In another thread:

http://www.greenbuildingforum.co.uk/forum114/comments.php?DiscussionID=9430&page=1#Comment_153897

JSHarrisThe latter point is key, as compensation control assumes a fixed linear relationship between outside temperature and heating requirement. This may or may not be the case, depending on the installation. For example, I tried (and failed) to get such a control system to work on the system I installed in our house in Scotland. The problem was that heat loss was far more dependent on local wind speed around the house than it was on outside air temperature.

My first thought is that just the effect of wind blowing over the surface of the building can't make that much difference. The thermal resistance from a flat surface to the air next to it is pretty small (of the order of 0.1 m²·K/W, I think). Wind will reduce this but even if it reduces it to zero it'll not make much difference compared with even a moderate amount of insulation.

What seems much more likely is that wind was blowing in to the insulation, bypassing it. The need to have a reasonably wind-tight layer outboard of the insulation is well understood (at least in theory). Blower door testing is used to check the in-board airtightness layer but has anybody given any thought to verification of the outboard air tightness beyond simple visual inspection?

The only test I know of which would capture this effectively is a co-heating test. This involves heating an unoccupied house to a constant temperature with metered electricity (fan heaters) over a number of weeks, so the relationship between heat loss and weather conditions can be established.

The results are often a lot worse than the advertised U values would imply due to a combination of air leakage & thermal bypass. However, the results can be difficult to interpret if the house is still drying out.

David

<blockquote><cite>Posted By: Ed Davies</cite>
What seems much more likely is that wind was blowing in to the insulation, bypassing it. The need to have a reasonably wind-tight layer outboard of the insulation is well understood (at least in theory). </blockquote>

Not as far as I could see. The house had a brick plinth to floor level and was rendered block above that, painted with the toughest paint I could find (and even that didn't stand up to salt spray that well). It was pretty well sealed externally, had a 50mm cavity between the brick/block skin and the membrane covered, ply faced, timber frame (which would itself have been fairly well sealed, I'd have thought) then a layer of rockwool or similar around 150mm thick, then there was a plasticised foil membrane of some sort, then plasterboard on the inner surface. The roof was a warm roof design, clad with ply inside and out, with a membrane over the outer skin. The window were all Norwegian made DG units, with double seals and multiple catches all around (they were laminated timber frames, tilt and turn type windows). The external doors were pretty massive timber affairs, internally insulated and again with multiple catches around to pull them tight to the seals. The choice of house design took into account the exposed location on the West coast, and used pretty much the best available techniques at the time for keeping draughts out.

Getting back to wind chill effects, the standard insulation calcs assume a fairly low air velocity across the surfaces. Increase the air velocity and rate of heat transfer increases, quite markedly if the outer skin of the building is at a significant temperature above ambient. Ideally the insulation would be perfect, and the building outer skin would be at ambient, but I'm not at all sure this can be achieved in practice even now, and certainly wasn't the case for my Scottish house built in 1993. Add in that weather variability can change the outer skin temperature up and down fairly quickly, aided and abetted by the wind, and you can have, as I found, some fairly big differences between the heating demand predicted by the outside air temperature and the heating demand that's needed in reality.

Posted By: SteamyTeaShould do a Wind Heating Degree Days and see how it correlates, could also do a Direct Beam Sunlight Heating Degree Days (though not nights) too. And then there is the effect of rain. Suspect what you end up with a a climate score based on all variables.

I log temp, solar and wind.
I prepare HDD data. It's definately not enough on its own.
A bit of solar can make a significant decrease in heating need[1], and a stiff breeze can make a significant increase.

Could you expand on the concepts of WHDD & DBSHDD? How do I include the wind/solar elemants. A quick google didn't find me any equations.

[1] A bit of sun helps both actual temps and the perceived need for heating (human response to radiant heat?).

Posted By: skyewrightCould you expand on the concepts of WHDD & DBSHDD?

We can all try:
http://www.greenbuildingforum.co.uk/forum114/comments.php?DiscussionID=9446

Our place needs much more heating when its windy (1800s vintage, airtighness slowly improved but still far from perfect). A fall of snow makes the place noticeably warmer, I guess the wind blows snow into all the cracks and gaps and blocks them up.

JSH, were there ventilated cavities in the wall, floor, roof? The rate of air changes in the cavities/roof/floor increases with the wind speed and direction. In calm weather the air layer in the cavity and the blockwork/slates act as additional insulation layers (OK maybe not much), but when its windy the cavity is flushed out with cold air - the tent flysheet effect. By convention the heat loss calcs assume that the ventilated spaces are always equal to external temperature, but in a ventilated loft in calm midsummer this can be felt not to be true.