I have planning permission to build and have been investigating different approaches. I have decided upon brick and block construction for various reasons and the Structural Engineer has specified a suspended ground floor due to heave potential. I have modelled my building in PHPP and the basic input data generates compliance so I am drawing up the details.
  
I have not found any suitable detailing for a suspended beam & block floor to achieve the sorts of U values that I would like. After lots of investigation I have drawn my own details for the floor and floor-wall junction based on the insulated floor system produced by Cube6. Their floor systems are “claimed” to achieve U <0.1 but I do not like their cold bridge detailing. The detailing in the attached drawing is my first attempt and I would appreciate peer review. The cavity is 250mm (was to be 300mm but PHPP is useful) and foundations have to be 1800mm (clay & trees!).

Some of questions:
1. Would I need to form a cavity tray across the XPS as I am in a low risk area for driving rain?
2. Does the concrete infill have to meet the XPS?
3. Does the concrete infill need a run-off slope and weep holes so deep down?

Is it possible to use a raft foundation in your situation?

Thanks for the responses, useful.
The Jetfloor system looks to be very similar to the Cube6. Spantherm is worth investigating as I intend to build to ground floor myself and their installation service (crane) would be useful.

I agree the Turbo blocks do not need to go further than two courses below ground floor level. I had thought about trying to keep them dry and am considering painting the sides with something (suggestions?) and using waterproofing agent in the cement.

I shall look in to using standard beam and block with insulation above followed by screed. But I suspect the whole floor thickness will increase meaning it will have to go further below ground level. I think this method would be slightly better than a Cube6 or Jetfloor system but it means I might have to excavate more earth away. A problem I am aware of from digging trial pits is they filled up with water out of the ground at about 0.8m, so I don't want to go to deep.


Raft foundation is not really practical as it could not be cast in one go because of the staged build - see below.

More back ground: The build is a bit complicated as it involves erecting a two storey extension in front, and separate, from an existing bungalow. When the extension is complete I will move in to it, demolish the existing bungalow and then rebuild. The rebuilt bungalow will be moved forward and joined to the new part to form one building. I know its complicated but it took a lot of pre-app advice submissions before discovering this would provide the space I want and be allowed!!!

I will post another floor idea shortly.
Thanks everyone.

Here is another idea, it will cost about the same as the Cube6 or similar but has a drawback - because PU foam has a conductivity of about 0.022 less thickness is needed to get the desired U value. The weakness being the surrounding walls because Turbo blocks only have a U value of 0.11 and I can only fit one within this thickness. Because there is no insulation under the beams there is little point carrying down the insulation below them - heat will conduct through the beam block floor to the supporting wall.

With the Cube6 floor the insulation is deeper allowing more Turbo block depth and a better overall U. As the beams are insulated underneath then there is merit in insulating the supporting wall below. I will draw another picture and post later.

Yep that's a similar floor buildup to mine, but 200mm of celotex (PIR) should get you better U value than 0.12, I'm getting same U value with 150mm.

Buildability wise its handly too, I did all myself, with no help. With a build up of 150mm I first laid 50mm, and my plumping pipes a beam Hoover ducting was cut out of this layer and then 100mm over the top, all joints staggered.


Check the BBA cert for the bead to ensure you can bump to that thickness.



If you are worried about cold bridging on the inner leaf you could line the exterior walls with PIR. I've 50mm of Quinn therm, and more than happy, friends have used 25mm.

Also have a think of your screed. 75mm or 100mm would be standard for sand/cement, but drying out time rule of thumb would be 1day/ per mm, could be more. A lot of guys are using quick drying screeds, cement or liquid based, then can also be laid thinner all allowing more insulation but at the cost of less thermal mass.

I checked the BBA cert for blown bead and they don't specify a maximum fill; lots on minimum cavity widths! They list a cavity fill of 215mm in one of their tables. I also got a quote from a local firm some time ago for a full fill 300mm cavity so I hope they don't change their mind.

I have been using PHPP to check the insulation factors and, yes, I have entered the data for a ventilated underfloor void (PHPP notes say it is not a recommended construction - but needs must). This building is right on top of a 60year old apple tree in medium expansive clay which is why the S.E has specified a suspended floor and a 175mm void.

I think I will have to have the suspended floor so I should show the telescopic ventilators on my drawing!
I will check with B.C before final decision.

They are an alternative brand to Quinn Lites and other AAC blocks (autoclaved aerated concrete), so lightweight and relatively low thermal conductivity, but this degrades when they get wet and they can soak up a lot of water.

Plasterers don't generally like them, but brickies do as they are so light and can be cut with a handsaw. Not very good for taking mechanical fixings though, so I wouldn't want to hang cupboards on them!

<blockquote><cite>Posted By: ward32</cite>I checked my calculation on Celotex @0.022w/mK and at 200mm U=0.11

I think any house built to the sorts of standards we are talking about will be excellent compared to 98% of UK stock. But I am looking for the best possible solution, for a reasonable price. I intend to do the floor element myself so I can get all the detail right.

The attached detail uses Cube6 system around the concrete beams, insulation below DPC and lateral insulation 300mm below the floor. I have calculated the U values of the insulated elements up to floor level and they make 0.2 so not too bad. I'm sure it will be fine!<div class="Attachments" id="Attachments_212585"><ul><div><img src="/newforum/extensions/InlineImages/image.php?AttachmentID=5431" alt="wall floor footings detail 2.JPG"></img></div></ul></div></blockquote>


Have a look at the kingspan tf70 150mm gives you U value 0.11-0.12 (dependant on perimeter ratio)

I'd include anything inboard of the insulation and PHPP method will include an internal and external surface resistance. I've seen some people include the layer of blinding etc under their insulation, but I think that's stretching it a bit!

@djh, how deep did you need to go, and did you consider putting in a basement?

Posted By: ringi@djh, how deep did you need to go, and did you consider putting in a basement?

Our raft needed a hole 1.2 m deep, so the strip foundations would have had to be 2.4 m (NHBC 'guidance'). I thought about a basement for about half an hour before I realized it would cost a boat load more money for something we would probably never use.

Hi,

picking up on a couple of points:
1) any jetfloor/cube6 type floor, with screed in direct contact with the beams will have more heat loss compared to putting the insulation on top of the beams. This is because you have a large surface area with high thermal conductivity in direct contact with the wall, which in turn is in direct contact with the ground. Putting your foam glass below the beam and block would help.
Putting the insulation above the floor reduces this problem, but you'll still have some losses where the internal load bearing walls puncture the insulation.
2) consider running the insulation on both sides of the internal skin of blocks as far into the ground do as possible, to make the heat path as long as possible, rather than filling between the inner and outer skins with concrere. You may need to lay the inner block on flat, or consider using ICF, up to floor level, to give the strength required.
3) word of caution re full fill insulation. NHBC are currently reviewing all full solutions, and are reverting back to 50mm ( or even 75mm) clear cavities over and above the insulation after a raise in the number of claims.
4) British +SAP flooring U values take into account the shape of the building in the U value, so you get a better figure that it would appear you should based solely on the thickness and quality of the insulation. In PHPP the u value entered in the spreadsheet, ignores the shape of the building, which is accounted for elsewhere in the sums. The heat loss is the same, just different accounting methods.

hope this helps.

Without full fill insulation how can we depend on the insulation being fixed correctly so does not move about and leave gaps?