Occasionally, you may spot something in an article or an equipment installation that you consider really clever and worth remembering for future use. One example is injecting gas into a centrifugal pump suction to reduce cavitation damage and improve pump performance. I came upon this years ago. It involved injecting air into a booster pump moving cooling water to the top of the cooling tower. This seemed odd. Engineers spend a lot of effort preventing air getting into pump suctions. What was going on?
[pullquote]
Cavitation is the physical response of a system to meet mass balance requirements. As any process engineer knows, a mass balance requires that flow out equals flow in. Cavitating systems have restrictions on flow in. The restriction may be in the feed piping or in a limit of liquid available. If the system downstream of the pump can accept more flow out than can flow into the pump, something must change to make flow out equal flow in.
When discharge flow exceeds suction flow, the upstream accumulation on top of the pump suction changes (out – in = Δ accumulation). The accumulation continues to drop until the pump capacity drops to out – in = 0. Cavitation accomplishes this. Vapor formation in the inlet line to the pump lowers the discharge head of the pump, decreasing the pump discharge capacity against the downstream pressure requirements. The out flow rate drops until it reaches the in flow rate.
We express the required accumulation to avoid cavitation as a liquid height over the pump suction, or a net positive suction head required (NPSHR). The NPSHR is determined by testing a pump and measuring the discharge pressure. The Hydraulic Institute defines NPSHR as the pump suction head that reduces the discharge head by 3%. NPSHR isn’t the NPSH that equals no cavitation. Rather, NPSHR is the suction head where a specific amount of cavitation already is occurring.
Cavitation damage comes from vapor bubbles in the pump collapsing and creating a damaging shockwave. The maximum damage happens at a suction head just above the NPSHR value.
I recently reviewed an installation having a cooling tower booster pump with an NPSHR of 12 ft and an NPSHA (available) of 14.5 ft. Maintenance history showed extensive and frequent repairs on the pump. The operation with the 14.5-ft NPSHA was very close to the worst operating point possible. Total collapse of vapor pockets was causing severe damage.
The solution was to use air injection. Air bubbles in the cooling water don’t completely collapse. This reduces the damage to the pump. You must take care to avoid too high an air rate, which can cause the pump to lose prime and stop working.
We hooked up instrument air to the cooling water pump suction. We installed two check valves in the instrument air injection line followed by a ⅜-in. needle valve and then a pressure regulator. This is more than strictly needed but dealt with some concerns from the plant operators. The pressure drop across the system was high enough for the orifice valve to operate as a critical flow device. The air flow rate didn’t vary with cooling water system pressure. We tried various settings to find a stable operating point and found that at a roughly 11-psig regulator pressure and a specific position on the needle valve, the system was very stable.
With the air injection, pump noise and vibration dropped dramatically. Over an extended period, pump maintenance needs fell and reliability improved. The best solution, of course, would be to have a system correctly constructed for reliable operation without this type of patch. Nevertheless, air injection did help substantially here.
Air injection is acceptable in this system because the water goes straight to the top of the cooling tower. The air requires no downstream handling. Process systems often are more difficult to address because of problems with injecting inert gas.
