Boat shaped underground home

Hello, I have been pondering about making a south facing hillside earth Bermed house for some time. I just read mike oehlers book and he is strongly against having the entrance / light coming from the south side because you will end up with water at the back of the building. Now I was wondering if anyone tried making an upside down boat shaped house with the tip being the backside of the house deepest in the hillside. In my imagination this will divert the water away reasonably good. It also seems like a solid construction (intuitively, could be dead wrong). I drew a very simple sketch to demonstrate the idea.
So basically I am looking for people who tried it, people who can tell me why it's a bad idea or if it might work.. I find a lot of books and plans on underground small houses but I can't seem to find regular people who have actually done it...

Welcome to permies!

Most of the "successful at staying dry" underground homes I've read about, build the home and then berm all around it so that in fact, the house is mostly above the original soil level and the water level. To get the soil for the berm, they dig a pond.

Water runs down hill - the more sloped the hill, the faster it will run. A house I once visited that had been built in such a situation, had finally spent a pile of money making a serious wall/swale combo to divert all the water coming off the hill away from their house.

So much depends on the exact ecosystem, and the exact hill. I wouldn't underestimate the amount of water you might have to deal with. Water is weird stuff - it can move up-hill if the water pressure is pushing it that way!

Similarly, storms are getting more extreme, and I know of 1 permie that had multiple days of 3 inches of rain with only minimal days in between this year.

You can build almost anything if you keep those principles in mind and plan for the extreme once in a 1000 year storm. I wouldn't do too much planning until you see a piece of land that has potential and someone very knowledgeable in soil and water movement assesses the land, and a good civil engineer advises on the structural side of things.

Beyond that, yes, the principle of making the design so that it will divert any water that gets to it, is sound - I'm suggesting that you design so there's little chance of the water getting there in the first place! Dry soil insulates. Wet soil does not. Modern underground houses are designed with an "umbrella" of various materials and membranes that keeps that lower soil dry. Then above the umbrella, there is wet soil grows grass.

I rather doubt the idea that Mike Oehler didn't want direct sunlight getting into underground houses. I think what he meant in what you read was that a roof should not slope back into the hillside, because that increases the amount of surface water that has to be kept out of the foundation. Making sure that all roof slopes lead downhill or to the side at worst will minimize the amount of water that has to be dealt with. An uphill patio will necessarily collect some water next to the foundation, but proper underground drain piping can direct much of that away downhill. There should be a swale above any uphill patio so that other surface water is diverted before it gets to the patio.


I agree with the idea that raising the ground level more than sinking a house into existing grade will minimize water problems. The more chance for rain to flow far away from the house in all directions, the less chance there is for it to get into the foundation.

"Boat shaped" is basically a version of dome, which is inherently the strongest underground structural form. So along with water diversion before it can get to the structure, doing this is a sound idea, as long as you arrange it so that the space is most usable. I think this would include making the walls expand from floor level to mid-height before curving back in to the ceiling, so that you do not have the space shrinking in the most useful height zone. A structural analysis of soil pressures at each level (which is greater at greater depths) would indicate the best shape for the curve. My sense is that it would want to have a curve in the "wall" section, then a somewhat sharper curve at the transition to "ceiling", then a more gentle curve across the ceiling as the soil load is fairly uniform and smaller than the side loading underground. This also depends on whether the side walls are straight (barrel vault) or curved (dome).