The Art of Maintaining Stand-by Pumps

I want to maintain the highest possible reliability of 2 parallel centrifugal pumps. One is the operating pump, and the other is a stand-by pump, required only as a back-up in case the running pump fails. What is the best run-time strategy to maximize the reliability of the pumps?

My first thought was, “50:50, of course!” That way the wear and tear on the pumps is spread out over both pumps, doubling the effective lifetime of the equipment. Seems reasonable to me.

Unfortunately, if you are using this strategy, there is a good chance you are significantly accelerating the wear on the pumps, resulting in increased downtime!! Read all about it…

One assumption that has to be made: The pumps are using mechanical seals. Pumps with packing glands normally are wetted by the working fluid. These pumps will probably require a set cycle schedule for packing maintenance. This has been the strategy (weekly pump shifts) for packed pumps for years, and it hasn’t necessarily changed for mechanically sealed pumps.

So why is a 50:50 run strategy bad?
First, the major wear and failure factor when considering a mechanical seal is the number of start-stop cycles, not overall run time. Starting and stopping the pumps solely for equal run time puts enormous stress on the seals
Next, 50:50 introduces many more failure modes than are present in a standby pump.
Finally, with perfectly even wear, both pumps (theoretically) should fail at about the same time. Not the ideal situation for an emergency standby pump!

It seems, then that the fewer start-stop cycles, the better. Ideally, as far as mechanical seal wear goes, the stand-by pump should never be started, maintaining it in pristine condition, ready to take over on the loss of the duty pump.
The problem with this, however, is that you no longer have confidence that the failure modes specific to the stand-by pump (fail to start, failure to reach full capacity) are not present.

A good compromise is a 90:10 ratio. For example, run the duty pump for 8 weeks, then run the standby pump for a full 8-hour shift. Then SHIFT BACK TO THE DUTY PUMP. This has several advantages:
– You have confidence the pump will run when needed.
– It will prove it can reach full load capacity
– It can be scheduled around your normal PdM periodicities. For example, conduct thermal and vibration analysis of the standby pump at the 2-month point, killing 2 birds with one stone.
– Most people shift pumps weekly, which is a total of 104 starts or stops for the 2 pumps over the course of a year. The 90:10 strategy lowers this to only 10 or 12 total.

Some facilities have actually color coded their pumps. The duty pump is green, and the standby pump is red. When somebody sees the red pump running, they can now question why we are not in the “reliable” line-up. The forces the operators to immediately report failures of the duty pump.

This philosophy may not “feel right”, but there is plenty of data to back it up.

Again, this assumes that the prevalent failure mode is seal failure. It also assumes that there are not other extenuating circumstances requiring pump shifting. For example, maybe you have a history of false brinelling of the standby pump bearings if the pump is idle for xx weeks.

Take a look at your strategy. You may find you are able to increase your equipment availability, reduce downtime, and limit repair costs, just by adjusting your pump switching schedule.