4-i. -.

______ -;
-. - -.- -

WATER HAM ME R/CAVITATIO N

Assessment & Prevention

"* WHAT IS WATERHAMMER 0 WHAT IS CAVITATION

"• WATERHAMMER CATEGORIES 0 WHERE IT OCCURS
"* WATERHAMMER PHENOMENON S HOW TO IDENTIFY
"* WATERHAMMER ASSESSMENT 0 EFFECTS OF CAVITATION
"* WATERHAMMER PREVENTION 0 CAVITATION DAMAGE
"* INDUSTRY EXPERIENCES DESIGN CONSIDERATIONS

SUBHASH C. KHURANA
FLORIDA POWER & LIGHT

L
FPL
 Cavitaiton Trng.
Page 1 of 6

CAVITATION

1.0 What is Cavitation

In a flowing liquid if the static pressure is dropped below the

vapor pressure corresponding to the liquid temperature (or the
temperature is increased above the saturation temperature
corresponding to the liquid pressure), vapor bubbles (cavities)
will be generated. If these bubbles are carried downstreams by
the liquid to a region of higher static pressure (or lower
temperature) then the bubbles will collapse. the region of the
flow where bubbles exist is the cavitating region, whereas the
observed damage is at the location of the bubble collapse.
Cavitation is distinguished from flashing or boi-in- flow in that
no pressure recovery occurs downstream in the latter type of
flow, and therefore no bubble collapse occurs.

2.0 Where Cavitation is likely to occur in Systems

a Orifices (Pressure profile across an Orifice, Figure-l)

* Valves
* Venturies
* Pump Suction
0 Pipe Fittings (Elbows, TEEs, etc.), (Figures 2 and 3)

0 Bumps

3.0 Identification of Cavitation

This process of bubble formation, growth, and collapse produces

oscillation in the flow regime that can be identified by distinct
sound. The sound may range from occasional popping (at the onset
of cavitaion) to a sound like frying bacon as cavitation level
increases. Extensive cavitation sounds like gravel flowing
through the system. Because the flow oscillation associated with
cavitation is broad band, pipe vibtation will also be associated
with cavitation.
 Cavitaiton Trng.
Page 2 of 6

4.0 Effects of Cavitation/Damage

Cavitation produces undesirable effects on piping system and

components. It generates undesirable noise, may cause erosion,
restrict flow capacity of the system, and may cause pipe failure
due to pipe failure.

The erosion observed in cavitating fluid is due to bubble

collapse near the pipe wall. Studies have shown that as a result
of bubble collapse the liquid jets with velocities of 300 to 2000
feet per second will impact the pipe wall. At this high impact
velocity, the local pressure at the wall could be as high as
10,000 atm. The continuous bomardment of the metal surface will
result in the pitted configuration observed in most damaged
components.

5.0 Design Considerations

For system susceptible to cavitation, the designer should ensure
that under operating conditions, the system is either free from
cavitation or has acceptable level of cavitation. Levels of
cavitation are defined below.

Incipient Cavitation The onset of cavitation. No

objectionable noise and no damage.

Critical Cavitation Continuous light cavitation usually

adopted as the design criterion. Noise
and vibtation are acceptable and only
minor damage after long periods.

Incipient Damage Onset of surface pitting after short

period of operation.

Choking Cavitation - Flow is not increased if downstream

pressure is decreased. Maximum vibration
and noise.

Cavitation at pump suction is well taken care by satisfying

Vendor's NPSH requirements. Valves and Orifices are genarally the
common source of cavitation problems encountered in power plants.
The following sections provide basic guidance on how to avoid
cavitation of orifices and valves.
 *m

Cavitaiton Trng.
Page 3 of 6
ORIFICES
(See Figure-1 for a pressure profile across an Orifice)
Cavitaiton Index

a = (Pd - PV)/(P- - Pd)

- (Pd - Pv)/AP (Reference 1)

Where: a = Orifice Cavitation Index

Pu = Upstream Pressure, psia

Pd = Downstream Pressure, psia

Pv = Vapor Pressure, psia

The above calculated "cavitation Index" is compared to

accepatable "Cavitation Level" (see Figure 5) which is
experimently determined . If the cavitaion index is equal to, or
higher than the cavitaion level, the orifice size is acceptable.
If a single orifice fails to satisfy cavitation criterion,
multiple orifices are required. The spacings between orifices
must allow complete pressure recovery, which is achieved by
placing the orifices 6-8 pipe diameters apart. The "f," ratio
(orifice diameter/pipe inside diameter) of orifices shall be in
increasing order along the flow. Sample Problem-I illustrates
selection of orifice(s) to have cavitation free operation.
When enough space is not available, multi hole type orifice plate
(perforated plate), or a cone type orifice is 'recommended.
Other definitions of Cavitation Parameter

(hu-h.)
a = and
U 2/2g

(hu-hv)
a = (Reference 2)
(hu-hd)
Where: a = Cavitation parameter
hu = Upstream head
hv = Vapor head
hv = Downstream head
U = Upstream velocity
g = Gravitational constant
 Cavitaiton Trng.
Page 4 of 6
Reference-2 utilizes reference velocities for a component size
and upstream head at which a type of cavitation commences. These
velocities are then corrected to other component sizes and
upstream head conditions using empirical relationships with
experimentally derived coefficients.

VALVES
Colorado State University tested some valves for Hydraulic and
Cavitation characteristics. The test data indicate that the
critical cavitation index does not only depend on the valve
opening (area ratio) but is also dependent on VALVE SIZE,
UPSTREAM PRESSURE, and VALVE TYPE. Figure-9 through-12 show
results of such tests. Figure-12 also includes a curve for the
critical cavitation index for a thin orifice plate for comparison
to that of a valve.
For a valve to operate completely cavitation free, operating
conditions should never approach those corresponding to the
critical cavitation index. Operation beyond critical cavitation
is never recommended.
Analytically, it may not always be possible to accurately predict
when a valve will start to cavitate. Therefore, monitoring for
noise and vibration levels at the required operating point is
recommended.
Reference 2 provides guidelines for determining cavitation
parameter for valves. Sample Problem-2 is used as an illustration
for a butterfly valve.

6.0 References
1. Technical Paper, "Eliminating cavitation from pressure
reducing orifices", CHEMICAL ENGINEERING DECEMBER 12,1983
2. INTERNAL FLOW SYSTEMS, by Donald S. Miller, Volume 5 in the
BHRA ( British Hydromechanics Research Association )Fluid
Engineering Series
 Cavitaiton Trng.
Page 5 of 6

SAMPLE PROBLEM-I -/- / &'-- - I0.

Pump discharges water to atmosphere (See Figure-4).

Pipe size: 14-inch, Schedul e 40

Flow: 4500 gpm
Pressure drop across orifice(s): 120 psid
Fluid temperature: 1600 F
Determine the size of orifice(s) for cavitation free operation.
Neglect piping friction loss. Use Figure 5 for comparing
calculated cavitation index (a).

SAMPLE PROBLEM-2
Water flows from a reservoir at higher elevation to the reservoir
at lower elevation, and flow is to be controlled by a butterfy
valve (See Figure-6). Determine the cavitation free range of
control for the A-type butterfly valve from fully closed
fully open position. Loss coefficients for butterfly valve totypes
are shown in Figure-7. Accepatance criteria is critical
cavitation velocity..

Pipe Size: 30-inch, Schedule-40

Valve Size: 30-inch diameter
Water Temperature: 900 F

K-for Piping between "A" to "B" : 17

K-for Piping between "C" to "D" : 7
K-for butterfly valves types: Shown in Figure-7
Based on Reference-2, Cavitation velocities for a butterfly valve
are determined as follows:
0
Ui or U, = Cl. (Ur or Ucr).[(hu-hv)/50] .39

and
Uch = Urch [ (hu-hb) /50] 0.5
 Cavitaiton Trng.
Page 6 of 6
Where:
U1 , U,,and Uch are incipient, critical or choking cavitation
velocity (to be determined) for a given butterfly valve.
C1 is correction factor for the valve size used, Figure-8

Ui,. U,,, and Uch are the reference cavitation velocities

(incipient, critical, or choking as applicable) for the base
valve, shown in Figure-B.
h = upstream head of the valve (meters)
h, = vapor head (meters)
"50" = "h,-hv" value (in meters) for the base valve (size
0.3m diameter)

CAUTION: USE DIMENSIONAL UNITS APPROPRIATELY

 ORIFICE
I
I

Pu Pd

PvP

PRESSURE PROFILE OF A SINGLE-ORIFICE SYSTEM

Pu = Pressure upstream of the orifice

Pc = Pressure at the vena-contracta A473

Pd = Pressure downstream of the orifice

Pv = Vaper pressure of the flowing fluid

FIGURE - 1
OA 7, VAPOR FILLED CAVITY

SB

ELBOW

A472

FIGURE - 2
Brim

I,
, !

FIGURE - 3

---------

ti
INCORRECT SOLUTION:
(1 stage orifice)

CORRECT SOLUTION:
(2 oriface stages)

2nd 1st

_7 -2P
Ii0 -2

A471

FIGURE - 4
 I-

Critical Cavitation Index

for
I.
Thin Plate Orifice

FIGURE - S

- I--t-I-t-4-I-I-I t-t-1-I I

- . - 1* 9 - I- -

NO CAVITATION /
---. 1L-J - + - 1 IiT -i

CAVITATION

- -...- . ..
L_~
- -- -- - •--

-.. -...- -•
. ... "- - .. ..-

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.

0.0

DIAMETER RATIO (d/D) -

 200 FT.

70 FT.
Pt.
RESERVOIR
D
RESERVOIR
B
REFERENCE
VALVE -0 FT.

FIGURE - 6

A477
 1000

100

50
" A

10

1.0
0.1
0 10 20 30 40 50 60 70 80 go
Vahv owniN (deg,)

Loss Coefficients for Butterfly Valves

(from Reference 2, Figure 14.19)

FIGURE - 7

lI
)

L=s coeffi~ent K

Cavitation Velocities for Butterfly Valves

(from Reference 2, Figure 6.22)

10.2
1.4 - 7 2m---" --

U
1.0 - m --
a oO." _. , , . ........
, .., - - -. , - - - - - " - -'- - - - "- - - - - " - - --
IA.
C
C

a
U

0.1 1 10

LOU coff @Mc

KV

Correction Factors for Valve Size

i (from Reference 2, Figure 6.23)

i
I FIGURE - 8
 6

I ....... 11. FPiGURE - 9

6 11

12-in. Rjuttes4L.VVv?! with

24-in. expansion downstream SiSr

5 Upstream pressure 00 psi

0 Normal tests
1 Air injection into stem
- -

- -J

S.... rnRf 70 t'0

(1
0 10 z0 iu "IU •,.Iv

"Test Valve Opening'- 9o

 , 4 .

)',C•,
MA 'z-
"FIGURE - 10
6

3
u
u;

u
z

0
0 to 20 30 40 50 60 70 80
Test Valve Opening - 7
 FIGURE- 12
3.0 - -

8-in. Nqg1j& .y• e

1.,xpansion Pu
o 14-in. 153 psi
* 14-in. 65 psi -

& 12-in. 153 psi

o 12-In. 65 psi
* 14- in. 153 psi (swOeep tests)
A 12-in. 153 psi (sweep tests) - S2.0
0 8-in. 65 psi

>0

u
"u . o.
."

F//
0 20 40 60 80 103

Test Valve Opening - I

 FIGURE - 12

U
La4

u 2

0
0 10 20 30 40 50 60 70 80

'rest Valve Opening - I

Orifice Area Ratio (d/D) - 0/0

 Cavitation Trng.
Attachment
Page 1 of 3

Eliminating cavitation from

pressure-reducing orices
Multiple orifices can avert the cavitation, and resulting noise
and vibration, caused by single orifices in hydraulic systems

P.ztiik C. Tuisg .. d Miltd MQiuie, Ouarw Hydro

o Fig I illustrates flow through a sing orifice in a

pipe: P. is the pressure upstream of the orifice; P, the
vena-contracta pressure; and P,. the pressure dow'n
stream of the orifice. The last pressure. Pj. must be
higher than the vapor pressure. P,. of the flowing fluid.
If PA falls belou P,. bubbles form. the first stage of
c,,vitation. As the fluid moves downstream, the bubbles
collapse under the higher pressure. Pg. in the final stage
of ca-itation. The implosions generate noise and vibra
tion. and accelerate the erosion of the piping []].
If P, is kept below P,. t'o-phase flow continues down P. I~P. p
stream as a flashing mixture without cavitation.
Cavitation criterion
To avoid orifice camitation. the orifice size selected for
the pressure-drop requirement must be checked against Pressure profile of a single-orifice system Fig. 1
an orifice cavitation index, which is then compared to a
cavitation level If the index is equal to. or higher than.
the cavitation level. the orifice size is acceptable.
The index was defined b. Tullis and Govindarajan 12]: is'ies the cavitation criterion. Calculate the cavitation
index. o'. for the orifice %ia Eq (1), and get the corre
0, - (P4 - PY(.W. - Pd) - (pd - P.-AP (1) sponding acceptable cavitation level, a,,. from Fig. 2.
Compare a' to a',: If a ý c',. the size of the single orifice is
They also experimentally determined acceptable cavita acceptable; if o < a,. multiple orifices are required
tion levels. These. ',. vs. orifice-to-pipe-dia. raitios. P. for
%anouspipe sizes are shou n in Fig. 2. At these cavitation Design of multiple-orifice systems
Levels, noise is io, and stead% and no damage will result. Each orifice in a multiple system (illustrated in Fig. 4)
When a single orifice fails to satisf' the cas'itation crite is sized to meet the c•vitation criterion
rion.two or more orifices in series are required. In multi I. Assume a P value for the fast orifice. OR1 (the one
pke-orifice systems. the spacings between orifices must furthest downstream). With this 9 value, determine the
allow complete pressure recover)', which is achieved corresponding resistance factor. K. from Fig 5.
when'the circumference of the orifice jet expanding to 2. Substitute the value for K obtained in Step I in'I
ward the pipe ^-all equals that of the inside pipe. Fig. 3 Eq. (2) to find the AP across OR,.
relates these distances to P [3]. 3. Substitute this calculated AP into Eq. (1) to deter
mine the ca2itation index. a,.
Design of single-orifice systems
4. With the Pvalue assumed in Step 1. find the accept
To design a single-orifice system free from cavitation: able caiLtation leveL o in Fig. 2.
1. Size the orifice for the required flowrate and pres 5. Compare oto o',. If ao:q the assumed 0 i satis
sure drop by means o7 !-.(2):
factory. (To have the minimum number of orifices for a
Ah- - X(V412)p (2) specific flourate and pressure-drop requirement. the
value of a should be kept as dose as possible to Or.) If
Determine the resistance factor. K. for the orifice by 0r< 0o, choose a larger p value and repeat Steps I
inserting the floeate and orifice pressure drop into Eq. through 5 until a'o'.r
(2). Having the K value, find the orifice size from Fig.1LX" 6. The downstream pressure. PIM. of the second ori
2. Check that the orifice size obtabwd from Step I sat-
69
CMUKAL LXWALLIUWG ZNWEMM U. Mo
 Cavitation Trng.
Attachment
Page 2 of 3
(Ca Tr1"I

i
A Pipe flow W=. V
I"\ I ' I
X Resistance factor P11
!n Man, Gowra. igh ---------
1 I
OR
I
oR,
p~d 3akpreuuur. L~a OR 3 2
P. OrirKr upavam prm=a. kPM
P. Vapor prmur. M•
A, Pipe internial radlus,
-V Tho.wVktiz). =Vs
to fm pipe & af a multiple-orifice system Fig 4
a Dinanc for orie jet drcuzmfu
¢mlference. =
I Du-ner rato o oroif/"tpipe fKce. OR-,. is set equal to the uptream pressure. PI.. of
• •-•). hem the first orifice. OR,. etc. (as indicated in Fig. 4)
4r Cavitation budes 7. The last orifice (the one funhest uPTreim) is sized
w, AmzpLabk tavisaog kevel as was described for a single-orifice system
8. After all the orifices have been sized. determine the
neccssarý distances between the orifices for cavitation
free operation b%means of Fig. 3. For example. find the
proper distance bet% een ORI and OR2 (Fig 4j. enter the
value of J for OR2 into Fig. 3 and obtain the correspond
ing value of x!Rr Having the pipe radius. R,. determine
the proper distance;-x, between OR, and OR.

Example illustrates the procedure

Design an orifice system for a 186 0-kg s floTate of
water through a NPS12 Sch. 80 pipe. The upstream
pressure and temperature are 10.340 kPa and 138'C.
respec-ve.. and the orifice backpressure is 690 IPi The
vapor pressure and density of water at I38'C: are 340 L)It
and 928.0 Ig'm'. respectivel.. The pipe's inside radius is
0.145 m. and its flou' area is 0.066 mi.
First determine the flow velocity through the pipe.
V - m'pA = 186.0:(928 00X0.066, - 3.0 m s
1. Trý a single orifice, sizing it via Eq. (21.
Acceptable cavitation level related - (10.340 - 690() I.SM = A(3.0"2] 92F 0
to orifice-to-pipe-dia. ratio 121 , . Fig. 2 I
SWith K - 2.300. P = 0.186. from Fig. 5.
- •. ... - .• , P
2. Check whether the orifice size = 0.186 satisfies
the cavitation criterion. Via Eq. (1):
arn (69 - 340)'00.340 - 690. = 0.036
With P - 0.186. Fig 2 gives: a, 1I.I.
As (a -- 0.036) < (ct- 1. 1). one onficr system does
.cL not meet the criterion, and a multiple s%stem must be
designed:
, .. : -- . 2• l. Assume that 1, = 0.4 for the first orifice. OR, (the
one furthest downstream). With this 191. A - 90
2. Determine the pressure drop via Eq (21.
I..:
AP - 90 1(3.0)2f2] 92.0 - 375.8 LRa
3. Substitute this AP value into Eq. (1):
a, - (690 -340)f375.8 - 0.93
0.2 - from Fig. 5.
4. With ,0.41 A, ,-.3,
5. Because (0, - 0.9) < (q, - 1.3). the cavitation cr•te
riOn has no( been met. Tr" a larger orifice size--e.g-.
1 -0 2~ Ps - 0.45. Repeating the "foregoing procedure tields
P-
" -,1.62-and a, = 1.6. which satisf% thecrflterion.
vistarne downstfeano -6The sizes of the remaining orifices are determined
for compltce pressure ri SPr:or'usl described-i.e. the downstream pressure.
 cavitation Trng.
Attachment
Page 3 of 3

10 0.1
O-OWAI DIM a 0

orificn 141 fig. S

RetationshiP of rnistame factor to cwific&-to-pipe-dia. ratio for "rp-edged

p2d. of the d OrifiCC- OR2. is Wt equal to dW UP Ibe, siý

ravick C lr;ýj'ý des;r eftw-M wW
SET'CaM pressure. PI.. of the first orifikx. Mi. CM Orwww Hvdro (700 L ftwrmn.-AC
7. Size the final oTifice. as his been dcscrR)ed. Torama Cku- MSG Dfi-CjMWA,,:,
I plimot 416-W--=W HOVý7.-ý - I
8. Lasd). find the dis=ccs beturctm the orifices. mpmuWam wchý V-d
J. Adadry. EAw Md Lr dnwn of PTUM %-k-p- -
and mWed eq-Prwww2HOkI" Of A
wwd'W&Cd earrwrmw from shi-ý
t;nrmm *(%%w=wn wWw- USA,;
References In= the Vwwrwn Of M-wt- Ow
a '70"'d CArrw" in Omano. a VwvIkw
1. Jaime. Ir-C&wwxan in eancral Vamjawww- ant Ca-d5rWars. VaL 34. of the Anwn= Sm 0( M-M--J
No.vmbrr 1961 J'r'w' WW the Othm O(wrrAI
2 TW6 j ?..NW GOVWAMYUL R. Cavaum WW SOW-5-k VISM SM Or6CCL:' muck, in wdwwW Pubb
- j df Piv"wfir Dn . Anwn= Sm. Ci Crei Lngwwcm MwCh 1273
T"Wd=. a G- MW %-*m U P. -1-M ri, ei L'w Ve" GWWN=7
AwwrIMA Sm. of MechanciJ Engvwo-L P*v he 73-%AITM-9
4. XaAwwkv- U- mid Twq. T C- Ssmg of TW*mbag Onfa=L. Saw%4PYOW WICU MAUAWOM S & wneac deW"
Ar GPft&bowwC. D"mub" IM ,pm-1 ý,h OnL&rw 14-ldfu H*
LwAvr=mn.Wj-wd 'Aabs.L W- nuid TlwwTbrwACIýTwoorifiaw is soon. map--b-b- widude drWn ww
Toom Awný Sm 4(XFAwmwd Lnmwm jounn 1157 .. "Mm of Vwjw-2' "Pup-"
OLIcnpffwL M 'Wý and MMn. j G- The Fkp- of Sowrxwd VAw TbrONO ,d d..---J P-- Wj
ThrombM CwVcm TvoL Amomm SK #f AdmArmait LegwAm JWY M 1 '(' IS " wwC"r-Kw CDC--q
imm aw U-V'w- -( bar'Ide
7. WummdL j ?- et aL. Cawwwwan Elko m WbcDodwav Ckeffamm Oddw TwjadrAL jw a a =,ed . S:
DwpEdCW Onf- "wNj t(Sa- X%--4 Man:h IM Onmb" of Lhe Lw=
TwIsLj. r- et aL hvkrwd Pbm a HvdrAK EmeThDo Im rni.-r-w b'r- a'd Ow a`
St a Undboa affhad r-,
=n Sm- a( 04 L-4wwm JOW T.Imbýwl MW O(bM--'
9 vAjwwL w. w., tvakwoona(Mubynsinve Wacj dfhýOý V&ýAwý
Sw- or Cr#j &Wmmjuw1977.
W TwLj.T
=n Sor- of 04 Lasww"L P-Oftf 1971.
t I. DrubeL L- CAwad Vshq Sm%W" ZA Fwmdm& few. raL. Jay "M
7
CKLUSCAL gw.&ýP*G DICU"" $2. IM3