-l

t r c T R 0 T E c H N tI 0
c 0 M M r s s ! 0E N N uT EE R N A T I 0 N A L E
(a ffi l i ddel ' 0 rg a n i s aIn - on
deNeormal i sati
ti otenrnati onal l S 0)
I l IE( lLNA C . E . I "
REC(]MMAiIDAT

EA
INTERNATI()N E C T R ( ) T E C H NG
LL I C( )AML M I S S I O N
(affiliated 0rganization
to theInternational - lS0)
for Standardization
I , E . CR
. EC(lMMENllATI(lN

4l
Publication
Deuxidm
6ed i t i o-n S e c o nedd i t i o n
1963
EpA/ . Jt,+

Code concernant
international surplace
derfception
lesessais
hydrauliques
desturbines

codeforthefieldacceptance
International turbines
testsofhydraulic

I
I flrstf trE*Ytrt0f(*rd .{Agcra-dr{fti$Tfi.fi {
{ **-***
I
'.---- "
t

| ".',.lgrn.jr:
I .
trug;tol -_ ia,!r'f
i
I
; O
!*!.4_.__^
IItD
I
r6serv6s
Droitsde reproduction - all rightsreserved
- Copyright

dela Commission
Central
Bureau Internationale
Electrotechnique
1, ruede Varembd

Suisse
0endve,

. *r=h
 - 3 -

CONTENTS

Pages

FoRervonp 9

PnEracr 9

PART 1 - GENERAL RECOMMENDATIONS

Csaprrn I - Oe.Ircr AND scoPE

1. Types of turbines l3
2. Acceptance tests l3
3. Excluded topics l3

CuRprrn II Tt'Rus, DEFINITIoNS, SYMBoLS AND uNlrs

1. Measuring systems l5
2. List of terms l5
2.1 Rate of flow or discharge l5
2.2 Area. 15
2.3 Mean velocity 15
2.4 Pressure 15
2.5 Specific weight of water . t7
2.6 Head 17
2.7 Input and output 25
2.8 Efficiency 25
2.9 Speed 25
2.lA Speed variations 27
2.1I Pressure variations 27
2.12 Cavitation 29
3. Clarificatioris 29
4. Nameplate 29
Tabie I: Variation of i of water 31
Table lI: Variation of the acceleration g JJ

Crinprrn III Na.runr AND EXTENT oF- TECHNICAL GUARANTEES

1. General 35
2. Main guarantees 35
2.1 Output. 35
2.2 Efficiency 35
3. Other guarantees 35
3.1 Speed and pressure variations 35
a t
3.2 Runaway speed . J I

3.3 Cavitation erosion 37

a n
4. Period of suarantee J I

Culprrn IV Tesr coNDITIoNS To BE FULFILLEI)

1. Adequate provision for the test when the plant is designed

2. Permissible deviation from specifled conditions
2 . 1 P r a c t i c a ip l a n t o p e r a t i o n
7.2 Fiuctuations during individual test run
2.3 Deviation from specified head
2.4 Tailwater level
3. Investigation at the site

---t' ,rl,
 Page
Cnnprrn V Trsr pRocEDURE

1. Personnel
43
1.1 Authority for test . 43
1.2 Chief of test 43
1.3 Personnel, number and competence. 43
1.4 Presenceat the tests
43
2. Preparations for test
43
2.1 Submission of drawings and relevant l o t u . 43
2.2 Inspection on site . 43
2.3 Measurements before test
45
2.4 Inspection after test . 45
3. Agreement on test procedure 4s
3.1 Consent to procedure
45
3,2 General programme 'test 45
3.2.1 Extent and estimated duration of 45
3.2.1 Units to be tested 45
3.2.3 Details of measurements 45
3.2.4 Kaplan Turbine, with or without cam 45
3.2.5 Procedure for weighted efficiencies 47
4. Instruments 47
4.I Provision of instruments A-f
4.2 Calibration of instruments 47
5. Observations
47
6. Signing of records
49
7. Procedure in case of dispute or repetition
49

Cn.qpran VI CovpurRTroN oF RESULTS

l. Reference to guaranteed conditions 5l

2. Inaccuracies in measurement 5l
3. Comparison with guarantee 53
4. Individual errors in measurement
59
4.1 Rate of water flow 59
4.2 Free level 59
4.3 Pressure 59
4.4 Output. 59
4.5 Head 6l
4.6 Eliciency .
61

CnaprER VII Frnnl REpoRr

1. Preparation of report
63
2. Content .
63

PART 2 METHODS OF MEASUREMENT

CslprrR VIII MprHoos oF DTscHARGEMEASUREMENT

l. Salt velocity method 67

2. Pressure-time method . 79
3. Currentmeter method . 95
4. Dilution t2l
5. Weir 123
6. Absolute measurement by weight and uotu-.tri" *""*i"* -"ihoo, 127
7. Travelling screen method 129
8. Pitot tubes l3t
9, Orifices, nozzles and Venturimeters 133

!!b---__
 Page
Cn.q.pren IX Hrno

l. References 135
2. Free water level 135
3. Pressure measurements 143
4. Vacuum measurements 159

Cn,tpren X MrnsunEMENT oF powER ourpur

1. General t6l
2. Indirect method of power output measurement t6r
3. Direct method of power output measurement 165

CHaprrR XI MEasUnEMENT AND DrsrRrBUTroN oF LossES

1. General . r69
2. Mechanical losses t69
3. Water consumption by auxiliary equipment
4. Determination of the efficiency of an electric generator by *"un, .f tfr" n.u.irlgof'the.ootingnoiO l:1
5. Thermodynamic method for measuring the efficiency of turbines 193

Cnaprrn XII Sprro

1. Speed measurements for direct measurement of output 201

2. Speed measurements in case of indirect measurement of the output 201

CsaprpR XIII Spnsp vARTATToNSDURTNG suDDEN LoAD REJECTToN

1. General remarks 203

2. Instruments 203

Cslprsn XIV PRrssuRr VARIATIONS DURING SUDDEN LOAD REJECTION

1. General remarks 205

2. Instruments 205
 - 9 -

IN T ER N A T IO N AL EL E C TR OTE C H N IC A L C OMMIS S ION

INTERNATIONAL CODE
FOR THE FIELD ACCEPTANCE TESTS
OF HYDRAULIC TURBINES

FOREWORD

r) The formal decisions or agreements of the I.E.C. on technical matters, prepared by Technical Committees on which all the
National Committees having a special interest therein are represented, express, as nearly as possible, an international
consensus of opinion on the subjects dealt with.

2) They have the form of recommendations for international use and they are accepted by the National Comrnittees in
that sense.

3) In order to promote this international unification, the I.E.C. expressesthe wish that all National Committees having
as yet no national rules, when preparing such rules, should use the I.E.C. recommendations as the fundamental basis
for these rules in so far as national conditions will permit.

4) The desirability is recognized of extending international agreement on these matters through an endeavour to harmonize
national standardization rules with these recommendations in so far as national conditions will permit. The National
Committees pledge their influence towards that end.

PREFACE

This recommendationhas been preparedby I.E.C. TechnicalCommittee No. 4, Hydraulic turbines.

It follows on the earlierI.E.C. Publications,No. 29 (1914),Prime moversfor electricalplant, Nomen-
clature of hydraulic turbines,and the first edition of PublicationNo. 41 (1928),I.E.C. Publicationon the
testing of hydraulic turbines, of which it constitutesthe secondedition. Neither of theseearlier ttublica-
tions, however,describecithe methods or limitations for guaranteetests.

When the work of Technical Committee No. 4 was recommencedin 1952,u'ith tlte United States
National Committee holding the Secretarratand Chairmanship,the immediate objective\\'asthe prepa-
ration of an International Field Test Code for Hydraulic Turbines. The generallines along rr hich this
work was to be carried out was discussedat the meeting held in Philadelphiain 1954 after a lapse of
27 yearc. The outline of the code and the various chapterswere decided upon at the nrer-tirrgheld in
Munich in 1956,when a number of Working Groups were set up to prepare the drafts of the ditlerent
chapters. Following the meetingheld in Zurichin l95l , a draft was submittedto the National Crrrrlntittces
for appr ov alun d e r th e S i x M o n th s ' R u l e i n A pri l 1961.
 The following countriesvoted explicitly in favour of publication:

Austria Japan
Australia Noru,aY
Canada R omani a
Czechoslovakia Su,itzerland
Denmark Turkey
Finland Union of Soviet SocialistRepublics
France IJnited Kingdom
Germany United Statesof America
Italy

Following amendmentsto the text to take into account comments submitted during the voting period,
Belgium and Sweden,which had cast a negativevote on the draft as circulatedunder the Six Months'
Rule, signified that they withdrew their nesativevotes and approved the publication as at present issued.

As with all codes,revisionswill be found desirableand, after the field test code for hydraulic turbines
has been in use for severaiyears,it wiil undoubtedly be desirableto issuea revisededition basedon the
experiencegained in its usc.
 - 1 3-

INTERNATIONAL CODE
FOR THE FIELD ACCEPTANCE TESTS
OF I{YDRATJLIC TURBINES

Part 1 - General Recommendations

CHAPTER I -- OBJECT AND SCOPE

The purpose of this code for field acceptancetests of hydraulic turbines is to define the terms and
quantitieswhich are used,to establishmethodsof testingand the way of measuringthe quantitiesinvolved,
so as to ascertainthe performanceof the turbine and to verify the guarantees.

l. Types of turbines

ln general the code applies to any size of any type of reaction or inipulse turbine. In particular
it applies to turbines directly coupled to electric generators.

2. Acceptancetests

2.1 This code covers the arrangementsfor acceptancetestsat the plant to determinethe attainlrent ol'
turbine guarantees. It determinesthe rules as to their conduct and prescribesthe measuresto be
taken if the conduct of the tests is disputed. It deals with the way in which the results are to be
computed ar-rdthe range, content and style of the final report.

2.2 Model tests for acceptancepurposesare dealt within another code.

3. Excluded topics

3 .1 T his c ode ex c l u d e sa l l ma tte rso f p u re l y c o ml nerci ali nterestexceptthosei rrextri cabl yboupd up ri i tl i
the conduct of the acceptancetests.

3.2 This code is not concernedwith the structural details of the turbines nor with the mechanicaloro-
per t iesof t he i r c o mp o n e n ts .
 I

_ 1 5_

CHAPTER II TERMS, DEFINITIONS, SYMBOLS AND UNITS

i. \Ieasuring systems

T he r t r et r icsy s te mo f units is adopted throughout the code but other systemsshall be

allowed if
I.l' r ' lc r r r ' c l.
T he t c r m s . d e fi n i ti o n s ,symbols and units adopted i n this code are listed below.

l. List of terms

Term Definition Symbol Units

- r l
Rate of Flow or
Disclnrge
Volume of water per second passing a
specified section.
o m3/s
(fr3/s)

2.1.1 Turbine discharge Volume of water used by the turbine per m3/s
second including leakage water in stuffing (ft3ls)
boxes and turbine thrust relief pipes but
excluding water required for operation of
generator, auxiliary machinery and for
cooling of bearings.

2.1.2 Guaranteed discharse Turbine discharge at guaranteed head, o- mt/s

output and speed. (ft3/s)

2.1.3 No-load discharse Turbine discharge at no-troad and normal Qo ffi3rS

speed and guaranteed head. (ft3is)

2. 2 Area Cross-sectional area normal to the seneral

(see Figures I to 4) direction of flow. (fr')
2 . 2. 1 Area at turbine inlet a) In plants with open flume, the area of A1 m2
the agreed section downstream of tur- (fr')
bine inlet screen.

b) In plants rvith closed conduit, the area

of section through the agreed measur-
ing point near the turbine casing down-
stream of inlet valve. \

2.2.2 Area at turbine outlet In reaction turbines, the area of draught A2 ll1:
tube cross-section normal to the flow at (ft:;
the outlet end of the draft tube must be
mutually agreed. In impulse turbines, the
area of cross-section at outlet cannot be
defined and is assumed infinitely large.

2. 3 Mean velocity Rate of flow divided by area of cross-section. mis

(fti s)
V:QIA

2.4 Pressure The pressure at any point in the system kp/cm2

expressedas force per unit area. (lbf/in'z)
 -I

l7 --

Term Dehnition Symbol Units

2.4.1 Gauge pressure The reading given by a gauge of the pres- n

rg kp/cmz
sure at any poirrt in the system. (lbfi in'9)

/--+.L Pressure at ir-rletto tur- T h e g a r - r g ep r e s s u r e a t t h e i n l e t ' m e a s u r i n g kp/cm!

bine point corrected for the elevation of gauge. (lbf/in'z)

2.4.3 Pressure at outlet from The gauge pressure at the outlet measur- kp/cmz
turbine ing point corrected for elevation of gauge. (lbf/in'9)

_l
l
2.5 SpeciJicweight oJ waterr) fne u'eight in air of unit volume of water kp/dm3
|
I used by turbine. (rbfifr3)

1
2.6 Heacl

2.6.1 Pressure head The head of water equivalent to the pres- l,o m
sure at any point in the system. (ft)

2.6.2 Velocitv head The head equivalent to the square of the l,u m
mean velocity divided by twice the acce- (ft)
leration of gravity.
..r)
r.e
lru -
r,

2.6.3 Potential head Eievation of a measuring point above mean m

(Geodetic head) sea le.,el or other reference datum. (ft;

2.6.4 Total head T h e s u m o f p o t e n t i a l h e a d , p r e s s u r eh e a d n1

and velocity head, in a given section. (ft)

h' -' t ' l'u '''

",
2.6.5 Net head The head available for doing work on the Hn m
turbine; it is the difference between total (ft)
head at inlet and outlet. (See illustrative
examples shown in Figures 1 to 4.)

**-l\'

Sl
xi
12 n
- /l\L W

Flc. 1. Reaction turbine (Francis, Propeller, Kaplan). R ectangul arspi ral casi ns
in concrete; draught tube bend.
( u'' 2 - u'r'
Hn : (zwt zw2) -'
k
1) See tabulated values of
T at end of this chapter, Table L
2) See tabulated values of s a t e n d o f t h i s c h a p t e r , T a b l e I I .
 - 1 9 -

Hn

v2
7q
:
,*Z
l/2

1
__----o_-

FIc. 2. - Reaction turbine. Spiral casing with circular section.

(y,z -
-n: -' . -
Y"27
Hn = (2, I et - z*z) r 10. i-
ro
(metric)

\-")l . \

Frc. 3. - Reaction turbine. Horizontal shaft.

Pe; (t',1_ urt)
H n : Q ,
,
l q t - z y y o ) ; 1 0 v ) o
t - o
(metric.)

---
 -21

--.--i:--v-
".*J^\
A \

q$lr

\i
o -- -1./-

Ftc. 4. - Impulse turbine.

A: Single nozzle
Hn:Qt*ot-"r)a 10.':t +
I -
::
6

B: Twin nozzle
Qr - Per lr'
tt ,t n _
-i Qrt \t t T o t zr) I 1 -o , - z r 1 1*) l o ,
e, oro;'erre, Y - 2 . s
(metric)

Term Definition Symbol Units

2.6.6 Guaranteed head The net head for which the turbine HI m
dered. (ft)

2.6.7 Gross head The difference in elevation between the H- m

water levels at the upstream and down- (fr)
stream limits of the installation when no
water is flowing.

2.6.8 Loss of head Loss of total head between anv two ses- m
tions. (fr)

2.6.9 Height of measuring The difference in elevation between the n-t

instrument datum for which the measuring instrument ( ft.1
is calibrated and the location of the meas-
ured point.

2.6.10 Barometer height The height of the water barometer corres- It6 n'l
ponding to the atmospheric pressure at rfrt
the corresponding temperature.

2.6.t1 Geodetic suction head Height of turbine or runner above tail- t,, n-l
water level (see Figure 5). ( f r)
 - 2 3 -

-T
I(^
a"

TI
*9
vV .L

Flc. 5. - Definition of geodetic suction head ft,.

a) Francis low specific speed
b) Francis high specific speed
c) Kaplan, Propeller
d) positive suction head
e) negative suction head

:::-.V
_::= Tailwater level

Symbol Units

2.6.12 The elevation of water surface above mean zw n]

sea level or other reference datum. (ft)
t
E
1<
i
;
I
'f

Term Definition Units

Input and Output

Turbine input The hydraulic power equivalent of the Pd

turbine discharge Q at net head I1',:
P d : k . ^ 1 Q H ,
, 1 000r
{A _ _ror)
1 356
(' o r k : - -- tbot pound system)
1000
Turbine output The mechanical power delivered by the Pt
turbine shaft. It is determined as the sum
of:
a) Output of the generator as measured
at the generator terminals.
b/ Mechanical and electrical losses in the
generator, except c7.
c) In case of common thrust bearings:
pro rata bearing losses due to the
generator.
d) Losses in the gear and flywheel.
e/ Power required by direct driven auxi-
liary machines.
f) If a brake is used instead of a) and l,)
the output is determined by means of
the brake.

2.7.3 Guaranteed output The turbine output at guaranteed net head Pr

11. and guaranteed speed r. for which the
turbine is ordered.

2.8 Efficiency

2.8.1 Turbine efficiency Ratio of turbine output to turbine input.

2.8.2 Turbine weighted The weighted average efficiency computed

average efficiency arithmetically from the single values:
'lp, 'f12, 'flB,
....
at the guaranteed output or discharge
values and at their respective rveights :
1 9 1 1W 2 t W 8 , . , , .

agreed for these outputs or discharges by:

WL.'ri -f Wz.'fp': Wt,. 4a I ....

4,*
wr f w2 t lre -t- ....

Turbine arithmetical Weighted average efficiency as defined in t'a %

averageBfficiency 2 . 8 . 2w i t h
W t : W z : W : t : I

Turbine planimetric The mean height of the efficiency curve 'ft

%
"pl
averageefficiency within the guaranteed range.

Speecl Number of revolutions per unit time. rev/min

2.9.1 Guaranteed speed The speed for which the turbine is ordered. rev/min

,'|
 -27 -

Term Defirritiolr

? o ) No-load speed Steady state speed w'ith the generator exci-

ted, but unloaded after disconnection from
network, with gor.'ernor connected.

Runaway speed The maximum speed attained when all load rev/min
is removed and when the supply of water
is uncirecked, in such an intermediate
position of guide vanes (and runner blades
of Kaplan turbines) wl,ich gives maximum
runaway speed.

2.r0 Speed variations

2.10.1 Initial speed The steady speed just before a specified n1 rev/min
load is removed.

2.10.2 Permanent speed The steady state speed variation from no- ^n
(),/
variation load to a specific load, expressed as a
percentage of the rated speeC.
('n" - ni o/

xn : 100
nr

2 .r 0 . 3 Maximum momentary The maximum momentary speed that is l?max rev/min

overspeed attained when a specified load is suddenly
removed (see Figure 6).

Ftc. 6. - Variation of speedafter a sudclenload rejection.

2.t0.4 Maximum nomentary The maximum momentary change of speed, ( . l ' , r) , . ' ' , , .
speed variation expressed as a percentage of the initial
speed, when a specified load is suddenly
changed.
()./
(J',',),.,.,u*: 100 . t-..tt..0ltt'---1!
17r

2.11 Pressure variations

2.11.1 Static pressure at stand- The pressure when water is not flowing. Po kp/cm?
still (lb/in'z)

2.11.2 Maximum momentary The maximum momentary pressure that is 2tnax kp/cmz
pressure obtained when a specified load is suddenly (lb/in'z)
rernoved (see Figure 7). or m water
column
 --29-

FIc. 7. - Yariation of pressure after a sudden load rejection.

Term Definition Symbol Units

2.1L3 Maximum momentary The maximum momentary pressurechange, (rp)-u* o/

/o
pressurechange expressed as a percentage of the static
pressure, when a specified load is suddenly
changed.
(P^u* - Po)
(zp)m,tx : 100 %
"( He

2.12 Cav,itation

2.12.1 Definition Cavitation is the formation of vapour bub-

bles due to local pressuredrop and their
subsequentcollapse.

2.12.2 Cavitation erosion Cavitation erosion over a certain period

can be measuredby the reduction in weight
of the constructionalmaterial or indirectly
by weighing the refill material necessary
for restoring the eroded surfacesor by the
working hours needed for this repairing.

3 . Clarifications

Clarification of any term, definition or unit of measurementin dispute shall be agreedto in u'riting
by the contracting parties in advance of the tests.

4. Nameplate

Each turbine shall bear a nameplate which provides the following data in a legible and durable
manner:

Nlur or MaxUFACTURER

Year . Serial No. .

Guaranteed head ... m (ft)
Guaranteed discharge .....mt/s (ft3/s)
Guaranteedoutput of turbine . . . . .k w
Guaranteedspeed . rev/min
 3l

Taers I

Variotion of y, the specffic weight of water 'r, tt:ith latitude, temperature and height above mean sea
level.

Specific weight of water 1 in kp/dm3

Latitude Water temperature in oC

10 30 40

At sea le,,,el
0 0.996 0.996 0.996 0.994 0.992 0.988
10 0.996 0.996 0.996 0.995 0.992 0.989
20 0.997 o.997 o.996 0.995 0.992 0.989
30 0.997 0.997 0.997 0.996 0.993 0.990
40 0.998 0.998 0.998 0.997 0.994 0.990
45 approx. o.999 o.999 0.998 0.997 0.994 0.991
50 0.999 0.999 0.999 0.997 0.99s 0.991
60 1.000 1.000 1.000 0.998 0.99s 0992
70 1.001 1.001 1.000 0.999 0.996 0.993

Ar 1000m
0 0.996 o.996 0.996 0.994 0.992 0.988
10 0.996 0.996 4.996 0.994 0.992 0.988
20 0.996 0.997 0.996 0.995 o.992 0.989
30 0.997 0.997 0.997 0.995 0.993 0.989
40 0.998 0.998 0.998 0.996 0.994 0.990
45 approx. 0.998 0.999 0.998 0.997 0.994 0.991
50 0.999 0.999 0.999 0.997 0.995 0.991
60 1.000 1.000 1.000 0.998 0.996 0.992
70 1.000 1.000 1.000 0.999 0.996 0.993

A t 2 000 nt
0 o.996 0.996 0.995 o.994 0.991 0.988
10 0.996 0.996 0.996 0.994 0.992 0.988
20 0.996 o.996 0.996 0.995 0.992 0.989
30 0.997 0.997 0.997 0.995 0.993 0.989
40 0.998 0.998 0.998 0.996 0.994 0.990
45 approx. 0.998 0.998 0.998 0.996 0.994 0.991
50 0.999 0.999 0.999 0.997 0.994 0 . 9 9r
60 1.000 1.000 1.000 0.998 0.995 0.991
70 1.000 1.000 1.000 o.999 0.996 0.993

At 4000 m
0 0.995 0.995 0.995 0.993 0.991 0.988
10 0.99s 0.995 0.995 0.994 o.991 0.988
20 0.996 0.996 0.996 0.994 0.992 0.988
30 0.996 0.997 0.996 0.995 0.992 0.989
40 0.997 0.997 0.997 0.996 0.993 0.989
45 approx. 0.998 0.998 0.998 0.996 0.994 0.990
50 0.998 0.998 0.998 0.997 o.994 0.990
60 0.999 o.999 0.999 0.997 0.995 0.991
70 1.000 1.000 1.000 0.998 0.996 0.992

* Pure distilled water values corrected for buoyancy of air corresponding to International Standard Atmosphere
(I.C.A.N.). This is for dry air, but corrections for humidity are not likely to alter the above values by more than :E 0.1%.
 -33

TesrE II

Variation oJ'the accelerationdue to gravity " g" tt'ith Eorth's latitude and height above mean sea level.

Height
Latitude

Degrees m/s' m/s' m/s2 m/st

0 9.78449 9.71740 9.77432 9 . 7 6 8t s

l0' 9.782A4 9 . 7 7 89 6 9 . 7 7 58 7 9.76970
20" 9.78652 9.78343 9.78034 9.77417
30" 9 . 7 9 33 8 9.79029 9 . 7 8 72 l 9 . 7 8 10 3
40" 9.80180 9.79B72 9 . 7 9 56 3 9.78946
50' 9.81079 9.80770 9 . 8 0 46 l 9.79844
60' 9.81924 9 . 8 1 6t 5 9 . 8t 3 0 7 9.80690
70" 9.82614 9 . 8 2 30 s 9 . 8 1 99 7 9 . 8 1 38 0

The international standard value of " g " adopted by the International

Commission for Weights and Measuresis 9.80665 m/s2 corresponding
approximatelyto latitude 45' and sea level.
 J ) -

CHAPTER III NATURE AND EXTENT OF TECHNICAL GUAR.ANTEES

l. General
A contract contains main guaranteesfor output and efficiencyof a turbine and sometimesadditional
guaranteescovering speed and pressurevariations, runaway speed and cavitation erosion.

The purchaser shall arrange for the supplier of the turbine to be provided with true, full and accep-
table data respecting the waterways from intake to the tailrace, and all parts arid equipment relating
thereto, and all the driven machinery whether electrical or not and the revolving parts thereof, and the
sovernor, valves and gates and all allied mechanisms.
'Ihe
supplier shall not be responsiblefor the values of head and discharge on which the outputs
and other guaranteesare to be based,even though he himself or his agent has made measurementsneces-
sary for them.
lf the electric generator is to be used as an indirect means of measuring output, then the electric
-ceneratorand its auxiliaries shall be given an appropriate acceptancetest at which the supplier or his
representativeshall have the right to be present. By agreement, a certified copy of the generator test
calcr-rlationsand results shall be given to the supplier.

2. Main guarantees

2. I O ut put
The guaranteeconcerning the turbine output is met if the turbine delivers not less thar-rits
guaranteed output at the specified guarantee net head and speed. The guaranteesmay be given
for different net heads.

2.2 Efficiency
2.2.1 The guaranteesregarding efficiencyare met if the efficienciesobtained are not less than those
guaranteed. The guaranteesmay be given for the guaranteed output or discharge and differer-rt
partial loads or discharges.
2.2.1.1 As weighted averageefficienciesover a load or dischargerange (1I.2.8.2).
2.2.1.2 As arithmetical averageefficienciesover a load or dischargerange (II.2.8.3).
2.2.1.3 As planimetric averageefficienciesover a load or dischargerange (1I.2.8.4).
These guaranteesmay be asked for and given for different net heads.
2.2.2 lf the guaranteedoutput or dischargeshould be exceededon test, the supplier shall have the
option of naming an output or dischargenot greater than 10\ above the said guaranteedoutput
or discharge,as a basisfor determiningaverageefficiency. He may use thesenewiy chosenvalues
of output or dischargeto reducethe amount of penalty; if premiums are concerned,the premiums
must be reduced to a minimum.
2.2.3 This newly chosen output or discharge shall be considered as guaranteed outptrt or discharge
for all other suarantees.

3. Other guarantees

3.1 Speed and pressure variations

The extent of speed and pressure variations when specified outputs are suddenly thrown off
or on may be the subject of an additional guarantee.
 3l -

3.2 Runaway speed

A guaranteemay be made that the runaway speedat a specifiedhead will not exceeda specified
value under the conditions of Clause II.2.9.3 and specifiedtailwater level.

3.3 Cavitation erosion

A guaranteelimiting the amount of cavitation erosion (see II.2.12.2) may be asked for and
given on the assumptionthat the turbine is operatedwithin the prescribedlimits of output, duration,
water temperature and tailwater level (Figure 5).

1. Period of guarantee

Unless otherwise agreed,the guaranteesoffered and given shall be in force continuously for a mini-
mum of 12 months to a maximum of 24 months from the agreed day on which the turbine is taken
i nt o ir r dus t r ialo p e ra ti o n .
 -39-

CHAPTER IV TEST CONDITIONS TO BE FULFILLED

l. Adequate provision for the test when the plant is designed

It is recommended that when the plant is being designed attention should be given to provisions
for testing. This applies particularly to the arrangementsfor measurementof the quantity of flow and
the head. The method to be used for measurementof flow should be consideredand chosenat this staee.

The test conditions should also be consideredwhen the intending purchaser,or his engineers,submits
an inquiry to possible suppliers. A statement should be inciuded that the acceptancetest shall be per-
formed in accordancewith the requirementsof this code. If there are to be any deviations from the code,
they should be clearly stated.

2. Permissible deviation from specified conditions

2.1 Practical plant operation

Practical plant operation as a rule involves some variation in head. It is therefore recom-
mended that, when a request is made for tenders, the data submitted to the supplier shall state the
anticipated normal, minimum and maximum heads to which the guaranteesof output and efficiency
are to apply. Should deviations from the specifiedconditions exist during the tests, they shall be
limited as follows:

2.2 Fluctuations during individual test run

Tests shall be made under conditions of constant head, constant load, constant speed and
with fixed gate openings and positions of runner blades within the followine limits of fluctuations
during an individual run.
2.2.I Fluctuations in load s h a l l n o t e x c e e d-! 1.5% of the averageload.

2.2.2 Fluctuations in head shall not exceed *[ 1l of the averagehead.

2.2.3 Fluctuations in speed shall not exceed + 0.5 % of the averagespeed.

2.3 Deviation from specified head

2'3.1 The actual averagehead (speed)for each individual test run during the whole acceptalicetest
may deviate from its specifiedvalue by not more than _[ 3% (+1.5%l if the speed (head) is
kept at its specifiedvalue as per Iy.2.2.

2'3.2 The head may deviate by not more than * l0%, if the actual speedn,, is adjustedto satisly the
relation :

l l n

; :
The subscriptsr and n refer to specifiedand actual conditions respectively.
2.3.3 If the speed cannot be adjusted to suit head deviations beyond ;[ 3\, the correction of test
resuits frLY, by mutual agreement,be made for head andfor speeddeviations by the use of model
test curves when they are available, or if sufficient guaranteed values for different heads are
available, then the corresponding guaranteed head and the corresponding guaranteed.values of
output and efficiency will be establishedby interpolation.

,,.1
 -41

2.4 Tailwater level

For reaction turbines, the taihvater level must not tall below the level which correspotldsto
the maximum suction head statedin the contract. If the actual head deviatesfrom the guaranteed
head according toIY.2.3, the cavitation conditions computed for this head must not be worse than
for the guaranteed head.
For impulse turbines, the highest tailwater level must not exceedthe rnaximum level stated in
t he c ont r ac t .

3. Investigation at the site

Conditions satisfyingthe above mentioneditem IY.2, as well as conditions for the planned measur-
ng methods according to Chapter VIII shall be investigatedat the site prior to the test. The condition
of the water shall be included in this investigation. The water must not carry uudue quantities of
bark, leaves,weeds or similar foreign elementswhich may cling to the edgesof the runner bladesand
unfavourably affect the flow in the turbine.
 43-

CHAPTER V - TEST PROCEDURE

l. Personnel

The selectionof personnelto conduct the test and proceduresin connectiontherewith are matters
of prime importance.

1.1 Authority for test

Unless otherwise provided in the contract, both parties to the contract shall have equal rights
in deterrniningthe test rnethods and proceduresand in selectingthe test personnel. It is preferable,
however,that the test be entrustedto expertswho hold an independentposition toward both parties.

1.2 ChieJ'of Test

A Chief of Test shall be appointed by agreement between the two parties. He shall exercise
authority over all observers. He shall conduct and supervisethe test and report on test conditions.
He shall be responsiblefor all measurements. He shall be responsiblefor the computation of results
and the preparation of the final report. On any question pertaining to the test or its execution,
his decision shali be final.

1.3 Personnel,number and competence

An experienced hydraulic engineer shall be selected for the measurement of flow and an
experiencedelectrical engineer for the measurementof power output. Other members of the test-
ing team will have the experiencerlecessaryto enable them to secureaccurateand reliable records
from the instruments entrusted to thern.

1.4 Presenceat tlte tests

The purchaser and tlie supplier of the turbine shall be entitled to have present members of their
staff.

2. Preparation for test

2.1 Submission of drawings and relevant data

All drawings of importance for the test and all relevant data, documents.specificationsand
certificatesand reports on operatingconditions, sl-rallbe placed at the disposalof the Cliief of Test.

2.2 Inspection on site

Within two days before the start of the tests, the turbine shall be subjectedto inspection.
The purpose of this is to verify:

2.2.I That the machinery is cgmplete according to specifications.

2.2.2 That the scales of guide vane or nozzle openings and blade angles correspond with actual
measurementstaken.

2.2.3 That water passagesare not obstructed or restricted by any foreign material.

2.2.4 That no undue wear has taken place on vital parts.

 1
I
j
_ 4 5 _

2.2.5 That all pressuretappings,tubes and connectionsare clear of obstruction,properly formed and
located.

2.3 Measurements before test

It is recommendedthat dimensionsand length of conduits betweenpressuretappings,in cases

where such information is required, be measured accurately before the unit is operated.

2.4 Inspection after test

After completion of the test, a check inspection of the water and piezometric passagesshall be
made to verily that these are free from foreign material.

3 . Agreement on test procedure

3 . 1 Consent to procedure

The procedure to be followed in the test shall be outlined by the Chief of Test u'der
co'sit
deration of plans made up previously according to IV.l. AII arralgements and plans
are to be
submitted to the purchaser and supplier in ample time for consideration and agreement.
Consent
or objection is to be given in writing.

3.2 Generalprogramme

The acceptancetest shall not take place until after the turbine, following due test running,
has
been handed over to the purchaser for commercial operation. lt is for the purchaser
to decide,
with respectto plant operation and flow conditions, on the date of the acceptancetest.
This should
be as soon as possible after the turbine is handed over to the purchaser, but within
tl-respecified
guaranteeperiod unlessotherwiseagreedby both parties in writing. The general
programme shall
be drafted by the Chief of Test and shall include particulars concerning the following
items:

3'2'l Extent and estimatedcluratictttoJ'test. This shall include a statementof the runs ancl
the loacl
points at which these are to be made.

3'2'2 Units to be tested. This shall state whether several or only one of a group of similar units
are to
be tested with regard to efficiency.

3'2'3 Details of measurement.The methods to be used for measurementof discharge,

output and head
shall be stated.

3'2'4 Koplan turbine, with or w'ithout cam. Tests for efficiency and output of Kaplan
turbines shall
be carried out' if possible with cam control between guide vane opening and blade angle
discon-
nected, as this offers a means of checking the correct relationship betweenthese settings.
A series
of not less than five efficiencytests for each fixed blade angle shall be run at different guide
vane
openings near the presumed correct opening for the selectedblade angle. A graph of
these tests
revealsthe best combination. Thesefr&y, in order to savetime, be carried out as " index
tests,,
with only relative values of the efficiency,if they are amended by suitable accepted.methods,
at
the determined correct guide vane opening. If the best blade angle relationshiphas previously
 A1

been determined by index test, the five test points

required herein above may be reduced to a
minimum of three points. By agreement,the
tests mav be run with the cam connected.

l ? i Pro(edure when weighted

fficiencies are used. when weighted efficiencies which give different
conrmercial importance to different values
of output andlor discharge are specifiedas the
for guarantees,the same procedure should basis
be followed and the maximum accuracy sought
all test runs and not only for those loads to for
which the weighted efficienciesrefer, sincereliable
graphs must be drawn covering the whole
range of tests. The chief of rest shall decide at
points repeat tests shall be made. which

+. Instruments

-r.l Provision oJ' instruments

Maker, manufacturing number and owner
of instruments shall be stated in the final record.
A1l level referencepoints in the head measurins
systen are to be reserveduntil the approval of
final test report. the

1.2 Calibration oJ' instruments

All instruments which cannot be calibrated

on site shall carry valid calibration certificates
an official institution. of
After completion of the test a repeat calibration
shall be made, unlessomitted by agreement.
The date and place of caribration sharl be stated
in the final record. This calibration institution
decide whether both calibrations lie within shall
normal limits. If such is the case, their arithmetical
mean shall be used. If not, a special agreement
must be made.

5 . Observations

5 . 1 A sufficientnumber of suitable observation

sheetswith a clearindication of the items to be
should be prepared for use at the various measured
observation points.
The observation sheets will be prepared in
a sufficient number of copies, so that each party
to the test and the chief of Test can immediately
be handed a complete set.
5 . 2 All readingsshould be taken in relation to time
and at regular intervals,or the readingscan be take'
on signalsto all observers.
5 . 3 All momentaty readingsshall extend over the
same interval of time at each measuri.g point.
arithmetical mean The
of the momentary readings will be used for
the test computations.
5.4 Before carrying out a test, the openings of the
needle nozzle or the guide vanes and, in the case
Kaplan turbines' the positions of the movable of
runner blades,for various positions of the operating
mechanisms' shall be accurately measured
and suitable scales provided. These scales must
accessiblein operation and recorded for each be
run. The test results can be referred to such recorded
p os it ions . ( S ee y .2 .2 .2 ).

5 . 5 Before carrying out the test, the necessary

correction and calibration curves of all instruments
employed shall be available, so that within a short time
following a test run, a sufficient number of
representativecalculations can be made. All
errors hereby determined shall be eliminated
possible or taken into account. only on where
this basis may the tests be terminated and the
ment s r em ov ed(s e eV .4 .2 ). test instru_

b--_---
 49-

5.6 All arrangementsnecessaryfor any method of measurementpror ided in the code are described
rn Part 2.

6. Signing of records

The test records shall be siened bv the Chief of Test.

7. Procedure in case of dispute or repetition

After completion of the test,the records shall be examined by both parties and representativeresults
shall be provisionally computed on site before dispersalof the testing staff. If there is any dissatislaction
ri ith the tests for clearly explained reasons stated in writing, the dissatisfiedparty shall have the right
to demand a new test.
lf final agreement as to the conduct of such test cannot be reached,,the matter shall be referred
to an independent arbitrator acceptable to both parties.
 - f I

CHAPTER VI COMPUTATIO\ OF RESULTS

l. Reference to guaranteed conditions

t . l Using the methods of measurementdescribedin Part 2, the test data shall provide, for each setting
of the turbine guide vanes or needlenozzle, the quantity of water (Q), the head (f1") and the tur-
bine output (Pt).

When weighted efficienciesare used and repeated tests are carried out for the more important
output and/or dischargewith unchangedturbine openings,two or more correlated sets of Q, Hn, i
and Pt will then be calculated for the same opening (see v.3.2.5). f
a
i

1.2 If the test conditions do not differ from those set forth in the requirernentsof Chapter IV, parti-
cularly with respectto the net operating head (F1"),the test results can be convertedto the guaranteed
conditions(fl,) (seeVI.3.1).

T he f oll o w i n g e q u a ti o n ss h o u l d b e u sed:
1/z
"1"
Q, 0 l' !H-nl l P,
\, F1, I
The turbine efficiency,which is taken over unchangedfor this conversion,
Pt 1 000
rvhere k =:
a Hn lU

r.3 If the test head (11")varies beyond the permissible limits of guaranteedhead (I1.), the speedcan be
corrected to guaranteedspeed as follows:
1/z

, r - - t t n (, ! L \
Hn,
cotlversely,this relation can be used to compute the test speed n,.,when the estirnatedtest head
(11') ditrers substantially from the guaranteed head (,I1.) (see rv.2.3).

1.4 Using the equationsin YI.l.2, the efficiencieslt for each point can be calculatedand plotted against
pt or a.

2. Inaccuracies in measurement

2 . 1 Every measurementis attendedby unavoidableinaccuraciesin measurement,eyel n'hen the methods.

instruments and calculationsemployed correspond exactly with this code. When comparing the
results of the tests with th: guaranteefigures, these inaccuraciesin measuremeptmust be takel
into account in a suitable mlnner. This can be done in various ways.

2.2 Experiencewith testson a large number of hydraulic turbines has shown that the inaccuracieswhich
occur when measuring head, flow, output and speedmay often offset each other, but in some cases
they can also be cumulative. If a test is carried out in accordancewith this code,the resultingaverage
overall error in the results for rlt should be between + I I and + 2%. The value of such overall
errors may be agreed upon when signing the contract, with due consideration to the measuring
methods to be adopted. When comparing the results of acceptancetests with the guaranteefigures,
the inaccuracyfor 1t in measurementshall be taken into consideration(see VI.3).
 - 5 3 -

2 . 3 The estimatedvalue of the averageinaccuracy for lt of test results may be arrived at by considering
the possible errors of each individual value (detailed data is given in VL4). The probable inac-
curacies in measurementfq, -f, and f, for the measurementof Q, H and P can, strictly speaking,
only be calculated when the results of the measurementsare complete; the inaccuracy in measure-
ment .fa for the efficiency?lt is then determined according to VI.4.6.

2 . 4 Inaccuraciesrelate only to the measurementsthemselvesand do not relate to the performance of

the turbine.

3 . Comparison with guarantee

3 . 1 In the case of power stations having heads which are subject to variations, guaranteeswith regard
to Pt and It are frequently given for several heads I1'.

3.2 In such instances the guarantee curve for the mean test head 1/,,,,' can be determined by inter-
polation. In this way the adjustment of the test speed nn to the various test heads Hn carr often
be dispensed with and thus facilitate and simplify the acceptancetests.

3.3 In Figure 8 the turbine efficiency is plotted against turbine output P1, converted to a constant
guarantee head H, or to f/,,,,'. For each test point a vertical line is drawn to show the inaccuracy in
measurementas agreed according to Y1.2.2 or as determinedaccording to VI.2.3. The upper and
lower ends of these vertical lines are connectedby two smooth curves. This gives a band of test
efficiency for the turbine. A11 efficiency points within this band are of equal value.

3.4 lf the curve of the guaranteed efficienciesis entirely within this band, as illustrated by Figure 8
then the guarantee as to efficiencyfor the respectivehead shall be regarded as having been fulfilled

34
o/o
3X

90

Il " -/J, ''

R-
84

8?

80

78
/5 MW 2n
L U

Ptn

Frc. 8. - Comparison of test values with guaranteevalues.

Guarantee within band of inaccuraciesof measurement.

3 . 5 lf the guarantee curve is entirely or partly above the upper limit of the efficiency band, then the
guaranteefor these ranges of output or dischargemust be regarded as not having been fulfilled.
 If a penalty is provided for in the contract"then it shall be calculatedonly for those sectionsof the
guaranteedcurve lying outside of the efficieiicvbanci (seeA in Figure 9).

94
%
32

30
i
4
I
I
rla rl
ir
l
sdd
B4
H n:consl
BT n = c o n s- t-

80

78
/0 /5 t4w t0
P7-.*

Ftc. 9. - Comparison of test values with guarantee values.

Guarantee partly above band of inaccuraciesof measurement(A).

3 . 6 If on the other hand the guaranteedcurve is entirely or partly below the lower limit of the efficiency
band, then the premium must be calculated for those sections only lying outside of the efficiency
band (see B in Figure 10).

91
ol
to
92

90

+88
I
*).QA
$wv

84

82

BO

78
/1
p,--
' t

Flc. 10. - Comparison of test values with guarantee values.

Guarantee partly below band of inaccuraciesof measurement(B).
 )/ -

3.7 If mutually agreed upon in the contract, a similar procedure can also be applied to each separate
guarantee point.

3.8 Full load output (100% output) can be designatedas any output Pt, actually attained by the turbine,
which lies between 100\ and ll0% of the output guaranteedfor the respectivehead fI' (see
111.2.2.2and IIL2 .2.3).

3.9 A similar designation is permissible when the efficienciesare guaranteedin relation to the quantity
of water.
3.10 Measured points lying outside the efficiency band must be considered invalid.

3.ll If more than/a of all measuredpoints are invalid, then the whole acceptancetest is to be declared I

invalid as the accuracv called for in this code has not been attained.

3.12 Another method to check the scatter of test points consistsin plotting the values of discharge
Q) and output (P,), adjusted to a constant net head (f1") (see VI.1.2,VI.3.1 and VI.3.2) with
respectto the guide vane or needle operating mechanism (seeVI.3.1). Similarly output shall be
plotted with respect to discharge.
Smooth curves shall be drawn to represent the means of the various observations and the
three curves (discharge-opening,output-opening and output-discharge) shall be compared and
adjusted until all are in agreement.
Using the last equation in VI.I.2, the efficiencycan be calculatedfor various points along
the whole range of these primary curves and a derived efficiency curve plotted against output
( F igur e 11) .

0/^
lri t
Brl
B6 RS
B5
M\//
16

14 4
^t/s
1I

t10 J
t
I
l 6
{
6

0 Z0 10 60 B0 100 lZ0 140 160 mm 100

jo*

F r c . 1 1 .- Curves of test values in function of gate opening ,to.

 -59-

4. Individual errors in measurement

The following table can be used as a guide for the calculation of a value for probable inac-
curacy./, (r.. VI.2.3) from the estimatedinaccuraciesof each individual item measured.

4.1 Rate of water flow Probable inaccuracyfO

4.1.I Currentmeter method in open channels . + r.s%

Closed turbine inlets and pipes from 1.2 to 1.5 m di ameter . + r.2%
In pipes of more than 1.5 m diameter . + r.0%

4.1.2 Weir measurement(without lateral contraction) + r.s%

4 . 1. 3 Res er v o i r. + 0.3%
4 . 1. 4 M ov ing sc re e n + 0.3%
4.1.5 Pressure-timemethod + 1.0%
4.I.6 Salt velocity method + r.0%
4 . 1. 7 P it ot t ube + t.s%
4 . 1. 8 Dilut ion i n p e n s to c k s + t.s%

4.2 Free level

Measurement of free water level differencesft (in m)

(Not for weir measurement).
Probable inaccuracy f1,
4 . 2. 1 W it h poi n te d h o o k ,
I
float gauge,
h %
compressedair

4.2.2 With plate gauge, 5 o/

fixed scales h ,/O

4.3 Pressure Probable inaccuracy f'n

4.3.1 Weight manometer ' :r 0'loo

0.1
4.3.2 Mercury column manometer (h in m) . - /,/,
Q
n

4.3.3 Spring pressuregauge, if calibrated at site . i 0.5%

4.4 Output Probableinaccuracy

Jo
4 . 4. 1 B y m eas u re me n o
t f to rq u e a n d s p e e d. + 1.0%
4.4.2 By measurement of the electric power at the terminals of :
- d. c . g e n e ra to rs
+ 1.0%
- a.c. generators ] 0.87.
Depending upon the method employed for the determination of the generator losses.and
in cases where shunts are used for high direct currents, .fp may be increased.
 - 6l

4.5 Head

This inaccuracyf, can be computed with a fair approximation

from the inaccuracy/r, of the
head h, from the inaccuracy .f* of the pressurep. and from
the specific weight T by
^ (.fh . h) - (.f, . rl^;_)
fs:
lt p!^;
4.6 Efficiency

4'6'l The inaccuracy-f'aof the turbine efficiencydetermination is

computed from the estimatedindi-
vidual inaccuraciesbv

J:n: a \76,, i
.fii' * -fp"
For careful measurement and normal testing conditions
an inaccuracy in rlt of + 1.0% to
+ 2.0% may be obtained; for unfavourable conditions, however, a higher
value has to be
considered.

4'6'2 Thermodynamicntethod. When applied with net head above

150 m
: + 1.5%
" f" q
 63-

CHAPTER VII FINAL REPORT

l. Preparation of report

The fi'al report shall be written by the chief

of resr. It shall be prepared in draft form ancl
i'itted to the purchaserand supplier of the sub-
turbine to obtain approval of both parties as
calculation and the presentationof results. Any to the details
"f differencesin viewpoint st att be resolvedby
Irartieswith each having equal rights in determining the final both
form.

'r1
I
l. Content l
The final report shall contain, but not necessarily
in this order, the following:

2 ,1 O bjec t of t es t .

I
2'2 Record of any preliminary agreementspertinent
to the test.
2.3 Personneltaking part in the test.

2'4 Description of whole generating unit

and particulars concerning service and load
the turbine to the time of the test, nameplate conditions of
d,ataof the turbine, efficiency,and output guaranteed
by contract.

2'5 Description of equipment and calibrations

used for load, head and flow measurements.

2.6 Procedure in making the test.

Company plant
Results of power tests of turbine No. from to
Determination of output and efficiencv

Flow Net Turbine Results at m (ft) head

No. Timing
Turbine head input
of of
test test opening
Flow Turbine
Dn Pd Pg..r. output
o Pt

Ftc' 12. - Tabulation of resurts of all measurements.

 65-

2'1 Sample calculation for at least one run in detail from ori-sinaldata to final result.

1.8 Daily Log of the events of the test.

l'9 Diagrams used for the calculated run (pressure-timediagrams. salt velocity curve,
currentmeter
velocity traverses,etc.).

I'10 Tabulation of results of all measurementsincluding the intermediate data, results of calculations
for the head, the flow measurements,through to final corrected. results with reduction
of the
tested values to guarantee conditions H, and,n, for
e and p1, ds well as the measurement of
electrical output and determination of the efficiency of generator (Figure 12).

2.11 Statement as to inaccuracy of each class of measurement and inaccuracy of data based
on com-
bined measurements. Unfavourable influences duritrg the tests shall be recorded.
 -I

- 6 7 _

Part 2 - Methods of measurement

CHAPTER VIII METHODS OF DISCHARGE MEASUREMENT

Methods of dischargemeasurementmay be divided into several

main groups as follows:
Lt./ Methods suitable for large rates of flow:
Salt velocity (Allen);
Pressure-time(Gibson);
Currentmeter;
Dilution.

b) Methods suitable for small rates of flow:

All the above methods where size permits,,with the addition of
:
Weir;
Gravimetric and volumetric methods;
Travelling screen;
Differential-pressuremeters.

c) Methods of water measurementwhich rnay not yet be generally acceptcd

for inclusion above but
which might, nevertheless,be used in specialcasesby agreement
between supplier and purchaser.

l. Salt velocity method') ,) u).)

1'1 The salt velocity method originally developedby Prolessor

c. M. Allen depends upon the fact
that salt in solution (brine) increasesthe conductivity of water and
that a slug of brineinaconduit
containing flowing water does not lose its identity. Salt solution is introduce
d at a section in the
conduit upstream from two or more sets of electrodesand its passagepast
the electrodestogether
with the elapsedtime are recorded graphically. The true dischargeis
found by dividing the volume
of the conduit between the planes of the electrodesby the mean
time of passagerequired for the
salt cloud to pass between the electrodes.
In order to meet the requirementsof this code the dischargeshall
only be measuredby the
differential method, that is between two sets of electrodes. The
measurementof dischargeusing
the volume betweenthe brine injection station and one set of
electrodesmay be used by mutual
agreement.

l'2 The brine is mixed in an open tank and continually agitated.

Its concentrationcan be varied to
suit the requirements of the recording system, between 120 gll
e.2 lblgal) to 3 gil ('/r, lb/gal).
A filter is fitted to prevent solid material from entering the injection
system (Figure l3).

l '3 The brine shall be injected into the conduit through a piping
system and injection valves which
discharge the brine in a flat sheet at right angles to the conduit flow.
These pop-valves shall be
spring loaded with a back pressureof about 0.35 kp/cm, (5 lbf/inr)
to provide positive closureand
t) Ttt"t*tions, Am' Soc.Mech.Eng.(A.S.M.E.)
volume 45, 1923:paperNo. 1902,pages2g5to 341inclusive, by
C. M. Allen and E. A. Taylor.
z; " Fieldchecksof the Saltvelocity
Method", Trans.A.s.M.E.Nov. 1g40,pages663-676inclusive by o. H. Dodkin.
3)" salt velocityMeasurements at Low velocities
in Pipes" by LeslieJ. Hooper.Trans.A.s.M.E.1g40,pages651-66I
inclusive.
a) " contribution d r'6tude
de ra mesuredes d6bits d'eau par la mdthodeAllen ", by Martin
de I'Hydraulique,Vol. V, 1939. Mason, Revue G6n6rale

j
I
l
I
)
 - 6 9 -

Flc. 13. - Arrangement of salt tank and piping for salt velocity method.

a) Fresh water supply

b) Mixing tank
c) Wooden box
d) Screen
e) Pressuretank
f) Blow off
g) Air supply
h) Quick acting valve
i) Short pipe
k) To pop-valve

to avoid leakage, and their opening shall be adjustable as shown in Figure 14. The injection pres-
sure of the brine shall exceedthe conduit pressureby at least 2.5 kp/cm, (36 lbf/inr). The injection
pipe must be as short as possible and air pockets shall be avoided.

n (t%')

FIc. 14. - Cross section of Allen pop-valve.

 _ il

t . 4 The electrodesmust give approximately equal conductir,'ityeffectsto equal areasof the pipe. In both
circular and rectangularconduits this may be done by dividing the sectioninto a number of equal
areas and locating an electrodeat the centre of each strip.
The electrodebars may be of two types parallel or 'o bowed " and shall be applied to the full
cross-sectionof the conduit. Only parallel electrodesconsisting of single bars shall be used for
rectangularconduits. Tl-reyshall be placed at the centres of each of 9 or 1l (an odd number is
required) strips of equal atea.
For circular conduits, parallel electrodesmay be used consisting of single bars placed at the
centre of each equal area strip (an odd number is required). It is also possibleand preferableto
install tapered electrodeswhich will in effect accomplish two traversesat right anglesacross the pipe
(Figure l5). In this casethe spacingof the bars is varied to give approximatelyequal conductivity
effectsto rings of equal area in the pipe cross-section.

Frc. 15. - Electrode arrangement.

A. Looking downstream
a) Electrodebars
b) Insulatedstiffeners
c) Penstockwall
d) Insulated connection to pipe wall
e) Insulated head guys with turnbuckles
f) Direction of flow
D) Diameter, to be checked in field
Note- - End spacing and distance between electrode bars : Dllo0

B.SectiottX-X

Not less than two sets of tapered electrodesshall be used at any measuringsection,each set
consistingof a pair of bars spacedtlrrth of the conduit diameter apart,at the centre of the conduit.
The spacing shall decreaselinearly from the centre towards the outer wall to a minimum of one
hundredth of the conduit diameter. Electrodes shall terminate at a distance from the conduit wall,
not less than this spacing.
The electrodesmust be strong, able to withstand vibration, and provided with upstream guys
to hold the electrodeif it becomesloose. The electrode bars must be carefully insulated from each
other and the conduit by materials such as phenolic compounds which have sufficiently high
 ---_:
i

mechanicaland electricaIstrength to prevent breakdou'n due to vibration. All surfacesnot intended

to be used as active electrodesshall be coated with insulatingpaint, including the edgesand outside
facesof the electrodes,the end supports,guys and the conduit wall, if conducting,withil 1 m of the
end of t he e l e c tro d e s .

1 . 5 The cotlnectionsbetween tl-reelectrical circuit and the electrodesshould be as short as possible.

All leads inside the conduit must be well insulated and rigidly supported, preferably in electrical
conduit. Alternate electrode bars shall be connected in parallel to the recording instrument, which
can be an ammeter or an oscillographas shown in Figures 16 and 17. In both casesit is desirable
to keep both sides of the electrica.lcircuit isolated from earth.

Ftc. 16. - Typical recording ammeter circuit

A) Recording ammeter
a) Two-circuittransformer
b) Electrodes
Note. - Tapped secondary winding which operates ungrounded.

Ftc. 17. - Typical oscillograph recording circuit.

a) 2 500Hz (c/s) oscillator
b) Electrodesconnectedin parallel
c) Matching transformer from 120 ohms impedance in
primary to variable secondary50,25 and 10 ohms
d) Oscillograph
 H
:r

The undamped natural period of a reccrding chart instrument shall be no greaterthan 1.5 s,
eitherdirect-writingor photo-
or one-fifthof the passagetime, whicheverbe the shorter. Oscillographs,
recording, are recommendedsince their sensitivitypermits the use of brine of lower density and
they have an extremely small undamped natural period.

An accuratetime baseshall be provided on the recordingchart by meansof reliablemechanical

or electrical timers which can be continuously calibrated durir-rgthe test.

1 . 6 The cross-sectionalareas of the interior of the conduit and the length betweenthe centre-linesof
the electrodestatigns shall be measuredin the field with sufficientprecision to determine the volume
of the test sectionwithin two tenths of one percent. Construction drawings shall be used only as
a check upon field measurements.

1 . 7 Before filling the conduit, the injection apparatus should be flushed out, adjusted and testedto make
sure that the pop-valves open simultaneously and close rapidly. Leaky pop-valves should be cor-
rected. The electrodes and circuits should be tested for continuity and insulation.

1 . 8 After filling the conduit, it is desirableto measurethe resistanceto each electrodebetweenbars and
also from each bar to earth to ensure that both electrode stations are operating similarly and all
faults corrected. At a convenient discharge,trial injections should be made to determine suitable
brine concentration, tank pressure,duration of valve opening (this should be about I s), electrode
voltage and instrument sensitivity to obtain chart records of suitable height for accurate evaluation.
A lower limit of conduit velocity for satisfactory operation is given by:

v2 : 0.2 gD ( ^ -_ t
\ I W )
wherev - mean pipe velocity in m/s (ft/s);
g : accelerationdue to gravity m/s'z(ft/s'z);
D : pipe-diameterin metres(feet);
^[ : specificweight of brine in kp/dm3(lbflft');
u
Yw: specificweight of water in kp/dm3 (lbf/ft).

l.g Each test run shall consist of at least five individual salt injections arrd the average of the five or
more times of passageof the brine shall be used in computing the flow rate for each test run.

1.10 In order to comply with the requirementsof this code, the following minimum conditions shall
be satisfied:

1.10.1 The salt velocity method can only be used in turbulent flow.

I.IO.2 The first set of electrodesshall be at least4 diameters(or 4 times the rnaximum conduit dimension)
from the brine injection valves. The second set of electrodesshall be at least 4 diameters(or
4 times the maximum conduit dimension)downstreamor l5 m (50 ft), whicheverbe the longer.

1.10.3 The whole test section, comprisir-rgthe entire length from the injection station to the further
electrodeshall preferably be of uniform or decreasingarea. If there are slight bends within the
proposedtest section,their centrelinelengthsshallbe added to the minimum lengthsof VIII.1.10.2,
provided such bends occur wholly betweenelectrodestations.

1
 ------------t

1.10.4 Pop-valvesshould be locatedas uniformly as possibleorer the cross-section

as shown in Figure 18.
There shall be at least one pop-valvefor each 2 mj (21.5ft.) of cross-sectionalareaat the injection
station and never lessthan 4 pop-valvesfor circtrlarconduits.

Ftc. 18. - Injection arrangement.

A. Looking upstreant
o) Injectionpiping systern
b) Welded joints
c) Supply pipe
d) Penstock wall
e) Head guys
./) Direction of flow

B. Side View
Dimensions valid for an area of p e n s t o c k : 1 1 3 f t 2 o r 1 0 . 5 m e
No. of pop-valves : 6.
Area for each pop-valve : 18.9 f t 2 o r 1 . 7 0 m 2 .

Injection pop-valvesare to be located at the centres of approximately square

eqllal subdi-
v is ionsof t h e c ro s s -s e c ti o n a l ' e a . In c i rc ul arcondui tsreqr-ri ri ng
4 pop-val vesthey shal l be equal l y
spaced on a circle 0.63 times the diarneter of the conduit.
l'10'5 For conventionalrecording chart instruments,the time of passageof the
brine betweenelectrodes
shall be no less than 9 secondsat full load. With oscillographs,shorter tirnes
are permissible
by agreement. In any case the trace from the upstream and dorvnstreamelectrodes
must not
overlap.
1'10'6 The timer shall be calibratedbefore and during the test so that its rate is
known within an absolute
error of 0.05/,.
1'10.7 Deviations from the above requirementsmay be permissibleby mutual
agreement.
1'11 The dischargeis equal to the volume of the test sectionbetweenthe two sets
of electrodesdivided
by the average time of passageof the brine.
The recordedcurve of brine passingeach electlodeshall be plotted on rectilinearco-ordinates
of time and current (Figure l9). Only that portior-rof the curve above the steady
value of the
initial current shall be used. The averagetime of passageis the distancemeasured
in seconds
 betweenthe centresof obtai ned injection usrng
the recording chart.

k .r

4=Az

Frc. 19. - Typical salt velocity curve.

4/ - Passage time between centres of gravity
kv k2 - Scale

If the curveshave not been recordedin uniform rectilinearco-ordinates,they

must be replotted
before evaluating the positions of the centres of area.

2 . Pressure-time method u) u)

2.1 This method for measuring the flow of water deviseclby N. R. Gibson is
applicable where the water
to be measuredflows through a closed conduit of either uniform or
varying (but not diverging)
cross-section.

2 . 2 It is based upon two fundamental principles, the first is Newton's Second

Law of Motion known
as the Equation of Impulse and Momentum, and the second,a corollary
of the first, which gives
the relation betweenchange of pressureand change of velocity of a column
of water expressedin
terms of the pressurewaves propagated during the change.

2'3 The method makes use of pressure-timediagrams of either the

simple or differential type.
2'3'l Simple diagramsare defined as those recording the pressurevariations
betweena singlepressure
measurementsection and the intake or first free surfacesection. They
are affectedby all variations
of level in the intake, surge tanks and other free surfacepoints in addition
to the basic pressure
variations due to friction and change in momentum.

2'3'2 Differential diagrams are those recording the pressurevariations

betweentwo pressuremeasure-
ment sectiouswith no intermediate free surface points of relief and
are affected only by the
differential friction and change in momentum between such two points.
The effect of conduit
friction outside of the differential test section, changesin intake or conduit
friction outside of
the differential test section and changesin intake or surge tank elevations
are present at both
pressuremeasurementsectionsand are thus eliminated.

5) Trans.A.S.M.E., Volume 45,

7923,Paper No. 1903,pages 343 to 376 inclusive,by Norman R.
6) " Experiencein the Use of the Gibson.
Gibson Method of Water Measurementfor EfficiencyTests of Hydraulic
by N. R. Gibson,Trans. A.S.M.E., 1959. Turbines,,
 8l

2.3.3 under this code, only differential diagrams are to be used.

2.4 The rninimal conditions for the use of this method under this code are:
a) For circular conduits of uniform diameter the length between the two pressure-measurement
sections for differential diagrams shall be not less than 9 m (30 ft) nor less
than twice the
maximum dimensionof the conduit cross-section,whichever be greater; and the produ
ct of L
aud vu shall not be less than 18.6,where vuis the mean velocity in the test section
in m/s when
the unit is carrying full load and I is the length of the conduit used for the
testsin metres.
The correspondingvalue in the foot-pound systemis 200, where lu is the mean
velocity in ft/s
and L is the length of the conduit in feet.

b) For conduits of rectangular cross-section,such as multiple intake sectionsor scroll

caseintakes,
the length of conduit between the two pressuremeasurementsectionsfor differential
diagrams
shall be not less than 18 m (60 f0 nor less than twice the maximum dimension
of the conduit
cross-section,whichever be greater; and the product of L and.ya shall not be less
than 3: .2,
where vu is the mean velocity in the test section in m/s when the unit is carrying
full load and
Z is the length of the conduit used for the tests, in metres. The corresponding
value in the
foot-pound systemis 400, where vuis the mean velocity in ft/s and L is the
length of the conduit
in feet.
ln multiple intake sections, simultaneous independent pressure-time diagrams
shall be
taken in each section of the intake.

2'5 It is recommended that the location of pressuretappings in any test section

be such that the most
reliable diagrams may be obtained, and the lengths substantially greater than
the minimum limit.

2 . 6 The leakage through the turbine guide vanes or other device used in producing
the pressure rise
must be measured separately with the closing device in the closed position.
Such leakage under
test conditions shall not be greater than 5 of full load discharge and
% shall be determined within
an error of 0.2 % of full load discharge.

2 . 7 The cross-sectionalareas and length of the interior of the conduit shall

be measured in the field
with sufficientprecision to determine the pipe factor F (seeVIll.2.l5 andVIII.2.l7)
within 0.2%.
Construction drawings shall be used only as a check on field measurements.

2 . 8 Two or more pressuretappings 9 mm (3/8 in) minimum diameter shall be installed

at each measur-
ing section. They shall be installed in positions diametrically opposed in
a plale normal to r6e
axis of the section, and shall be connectedto a common header suitably located
for each section.
Pipes leading from the headers of the two sections are to be connected
one to each side of the
Gibson apparatus U-tube manometer. For circular conduits, the pressuretappings
sl-rouldbe located
at 45' to the horizontal centreline of the section and shall be connectedto
a common header suitably
Iocated for each section. For rectangular conduits, the pressure-tappings
should be located in the
vertical walls.

2'9 All pressuretappings shall be so connected with individual valves

that they may b9 read separately.

The wall of the conduit shall be smooth and parallel with the flow in the
vicinity of the orifices
for a distance of at least 450 mrn (18 in) upstream and 150 mm (6 in)
downstream. The bores of
the pressuretappings shall be not less than 9 mm (3/8 in) in diameter, and
their edge shall be free
from burrs or irregularitiesand shall preferably be rounded to a radius
of 1.5 mm (l/16 in) or at
least an equivalent chamfer.
 -81 --

The pressuretappings shall be flush with the surlaceof the conduit wall; their bore shall be
normal to the wall. and uniform for a depth of at least twice the bore.

2.10 All g.:ugeconnectionsshall be tight againstleakage,and shall be flushedout and checkedbefore

and after the test. Pressurereadingsshall be checked before and after the test by comparison
with headwaterelevationwhen the turbine is shut down and static conditions exist in the conduit.
lf, during the test, any pressure tapping be obviously in error due to some local cause or
other condition showing a total head differing from initial available head corresponding to the
elevationof the surface of headwaterless the conduit head loss, the source of discrepancyshall
be found and removed, or the tapping shall be eliminated.

Measurementsof the head acting upon the turbine are describedin Chapter IX.

2.II To ensureaccuracyof the pressurerecording of the Gibson diagrams,conduit conditions,includ-

ing velocity distribution, length of straight run of conduit, and conditions of tapping orifices
shall be such that no pressuretapping shall vary in its readings by more than 20 % of the velocity
head from the average of all tappings in the cross-section. The average of readings of any pair
of oppositetappings shall not vary from the averageof any other pair by more tl-ranl0/"of the
velocity head.

2 . 1 2 The rate of flow which is to be measuredin the conduit shall be fixed bv lockine the guide vane
or nozzle opening of the turbine in the required position.

f"
I
I
^ - / . J J - tt.tru
L --7;i
uay
zc
I
=l
-s
Jl-Xy^n, q
n rrr /l.l
=u..Jtu \
tU/
jf I
I
- (a)
\
"o.Ael 0.00r)=0.125"
0.5/5
N
9)
cd
0.125't

&-,F;;I

Ftc. 20. - Typical pressure-timediagram.

a) Diagram end point value
b) Harmonic ratio (see FiS. 22)
c) Ratio of wave heights

2.13 A pressure-timediagram (Figure 20) shall then be obtained by gradually closing the guide vane
or nozzle in one continuous movement and graphically recording the resultant change in pressure
on the chart of a recording device such as the Gibson apparatus (Figure 21) which photographs
 -85-

on a sensitizedfilm moving at a uniform rate, the movement caused by the pressure change of
the top surface of a column of mercurv in a U-tube manometer.

) )

/'ii
S---1

Ftc. 21. - Arrangement of Gibson apparatus.

A) Downstream pressure measuring section
B) Upstream pressuremeasuring section
C) Light source
D) Glass tube leg of manometer
E) Stainlesssteel leg of manometer
F) Seconds pendulum
G) Photographic lens
H) Photographic film on revolving drum
D Conduit

Procedure for delineating and measuring the diagram is as follows:

A horizontal line AA, called the running line, is drawn coincident with the position on the
diagram representingthe level of the surface of the mercury in the glass tube of the apparatus
under running conditions before the turbine guide vanes begin to close. If the running position
of the mercury is wavy, the running line should be taken as the mean position of its peaks and
valleys.
A horizontal line F, called the static line, is drawn coincident with the position on the dia-
gram representing the level of the mercury in the glass tube of the apparatus under static condi-
tions after the turbine guide vanes have been closed. Usually this line is coincident with the
median line of the afterwaves shown at PP. The median line is obtained by bisecting the envelope
curves shown at HH and GG.

,1
Y
l$
i

s7-

T he v e rti c a l l i n e KN ' I n ra rk i n g th e end of the di agram i s l ocated as fol l ow s:

a) Measure the horizotttal clistaltcealong the static line from its intersectionat V with the
rising edge of the first afterrvaveto its ir-rtersection
at R with falling edge.

b) S ubt r a c t th e s l o t c o rre c ti o n fro rn thi s rneasurement.

C) Measure the heights above the static line of the peaks of two adjacent waves apd colnpute
the ratio c, of wave heights as shown on Figure 20.

d) Determine the quantity 1- -tt- fr-om Figure 22 and multiply this quantirl, by, rhe
;
dim en s i o no b ta i n e d i n i te m D ).

/ r
/ 0 e B 7 2 / 0 U { tgnt
155 +
r.. Jt
A
J I -
I
I
l s + \Vi l evssinr
=l \v /
st ,/ _r
tQl\." Ag e lostnT
.I I JJ
N I

t 4
t &
i tsnr
/ g -e

/ ., 2 3
l t m o x '. pU I lL n _ +
-/
t2 no^

Ftc. 22. - Study of damped harmonic wave.

Curve showing horizontal distance from point of inflec-
tion to vertical line through peak of wave
(: tr - Xr',.,or)

The numerical valuesfor items a, b, c, are shown on Figure 20. The distance so determined
is laid off along the static line from the point W where it intersectsthe falling edge of the diagram
following closure to a point M. Draw a vertical line MK. This line determinesthe end of the
diasram.

2.14 The recovery of friction and velocity heads on the diagram shall be evaluatedby dividing the
diagram into several parts and by trial and error determining the friction and velocity heads
destroyedthroughout the period of closr.rre.At any time during the closing period, the ratio of
any part of the net area, from the beginning of the diagram to the whole net area, is equal to
the ratio of the velocity destroyedup to that time to the initial velocity.

i
j
)
 _ 8 9 _

The diagram is divided into small segmentalareasAr, Ar, A" etc. A trial recovery line ONM
is drawn as shown by the dotted line on Figure 20. Each of the small areas above ONM is then
measured by planimeter. From each of these areas the slot correction is deducted.

Figure 23 is a tabular form which has been designedso that the successivestepsto be taken
in measuring a diagram may be taken in a systematic manner.

a) The data entered at the top of the form for identification and convenient referenceare self-
explanatory.

b) The first five columns of the table show respectively:

Column l. The designating letters of the segmental areas A r, A r, A " etc.

Column 2. The vertical height of each segmentalarea to be multiplied by the effective

slot width to obtain the slot correction.
Note. - A scale can be constructedwhich will read the area of slot correction directly,
in which case the words " height for " at the head of the column may be eliminated.

Column 3. The planimeter readings for each segmental area.

Note. - Satisfactory measurementsof 4rea are obtained by using a precision type
planimeter rated to have an accuracy of 1 : 4 000. This instrument may be set to
half scale. Since at least two separate determinations of each area should be
made to ensure the necessaryprecision and minimize personal errors of observation,
the sum of differencesbetween two separatereadings after the first setting gives to full
scale the mean of the two determinations of area.

Column 4. The difference between planimeter readings.

Column 5. The mean area of each small segmentof the diagram and the slot correction
which is subtracted to give the net area.

c) The summationfiguresin Column 6 are obtained by adding together successivelythe net areas
in Column 5. When the last of the segmental areas has been added in, the total will be the
net arca of the diagram above the assumed recovery line. It is not necessaryfor the first
trial to check the sum of the small areas,by measuring the total area, before proceeding
with the computations but, after the final recovery line has been located, the final total area
of the diagram is then measured by planimeter.

2.15 That part of the grossarea of the diagram produced by the recoveryof friction and velocity head
shall therefore be eliminated by consideringthat the value of the friction and velocity headsremain-
ing is equal to
(l - r)'
where a
AT+AL
c - sum of the friction and velocity heads recorded on the pressure-timediagram

and
a : the net area of the diagram measured up to a given segmentalarea x
(a :Ar*A " * A " * 4" )

Ar: the total net area of the diagram for all the x segmental areas

Ar : )tr, * A,* A, .... a A*)

,1
 -91

Efficiency Test by Gibson Method

Plant . Unit No. Date .

Proprietor Run No.
Apparatus No. T r - r b eN o . Riser No. K :
q.F.S
Pipe FactorF: . . . L e a k e t g e( t ' - . . . m 3 / s( f t : ' / s )l L n ' r -( r n - ,
K
P r e s s u r e - T i m e - D i a g r a n r S c i t l c s:
Vert. 1
Horiz.
h, lhr
Width for Slot Correction

, l 8 l
Computation of Recovcrr
Segment Flanimcter Measurements

Lleight
for Slot
Corr. ( l-r'1:16'

A1

4.,
--l1
A3

Al

AJ

: A"l
A7 : At,
:Ar * Ar

Total x Corr. for Print r n 21 i n 2 )

Area

K.A
O
' m3/s (ft3/s)
f.s
Leakagc 17 m31s (ft3/s)

Total Q

Frc. 23.- Sampleof a data and computation sheetof an efficiencytest by the pressure-
time method.

a
 - 9 3 -

/ -
,tL - the area of the diagram corresponding to leakage flow
q ' F S
/ -
" L - k
where 4 .: leakage flow past the turbine inlet control device used to create the pressurerise
in m3/s (ft'l$.
tr constant of conduit called the pipe factor. If the pipe is of uniform cross-section
!-.
F:
A
r - distance in metres (feet) between the two pressuremeasuring sectionsfor differential
diagrams, measured along the centreline of the test section.
A cross-sectionalarea of the conduit in square metres (ftr).
k calibration constant of the recording apparatus : gly
v : vertical height in centimetres (inches) on the diagram corresponding to 1 m (1 ft)
of pressure head change in the conduit. This calibration constant is constant for
any particular set of apparatus.
,S : time-constant of diagram. It is the horizontal length in centimetres (inches)
correspondingto one second of time.

When the value of c is small, as it is when the upper and lower piezometer section have the
same cross-sectionalatea and consequentlythe value of h, is the same at both sections,the reco-
very line can, frequently, be located with sufficient precision by eye and the process of subdi-
viding the area of the diagram into segmental parts can be omitted or at least limited to a
check at one or two points.

2.16 The rate of flow shall be computedby the equation

A a
o : : ' 1- q
J" f' -
where Q : discharge in m'/s (ft'/s).
a
6
: acceleration of gravity in m/s'9(ft/s') adjusted for latitude and elevation above sea
level for the power station.
Aa : final net area of the diagram after the recovery line has been established,measured
by planimeter and corrected for width of slot and shrinkage or stretching of the
print of the diagram.

The area of the pressure-timediagram is a measure of only that quantity of water actually
shut off. If any quantity remains flowing in the conduit after closure of the turbine guide vanes
it must be measured separatelyand the amount of it included to obtain the total dischargeat the
moment the guide vanes began to close. The leakage shall be measured under the same condi-
tions of closure attained during the test period.

When the leakage quantity is thus relatively small, no appreciable error is introduced in the
computation of dischargefrom the diagrams by assuring that the leakage velocity in the conduit
is zero. The leakage quantity is simply added to the discharge computed from the diagram.

2.17 If the conduit be of non-uniform cross-section,

F is the integration of length divided by area:

where I -
F -
f+
length along the conduit in metres(feet).
 -9.5-

3 . Currentmeter method

3 . 1 General
T he c ur re n tme te rme th o d e mp l o y sa number of i ndi vi cl ualcurrenrmetersproperl y pl aced i '
a c los edor op e n c o n d u i t to re g i s te rth e fl o u' . Thei r nuntberand l ocati onsshal l be such as to gi ve
recordsat a sufficientnumber of individual points to permit a precisedeterrninationof the averase
f l o w i n t h e t e s t s e c t i o nd u r i n g t h e r e c o r d i n gt i m e ( s e eV I I I . 3 . 3 . 6 a n d V 1 I I . 3 . 5 . l ) .

3 .1 . 1 T he c ur r en tme te rm e th o d c a n b e a p p l i e d to the fol l ow i ng l ocati onsof test secti onsi p a hydro-

elec t r icplan t:
Type (a) Open canals;
Type (6) Turbine or penstock intake structures;
Type (c) Closed conduits.
3.1.2 Each location will be describedand its limitations set forth in the applicable sectior1s
of this
code. Where the description of apparatus,proceduresor limitations apply to two or more test
s ec t ionloc a ti o n s ,th e y w i l l b e s o s ta te d by referenceto V III.3.l .l (a), (& ), or (c).

3.2 Provisions applicable to all test locatiotts

3'2.1 Currentmeter measurementsshould only be carried out if steady flow is maintained durilg
the
whole of a measuringrun. If any change of load or flow occurs during any measuringrun, the
run shall be repeated (see IV.2.Z).

3'2.2 Types of currentmeters.Only axial flow electric signalling meters shall be used. The
bearing
arrangementis of specialimportance. The nreter shall be so designedthat changesin temperature
aud minor impurities ir-rthe water do not influence appreciably the firlctioning of the meter.

Generally,currentmetersshall have a maximum tip diameter not lesst6an 100 mm (3.93 in).
For measurementsin a penstock of a diameter smaller tltan 2 m (6.6 ft), currentmetersof a
tip
diameter of 50 to B0 mm may be used in the peripheral zone.
The currentmetermay be fitted with directional flow devicesto detect aly possiblereverse
currents (see VIII.3.2.8).
3'2'3 Mounting of currentmeters. Currentmetersshall be accurately locateclwith their
axis parallel
to the canal or pipe axis and fastenedto the mounting rocl as shown in Figure 24 so as to
avoicl
deflection and minimize vibration.

3'2.4 Staggeredlayout- In so far as local conditions do not permit the use of currentmeters
as in
VI1I.3.2.3,they can be arranged staggeredas in Figure 25. The axis of the currenuneters
shall
neverthelessbe parallel to the conduit axis.
A staggeredarrangementcan therefore only be used when there is a sufficientlelgth of uniform
sectionabove and beiow the gauging section to ensurethat the velocity distribution is
the same
at the entrance and exit of the section. The angle of staggershall not exceed30'.

3'2'5 Calibration and cltecking of currentmeters. The currentmeters shall be calibrated

with the same
type of mounting as that used during the tests. The calibration shall preferably include
oblique
flow up to 10'- Any deformation of the blades subsequentto calibration may readily
be checked
by using plaster moulds or metal forms. The rrrould must correspond with the meter
as calibrated
and should have the date of calibration marked on it. The bearing friction and condition
of the
bearings of the currentmeter shall be checked before and after the test.
 -'l
-97 -

F rc . 2 4 . - Types of currentmeters for axial and oblique flow.

a) Stoppani bladed meter with protecting ring of 120 mm
diam., impeller diam. 90 mm
b) Ott-3-BladeMeter, 120mm diam.
Note. - Both types are suitable for clean water but easily
clogged by grass,leaves,algae. Type (a) is of advantagefor
measurements in ill determinedsections).

c) Ott curved vane n"reter,100 mm

d) Dumas/Neyrpic curved vane meter, diam. 100 mm

e) Ott miniature curved vane meter, diam. 50 mm

f) Amsler curved vane meter, diam. 130mm
Note. - All types( a) to (f) are for axial flow measurements;
seeFig. 26, a within about + 10'.

c) Ott curved vane meter, type F, diam. 125 mm, for

oblique flow, a within about + 20'
h) Ott curved vane meter, Type A, diam. 100 mm, for
oblique flow, a within about l, 45o

\-'

Ftc. 25. - Staggeredmounting of currentmeters when the measuring section is not

perpendicularto the general direction of flow, or the axis of the canal or conduit.
 'l
99-

3.2.6 Effect oJ obliqueflo*. Currentmeterscan be used for oblique flows provided:

a) That the currentmeters used are designed to measure accurately the true axial component
of velocity, this being checked by suitable calibration;
b) The maximum deviation of the flow from the meter axis is not in excessof 10" (Figure 26).

Note. - The usual type of currentmeter will give correct registration for angles less than 10', with an accuracy
of 1l or less as related to the cosine law. There are types of so-called " component " currentmeters
available that show departures from the cosine law of less than l)( for angles greater than 10'. Under
conditions where suctr currentmeters are used, they must be placed to itvoid any effect on their accuracy
due to the frame or supports.

\ 0.9
*{t <!t
/ac /00
0.8
I n,
I
I n,
g '''
a
' n 6

0,4
I
60" 'f',- 4f 30" 0'
-<-=- u -+

Ftc. 26. - Comparison of the calibration curve with the cosine law in oblique flow.
v velocity of flow at angle a
vs component of the velocity v, perpendicular to the mea-
suring section, that is parallel to the axis of the nleter.

This can be expressed as : coS cr :

.:'
, l c

Actually the currentmeter indicates ve': COS .{

v

3.2.7 Velocity limits. The use of currentmetersfor high velocitiesis lirnited only by the abilitl' to
provide mounting of the meters in accordancewith VIII.3.2.3. In no case ntay the ratin-ecur\;e
be ex t r ap o l a te d .T h e me a n v e l o c i tyQIA rn the gaugi ngsecti onshoul dbe at l east0.4 In' s (1.3ft/s).

3.2.8 Velocity distribution. Every effort is to be made to secure as uniform a velocity distributiou as
possible. The velocity distribution in the gauging section must be investigatedbefore the main
test.
If the velocity distribution is found to be very irregular in open channelsor intakes,it shall
be improved by installing appropriate guide walls, submergedroofing, stilling, racks, raflts,etc.,
until the desiredflow pattern is secured(see Figures 27 and 28). If there are counter currents
or reversalsof flow in any test section,measurementsshail not be carried out, evclt thougl-rthe
currentmetersare equipped with directional flow contacts.
 I
- l0l

F tc . 2 7 . - Means for stabilizing flow in an open canal.

R) Racks
F) Rafts
f) Submergedroof
M) Measuring section

-t-

I
0'l

\
/.,:

Ftc. 28. - Temporary nozzle or bell mouth placed in the intake of a low head turbine.
a) Measuring section

3'2'9 Pulsation In order to discloseany possible pulsation in the velocity, currentmetersshall

be used
which make a signal at least every 20, and preferably every 10 revolutions of the propeller.

3.2.10 I'{umberof currentmeters. The number and location of currentmeterswill vary. according
to the
type of test section and conduit. The required number is specifiedfor each particular type of test
s ec t ionunde r V III.3 .3 .6 a n d V III.3 .5 .l .
 1
l 0_j -

3'2'11 Measurementof time and of rotational speec{o.f ctrrrentmeter.

The time rate measurement shall
be accurateto at least 0.05%. If severalcurrentmetersare useclsimultaneously,
indications may
be registeredby means of an electric recordin-echroncmeter or
by coulters.
Whicheverrnethodis used,the rotational speedof the currentmetersshall
be measuredwithin
an error of + 0.1%.

3'2.12 Measurementprocess. Each flow measurementshall cover the whole area of

the gaugingsection.
The mean water velocity,and consequentlythe discharge,must not
be computedfrom the velocity
measuredat only a single point of the section.
If the conduit is divided into several sections,the flow must be computed
for each secti on
individually.

3'2'13 Duratiort of measurernents.Measurementsfor each currentmeter position

shall last at least two
minutes.
Should the water velocity be subjectto perioclicpulsations,the duration
of the measurement
shall be prolonged so that at least two periods of the pulsation
are availablefor tl-recalculatior-r.

If a group of fixed-positioncurrentmetersare used,the measuring period

shall be at least
5 minutes, iu order to provide sufficienttime for the other measurements.

3.3 currentnteter measurement test sections in canals. Type (a)

The location of the test section for currentmetermeasLrren'lent

shall be chosen witir particular
care to ensure the greatestpossibleaccLrracy.The following poi6ts
should be observed:
3'3' l The minimum dimensionsfor rectanguiar or trapezoidal gauging
sectiols are as follows :

Minimum width of canal B : 0.8 m (2.6 ft) or 8 x cnrrentmeter tip diameter

M inim u m d e p th o f w a te r Z : 0 .8 m (2.6 ft) or B x currentmeterti p di ameter

3'3'2 The sidesof the canal should be straight and parallel

and the bottom should be as nearly hori-
zontal as possible for a length of at least 2 m (6.6 ft) before
entering the test section. Both the
approach and downstreamportions of the canal shall be free
from any obstructionswhich are
liable to cause flow disturbance.

3'3'3 The test section shall be perpendicularto the direction

of flor,vand the currentmeteraxis parallel
to the flow' If local conditions do not permit such an arrangement,
a staggeredmou'tir,s arrar-
gement of the curre'tmeters may be used (vrrr.3.2.4).

3'3'4 No leakageor diversion of water shall be permitted

betweenthe test sectioirancl the tLrrbineinlet
or outlet.

3'3'5 Devicesto improyefiow conclitions.If devicesto stabilize

flow, such as baffiesetc. are r-rsed,tirey
should be placed not less than 3 rn (10 ft) upstreamof the gauging
section. J-hesemay servenot
only to improve the velocity distribution but also to obtain
better conditions f-or an accurate
measurementof the depth of water.

3'3'6 Number o.f gaugingpoittts. For canals of rectangular

or trapezoidal section the number of gaug-
ing points z in a given section A can be calculatedfrom
the following formula :
3,-
^, ^',- -
24\ A t-_: 136
\',{
rvhere A (the area of cross-section) is in square metres.
 -105_

3'3'7 Distribution of gaugingpoints. The gauging points shall be arran-eed

closer to one another in the
zones where the greater velocity variations are to be found, i.e. near the walls,
the bottom of
the canal and the water surface.

3'3'8 Distancefrom walls and water surfoce. The distance of the axis of the
nearestcurrentmeter from
a wall or t h e b o tto m o f th e c a n a l s h a l l be kept w i thi n the fol l ow i ng l i mi ts:
minimum : 0.75 x tip diameter of propeller
maximum : 0.2 m (8 in).
The topmost currentmeter must be entirely submerged and placed as close
as possible to
the surflace,keeping a depth so that the influence of the water surface will
not affect the meter.
3'3'9 Constancyof flow. As the water level in the gauging section is used to
determine the water depth,
it shall be constantly checkedduring the time of gauging. The variations of the
water depth during
measurementsshall not exceedf ll of the mean depth.
The water depth can be checked by means of a separatepointed measuring
rod (Figure 29)
or by raising the currentmeter supporting rod and, setting the axis of
the uppermost meter at
the water surface (Figure 30). Any variations in the average velocity should
be checked at the

t same time as the test run, with one or more currentmeters.

I
i
I

-I

Ftc. 29. - Measurementof water level with a point gauge.

3.4 Turbine or conduit intakes. Type (b).

With low head power stationsthe measuringsectioncan also be located

i1 the turbine ir-rtake.
It is essentialin such casesto ensure that a straight and parallel flow
is obtained over the whole
gauging section by means of temporary guide walls and a submerged
roofing (Figures 27 and,2g).

3'4'l Converging oblique conduits. Measurement with converging flow

over an oblique measurin,s
section should wherever possible be avoided becauseit is difficult to
allow for systematicerrors

i
; i
il
I
 -107-

FS

- A
1t- 1

l / _
4

Ftc. 30. - Measurementof water level using the tip of the uppermost meter and raising
the supporting frame.

which can arise. In any case,witl-rsuch a measurementonly true cosinecomponent currentmeters

are to be used as describedin YIII.3.2.6.

3.4.2 Bell mouth nozzles. Tl-re provision of a bell mouth nozzle (Figure 28) has the advantage of
per m it t ing:
- An increasein the velocity that can be measured;
- Improved velocity distribution;
- Elimination of water deptli measurementsby providing a constant area at the test section.

FIc.3l. Currentn.retersupporting frames for measuringflow in the intake of large

lorv-headturbines (Bittcrli-Typeframe).
 1 0 9-

3.4.3 The provisions of the following paragraphs shall apply equally to intakes type (b).

VIII.3.3.I
VIIT.3.3.6
VIII.3.3.7
vilr.3.3.8
VIII.3.3.9
3.4.4 simultaneous measurementwith severalcurrentmetersis an advantagewith large gauging sections
in intakes. In this way the duration of the test can be appreciably reduced. According to the
local conditions, simultaneous reading can thus be taken for complete horizontal or vertical
traverses(Figure 3l).

3.5 Closed conduits of circular section. Type (c).

Currentmeters shall only be used for measurementsin penstock or closed conduits when they
have an internal diameter of at least 1.2 m (3.9 ft). If a measurementis carried out with a fixed
supporting frame and a larger number of currentmeters, then the minimum diameter of the pipe
should not be less than 1.4 m or 4.6 ft, or 14 x currentmeter diameter.
For large conduits, special referenceshould be made to the provisions set forth in paragraph
V I I I . 3. 5. 2.

3.5.1 When making measurementsin a conduit with a fixed currentmeterframe (e.g.as in Figure 32
with 4 arms) and with currentmeters of about 100 mm (3.9 in) tip diameter, the number Z of
currentmeterson each arm of the support is calculatedfrom the following formula :

4 1 ' ,R' . - Z < 5 \ R

where R is the internal radius of the conduit in metres.

FIc. 32. - Battery of stationary currentmeters on fixed cross bars in a closed conduit.

3.5.2 A minimum of l3 measuringpoints must be used at any sectionand whatevertype of supporting

frame is used. A currentmeter should be mounted in the centre.
 - l l 1 -

Figure 33 shows al'l example with a diagram and tabulation for the location of currentmeter
measuring points in a circular conduit. The positions of the currentmeters on the radius arms
of the supports are arrangedto give, as closelyas possible,equal weight to all measuringpoints.

D m m
(in) r.J rr ,,
l,
I

| 200-24AA
13
(48-e5) 0 . 7 5D r Dt2-x o.sra
I
____t__
I t---
17
2200-3 240
(87-127) 0.75Dn

2 900-4500
Dl2-x
----i i
0.866
I
2l
( l 1s - 17 9 ) 0 . 7 5D p D,t2--r o.sl+ l-"ot-
i I
3 800-5500
25
( 1 5 i - 2 1 8 )0 . 7 5D r ;:,","- 0.8r6 | 0.707 0.408

29 5 000-7000 0.535 0.378

(r9B-278) 0 . 1 5D p 0.845 0 . 7 5 61 0 . 6 s 5

a a 6 300-8s00 0.354
J J
(2s0-337) 0 . 7 5D e 0.613 0.500

=n zq/l-

Ftc. 33. Table and diagram for selectingthe number of currentneters in a circular
conduit, the location of the meters or points of measurement.
z) Nuurber of meters, inclr-rding the centre location
D) Inside diameter of the conduit
Dp) Diameter of the currentmeter
x) Distance from the axis of the meter to the wall of
the conduit
r) Radius at the axis of the currentmeter

3.5.3 Gauging sectionsshall lie in a straight stretch and at least 20 pipe diametersdownstreamand at
least 5 pipe diametersupstreamfrom the nearestbend. Departuresfrom this rule are permissible
only when tlie flou' conditions at the proposedmeasuringsectionhave been examinedin advance.
In pipes. nrcasurementsare normaliy taken along at least two diarnetersperpendicularto each
other.
 - ll3 -

The mean diameter of the pipe shall be taken as the arithmetical mean of six measured
diameters. Measurententsof flow using two diameters perpendicular to each other are consid-
ered satisfactory if the florvs computed for each diameter do not differ by more than * 2\ of
their mean.

If the distancesdownstream or upstrearn from a bend are less than those speeifiedabove,
the purchaser and the manufacturer must agree on a greater allowance for inaccuraciesin the
flow measurement or, if possible, use traverses on more than two diameters.

Providedthe requirementsof YIII.3.2.3 are satisfied,the gaugingcan be carried out as follows:

with several currentmeters permanently mounted in the penstock (Figure 32);

with a single currentmeter and valves (Dufour method, Figure 34);
with one or more currentmeters sliding along a fixed bar.

I
t--
i
--Jlm'\-
- "-v,,"
+-
FIc. 34. - Traverse acrossconduit by means of a single currentmeter with a lock valve
on two perpendicular diameters (Dufour Method).
 _ lt5 _

Should local conditions not permit the use of one of the above mentionedmethods,gauging
can also be carried out. if the two oarties agree,by means of a currentmeter mounted on a rod
which can be moved both arially and swung from side to side over the whole gauging section to
approximate two diameter traverses.

Another arrangement uses a rotating support with a number of currentmeters mounted on

the arms (Figure 35).

Jt

Ftc. 35. - Measuring section of a conduit with currentmetersmounted on rotatins arms.

a) Direction of flow
b) Currentmeters
c) Rotating arm
d) Tripod support
e) Attachment at wall
.f) Driving gear of rotating arm
d To the recorder

3.6 Computation of results

3.6.1 Calculation of mean velocities. The evaluation of the currentmeter readingsshall be done graph-
ically by plotting the measured velocity against measurementsection areas and with a plani-
meter determining the dischargeas the product of velocity and area. The resultsshould be checked
arithmetically.
Graph or cross-sectionpaper shall not be used on account of inaccuracv and irresularitv
of printing. Account must be taken of the shrinkage of the paper.
The planimeter used shall be checked before, during and after the computations. A linear
planimeter, with rollers on rails, is preferred. The scales shall be so chosen that the accuracy
of the graphical integration is at least equal to or better than that of the time, length, depth and
 - Ilt -

velocity measurements. Practical experiencehas shown that the most appropriate scalesare those
giving approximately sqLlarediagrams with sides about 150 to 200 mm (6 in to 8 in) long.

3.6.2 Peripheral zone

a) Open canals or turbine intakes. When turbulent flow is present, as in most cases,the extra-
poiation of the velocity curve from the last measuring point to the wall of the measuring section
can be made using the formula:
I
/o*r n
Y rl - l
\ 4 '

where v* and vt are the velocities in the extrapolation zone (at any distance a* from the wall)
and at the measuring point nearest to the wall (at the distance a) and exponent n :
\n may
vary between 5 and 10).
If point by point velocity measurementsby pitot tubes or other devices show substantial
departures from the relation using exponent fi : 7, then the corresponding exponent shall be
us ed.
The computation can be simplified by graphical methods as shown in Figure 36 or by inte-
gration. In the latter case the mean velocity v*- in this zone is given by the relation :

r* ,,.: v,
(;-)

--a
l/o

Ftc. 36. - Graphical method of determining velocity at the periphery of the measuring
sectlon.

b) Measurementin pipes and closed conduits. The velocity distribution in pipes is to be deter-
m ined as in V III.3 .6 .1 .

3.6.3 Sequence of integrations for closed or open conduits with trapezoidal sections. The first inte-
gration shall be carried out for the vertical or horizontal line along which the velocities have
been measured. For instance, with a vertical line (Figure 37), the discharge is determined
according to the formula:
B T
/ t l r

Q : l d B, J l vdT
tl
o o
wher e B is th e ma ri m u m w i d th o f th e gauging sectron and T the water depth.
 - ll9 -

P,
llt
I

|).t
I
pz
I
llz

Ftc. 37. - Example showing flow measurementin a slightly trapezoidal canal.

Successivegraphic integration.

3.6.4 Closed conduits of circular cross-section. Both graphical methods shown in Figure 38 shall be
used for the calculation of dischargewith co-ordinates:

a) The product of rv as a function of the distance r between the point of measurementand the
centre of tlre section:

b) The velocity v as a function of the square rz of the distance r.

The difference between the two dischargescomputed by the two graphical methods should
not exceed I%. The mean value of both is consideredas the final one.

Ftc. 38. - Example showing flow measurementin a circular closed conduit.

Graphical integration methods.

3.6.5 Direct integration methodfor intake sections. The use of the direct integration method where
currentmeters are moved with a constant speed over the whole gauging section leads to certain
difficulties and calls for very costly equipment, so that, as compared with normal currentmeter
measurement,the only advantage gained occurs when measuring very great water quantities in
an extretnely large gauging section. If the direct integration method is used, a sufficient number
of measurementsmust be taken with a stationary frame to show a satisfactory agreement.
 -l2t-

,4. Dilution

4 . 1 This method, suitable for long len-uthsof pipe and canal, consists
of introducing continuously a
known concentrated solution of chernical at a steady measured rate
into the main water flow and
determining the resuiting concentration at a point far enough
downstream to ensure thorough
mixing' The main water flow can be computed from the following
expression:

Cr-C,
o = :q Cr-Co
where: Q rate of flow to be measured;
cl rate of flow of chemical solution injected;
co concentration of chemical in the natural water;
c1 concentration of chemical solution injected;
C, concentration of chemical in the resulting mixture.

4 . 2 The section along which it is proposed to inject the dosing solution

shall be previously explored
to check that reversecurrentsdo not exist,which may carry any proportion
of the chemicalupstream.

4.3 The dosing solution may consist of chemicalssuch as:

Solution Finql concentration

recommended
Sodium chloride 2 mgll
Sodium dichromate 0.2 msll
4'4 Ideally, chemicals should be chemically stable, non-toxic and not
normally found in natural water.
In addition, they should be such as to be easily detectedand their concentrations
measuredto a
few parts in 10-8by titration, conductivity tests, or some other well proven
chemical technique.

4'5 The rate of the injected solution shall be uniform and accuratelycontrolled
and measured. Depend-
ing on the type of installation, constant flow devices such as syphons,
Mariotte vessel,metering
pumps, may be used. The chemical solution shall be thoroughly
stirred to secure homogeneity
before it is injected into the main flow.

4 . 6 Samplesshall be taken at points located on the centresof equal areas

over the measuringsection.
Samples shall be taken continuously from not iess than five points, for
a minimum time of fir.e
minutes. The samplesmust show at each point uniformity of dilution with
respectto time and also
o v er t he whole s e c ti o n .

4 . 7 Samples shall be taken regularly:

a) of the water at least three diameters upstream of the injection poi nt;
b) of the mixture of natural water and solution downstream of the injection point.
In this case
the uniformity of the mixture must be verified;
c) of the concentrated solution at the beginning and end of each injection.

4'8 For dischargecalculations,it is recommendedthat the " compared

dilutions method,, should be
used' This method consistsin comparing samplestaken from the flow with samples
of the injected
solution mixed with water taken out upstream of the injection point.
Since it is not necessaryto measure the strength of solution, the colorimetric
method of com-
parison may be used and thus large systematicerrors avoided.
 -r23-

4 . 9 For each test, a concentration curve shall be made of the water downstream of the injection point.
This curve shall have time plotted as the abscissa,and quantity flowing as ordinates. It should have
a section of practically cor-rstanthei-ehtcorresponding to the period during which the strength of
the solution is constant. It is this strength or degreeof concentrationthat is used in the previous
formula.
As it is imperative to verify the uniformity of strength of the solution, great care should be taken
with this part of the test.

5 . Weir

5.1 In order to have the best known head dischargerelationship, the weir shall be rectangular without
end contraction, sharp-crested,with a smooth vertical crest wall, complete crest contraction and
free overflow.
The crest shall preferably be formed of rustproof metal about 5 mm (0.2 in) thick, with a sharp
right-angled corner on the upstream edge, an actual crest width of 2 mm (0.08 in) and bevelled at
an angle of 45" on the downstream face.
The crest edge shall be level, smooth and free frorn rust; its height s above the bottom of the
approach canal shall be not less than twice the maximum head to be measured.
Complete aeration of the nappe shall be secured. The ventilation must be sufficient to keep
the air underneath the nappe at approximately atmospheric pressure. The cross-sectionalarea of
the ventilation holes must be at least0.5 % of the product of the width of the weir times by the height
of the crest above the upstream canal floor (Figure 39).

(0'08")
'j 'r1tt
- M> 1 llnor)
(6 lLnox
D1

---.-.- m^(o.zo,,)
lehil A

FIc. 39. - Weir dimensions.

s) Height of weir crest
ft) Overflow head

Observation of the crest conditions and form of nappe shall be made during the test to avoid
unsuitable conditions such as adhering nappe, disturbed or turbulent flow, or surging.

The side walls of the canal shall be smooth and parallel and shall extend downstream beyond
the affiux and above the level of the crest. The weir shall be setperpendicular to the canal walls and
bottom, and its upstream face shall be absolutely smooth and free from projections of any kind,
even bolt and rivet l-reads.
 1
I
j

-r25- i
I
i
5.2 Conditions of installatiott ancl use I
!

The weir is commonly located on the tailrace side of the turbine, and care shall be taken to ensure
that smooth flow (free from eddies, surface disturbancesor the presenceof considerableamounts
of entrained air) exists in the approach canal.
If it is necessaryto locate the weir in the tailrace, it shall be far enough from the turbine outlet
to enable the water to releaseits air bubbles before reaching the weir. Stilling screensand baffies
should be used when necessaryto give a uniform velocity distribution over the whole of the cross-
section. Disturbed surface or undercurrents, or asymmetry of any kind, must be corrected by
suitable screens.

The approach canal shall be straight, of uniform cross-section,and unobstructed by screens

and baffies for a length upstream from the crest of at least 20.h^u" and downstream for 5.l.r^u*.
Along this length the bottom slope must be very small (< 0.005). A desilting sluice can be installed
if required, but it shall not disturb the regular flow of water along the upstream face of the weir.

5.3 Measurementof head

The head h above the crest shall be measured upstream of the weir at a distance not less
than 4 nor more than 6 times the maximum head. For measuring h the number of measuring
points uniformly spaced across the weir canal shall be as follows:
((
Length oJ' crest b )) Number of measuring points
For b <2 m (6.5 ft) 2
2 m ( b < 6 m (2 0 ft) 3
b) 6 m 4 at least.

Before and after a series of point measurements,the zero height shall be accurately checked.
Measurements of the head at each side of the approach canal shall not differ more than A.5"1.
For calculating the mean flow, the arithmetic mean of all the head measurementsshall be used.

For head measurementsit may be convenientto use hook gauges,either directly or placed in
stilling boxes at the side of the approach canal, communicating through specialpressureconnections
terminating in orifices 3 to 5 mm (0.12 in - 0.20 in) in diameter. Severalreadings at regular time
intervals shall be taken to check that steady flow conditions have been reached.

5.4 Limits

The dimensions of the weir shall preferably be arranged so that, for full load on the turbine,
the following limits apply:

Width of weir b > 0.25 m (0.8 ft)

Height of weir to crest s ) 0.30 m (1.0 ft)
Head h < 0.75 m (2.5 ft)
Velocity of approach < 0.45 m/s (1.5 ftls)

Under these conditions, measurementsby weir will be within the range covered by the experiments
upon which the weir formulae are based.
 1 2 7-

5.5 Discharge formulae

The general formula for calculating the discharge of a weir is the Poleni Formula:

0:2'3.'t.h \t h't.
where : Q : rate of flow m3is (ft3,ts);
'rt : coefficient;
b : length of weir crest in metres (feet);
h : observedhead above crest in metres (feet).
Among many formulae which have been proposed to expressmore exactly from experimental
results the head-dischargerelation, the following formulae appear to be the best recommended.

Bazin formula (1898):

t. ,l -

a : ( o . o o r + ! ' o o 3 i r + o s s(\ J/ z- )* s ' |I . b . \ , z s h ' t ,

\ h )
where: s : height of weir crest above bottom of canal of approach (in the above formula the

term 9i03 becomes

ry h in the foot-poundsystemof units).
h
Limits of applications: (0.3 f0 0.1m < h < 0.6 m (2.0 f0
(0.6s ft) 0.2m ( s ( 2.0 m (6.5 ft)
Length between the screensand the weir: \ 80 . b.

Sv',fusSociety of Engineers' and Architects' .formula ( 1924):

1 \ | , h \zl
Q : o.+ro
(r + t oooi=F ri) t - o'5t,' -*) b \ti h't,
| l
Limits of applications: 0.025m < h {0.80m
s )0.30m
hls{l
I 3.28
(In the above formula the term becomes in the foot-pound
1 0 0 0 .h + 1 . 6 1 0 0 0 .h + s . 2 5
systemof units.)
Rehbockformula ( 1929):
a : (0.4023 -F 0.0s
42 h"ls)b . \,i tt,'i,
wher e: h. : ft + 0.0011m (h + 0.0036f0
Limits of applications
: 0.1m <.- /z ( 0.8 m
s )0.3m
hls{l
It must be noted that irregular flow in the approach canal, as well as a rounded crest, or a
rough upstream face of the weir, or unsatisfactory aeration of the underside of the nappe, may
alter sensibly the head-dischargerelation.

6. Absolute measurementby weight and volumetric gauging methods

6.I Flow measurement by weighing and by volumetric tank, preferably calibrated by weighing, are
among the most accurate methods, but generally are only convenient for relatively small rates of
flow, becauseof the size of collecting tank required. They are thus particularly suitable for tests
in the laboratory and on small installations but can seldom be used on larse turbines.
 129 -

6.2 The measurementof water flow by collecting in a tank for a measured time and weighing is an
accurate absolute method. With suitable methods of diverting the water into and out of the tank,
weighing, and time measurement,high accuracy can be obtained.

6 . 3 The volumetric method is the next most accurate to weighing, provided the capacity of the tank
can be determined accurately. Small tanks are best calibrated by filling with accurately weighed
amounts of water. Larger tanks and reservoirs must be calibrated from geometrical measurements.
In all casesthe walls of the container must be rigid so that the volurne is reliably determined for
each water level. Levels are measured by piezometer tubes (bore 20 mm [0.80 in]), hook gauges,
floats or measuring rods. In large reservoirs,water levels must be measuredat severalpoints simul-
taneously.

6 . 4 The following factors must be carefully guarded against:

- Deformation of the walls of the tank (for small tanks);
-- Rain, evaporation, wind, leakage or inflow (for large reservoirs).

7 . Travelling screen method

7 . 1 Apart from exceptional cases,flow measurement by means of a travelling screen is only used in
test laboratories, as the method requires relatively lengthy and expensiveapparatus. The method
consists of an accurate flow measuring device in which a long, open canal of uniform rectangular
cross-sectionis used to carry the flow and a light-weight vertical travelling screen with minimum
friction and leakage is used to measure its speed. Its merits are its simplicity and precision, but it
is only convenient for small flows.

7 . 2 The method is analogousto the volumetric method in that the dischargeis computed as follows:
A
o : L
t
where : Q : discharge in m3/s (ft3/s);
L : length of test track in metres (feet) between the marked sections;
A : water cross-sectionarea of canal in square metres (ftt);
t - measuredtime for screen to travel alons the test lensth in seconds.

7.3 The flow canai shall be open and of sufficient length to securea uniform flow of water. The wetted
walls and floor shall be perfectly straight and level respectivelyand the floor shall be maintained
free from sand and silt. The test length shall be visibly marked so that it can be checked. The
cross-sectionshall not vary in the test length, nor in the adjoining 2 m (6.5 ft) at each end by more
than 0.5\. Suitable pockets shall be provided at three points along the test length to provide
lacilities for measuring the depth of water. A stilling device shall be used to securea smooth water
surface free from waves or ripples. An effectivecontrol-gate, preferably automatic, should be used
to maintain the level of flow constant durins a test.
/

7.4 The travelling screen shall fit the cross-sectionalplane of the canal with a clearancebetweenfloor,
walls and screen edges of not more than 5 mm (0.20 in). It shall be made of rigid light-weight
material and suitably supported on a wheeled carriage running on a suitable track. The screenor
its carriage shall be fitted with an effective and reliable timing device. To avoid windage the top
edge of the travelling screen should never project too much above the smooth water surface. To
enable the screen to acquire the speed of the water quickly, there is no objection to providing an
initial accelerationto counteract the initial resistanceof the flowins water.
 7'5 The main track shall be so made and maintained that the suspendedscreen is free to move
trans-
versely with a maximum variation of 2 mm (0.08 in) in any direction from its true position
throughout its travel.
An auxiliary track should be fitted at the upstream end to retain, detach and introduce the
screensuspendedon its carriage to the test canal without the formation of surfacewaves. A similar
auxiliary track would facilitate the screen's exit and retention.

7.6 Desirable features

7'6'l Measurementof flow. The screen should be carried on a three or four wheeled frame equipped
with ball-bearing light-weight wheels of about 600 mm
Qa ifi diameter. The screenmusr travel
easily and there must be no tendency either to over-run or under-run the speed of the water.
The maximum difference in level between the water on opposite sides of the screen should
not
exceed0'2 mm (0.08 in). The maximum speed should be about 1 m/s. The water surface should
remain constant during the test flow, and this shall be verified before and after.

The length of the test canal should be at least r 5 m (50 ft).

The test length should be at least g m (26 ft).
The width may range from I to 3 m (3 to l0 ft).
The depth may range from I to 3 m (3 to l0 ft).

7'6.2 Measurementof speed. Electrical timing devices or chronographs are desirable

and they shall
be checked by calibration. Three or more timing contacts should be fitted at.equal intervals
in
the test lensth.

8. Pitot tubes

8'1 The method consists of determining the velocity by means of Pitot tubes at a sufficielt
number of
points over a measuringsection to determine by integration the rate of flow. It
is similar in principle
to the currentmeter method, but more limited in the flow range covered becausethe
reading of a
Pitot tube varies with the square of the water velocity. It is essentialto calibrate the pitot
tube.
The method is suitable for closed conduits running full, but is not recommendedfor
ca'als and
free surface flows.

8'2 To measure the rate of flow of water in a closed conduit, the measuring section shall
be located in
a straight length at a distance not less than ten pipe diameters from any upstream bend,
elbow,
branch-pipe, valve or other disturbance to uniform flow.

8'3 The Pitot tube shall be rigidly fastened to a movable rod and shall at all times point
upstream
normal to the plane of the section.

8 . 4 For Pitot tubes consistingonly of a singleopening impact tube, the static pressureshall be
obtained
by pressuretappings, installed flush with the interior wall of the conduit, and so located
as to be
unaffected by the supports (see IX.3.3).

8 . 5 Two or preferably four static pressuretappings shall be installed symmetrically around the measur-
ing section on the same transverseplane as the Pitot orifices. Each pressuretapping shall
be con-
nected to a separatemanometer. The static pressurehead at the measuring section shall
be taken
as the mean of all pressuretapping readingswhich shall be made simultaneously.
8.6 At least two Pitot tubes shall be arranged to traverse at least two diameters of the penstock. The
diameters at which traversesare made shall be equally spacedaround the penstock. Two complete
traversesshall be made alor.r-eeach diameter for each test run.

8.7 The velocity head hu for each point of the traverse shall be obtained by subtracting from the pitot
tube reading for that point the average static pressure head as determined above. The velocitv v
at each point of the traverse shall be cornputed by the formula:

Yi : C . \"2g ftvi
where: yi : the velocity at the point i ;
7,i : the differential pressure head at point i as determined above;
C : coefficient of a Pitot tube of the single-openingimpact type combined with pres-
sure tappings, so long as flow is not very turbulent, is generally near C - l.

8.8 Traversing and integrating methods are similar for Pitot tubes and currentmeters.

8.9 Pitometer-type instruments.

Instead of the single-opening impact type, the Pitot tube may be provided with an
attached or coaxial tube containing tappings for measuring static pressure head local1y. Pitot
and static tubes are connected to a differential manometer.
Alternatively, the second tube may have, instead of a static orifice, a trailing orifice pointing
downstream to create a certain suction, which assistsin increasinghu. Such varieties of Pitot tube
include Darcy, Frank, Prandtl, Cole pitometers. Their coefficientsmust be carefully established
by calibration over the range of velocities covered.

9. Orifices, nozzles and Venturimeters

9 . 1 The method of flow measurement by orifice, nozzle or by Venturimeter is the subject of recom-
mendations by the International Organization for Standardization. Whenever it is possible to
satisfy these recommendations it is unnecessaryto calibrate the apparatus as the coefficientsindi-
cated in the ISO recommendations may be used provided their resulting accuraeyis considered
sufffrcient.

9 . 2 If the conditions of use of such apparatus, particularly installation conditions, make it impossible
to satisfy all the ISO requirementsthe apparatus shall be calibrated on site under conditions of use.

9.3 The flow is given by the equation :

Q: a m A \rgt,
where : Q : r&ts of flow m'/s (fts/s);
A : cross-sectionalarea of the conduit in square metres (ftr);
tn : ratio of throat/main arca;
h : differential pressure expressedin metres (feet) of water column;
: discharge coefficient depending upon the type of apparatus, area ratio m, the
Reynolds Number and the absolutedimensions.
 - 1 3 5-

CHAPTER IX HEAD

l. References

l.l Moin fixed points, derivedfixed points, reference points

In each water power installation, a main fixed point or bench mark should
be provided, the
level of which must be accurately determined, preferably in relation to
some establishedsurvey
system (mean sea level). This main fixed point must be clearly marked
to avoid any possibility
of error.
The levels of the derived fixed points for free water levels and the referencepoints
for pressure
gaugesmust be accurately determined in relation to the main fixed point
prior to the start of the
acceptancetest.

1.2 Steady conditions and number of readings

Readings required to determine the turbine head shall be taken only after
steady conditions
prevail' The head during tests must be maintained as constant as possible.
Readingsmust be take'
at uniform intervals at a minimum of 5 readings for each load point;
the number of readings
and intervals between readings depend on the magnitude and the frequency
of the variations
(seeIV.2.2).

2. Free water level

2 . 1 Choice of measuring sections

For the choice of measuring sections in which a free water level must
be determined. the
following points must be considered:
a) the flow should be steady, and there should be no local disturbances;
b) the cross-sectionalarea used to determine the mean water velocity must
be accurately defined
and readily measurable.
The headwater level in installations comprising open canals or intakes
must be determined
downstream of the turbine trash rack. The tailwater level should be determined
in the section of
the draft tube outlet. If this is not possible,either a closed draft tube section
immediately upstream
from the end of the draft tube or an open section in the tailrace in the immediate
vicinity of the
draft tube outlet may be used. In the case of turbines operated without a
characteristicdraft tube,
the turbine outlet section must be agreed upon when the equipment is purchased
(see 1I.2.2.2).

Measurement of free water lever must be obtained for at least two points
of the measurins
section and the readings averaged to obtain the true level.

2.2 Measuring wells and stilling boxes

If the free water surface is not accessibleeither at the turbine inlet or at

the turbine outlet,
measuring wells about 300 mm (l f0 square, which permit accurate and
convenient measurements,
must be provided for at the planning stage. The connection canals or tubes
must be constructed
at right angles to the well of the measuring section and should preferably
be covered with smooth
perforated plates (perforations of 3.5 mm
[ru l/S in] diameter). Such cover plates must be flush
 1
I
I
-r37- I

with the wall of the measuringsectionto eliminate any velocity effects(Figure 40). It is recommended
that at least two measurin-s*ells be provided in each measuring section on opposite sides of
the canal.

Frc. 40. - Measuring well.

a) Perforated plate
b) Flushing valve

Where the water surface is accessible,but not sufficiently calm for accurate measurement,a
stilling box, similar to the measuring well, should be installed.

2.3 Measuring apparatus

2.3.1 plate gauge. A plate gauge, consisting of a metal disk suspendedfrom a steel-ribbontape, will
be found convenient to determine the level in relation to the derived fixed point in the measuring
section (Figure 41).

2.3.2 point or hook gauge. Pcint or hook gauges may be used to determine the level of calm water
for example inside stop-log grooves, measuring wells or stilling boxes (Figure 42).

2.3.3 Float gauge. Float gauges, properly calibrated and in good working order, may be used and
are recommendedwhere the water level is variable. The float diameter should be at least 200 mm
(nv 8 in). Such gaugesshould be sensitivewithin 5 mm (x 3116in) when manually displaced
from the true reading.

A float diameter of 200 mm (8 in) is consideredadequate for use with stilling boxes 250 mm
(10 in) square, which often is the largest size suitable for installation in stop 1og slots.

2.3.4 Fixed (board) scales. Fixed (board) scales,installed flush with the wall of the measuring section,
may be used where the head is greater than l0 m (32.8 ft).

2.4 Liquid manometers

If measuring sectionsare inaccessible,the free water level may be determined by means of two
or more liquid manometers. The pressuretappings at the wetted periphery of the measuring section
must be constructed in the same way as the connection of measuring wells (seeFigure 40), to avoid
velocity effects.
 - 1 3 9-

I
I
;

- =---: ---- -- -:-'

Ftc. 41. - Pl a te g a u g e for measuring free water level from a lixed point A.
a) Constant length
h) Steel measuring tape
c) Metal plate

Frc. 42. - Point and hook gauges.

Al Fired point
 -_l4l-

Ftc. 43. - Inaccessiblefree water level.

a) Reference vessel
b) Compressed air in differentiai manometer
c) Level to be measured

Ftc. 44. __ Inaccessiblefree water level.

a) Reference vessel
b) To vacuum pump
c) Level to be measured
 143 -

The liquid manometers are often of the differential pressure gauge type (inverted U-tube).
In this case, the second leg of the U-tube is connected to a reference vesselfilled with water in
which the water level is maintained at a fixed level. If the free water level to be measuredis above
the manometer, the upper portion of the U-tube must be unwatered by means of compressedair
(Figure 43). If, however, the free water level to be measured is below the manometer, the levels
in the two U-tube legs must be raised by suction (Figure 44). The connection tubes to the measur-
ing instrument must allow for ready purging to remove any air pockets, and to maintain the same
water temperature. They must be sufficiently air-tight to avoid leakage of air into sections below
atmospheric pressure.

2.5 Measurements by means of compressedair

The free water level may also be determined by means of the pressureinside a tube filled with
compressedair; one end of this tubing being connectedto a small compressoror to a pressurechamber
by means of a reducer valve while the other end is located at a well-defined level below the water ;
surface to be measured (Figure 45). As the pressureloss in the pressuretube supplying air falsifies
the measurements,this pressure loss must be kept to a minimum (< 5 mm or 3/16 in). For this
reason, the cross-sectionalarea of the tube should be liberally dimensioned, the tube should be
as short as possible and the air volume which flows constantly through the tube and escapesunder
water, should be as small as possible. The air tube should be introduced only in still water, other-
wise dynamic effects are liable to cause appreciable errors.

Ftc. 45. - Inaccessiblefree water level.

a) Compressed air
b) Tube end, at fixed level
c) Level to be measured

a
J. Pressure measurements

3 . 1 Choice of measuring sections

The following should be consideredwhen selectinga measuring section to obtain water pressure
measurements:
a) There should be a minimum of disturbance in the flow; sections where the velocity pattern is
seriously distorted by an elbow or valve should be avoided;
 -t45-

b) T he c r o s s -s e c ti o traalre a . * h i c h required for computing the mean water velocity must be

readily measurable.
The measuring section should preferably be arranged in a straight pipe section (which need
not be cylindrical)extending5 diametersupstream and 2 diarnetersdownstreamfrom the measuring
section.

3.2 lr{umber and location of' pressure tappittgs

Four pressuretappings arranged on two diametersat right anglesto each other should be used.
The connectionsshould not be located at or near the highest point of the measuring section in
order to avoid air pockets,and shouid not be placed at or near the lowest point becauseof the risk
of dirt accumulation in the pressureconnections. With favourable flow conditions. only two taps
located diametrically opposite rnay be used by agreement.

As the level of the measuringpoint will be assumedto be that of the centre of gravity of the
gr oup of c on n e c ti n gp re s s u reta p p i n g s ,th i s l evelshal l be that of the centreof the measuri ngsecti on.

3.3 Gauge tappings

The cylindrical bore of the pressuretapping opening in the conduit boundary should be 3
to 6 mm (1i8 in to ll4 in) in diameter and should have a minimum length of 6 to 12 mm (ll4
in to ll2 in). It must be drilled at right anglesto the pipe wall, and to avoid velocity effects burrs
shail be removed and the bore shall be rounded off by a radius 4 dl4 (Figure 46), or at least
a small chamfer.
T he s ur l a c eo f th e c o n d u i t s h a l l b e s mootl -rand paral l elw i th the fl ow i rr the vi ci ni ty of the bore
f or at leas t 2 5 0 m m (1 0 i n ) u p s tre a ma n d 100 mnr (a i n) dow ' r-tstream.

d= 3-6nm
l = zd= o-/zrnn
r=d14

Ftc. 46. - Pressuretapping.

3.4 Gaugepiping

Each connection, to allor,v individual measurement of the pressure at each tap, must
be connected to the pressure gauge through a separatepipe which can be closed off individually
(Figure 47). lf a ring manifold is used, a shut-off valve must be provided between each connection
and the manifold (Figure 48). The piping should be continuouslyrising without any loops or sags
which are liable to trap air during the measurements.In addition, it is necessaryto provide for
purging air from the piping, So that a flushing valve of adequate size shall be provided at the top
of the piping system,close to the pressuregauge. Piping should carefully be checked to avoid any
 -r47-

leaks, no matter how small. If leaks develop during the tests,the tests should be stopped until the
leaks are repaired [transparent conduits of plastic material have been found most satisfactorYfor
use with pressuresup to 6 kpi'cm' (90 lb/in')].

Frc. 47. - Pressuretappings connected through separatepipes to manometer.

a) Purging
b) Pressure gauge

Frc. 48. - Pressuretappings connected through ring main to pressure gauge.

a) Purging
b) Pressuregauge
c) Drainage

In the proximity of the pressure gauge a fixed reference point must be provided, so that the
level of the zero point of the measuring instrument can be determined.
 r29 -

for a measured time and weighing is an

6.2 The measurement of water flow by collecting in a tank
the water into and out of the tank,
accurate absolute method. With suitable methods of diverting
weighing, and time measufement, high accuracy can be obtained'

provided the capacity of the tank

6.3 The volumetric method is the next most accurate to weighing,
by filling with accurately weighed
can be determined accurately. Small tanks are best calibrated
from geometrical measurements'
amounts of water. Larger tanks and reservoirs must be calibrated
the volume is reliably determined for
In all casesthe walls of the container must be rigid so that
(bore 20 mm [0.80 in]), hook gauges,
each water level. Levels are measured by piezometer tubes
be measuredat severalpoints simul-
floats or measuring rocls. In large reservoirs,water levels must
taneously.

6.4 The following factors must be carefully guarded against:

- Deformation of the walls of the tank (for small tanks);
-_ Rain, evaporation, wind, leakage or inflow (for large reservoirs)'

7 . Travelling screen method

a travelling screen is only used in
7 . 1 Apart from exceptional cases,flow measurement by means of
expensiveapparatus' The method
test laboratories, as the method requires relatively lengthy and
open canal of uniform rectangular
consists of an accurate flow measuring device in which a long,
travelling screen with minimum
cross-sectionis used to carry the flow and a light-weight vertical
are its simplicity and precision, but it
friction ancl leakage is used to measure its speed. lts merits
is only convenient for small flows.

dischargeis computed as follows:

7 . 2 The method is analogousto the volumetric method in that the
L A
Q :

wher e: n : discharge in m'/s (ft'/s;;

t - length of test track in metres (feet) between the marked sections;
, t _
,17. water cross-sectionarea of canal in square metres (ft');
measuredtime for screen to travel along the test length in seconds'
a uniform flow of water' The wetted
- a
The flow canal shall be open and of sufficient length to secure
and the floor shall be maintained
walls and floor shall be perfectly straight and level respectively
so that it can be checked' The
free from sand and silt. The test length shall be visibly marked
2 m (6.5 ft) at each end by more
cross-sectionshall not vary in the test length, nor in the acljoining
points along the test length to provide
than 0.51. Suitable pockets shall be provided at three
be used to securea smooth water
facilities for measuring the depth of water. A stilling device shall
preferably automatic, should be used
surface free from waves or ripples. An effectivecontrol-gate,
to maintain the level ofl flow constant during a test.
.,

canal with a clearancebetweenfloor'

7 . 4 The travelling screen shall fit the cross-sectionalplane of the
shall be made of rigid light-weight
walls and screen edges of not more than 5 mm (0.20 in). It
on a suitable track' The screen or
material and suitably supported on a wheeled carriage running
device. To avoid windage the top
its carriage shall be fitted with an effective and reliable timing
the smooth water surface' To
edge of the travelling screen should never project too much above
there is no objection to providing an
enable the screen to acquire the speed of the water quickly,
flowing water.
initial acceleration to counteract the initial resistanceof the
 149 -

3.5 Pressure gquges

3.5.1 Manometers. Liquid column gaLrges are used to measurecomparativelylow pressures.In order
to avoid capillary effects,the tube of the water column gauge must have an inside diameter of
at least 12 mm (ll2in) in the measuringrange; with mercury column gauges,this diametershould
be at lea s t 8 m m (5 /1 6 i n ).
Mercury pressuregaugesare available as:
a) U-tubes. The height of mercury columns in the ti.vo legs must be read off simultaneously
(Figure 49).
b) Measuring pot ruith stand pipe. In this case, only the height of the mercury column in the
stand pipe need be read, the small variation in the mercury level inside the measuringpot
being accounted for b5za multiplying factor almost equal to unity, to allow for the ratio
of areas of the stand pipe and pot (Figure 50).
The diameter of the measuringpot should be at least l0 times the bore of the stand pipe.
c) Calibrated U-tube. The height of the mercury column is read from only one leg, while the
displacement of the mercury in the other leg (whose cross-sectionalarea is, say l0 times
that of the measuring leg) is measured by a suitable separatescale.

For all these instruments only pure mercury may be used, which at certain intervals must
be cleaned. Removal of foreign material by bubbling of air through the interface betweenmercury
and clean water in a dish often is effective"

Ftc. 49. - U-tube rnercury differentialrlanorneter.

a) Water
h) Air
r') Mcrcury
d) Zero point
e) Measuring point level

1 [ I
ps:10 :(i'ue-'[l hj l
l't.f
I
(metric)
 -- l5l -

Flc. 50, - Mercurymeasuringpot as differentialmanometer,

a) Water
b) Air
c) Mercury
d) Zerc point
e) Purging
I | . l
Pe: t0 +vgr'ftJ
lr.l (metric)

3.5.2 Dead weight manometers. Dead weight manometers may be provided with simple or differential
piston. The effectivepiston diameter d" of the former may be determined as the arithmetical mean
value of the directly measureddiameter dp and bore diam eter da and used for pressurecalculatiol
without further calibration :

dnidr o: n:
d e : if z. o.t%
d a l d p

This quantity must be checked before starting the acceptancetest.

Generally it is preferable, however, to calibrate a dead weight manometer and determine

thereby indirectly its effectivepiston area over the whole measuring range against a simple tiquid
column manometer.

The friction between the piston and bore can be practically eliminated by slowly rotating
the piston.

A dead weight manometer which is connectedin serieswith a mercury manometer is suitable

also for measurementsof varying pressures,for which purpose dead weight manometers alone
are not practicable (Figure 51).
 1 5 3-

p
' v ,t

ZM

t v f
Z

Ftc. 51. - Dead-lveightrnanometer,combined with mercury pot manometer.

a) Weights G
b) Mercury
c)Water
e) Zero point
f) Measuringpoint level
d Oil filled
i) Purging valve
l l
Ps: (v.k Y H s l t )- Pc- \ si1 ' h,
l0 l0
G
whcrein I.' G : i t
0e-
4 (rnetric)

3.5.3 Spring pressuregauges. For acceptancetests oniy precision spring gaugeswith pointers on knife-
edge supportsmay be used,so that the samepointer positionsfor identical pressuresare obtained
by calibration against weight manometers. The optimum measuringrange for a spring pressure
gauge is between 50 and 70 % of full scale value.

Spling pressuregaLlgesmust be calibrated against reliable standardssuch as weight lnano-

meters,in the same positiorl and over the same range as used for the test, before and after the
tests.

3.5.4 Dffirential manometers. With installations comprising several penstocks,the pressurein the
measuringsection of the water passageof the turbine under test should be determinedfrom the
pressuredrop as compaled lviih DressLlre in a secondturbine penstock while the secondmachine
is shut down. The differenceiu pressurein the two penstocksmay be determinedby means of
a mercury U-tube (Figure 52) or with an inverted U-tube filled with a sealingliquid that is lighter
than water (Figure 53), when small diflerencesin pressLrre are involved. Recommendedsealing
liquids for this purpose are toluene with a specific weight of y : 0.867 kp/dm3 or compressed
air with an appropriate specificweight.

The loss of head should be corrected by the differencesof the water levels at the penstock
entrances.
 -l

-155-

a
IL
rl nt

Frc. 52. - Mercury differential manometer.

a) Water (y)
b) Mercury (ygu)

Hr * : (vue - v) . A/r
::
zg
Hl : Pressureloss between forebay level and section of pressure measurement.

Frc. 53. - Differential manometer with inverted U-tube.

a) Air (ya)
b) Water (y)
0

Hr t: : (y - yo) rft
zg
Hl : Pressureloss betweenforebay level and section of pressurerneasurement.

3.6 Damping devices

When the pressureto be measured is pulsating, it may be difficult to obtain correct readings
on the pressuremanometer. In order to improve such conditions, suitable damping may be pro-
vided. This requires special care, however, becausea proper damping device depending on viscous
resistanceshould be fully symmetrical with equal resistanceto flow in both directions. This may
be secured by use of a capillary tube or a speciallydesigned plug valve; the latter providing for
easy flushing.

For pulsating spring pressuregauges,it may be desirable to connect an air or surge chamber
in the pressure line ahead of the gauge.
 -157-

3.7 Calibration of manometers

3'7'l Comparisonof readingsw'ith the static pressurepo. Before and after the acceptancetests, the
read-
ings of the pressuregaugesshould be compared with the static pressurepo in the measuring
section
of the turbine supply conduit for e - 0.

For calculation of po, the difference in barometric pressuresacting on the head water level
and at the pressuregauge should always be taken into account; with very high heads it is further-
more necessaryto consider the effect of water compressibility.

The static pressurepo at standstill, as gauge pressurein kp/cm2 at the measuring point (m)
z
above sealevel and with a barometric reading of h6o (mm) mercury column, is given in the
follow_
ing formula for a head water level at elevation z" above sea level and at a barometric
reading ha
mercury column (Figure 54).

Frc. 54. - Calculation of pn.

Po + Lpo
1oQs-z)

T
'0 o : 1 0 ( z w - z ) + aP"

z
(l ta - ha" ) PH
where: Ap o : r.3 5 4 6 _. o.l v@) dz
I 000 I
, l
Z

For a location z up to about 3 000 m above sea level and for a water temperature
of 10 "C and
g: 9. 81 m/s 2 th e c o rre c ti o n te rm A p o as a functi on of (2, - z) is given in the table
below:
 - r 5 9-

Apo (kp/cm':)

200 -0.0147 - 0.0114 - 0.0094 - 0.0071

400 - 0.0088 - 0.0031 0.0020 0.0061
600 0.0178 0.0263 0.0335 0.0399
800 0.0651 0.o769 0.0856 0.0940
000 o.1327 0.1464 0 . 15 8 2 0.1686
200 0.2212 0.2371 0.2s12 0.2634
400 o.3298 0.3486 0.3645 0.3788
600 0.4s9s 0.4804 0.4986 0.5146
800 0.609s o.6328 0.6530 0.6709
2 000 0.7802 0.80s7 0.8279 0.8475

The accuracy of po dependsmainly on the accuracy with which the elevations z and zrhave
been determined.
From po and the values p measured during the acceptancetests for various Q values the
head loss F1, can be determined as a function of Q:
10
H t - - ( P " -p)
) r - : _I

3.7.2 Comparison with other measuring instruments. Spring pressure gauges must be checked over
their whole measuring range against other measuring instruments of higher accuracy such as
dead weight manometers.

4. Vacuum measurernents

4.1 General remarks

For acceptancetests carried out on water turbines, vacuum measurementsare rarely required.
Such measurementsmay, for example, be required in installations without a characteristicdraft
tube. In such cases,a section of the dischargepipe must be agreed upon as representativeof the
outlet cross-section,in accordancewith Sub-clause1I.2.2.2.

For vacuum measurementsall remarks made in section IX.3 are applicable,except as noted
in the followine clausesIX.4.2 and IX.4.3.

4 . 2 Toppings
A tapping of 6 mm Qla h) diameter should be provided for the vacuum measuringconnections.

4 . 3 Gaugepiping
The gauge piping must either be completelyfilled with water or, if air is used, shall be trans-
parent to permit observation of the water level. Such pipes, when filled with water, must carefully
and frequently be flushed with water between test runs to remove any air coming out of solution
or leaking through the connection, and to maintain the water in the gauge pipe at the same
temperatureas at the penstock. The gaugepiping and its connectionto the gauging instrumentsas
well as the measuring connections to the pipe must be air-tight.

Flexible hoses may be used as gauge pipes only if they cannot be collapsed by atmospheric
pressure.
 t6l

CHAPTER X MEASUREMENT OF POWER OUTPUT

l. General

The power on the turbine shaft can be measured:

Either directly by means of brakes (mechanical, hydraulic or electric) or by means of dynamometers
(usually torsion dynamometers);
or indirectly, when the turbine is coupled, or can be coupled, to an electric generator, by adding
to the electric power output of the generator the various electrical and mechanicallossesspecified
in Sub-clauseII.2.J.2.

2. Indirect methods of power output measurement

In the indirect methods, the electric generator coupled to the turbine acts not only as an absorption
brake but also as a device for measuring the power. Becauseof this, its efficiencymust be determined
accurately under the same conditions of load, voltage, power-factor, speed,temperature, etc., as occur
during the testing of the turbine. This efficiency may therefore vary slightly from that obtained for
standardizedconditions.

2.1 The efficiency of the electric generator shall be determined taking into account all of the losses
specified in I.E.C. Publication 34-2 (1960), Recommendations for rotating electrical machinery
(excluding machines for traction vehicles). Part 2: Determination of efficiencyof rotating electrical
rnachinery,So long as that text does not contradict the stipulations of tl-represent code.

When the above-mentioned document does not give details of the procedure to be followed
for these measurements,the procedure shall follow the regulations of one of the official national
codesin existence,chosenby common accord betweenthe parties at the time the turbine is ordered.
Generally, the efficiency of an electric generator is determined by indirect measurement,con-
sisting essentially of measuring the aggregatelosses of the machine (total losses),and in assuming
that these lossesrepresent the difference between the mechanical porver absorbed and the electric
power output. In practice,the determinationof the above " total losses" is generallyaccomplished
by the individual measurementor calculation of the various categoriesof lossesand by their sum-
mation. Thus, this method of measuring the efficiency is said to be " separate losses".
Two main methods of measuring lossesare currently used. The deceleration(retardation)
method and the calorimetric method based on the heating of the cooling fluid. The deceleration
method shall be performed in accordancewith I.E.C. Publication 34-2 (1960),Appendix II, page 33.
For the calorimetric method see XI.4.

2.2 Measurement of effective pov:er at the generator terminals

Measurement of the effectiveoutput must be made at the generator terminals if at all possible.
If this cannot be done, the measuredpower must be increasedby the lossesoccuring betweenthe
generator terminals and the location of the measuring device.
In the absenceof an international code for electric measurements,these power measurements
shall be made in accordancewith one of the national codes in existence,chosen by common accord
between the two parties when the turbine order is placed, and taking into account the following
obligat or y re g u l a ti o n s .
 -1

- 163-

During the turbine tests the -eeneratormust operate,insofar as possible,at about its nominal
vgltage, and in the case of an alternating current machine at its nominal frequency and at a power-
factor equal to unity. If theseconditior-rscannot be fulfilled, due allowance shall be made in calculat-
ing the lossesand correspondingoutputs, and in estimatingthe probable error in the measurements.

The electric power shall be measuredby means of wattmeters or by watt-hour metersof precision
class0.2 and shielded;they shall be calibratedas indicated hereinafter. The instrumenttransformers
feeding the meters shall also be of precision class 0.2.
If the neutral of a three-phasegenerator is brought out and is connectedto the network or to
earth, it will be obligatory to use the three-wattmeter (or watthour-meter) method (Figure 55).

Pl r2 R(Unln)
S(UsIs)

Frc. 55. - Measurementof the power at the terminals of a generator,three-wattmetermethod.

I Potential commutator
II Current commutator

If the neutral of the generatoris brought out but not connectedto tl-reuetwork, or to earth
during the test, either the three-wattmeter (or watthour-meter) method or the two-wattmeter (or
watthour-meter) method may be used. The three-wattmetermethod is recommended. If the neutral
of the generatoris not brought out, the two-wattrneter(or wattl-rour-meter) method shall be used
( F igur e 56).

U R (ln)

E__
W
Uns
r T(Ir) Usr

Frc. 56. - Measurement of the power at the terminals of a generator, two wattmeter method.
I Potential commutator
II Current commutator
 1I
I
_ 1 6 5_
II
I

It may be necessaryto allow for temperature correction in the above instruments. In the case
of very great temperature fluctuations, it may be advisable (particularly for watthour-meters) to
place the instruments in a constant temperature enclosure.
In addition to the measurementof output at the terminals, voltage, current per phase,excitation
voltage and excitation current shall be measuredby suitable precision instruments. The temperature
in the stator winding and other factors necessaryto establishthe lossesand output of the machine
under the test conditions shall also be measured.
If the power is measured by direct-reading instruments, the number of readings will depend
upon the duration of the test and fluctuations in the load. The number of readingsshall be sufficient
to permit as accurate a calculation as possible of the mean power over the duration of the test.
It is recommended that the various wattmeters used in the measurementof the power should be
read simultaneously.
If the power is measured by watthour-meters, they shall be used simultaneouslywith direct-
reading wattmeters and checked against them in the course of each measurement. The duration
of the recording of the integrating devices shall be measured by means of stop watches or other
time-measuringdevicessufficientlyaccurateto permit the determination of this factor to an accuracy
of at least0.1%. The power shall be measuredduring the sameperiod during which the flow through
the turbine is measured;for example,if the pressure-timemethod is used,the power shall be the mean
of the readings taken during a period immediately preceding the recording of the " pressure-time"
diagram.
Because of the possible inaccuracies in shunts used for measuring direct currents of high
amperage, the degree of accuracy obtained in the measurement of direct current power may be
considerably lower than that achieved in the measurement of alternating current.
At a date as close as possibleto the start of the tests,the measuringinstruments to be used shall
be specially calibrated over the entire range of readings expected, by an organization chosen by
agreement between the parties.
In the course of the turbine tests, every effort shall be made to compare as often as possible
the readings of the various instruments among themselvesto check that they remain in calibration.
In casesof dispute, a second calibration shall be effected as soon as possible after the end of the
tests.
Becauseof the accuracy required for the electric measuring instruments (including the instru-
ment transformers), the usual switchboard ones are not to be used for acceptancetests. Similarly,
the instrument transformers supplying the wattmeters, watthour-meters, or other electric measuring
instruments specially installed for the acceptancetests, shall be used solely for that purpose and
may not be used to supply at the same time the measuring and protective circuits, etc. of the plant.
An exception may be made to this rule only if, by agreement between the parties, it is decided,
and possibly checkedby calibration, that the same degreeof accuracy can be obtained with the use
of certain switchboard instrument transformers (in view of their precision class,their effectiveload
and their conditions of use),as could be obtainedwith the use of the independentset of instruments
and transformers specially reserved for the test specified above.

3. Direct method of power output measurement

3.1 The direct method of measuring power on the turbine shaft by means of a brake is relatively easy
and accurate only in the case of small units of low capacity. lt may have to be used, however, for
turbines which are not intended for direct coupling to an electric generator.
 167 -

The torsion dynamometerseelnsto be more generallyapplicable,but in practice the accurate

calibration of such an instrument for high torque imposes very difficult problems.
the speedof the machinewith accuracy(seeChapterXID.
to measLlre
ln either case,it is necessary

3.2 Measurement by means o.f a bralce (mechanical, hydraulic or electric:)

These measurementsentail the accurate determination of the following features: l

- speed of rotation of the brake shaft;
- length of the lever arm of the brake;
- tare weight of the brake.

The brake shall be mounted ir-rsuch a way as to minimize axial or lateral thrusts on the turbine
shaft and bearings. The power absorbedby friction in these bearings,becauseof the existenceof
such loads, cannot be charged against the turbine but must be placed to its credit in calculating
output.
If the turbine has a vertical shaft, the brake must be suspendedso as not to inrposeany bending
stresson the shaft of the machine.
Care must be taken to ensure that the brake cooling fluid does not cause a parasitic torque
reaction; should it do so, it must be taken into account in the calculations.
Mechanical braking on a secondary shaft driven by a belt or a cable is dangerous ar-rdshall
not be used.
Every effort should be made to measure simultaneously and over the same time interval the
hydraulic input to the turbine and the balancing brake load.

3.3 Measurement by means of a torsion dynamometer

A torsion dynamometer may consist of a special instrument introduced into the mechanical
transmissiondriven by the turbine or of a length of the turbine shaft itself, the torsion angle of
which is measuredunder the action of the transmitted couple.
In either case, the apparatus must be calibrated before the start and at the end of the tests,
if possible in situ.
 -169-

CHAPTER XI MEASUREMENT AND DISTRIBUTION OF LOSSES

l. General

equal to the power absorbedby the

As stated in IL2.7.2, the turbine output Pt shall be considered
electric generator shaft Plus:
the output of the generator;
a) the various losses not taken into account in calculating

the power absorbed by certain accessories'

b)
source of error, all necessarymachinery
ln order to simplify the tests and to eliminate every possible
possible during the course of the tests'
driven by the turbine shaft shall be disengagedinsofar as
efficiency,which shall, if considered
The lossesof the electric generator will be calculated from its
special test. The test method will be chosen
necessaryby one or other of the parties, be the object of a
by agreementbetween the supplier and purchaser of the turbine'
power station. In the first case,all losses
Tests can be made in the manufacturers' works or in the
l during the turbine tests. Special care must
are known and can be adjusted to the conditions existing
testsare similar to those in the power plant'
I be take' to ensurethat the windage lossesduring the work

the generator test' Failing this, the

In the second case the turbine runner should be removed for
losses; for further particulars see
turbine runner maY be allowed to turn in air producing windage
X1. 2. 4. I . ln both c a s e sthe deceleration method may be applied'

method may be applied'

If it is impossibleto run the turbine runner in air, the calorimetric

2. Mechanical losses

2 . r Measurement of various mechanical losses

insofar as possible through the
The lossesin the thrust and guide bearings shall be measured
measurementis not possible,the
flow and the heating of the cooling fluid used for them. If direct
on the basis of existing empirical
lossesshali be estimatedby agreement,either by comparison or
formulae ?).
governor,etc., driven by electric
The power absorbedby the various cooling pumps, tachometer,
by thesemotors. In more difficult
motors shall be determinedby measuringthe electricpower used
their flow and delivery pres-
cases,the power absorbed by pumps shall be estimated by measuring
instance, by test-stand results'
sure, taking into account their overall effi.ciency,as obtained, for
power absorbedby other accessoriesmay in most casesbe estimated,in view of its small magnitude;
considered insufficiently accurate,
in the very exceptional caseswhere such an estimate might be
should be made'
direct measurementsby means of a torsion dynamometer for instance,

2.2 Lossesfrom v'indage oJ' open flywheel

by means of the follow-
Windage lossesfrom open flywheels can be calculated approximately
ing formula:
P v : 0 . 3 5 n t D u 5 ( l + 1 . 8B I D ^ ) . l o - e
?) See, fot' examPle, ASEA Journals5 and 6, 1947.
 - l7l -

-- the windage loss in kilowatts;

where 1 Pv
minute;
n the speed of the flywheel in revolutions per
the outside diameter of the flywheel in metres;
Du
B the width of the rim in metres'

2.3 Lossesin main mechanical transmissions

locatedbetweenthe shaft bearing,the turbine
The power lossesin thoseparts of the transmission
power output (an electric generator, for instance)
wheel and the system fo, -eu.uring the effective as
insofar as possible, under the same conditions
shall be the object of a separatemeasurement, guide bearings'
in a step-up gear' or in thrust or
those for the turbine acceptancetests. The losses
absorbed by the lubrication oil (or by its cooling
etc. may be determined by measuring the heat
water), and by the surrounding air'

b)' the turbine runner turning in air .for certain

2.4 Correction to take into account the power absorbed
methods of measuring generator losses
the turbine from the generator which it drives
It is often impossible or undesirableto uncouple
generator itself. In such cases,certain test methods
during testsmade for measuringthe lossesin the
generator as a motor without load' a deceleration
(such as those comprising the operation of the "'
taking into account of certain " parasitic losses
test without load and a short-circuit test), entail the

compared to the friction and windage losses

often these parasitic losses are relatively small
to take them into account by means of a suitable
of the generatoritself,,but it is neverthelessdesirable
correction.
in three groups:
The parasitic losses referred to above may be classified
wheel turning in air;
a) Windage and friction lossesof the turbine
of the weight of the turbine runner;
b) Increasein thrust bearing lossesby reason
c)Lossesinthebearingorbearings'andpackingoftheturbine.
of
turning in air. windage lossesPu in kilowatts
2.4.1 Friction and windage lossesoJ'ttte turbine runner applicable
calculated by the following formula,
a runner turning in air can be approximately
t o all t Y Pe so f tu rb i n e s :
\f
l a

P,.: (,
d
in which:
at 20'C);
1 ulg is the specificmass of air (0'132 kg per mB
on the type of turbi ne;
K, a coefficient determined by tests and depending
n : the speed of the turbine in revolutions per minute;
runner' for Kaplan and Pelton
D : the external diameter, in metres, of the turbine o'
the throat " of the runner'
runners ; for Francis turbines, D is the diameter of

Tlre coefficient K, can be evaluated as follows:

A. Kaplan turbines (blades in closed position)

Kur:ryk to ,:%"
seals)
B. Francis turbines (without wetting of the labyrinth
K"r is given bY the formula:
I-:Air | /n" 140'tz
K,r : -{' t + o'03
\ t* /
Lo
 _r73_

in which :
n . Pro't
t'ts : which is the sPecific sPeed of a Francis turbine;
H:.25
Pt : the power of the turbine at maximum efficiency in horse Power;
Hn : flet head on the turbine, in metres;
n : speed, in revolutions Per minute'

C. Pelton turbines e)
In the absenceof housing, the following formulae can be applied:
For integrallY cast runllers
l B u \ n
4"o:0'01 +12'40 (,, I

F or w h e e l sw i th d e ta c h a b l eb u c k ets:
'\
,' Bu
Kuro : 0'035 + 0'25 \ l
I "'
I ln these formulae :
I Bu : maximum width of bucket, in metres;
Da : maximum diameter of wheel, in metres'
for-
The lossesfor a similar wheel under a housing are found from those calculatedby the
coefficient K,pf K,po smaller than unity. The relative
mulae above by multiplying them by a
in metres,
proportions of the housing are characterizedby the dimensionsBi and Ri, measured
part and the index u
as indicated on Figures 57 and 58. If the index o is relative to the upper
to the lower Part of the machine:

* Bi^-
bL;'-lb
)b?l

Frc. 57. - Dimensions of the housing for Pelton turbines with horizontal runner axis'
.lRio Distance between the axis and the highest point
of the housing
JRiu Distance between the axis and the tail water level;
actually only down to the rounding of the diver-
gent to the lower half of the housing
(usually Ri,riRio )2)
fiio or ,8i., in the highest and the lowest point of D,
(opposite ,Ba).

-r) .. Ventilations-Verlustevon Freistrahlturbinen-Laufrddern". Bulletin Schweiz. Elektrotechn. Verein'

Gr*""" -
1956, o. 9.
N
 -r75-

Frc. 58. - Dimensions of the housing for Pelton turbines with vertical runner axls.
lRio Riu at the height of the middle of the wheel
I
bio on D, (opposite Bu)
6;u practically to the tail water level

Bi b Bi (Bi - B^)
wit h ;Da :
Du Bu 2D^
and Rr - Ru :
Ri 0.5 Du
the values of K"r/Ku"o ur. given by Figures 59 and 60 for single piece runners and by Figures
61 and 62 for runners with detachable buckets. The average coefficient is calculated by the
followins formulae:

/.0
___i K,; - f n-naI -
I
. - - !
=;,-F--:
-,----+ h/ | t t x a .
_\_ - ll --i;1 -_: t_!tUp^)=-;;i- - - ;= _ _
\ : : .: r :
-
u l
/
/ l
l
=--f----] I
-trti---
U,OU'.
----1--it t
_u.tu l_

nl
I -;=
t-_I:: 0.70i

\:
'./,
71
/--1
'---1.
--1

-----
-_ --*
-
--]
l
n E 4i i
u.vu

n 6nl
-:=
lo' -7
---.----l
r l

ffi
::. : - : = = _ i U.rl
nqn

I l-'-l
U

vT
- l
-:-: - - - - - - U.+J
--_
S
' l

a0/ , -T
t-
I i _
. 1
I
0.003
n1 0.? 0.3 0.1 0.5 0.6 0.7 0.8 0.9 /.0 /.7 /.Zn 7.3
frt-fro

Frc. 59.- Influence of the housing with integral cast runners.

- measured values
*'
l'"""';fl"
il;-; J* "1. J'
 -111-

6 ,vp
7 K
^upo

+,
l
I
+

t ^
I o.uo
\.J't LU
qs' 0.70
ooi

0.7 0.2 0.3 0.1 0.5 0.0 0.7 0.8 0.9 /.0 /./ l.zm/.3
frt- fro#

Frc. 60. - Innuence rl'r* t j t cast runners'

1 ":::: :..: i:, l,t"' "
extraPolated values
both for K"plKvpo : constant

t.0 ---i :=1--: Krit

61 fff
-=
'
-?f =.{: U.dU
+ - -t
= _0.u_
I - f
----..t
.--t- -.1 nan
U.IU

T=j -t 0.65

IJ,2 I

,#
..- i--1 u.ba
I
n6E
0.1
--i =-*-: 0.50
,
\---+
II 0.15
I
/ 0.u
I a \
\-1

S 435

a0/
| ,/ o.i,o't'
i z T r I\
: t I \ il 0J0
rr-
fi ]t- rl
0.25

t r r t t '
Q.QQltt
0 0.1 0.2 o,l 0.,, 0.5 0.6 0.7 as 0'9 /'0 t't l'2n l3
fli - fta--

Frc. 61. - Influenceof the housingwith runnerwith boltedbuckets.

measured values
extraPolated values
both for KuplKupo - constant
 1t t7 t 0 -

--+---
:_T_:
0./, 0.5 0.6 0.7 0.8 ag /.2n /.3
fti- flo

Ftc. 62.- Influence of the housing for runner with bolted buckets'

- - ---: ffil:::LJ';;i,*,
'- constitnt
both for K'p/Kp,,

Peltott tt'ith horizontal sha./i:

(K"rlKur.,)n, : 0'5 (K,rll(up,,) lower
-, 0.5 (K " r/K rn,,)LrpP er

Pelton t+'ith vertit'al sltaJ't:

(K,rlKuno)n',: 0.3 (K.,n/Kur.,)lower
-r 0.7 (K ,./K " no) uP P er

the windage losses,Pu, of the runner turn-

In certain cases,for Kaplal anci Francis turbines,
power absorbed by the same runner turning in
ing in air, can be determinedby rneasuringthe
by the generatorunder theseconditionsrepre-
water, the gatesbei'g closed. The power absorbed
runuer turnilg in water'
sents,less its own losses,the power lost by the

and kinematic viscosity of air and water, the

Allowing for the differencein specificweight
fromthe val ue thus obtai nedby di vi di ng i t by 500'
los s eso f th e ru n n e r tu rn i n g i n a i r a r e deduced

tttrbitte
bearing o-f ttte generator clueto the vt'eightoJ the
2.4.2 Correctionfor Jriction lossesin tlt tltrust
of measuringthe generatorlossesby the
(itt the caseof a unit ylithvertical haJt). In tl"recourse
jointly by the weight of the generatorand of
method describedabove,the thrust bearing is loacled
and its shaft increasedby the hydrodynarnic
its shaft, Gr, and by the weight of the turbine rlulrler
therefore G' * G'' As mentioned previously' the
thrust on the runner, Gt. The total load is
preferably by the heat dissipatedby the
friction lossesin the thrust bearing should be measured
However, theselossesmay be calculatedwith
cooling fluid and through the walls of the bearing.
of the oil are known' as follows :
sufficient accuracy if the temperature and viscosity

For a bearing the segmentsof which have:

- an outside diameter do, atd
- an inside diameter di

SO that the radial width of a segment 1 3 :

(in metres)
b : 0 . 5 (d" _ di)
 and its tangential mean length:
0.85 ;r (d" + r/t)
I - -
2 N
(1/ being the number of segtnents,generally between 8 and 10); the friction coefficientis given
bv the formula:

f :25\/#
in which :
'4 is the coefficientof absolute viscosity of the oil (in kp.s.m-2)

v''-'is the mean speed of the thrust bearing runner :

nn (ds :, di)
vm: (inm/s)
60x2
p is the specificpressureon the thrust bearing:
G.+Gt
p : ^, ,, , (in kp/mr)

The correspondinglossesby friction can be expressed:

Pu (r * t) : (G. + Gr) f . rm . 9.81 10-3 (in kW)

If the loss consideredabove is to be divided pro rata between the loads G. and Gt imposed
on the thrust bearing by the generatorand the turbine respectively(accordingto 1L2.7.2),it will
be allotted:
G,
- to the electric generator: Pb, : PaU+ t)
G'+Gt
Gt
- and to the turbine: Pa, : Pu(r * t)
G'+Gt
Note. - The lossesin the turbine packing on the turbine shaft may be neglected.

2.4.3 Friction lossesin the guide bearings of the turbine. These losses can be calculated bv the follow-
ing formula:
1.6rn
1zdrl -4
Pea.: -,. - . l0 (in kW)
()

in which:
"q : viscosity of the oil 1in kp.s.m-2);
d : the diameter of the journal in metres;
/ : the length of the journal in metres;
d : the total diametral play in metres.

3. Water consumption by auxiliary equipment

3.1 All water consumption and corresponding operating pressuresfor the auxiliaries and instruments
necessaryfor the proper operation of the turbine must in all casesbe specifiedat the time the order
is placed, and they may be verified in the course of the acceptancetests.
 183--

fl ow requi redfor the operati onof the el ectri c

3 .2 A c c or dingt o th e s ti p u l a ti o n so f S r-rb -c l a u seII.2.1.1,the
generator,the auxiliary machinery and for the cooling of the bearingsshould not be added to that
used by the turbine; in particular, if this water is taken front the turbine flow downstream of the
measuringpoint, it rnurstbe measuredseparatelyin the coLlrseof the acceptancetestsfor the turbine
and deduc t e dfro m th e fl o w d e te ri u i n e da t the measuri ngpoi nt duri ng the courseof the test.

4. Determination of the efficiency of an electric generator by means of the heating of the cooling fluid

4.1 General

ln machinesdesignedto be cooled try the circLrlationof a fluid brought in from the outside,
the greatestpart of the lossescan be determinedby measuringthe quantity of heat carried away
by the fluid.
entails the measurement,once thermal equilibrium is reached,
Ip all cases,this detern-rination
of the differencein temperatureof the fluid at the inlet and outlet. Of the two methods explained,
only the secondone, however,invoivesthe measuringof the flow or of the quantitiesthat determineit.

The lossesof the machine that are not carried away by the cooling fluid consist of :
- lossesdissipatedto the foundatiorrby conduction; not measurablebut negligible;
- losses compensated by the variation in kinetic energy of the fluid through the machiue,
when this fluid is a gas. These lossescan be calculated but are almost always negligible;
- lossesby convection as the result of the contact of the outside surface of the machine with
the ambient air. These can be evaluated by recognizing that each square metre of surface
dissipatesby radiation 10 watts per degreeof differencebetween its temperature and that of
the ambient air.
For generatorsof large power, these lossesare always very small (1 to 2\ of the total
losses,maximum).
- lossesoutside the air circuit, comprising in particular the lossesin the exciters(rvhich can
be determined with sufficient accuracy from the cl-raracteristicsof the machines) and the
lossesby friction in tl'rebearings,in whole or part, accordingto whether tl-reyare wholly or
partly outsidethe ventilatingcircuit. [n ,qeneral,thesebearingsthemselvesare cooled by an
independentcircuit, in which the power carried away by the cooling fluid can be measured.

The two methods describedbelow do not require any difficult set-up. The electro-magnetic
lossesare measured under normal operating conditions of the rnachirre(flux distribution, eddy-
current loss, temperaturesof laminations and conductors). They can be carried out without
uncoupling the turbine or even without requiring the unwatering of it.

4.2 Method of calibrating by the heating o.f the cooling air

4.2.1 Principles. Tlie volume of coolir-rgair of a generator operating at its nominal speed is fairly
constant, regardlessof its power output. Becauseof this fact, the temperaturerise of this air
betweenthe inlet and the outlet of the generatoris closelyproportional to the internal lossesof
the machine.
 - 1 8 5 -

of the heating of the cool-

These lossescan thus be determinedby means of the measuring
the magnetic losses of the
ing air, subject to prior calibration, accomplished by determining
machine running without load.

The application of this method thus requires:

no-load;
- the recording of the magnetic lossesof the generator at
-. the calibration of the heating of the cooling air;
- runs under various loads to determine the lossesunder load'

load are deterrnined by measuring the

4.2.2 Test .for lossesat no-load. The magnetic losseswithout
when operating as a motor
variation of the l0ssesof the machine as a function of the voltage,
nominat speedwith a power-factor
or as synchronousgeneratorand running without load at its
close to unitY.
turbine are closed, or adjusted
ln order to effect such an operation, the guide vanes of the
for running speedat no-load)
to a very small and strictly constant opening (close to the opening
voltage of which can be varied'
and the generator is coupled to another generator, the terminal
and the variation in the power
After deducting the .I2R losseswhich, moreover, are negligible,
between the power at the
required for excitation, which can be easily measured, the difference
at any voltage, representsthe
generator terminals for two runs, one at zero voltage, the other
magnetic losses of the alternator without load for that voltage.
function of the voltage can be
The ordin ate at the origin of the power variation curve as a
it as small as possible,ir-rorder that
set at any arbitrary figure. It is desirable,however, to make
power be maintained within a range of precisemeasurement'

measuredpoints be re-run,
It is recommended,also, that at the end of the test one of the first
(or any other parameter of operation)
in order to check that, during the test period, the opening
absorbed by the machine did
of the turbine did not chanle and that the power supplied or
not vary during the course of the test'

heating of the cooling air is accom-

4.2.3 calibration by heating of the cooling air. The calibration by
its nominal speed,and
plished by operating the machine as synchronous generator, turning at
of the air at the inlet and outlet
by measuring simultaneously the difference in the temperature
on the other hand'
of the machine on one hand, and the voltage at the terminals
or noted by means of
The heating can be measured by means of standard thermometers
last mentioned case' identical
thermocouplesor by resistancetemperatrrredetectors. ln the
at the outlet of the cooling circuit, to form the
detectorsare installed,one at the inlet, the other
bridge can be achieved by inserting
two arms of a Wheatstone Bridge. The equilibrium of the
practice this resistancewill be
an auxiliary resistancein series with the cold resistance.In
proportional to the temperature difference'
The calibration testsconsistin making successively severalruns at differentterminal voltages,
the value of the auxiliary resist-
and in measuringfor each, once thermal equilibrium is attained,
ance.
note is made of the power
For each voltage obtained in the course of these calibration tests,
is separate from that of the
consumed at no-load, less the losses in the exciter (if its cooling
coils and possibly also the power
alternator), less the 12R(in practice negligible) in the armature
The powers thus obtained
radiated to the outside by the ventilation ducts (generally very small).
machine that tend to increase
represent,within a constant small difference,the internal lossesof the
the temperature of the cooling air'
 t87 -

in temperature (or of the

The power thus obtained, when plotted against the differences
detectorsare used), will yield a
values of the auxiliary resistance,if electric variable resistance
of the heating of the cooling air
straight line which constitutesthe straight line of calibration
of the alternator.
trine of calibration of the heating
4.2.4 separation o/' lossesof alternator ot no-loacl. The straight
at no-load'
makes possible the separation of certain lossesof the alternator
: 0 gives a value for extrapolation of the
From the test at zero voltage, the abscissa[/
for the start of the heat rise
curve of losseswithout load, Figure 63. In fact, the value obtained
tests is establishedby extrapolatiol of a straight line'

n
rv V i

(u-0)
Frc. 63. - SeParationof losses.
Pu; Internal lossesof generator
o)' Differencein resistance
a) Straight line calibration curve
b:) Poini of operation at zero voltage
c') Calibration points at the voltages U
il Magnetic loises without load at the voltage Un'
L I'2 R losses in the rotor
e) Windage losses of generator

arbitrary power, in which

The ordinate at the origin of the calibration curve representsan
cooling air of the getlerator'
are included the lossesthat do not contribute to the heating of the
the point correspondingto
The differencein ordinates between the point of zero heating and
therefore, the lossesof the generator by
operation at no-load and at zero voltage represents,
windage.
lossesat no-load dependsupon
4.2.5 Remarks. The power obtained in the course of measuring the
other wcrds, the ordinate at
the opening of the guidc vanes of the turbine during the tests. In
but in order to improve the
the origi' of the straight line of calibration is completely arbitrary,
precision of the readings, its value should be kept as low as possible'
lossesthat do not cotl-
It is necessary,therefore,to cletermineindependentlythe mechanical
that arise from friction in thrust
tribute to the heating of the cooling air, that is to say, of those
by determining the heating
and guide bearings. This auxiliary measurementis generally made
of the cooling fluid in tl-reseparts'
 1 8 9-

4.2.6 Measurementof the lossesoJ' the generator under load. The calibration curve of the heating of
the cooling fluid of the generator makes it possible to determine the value of the lossesunder
;
load for any operating condition.
I Once such an operating condition is established,the value of the resistanceobtained after
thermal equilibrium is established(taking into account the correction for radiated power) makes
it possible to determine the total interr-rallossesof the machine under such operating condition.
After deducting the losses at zero load at the test voltage and the power supplied by the
exciter, the balance representsthe lossesof the generator under load, that is to say, the I2 R losses
and the supplementary magnetic and electric losses. The determination of these lossesis more
accurate if the readings of the difference in resistanceare as close as possible to those obtained
in the course of calibration. Becauseof this fact, it is advisable to carry the calibration tests up
to voltages as high as possible, provided they are compatible with the normal safety margin of
t he m ac hin e .
In order to obtain the total lossesof the generator,it is advisableto add to the internal losses
determined as specifiedabove the lossesthat do not contribute to the heating of the cool air,
that is to say :
- the power lost to the outside by convection through the casing;
- the lossesin the exciters;
- the lossesin the bearines.

4.3 Precautions to be taken in the course of the tests

4.3.1 Recording of the temperoturedifferences. The method is based on the principle that the mixing
of the air, both at the inlet and at the outlet of the machine,is such as resultsin virtually uniform
temperaturesacross the measuring sections.
So as to make possiblea reduction of errors due to non-uniformity of temperaturesin the
measuringsections,it is advisableto take the mean temperaturein the measuringsections. This
can be done, for example, by means of resistanceframes, the elementsof which are distributed
as uniformly as possible within these sections.
It is also advisableto arrange one or severalpairs of detectorsrecording the local temperatures,
with a view to checking the consistencyof the various instruments recording the differencesin
temperature between the inlet and the outlet of the machine.

4.3.2 Correctiondue to Jiiction loss in the coolingfluids. When the generator operatesin a closedcircuit
and becauseof that fact the differencein temperature between the inlet and outlet of the cooling
fluids is measured,it is necessaryto add to the lossesby windage,determinedas indicatedabove,
those correspondingto the friction lossesof the air dissipatedin the cooling fluid;

4.3.3 Correctionfor air intake. When the generator operates in open circuit, then the intake of the
air may be subject to pressureand temperaturedifferences. If the variations turn out to be impor-
tant, then it is necessaryto correct the results of the measurementsof the temperature differences
(or of the auxiliary resistance)by multiplying each of the temperature figures obtained by the
factor
p v
T

p designatingthe absolute mean pressureof the fluid;

u designating its mean velocity and T its absolute temperature.

4.3.4 Determination of the straight line of calibrating.for heating. Although this curve may be obtained
by means of two measured points only, it is advisable to provide four to five calibration points,
 - l9l -

both to ascertainthe dispersionof tl-rerecordedpoints and to establishthe slope of the straight

line, so as to reduce extrapolation en'ors to the minimum, for determining lossesby windage,
as well as electric lossesin the r,'icinityof full load.

4 . 3 . 5 D u r a t i o n o f t e s t . T h e m a j o r d i s a d v a n t a g e s o f t h e c a l i b r a t i o n m e t h o d a r e t h e l a r g e n u m b e rt o
e fs t s
required (from eight to trvelve according to the circumstances)and that each requires steady
operation for severall^rours,in order that the thermal balanceof the machine is established.It is
possible to reduce the duration of each test appreciably if the temperatureequilibrium of the
machir-re for a giveir operatingconditicn is not sought,but on the contrary, the operatingcondition
heati ng.
is v ar ied s o a s to o b ta i n a p re d e te rm i necl

4.4 Method employing the measuremento.f the flow ancl fhe heotirrg of the cooling Y'ater

This method is employed only when the ventilation of the generator is effectedby a closed
circuit and when the heat corresponclingto the power dissipatedis carried away by the cooling
water. In such a case this is the most accuratemethod of measurement,since it results in the
direct determinationof a quantity of heat by means of the flow and heating of the cooling water.

The power carried away by the water is given by the formttla:

AP : 4.2 q Lt 103

in which the .,zarioussymbols have the following meanings:

AP : power carried away by water in kilowatts:

q : flow of cooling water in mt/s;
Lt : the difference in the temperature of the cooling water between the inlet and the
outlet of the machine in degreesCelsius.

4.4.1 h,Ieasurementof the t,olwmeof v'ater. Tl-revolr,rmeof water can be measuredby means of one of
the classicalmethods such as Venturi tubes, orifice, nozzle, volume counters, etc., accurateto
within I to 2l[.

4.4.2 Measurentent of the heatingo.fthe v'ater. The differencein ternperatureof the cooling water between
the inlet a1d the outlet of the generator can be measuredby means of mercury thermometers
(which it is advisableto interchangeoccasionallyso as to eliminateany risk of systematicerrors),
by means of thermocouples,or by means of resistancetemperaturedetectors,with an accuracy
of + t%.

4.4.3 CalculationoJ'losses.The power correspondingto the quantity of heat carried away by the cool-
ing of the water, corrected as indicated below, representsthe internal lossesof the machine.

The total lossesare obtained by adding to tl-ratpower:

- the lossesof the exciters.
- tl-relossesin the bearings;
- the power exchangedwith the outside by convection from the casing of the generator and
the piping of the cooling water on the outsideof the generator.

4.4.4 Correction. When the determinationof lossesin the generatoris used for measuringthe efficiency
of a turbine, it is advisableto subtract from the above mentioned total lossesthe power corres-
ponding to the friction lossesin the cooling water circuit (ordinarily fed independentlyfrom the
turbine). This correction is frequently negligible.
 r93 -

turbines n) 'o)
5. Thermodynamic method for measuring the efficiency of

method is particularly suited to heads higher

5.1 In the present state of practice, the thermodynamic
down to 100 m, by agreementbetween
than 150 m. In favourable cases,it can be used for heads
the purchaserand the suPPlier'

5.2 Formulae used

the principle of the conservation
The basic formula for the application of this method expresses
of energY. It maY be written :
- u \
l ' t u : A ( ht + (l)
g
energy'
It.., : effective head converted by the runner into mechanical
: total head in a section of penstock or canal, at
steady flow'
lh

u 1 of water),
u(0, P) internal energy per unit weight of water (equivalent to a head
g g " g
at temperature and PressureP'

I u \ t Ll \ between two sections, above and below the

1 (\ / r t | : difference of (iz, + respectively.
'r (
t)
t r )turbine
the exterior in the interval betweerl
Equation (I) assumesthat there is no thermal exchangewith
the two sections.

0.06

0.05
)
7
0.04
./
,/
0.03||
I
I I
I 0.0?l

0.01
I r'1
---4 ----_l
0
i-hr= om
-----_ 500m
0.01 -- looom
- 1500m
10
"C 30
0t --

Frc. 64. - Thermodynamic method for measuring efficiencies'

'C
Values of k plotted against values of G)r in

lt) G. Wtt-t-tr.t - Measurementof Efficiency of Hydraulic Turbines by the Poirson Thermodynamic

and P. C,q.N,Iprvlrs
Method - Houille Blanche - July/October' 1954'
Application
tol G. Wtt-t-tvt - O n the precisionof the ThermodynamicMethod in the Light of rwo Year's Industrial
Houille Blanche - October 1956.
 - 1 9 5 -

and conducted without exchange of

lf water is taken from the intake manifold of a tr.rrbine
'Et' at temperature 9., under a head ltot
energy with the surroundings, into an insulated enclosure
Ihe water there at temperature 0r, the
at elevation zr, and if the tailrace level is at elevation 2,,
then :
velocity head at E, and in the tailrace being zero or negligible,

It,,: lrp'(l - k) I z1 - Zz- (g, - g,) (ll)

:g-

gr, given by the graph on Figure 64'

k is a coefficient, a function of ho, and

temperature'
is the calorific energy of water per unit of weight and of
; 6

is :
|f H, is the net head, the hydraulic efficiency of the turbine
h,
.fth .:
H"

the guide and thrust bearings' These

This efficiencydoes not include the mechar,icallossesin
guaranteedefficiency'
lossesmust be evaluateddirectly for calculating the

i* 5.3 Practical methoclsfor the measurementof h"

i Methods for the measurement of hu which would seem
most appropriate are the following
I (see schematicdiagram, Figure 65)'
the intake manifold of the turbine
water extracted through a dynamic withdrawal probe in
adjustablefor loss of head'
(or possibly in the penstock) is passedthrough a pressurereducing valve,
measuringenclosure'also insulated' in
and through a system of carefully insulated piping into a
gauge'
which is immersed a platinum resistancetemperature
a second identical temperature
This gauge is balanced through a wheatstone Bridge against
bridge, for equality of temperature
gauge immersed in the tailrace water. The calibration of the
in which the two were immersed
between the two gauges,is known from a previous measurement
head loss through the insulated system
in the same water at a thoroughly uniform temperature. The
the insulated measuring enclosure equal
can then be regulated so as to make the temperature in
of the bridge (no deviation of
to that in the tailrace; this is inclicatedby the return to equilibrium
the galvanometer).
this equilibrium to be obtained
lt is possibleto obtain equipment of a sensitivitypermitting
with a head differenceof only 0'2 to 0'4 m'

Under these conditions:

',
hu : i rp, (l - A) (2, - z" )
gz - 0t :0, that is to a zeto rcading of the
In practice, the value of h, correspolldingto
of ho, correspondingto small
galvanometer,will be obtained by linear interpolation of many values
positive and negative deviations of the galvanometer'
by checking its zero periodically
The stability of the equilibrium of the bridge must be verified
quality of the thermometric gaugesand of
during the course of the measurements. Normally, the
correspondingto changesof from 0'2
the bridge itself shor.rldmake it possibleto obtain a balance
to 0.4 m of water.
Hn, calibrationerror is eliminated'
By usingthe sameapparatusfor the measurementof hrrand of
the atmosphere results in very
The flow extracted must be such that thermal exchange with
l/s can be obtained in practice and is
small changes of temperature. A take-off flow of about 0.4
generally sufficient, with proper insulation'
 197 -

Frc. 65. - Diagram of the completeapparatusfor the thermodynamicmethod.

a) Insulated tube
b) Adjustable throttle v a l v e
c) Manometer
d.) Potentiometer
e) Galvanometer
f,t Wheatstone Bridge

/tu.
Nevertheless,e!.en under these conditions, it is often necessaryto make a correctlon ln
exchanges with the atmosphere from the take-off
in order to take into account the fact that energy
florv
circuit are not entirely negligible. This correction can be calculated by varying the take-off
proportional to the flow'
and by assuming(as verified by experience)that the correction is inversely

The upstrearn take-off must be made by means of' a dynamic lvithdrawal prob (that is, one
by
whose orifice faces upstream). Experienceshows that the measurementof /2" is little influenced
peculiarity. For instance,down-
the choice of draw-off point, exceptfor some especiallyimportant
stream of a butterfly valve, measurementsremain valid provided the withdrawal is not affected
in the wake of the disk itself.
In general,it is advisableto place the draw-off point at a distancefrom the wall of about one-
seventh of the conduit diameter.
the
The downstream measuring section (in the tailrace) must be located at a distance from
turbine outlet sufficientto ensurethat the mixture of the various streamlines(at different tempera-
to render
tures because of having worked with different efficiencies)has been completed so as
not be excessive,however, so as to avoid
the temperature homogeneous. This clistanc must
 t99 -

disturbing thermal exchangesbetweenthe water that has passed through the turbine and the mass
surrounding the tailrace. In general,the outlet of the draft tube is suitable for Francis turbines,
and a distance of between 5 and 15 m from the dischargepit for Pelton turbines.
to ascertain,by checking severalpoints of the section chosen in
It is advisable,nevertheless,
the tailrace, the magnitude of the deviations in temperaturethat could exist betweenthese points
and to choose a measuring point truly representativeof the average.
The measurementsmay be upset by a variation in the temperature of the water supplying the
turbine (heating of the pipeline by the sun, multiple intakes with water at different temperatures,
insufficiently rnixed or arriving in variable proportions, etc.). Also the constancy of the temperature
of the water drawn off must be checked by observing it with a thermometer graduated in tenths
of a degree Celsius. In tl-re case of a slow, regular variation of temperature (a few tenths of a
degree per hour), a correction can be made taking into account the variation in the respective
times of flow of the water between the draw-off point and the thermometric gauge upstream on the
one hand, and between the draw-off point and the thermometric gauge downstream on the other.
In unfavourable circumstances(pipeline exposed to the sun), measurementscould be taken
at night when the temperature of the water is at its daily minirnum.
Experience shows that in general there is no need to fear any calorific exchange capable of
introducing appreciable errors, between the water in the turbine and the exterior, provided that,
in the caseof Pelton turbines,it is ensuredthat there is not a strong current of air acrossthe meas-
uring section.
 -201-

CHAPTER XII SPEED

of output
1. Speett measurementsfor direct measurement
I
I of a brake, a torsion-
by measuring the torque by means
when the turbine output is determined by means of a precision speed
dynamo, the speedmust be measured
dynamometer or a torque-reaction or the average of several
means of tachometer and chronograph
measuring device, for example, by -.u"'red at the turbine shaft'
stop watches. The numt.. of revolutions should preferably b.
calibrated torque measuring instrument
the same time as the readings on the
and the readings should be taken at
are observed.

2.Speedmeasurementsincasesofindirectmeasurementoftheoutput
at the generator
o' the basis of the output as measured
when the turbine output is determined frequencymeter duri'g
turbine speedby means of the '*it"hbourd
terminals, it wilr sufficeto measurethe be checked against a suitable
period when the system road is steady. This frequency meter should
a
precision instrument'
 -203-

REJECTION
CHAPTER XIII SPEED VARIATION DURING SUDDEN LOAD

t. General remarks

load rejection is determined from the speed

The maximum instantaneousspeed n^u*, during sudden
Pt :0 (see II.2.10.3). since the speed variation
nr, when the power p, is being reduced to no-load
movement of the closing devicesof the turbine should
depenclson the closing operation of the turbine, the
be recorded simultaneouslywith the speed'

2. Instruments

tachograph should be used' The tacho-

For the recording of the instantaneousspeed a calibrated
record the speedvariations accurately during
graph should be constructed with a low inertia in order to
fulfills the requirements for this type of test'
the test. A suitable self-registeringelectrical tachometer
 -- 20s-

CHAPTER XIV PRESSUREVARIATIONS DURING SUDDEN LOAD REJECTION

l. General remarks

Since the pressurevariations in the turbine conduit are dependenton rate of guide vane closure
and affect the speedduring the closing operation, the pressuresar-rdthe speedsmust be recordedsimul-
taneously.
The maximum momentary pressureP-u* is determined from the pressureswhich occur during the
shut-down operation when a specifiedoutput Pt is being reduced to no-load Pr : 0 (see ll.2.ll.2 and
r r . 2 .r1. 3 ) .

2. Instruments

For recording the pressurevariations calibrated spring type indicators shall be used. With higher
pressuresthese indicators should be fitted with pre-loaded springs. The natural frequencyof the measur-
ing instrument should be at least 3 to 4 times higher than the frequency of the pressurevariation to be
measured.
The pressuremeasuringinstrument should be connected,if possible,directly to the penstockto avoid
lag causedby inertia and friction. If the instrument cannot be directly connectedto the penstock,the
piping should be as short and as straight as possibleand should be made of metal tubing, never of
rubber or flexible material.

Other methods of obtaining measurementsof pressure rise, such as the use of calibrated electric
resistancestrain gaugesin combination with suitable electric amplifying and recording equipment, may
be permitted.