9 replies to this topic

[Propacket]

I am facing a difficulty in simulation of a gas treatment plant comprising of followings:
1-Three Phase Separator
2-Dew Point Control Unit
3-Condensate Stabilization unit
20 MMSCFD gas is injected into three phase separator, then through dew point control unit consisting of gas /gas exchanger, propane chiller and cold separator. We have two types of condensate products:
-From three phase separator (RVP 18 psia)
- From cold separator (RVP 220 psia)
Condensates from three phase separator and cold separator are mixed and then routed to a stabilizer column (without condenser) where it is heated to achieve 10 psia RVP. 5 MMSCFD overhead gas is generated which is compressed to the pressure of inlet gas and then mixed with inlet gas prior to entering the three phase separator.
Problem is that when overhead gas is mixed with inlet gas, hysys shows a sharp increase in flow rates of all streams, for example, Overhead gas from stabilizer is 250 MMSCFD which is not possible by 20 MMSCFD inlet gas.
Can anybody guide me? I am unable to attach the hysys PFD due to company restrictions.

[Zauberberg]

Maybe you can't attach the Hysys file but definitely you could upload a PFD/sketch of your system, including the main parameters so that we can comment on it.

[Srikrishna Chaitanya]

P.Engr,


For me it seems like a recycle / flow spec problem which is causing instability in column ,From my wild guess I feel that you are giving column liquid outlet as flow spec ( In hysys we give specifications to make column degrees of freedom as zero , once degrees of freedom becomes zero we can run column and get the results . These specs can be flow , recycle ratios , heat duty etc.), In case if you specify column bottom flow then simulation can adjust only the overhead flow to converge with your specs . pl check this and if this is the case then vary heat duties of the column provided you will not deviate your column temp profile and see whether you get some improvement . In other words pl don’t give flow specs especially column bottom flow ,You can give heat duty as spec .In case if this doesn’t help pl let us know more details to analyse it better as said by Zauberberg.

Regards
Satish

[Propacket]

Maybe you can't attach the Hysys file but definitely you could upload a PFD/sketch of your system, including the main parameters so that we can comment on it.

Zauberberg
I have attached the PFD and material streams.Please see if you can help me. Feel free to ask me any question.

Attached Files

•  PFD and Material Streams.pdf   61.02KB   295 downloads

Edited by P.Engr, 05 April 2010 - 02:30 AM.

[Propacket]

P.Engr,


For me it seems like a recycle / flow spec problem which is causing instability in column ,From my wild guess I feel that you are giving column liquid outlet as flow spec ( In hysys we give specifications to make column degrees of freedom as zero , once degrees of freedom becomes zero we can run column and get the results . These specs can be flow , recycle ratios , heat duty etc.), In case if you specify column bottom flow then simulation can adjust only the overhead flow to converge with your specs . pl check this and if this is the case then vary heat duties of the column provided you will not deviate your column temp profile and see whether you get some improvement . In other words pl don’t give flow specs especially column bottom flow ,You can give heat duty as spec .In case if this doesn’t help pl let us know more details to analyse it better as said by Zauberberg.

Regards
Satish

Satish,
I have given the column only reboiler temperature spec.

[Zauberberg]

Problem is that when overhead gas is mixed with inlet gas, hysys shows a sharp increase in flow rates of all streams, for example, Overhead gas from stabilizer is 250 MMSCFD which is not possible by 20 MMSCFD inlet gas.

Hello P.Engr,

The problem is probably due to the fact that you have generated large recycle stream from the stabilizer overhead, that spins on and on between the inlet reception facilities and the tower. I don't know the compositions here but very likely you are condensing a good portion of compressor discharge and re-introducing it again as column feed. That doesn't sound as a good option.

You should look at your gas specs (i.e. C5+ content) and if significant, you can reconfigure the tower to give you better separation and less condensate losses with the overhead gas. That could be done by changing the number of stages, operating pressure, or changing the tower configuration (try refluxed tower instead of conventional stripper). Recycle stream shouldn't be that much big.

Good luck,

[Zauberberg]

I forgot - and when you do that, simply export the discharge of compressor. Do not mix it with the plant feed again.

[Propacket]

Problem is that when overhead gas is mixed with inlet gas, hysys shows a sharp increase in flow rates of all streams, for example, Overhead gas from stabilizer is 250 MMSCFD which is not possible by 20 MMSCFD inlet gas.

Hello P.Engr,

The problem is probably due to the fact that you have generated large recycle stream from the stabilizer overhead, that spins on and on between the inlet reception facilities and the tower. I don't know the compositions here but very likely you are condensing a good portion of compressor discharge and re-introducing it again as column feed. That doesn't sound as a good option.

You should look at your gas specs (i.e. C5+ content) and if significant, you can reconfigure the tower to give you better separation and less condensate losses with the overhead gas. That could be done by changing the number of stages, operating pressure, or changing the tower configuration (try refluxed tower instead of conventional stripper). Recycle stream shouldn't be that much big.

Good luck,

Hi Zauberberg,

Zauberberg,
I have reworked on the simulation. I have used refluxed tower, increased number of trays, reduced condenser operating pressure and result is C5+ free Overhead gas. At compressor discharge, there is no condensate with overhead gas. If I do not mix the overhead gas with plant feed and simply export it, overhead gas flow is only 3.5 MMSCFD. No problem is encountered in exporting the overhead gas. But even if this gas (which is free from C5+ and there is no condensate with the gas) is mixed with plant feed, hysys generates huge flow rate of overhead gas.
There is a process design constraint in this simulation that multiple gas or liquid streams cannot be exported from the plant. I have to generate a single gas stream with the required cricondentherm and a single condensate stream with the required RVP.
I have simulated condensate stabilization several times and this is a common practice that overhead gas (free of C5+) is mixed with plant feed. I have never experienced any problem like in this simulation.
Please advise any alternative process scheme that will not pose this problem.
I have attached the composition of feed gas and overhead gas.
Thanks

Attached Files

•  Composition.pdf   33.57KB   132 downloads

[Zauberberg]

Speaking without going into too many details - there are two export streams you have to make:

1. Gas (Cricondentherm specification)
2. Condensate (RVP specification)

Looking at individual components, all of them will be distributed between the two product streams. If, for example, there is an excess of C4's that cannot go with the Condensate (RVP limits) and they also cannot go with the Sales gas (Cricondentherm limit), obviously there should be another export stream (e.g. LPG) where these components will be accumulated. I think it's a fairly simple material balance issue.

If you have to process the overhead gas further in order to make it suitable for Export, I would suggest you to do it separately from the feed gas - this way you are simply loading the unit with no reason. But again, if there is an excess of any component that cannot be redistributed between the two product streams while still their specs are OK, you have to make additional product stream from the plant - it's as simple as that.

Imagine distilling crude oil and having only Naphtha, Diesel, and Residue as allowable products. Obviously, you would contaminate your Naphtha with C4- components if there is no processing option for removing them from Naphtha stream.

[Propacket]

Speaking without going into too many details - there are two export streams you have to make:

1. Gas (Cricondentherm specification)
2. Condensate (RVP specification)

Looking at individual components, all of them will be distributed between the two product streams. If, for example, there is an excess of C4's that cannot go with the Condensate (RVP limits) and they also cannot go with the Sales gas (Cricondentherm limit), obviously there should be another export stream (e.g. LPG) where these components will be accumulated. I think it's a fairly simple material balance issue.

If you have to process the overhead gas further in order to make it suitable for Export, I would suggest you to do it separately from the feed gas - this way you are simply loading the unit with no reason. But again, if there is an excess of any component that cannot be redistributed between the two product streams while still their specs are OK, you have to make additional product stream from the plant - it's as simple as that.

Imagine distilling crude oil and having only Naphtha, Diesel, and Residue as allowable products. Obviously, you would contaminate your Naphtha with C4- components if there is no processing option for removing them from Naphtha stream.

Zauberberg,
I got your point. The gas is rich in C3 and C4. In order to achieve the required cricondentherm, some C3 and C4 must be removed from the gas along with higher ends. Therefore, at operating temperature of LTS, lots of C3 and C4 are condensed. I made the mistake that I mixed all of the stripped C3 and C4 from stabilizer again with the feed gas. It was my misconception that C3 and C4 usually do not cause high dew points and are carried with the treated gas even after dew point control.
Instead of condensate stabilization, I used DEC2 and DEC4, recovered the C3 and C4 as LPG. Lighter ends (C1 and C2) recovered as DEC2 overhead gas which i mixed with treated gas because it has no effect on the cricondentherm of treated gas. C5+ recovered as stabilized condensate.
All the specifications were complied with the recovery of an additional product (LPG).
Thanks a lot for your time and valuable advices.