My roof insulation was based on the knauf insulation spec for
between-joist-only insulation, the example of which was 225mm wool
between 225mm joists, with 6.2% bridging due to the joists,
which gave a sufficient U-value. After refusal, I checked mine and it
was 10% bridging due to
joists being thicker although further apart. So, I need to fix it.
Also, I could replace some of the wool with celotex so that the
bridging is OK. I need to keep
the roof a certain thickness, thus not wanting to add celotex under or
over the joists.
How do I calculate U values when there are parallel conductivity paths
(joist / wool) and serial
(if I add a thin layer of celotex) in the same structure ?
I need calcs to be able to prove that my fixed insulation spec is OK.
Thanks,
Simon.

Also, I could replace some of the wool with celotex so that the
bridging is OK. I need to keep the roof a certain thickness,
thus not wanting to add celotex under or over the joists.

Work with conductivities, which is the reciprocal of the uValues. Then you
can multiply them by the area, add them up and take the reciprocal again to
get the final value.

Finally, are you sure you can't get anything under the joists? Cold bridging
has deleterious effects in addition to the overall energy loss, which is all
the BCO will be interested in.

In particular, with cold bridging, you will get cold lines on the
plasterboard which will discolour and attract mould. You would be far
better, if at all possible, to install even a little insulation underneath,
even if it is only 10mm or 12mm to eliminate this possibility and provide a
uniform warm surface to the plaster.

Replacing rockwool with celotex is unlikely to be a bad idea, even with the
underboarding. The building regs minimum is a minimum. It would be to your
own benefit to improve on it.

Christian.

wrote:
My roof insulation was based on the knauf insulation spec for
between-joist-only insulation, the example of which was 225mm wool
between 225mm joists, with 6.2% bridging due to the joists,
which gave a sufficient U-value. After refusal, I checked mine and it
was 10% bridging due to
joists being thicker although further apart. So, I need to fix it.
Also, I could replace some of the wool with celotex so that the
bridging is OK. I need to keep
the roof a certain thickness, thus not wanting to add celotex under or
over the joists.
How do I calculate U values when there are parallel conductivity paths
(joist / wool) and serial
(if I add a thin layer of celotex) in the same structure ?
I need calcs to be able to prove that my fixed insulation spec is OK.
Thanks,
Simon.


You can sum U values over parallel paths easily enough.

U value = sum (U value times area for all the bits)/total area.


Thats is basically saying that for a given section there will be a
certain heat loss in watts per degree C..and summing all those losses
gives you an overall watts per degree C of the whole roof..dividing that
by the total area nets you an 'average' U value..don't forget that if
you e.g. put a skin of plasterboard on the bottom, and a skin of e.g.
chipoard on the top, you need to do the other calculation for U values
in series.. which is that 1/Uvalue final = 1/U value plasterboard +
1/Uvalue chipboard + 1/Uvalue calculated as above for beams and rockwool.

I can't be arsed to do the calculations for you..but they should be
available online.

Work with conductivities, which is the reciprocal of the uValues. Then you
can multiply them by the area, add them up and take the reciprocal again to
get the final value.

Finally, are you sure you can't get anything under the joists? Cold bridging
has deleterious effects in addition to the overall energy loss, which is all
the BCO will be interested in.

In particular, with cold bridging, you will get cold lines on the
plasterboard which will discolour and attract mould. You would be far
better, if at all possible, to install even a little insulation underneath,
even if it is only 10mm or 12mm to eliminate this possibility and provide a
uniform warm surface to the plaster.

Replacing rockwool with celotex is unlikely to be a bad idea, even with the
underboarding. The building regs minimum is a minimum. It would be to your
own benefit to improve on it.

What about conductivities in series ?
I guess these operate in the reciprocal sense to the reciprocal rule,
if that makes sense.

I also need calcs for the roof, so I am seeing a structural engineer
anyway next
week. I will see if he can get the roof thinner and allow for the
celotex.
I would still need to work out the thickness of celotex to get the
U-value acceptable.
By the way, the reason for the roof thickness fuss, is that I need the
pitch over
12.5 degrees for the tiles, I need headroom at the eves and I have a
mono-pitch roof on the party wall to line up with. And ... I'm keeping
the roof open nearly to the apex.

Simon.

What about conductivities in series ?
I guess these operate in the reciprocal sense to the reciprocal rule,
if that makes sense.

Actually, that's what I accidentally described, due to a brain fart. To
calculate the overall uValue, just multiply the %age of the material by the
uValue of that part and sum them all together. Only take reciprocals for the
series calculation.

Christian.

On Mon, 30 Jan 2006 13:42:01 +0000 The Natural Philosopher wrote :
You can sum U values over parallel paths easily enough.

U value = sum (U value times area for all the bits)/total area.

Not that easy unfortunately: you need to do R/upper and R/lower
calculations. Example on our website at
http://www.sda.co.uk/shuvalpr.htm

--
Tony Bryer SDA UK 'Software to build on' http://www.sda.co.uk
Free SEDBUK boiler database browser http://www.sda.co.uk/qsedbuk.htm
[Latest version QSEDBUK 1.12 released 8 Dec 2005]

Not that easy unfortunately: you need to do R/upper and R/lower
calculations. Example on our website at
http://www.sda.co.uk/shuvalpr.htm
No idea what uppers and lowers we are talking about - for doing a whole
room ??
But, it leads me to part L appendix B which has a nice example of studs
/ insulation /
sheathing that I can apply. The sums aren't to bad really, so no need
to buy superheat !
Thanks,
Simon.

Hah, BCO is wrong. He probably looked at the cold bridging and got
scared
I calculated for my case and the one from the knauf website.
My figure for the case on the website came up with 0.18 which matched
their figure, so sums OK.
My figure for my case came up with 0.199. Less than 0.2.
And that's without including the plasterboard or boundary layer.
So, I can prove it if I want. If he disagrees, he will have to show me
his calcs !

But, I take on board the cold lines factor. But, 225mm of wood is a
relatively good
insulator I would have thought, so not sure this would be a problem in
practice.

But I need to keep on good terms, I may have to confront him with
legalities if
he wants me to pay for the electrical inspection. They want a note in
the spec
that "a test certificate to BSxxxx will be given to building control".
I'll have to be
careful how I word that bit.

Simon.

On 31 Jan 2006 03:38:51 -0800, a particular chimpanzee named
randomly hit the keyboard and produced:

Hah, BCO is wrong...
So, I can prove it if I want. If he disagrees, he will have to show me
his calcs !

Calcs! You think the BCO hasn't got more to do than calculate the
heat loss through your roof. He'll have just looked up your spec on
the Knauf website, noted that their spec is based on 38mm wide trusses
at 600mm centres, and that yours are obviously thicker and closer, and
asked you to prove your spec.

BTW, have you allowed for ventilation? Either you need to leave a
50mm gap over the insulation, or counterbatten over. Maybe he's
assumed that the thickness of insulation will need to be reduced by
50mm.
--
Hugo Nebula
"If no-one on the internet wants a piece of this,
just how far from the pack have you strayed?"