Flare Modelling.

[AB6005]

Guys,

I just started to check the adequacy of the existing flare system which was built way back in 1950's.

Initially i started doing the adequacy check for the existing PSV's so that i can make sure that i dont have any in adequate PSV's or inlet pressure drop offering valves.

but my doubt is as i dont have any values of back pressure right now should i need to check for outlet pressure drop by considering zero for the back pressure there by i can validate the PSV's though i have outlet pressure drop ?????

or

please help me with a way to proceed with him.


i am making use of flarenet for modelling and simulation.
i have already freezed the controlling scenarios.

 

[MikeClay]

You're in for a LOT of fun. My guess is that at best all that you have for the relief system, assuming a pipe flare, are the PSV datasheets and maybe the flare design capactiy.

This information should give you an idea of the expected backpressure and selected relieving case. As long as the expected backpressures don't effect valve capacity, I'd feel comfortable sizing the relief valves with no backpressure. Backpressures still have to be confirmed later.

Once you have the expected flows, the relief header can be modeled. Lay out system. Hopefully you have access to a hydraulic package such as Flarenet or Visual Flow. Spreadsheeting these sysetms can be a nightmare. Having isometric drawings would be nice, but they are not normally available for a 50 year old plant. You may be doing a lot of measuring and fitting counts, even if only to veryfy the iso's.

One the system is modeled, check the backpressures. Normally the PSV tail piece is modeled at full relieving rate while the headers use the required flow. Make sure to consider credible simultaneous reliefs (e.g. Fire effecting common equipment; Feed failure overpressuring multiple vessels...).

Then compare these backpressures to the PSV calculations to see if the backpressures effect the PSV's.

I know this is only a brief overview, but depending on the size of the flare system, you will be having a lot of fun.

--Mike--

 

[MortenA]

Start by making a special PFD (if this dosnt already exists) mapping the entire flare system and all sources.

For each source mark the design case and group them for simultaneous relief (PSV, BDVs etc).

Then for each relief/BD scenario set up a flow model and calculate back pressure at each source.

Then double check that your PSV's, BDVs etc meets the required capacities

Thats more or less the task.

Best regards

Morten

 

[JoeWong88]

Huh...great fun for you...
There are many thing to be checked other than backpressure.

Minimum back pressure to PSV
In regards to your query on backpressure, you may consider ATM for conservativeness (from Mach number aspect) and save time. If your results shown failure to meet criteria, then start to consider built-up backpressure of individual PSV relief by modeling using Flarenet.

If you have conventional PSV, you have to consider rated flow for tail pipe calc. If you have modulated type PSV, API allow relieving flow for tail pipe calc. For inlet section, you have to use rated flow.

Maximum back pressure to PSV
You have to check for multiple PSVs relief scenario i.e. total power failure, fire, Instrument air failure, etc. Accumulated flow will give higher back pressure to PSV. Make sure it meet the 10% requirement for conventional type.
.
.
.

Don't miss the PSV potentially relieve boiling liquid, 2 phase flow and subcooled liquid but flash across PSV. These kind of PSV may be undersized.

The flare header (some part) may have been designed for MACH 1. Vibration and noise check may required.
.
.
.


JoeWong
Chemical & Process Technology

 

[AB6005]

@ all thanks for your replies.

@ Joe wong...

joe is there any limits for Mach number in flare header...i have heard that there are some different mach number levels for main header,sub-header and psv tail pipe..... is it so ?????

then how about noise levels ??? and vibration criteria ???

joe whether i need to validate my PSV's and then go for flare header sizing....plz help me out.......

 

[JoeWong88]

There is no fix rule on limit of Mach no. Based on my experiences, there are many practices. Typically, PSV tail limit to 0.7 (some goes up to 0.8). For sub-header and main headers, typically 0.5.

Noise level for continuous service is about 80-85 dBA. Generally limit to 115 dBA.

Vibration is more complicated. There are high frequency vibration, low frequency vibration, etc. Tail pipe, sub-header and header integrity shall not susceptible to vibration and it is subject to wall thickness, location of support, type of support, small bore piping, etc.

I suggest you tackle the Mach no and backpressure to PSV, follow by noise level. Then tackle the vibration issue with your piping specialist or external consultant.

Hope these information helps.

JoeWong
Chemical & Process Technology

 

[797989]

Thanks joe for the valuable tips.

do you have any idea on zone classification of PSV's ???

how do they do it ????

 

[AB6005]

Thanks joe for the valuable tips.

do you have any idea on zone classification of PSV's ???

how do they do it ????

Is that LOPA is any way related with zone classification.

 

[JoeWong88]

Zone classification is subject to release point, gas characteristic, etc. Don't think it link to valve itself.

JoeWong
Chemical & Process Technology

 

[786392]

NFPA 30 and NFPA 497 should be helpful on this issue.

Best Regards
Qalander(Chem)