9 replies to this topic

[Erik Tang]

Hello Everyone,

I apologize in advance for the long post. I just want to be clear.

I have read the following link regarding estimating insulation thickness and will reference the equations in the article:

http://cheresources....insulation?pg=3

I need to determine how much insulation is needed to obtain a safe outer insulation temperature. I have therminol66 at 343C going through a 2" NPS carbon steel pipe. I am ignoring the most outer aluminum layer.

The article states, "Since the heat loss is constant for each layer, use Equation 4 to calculate Q from the bare pipe, then solve Equation 6 for T4 (surface temperature)."

Now to solve for Q of the bare pipe, I need to know the temperature of the pipe surface itself with the insulation touching it (T2). This is a value I do not know. The only temperatures I know are the fluid temperature (I will approximate this to be the temperature of the pipe inner diameter as well) and the ambient temperature. How can I solve for the insulation thickness without knowing this surface temperature. In other words, how can I get a Q value?

Or is the author suggesting that I calculate Q as if the bare pipe was not insulated and open to ambient air (call it Q_air) then use Q_air in equation 5. This approach seems incorrect to me because Q_air has a much larger delta T than through just the pipe. In other words the temperature at the surface of the pipe is much higher than the ambient temperature which in turn will produce different heat fluxes.

I have also looked at equation 3 and equation 2. If I consider convection in the fluid, conduction through the pipe and insulation only, I could potentially solve for R3 (insulation radius). delta T would come from my working fluid and desired outer surface temperature. Once again I am stuck not know the heat flux, Q.

Any feedback and help would be appreciated.

Thank You!

[ankur2061]

Eric Tang,

For most piping insulation applications there is the core insulation material (mineral wool, glass wool, perlite etc.) covering the pipe surface circumferentailly and over this core material a thin metal cladding (generally aluminum) which is basically for the purpose of holding the core insualtion material in place and protecting it from weathering.

The heat loss calculations from the pipe are considered based on a surafce temperature of 60°C on the cladding metal surface or in other words the delta T for heat loss is the temperature of the fluid in the pipe minus 60°C.

Now coming to the question about the basis of choosing 60°C. It is the maximum surface temperature up to which any accidental and instantaneous skin contact to the pipe will not cause a 1st degree skin burn. Please note that I have used the word "instanatneous". If the skin is in contact for a longer period then a burn can occur.

It is also important to remember that any insulated surface exposed to solar radiation can cause the surface temperture at the cladding to be more than the value of 60°C.Black body temperture due to radiation is frequently considered to be 82°C in the deserts of the midle-east, which definitely can cause a burn.

I have sometimes also used a pipe surface temperature of 55°C for heat loss calculations and insualtion thickness determination.

Hope this helps.

Regards,
Ankur.

[Erik Tang]

Thank you for the reply and input thusfar. I am still unclear though.


The heat flux for the entire pipe+insulation+aluminum would be based on delta_T=Fluid_Temp-60C, which I understand. There are still two unknowns in eqn 5, the heat flux, Q, and the insulation thickness, R3, needed to achieve my delta_T.

I am after the insulation thickness (R3) so to do this, Q must be known.

How do you determine the correct Q to get an insulation thickness without knowing T2 as suggested in equation 4?

On a side note, you mention the pipe cladding temperature (T4) to be 60C and the pipe surface temperature (T2) to be 55C. Are you suggesting I just set T2=55C or 60C to solve for Q in eqn 4?

Thank you for all your time and effort!

[ankur2061]

Eric Tang,

Some of the earliest hand calculations I have done based on Fourier's law of conductivity for heat loss from a insulated pipe surface had the following form:

q / l = 2*Πi*K_m*ΔT / ln (D_o / D_i) --------------------(1)

where:
q / l = heat loss per unit length Btu / hr / ft

K_m = Thermal conducivity of the insulation material, Btu / ft.hr.˚F

ΔT = Temperature difference between fluid and surface temperature on cladding, ˚F

D_o = Outside Diameter of the pipe including insulation thickness, inches

D_i = Outside Diameter of the bare pipe excluding insulation thickness, inches

A thumb rule that was used and still is conservatively adequate was that the insulation thickness should be chosen such that the heat loss from a unit pipe surface area should be less than equal to 37 Btu /hr.ft²

For an insulated pipe the surface area is

A = Π*D_o*l

or

l = A / Π*D_o

where:

A = surface area of unit length of insulated pipe

Eqn (1) than can be re-written as

q / A = (2*Πi*K_m*ΔT / ln (D_o / D_i))* Π*D_o -----------------(2)

The insulation thickness is now required to be chosen such that the value of q / A is less than equal to 37 Btu /hr.ft²

The above method did not consider heat losses due to convection (air currents) but was still adequate enough to provide a reasonably conservative pipe insulation thickness.

Hope this helps.

Reagrds,
Ankur.

[Erik Tang]

Ankur,

Thank you for your helpful reply. I now have one final question.

Where did this 37 Btu /hr.ft² come from^?

Is it from your own personal experience?

[S.AHMAD]

Dear Erik
1. See attached file.
2. This is my lecture slide for the same subject. Hope it helps.

Attached Files

•  Session 12 - Insulation.pdf   275.65KB   218 downloads

[Erik Tang]

S.Ahmad,

Thank you! Out of curiosity, is Q_convec (on slide 7) based on empirical data or some sort of correlation?

Do you have the source of these equations so I can perhaps find the proof of them?

Thank you very much.

[S.AHMAD]

1. The correlations were extracted from an article in Hydrocarbon Processing sometime ago (sorry I don't have the detail references since I am outstation at this moment)
2. The equations are empirical (if I am not mistaken).

Edited by S.AHMAD, 11 April 2012 - 08:50 PM.

[ankur2061]

Ankur,

Thank you for your helpful reply. I now have one final question.

Where did this 37 Btu /hr.ft² come from^?

Is it from your own personal experience?

This was a rule of thumb very much practiced during the nineties when I started my consulting career. The heating systems used to be designed based on this heat loss value from hot systems provided with insulation.

Regards,
Ankur.

[Erik Tang]

All, thank you for all your help!