Tube in Tube Coiled heater

[Statesman]

I am looking at a requirement we have, possibly for a tube in tube coil heater, for a closed lube oil system we have. (To be cooled by water). We looked at sizing a shell & tube using typical design values of U from Perrys Chem Handbook of approx 30-50 BTU/°F ft² h. Which gave us a decent sized heater (8" Shell by 6ft Lg) but, when I worked out my Ho (Shell side co-efficient) due to the really low flow of oil through my shell side (10 Gal/min giving me a Reynolds No. of approx 9 !!) my calculated U came out about 5 BTU/°F ft² h, which obviously gives us a tube length that wont fit in the space we have. So I have had to look at various other coolers, the on ethat seems most suited to our requirement is a coiled tube-in tube design, I have a couple of questions regarding this which I am hoping someone can help me with. :-

1) Obviously I can work out the inner tube (water) hi value quite easily as we have turbulent flow going through and I can work it out from the Prandtl No. I plan on having the outer tube carry the oil (I will try and make the gap pretty tight to try and promote turbulent flow) However, when I come to work out the ho for this inner tube, what is the best way to calculate it ? (Should I work out my Nusselt from Nu=0.24Re^0.6 and then my ho from Nu=ho.do/k
or is there another way that should be used ?

2) This is more of a construction question. Obviously one tube sits inside another tube and is then coiled, how is the inner tube seperated from within the outer tube (Obvioulsy we dont want one tube to touch the other) Are they separated by some kind of fin/baffle attached to the inner tube ? Does anyone know a formula for working out the size of the coil knowing the tube lenght and required radius of the coil. i.e. I could work out the circumferential distance but how do I incorporate the 'pitch' of the coil into this ?

Thanks in advance

 

[ardilesd]

1) For the outer tube, take the hydraulic radius to determine the Reynolds number (See Kern )

De= 4 x Flow Area/Wet perimeter
= 4 x (do^2-di^2) /(4 x di) = (do^2-di^2)/di

Use this De for correlations (Sieder-Tate) , extract jH from charts as a function fo Re

For laminar flow ==> Nu= 1.86 x (Re x Pr x D/L)^(1/3)

2) Not sure, but you can use some helicoidal spacer between both pipes, ensuring that the pitch is long enough, avoiding restriction to flow area

 

[chaucer]

Regarding the pitch I would start with finding the surface area needed to transfer the heat [A = Q/U/LMTD]. Then you can vary the exchanger length, coil diameter, and pitch that gives you the best configuration to satisfy the surface area requirement. It's an iterative procedure.

The way that you suggest calculating the outside ho based on a Nusselt Number is correct. You need to be sure to use an appropriate correlation for the configuration that you select.