Earth Sheltered Umbrella Basics *
The idea of extending the insulation out horizontally from the roof line of an earth sheltered home (an earth sheltered umbrella) was first presented to me by Earth Sheltered Housing Design, prepared by the University of Minnesota. (In Minnesota, with its very cold, but very sunny, winters, passive solar earth sheltering makes a lot of sense. This lead to government incentives and funded research which further increased its popularity.) The idea is very similar to the “Frost Protected Shallow Foundations” idea that has been widely used in Scandinavian countries since the 1940’s. John Hait is the most famous for developing and popularizing this idea, I think he also coined the “Umbrella” name. In his book, he talks about originally coming upon the idea as a way to simplify the application of insulation to his geodesic dome home. I have further improved on the concept with my “by-passive” heating idea that avoids overheating the home.
The basic idea is to use a few inches of rigid insulation to form an underground thermal umbrella that further separates the home and the soil around it from the above ground air temperatures. It is also important that the mass of earth be thermally connected to the home its self. When summing up static R-values, the order of the layers is not very important, but by placing a thick layer of earth between the home and the insulation, the thermal inertia and dynamic R value increases dramatically. This is simply taking the basic concepts and components of an earth sheltered home and rearranging them for even better performance.
Soil is a poor insulator compared to modern expanded or extruded polystyrene (EPS or XPS), but it does have some R value (which is why it was used historically and by animals). The R value of soil is proportional to many things, including the distance the energy must travel thru the soil, the moister content, density, soil type, etc. Above the soil, vegetation, and even snow, can add additional insulation. In terms of “dynamic” R value, soil brings low cost thermal inertia to the design.
Soil can also absorb and retain warmth like a capacitor. You know that heat flux thru a medium is inversely proportional to its R value… but Fourier’s law also states that it is proportional to the temperature difference across the medium. Just as pressure pushes flow or voltage pushes current, it is the temperature difference that pushes the heat thru a conductive medium. However, in a material with high thermal capacity, such as soil, some of the heat is absorbed instead of transmitted. As the material heats up locally, there is less and less local temperature difference remaining to drive the heat transfer. In above ground cement walls, this is known as “dynamic R value” and can dramatically improve the thermal comfort of a building in certain situations. Below ground, the much greater mass leads to a much longer cycle and an almost steady state R value improvement. This phenomenon is the same one that leads to the soil temperature profiles we discussed in the Earth Sheltering Basics.
Above Ground Note: Dynamic R value is often “sold” by ICF dealers and proponents. It basically describes the additional thermal inertia (resistance to change) in the wall during temperature swings from one side of room temperature to another. For instance, in the southwest, you may have a very cold night followed by a very hot day. High mass walls cool down during the night and provide dynamic insulation into the hot day as they warm up again. If the following night is cool, the warmed up walls continue to stabilize the home temperature as they cool down. However, this benefit is dramatically reduced in areas that are cold both day and night. In temperate areas like Michigan, air temperatures do fluctuate over any 24 hour period, and dynamic R value can bring some moderation, both daily extremes are typically on one side of the desired internal room temperature. In temperate regions, wide temperature changes are seasonal and the Static R-value and the temperature outside the walls is much more more important for calculating heat loss. On the other hand, an earth sheltered home provides a lot more mass than an ICF wall, enough to moderate seasonal changes. Also, 6 inches of insulation with static R-values of 60 to 100 are typical for an earth sheltered umbrella and the dynamic-R value of the soil on both sides is a bonus.
When you bury a wall, the deeper portions are further from the surface. The heat must travel further to reach the surface, this increases the R-value of the soil. It increases the resistance to conductive heat flow presented by the soil…
Due to the longer path to the surface, the total resistance (R-value) is greater near the bottom of a fully insulated wall. Taking the path of least resistance, heat flux is therefore greater at at the top of the wall. We could reduce the heat flux at the top by adding insulation there, or we could make better use of the same amount of insulation by taking some from the bottom and using it higher up the wall where the heat flux is greater. (Again this is similar to what many smart above ground builders have done for years in terms of concentrating insulation under the perimeter of the slab.)
Re-positioning the insulation further up the wall can balance, and overall reduce, the heat loss without increasing the quantity or cost of materials. It is important to note that, for conduction, The R values are simply added up (commutative and associative properties of addition) so it doesn’t matter what order the R value components come in, or what materials they come from, the total R value is the same regardless of the order of the layers. These two concepts open a world of possibilities for optimizing both the insulation and the thermal mass.
We can increase our thermal mass simply by including some soil under the insulation. Heat energy follows the path of least resistance and will either pass thru the the rigid insulation or take the longer path down and around the insulation to the surface. The parameters of the design (thickness and extent of insulation) can be adjusted to minimize the heat loss for a given quantity (cost) of insulation. One major flaw with this idea is that it could be difficult to install.
Another idea is to take the same quantity of insulation and extended it out horizontally from the edge of the buried roof. The R value of the earth plus insulation can be the same or greater, but the thermal mass is definitely maximized. This can be done after the walls are backfilled so that construction is as straight forward as laying insulation on the ground. Tilting the edges down a little at the perimeter would help with diverting any sub surface water away from the home.
The University of Minnesota wanted to use about the same amount of insulation as a regular home, so they tested this with a 5.6ft extension and showed overall thermal performance, particularly for the summer time. John Hait was trying to insulate a geodesic dome, so he included some soil over the dome to simplify the shape and then extended the “umbrella” 10 feet past the sides. He also saw very good results, but wished that he had extended 20 ft.
One topic often discussed with these earth umbrellas is water penetration. The concern is that cold water (from cold rain or snow melt), with its greater thermal capacitance, would percolate thru the soil, steal much of the absorbed heat from the earth, and dump it into the water table. This is why John Hait used his three layers of plastic sheeting to waterproof his umbrella. I also plan to layer my insulation with waterproofing, but I am less concerned about this theory since conducting a full scale experiment. In my experiment, I buried temperature and moisture sensors at various depths below ground. I had two sites. One was the control site and the other was covered in a few hundred feet of plastic water proofing buried 1 ft below the ground. I expected to show how much heat was lost from the earth due to this percolation. However, watching the data come in over the past year, it is clear that the effect of the rain on the soil temp is only in the first few ft near the surface.
The waterproof umbrella does have several other benefits. Primarily, it helps to reduce the amount of water in the soil around the structure. Any time you build below the earth, you need to be very concerned about water in the soil. Reducing the water content of this soil reduces the lateral pressure on the structure, it reduces the chances of water penetrating the cement and causing other damage, etc. Traditionally, earth sheltered home builders dealt with this water by severely over draining the soil around and above the earth shelter. In many cases, this actually lead to surface conditions that were too dry to support plant life. In contrast, the umbrella serves as a divide between the moist plant bearing soil above and the well drained earth shelter friendly soil below. As a bonus, the dry and almost airtight soil under the umbrella is also unattractive for most digging pests, tree roots, etc.
It may be a good idea to provide a vent in the top of the umbrella to let radon (if any) escape…
The robust umbrella construction recommended by John Hait includes 2 layers of XPS rigid foam insulation (thickness tapers off toward the edges) between 3 layers of 6 mil plastic sheeting (or pond liner in some cases). The XPS foam is recommended because water absorption tests show that it will absorb no more than 0.3 percent (by volume), this is 1/1oth of what the EPS foam will absorb. Water absorption reduces R value.
Note: I have since discovered that EPS may actually be a better plan… For more info, see my page on insulation. A switch to EPS cuts the cost of the umbrella in half, so it is well worth considering.
I suspect that thinner plastic (4 mil or 2 mil) could be used on the lower layers. The layered construction is intended to make it difficult to puncture the umbrella. If one or more layers are punctured, any drips that pass thru tend to run along the underside of the plastic which is sloped towards the perimeter. The primary downside to this plan is that if moisture does get thru, it becomes almost impossible to determine where a leak has happened because it runs along and enters somewhere else.
I discovered that there are many places selling used billboard vinyls. These are 15 to 17 mil thick, waterproof and tough. They are way over designed for their short life on the side of the highway and still have a lot of life in them when they are taken down. Some companies will even glue them together into large 200’x200′ pieces… The prices are great, but shop around a little. I found that the store I linked to in this section had prices about half of what other more generalized building material re-sellers were charging.
The layer of soil below the umbrella is there to facilitate installation (easier fit was one of the reasons the idea came to John Hait in the first place) and to keep the umbrella from rubbing against the cement. It will be important to keep sharp rocks and other “dangers” out of these early layers.
Our early plan was to waterproof with Zypex (or equivalent) and perhaps even a layer of bentonite clay in some areas. However, I am now more likely to go with a painted on rubberized waterproofing… Additionally, many earth sheltered home builders also add a layer of used carpet over the top of the umbrella for further protection. I have not tried it myself, but I imagine that it is actually quite difficult to accidentally put your shovel thru polyester carpet. And the synthetic nature of most carpets ensures that the protective barrier will have a very long life.
Initially, my design also included some cellular concrete (aircrete) on the roof between the vaults to fill in the gaps without adding too much weight. This is not shown in the above image, but would be found below the top layer of cement. Cellular concrete has a pretty decent R-value and can also be formulated in a way that makes it naturally moisture resistant. However, due to the lack of availability of reasonably priced cellular concrete in my area, I briefly considered starting my own company and then decided I had enough on my plate and would just fill the gaps with EPS rigid insulation. Either method could be covered with a thin (2″) topper layer of waterproof cement.
So, in summary, the umbrella is an alternative placement of insulation that allows the many tons of earth around the home to be incorporated as part of the thermal mass of the home. An umbrella extending 20 ft around the home would be similar to an additional R100 (20 ft x R5/ft) against a normal wall. The high thermal capacity of the soil also stores heat and increases resistance to heat flux. While slightly reducing the volume of thermal storage under the umbrella, the edges are sloped down to enhance runoff (water management).