Common Causes of Waterhammer
As discussed in the article What is Waterhammer?, there are many different ways a surge event can be initiated. Most generally, any time there is a rapid change in fluid velocity there is a high likelihood for a waterhammer event.
Below outlines some of the most common causes engineers should be aware of when designing or troubleshooting a piping system to avoid and mitigate waterhammer.
Valve closures are the most intuitive cause of a waterhammer event, and often the most considered. While closures are defined as a complete stoppage of flow, similar waterhammer consequences can be found just by rapidly changing a valve’s position. In both cases, any rapid and drastic change in the fluid’s velocity will create a surge event as the fluid’s kinetic energy is converted into potential energy in the form of pressure.
While words like “rapid” and “drastic” are used to describe closures that can result in waterhammer, these concepts are formalized with communication time and instantaneous transients. The valve’s closure time, effective closure time, and the system’s communication time can all influence how severe a waterhammer event can be. Most generally, think the faster the closure the worse the consequences.
Note that not all valve closures will create a waterhammer event. Slowing flow gradually allows the generated pressure to disperse through the system and approach intermediate operating points, spreading the pressure surge out over time. However, in many cases a valve closure does not have the freedom to gradually slow the fluid. Emergency shutdown valves for example have the very clear purpose of stopping flow as quickly as possible, antithetical to avoiding a waterhammer event. In that case, rather than avoiding the waterhammer event, analysis would likely focus on mitigating the waterhammer event.
Check Valve Slam
Check valves are added to a system to avoid reverse flow. Reverse flow can be caused by a pump tripping, or if there are several pumps in parallel with one shut down. Reverse flow through a pump should generally be avoided, hence the addition of a check valve downstream to protect the pump’s internals.
Check valves extend the foundational understanding of an intentional valve closure to a reactionary valve closure. A reactive check valve is difficult to control since its closure time depends on the check valve’s construction and the rate of reverse flow. If the check valve’s closure allows significant reverse flow only to halt it rapidly, the check valve will slam and cause a waterhammer event.
There are two schools of thought when closing check valves to avoid check valve slam. The first is to close the check valve very gradually to halt an allowable amount of reverse flow. Conceptually this is similar to the gradual closing of a normal valve. The second option is to close the valve as soon as flow reverses, preventing any significant reverse flow through the closing check valve. Closing a check valve instantly may seem antithetical to avoiding waterhammer, however the instantly closing check valve prevents any reverse flow before it can happen. By avoiding a large reverse velocity, the closing valve doesn’t need to rapidly decelerate the fluid, thus avoiding a waterhammer event. Check valves can prove useful for mitigating waterhammer as well.
While pump trips may cause a waterhammer event through check valve slam, starting up a pump can cause similar waterhammer consequences. Generally pumps should be started-up against a partially closed valve. This not only helps the pump avoid burning out its motor without a load, but it also helps avoid pressure waves created by a starting pump. These pressure waves can travel through a system and cause similar waterhammer consequences.
The shape of a pump’s curve can also impact its potential for waterhammer. For example, a flat pump curve will have more drastic waterhammer effects as small changes in pressure requirements cause significant changes in flow. This is different from a steep pump curve which requires massive changes in pressure requirements to significantly change its flow. This is demonstrated visually in Figure 1 below.
Figure 1: For similar reductions in flow, a steep pump curve requires the system to cause much more pressure drop than a flat curve. This makes the steep pump curve less sensitive to slight changes in valve position, making waterhammer less likely.
As with valves, pump transients should avoid rapidly accelerating or decelerating the fluid. Most commonly this is due to pump start-ups aggressively approaching steady state, rather than a smooth gradual approach. Similarly, during pump trips the fluid can be rapidly decelerated resulting in a waterhammer event. There are several ways to slow these transients. The addition of a flywheel would increase the pump’s rotational inertia, slowing changes in momentum. Variable Frequency Drives (VFDs) can gradually bring the pump to its steady state speed or slow it toward shutdown. Soft starters, which focus on pump start-up, can also make the pumps rise to steady state more gradual, avoiding waterhammer events.
Of course, many of these causes can be interrelated, and the consequences of their waterhammer events compounded. A pump may trip from insufficient NPSH due to an upstream valve closure, causing reverse flow to slam a check valve. While many of these may be checked easily independently, it is also important to understand how interaction between components impact wa