7 replies to this topic

[daryon]

Hi All,

I'm looking for some advice/opinions on relief rate calculation for a full flow PSV on a production header. The production header recieves well fluids from multiple subsea flowlines that varry in length from 2km to 13 km. Flow assurance performed on the flowlines predicts large uncontrolled liquid slugs (upto 50m³)arriving at the production facilties fairly frequently (every 2hrs) from each of the flowlines.

We have full flow relief valves on the production header which are installed to protect the piping and downstream equipment (HTEX & Separator)in the event it is shut-in and overpressured by the well streams. The flowlines, risers and turret piping is all 600# and the spec break to 150# is at the turret/topside interface. The liquid design capacity for the process is 45,000 blpd, however when a liquid slug is being recieved into the process the liquid flowrate can spike severely (well over 100,000 blpd).

My question is; when considering the full flow rate that the production header PSVs must relief would you take into account the peak transient liquid flowrate that can occur when a slug is being recieved? This is basically saying that a liquid slug arrives simultaneously with a accidental shut-in of the front of the process.

Would really appreciate your advice and thoughts on this.

Thanks

[shan]

PSV’s are designed for releasing vapor to flare not for liquid. Even if you have large enough PSV’s to relieve the liquid, do you have large enough vessel/container volume to hold the liquid? The only way to deal high pressure from liquid slugs is to shut your ESV boarding valves.

[daryon]

Hi Shan,

The FKOD is sized for 15 minutes of full flow from one flowline or 90 seconds of flow from all flowlines whichever is greater. The sizing basis is from NORSOK. The FKOD will fill up pretty quickly.... probably under a minute to maximum allowable liquid level if we consider the liquid relief at the peak transient flowrates. But your saying so would not consider this??

There are high pressure trips that close the riser ESDVs on high production header pressure, but this is only a single level of protection. If you are designing to API RP 14C then you need a full flow PSV also. ( 2 levels of protection)

[ashetty]

Hello Daryon,
What about offshore platforms which export gas to onshore processing terminals? These offshore platforms do not have a flare or flare KODs. In my opinion if the flowline and production header are designed for shut-in pressure of the well & you have PSHL`s which close the riser ESDVs you might not require the PSVs, but you need to follow the API or relevant standards/codes...can anyone give inputs??
Best Regards.

[ashetty]

You might consider posting this in the relief device forums to get more response...

[shan]

Hi Shan,

The FKOD is sized for 15 minutes of full flow from one flowline or 90 seconds of flow from all flowlines whichever is greater. The sizing basis is from NORSOK. The FKOD will fill up pretty quickly.... probably under a minute to maximum allowable liquid level if we consider the liquid relief at the peak transient flowrates. But your saying so would not consider this??

There are high pressure trips that close the riser ESDVs on high production header pressure, but this is only a single level of protection. If you are designing to API RP 14C then you need a full flow PSV also. ( 2 levels of protection)

Hi Daryon,

I always size the flare knock-out drums based maximum relief gas flow not even single time for liquid retention time, because I never know how much liquid carried by the released gas. When the well fluid arrive the platform, they are two phase flow in most cases. If the free gas is released from PSV, the flowline overpressure will be relieved.

It is the same concept as relieving beer bottle pressure by releasing gas through top cap. You do not have to empty all the liquid (beer) to relive pressure. Otherwise, nobody has taste of beer. Your flare system is not designed for relieving full liquid flow. If you introduce that much liquid into the flare header, the other PSVs will be unable to open at all because the backpressure from the liquid.

Overall, PSV’s on the flowlines are the secondary protection for overpressure, but they are not designed for relieve the liquid slug volume.

[daryon]

Hi ashetty

We didn't want to design our production header and front end equipment (HTEXs and separator) with a pressure rating suitable to cope with the shut-in pressure of the wells. So we have a spec break and overpressure protection (pressure trips and PSV) for equipment and piping with the lower pressure rating. Sometimes I agree it is more economical and simplifies design if the production header is designed for well shut-in pressure, but it depends on the process configuration and pressures involved.

thanks i'll post this in the relief forum too.



Hi again Shan,

Thanks for the response i especially like the beer example its something i can relate too. I actually disagree a little bit with what you are saying; my experience allthough fairly limited is to always consider liquid relief when sizing the FKOD. When sizing a drum you have to ensure that there is sufficient gas space above a maximum allowable liquid level to ensure you can slow the gas down and allow disengagment all the liquid droplets of a certain size (normally 400- 450 micron) and above. This should stop liquid fire raining down on all below!

My intepretion of both API RP 521 and NORSOK Std. P-100 is that you have to design the FKOD liquid capacity to handle full flow from one (or more) flowlines for a certain period of time. Full flow is the liquid and gas. If you block the production header and the pressure relief valve lifts you will have 2-phase flow that was going to your production separator going to flare system for a period of time. Why would the liquid not flow? I like the beer example but its not quite the same deal when you're considering piping systems with high pressure upstream and low pressure downstream.

I have designed flare systems in the in the past using FLARENET and one the first objectives is to sit and decicde which relief scenarios can occur co-incidently. A full flow relief from the production headers is very unlikely (if even possible) to occur simulatenously with any other reliefs and therefore the builtup back pressure on other PSVs is not of concern. The full flow PSV discharge piping is sutiably sized to ensure that a full flow relief does not build up backpressure on the valve that exceeds 40% of the set presssure (balanced bellow valve) or 10% for a conventional valves.

So my question is when deciding what the FKOD liquid capacity must be and when considering the PSV sizing should I worry about the peak transient flowrate that occurs when a liquid slug arrives in the production header? The more I think about it the more I think this is unnecessary and over conservative, for the full flow PSV on the production header the sizing should be based on the plant process capacity for the full flow relef scenario not peak transient flowrates caused by liquid slugging.

Thanks Shan

Edited by daryon, 02 September 2009 - 07:59 PM.

[shan]

Hi ashetty

We didn't want to design our production header and front end equipment (HTEXs and separator) with a pressure rating suitable to cope with the shut-in pressure of the wells. So we have a spec break and overpressure protection (pressure trips and PSV) for equipment and piping with the lower pressure rating. Sometimes I agree it is more economical and simplifies design if the production header is designed for well shut-in pressure, but it depends on the process configuration and pressures involved.

thanks i'll post this in the relief forum too.



Hi again Shan,

Thanks for the response i especially like the beer example its something i can relate too. I actually disagree a little bit with what you are saying; my experience allthough fairly limited is to always consider liquid relief when sizing the FKOD. When sizing a drum you have to ensure that there is sufficient gas space above a maximum allowable liquid level to ensure you can slow the gas down and allow disengagment all the liquid droplets of a certain size (normally 400- 450 micron) and above. This should stop liquid fire raining down on all below!

My intepretion of both API RP 521 and NORSOK Std. P-100 is that you have to design the FKOD liquid capacity to handle full flow from one (or more) flowlines for a certain period of time. Full flow is the liquid and gas. If you block the production header and the pressure relief valve lifts you will have 2-phase flow that was going to your production separator going to flare system for a period of time. Why would the liquid not flow? I like the beer example but its not quite the same deal when you're considering piping systems with high pressure upstream and low pressure downstream.

I have designed flare systems in the in the past using FLARENET and one the first objectives is to sit and decicde which relief scenarios can occur co-incidently. A full flow relief from the production headers is very unlikely (if even possible) to occur simulatenously with any other reliefs and therefore the builtup back pressure on other PSVs is not of concern. The full flow PSV discharge piping is sutiably sized to ensure that a full flow relief does not build up backpressure on the valve that exceeds 40% of the set presssure (balanced bellow valve) or 10% for a conventional valves.

So my question is when deciding what the FKOD liquid capacity must be and when considering the PSV sizing should I worry about the peak transient flowrate that occurs when a liquid slug arrives in the production header? The more I think about it the more I think this is unnecessary and over conservative, for the full flow PSV on the production header the sizing should be based on the plant process capacity for the full flow relef scenario not peak transient flowrates caused by liquid slugging.

Thanks Shan

When you size a separation vessel (separator, scrubber or flare knock-out drum), you have to consider vapor capacity and liquid capacity. Usually, vapor capacity and liquid capacity result different vessel diameter and length. In most cases, liquid capacity is the governing factor for the oil separator and vapor capacity for the scrubbers and the flare knock-out drums.

A slug catcher is supposed to handle liquid slugs. However, usually there is no slug catcher on the offshore platform because of limitation of space.

I consider the scenario of slug arriving and outlet block simultaneously as double jeopardy.