Heat Convection Inside an Autoclave

Stainless steel or cast iron turbines, but it looks like your problem is the hysteresis of that huge cylinder, your temp controller might need to be setup accordingly but your cycle time will increase considerably, I don't see a workaround.

Uranium right ?, not plutonium...

The autoclaves are in operation and I don't want to add a port to them for the air movement due to recertification of the pressure vessel, otherwise the turbine or any fan would work. I have tried relocating all the thermocouples and added a cascade loop for better dynamic control of the heat. Heating the 14 ton DUF6 cylinder quick enough to generate the process pressure to feed the reactor is the most difficult thing I have ever experienced in my career. Thanks for the reply.

Have you considered band heaters? Then you take convection out of the equation and replace it with conduction. But you would probably need safeguards against overheating. From your response to my earlier post, it sounds like equilibrium was assumed and heat transfer was not considered. Since heat transfer is a time-rate process, time is an important element. And when you are heating a static mass (also encountered in food and other processes) other factors are brought in. Sound like you may need to bring in an expert in heat transfer.

An after thought, would it be possible to use multiple, smaller containers? This would increase the surface/volume ratio.

BWCS (Babcock & Wilcox Conversion Service) seems to not have any particular issue getting depleted UF6 conversion to work, why don't you ask them? Are you using induction or radiant heat? Direct heat conduction? Are you sure?

You say you want more conduction. Could you use heat baffle to create a thermal siphon within the autoclave? Could you somehow use magnetic coupling to drive a fan? IMHO - you really need to hire a firm that can engineer this correctly from the get go. You need to define what it is you need to accomplish to them (not on this forum).

James I am the Vaporization Engineer here at the nuclear site I work for B&W. Thanks for the response.

If you could raise the autoclave temperature (still within operating certificate), and reach liquidus pressure, then I presume a lot of the issue with differential heating of the cylinder might be resolved, but you still need to provide for better convection, by using non-moving baffles within the autoclave. Otherwise - I hope you can find a fan that can operate within the autoclave with only an electrical feedthru, and magnetic bearings.

Sure, put the motor on the outside of the autoclave and run the fan shaft through the wall of the autoclave with a suitable seal. At high temperatures some shielding is used, but at 140 may not be required.

But depending on the form of the contents, a convection fan may not help much. For example, if the contents is pellets submerged in air the heat transfer rate will probably be controlled by the contents and not by the outside of the container. How was 70 kilowatts selected?

The heating capacity was calculated based on the ideal gas law and the properties of DUF6

Obviously your calculations are wrong.

I am not sure I understand what you are doing. DUF6 is a solid at room temperature and at atmospheric pressure sublimes to a gas as it is heated. If it is stored in a pressurized steel container, when you heat it, it would also increase the pressure. The phase diagram shows that it can be a liquid (like CO2) under pressure.

As it crosses the phase boundary, additional heat is required (heat of fusion) in addition to the specific heat needed to increase the temperature.

Ref: http://en.wikipedia.org/wiki/Uranium_hexafluoride

Are you sure of the phase diagram of UF6? Are you sure you are not having to vaporize it first?

http://web.evs.anl.gov/uranium/guide/uf6/propertiesuf6/index.cfm

As long as you stay under the triple point, then you are subliming the DUF6. That will be the absolute upper pressure you would want. I think there is a slight problem getting the heat to equilibrate through out the material in the cylinder since this is a temporal issue more than a control temperature issue.

Heat baffles might be the simplest means of controlling the heat to the area near the cylinder such that baffles placed near the cylinder should induce a downdraft in the furnace to "bathe" the cylinder top-down in hot gas. That still will not really speed things up all that much IMHO, because the mass of DUF6 is large and of a thick diameter within the cylinder. If different parts of the material are at different temperatures, the hotter sublimes the faster, and probably separates from the walls of the cylinder. UF6 is not something I would pick for a heat transfer gas in general (due to the mass of the molecule). You are probably wanting this material to do something it cannot.

The other option is to control the heating up past the triple point into the liquidus, and hold pressure and temperature such as draw-off will be based upon the "boil-off rate", whatever that is. A liquid will at least stay in contact with the portion of the cylinder where it rests. Can your system take the pressure? Your call.

Just put a slightly subcritical mass of U-235 inside the cylinder....

My convection oven that I cook my food at home cannot be set any lower than 300 °F (150 °C) in any of the convection settings. A depeleted uranium hexaflouride should only be handled by qualified and authorized people. You should not be touching this cylinder.

redfred I work at a nuclear facility and we are well versed in safe handling practices of UF6 cylinders. We are just attempting to speed up the heating process for a higher throughput.

Ok, now you're being clear. Got a question for you: Do you know if Eddy currents can be induced into the UF6 cylinder? If so, then go for an induction heater, and forget about the hysteresis inherent to a heat convection system.

Have to do it without modifying and having to recertify your equipment ? then do magic.

I am most likely unclear on this, but would some form of stirling cycle device be practical?

The original calculations were fine the heating process works as designed we are just trying to improve the cycle time without major redesign diminishing returns is not an option LOL. Thermal imaging shows a much higher temperature in the vapor space above the cylinder just thought there may be someone out there who might know of some way to move the air around in the autoclave.

The good folks at Getinge use fans in their steam autoclaves, perhaps they could help?

That is why I mentioned a stirling cycle heat engine earlier, driving a fan. It would work if there was enough differential. Also, remember the principals of sodium cooled exhaust valves. I offer these ideas in keeping with your requirements for minimal invasion of the chamber.

But you have tried everything that has no moving parts? Such as heat baffles that will force convection within the autoclave?

The system can be seen -if I am not wrong- as 2 cylinders one in the other. I think that in actual situation two air flow loops (as 2 kidneys) are build one on each side of the central cylinder with a free convection circulation: near to outer wall hot air goes up and along internal wall cooler air goes down. If the distance between cylinders is too small those loops cannot develop completely and heat is partly transferred by conduction. In both cases upper space will be hotter than lower space thus limiting the surface transfer capacity. In any case this flow will NOT heat the lower part of the inner cylinder, heat transfer will be limited to only an upper part of it even if the 2 loops are fully developed.

I presume the computations were based on an uniform air temperature which is wrong and demonstrated by made measurements.

The solution is a better hot air flow and if possible going from free convection to forced convection by a circular movement imposed by a fan (or better several if the cylinder is long). Baffles could also be used to lead the flow in a circular AND axial path.

I want to add a remark. The inner cylinder wall conductivity will transfer part of heat from top to bottom but how much it is depends on several factors as the ratio wall thickness/cylinder diameter, the material conductivity and the heat transfer from wall to UF6. The problem is much more complex than it seams to be if detailed.

As for other suggestions may I make following comments:

- baffles are ONLY effective if a flow can be generated and guided by baffles they are ineffective if transfer is only by free convection because flows go also against the free convection trend.

- it is not compulsory to use motors designed for high or very high temperature. It seems to me that this autoclave is dedicated so that it can accept a modification. I would suggest to place the "propellers" inside and use a long, air cooled , shaft to connect them with external motors. The only constrain will be to design the shaft with a stiffness high enough in order to avoid resonance and allow to place bearings outside where the temperature will assure the necessary life expectancy.

I would appreciate an information from the OP as inner cylinder diameter autoclave internal diameter and length so that I can confirm or infirm my hypothesis of the "kidney flows".

Heat baffles for the third and last time.

What?

Moving air isn't your problem, getting enough heat into the mass is. You're trying to heat a 14 ton mass that's roughly 13 ft long by 5 feet in diameter with a 70kW heater.

That's only 239,000 BTU/hr for a 28,000 lb mass, or less than 8.5 Btu/hr/pound. If it were water it would take around 20 hours to raise it 160ºF, and that does not account for heating the autoclave atmosphere and metal (which I would guess weighs much more than the spent fuel), any heat of fusion, plus all the losses through the autoclave walls.

Air is a poor heat transfer medium, how about placing your heaters in tight contact with the cylinder or using an intermediate heat transfer media, but i somehow doubt that you want the problem of disposing of all that contaminated fluid.

You have bumped smack into the laws of thermodynamics and are faced with the inevitable fact that you may be at the the thermal limits of your autoclave and the load you can heat with it.

Mother Nature is a Bitch and you can't fool, or fool with, her.

Ricfre,

I'm glad you're working with this and not myself. DUF₆ is some incredibly nasty stuff. Biggs had a good suggestion of putting band heaters on the tank and doing the heating by direct contact. If the band heaters are placed on the lower portion of the tank and were connected in layers from the bottom upwards on the sides, then your heat is focused on the solid/liquid portion and will be more effective (and the heaters can be turned off in layers/zones as the rising surface temperature in each zone shows that that level is now occupied by vapor.

If band heaters are not possible, then another possibility is infrared (radiant) heat. This may work if the outside of the tank is not polished or shiny. I'm not sure what the tank walls are made of, or coated with--Nickel?? If it is, then IR won't be of much help.

--JMM

Sounds promising. I also wonder about induction heating possibilities, but I don't know if the configuration would lend itself to that.

Your problem depends on 3 heat transfer zones:

-from heater to cylinder wall by radiation and convection

- cylinder wall it self by conduction

- from cylinder wall (inside) to the DUF6 by conduction and/or convection depending on the UF6 structure (conduction if it is solid, convection if it is a fluid)

One thermal resistance cannot be modified : the cylinder wall itself

The outside transfer can be improved but only in a limited way since radiation will depend on the cylinder absorption properties and will decrease as its temperature goes up (you should consider that in the radiation transfer temperature is at the 4th power).

You could increase it only by increasing the temperature at the heating element surface.

You could increase convection, as you intend, by a higher hot air velocity in the space between heater and cylinder. If it this is as efficient as you expect it depends on the geometry of this space which I do not know to appreciate and of course of the heaters position.

This brings us to the internal heat transfer. If you look at the graph Rixter brought you see that at a higher pressure the solid will become fluid at a rather low temperature so that the heat transfer will increase since a liquid will better take and move (convect) the heat from a hot wall. This is not valid for gases it can be seen in the big difference between convection coefficients for gases and liquids.

Your interest would be to start at as low a temperature as possible to heat a fluid UF6 in order to reduce the cycle time.

It is qualitative estimation and it would be good to try to quantify it. If you want to continue the discussion you may use the private channel.

To go further in this analysis you should have heaters at the lower part of the cylinder to increase as much as possible the internal convection which is a "free convection" thus amplified if the heat source is a the bottom.

Back to your original inquiry. This links to an article on high temperature motors used for submersible down-hole motors. You could contact the author, Tom, and see if he knows of any motor that might withstand your conditions.

Would magnetic coupling of a rotor inside the autoclave by a stator outside be possible?