstration projects of Dow Co. have shown this to be a difficult
but solvable problem. To prevent corrosion-type failures, careful consideration
should be given to material selection and insulation, since the hot halogen gases
and condensate have demonstrated their highly corrosive nature in several field
applications. The possibility of changing the stripping acid from 11C1 to ll^SO ^
should also be examined as a method of eliminating or minimi/Jng corrosion
problems. Problems identified in the field applications will be addressed in
the development of a storm and combined sewage monitoring system.
30
-------
section vrr
PERFORMANCE SPECIFICATIONS
Most of the design goals as delineated in subject "Scope of Work" address the
mechanical and environmental aspects of an organic monitor suitable for analyz-
ing storm and related wastewaters. In this section, Raytheon's project staff
expands on the performance design goals in light of the previous discussion on
storm water characteristics. The important o r ga ni c -mon i to r i ng~ sy s te m speci-
fications are presented in chronological order of a storm and subsequent TOC
analyses.
START-UP
Because storms are random events with a conditional probability of occurrence
based on weather forecasts, seasons and geographical location, it is essential
for a stormwater-organic-monitoring system to be on-line, producing valid data
within five minutes or less of the command signal. The slart-up command can
be generated by liquid level sensors or rain gages,
SAMPLE PROBE
Available storm and combined-sewage data-indicate that the vast majority of
the organics present arc soluble and small particulate matter (i.e. , less than
10 mm). Accordingly, the sample probe should accept organic particles
31
-------
less than or equal to 10 mm, reject larger malarial, and exclude grit and other
dense particles. Moreover, the sample probe should take samples from the
midstream section of the storm water or combined sower,
SAMPLE CONDITIONING
It is essential that the suspended matter contained in the sample be reduced to
particles less than 0.2 mm {specific gravity --S1.1) in any direction prior to
induction into the analytical section of the TOC monitoring system because the
particulate material should be within the region of Stokes Law; otherwise it is
difficult to transport and analyze a representative sample. In essence, the con-
ditioning' equipment must be capable of reducing 10-mm particulate matter to
less than 0.2 mm {maximum dimension} suspensions. All stringy, fibrous
material should be suitably "chopped" to conform to the 0.2 mm specification.
SAMPLE TRANSPORT
To effectively transport storm and combined sewage samples, a minimum veloc-
ity of Glcm/sec (2 f-t/sec) is needed. Additionally all connections, coupling, etc. ,
should use sanitary type fittings. A sampling line should never be dead ended.
SAMPLING RATE
Because the organic concentration of storm and combined sewage changes very
rapidly, it. is essential to have sampling rates greater than one sample per min-
ute; it is desirable to continuously analyze the organic content. (Raytheon's
proposed TOC-monitoring system utilizes continuous monitoring practices. }
Slow sampling rates lead to loss of meaningful high-frequency information.
32
-------
RA NGE
During the initial phase of storm events, the increased How scours the eoliee-
tion system and the organic concentration increases considerably above normal
levels—the "first flush" phenomenon. The analytical components of a stormwater
and combined sewage monitor should be capable of analyzing samples with TOC
concentrations from 0 to 1,000 mg/L As the storm progresses the TOC con-
centration usually decreases to about 200 mg/L Wide, but time-phased TOC
variations associated with storm-related waters may make a two-range instru-
ment desirable: 0—400 mg/I TOC and 0—1,000 mg/1 TOC. When the TOC
exceeds 70 percent of the first range, 280 mg/1, the instrument should auto-
matically switch to the higher range; conversely, when TOC concentrations
decrease below 20 percent of the high range, 200 mg/1, the device should switch
into the lower range," However, consistent with the theory of keeping the instru-
ment as simple as possible, the feasibility of a single range (0—1000 mg/1)
instrument should first be investigated. In this single range instrument, it is
expected that, the sensitivity will be adequate to provide the necessary precision.
SUSPENDED SO [JDS
Because of the first Hush phenomenon, the TOC-monitoring system must accept
and produce valid data on samples with a suspended-solids concentration from
0 to 8,000 mg/1. This is consistent with the 0—1000 mg/1 TOC table, since
many of the suspended solids are inorganic in nature. The monitoring system
should be free from plugging by variously sized suspended matter.
SENSITIVITY
During storm events, a 15 mg/1 change in TOC represents a meaningful change
in composition. Accordingly the TOC-measuring system should have a 15 mg/1
33
-------
limit oi" detectability and sensitivity in the low range. For the high range, 1000
a 50 mg/1 sensitivity is sufficient. For a single-range instrument, a 20
mg/1 sensitivity should suffice.
ACCURACY
Whenever a measuring device is used for historical-data acquisition or in feed-
forward control, the highest degree of accuracy possible is usually expected.
Since many stormwater-organic-monitoring applications involve measuring of
the organic loarl, and sometimes address feed-forward applications, obtaining
high accuracy is justified. The best available technology should yield r/?f' of
full-scale accuracy for the organic-monitoring system.
REPEATABILITY
The ability to obtain repeated readings under the same conditions is essential
for feedback, and other control systems. Because many stormwater-TOC-
monitoring applications encompass control of treatment processes, a precision
of :3V;' of full scale is required. Additionally, the system output should be linear
over the range.
RESPONSE TIME
Since the aqueous organic concentration fluctuates rapidly during storm events,
response Lime—especially in real-time control applications—should be as short
as possible. Presently, a 5-minute response lime appears to be desirable.
The principal lag is the inorganic-stripping devices that are necessary for TOG
determination; on the other hand, total carbon determinations should have
shorter response time, since a stripping device is unnecessary.
34
-------
ALARM FUNCTIONS
Stormwater and combined-sewage-monitoring stations are frequently located in
remote areas. Ah such, a TOO instrument may not receive periodic inspections,
and alarms are necessary to alert a central monitoring location to dispatch a
repair crew. Alarm systems should monitor the temperature, sample flow,
oxygen flow, and water flow. If an alarm condition exists, the instrument should
automatically institute corrective action to prevent damage to these devices.
AUTOMATIC ACCESSORIES
The remote location and unattended operational goals of a stormwater and
com billed-sewage-organic-monitoring system clearly dictate automatic calibra-
tion of the TOC analyzer. Accordingly, the system must be capable; of auto zero
and span adjustments on a timed or external command basis.
SHUT DOWN
After the cessation of significant flows, the TOC system should shut down on
llow rate or liquid level signals. The analyzer, moreover, should be readied
for the next storm event. Shut-down procedure's would include automatic cali-
bration, flushing with purge water, and system calibration. The TOC system
should be in stand-by mode, ready for the next storm event.
DATA INDICATING, TRANSMISSION AND RECORDING
To provide information for management and control, the TOC rlata, an analog-
signal, must be capable oi being transmitted to a cenLra 1 location, usually located
several miles from the monitoring station. Accordingly a standard instrument
35
-------
output of either 0—5 volts dc, or 0—100 m Vde, or 4—20 milluimps is necessary
so as to interlace with standard telemetry systems {i.e., tone, frequency shift,
or pulse duration transmitter/receiver systems). In most cases, a local record-
ing device is unnecessary, because of the remote recorders. Local indicators,
however, are desirable to facilitate TOC-analyzer and transmitter calibrations.
PERIODIC EXERCISING
As with other storm-related equipment, it is good practice to periodically exer-
cise the TGC system during extended dry intervals. The developed TOC system,
accordingly, should be capable of operating on a simulated sample upon command.
M A IN1TENA N C E R E QUIR E M E N TS
Most commercially available TOC analyzers require an analytical chemist or a
very highly trained technician for proper operation and maintenance. Many
municipalities simply do not have these highly qualified personnel. Consequently,
the developed TOC system should be designed so an alert operator, plant
mechanic or electrician can service the system with a minimum of training
(about a week). Inevitably, simple maintenance requirements dictate a well
thought out, but simple design which uses ruggedized components.
-------
SECTION VIII
ESSENTIAL COMPONENTS FOR A STORM WAT EK ORGANIC
MONITORING SYSTEM
The following unit operation and hardware describe the "best" choice assembly
that is necessary for the automatic determination of TOO concent ration of
storm and combined sewage.
Table 5, OPERATIONS AND COMPONENTS NECESSARY FOR
TOC DET K RM IN AT ION
Sample Unit Operation
Component
Sample Taking
Sample Transfer
HydroSpaee—Challenger System
Homogunization
Raytheon's liomogenixcr
CO2 Stripping
Raytheon's Simultaneous Delivery
Sample Delivery
and Sparge System
Unit Processes
Combustion
950° Reactor (oxygen)
Reactor Regime ration
Flushing
Combustion Product LJnil Operations
Condensation & Cooling
Conventional Ileal Exchanger
CO2 Detection
Infrared Detection
Automatic Accessories
Sta rt-up/Shut flown
151F Cronoflow Level Detector
Zero & Span Adjust
State of the Art Engineering Design
Data Acquisition
Analog Recorder or Digital/Signal/
Processing
37
-------
Figure 8 illustrates the series nature of these processes. Although a detailed
discussion with regard to the reliability and availability of suitable components
follows, it is profitable to review each function at this point.
In the case of a sample system, the sample preparation requires that the grinding
and blending of the raw sewage should accommodate all the varied materials that
are present in sewage. This must be done with continuous supply of sample and
with homogenization of the components of the Total Organic Carbon that would
normally contribute significantly to the information of interest. The processing
of sample at this point can introduce non-representative signal, or noise to the
system. The reactor system supplies heat and oxygen to support combustion of
the sample. The geometry and material used in this design must consider effective
heat transfer and corrosion. Because of the buildup on refractory materials,
the reactor should be regenerated periodically. The condenser system can be
expected to operate routinely. Fog occurs when inefficient cooling is provided,
and condenser design should maximize heat removal,
'Hie sample system is an assembly of components designed to grind and blend
the raw sewage, transport it to the area of analysis and deliver a quantitative
sample to the area of reaction in a composition that is characteristic of the
material entering the system. A system of this type must be free of hangups and
operate so as to maintain the sample in a liquid or suspension up to the point of
delivery to the reactor.
The reactor is the component in which the quantitative sample is mixed with
oxidizing gas and brought to temperature. At about 950° C, there occurs rapid
evaporation of water and oxidation of carbon to CO^- This process results
in the accumulation of refractory solids that must periodically be purged from
the reactor,
3S
-------
nrE
REACTOR SYSTEM
"1 ("separation
i r7E
DETECTION
L
J L
DRAIN
J L
SAMPLE S VST EM
CONDENSATION
SEPARATION
DETECTOR
co j
EXCESS O.
PERIODIC
FLUSH Of
SOLIDS
RECORDER
WATER TO
DRAIN
Figure S. Block diagram of TOC
-------
In the separation process a mixture of gases is treated so as Lo isolate for
measurement the component of inlerest, carbon dioxide. In order to do this,
the gases arc passed through a water-cooled condenser where water is separated
to a drain. Fog is coalesced and the carbon dioxide passed to the detector with
the unreacted oxygen.
In the detector, a signal is developed. The signal is delivered to a meter and
recorder to provide a permanent record of the Total Organic Carbon in the
sample stream.
The choice of components to be used in the design of an on-line organic monitoring
system involves a series of tradeoffs. These tradeoffs are dictated by concern
for cost, reliability, maintainance requirements, economy of design and the
need to interface with other components.
THE SAMPLING SYSTEM
Analytical systems are designed to operate on either continuous or discrete
samples. Continuous samples are delivered to the sample process at a constant
rate; whereas, discrete samples are generally delivered at a constant volume.
While sampling of a constant mass might be of considerable value in the case of
suspensions, the technology necessary to accomplish this has not been developed.
Even though Raytheon plans to utilize continuous sampling principles, a brief
description of a discrete system was included for completeness.
CONTINUOUS SAMPLE SYSTEM
In the case of stormwatcr TOC, where uninterrupted information is desirable,
a continuous sampling process is necessary. Continuous sampling has the
advantages of delivering an analog signal Lo the readout proportional to the TOC
of the system, and has a better sensitivity due to a constant background.
40
-------
For the; continuous analysis of stormwater, it is necessary to provide means to
transport a representative sample from the sampling point to the TOC system.
The automatic sampling system required for this purpose must have the eapabiii ty
of operating continuously in a stormwater stream containing varying amounts of
suspended solids. To prevent sample line plugging or blockage, a grinder is
required to break down the debris or suspended solids to a pro-determined size
such that unobstructed sample flow can occur.
Commercial units arc available that are suited for this function. They contain a
sufficiently powerful pump to maintain a high flow {typically 40-80 liters/mm)
which prevents solid settling in a correctly sized sample line (nun. velocity iilcm/sec,
2ft/sec). The best grinder pump, based on our experience is the Hytiro-O-Grind
made by tlx) Ilydromatic Pump Co., because it has the ability to grind suspended
solids to approximately-G. 5mm (1/4 ") with an adjustable flow rate.
The ground sample is then passed to a homogenizer (Figure 9) for further pro-
cessing, to fine particles (less than 100 microns).
Although several batch homogenizers are marketed, virtually no small size
continuous duty homogenizers are commercially available. Raytheon's Knviron-
mental Systems Center, recognizing the need, developed and tested a suitable
homogenizer. This device contains cutters as well as grinding surfaces. To
date, Raytheon's team has experienced good particle-size reductions in a
local waste-treatment facility while operating on raw sewage (1 able (i). I he
in-line blender, Figure 5), has operated several months without failure. Raytheon's
experience with this unit has been quite favorable. We have operated it contin-
uously for several weeks.
41
-------
igurc 9, .Raytheon's blender
42
-------
Tabic 6. PARTICLE SIZE REDUCTION
Source
Batch
% <110,^
llomogenizer
Effluent
% < I10(i
Not
(deduction
Secondary
CIarifier Effluent
100
100
—
Haw Sewage
58.2
81. 1
41.1%
Primary Effluent
66.8
85, 5
22. 0'1<
Mixed Liquor
82. 6
87.9
fi.9%
Se condary Slu dge
-------
CARRIER £
SAMPLE £
VALVE
777
£jLL
POSITION 1
I SYSTEM
1 DRAIN
POSITION 2
Moving valve from 1st position to 2nd position for time t, then return
introduces fixed volume V2 of sample into system.
Figure 10. Discrete Sampling Valve Diagram
concentration separately, and the second removes the inorganic carbon from
the sample stream prior to entering the high-tempcrature-oxidation furnace.
To obtain measurements of inorganic carbon and organic carbon, designers
have provided for independent measurements of each, A furnace maintained
at 150° C with an aeid reactor is used to evolve carbon dioxide from the
inorganic carbonate. A 950 0 C furnace converts boLh organic and inorganic
carbon to carbon dioxide. Subtraction of the independent measurement gives
TQC, The use of a low-temperature reactor to directly measure the inorganic
carbon concentration in stormwater is not recommended because of the higher
probability of solid plugging in the low temperature reaeLor. Suspended organic
ir.aLter will noL combust and accumulates in the reactor bed. This will require
frequent maintenance which is generally Lo be avoided in automated instrumenta-
tion.
44
-------
In the continuous removal of the inorganic carbon from the sample before oxida-
tion of the organic carbon, the technique commonly employed involves acidifica-
tion of the sample stream to a pH of approximately 1.5. The inorganic carbon-
ates are converted to COg gas which is stripped from the liquid phase by
vigorous flow of CO free gas.
Of Lhe two methods, the CO stripping technique is preferred because of its
simplicity. One procedure successfully employed by Raytheon requires only u
pump to add the acid and a source of CO free sparging gas. Because of the
z*
extremely small demand, no commercial source of suitable sparging appartus is
available. Raytheon plans to use a specifically designed proprietary device which
has demonstrated its ability to accept sewage with a high suspended solids
content.
SAMPLE AND HEAGENT METERING (PUMPS)
The sample delivery system contains those elements Lhat accurately mix the
prescribed quantity of oxygen with the sample for conveyance into the combus-
tion chamber. The need for continuous operation at n re mole location in the
presence of suspended solids in the sample places restrictions on Lhe selection
of these components. One of the components of primary interest is the pumping
method of transferring liquids quantitatively and continuously. To do this
reliably, the choice of a suitable pump design is of primary concern.
Pumps used in the transfer and precision blending of liquids are of essentially
two types, either centrifugal or positive displacement pumps. Those two
divisions can be further subdivided with advantages and disadvantages listed in
Table 7.
45
-------
Tabic 7, PUMPING SYSTEMS
I'f-.M'lll HK 1A |, l'l lU'lNfi
Advimtu^i-s
3 ^Nailvjjjua^es
iVrurjH^.P
*i h<* rt'nU pump is t5->>-d (nr hitfh £:ilUi!i;t£«' at low piesnures
It 3N noi jn-iJ-priming ufifJi-t' nufftiii* I'lrciunsSiiiKTH, It wiJI pump
abrasive-lnden i. ¦ he - in n:; L1 as well watei'i SJsllen'r.t niatr rials
are iiM'tl in its con^t rurlinn, as well as different typos of sba-Ft
seals for a wide nu^o uf <;h«*mn"a) handling ability.
Hh^ fk-lu-crv H.y-Urm is susvi-ptil>N- !o vupor Inek us well na
Uiput Jf,d iniEjjut slM'iim pressure.
Turbine
"Ibe Ujrbitne pump is sniubiv U) I he cent filial exrept that it \r
;«ble ui produce bi^h pressures 0 f>0 |?si V and low ^alUma^e.
Available in bronze ;»ml easl-mm construction. *!s«l far ehem-
u:;il awl walrr .sjirayihn, pressure booMer iserviee and boiler
feed* II is licit sr 11 -pruning under nnrnul uirciumstanees.
'l lie Cur'biin1 hlnrles iimiirj )«; exfaeuled to wl-h r m ts ^ r'i l-be:s r-
»ntf flow «>'stem.
ixfsrrjvk 133ss11,a
TMl'NT !3i MPlNfi
Advn ntages
aUi'.s
Usilpb runrii
'['he duqibra^m pump tiscs ihe same prim jpU" ;ts the i»irl bund
bellows. On the ihswnU";ird stroke the suehoas valve is oprwd t>s
allow inpuds :»nil riohd.s Ui enU'ti, <•» the tot ward .stroke the dis-
charge valve opens jsurhon elosfsj and the Lacildc] is forced old.
The preer.smn r>f the fjehvery sysfeni is hmifefl when useil with
suh.ilions lh.it i:oHtnm anv ui aft. 'Mus ik f^*eause the
so ititiju!; of bull \Mlvtis are I He main r-nimnieii'ation oi the Ui'hv-
erv j'att:.
1'isUm
'Hsi' piston pump vi nrner:illv u-^*d for pumpinu Sow to medium
viscous liquids id high p c i • s s.ia n-s imuJ p.m i. 11 is sett-
jii'tnunE iU short distances This type of pump is universally
o.^ed m prciH^uri' wash appheaUons,
Wi'i»r ;is Wi'M ;is v,» 1 for machine coolant,,
transform^ vjseous materials from urn- place to nnnlher, or fur
handling materials that will attack rubber ur plasite, ft will nc>i
hnndh- nhrnsivrs.
Hei-;»us«' it will m>t hajidle abrasives such as Munti and j^iavel,
il i;aunifl tn1 used wilb sew:i^«\
KJoxiIjU* UufielhT
The tlrxihle iroiR-Her pmnp h;is nnire uses than anv nfiu-i' lyjw
at /mmjt. U is ^elf-pi ihiuuu will luuvllt- vjsi-uu^ anr! abrasive
briunl-i, will deliver up tu UlU (! I'M at pres'suri'-s to lift p)S5 :nxd
i,s avaiUiSvle in tuiinv ebun'-efi of niatenals* }t may bt' u.sed ;is a
m*nt'r.il irrtiisler pump i*nr WHlfj' urnl iibi'nueuKHf as ^<-1] as sistu-
tary appluiaUonJii.
This wsH poH.sib'.i' wutk to* hi^h floA1 rates bt»l it exjxKses Jh<«
mieinals of the *v*1em ii» the sand and ^jut in u
J'oj'iKtalc ic
Use pi-'n^tuUic pui»p in «.setj fur luw galionaKc extreme
HnmliSrv npjHu'atjori-s. Kulhiiii; comes jn eontnet with the 3u[i3irf
other than 1 he Uahsn^; ilself. There ure no H<-als or gaKkW-s lu
le;iS{ or riiiit;»nsin:iie fhe flmda. ft 5h and will
bandle ti^SH ahrasU'es,
Ui'placetiU'nt of 1 he flexible tubing is L hf L a tin (tnn fnvtor.
46
-------
After a careful consideration of the various alternatives, it is appa runt, that only
peristal tic pumps arc compatible with the precision delivery of streams con-
taining large amounts of suspended solids. For this purpose, we have; used
peristaltic pumping. An examination of the many commercially available
peristaltic pump manufacturers leads Raytheon to choose Masterilcx pumps
because of the high quality, long life and ease of maintenance.
In the case of peristaltic pumping, wear and embritUement of the flexible plastic
is likely to be the limiting factor. The tubing might be expected to last from ;j
months to a year. Maintenance should be provided during intervals of this dura-
tion.
THE REACTOR SYSTEM
*
In order to determine Total Organic Carbon in sewage, it is necessary that the
water be rapidly removed from the sample and that the carbon compounds pre-
sent in the sewage be volatilized, mixed with oxygen and allowed to undergo
complete combustion. This operation requires a great deal of energy delivered
rapidly and continuously. This process occurs in the reactor. Most TOC instru-
ment suppliers use packed-bed reucLors with various design configurations. In
response to the need of high-heat transfer rates, non-plugging and complete
combustion, a unique packed-bed reactor was designed and tested (see Available
Tost. Data section). During the field survey, reactor design deficiencies wore
frequently cited as the principal problem area.
Some reactors employ catalytic surfaces such as cobalt oxide or copper oxide
which are designed to hasten the rate of oxidation. Catalytic systems can ram-
plicate the problem of oxidation by promoting or inhibiting processes depending
47
-------
on changes in temperature and catalyst poisoning due to surface contamination.
For these reasons, catalytic oxidation appears impractical, and a non-catalytic
fixed-bed reactor was designed, built and tested.
Ill our proprietary reactor design, the sample enters the water-cooled
sample tube and is rapidly exposed to the high-temperature environment
of the reactor. Droplets of water are evaporated. Energy is also required
to superheat these gases. Because of the heat requirements of these processes
a thermal gradient develops. This has been studied by making a number of
temperature measurements at different positions in the reactor. These data
are reported in Figure 11.
It is important to provide for efficient mixing and heat transfer to obtain rapid
and complete combustion of the sample entering' the reactor. In order to do this,
a number of compounds were studied and data collected in an effort to recover
total carbon in the sample. These data are reported in Table 8,
In view of the thermal gradient that can be expected in the reactor, it is rea-
sonable to assume that the residence time is somewhat longer than would be
expected from the equilibrium condition of the products at 050°C.
It is necessary that we demonstrate that our reactor design is effective for the
purpose of total oxidation of the carbon content of the sewage sample. In order
to measure combustion efficiency, several pure compounds were placed in solu-
tion at known concentrations. The total carbon was determined and compared
with the theoretical value In Table 8. The results expressed give the percentage
of theoretical organic carbon for the concentration indicated.
48
-------
~ DENOTES CONTROLLER THERMOCOUPLE
95D°C
900DC
850°C
S BOO°C
750 C
700°C
B50°C
SAMPLE IN
SAMPLE OUT
CENTIMETERS
Figure 11. Temperature profile of the continuously operating reactor
-------
Table 8. PERCENTAGE OF ORGANIC CARBON RECOVERED
From
. 1% of Solution
5% Solution
Resocinol
98.6
99.0
Ethylene Glycol
99.4
98.3
Potassium Acid Pthalate
102.0
96.8
Cyclohexanol
100.7
97.3
2 Methal—3, 4, pentanediol
100. 6
NR
Ethanol
99.7
Nil
Oxalic Acid
103.3
96. 5
II is only by virtue of the efficiency of mixing and heat transfer that Raytheon in
able to achieve continuous combustion of samples under these conditions. Because
a high combustion efficiency has been achieved at a flow rate of 3cc/minutc, it
is reasonable to expect that a safety margin exists for delivery at a lower rate. With
delivery at a lower rate there would also be a lesser rate of buildup of refractory
solids which would require proportionally less frequent flushing cycle.
OXIDIZING GAS
hi its present design, the TOC uses oxygen as the oxidizing gas. The oxygen is
introduced into the reactor as a pure gas. Future work is intended to explore
the substitution of air for oxygen.
The oxygen obtained from either a compressed air cylinder (CO^ free) or an air
compressor requires a flow regulator to maintain a constant and precise gas
flow. If an air compressor is used, the CO present in the gas phase must be
removed before the gas can be used as an oxygen source. A caustic liquid
scrubber or solid phase absorption train can be used for this purpose. The
50
-------
absorption technique might be best suited far the stormwater application because
of the potential for automatic regeneration of Lhe system. One possible absorp-
tion train consists of two separate columns of molecular sieves. The first
(Linde 3A) would selectively absorb the water vapor, the second (Linde 4A)
would absorb the carbon dioxide. Periodic heating of the absorption columns to
about 150°C would be necessary.
COOLING AND CONDENSATION
Hot gases pass through a heat exchanger to reduce the water content in the gas
phase, to prevent subsequent condensation in the sensor system. The cooling
capacity of the heat exchanger has to be designed to maintain a constant humidity
to prevent variation of the mole fraction of water in the cooled gas phase. A
condensation trap is an integral part of the cooling system. The trap is designed
to allow continuous condensate flow to waste while at the same time maintaining
a sufficiently high head of condensate water to prevent the gas phase from
escaping from reactor system to the atmosphere. Only a few inches of water
are required for this purpose because the head is sufficiently small to act as a
pressure relief value if excessively high pressures build up in the system. The
actual volume of water required to maintain this lie ad should be kept to a mini-
mum to lessen Lhe error because of the solubility of the CO in the liquid con-
densate.
The products of combustion in the reactor are water and CO . This mix Lure
-------
velocity at the inlet is 18.56 liters/minute. At the outlet, it is about equal to
the velocity of the oxidizing gas, 0.4 liters/minute. Consequently, large
amounts of water must be removed from the stream at the condenser.
Residual water droplets and suspended particulate matter are frequently found
in the flowing gas stream after it leaves the condenser assembly. Filters are
required to remove these suspensions to prevent fouling of the C02
sensors. It is recommended that a vertical cylinder filled with a spherical
solid of approximately 1/4" diameter precede the cartridge filter. It has been
determined that the life of the filter is considerably lengthened because the
major portion of the water droplets are mechanically removed in the vertical
cylinder,
THE DETERMINATION OF CARBON DIOXIDE
CI assical chemistry offers us a number of options for the determination of car-
bon dioxide. Carbon dioxide is an acid which has some .solubility in water and
base. It can be determined by measuring the change in conductivity of alkaline
solutions. It can be used to precipitate barium and calcium salts. Carbon
dioxide transmits visible light but absorbs infrared energy in the same way as
most organic compounds. In contrast to similar compounds such as NO^ and
SO , it is an unreactive gas with a relatively simple chemistry.
INFRARED DETECTION
Organic compounds all have a distinctive infrared spectrum. The spectrum is
more or less complicated as the compound is complicated. The simple com-
pounds such as methane or carbon dioxide have a simple infrared spectra.
Because of this, they are easily determined by Infrared spectroscopy. There
52
-------
are several approaches to the determination of various species by infrared
spectra. These will be reviewed later. For our purposes, non-dispersive
infrared detection was preferred. Non-dispersive infrared (NJ)IR) is an optimal
technique for the measurement of a single component in a uniform and fairly
simple gas stream. The principle of selectivity is in the design of the sensor.
The gases flow through the light path resulting in energy changes as a result of
absorption in a closed capacitance system.
The usual complaint about non-dispersive infrared (NDIIt) is that it is not suffi-
ciently versatile for the needs of a spectroscopic analyst. This, however, is
not the problem in our analysis. For the purposes of an instrumental method
for the determination of TOC, there is an interest in the analysis of one and only
one component to the exclusion of interferences. NDIll gives us the method fo r
obtaining the maximum signal from a minimum component concentration. The
method of NDIli detection lends itself very well to the determination of CO^ in
our system.
The choice of the MSA Lira as an infrared detector represents a compromise
among sensitivity, versatility and ruggedness of design. It provides an accept-
able signal and sensitivity. This makes it possible to obtain a signal and to
develop the data on concentration of TOC as low as 72 irtg/1 full scale with our
present assembly. This is substantially better than necessary for sewage of
almost any kind. Deionized water contains as much as 7 mg/l organic carbon as
indicated by 10fjf full scale. The system has an easily modified range of opera-
tion and is easily adapted to the variety of operations and measurements that
are a part of an experimental program.
53
-------
The classical approach to infrared spectroscopy consists of the measurement of
monochromatic signal from a black body radiator such as a i>lobar. The mono-
chromatic energy is developed through an alkali halide prism or by adjusting the
spacing of an interference grating. This permitted a monochromatic infrared
signal to reach the sample. The grating or prism is programmed so that a con-
tinuous spectrum of energy reaches the sample in time. The absorption in the
sample is indicated by a measurement of photoconductance of one of several
possible infrared sensors, Solid state narrow band gap materials such as PbS,
PbSe and InSb are common sensors; thermistors, thermocouples and other heat
detectors are also used. The absorption of energy by the sample results in a
decrease in signal at the sensor and a consequent change in some electrical
parameter of the sensor. There are several circuit designs in which a change
of this kind can be monitored. The signal is usually delivered to a recorder.
Continuous spectra are used for both qualitative and quantitative signal measure-
ment. The optics of systems of infrared spectra are frequently crystals of
alkali halide. They tend to require continuous maintenance and freedom from
moisture. The instrumentation necessary to implement the classical measure-
ment is expensive and does not lend itself to continuous monitoring equipment.
There are exotic approaches to the measurement of small quantities of hydro-
carbon species in an atmosphere. Present designs available from Diax Corpora-
tion use monochromatic infrared lasers to energise gases flowing in a tuned
cavity. There is a sonic source and a microphone in either end of the tuned
cavity-—all are tuned to the same sonic frequency. When the gas flowing through
the tuned cavity absorbs energy of the monochromatic laser source, detuning
of the cavity and decay of the microphone response result. The results
can be used with several species and several lasers. Any species in the gas
54
-------
stream can be effective in detuning the cavity. This system with readout can be
obtained for $40, 000. As such it appears to be prohibitively expensive for sewer
monitoring requirements.
By contrast, there are relatively simple systems available from organizations
such as Infrared Industries and Wilks that can be obtained in useful form for
less than $1000. Instead of elaborate systems of spectral dispersion and selec-
tion they use filters. They have a relationship to infrared spectroscopy that
colorimetry has to visible spectrophotometry. One often sacrifices sensitivity
in this type of system, but sewer monitoring may be one of those areas where
sensitivity may not be the primary concern in the infrared detector. Our choice
of the Lira NDIR might yield to cost considerations and size limitations in future
design. For the time being, however, the MSA Lira serves the purpose.
Experience indicates that it lends itself well to this analysis. It is relatively
trouble free.
ALTERNATIVES TO INFRARED DETECTION
Wet Chemistry
For the determination of C02> infrared analysis would appear to be the optimal
approach. Several alternatives are possible and have been used in various appli-
cations.
The Wosthoff instrument for the determination of TOC provides a solution of
sodium hydroxide that is passed through a system of differential conductivity
55
-------
cells. The CO^ evolved is passed into the flowing alkali system. The change
in conductivity resulting from the formation of ions of CO is monitored and
reported as proportional to the total organic carbon burned in the sample.
Delta Scientific collects CO^ in barium hydroxide solution and determines the
light dispersed by the precipitate.
These are essentially wet chemical methods well known to ancient arts, such as
winemaking. They suffer from the need for standardized reagents isolated from
CO^ in the atmosphere. The solution must be delivered to known concentrations
by skilled technicians. They must be standardized and given more attention than
is usually desired for maintenance of automated instrumentation.
Vapor Phase Chemistry
There exists a system (available from Lion Research Corporation) for the
determination of CO by passing a stream of unknown gas in a constant background
into an ionizing radioactive (americium) environment. The higher cross section
of CO causes a greater ionization in a counter of suitable design. This is eom-
2
pared with the ionization in an input stream of unconLaminated gas. This is a
technique that is used at fairly high levels of CO^ and is employed in Gemini
space flights as a means of maintaining the level of COg in the confined atmos-
phere.
In a private communication from Dr. John Litzkowitz, we learned that a patent
application on a C09 detecting device has recently been delivered to the U.S.
Patent office. Because the detail was proprietary and not disclosed, it must
be inferred from similar technologies. A crystal is made to oscillate at its
characteristic piezoelectric frequency which is in part determined by its total
mass. The adsorption of CO on a resin-coated crystal changes it.s mass and
56
-------
consequently its frequency. By monitoring the frequency, it is possible to
monitor the CO . The resin is periodically recharged by flushing with CO
z • 2
free gas.
Previously, in the early 30s, Zisman of the Naval Research Laboratories
implemented this technique to detect adsorption on oscillating charged metal
electrodes. His work wan referenced by Chambers and Eaton in an effort to
design an odor deteetor in the late 50s. The quartz crystal was used by King
in his effort to design an organic and water monitor. It has been commercially
employed by Dupont in the Dupont water monitor.
The problem with all these devices is that they depend upon a change in mass.
Any mass will bring about a change in frequency. They are as good as they are
selective. Oil and hydrocarbons will absorb and change oscillating frequency,
also they require frequent calibration and recalibration. Since they are essen-
tially integrating devices, the absorbing surfaces become saturated resulting
in non-linearities and the need for frequent purging.
These problems may be overcome by careful design. In view of the long suc-
cessful experience with other systems, the measuring principle will have to be
confirmed by the test of experience before it can be recommended.
(19)
Recent publications suggest tlu; possibility of a reducible gas sensing elec-
trode system. There is not a great deal of experience to establish these as
useful techniques. As it appears, lanthanum fluoride was deposited by means
of vacuum deposition on metal substrates of various materials. Signals were
developed across this material in the presence of several oxidizing gases. The
current measured from anode to cathode was found to be proportional to the
57
-------
oxidizing gas. C09, NO^ and SO^ were among the gases with a reasonable sig-
nal.
There is hardware available from Johnson Laboratories for the deLermination
of carbon fourteen in CO^. If the carbon dioxide is enriched with carbon four-
teen or concentrated in someway, it is possible to determine carbon by the com-
bustion of any organic material. In the case of sewage, a measurement of this
sort would be of considerable interest because it could bo used to determine
energy source products as a source of pollution. Petroleum and coal are radio -
actively dead carbon sources but living matter has a representative carbon four-
teen concentration. The difference between to till carbon and radioactive carbon
would be the energy source carbon.
There are other possibilities that are even more speculative than the above but
it is evident from the review above thai infrared is the only practical choice and
among the various techniques, non-dispersive infrared appears to offer the most
practical compromise at the moment.
For the purpose of providing a comprehensive overview of the information in the
previous paragraphs, we have prepared Table U. This also includes some addi-
tional methods employed in carbon determination. Here we have addressed the
various options and evaluated these options as they apply to sewage,
SAFETY AND ALAKM SYSTEMS
It is possible to anticipate several problems in the abnormal operation of the
TOC instrument. The occurrence of these events should give cause for alarm
and attention. It is a characteristic of a well designed automated instrument
that the alarming event should be signalled and corrected by automated instru-
mentation.
58
-------
Table 9. THE DE T E li M IN AT ION OF CARBON IN S E WA G K
Phasi-
Ounpimwi
Method i>f DulenrtiruiUon
Commem
L itjuiii
Ca rbon t'omiiuuntis
Eik soluli in llarium
IIroxuh' measurer! bv
nepluilomeLry
II) Utile rentiul CortffiK'iiviLv
Nol for Uremics
Solids dispense ilit1 energy<
Hequi res supply ami mainte-
nance of roa^r'nt Lsohilet f fmm
i"0;j in nil'. riubicuL to numer-
ous iiUrj'JtMvmvH,
|{2 in ;ti r< .Sub|ret to numer-
ous ipieri'ei'enires.
llci*
IVU-t ha ne Cli^
CnrNm fHoxuIe
\) Flame soni nation tleleKinn
li) Temperature f3g fornbjisiiblr
mi m lure
1 } lofiMs rt-el I >eter(or
a) PrismT ^fatihiter
devices
to) IS'ondi spri'si ve I til rarer I
iMHlt)
rj Hiiix l?t*sf rarbim.
Requires fvtk- rerlnolion oj'
l'br»H f>|HM«!c.
i t'S ct>jiibusts!j] r s
which h:is 1 c> hi- tbrmetl lirsl.
I'^KSJVi" iiclrt'tor HVSlf'Ill <*\'
hi^h \ i ¦ r^ii ti 1 iiv of ennsirtera b3t*
l*ai-sivr Hrfeeho' Nvsti-m oj
hiu,h M'nMS i eit v ;jtH seler! ivitv
< ii ;i n-eiJ jr In Hid si ruel n re .
hMreme sensitivity li> im\-
liis'c^ n| rsti'rnir-Iv ^mall rim-*
rent fit!aorss U pari per bMbmi}
lnif expensive.
Vitv reliable nivi pnssivt* svs-
Icm . ]{< '<- pern Is
mi Chi- sit lion i*l ihi' rtju--
u» ratiuitivr
Ni'W U'Hini'juc liUt'lv In lie Hon-
sclectm' H5 ap|)HrniaLion to estimate nppH-
i• :i 11j I li V-
There are several sources of difficulty in which the operator might wish to shut
down the instrument. One of these is a thermal runaway where the temperature
controller is open circuited or disconnected resulting in a signal that calls for
continuous heating of the reactor. This would cause destruction of the instru-
ment. We have provided for this eventuality by breaking the main line power
59
-------
when the reactor reaches 1000°C, which is beyond the range of the control
thermocouple.
A second source of difficulty might result from corrosion of components in the
reactor. High temperatures and undefined acid conditions or the reactors can
result in corrosion of the water-jacketed sample tube. The offset would be to
deliver water from the cooling tube to the hot reactor resulting in a rapid cool-
ing of the reactor and steam generation. This would cause high pressure and
vapor evolution. Here a low temperature shutoff, would be of value, closing
down power and flow to the water line. This should bo associated with an alarm
indication.
if sample flow is discontinued for any reason it would bo useful to have this
information available. In this case, the graphical readout would tend to zero on
a continuous basis. It would be sufficient withouL shutdown to have a visual
indication of the effort. A signal could be l.;ikcn from the readout Utsclf .simply
by placing a low level signal switch on the recorder -
A more serious condition would exist if the sample was delivered Lo a plugged
reactor. This would result in tin explosive buildup of pressure. A pressure
gage would be a good indication of this condition. The'present design incorpor-
ates a tygon tube connection that would serve as a blowout under a condition of
excess pressure. There is also a pressure switch.
If oxygen were to stop flowing, a rotomeler flow indicator could be used to indicate
this and the signal developed to give a readout and Lo shut down the power Lo the
instrument.
Water flow to the heat exchangers is about as important as oxygen flow and
might be indicated in the same manner.
GO
-------
in spite of the importance of automatic conu olo md safety alarm, it is essential
that we establish some routine preventive maintenance and inspection. The
failure of a plastic tube or a pump might result in a serious nuisance such as a
sewage spill. This prospect could be eliminated or greatly reduced by adherence
to a routine inspection and replacement schedule, similar to the requirement of
automobile and airplane maintenance.
The alarms, delineated in Table 10, arc commercially available from several
commercial suppliers at this Lime. Raytheon's staff has selected only the thermal
runaway alarm system. As the laboratory test program matures, the remaining
alarms will be appropriately selected.
hi addition to alarms, automated features can be incorporated with digital con-
trol and signal delivery as described in Figure 12.
PAPER TAPE
PRINTED DATA
TOC
100 MICRON
PARTICLES
STORMWATER
TOC
COMBINED
STORM
OVERFLOW
AUTO
ZERO /
SPAN
AUTO
START/
STOP
HQMOGENIZER
20 CHANNEL
AUTOMATIC
RATA ACOUISTION
S VST EM
AUTO
CALIBRATION
HYOROSPACE
CHALLENGER
SAMPLE DELIVERY
SYSTEM
SIGNAL FROM
BU CONTROL
PANEL
Figure 12. Field system diagram
n
-------
Table 10.
ALARMS
Alarm
Condition
Sensor
Action
Hazards Due
to Failure
Failure of temper-
Thermocouple
Shutdown—repair or
Thermal overrun
ature control of
replace tempera-
will melt the In-
reactor
ture control circuit
conel.
Corrosion of sam-
Diaphragm pres-
Increased pressure
Blowout of sample
ple cooled water
sure switch
causes automatic
line
jacket
shutdown
Sample flow inter-
Recorder
l,ow level recorder
Uncontrolled sew-
rupted
switch causes sam-
ple pump shutdown
age spill
Ileactor plugged
Diaphragm pres-
Increased pressure
Blow off of the
sure switch
causes automatic
shutdown
sample tube
( )xvgen iluw stopped
Capacitor sensor
Sit;ii;i i light and
Reactor will have
on rotometer
shutdown
incomplete com-
bustion distilling
carbon compounds
which plug the
reactor and con-
den sor
Water flow stopped
Capacitor sensor
Signal light and
Hot gases will flow-
-
on rotometer
shutdown
to the ND1R contam-
inating the oplical
system
-------
SECTION IX
A SYSTEM DESIGN FOR SAMPLING AND TOC MEASUREMENT
OF THE STORM EVENT
In order to fully understand the requirements of a successful, rapid on-line
storm water organic monitoring system, it is advisable to review the proposed
operating objectives;
¦ Large organic debris such as mattresses, rubber tires and two by fours
cannot practically be included in the sample system and must be elimi-
nated from sample intake orifices,
¦ A statistics Sly representative sample containing suspended solids must
be delivered from the sewer to the instrumentation without solid segre-
gation because of settling or high velocity ontrainmcnt.
¦ The sample must he put in a condition for acceptance bv the instrument-
ation that is ready to proceed with uninterrupted, continuous analysis,
¦ When the storm even! is over, it is necessary to flush the equipment
and shutdown the system so that it will be left in a condition to be
turned on when necessary at a later storm event. In this section, we
will consider these problems and reference the available technology to
implement the design.
83
-------
SAMPLE SYSTEM
Because of the significance of the suspended solids a carefully selected sump ling
system will be required for the delivery of a representative sample to the TOC
system. This has been a problem of great concern to a number of people and
there are a number of devices that have been used to sample sewers. Most of
them are intrusion devices that would be demolished by the debris of a storm
event. They have been evaluated in some detail by Shelley and Kirkpatrick^^
and are reported as inadequate for the purpose of making measurements on
storm and combined sewage. With this problem in mind, Dr. Shelley of
Hydrospaee-Challenger has recently been contracted by the Storm and Combined
Sewer Technology Branch of the EPA to design a sample transport and handling
system that would be useful for the purpose of taking representative samples
during storm events." The system which Dr. Shelley is assembling for this
purpose is designed for storm sewer sampling and is intended to be adapted for
use in the rapid, on-line organic monitoring field demonstration at Boston
MDC-BU Station.
In a private communication with Dr. Shelley, it was learned that the system he is
considering at present is as follows;
A number of tubes are placed along the perimeter of the sewer
oriented so as to present a downstream intake to the system with a mini-
mal intrusion into the pipe. This arrangement eliminates gross solids
and transfers the significant small size particles and organic liquids
to the sampling system. Sample entering the system is transferred at
velocity of 305 cm/sec(10 ft./sec.) by means of peristaltic pumping.
Fluidic diverters are used to switch streams. The control and elec-
tronic system are completely solid state. Each sample intake has an
64
-------
independent pumping system and stream divertnr. The samples are
delivered to a central compositing receiver which stores the sample.
After several minutes of this operation, water is passed through the
lines and a flush cycle is operated. This is completed end a new
sample composite is synthesized.
The system as presently designed is intended for the preparation of discrete
samples. It will be necessary to make several accommodations for the con-
tinuous sample delivery. The present system has a redundancy of sample
delivery lines to provide a statistical sample. It should not be too difficult to
program a portion of the system to deliver the sample while the other portion
flushes. This should meet the needs of the continuous sample delivery
requirement.
Since this sample delivery system has been designed for the purpose of pre-
paring samples for analysis of data associated with the storm event and has
been funded to meet the need for such a device, there is no reason to look for
another.
TOC ANALYSES
After extensively reviewing the several alternative systems and selecting the
components for the monitoring of storm and combined sewage organic content,
it is appropriate to generalize the observations.
All carbon compounds oxidize, When this oxidation is complete there Is only
one product, carbon dioxide. The universality and simplicity of this chemistry
is well suited to the principle of design of a Total Organic Carbon analyzer.
65
-------
Because this product is stable, easily handled, non-combustible and generally
well behaved, it is unnecessary to look to other compounds on which to develop
an analysis of TOC.
Several TOC analyzers are designed around the catalytic conversion of carbon
dioxide to methane and the further determination of methane by flamc-ionization
detection. The only justification of this approach would have to be based on the
need for sensitivity to carbon concentration less than 10 mg/l. Sewage rarely
contains so little carbon. The additional complication required for this approach
are .sufficient to exclude it from serious consideration as a means of determining
Total Organic Carbon in storm sewage. The direct determination of the oxidized
product of combustionf carbon dixoide is preferred for the purpose of continuous
measurement of organic carbon in sewage. The greatest problem that had to be
solved is sample handling and delivery. A considerable study has demonstrated
that it is possible to prepare the sample to the necessary degree of uniformity
Lo obtain totally acceptable data on a long-term basis with a dynamically flowing
sample (see Figure 13). The present TOC has been designed with a ND1K
detector that is capable of producing an uninterrupted signal to a recorder. For
our laboratory model, a number of automated functions have been programmed
into the internal control, operation and protection of the instrument. This
arrangement is diagramatically presented in Figure 14.
In summary, the presented literature clearly shows that from all the available
methods of measuring the organic content of wastewaters, vapor phase oxidation
is the only method suitable for reliable, on-line, unattended operation. Among
the options of monitoring the products of combustion, infrared detection of COQ
£s
is the most desirable technique because it is straightforward, reliable and
66
-------
addressable by ruggedized equipment. However, the inability to accept samples
with suspended solids and the accumulation of refractory material in the com-
bustion reactors have impeded the development of a reliable, on-line TOC
suitable for unattended monitoring of the organic content of wastewaters.
Based on available technology and Raytheon's experience, (see Prototype Data
section}, development of a suitable stormwater and combined sewage organic
monitoring system is feasible, if the suspended solids and refractory material
problems are specifically addressed during design stages. The remainder of
this report discusses Engineering Model Data and our schedule of testing and
reporting on the developed system.
67
-------
7DQ
E30
5G0
490
420
o
o
f—
=*
BO
E
280
210
140
70
0
¦ ' f f "
i - -
.1 :
- ; t 12 3 .
- 1 " I j i
t ; i
i,.
: j
¦i J -
i
_ I .
--
: ~!*
i . i .. 4....... i
; j 1
I : ! i
*
h
l
. - i i i
t... , ... i i
'.! i j
i i
GROU
NO SAMPL
E L.
-
1
|
t
...
" T ! :
1 i
f
~~r
i ¦ i
* 1
- ! : j
I
!
-
»
!
¦¦¦"'
4 - -
.' V 1 : ~
• : f
i
- :
i f
j ;
1
" » -
1
V:
i-
H - -
x ... r
, -1 i
,ftm i: DDno i ri m.n 1 '
i
—i—
; ! :
! i
t I
! 1
-
1
-
-t
: I |.!l!
!
Mil:
I -
t -
1 lit
-4
700
630
560
<19G
CJ
o
420
h—
E
350
280
210
110
70
0
UNGROUNDSAMPLE
DATA ERROR 15-25 nig/I
Figure 13. Continuous data on a flowing sample
68
-------
o
AUTOMATION OF OPERATIONS. ALARMS
1. Push button startup operation. TOC operation, TC operation,
zero and span.
2. Automatic sample handling and delivery system. Operates on
suspended solids.
3. Automatic temperature control.
4. Precise and reproducible peristaltic blending.
5. Precision flow .ontrul on oxygen supply.
6. Protection circuits.
7. Overtemperature indicator light on reactor.
8. Push button automatic reactor flush.
9. Operator resupply of reagent required no more than once a
month.
SAMPLE PROCESSING
TRANSDUCER
SIGNAL PROCESSING
Inorganic Carbonate is removed by
acid. Sample blending operations
controlled by Peristaltic pumping.
Sample degassing achieved by pro-
prietary high velocity gas stream
transfer. High temperature com-
bustion forms C02 Water and
solids removed from samp1'1 .neam
to deliver C02 and Oj tn Lirj.
NDIR Detector (Lira)
Operates using a continuous stream CDj. The
optical signal is chopped at two cps, filtered
and sensed by the expansion chamber detector.
Changes of CO2 concentration cause the opti
cal signal to the expansion chamber to be
attenuated and the capacitance of the expan-
sion cell to change as the concentration changes.
The ac signal from the Lira is con-
verted to a dc signal and delivered
to an Esterline Series S recorder
havtng an impedance of 10 meg-
ohms.
SIGNAL APPLICATIONS. FEEDBACK CONTROL
Signal used for information.
Figure 14. Proposed TOC
-------
SECTION X
PROPOSED OPERATING MODE TEST SITE INTERFACES AND
FUTURE ACTIVITIES
START-UP
The start-up of the analytical and sampling equipment can be sensed by a suit-
able level indicator as is presently done at Boston's MDC-BU installation. A
bubbler tube device measures the wet-well level (loop 2, Figure 15), When this
level reaches Home, predetermined value, a pressure signal is delivered to
indicate this event and to start up the treatment processes. A device to deliver
this signal is the BIF Cronoflow Transmitter Model 2.'il-2Q which provides
signal telemetering where that is of interest. Raytheon plans to use the wet-
well level signal to start storm water sample flow to the organic monitoring
system.
Specifically, the TOC analyzer will be standing by, already at operating tem-
perature; the sample delivery and conditioning system will contain clean flush
water (see stand-by discussion). When the wet well reaches a predetermined
level, the sample delivery pump {Figure IB) will start transporting storm water
samples to the analytical train. The data acquisition system, automatic sample
takers, and strip chart will also be started up at this time.
A signal received from the BIF transmitter might he used to operate a sequence
timer that could be programmed to perform the necessary start-up sequence.
70
-------
©-
STATION
STARTING
SIGNAL
AT
fltMOTf
MANNED
STATION
RESIDUAL
CHLORINE
NtQCl
SOL
FEEDER
WATER
FILTER
PUMP
CONTROL
OVfflFLOW
[4FUMP
« FEEDERS)
DETENTION IKS S
TO RfvER
CONTACT
PROBE
WET
ttfLL
COARSE FINE
SCRftNS
RECIRCULATE
D«A N
COMBINED
SSWRS
Nare Slaiidud linttumrnunnn S'fiH&oSs and WMtitoMiiuri u\ \ht ISA
Figure 15. BU stormwater detention facility
-------
COMPOSITE
SAMPLE
refrigerator
STORAGE
OXYGEN
AIR
COMPRESSOR
Hv3RG$PACE-challenger
SAMPLING SYSTEM
D^AiN
TOG
analyzer
HCMGGEN12ER
DRAIN
H yO FLU SH
FLUIDIC
DIVERTER
V2
V3
PERISTALTIC
PUMP
DRAIN
COMPRESSED AIR
V 4
dump to
DETENTION TANKS
AP
MULTIPLE
SAMPLING
POINTS
COURSE FINE
SCREENS
Figure 10. Application of sample system and TOC monitor to MDC-EU installation
-------
A sensor of this sort would respond whenever the sewer level exceeded some
predeLermined level.
OPERATION OF TQC
When a suitable sample has been taken it must be delivered to an operating TOC
instrument. The operation of the TOC instrument requires a continuous .supply
of air, cooling water and power. For the purpose of rapid response to the
storm event, the furnaces must be at operating temperature, the sample tube
and condenser must be cooled, and air or oxygen must be supplied for the
exercise of the delivery system,
There is a need for peristaltic pumping throughout this instrumentation, if tygon
pump vanes are allowed to remain compressed for an extended period of time,
as would be the ease of an inoperative peristaltic pump, they deform and even-
tually become brittle and crack. Since both the sample delivery system and the
TOC instrumentation make extensive use of this type of pumping, our present
design commits us to some type of continuous sample processing. If continuous
sample processing is necessary, it is convenient to use water from the eon-
denser-cooling operation as a sample, in a ease of a saniLury system where
relatively clean water is available at a short distance, only pumping would be
necessary.
PERIODIC OPERATION
With our present experience in the long-term operation of the stormwater TOC
and sampling systems, it is too early to predict whether it would be more desir-
able to operate continuously with sample delivery or remain idle with periodic
exercise, It is evident that continuous furnace and infrared detector operation
73
-------
will be necessary simply to eliminate the Lirno to arrive at opt;rating temperature
when the instrument is needed for stormwater monitoring. This should not be
a problem and can be implemented without serious inconvenience.
The exercise of the sample delivery system is possible by delivery of a signal
from a 12-day clock timer. These nre in common use on home water condi-
tioners and have a good record of reliability. The clock timer can be used to
activate a program timer that will autozero, auto.span, and deliver water to the
operative peristaltic pumps, A two-hour program timer such as those avail-able
from Singer's Industrial Timer Division should be adequate for this purpose.
While our ultimate design is expected to operate at a variety of locations, we
intend to conduct our first tests in the Boston MDC-BU installation. This is,
in fact, a well designed and well maintained stormwater faei lily where personne1
might be expected to observe, report and provide data and operations so that
maintenance etui be supplied as needed.
Figures 17and 18 :iro illustrations or the MDC-BU operation. Figure 16 gives
a diagram of how this facility might be brought to interface with the Ilydrospaee-
Challonger sample system. The diagram contains further description of how
the sample system might interface with the stormwater TGC, Valves that inter-
face the operation of the TOC with the sampling system are described in Figure
16. These valves are inserted to permit the diversion of air and water, to flush
the sample delivery system and to deliver exercise sample to the TOC when it
is intended to operate the instrument in the exercise mode. They can also be
programmed to flush the sample system with both air and water as required
by the Ilydrospaee-Challonger design.
74
-------
Figure 17. Boston's BU storm detention and chlorination station
75
-------
CHLORINE DETENTION
tanks
Outfall
STORM
overflows
OWNER
COMMONWEALTH OF MASSACHUSETTS — FRANCIS W, SARGENT, GOVERNOR
METROPOLITAN DiSTRICT COMMISSION — JOHN W. SEARS COMMISSIONER
ASSOCIATE COMMISSIONERS: J.F, HAGGERTY, A,T. LYMAN JR., V.R O'BRIEN, M. ROSENBLATT
CONTRACTOR LOCATION
DE MATTEO CONSTRUCTION COMPANY NEAR BOSTON UNIVERSITY BRIDGE
ENGINEER CAPACITY
CHARLES A. MAGUIRE a ASSOCIATES, INC. 233 MILLION GALLONS PER DAY
Figure IS. Boston's BL" storm detention find e hi urination station, flow diagram
-------
SHUTDOWN
If it should become necessary to shut down the instrument, as might bo the: case
through a long winter when storm flows might bo moderated by long freezing
conditions, the instrumentation should bo operated on water continuously for a
period of at least an hour. During this interval, lines will be flushed and samples
purged from the delivery system to eliminate settling and coagulation that could
be expected to plug the system on start-up. Start-up ;iftor a prolonged shutdown
should be programmed Lo follow an inspection and maintenance routine.
ENGINEERING MODEL DATA
Components were selected for the design and assembly of an instrument for the
reliable measurement of Total Organic Carbon (see Figure 13). Data (see Table
11) were collected in Raytheon's TOC development laboratory. After the instru-
ment was found to give a satisfactory determination of Total Organic Carbon
concentration, the instrument was moved to the municipal wastewater LreaLment
plant at Cranston, Rhode Island (Figure 19) where it was tested con-
tinuously in unattended operation with raw sewage for a period of sovertd weeks.
In the above tests we have shown that sample handling, oxidation reactor and
hardware assembly are well designed for the purpose of monitoring Total Organic
Carbon, TOC. Early problems with signal noise have been reduced substantially
by processing the sample with continuous blending through Raytheon's blender
(Figure 9). The response time of the present instrumentation is less than five
minutes when going from a signal that is indicating GOO mg/i of carbon present as
cyclohexanol to water and then back Lo sewage. This is seen in Figure 20.
77
-------
Figure 19. Municipal wastewater treatment plant, Cranston, RI (sheet 1 of 3)
-------
. ¦ , owner treatment plant, Cranslon, M
y[«ure 19- Municipal ^t.wei
I (sheet 2 of 3)
-------
Figure 19. Municipal wastewater treatment plant, Cranston, RI (sheet 3 of 3)
-------
Table 11. DATA SHi:KT FOR CHARACTERIZATION OF SEWAGE AND CORRELATION WITH TOC ANALYSIS
1ST DAY 2ND DAY
Tt
T2
T3
T4
T1
T2
T3
T4
MEAN SO
PROPOSED
TOC
TC
Tt C
TOC
TC
TIC
TOC
TC
TIC
TOC
TC
TIC
TOC
TC
TIC
TOC
TC
TIC
TOC
TC
TIC
TOC
TC
TIC
TOC
REPLICATES
BECKMAN
TOC
1
2
,
3
4
5
MEAN
STAND DIV.
TOC
SOLUBLE
SETTLEABLE
SOLIDS
SUSPENDED
SOLIDS
VOLATILE
SUSPENDED
SOLIDS
><
X
TOTAL
DISSOLVED
SOLIDS
X
X
MICROSCOPIC
EXAMINATION
X
X
-------
600 (Wl CARBON AS EVCLOHEXANDL
600
500
400
SEWAGE
cc
300
2G0
WATER TO CRANSTON
SEWAGE
100
6
7
12 13 14 15
8
9 10 II
5
3
4
0
a
01 23456789 10
ELAPSED TIME IN MINUTES
Figure 20, Response time of organic cyclohoxanol solution to water and to sewage
-------
FUTURE ACTIVITIES
Our immediate effort will be to operate the present design continuously and with-
out attention on Lhc ground, blended and flowing raw sewage at the Cranston planL.
Instrumentation associated with municipal sewage has been plagued by assembly
failures, breakdowns mul unreliable performance. With this background, it is
essential that data be obtained with a view toward establishing a sense of con-
fidence in the assembly and its design. The forthcoming laboratory evaluation
program provides lhc opportunity to document the following important instru-
mentation parameters:
¦ Accuracy
¦ Repeatability
¦ Long-Term Drift
¦ Signal-to-Noise Ratio
* Sensitivity
is Correlation of the above with Sower System Characteristics of
• TOC Soluble
• Set tic able Solids
• Suspended Solids
• Volatile Suspended Solids
» Total Dissolved Solids
o Microscopic Kxamination.
In order to obtain information on the instrumentation parameters listed above,
Raytheon's staff has designed the experiment in Table 11.
Once Liie technology lias been established Cor the above characterization, samples
will be obtained from MDC-BU for the characterization of storm water during
83
-------
the storm event. With the information obtained from these data we will be in a
position to design instrumentation for monitoring stormwater TOG.
Because combined sewage is somewhat similar to dry weather raw sewage, the
immediate objective of our program is to obtain information on the reliability
and reproducibility of prototype TOC instrument on the samples obtained from
the grit chamber of I.he primary settling basin at Cranston. Conducting our
experiments at this facility will insure a "fresh" supply of samples that contain
similar suspended solids.
In order to do this, it is necessary that stationary sample be delivered to the
instrument so that a constant sample concentration can be drawn into the Total
Organic Carbon analyzer. The sample will then be stabilized by bringing the
sewage to a pll<2 so that samples are continuously analyzed over a period of
time from the same stabilized, mixed reservoir. Results of this effort will
provide an estimate of instrument variation over the period of time (drift) and
the reliability of the proposed TOC instrument. During this phase, hand-con-
ditioned (homogenized) samples analyzed with sufficient number of repetitions
will serve as the primary standard. Table 13 shows a sample data form that
will bo used in accumulating the laboratory data.
As has been mentioned previously, the reason for assembling the data in Table
11 is to determine the precision of the instrument in making the measurement of
interest. This may also be described as repeatability. The reason for running
the same sample for two days is to obLain an estimate of the long-term drift of
the measurement. These tests and several others to be described later may be
used to give a sense of the accuracy of the instrument.
Accuracy is the ability of a system to approximate a true value. In general,
accuracy is estimated. Tor our purposes, the standards of quantitative analysis
-------
should be a suitable method of standardizing our equipment. Potass in m acid
p thai ate, when properly handled, is a standard of material from which a solution
of known organic carbon composition can be prepared. There is a considerable
history of application of this material in quantitative analysis. Solutions cottld
be standardized by titration with sodium hydroxide. Ill is procedure will dem-
onstrate the ability of an instrument to measure Total Organic Carbon in a solu-
tion of known composition, or serve as a method of calibration.
There is a large problem in solutions of unknown compositions containing
some carbon as suspended solids. For some compounds {e.g., anthrocite),
residence time in the reactor might be too short to give complete combustion.
For similar compounds it may be impossible to grind them to a particle size
sufficient to pass Lhrough a syringe in the Beckman even though the residence
time might be essentially infinite. One would expect distortion of the Beckman
signal where com bus I ion residence time is variable. In the case of incomplete
combustion, the meanTOC value obtained from the Beckman should be higher
than that from the proposed TOC. If the solid sample cannot be passed through
a syringe, the Beckman should produce a lower mean value Lh;m the proposed
TOC analyzer.
Observations of this kind will make it possible to discuss the accuracy of the
instrument.
Since the ultimate objective of this program is to design a device for the auto-
matic, rapid on-site moniLoring of I he organic loading of storm and combined
sower flows, some additional modification of the prototype instrument is neces-
sary for unattended operation in storm conditions.
85
-------
Several features remain to be automated in the design of a storm water TQC.
The start-up and shut-down operation will require automated signaling as a
result of certain occurrences such as storm events or sewer flooding.
The present design which incorporates Og as the reactive gas should be adapted
for use with COt> free air. This will require some modification of the present
design.
86
-------
SECTION XI
BIBLIOGRAPHY
1. Gamoson, A. L. U. and H.N. Davidson, "Storm Waiter Investigations ut
Northampton," Inst. Sow. Parif., Conf. Paper No. 5, Annual Conf.,
Llandudno, Eng. (June 19-22, 1962),
2. Field* R. And E.J, Struzcski, "Management and Control of Combine Sewer
Overflow," WPCF Journal, 44, 7 pg. 1393 (.July, 1972),
3. Fcverstein, D. L., "In-Sewer Fixed Screening of Combined Sewer Overflow, "
FWQA Program 1J024FKJ Contract 14-12-180 Pages 116-120,
¦1. Gent he, W. K. and K.M. Srinivasarughaven, "Go On-Line?; It's Vital for
Automation," Water & Wastes Engineering Vol. 42-44, (July 19 72).
5. Montgomery, II.A,C, and Peirdue K. Gardiner, "Experience with a Baeteria
Inoculum for Use in Rcspirometric Tests for Oxygen Demand," Water
Research 5 147-1G3 (1971).
0, Bridie, A. L. A. M., "Determination of Biochemical Oxygen Demand with
Continuous Recording of Oxygen," Water Research 3_ 157-1(5:1 (1909).
7. Molof, A.II. and N.S. Zaleiko, "The Detection of Organic Pollution by
Automated COD," 19th Purdue Industrial Water Conference, (May 7, 19(34).
S. Adel mail, M. 11., "Automated Measurement of COD," Water & Wastes Engi-
neering Vol. 02-55 {June 1968).
9. Bleier, II., "An Automatic System for the Continuous Determination of
Organics in Water & Wastewater," Water Research (j_ (it) 5-009 (1972).
10, Van I tall, C. E., John Hufrunko, and V.A. HlenR'or, "Rapid Combustion
Method for the Determination of Organic Substances in Aqueous Solutions, "
Analytical Chemistry 35 (3) 315-323 {1903}.
87
-------
11. Emery, R.M., E.B, Welch, and li.F. Christman, "The Total Organic
Carbon Analyzer and its Application to Water Research," J. Water Pollu-
tion Control Federation 43 (9) 1834-1844 (1971),
12. Jones, R . H., ''Total Organic Carbon Analysis and its Relationship to
Biological and Chemical Oxygen Demand, " Total Organic Carbon Analysis,
Special Publication. Edited by R. Chapman 118-128,
13. Van Hall, C.E. and V.A. Stenger, "Use of Infrared Analyser for Total
Carbon Determination," Water & Sewage Works Vol. 267-270, (June 1964).
14. Stenger, V.A. and C.E., "Analysis is Municipal and Chemical Wastewaters
by an Instrumental Method for COD Determination," J. Water Pollution
Control Federation Vol, .1755-1767 (1968),
15. Dobbs, R.A.} R.II. Wise and R.B. Dean, "Measurement of Organic Carbon
in Water Using the Hydrogen Flame Ionization Detector," Anal, Chem 30
(9) 1255-1.258 (1.967).
16. Stoss, F.H. and F.T. Eggertsen, "Flame Detection Method for Determining
Organic Carbin in Water," Anal, Chem. 44 (4) 709-14 (1972),
17. Lung, C.A. and R.B, Roberts, "Analytical. Instrumentation tor Determining
Low Levels of Total Organic Carbon in Water,11 Acrvoet General Report No.
F-1275 to Fcdera! Water Quality Administration (Juno 1970).
l*i Murka, M., "Investigation of Organic Pollution of Surface Waters by Ultra-
violet Spectrophotometry," J. Water Pollution Control Federation, Vol. 41,
No. 11, Par * 1 pagos 1923-1931 (Nov 19G9).
19. La Roy, B.C., A.C. Lilly and C.O. Tiller, "A Solid-State Electrode for
Reducible Gases," J. Electrochom SOC. 120 (1606-1073) (1973).
zo Shelley, P.E. and C.A. Kirkpatriek, "An Assessment of Automatic Sewer
Flow Samplers," EPA-R2-73-261, June 1973.
88
-------
TECHNICAL REPORT DATA
/i'lcijsc rend [itslniclious on ihr reverse before i-i»nplcliiif:j
1. HLI'OHT NO. 2.
OPA-670/2-74-087
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
ASSESSMENT AND DHVT.LOPMENT PLAN FOR MONITORING OP
ORGANICS IN STORM FLOWS
[i. REPORT DATE
November 1974, Final
6. PERFORMING ORGANIZATION CODE
?, AUTHQRISI , ,, . ... ..
AX1en Molvar, Ph.D.
Angelo Tulumello, Ph.D.
8. PERFORMING ORGANIZATION REPORT NO.
R 1382 (Internal)
9. PERFORMING ORG \NIZATION NAME AND ADDRESS
Raytheon Company
Portsmouth, RI 02871
10. PROGRAM ELEMENT NO.
1BB034 R0AP 21-ASY; Task-0 38
11. CONTRACT/GRANT NO.
12, SPONSORING AGENCY NAME AND ADDRESS
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 4526S
13. TYPE OF REPORT AND PERIOD COVERED
Interim
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Sewer line scouring, urban runoff, and combined sewage associated with storm events
represent a substantial organic pollution load. Since storms usually exhibit high
flow rates over a short, period of time, the treatment facilities become overloaded
and deliver an organic pollution load to receiving water bodies. Many times a signif-
icant amount of the combined sewage bypasses the treatment plant and is discharged un-
treated,
A method for assessing the organic content of storm related wastewaters would permit
programming discharges, and monitoring and controlling treatment processes, A variety
q£ laboratory techniques have been employed to measure this organic loading, but only
an oil-line technique such as continuous TOC will provide the necessary information on
a real or quick-time basis.
Experience with currently available commercial TOC units has not resulted in a sense of
confidence in the hardware. An evaluation of the instrumentation necessary for a reli-
able TOC in the stormwater environment leads to the selection of a measurement system
based on total combustion of sewage and detection of CO2 by infrared methods. Tests
are presently under way to establish samp1e processing, modifications of the engineer-
ing model, and accumulation of the continuous monitoring data on total organic carbon
mntfTit of storm nnd combined sewage.
17 KEY WORDS AND DOCUMENT ANALYSIS
a. DfcSCRIf'TQBS
b.IDENTIFIERS/OPEN ENDED TERMS
l\ COSati I'icki/tJrmip
infrared on-line C02
TOC alarms sample taking
continuous
monitoring combustion sample transpo
IR instruments homogenisation
Automatic recording
stormwater TOC
organic pollution
combined sewage
"t storm related waste-
waters
stormwater environment
13 B
13, DISTRIBUTION STATEMENT
NTIS ONLY
19. SECURITY CLASS (This ReportJ
UNCLASSIFIED
21. NO. OF PAGES
20. SECURITY CLASS (This page)
UNCLASSIFIED
22, PRICE
EPA Form 2220-1 (9-73)
-------
INSTRUCTIONS
1. REPORT NUMBER
Insert the iil'A report number as it appears on the cover of tits publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
4. TITLE AND SUBTITLE
Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subline, if used, in smaller
type or otherwise subordinate it to main littc. When a report is prepared in more than one volume, repeat the primary title, add volume
number and include subtitle for tlie specific title.
5. REPORT DATE
ICacIl report shall carry a date indicating at least month and year. Indicate the bask ok which it was selected (e.g., Jaw af issue, date of
approval, dote of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOR(S)
Give narae(s) in conventional order (John R. Doc, J, Robert Doc, etc.}. List author's affiliation if it differs from the performing organi-
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
{jive name, street, city, s laic, and ZIP code. List no more than two levels of an organizational hirearehy.
10. PROGRAM ELEMENT NUMBER
Use the program element number under winch the report was prepared. Subordinate numbers may be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant number under which report was prepared.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGENCY CODE
Leave blank.
15. SUPPLEMENTARY NOTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation with, Translation of. Presented at conference of,
To be published in, Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (2(H) words or less,! tactual summary of the most significant information contained in the report, if the report contains u
significant bibliography or literature survey, mention il here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS • Select from the Thesaurus of Engineering and Scientific lerins the proper authorised terms lhal identify the major
concept of the research and arc sufficiently specific and precise to be used as index entries for cataloging.
(b) IDLNTJHKKS AND Ol'l-'N-liNPLD Tl-RMS • Use identifiers for project names, code names, equipment designators, etc. Use open-
ended lenns written in descriptor form for those subjects for which no descriptor exists.
(c) COS ATI ITliLI) (.ROUP - Held and group assignments are to be taken from the 196.*^ C OSA 11 Subject Cj[egory List, Since the ma*
jority of documents are miilltdisciplinary in nature, the Primary Meld/Group assignment! s) will be specific discipline, area of human
endeavor, or type of physical object. The application^) will be cross-rcferenced with secondary Held/Group assignments iliai will follow
the primary posting(s).
18. DISTRIBUTION STATEMENT
Denote reusability to the public or limitation for reasons other than security for example ""Release Unlimited." file any availability to
the public, with address and price.
19. & 20. SECURITY CLASSIFICATION
IM) NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert llie toiai number of pages, including litis one and unnumbered pages, but exclude distribution list, if any.
22. PRICE
Insert the price set by the National Technical information Service or the Government Printing Office, if known.
EPA Form 22ZO-1 (3-73) (Reverse)
------- |