3 replies to this topic

[raxza]

I want to check my existing horizontal vessel size due to increasing gas flow rate. In API 12J and GPSA Section 7 the vessel diameter size is determined from the maximum gas superficial velocity. With the old gas flow rate the velocity is 2.3 ft/s, after I recalculate withe\ the new gas rate (double from the old rate but the pressure is increased up to 60%), the velocity become 6.3 ft/s.

Is there any criteria for maximum velocity gas in a vessel? Or can i assume the vessel as a pipe (due to small amount of liquid) sot he maximum velocity would be 60 ft/s (as per API 14E)?

Or the criteria will be based on maximum liquid droplet? The old design is 150 micron, the gas will be delivered to costumer using burner, so is there any requirement for liquid droplet for the burner (the vessel has no demister and this is the last vessle before sent to the customer) .

Thx.

[Art Montemayor]

raxza:

Some of the information you give allows me to assume several items that you don't state or furnish. However, some of the other information gives me ground to suspect there is an error in your numbers or that I don't interpret your description correctly. Allow me to detail the above:

1) I assume we are dealing with a 2-phase, horizontal, API-type vapor-liquid separator.

2) The separator could be a conventional one for hydrocarbon vapor and liquid water (or oil); it could also be a "3-phase" separator: the liquid portion could be an immiscible mixture of oil and water;

3) You don't state whether the separator is a conventional one as seen in Fig A-1 in API 12J or a "double-barrel" design as also seen in the same Fig.

4) The separator type or design doesn't matter as long as the vapor flow area is kept constant - which I assume is being done;

5) The design criteria used is a variation of the famous Brown-Souders equation: Va = K [(dL - dG)/dG]^0.5

6) If you have increased the pressure, then you have decreased the specific volume of the vapor involved; the relationship between velocity and area is:

Superficial Velocity = (vapor mass rate)(specific volume)/(area)

since the area is constant and the specific volume is reduced, the superficial velocity should also be reduced by a proportional amount. But you state that you have a 3x increase in the velocity. If the flow increases by 100% (double) the velocity should also double (& this is not allowing for a decrease in specific volume). So, our thinking is not on the same plane. Either I have made an error or you have one in your calculations - but both of us can't be correct. Can you straighten me out on this?

7) Stokes law is incorporated in the Brown-Souders relationship and the time of residence in the separation zone is important - it allows for the liquid droplets to separate and drop out by gravity. The length of the vapor trajectory (which sets the residence time) is critical to separate out the liquid droplets.

8) The criteria for maximum vapor velocity in a horizontal separator is clearly stated in API 12 J, so I don't understand your question. If you exceed this superficial velocity, you will start to entrain liquid particles in your outlet stream. Therefore, it is imperative that you stay within the confines of the recommended design superficial velocity for the conditions the vessel was designed for - both temperature and pressure. Do not believe that since you have lowered the vapor specific volume, you can proportionately increase the mass flow rate directly. The "K" values you see published (& those used in the design) are specific to a vapor pressure condition and they will vary with varying conditions.

9) If you purchased your separator from an engineering/fabrication company, then the design conditions were specific to the original specifications. I would not expect to get any appreciable capacity increases merely by increasing the pressure of the system. You will get some increase, but how much is up to the criteria used in the original design. You will have to make that accurate calculation and decision yourself (& take the liability and responsibility for the results) if you make any process modifications that are different from the original specifications.

10) The "correct" K to use for the new operating conditions depends on the system, the fluids, and your expertise and experience in this field. The K is a constant that is empirically obtained and subject to a lot of variables that can't be defined in writing.

11) You should contact your burner supplier and find out if there are any specifications on the maximum size of liquid droplets that the burner can handle without a process upset.

I hope the above comments are of some help.

Art Montemayor
Spring, Texas

[raxza]

Thx for your comments.
Let me explain about your comments:

1,2,3,4. The vessel is conventional Horizontal vessel for 2 phase (Vapor-Liquid). The vapor area is kept constant.

5,6. You were right, I misplaced other vessel area...The velocity becomes double (4.07 ft/s).

7,8,11. Let’s make this more straight. With the new calculations, the velocity is exceeds the maximum vapor velocity of the existing vessel, so is there any tolerance for this, I mean if the velocity is increase about 100% is the vessel still adequate since the liquid probably constant or might be decreased. And I thought the velocity will affect to the separation of the liquid (liquid carry over on the outlet stream), so If this is true I will contact the burner vendor to find out the specification of liquid droplets handling

9,10. The design pressure and temperature is still the same (the design is made to account the future pressure and temperature) BUT the design is not take the account of increased gas flow rate (it’s my old colleague work, now he isn’t here anymore—and I don’t know him all I have is detailed engineering document and this is unexpected gas supply but we MUST take this gas, the agreement is already signed and is expected to flow in the next 2 years). So instead of buying a new vessel (it;s a big vessel) , I want to use the existing vessel.

[Art Montemayor]

raxza:

You haven't furnished any specifics, so all I can offer is general and qualitative advice.

First, if you have to increase the gas flow rate by 100% (double the design rate), then you can surely expect to have liquid droplet entrainment in your outlet. I can't be any more specific, since I don't even know the liquid in question, the pressure, temperature, physical properties, etc. But I would prepare myself for have liquid entrained downstream.

If your downstream burner is a critical service, then I would take precautions to try to reduce (or, hopefully eliminate) droplet entrainment. This is a fair and equitable result since you can't justify doubling the capacity rate to an existing piece of equipment without paying back with some trade-off. In other words, you're going to have to spend some money over and above what was budgeted or proposed - due to this last -minute change in your scope of work. What you describe is a BIG TIME scope change, and it will cost money to resolve. This is to be expected in engineering and in doing business. Having said that, what can be done to alleviate the situation? I would do the following:

1) fabricate and install a second, smaller vessel under the main separator - exactly like you see in Fig A-1 in API 12J as a "double-barrel" design. What this is, is nothing more than a large piece of pipe connected to the belly of the separator with two large nozzles. Make sure the nozzles are liberally sized to allow for self-venting of the liquid as it drains from the separator to the smaller vessel. Control the liquid level in the second, smaller vessel and this allows the separator to operate drier and with no liquid level in it. Of course, you're going to have to cut and weld into the existing separator and will require an "R" ASME stamp. You also will have to raise the separator to a higher location, due to the second vessel underneath. Concrete piers under your separator steel saddles might be an answer.

2) If you can tolerate a reasonable pressure drop (say, 5-10 psig), then I would install internal "chevron baffles in the vertical orientation. These baffles allow the impingement of liquid droplets and their conglomeration into large liquid masses which will drop out at the baffles or immediately afterwards. Peerless Separators made a business of fabricating this type of mechanical separator. They may still be around. Perhaps you could have this type of internals designed, fabricated (& warranted?) for you by others. You might also employ something like York Teflon mist eliminator pads inside the vessel. You don't give us the size of the separator, so I can't visualize any problem doing this.

You have 2 years to resolve this situation and I would start out immediately. Since the separator is already built (& undersized), it is your responsibility at this point and I would take no chances or spare anything that could help ensure that you don't have any future separation capacity problems.

I wish you the best of luck.

Art Montemayor
Spring, Texas