# A semi-detached house

Hi everyone,

Do you have any idea how to create a semi-detached house with one or more walls without losing heat?
None of the strategies outlined below have been successful.

I have heard @niraj.poudel mention that, when walls are missing like that, and you have perfectly aligning floors, Sefaira will project wall surfaces between and design as if there were walls completing the envelope of the building.

1 Like

Yes, I know, but I wanted to check it with walls tagged as â€śignoredâ€ť or â€śinternalâ€ť and as you can see it actually doesnâ€™t work. So there is a way to do this? We often design such buildings.

Hi @igmalion,

Also I might point out is that if you want one of the surface boundaries to not lose heat then please tag that boundary as a â€śshadingâ€ť surface instead. Sefaira considers any shading surface as an adiabatic surface and will not have any heat transfer through it.

Iâ€™m afraid it doesnâ€™t work Or to say it another way, during the wall tagged as â€śshadingâ€ť is adiabatic Sefaira still creates a new wall between the floor and the roof.

Oops, yes, I forgot that that happens due to the automated wall generation algorithm. Let me check something and will get back to you.

Hi @igmalion,

So what you would do instead is to draw your adjacent space and tag that as â€śshadingâ€ť instead. Thereby preventing any heat flow through the adjacent surface so that the outside boundary condition to the shared wall is one that prevents any heat transfer to the outside. In the example below (a more exaggerated version) I just drew a shading surface 1" off of the wall and connected that shading surface to the building and tagged those surfaces as shading as well.

Thanks @niraj.poudel for this example.
Iâ€™ve tried this before and the results were similar to yours, but Iâ€™m having trouble understanding - shouldnâ€™t the energy transfer through the walls be less because the surface area of the outer wall is smaller?

Hi @igmalion,

I canâ€™t tell for sure without diving into the hourly gains + losses but if I were to take a stab at what might be going on is that, the shading surface in the analysis above is on the south facade and therefore the facade that would otherwise be receiving incident solar radiation isnâ€™t anymore. The analysis is set in New York and the heating loads are higher than cooling. The heating loads arenâ€™t offset by the solar gains from the south facade because of the shading surface blocking any incident radiation on that facade. The heating requirement is therefore going up and cooling is going down. I have the default ASHRAE baseline set as a template and therefore do not see conduction driven by the dry bulb temperature differential to be a major driver of gains/losses in this building.

Either way, if you want to test out different scenarios and look at hourly outputs then you can also take the .idf file and test it out further.

Thanks @niraj.poudel for sharing the idea. I will investigate it in days.