Flash Vessel Sizing : Max Flash Steam Velocity Criteria- GPSA Engineering Data Book vs EPRI CS 2251 1

[Vicker85]

Hello Experts,

I from a mechanical background but seeking some advice on basic Sizing of a Vertical Flash Tank/Vessel WITHOUT Mesh Pad. The flow of incoming condensate is just 3000 lbs/hr (so low!). The vent and drain both go to a Surface Condenser of a power plant.

I searched on this forum and found out that to size the ID of the vessel, we need to have the max permissible velocity of flashed steam. And this as per GPSA Engineering Data Book is :
Vmax = (K) √[ (dL - dV) / dV ]

where:
Vmax = maximum vapor velocity, ft/sec
dL = liquid density, lb/ft3
dV = vapor density, lb/ft3
K = 0.1 (??), ft/s

BUT as per EPRI CS 2251 (Recommended Guidelines for the Admission of High Energy Fluids to Steam Surface Condensers), the max permissible velocity of flashed steam is :

Vmax = 5 x √Vs
where:
Vs = Specific Volume of Flashed Steam, ft3/lb

For the data that I have, I am getting significantly different results!!

Could someone guide on what would be the more suitable criteria for max velocity as this dictates my tank ID for the given flow.

 

[katmar]

Your second reference is for sizing of piping and has nothing to do with sizing of a flash vessel. It will give a much higher velocity than would be suitable for a flash vessel.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[Vicker85]

@ katmar:

Thank you for your valuable insights. You are correct about the higher max velocity as recommended by EPRI.

However, per your comment about the max velocity being for sizing pipes - could you please refer to the snap I have attached here and look for the text I have highlighted in red.

What I infer from the V[sub]s[/sub] is that it’s the velocity of the flashed steam in the vessel and not in the pipe. In fact, the piping max velocity is also marked as V[sub]1[/sub].

Could you provide your comments on the same?
Also, if we were to size using GPSA guidelines, what would be the preferred value of K for this application (with NO mesh pad) ?

 

[georgeverghese]

So, the GPSA says K = 0.09 - 0.18fps for vertical KOD, while the Shell K value for the same application is 0.2fps.
The EPRI diagram shows a special compact cyclonic type KOD, since you've got a 45deg feed ( near tangential) here - typically, you can use much higher V for these since there are centrifugal forces acting on the liquid droplets. The GPSA KOD is for a conventional type KOD with no internals and conventional feed.

 

[Vicker85]

@ georgeverghese:

Thank you for your valuable reply!

Since for my present application, the total flow rate is quite low (3000 lb/h) and that too is distributed among different incoming lines - would it be Ok if we use radial nozzles (with impingement plate)instead of tangential ones as shown in EPRI ?

Or should I size the tank as per GPSA guidelines which will give me substantially bigger sized tank ?

Any other point which you think I should take care while sizing

 

[georgeverghese]

Since your flows are low, would suggest using a regular vertical KOD with K=0.2fps as in the GPSA and install an impact plate on the feed nozzle in the vessel as suggested. I dont see the justification for a cyclonic type tangential feed KOD for this service with low flows. Add a flow design margin to account for peak transient flows (at least 10% above max ) when calculating vessel ID. If flow is unsteady due to slugging (or some other reason) in some of these feeder lines, margin of up to 40-50% may be required.
What sizes do you get for these 2 configuration options ?

 

[Vicker85]

@ georgeverghese:

Thank you for you reply!! That explanation was very helpful.

What sizes do you get for these 2 configuration options ?

For Congigurations of GPSA vs EPRI (with the actual flow provided) :

Per EPRI : ID = 20"

Per GPSA (K=0.2) : ID = 32"

Any observations or comments on those values ?

 

[georgeverghese]

So, with a 36inch pipe acting as your tank, you should be able to accommodate up to 25% flow design margin with a standard vertical KOD (no tangential inlet or internal cyclone).

 

[Vicker85]

@ georgeverghese:

Thanks a lot for your valuable insights on the topic! Very helpful.

 

[pierreick]

Hi,
To support your work :

https://checalc.com/calc/vertsep.html

https://www.red-bag.com/engineering-guides/249-bn-eg-ue109-guide-for-vessel-sizing.html

Good luck
Pierre

 

[Vicker85]

@ pierreick:

Thanks a lot for those links! Helpful to double check the sizing.

 

[pierreick]

Hi ,
additional resource :

http://www.pennseparator.com/fssel.htm

Good luck
Pierre

 

[don1980]

If you're designing your own vap-liq separator, it's essential that you understand the variables that fundamentally affect the design (process and mechanical design). From the discussion above, I sense that you're not currently familiar with these. My advice is to purchase this separator from a company that specializes in separator design - there are a number of them (ref: Google liquid vapor separator). And BTW, even if you were knowledgeable about the key design variables, I'd still suggest purchasing from a vendor. It's very easy to overlook a very important mechanical design detail that leads to the separator not performing as planned, and/or leads to wearing a hole in the separator after a relatively short time in service.

The required diameter has everything to do with the design of the inlet. The first step in separating the flash vapor from the liquid droplets is preventing the production of even more droplets, and to start the coalescing of the existing ones. That's the reason from the tangential entry nozzle and the internal channel around that inlet stream. The separator diameter is designed based on the amount of vapor load. Specifically, it's designed based on the superficial vapor velocity limit. If the separator does a good job of conditioning the inlet stream, then the vessel diameter can be smaller, as compared to a separator with a poor inlet design. Thus, the inlet design and the vessel diameter are inextricable linked. The vessel diameter and height are also a function of the tolerable droplet diameter in the outlet stream. If, for example, you need to limit the droplets to 300 microns instead of 600 microns, then you'll need a larger vessel diameter, more height above the inlet nozzle, and/or a demister in the top of the vessel (mesh or vane type).

Another important detail is the vessel height and the height between the liquid level and the inlet stream. As previously stated, you don't want the inlet stream to create more droplets. That can happen with a poor design, where the high velocity inlet stream impinges on the liquid level, creating more spray.

Then there's the metallurgy design considerations. A hard SS material and proper welding procedures are needed for the inlet nozzle and the inside protective wear-belt. Otherwise, the high velocity flashing condensate will quickly erode a hole in the vessel.