Stop Swapping Your Pumps and Start Testing Them Instead

Anybody have two pumps that are installed in parallel so that if one fails you can swap to the other pump with little or no consequence to operations?  I know most of you do, and if you have read the title of this article you already think I’m crazy.  Because, if you don’t swap your pumps then how do you know the “stand-by” will work when you need it?  Well, I’m going to tell you how, but first I’m going to explain why most folks are swapping their run time between pumps and why this is wrong.

Over my years in the paper industry, I’ve only seen two reasons for an installed spare pumping arrangement: 

1. The designers felt that the consequence of failure was too severe to only have one pump, and only with a backup pump could the availability of the function be secured to a tolerable level.  Or,

2. Money was really no object.  I never actually experienced this phenomenon, but over the years, the more seasoned folks I’ve run across talk about a time when capital expenditures were paid for with pure cash flow.  They literally couldn’t spend the money faster than they were making it.  Consequently, they either installed back up pumps for a rainy day or because cost benefit justified it.

Now, there are other reasons to have an installed spare pumping arrangement.  Waste water pumping station come to mind as a very unique example, but for the majority of the pulp & paper industry those installed spare pumping systems are there to maintain process availability. 

Here comes a tricky part, because in most cases, very few of us have seen a green field mill come up out of the mud.  We have learned our maintenance & operational practice through tribal knowledge, but in most cases, we have not been provided knowledge at all.  We have just been given tasks to do with no real explanation as to why.  As it relates to an installed spare pump arrangement, I dare say that the designers of those systems almost never intended for those pumps to be operated in an alternating run cycle.  I bet most designers intended for the installed spare pumping set to be operated in a true duty / stand by pump arrangement.  Where the duty pump is the primary running component and the stand by pump sets idle next to the duty pump and is ready, willing, and able to take over in the event the duty pump fails. 

So then why do we swap pumps?  Because, one day the duty pump failed and the standby would not start up.  When we opened up the stand by pump we found that it was “mudded in” because the isolation valves leak by.  So instead of fixing the isolation valves we began a routine of alternating from one pump to another.  I have seldom seen an organization that was really that good at swapping their pumps (exception being very critical systems like boiler feed water).  To be fair, they were really good at documenting the swap, just not performing it. 

Over the years I have seen many frustrated maintenance folks being called to fix both the duty pump and the stand by pump because neither was operational.  I actually know one maintenance superintended who had all the stand by pumps removed from his operational area.  Because in his words, “they (the stand by pumps) were always broke and maintenance doesn’t find out about it until the duty pump breaks as well.  So they (operations) really don’t have a stand by pump so they must really not need them.”  This is an extreme example, but I can tell you that their uptime actually increased.  They were forced to focus proactive maintenance on what became a standalone pumping system.  Now I’m not suggesting you remove your stand by pumps, but I would say that a strong proactive maintenance program is in order and alternating run time is not it.

So what is wrong with alternating pumps?  You are stressing the ware components equality.  So if (when) you have a ware related failure of the first alternating pump you only have a remaining life on the other pump equal to something less than the alternating frequency.  Most of the time, this remaining life will not fit into a standard planning & scheduling window so it becomes an emergency repair.  So, that’s what’s wrong with alternating pumps.  You wear them equally and they fail at about the same time. 

So what are you to do?  Pick one of the alternating pumps as your duty pump and run it, and periodically perform a full function test to insure the standby pump is available to take over.  I know this is easy to stay, but it is fairly straight forward to put into practice.  Beware though; you need to keep in mind that the failure modes of the stand by pump are different than those of the duty pump.  For instance, since the standby pump is setting idle most of the time, it is susceptible to failure modes the duty pump is not; such as, bernilling of the bearing, slumping of the grease, being “borrowed,” mudding in, etc.  So if any of these failure modes are likely, then you must put proactive plans in place to deal with each one of them. 

For example:

• bernilling you may just “bump” the motor on/off to set the bearing in a new position,
• the same procedure may address the slumping of grease,
• as for the “borrowing,” you may just have to do a visual check to make sure it’s still there,
• Mudding in, fix the leaking valves.  I know that’s easier said than done, so maybe run the standby pump so that you restore its inlet and outlet and then shut it back down.

Now, the one thing I have not addressed is how often these tasks should be performed.  For bernilling, slumping, and mudding you just have to have a reasonable idea how long each of these degradation processes take.  Then perform the restoration at an interval that is less than the degradation process.  As for the “borrowing” and the full function test there is a way to calculate the correct Failure Finding Interval, but that explanation is best covered in a more technical article.

To close, I would say I have lost this battle more times than I have won it.  This is a tough fight.  Typically, the folks in senior leadership will not take the time to fully understand this logic and are frighten by a change that has been culturally ingrained in their sole.  Also, most managers are very risk adverse, so unless there is a compelling reason to change they want.  In the case of an alternating pump arrangement, they have very little to personally gain from making a change.  In most cases, everyone is typically satisfied with the availability of these systems.  So why change, even if the system uptime could be driven much higher.  The perceived risk of the change does not warrant the effort in their minds.  The bottom line is however, alternating pumps is rarely the appropriate maintenance strategy and it puts your operation at risk of unnecessary downtime and possibly safety & environmental consequences.