11 replies to this topic

[bpc]

I have one query on the Centrifugal Compressor blocked outlet scenario.

 

Please refer attached sketch.

 

This is regarding Centrifugal compressor operating on the onshore oil and gas separation facility (in design/Feed stage).

 

As per blocked outlet scenario, the blocked outlet flow rate for the PSV-101 is 12.5 MMSCMD (set at 4200 kpa). However, this new compressor will be connected to existing flare system which has maximum capacity of 8 mmscmd.

 

Therefore, in the feed document, one additional PSV-102 (“Recycle PSV”) is provided downstream of air cooler, which is set at slightly lower set pressure (3990 kpa) . But this PSV discharge back to inlet of the suction scrubber. This PSV (PSV-102) sized for 4.5 mmscmd and PSV-101 is for 8 mmscmd. This way feed contractor has justified that maximum 8 mmscmd flow will go to the flare.

 

I am not able to understand how this arrangement will help to provide enough protection in the compressor blocked outlet scenario.

 

I'd be glad if any one share his opinion/experience.

Attached Files

•  compressor.pdf   198.82KB   248 downloads

[fallah]

bpc,

In fact, the PSV-101 and PSV-102 are to be set almost in staggered configuration (set pressure of PSV-101 is almost 1.05 times that of PSV-102) but with different destination in back pressure standpoint. Then, the back pressure of flare network and suction scrubber should be such that wouldn't affect the capacity of these PSV's at relevant set points, otherwise the type of mentioned PSV's should be selected such that not being affected by back pressure in capacity standpoint.

[Sathya R]

Review the relief capacity inrelation to the performance map of the compressor.

 

Recycling a part of the gas to the suction will result in increase of the compressor suction pressure resulting in increase in the compressor capacity which has to relieved to flare anyway. 

[MYI]

Dear BPC,

Good Morning,

 

In my opinion, PSVs are not the good solution for the over pressure protection system even in case of block outlet; please review following design strategy & could be optimum for your compression system.

 

1- Installed the Bleeder Valve (PCV) on the down stream of the air cooler & set the pressure of pressure controller; in case of pressure buildup in the system, this bleeder valve will open through the actuator & continuous release the mass flow of gas until it will reach again on the operating pressure condition.

2- Consider the HIPP (High integrity Pressure Protection), three pressure transmitter, in case of pressure build up (Due to any reason) but more than the bleeder valve (PCV) set point; this HIPP loop will activate & send the signal to the shutdown panel (Pressurized Shutdown), to trip the compressor. 

3- Consider the higher design pressure of compressor body & down stream piping & equipment even more than the compressor shutoff pressure; so, if compressor will reach upto the shutoff pressure than there will be no effect on the equipment & piping, & system will be inherently safe.

 

Thanks for understanding.

 

MYI

[Dacs]

My understanding is that when you have a blocked-in scenario, you have to avoid compressor surge.

 

Normally you'd have anti-surge lines from the discharge back to suction and you'd operate your compressor just above the compressor surge point.

 

In this case however, it seems that your system can't afford to totally block in the flow (which I presume might upset the processes upstream of compressor) so the designer opted to send a certain amount of flow to the flare header (which is already fixed in capacity).

 

My impression is that the compressor is capable to pump out13 MMSCMD of fluid at the expected relieving conditions (ie compressor discharge pressure) and the designer placed this scheme to accommodate 8.5 MMSCMD of fluid going to flare and the rest (4.5 MMSCMD) as a recycle to avoid compressor surge.

 

Granted that I got everything correctly, I'm wondering how the compressor can afford to send higher flowrate despite of the increase in differential pressure (DP) (from 3400 kPa /9 MMSCMD to 4200 kPa + accumulation/13 MMSCMD) because I'm expecting the other way around (higher DP -> lower flowrate), unless the designer has considered the suction pressure at maximum (2000 kPa vs 800 kPa min), which may have resulted in an effective lower DP for the compressor.

 

My 2 cents.

Edited by Dacs, 02 April 2013 - 12:48 AM.

[bpc]

Dear Fallah,

 

Thanks for your reply on post. 

When reycle PSV will open,  it will see the backpressure of scrubber which is ultimately suction pressure of compressor.

This is going to release 4.5 mmscmd to the suction that will increase the suction pressure as discharge is blocked.

But when the PSV to the flare will open, the suction pressure will also stabilize to certain value and hence backpressure is expected to be constant. But the concern is since the compressor is required to relieve 12.5 mmscmd and it is only relieving 8 mmscmd, ultimately the recycle psv does not serve purpose. That's what I feel.

 

 

 

 

bpc,

In fact, the PSV-101 and PSV-102 are to be set almost in staggered configuration (set pressure of PSV-101 is almost 1.05 times that of PSV-102) but with different destination in back pressure standpoint. Then, the back pressure of flare network and suction scrubber should be such that wouldn't affect the capacity of these PSV's at relevant set points, otherwise the type of mentioned PSV's should be selected such that not being affected by back pressure in capacity standpoint.

[bpc]

Dear MYI,

 

Thanks for your reply.

I am not sure about 1st point, I think even bleeder valve will relieve the maximum estimated relieving rate which is 12.5 mmscmd for which flare is not sized?

HIPs system is not considered.  Agreed that can be the Solution here instead of recycle PSV.

Designing at higher design pressure is not feasible here as this train will be connected to existing system and discharge design pressure is based on the existing system design pressure.

 

 

 

 

Dear BPC,

Good Morning,

 

 

 

In my opinion, PSVs are not the good solution for the over pressure protection system even in case of block outlet; please review following design strategy & could be optimum for your compression system.

 

 

 

1- Installed the Bleeder Valve (PCV) on the down stream of the air cooler & set the pressure of pressure controller; in case of pressure buildup in the system, this bleeder valve will open through the actuator & continuous release the mass flow of gas until it will reach again on the operating pressure condition.

 

2- Consider the HIPP (High integrity Pressure Protection), three pressure transmitter, in case of pressure build up (Due to any reason) but more than the bleeder valve (PCV) set point; this HIPP loop will activate & send the signal to the shutdown panel (Pressurized Shutdown), to trip the compressor. 

 

3- Consider the higher design pressure of compressor body & down stream piping & equipment even more than the compressor shutoff pressure; so, if compressor will reach upto the shutoff pressure than there will be no effect on the equipment & piping, & system will be inherently safe.

 

 

 

Thanks for understanding.

 

 

 

MYI

[fallah]

Dear Fallah,

 

Thanks for your reply on post. 

When reycle PSV will open,  it will see the backpressure of scrubber which is ultimately suction pressure of compressor.

This is going to release 4.5 mmscmd to the suction that will increase the suction pressure as discharge is blocked.

But when the PSV to the flare will open, the suction pressure will also stabilize to certain value and hence backpressure is expected to be constant. But the concern is since the compressor is required to relieve 12.5 mmscmd and it is only relieving 8 mmscmd, ultimately the recycle psv does not serve purpose. That's what I feel.

 

bpc,

 

Appears the compressor will adjust itself to admit just 8 mmscfd at incoming line to suction drum before anti surge return connection in order to balance the input/output flows. Now it should be investigated what would be happened in upstream with this reduction in flow and if it is possible for upstream facilities?

[bpc]

Fallah,

 

Upstream system is flowline manifold preceded by flowlines from 15-20 wells. Maximum total flow from all flowlines is 18 mmscmd.

Based on this flow 3x 33% compressor trains are provided to compress the gas from 8 bar to 20 bar. But it will vary between 20 bar to 4 bar as resorvoir pressure deplete. This 12.5 mmscmd flow for relief flow is estimated based on 21 bar maximum suction pressure, discharge pressure based on PSV set pressure and rated power of the compressor which is 12MW. Flowline maximum pressure can go upto 35 bar. So I guess upstream facility is capable of sending flow. That's what you are asking?

 

 

 

 

 

Dear Fallah,

 

Thanks for your reply on post. 

When reycle PSV will open,  it will see the backpressure of scrubber which is ultimately suction pressure of compressor.

This is going to release 4.5 mmscmd to the suction that will increase the suction pressure as discharge is blocked.

But when the PSV to the flare will open, the suction pressure will also stabilize to certain value and hence backpressure is expected to be constant. But the concern is since the compressor is required to relieve 12.5 mmscmd and it is only relieving 8 mmscmd, ultimately the recycle psv does not serve purpose. That's what I feel.

 

bpc,

 

Appears the compressor will adjust itself to admit just 8 mmscfd at incoming line to suction drum before anti surge return connection in order to balance the input/output flows. Now it should be investigated what would be happened in upstream with this reduction in flow and if it is possible for upstream facilities?

[fallah]

bpc,

 

Suction pressure could be 21 barg and you want to compress the gas to 20 barg! Anyway, 12.5 mmscmd is estimated based on this condition and should be clarified firstly.

Edited by fallah, 03 April 2013 - 04:32 AM.

[bpc]

Fallah,

 

Please note that normal design case for the compressor is

Ps = 8 bara

Pd = 20 bara 

Flow = 8 mmscmd

 

However at higher suction pressure which can go upto 21 bara, discharge pressure and flow will be higher.

 

For PSV relief calculation, suction max pressure considered as 21 bara. 

 

As stated earlier, 12.5 mmscmd is based on compressor curve with 21 bara suction pressure and max power of 12 MW. 

 

 

Edited by bpc, 03 April 2013 - 06:36 AM.

[fallah]

bpc,

 

If the flow from wells toward the compressor station could be limited to 8 mmscmd during blocked outlet, then the flow passing through the compressor itself would be 12.5 mmscmd and the PSVs system don't allow the compressor discharge to be overpressured. Otherwise, the flow through the compressor would be higher and existing PSVs cannot handle excessive flow which is higher than 12.5 mmscmd....