8 replies to this topic

[joelbr]

At our plant, we have a HP Steam Separator that will have increased inlet flow in the next coming year. I am trying to determine whether the Separator is sized adequately for such an increase. It used to be a vertical, cyclonic separator used to separate water and steam from 70% quality to ~99% quality steam.

Due to damages internally, the separator now is a vertical, gravity separator and most likely can only provide a maximum 90% quality steam. Using the GPSA handbook, the given dimensions of the vessel, and the process condtions, I was able to determine the maximum droplet size entrained within the vessel.

My problem is how to use this infromation to determine steam quality (i.e. liquid carryover). Anyone have any ideas? I am guess I would need a droplet size distribution as well as some other information.

My findings were:

With a Separator of ID = 60" and L = 240" operating at 14,000 kPa and 337 deg C:

Case 1: 4,900 Sm3/d of dry steam (7,000 wet) has water particles less than or equal to 54 microns
Case 2: 6,300 Sm3/d of dry steam (9,000 wet) has water particles less than or equal to 78 microns
Case 3: 8,820 Sm3/d of dry steam (12,600 wet) has water particles less than or equal to 133 microns

Like I stated earlier, all I want to know is the approximate liquid carryover in the vapour stream.

Thanks for the help

[Art Montemayor]

Joel:

To my knowledge and in my experience, there is no manner in which you can determine the amount (or particle size) of liquid entrained in the overheads of any separator. You can, of course, use theoretical math calculations to estimate these values, but those are simply estimates – and you will find no one can confirm or give you the guaranteed accuracy of these theoretical calculations – especially on the particle sizes. The simple reason for this is that there is no known “particle size” meter available to use on the overheads stream and analyze or monitor the overheads. If there is one that has been developed, tested, and accepted in industry I would be very interested in finding out about it.

I have designed, built, and operated both 2-phase and 3-phase separators in the field and I never had carry-over problems with any of the products. I used conservative, empirical values in my calculations.

[Pilesar]

joelbr,

Steam separators are usually sized smaller than process separators because the steam separator is meant to be a rough cut. You can usually use a larger droplet size like 300 microns and then check to see if the separator will meet your maximum carryover criteria. Note that steam quality is not static and will be greatly affected by heat loss in your downstream piping. If you have no carryover, you can still end up with liquid in your steam. Proper steam traps are still needed and may ultimately have a greater effect on steam quality.

[joelbr]

Joel:

To my knowledge and in my experience, there is no manner in which you can determine the amount (or particle size) of liquid entrained in the overheads of any separator. You can, of course, use theoretical math calculations to estimate these values, but those are simply estimates – and you will find no one can confirm or give you the guaranteed accuracy of these theoretical calculations – especially on the particle sizes. The simple reason for this is that there is no known “particle size” meter available to use on the overheads stream and analyze or monitor the overheads. If there is one that has been developed, tested, and accepted in industry I would be very interested in finding out about it.

I have designed, built, and operated both 2-phase and 3-phase separators in the field and I never had carry-over problems with any of the products. I used conservative, empirical values in my calculations.

Thanks for the advice. I'm curious then as to why it is common to specify the removal of particle sizes greater than a particular value if one does not know whether this level of separation will provide no liquid carryover/liquid retention. My ultimate goal is to estimate the quality at high inlet flow rates to the separator with known dimensions to ensure adequate separation. Is there possibly an alternative method one could use to determine how poor the steam quality will become at increased inlet flow without doing a rate test in the field?

[joelbr]

joelbr,

Steam separators are usually sized smaller than process separators because the steam separator is meant to be a rough cut. You can usually use a larger droplet size like 300 microns and then check to see if the separator will meet your maximum carryover criteria. Note that steam quality is not static and will be greatly affected by heat loss in your downstream piping. If you have no carryover, you can still end up with liquid in your steam. Proper steam traps are still needed and may ultimately have a greater effect on steam quality.

Thanks Pilesar,

I understand the dynamics of steam quality and heat loss downstream. What I am considering is purely the outlet quality of steam directly at the HP Steam Separator. Your suggestion is to set the particle size to 300 microns and compare calculated vessel dimensions to the actual, correct?

[Bobby Strain]

You can go to www.bobby-strain-group.com and find a rating program to check for theoretical separation. But it's usually the drum vendor who can best advise you.

Bobby

[joelbr]

You can go to www.bobby-strain-group.com and find a rating program to check for theoretical separation. But it's usually the drum vendor who can best advise you.

Bobby

Thanks for the resource, Bobby. I did initially ask the drum vendor to adivse, but since the top portion of the HP Steam Separator was abandoned (the cyclonic portion - due to supposed performance issues), they were reluctant to provide information.

[Bobby Strain]

You will probably need to install new internal demisters in the steam drum to get satisfactory performance. The drums are designed for economy and rely on demisting internals to produce good quality steam. You won't find any way to predict performance without internals. Theoretical calculations are of little use. Velocities are high to the point of re-entraining liquid. So, best to start looking for a solution sooner than later.

Bobby

[Pilesar]

If you are constrained to use gravity settling only without a mist extractor, you should not count on the ability to efficiently remove particles much below 300 microns. So assuming 300 micron particle size will let you get a size required for 1-2% entrainment. This would be a 'theoretical' answer to use in comparing your existing drum.

Large particle sizes are certainly appropriate for a condensate flash drum scenario. However, after rereading your question, I think you should not rely on gravity separation without vanes, cyclones, or some other means of handling fog and mist droplet sizes.
Figure 7-6 of the GPSA handbook suggests the expected particle size for condensation from a saturated vapor is in the range of 0.1 microns to 50 microns while condensation on a surface would extend the particle size range to about 800 microns.

I recommend you work toward getting new internals and letting the specialty vendors help. Otto York is now a part of Koch-Glitsch. ACS and AMISTCO combined to form AMACS. These companies have many years of experience.
http://www.amacs.com/
http://www.koch-glitsch.com/
They will propose solutions based on testing their proprietary designs.

In theory, there is no difference between theory and practice, but in practice, there is.