6 replies to this topic

[chem101]

Hi every One,
I was involved in one of the separation of gas, water and condensate project. I went through the separator package and found that the all separator looks alike. Please see attached P & ID.
I have founded following error but no one agree with me.
(1)Mist Eliminator should be installed on other chamber (bucket side) so all gas can pass through the mist eliminator.
(2) We don’t need bucket if we are providing Boot OR don’t need boot if providing bucket.
(3) LT-904 signal going to PLC-001 is not a correct symbol. It needs to be triangle shape like of LT-903.
(4) Don’t require 2 x LT on BOOT side. One LT should be enough to turn on or off the pump; also alarm can be set on the same LT.
(5) Some of the gas will bypass mist eliminator. As liquid level can only be maintained in boot section.
Please let me know if I am mistaking or is there any impact on process. As my PM do not want make any changes.
Thank you in advance

Attached Files

•  Three Phase Separator.pdf   316.94KB   304 downloads

Edited by chem101, 27 January 2012 - 12:37 PM.

[fallah]

(1)Mist Eliminator should be installed on other chamber (bucket side) so all gas can pass through the mist eliminator. (2) We don’t need bucket if we are providing Boot OR don’t need boot if providing bucket. (3) LT-904 signal going to PLC-001 is not a correct symbol. It needs to be triangle shape like of LT-903. (4) Don’t require 2 x LT on BOOT side. One LT should be enough to turn on or off the pump; also alarm can be set on the same LT. (5) Some of the gas will bypass mist eliminator. As liquid level can only be maintained in boot section.

chem101,

My comments on your found errors with lack of adequate information:
1) As far as i know the mist eliminator should be as close as possible to gas outlet. Hence, seems is located in right place (with correction as per item 5).
2) Reason of simultaneous using of bucket and boot is probably due to lesser water flowrate respect to condensate flowrate (it is understood from LT's configuration along the boot) that in steady state condition it is needed to have adequate water storage for P-909 start-up otherwise the lenght of the vessel after weir should be increased.
3) LT-904 is considered for P-909 start/stop and LT-903 for pump shut-down. Hence, there is no reason these different LT's have similar symbols for their relevant controllers.
4) Seems alarme for H and L liquid levels is already considered on LT-904 and LT-903 is a dedicated level transmitter for pump shut-down on LL liquid level in order to have a separate layer of protection.
5) Seems you are right and mist eliminator should be located below gas outlet nozzle. Of course, PID is schematic drawing and for precise checking of this matter you should refer to vessel engineering drawing.

Fallah

Edited by fallah, 28 January 2012 - 04:40 AM.

[Art Montemayor]

I think I know what you call a "boot".

I have never heard the term "bucket" used in conjunction with a 3-phase separator. Can you please tell me what you mean by "bucket"? Is it the liquid chamber upstream of the internal weir?

[chem101]

Art,

The Term Bucket and weir is been used in artical “Successfully Specify Three Phase Separator" by Wayne D. Monnery and William Y. Svrcek.

Bucket is mainly use to collect hydrocarbon, where weir is use to main heavy liquid level in upstream of the internal weir. This helps hydrocarbon to spill into bucket.

I hope i clarify the term Bucket.

Edited by chem101, 31 January 2012 - 11:33 AM.

[Dacs]

To answer your question,

1. If I were the one designing the vessel, I'll place the demister pad adjacent (and covering) the vapor outlet nozzle. From how it's designed as shown, I agree on the location (close to the outlet as possible)
2. Expected water flowrate is small enough that a boot is needed to have a workable level on such expected volume. Alternatively, we can adjust the baffle location to give a comparable holding volume as the boot and remove the boot altogether.
3 and 4. PLC-001 with LT-003 is intended with safety function in mind (to stop the pump at LLLL). The diamond symbol is normally associated with Emergency Shutdown Device (ESD). The other PLC takes care of the normal discharge of condensed liquid at the boot, which is not considered for safety purposes. Such redundancy is usually seen in plants (independent instrument/control loop handles operation and shutdown of the plant).

For instance, if we were to install one LT for both functions, what will happen to the plant if the LT fails? It will take down both operability and safety of the plant in question.
5. While the gas may bypass below the mist eliminator, larger liquid droplets won't go "up" with the gas (it will just fall down to the liquid inventory). I'm starting to think that the mist eliminator is placed in that location to handle the smaller liquid droplets that may entrain otherwise with the gas.

Edited by Dacs, 31 January 2012 - 12:41 AM.

[chem101]

To answer your question,

• If I were the one designing the vessel, I'll place the demister pad adjacent (and covering) the vapor outlet nozzle. From how it's designed as shown, I agree on the location (close to the outlet as possible)

I agree with you that Mist Eliminator need to be placed close to the gas outlet. Only, if the Eliminator is installed horizontally. However, if Mist eliminator is installed vertically. In this case preferred location is close to Bucket in 3-phase separator. The reason is that “by maintain liquid level in this chamber will provide a good gas seal. By installing close to out let, gas will be by passed the mist eliminator since there is no liquid level maintained.

• Expected water flow rate is small enough that a boot is needed to have a workable level on such expected volume. Alternatively, we can adjust the baffle location to give a comparable holding volume as the boot and remove the boot altogether.

Expected water flow is higher in this case that’s why we have bucket for hydrocarbon.
3 and 4. PLC-001 with LT-003 is intended with safety function in mind (to stop the pump at LLLL). The diamond symbol is normally associated with Emergency Shutdown Device (ESD). The other PLC takes care of the normal discharge of condensed liquid at the boot, which is not considered for safety purposes. Such redundancy is usually seen in plants (independent instrument/control loop handles operation and shutdown of the plant).
In my opinion One LT should be enough. The PLC-001 has high and low alarm it can also should down the system when liquid level is low low.

For instance, if we were to install one LT for both functions, what will happen to the plant if the LT fails? It will take down both operability and safety of the plant in question.
You mean redundant LT
5. While the gas may bypass below the mist eliminator, larger liquid droplets won't go "up" with the gas (it will just fall down to the liquid inventory). I'm starting to think that the mist eliminator is placed in that location to handle the smaller liquid droplets that may entrain otherwise with the gas.
If gas by pass below the mist eliminator. Then it will defeat the purpose of mist eliminator.

[Dacs]

I agree with you that Mist Eliminator need to be placed close to the gas outlet. Only, if the Eliminator is installed horizontally. However, if Mist eliminator is installed vertically. In this case preferred location is close to Bucket in 3-phase separator. The reason is that “by maintain liquid level in this chamber will provide a good gas seal. By installing close to out let, gas will be by passed the mist eliminator since there is no liquid level maintained.

My thinking is when the gas goes below the demister, the entrained liquid droplets will tend to go straight to the demister because their movement won't be as flexible as the gas itself (due to momentum). Droplets big enough will fall down. Smaller droplets will tend to cover a straighter path (going to the pad). While I say that the current arrangement shown in P&ID is not optimal as a horizontal pad covering the outlet nozzle, I would somehow get how this arrangement might work.

Expected water flow is higher in this case that’s why we have bucket for hydrocarbon.

Drum size for three phase separators is determined by vapor-liquid disengagement as well. (Without supplement information at hand) Maybe this drives up the vessel diameter such a way that the expected liquid flows (especially the heavy liquid which is water) are small enough that changes in liquid levels would result in (designed) holdup time is too small for a reliable level readout. Hence the need for a boot.

In my opinion One LT should be enough. The PLC-001 has high and low alarm it can also should down the system when liquid level is low low.

This will work if we can guarantee that the instrument will never fail on the duration of its life.

Thing is, if this fails during operation such that it doesn't give the proper level (say stuck in NLL). Without the other LT, how'd you shutdown the plant if both relies on a single LT?

This is usually being discussed during HAZOP and most probably the requirement of separate LT's came from the HAZOP analysis.

Edited by Dacs, 31 January 2012 - 07:43 PM.