Design advise on using a vaporizer to pump liquid.

[Duane365]

Has anyone ever used a vaporizer to pump liquid? The liquid to be pumped is a highly corrosive toxic acid. Currently we block-in a tank heat the tank and boil some of the liquid. We then use the pressure in the tank to pump the liquid out. The problem is that the liquid is at its boiling point and as it flows through the line the pressure drops causing it to flash. It is then hard to increase flow. It also takes to long to pressurize the tank

"Plant operations" does not want to use a pump or Nitrogen to push the liquid.

My solution is to use a vertical vaporizer attached to both the liquid and vapor space of the tank. This would allow the tank to be pressurized without increasing the temp of the tank or introducing a foreign material into the tank. I know some of the vapor would condense on the surface of the liquid, but what other things could cause problems? I can not believe this has not been done before, but I can not find any information on the Internet for this type of arrangement.

I can’t seem to paste my diagram so I hope I have explained this well enough to understand. My apologies if I have not. Any help would be appreciated.

 

[Ashereng]

Duane365,

I don't have any experience doing what you are proposing.


Let me clarify what I think you are trying to do: are you trying to increase the pressure above the liquid level to "pump" the liquid out of the tank? (assuming that the hydrostatic head alone is not enough?).

I am not quite sure what you mean by a vertical vaporizer. I am imagining a stand pipe where you vaporize (or boil) a small portion of the liquid to increase vapor pressure above the liquid.

If yes to the above assumptions, I would think that this should work (please see first sentence above).

However, are there any alternatives, such as:
1) elevate the tank and use larger diameter pipe (increase head)
2) lower destination point and use larger diameter pipe(increase head)
3) transform liquid corrosive toxic acid to a solid and transfer (we transfer adipic acid as a solid)


Anyway, that's my 2 cents worth. Hope it helps.

 

[pmover]

Duane365,

there are pumps that can be designed to pump the corrosive toxic acid - magnetic sealless pumps is one.

i understand (at least i think i do) the current operation and certainly understand that it takes significant time to heat the fluid in the tank. coupled that with the fact of the fluid flashing downstream in the piping (due to reduced pressure, i.e. head loss), the process to remove the fluid can be tedius and maybe even hazardous (elevated temperatures).

how about pressurizing the tank with N2 to transport the fluid? is N2 compatible (i.e. non-reactive) with the fluid? it seems that if the tank is pressurized with the fluid, certainly pressurizing with N2 is an alternative. of course, it depends upon how big the tank is. if tank is large, perhaps a surge tank (smaller) can be used like a batch process to remove the fluid. for example, via piping connections, fill surge tank by means of gravity flow, isolate from main storage tank, slightly pressurize with N2, and then remove/transport the fluid. repeat the process until fluid is removed from the storage tank.

hope this helps. good luck!

-pmover

 

[Ashereng]

The basis of the problem stated is:

"Plant operations" does not want to use a pump or Nitrogen to push the liquid.

 

[Duane365]

Ashereng,

You are correct in you assumptions of my “vertical vaporizer” boiling liquid in a standpipe. The important part is that the vaporizer extent above and below the level of the tank and boil a small (compared to the tank) amount of the liquid. The tank is actually a pressure vessel. We normally operate at 125 psi but it is ratted for 250.

Changes in elevation would not be a practical solution to the problem. This is used in a chemical plant therefore changing its physical form to a solid is out as well.

Thanks for your help

 

[Duane365]

pmover

To clarify, there are actually two of these tanks in service. One tank is being filled and heated while the other is in operation. They then switch for a continuos process.

N2 was my first thought as well, but the fluid is being pumped into a reactor. Increased devolved N2 will slow the reaction.

We use a number of sealed pumps some mag drive, some canned, many with hastolly internals. We replace them all the time.

This system is very reliable we would like to keep it that way. If it were not for increased rates it would not be an issue. We could increase pipe size, but this would be expensive. We could increase temperature but at some point the liquid starts to degrade. This does not address heat-up time either. We could put in a third tank and larger pipe but this is an expense we would like to avoid.

We are not at the limit of flow and heat-up time and it is the choke point for our plant. There are many ways to fix the problem, but is the method I presented a reasonable option?

 

[Ashereng]

Duane365,

I think it should work. I drew it out on my board, and it looks right.

In a closed system, the temperature may also be used to control flow rate as increase in temperature will raise the the pressure in the vapour space (because you will also elevate boiling point). An increase in pressure, assuming all else constant, should result in higher flow rate.

 

[Zoobie]

I agree with Ashereng. Since we don't know your process conditions it would be difficult to speculate further.

The only thing that occurred to me is that your vaporizer could act like a thermosyphon which means that you may end up at close to square one with regards to total heat input...probably mitigated with attention to design. Though presumably a recirculating system would be 'faster' than if you had a simple immersion or jacket heater in the tank.

 

[scalleke]

Duane365,

Question I have is "What is the vapour that you have above the liquid phase in the vessel?".

I do believe you are on the right track.

Heating up the main volume of liquid that you have gives the nasty side effect that all the liquid you are trying to move is being heated up ( I understand that you do not want this).

Would it be possible to heat up only the vapour phase?

Best regards.

Scalleke

 

[sterl]

Compressed air diaphragm pump might be a potential selection for the seal and shaft problems associated with momentum effect or rotary PD pumps. They would isolate your fluid from contamination and not introduce bearings and seals and so on...They can be monitored and managed to give almost any level of protection.

As to vapor displacement: Most systems involve a compressor for the vapor...LNG systems for tank car transfer, certain storages for NH3 and similar. Corken makes "oil free" compressors just for LNG.

Doing it with heat alone, without appreciable warming of the bulk of the liquid, you can displace something less than 50% of the fill volume with an "inherently" safe/ self managing arrangement. At the delivery pressure, is the ratio of Liquid to Vapor Density above, say 300? and is there more than 30-deg. F. Temperature Difference between the Vapor and the Cooler Liquid? If so, you will need to consider how to avoid "condensation pinging". At the limit, the latter effect will easily rupture your PV.

 

[Duane365]

sterl
I googled "condensation pinging" and found nothing. Please explain further.

 

[sterl]

In a single component system, if you introduce vapor as bubbles into free standing liquid that is considerably colder than the saturation temperature, the bubbles collapse or implode. With Sparge Pipe type vessels, this often heard as little pings...Even with well distributed introduction of vapor, the mechanism will erode ordinary surfaces within pressure vessels; various pressure vessel internals are incorporated to mitigate the effects. Burping one vessel full of near saturated vapor into another full of cold liquid, represents the classic demonstration and can create some pretty wild reactions.

In a near-horizontal pipe, entrapping vapor such that it is surrounded by cold liquid results in Condensation (or Condensate) Induced Shock, which has all kinds of documentation in the literature, most of it for Central Steam Plants. See

http://www.kirsner.org/pages/condInduceWatHamText.html

and others by the same author.

Transfer tanks or tranfer arrangements based on these premises are utilized in many industries. The tanks and the piping have to be designed together: the detailing and design issues are discussed in the literature. See also:

http://catalog.asme.org/Education/ShortCourse/

Predict_Thermal_Hydraulic.cfm?TAXONOMYITEMID=3029