15 replies to this topic

[ogpprocessing]

Dear CHE Experts,

As you all know low poionts shall be avoided at flare lines. I want to know if due to plant layout limitation low point at flare lines can not be avoided and flare network simulation shows that condensation will not occure at any case within flare network, then is it allowed to have low point?

Edited by ogpprocessing, 08 November 2010 - 05:41 AM.

[engg]

just a thought.... would it be feasible to put a low points drain on the line and connect it to maybe a condenstae tank or closed drain header? the problem i see is the flow of flare gas when the drain will be open...is the flare continuous?

Edited by engg, 08 November 2010 - 06:32 AM.

[ogpprocessing]

just a thought.... would it be feasible to put a low points drain on the line and connect it to maybe a condenstae tank or closed drain header? the problem i see is the flow of flare gas when the drain will be open...is the flare continuous?

The flare is not continous. This option proposed by you is not feasible due to lots of concerns on having low point at flare line even with drain.

[ankur2061]

ogpprocessing,

Some company standards permit provision of small local drain pots or drip legs in the low point of the flare piping if it is evaluated during the design stage of the flare system that the quantity of liquid that can be condensed or formed is very small. However, an operating procedure needs to be developed for regular inspection and periodic emptying of these local drain pots in order to prevent liquid accumulation in the flare piping.

Regards,
Ankur.

[ogpprocessing]

ogpprocessing,

Some company standards permit provision of small local drain pots or drip legs in the low point of the flare piping if it is evaluated during the design stage of the flare system that the quantity of liquid that can be condensed or formed is very small. However, an operating procedure needs to be developed for regular inspection and periodic emptying of these local drain pots in order to prevent liquid accumulation in the flare piping.

Regards,
Ankur.

Dear Ankur,
In the specific project that I am involved at most of the governing cases for safety valves is liquid blocked outlet. So is it still allowed to have low point with such a drain pot or drip leg? Is this drain pot or drip leg evacuated manually or with pump?

[ankur2061]

ogpprocessing,

For liquid relief service the system is different from the vapor relief system.

The ones I am familiar with are all the liquid relief valves hooked up to a gravity drain header running to a liquid blow down vessel whose liquid relief inlet nozzle is at an elevation lower than the lowest liquid relief valve in the plant/unit. Additionally a separate vapor relief valve is provided on the liquid blowdown vessel which hooks up to the flare or vent. The sizing philosophy of the vapor relief valve on the blowdown vessel is based on any one of the liquid liquid relief valves connected to the blowdown vessel which provides the largest liquid relief capacity. This is based on the engineering logic that simultaneous liquid relief from more than one relief valve is not credible. However, This is again on a case-to-case basis and you need to arrive at your own credible scenario commensurate with your company's relief system guidelines.

If there are other vapor relief valves in the plant / unit they are hooked up to the flare or vent header along with the vapor relief valve from the blowdown vessel. This system is a tried and tested one in some of the plants and units that I have been associated with.

Regards,
Ankur.

[fallah]

The sizing philosophy of the vapor relief valve on the blowdown vessel is based on any one of the liquid liquid relief valves connected to the blowdown vessel which provides the largest liquid relief capacity. This is based on the engineering logic that simultaneous liquid relief from more than one relief valve is not credible.
This criteria relates to sizing the vessel not PSV.
If there are other vapor relief valves in the plant / unit they are hooked up to the flare or vent header along with the vapor relief valve from the blowdown vessel.
If the vessel would be connected to flare system by a line including a CSO valve no need to pressure protection by a PSV.

Edited by fallah, 11 November 2010 - 12:38 AM.

[ankur2061]

The sizing philosophy of the vapor relief valve on the blowdown vessel is based on any one of the liquid liquid relief valves connected to the blowdown vessel which provides the largest liquid relief capacity. This is based on the engineering logic that simultaneous liquid relief from more than one relief valve is not credible.
This criteria relates to sizing the vessel not PSV.
If there are other vapor relief valves in the plant / unit they are hooked up to the flare or vent header along with the vapor relief valve from the blowdown vessel.
If the vessel would be connected to flare system by a line including a CSO valve no need to pressure protection by a PSV.

Fallah,

For the first point I do not agree that it is just for sizing of the vessel, because liquid accumulation in the vessel needs to be considered for sizing the vessel as well as the liquid pump-out capacity when deciding the vessel capacity. The case for sizing the PSV is considered based on the vapor displaced (or blanketing inert gas if vessel is blanketed) by the liquid flowing into the vessel from the largest liquid relief scenario.

The second point is acceptable of hooking up the vapor line from the blowdown vessel to the flare or vent by a CSO valve. However, the blowdown vessel would still require an additional overpressure protection and maybe a different case such as fire case needs to be considered for the relief valve sizing.

Regards,
Ankur.

[fallah]

Fallah,

For the first point I do not agree that it is just for sizing of the vessel, because liquid accumulation in the vessel needs to be considered for sizing the vessel as well as the liquid pump-out capacity when deciding the vessel capacity. The case for sizing the PSV is considered based on the vapor displaced (or blanketing inert gas if vessel is blanketed) by the liquid flowing into the vessel from the largest liquid relief scenario.

The second point is acceptable of hooking up the vapor line from the blowdown vessel to the flare or vent by a CSO valve. However, the blowdown vessel would still require an additional overpressure protection and maybe a different case such as fire case needs to be considered for the relief valve sizing.

Regards,
Ankur.

Dear Ankur,

About the first point due to all time connection between vessel and flare system,vapor displacement would be handled by that connection line which to be sized such that be able to handle largest diplacement without pressurizing the vessel above its design pressure.

About the second point,because mentioned vessels are usually buried they don't need to PSV,but if it is not the case and fire case could be applicable (vessel not buried) and also vent line couldn't be handled fire case relif load,you are right and the vessel need to be protected by a PSV.

Best Regards,
Naser Fallah

[Qalander (Chem)]

Fallah,

For the first point I do not agree that it is just for sizing of the vessel, because liquid accumulation in the vessel needs to be considered for sizing the vessel as well as the liquid pump-out capacity when deciding the vessel capacity. The case for sizing the PSV is considered based on the vapor displaced (or blanketing inert gas if vessel is blanketed) by the liquid flowing into the vessel from the largest liquid relief scenario.

The second point is acceptable of hooking up the vapor line from the blowdown vessel to the flare or vent by a CSO valve. However, the blowdown vessel would still require an additional overpressure protection and maybe a different case such as fire case needs to be considered for the relief valve sizing.

Regards,
Ankur.

Dear Ankur,

About the first point due to all time connection between vessel and flare system,vapor displacement would be handled by that connection line which to be sized such that be able to handle largest diplacement without pressurizing the vessel above its design pressure.

About the second point,because mentioned vessels are usually buried they don't need to PSV,but if it is not the case and fire case could be applicable (vessel not buried) and also vent line couldn't be handled fire case relif load,you are right and the vessel need to be protected by a PSV.

Best Regards,
Naser Fallah

Dears,It is indeed a very interesting and informatory discussion and feel sorry to miss this early-on.

My concept is more of what ankur has said in above last post and

• one thing which has somehow not surfaced during discussions is the 'prevalent' or 'dictating'
• ambient conditions consideration for liquid volumes generation(s); if any. since this will
• heavily impact the pump-out capacity requirements, Flare Knock Out drum's capacity and if necessitated
• liquid handling additional systems& vessel if any
• with the protection discussed by our friends "ankur"&"fallah" in preceding post above.

Hopefully this helps envisaging another important scenario needs to have due safeguards in-place.

Welcome if needed further discussion of this issue etc.

[ogpprocessing]

ogpprocessing,

For liquid relief service the system is different from the vapor relief system.

The ones I am familiar with are all the liquid relief valves hooked up to a gravity drain header running to a liquid blow down vessel whose liquid relief inlet nozzle is at an elevation lower than the lowest liquid relief valve in the plant/unit. Additionally a separate vapor relief valve is provided on the liquid blowdown vessel which hooks up to the flare or vent. The sizing philosophy of the vapor relief valve on the blowdown vessel is based on any one of the liquid liquid relief valves connected to the blowdown vessel which provides the largest liquid relief capacity. This is based on the engineering logic that simultaneous liquid relief from more than one relief valve is not credible. However, This is again on a case-to-case basis and you need to arrive at your own credible scenario commensurate with your company's relief system guidelines.

If there are other vapor relief valves in the plant / unit they are hooked up to the flare or vent header along with the vapor relief valve from the blowdown vessel. This system is a tried and tested one in some of the plants and units that I have been associated with.

Regards,
Ankur.

Dear Ankur,

By liquid relief I did mean the liquid blocked outlet is the governing case of the PSV and not the sole credible scenario and still other cases of vapor relief such as fire is credible. So your preposed configuration to segregate liquid relief streams from vapor relief streams is not applicable for my plant. Now for my case keeping in mind that liquid bloacked outlet is the governing case for most of the safety valves, is it allowed based on good engineering practice (as you said) to have low point drains or drip legs at flare lines?

Let me know your idea.

[ankur2061]

ogpprocessing,

Some company guidelines explicitly mention the desirability of segregating the liquid and vapor relief system. I would also recommend that. I would provide a small blow down vessel for collecting the liquid relief and connect the vent of this vessel to the vapor relief header.

Practically also I haven't seen a common relief header for liquid and vapor relief. The low point liquid collection is for vapor condensation in long vapor flare headers and is basically meant for handling very small quantities of liquid and not for handling a relief valve discharging liquid into the relief header.

I am not sure the configuration you are mentioning is practical & I won't be able to give any advice about such a common liquid / vapor flare relief header.

Regards,
Ankur.

[daryon]

Dear CHE Experts,

As you all know low poionts shall be avoided at flare lines. I want to know if due to plant layout limitation low point at flare lines can not be avoided and flare network simulation shows that condensation will not occure at any case within flare network, then is it allowed to have low point?

Are you talking about a single PSV discharge line prior to entering a main flare header or sub-header? or are you talking about a main header or sub-header?

In my opinion having a low point in a main flare header or sub-header is very poor engineering practice for the following reasons:
1. Liquid accumulation in low point or pocket may result in high velocity liquid slugs entering the FKOD when a vapour release is observed. The FKOD drum may not be sufficiently sized to knock out the liquid droplets and carryover to flare may result in burning droplets failing near or on the facility.
2. Liquid accumulation may result in accelerated corrosion.
3. If water is collecting at the low point you have a prime area where hydrates may agglomerate, build up and potentially block the pipe if cold vapour releases are observed.

Most (if not all) companies will not allow a low point in main or sub-headers, the flare lines should slope without pockets all the way to the FKOD.

If you are talking about having a low point in a single PSV discharge line, then this may well be acceptable. Sloping down from the PSV discharge to a low point which should be fitted with a standpipe and equipped with a level gauge / level switch and drain is o.k.From the the low point the pipe work should then rise without pockets and enter the a flare header or sub-header at the top of the pipe. I don't see it as a problem if there is a liquid relief case as well as other scenarios (e.g fire case) provided that the built up back pressure (including static head) does not exceed allowable back pressure for the selected PSV. If the stand pipe at the low point fills up during a liquid relief operators can drain it.

Edited by daryon, 25 November 2010 - 03:22 AM.

[ogpprocessing]

Dear CHE Experts,

As you all know low poionts shall be avoided at flare lines. I want to know if due to plant layout limitation low point at flare lines can not be avoided and flare network simulation shows that condensation will not occure at any case within flare network, then is it allowed to have low point?

Are you talking about a single PSV discharge line prior to entering a main flare header or sub-header? or are you talking about a main header or sub-header?

In my opinion having a low point in a main flare header or sub-header is very poor engineering practice for the following reasons:
1. Liquid accumulation in low point or pocket may result in high velocity liquid slugs entering the FKOD when a vapour release is observed. The FKOD drum may not be sufficiently sized to knock out the liquid droplets and carryover to flare may result in burning droplets failing near or on the facility.
2. Liquid accumulation may result in accelerated corrosion.
3. If water is collecting at the low point you have a prime area where hydrates may agglomerate, build up and potentially block the pipe if cold vapour releases are observed.

Most (if not all) companies will not allow a low point in main or sub-headers, the flare lines should slope without pockets all the way to the FKOD.

If you are talking about having a low point in a single PSV discharge line, then this may well be acceptable. Sloping down from the PSV discharge to a low point which should be fitted with a standpipe and equipped with a level gauge / level switch and drain is o.k.From the the low point the pipe work should then rise without pockets and enter the a flare header or sub-header at the top of the pipe. I don't see it as a problem if there is a liquid relief case as well as other scenarios (e.g fire case) provided that the built up back pressure (including static head) does not exceed allowable back pressure for the selected PSV. If the stand pipe at the low point fills up during a liquid relief operators can drain it.

In my case the low point concern refers to flare main header. I fully agree with your explanation but here the low point may not be avoided due to layout limitation because the flare network pipes up to flare KO drum are located at one side of the road and the flare stack is located at the other side of the road and the flare main header outgoing from flare KO drum should pass the road from below (underground) and then joints to flare stack.

[daryon]

This design should be approached with much caution. Is it not possible that liquid will condense in this section of pipe downtream of the FKOD and upstream of the flare stack? What about purge gas condensing? Or if a PSV is leaking small flowrates of 'heavier' gas from somewhere in the process. You need to be sure you will not get liquid accumulation in this section of pipe, if you do, you risk liquid slugs being pushed up the stack. How will you know if liquids are accumulating? Regualar checking and putting procedures in place? What if someone forgets?

Much better if this line slopes back to FKOD from the stack riser with no pocket. Can you not run the pipe over the road?

[Qalander (Chem)]

Dear CHE Experts,

As you all know low poionts shall be avoided at flare lines. I want to know if due to plant layout limitation low point at flare lines can not be avoided and flare network simulation shows that condensation will not occure at any case within flare network, then is it allowed to have low point?

Are you talking about a single PSV discharge line prior to entering a main flare header or sub-header? or are you talking about a main header or sub-header?

In my opinion having a low point in a main flare header or sub-header is very poor engineering practice for the following reasons:
1. Liquid accumulation in low point or pocket may result in high velocity liquid slugs entering the FKOD when a vapour release is observed. The FKOD drum may not be sufficiently sized to knock out the liquid droplets and carryover to flare may result in burning droplets failing near or on the facility.
2. Liquid accumulation may result in accelerated corrosion.
3. If water is collecting at the low point you have a prime area where hydrates may agglomerate, build up and potentially block the pipe if cold vapour releases are observed.

Most (if not all) companies will not allow a low point in main or sub-headers, the flare lines should slope without pockets all the way to the FKOD.

If you are talking about having a low point in a single PSV discharge line, then this may well be acceptable. Sloping down from the PSV discharge to a low point which should be fitted with a standpipe and equipped with a level gauge / level switch and drain is o.k.From the the low point the pipe work should then rise without pockets and enter the a flare header or sub-header at the top of the pipe. I don't see it as a problem if there is a liquid relief case as well as other scenarios (e.g fire case) provided that the built up back pressure (including static head) does not exceed allowable back pressure for the selected PSV. If the stand pipe at the low point fills up during a liquid relief operators can drain it.

In my case the low point concern refers to flare main header. I fully agree with your explanation but here the low point may not be avoided due to layout limitation because the flare network pipes up to flare KO drum are located at one side of the road and the flare stack is located at the other side of the road and the flare main header outgoing from flare KO drum should pass the road from below (underground) and then joints to flare stack.

Dear Hello/Good Night,
In Case of the limitations mentioned by you

• I can envisage one thing to be supportive
• as a possible remedy is having the subject underground segment double steam traced and thoroughly insulated;
• however this may not all-seasons/all times 100% fool proof solution
• but definitively helping minimizing K.O.D. downstream liquid slug possibilities
• thereby avoiding any unsafe conditions upstream side including the K.O.D. in question.

Hope this proves helping.