High Temperature Heat Transfer Fluid 3

[DrSmile]

Hello,

I am trying to design a Fluidized bed reactor where i have to supply large amounts of heat to the solid bed at high temperatures : 600-800 C.
A large part of the heat has to be supplied indirectly, through immersed in bed heat exchanger tubes, and will be supplied either by combustion of hydrogen or of coal.
My question is in what conditions steam can be used for heat exchange at these temperatures? Does it imply a special choice of materials for the exchangers? Thank you very much as i am new to the job.

Vlatko

 

[thermcool]

If you are thinking of using steam at such high temperatures (600-700 C), you have to choose the material that can contain high pressure steams. May be stainless steel can solve the purpose but you have to choose thickness of the tube accordingly. See from the steam table what is the pressure for the steam at saturated/superheated state for 600-700 C temperature. For the obtained pressure you can choose the appropriate tube thickness from the SS tube data. Also, you can use the hoop streaa model to calculate the minimum required thickness for the tube.

 

[quark]

Forget about steam. The saturation pressure at 374⁰C is 225.5 kg/cm². There may be better fluids than superheated steam. I used Dowtherm (Dyphil) upto 400⁰C by pressurising to 10 bar. You should better speak to Dow. Also, wait for some more suggestions, for there are better engineers here.

 

[25362]

Upon burning hydrogen with air, nitrogen-diluted superheated steam would be generated, right ? Are you trying to substitute superheated steam for combustion gases ?

 

[btrueblood]

If you need a fluid, and not a gas as 25362 suggests, then at these temperatures you will need to consider molten metals. I've seen and analysed molten tin, sodium, and potassium systems. You then have an issue about how to handle start-up and shut-down transients (i.e. how to keep the pipe from freezing full of once-liquid tin).

 

[thermcool]

Metals like mercury and sodium are used in the heat pipe for high temperature application. The startup problem are normally less than liquid working fluid like mercury.

 

[DrSmile]

Hello everyone,

thanks for the very quick replies.
To answer 25362's question, yes i was considering burning hydrogen and using the resulting gases to supply the heat, but was wondering about the material constraints on the heat exchangers.
Hydrogen burning not being ideal for this application, i was also looking for anotther heat transfer medium at these temperatures.
What are the advantages of molten salts over superheated steam for example?? Just the specific volume? or something else..
Does anyone know of any other high-temperature heat transfer fluids? Thanks very much

Vlatko

 

[unclesyd]

The Diphenyl and Diphenyl Oxide Heating fluids have one bad habit, the will detonate around 495°C.

The normal heat transfer salts (nitrates/Nitrites) generally have a max operating temperature of approximately 600°C.
The neutral salts will go much higher. I have never seen them used for any thing other than for heat treating.


Here is the salt we built our process, 450°C, on. We have made a couple of modifications to the composition.

U-TEC-TIC Salt

http://www.hummelcroton.com/utt/utt.html

We have used these salts in the bench scale and pilot plants with good results. Park has "Neutral Salts and Tool Steel Salts (Chlorides) that operate much at higher temperatures. We have use them but have had corrosion problems.

Partherm 290 and others

http://www.heatbath.com/Products-link-page.htm

Here is another HTS that we have used for process cooling.

HITEC

http://www.coastalchem.com/PDFs/HITECSALT/HITEC%

20Heat%20Transfer%20Salt.pdf

http://www.coastalchem.com/index.html

I wasn't directly involved with this project but years ago we did have a shallow fluidized bed, once through, that took it's heat, around 1300°C, from a NOX incinerator. It operated for a couple of years.

 

[davefitz]

If you can allow contamination of the bed material with the products of combustion of coal ( ash, sulphur, etc), then you can easily raise the bed material temp to 800 C by using a cascading fluid bed system. One fluid bed is used to combust the coal ( and heat the bed material), and this bed material can cascade doen across a weir or thru a spout inot a parallel chamer. The other fluid bed activity ( which is the object of your process) would occur in the parallel bed.

As I recall, the US DOE had proposed such a process in order to simulate the activity of the pebble bed reactor- they wanted to heat up high pressure helium in the bed tubes that are placed in the parallel fluid bed.

 

[DrSmile]

Hello,

Thanks for the replies. Unfortunately i can't allow any pollution.
I have read the numerous threads concerning the uselessness of superheated steam as a heat transfer fluid.
However, maybe in the case i am working on it may prove to be the right answer because:
I have 2 reactors directly using steam at around 800C. Another reactor generates heat at 350C.

Furthermore,for the reactor i wish to heat indirectly the heat transfer coefficient of the fluid bed to the immersed exchangers is expected to be low (app. 200 W/m2/K) similar, from what i gather, to the coeff. of superheated steam.

Do you think these reasons could justify the use of superheated steam in heat transfer? If yes, does it require any special materials??

Thanks,
Vlatko

 

[25362]

Perhaps the following threads may shed some light on the metallurgy for high temperatures:

thread725-37130
thread367-58995
thread330-62473
thread330-55502

As for htc the steam-side coefficient appears correct, what about the process side ?

 

[DrSmile]

Hey thanks for the links,

On the process side, the fluidized bed-to-wall htc is expected to be 200-400 W/m2/K so it should be quite similar. I did the calculation and it apears that to achieve the reaction in the time i need, it requires 1 m2 of exchanger / 0.185 m3 of fluidized bed reactor in the worst case.. in the best it's approx 1 m2 / 1.4 m3.
Since i have little practical experience i don't quite realize if it is a lot or not.. What do you think?

Thanks a lot,

Vlatko

 

[25362]

Probably a good reading material with plenty of references would be Ullmann's Encyclopedia of Industrial Chemistry Vol. B4 (VCH).

 

[25362]

If the process side fluid is a hydrocarbon, heat fluxes may be greater than the endothermicity of the reaction and coke may form on the tube walls. Experts also say that wall-originated Fe can catalyse coke formation. Coke would then act as a thermal blanket.

If the coke barrier is porous enough, coke precursors could still diffuse to form new coke at a self-accelerating rate. Since no details were given on the process in question, this is pure speculation.

 

[Jingyuan]

Liquid sodium can be used as a heat transfer medium. It is used in nuclea power plant. I involved with project using liquid sodium to heat fluidized bed calciner. The sodium has a melting temperature less than 100 degree C and remains liquid up to a temperature greater than 800 degree C. The material of the immersed coil was inconel

The advantages are: high heat transfer coefficient from sodium to tube wall, steady temperature so no overheating problems etcm high heating capacity.

The combustion gas can be used as heating medium, but the disadvantage is that the heat transfer coefficient is low. Large amount of transfer surface has to be used.

 

[PropulsionGuy]

I would like to learn more about molten tin as a heat transfer fluid. I don't mind preheating the tin tankage to keep it molten and I would like to use it to temperatures in the 1000C range. Is there data available to predict heat transfer in cylindrical tubes and also pressure drop as well as materials compatiblity? I understand that Sn corrosion rates with refractory metals are relatively low. What is the appropriate pump type, flow measurement and pressure sensor approaches?