Adiabatic Gas Cooler Concept 3

[rjw57]

Hi,

I am currently working on a scheme to adiabatically cool a gas stream (high thermal conductivity gas (50% mol fraction) mixed with water vapor, nitrogen and CO2) from approx 350degC to 300degC using a stream of saturated water liquid/vapor injected into the gas stream at the face of an open cell nickel metal foam cylinder (gas stream flows along the axis of this cylinder). I cannot guarantee the mass quality of the water, just that it will be saturated. I am trying to make certain that the saturated liquid/vapor combo (~120degC) be injected at multiple locations on the face of a cylindrical metal foam (depth in direction of flow unknown), that it evaporate completely and obviously since the combined mass of the injected water and gas stream being cooled goes to 300degC, that the water becomes superheated. It is most important that the water is fully vaporized since any liquid downstream can affect the next unit op. The foam I am considering is ~ 90ppi open cell nickel made by INCO. Since this is only a concept at the moment, can anyone bounce any ideas on why this might/might not actually work. My two biggest concerns are: 1) distribution (ideally, the water vapor would be well mixed with the gas stream) and 2) total vaporization of flow for the smaller flow stream, the water used for cooling the gas stream. Flow is vertical down so that gravity is causing the water to flow in the direction of the gas stream.

To end, I should ask an even more important question - has anyone reading this done anything like this and would you be willing to share results?

Thanks,
Bob

 

[Montemayor]

Bob:

I’ve used the Unit Operation of simultaneous heat and mass transfer for heat reactor feeds as well as cooling reaction effluents. This is what I believe you are proposing. The adiabatic term is only to describe that there is no heat transferred beyond your thermodynamic system. I’m going to list those details that I can comment on as relates your stated basic data:

1) Although you don’t state it, I’m pretty sure that you are dealing with the following components:
Gas Thermal Conductivity @212 oF
Hydrogen 0.1290
Nitrogen 0.0180
Steam 0.0137
Carbon Dioxide 0.0133

The reason I think you’re dealing with Hydrogen is that it’s the highest thermal conductivity I’m familiar with – and one I’ve dealt with in the past. I don’t know why you just don’t state it, because it would help to know as much basic data as possible. I don't have the Thermal Conductivity properties at 350 oC, but they are relatively higher.

2) I don’t know why you use the term “foam”. I believe you mean sintered metal (Nickel, in this case). I believe you intend to employ the sintered metal as a surface that breaks up the injected water into super fine droplets to accentuate the much-needed intimate vapor-liquid contact. This is a good start. I have found that more is needed to ensure that the desired heat and mass transfer takes place: I would use a packed bed section to add additional opportunity for the injected liquid to thoroughly mix with the hot gas. The reason for this is that, as you correctly identified, this is a Unit Operations step. Chemical Engineers design this type of Unit Operations utilizing a step-by-step type of solution that employs stages of intimate liquid-gas contact – such as a distillation tower or an absorber. That’s why the packed bed section is recommended.

3) You say you “cannot guarantee the mass quality of the water”; well, I have a lot of difficulty with this statement and I am willing to bet you will have the time to regret it if you can’t guarantee the water quality that is being injected into your porous medium. If you don’t ensure that the saturated water being injected is of the highest possible purity (100% free of dissolved solids), your operation is going to come to a screeching halt long before you expect it to. You’ll wind up with a sintered Nickel section plugged solid like a concrete block. I’m willing to bet my year’s salary on that happening if you allow any dissolved solids in the injected water. I have the experience, the background, and the chemistry to prove that it will happen that way. Bottom line: inject only the purest saturated water available – preferably double-distilled water.

4) Your explanation about what happens next is not too clear. I think you mean that you will cool the hot gas from 350 oC to 300 oC and you allege that the subsequent product mixture will not be saturated with water, but that it will contain superheated steam. I assume you have made these calculations already that prove that that is the case. Otherwise, I wouldn’t make the statement without calculation proof at least.

5) If you can’t tolerate any liquid water in your 300 oC outlet cooled gas, then I would not carry out the operation in parallel flow as you describe it. I would use a counter-current packed bed apparatus with the water injected at the top (below a disengagement section) and the hot, 350 oC inlet gas entering the bottom of the bed. The bottom section of the packed tower would have a sump for collection of any excess liquid water to collect there and be drained off as collected.

What I’ve described is what is normally done in this type Unit Operation involving simultaneous heat and mass transfer. Process calculations must be done to identify the amount of liquid water required to do the desired cooling. A heat and mass balance around the system is done to find that out as well as to identify the “humidy” (% of water vapor) in the exit, overhead gas stream.

I hope this helps you out.


Art Montemayor
Spring, TX

 

[rjw57]

Hi Art,

Thanks for the reply. Here are my responses:

1) Yes, it is hydrogen.
2) It is truly a "foam" since it is not a product manufactured using the same techniques as sintered materials. As mentioned, the foam I will be using is 90 pores per inch with a geometric surface area of ~2000 m2/m3, so it should very well simulate a packed bed design. See the related website:

http://www.incosp.com/products/incofoam.asp

3) My term was misleading. In this concept, I have preheated the water stream using a coil at the inlet to the foam section prior to actually delivering it to the face of the foam. What I actually meant was that I could not guarantee the percentage (mass based) of saturated liquid versus vapor in this heat transfer coils outlet. The liquid delivered is of RO/DI >20 MOhm so no dissolved solids problems here.
4) Aspen calcs (process simulation) indicate that the stream is not saturated at these temperatures/pressures (assuming that evaporation and mixing have taken place perfectly). The steam must exist in a superheated state if it truly does go to 300degC.
5) I am aware of the use of packed beds/towers. I am not however terribly familiar with the design/calculation of this equipment. I understand your recommendation to have a sump, but this is precisely what I am trying to avoid. This adds a great deal of complication to the equipment design and operation (I would be retrofitting something in which the gas stream currently flows vertically downward) as well as additional cost and if at all possible, I would prefer to avoid it. I suppose that if I could "distill" my question even further, I would like to know what it would take to be certain that there is never any liquid present at the outlet of the foam. If not, then a sump it is!

Again, thanks for your reply.

Bob

 

[25362]

When speaking of gas simultaneous humidification and quenching, the heat transfer coefficient would differ from the one expected from heat transfer alone without mass transfer.

If having liquid remnants is considered disastrous, thus prohibitive, a "safer" option for quenching the gas mix is by diluting it with the same gas pre-cooled by indirect heat transfer using a "cold" wall.

In this case no sensible heat would flow from the gas to evaporate water, and no simultaneous (co-current) mass flow of water into the gas would take place.

The relevant properties of the gases in question at 600K and atmospheric pressure taken from J.P. Holman's Heat Transfer (McGraw-Hill), are:

Cp, th. cond.
kJ/kg.^oC W/m.^oC Pr

hydrogen 14.537 0.315 0.664
carbon dioxide 1.076 0.04311 0.668
nitrogen 1.0756 0.0458 0.686
water vapor 2.026 0.0422 0.986


Good luck.

 

[rjw57]

25362,

Thanks for your input. I have two concerns with your suggestion (not that your suggestion isn't spot on). First, the size of equipment required to do non-evaporative cooling is significant. Also, any energy removed from the process at this point is not available downstream for recovery (later in the process stream). This will have a direct effect on system efficiency, something of great importance. I will give some thought as to how I could possibly use your suggestions in as compact a cooler as possible WITHOUT giving away the energy to the surroundings. My desire is to stay with coolant injection if possible, but I'm not dead set on it if something is more reliable and meets design goals.

Thanks again,
Bob

 

[25362]

To rjw57, if the evaporative cooling is so important, one can do it outside the foamy metal environment on a slip stream and then mix the cooled gas with the main stream to ensure no liquids are being carried over.

Of course, the cooling effect on the slip stream should be larger to attain the final desired result, and the streams' moving and mixing pressures should be made available.

 

[EGT01]

Can't say I've ever done what you propose but what you propose sounds a lot like a desuperheater, so why not just pick a suitable off-the-shelf model. Is the foam cylinder required for another reason?

However, you seem to be concerned about the potential for liquid phase downstream, from that perspective it would seem no matter what you install (custom design or off-the-shelf) you may want some type of equipment downstream to act as a safety net. Doesn't seem like it would have to be much more than a wide spot in the line with maybe wire mesh coalescer.

Of course I may be oversimplifying the needs.

 

[rjw57]

Thanks to everyone that has replied so far. EGT01's response was something I had not even considered. After looking into this, I may still want to do my own thing, but this is in concept what I am attempting to do. Based on my search, it seems as though many folks would consider this an attemperator since the intent is not to reduce the gas stream to anywhere near dewpoint. As mentioned, I am going to ~300 degC while the dewpoint of the stream is ~90degC. Again, thanks for all the help and wish me luck!

Bob

 

[sterl]

The flow path of a refrigerated compressed air drier applies heat from the immediate discharge gas exiting the compressor to a near-saturated cooled stream as it exits the "Chilled" section, thus reducing the delivery temperature to the refrigerated section while re-heating the delivery to the point of consumption, very effectively ensuring that the "plant air" is well above saturation...

To effect what you're trying to do without accumulating the water, this counterflow to reheat arrangement might be dead simple, particularly if you can tolerate the gas stream geometry involved...

Flow path can be found on Hankinson or Aerotek web sites...