Causes of Water Hammer
Causes of Water Hammer
Causes of Water Hammer
pressure surge that can arise in any pumping system that undergoes an abrupt change in its rate
of flow and usually results from pump starts and stops, the opening and closing of valves, or
water column separation and closure. These abrupt changes can cause all or part of the flowing
water column to undergo a momentum change. This change can produce a shock wave that
travels back and forth between the barrier that created it and a secondary barrier. If the intensity
of the shock wave is high, physical damage to the system can occur. Oddly enough, it can be
more of a concern in low pressure applications.
Water hammer is yet another example of conservation of energy and results from the conversion
of velocity energy into pressure energy.
Since liquids have a low compressibility, the resulting pressure energy tends to be high.
Perhaps the best way to visualize this action is to start with a hypothetical example. Figure 1
below shows a pump pumping water into a pipe that was empty when the pump started. The two
valves, located at the pump discharge and the far end of the pipe, are fully open and have the
ability to close instantaneously. The pipe, valves and other fittings are entirely inelastic and no
volume change can occur, regardless of the pressure. The column of water flowing through the
pipe also has a perfectly flat leading edge that matches that of the cross sectional ID of the pipe.
When the leading edge of the water column reaches the downstream valve, it closes at nearly the
speed of light and entraps no air ahead of the water column.
Figure 1
Even though the leading edge has struck the closed valve, flow into the pipe continues for the
next few milliseconds. Just as flow ceases, the upstream valve closes (this time at the true speed
of light), and the water column is completely isolated between the two valves. What events occur
as the column strikes the closed, downstream valve and why does water continue to enter the
pipe even though the valve is closed?
If this moving column was a column of metal instead of water (hypothetically, of course), a
couple of things could occur. Depending on its coefficient of restitution (its ability to avoid
permanent damage), the kinetic energy due to flow (motion) could be transformed into
mechanical energy as the leading edge of the metal column is crushed against the closed valve. If
this occurred, the column would come to rest and remain motionless at the valve. If its restitution
is high enough to prevent crushing, that same kinetic energy could be used to reverse its
direction in the form of a bounce. Regardless of the outcome, the "entire" metal column would
either come to rest or bounce in the opposite direction. Neither of these events occurs when
water is involved.
Water is a nearly non-compressible liquid, which seems to suggest it is slightly compressible. At
ambient temperature, 1-psi will decrease its volume by about 0.0000034 percent. That seems
pretty small, but the larger the volume, the easier it is to see the effect. For example, if water did
not compress, sea level would be roughly 100-ft higher than its current level! At very high
pressures, say 40,000-psi, its compressibility is increased to about 10 percent. But, most water is
not just water-it also contains air, which is primarily nitrogen (78 percent) and oxygen (21
percent). Otherwise, fish could not survive! Dissolved air composes about 2 percent of a given
volume of unprocessed water, and adds substantially to its compressibility.
Why
It is water's (and dissolved air's) compressibility that causes water to act differently than the
metal column. Were it not compressible, its leading edge would be permanently crushed or the
entire column would bounce backward. When the leading edge of a water column strikes the
closed valve it abruptly stops. Since the water behind the leading edge is still in motion, it begins
to compress. This compression along the entire length of the column allows a small amount of
water to continue to flow into the pipe even though the leading edge has halted. When flow
ceases, all of its kinetic energy of motion and that due to compression is converted into pressure
energy.
Compression begins at the leading edge of the water column and since the additional energy it
produces cannot continue past the closed valve, a pressure or shock wave is generated and travels
along the path of least resistance which, in this example, is back upstream. Its inception is similar
to the echo produced when a sound wave, traveling through air, strikes a similar barrier. When
the wave hits the upstream valve, it is reflected back downstream but with a diminished intensity.
This back and forth motion continues until friction and reflection losses cause the wave to
disappear. The speed at which a wave travels and the energy it loses during travel depends on the
density and compressibility of the medium in which it travels. The density and compressibility of
water make it a good medium for shock wave generation and transmission.
The pressure waves created by hydraulic shock have characteristics similar to those of sound
waves and travel at a similar velocity. The time required for a water hammer pressure wave to
negotiate a length of pipe is simply the pipe length divided by the speed of sound in water
(approximately 4,860-ft/sec). In water hammer analysis, a time constant that is often used
describes the progression of the wave from its inception to the secondary barrier and then back
again. It takes the form of Tc = 2L/a (where L is the pipe length and a is the velocity of the wave,
which is the speed of sound). In a 1,000-ft pipe, the wave can make a complete round trip in less
than one half second.
The pressure created by this shock wave is directly proportional to both the wave velocity and
the velocity of the water flowing in the pipe. Although the equation below does not take into
account the effect of pipe length, diameter and elasticity, it will provide some insight as to the
additional pressure created by a water hammer pressure wave.
P(additional) = aV / 2.31g
P is the additional pressure the shock wave creates, a is wave velocity, V is the velocity of the
flowing water in the pipe in feet per second, g is the universal gravitational constant @ 32-ft/sec2
and 2.31 is the pressure conversion constant. At a pipeline velocity of 5-ft/sec, the additional
pressure created by the shock wave is approximately 328-psi. Increasing that velocity to 10-ft/sec
increases the additional pressure to about 657-psi. Obviously, systems that are not designed to
accommodate such an increased pressure are often damaged or even destroyed.
Both columns of the table illustrate the proportional influence of velocity and length—pressure
increases as they increase. The lower values in the right hand column illustrate the inverse
relationship of time; these pressures are half those in the left hand column because the closure
time has doubled. The value of L is often fixed and depends upon the application, but we can
exercise substantial control over the other two variables. By doing so, we can eliminate or
greatly reduce water hammer's effect.
I mentioned earlier that pipe diameter and the elasticity of its material also influence the pressure
generated. Larger diameters and more elastic materials absorb some of the intensity of the shock
waves and therefore reduce the pressure generated. Several pipe manufacturers publish curves or
tables that show the potential water hammer pressure increase for various pipe diameters and
materials.
Suppose that the branch line valve is closed. If it is opened quickly the effect is similar to that of
quick closing. When the valve is opened quickly, the branch line sees an immediate drop in
pressure, and incoming water from the main line accelerates the previously static column. As
friction and other factors restrict its flow, the forward portion of the column can act as the initial
barrier and give rise to water hammer. Usually its effect is much smaller than that of valve
closure and is often referred to as a "surge." Still, under certain conditions, this surge can be
damaging.
Water hammer’s effects can be more significant in low pressure systems. The additional pressure
generated by a shock wave is proportional to the length of the pipe and velocity of the water
flowing in it and is completely independent of its operating pressure. Therefore, the shock wave
created in a 1,000-ft pipe flowing at 5-fps will be the same whether the operating pressure is 50-
psi or 200-psi. The difference is the ratio of shock pressure to design pressure can be
significantly higher in the low pressure system; therefore the potential for damage will also be
greater.
Water column separation can occur when a pump is stopped and the water column reverses
direction or in condensate lines where high temperatures can mitigate the need for a large
pressure drop. Although both forms can be extremely damaging, condensate lines tend to be far
more dangerous. The shock waves generated by column closure can travel in opposite directions,
and if they hit secondary barriers they can be redirected back toward one another. It would not be
unusual for these reflected waves to increase in intensity when they collide. This is certainly the
case with water and voltage waves and may account for the often greater damage resulting from
closure-initiated water hammer