Consider a pre-engineered flash tank for steam. The overall cost may be cheaper since the engineering is in the price and the sizing has been tested in multiple industrial installations. See for example: http://www.spiraxsar...Information.pdf
You can design steam flash tanks yourself, of course, but I have always calculated much larger tanks when using standard vapor-liquid separator guidelines such as found in Chapter 7 of the GPSA Engineering Data Book. I think the separation criteria is different for steam than for process streams since you will always have condensate in the saturated steam lines which have traps anyway. So what if you get wet steam out the top of the tank? The economics for larger 'normal sized' flash tanks may not justify them. I have tried to back-calculate how the expert steam specialists get such small diameters for their steam flash tanks, but there must be some engineering principle I am not taking into consideration.
If you have tangential inlet in your design, expect wall erosion and include enough material in the target area to account for severe wear.
Hi Pilesar,
I realize that this is an old post, but I was searching the site for steam flash tank sizing information, and I found your post. I am currently troubleshooting a steam flash tank, and I too have noticed the discrepancy for the required diameter between the common methods (e.g., API 12J) and the results of steam/condensate flash tank manufacturers.
In my mind, the difference must lie in the K-factor (as in, Va = K * [(rhoL-rhoG)/rhoG]^0.5 ). Although many references commonly refer to the K-factor as a constant, it is really a velocity term with 0.35 ft/s or 0.107 m/s as "typical" values. The assumed maximum allowable velocity for a droplet to settle is then modified by the density term.
This in turn, must be fundamentally related to the droplet size assumed in standard oil/gas processing references. Since organics typically have a much lower surface tension than water, they make smaller (finer) droplets; thus requiring a lower maximum allowable velocity for them to gravity settle. The problem is, I cannot find the stated K-factor for steam/condensate systems. It must be a much higher number.
For the unit I'm currently troubleshooting, it was apparently designed with a vapor velocity of 30 ft/s in the flash tank (i.e., not the vent line). This seems ridiculously high to me, but the unit was designed by a well-known supplier.
Have you been able to find any further references to an acceptable K-factor for steam/condensate systems?