Taking the ONEPROD Falcon to the field is a big part of the 2-day training we provide on our entire Falcon line of condition monitoring tools. When we go to the field we always find interesting issues to analyze and discuss. A recent check of a centrifugal pump yielded one of those issues.
The pump in question always has high vibration and has a mechanical seal failure every other week. Total cost for this is around $1,700 per seal and 6 total man hours between two techs. At the end we will do a little math and see how much we can save if we can resolve the cause of these failures. (Spoiler alert, we did!)
The pump did in fact have high vibration when we checked it. The dominant vibration is at 1x running speed and it is highest in the pump. Knowing that looseness may be a factor we used the live mode on the Falcon to see if we could find a vibration source or location. We checked various locations around the pump and found the vibration to be about equal.
We worked our way down towards the base and noticed a change as we went from pump to base. Turns out the pump drive side support was loose. After tightening it the amplitude did go down a bit, but still quite a bit of vibration remained. We found a symptom not the root cause. We took more readings on the base itself and noted very high amplitudes which then dropped off on the grout base riser.
So, going back to the base we noted that amplitudes varied greatly across the pump and were very high directly under the wet end and at the support we found loose in the beginning. We theorized that the grout may have deteriorated, separated or maybe didn’t even exist under these locations. We then employed a very low-tech tool to see if the base sounded solid or hollow. We found an old ¾ 10x 6 bolt on the floor and used it to “hammer” on the base listening as we went. As we suspected the base sounded very hollow exactly in the spots of elevated 1xRPM vibration.
As you can see in the image below the grout was not supporting the pump base very well. Whoever poured the grout didn’t allow it to flow and fill in all the voids. These pockets allowed the base to flex in the areas it needed to properly support the pump.
Remember that seal we talked about earlier? Well after this issue was corrected it hasn’t been changed and no one has needed to work on it since.
Like we said, a little math will tell us how we did. If the seal failed every other week, that’s 26 events over a year and cost around $44,200 in seals alone. Throw in around 156 hours of labor at $50.00 cost and you can add in $7,800 there for a grand total of $52,000 savings over a years’ time. That’s a good amount of savings for one find in a new Reliability Program, plus I’d also say most mangers would love to have a reoccurring failure issue fixed and be able to claim that kind of cost savings.