Hi All,

. . . to build myself a small, (relatively) cheap house. Given modern chemistry . . . 

. . . I can mix/pour small batches of concrete as I have time and cash available, yet still end up with a structurally-sound end-product.

So that part of the project is clearly feasible.

But I've recently had a thought that prompted this post, seeking the collective wisdom of the list's resident engineers and scientists:

Assuming a structure built almost exclusively of steel-reinforced concrete, would it be practical to use a vacuum to hyper-insulate the entire house?

What I'm considering is pouring inner and outer walls separated by a large gap, chemically sealing the walls inside and out against air/water infiltration, then pumping the air pressure within the gap down toward a vacuum.

Essentially, I'd be building the house as an ENORMOUS Dewar flask.

I expect to need THICK steel-reinforced walls to withstand the pressure differential. I was thinking that each wall - as well as each surround for the (minimal) windows and doors - would be 16" thick, with an 8" space in between (e.g. large enough to easily apply sealant internally using a roller brush, at each slip-forming stage). I expect to cast a welded heavy-gauge-steel vacuum-application point into the structure's poured-concrete "attic" floor (with a somewhat-more-conventional roof over all). This vacuum point would incorporate a powered air-compressor to pull air out of the Dewar space. I expect to leave the compressor in place as part of the desigh, hooked up to a manometer to kick it on whenever the air pressure in the Dewar gap climbs above a certain level. 

Does this sound like a reasonable (if odd) idea?

Or am I missing some point of physics (such as the compressor needing to run so often that the power bill will make this more expensive than conventional insulation design) which makes my idea ludicrous, rather than merely odd?

--