11 replies to this topic

[Anup87]

Dear all,

I am working on Debutanizer column safety analysis. In this column system previously they have provided two PSVs. One on column and other on accumulator (or reflux drum). Column PSV is relieving to atmosphere while accumulator PSV is relieving to flare. And I am not getting any clue, why they have two different relief locations for same system. Please help me to understand this system, before moving into real analysis.

Note that accumulator and column are in open communication and it is expected that both the PSVs will pop in all pressure events.

Thanks,
Anup.

Edited by Anup87, 14 September 2012 - 05:14 AM.

[flarenuf]

Anup,

Are the PSV's set at the same set pressure ?
Do you know the relative sizes of each ?

flare e nuf

[Anup87]

Flare,

Column PSV is set at 200 psig (lowest MAWP of the system), while accumulator PSV is set at 195 psig.

Column PSV is of 4" X 6" (Orifice letter "N") while accumulator PSV size is 3" X 4" (orifice letter "K").

Anup.

[Bobby Strain]

Don't get sidetracked. Keep working on your relief analysis.

Bobby

[gegio1960]

Just for brainstorming....
Maybe there is a limits in the flare loads.
The psv on the accumulator should open first (set at 195 psig), relieving to flare, at a lower rate (it's the smaller size psv) but enough to cover most part of the relieving cases.
This protection could not be sufficient for the major relief scenarios.
In these cases, the pressure will increase up to the opening of the 2nd psv that will relief to atmosphere (at safe location!) saving both the protected equipment and the flare operation, that won't be overloaded.

[fallah]

Dear all,

I am working on Debutanizer column safety analysis. In this column system previously they have provided two PSVs. One on column and other on accumulator (or reflux drum). Column PSV is relieving to atmosphere while accumulator PSV is relieving to flare. And I am not getting any clue, why they have two different relief locations for same system. Please help me to understand this system, before moving into real analysis.

Note that accumulator and column are in open communication and it is expected that both the PSVs will pop in all pressure events.

Thanks,
Anup.

Anup87,

What do you mean by "open communication"? As far as i know there are some PV and isolation valves between debutaniser and reflux drum forced you to consider one dedicated PSV for each of them in order to code compliance. Obviously, PSV sizes and set pressures are another stories and related to relief loads of relevant governing scenarios and design pressure of the column/vessel, respectively. Indeed, in no case of safety standpoints you are allowed to release the outlet of the PSV on the debutaniser toward atmosphere while having flare network in vicinity.

Fallah

[kkala]

Interpretation by gegio1960 is judged reasonable to me; besides debutanizer PSV can be directed to flare header when its size gets high enough. Clarification is needed on the point raised by fallah in the preceding post, but let us presently assume no isolation between debutaniser and reflux drum (being "in open communication") during the scenaria involving the pair of PSVs.
Under these conditions, distinct functions of the PSVs are (α) discharge only from reflux drum PSV, (β) discharge from both PSVs simultaneously. Drum's PSV is set at 195 psig, debutaniser's at 200 psig. Can a set pressure difference of only 5 psi (2.5% of gauge set pressure) be enough for events (α) and (β) to be timely dinstict, or only (β) will occur in practice? Advice on the point is welcomed. Difference between set pressures may need increase, depending on the answer.
An example of similar case concerns two PSVs on a (ASME) boiler steam drum. First PSV is set at drum gauge design pressure, second at 3% higher (rounded to nearest 5 psi), according to instructions found (~1985). In an actual boiler (ordered in 2000) set pressures are 54.00 and 55.00 kgf/cm2 g (difference=1.8%), while allowable overpressure is only 3% for both PSVs. This may indicate that set pressure difference of 2.5% can be adequate, even though overpressure is not well understood (full relief of PSV1 at 54*1.03=55.62 kgf/cm2 g, while PSV2 starts opening at 55.00 kgf/cm2 g). But it concerns ASME boilers, not process equipment.
At any case clarifying allowable overpressure by Anup87, from the data sheets of two PSVs (set at 195 and 200 psig), would be useful.

Edited by kkala, 16 September 2012 - 01:49 PM.

[Anup87]

Thanks All.

Fallah you are 100 % correct. There is one control valve on pressure equalisation line. Also between condenser and reflux drum there are isolation valves. For you guys reference I have attached hand drawn sketch (hope it will help you). And normally these isolation valves will be open to carry column operation, that why we consider them in open communication.

KKala, API-520 says take 16 % of overpressure over lowest MAWP for multiple PSVs. So here relief pressure will be 232 psig.

Regards,
Anup

Edited by Anup87, 17 September 2012 - 07:12 AM.

[kkala]

The set pressures of two PSVs have a difference of only 5 psi, it is useful to know at what pressure each of them gets fully open (relieving pressure).
Overpressure for this condition (accumulation) has been requested from the data sheet of each PSV, http://process-eng.blogspot.gr/2012/03/accumulation-and-overpressure.html, trying to distinguish overlapping between (α) and (β) per post No 7. Do data sheets indicate relieving pressure of 232 psig for these two PSVs? Data sheets often show relieving pressure as % pressure increase over (gauge) set pressure.
Besides hand drawn sketch seems to be missing, please attach.

Edited by kkala, 17 September 2012 - 04:40 PM.

[fallah]

Kkala,

As OP mentioned, each of two vessels (column and drum) with some valves in between will be protected by its own PSV ; means there is the possibility of mentioned valves closing inadvertently or due to failure. Hence, the rules for "multiple valves installation" in API 520 wouldn't be applicable in this case.

Fallah

Edited by fallah, 18 September 2012 - 05:59 AM.

[kkala]

Indeed, Fallah, Annup87 (OP) confirms isolation valves between debutaniser and reflux drum; yet these are considered open in OP's safety analysis (post No 8, locked opened?) and "both the PSVs will pop in all pressure events" (post No 1). This indicates no practical isolation between column and drum during all considered scenaria of discharging from either one or both PSVs. Posts No 7 and 9 are based on this ''assumption''; if not valid, explanation to post No 1 might be (temporary?) inadequacy of existing flare header to transfer max discharge from debutanizer PSV.
Nevertheless OP can clarify mentioned "assumption". Process data sheets of two PSVs and sketch (posts No 8, 9) could be helpful too.
Note: Although usually considered isolated, two vessels can be "in open communication" during fire, http://www.cheresour...y-a-common-psv/ '> http://www.cheresour...y-a-common-psv/ .

[Anup87]

Thank you all.