Disclosure to Promote the Right To Information

Whereas the Parliament of India has set out to provide a practical regime of right to
information for citizens to secure access to information under the control of public authorities,
in order to promote transparency and accountability in the working of every public authority,
and whereas the attached publication of the Bureau of Indian Standards is of particular interest
to the public, particularly disadvantaged communities and those engaged in the pursuit of
education and knowledge, the attached public safety standard is made available to promote the
timely dissemination of this information in an accurate manner to the public.

1 +, 1 + 01 ' 5
Mazdoor Kisan Shakti Sangathan Jawaharlal Nehru
The Right to Information, The Right to Live Step Out From the Old to the New

IS 14371 (1996): Measurement of liquid flow in open

channels - Parshall and SANIIRI flumes [WRD 1: Hydrometry]

! $ ' +-
Satyanarayan Gangaram Pitroda
Invent a New India Using Knowledge

! > 0 B


BharthariNtiatakam
Knowledge is such a treasure which cannot be stolen
 IS 14371 :I996
( Reaffirmed 2003 )

Indian Standard
MEASUREMENT OF
LIQUID FLOW IN OPEN CHANNELS -
PARSHALL AND SANIIRI FLUMES

ICS 17.120.20

0 BIS 1996

BUREAU OF INDIAN STANDARDS

MANAK BHAVAN, 9 BAHADUR SHAH ZAFAR MARG
NEW DELHI 110002

August 1996 Price Group 7

Fluid Flow Measurement Sectional Committee. RVD 1

FOREWORD

This Indian Standard was adopted by the Bureau of Indian Standards, after the draft finalized by the Fluid
Flow Measurement Sectional Committee had been approved by the River Valley Division Council.

Various methods are adopted for measurement of flow of water in open channels like velocity area method,
slope area method, etc, depending upon the channel, flow conditions, measuring equipment available, etc.
Parshall and SANIIRI flumes are used for flow measurement in open channels, particularly in irrigation
canals, under steady or slowly varying flow conditions which may be either free-flow or submerged-flow.
This Indian Standard on liquid flow measurement by Parshall and SANIIRI flumes has been prepared based
on IS0 9826 : 1982 Measurement of liquid flow in open channels - Parshall and SANIIRI flumes and,
therefore, is technically equivalent to the IS0 Standard.

For the purpose of deciding whether a particular requirement of this standard is complied with, the final
value, observed or calculated, expressing the result of a test or analysis, shall be rounded off in accordance
with IS 2 : 1960 Rules for rounding off numerical values (revisetl). The number of significant places retained
in the rounded off value should be the same as that of the specified value in this standard.
 IS 14371 : 1996

Indian Standard
MEASUREMENT OF
LIQUID FLOW IN OPEN .CHANNELS -
PARSHALL AND SANIIRI FLUMES
1 SCOPE 4.2 Parshall flumes have a rectangular cross-section
and a wide range of throat widths varying from very
This standard specifies methods of liquid flow
small (0.025 4 m) to large (15 m and greater).
measurement in open channels (particularly in
irrigation canals) under steady or slowly varying flow
Medium-sized Parshall flumes, with throat widths
conditions, using Pal-shall and SANIIRI flumes.
between about 0.15 m and about 2.5 m which are
These flumes are designed to operate under both free- suitable for measuring discharges in the range from
flow and submcrgencc conditions. 0.001 5 ml/s to 4.0 ml/s, are those most commonly
used for flow measurements; they are thus
2 REFERENCE recommended in this standard as standard structures.
The Indian Standard IS 1191: 197 1 Glossary of terms Large Parshall flumes with throat widths hetween
and symbols used in connection with the measurement about 3 m and about 15 m, the design of which varies
of liquid flow with a free surface (first revision) depending on the size of the flume, are suitable for
contains provisions which, through reference in this measuring discharges in the range from 0.75 IN/S
text, constitute provisions of this standard. At the to 93 ml/s.
time of publication, the edition indicated was valid.
One of the most desirable features of the Parshall
All standards are subject to revision, and parties to
flume is that it operates satisfactorily at high
agreements based on this standard are encouraged
submergence ratios with low head loss, this makes
to invcstigatc the possibility of applying the most
it especially suitable for flow measurements in
recent edition of the standard.
channels having small hcd slopes. Howcvcr, the
complicated design of this flume offsets somewhat
3 DEFINITIONS AND SYMBOLS
the advantages that it oflers.
3.0 For the purpose of the standard, the definitions 4.3 SANIIRI flumes are rectangular in cross-section,
and symbols given in IS 1191 :I971 and the following level-floored and have an exit cross-sectional width
shall apply. between 0.3 m and 1.O 111. They are suitable for
measuring discharges in the range from 0.03 m/s
3.1 Parshall Flume and 2.0 ml/s.
Measuring llume having a converging entrance section SANIIRI flumes are simple in design and construction,
with a level lloor, a short throat section with a floor with the exception that a small fall at the downstream
inclined downwards at a gradient of 3:8, and a end of the floor of the llumc has to hc provided.
diverging exit section with a floor inclined upwards
at a gradient 01 1:6. 5 INSTALLATION

3.2 SANIIRI Flume 5.1 Selection of Site

Measul-ing flume with a converging entrance section 5.1.1 The llumc shall bc located in a straight section
having a level floor with a vertical drop at its of the channel, avoiding local obstructions and
downstream end and perpendicular walls to join it roughness or unevenness of the bed.
IO the downstream channel. 5.1.2. A preliminary study shall be made of the
physical and hydraulic features of the proposed site,
4 SELECTION OF FLUME TYPE to check that it conforms (or may he constructed or
modified to conform) with the requirements necessary
4.1 The choice as to whether a Pal-shall or a SANIIRI for discharge measurement by the flume. Particular
Ilumes should be used depends on several factors attention shall be paid to the following Features in
such as the range of discharge to be measured, the selecting the site:
head available, the modular limit and the maximum
submergence ratio, the channel or canal characteristic, a) The adequacy of the length of channel of
the amount of head loss which can be allowedthrough regular cross-section and slope available;
the flume, the possibility of deepening the bed and
providing a drop therein, the accuracy of measurement
b) The uniformity of the existing velocity
distribution;
required, whether or not the flow carries sediment,
the operating condition that necessiate the use of c) The conditions downstream (including
either stationary or portable flumes, and economic influences such as tides, control structures,
considerations. etc);

1
IS 14371 : 1996
The impermeability of the ground on which 5.2.2.2 The flow conditions and the symmetry 01
the structure is to be founded and the necessity the velocity distribution in the approach channel shall
for pilling, grouting or other means of be checked by inspection and measurement using for
controlling seepage; example, current meters, floats, velocity rods or dye.

The stability of the banks or side slopes of NOTE - A complete assessment of the velocity distribution
the channel, and the necessity for trimming may be made by using a current meter.
and/or revetment;
5.2.3 Flume Structure
The necessity for flood banks, to confine the
maximum discharge to the channel and the 5.2.3.1 The structure shall be rigid and watertight
backwater caused by the installation of the and capable of withstanding flood-flow conditions
ll u me; without damage from out-flanking or from downstream
erosion. The axis shall be in line with the direction
The effect of wind on the flow over the flume, of flow in the upstream channel, and the geometry
especially when the flume is wide and the shall conform with the dimensions given in 8 or 9
head is small and when the prevailing wind as appropriate.
is in a direction transverse to the direction
of flow; 5.2.3.2 The surfaces of the flume, particularly those
of the entrance section and throat, shall be smooth.
Aquatic weed growth; and The flume may be constructed of concrete with a
smooth cement finish or may be surfaced with a
Sediment transported by the flow. smooth non-corrodible material. In laboratory
installations, the finish shall be equivalent to that of
5.1.3. If the site does not possess the characteristics rolled sheet metal or planed, sanded and painted
necessary for satisfactory discharge measurements, timber. The surface finish is of particular importance
it shall not be used unless suitable improvements are within the prismatic part of the throat, but the
practicable. requirements may be relaxed beyond a distance along
the profile 0:5hlllilI upstream and downstream of the
5.2 Installation Conditions throat proper.

5.2.1 GLVWUI~Rquirernmts 5.2.3.3. To minimize uncertainly in the discharge

measurement the following tolerances shall be satisfied
The complete measuring installation consists of an in construction:
approach channel, a flume structure and adownstream
channel. The condition of each of these three 4 On the bottom width b of the throat - 0.2
components affects the overall accuracy of the percent of h with an absolute maximum 01
measurements. In addition, features such as the surface 0.01 m,
finish of the Ilumc, the cross-sectional shape of the
b) On point deviations from a plane surlacc in
channel and the chunncl roughness shall be taken into the throat - 0.1 percent of 1,
consider-ation.
c) On the width between vertical surface in the
throat - 0.2 percenl of this width with a
maximum of 0.01 171,
5.2.2.1 The approach channel shall comply with the On the. average longitudinal and transverse
following rcquircmcnts: slopes of the base of the thl-oat - 0. I percent,
11 shall hc straight and uniform and have a On the slope of inclined surface in the throat
constan\ slope for a length equal to five to - 0.1 percent,
ten limes the water surlacc width at maximum
On the Icngth of the throat - I pcrccnt ol
IIUW.
1.
The bed slope shall bc such as to ensure
subcritical Iluw with a Froude numher Fr of On point deviations tiom a plant surf;lcc in
Icsx than 0.5 (or 0.7), whcr-c the entrance transition to the throat - 0. I
percent of 1.

h) On point deviations Il-om a plane surlacc iii

the exit transition 1rom the throat -- 0.3 pcrccnt
OS 1.
whcrc
i) On dcviationb iron7 a plant UI-CIIWV on othcl
Q,,,,, =
Ihe Inaximum discharge in cubic vertical WI-inclined surL1cc.s - I pctwnl 01
inclrc\ per second, 1, and
/t = the cl-ass-sectional ar-en of the channel
in square nictr-cs, and k) On deviation from a plane of the bed 01 the
lined approach channel - 0. I percent (11 1
L =
the nI;IxinIunl water depth in metres.

2
 IS 14371 : 1996
The structure shall be measured on completion of 6. MAINTENANCE - GENERAL
construction and average values of relevant dimensions REQUIREMENTS
and their standard deviations at 95 percent confidence
limits shall be computed. The average values of 6.1 Maintenance of the measuring structure and the
dimensions shall be used for computation of the approach channel is important to secure accurate
discharge and their standard deviations shall be used measurements.
to obtain the overall uncertainty in the determination It is essential that, as far as practicable, the approach
of discharge. channel to flumes be kept clean and free from slit
and vegetation for the minimum distance specified
5.2.4. Downstream of the Structure in 5.2.2.1.
6.2 The float well, the connecting pipe and the inlet
The flow conditions downstream of the structure are from the approach channel shall be kept clean and
important in that they control the tail-water level free from deposits. The throat and the curved entry
which may influence the operation of the flume. The to a flume shall be kept clean and free from algal
flume shall be so designed that it cannot become growths.
drowned under normal operating conditions except 7 MEASUREMENT OF HEAD(S)
for a limited period of time, for example, during
floods. The construction of a flume ,in a river or Requirements on head measurements for particular
stream may alter the flow conditions upstream and types of flume are dealt with in 8 and 9.
downstream of the structure. This may result in the
8 PARSHALL FLUMES
accumulation of river bed material further downstream
which, in time, may cause the normal water level 8.1 Description
to rise sufficiently to drown the flume, particularly
at low rates of flow. Any such accumulation of material Parshall flumes have a rectangular cross-section and
shall be removed before it becomes excessive. consist of a converging entrance section, a throat and
a diverging exit section (XC Fig. I).

V 1A Plan View

DIRECTION
OF FLOW

1B Sectional View

FIG. 1 PARSHALLFLUME
IS 14371 : 1996
The floor of the entrance section shall be truly level for the measurement of the head h, (co1 4 and 5),
both longitudinally and laterally. The side walls shall the axial length of the exit section I, (co1 12), the
be vertical and disposed at a constant angle of height hp2 (co1 13), the slope of the throat floor (3:8)
convergence of 11 19 or shall have a 15 contraction and the reverse slope of the exit section floor (1:6).
in plan with respect to the flume axis.
The other dimensions of these flumes (No. 2 to 13)
The side walls of the throat shall be parallel in plan. are calculated using the followed equations:
The floor shall be inclined downwards with a gradient
of 3:8; this applies of flumes of all sizes. The line a> Width, in metres, of the entrance cross-section
of intersection of the entrance section floor with the of the flume
throat floor is known as the crest of the flume. The b,= 1.2b + 0.48 . . . (1)
elevation of the crest above the throat invert is referred
to as the height of the flume crest hD,. b) Axial length, in metres, of the entrance section
1, = OSb + 1.2 . (2)
The side walls of the exit section shall be vertical
and disposed at a constant angle of divergence of Cl Converging wall length, in metres
9 28 or shall have a I:6 expansion in plan with lo= 1.021, . . . (3)
respect to the flume axis. The floor shall be inclined
upwards with a reverse gradient of 1:6; this applies 4 Wall length, in metres, between the crest and
to flumes of alI sizes. the head 11,~measurement section
l;, = &c/3 . . . (4)
To ensure a smooth entry of the flow into the flume
and to prevent surface disturbance at the exit of the e) Width, in metres, of the exit cross-section
flume, the entrance and exit cross-sections shall be of the flume
connected to the natural channel banks or the artificial b, = b + 0.30 . (5)
channel side slopes by means of vertical wing walls
disposed at 4.5 to the flume axis or curved in plan f) Side wall height, in metres, in entrance
with a radius R > 211,,r,x(see Fig. 1). For smaller sizes section
of tlumes with throit widths less than 0.5 m, the + (0.15 to 0.20)
hc = h:, 1,1:1x (6)
wing walls may be placed at right angles to the flume
axis. It is recommended that an additional allowance of
up to 1 m be provided in the height of the side walls
Parshall Ilun~es may be constructed of wood, stone, to avoid the risk of overtopping when flows through
concrete, reinforced concrete, or any other material the flume are in excess of the maximum design
depending on the prevailing conditions. Small Parshall discharge.
Ilumes may bc built of sheet metal and used as portable
structures. Flumes made of reinforced concrete may The lengths I, and I, of the wing walls vary with
be prefabricated for assembly in the field. the width of the natural or artificial channel (see
Fig. I) To ensure proper connection to the channel
8.2 Dimensions banks or the artificial channel side slopes, the wing
walls shall extend a distance of at least 0.4m to OSm
Parshall flumes have a specific feature in that the into the channel banks.
flumes are not geometrically similar models of each
other. The throat length, crest height and length of 8.2.2 Large Pat-shall Flmnes
the exit section remain constant for a series of flumes
while other dimensions vary as a function of the In contrast with standard Parshall flumes, the
throat width; these other dimensions may be dimensions of large Parshall flumes shall bc
determined analytically. determined independently for each particular design
as a function of the throat width. For the determination
It is thus essential to use calibrated flumes constructed of the leading dimensions of large Parshall flumes,
in accordance with the dimensions specified in Tables the values specified in Table 2 shall apply. These
1 and 2 for standard and large Parshall flumes values shall be neither varied nor rounded off without
respectively. additional calibration of the flume.

8.2.1 Stmrdud Parshall Flumes
The size of a particular standard Parshall flume is
denoted by its throat width b (see Table 1, co1 2).
For the series of standard Parshall flumes having
throat widths 19from 0.250 m to 2.400 m (see Table
1, col 1. No. 2 to 13) the leading dimensions are
identical, that is the throat length 1 (co1 3), the height
01 the crest /zr, (co1 6), the coordinates X and Y of
the throat cross-section at the stilling well pipe used
Table 2 gives the leading dimensions of large Parshall
flumes with throat width between 3.05 m and 15.24 m,
for measuring discharges in the range from 0.16
m3/s to 93 ml/s. It may be seen in Table 2 that 1,
X, Y h,,, and /I~, remain constant for a series of
flumes.
In addition, the slopes 3:8 and i:6 of the throat floor
and the exit section floor respectively, and the angles
of convergence (11 18) and divergence (9 28) of
the side walls of the entrance and exit sections also

4
 IS 14371 : 1996
Table 1 Dimensions for Standard Parshall Flumes
( Clauses 8.2 and 8.2.1 )
All dimensions in metres.

Parshall Throat Entrance Section Exit Section Side Wall

Flume
A
No. - I F-----l Height
b I X y hp, b, 4 ie Ia b, 1 h hc
12

1 2 3 4 5 6 7 8 9 10 11 12 13 14

I 0. is2 0.305 0.05 0.075 0.115 0.40 0.610 0.622 0.415 0.39 0.6 I 0.012 0.60
2 0.250 0.600 0.0s 0.075 0.230 0.78 1.325 I .352 0.900 0.55 0.92 0.072 0.80
3 0.300 0.600 0.05 0.075 0.230 0.84 1.350 1.377 0.920 0.60 0.92 0.072 0.9s
4 0.450 0.600 0.05 0.07s 0.230 1.02 I.425 1.454 0.967 0.75 0.92 0.072 0.9.5
5 0.600 0.600 0.05 0.075 0.230 I .20 1.500 1.530 1.020 0.90 0.92 0.072 0.95
6 0.750 0.600 0.0s 0.075 0.230 1.38 1.575 I.607 1.074 I .os 0.92 0.072 0.95
7 0.900 0.600 0.05 0.075 0.230 1.56 I.650 I.683 I.121 I.20 0.92 0.072 0.95
8 I .ooo 0.600 0.05 0.075 0.230 I .68 I .700 I .I34 I.161 1.30 0.92 0.072 1.oo
0 1.200 0.600 0.05 0.075 0.230 I .92 1.800 1.836 1.227 I.50 0.92 0.072 I .oo
IO 1.soo 0.600 0.05 0.075 0.230 2.28 1.950 1.989 1.329 1.80 0.92 0.072 1.00
II 1.800 0.600 0.05 0.075 0.230 2.64 2.100 2.142 1.427 2.10 0.92 0.072 1.oo
12 2.100 0.600 0.05 0.075 0.230 3.00 2.250 2.295 I .534 2.40 0.92 0.072 I .oo
13 2.400 0.600 0.05 0.075 0.230 3.36 2.400 2.448 I .632 2.70 0.92 0.072 I .oo

Table 2 Dimensions for Large Parshall Flumes

( Clauses 8.2 and 8.2.2)
All dimensions in metres.

Parshall Throat Entrance Section Exit Section Side Wall

Flume
No. ,A A /--+--I Height

b I X y $1 b, 1 A b, 4 $2 4

I 2 3 4 5 6 7 8 9 10 II I2 I3

14 3.0s 0.9 I 0.305 0.23 0.343 4.76 4.27 I .83 3.66 1.83 0. IS2 I .22
15 3.66 0.9 1 0.305 0.23 0.343 5.61 4.88 2.03 4.47 2.44 0.1.52 I .52
I6 4.57 1.22 0.305 0.23 0.457 7.62 7.62 2.34 5.59 3.05 0.203 I .83
17 6.10 I .x3 0.305 0.23 0.686 9.14 7.62 2.84 7.32 3.66 0.305 2.13
IX 7.62 I.83 0.305 0.23 0.686 IO.67 7.62 3.35 8.94 3.96 0.305 2.13
19 9.14 I .83 0.305 0.23 0.686 12.31 7.93 3.86 10.57 4:27 0.30s 2.13
20 12.19 1.83 0.305 0.23 0.686 15.48 5.23 4.88 13.82 4.88 0.305 2.13
21 15.24 1.83 0.305 0.23 0.686 18.53 8 23 5.89 17.27 6.10 0.305 2.13

remain constant for all Parshall flumes. The only section (upstream head, hi,) and throat section
dimension that may be determined analytically is the (downstream head, hJ. Whether one or both heads
wall length between the crest and the entrance cross- have to be measured depends on the flow conditions
section of the stilling well pipe used for the in the flume.
measurement of ha.
For free-flow conditions, only the head h needs to
The length is given, in metres, by the, equation
be measured. The section for measuremgnt of the
l;, = b + 0.813 . . . (7) head hS shall be located at distance ia measured along
the obhque wall upstream from the crest of the flume.
3
la may be calculated using formula (4) (see 8.2.1)
It is recommended that the throat width b be equal and formula (7) (see 8.2.2). The recommended range
to from one third to one-half times the bottom width of head ha is specified in Tables 3 and 4.
bL of tVie natural or artificial channel (see Fig. 1).
Where high accuracy is not of great importance, a
8.3 Measurement of Head and Limits of staff gauge, set vertically in the head measurcmcnt
Application section on the inside face of the converging entrance
The discharge through a Parshall flume is wall, may be used to determine the head ha. The staff
determined by measuring the heads in the entrance gauge shall be zeroed carefully with respect to the

5
IS 14371 : 1996
Table 3 Discharge Characteristics of Standard Parshall Flumes
( Clause 8.3 )

Parshall Throat Discharge Head Range Discharge Range) Modular Sub

klume No. Width Equation) Limit mergencc
b Q = Ch; ha Q O,: Ratio
m k IO,,;,< CJ
A A
I mls %firr MUX -Min MU.? (Experi- Recom-
mental) Illelldc~l

I 2 3 4 5 6 7 8 9

I 0.152 0.38 I hu W 0.03 0.4s 1.5 100 0.55 0.6

2 0.25 0.561 h .sl 0.03 0.60 3.0 250 - 0.6
3 0.30 0.679 I; IV_ 0.03 0.7s 3.5 400 0.62 0.6
4 0.45 1.038 h: .w 0.03 0.75 4.5 630 064 0.6
5 0.60 1.403 ha +w 0.05 0.75 12.5 850 0.66 0.6
6 0.75 1.772 hd s 0.06 0.75 25.0 I ID0 0.67 0.6
7 0.90 2.147 ,, I..(,*( 0.06 0.7s 30.0 I 2.50 0.68 0.6
8 I .oo 2.397 h I Sw 0.06 0.80 30.0 1500 - 0.7
9 I .20 2.904 hY . 0.06 0.80 3.5.0 2 000 0.70 0.7
IO I.50 3.668 h:, 0.06 0.80 45.0 2 500 0.72 0.7
II 1.80 4.440 ha w 0.08 0.80 80.0 3 000 0.74 0.7
12 2.10 5.222 h, .w 0.08 0.80 95.0 3 600 0.76 0.7
13 2.40 6.004 h: .fds 0.08 0.80 100.0 4 000 0.78 0.7

I) C = C,,b x 3.279

where
C,, = the coefficient of discharge, and
II = an exponent dependent on b.

2) Rounded to the nearest rationalized value

Table 4 Discharge Characteristics of Large Parshall Flumes

( Clause 8.3 )

Parshall Throat Discharge Head Range Discharge Range) Suhmer- Sulnner-

Flume No. Width Equation) gence $ence
for Free-flow Ratio Coefficient
Conditions (Correction
Factor)
b Q =C /, .h
I u h& Q 0 =s

(Recommentlctl)

I 2 3 4 5 6 7 8 9

I 3.05 7.463 C: 0.09 1.07 0.16 X.28 0.80 I.0

IS 3.66 8.859 h:. 0.09 I .37 0.19 14.68 0.80 1.2
16 4.57 10.96 h: 0.09 I .67 0.23 25.04 0.80 I..5
17 6.10 14.45 h: A3JJ9 Ed3 ~LL31 37.97 0.80 2.0
IX 7.62 17.94 11: 0.09 I .83 0.38 47. I6 0.80 2.5
19 9.14 21.44 11: 0.09 I .83 0.46 56.33 0.80 3.0
20 12.19 2X.43 h:. 0.09 I .83 0.60 74.70 0.80 4.0
21 15.24 35.41 h:. 0.09 1.83 0.75 93.04 0.80 5.0

I) C, =C,,b, where C, is the coefficient of discharge.

elevation of the flume crest, which is the elevation position for the measurement of head, near the floor
of the horizontal flume floor at the downstream end of the entrance section (see Fig.1).
of the entrance section.
If a Parshall flume is to be operated under submerged-
Where greater accuracy is required or where flow conditions, measurement of both heads h;, and
continuous recording instruments or stage-sensing h, is required. -The section for the measurement of
devices are to be used, consideration shall be given h, shall be located in the throat, a distance X from
to providing a stilling well. To connect the stilling the throat invert. Since the flow in the throat is quite
well to the flow in the flume, a length of pipe is turbulent, which causes considerab1.e fluctuation of
usctl. its inlet being located at the recommended the water surface, it is undcsirahlc to use a staff gauge

6
 IS 14371 : 1996

for the measurement of h,. Consequently, a stilling It shbuld be noted that a flume operating undcl
well is necessary. submerged-flow conditions offers the advantage of
the lowest head loss. However, submerged Ilow
Tables 1 and 2 give values of X and Y, which are conditions make discharge measurements less accurate
the coordinates of the entrance cross-section of the than those carried out under free-flow conditions. It
connecting pipe, for various flume sizes. The stilling is thus advisable to choose the dimensions of a flume
well may accommodate a staff gauge, a stage-sensing so that it operates under submerged-tlow conditions
dcvice or a continuous-recording instrument which only for a limited period of time, for example, during
shall be zeroed accurately to the elevation of the floods.
flume crest.
8.5 Determination of Discharge
The design of stilling wells and connecting pipes
shall comply with the requirements specified in 7. 8.5.1 Determination of Discharge Under Free-flaw
Conditions
Stilling wells for the measurement of head /I~ aid
/I,> shall preferably be placed adjacest to one another The discharge through a Parshall flume operating
50 that the complete installation is located in one under free-flow conditions (that is, CJ< oc) is obtained
place (either outdoors or indoors). from the following general equation:

The recommended range of heads that can be Q = Cobha . . (9)

measured by various sizes of Parshall flumes is from
where
0.03 m to 0.8 m for standard flumes and from
0.09 m to 1.83 m for large flumes (see Tables 3 and Q = the discharge, in cubic metres per second;
4 respectively).
b = the throat width, in mctres;
21.4 Free-flow and Submerged-flow Conditions ha = the head in the entrance section, in
metres;
The discharge through a Parshall tlume is considered the coefficient of discharge; and
CD =
to be free flow when it is independent of variations
in tail-water level. In a Parshall flume operating under n = an exponent deponent dependent on b.
lrcc-flow conditions, flow in the entrance section is
subcritical, with depths decreasing in the direction The discharge through standard Parshall flume No.
of flow until the critical depth is reached near the 2 to 13, operating under free-flow conditions, is
flume crest. Beyond the crest, in the throat section, obtained from the following equation:
depths are subcritical (see Fig. 1). Free-flow / h \ 1.569b
conditions will exist until the downstream head . . (10)
increases to the point where it causes the submergence
ratio (0 = /lb//la) to become equal to the modular
(that is, C = 0.372 and !z = 1.569Vzh; for the standard
limit oc) that is,
Parshall &nnc No. 1, C,, = 0.384 and II = 1:659bAzh.
oc = h blh a
The discharge equations for each of the standard
When this happens the flow in the exit section and Parshall flumes are specified in Table 3, C =
in the greater part of the throat becomes drawned C,,b(3.279).
(.see Fig. I).
The discharge through large Parshall flumes (see Table
With an even greater downstream head, submerged- 4, co1 1, No.14 to 21) operating under free-flow
flow conditions will extend further upstream to the conditions (that is 0 < 0,) is obtained from rhe
entrance section and will thereby reduce the discharge following equation :
through the flume. In a flume operating under
submerged-flow conditions, the discharge to be Q = (2.2921, + 0.48) h,.6
measured depends on the submerged ratio (T. = (2.3h + 0.48) ha.6 . . (1 I>

Calibration tests indicate the modular limit for (that is, C,, = 2.3 + 0.48/b and rl = 1.6)
Standard Pal-shall flumes to be from 0.55 to 0.78 (see
Tnhle 3, co1 8). The recommended average value of The discharge equations for each of the large Pat-shall
the submergence ratio is 0.6 to 0.7 (see Table 3, flumes are specified in Table 4, co1 3, where C, = C,,.
col 9) and 0.X (see Table 4, co1 8) for standard and
large Pal-shall flumes respectively. Tables3 and 4 also give values of the range of free
discharge [computed from formulae (10) and (I 1)]
The determination of discharge under submerged- applicable for all flume sizes.
flow conditions is possible provided, that the
submergence ratio does not exceed 0.95 8.5.2 Determination of Discharge Uuder Suhmergetl-
flow Conditions
With higher submergence ratio the flume ceases to The discharge through a Parshall flume oper;lting
operate as a flow-measuring structure. under submerged-flow conditions is affected by the

7
IS 14371 : 1996
downstream head and is thus obtained by means of The absence of a throat and a diverging exit section
an adjustment to the free discharge by the equation: makes SANIIRI flumes simpler in design than Parshall
flumes.
Q,,,. = Q - Q, The side walls shall be vertical and shall converge
where (in plan) at an angle of convergence of 11; this
Qd, = the submerged discharge; applies to all sizes of flume.

Q = the free discharge obtained from either The fall, that is the elevation of the flume floor above
formula (10) or formula (11); and the bottom of the downstream channel, is referred
to as the sill of the flume h When a flume with
QE = the reduction in discharge as a result of a sill is constructed in a naturalPbhannel, it is necessary
submergence. that the bed and banks downstream of the sill be
lined for a distance 1,.
To evaluate QE for standard Parshall tlumes (that is Where desired or necessary, the floor of the flume
No. 1 to 13) the following empirical equation shall may be elevated in comparison with the bottom of
be used; the upstream channel, thus producing a sill of height
4.57~3.14-J) ,x,s
hn h
Q,= 0.07 + CJ b ... (13) PI
[( 1.8/0).~-2.4610.305 I
The entrance to and exit from the flume shall be
connected to the channel banks by means of vertical
For large Parshall flumes, the procedure for
walls disposed (in plan) at right angles to the flume
determining QE is as follows:
axis (see Fig. 3).
From Fig. 2, select the value of Qi,? (for the throat
width b = 3.05 m) corresponding to the submergence SANIIRI flumes may be constructed of concrete,
ratio o and the upstream head hs for the flume. reinforced concrete or concrete block, or as a hollow
For throat widths b other than 3.05 m, multiply the structure made of sheet metal with stiffened angles
value of QE,lobtained from Fig. 3 by the submergence and filled with cement mortar.
coefficient C, corresponding to the actual throat width
(see Table 4, co1 9), that is, 9.2 Dimensions

QE = Q,,& ...(14) SANIIRI flume designs are geometrically similar

models of each other, their dimensions being a function
Substitute the calculated value of QE into formula of width b of the exit cross section of the flume.
(12) to determine the value of submerged discharge The other dimensions of these flumes are calculated
QdY using the following equations:

9 SANIIRI FLUMES a) Width, in metres, of the entrance cross-

section
9.1 Description b, = 1.7b . . . (15)
SANIIRI flumes have a rectangular cross-section and
consist of a converging entrance section with a level b) Length, in metres, of the flume
floor which has a fall at its downstream end. There 10 =
1, = 2b . . . (16)
is then an abrupt expansion (in plan) of the flume
cross-section to join the downstream channel (see c) Height, in metres, of the sill
Fig. 3). hD 2 0.5h B,,~ . . . (17)

Table 5 Dimensions and Capacities of Standard SANIIRI Flumes

( Clauses 9.2 and 9.3 )

Free-flow
SANIIRI h 1, = I, b, h hc 15 Head Range Discharge Range, Q
P
Flume m m%
No. A A
m m m m m m /Mi, M> /Min M&Z
I 2 3 4 5 6 I 8 9 10 11

I 0.3 0.6 0.51 0.40 0.7 1.8 0.14 0.55 0.03 0.25

2 0.4 0.8 0.68 0.50 0.8 1.8 0.14 0.60 0.04 0.40
3 0.5 1.0 0.85 0.65 0.9 2.0 0.15 0.70 0.06 0.63
4 0.60 1.2 1.02 0.80 1.0 2.5 0.20 0.85 0.10 1.oo
5 0.75 1.5 I .275 1.00 1.2 3.0 0.22 1.0 0.16 1.60
6 1.0 2.0 1.70 1.20 1.3 3.0 0.24 1.1 0.25 2.50

8
 IS 14371 : 1996

0,Ol 0,02 0.03 0.05 0,07 0,l 0,2 0.3 0,s 0,7 1 2 3 5 7 10 20
P,,, (m3/sJ

FIG. 2 DIAGRAM FOR DETERMININGTHE DISCHARGECORRECTIONFOR LARGE PARSHALLFLUMES

d) Length, in metres, of the lined downstream to the elevation of the fldme floor, shall be set near
channel the entrance to the flume.
15 = %mx . . . (18)
For submerged flow both heads ha and h, need to
e> Height, in metres, of the side walls be measured.
hC = Lllax + (0.15 to 0.20). . .(19) The location of the measurement section
0 Range of width, in me&es, of the exit cross- accommodating the inlet (7 in Fig. 3) of the connecting
section pipe to the stilling well (6 in Fig. 3) for the
0.2 5 b < 1.0 . . . (20) measurement of h, coincides with the exit cross-
section of the flume. The inlet of the connecting
Table 5 gives a summary of the dimensions and
pipe shall be located at the elevation of the flume
capacities of all standard SANIIRI flumes. floor. If no stilling well is provided, a staff gauge
The mean width F of the natural or artificial channel (8 in Fig. 3) set vertically on the exit wall may be
shall be greater than or equal to 1.46,. that is, used.
The design for stilling wells and connecting pipes
b, I 0.7 b, . . . (21) shall comply with the requirements specified in 7.
This shall be taken into considerastion when selecting The range of heads that can be measurement by various
the size of flume to be used, in any particular channel. sizes of SANIIRI flumes is from 0.1 m to 1.1 m
(see Table 5).
9.3 Measurement of Head and Limits of
Application 9.4 Free-flow and Submerged-flow Conditions

The discharge through a SANIIRI flume is determined
by measuring the heads (water depths) in the entrance
section (upstream head, h ) and exit section
(downstream head, hJ (see gg. 3).
Whether one or both heads have to be measured
depends on the flow conditions in the flume.
For free-flow conditions, only the head hil needs to
be measured. The location of the measurement section
accommodating the inlet (2 inFig. 3) of the connecting
pipe to the stilling well (3 in Fig.3) for the
measurement of h coincides with the entrance cross-
section of the tlime.
If no stilling well is provided, a staff gauge (1 in
Fig. 3), which has been zeroed carefully with respect
The discharge through a SANIIRI flume is considered
to be free flow until the modular limit oc = 0.2. When
the submergence ratio (CJ= h /ha) is greater than the
modular limit, the flow in t!le flume will become
drowned. An additional height of the sill may be
provided to extend the free flow range.
The determination of discharge under submerged-
flow conditions is possible provided that the
submergence ratio does not exceed 0.9.
9.5 Determination of Discharge
9.5.1 Determination of Discharge Under Free-flow
Conditions
The discharge through a SANIIRI flume operating

9
IS 14371 : 1996
under fret-flow conditions (that is o 5 0.2) is obtained 10 UNCERTAINTIES IN FLOW
from the following equation: MEASUREMENT

Q = C,,b 4% hi, 1.5 . . . (22) 10.1 General

where C,, is the coefficient of discharge obtained from
10.1.1 In general, the component uncertainties
0.109
c,, = 0.5 - . . . (23) arising from various sources of error may be assessed
6.2612;,
+I (see 10.4and 10.5) and combined (see 10.6) to obtain
an estimation of the total uncertainty in the discharge
9.5.2 Detcrmincrtion qf Discharge Under Submerged- measurement. This total uncertainty will allow
,flow Cnrlditinrls judgment or whether the~discharge can be measured
with sufficient accuracy for the purpose in hand. This
The discharge through a SANIIRI flume operating clause is intended to provide sufficient information
under submerged flow conditions (that is cr > crc = for the user to estimate the uncertainty in a
0.2) is obtained from the empirical equation. measurement of discharge.

Q,,,
= Qc, ...(24) 10.1.2 The total uncertainty may be defined as the
difference between the true discharge and that
where
calculated in accordance with the equations used for
Q,,,. = the submergence discharge,
calibrating the flume, which is assumed to be
C, = the submergence coefficient or correction
constructed and installed in accordance with this
factor obtained from
standard.
1
c* = 1.085 1 - (25) The term uncertainty is used to denote the range
[ 11.7(1 - CT) + 1 I . . . of values, around the measured value, within which
Values of C, for the range of submergence ratios o the true discharge is expected to lie 19 time out of
from 0.20 to 0.90 are specified in Table 6. 20 (that is will1 95 percent confidence limits).

12 5 6 18
\
\

1A Plan View

DIRECTION
OF ~FLOW

6 4 15
1 B Sectional View r L
FIG 3 SANIIRI FLUME

10
 IS 14371 : 1996

10.2 Sources of Error in setting the zero of a water-level gauge to the

elevation of entrance floor produces a systematic
10.2.1 The sources of error in the discharge difference between the true mean of the measured
measurement may be identified by considering a head and the actual value. As repetition of the
coeneralized form of the discharge equation for flumes. measurement does not eliminate systematic error, the
actual value can only be determined by an independent
Q = C,,C,,I? $-h . . 9 (26) measurement known to be more accurate.

where 10.4 Uncertainties in Coefficient Values

a numerical constant not subject to error, 10.4.1 The values quoted in this standard for the
c,, =
and various coefficients in the discharge equations for
g = the acceleration due to gravity, which Parshall and SANIIRI flumes were obtained
varies from place to place, but the empirically on the basis of experiments, which have
variation may be neglected in flow been carefully carried out with sufficient repetition
measurements. of the readings to ensure adequate precision. However,
when measurements are made on other installations,
10.2.2 The only sources of error which need to be systematic discrepancies between coefficients of
considered are: discharge may occur owing to variations in the surface
finish of the device, its installations, the approach
a) The discharge coefficient C,,; flow conditions, the scale effect between model and
site structures, etc.
b) The dimensional measurements of the flume,
for example the throat width, b, of the flume; 10.4.2 The uncertainties in the discharge coefficients
and quoted in this standard are calculated on the basis
c) The measured head h. of the deviation of the experimental data (from various
sources) from the theoretical equations given and are
10.3 Types of Error on the whole systematic. The percentage systematic
uncertainty in C for Parshall flumes X; is between
10.3.1 Errors may be classified as random or 2 percent and 4 percent and for SANIIRI flumes
systematic, the former affecting the reproducibility X; = 3 percent.
(precision) of measurement and the latter affecting
its true accuracy. 10.5 Uncertainties in Measurements Made by the
User
10.3.2 The standard deviation of a set of n
measurements of a quantity y under steady conditions 10.51 Both random and systematic errors will occur
may be estimated from the equation. in measurements made by the user.

r ,;, (Yi - 3* 1% 10.5.2 Since neither the methods of measurement

.y= ___ nor the way in which they are to be made is specified,
y n-1 1 i
. . (27) no numerical values for uncertainties in this category
can be given: they shall be estimated by the user.
where For example, consideration of the method of
measurement of the width of the flume should permit
y = the arithmetic mean of the n measurements, the user to determine the uncertainty in this quantity.
and
10.53 The uncertainty in the value of the gauged
,; = the result of a single measurement.
head shall be determined from an assessment of the
The standard deviation of the mean is then given individual sources of error, for example the
hY uncertainty. In the determination of the gauge zero,
s the freedom from bias and the repeatability of the
*s ZZ- Y . . . (28)
, measuring device (of which the mechanical backlash
711;- of the equipment is an important element), the
and the uncertainty of the mean (see note) is 2sy (at fluctuations of the level to be measured, etc. The
the 95 percent confindence level). This uncertamty uncertainty in the gauged head is the square root 01
is the contribution of the uncertainties in the the sum of the squares of the individual uncertainties.
observations of J to the local uncertainty. This uncertainty may be small if a vernier or
micrometer instrument is used, with a zero
NOTE - The factor of 2 is applicable where n is large. determination of comparable accuracy.
For II = 6 the factor is 2.6; for II = 8 it is 2.4; for n = 10
it is 2.3; for I) = 15 it is 2.1. 10.54 The uncertainty in dimensional measurement
of the flume (essentially the width I?) will depend
10.3.3 A measurement may also be subject to on the accuracy to which the device as constructed
systematic error; the mean of very many measured can be measured. In practice, this uncertainty may
values would thusstill differ from the true value of often prove to be insignificant in comparison with
the quantity being measured. For example, an error other uncertainties.

11
IS 14371 : 1996
Table 6 Submergence Coefficients (Correction Factors) for SANIIRI Flumes
( Clause .9.5.2)

0- Cs 0 Cs o Cs G Cs

0.20 0.98 050 0.92 0.72 0.83 0.81 0.75

().26 ().97 055 0.91 0.74 0.82 0.82 0.73
().32 ().96 0.5X 0.90 0.75 0.81 0.83 0.71
0.60 0.89 0.76 0.80 0.85 0.69
() 38 ().95 062 0,88 0.77 0.79 0.86 0.67
0.78 0.78 0.87 0.65
().42 ().94 065 0,87 0.79 0,77 0.88 0.63
0.67 0.86 0.80 0.76 0.89 0.61
().47 ().93 0.70 0.84 0.90 0.58

10.6 Combination of Uncertainties X,, = the percentage systematic uncertainty in

}1:,,
10.6.1 The total systematic or random uncertainty
is the resultan[ of several contributory uncertainties, 10.7 Presentation of Results
which may themselves be composite uncertainties.
Although it is desirable, and frequently necessary,
Provided th~[ the contributing uncertainties are
independent, small and numerous, they may be to list the total random and total systematic
uncertainties separately, it is appreciated that a simpler
combined together to give an overall random (or
presentation of results may be required. For this
systematic Llncertainty at the 95 percent confidence
purpose, random and systematic uncertainties may
level.
be combined using the following equation:
10.6.2 All sources contribLlting uncertainties will XQ=* ~ X12Q + Xflz
(31)
Q
have both random and sys[cmatic components,
However, in some cases, eitbcr the random or the 11 EXAMPLE
systematic co]nponent may bc predominant and the 11.1 The following is an example of the computation
other component can be neglected in comparison. of the discharge and the associated uncertainty in a
single measurement of flow using a Parshall flume
1().6.3 Because of the different nature of random
operating under free-flow conditions. The throat width
and systematic Llncertainties, they should not normally b = 1.0 m and the gauged head hti = 0.6 m.
be combined with each other, However, if the
provision of 10.6.1 is taken into account, random The other dimensions of the Parshall flume are as
Llncel-tainties from different sources may be combined specitled in Table 1 for the ilume No. 8.
together by the root-sum-of-squares rule and
11.2 The discharge is calculated using the equation
systematic uncertainties from different sources may
given for tlume No. 8 in Table 3:
be similarly combined.
Q = Z,397A,,5) = 2.397 X 0.61S69 =1 .075 lllT/s
10.6.4 The percentage random uncertainty X~ in the
discharge may be calculated from the following 11.3 Since the random uncertainty is negligible, the
equation: uncertainty in this value of Q is dependent only on
the systematic uncertainty.
y,, = * J Xc + ) WI, + /lx*,,,l . (29)
Let us assume that X, = 3 percent (see 10.4.2).
where
11.4 If it is assumed that several measurements of
XC = the percentage random uncertainty in c,
x,, = the width are taken, the random component of
[be percentage random uncertainty in b,
uncertainty in the width measurement is likely to be
and
negligible. The systematic uncertainty in the width
X,,,,= the percentage random uncertainty in h,,
measurement is assumed in this case to be 0.01 m.
j and Jj are exponents of b and h respectively and
Accordingly,
are dependent on the type and di mcnsions of the flume.
10.6.5 The percentage systematic uncertainty Xu Xl, = o
in the discharge may be calculated from the following y 0.01
cc]Llation: lh3=+ x 100
1.0
= * 1 percent
where 11.5 The magnitude of the uncertainty associated
X, the percentage
= systematic uncertainty in c, with the head measuring device is related to the
X,, = the percentage systematic uncertainty in b, particular equipment used. It has been demonstrated
and that the gauge zero of a water-level recorder can be

12
 IS 14371 : 1996
set to an accuracy of + 0.003 m. This is a systematic 11.7 The combination of individual uncertainties to
uncertainty. There is to random uncertainty associated obtain the overall uncerstainty in head is carried out
with the zero setting because, until the zero is as follows:
reset, the true zero will have the same magnitude
It is assumed that X,,,is negligible, the uncertainties
and sign.
in the water-level measurements are :
Therefore,
x,;! = + (xQh,, + X2,)
x 111) =o
0.003 = -t (0 + .I*)
x,,<, = -+ x 100
= f 1 percent
h;,
0.003 and
=f x 100
0.6 Pha = + (YZh + X2,1)%
= -t 0.5 percent = f (0.52 + 0.422)
11.6 Uncertainties associated with different types
= 2 0.65 percent
of water level observation equipment can be
determined using careful tests under controlled 11.8 The combination of individual uncertainties to
conditions, The random component of uncertainty obtain the overall uncertainty in discharge can be
can be determined by taking a series of readings at carried out as follows:
given water level. However, to distinguish this
The total percentage random uncertainty in the
uncertainty from other sources of uncertainty it is
discharge measurement is
necessary that these readings be carried out with the
water level always rising (or falling). For the x, = f (XQC +- y X2, + n*X*,,a)~
equipment used in this example, the standard deviation
of the mean is assumed to be s,, = 0.003 m. Systematic =- + (12 + 0 + 1.569 x I>
uncertainties in water-level measurement occur owing = +- 1.86 percent
to backlash, tape stretching, etc. Where possible,
corrections should be applied, but controlled tests for The total percentage systematic uncertainty in the
given types of equipment will indicate the magnitude discharge measurement is :
of the residual systematic uncertainty. In this case,
X, = & (x, + y2 x"$ + n*pJ
when a water-level recorder is used, the value is
approximately + 0.002 5 m. = it (32 + 1.O52 x 1 +1.5692x 0.652)y2
Accordingly, = + 3.34 percent

SK = 0.003 m 11.9 To facilitate a simple presentation, the random

= 0.006 m and systematic uncertainties can be combined by the
2%
root-sum-of-squares rule as follows:
x,, =+ 2% x 100 = + (X2o + XQD
h Xu
0.006 = + (1.862 + 3.342)
=& x 100
0.6 = rt 3.82 percent
= f 1 percent
The discharge Q is therefore 1.075 mVs + 3.82 percent
0.002 5
x, = & x 100
0.6 Tt.034 I Q < 1.12) m%.

= f 0.42 percent The percentage random uncertainty is + 1.86 percent.

13