Spiral Wound Gaskets - Poor Pump Performance (Update from 2015) 2

[JJPellin]

Three years ago, I posted a question about the use of spiral wound gaskets pressurized from OD to ID. The real issue I was exploring was a set of large barrel pumps that were chronically underperforming. I was speculating that the gasket stack between the outer head and inner casing was allowing leakage from discharge back to suction as a result of the configuration of the gaskets.
Refer to thread378-381741.

This question went unresolved for three years until last week. The problem was recirculation from discharge back to suction. But, it was not occurring by leakage past the spiral wound gaskets. The problem ended up being a piping error that occurred in 1994 when the pumps were originally installed. This is a large barrel pump with an axially split inner casing. The outer barrel has two case drains. The cross sectional drawing suggested that both of these drains were connected to the portion of the casing normally under discharge pressure. This drawing was wrong. One of these drain connections actually breaks through into the suction chamber inside the casing. These two case drains were piped together with a common drain line out to a single valve at the edge of the skid.

The arrangement of this case drain piping had created an accidental spill-back path for discharge back to suction. For these particular pumps, the differential pressure is in the range of 1750 psi (4,900 feet or almost 1500 meters TDH). This spill-back created two problems. First, it reduced the developed head of the pump by almost 20 percent. The other problem was a result of the fact that these pumps pass a small amount of abrasive solids in the stream. The location of the discharge drain connection was at the bottom of the case and at the largest possible diameter. This was acting like a centrifuge, concentrating the abrasive solids and returning them to the pump suction. This resulted in accelerated wear to bushings and wear rings.

A drawing error in 1994 showed an inaccurate representation of the location of the case drains. A piping designer created a drain manifold based on this inaccurate drawing. The result was 24 years of wasted energy and at least half a dozen pump overhauls that would have otherwise been unnecessary. I am estimating the cost of that error in the range of $6 million (US). The small details really do matter.


Johnny Pellin

 

[Latexman]

Amazing. Thanks for the follow-up Johnny.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[LittleInch]

Somewhere I worked a good few years ago on a hot oil line, they never made the planned capacity, but basically put it down to poor assumptions / basic cool down calculations (days of hand calculations).

Fast forward 20 years and the maintenance group had a new machine and set of devices to measure pump efficiency to track wear and drop off in efficiency. Efficiency of the pumps was very bad compared to pump curve. Strip down of pump - All checked out. Put pump back together again - still poor efficiency. Bit of head scratching. Someone then noticed motor going round the wrong way. Swapped two three phase leads - instant increase in flow of 15%.....

If there was dirt in the line, I'm a little surprised the drain lines didn't wear out quite quickly??



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[JJPellin]

LittleInch,

Actually, the drain lines have worn out at least once or twice. Thin pipe is managed by a different group. They replaced the lines, in-kind without consulting with the rotating machinery group. Someone could have recognized the problem by noting that we had high velocity erosion in a line that is supposed to be a dead-leg. But, not one put two and two together. Instead, they just increased the inspection frequency on those drain lines.



Johnny Pellin

 

[LittleInch]

JJ,

Wow, sometimes you just have to shake your head....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[TugboatEng]

You're lucky the pump case didn't get pressurized to the discharge pressure though that would have made the problem evident very quickly and possibly saved the company some money.

 

[racookpe1978]

I appreciate the feedback. We do field mods of turbines, GT and ST, bearings - it shows the care we (my crews) need to cut and drill exactly right.

 

[GD2]

JJPellin,
The Vendor usually supplies all piping drawings in a skid with all tie-in points at the skid-edge. Are you saying it was a wrong design of drain locations on the barrel by the pump manufacturer or piping design fault after the skid which is outside the scope of the manufacturer?

GDD
Canada

 

[JJPellin]

This piping was designed by our designer and was not designed or provided by the pump manufacturer.

Johnny Pellin

 

[JJPellin]

The pump manufacturer did provide the inaccurate cross-sectional drawing that contributed to the error by our piping designer.

Johnny Pellin