Cold bridges at the founds/ground slab interface and Sleeper walls

[Mark Siddall]

Has any body done much research into how best to avoid cold bridges at the founds/ground slab and ground slab/sleeper wall interfaces? Are there any innovative products/techniques being imported from other countries with a more enlightened perspective?

Mark

[Anonymous]

Mark
– David Olivier's article in Feb 07 Self Build & Design mag deals with this quite neatly, I feel.
Steve

[Mark Siddall]

Steve,
Had a look at David O’s details. The optimal one was along the lines of what I have been researching (see 2). This said, for those not comfortable with that solution there is another option that I am aware of (not quite as optimal as David’s option though): –

1) THE OVER SLAB INSULATION OPTION:
75mm screed on over slab insulation to extent of floor area. Beneath load-bearing (internal) wall provide butt jointed 65mm thick Foamglass PERINSUL SL insulation by Pittsburgh Corning as starter block. Cavity insulation to be taken below PERINSUL SL by min 150mm NOTE: Min width of Foamglass PERINSUL SL to match that of blockwork over i.e. 100mm, insulation to be installed in accordance with manufacturers recommendations. Hollow or flat indented blocks should not be used on the first course above the Foamglass PERINSUL SL as a uniformly distributed load is required. (This detail can be used on sleeper walls also.)

2) THE UNDER SLAB OPTION:
In the UK we take the founds down to 600mm to avoid frost/ground heave, I have heard that the Nordic states have a slightly different approach whereby they use a raft foundations with much greater regularity: concrete slab, on 300mm load-bearing insulation, on 500mm hardcore, an insulated edge beam is formed by reducing the insulation to a depth of 100mm and turning it up to DPC level at the perimeter (this insulation is the tapered at the head, from falling outside to inside at 45 deg so the external leaf can be aligned with the instu slab edge/insulation). As a consequence the hardcore layer remains level.
At level of the continuous 100mm insulation layer a leg of insulation is extended beyond the building by 600-1200mm so as to form a frost protection (this forms a distortion in the isotherms within the earth preventing frost forming below the slab).
As the UK doesn’t suffer from the same frost problems the 600-1200mm leg could conceivably be cut back.

Mark

[Mark Siddall]

Steve,
The way that you are incorporating the UFH into the structural slab is what they are doing in Sweden and Norway. In cutting out labour and material it is, as you say, cost effective. In this respect I admit that the option screed, insulation, slab option I posted is more costly however my key concern is ensuring the absolute continuity of insulation at this detail in order to avoid linear thermal bridging at the load bearing interface.

The use of thermalite can help to reduce thermal bridging but not obviate it, as a consequence the junction remains a fairly serious cold bridge compared to what is possible; though it is an improvement upon a block with a lower resistance.

One option to develop your detail a stage further would be if you place the Foamglass PERINSUL SL on top of the Thermolite (so that the top edge of Foamglass and the top edge of polystyrene insulation are in good horizontal alignment). This will help to minimise thermal bridging even further. (Having said this check with the rep before you progress this type of detail, I don’t know how suited PERINSUL SL is to use in this condition, and recalling the brochure, I see no reason why it shouldn’t be up to the job….but I would still ask the question none the less.) NOTE: I would not opt to use Foamglass as a complete under-slab insulation solution, in its own right its quite pricy so you need to specify it with care.

Is there a reason why you are not adopting the raft solution, using a similar detail to David O’s perimeter detail, ground beams could be accommodated where there are load bearing internal walls, thus ensuring that the volume of concrete is minimised (saving cost and labour compared to a mass concrete slab). This option would avoid sleeper walls (materials and labour) and a slightly reduce the construction programme (time and prelims).

Mark

[Mark Siddall]

David,
I don't mean to put you on the spot but how is the insulation protected in the details shown in Self Build & Design? The under slab insulation detail show is equally subject to damaged by vermin is it not?

Mark

[David Olivier]

That drawing was schematic only.

One way to protect it is a plinth wall of masonry up to DPM level.
Another would be to continue the layer reinforced render below DPM, as done by some suppliers of external insulation system such as Sto.

I'd like to be sure of the rodent-resistance of the 2nd method. On my house I had some light stainless steel insect mesh actually damaged by wildlife … presumably sharpening their teeth!

The reports suggest a risk of damage up to 700 mm deep in open ground or next to a building, but apparently not below houses. I don't think insulation manufacturers want to know about this issue, since none have ever issued warnings.

BTW Celcon gave me a conductivity figure of 0.18 W/mK below DPM for ultra-lightweight blocks which in dry conditions are 0.11 W/mK.

David.

[Mark Siddall]

Thanks for clarifying that.

Based on what you say the Foamglass could be located out of harms way without to many worries. The potential leaching from bitumen remains. Is the leaching a greater threat than the thermal bridging? I don't know…. Ideally both problems would be addressed, which is why I rather like your raft solution aas other insulants can be used to avoid both leaching and avoid cold bridging.

Mark

[David Olivier]

Mike

Currently not. May be in future.

Also AECB will be publishing some (electronic) design advice anyway on meeting Silver, likely to include a few ground floor details for masonry and timber.

David.

[Anonymous]

Thanks David, I have seen the [draft] Silver standard details as I went to a very good workshop in Cardiff presented by John Willoughby and Peter Warm. Is there much difference between these and your published details?

[David Olivier]

Mike

Yes; the published details refer to situations which may be applicable to “fairly advanced self-builders”. The AECB guidance is more focussed on trying to meet the U-values of 0.15/0.25/0.2/1.5 as in the Silver Standard (NB the 1.5 includes impact of thermal bridge between window and wall) and showing what the thermal bridges are.

David.

[Anonymous]

Thanks David

[David Olivier]

It's not normally recommended for the heat emitters to have a similar or higher thermal capacity relative to the building which they heat. The nearest I can think of to this in the UK which succeeded is the Elizabeth Fry Bldg at UEA whch does have radiant floor heating using an all-air system, and hence a high thermal mass in emitters, but it has very high thermal capacity in the rest of the building, higher than in the cores of the concrete floors. Also it had M&E engineers on hand to design a control system, rectify gross installation errors and fine-tune it long after initial occupation.

Putting heat into a concrete floor is only usually viable iif it's used as a heat sink/store for ambient energy; e.g. solar heat, i.e., on winter days when solar collectors can't produce hot enough water to store in the insulated tank (was done in Switzerland in a pioneering project 17 years ago). Again, it needs a high thermal capacity building, in order to minimise the temperature fluctuations arising from the occasional uncontrolled heat inputs into the concrete slab. You'd still be advised in that type of house not to put your backup heat into the heat store but to put it straight into the house interior (i.e. radiators or fan coils).

I think you need to find a good M&E consultant to design a control system. Inevitably it will need to be continuously-heated with weather compensation, which is fine in priciple – indeed used in rest of Europe – but very different from normal UK practice.

David.

[Author]

Posts