Mechanical Seal Damage Due to Suction Pressure Spike

[swern]

The plant I work at has two centrifugal compressors in parallel. The two compressors have a poorly designed surge control system such that the recycled discharge volume is considerably greater than the suction volume. Because of the volume disparities, the suction pressure spikes during a trip which momentarily creates a negative DP across the mechanical seals. The mechanical seal faces become damaged after only a handful of trips. One option is to modify the anti-surge loop; however, for various intercompany reasons this will probably be shot down. Another option is to provide a backpressure control valve on the volume between the inboard "oil guard" laby seal and the mechanical seal to diminish the pressure spike seen by the mechanical seal. Any suggestions?

Thanks,
Swern

 

[mucour]

I am not clear with your description. If the compressor anti-surge control valve opens and it causes the suction inlet volume flow to increase because of a poor antisurge control design, then I do not see how this leads to a pressure spike. What I see is infact a reverse, ie. the suction pressure will have to reduce to cause an increase in flow.

How do you intend to provide a backpressure control valve (BPCV) on the volume between the inboard "oil guard" laby seal and the mechanical seal to diminish the pressure spike seen by the mechanical seal?
What kind of mechanical seal do you have in the compressor?

How will you install a BPCV between the laby seal and the mechanical seal? Please expain. My understanding is that BPCV can only be installed on a piping. The piping connections to a compressor mechanical seal are oil or gas barrier supply inlet line and drain or exhaust (if gas). So I am just wondering how you intend to install the BPCV between the laby and the mech. seal?

The sytsem resistance curve will move to the left of the performance curve during surge. Also note, if the compressor is surging and it is not averted, it leads to axial movement of the rotor which can damage the bearings, seals etc. The gas pressure is increased during surge. If the barrier oil or gas used for the seal is not high enough in pressure to prevent the casing gas from escaping, it will pass through the mechanical seal faces. If this gas is dirty, it can damage the seal faces. Those are what I identified has possible root causes of your mechanical seals failure.

 

[pmover]

swern,

sorry, but if the problem is the anti-surge control system or a component thereof, then i'd highly recommend fixing the problem. i will assure that continued operation of "questionable" anti-surge control system may very well lead to further damage to compressor. thus, increased repair cost and downtime (lost revenue). i'd definitely spend money where the problem is and not otherwise.

short of that, will a check valve work? thoroughly investigate the piping/seal arrangement and conduct an analysis of what ifs!

mucour,
inlet/suction pressure of a compressor does not necessarily decrease when anti-surge control systems are activated (recycling gas). the purpose of anti-surge is to reduce the head across the compressor at the particular unit rpm such that a stable flow exists within the compressor. i suppose it depends upon system configuration.

an aside, i wonder where the compressor inlet pressure is monitored at this particular unit.

good luck!
-pmover

 

[rzrbk]

I am unclear on how the recycle loop would cause a pressure spike in the suction; the compressor should not be able to pressure its own suction. I am assuming that the suction pressure rises due to a settling out effect (averaging the discharge and suction pressures). If this is the case, it the seal gas injection should not allow a pressure reversal.

If I am reading correctly, the recycled (spillback) volume much grater than the incoming gas seems to indicate a grossly oversized compressor, although if this was the case, you would not be running two in parallel.

Of greater concern is the 'handful of trips' comment. This is no small thing on a centrifugal compressor.

My main suggestion would be to concentrate on the cause of the trips.
Another would be to look at the way your seal gas is injected into the seal; if it is on diff pressure control, swapping to a flow control may help it ride out any pressure swings.

 

[swern]

Let me add a few more pieces of the puzzle. Management has decried that the recycle valve will go 100% open on a trip. Based on the discharge volume being significantly larger than the suction volume, as well as, the valve throughput, the suction pressure climbs ~ 200 psi and then decays.

The seals are balanced mechanical contact seals without buffer gas. The process gas which leaks past the upstream laby seal exits via the compressor end plate. The seal oil is then maintained at 50 psi above the "leaked" process gas pressure (i.e. cavity between mechanical seal and laby seal). The PCV would be mounted on the piping connected to the compressor end plate.

Mucour - our thought is as you mentioned...."dirty" gas is flowing across the seal face. However, we do not have evidence of a surge. It is believed to be caused by the pressure spike which occurs more rapidly than the PDCV can accomodate.

The number of trips is definitely a concern which we are addressing. One of the compressors is driven by a steam turbine. The boiler which drives the turbine is currently reliant on purchase power. The purchase power is causing the problems. This is in the process of being addressed.

My current plan is to try the PCV (quickest/cheapest modification)until I can get management to buy into a recycle piping modification.

Thank you for the comments!

 

[bvsrtt]

Hi Swern,

The problen is really interesting ! which is the compressor service make and model of machine, is it motor dtriven ?

The phenomenon explained by you is a normal feature in high pressure centrifugal compressors. When Antisurge valve opens, the down stream of ASV or in other words suction pressure rises. This pressure rise could be quite normal and not noticeable in low pressure compressors and very high in compressors having higher presure ratio.

AA) when we have compressors in a train with 2 , 3 or 4 casing configurations with a high pressure rise across suction and discharge such as 165 bar, or 100 bar, it is a practice to provide a vent valve in discharge ( upstream of ASV)which will open alongwith ASV at the same instant of trip and remains open for 3 to 10 sec and closes. This will remove the large pressure pulse in discharge line. In one of our Syn gas compresors with 3 casing design with discharge pressure of 220 bar and suction pressure of 25 bar, we could resolve by suitably controlling the opening of vent valve timing ( time it remains open after trip )increase from 3 sec to a higher valve.

BB)you can examine if the reference gas pressure tapping from the cavity you mentioned is sluggish and blocked . unable to communicate the quick changes in suction pressure to oil inlet(i.e. cavity between mechanical seal and laby seal).

I do not know in your present case what are the pressure levels and total pressure levels and service application like FCC, gas or H2 etc. It will be appropriate to know the compressor parameters to appreciate what is happening.
Pl examine the aspects suggested by me.

"bvsrtt" has worked in BHEL, India as a compressor designer for centrifugal compressors dept at Hyderabad, India . If you like u nay contact