Triple glazing u value effect

[SBMS]

Does anyone know if there is a general rule

of thumb for how the u value of the glazing unit affects the overall window u value. This is just for some rough calculations whilst doing some cost benefit analysis for different windows. 
 

In particular, say that a windows total u value is 1.4 based on a window u value of 1.0. If that window was changed for a triple glazed unit with a u value of 0.6, what might one expect the total u value to decrease to? 
 

My simple logic (probably wrong) makes

me think that as proportionally the majority of the surface area of the window is glass, the majority of this 0.4 reduction in u value is conferred across the whole unit, possibly bringing it down to 1.0 overall u value. The bigger the window, the more pronounced this might be. 
 

Is this in any way correct, is there a way to approximate how lower u values of glazing might affect the overall frame u value?

[G and J]

Having been reading about this stuff incessantly of late I think I know how this works but how about you post some dimensions and I’ll use it as a revision question and work some numbers?

 

(and in the meantime I’ll revise!)

[SBMS]

Just now, G and J said:

Having been reading about this stuff incessantly of late I think I know how this works but how about you post some dimensions and I’ll use it as a revision question and work some numbers?

 

(and in the meantime I’ll revise!)

We’ve got windows that range from 2300x1300 to 600x1660 So It’s a bit tricky to post particulars @G and J as we’ve got all sorts of different size windows, I was just wondering if there’s a rough formula?

 

You can end up obsessing about u values…

 

[G and J]

14 minutes ago, SBMS said:

You can end up obsessing about u values…

Yes I do!

Go on, pick one window to use as an example.  I’ll be sitting round waiting for heavy drilling equipment to come available (yep, the dentist) and doing a worked example will distract me.

[JohnMo]

57 minutes ago, SBMS said:

Does anyone know if there is a general rule

of thumb for how the u value of the glazing unit affects the overall window u value

Rules of thumb are generally wrong when you go into detail. Will all depend on the ratio of frame to glazing unit. If you know the actual glazing unit U value and the ratio of frame to glass you could work out the frame U value. Then back calculate to the new glazing and U value.

 

Suspect it will be a bit of rubbish in rubbish out.

 

 

[Alan Ambrose]

OK here's a few numbers - from my window schedule in PHPP nominally using Rationel Auraplus & Velux. S/W/E windows using Rationel Sun SKN176; N wndows using Rationel Auraplus 4/20/4/20/4; Veluxes using Velux Heat Protection Glazing 69:
 

Average Uframe=1.132, Uglazing=0.522, Uwindow=0.712, Uinstalled=0.822

 

If I change all the glass for better stuff with 90% of the U-value (e.g. 0.5 glass becomes 0.45 glass etc), the numbers become:

 

Average Uframe=1.132, Uglazing=0.470, Uwindow=0.675, Uinstalled=0.785

 

So, sounds like the Uwindow/Uglazing factor (the discrete derivative? ) might be around (0.712-0.675)/(0.5-0.45) i.e. 74%. Hope that helps .

 

I guess you can also sometimes get this from the window frame supplier's various double/triple glazed offerings - if you believe their numbers...

 

You can play these type of games fairly easily by mangling PHPP if you choose to.

 

Edited July 26 by Alan Ambrose

[G and J]

5 hours ago, Alan Ambrose said:

Average Uframe=1.132, Uglazing=0.522, Uwindow=0.712, Uinstalled=0.822

So, an installed window with a total of 1m2 showing (frame and glass) would lose 0.712W for each degree in temperature difference.   I’m assuming Uwindow means the overall u value of the window when it’s installed.

 

if Z fraction of the window is glass then:

 

 

(Z x 0.522) + ((1 - Z) x 1.132) = 0.712

 

Therefore:

 

(Z x 0.522) + 1.132 - (Z x 1.132) = 0.712

 

so….

 

Z x (0.522 - 1.132] = 0.712 - 1.132 = -0.42

 

thus….

 

Z = -0.42 / -0.61 = 0.689 AKA 68.9%

 

5 hours ago, Alan Ambrose said:

Average Uframe=1.132, Uglazing=0.470, Uwindow=0.675, Uinstalled=0.785

therefore :

 

0.689 * 0.470 + (1 - 0.689) * 1.132 = 0.32383 + 0.3521 = 0.676.

 

Wow.  I did it!   Do I get a sweety?

 

 

[Alan Ambrose]

>>> Do I get a sweety?

 

Certainly for effort .

 

Well yes, given the data you were given, you're right. PHPP, for my windows, says Z is between 62% and 82% with an average of 71%. Not sure where the discrepancy is - I guess I was giving averages over a bunch of windows, so that's it.

 

 

 

[SteamyTea]

Not read the above as had a hard day at work.

But as a general rule, for every double of thickness, you halve the power losses (note power, not energy).

A window works the same as any other thermal barrier, except it allows some wavelengths of light though it (light has difference energy levels depending on the wave length, note energy here).

It also has internal convection currents that are temperature dependant, so not linear.

But for any given static temperature difference, you can use first order physical properties to get a good estimation of the losses.