Professional Documents
Culture Documents
This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in Denver, Colorado, USA, 30 October2 November 2011.
This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.
Abstract
Water hammer is a known pressure pulse or surge that may occur by the instant shut-in of a valve in a flow line. Sudden
momentum change may create a pressure cyclic pulse that could cause damage to valves, bending parts in tubing, and/or
joints. Usually this effect has been well managed in surface facility design; however, it tends to be overlooked in subsurface
well design. Additional possible impact by water hammer in subsurface wells could be on the sandface completions. The
severe water hammer could cause failure of formation integrity, resulting in sand production. It may also damage the
wellbore and downhole completions. Especially for deep sea water injection and/or production operations, water hammer
effect needs to be thoroughly investigated and properly managed because it could be more severe due to longer flow line and
higher flow rate.
The purpose of this study is to have a comprehensive investigation on water hammer effect for an actual water injection
well in Chevrons deep water project with different design parameters and operating parameters. The design parameters
include a) height of vertical riser; b) tubing diameter; c) injectivity index (skin or completion type); d) sandface wellbore
length; and e) well deviation. The operational parameters include a) injection rate; b) closing time; and c) injection water
temperature. Multiphase transient fluid flow model OLGA is used for the water hammer simulation.
Results of the water hammer parameter study for optimum well design and operating strategy are reported here. It is
shown that the impact of water hammer can be significantly mitigated or eliminated at well design stage or by adjusting the
operating parameter(s).
Introduction
The importance of water injectors has increased as waterflooding is entrenched as the most popular secondary recovery
process in matured reservoirs. However, inappropriate shut-in operation and well design of water injectors frequently causes
water hammer, resulting in severe damage by sand production, detrimental accident by tubing or valve rupture, and
premature abandonment by wellbore collapse. Water hammer effect is a well known phenomenon that can occur at the
pipeline where water is being transported. A transient nature of pressure wave occurs when water in motion is forced to stop
or change its direction suddenly. Two different mechanisms can explain the causes of water hammer downstream and
upstream of the shut-in valve. Upstream of the valve, when the valve is rapidly closed, the mass of water (that is moving
forward) builds up to a high pressure and shock wave at the position of the valve and the pressure wave travels back and forth
until the energy dissipates due to friction. On the other hand, downstream of the valve, the mass of water tends to continue
flowing by law of inertia, when the valve is instantly closed. It creates a temporary vacuum just below the shut-in valve,
pulling the water body back to hit the shut-in valve and then rebound again. It is a cyclic process that gradually dissipates
due to friction. Water hammer intensity is at the peak at the shut-in valve and it tends to fade away farther from the valve
because the pressure wave loses its energy during propagation due to friction.
The study of water hammer was first initiated in the nineteenth century and many researchers have put forth considerable
effort to understand the principle of water hammer and develop appropriate expressions1-9. A big milestone was made by
Joukowsy4. He developed a well known analytical equation called fundamental equation of water hammer, which is still in
popular use. The Joukowsys equation is:
2 SPE 146300
a V
P = a V or H = (1)
g
where = fluid density, a = acoustic wave speed, V = cross-sectional average velocity, H = piezometric head, and g =
gravitational acceleration. The acoustic wave speed is defined as:
1 d dA
= + (2)
a 2 dP A dP
The first term in the right hand side indicates fluid compressibility effect and the second term represents pipe flexibility
effect. In the oil and gas industry, only a few papers10-16 have been published on water hammer and corresponding sand
production issue, which has recently become recognized as a critical area.
The objectives of this paper are (a) to have a comprehensive investigation on water hammer for an actual water injection
well in Chevrons deep water project in order to understand possible causes of water hammer in (deep sea) water injection
operations, and (b) to run an extensive sensitivity study to explore the opportunities to reduce the water hammer impacts with
different design parameters and operating parameters. It will provide a useful and practical guideline to mitigate or eliminate
possible water hammer at the well design stage as well as in field operations. The water hammer (pressure wave) data could
also be used as key input parameters for geomechanical simulations to evaluate the possibility of sand production and
determine the sand volume, if any.
Methodology
A base case model was first developed and extensive sensitivity studies were performed by switching the model parameter
one by one in order to investigate how each parameter affects water hammer. The base case represents the current design and
operation conditions of Chevrons actual water injector. A transient wellbore and flowline simulator called OLGA (SPT
group) version 6.2.7 was used in this study. OLGA numerically solves mass conservation, momentum conservation, and
energy conversation equations, which enables it to handle various transient and steady state problems with multiphase flow.
The PVT file of water properties were generated by water package in PVTsim (Calsep Inc.) version 19.2.0 and imported
into OLGA for water PVT calculation.
Model Setup and Key Input Parameters for the Base Case
Figure 1 shows a schematic diagram for a water injector in scope and the well trajectory is illustrated in Figure 2. The well
is drilled in a vertical direction to approximately 2,200 ft from the sea bed and starts to deviate for the horizontal section.
The sandface wellbore is exactly horizontal with the lateral length of 2,000 ft. A Sub-Sea Test valve (SSTV) is installed as a
major shut-in valve in the Christmas tree at the sea bed and a Surface-Controlled Subsurface Safety valve (SCSSV) can be
located in either one of three possible locations (300, 1,718, and 7,753 ft MD). The major input parameters for the base case
are given below.
Injection Rate : 50K BWPD
Closing Time : 30 seconds
Injection Temperature : 32 F
Shut-In Valve : SSTV
Tubing Diameter : 7 / #29
Injectivity Index : 7 x 10-4 kg/s/pa
Sandface wellbore : 2,000 ft horizontal
The captured screen of OLGA model is shown in Figure 3. The model covers the fluid flow from SSTV on the sea bed all
the way down to total depth (TD). The wellbore is divided into 102 segments; each represents about 100 ft of the pipe. Out
of 102 sections, the 2,000-ft sandface lateral section (7,753 9,753 MD) is divided into 21 sections representing the well
outflow to the formation. The heat transfer through the tubing walls by temperature difference is considered by
implementing wall layers with the corresponding heat properties into OLGA.
SPE 146300 3
Displacement,ft
0 1000 2000 3000 4000 5000 6000 7000
0
1000
2000
TVD,ft
3000
4000
5000
6000
Figure 1. Schematic Diagram of a Water Injector Figure 2. Well Trajectory of a Water Injector
PressureDecline(psi)
TransientDuration (sec)
BackPressurePulse(psi)
Case Definitions
A total of nine parameters (considered to affect water hammer) were carefully chosen and studied as sensitivity variables.
Table 2 shows the case definition of the parameters and their study ranges. Note that the values in yellow boxes are the base
case values. Three operating variables (that are controllable in field operations) are (a) injection rate; (b) closing time; and
(c) injection water temperature. The well design parameters (that need to be determined before drilling) include (a) shut-in
valve location; (b) height of vertical riser; (c) tubing diameter; (d) injectivity index; (e) sandface wellbore length; and (f) well
deviation. The case studies were performed by changing only one parameter while leaving other parameters the same as the
ones in the base case, except in the case of sandface wellbore length. Since the open area to reservoir changes with the
sandface wellbore length, the injectivity index was adjusted accordingly (Table 2). In the well deviation parameter, the build
section was artificially corrected as needed to achieve realistic wellbore trajectory, as shown in Figure 5.
4 SPE 146300
Displacement,ft
0 1000 2000 3000 4000 5000 6000 7000
0
2000
4000
TrueVerticalDepth,ft
6000 90deg
8000 45deg
0deg
10000
12000
Figure 5. Well Trajectories (True Vertical Depth versus Displacement) for 0, 45, and 90 deg Deviation Cases
SPE 146300 5
Terminology Nomenclature
The following four terminologies are defined to help with understanding the study results. Figure 4 displays the schematic
definitions of these four terms.
Pressure Decline (psi) : Maximum pressure Minimum pressure
Back Pressure Pulse (psi) : Final stabilized pressure Minimum pressure
Transient Duration (sec) : Time that it takes from the start of shut-in until the pressure fluctuation diminishes
within a certain specific amplitude
Water hammer : Pressure fluctuation or wave as shown in the shaded oval area in Figure 4
In particular, back pressure pulse is an important variable to help estimate sand production. When the back pressure pulse
occurs locally (the sandface in the well), it changes the flow direction and increases the instant velocity, creating a drag force
in the pores of the sandface. The drag force causes dismantling rock formation, enhancing sand production if it exceeds a
failure criterion. Therefore, the back pressure pulse will be the main parameter with transient duration for subsequent
geomechanical simulations.
Base Case
Figure 6 shows pressure response, valve opening, and water volume flow at standard condition for the injector in the base
case. The valve position (in blue) initially indicates one (fully open position) and goes to zero (fully close position) with the
designated closing time of 30 seconds. Accordingly the water flowrate (in red) decreases by the specified valve characteristic
curve and reaches zero when the valve is fully closed. The pressure at the top of the sandface (in black) decreases, shoots
down to a minimum, recovers a little bit and then fluctuates until it finally stabilizes at the end, which is water hammer. In
the base case, the pressure decline is 32 psi and the back pressure pulse is 7.3 psi with approximately 149 seconds transient
duration.
Figure 7 illustrates the pressure responses at the different locations: (a) just below at SSTV (0); (b) at SCSSV (1,718);
and at the top of the sandface, when the SSTV is closed. The biggest pressure pulse is observed at the place just below the
shut-in valve (SSTV) and the magnitude of pressure pulse tends to decrease as the measured point is deeper from the shut-in
valve. As mentioned earlier, pressure wave (that occurs at the shut-in valve) dissipates by friction along the pipeline (tubing)
during the propagation.
Figure 6. Pressure Response, Valve Opening, and Water Volume Flow at Standard Condition (Base Case)
6 SPE 146300
AttheTopofSandface
AtSCSSV(1718)
JustBelowatSSTV(0)
Figure 7. Pressure Responses at the Different Locations: (a) just below at SSTV (0); (b) at SCSSV (1718); and at the Top
of the Sandface, when SSTV is shut in
Velocity Mechanism
Injection rate and tubing diameter are the variables depending on velocity mechanism for water hammer. Higher velocity (by
increasing injection rate or reducing tubing diameter) increases the difference of cross-sectional average velocity (increases
momentum change), resulting in higher water hammer effect. It can be easily explained with Joukowskys fundamental
equation, Equation (1): P is proportional to V. Figure 8 illustrates the pressure responses of three different water
injection rates: 25K, 50K, and 75K BWPD. As the injection rate increases, the pressure decline increases with the required
well head pressure. Accordingly, the back pressure pulse also increases proportional to injection rate and the pulse lasts
longer for higher injection rate.
The influence of tubing diameter on water hammer is shown in Figure 9. Increasing tubing diameter is a practical way to
decrease water velocity, reducing water hammer. In particular, 9-5/8 diameter can reduce back pressure pulse by more than
one third, compared to 5-1/2 diameter for the same water injection rate. As expected, the well head pressure and the
pressure decline increases as the tubing diameter decreases. Transient duration also increases for smaller tubing diameter.
32.0psi
25KBWPD
15.5psi
Time[s]
Figure 8. Pressure Responses for Different Injection Rates (25K, 50K, and 75K BWPD)
SPE 146300 7
28.5psi
32.0psi
38.8psi
95/8
51/2
Time[s]
Figure 9. Pressure Responses for Different Tubing Diameters (5-1/2, 7, and 9-5/8)
28.2psi 26.6psi
32.0psi
CASE: WellHead Pressure BackPress. Transient
ClosingTime Pressure Decline Pulse Duration
(seconds) (psig) (psi) (psi) (seconds)
5secs 5 2368.0 78.9 49.9 146
30 2368.0 32.0 7.3 149
78.9psi
60 2368.0 28.2 3.5 165
100 2368.0 26.6 1.9 192
Time[s]
Figure 10. Pressure Responses for Different Closing Time (5, 30, 60, and 100 seconds)
(Figure 11). The range of temperature studied here covers possible field water temperature (32 to 104 F). The main
reason of temperature effect on water hammer is water compressibility. Higher water temperature increases water
compressibility, which enables it to absorb the pressure wave and consequently reduce water hammer. It can be supported by
Joukowskys fundamental equation (there is a fluid compressibility term in the right hand side of acoustic wave speed
formula). The observation that P is inversely proportional to the square root of the fluid compressibility can also explain the
reason of weak temperature dependence on water hammer. Less water hammer according to the increase of injection water
temperature is displayed by the decrease of the pressure decline, the back pressure pulse, and the transient duration, as shown
in Figure 11.
28.6psi
29.7psi
32.0psi
Time[s]
Figure 11. Pressure Responses for Different Water Injection Temperature (32, 68, 104 deg F)
29.8psi
30.8psi 0ft
32.0psi 1718ft
7753ft
Time[s]
Figure 12. Pressure Responses for Different Shut-In Valve Locations (0, 1718, and 7753 ft)
2218ft
29.8psi
32.0psi
35.1psi
0ft
5000ft
Time[s]
Figure 13. Pressure Responses for Different height of Vertical Riser (0, 2218, and 5000 ft)
10 SPE 146300
32.0psi 90
45.0psi 45
49.8psi 0
Time[s]
Figure 14. Pressure Responses for Different Well Deviation (0, 45, and 90 deg)
CASE:Inj.
WellHead Pressure BackPress. Transient
78.9psi
3x104 kg/s/pa Index Pressure Decline Pulse Duration
(104kg/s/pa) (psig) (psi) (psi) (seconds)
0.2 2989.0 670.5 13.1 57
3 2390.0 56.1 9.5 136
28.2psi
7 2368.0 32.0 7.3 149
12 2359.7 22.5 6.3 154
Pressure[psig]
7x104 kg/s/pa
12x104 kg/s/pa
26.6psi
Time[s]
-4
Figure 15. Pressure Responses for Different Injectivity Indexes (0.2, 3, 7, and 12 x 10 kg/s/pa)
35.7psi 2000ft
32.0psi 3000 2365.0 29.8 9.1 161
3000ft 200ft
29.8psi
Time[s]
Figure 16. Pressure Responses for Different Sandface Length (200, 800, 1400, and 2,000 ft)
12 SPE 146300
16
14 13.8
BackPressurePulse,psi
12
10 10.1
9.1
8 7.3
5.9 7.3
6
6.0
3.5 5.2
4
VariedInjectivityIndex
2
2.2 ConstantInjectivityIndex
0
0 500 1000 1500 2000 2500 3000 3500
WellLength,ft
Figure 17. Pressure Pulse and Transient Duration for Different Sandface Length (200, 800, 1,400, 2,000 and 3,000 ft)
Tornado Chart
Figure 18 shows a Tornado chart displaying the sensitivity trends for the back pressure pulse. Note that the trends are only
applicable to the water injector in this specific project: the trends may be different for different injector cases. Although the
parameter ranges are arbitrary, it is worthy of preparing such a chart to understand the impact of variables on water hammer,
which can help determine optimum operational and design strategy of water injector.
In this project, the closing time is the most sensitive parameter and the injection water temperature has the least impact.
Only increasing the closing time from 5 to 60 seconds enables a reduction in the potential water hammer by approximately 25
times. If this well is in the planning stage for drilling, it would help to considerably reduce or eliminate the potential water
hammer by using larger tubing size (9-5/8) and placing the shut-in valve at 7,753 ft (just above the top of sandface). It is
also highly recommended that the valve be closed for more than 60 seconds.
Conclusions
This paper discussed the potential water hammer with an actual water injection well in Chevrons deep water project and
showed how operating parameters and design parameters can affect water hammer. The following observations are made.
Deep sea water injection well operations have a higher possibility of water hammer than shallow well operations.
Water hammer could cause failure of formation integrity, resulting in sand production. It may also damage the
wellbore and downhole completions.
Back pressure pulse and transient duration are the major parameters to evaluate the possibility of sand production
and determine the sand volume, if any, in geomechanical simulations.
Water hammer is most pronounced at the location of the shut-in, and tends to fade away father from the shut-in
location due to friction.
From the side of operating parameters, (a) reducing injection rate; (b) increasing closing time; and (c) increasing
injection water temperature would help to reduce or eliminate water hammer. The most significant variable
affecting water hammer would be the closing time.
For well design parameters, water hammer can be reduced or eliminated by (a) placing the shut-in valve as deep as
possible; (b) reducing the straight vertical section of the well; (c) increasing tubing diameter; (e) increasing
injectivity; (f) increasing the sandface length (up to the optimum length); and (g) drilling a horizontal well rather
than vertical.
The results would be different case by case. Therefore, water hammer needs to be thoroughly investigated and
properly managed for each water injector at the well design stage as well as in field operations.
Nomenclature
a Acoustic wave speed
d Derivative
g Gravitational acceleration
H Piezometric head
V Cross-sectional average velocity
Fluid density
Difference
Acknowledgements
The authors would like to thank Chevron Energy Technology Company for permission to publish this paper.
References
1. Weston, E. B.: Description of Some Experiments Made on the Providence, RI Water Works to Ascertain the Force of
Water Ram in Pipes, Trans. Am. Soc. Civ. Eng. 14, p. 238, 1885.
2. Carpenter, R. C.: Experiments on Waterhammer, Trans. ASME, 15., 1893.
3. Frizell, J. P.: Pressures Resulting from Changes of Velocity of Water in Pipes, Trans. Am. Soc. Civ. Eng. 39, pp. 1
18., 1898.
4. Joukowski, N. E.: Memoirs of the Imperial Academy Society of St. Petersburg, 9(5) (Russian translated by O Simin
1904), Proc. Amer. Water Works Assoc. 24, pp. 341424, 1898.
5. Wood, F. M.: The Application of Heavisides Operational Calculus to the Solution of Problems in Waterhammer,
Trans. ASME 59, pp. 707713., 1937.
6. Rich, G.: Waterhammer Analysis by the Laplace-Mellin Transformations, Trans. ASME, pp. 194445., 1944.
7. Jaeger, C.: Engineering Fluid Mechanics translated from German by P.O. Wolf, Blackie, London, 1956.
8. Parmakian, J.: Water-Hammer Analysis, Prentice-Hall Englewood Cliffs, N.J., 1955 (Dover Reprint 1963).
9. Streeter, V. L., and Lai, C.: Waterhammer Analysis Including Fluid Friction, Trans. Am. Soc. Civ. Eng. 128, pp.
14911524, 1963.
10. Morita, N., Davies, E. and Whitebay, L.: Guidelines for Solving Sand Problems in Water Injection Wells, SPE 39436
presented at SPE Int. Symp. on Formation Damage, Lafayette, Louisiana, Feb. 18-19, 1998.
14 SPE 146300
11. Santarelli, F.J., Skomedal, E., Markestad, P., Berge, H.I. and Nasvig, H.: Sand Production on Water Injectors: How Bad
Can It Get? SPE Drill. & Completion, 15, no.2, 132., 2000.
12. Hayatdavoudi, A.: Well Sanding and Lost Production Due to Cyclic Water Hammer, SPE 100928 presenated at SPE
ATCE, San Antonio, Texas, Sept. 24-27, 2006.
13. McCarty, R. A. and Norman, W.D.: The Resiliency of Frac-Packed Subsea Injection Wells, SPE 102990 presented at
SPE Annual Technical Conference and Exhibition held in San Antonio, TX, Sept. 24-27, 2006.
14. Vaziri, H., Nouri, A., Hovem, K., Wang, X.: Compution of Sand production in Water Injectors, SPE 107695 presented
at the European Formation Damage Conference, Scheveningen, The Netherlands, May 30-Jun. 1, 2007.
15. Wang, X., Hoven, K., Moos, D., Quan, Y.: Water Hammer Effects on Water Injection Well Performance and
Longevity, SPE 112282 presented at SPE International Symposium and Exhibition on Formation Damage Control,
Lafayette, Louisiana, Feb. 13-15, 2008.
16. Tang, Y., Ouyang, L.: A Dynamic Simulation Study of Water Hammer for Offshore Injection Wells to Provide
Operation Guidelines, SPE 131594 presented at CPS/SPE International Oil & Gas Conference and Exhibition, Beijing,
China, Jun. 8-10, 2010.
17. Lahlou, Z. M.: Water Hammer, Tech Brief National Drinking Water Clearing House, 2003.
18. Ghidaoui, M. S., Zhao, M., Mclnnis, D. A., Axworthy, D. H.: A Review of Water Hammer Theroy and Practice,
Applied Mechanics Reviews, 58, pp. 49-76, 2005.