u/g tank heat loss calculation

[jmelia]

I need to evaluate heat loss in an uninsulated underground (1 m) tank (20 m3) due to freezing problems.
I have some values for apparent thermal conductivity for different soils based on soil type and moisture content, but I don't know how to estimate the equivalent thickness to calculate the overall heat transfer coefficient.
Any idea?

thanks

j

 

[TD2K]

Pipephase does heat loss calculations for buried pipelines and includes the correlations they use (or used to at least, it's an older manual I have) which I can send to you. Drop me a line at testdog2000@yahoo.com with either an email address and I'll send you a copy.

 

[jmelia]

thanks,
i'll drop you a later at night

j

 

[roaling]

Is the tank jacketed ? I expect yes because underground.

 

[jmelia]

yes, the tank is jacketed (double wall)
i've already take it into account in my calculations, i fixed constant inner temperature, evaluating heat conduction across inox wall, intermediate natural turbulent convection (air chamber) and conduction across carbon steal and coating.... and now the soil, my problem

thanks
j

 

[roaling]

jmelia, thank for your clarification.
Since I have a problem similar to yours, would you please specify how did you treat the air resistance and what is the gap thickness ? Apart from the soil, I think that the air chamber has a significant heat resistance.
Thanks, Roal

 

[jmelia]

sorry roaling...
i looked for natural-convection coefficients in enclosed spaces (Perry 7th 5-13) but my case was out of range...
so i use a 50 W/m2K coefficient as a very conservative value (between 10-50W/m2K are reasonable values, i think )

however the heat resistance in the soil is the most significant

j

p.d.: 8mm is the gap tickness in my case

 

[roaling]

j,
I think you are wrong.
If gap thickness is only 8 mm, I should consider only air conductivity, so the coefficient is much lower. Unfortunately, I have Perry 6th and cannot check your reference.

 

[jmelia]

roaling (surely you're right),
i've also considered only heat conductivity and numerical results (inox-air-c.steel) differed less than 10% with a deltaT= 25ºC as i really considered h(air)=17 W/m2K

j

 

[roaling]

Sorry j I insist.
Assuming 10 mm SS wall: 0.01/15 = 0.00067
8 mm air gap: 0.008/0.025 = 0.32
Assuming 10 mm CS wall: 0.01/40 = 0.00025
1/K = 0.00067 + 0.32 + 0.0025 = 0.32092
K = 3.12 W/sqm°C
This figure differs from that calculated by you (17).
Am I wrong ?