When a pump stops, the water it was pushing does not stop with it. The column in the discharge pipe keeps moving forward for a moment, then falls back and reverses toward the pump, and the discharge check valve has to catch that reversal. If the valve is still partly open when the flow slams into it, the disc snaps shut against a moving column of water and produces a loud bang and a pressure spike. That event is check valve slam, and it is one of the most common sources of surge damage in a pump station. This guide explains why it happens, why a sudden power loss is worse than a controlled stop, and how valve selection and a soft ramp-down keep the valve from slamming.
Check Valve Slam in one line: Check valve slam is the sudden banging and pressure surge that happens when a discharge check valve snaps shut on reversing flow after a pump stops. The forward water column decelerates, reverses, and drives the check disc closed hard, sending a pressure wave back into the pump and pipe. It is worst when the pump loses power abruptly, because the flow reverses quickly while the disc is still open, and it is reduced by using non-slam or controlled-closing check valves and by ramping the pump down gradually rather than tripping it.
Which page do you need? This page focuses on causes, damage, and prevention. For what check valve slam is and how it is detected, see Pump-Station Check Valve Slam.
A discharge check valve exists to stop flow from running backward through the pump when the pump is off. While the pump runs, forward flow holds the disc open against its seat stop. When the pump stops, the energy driving the flow disappears, the forward column decelerates, and gravity or the downstream static head begins to pull the water back toward the pump. The check valve is supposed to close during that brief window as the flow crosses through zero, sealing before any real reverse velocity builds up.
Slam occurs when the valve is too slow, or its disc too heavy, to close in that window. Instead of sealing at the moment of zero flow, the disc is still hanging partway open when the reversing column arrives, and the water accelerates backward until it slams the disc shut against its seat. The impact of the disc and the abrupt stop of the reverse column together generate a pressure wave, and that wave is what you hear as a bang and what the pipe, valve, and pump feel as a mechanical shock. The heavier and slower the disc and the faster the flow reverses, the harder the slam.
The severity depends on how quickly reverse velocity develops before the disc closes. A short discharge line into an open tank with modest head reverses gently, and even an ordinary swing check can catch it without a violent slam. A long force main with high static head, or a system with significant elevation, reverses hard and fast, and the same swing check will slam badly. This is why slam is a station-by-station problem that depends on the piping and the head, not just on the valve itself.
The single biggest factor in how hard a check valve slams is how fast the pump loses its ability to push. When a pump is stopped normally through its drive, it can be ramped down over several seconds, so the forward flow bleeds away gradually and the reversal, when it comes, is slow and the disc has time to seat gently. When the pump instead loses power abruptly, on a utility outage or a breaker trip, the motor torque vanishes almost instantly, the flow collapses far faster, and the reversal arrives before a slow disc can close. That is why an uncontrolled power failure is the worst case for check valve slam.
The pump's own inertia matters here too. A pump and motor with significant rotating inertia keep spinning briefly after power is lost and continue to push a little flow as they coast down, which softens the deceleration. A low-inertia unit stops almost at once, so the flow collapses and reverses very quickly. Small, fast pumps on high-head systems are therefore particularly prone to hard slam on power loss, because they combine the fastest flow collapse with the strongest reversing head.
This distinction is why mitigation focuses on two fronts. On a normal stop, the control system can and should ramp the pump down so the check valve is never asked to catch a fast reversal. But the control system cannot ramp down a pump that just lost power, so for the power-failure case the protection has to come from the valve itself and from surge equipment, since there is no opportunity to slow things down electrically once the supply is gone.
The valve-side fix is to pick a check valve designed to close fast and quietly. Non-slam check valves, such as spring-assisted or nozzle-style silent checks, use a light disc and a spring that begins closing the disc as soon as forward flow decays, so the valve is nearly seated by the time flow reverses and there is little reverse velocity to slam against. Where slam is severe, a controlled-closing check with an external oil dashpot lets the disc close quickly for most of its travel and then cushions the final seating, absorbing the impact. Matching the valve type and its closing characteristic to the station's head and reversal speed is the core of the mechanical solution.
The control-side fix is to remove the reversal on any stop the drive can manage. A variable frequency drive can ramp the pump speed down over a set deceleration time so flow tapers off smoothly and the check valve seats before any reverse column forms, turning a slam into a soft close. The station logic sequences this as a soft stop: on a normal or lag-pump stop, command the ramp-down rather than an instant off, and only allow the check valve to take a hard reversal when there is genuinely no other option, such as a real power failure. On systems with a pump control valve, the logic can also stroke that valve closed on a timed profile ahead of stopping the pump so flow is already throttled to near zero.
A cloud SCADA platform such as Merobix ties this together for an unmanned station by both driving the soft stop and watching for the slam it is meant to prevent. The control logic issues the ramp-down and any valve stroking in the right order on every commanded stop, and it captures the utility-power and pump-status signals so it knows when a stop was a controlled one versus an abrupt trip. Trending discharge pressure across stops lets operators see whether a surge spike is appearing on power-fail events, and repeated hard slams or a rising spike often flag a stuck disc, a failed dashpot, or a spring-check that has weakened. Catching that pattern remotely turns a slow, damaging problem into a scheduled valve repair rather than a burst pipe.
A check valve slams when it is still partly open at the moment discharge flow reverses after the pump stops, so the reversing water column drives the disc hard against its seat. The forward flow decelerates when the pump loses drive, then falls back under the downstream head, and a slow or heavy disc cannot close in time. The impact and the abrupt stop of the reverse flow create a bang and a pressure surge. It is worst on high-head force mains and on abrupt power losses, where the flow reverses fastest.
On a normal stop the drive can ramp the pump down over several seconds, so forward flow tapers off and the check valve seats gently before any reversal builds up. On a power failure the motor loses torque almost instantly, flow collapses and reverses far faster, and there is no way to slow it electrically because the supply is gone. A slow disc cannot catch that fast reversal, so it slams. This is why power-fail slam has to be handled by the valve and surge equipment rather than by control logic.
Use a non-slam or controlled-closing check valve that begins closing as forward flow decays so it is nearly seated before flow reverses, and on any stop the drive can manage, ramp the pump down with a VFD so the reversal never forms. The station logic should sequence a soft stop on normal and lag-pump stops and reserve the hard reversal only for genuine power failures. Where slam is severe, a dashpot-cushioned check or a timed pump control valve absorbs the remaining impact.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.