EPASOO/
2-84-183
PBS5-121663
RECOMMENDED PRACTICE FOR FLOW MEASUREMENT IN WASTEWATER
TREATMENT PLANTS WITH VENTURI TUBES AND VENTURI NOZZLES
National Bureau of Standards
Washington, DC
Nov 84
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
-------
-------
c
EE85-121b65
EPA-600/2-84-185
November 1984
RECOMMENDED PRACTICE FOR FLOW MEASUREMENT
IN WASTEWATER TREATMENT PLANTS WITH
VENTURI TUBES AND VENTURI NOZZLES
by
Gershon Kulin
Fluid Engineering Division
National Bureau of Standards
Washington, D. C. 20234
EPA 78-D-X0024-1
Project Officer
Walter W. Schuk
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
REPRODUCED BY
NATIONAL TECHNICAL
INFORMATION SERVICE
U.S. DEPARTMENT OF COMMERCE
SPRINGFIELD. VA 22161
-------
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-34-185
[3. RECIPIENT'S ACCESSION NO.
121663
4. TITLE AND SUBTITLE
RECOMMENDED PRACTICE FOR FLOW MEASUREMENT IN
WASTEWATER TREATMENT PLANTS WITH VENTURI TUBES
AND VENTURI NOZZLES
5. REPORT DATE
November 1984
|6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
[8. PERFORMING ORGANIZATION REPORT NO.
Gershon Kulin
9. PERFORMING ORGANIZATION NAME AND ADDRESS
National Bureau of Standards
Fluid Engineering Division
Washington, DC 20234
10. PROGRAM ELEMENT NO.
B113, CAZB1B
11. CONTRACT/GRANT NO.
IAG No. EPA-78-D-X0024-1
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory--Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Handbook—10/1 /78-9/30/81
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer: Walter W. Schuk Telephone - (513) 684-2621
16. ABSTRACT
Venturi tubes and venturi nozzles are suitable for in-plant flow measurement
of raw influent, treated effluent, return activated sludge, certain digested
sludges, and for air and gas flows. However, they are not generally recommended
for measurement of raw primary sludge.
For classical venturi tubes which operate under optimum prescribed conditions,
the primary-element discharge coefficient is predictable to within a basic uncer-
tainty of about 1 percent. For standard venturi nozzles this basic uncertainty
ranges from about 1 percent to 2 percent, i Errors in the secondary system must be
considered in addition to estimate the uncertainty of the output measurement.i The
primary-element uncertainty increases to over 3 percent in venturi tubes for a
variety of Tess-than optimum conditions such as insufficient upstream approach
length, roughening and aging, and low Reynolds number. For cases in which the
"standard" uncertainty is acceptable to the user, the initial performance check
consists of the calibration of the secondary system as described in section 11.2.
However, non-standard tubes and/or installations may require an initial field cali-
bration of the primary-element coefficient as well as the secondary system. The
primary calibrations are described.in section 11.3 and usually involve either
volumetric or dilution methods, or comparison with a reference meter.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI I ield/Group
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Kcparll
UNCLASSIFIED
21. NO. OF PiXGES
2O. SECURITY CLASS (This
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (R«v. 4-77) PREVIOUS EDITION is OBSOLETE .
-------
DISCLAIMER
Although the information described in this document has been funded
wholly or in part by the United States Environmental Protection Agency
through assistance agreement number EPA 78-D-X0024-1 to National Bureau of
Standards, it has not been subjected to the Agency's required peer and
administrative review and therefore does not necessarily reflect the views
of the Agency and no official endorsement should be inferred.
-------
FOREWORD
The U. S. Environmental Protection Agency was created because of in-
creasing public and Government concern about the dangers of pollution to the
health and welfare of the American people. Noxious air, foul water, and
spoiled land are tragic testimonies to the deterioration of our natural
environment. The complexity of that environment and the interplay of its
components require a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem solu-
tion; it involves defining the problem, measuring its impact, and searching
for solutions. The Municipal Environmental Research Laboratory develops new
and improved technology and systems to prevent, treat, and manage wastewater
and solid and hazardous waste pollutant discharges from municipal and communi-
ty sources, to preserve and treat public drinking water supplies, and to mini-
mize the adverse economic, social, health, and aesthetic effects of pollution.
This publication is one of the products of that research and provides a most
vital communications link between the researcher and the user community.
This document, the first of a series, represents an effort to provide
improved guidelines for the selection, installation, calibration and main-
tenance of instruments used for monitoring and process control in wastewater
treatment plarfts.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
iii
-------
ABSTRACT
Venturi tubes and venturi nozzles are suitable for in-plant flow measure-
ment of raw influent, treated effluent, return activated sludge, certain di-
gested sludges, and for air and gas flows. However, they are not generally
recommended for measurement of raw primary sludge.
The classical venturi tube and the standard venturi nozzle have very
specific requirements for construction and installation, which are described
in sections 4, 5, 6 and 7 of this document. There are also specific require-
ments on the secondary system which are described in section 8.
For classical venturi tubes which operate under optimum prescribed con-
ditions, the primary-element discharge coefficient is predictable to within
a basic uncertainty of about 1 percent. For standard venturi nozzles this
basic uncertainty ranges from about 1 percent to 2 percent. Errors in the
secondary system must be considered in addition in order to estimate the un-
certainty of the output measurement. The primary-element uncertainty in-
creases to over 3 percent in venturi tubes for a variety of less-than-optimum
conditions such as insufficient upstream approach length, roughening and
aging, and low Reynolds number. Similar effects can be expected for venturi
nozzles although they are not as well-documented.
*
For cases in which the "standard" uncertainty is acceptable to the user,
the initial performance check consists of the calibration of the secondary
system as described in section 11.2. However, nonstandard tubes and/or in-
stallations may require an initial field calibration of the primary-element
coefficient as well as the secondary system. The primary calibrations are
described in section 11.3 and usually involve either volumetric or dilution
methods, or comparison with a reference meter.
Methods of estimating the uncertainty of the calibration (reference)
measurements are given (section 11.4) so that the performance of the on-line
system can be equitably evaluated.
This report was submitted as part of Interagency Agreement No. 78-D-
X0024-1 between the Environmental Protection Agency and the National Bureau
of Standards.
iv
-------
CONTENTS
Foreword ii
Abstract iv
1. Scope of Recommended Practice 1
2. Nomenclature and Definitions 2
2.1 Nomenclature 2
2.2 Definitions 2
3. Principles 4
4. Specifications for Classical Venturi Tubes 6
4.1 General 6
4.2 Overall Geometry 6
4.3 Pressure Taps ; 8
4.4 Materials and Construction 9
4.5 Discharge Coefficients 10
5. Installation Requirements for Venturi Tubes 12
5.1 General 12
5.2 Valves 12
5.3 Pumps 12
5.4 Bends and Other Fittings 12
5.5 Pipeline 14
5.6 Alignment 15
5.7 Other Considerations 15
6. Specifications for Venturi Nozzles 17
6.1 General 17
6.2 Overall Geometry 17
6.3 Pressure Taps 19
6.4 Materials and Construction 21
6.5 Discharge Coefficients 21
7. Installation Requirements for Venturi Nozzles 25
7.1 General , 25
7.2 Valves 25
7.3 Pumps 25
7.4 Bends and Other Fittings 25
7.5 Pipeline 28
7.6 Alignment 29
7.7 Other Considerations 29
8. Specifications for Secondary Systems 31
8.1 General 31
8.2 Location Requirements 31
8.3 Transmission 31
8.4 Connections Between Primary and Secondary 31
8.5 Purging 33
9. Error Sources 35
9.1 Primary Elements 35
-------
9.2 Installation 35
9.3 Pulsations 36
10. Operation and Maintenance Requirements 37
10.1 Secondary System 37
10.2 Primary 38
10.3 Sludge Flows 38
11. Performance Checks and Calibrations 41
11.1 General 41
11.2 Calibrating Secondary System with Manometers ... 42
11.3 Calibrating Secondary System with Standpipes ... 44
11.4 Calibration of the Complete System 44
11.5 Approximate Methods 49
11.6 Estimating Errors 51
12 References 53
Appendices
A. Footnotes 54
B. Expansibility Factors 56
vi
-------
1. SCOPE OF RECOMMENDED PRACTICE
1.1 This practice covers classical venturi tubes, truncated classical venturi
tubes, and venturi nozzles in circular pipes flowing full.
1.2 This practice covers venturi tubes and nozzles for use in wastewater
treatment plants, i.e., for flowrate measurement of influent wastewater,
treated effluent, air to aeration tanks, raw sludge, digested sludge,
and activated sludge.
1.3 This practice covers:
- Meter (primary element) construction and configuration;
- Meter (primary element) installation requirements;
- Secondary element installation requirements; and
- Performance checks and error estimates.
1.4 The purpose of this practice is to provide users with a technical base
that enables them to:
Specify the proper instrument (type and size) for the various appli-
cations;
- Check the measuring system after installation; and
- Monitor subsequent performance as necessary.
-------
2. NOMENCLATURE AND DEFINITIONS
2.1 Nomenclature. Terms are defined here and where they first appear in the
text.
c = tracer concentration, in flow measurement by dilution
d = throat diameter
g = acceleration due to gravity
Ah = differential head on meter in terms of height of the flowing fluid
Ap = differential pressure on meter
q = flowrate of added tracer, in flow measurement by dilution
A = throat area, ird2/4
B
C = flow coefficient
D = inlet diameter
M = geometric constant in equation [6]
N = geometric constant in equation [6]
Q = volumetric flowrate
Q = mass flowrate of air
m
R = Reynolds number, VD/v
U = average velocity
3 = diameter ratio, d/D
& = uncertainty, in equation [6]
e = expansibility factor, in gas flow
p = density of flowing fluid „
v = kinematic viscosity of flowing fluid (length /time)
2.2 Definitions.
2.2.1 Density — Mass per unit volume of fluid, or weight per unit
volume divided by the value of acceleration due to gravity.
2.2.2 Hydraulic grade line — A profile of the piezometric or static
pressure level of the fluid at all points along a line; in a
liquid flow, the height to which the liquid would rise in a
piezometer tube.
2.2.3 Non-Newtonian fluid — A fluid which does not exhibit the simple
linear Newtonian relation between shear stress, laminar velocity
gradient and viscosity. For example, a threshold yield stress
may have to be exceeded before flow can start (plastic fluid)
or the behavior may depend upon prior history of motion
(thixotropic fluid) .
-------
2.2.4 Primary element — In this context the primary element is the
venturi device itself, which generates the pressure difference,
the measurable parameter that characterizes the flowrate.
2.2.5 Purge — A continuous flow of external, clean water inward to
the venturi tube through the pressure taps to clean and flush
the taps.
2.2.6 Reynolds number — A dimensionless number characterizing the
ratio between inertial and viscous effects in a flow. Low
Reynolds numbers (below about 2000) describe laminar flows,
which are dominated by viscous effects. See section 4.5.2.
2.2.7 Secondary element — The device which measures the differential
pressure generated by the primary element.
2.2.8 Specific gravity — Ratio of the density of a fluid to that
of pure water at 4 degrees Celsius.
2.2.9 Uncertainty — Twice the standard deviation of a number of
points scattered about an average value. The true value of a
measurement has a 95 percent probability of falling within this
band.
2.2.10 Venturi tube — A device which, by a relatively long converging
section, gradually contracts an originally parallel flow to a
higher velocity (smaller area) parallel flow and then diverges
it gradually back to a lower velocity. (See Figure 1, 4.2.1).
•
2.2.11 Venturi nozzle — A nozzle-type curvilinear and relatively abrupt
contraction followed by a gradual divergence similar to that of
the venturi tube. (See Figure 2, 6.2.1.)
-------
3 . PRINCIPLES
3.1 The volumetric flowrate, Q, is given by
1 I1) -
Q = CAB(2gAh)1/Z [1]
where, in compatible units:
A = ird2/4
d = throat diameter
B
3 = ratio of throat to inlet diameter, d/D
Ah = pressure difference between inlet and throat pressure taps
in terms of height of flowing fluid
g = acceleration due to gravity
C = flow coefficient, a function of geometry, roughness and
Reynolds number.
3.2 Alternatively, Ah can be expressed as Ap/pg, where Ap is the pressure
difference in terms of force per area and p is the density of the
flowing fluid. Equation [1] then becomes
Q = CAB(2Ap/p) [2]
3.2.1 For practical purposes the density of influent or effluent
sewage can be considered to be the same as that of water.
3.2.2 For sludge flows the density of the sludge may differ enough
from that of water to be taken into account in the equations .
See section 10.3.
3.3 For air (or other gases) the flowrate is determined from
Q = CeAB(2AP/Pl)1/2 [3]
where
PI = air density at inlet
e = expansibility factor given in the Appendix
3.4 The recommended values for C will be given in sections 4.5.2 and
6.5.2 for tubes and nozzles which meet the requirements of sections
4 through 7 for standard conditions.
-------
3.4.1 For tubes and nozzles which do not'meet standard requirements
but for which equations 1 and 2 are still valid, see sections
4.5.3 and 6.5.4.
3.4.2 If a standard value of C can be applied, it is only necessary
to use equation [1] and [2] with an independent pressure
difference measurement to check the performance of the
secondary system. See section 11.2.
-------
4. SPECIFICATIONS FOR CLASSICAL VENTURI TUBES
4.1 General. Adherence to the following specifications is required only for
those meters which are to be used with the standard coefficients of
section 4.5.2 (footnote 1).
4.2 Overall Geometry.
4.2.1 The venturi tube consists (progressing downstream) of a
cylindrical inlet section of the same diameter of the upstream
pipe, a conical convergent section leading into a cylindrical
throat section (throat-to-inlet diameter ratio is 3), and a
conical divergent section. See Figure 1.
Comment: A meter geometry of course can be specified for
procurement purposes simply by referring to the appropriate
published standard. However, important details of the geometry
are furnished in the following so that the user can check
critical dimensions before installation.
Comment: The geometry requirements cited in the following
have been adapted from ISO specifications (1).* Of the three
options included by ISO, only the rough-cast convergent is
Figure 1. Classical venturi tube.
*Numbers in parentheses are references listed in section 12.
-------
considered here, partly because it is the most common type
in the United States (2, p. 230) and also because exposure
to sewage is likely to roughen any smoother configuration
anyway.
4.2.2 Mark the direction of flow clearly with an arrow on the outside
of tube to prevent installation error.
4.2.3 Inlet Section.
4.2.3.1 The diameter, D, of the inlet section shall be within
0.01D of the diameter of the upstream pipe.
4.2.3.2 The roundness of the inlet section shall be such that
no diameter differs from the mean diameter by more than
0.004D.
4.2.3.3 The length of the inlet section is preferably ID.
However, for D larger than about 35 cm (14 in) the
length may be reduced to 0.25D + 25 cm (10 in).
4.2.3.4 The inlet section should be faired into the convergent
section with a radius of 1.375D (+ 20 percent).
4.2.3.5 The inlet section should be no rougher than the
convergent section.
4.2.4 Convergent Section.
•
4.2.4.1 The convergence angle shall be 21 degrees + 1 degree.
4.2.4.2 The convergence angle shall be conical to within
0.004D when checked with a template.
4.2.4.3 The conical convergent section shall be faired into
the throat with a radius of between 3.625d and 3.75d,
where d is the throat diameter.
4.2.4.4 The surface of the convergent section should have a
sand cast finish, but it should be free of cracks,
protuberances, depressions or other obvious irregulari-
ties which could still satisfy paragraph 4.2.4.2.
4.2.5 Throat.
4.2.5.1 The length of the throat shall be Id. This is measured
from the intersections of the extended conical convergent
and divergent sections with the extended throat cylinder.
4.2.5.2 The throat roundness shall be such that no diameter in
the plane of the pressure taps differs from the mean of
at least four equally spaced measurements by more than
-------
O.OOld.
4.2.5.3 The average throat diameter, d, shall be such that the
ratio g = d/D is between 0.30 and 0.75.
4.2.5.4 The downstream end of the cylindrical throat should be
faired into the conical divergent section with a radius
of approximately lOd. See also paragraph 4.2.6.1.
4.2.5.5 The throat section shall be machined or of equivalent
smoothness.
4.2.6 Divergent Section.
4.2.6.1 The conical divergent section can diverge with an
included angle as large as 15 degrees, but 7 or 8
degrees is preferable. If a 15 degree angle is used,
the radius cited in paragraph 4.2.5.4 should be
decreased to about 5d.
4.2.6.2 The divergent section can be truncated if necessary
by up to 35 percent of its length without changing
the flow coefficient or substantially changing the
head loss.
4.3 Pressure Taps.
4.3.1 The upstream pressure taps shall be located upstream of the
intersection (-projected straight lines) of the cylindrical en-
trance and conical convergent sections by a distance of 0.5D
+ 0.25D for D up to 15 cm (6 in.) and 0.5D - 0.25D, +0 for larger
sizes.
4.3.2 The throat pressure taps shall be located downstream of the
intersection (projected straight lines) of the conical convergent
and cylindrical throat sections by a distance of 0.5d + 0.02d.
4.3.3 In general (for clean fluids) there should be at least four
pressure taps in a plane perpendicular to the tube axis at each
of the above locations, with the taps evenly distributed around
the periphery. See sections 4.3.5 for dirty fluids.
4.3.4 When multiple taps are used as in paragraph 4.3.3 they should be
connected by an annular chamber of which the cross-sectional area
is more than half the total area of the tapping holes. Double
this area if asymmetric flows are suspected. The connection from
the annular ring to the secondary tubing should not be directly
opposite any of the taps.
4.3.5 For metering dirty fluids and sludge, use only one pressure tap
at each measuring station (3). An exception can be made if it
can be shown to the satisfaction of involved parties that clogging
8
-------
of annular rings can be avoided. See section 5.4.4 for possible
effect on accuracy. See also section 4.3.7 for sewage and
sludge lines.
4.3.6 Pressure taps shall be smooth and burr-free on the inside tube
surface. The maximum tap-hole diameters shall be 0.1D and
0.13d for inlet and throat taps, respectively, but not larger
than 1 cm (0.4 in) nor smaller than 0.4 cm (0.16 in). The
diameter of the tap holes shall remain constant for a distance
of at least 2-1/2 diameters.
4.3.7 Special considerations for sewage/sludge lines.
4.3.7.1 The following precautions pertain to sewage in all
stages of processing including treated effluent as well
as to sludges.
4.3.7.2 Pressure taps for use with sewage and sludge should have
built-in capability for manual rodding of holes. See
section 8 for details.
4.3.7.3 For sewage and sludge the tap hole diameters should be
at least the maximum size recommended in paragraph
4.3.6.
Comment: For larger venturi tubes, maximum diameters of
1.9 cm (3/4 in) have been found advantageous in reducing
clogging without materially reducing accuracy.
4.3.7.4 See also paragraphs 4.3.5 and 5.5.5.
4.3.8 Proprietary systems employing flush diaphragms at the taps rather
than open taps are also acceptable provided the following con-
ditions are met.
4.3.8.1 The diaphragms must always be flush with the surface.
4.3.8.2 The system must meet accuracy requirements for secondary
systems as cited in paragraph 8.1.
4.3.8.3 Provision must be made for the user to check the system
accuracy, preferably by providing an alternate set of
conventional open taps.
4.4 Materials and Construction.
4.4.1 In principle the venturi tubes can be made of any material which
can be formed or machined to geometric specifications, is stable,
and can meet the surface requirements of paragraphs 4.2.3.5,
4.2.4.4, and 4.2.5.5.
4.4.2 In practice, venturi tubes for clean fluids are generally made of
cast iron with bronze-lined throat. These materials may also be
-------
adequate for influent sewage and treated effluent (footnote 2).
However, bronze or stainless-steel lining of both the convergent
and throat sections is recommended for this application.
4.4.3 Grease-resistant linings are recommended (but are not compulsory)
for lines carrying sludge (4, p. 69) (footnote 3).
4.4.4 The fabrication method must provide for smooth transitions
throughout the venturi tube; that is, if the tube is not
fabricated in one piece, provision must be made to avoid
offsets at the assembly joints.
4.5 Discharge Coefficients.
4.5.1 This section provides information for:
- Determining the coefficient C for use in equation [1], [2] or
[3] for tubes which meet the fabrication specifications of
sections 4.1 through 4.4, and which further meet the
installation specifications of section 5; and
- Ascertaining the validity of manufacturers' values of C for
tubes which differ from standard tubes.
4.5.2 Standard venturi tubes fabricated and installed in conformance
with sections 4 and 5 can be assigned a basic coefficient C of
0.984 for use in equation [1] and [2] when the following restric-
tions are observed:
(a) Pipe diameter, D, between 10 and 80 cm (about 4 to 32 in);
(b) 6 between 0.3 and 0.75; and
(c) Reynolds number, R, above 200,000. Here the Reynolds number
is defined as
R = UD/v
where U is the average velocity in the inlet pipe of
diameter D and v is the kinematic viscosity of the flowing
liquid. See also paragraph 4.5.2.3.
4.5.2.1 The value of C with the restrictions above has an
uncertainty of about 0.7 percent (1).
4.5.2.2 For flows which do not meet the Reynolds number
requirement of 200,000 the following estimates can be
made:
R £ Uncertainty, %
150,000 0.982 1.0
100,000 0.976 1.5
60,000 0.966 2.0
40,000 0.957 2.5
10
-------
4.5.2.3 The Reynolds number condition of paragraph 4.5.2(c)
above is considered to be firmly supported by data up
to a Reynolds number of 2 x 10 . Also, the preponderance
of evidence suggests that C remains the same for even
higher values of R, so for the purposes of this practice
no restriction is placed on maximum R.
4.5.2.4 Before applying the coefficients cited in this section,
the user should examine sections 5 and 9 for estimate
of effects of on-site conditions and section 10.3 for
sludge flows.
4.5.3 Nonstandard venturi tubes, i.e., those which do not conform to
the specifications of sections 4.1 through 4.4, can still be
acceptable provided the following conditions are met.
4.5.3.1 The manufacturer shall furnish detailed information on C,
which shall have been determined from laboratory experi-
ments or referenced to earlier standards. This informa-
tion should include Reynolds number dependence, estimated
roughness dependence and uncertainty of C, along with the
tube dimensions necessary to achieve the given value of
C. The manufacturer must be able to document upon re-
quest the number and type of experiments performed along
with enough related information to establish for the in-
volved parties the validity and stability (against abrupt
shifts due-to hydrodynamic causes, for example) of the
discharge coefficient, C. Similarly, if the values of C
are based on adaptations of existing values rather than
on experiments, the rationale for these values will be
made available to the user on request.
4.5.3.2 A manufacturer may find it necessary or desirable to use
a form of equation different from those given in section
3, or graphs and tables rather than equations. In any
event he shall furnish information equivalent to that of
paragraph 4.5.3.1.
4.5.3.3 The manufacturer shall inform the user of any changes in
the standard installation requirements of section 5 and
shall be able to document the reasons for such changes
on request. The absence of notification of such changes
implies the applicability of section 5.
4.5.3.4 Failure of a nonstandard venturi tube to qualify under
section 4.5.3 requires an in-place calibration of the
meter system. See section 11.4.
11
-------
5. INSTALLATION REQUIREMENTS FOR VENTURI TUBES
5.1 General.
5.1.1 Section 5 describes installation conditions which insure that
flows entering the venturi tube are of sufficient quality for the
discharge coefficients of section 4.5 to be valid. See paragraph
4.5.3.3 for non-standard venturi tubes.
5.1.2 If any of the following installation requirements cannot be met,
or if fittings are used which are not covered in this section,
the system may still be acceptable without a full calibration if
it can be shown independently to the satisfaction of the involved
parties that acceptable entry flows exist.
5.2 Valves.
5.2.1 If a flow control valve is necessary in the line, place it down-
stream of the venturi tube. See paragraph 5.4.5 for downstream
distance.
5.2.2 If an isolation valve is necessary upstream of the venturi, use a
gate valve and make certain that it is fully open during flowmeter
measurements. See Table 1 for minimum upstream distances.
5.3 Pumps.
5.3.1 In the case of centrifugal pumps, locate the venturi tube on the
inlet (suction) side (2) whenever this can be done without intro-
ducing subatmospheric pressure in the throat. If location on the
discharge side is unavoidable, allow a minimum length of 10D
between pump and venturi (5).
5.3.2 If a reciprocating pump is used, e.g., in a sludge line, the
venturi tube cannot be recommended for accurate measurement. To
minimize the error (which usually will be such as to indicate an
apparently higher flow) locate the venturi tube as far as possible
downstream of the pump (2).
5.4 Bends and Other Fittings.
5.4.1 Table 1 gives recommended minimum straight lengths between the
closest upstream fitting and the upstream pressure taps of the
venturi tube. These lengths are the minimum for which the
uncertainties in C given in paragraphs 4.5.2.1 and 4.5.2.2 are
valid (1).
12
-------
5.4.2 Table 2 gives shorter allowable straight lengths between the
closest upstream fitting and the venturi tube for which an
additional 0.5 percent uncertainty in C must be considered.
Add this 0.5 percent arithmetically to the uncertainties given
in paragraph 4.5.2.1 and 4.5.2.2 (1).
5.4.3 If several fittings (other than 90-degree bends) are in series
upstream of the venturi tube, the minimum straight length be-
tween the second and first (closest) upstream fitting should be
equal to one-half the Table 1 value for the second fitting with
3 = 0.7, regardless of the actual g value. This length causes no
additional uncertainty in C. If one-half the corresponding Table
2 value is used, add another 0.5 percent to the uncertainty in C.
5.4.4 Single-tap venturi tubes.
5.4.4.1 If a single-tap venturi tube is downstream of a single
bend, orient the tap perpendicular to the plane of the
bend whenever possible.
5.4.4.2 The distances in Tables 1 and 2 pertain to tubes with
multiple taps and annular chambers. Use Table 1 dis-
tances for single taps. There is very limited evidence
available (footnote 4) which suggests an additional 1
percent uncertainty in C for single-tap tubes close to
asymmetric disturbances such as upstream bends.
TABLE 1.
MINIMUM NUMBER OF PIPE DIAMETERS BETWEEN SELECTED
FITTINGS AND VENTURI TUBE
Reducer, 3D
Two or More to D Over Expander, 0.75D
Single 90° Bend, 90° Bends in Length of to D, Over Gate Valve
B Radius < D Same Plane 3.5D* Length of D Fully Open
0.30 0.5 1.5 0.5 1.5 1.5
0.35 0.5 1.5 1.5 1.5 2.5
0.40 0.5 1.5 2.5 1.5 2.5
0.45 1.0 1.5 4.5 2.5 3.5
0.50 1.5 2.5 5.5 2.5 3.5
0.55 2.5 2.5 6.5 3.5 4.5
0.60 3.0 3.5 8.5 3.5 4.5
0.65 4.0 4.5 9.5 4.5 4.5
0.70 4.0 4.5 10.5 5.5 5.5
0.75 4.5 4.5 11.5 6.5 5.5
*Abrupt symmetrical reductions with diameter ratio larger than 1/2, use 30D;
for entrance from large reservoir, total distance to primary should exceed
30D even if there is an intervening fitting that allows a smaller value.
13
-------
TABLE 2.
MINIMUM NUMBER OF PIPE DIAMETERS BETWEEN SELECTED UPSTREAM FITTINGS
VENTURI TUBE FOR 0.5 PERCENT ADDED UNCERTAINTY
Single 90°
Bend
Radius > D
Two or More
90° Bends
in Same
Plane
Two or More
90° Bends,
Different
Planes*
Reducer, 3D
to D Over
Length
of 3.5Dt
Expander,
0.75D to
D, Over
Length
of D
Gate Valve
Fully Open
0.30
0.35
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
+
+
+
0.5
0.5
0.5
1.0
1.5
2.0
3.0
0.
0.
0.
0.
1.5
1.5
2.5
2.5
2.5
3.5
0.
0.
0.
0.
8.
12.5
17.5
.5
.5
.5
.5
.5
23,
27,
29.5
0.
0.
0.
0.
0.
0.
1.5
2.5
3.5
1.0
1.5
1.5
1.5
2.5
3.5
4.5
0.5
0.5
1.5
1.5
1.5
2.5
2.5
2.5
3.5
3.5
*These fittings have an effect even 40D downstream; hence no entry in Table 1.
+Since fittings cannot be closer than 0.5D to venturi tube, only the Table 1
values are valid here.
tAbrupt symmetrical reductions with diameter ratio larger than 1/2, use 15D;
for entrance from large reservoir, total distance to primary should exceed
15D even if there is an intervening fitting that allows a smaller value.
5.4.4.3 Do not use lengths smaller than l.OD with single-tap
tubes.
5.4.5 Downstream fittings. Conditions downstream of a venturi tube
have comparatively little effect on its performance, but
fittings should not be placed closer than 4d downstream of the
throat taps.
5.4.6 Straighteners. In clean fluids only, flow straighteners can be
installed upstream of the venturi tube in cases where there are
fittings not covered by Tables 1 and 2. Standardized straightener
designs are available (1, 2) but require long straight pipe
lengths (20D and 22D upstream and downstream, respectively).
Therefore, it is likely that an independent demonstration of
suitability with shorter lengths will frequently be resorted to
in accordance with paragraph 5.1.2.
5.5 Pipeline.
5.5.1 Size upstream. At the junction between the pipe and the venturi
tube inlet section, the mean pipe diameter must be within + 1
14
-------
percent of the inlet diameter D. Further, no measured pipe
diameter for a distance of 2D upstream should differ from this
measured mean by more than + 2 percent.
5.5.2 Size downstream. The downstream pipe diameter should not be less
than 90 percent of the diameter of the end of the divergent sec-
tion of the venturi tube.
5.5.3 Roughness. For a distance of 10 diameters upstream or for at
least the distances given in Tables 1 and 2 the pipe surface
should be the equivalent of smooth new commerial pipe. Further,
for the two diameters immediately upstream of the venturi tube
the pipe surface should be as smooth as the cast convergent.
There should be no pitting, incrustations or deposits. See
section 9.2.2 for effects on C of rougher pipes.
5.5.4 Gaskets. Do not allow gaskets to protrude into the interior,
particularly at joints close to the entrance to the venturi tube.
5.5.5 Orientation. In sewage and sludge flows, or for any flows in
which there are solids or condensate, it is recommended that the
venturi device be placed in a horizontal line.
5.6 Alignment.
5.6.1 The angular alignment of the pipe axis and venturi tube axis
should be within 1 degree.
5.6.2 The offset between the pipe and venturi tube centerlines at the
junction plane should not exceed 0.005D.
5.7 Other Considerations.
5.7.1 Drain holes. Install drain holes and vent holes in the .pipe close
to the meter as appropriate, but be certain that the holes are
closed while flow measurements are being made.
5.7.2 Tube selection considerations.
5.7.2.1 When a completely clean fluid is flowing, select the
venturi tube size so that, for the minimum flowrate
that is to be accurately metered, the head differential
is at least 2.5 cm (1 in) of water. Substantially larger
deflections will usually prevail for sewage and sludge
flows because of the upstream velocity requirements.
See section 10 for sludge flow velocities.
5.7.2.2 Avoid sub-atmospheric pressures in the throat.
5.7.3 Accessibility.
5.7.3.1 Design the installation so that the primary and secondary
15
-------
devices are accessible with reasonable effort.
5.7.3.2 It is desirable that, whenever possible, by-passes be
built into the flow circuits so that the primary devices
can be removed for examination. Hand holes that are loca-
ted and designed so that no other specifications are
violated can also be useful for the examination of
venturi-tube interiors.
16
-------
6. SPECIFICATIONS FOR VENTURI NOZZLES
6.1 General. Adherence to the following specifications is required only for
those meters which are to be used with the standard coefficients of
section 6.5.2.
6.2 Overall Geometry.
6.2.1 The venturi nozzle consists (progressing downstream) of a con-
vergent section with rounded profile leading into a cylindrical
throat section (throat-to-inlet pipe diameter ratio is 3)» and
a conical divergent section. The shape of the convergent section
depends upon whether B is greater or less than 2/3. See Figure 2.
Inlet
Piezoneter ring (slot)
Slot _
,* r
a
I
Throat
Divergent
Tap-
Piezoueter ring (tap)
Figure 2. Standard venturi nozzle (shown for d/D < 2/3)
17
-------
6.2.2 Inlet section, 3 < 2/3.
6.2.2.1 The upstream portion of the convergent section consists
of a flat face perpendicular to the tube axis. The
outer diameter of this flat surface is equal to the in-
side diameter of the upstream pipe and the inner diameter
is equal to 1.5d, where d is the throat diameter. The
radial width of this flat face approaches zero as 3
approaches 2/3.
6.2.2.2 The flat face of 6.2.2.1 is tangent to a circular arc
section of radius R equal to 0.2d + 10% (for 3 < 0.5) or
+ 3% (for 3 > 0.5. The center of this arc is 0.2d from
the inlet (flat face) and 0.75d from the nozzle axis.
6.2.2.3 The circular arc of 6.2.2.2 is tangent to a second cir-
cular arc radius R equal to d/3 + 10% (for 3 < 0.5) or
+ 3% (for 3 > 0.5)7 The center of this arc is 0.304d
from the flat face and 5/6 d from the nozzle axis.
6.2.3 Inlet section, 3 > 2/3.
6.2.3.1 When d is larger than 2D/3, there can be no flat face
and the curved section extends to the wall. The inlet
is fabricated as though 3 < 2/3 and the face is machined
down until the inner diameter of the flat portion is
equal to D. See Figure 2.
6.2.4 Throat section.
6.2.4.1 The total length of the throat section is 0.70d to 0.75d.
See 6.3.3.2 for location of pressure taps.
6.2.4.2 The diameter d is determined as the mean of four diameter
measurements made in different axial planes at approxi-
mately even intervals. The diameter at any throat cross
section shall not differ from this mean by more than 0.1
percent. (See paragraph 7.5.1 for determination of
diameter D.)
6.2.5 Divergent section.
6.2.5.1 The total included angle of the divergent section shall
not exceed 30 degrees. Within this limit the divergence
angle affects the pressure loss but not the flow coeffi-
cient.
6.2.5.2 The divergent section may be truncated similarly to the
classical venturi tube of paragraph 4.2.6.2.
6.2.5.3 There is no fairing at the junction of the cylindrical
throat and the divergent section. However, any burrs
18
-------
should be removed.
6.3 Pressure Taps.
6.3.1 Upstream pressure taps.
6.3.1.1 Piezometer ring. The diameter of the piezometer ring
must, be no less than l.OOD and no greater than 1.04D, with
the thickness not to exceed that shown in Figure 3.
0.6
o.s
0.4
•2 0.3
g
3
0.1
0.2
-1.02D
Ring Diameter
/- 1.010
0.4
0.5 0.6
d/D
0.7
0.8
Figure 3. Allowable piezometer (carrier) ring thickness for venturi nozzles.
6.3.1.2 Upstream taps are always corner taps and may be in the
form of individual taps or an annular slot as in Figure
2. See also paragraph 6.3.4.1 for dir^y fluid cases.
6.3.1.3 The diameter of the individual taps or the width of the
annular slot must be between 0.005D and 0.03D for
6 < 0.65; and between 0.01D and 0.02D for g > 0.65,
except that the limiting values shall be between 1 mm
and 10 mm for clean fluids.
6.3.1.4 Individual corner tappings should be, as closely as
possible, perpendicular to the nozzle axis.
19
-------
6.3.2 Throat pressure taps.
6.3.2.1 The throat pressure taps are always individual taps
leading into an annular chamber or ring. There should be
at least four of these taps, evenly spaced around the
throat circumference. See paragraph 6.3.4.1 for except-
ion.
6.3.2.2 The throat pressure taps are located 0.3d downstream from
the beginning of the cylindrical throat section.
6.3.2.3 The diameter of the throat taps must be less than or
equal to 0.04d and between 0.2 and 1.0 cm.
6.3.3 General pressure tap conditions.
6.3.3.1 The face of the taps must be circular, smooth and burr-
free.
6.3.3.2 Length of slot should be at least twice its width; length
of individual taps at least 2.5 times diameter. See also
paragraph 4.3.4.
6.3.4 Special considerations for dirty fluids and sewage or sludge lines.
6.3.4.1 In order to avoid geometries that may encourage fouling,
it is recommended that single rather than multiple taps
be used in dirty fluids unless it can be shown that annu-
lar rings can be prevented from clogging. See section
7.4 for effect on accuracy.
6.3.4.2 Taps for use with dirty fluids, and particularly with
sewage and sludge, must have built-in capability for
manual "rodding" of the holes.
6.3.4.3 For sewage and sludge, always use the maximum hole dia-
meter permitted under paragraphs 6.3.1.3 and 6.3.2.3.
6.3.4.4 Observe the precautions of this section for all stages of
in-process sewage up through and including treated ef-
fluent. See also paragraph 7.5.4.
6.3.6 Proprietary systems employing flush diaphragms at the taps rather
than open taps are also acceptable provided the following condi-
tions are met.
6.3.6.1 The diaphragms must always be flush with the surface.
6.3.6.2 The system must meet accuracy requirements for secondary
systems as cited in paragraph 8.3.
20
-------
6.3.6.3 Provision must be made for the user to check the system
accuracy, preferably by providing an alternate set of
conventional open taps.
6.4 Materials and Construction.
6.4.1 Venturi nozzles may be made of any material which can be formed or
machined to geometric specifications, is stable, and which also
can conform to paragraphs 6.4.2 and 6.4.3.
6.4.2 The roughness of the convergent and throat surfaces should not
exceed the equivalent of smoothly finished cast bronze.
6.4.3 The material shall have corrosion resistance appropriate to the
intended use of the nozzle.
6.4.4 See also paragraphs 4.4.3 and 4.4.4.
6.5 Discharge Coefficients.
6.5.1 This section provides information for:
- Determining the coefficient C for use in equation [1], [2] or
[3] for venturi nozzles which meet the fabrication specifica-
tions of sections 6.1 through 6.4, and which further meet the
installation specifications of section 7; and
- Ascertaining the validity of manufacturers' values of C for
nozzles which differ from standard ones.
6.5.2 Standard venturi nozzles fabricated and installed in accordance
with sections 6 and 7 have a basic discharge coefficient given
by (see also Figure 4)
C = 0.986 - 0.19664'5
when the following restrictions are also observed:
(a) Pipe diameter, D, between 6.5 and 50 cm (about 2-1/2 to 20
inches);
(b) Throat diameter larger than 5.0 cm (about 2 inches);
(c) B between 0.316 and 0.775;
(d) Reynolds number, R, between 150,000 and 2,000,000.
(e) The relative roughness of the pipe, k/D, for a distance of at
least 10D upstream of the nozzle, is within the following
.limits.
for 6 = 0.35, k/D < 25 x 10~£
0.40, < 10.6 x 10_^
0.45, < 7.1 x 10_?
0.50, < 5.6 x 10_
0.60, < 4.5 x IQ_
> 0.75, < 3.9 x 10
21
-------
1.00
0.98
0.96
- 0.94
o.g;
0.3
0.4
0.5
0.6
0.7
0.8
d/D
Figure 4. Venturi nozzle discharge coefficient.
Here k is a roughness height for which guidelines are given in
Table 3. Before applying the foregoing coefficients, the user
should examine sections 7 and 9 for possible effects of on-site
conditions.
6.5.3 Within the foregoing limits the uncertainty of the coefficient
ranges from a little over 1 percent for the smallest 6 to almost
2 percent for the largest g. (This uncertainty is actually based
, on C//1 - g1*, but is given here for convenience on C.) Also,
although published standards do not cite values of venturi-nozzle
C for Reynolds numbers lower than above, comparison, with nozzle
data suggests an additional uncertainty of at least 1 percent
for R as low as 50,000.
6.5.4 Nonstandard venturi nozzles, i.e., those which do not conform to
the specifications of sections 6.1 through 6.4, can still be
acceptable provided the following conditions are met.
22
-------
TABLE 3.
EXAMPLES OF PIPE ROUGHNESS
Material
Condition
*Roughness height.
k,* mm
I . Aluminum
Brass
Copper
Glass
Plastic
II. Steel
Smooth
Smooth
Smooth
Smooth
Smooth
New, seamless
New, seamless
New, seamless
, cold drawn
, hot drawn
, rolled
New, longitudinal welds
New, bituminizing
New, spiral welds
Slight rust
Rusty
Encrusted
Heavily Encrusted
III. Cast Iron
IV. Asbestos Cement
Galvanized
Bituminized,
New
Rusty
Encrusted
New
normal
Normal, not insulated
0.03
0.03
0.03
0.03
0.03
0.03
0.05-0
0.05-0
0.05-0
6.03-0
0.10
0.10-0
0.20-0
0.50-2
2.0
0.13
0.10-0
0.25
1.0-1.
1.5
0.03
0.05
.10
.10
.10
.05
.20
.30
.0
.20
5
6.5.4.1 The manufacturer shall furnish detailed information on C,
which shall have been determined from laboratory experi-
ments or referenced to earlier standards. This informa-
tion shall include Reynolds number dependence, roughness
dependence and uncertainty of C, along with the meter
dimensions necessary to achieve the given values of C.
The manufacturer shall be able to document upon request
the number and type of experiments performed along with
enough related information to establish for the involved
parties the validity and stability (against abrupt shifts
due to hydrodynamic causes, for example) of the discharge
coefficient, C. Similarly, if the values of C are based
on adaptations of existing values rather than on experi-
ments, the rationale for these values shall be made
23
-------
available to the-user on request.
6.5.4.2 A manufacturer may find it necessary or desirable to use
a form of equation different from those given in section
3, or graphs and tables rather than equations. In any
event, he shall furnish information equivalent to that of
paragraph 6.5.4.1.
6.5.4.3 Install nonstandard venturi nozzles in conformance with
section 7, unless the information provided by the manu-
facturer includes documentation which either requires or
justifies exceptions.
6.5.4.4 Failure of a nonstandard venturi nozzle to qualify under
section 6.5.4 requires an in-place calibration of the
meter system. See section 11.4.
24
-------
7. INSTALLATION REQUIREMENTS FOR VENTURI NOZZLES
7.1 General.
7.1.1 Section 7 describes installation conditions which insure that
flows entering the venturi nozzle, are of sufficient quality for
discharge coefficients of section 6.5 to be valid. See para-
graph 6.5.4.3 for nonstandard nozzles.
7.1.2 If any of the following installation requirements cannot be met,
or if fittings are used which are not covered in this section,
the system may still be acceptable without a full calibration if
it can be shown independently to the satisfaction of the involved
parties that acceptable flows exist.
7.2 Valves.
7.2.1 If a flow control valve is necessary in the line, place it down-
stream of the venturi nozzle. See Tables 4 and 5 for minimum
downstream distance for gate and globe valves.
7.2.2 If an isolation valve is necessary upstream of the nozzle, use a
gate valve and make certain that it is fully open during flow mea-
surements. See Tables 4 and 5 for minimum upstream distances.
7.3 Pumps.
7.3.1 In the case of centrifugal pumps, locate the venturi nozzle on
the inlet (suction) side whenever this can be done without intro-
ducing subatmospheric pressure in the throat. There are no pub-
lished guidelines for cases where nozzle placement on the dis-
charge side of the pump is unavoidable. However, it is reason-
able to assign longer minimum distances than for the venturi tube
(paragraph 5.3.1) and a minimum length of 20D is recommended.
7.3.2 See paragraph 5.3.2 for reciprocating pumps.
7.4 Bends and Other Fittings.
7.4.1 Table 4 gives recommended minimum straight pipe lengths between
the closest upstream and downstream fittings and the venturi
nozzle. These lengths are the minimum for the uncertainties of
paragraph 6.5.3 to apply.
7.4.2 Table 5 gives shorter allowable straight lengths between the
closest upstream fitting and the venturi tube for which an
25
-------
TABLE 4.
MINIMUM NUMBER OF PIPE DIAMETERS BETWEEN SELECTED
FITTINGS AND VENTURI NOZZLE
e
0.30
0.35
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
Single
90°
Bend*
10
12
14
14
14
16
18
22
28
36
46
Two or More
90° Bends,
in Same
Plane
16
16
18
18
20
22
26
32
36
42
50
Two or More
90° Bends,
in Differ-
ent Planes
34
36
36
38
40
44
48
54
62
70
80
Reducer,
2D to D,
Length
1.5D to
3Df
5
5
5
5
6
8
9
11
14
22
30
Expander ,
0.5D
to D,
Length
ID to 2D
16
16
16
17
18
20
22
25
30
38
54
Globe Valve
Fully Open
18
18
20
20
22
24
26
28
32
36
44
Gate Valve
Fully Open
12
12
12
12
12
14
14
16
20
24
30
All listed
Fittings,
When Down-
stream
5
5
6
6
6
6
7
7
7
8
8
*Includes tee with flow from one branch only.
tAbrupt symmetrical reductions with diameter ratio larger than 1/2, use 30D; for
entrance from large reservoir, total distance to primary should exceed 30D even
if there is an intervening fitting that allows a smaller value.
-------
TABLE 5.
MINIMUM NUMBER OF PIPE DIAMETERS BETWEEN SELECTED FITTINGS AND VENTURI NOZZLE
FOR 0.5 PERCENT ADDED UNCERTAINTY
NJ
B
0.30
0.35
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
Single
90°
Bend*
6
6
7
7
7
8
9
11
14
18
23
Two or More
90° Bends,
in Same
Plane
8
8
9
9
10
11
13
16
18
21
25
Two or More
90° Bends,
in Differ-
ent Planes
17
18
18
19
20
22
24
27
31
35
40
Reducer,
2D to D,
Length
1.5D to
3Dt
5
5
5
6
7
11
15
Expander,
0.5D
to D,
Length
ID to 2D
8
8
8
9
9
10
11
13
15
19
27
Globe Valve
Fully Open
9
9
10
10
11
12
13
14
16
18
22
Gate Valve
Fully Open
6
6
6
6
6
7
7
8
10
12
15
All listed
Fittings,
When Down-
stream
2-1/2
2-1/2
3
3
3
3
3-1/2
3-1/2
3-1/2
4
4
*Includes tee with flow from one branch only.
tAbrupt symmetrical reductions with diameter ratio larger than 1/2, use 15D; for
entrance from large reservoir, total distance to primary should exceed 15D even
if there is an intervening fitting that allows a smaller value.
-------
additional 0.5 percent uncertainty in C must be considered. Add
this 0.5 percent arithmetically to the uncertainties given in
paragraph 6.5.3.
7.4.3 If several fittings (other than 90-degree bends) are in series
upstream of the venturi tube, the minimum straight length between
the second and first (closest) upstream fitting should be equal
to one-half the Table 4 value for the second fitting with B = 0.7,
regardless of the actual 3 value. This length causes no addi-
tional uncertainty in C. If one-half the corresponding Table 5
value is used, add another 0.5 percent to the uncertainty in C.
7.4.4 Single-tap venturi nozzles.
7.4.4.1 If a single-tap venturi nozzle is downstream of a single
elbow, orient the tap at a right angle to the plane of
the bend whenever possible.
7.4.4.2 The distances in Tables 4 and 5 pertain to tubes with
multiple taps and annular chambers. Use Table 4 dis-
tances for single-tap nozzles. Allow 0.5 percent added
uncertainty.
7.4.5 Straighteners. In clean fluids only, flow straighteners can be
installed upstream of the venturi nozzle for cases where the
lengths required by Tables 4 and 5 are not available or where
fittings other than those listed are used. Straight pipe lengths
are required for 20D and 22D upstream and downstream of the
straightener, respectively. Standardized straightener designs
are available (1,2). If these lengths are not available, see
paragraph 5.1.2.
7.5 Pipeline.
7.5.1 Size. For at least 2D upstream of the nozzle, the pipe should be
cylindrical to the extent that no diameter differs from an aver-
age diameter by more than 0.3 percent. This average diameter is
the mean of at least four diameter measurements made in at least
three planes in the first 0.5D upstream, including the inlet
(O.OD) plane. The average diameter so obtained is the value of
D used to determine 6. The downstream diameter should be within
3 percent of the divergent-section downstream diameter for a dis-
tance of at least 2D measured from upstream face.
7.5.2 Joints. The foregoing 2D pipe length should be a single pipe;
i.e., there should be no joints in it. Pipe joints farther up-
stream, up to the first fitting, are permitted, but they should
have steps or offsets not exceeding those indicated in Figure 5.
7.5.3 Roughness. The pipe must be clean, free of pits and deposits,
and for at least 10D upstream of the nozzle should have maximum
roughnesses as cited in paragraph 6.5.2. See paragraph 9.2.2.2
28
-------
0.05
0.04
0.03
0.02
0.01
/70
— Limits for 0.24
added uncertainty
Limit for Ot addet
uncertainty ->
\
d/D
10 20 30
Distance upstreaa of inlet taps, s/D
Figure 5. Upstream joint steps for venturi nozzle.
for effects of rougher pipes.
7.5.4 Orientation. See paragraph 5.5.5.
7.5.5 Gaskets. Gaskets installed near the nozzle should not protrude
beyond the inner pipe surface (or into the annular ring) and
should be thinner than 0.03D.
7.6 Alignment.
.»
7.6.1- The nozzle axis should be aligned with the pipe axis to within
+ 1 degree.
7.6.2 The piezometer ring must not protrude inside the pipe diameter at
any point. The permissible lack of concentricity between the
nozzle and pipe centerlines at the junction plane is given in
Figure 6.
7.7 Other Considerations.
7.7.1 Drain holes. Install drain and vent holes upstreaip of the nozzle.
If they are close to the nozzle their diameters should not exceed
0.08D, but in no case should they be closer than 0.5D to the
nearest pressure tap or longitudinally aligned with a pressure
tap. Be certain that drains and vents are closed off while mea-
surements are being made.
7.7.2 Nozzle selection considerations. See section 5.7.2.
29
-------
0.05
0.04
- 0.05
0.02
0.01
I
0.3* added
uncertainty -
0.3 0.4 O.S 0.6 0.7 0.8
d/D
Figure 6. Allowable lack of concentricity for venturi nozzles.
7.7.3 Accessibility. See section 5.7.3.
30
-------
8. SPECIFICATIONS FOR SECONDARY SYSTEMS
8.1 General.
8.1.1 The accuracy of each device within the secondary system (e.g.,
transducer/transmitter, receiver, recorder) shall be within 1/2
percent of full scale (footnote 1). Alternatively, this speci-
fication can be replaced with an accuracy requirement on the
entire system.
8.1.2 The manufacturer must furnish with each secondary device instruc-
tions for maintenance and for servicing that can be done in the
field.
8.2 Location Requirements.
8.2.1 Locate an indicator gage in the immediate vicinity of the pri-
mary element for convenience in performance checking.
8.2.2 Place the differential pressure transmitter below the hydraulic
grade line, to facilitate positive bleeding of lines.
8.3 Transmission.
8.3.1 Do not use pneumatic transmission unless the distances involved
are less than 300 m (1000 ft) and temperatures are always above
freezing.
8.3.2 The pressure differential shall be transmittable in computer-
compatible form or be capable of future conversion to such a
form.
8.4 Connections Between Primary and Secondary.
8.4.1 General. A major object of this section is to avoid accumulations
of gas or sediment in the connecting lines. To this end, arrange-
ments other than those cited here are acceptable where it can be
shown that the objective is accomplished. This section does not
apply to enclosed proprietary systems of sections 4.3.8 or 6.3.6.
8.4.2 Air flow. If air (or other gas) is being measured, connect the
secondary tubing near the top of the vertical meridian plane.
If there is an annular ring, include provision for occasional
(manual) draining of condensate from the bottom.
31
-------
8.4.3 Liquid flow.
8.4.3.1 If a clean liquid is flowing, the secondary tubing
should connect to the primary in the lower half of
the periphery (6) in order to prevent the trapping of
gas. Preferably make the connection in a zone about
45 degrees below the horizontal meridian (6). Treated
effluent is not a clean liquid for this purpose.
8.4.3.2 For dirty (i.e., containing solids or gas bubbles)
liquids, where there is only one pressure tap at each
station as recommended in section 4.3.5, the secondary
tubing connects directly to that tap. Orient the tube
so that the tap is in the upper half of the meridian
plane, preferably at about 45 degrees from the horizon-
tal. If there are multiple taps and an annular ring,
use a settling chamber and/or gas collector (or equiva-
lent devices) at the bottom and top of the ring,
respectively, in addition to the devices recommended in
section 8.4.4.
8.4.4 Connecting tubing.
8.4.4.1 Connecting tubing should be installed so that it has a
slope of at least 1 on 12 relative to the horizontal.
It is preferable that this slope be continuously upward
or continuously downward; however, this is often not
possible where it causes, for example, placement of the
secondary device in an inconvenient position or above
the hydraulic grade line. In any case the highest and
lowest points in the connecting tubing should be equip-
ped with gas collectors (or bleed valves) and sediment
chambers or condensate collectors as appropriate. Ex-
amples are shown in Figure 7.
8.4.4.2 The connecting tubing should be valved or otherwise
fitted so that all portions of the lines can be flushed
as necessary.
8.4.4.3 The tubing material should be resistant to corrosion.
The tubing bore should be at least 1 cm or 3/8 inch (6).
8.4.5 Other considerations.
8.4.5.1 Place shut-off valves in each line next to the primary
and preferably also next to the secondary, if such
valves are not already an integral part of the secondary.
8.4.5.2 Include a valved by-pass across the secondary instru-
ment, located between the secondary instrument and its
shut-off valves, for use in zero checks, unless such a
by-pass is furnished as an integral part of the secondary
32
-------
Cas collectors-
* V.lve
Secondary
, . equal iler-7 •> >
Secondary
- Bleeds •
-Primary
Figure 1, Typical secondary connections.
8.4.5.3
instrument.
Incorporate valved tees in each of the pressure lines,
preferably near the primary, so that an independent
pressure-differential measuring device can be installed
for the performance checks of section 11. The geometry
between this connection and the pressure tap should be
the same in both pressure lines.
8.5 Purging.
8.5.1 Continuous purging of pressure taps is required in sewage and
sludgeflows. This purging is preferably done with tap water.
8.5.2 The head loss in the tubing between the purge water connection
and the tap should be the same in both of the lines so that the
pressure differential is essentially unaffected. This can be
accomplished by making the two paths geometrically similar and
by keeping the purge flowrate the same in both legs. Install a
variable area flowmeter or equivalent and a flow control valve in
each of the purge water lines for flow adjustment.
33
-------
8.5.3 The purge supply pressure should be at least 10 psi higher than
the highest pressure anticipated at the tap. When flow through-
out the venturi is steady, the purge flow should be valved down
to the lowest flow consistent with good control. If the flow
through the venturi is unsteady, the purge flow should be at
least fast enough to keep dirty fluid from flowing into the
tubing as the capacity of the secondary device changes. In any
case purge flow should be made high enough during bleeding of
the secondary lines to keep clean fluid in the lines.
34
-------
9. ERROR SOURCES
9.1 Primary Elements.
9.1.1 Geometry. The effect of fabrication errors resulting in geometry
differing from that of sections 4.2, 4.3, 6.2 and 6.3 cannot be
quantitatively estimated. Treat such cases as nonstandard meters
in accordance with sections 4.5.3 and 6.5.4 for venturi tubes and
nozzles, respectively.
9.1.2 Roughness.
9.1.2.1 Tube. Roughening the meter itself is known to reduce C.
There is insufficient data to quantify this effect, but
it appears not to exceed 2 percent (1). Because meters
with rough-cast convergents are less susceptible to
roughness effects than those with machined convergents,
an estimated reduction of 1 percent in C is suggested
for venturi tubes that have been in routine treatment
plant use, with an added tolerance of 0.5 percent. See
also paragraph 10.3.4.4.
9.1.2.2 Nozzle. No experimental information is available on the
effect of nozzle roughness on the discharge coefficient.
In the absence of any guidelines, treat it the same as
the venturi tube of paragraph 9.1.2.1.
9.1.3 Taps. Substantial errors can be introduced by imperfections in
tap geometry. The throat taps are more critical in this regard
than the inlet taps because of the high throat velocities, i.e.,
similar but nonstandard inlet and throat taps will not necessarily
cancel their errors in a differential pressure measurement. Burrs,
corrosion and incrustations can have serious effects on the pres-
sure measurement, particularly in single-tap Venturis. Monitor-
ing the tap hole condition of meters which have been on-line in
adverse environments is therefore particularly important.
9.2 Installation.
9.2.1 Upstream lengths which do not meet the minimum conditions of
Tables 2 and 4 cause errors which cannot be estimated. An
in-place calibration generally is required, with the following
exceptions.
9.2.1.1 In-place calibration requirements can be waived if in-
flow of adequate quality is demonstrated under paragraph
35
-------
5.1.2 or 7.1.2.
9.2.1.2 For venturi tubes which have multiple taps and annular
chambers, the installation error is unlikely to exceed
3 percent (3). The in-place calibration requirement can
be waived if this is an acceptable figure. This para-
graph does not pertain to venturi nozzles.
9.2.2 Roughness of upstream pipe.
9.2.2.1 Venturi tubes. The values of C in section 4.5.2 are
premised on the use of commercially smooth pipe up-
stream. It is generally accepted that upstream pipe
roughness, by virtue of its effect on velocity distri-
bution, increases C by an amount which increases for
larger values of the diameter ratio, 3. There is in-
sufficient data available to quantify this error. How-
ever, an increase of 1 percent for large-3 tubes in
treatment plants is a reasonable estimate, along with
an additional 0.5 percent uncertainty. It is likely
that the smallest-g tubes will be relatively insensitive
to this effect. See also paragraph 10.3.4.4.
9.2.2.2 Venturi nozzles. There is insufficient experimental in-
formation to quantify errors due to pipe roughness. The
roughness requirements of paragraph 6.5.2 show that sen-
sitivity to this effect increases with b ratio. For 6
larger than 0.6, relatively smooth pipes are specified
and the effect of actual in-plant roughness can be ex-
pected to be larger than the corresponding effect for
venturi tubes.
9.3 Pulsations. Pulsating flow, such as that caused by reciprocating pumps,
will cause the meter reading to be too high, in part because the true
discharge is proportional to the average of the square roots of the in-
stantaneous pressure differentials while the indicated flow usually is
proportional to the square root of the average of the instantaneous
pressure differentials. See paragraph 5.3.2.
36
-------
10. OPERATION AND MAINTENANCE REQUIREMENTS
10.1 Secondary System.
10.1.1 Regularly bleed the secondary lines and/or vent the gas collec-
tors at the high points of the lines as provided for in para-
graph 8.4.4.1. The frequency of this operation can be deter-
mined only from experience with each situation. When flows are
expected to contain substantial amounts of gas or entrained air,
the secondary lines should have been equipped with gas collec-
tors rather than bleed valves. If these collectors do not have
automatic venting capability, they should be manually vented
once per shift to start with, until the appropriate frequency
is determined from experience.
10.1.2 Regularly check the sediment chambers or remove sediment through
valves at the low points of the secondary lines. If the second-
ary tubing is connected at the recommended position along the
primary (section 8.4.3) and purge flows are always in use, this
check probably can be made much less frequently than that of the
preceding paragraph. Again, the frequency will be determined
from experience gained from initial monitoring.
10.1.3 It is recommended that all secondary lines occasionally be
thoroughly flushed with purge flows higher than those normally
used. The frequency of such flushing will depend upon the fre-
quency with which solids enter the secondary lines and on the
cleaniness of the purge flow.
10.1.4 Check the zero reading of the secondary device by closing the
connections to the primary and opening the equalizing valve be-
tween the two sides of the secondary instrument (paragraph
8.4.5.2), or as otherwise provided for by the manufacturer.
This check should be made daily at first until more experience
is gained on the drift behavior of the system.
10.1.5 Follow manufacturers' instructions for routine instrument main-
tenance.
10.1.6 Make periodic checks of the secondary system using independent
manometry as described in section 11.2. This should be done
when the system is first installed and periodically (order of
weekly) thereafter until a final interval is determined from
experience.
37
-------
10.1.7 Check purge flows at least once each shift and more frequently
if there are flowrate changes through the primary. Equalize
as necessary.
10.1.8 Before resumption of flow following a period of shut-down, be
certain that purge fluid is flowing. Bleed the secondary lines
before resuming measurements.
10.2 Primary.
10.2.1 Use the rodding device on the pressure taps daily.
10.2.2 If the primary has annular rings, check them periodically for
gas, sediment or condensate accumulation as in paragraph 10.1.1.
10.2.3 Using the vent valve upstream of the primary, check periodically
for gas accumulation.
10.2.4 Long term maintenance on venturi tubes and venturi nozzles con-
sists primarily of examination of the interior of the primary
and the immediate upstream pipe in cases where this is possible
through diversion of flow elsewhere. Upon examination a deci-
sion can be made as to whether use of the venturi can continue
with the same or adjusted (estimated) coefficient or a complete
calibration (section 11.3) is necessary. The interval between
such examinations depends largely upon the hostile nature of
the flowing fluid, with sewage and sludge flows requiring the
most careful monitoring.
10.3 Sludge Flows.
10.3.1 General.
10.3.1.1 The hydraulic characteristics of sludge are not well-
defined, not only because most sludges exhibit non-
Newtonian behavior to some degree but also because
their make-up differs from plant to plant. The rec-
ommendations in this section have been assembled from
experiences reported in the literature and should be
regarded as estimates only. Further research on the
behavior of sludges in measuring devices is required
before standard practices can be outlined for this area.
10.3.1.2 See section 8.5 for purge flow requirements, which are
especially important in sludge applications.
10.3.2 Raw primary sludge.
10.3.2.1 Venturi tubes and venturi nozzles are not recommended
for accurate measurement of raw primary sludge be-
cause of the difficulty of attaining in this sludge
a fully turbulent flow with the effective Reynolds
38
-------
numbers cited in sections 4.5 and 6.5.
10.3.2.2 If use of a venturi device is unavoidable:
- Keep minimum velocities high, preferably about
8 ft/sec (245 cm/sec).
- Use large values of 0, particularly in the smaller
pipe sizes, so that the throat will pass large pri-
mary particles.
- Use a density in equation [2] based on a specific
gravity of 1.02, if a density correction is not
built in to the secondary recorder.
- Treat the recorded flowrate as an approximate value
since there is no basis for estimating the error
limits without a field calibration. The volumetric
calibration method of section 11.4.2 is generally
the most appropriate here.
- Pay particular attention to the internal inspection
recommendation of paragraph 10.2.4, since the grease
content of raw primary sludge is very high.
10.3.3 Activated sludge.
10.3.3.1 In general, fresh activated sludge has a solids con-
tent that is low enough for its hydraulic behavior in
turbulent flow to be similar to (but slightly more
viscous than) that of water and it can be metered
accordingly, as described in the following.
10.3.3.2 For the purposes of applying equation [1] or [2],
consider the density of fresh activated sludge to be
no more than 1 percent greater than that of water.
10.3.3.3 The effective viscosity of fresh activated sludge
is difficult to predict, but it may be in the
neighborhood of 1-1/2 to 2 times that of water. To
estimate a value of C, it is suggested a pipe Reynolds
number be first calculated from the meter reading as
though the fluid were water, and then divided by 1.5
or 2 to obtain an estimate sludge Reynolds number.
For venturi tubes use section 4.5 to estimate the
adjusted C. No adjustment is warranted if the esti-
mated sludge Reynolds number is greater than 150,000.
(Note that the standard venturi nozzle does not have
recommended values of coefficients for Reynolds num-
bers less than 150,000. It is suggested that, down
to Reynolds numbers of 40,000, the standard coeffi-
cients be used but with additional uncertainty of 1
percent, an estimate obtained by assuming that the
Reynolds-number dependence of the venturi nozzle
is similar to that of the nozzle.)
39
-------
10.3.3.4 Maintain velocities greater than 60 to 90 cm/sec (2
to 3 ft/sec).
10.3.3.5 The foregoing parts of section 10.3.3 do not pertain
to thickened activated sludge, for which the effec-
tive viscosity is substantially higher and less pre-
dictable.
10.3.4 Digested sludge.
10.3.4.1 Flow of well-digested sludge in concentrations (solids
content) up to approximately 4 percent can be measured
with venturi devices. The digestion process apparently
breaks up the raw sludge solids into particles that are
small enough so that at high velocities there is essen-
tially a turbulent Newtonian flow.
10.3.4.2 Maintain average velocities (in the pipe) of at least
50 cm/sec (5 ft/sec).
10.3.4.3 The effective viscosity of the digested sludge cannot
be accurately predicted. If, for estimating purposes,
a value of about 10 times that of water is used for
4 percent sludge; the effective Reynolds numbers are
likely to be in the lowest range cited in paragraph
4.5.2.2 for venturi tubes, suggesting a C of 0.96 with
an uncertainty of +_ 3 percent. (Corresponding esti-
mates for 3 percent sludge would be C = 0.97 + 2-1/2
percent.) A field calibration would be necessary to
obtain more precise values. See paragraph 10.3.3.3
for venturi nozzles.
10.3.4.4 For Reynolds numbers as low as 40,000 upstream of a
venturi tube (only), the pipe-roughness correction of
paragraph 9.2.2.1 should be omitted. If the Reynolds
number based on the throat diameter of the venturi
tube is less than 100,000, omit the tube-roughness
correction of paragraph 9.1.2.1.
10.3.4.5 For 4 percent sludge, use a density in equation [2]
that is based on a specific gravity of about 1.02.
Adjust this value for lower concentrations where
warranted.
10.3.4.6 Because digested sludge is notably high in gas con-
tent, precautions against gas accumulation in the
lines should be emphasized. See sections 10.1 and
8.5.
40
-------
11. PERFORMANCE CHECKS AND CALIBRATIONS
11.1 General.
11.1.1 Standard venturi tubes and venturi nozzles.
11.1.1.1 Section 11.1.1 pertains to venturi tubes and nozzles
which conform to all fabrication and installation
requirements of sections 4 and 5 or 6 and 7.
11.1.1.2 A newly installed venturi system which meets the re-
quirements of paragraph 11.1.1.1 requires only a
check on the secondary system, provided that the un-
certainties in C cited in sections 4.5.2 and 6.5.3
(or as increased for specified conditions cited in
subsequent sections) are acceptable. See paragraph
11.1.1.4 for follow-on monitoring.
11.1.1.3 To check the performance of the secondary system,
make an independent measurement of the pressure using
equation [1] for liquid flows and equation [3] for
compressible flows, and compare it with the output
reading of the installed system. See section 11.2
for details.
11.1.1.4 The performance of the venturi system of paragraph
11.1.1.2 must be checked periodically after it has
been in use. Again, a check of the secondary system
alone can be acceptable provided that:
- The combination of fluid properties, velocity and
venturi material is such that severe corrosion and
grease coatings can be ruled out; and
- A value of C is selected consistent with normal
roughening of the tube in routine use. See section
9.1.2.
Comment: There is an element of risk in this pro-
cedure in that any errors caused by small irregulari-
ties which may have developed near the edge of the
pressure taps will go undetected. This is of particu-
lar concern in single-tap Venturis. Make periodic
internal inspections of the tube where feasible.
11.1.2 Nonstandard venturi tubes and venturi nozzles.
11.1.2.1 Venturi tubes or venturi nozzles which are nonstandard
only in their dimensions but which meet the information
41
-------
requirements on C described in section 4.5.3 or 6.5.4
can be checked in the same manner as the standard
tubes of section 11.1.1.
11.1.2.2 Venturi tubes or venturi nozzles which are nonstandard
because they fail to meet the installation specifica-
tions of section 5 or 7 require a complete calibration
after installation (see section 11.4), unless it can
be shown to the satisfaction of all parties concerned
that adequate inflow conditions exist or unless the
conditions of paragraph 9.2.1.2 prevail. Once C has
been established with a full calibration over the
anticipated range of flow, future checks can be made
in accordance with paragraph 11.1.1.4.
11.2 Calibrating the Secondary System with Manometers.
11.2.1 Install a manometer (or its equivalent according to paragraph
11.2.1.3) at the connections provided under paragraph 8.4.5.3.
11.2.1.1 Do not use mercury manometers for differentials less
than 5 cm (2 in) of mercury; below this level use air-
water manometers.
11.2.1.2 The manometry used for this purpose must conform to
accepted good practice.
- As was the case for the secondary systems in section
8, the manometer tubing must provide for gas bleed-
ing and for zero checks as needed.
- The scale should permit reading the meniscus posi-
tion at least to the nearest 0.5 mm (0.02 in).
- Use glass tubing of large enough bore to minimize
the effect on the meniscus of dirt deposited on the
wall.
11.2.1.3 If a differential pressure transducer is used instead
of a manometer, it must be of demonstrable accuracy
to the satisfaction of the involved parties. Informa-
tion on its measurement uncertainty must be available
for later use in paragraph 11.6.2.1.
11.2.2 Use the purging system (section 8.5) to keep clean water in the
manometer tubes.
11.2.2.1 Check to see that the purge flows are small and equal
in both legs.
11.2.2.2 Where necessary, adjust the manometer reading for the
difference in density between the purge water and the
flowing liquid.
42
-------
11.2.2.3 Any uncertainty in the density of the latter should
be taken into account in the uncertainty estimate in
section 11.4.
11.2.2.4 After the manometer readings of paragraph 11.2.3 below
have been made, it is advisable to stop the purge flow
just long enough to make another set of manometer
readings in order to ascertain whether the pressure
differential calls for adjustment in purge flow rate
to one of the legs. Flow through the primary must
remain constant during this process.
11.2.3 Use the manometer to check the secondary system in the follow-
ing manner.
11.2.3.1 Before each series of measurements check or bleed the
manometer lines, check the manometer zero and purge
flows.
11.2.3.2 When the line flow appears to be steady, make several
manometer readings in fairly rapid succession. In-
crease the number of readings if the manometer columns
are oscillating. Use the average of these readings,
adjusted per paragraph 11.2.2.2, for the head differ-
ence. In general it is not advisable to dampen os-
cillations by closing down on valves in the manometer
line as error can be introduced in that way.
11.2.3.3 For liquid flows, compute the flowrate from equation
[1] using the head differential from paragraph 11.2.3.2
and either the standard or the manufacturer's value
of C.
11.2.3.4 Repeat this process several times for the same flow-
rate. Compare the results with the flowrates indica-
ted and/or recorded simultaneously by the secondary
system. To determine whether the differences are
within agreed-upon limits, refer to section 11.6.2.
11.2.3.5 Recheck the manometer zero after each series of mea-
surements at a given flowrate has been completed.
11.2.3.6 For compressible flows it is also necessary to mea-
sure the absolute pressure and temperature near the
inlet section so that the inflow gas density and ex-
pansibility factor can be determined for use in equa-
tion [3].
43
-------
11.2.3.7 If the venturi device is to measure a range of flows,
perform the foregoing procedure for at least three
flowrates—low, medium and high.
11.2.3.8 The foregoing paragraphs have assumed that the secon-
dary system output is in terms of flowrate. If the
readout is directly in terms of differential head or
pressure, the evaluation can be made after paragraph
11.2.3.2.
11.2.3.9 After examining the results of paragraph 11.2.3.7,
adjust or repair the secondary device as necessary.
11.3 Calibrating the Secondary System with Standpipes.
11.3.1 Instead of installing a manometer across the primary device and
in parallel with the secondary device, differential pressures
can be applied directly to the secondary device using water
standpipes.
11.3.2 The usual practices of good manometry should be observed in
order to obtain accurate measurements of the applied heads.
11.3.3 This method is most convenient for relatively small differential
pressures.
11.4 Calibration of the Complete System.
11.4.1 General.
Section 11.4 pertains to complete, in-place calibrations of
those venturi measuring systems which do not qualify for a
secondary-only calibration under sections 11.1.1 and 11.1.2.
11.4.1.1 The purpose of section 11.4 is to provide a general
overview of methods for determining in-place values
of the coefficient C so that, coupled with a separate
calibration of the secondary system, a complete cali-
bration of the measuring system is accomplished.
11.4.1.2 Therefore, as part of the complete calibration, check
the secondary system separately (and simultaneously,
for convenience) in accordance with section 11.2. In
this way, those differences between the reference and
recorded flowrates which are chargeable to the primary
device can be assigned to it, and future monitoring
of meters which qualify under paragraph 11.1.1.4 can
be restricted to the secondary system.
11.4.1.3 During the tests to determine C, use a manometer or
equivalent to measure the pressure difference. See
sections .11.2.1 and 11.2.2 and paragraphs 11.2.3.1
44
-------
and 11.2.3.2.
11.4.1.4 If the venturi is going to measure a range of flows,
perform the calibrations for at least three flowrates
—low, medium, and high.
11.4.1.5 Repeat the calibration process several times if possi-
ble at each flowrate and use the average measured
flowrate to determine C from equation [1] using the
manometer measurement and the measured Q.
11.4.1.6 There is no single calibration method applicable to
. all situations. The choice may depend not only on
technical factors described in the following sections
but also on such factors as availability of skilled
manpower, funds, in-plant laboratory capability, etc.
The purpose of the following sections is to point out
some advantages and disadvantages of several common
calibration methods and conditions for their use.
The major calibration methods are:
- Volumetric
- Comparison with reference meter
- Dilution
- Salt velocity
- Velocity-area traverse
11.4.2 Volumetric calibration.
11.4.2.1 Volumetric calibration can be used for all liquid and
sludge flows. Its feasibility depends upon the avail-
ability of suitable tank space and connecting piping.
The potential accuracy is high, provided that:
- The tank is regular in configuration so that its
lateral dimensions can be measured within acceptable
limits of accuracy.
- The tank is large enough to permit a test run of
sufficient length for the effect of timing errors
at the start and finish to be kept within accept-
able limits.
- The change in liquid level during the run is large
enough so that the starting and finishing depths
(probably obtained by the "on-the-run" method) can
be measured within acceptable relative error limits.
- The flowrate remains relatively constant during the
run.
11.4.2.2 The volumetric method can be used to calibrate venturi-
type meters in intermittently operating pumping sta-
tions by using the fall of sewage level in the wet wall
during a pumping cycle. It is necessary to correct
for the inflow occurring during the pumping-out pro-
cess. However, the standard open-tap venturi tube is
45
-------
not suitable for intermittent flows because of the
likelihood of trapping air in the secondary lines.
11.4.2.3 The volumetric method is often the only basic calibra-
tion method practical for raw primary sludge. See
section 10.3.2.
11.4.2.4 Estimate the uncertainty of this method as a combina-
tion of the estimated uncertainties of the measure-
ments of the lateral area, the depth change and the
time.
11.4.3 Comparison with a reference meter.
11.4.3.1 In this context a reference meter is a flowrate mea-
surement device whose performance can be referenced
to published standards or recommended practices that
are acceptable to the parties involved. Examples
include:
- Standard venturi tubes and venturi nozzles as des-
cribed in this document
- Orifice plates (1)
- Thin plate weirs (7) (footnote 5)
- Venturi flumes (7) (footnote 5)
- Parshall flumes (8) (footnote 5)
11.4.3.2 Such meters must meet all requirements of accepted
standards in fabrication, installation and use.
11.4.3.3 The performance information available or obtainable
for the reference instrument must include estimates
of uncertainty so that error estimates can be made
for the purposes given in section 11.6.3.
11.4.3.4 When a differential-pressure type of meter is used as
a reference device, measure the pressure differential
with a U-tube manometer. If necessary, a transducer
can be used under the terms of paragraph 11.2.1.3.
11.4.3.5 When a critical-flow type of open-channel meter is
used as a reference device, measure its head with a
point gage or similar direct depth measuring instru-
ment after a careful determination of the zero-depth
condition. If it is necessary to use a float gage or
other commercial instrument for this purpose, infor-
mation must be available on its measurement errors so
that uncertainties can be estimated.
11.4.3.6 If it is impossible to meet the requirements of para-
graph 11.4.3.1, it may be acceptable to use as a ref-
erence meter a device for which there are no published
standards provided that:
46
-------
- The device has had a recent calibration and/or its
current accuracy can be otherwise demonstrated
satisfactorily;
- The device is used under the same conditions for
which it was calibrated or for which its accuracy
was otherwise demonstrated;
- Sufficient information is available to permit the
involved parties to agree on its uncertainty.
Examples of such devices are:
- Propeller meters (footnote 6)
- Segmental orifices
- Electromagnetic flowmeters
11.4.4 Dilution method.
11.4.4.1 In the dilution method the flowrate is deduced from
the dilution of measurable properties, e.g., color,
conductivity or fluorescence of tracer chemicals add-
ed to the flow in known amounts. The calibration can
be done by either the constant-rate injection method,
or the slug injection method. The constant-rate
method is recommended here because it appears more
practical for in-plant use and because documentation
on it is available in the form of published stan-
dards, e.g. (1, 10).
11.4.4.2 In the constant-rate injection method, a tracer solu-
tion of accurately known concentration is injected
upstream at a rate which is constant and accurately
measurable. At a downstream distance long enough for
complete mixing, the flow is sampled and the concen-
tration determined after a steady state or concentra-
tion "plateau" is attained. The flowrate, Q, is then
determined from
Q = q(C;L - c2)/(c2 - CQ) [4]
where: q is the rate at which the sample of concen-
tration c. is injected; c~ is the measured "plateau"
concentration downstream; and c (which may be close
to zero) is the background concentration of the tracer
chemical existing in the flow.
11.4.4.3 This method requires accurate measurement of q and of
all concentrations. Skilled personnel and specialized
equipment are needed. The potential accuracy is high
under optimum conditions; see references (9, 10) for
methods of estimating errors.
11.4.4.4 The tracer chemical must be conservative, since losses
by absorption to the solids component will be reflect-
ed as an apparent reduction in c . The fluorescent
47
-------
dye Rhodamine WT has been successfully used in sewage
without losses. There are no reports of its applica-
tion to sludge flows and this use is not recommended.
11.4.4.5 This method requires fully turbulent flow.
11.4.5 Salt-velocity method.
11.4.5.1 In the salt-velocity method, brine is injected sudden-
ly at an upstream station in such a way that it be-
comes well distributed across the section very rapid-
ly. The time of passage of the salt pulse between
two downstream stations is measured by means of elec-
trodes which detect the increased conductivity assoc-
iated with the passage of the brine. The flowrate
then can be determined provided the volume of the
pipe between the electrodes is accurately known.
This method has a potential for 1 percent accuracy
under optimum conditions. The accuracy actually ob-
tained depends upon the tranverse mixing and coherence
of the injected brine slug, upon the accuracy of deter-
mination of the centers of gravity of the tracer con-
ductivity and the time separating them, as well as
upon the accuracy of the aforementioned volume deter-
mination.
11.4.5.2 This method requires a length of (preferably straight)
pipe upstream of the first electrode sufficient to in-
sure complete lateral mixing of the salt. This length
can be as short as 4D when the injection is accom-
plished internally in the standard manner (2). How-
ever, for injections from the pipe periphery a sub-
stantially greater length is required. The distance
between electrodes must be a minimum of 4D.
11.4.5.3 Brine injection must be sudden, with the injection
interval of the order of 1 second with no leakage
thereafter.
11.4.5.4 The electrodes must provide equal increments of con-
ductivity for equal segments of cross-sectional area.
Because the electrodes are intrusive the method is
not suitable for flows in which there are fouling
solids.
11.4.5.5 This method requires fully turbulent flow.
11.4.5.6 This method requires the liquid being measured to have
a significantly smaller electrical conductivity than
the brine.
48
-------
11.4.5.7 See reference (2) for further details of this method.
11.4.6 Velocity-area method.
11.4.6.1 The velocity-area method involves the measurement of
a number of velocities, each representative of the
velocity within an incremental area, and summing the
resulting velocity-area products over the flow cross-
section.
11.4.6.2 Because most velocity-measuring instruments are in-
trusive and because a substantial amount of immersion
time is needed, the velocity-area method cannot be
used in flows containing fouling solids, nor can it
be used in conduits where the blockage effect is ex-
cessive.
11.4.6.3 In view of paragraph 11.4.6.2 there are likely to be
few in-plant situations where this method can be used.
However, for those cases where the method is suitable:
- Use only velocity-measuring instruments which have
been calibrated recently and whose performance in
regard to uncertainties can be documented.
- Consult reference (11) or (12) for distribution of
velocity-sampling points in the cross-section, des-
criptions of apparatus, and for other conditions on
the measurements.
11.5. Approximate Methods.
11.5.1 It may be useful on occasion to have a relatively quick and in-
expensive way of knowing whether or not a flowmeter is even
approximately correct. Measurements in the + 10 percent un-
certainty range are adequate for this purpose. Some examples
are given in this section; all involve measurement of a head
difference with a manometer or its equivalent.
11.5.2 Elbow meters.
11.5.2.1 The accelerations associated with flow around a curve
of radius r cause a pressure difference in the radial
direction which can be used to deduce a flowrate in a
full conduit. This method, which has been investiga-
ted mainly for 90-degree pipe elbows, requires mea-
surement of the head difference between two diametric-
ally opposite pressure taps drilled in the plane of
the bend and half way between the end flanges of the '
elbow, i.e., at the 45-degree position.
11.5.2.2 The flowrate can be estimated from the following
analytically determined expression.
49
-------
Q = (r/2D)1/2(nD2/4)(2gAh)1/2 [5]
Here, r is the radius of curvature of the elbow
centerline, D is the elbow (and pipe) diameter
and Ah is the measured head difference as pre-
viously defined.
11.5.2.3 The elbow performance is more sensitive to the radius
of curvature of the inside bend than to that of the
outside bend. Therefore it is desirable, when prac-
ticable, to determine r by measuring the inner bend
curvature and adding half the diameter rather than to
use a nominal value of r.
11.5.2.4 The elbow should be preceded by about 10 diameters of
straight upstream pipe. There is insufficient infor-
mation with which to evaluate such effects as pipe
roughness and Reynolds number, except to note that
with decreasing Reynolds number the flow is less than
that predicted by equation [5]. Therefore added
caution must be exercised in its application to
sludge flows.
11.5.2.5 Examination of published experimental results suggests
that equation [5] cannot be depended upon for accu-
racies better than roughly + 10 percent (see footnote
7 for possible exception).
11.5.2.6 However, it should be noted that an elbow meter that
is carefully fabricated and installed and properly
calibrated can be as effective a flowmeter as other
types of pressure differential devices. The fore-
going paragraphs of section 11.5.2 pertain to un-
calibrated elbows only.
11.5.3 Valves. Butterfly valves are sometimes furnished with flow-
rate vs. angle-of-opening data which can be used for approxi-
mate checks on flowmeters. Such measurements would of course
be affected to an unknown extent by upstream conditions such
as presence of fittings and roughness and by Reynolds number.
11.5.4 Measurements for monitoring.
11.5.4.1 Differential-head measurements can be used, without
regard to their absolute accuracy, to monitor changes
in the system and to observe whether the flowmeter
is responding in a consistent manner.
11.5.4.2 In addition to the measurements heretofore cited in
section 11.5, the head difference between the suction
and discharge sides of a centrifugal pump can be used
50
-------
for monitoring, provided that (for a given speed of
rotation) its head decreases continuously with in-
creasing discharge over the range of interest. In
principle this method can be used directly for approxi-
mate flow measurements provided that the»head differ-
ence (which includes the difference in V /2g) is deter-
mined in the same way it was done to establish the
head-discharge curves. Generally head-discharge
curves are presented for cold water and would not
necessarily be valid for sludges of substantially
higher viscosity.
11.5.4.3 Monitoring the head loss in an upstream (or similar)
section of pipe from the time of its installation can
provide a basis for making pipe-roughness adjustments
to C.
11.6 Estimating Errors.
11.6.1 Error estimates provide an assessment of the uncertainty of a
measurement. One method of estimating the uncertainty of a
flowrate determination based on equation [1] is to combine in-
dividual uncertainties as follows for a venturi tube.
77 7 ? 7 71/7
6Q/Q = [(6C/C) + M (6D/D) + N (6d/d) + (1/4)(6Ah/Ah) ]Ll [6]
with M = 264/(l - 34)
N = 2/(l - g4)
6 = uncertainty
The second and third terms inside the brackets are usually
small, since the diameter uncertainties are limited, e.g., by
paragraphs 4.2.3.2 and 4.2.5.2.
11.6.2 Application to secondary-system.
11.6.2.1 In order to compare the indicated flowrate with the
flowrate computed from manometer measurements (para-
graph 11.2.3.4) it is first necessary to estimate
the uncertainty of the latter, or "reference," flow-
rate.
11.6.2.2 For this purpose often only the first and last brack-
eted terms in equation [6] need be considered.
The first term is either the standard uncertainty
(paragraph 4.5.2.1 or 6.5.3) or the uncertainty fur-
nished by the manufacturer. The last bracketed term
allows for an estimate of the manometer reading error,
typically 1 percent and preferably not more than 2
percent at low flowrates. Any uncertainty in the
density of the flowing liquid compared with the mano-
meter liquid should be included here.
51
-------
11.6.2.3 The tolerance estimated from equation [6] represents
a band (its width may increase at low flows) of un-
certainty about the flowrate determined in paragraph
11.2.3.3.
11.6.2.4 Compare the flowrate recorded by the on-line mea-
suring system at the time of the manometer readings
with that determined in paragraph 11.2.3.3. The re-
corded flowrate should not fall outside of the toler-
ance band by more than the errors allowed in trans-
mitting, receiving and recording the differential
pressure signal in paragraph 8.1.1. For example, if
the uncertainty in C and in Ah are both 1 percent at
a particular flowrate, the resulting uncertainty in
Q is 1.1 percent. If the specifications (section
8.1) require a 2 percent accuracy for that flowrate,
the allowable difference between recorded and refer-
ence flowrate could be slightly over 3 percent.
11.6.2.4 The above method of comparison is an illustration
only. Other methods agreed on by the involved parties
are acceptable. The important point is that the un-
certainty of the reference measurement should be
taken into account when evaluating another measure-
ment.
11.6.3 Other error estimates.
11.6.3.1 To estimate the uncertainty of a flow measurement,
it is necessary to combine the error estimated from
the secondary-system calibration with an estimate of
the uncertainty in C. Such uncertainties have been
cited earlier in this document for a limited number
of specific conditions for which some information is
available.
11.6.3.2 When it is necessary to use one of the calibration
methods of section 11.4 to determine C, an uncertainty
estimate satisfactory to all parties must be derived
based on the quality of the tests. Guidelines for
estimating uncertainty are available in the cited
references.
52
-------
12. REFERENCES
(1) International Standards Organization, "Measurement of Fluid Flow by
Means of Orifice Plates, Nozzles and Venturi Tubes Inserted in
Circular Cross-Section Conduits Running Full," ISO/DIS 5167, 1976,
draft revision of R781.
(2) American Society of Mechanical Engineers, "Fluid Meters - Their
Theory and Application," 6th ed., 1971, 345 E. 47 St. New York,
NY 10017.
(3) American Society for Testing and Materials, "Standard Method of Flow
Measurement of Water by the Venturi Meter Tube," ASTM D2458-69.
(4) Water Pollution Control Federation, "Wastewater Treatment Plant
Design," WPCF Manual of Practice No. 8, 1977.
(5) Hydraulic Institute, "Standards for Centrifugal, Rotary and
Reciprocating Pumps," 12th edition.
(6) International Standards Organization, "Fluid Flow in Closed Conduits—
Connections for Pressure Signal Transmissions Between Primary and
Secondary Elements," ISO 2186 - 1973.
(7) British Standards Institution, Standard No. 2680-4A, "Methods of
Measurement of Liquid Flow in Open Channels: Part 4A, Thin Plate
Weirs and Venturi Flumes," 1965.
(8) American Society for Testing and Materials, "Standard Method for
Open Channel Flow Measurement of Industrial Water and Industrial
Waste Water by the Parshall Flume," ASTM D1941-67.
(9) International Standards Organization, "Measurement of Water Flow in
Closed Conduits—Tracer Methods, Part I; General," ISO No. 2975/1,
1974.
(10) International Standards Organization, "Measurement of Water Flow
in Closed Conduits—Tracer Methods, Part II; Constant Rate
Injection Method Using Non-Radioactive Tracers," ISO DIS 2975/11.
(11) British Standards Institution, Standard No. BS1042-2A, "Methods for
The Measurement of Fluid Flow in Pipes, Part 2, Pitot Tubes; Part 2A,
Class A Accuracy," 1973.
(12) International Electrotechnical Commission, "International Code for the
Field Acceptance Tests of Hydraulic Turbines," Publication 41, 1963.
53
-------
APPENDIX A. FOOTNOTES
(1) The specifications of sections 4 and 6 may appear unrealistically rigid
for sewage plant application. However, the quality of the final mea-
surement depends upon the performance of the complete measurement system,
i.e., primary plus secondary elements. Because substantial errors can
be introduced in the sensing, transmission and recording of the pressure
differential, it is to the user's advantage not only to give the second-
ary unit extensive attention, but also to minimize errors in the primary,
thereby keeping the total error within reasonable limits.
(2) Conflicting experiences are reported here. Keefer ("The Effect of
Sewage on Cast Iron Venturi Meters," Eng. News-Record, 112, Jan. 11,
1934, p. 46) reported that domestic sewage did not adversely affect
cast iron Venturis after 12 years of use. These 42 x 21 meters had been
coated originally with a coal tar varnish and the throats were bronze
lined. Also, the meters always had been kept filled so there was no
alternate wetting and drying. On the other hand, Richardson ("Venturi
Meters for Sewage," Eng. News-Record, 112, Apr. 12, 1934, p. 482) re-
ported gradual but serious accumulation of deposits, although the type
of deposit was not specified. Crossley ("Has Your Treatment Works Too
Many Instruments?," Progress in Water Technol., (±, 1974, Pergamon
Press) cites a 36 x 27 inch mixed liquor venturi tube which was examined
after 30 years of service and, apart from a few barnacles on the inlet
cone, the throat was smooth although slightly pitted.
(3) Again conflicting experiences are reported. Scott ("Magnetic Flowmeter
—A New Sludge Meter," Prog, in Water Technol., jj, 1974, Pergamon Press)
lists several plastics used in magnetic flowmeters to prevent grease
build-up. However, a discusser of this paper claimed that only glass
had been proven effective.
(4) Experiments reported by Halmi ("Practical Guide to the Evaluation of
the Metering Performance of Differential Producers," ASME, Jour. Fluids
Eng., March 1973) showed an effect of tap orientation up to about 8D
downstream of a short radius elbow for a B = 0.75 meter. Although this
meter was a proprietary one with convergence shape different from that
of the classical tube, it appears that a conservative approach should
allow for additional uncertainty.
(5) Open channel measuring devices are included in this list because it may
be possible, for example, to calibrate an influent venturi tube against
a good effluent weir system (provided that change of storage within the
plant can be avoided) or against a flume elsewhere in the plant.
54
-------
(6) There is an American Water Works Association standard, AWWA C704-70 for
propeller-type cold-water meters for main line applications.
(7) Replogle, et al., ("Evaluation of Pipe Elbows as Flow Meters," Proc.
Amer. Soc. Civ. Eng. 92, IRS, 1966) indicated that uncalibrated elbow
meters could be accurate to within + 3 percent if empirical coefficients
were used to modify equation [5] as shown below. The empirical coeffi-
cients listed here were obtained from and should be used only for com-
mercial cast flanged elbows.
- Multiply the right hand side of equation [5] by the following
coefficients for specific elbow sizes.
12-inch . coeff. = 1.048
10-inch coeff. = 1.021
6-inch coeff. = 0.983
3-inch (long) coeff. = 1.014
3-inch (short) coeff. = 0.994
- Also, instead of applying the exponent 1/2 to the measured head,
use (for the above range of diameters)
Exponent = 0.489 + 0.038D
- Extrapolation of these results to larger sizes is not recommended.
- Access to the inside of the elbow is necessary to achieve this
accuracy. Dimensions must be carefully measured for use in
equation [5]. Use the inner bend surface radius (obtained from
a plaster cast if necessary) as a base for determining "r" and
check to see that the measured radius is constant.
- Maintain at least 20 straight pipe diameters upstream for this
accuracy.
55
-------
APPENDIX B. EXPANSIBILITY FACTORS, e
Adiabatic ,d,4
Factor, k V
1.2 0
0.1
0.2
0.3
0.4
0.41
1.3 0
0.1
0.2
0.3
0.4
0.41
1.4 0
0.1
(air) 0.2
0.3
0.4
0.41
1.66 0
0.1
0.2
0.3
0.4
0.41
Ratio of
0
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
.98
987
986
983
980
977
976
988
987
985
982
979
978
989
988
986
983
980
980
991
990
988
986
983
983
0.96
0.975
0.971
0.967
0.961
0.954
0.953
0.977
0.973
0.969
0.964
0.958
0.957
0.978
0.975
0.972
0.967
0.960
0.960
0.982
0.979
0.976
0.972
0.966
0.966
absolute
0.94
0.962
0.957
0.950
0.942
0.932
0.931
0.965
0.960
0.954
0.947
0.937
0.936
0.967
0.963
0.957
0.950
0.941
0.940
0.972
0.969
0.964
0.958
0.950
0.949
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
pressure, throat to inlet
92
949
942
934
924
911
909
953
947
939
929
917
915
956
950
943
934
922
921
963
958
952
944
934
932
0.90
0.936
0.928
0.918
0.905
0.890
0.888
0.941
0.933
0.924
0.912
0.897
0.895
0.945
0.938
0.929
0.918
0.904
0.902
0.953
0.947
0.939
0.930
0.918
0.916
0.85
0.903
0.891
0.877
0.860
0.839
0.837
0.910
0.899
0.886
0.870
0.850
0.847
0.916
0.906
0.893
0.878
0.859
0.857
0.929
0.920
0.909
0.895
0.878
0.876
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
80
869
854
837
816
791
788
878
865
948
828
804
801
886
873
858
839
815
813
903
892
878
861
840
837
0.75
0.834
0.817
0.797
0.773
0.745
0.742
0.846
0.829
0.810
0.788
0.760
0.757
0.856
0.840
0.822
0.800
0.773
0.770
0.877
0.863
0.846
0.827
0.802
0.799
56
------- |