PSV for deaerator blocked discharge: set point? discharging to a vessel drained at grade?

[EFTCAT]

Dear All,
Thank you in advance if you can guide me for a PSV (set at 300 kPag) sized for blocked discharge protecting a BFW deaerator designed at 250 kPag. The suction nozzle of PSV is at 5000 mm below the suction PSV lateral nozzle at the drum...Client complained that this PSV discharges at grade where during winter hot water gets frozen and the splashing creates a hazard, so client wants PSV to discharge in a vessel and then drained to a trench...What would be the basis for sizing this vessel and the drain line to trench by gravity flow?
Thank you in advance.

 

[Latexman]

I can’t picture what you describe, so a dimensioned sketch or drawing would help.

PSV (set at 300 kPag) sized for blocked discharge protecting a BFW deaerator designed at 250 kPag.

Can you explain how this complies with Code? I don’t think it does.

Good Luck,
Latexman

 

[EFTCAT]

Hi Latexman: thanks for your reply.
I just attached the PID for a better visual of the system. Client wants to discharge not directly at grade but rather first in a tank, then in a drain trench...
There you can see the same info I specified above about the set pressure and elevation of connection nozzles...
So any idea if it makes sense to add a discharge tank and how to specify it? The PSV datasheet shows a required rate of 616 m3/hr water liquid and a rated flow of 866.7 m3/hr for SG 0.951 @ 110 C relief
Thanks

 

[LittleInch]

it took me a while to find your valves - a circle round then next time might be useful...

Well you vessel / tank etc just needs to be big enough to create a head that allows the drain out volume equal to the drain in volume.

Purpose of this tank is not clear to say the least, though water at 110C emitting to atmosphere will not be water , but some unholy two phase mixture of steam and water so maybe the client thinks this needs to be calmed down, let the steam escape and the water to now flow gently, so in essence a degassing drum,

So now you need to make sure that the vent pipe to atmosphere is big enough to handle the flash steam without creating any real pressure in the drum / tank.

Latexman, I assume they are allowing for the 5m drop from the vessel to the relief valve mounted 5m below the vessel being full of water? But I agree - If it isn't full of water and NLL is below the entry then it is not allowed. All very odd. EFTCAT - Can you explain more??

And I've never heard inlet flange into a PSV called a suction nozzle before???

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[EFTCAT]

Hi LitteInch, you got me here....suction flange, not nozzle for the PSV. I was referring to nozzle at the BFW tank rather. But at 3am in the morning, you can excuse me for above. Anyway, the isometrics attached also explains what I meant by 5 m drop from the vessel, and also the PSV data sheet for the liquid water, because at relief pressure and 110 C, the water is liquid not steam...The verbal scope of work from client was that operation people had problems with freezing water splashing at grade discharged from this PSV and getting frozen during winter...So client wanted to add a vertical (?) vessel to capture this PSV discharge, then drain it to a trench cooler at lower speed...It is not clear to me the basis to size this vessel: is it a 10 min relief which would determine capacity of vessel? what would be the discharge line sizing criteria: a max backpressure determined for bellows type and max 0.7M at rated flow in the lateral discharge? The BFW storage vessel to which PSV is attached is at 14.7m AG...The deaerator overflow vessel disch line is 12" and same size and pipe spec as the current PSV discharge, so I would rather tie-in the disch PSV line into the overflow vessel drain line rather than adding a new vessel, isn't it? The drain line shall be gravity flow. Any idea from your side? Appreciated, before I get into too much time spent on flarenet or Hysys hydraulics...Thanks again.

 

[Latexman]

Can 521-C-12"-N4A-4318 just be redirected into the trench? No tank.

If not, the basis would be an open flash tank to atmosphere; a steam/water separator. The basis for drain line sizing would be for flashing condensate (2-phase) flow. As LI said the liquid head in this new tank is the driving force to get the inflow out.

Good Luck,
Latexman

 

[LittleInch]

It's not easy to get an idea of what is going on here, but the key points seem to be:

The set pressure of the PRVS seems strange, being above the design pressure of the vessels, but maybe their MAWP is higher?

Anyway, what is coming out of the PSVs is a mixture of steam and water so whoever decided to dump this "at grade" needs their head seeing to.

The other very worrying thing is that this PSV seems to be going off. PSVs should NEVER go off if the process is being controlled properly. Sounds' like someone is normalizing failure here.

but yes, you've caught the issue - how long a period are you going to get more flow in versus less flow out, where does the steam go and why can't you ust tie it into the drain line from the overflow vessel.

BTW what are all those "vents" on the DA vessel set at?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[EFTCAT]

Hi LittleInch: unfortunately, I do not have any more data than what is shown on P&ID. Plus I do not have datasheets of vessels. The main client concern is the splashing of a big volume of hot water liquid on the floor outdoor creating a potential hazard. So what you would propose as technical solution? And I do not understand why the client prefers a vertical surge vessel instead of a horizontal vessel since most of the release at blocked discharge is water liquid...

 

[Latexman]

I wonder if a tee separator would satisfy the client? A tee separator is not highly efficient. It allows most of the vapor to flow up, and the liquid to drain down. With flashing flow, some vapor will exit out the drain. But if the drain is down in the trench and the vapors are water vapor (i.e. not flammable or toxic), maybe that is good enough. It'll definately direct the liquid to the trench, and it's cheap!

Good Luck,
Latexman

 

[LittleInch]

The client should be more concerned with why the PSVs are going off and dumping this water. IMHO.

Technical solution is pipe it somewhere else, but the problem there is will you ever know if the valves are lifting or not? At least this way you'll notice it, but it seems a very strange way to deal with very hot two phase liquid to squirt it onto the floor in large quantities where people are.

we can't do this sort of thing remotely as we can't see what you can see in front of your eyes. Why can't this pipe just be extended to wherever the "safe location" / trench the client wants you to us is located?

I think the client is just thinking of space and practical issues such as how to support it. I'm not a mind reader - why don't you ask him or her?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[EFTCAT]

Yeah, but the trench is just 700 mm below grade as I understood. I do not have direct contact with the client, management has, so they take as a law whatever client pays for...I am more concerned about safety issues and how to design it. Yes, initial thought was to include a vertical section of 30in vertical pipe as a surge volume for liquid to reduce relief velocity, then discharging to grade/trench via a 22in section...Velocity is an issue for erosion aspect. Finally client decided that a vertical vessel will help as surge volume to calm down the relief velocity, cool down a bit and discharge it safely to trench. The distance from trench is about 30 m horizontally. Now design wise, how much would be the residence time? 10 min...15 min? even so, the volume is huge...As for velocity, this is very high in a 12" disch pipe (existing), so I can understand why is desired to slow down this big volume of hot water...maybe 10% vapor and 90% liquid water. Corrosion/toxicity is not an issue here since it is basically BFW...Any ideas? Thanks.

 

[Latexman]

LI, I agree 100%. It's a funky set up with varying liquid levels and DA pressures complicating the situation and making analysis difficult or impossible. A pilot operated PSV (Set @ 250 kPag) with remote sensor into the top vapor space of the DA would eliminate those potential interactions.

The effluent mess/situation will still need to be addressed, even if the frequent PSV opening is solved. Right now it doesn't sound like it is piped to a "safe location".

Good Luck,
Latexman

 

[georgeverghese]

No details on normal pressure control at this de aerator.

Looks like these PSVs' in question are configured purely for liquid relief, which is why the inlet nozzle at 2973mm above vessel bottom is well below LAHH at the BFW storage tank ( PSV set at 300kpag implies the PSV is some 2m below the bottom of the storage tank assumes the inlet line is liquid full). Why these PSVs' would be lifting frequently is probably something to do with some instability with normal pressure and level control. The PSV datasheet says normal op pressure is 40kpag, while max op pressure is 13% less than 250kpag. May help if this root problem is addressed first before addressing the consequences of the poor pressure control. And no details provided on vapor pressure relief. There may be instabilities with feed flow to the deaerator also.

Suspect the plan is to comply with environmental discharge guidelines on upper limit on temperature for liquid emissions to public drainage, hence the plan for a relief stream cooler. And hence they need to smoothen out the flow to this cooler with a surge vessel upstream to receive these intermittent relief slugs from the PSV and degas this 2phase stream from the PSV. It is anybody's guess what the surge volume for this intermediate degassing - surge drum should be. Some feedback from Operations on level / pressure trends in these vessels should be the basis for deriving the size of this drum.

 

[LittleInch]

I am more concerned about safety issues and how to design it.

OK, so you're a professional engineer and need to do things properly.

So I would write a report noting down everything you don't know, what the safety issues are ( PSVS going off should never be regarded as "normal" operations) and possible ways to fix it so that the operation is as safe as possible.

From the minimal information we have it seems to me that the plant as designed and working is not safe as

a) the drain of scalding hot water and steam which explosively erupts from the end of the PSV line is not routed to a "SAFE LOCATION"
b) The PSVs are intended as a last line of defence safety relief and any event which causes them to go off needs to be investigated, studied and recommendations made / changes made to the operating system to prevent it happening again as far as possible.
c) the information supplied and written on the documents is inadequate for anyone to work out what is supposed to be happening or indeed monitor what is actually occurring. A HAZOP review would seem to be in order to me.

"I do not have any more data than what is shown on P&ID. Plus I do not have datasheets of vessels." Then you can't do your work to the standard required of a Professional Engineer.

Make sure you keep a copy of everything secure somewhere so that when OSHA or whatever safety man turns up after some gets seriously injured, you have your defence.

All IMHO.

LI

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

@LI/ Latexman, Suspect this steam condensate arriving at this de aerator is partly foul contaminated recycled process condensate - it may be foaming in this deaerator, which would result in poor level control, poor pressure control with 2 phase feed to the PCV.
Plan for a cooler may not be a good idea - what if it overcooled and blocked the tubes in the cooler? Alternatives may be to install a safety overflow line at the degasser or use a passive form of cooling.
This foul steam condensate may contain large amounts of high boiling hydrocarbons ?? If so, disposal at grade at new location may need to account for safe disposal of residual flammables, volatile components, and maybe even H2S after degassing.