dry gas seal contamination 2

[lahcen]

We have a compressor that compresses an associated gas (gas separated from oil). But although all the filtration we always have contamination of the dry gas seals by oil.

What are the possible causes of this contamination knowing that the filtration is by the standards?

Why always the failure happens on the gas seal of the output side?

thanks in advance for your help.

 

[LittleInch]

My guess, because you've given us very little information, is that your rich gas is being either cooled or depressurised and some of the liquid is coming out of the gas which your filters are not picking up.

What pressure and temperature is the seal gas to start with and when it enters the seals?

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[rotw]

The OP question is not clear and poor information is given.
What type of compressor are you talking about ?
Suppose you refer to a centrifugal.

- The seal gas is contaminated by oil, which oil ? lube oil? Filtration will remove solid particle and liquids (water, gas heavy HC condensate, etc) but is not supposed to remove lube oil, which is also not supposed to leak over there.

or do you mean ''real oil'' carried over with the process gas. in this case, if you have an autobuffer system you get that oil back to the DGS. Having oil carried over with the process gas is also odd.

- What type of separation seal do you have, labyrinth or carbon ring ? I suppose you refer to a tandem DGS ?not clear
- Have you checked the Lube oil header pressure and the separation seal pressure, is it all OKay ?
- What do you mean by compressor output side (is this the discharge side of the compressor) ?
- Is that a drive through compressor or end drive ?
Driven/Non driven side may have an effect due to the rotating shaft/coupling that possibly effect the pressure differential on the bearing and therefore the oil.

 

[lahcen]

It is a centrifugal compressor. Its suction pressure is 55 bars ( T=56°C) and the discharge pressure is 118 bars (T= 120°C).
The DGS is tandem with no intermediary labyrinth.
The DGS is contaminated with the oil carried over with the process gas; this gas is taken from the discharge of the first stage to be used as the sealing gas in the DGS
For a long period of time we have not opened the purging valves scaling of the filters. Could that lead to this contamination? If so, why does the contamination happens always in the drive through side?

 

[rotw]

Seems also drive through terminology is here misused. Drive through means if you have two or more body in tandem arrangement, one casing or more will have to be drive through .
If you have only one casing than you could indicate simply if that happen on the drive end or non drive end.
The aero path is standard/common for inline (supposedly inline) but can be reversed so you still have to indicate if the concerned side is suction or discharge.

The pressure on the primary seal should be equal for both side due to the balance line. The main difference between discharge and suction is the temperature of the casing during operation. The primary seal gas is feeding normally both sides of the casing at same pressure and temperature (more or less). On the discharge (hot) side, the final temperature of the seal is somewhere between the primary seal feed temperature and the wall temperature. Hot temperature at discharge side may prevent the gas and nasty components it contains from condensing. So my guess is that your drive through side is suction side more prone to have condensation.

During stand still however, if you wait a while, both sides of the casing are at same temperature (ambient) and both seals will be in principle at same conditions. So my guess is that during stand still you always depressurize the machine.

If the assumption above makes some sense, than you may try to put a heater to superheat the primary seal gas. This will not solve the fundamental problem that oil should not be carried over with process gas. If this was by design or not is not clear, but it is definitely unusual but for me the seal is just a warning that something is going wrong at higher level. I am not aware that coalescing filters are designed for catching oil droplets but I might not know enough.

Finally all the above is speculative. Obvious. The amount of info is very limited so all what you get here is assumptions that's it.

 

[rotw]

I would add this different possibility : contamination happening because the leakage is not going from the seal to process but the other way around, due to some sealing pressure problem.

 

[lahcen]

thank you very much for your help