Oversized Flash Vessel 4

[esterplant]

I have sized a flash vessel with operating pressure 14 bar and design pressure 21 bar. The actual vessel in the plant seems to have diameter twice of what I sized but height more or les the same. However this flash vessel have a great water carryover. A decision is made to install a separator after the flash vessel but there are still problem because the condensate of seperator will have steam with it and this is a loss to us. Can anyone suggest to me why my flash vessel have such a great carryover?

 

[samuelliu]

It is undersized, perhaps.

 

[StoneCold]

esterplant
Have you tried calculating the vapor velocity in the existing vessel and comparing it to the entrainment velocity of the liquid droplets? Why add a second drum instead of adding a demister or coalesser pad?

 

[BenjyMac]

Possibly your condensate removal from the bottom of the flash vessel is undersized or the steam traps(if used are undersized)

Regards

Benjy

 

[quark]

May be the pressure at which you are flashing the steam is not sufficient. However data like the condensate flow rate, condensate pressure and flash steam pressure, if provided, will give better understanding of the process.

 

[esterplant]

Steam(water) enter the flash vessel at 1.5 Metric tonne per hour(MT/hr),65bar,100% saturated liquid. Dry steam, come out at 14 bar at 0.3145 MT/hr.Liquid flow out at 1.1855MT/hr,14bar. Actual dimension of flash vessel in plant:Height=1.645m,diameter=0.8m,steam outlet,condensate inlet and outlet diameter=50NB.There was no demister pad in this flash vessel. Why serious carryover happened in my flash vessel?Why not install a demister pad but install a separator?Is it cheaper and more troublesome to install a demister pad in it?

 

[jproj]

I think everyone is saying that you should install a demister pad or some other type of vapor/liquid separator inside the flash tank instead of downstream of the steam outlet. Demisters are cheap (couple hundred dollars), but welding them into a 2.6m diameter existing tank would be extremely difficult if not impossible.

A larger diameter tank will reduce the vapor velocity within the tank and will reduce liquid carry over to the steam outlet. Your flash tank's diameter is too small and the inlet velocity is causing liquid carry over in the steam line.

 

[quark]

I don't think your flash vessel size is not adequate. Infact the diameter is so big that I am thinking my calculation is wrong.

Presuming your numbers are correct the calculation goes like this.

The enthalpy difference of hot condensate at 65 bar and 14 bar is (295-197) kcal /kg and latent heat of steam at 14 bar pressure is 665.7 kcal/kg.

So the percentage of flash steam is (295-197)x100/665.7 = 14.72%

Total condensate flow rate is 1500 Kgs/Hr, so flash steam generation rate is 1500 x 0.147 = 220.5 kgs/hr. The specific volume of steam at 14 bar is 0.143 cu.m/kg, so the volume flow rate of steam is 31.58 cu.m/hr or
18.57 cu.ft/min. Now considering the recommended safe velocity of flash steam (which will reduce carryover)at
600 fpm (10 fps) equation for the diameter in feet is

[(18.57 x 4)/(3.142 x 600)]^1/2, well I am afraid to give the answer. check it on your own.

Now you can calculate other way round to know the velocity of steam considering 0.8 m as diameter. It is coming out to be exceptionally low.

Now my questions are,

1. Is it a continuously drained system? If so check the condensate flow rate with the calculated one.

2. what are you doing with that high pressure condensate? (for a weaker soul like me 14 bar is still high pressure) I just want to know about back pressures.

3. Is your flash tank insulated?

3. For more learned gentlemen, what if the velocity of steam is very low? None of my books addressed this issue.

Regards,

 

[StoneCold]

I agree with quark. I calculated the numbers as well and came out with values that are similar to quarks. Have you looked at the nozzel placement? Have you looked at the liquid level in the vessel? Are you sure about the pressures? Are you sure that you do not have steam flooding into the condensate header and drastically changing the conditions in the separator? The vessel appears to be much larger than necessary so I tend to think that if the nozzel locations appear ok then the vessel is not the problem and you need to look at insulation problems downsteam where the condensate is developing.

 

[jproj]

quark / BradStone:

You guys are calculating the percent flash incorrectly. By definition:

h(vl) = X*h(v) + (1-X)*h(l)

where
h(vl) = two phase enthalpy
h(v) = saturated vapor enthalpy
h(l) = saturated liquid enthalpy
X = quality

If you re-arrange the above equation, you can calculate the % flash (quality):

X = [h(vl)-h(l)]/[h(v)-h(l)] = [295-197]/[665.7-197]

X = 0.209 or 20.9% flash which gives a 313.6 kg/hr steam outlet flow.

...Not like it matters to terribly much. His vessel is plenty big. I'm not really sure what the velocities need to be, but if they are not high enough, the centripetal force will not separate the vapor from the liquid. Either a mist eliminator (between the flash inlet and the steam outlet line) or a baffle system is needed to separate the entrained liquid. If the condensate level is low and the nozzle is in the lower portion of the tank, there may be enough vapor space to keep entrained liquid from carrying over into the steam line.

jproj

 

[saxon]

esterplant, Just a couple of questions, 1. Is the condensaste removal system sized large enough to handle the condensate load? 2. Is it operating correctly?. Solving one or both of the problems will in all probability solve the carry-over problem in the flash steam.

saxon

 

[StoneCold]

jproj
You are correct in saying the percentage flashed is really 20% not 14% however for this vessel it really does not make any difference. You bring up a good point though about asking if the vessel has either a tangential feed nozzel or baffles above the feed nozzel to help with disengagement. That could be another helpful piece of information?

 

[quark]

Thanks for correction Jproj and a star for you.

 

[esterplant]

I have the actual dimension and drawing to indicate the location of all nozzles. The location of condensate inlet is only 250mm from the vapor outlet. To me this top entry is causing serious carryover. What do you think? Anyone interested to have a look at the drawing? Just provide me with your email address.

 

[quark]

builblock1@yahoo.com

 

[StoneCold]

colorado_stone@yahoo.com

 

[quark]

esterplant!

I went through your drawing but without much result. Everything seems to be ok. but few concerns.

1. As you are pumping condensate inside the vessel almost tangentially, lighter water particles may raise and enter steam outlet. (steam also froms from the water collected below, still it is a speculation)

2. I checked for splashing of condensate due to striking on the other side of the vessel. The distance of other end of vessel from condensate inlet is 550 mm approx. (from your drawing) If I consider velocity of condensate at 2.2 m/sec (15000 kg/hr, but loalised velocity at exit may be high) it takes 0.25 sec to reach the other side of wall. So water would fall vertically and strike the wall at 307mm below the condensate inlet (opposite to it). So no problem of water carryover because of flashing.

3.You didn't tell us whether the condensate drain is continuous or intermittent.

4. Is it possible for you to provide a downward elbow at condensate inlet to the tank?

Beyond this I am exhausted.

Regards,