EPA-600/4-75-012
November 1975
Environmental Monitoring Series
RECOMMENDED DESIGN OF
SAMPLE INTAKE SYSTEMS FOR
AUTOMATIC INSTRUMENTATION
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development,
U.S. Environmental Protection Agency, have been grouped into
five series. These five broad categories were established to
facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL MONITORING
series. This series describes research conducted to develop new
or improved methods and instrumentation for the identification
and quantification of environmental pollutants at the lowest
conceivably significant concentrations. It also includes studies
to determine the ambient concentrations of pollutants in the
environment and/or the variance of pollutants as a function of
time or meteorological factors.
This document is available to the public through the National
Technical Information Service, Springfield, Virginia 22161.
-------
EPA-600/4-75-012
November 1975
RECOMMENDED DESIGN OF SAMPLE INTAKE.SYSTEMS FOR
AUTOMATIC INSTRUMENTATION
by
Richard P. Lauch
Instrumentation Development Branch
Program Element No. 1HA327
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Environmental
Monitoring and Support Laboratory - Cincinnati, U.S.
Environmental Protection Agency, and approved for
publication. Mention of trade names or commercial
products does not constitute endorsement or recom-
mendation for use.
11
-------
FOREWORD
Environmental measurements are required to determine the quality
of ambient waters and the character of waste effluents. The Envi-
ronmental Monitoring and Support Laboratory - Cincinnati conducts
research to:
• Develop and evaluate techniques to measure the presence
and concentration of physical, chemical, and radiological
pollutants in water, wastewater, bottom sediments, and
solid waste.
• Investigate methods for the concentration, recovery, and
identification of viruses, bacteria and other microbio-
logical organisms in water. Conducts studies to determine
the responses of aquatic organisms to water quality.
• Conduct an Agency-wide quality assurance program to assure
standardization and quality control of systems for moni-
toring water and wastewater.
The Instrumentation Development Branch, EMSL, has provided func-
tional designs relating to water quality instrumentation systems.
This report, which discusses a variety of water sample intake
designs, provides considerations for field personnel in acquiring
samples for quiescent monitoring.
* Dwight G. Ballinger
Acting Director
Environmental Monitoring and Support Laboratory
Cincinnati
111
-------
ABSTRACT
Pumping systems for automatic water quality monitors are discussed,
and recommendations on sample change, residence time, site selec-
tion, pipe size, pump selection, system cleaning, and overall
design are given. Experimental data showing sample degradation
because of biological metabolism, cavitation, and aeration are
presented. A recommended system to overcome past problems is
presented and alternative approaches for system installation are
also shown.
IV
-------
CONTENTS
Page
Abstract iv
List of Figures vi
List of Tables vi
Sections
I Conclusions 1
II Recommendations 2
III Introduction 3
IV Problems that have Occurred Within Pumping Systems
Biochemical Sample Degradation 4
Mechanical Sample Degradation 6
Maintenance 11
V Design Approaches
Existing Structures 13
No Existing Structions 13
VI Sizing Pumps and Pipelines 18
VII Automatic Cleaning 20
VIII Discussion 21
IX References 22
-------
FIGURES
No.
1 Well-Type System 10
2 Example of Samping (Complete Mixing Assumed) 10
3 Site Selection 14
4 Type-1 System (Low Residence Time Position
Pressure) 15
TABLES
No. Page
1 Sludge Accumulation and Demand 5
2 Dissolved Oxygen Change Resulting From
Deposits and Slime Growth 7
3 Dissolved Oxygen Depletion Resulting From
Cavitation 8
4 Reaeration: Complete Tabulation of Dissolved
Oxygen Data Before Cleaning 9
5 Average Dissolved Oxygen Error with Screw-Type
and Centrifugal Pumps 17
6 Velocity Increase with Constant Reynolds Number 18
VI
-------
SECTION I
CONCLUSIONS
Considerable engineering time and money are spent on the design of
water quality monitoring instrumentation, including functional
components such as computers and telemetering. All of this is in
vain if the water sample delivered to the instrument is not repre-
sentative. Therefore, adequate engineering effort, time, and funds
are required to design an accurate intake system.
Normally, the basic objective for ambient monitoring is to observe
a representative stream cross section. To fulfill this objective,
pumping systems should be designed to deliver a sample that is
representative of most of the water in the stream. This is accom-
plished when the sample is drawn from the river channel.
Because sample degradation resulting from the biological metabolism
of sludge and slime microbes within the pipeline should be insigni-
ficant, high sample velocity through a pipeline of consistent cross
section is required. The system should be designed so that raw
water flows directly from the river through a pump and to the
instrumentation shelter.
Mechanical sample change because of cavitation, reaeration, and
damping can be avoided by using a positive pressure system with
low residence time.
Automatic cleaning may not be required on a system that is designed
for minimal biological sample degradation. Obviously there is a
sample velocity for a specific line length and exposed internal
surface above which biological degradation would be insignificant.
This velocity may be so high when sampling polluted water that
cleaning with a bactericide during warm weather would always be
required. Cursory results show that for lines of specific length
and exposed surface, an optimum velocity is attainable that would
eliminate the need for automatic cleaning. More research is needed
in this area, but initial design should be aimed toward this opti-
mum system without automatic line cleaning. Periodic manual
cleaning of the intake strainer is required, and system design
should include easy access to this component.
System maintenance must be considered in the design. Access to
all components from the river bank during the most adverse stream
conditions is necessary.
-------
SECTION II
RECOMMENDATIONS
Automatic instrumentation can be installed at existing facilities
such as water, power, and industrial plants along the stream if
these locations provide meaningful data. A thorough study should
be made, however, to prove that the sample, as received by the
instrument, is representative and nondegraded before locating at
these facilities. Pumping systems may also be located at bridges
and docks along the stream, but again, a thorough investigation is
required to insure that meaningful and nondegraded data will be
obtained.
When no existing facilities are available, the type-1 system
described in this report is recommended. With this system, water
travels directly from the river through a pump and to the monitor.
The intake should be located up from the river bottom sludge and
should draw water from the stream channel. Water velocity within
all parts of this system should be greater than 4.7 feet per second
for minimal sample degradation. Positive pressure throughout all
parts of this system is recommended. This system is designed so
that every component can be maintained from the river bank during
all stream conditions. The intake system should be designed so
that sample degradation is minimal or insignificant, thereby elimi-
nating the need for an automatic intake cleaning system. High
sample velocity along with nonporous pipeline materials (such as
Celanese ultrahigh molecular density type) will minimize sample
degradation.
Type-2 and type-3 systems are described in this report as alterna-
tives if the type-1 system is not feasible at a specific location.
The type-2 system operates under negative pressure and therefore
requires the use of a screw-type pump to prevent dissolved oxygen
(DO) loss because of cavitation. Aeration will take place at
loose connections. Therefore, this system will require frequent
data verification to prove that DO readings are accurate. The
type-3 system is a well, and cleaning with a bactericide is recom-
mended twice a week during warm weather to keep sample degradation
minimal. Removal of sediment is also required. A larger-than-
normal pump is recommended for the well to minimize the effect of
damping.
-------
SECTION III
INTRODUCTION
Small, reliable pumping systems are needed to bring sample water to
automatic instrumentation. In most cases, it is not practical to
place sensors in the stream because thay are inherently delicate
and should not be exposed to rough water conditions where sand,
debris, turbidity, high velocity, biological growth, and other
factors can introduce spurious signals.
In the past, everything imaginable has happened to intake systems
for instrumentation. At existing facilities, monitors have sampled
from dead water mains, lagoons, recirculating pools for power plants,
and other undesirable places. Samples have been drawn from long
pipelines of low velocity, from clogged wells, cavitating pumps,
reaerating systems, lines where chemicals were being added, and
other meaningless locations. Pumping system components have failed,
and intakes have been washed out and covered with sandbars or debris.
There have been times when service was impossible because of high
water.
This study determines, in theory, the biochemical and mechanical
reasons for sample change. A method for obtaining representative
samples at remote locations is given. During the investigation,
tests were made on different types of pumping systems, and data
were obtained to confirm the theory for sample change. Recommenda-
tions are given to overcome past difficulties, and a design is
presented that gives a representative sample, minimizes degradation,
and is serviceable under all river conditions.
-------
SECTION IV
PROBLEMS THAT HAVE OCCURRED WITHIN PUMPING SYSTEMS
BIOCHEMICAL SAMPLE DEGRADATION
Loss in DO has frequently been detected across intake systems
pumping polluted water during warm weather periods. When oxygen
loss is due to microbial metabolism, other less conspicuous pa-
rameters will also change; but since oxygen change is the most
noticeable and easily detectable it is the most frequently dis-
cussed. Oxygen depletion is due to one or more of the following:
biochemical oxygen demand (BOD) of the water sample during travel,
BOD of the sludge deposits within the pipeline, and BOD of the
slime growth on the inner pipe walls.
BOD of Water Sample - The BOD of a slug of water as it travels
from the river to the sample probe is conventionally formu-
lated as:
BOD = L (l-10~kt) (1)
cL
where BOD = 02 consumed during travel
L = ultimate BOD of the water sample
t = travel time, days
k = reaction velocity constant
Oxygen deficit resulting from the BOD of the sample should be
insignificant, because k is low for river water, and t should
only be a few minutes or less for a properly designed system.
BOD of Sludge Deposits - These sludge deposits may form at an
enlarged section of pipe where the velocity of the water is
low; also, if the intake pipe is resting on the river bottom,
the sample may pass over a pile of sludge just before enter-
ing the intake. Velz* discusses the effect of sludge deposits
in streams and presents theoretical information that shows
that DO can be depleted significantly because of the BOD of
accumulated sludge. His reasoning is here applied to pumping
systems. The BOD of sludge deposits in streams can be defined
by equation 2.
-------
(2,
where
L, = cumulative BOD of the deposit, pounds
P, = BOD added to the deposit, pounds per day
k1 = specific rate of oxidation of the deposit
(usually 0.03 per day)
t = time of accumulation, days
Velz tabulated Table 1 from equation 2.
Table 1. SLUDGE ACCUMULATION AND DEMAND*
Time
in
days
2
3
4
5
10
20
30
40
50
60
70
80
90
Ultimate
Accumulation
as a percentage
of the BOD of
the daily deposit
187
271
350
423
724
1,086
1,267
1,359
1,404
1,427
1,438
1,444
1,447
1,450
Daily demand from
the accumulation as a
percentage of the BOD
of the daily deposit
12.9
18.7
24.1
29.2
49.9
74.9
87.4
93.7
96.8
98.4
99.2
99.6
99.8
100.0
*Source: C. J. Velz1.
Table 1 shows that after sludge has accumulated for a long
enough period of time, the rate of biological oxygen demand
from the sludge pile is equal to the BOD of the daily deposit
(equilibrum). Therefore, oxygen within the sample could
eventually be depleted by an amount equal to the BOD of the
suspended solids within the water sample. The table shows
-------
that after only 10 days, DO is being depleted at a rate of
49.9 percent of the BOD of the suspended material that is
settling. This reasoning is theoretical and assumes constant
and aerobically decomposable suspended solids conditions. It
is idealistic to think that sludge will form in a pumping
system exactly according to the equation. The equation does,
however, clearly point out the fact that sludge deposits in
sampling systems could significantly lower the DO reading.
BOD of Slime Growth on Pipe Walls - Organic materials (sub-
strate) in the water sample are consumed by the thin coating
of slime bacteria on the inner surface of the pipeline. If
this substrate upon which they feed is sufficient, then the
number of these bacteria is proportional to the surface area
upon which they live; therefore, DO depletion within the
sample would increase with sample residence time as long as
the required substrate were present. Hence a long pipeline
exposed to an enriched sample with sufficient residence time
could deplete dissolved oxygen significantly.
Data showing oxygen losses occurring because of sludge deposits
and slime growth mentioned earlier are included in Table 2. Sta-
tion 1 was located at the intake in the river and Station 3 was
located just ahead of the monitor. The two stations were connected
by 450 feet of 1-1/4-inch diameter plastic pipe. The table shows
significant DO loss before cleaning, but DO loss after cleaning
was insignificant. Cleaning consisted of removing a sludge deposit
from within the inlet strainer and flushing the system with a
solution of chlorine to kill slime bacteria.
MECHANICAL SAMPLE DEGRADATION
Mechanical sample degradation includes DO change because of cavi-
tation, reaeration, and damping.
Cavitation
Cavitation can occur when the pressure within the pumping system
drops below the vapor pressure of the dissolved or condensed gases
within the sample. Table 3 shows DO loss across a centrifugal pump
operating at suction lifts of 20 feet and 2 feet of water, respec-
tively. Station 1 was the river intake, station 2 was the high
pressure side of the pump, and station 3 was located just ahead of
the monitor. There were 350 feet of 3/4-inch plastic pipe between
stations 2 and 3. The table shows insignificant DO losses at
-------
Table 2. DISSOLVED OXYGEN CHANGE RESULTING FROM
SLUDGE DEPOSITS AND SLIME GROWTH
Suction
Date and
time
lift
Cf
to
at
1
DO
Cppm)
station
no.
3
Before
8/25/69
11:45
12:30
12:55
1:30
3:07
3:25
8/26/69
8:17
8:53
9:38
10:06
10:36
11:02
12:07
12:33
12:59
1:25
3:03
3:40
4:17
8/27/69
8:13
8:55
9:48
11:00
9/24/69
11:54
12:00*
3:00*
3:05
3:15
3:30
3:55
9/25/69
8:15
8:30
8:45
9:00
13
13
13
13
0
0
0
0
0
0
0
0
0
0
0
0
13
13
13
13
13
13
13
13
-
-
13
13
13
13
12
12
12
12
.4
.4
.4
.4
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.4
.4
.4
..4
.4
.4
.4
.8
After
--
--
.6
.6
.6
.6
.1
.1
.1
.1
10.
11.
11.
11.
11.
11.
8.
8.
9.
10.
11.
9
0
6
3
3
3
3
9
8
7
5
11.9
12.
12.
12.
12.
13.
14.
14.
8.
9.
10.
12.
5.
7
9
7
6
3
25
85
45
0
4
5
2
10
10
10
8
9
9
7
8
9
9
10
10
10
10
10
10
9
8
8
7
7
8
11
4
.1
.1
.2
.3
.6
.7
.5
.1
.4
.6
.0
.3
.7
.7
.3
.0
.1
.85
.9
.0
.65
.8
.0
.6
Error
(ppm)
Temperature
station (°F)
(1) minus (3)
cleaning
0.
0.
1.
3.
1.
1.
0.
0.
0.
1.
1.
1.
2.
2.
2.
2.
4.
5.
5.
1.
1.
1.
1.
0.
cleaning and backf lushing
--
--
7.
7.
7.
7.
6.
6.
6.
7.
-
-
0
3
35
4
65
8
9
0
-
-
7
7
7
7
6
6
6
6
--
--
.0
.25
.30
.4
.6
.7
.8
.85
__
--
0.
0.
0.
0.
0.
0.
0.
0.
8
9
4
0
7
6
8
8
4
1
5
6
0
2
4
6
2
4
95
45
35
6
5
6
Flow
station 3 (gpm)
76
76
77
78
79
80
74
75
75
75
76
76
77
78
78
79
80
81
-
74
-
75
76
67
.0
.5
.5
.0
.5
.0
.5
.0
.5
.5
.0
.5
.5
.0
.5
.0
.5
.0
--
.5
--
.0
.5
.5
12.3
10.1
8.5
8.8
with chlorine
_
-
0
05
05
0
05
10
10
15
-
-
68
68
-
-
65
-
-
65
--
--
.5
.0
--
--
.5
--
--
.5
11.9
*System flushed with chlorine solution.
-------
suction lifts of 2 feet. DO losses were significant at suction
lifts of 20 feet; hence, cavitation was experienced at the eye of
the impeller, and oxygen came out of solution, passed through the
impeller, and was severely agitated at the periphery. Table 3
shows that DO loss was most critical at station 2, with some oxygen
being redissolved during travel between station 2 and 3. Oxygen
bubbling out of solution (similar to C02 loss from a carbonated
beverage) was visible when drawing the DO sample at stations 2 and
3.
Table 3. DISSOLVED OXYGEN DEPLETION RESULTING FROM CAVITATION
Date and
time
1968
10/22
4:45
10/23
7:35
8:20
10/25
4:25
10/28
8:55
11:20
2:20
Suction
lift
(ft)
20
20
20
20
20
2
2
Station
1*
9.40
8.60
8.70
9.70
10.00
10.60
10.80
DO (ppm)
Station
2+
8.25
7.95
7.50
8.50
8.80
10.50
10.75
Station
3*
8.71
8.39
8.30
9.32
9.64
10.50
10.75
Temp.
(F)
61.5
54.0
55.5
52.5
49.0
49.5
49.5
Saturation
level of
DO (ppm)
10.0
10.8
10.6
11.1
11.3
11.3
11.3
*River intake.
"*"High pressure side of the punro.
^Located just ahead of the monitor.
Similar testing with a positive displacement screw-type pump
showed insignificant DO losses in most samples when the system
was operated carefully and conscientiously. Data on this type of
system are presented later in the report.
Reaeration
An example of reaeration is given in Table 4. These data were
collected during the evaluation of a well-type pumping system.
Figure 1 depicts the well-type system and shows the station
-------
Table 4. REAERATION: COMPLETE TABULATION OF DISSOLVED OXYGEN DATA BEFORE CLEANING
1
Time
Date
8/4/70*
8/5/70*
8/6/70*
Average
8/10/70f
Start
8:30
9:00
10:00
10:30
11:00
11:30
8:3C
9:0f
9:30
10:00
10:30
11:00
11:30
1:00
11:30
1:00
1:30
2:00
2:30
3:00
3:30
4:00
—
—
Stop
8:36
9:06
10:06
10:36
11:06
11:36
8:36
9:06
9:36
10:06
10:36
11:06
11:36
1:06
11:36
1:06
1:36
2:06
2:36
3:06
3:36
4:06
...
—
Station 1
7.60
9.25
10.05
10.85
12.00
7.35
7.50
7.70
8.05
8.50
9.10
9.40
9.80
10.30
11.75
12.05
12.80
13.25
13.95
14.45
14.95
10.51
6.10
DO (mg/1)
Station 2
2.75
2.60
3.20
3.65
3.70
3.85
3.60
2.90
2.65
2.85
3.15
3.05
3.25
3.80
3.50
4.85
4.75
4.85
5.05
4.90
5.25
5.30
...
0.50
Station 3
7.45
7.40
7.35
7.25
7.60
8.15
7.90
9.30
9.75
10.65
7.10
9.50
10.35
9.10
5.90
6.60
7.70
6.80
7.00
7.85
8.20
8.75
10.90
Station 4
7.45
7.35
7.70
7.30
8.05
8.55
7.75
8.75
9.10
9.10
7.10
8.85
8.95
9.80
5.15
6.20
6. 25
6.55
6.70
7.80
8.30
8.30
.._
Inlet
Pipe
(2 minus 1)
-5.00
-6.05
-6.40
-7.15
-8.15
-3.75
-4.60
-5.05
-5.20
-5.35
-6.05
-6.15
-6.00
-6.80
-6.90
-7.30
-7.95
-8.20
-9.05
-9 . 20
-9.65
-6.66
-5.60
DO difference
Well
(3 minus 2)
4.70
4.80
4.15
3.60
3.90
4.30
4.30
6.40
7.10
7.80
3.95
6.45
7.10
5.30
2.40
1.75
2.95
1.9S
1.9S
2.95
2.95
3. 45
4.28
10.40
(mg/1) across
Pressure
Pipe
(4 minus 3)
0.00
-0.05
0.35
0.05
0.45
0.40
-0.15
-0.55
-0.65
-1.55
0.00
-0.65
-1.40
0.70
-0.. 75
-0.40
-1.45
-0.25
-0.30
-0.05
0.10
-0.45
-0.30
Overall
(4 minus 1)
-0.25
-1.55
-2.75
-2.80
-3.45
0.40
1.25
1.40
1.05
-1.40
-0.25
-0.45
0.00
-5.15
-5.55
-5.80
-6.25
-6.55
-6.15
-6.15
-6.65
-2.71
*Pumn spraying water and sucking air.
+n sucking air.
-------
STATION 1
STATION 2 STATION 3
STATION 4
WELL
Figure 1. Well-type system.
MONITOR
ENLARGED PIPE SECTION
iot "xr — ^
9
3 8
a.
a..
7
0°
6
5
4
\ C°T c
_ \
\
\
• ^^^
— ^^P^^^
i i i i
0 5 10 15 20
r — i
i ^
c
c
Q
V
•^~
^^++
j L
25 30
— CR
j = 10 ppm
R = 4 ppm
= 17.6 gpm
= 28.3 ft3
-C0 = C»+(Cx-
i l i — i
35 40 45
50
TIME (MINUTES)
Figure 2. Example of damping
(complete mixing assumed)
10
-------
locations. The DO readings at station 3 were higher than those at
station 2, indicating reaeration. A visible inspection revealed
that the inlet pipe was clogged, the flow was very low, and the
well was drawing down below the pump inlet; hence the pump was
sucking water and air. The tremendous oxygen loss between station
1 and 2 is an excellent example of all three types of biochemical
sample degradation taking place.
Damping
Travel time of the sample from the river to the monitor should be
minimal. The velocity of the sample water should be as rapid as
possible without causing an excessive pressure drop in the pipe
lines. Low velocity of flow and enlarged pipeline sections should
be avoided. An example of damping that could take place in a sys-
tem with an enlarged section is given below:
Assume an enlarged cylindrical section with the dimensions given in
Figure 2. The initial DO concentration in the enlarged section is
10 ppm (Ci = 10 ppm) . A slug of highly polluted water (CR = 4 ppm)
comes down the river. How long will it take for the outflow from
the enlarged section (Co) to indicate the river condition? If
complete mixing can be assumed, then the following fundamental
charge/discharge equation can be used:
A plot of this equation for the condition stated is given in
Figure 2. As shown, it will theoretically take 20 minutes for
the monitor to read 5 ppm. It takes 40 minutes for the monitor
to read 4.2 ppm, or within 5 percent of the true condition. By
this time, the serious pollution may have passed. Modern instru-
mentation is built for fast response, and components of the
pumping system should not significantly impair the overall
response time.
MAINTENANCE
Maintenance problems have included pump failure, clogged intake,
equipment covered with sandbar, washed-out equipment, damage from
floating debris, high water, etc. Most of these problems are
related to poor design and/or inadequate funds. Most of the
11
-------
engineering time and money for an automatic monitoring facility is
put into the instrumentation and system components such as tele-
metering, computers, etc., with little attention paid to the system
that delivers the water sample. A continuous, representative, and
nondegraded sample is required; and therefore an appropriate pro-
portion of engineering hours and funds must be allocated to the
intake system.
12
-------
SECTION V
DESIGN APPROACHES
EXISTING STRUCTURES
Many water quality monitors have been installed at water, power,
and industrial plants along a stream. The raw water sample to the
monitor is tapped off the plant's supply, a practice that elimi-
nates the need to install and maintain a separate pumping system.
This approach is good if it is certain that the sample is repre-
sentative of the desired river location and if the sample is not
changed within the pipeline before reaching the monitor. The
object is to obtain meaningful water quality data. An easy instal-
lation is fine if it is not made at the expense of obtaining good
data.
The pumping system may also be located from a bridge pier or dock
if it is situated where meaningful data can be obtained. In such
cases, some engineering design is still required; but the time and
expense should be less because a substantial structure is already
provided for mounting the system.
NO EXISTING STRUCTURES
When existing structures are not available, conscientious engi-
neering time and adequate funds are required to design and install
a dependable sampling system.
Good engineering judgment is required for site selection. One
must be certain that the intake is not located where the river
could deposit a sandbar. A homogeneous sample that is representa-
tive of most of the water in the river at the specific location is
desired. This type of sample can most likely be collected from
the river channel; therefore, an economical method of sampling
the channel is required. The channel is not always in the center
of the river, but in many cases is close to one bank. Figure 3
shows one instance where nature brings the channel close to the
bank, and this is the logical location of an intake.
In Figure 3, the inset of section A-A illustrates a river cross
section with the channel nearly centered. At section B-B, the
channel is much closer to the bank; hence it is easier and more
economical to locate the intake at section B-B. The river will
13
-------
SLANT TO
DEFLECT DEBRIS
LOCATE INTAKE AT B-B
100
SECTION B-B
SECTION A-A
Figure 3. Site selection: note that the channel is naturally
close to one bank and is thus the logical location
of an intake.
tend to score at section B-B. This tendency is advantageous in
that it will keep settlement and debris from covering the intake,
but it may be necessary to place heavy riprap on this bank to pre-
vent erosion.
The above conclusions are logical, but it is still necessary to
obtain confirming proof of sample representativeness before intake
installation by making both a longitudinal and cross sectional
sample study.
Intake system maintenance along with sample representativeness
must be considered in the design. The system should be designed
so that pumps and intake strainers can be serviced and cleaned
during the most adverse stream conditions. Three types of systems
14
-------
that are rugged and allow maintenance during most river conditions
are presented. The type-1 system is recommended first for ease of
maintenance and minimal sample degradation, and types 2 and 3
could be used as alternatives if for some reason the type-1 system
is not feasible.
Type-1 System: Low-Residence-Time Positive Pressure
The type-1 system is shown in Figure 4. Site selection was made as
described above. The system consists of a pump that slips from
the shelter, through a casing, into the river. The intake strainer
protrudes 8 to 12 inches beyond the end of the casing. A winch can
be used to pull the pump and strainer back to the shelter for
servicing. This system has no enlarged sections, and the residence
time is very low. Note that a pile supports the end of the casing
and keeps it up out of the river bottom sludge. A small contrac-
tor's pump could be used at the river. These units are rugged,
but shut off head is only about 45 feet of water. Therefore a
booster pump may be required in the shelter or midway between the
river and shelter. A positive displacement screw type pump could
also be used at the river. These units can overcome considerable
INSTRUMENT
SHELTER'
MARKER BUOY
IF REQUIRED
CASING
Figure 4. Type-1 system (low-residence-time positive pressure),
15
-------
head and a booster pump would not be required; however in the past
this pump has been known to fail. Cause of failure has been traced
to the thrust bearing within the electric motor and the universal
joint connecting motor and pump. The unit has head vs. flow char-
acteristics that are very desirable for a monitoring installation,
and the pump head can withstand the mild abrasives found in raw
river water. The author feels that the past problems of failure
with this pump can be solved.
The type-1 system can be designed with a high velocity of flow
from river to shelter, and therefore backflushing with a chlorine
solution is probably not required. This system has many attributes,
and it is highly recommended by the author.
Type-2 System: Low-Residence-Time Negative Pressure (Alternate)
This system is similar to the type 1, except that only an inlet
strainer is pushed through the casing to the river. A positive
displacement screw-type surface pump is installed within the
shelter. Table 5 shows that it is possible to pump at pressures
less than atmospheric with insignificant DO loss if the screw-type
pump is used. It is recommended that the pressure at the pump
inlet be no less than 1/2 atmosphere. The suction line must be
tight so that no air leaks into the system. DO data from this sys-
tem are easily subject to suspicion, and for this reason it is best
to use the type-1 system, if possible. Advantages of the negative
pressure system are the ease with which a small strainer could be
slipped through the casing, reduced weight (which eliminates the
need for a winch), the fact that no electric cable is required, and
the smaller casing.
Type-3 Well System (Alternate)
A sketch of this system was shown in Figure 1. Conventional, sub-
mersible, centrifugal, clear well pumps have operated longer without
failure in the well than other type of system. These pumps are
conventionally mounted in a vertical position, as their design
intended, and sand, which is frequently encountered in river flow,
settles within the well before passing through the pump; therefore,
pump life is prolonged. Tests have shown that good data with
little sample degradation can be obtained from the well system if
a frequent cleaning schedule is maintained. Cleaning requires
flushing the lines with a chlorine solution twice a week during
warm weather and removing sediment from the well about once a week.
Well-type systems are relatively inexpensive.
16
-------
Table 5. AVERAGE DISSOLVED OXYGEN ERROR WITH SCREW-TYPE AND CENTRIFUGAL PUMPS
Station 1 minus station 3
Pump
number
it
1
1
2+
Suction
lift (ft)
2
8
14
20
20
For all readings
Pumps 1 § 2
st
3
20
2
Number of
readings
6
9
12
6
17
50
5
2
Avg . error
(ppm)
.05
.05
.02
.00
.03
.03
.41
.08
Std. dev.
.05
.10
.10
.08
.05
.08
.18
.04
Number of
readings
6
8
10
5
16
45
5
2
Station 1 minus station 2
Avg . error
(ppm)
.08
.05
.05
.07
.01
.04
1.08
.08
Std. dev.
.05
.16
.08
.09
.05
.09
.24
.04
Station 1 was the river intake.
Station 2 was the high pressure side of the pump;
station 1 and 2.
Station 3 was within the instrumentation shelter,
station 2 and 3.
"^Positive displacement screw-type, 1/2 hp.
centrifugal, 1/3 hp.
there was 150 feet of 3/4-inch plastic pipe between
there was 300 feet of 3/4-inch plastic pipe between
-------
SECTION VI
SIZING PUMPS AND PIPELINES
Pipelines should be sized for a high velocity of flow to prevent
settling and to keep slime growth scored out. There are practical
limits for velocity, since increasing it causes greater friction
loss and thus puts greater pressure strain on the system and demands
more pump horsepower. Stierli3 et al show minimal DO loss if the
Reynolds number is kept above 25,000. Recent DO data taken across
pumping systems indicate that velocity of flow should also be con-
sidered. Table 6 rationalizes this point.
Table 6. VELOCITY INCREASE WITH CONSTANT REYNOLDS NUMBER
D
Pipe
diameter
Cin.)
1.0
1.5
2.0
2
Flow
(gpm)
9.0
13.5
18.0
A
Surface
area
(in.2/in.)
3.14
4.71
6.28
R
Reynolds
number
29,970
29,970
29,970
V
Velocity
(ft. /sec.)
3.70
2.44
1.83
Residence
time per
100 ft. (sec.)
27.0
41.0
54.6
A/Q
Number
bacteria per
rate of flow
.349
.349
.349
The table illustrates three pipelines with internal diameters of
1, 1.5, and 2 inches. The Reynolds number* is held constant as the
diameter increases by increasing flow proportionately. The table
shows velocity decreasing from 3.7 to 1.83 feet per second and a
proportionate increase in residence time. Assume that a certain
number of slime bacteria can occupy a unit area of pipe surface;
then the A/Q column shows that the number of bacteria on the pipe
surface per rate of flow is constant. Hence, with increasing
R
- 4Q_
ITDv
where R = Reynolds number
Q = flow, feet3 per second
D = pipe diameter, feet
v = kinmatic viscosity of water (0.00001 feet2 per second
for water at 25C)
18
-------
residence time, the same number of bacteria are given a longer
period to oxidize the substrate within the water sample, and the
DO consumption will be greater. Therefore it is better to use the
1-inch line with a 3.7-feet-per-second velocity than the 2-inch
line with a velocity of 1.83 feet per second. If the Reynolds
number were the only criterion, then all three lines would be
equal. More tests are needed in this area, but cursory results
indicate that velocities of 4.7 feet per second or higher are
required to give accurate DO readings without cleaning for pipe-
lines 400 feet long. Lower velocities and/or longer lines may
require automatic cleaning with a low concentration of a strong
oxidant such as chlorine. The following example for sizing a
pump and pipeline is given:
Required raw water flow to monitor = 18 gpm (0-04 feet^
per second)
Pressure drop across monitor = 15 psi
Elevation of monitor above river = 30 feet (13 psi)
Total length of pipeline = 400 feet
For velocity > 4.7 feet per
second
D = 4Q
TV
D = 2 '°4
4.71F
D = .104 feet
D = 1-1/4 inch
Pressure drop in 400 feet of
1-1/4-inch smooth pipe
at 18 gpm = 14 psi*
Total pressure drop = 43 psi.
Therefore a pump that will deliver 18 gpm at a total dynamic head
of 42 psi should be purchased.
*Bureau of Standards Report BMS 79.
19
-------
SECTION VII
AUTOMATIC CLEANING
Automatic cleaning is probably not required with a properly
designed pumping system such as the type-1. There may be some
cases, such as the well system or a system having very long pipe-
lines, that require cleaning. Eckoldt^ discusses backflushing
with compressed air, and an automatic cleaning system that peri-
odically flushes the lines with a low concentration of chlorine
has been discussed by this author.^ Systems requiring an inlet
strainer, such as the type-1, should be designed so that manual
cleaning from the river bank is possible.
20
-------
SECTION VIII
DISCUSSION
Pumping systems for automatic water quality instrumentation have
presented many problems in the past. This report describes the
cause of these problems and gives recommendations for avoiding
them. These recommendations are based on the experience of the
author and others with intakes and includes tests on the well^*^
negative lift systems', various pumps, and automatic chlorina-
The type-1 system described in this report is recommended as a
method to overcome past difficulties . This system presents a new
technique in that it has a high sample velocity with no enlarged
sections, and it is completely serviceable from dry land during
all stream conditions. If built as described, the type-1 system
is rugged, could not get washed out or buried in sand, and would
provide a continuous representative and nondegraded sample.
Installation of the system is made from the riverbank, and costs
are relatively low.
21
-------
SECTION IX
REFERENCES
1. Velz, C. J., "Oxygen Relationships in Streams, Significance of
Organic Sludge Deposits," Robert A. Taft Sanitary Engineering
Center, Technical Report W 58-2, U.S. Department of Health,
Education, and Welfare, Public Health Service, 1958.
2. Mentink, A. F., and Griffith, J., Unpublished quarterly reports,
Analytical Quality Control Laboratory, U.S. Environmental Pro-
tection Agency, National Environmental Research Center,
Cincinnati, Ohio, 1972.
3. Stierli, H., Weeks, J. D., and Buck, R. A., "An Experiment to
Determine the Relationship Between Dissolved Oxygen Change in
Pipeline Flow and Reynolds Number," National Water Quality
Network Applications and Development Report #8, Division of
Water Supply and Pollution Control, U.S. Public Health Ser-
vice, Cincinnati, Ohio, 1963.
4. Eckoldt, M., "Measuring Stations for Recording of Water Quality
in Flowing Waters," German Hydrologic Reports, 11(1):3-11, 1967.
5. Lauch, R. P., "An Automatic Chlorination System for Eliminating
Biological Growth in Pumping Systems for Automatic Instrumen-
tation," U.S. Environmental Protection Agency, National
Environmental Research Center, Cincinnati, Ohio, Unpublished
report, 1972.
6. Lauch, R. P., "An Evaluation of a Well-Type Pumping System for
Automatic Water Quality Monitors," U.S. Environmental Protec-
tion Agency, National Environmental Research Center, Cincinnati,
Ohio, Unpublished report, 1971.
7. Lauch, R. P., "Pumping Systems for Automatic Water Quality
Monitors," U.S. Environmental Protection Agency, National
Environmental Research Center, Cincinnati, Ohio, Unpublished
report, 1970.
22
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/4-75-012
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
RECOMMENDED DESIGN OF SAMPLE INTAKE SYSTEMS FOR
AUTOMATIC INSTRUMENTATION
5. REPORT DATE
November 1975 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Richard P- Lauch
8. PERFORMING ORGANIZATION REPORT NO,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
10. PROGRAM ELEMENT NO.
1HA327; ROAP 24 ALE; TASK 03
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Same as above
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Pumping systems for automatic water quality monitors are discussed, and recommenda-
tions on sample change, residence time, site selection, pipe size, pump selection,
system cleaning, and overall design are given. Experimental data showing sample
degradation because of biological metabolism, cavitation, and aeration are presented.
A recommended system to overcome past problems is presented and alternative approaches
for system installation are also shown.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Intake systems, Water intakes, Pumps,
Waste water, Water pollution, Water
quality, Instruments
Intake, Water quality
instrumentation intakes,
Water monitoring intakes,
Wastewater instrumenta-
tion intakes, Water
quality instrumentation
pumping systems
13B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
29
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
23
*USGPO: 1976 — 657-695/5337 Region 5-11
------- |