1975
PROCEEDINGS OF
THE AMERICAN SOCIETY
OF CIVIL ENGINEERS
11142
FEBRUARY 1975
HY2
JOURNAL OF THE
HYDRAUL~CS DIVISION
REMOVAL OF
AIR
BY HYDRAULIC MEANS
By Paul E. Wisner, 1 M. ASCE, Farrukh N. Mohsen,2
and Nicholas Kouwen,3 A. M. ASCE
INTRODUCTION
A survey of case histories involving the causes of air presence in water lines
and its consequences was conducted under the auspices of the Water Research
Section of the University of Waterloo at Waterloo, Ontario, Canada. This survey
and the available literature in this field indicate that air may enter a water
transmission system in a variety of ways and its effects may vary from insignificant
to severe operational difficulties. The significant ways by which air may be
taken into a water line are : vortices at intakes, pumps, turbulence in shafts,
filling of lines , air separation at low pressures, vents, etc. (Fig. 1). Free air
in water lines may give rise to a variety of problems. Some of the more important
of these are reduction of capacity, surges and blow backs, white water, difficulties
in filter operation, reduction of pump efficiency, and corrosion of lines.
The best and the most obvious mea~ure that may be adopted against air
presence in water lines is to remove the cause of air introduction. In many
cases this is not possible, e.g., it is difficult to design a submerged entrance
to avoid vortex action under all heads.
In general removal of air from water lines may be accomplished by two
methods: (I) Mechanically by means of air relief valves, vacuum pumps, and
air traps; and (2) hydraulically by the inertia of flowing water. This report
summarizes the results of an experimental investigation carried out to examine
removal of air by hydraulic means. Although removal by mechanical means
was not investigated, some general indications are provided as to when hydraulic
means are incapable of effective air removal and mechanical methods are to
be adopted.
Note.-Discussion open until July I, 1975. To extend the closing date one month,
a written request must be filed with the Editor of Technical Publications, ASCE. This
paper is part of the copyrighted Journal of the Hydraulics Division, Proceedings of the
American Society of Civil Engineers, Vol. IOI, No. HY2, February , 1975. Manuscript
was submitted for review for possible publication on May 13, 1974.
1
Mgr., Hydrologic Sciences Div., James F. MacLaren Ltd., Environmental Engrs. and
Scientists, Willowdale, Ontario, Canada.
2
Research Asst., Univ. of Waterloo, Waterloo, Ontario, Canada.
3
Asst. Prof. of Civ . Engrg., Univ. of Waterloo, Waterloo, Ontario, Canada.
243
FEBRUARY 1975
244
(b)
(c)
A1
&
""'fj
4,
SECT I ON 1- 1
HY2
HY2
WATER LINES
245
OBJECTIVES
After an extensive literature survey on air problems in water lines , the objectives
of this research were set up as follows: (1) Simulate, if possible, in a physical
model, all the conditions under which different investigators worked in order
to appreciate each one's recommendations ; (2) investigate the scale effect on
" clearing velocity" ; and (3) recommend some tools to enable practicing engineers
to identify the different aspects of air presence and methods for eliminating
air, or adopting some remedial measures , or both.
0EANITIONS
246
FEBRUARY 1975
HY2
Q;
= (0.707) tan6 .
gDs
. . .. . . . . . . . . . . . . . . . . .. .. (1)
HY2
247
WATER LINES
higher than the highest of those listed. On the other hand, an example may
be cited where an investigator found that Kent's proposed value is on the safe
side. Babb (1) took measurements on a prototype siphon and a model and
TABLE 1.-Velocities Proposed by Various Investigators
Investigator
(1)
(2)
Ganderberger (2)
Veronese (9) and Lara (6)
Kalinski-Bliss (3)
Kent (5)
3.7
( 1.13)
1.92
(0.586)
2.04
(0.622)
2.46
(0.750)
57
25.4 mm)
showed that the prototype cleared satisfactorily at a velocity lower than that
proposed by Kent.
Two possible reasons may account for the apparent disagreement noted : (1)
248
HY2
FEBRUARY 1975
Conditions adopted by different investigators were not general ; and (2) scale
effects. From these conclusions and from the fact that all the previous investigators
worked with relatively small diameters and prototypes were comparatively large.
This leads to the conclusion that an investigation is needed: (1) Into the diversity
of results of previous investigators; and (2) to establish whether or not a similitude
exists in the matter of " clearing velocity " , i.e. , whether or not laboratory results
could be extrapolated for prototypes.
DIMENSIONAL ANALYSIS OF MECHANICS OF STABLE ISOLATED POCKETS IN PIPELINE
. . . . . . . . . . . . . . . . . . . . . . . . . . . (3)
Replacing
gD
v,
LB V,D
vw
-0=0
f ( -2,
'
'
By rearranging
. . . . . . . . . (5)
V,
Vgt5
= f (LB
D ,R ,0 ) . . . . . . . . . . . . . . . . . . . . (6)
v,
r-r.
v gD
= f (R) . . .
. . . . . (7)
v,
r-r. =
v gD
. . . . (8)
249
WATER LINES
HY2
EXPERIMENTAL SETUP
The size, length, slopes, and summit bend geometry of the experimental set
up was based on the following.
Size.-Probably the most important aspect of this research was to investigate
- - - -----$ -
EL
4 71 M
39.37 in.; 1 cm
0.39 in.)
the scale effects related to the mechanics of air pockets. Previous investigators
worked mostly with I-in. (25-mm-4-in. (100-mm) diam pipes and, in one case,
6 in. (150 mm) . Modern prototype water transmission lines are often very large
in diameter. Thus an investigation of the scale effects by experimenting on
larger pipes is necessary. This approach has been recommended by a number
250
FEBRUARY 1975
HY2
/(6) . . . . . . . . . .. . . . . . . . . (9)
The variation of the rising velocity with slope is complicated by the change
shape with slope. However, Fig. 10 shows the variation of
[ V,/vgD JR.,, io5,n > 0_8 with 6, based on Gandenberger's and the writers results.
Moving Water .-An experiment was performed to investigate Veronese's ''limit
bubble. " A large pocket was introduced in flowing water. The water velocity
was changed continuously to keep the pocket in equilibrium as disruption
of
WATER LINES
HY2
251
3 0
...
>
..
20
0 10
0 20
0 30
0-40
~ 0-40
o "'o AUTHOR
* o GANDENBERGER
>
0 30
0 20 2L_O_O_ _ _ ___J.3_D_0 _ _ _ __
Log
4~
-o
-o-----~5-0....,....,0_
_ _ _ ____,....6DO
(R )
..,
0
" ' 06 ~
a::
CD
P---~---
05 ~
"'c:
I1
'
~ OA ~
>
e-
252
FEBRUARY 1975
HY2
progressed. It was observed that the pocket was finally reduced to a small
stable size and that any increase of the velocity does not further disrupt the
pocket but sweeps it out. The successive stages of disruption and the final
stable pocket are shown in Fig. 11. The results extend Veronese 's results on
limit length and limit velocity (Figs. 12 and 13). The results confirm that, in
the range of Veronese's and the writers' experiments at least, stability may
be achieved for a particular size of pocket and flow velocity. The values are
unique, i.e., for a particular diameter and slope there is one stable length and
corresponding velocity. From a study of the mechanics of isolated air pockets
in still water it was expected that since Lb/ D (consequently n) for limit bubbles
varies with pipe size, the parameter, V,/'\/iD, will not be independent of R.
This is indicated by the plot of Veronese 's and the writers ' experimental data
on limit bubbles (Fig. 14). Although no general rule can be offered unless more
experiments are performed for a greater range of Reynold 's number it may
be concluded that the parameter, VJ'\/iD, is a function of Reynold 's number
V, / '\/ii5 decreasing with R beyond R = 4.2 x 104
The experimental results clarified two important points. (1) The limit velocity
does not become a constant quantity with increasing diameter as suggested
HY2
253
WATER LINES
by Veronese and Lara, but it decreases with diameter (at least in the range
of writers ' and Veronese's experiments; and (2) the limit length does not become
a constant beyond 3.94 in. (100 mm) in diameter, but decreases at a decreasing
8=18 5
150
E
u
VERONESE
AUTHOR
50
50
10 0
150
200
250
D (cm)
c.
E
0.39 in.)
8:19 .5
!:: 0 6
VERONESE
:::;:
AUTH/"
_J
>
OA
/
D (cm)
0.39 in.)
rate. For the 9-5 / 8-in. (244-mm) pipe the limit velocity and limit length were
found to be 2.35 fps (0 .72 m/ s) and 1.8 in. (46 mm), respectively.
A more comprehensive research program on larger pipe diameters would be
254
FEBRUARY 1975
HY2
7 18
14 47
...
"'
10
09
08
l<ALINSl<E
x AUTHOR
07
(RANGE OF KENT'S
RESULTS)
( FIGURES BESIDE EACH
POINT INDICATE
CORRESPONDING Y1/D )
'-.. 06
u
>
05
OA
"''o "'
HY2
255
WATER LINES
10 ~----------,,----r----------,
0 9
0 8
0 7
06
I~
0 5
~ 0-4
0 3
02
OI
05
l O
V ( mps)
15
20
256
FEBRUARY 1975
HY2
The study was supported by the University of Waterloo and the National
Research Council of Canada. Their support is gratefully acknowledged.
APPENDIX !.-REFERENCES
WATER LINES
HY2
257
F
g
LB
n
Q,
R
T1
v,
v,
vmin
v,
v,
Y.
v1
= A 1 / T1
Y1