calculating heat loss from window u values and degree days

[rolling-orbit]

Hi All

 

Is this a correct way to work out the approximate heat loss from central heating through windows?

 

Average U-value of windows: 1.0 W/m²K

Area of windows: 10 m²

Degree days in London[1]: 1889

 

(1.0 x 10 x 1889 x 24) / 1000 = 453 kWh a year heat loss through windows.

 

I'm a little unsure about it, especially degree days. All else held the same, I'm trying to quantify the difference in heat loss between different window u-values (e.g. 1.2, 1.0, 0.8), very approximately. I understand perception of heat in the room via radiation is also an important subjective benefit of lower u-value windows, however I wanted to start with the above heat losses first. Thanks! 😁

 

[1] https://www.open.edu/openlearn/nature-environment/energy-buildings/content-section-2.4.2

[2] https://www.heatspaceandlight.com/calculate-much-energy-windows-will-save-u-values/

 

Edited February 14 by rolling-orbit
silly brain not working

[RobLe]

HDD generally has a "base temperature", which will affect the results.  When considering how much heating your home will take to be comfortable, you need a base of maybe 15C - that's the outdoor temperature at which no deliberate heating is required inside.  Ie you don't include accidental heating sources such as body heat, incidental electricity use, sunlight, cats n dogs.

I think the subtly different question of:  "How much energy is lost per year out of the windows" would need a base of 21C.

[JohnMo]

Not sure I would use heat loss days. You would be better using the difference between your average outside temp (say 7) during the heating season, and average inside temp (say 20), as your dT.

 

For you windows you need to use U_W, not U_g for the calculation.

 

Then it's

area x U_w x 24 x number of heating season days (say 180)

[rolling-orbit]

Hi Both

 

Thanks for having a look at this!

 

So instead of looking up degree days from the table on source 1, I could use different between "minimum internal temperature" and "average outside temperature" for the days of the year that I heat the house.

 

Average U-value of windows (glazing and frame): 1.0 W/m²K

Area of windows: 10 m²

Minimum internal temperature: 19°C.

The mean annual outside temperature[3]: 10°C

Temperature different: 9°C

Heating days: 180

 

(1.0 x 10 x 9 x 180 x 24)/1000 = 389  kWh a year heat loss through windows.

 

I've added a table below in case its useful for other people.

 

You both can't see anything crazyly wrong with either approach as a first approximation heat loss from central heating through windows however?

 

U-value of glazing and frame (W/m²K)	Heat loss through windows using annual mean outside temperature (kWh)	Heat loss through windows using degree days table (kWh)
1.2	467	544
1	389	453
0.8	311	363

 

[3] https://www.arcgis.com/apps/dashboards/d863748b238c4f72b2aed0718591b66b

Edited February 14 by rolling-orbit

[SteamyTea]

1 hour ago, rolling-orbit said:

can't see anything crazyly wrong with either approach as a first approximation heat loss

Seem a good place at start from.

 

I generally find that if the mean outside air temperature (OAT) is above 9°C for 3 days, I don't need any heating.

This does not work in December and January though as there is so little solar input (both in intensity and duration) that even my heavy breathing does little to warm the place up (even today, valentine's day).

Generally if my house is unoccupied and unheated it will always be 3 or 4°C above OAT in the winter.

 

Edited February 14 by SteamyTea

[IanR]

2 hours ago, rolling-orbit said:

Is this a correct way to work out the approximate heat loss from central heating through windows?

 

As long as only the U Value differs. A U_W value of 1,0 W/m²K is getting towards high performing windows where the Ψ value of the connection into the building becomes enough of a portion of the heat losses that it shouldn't be ignored. Then there's glass coatings that effect solar gain and air tightness.

I found PHPP really helpful in this respect, all the windows I was considering were available as pre-sets within the spread sheet, so easy to compare one with another. Of course, if you're not calculating losses for a high performing new build then PHPP is over the top, so your ball-park calculations are as good as any.

Edited February 14 by IanR