In previous posts, we have talked about the best layout for the house, base on its solar orientation, opportunities for natural ventilation, privacy considerations, engagement of the landscape and relation to the surrounding architectural context. This exploration led to the conclusion that an L-shaped floorplan ‘pointing’ South was the best suited overall.
There are many massing considerations to explore in three-dimensions – an we will start with the roof. A good place to start because the shape of the roof definitely helps determine the architectural character of the building AND because the roof is generally a good place to mount solar panels for generation of electricity and hot water. The roof is high enough to avoid being shaded by the landscape and other buildings (at least in a residential neighborhood), and this space is often unused for other functions (except where green roofs and terraces are part of the program). The panels are also relatively protected from vandalism and theft, which unfortunately can be a consideration even in relatively safe urban neighborhoods. Finally, the panels CAN help shade the roof, reducing the cooling load on the house in the summer (depending on the details of the panel mounting). We shouldn’t rule out some wall-mounted solar panels either, but we’ll get to that later.
In this project, as noted above, the CORNER of the house points south. What form should the roof take? To answer that question, we will consider the follow factors:
- Power and heat generating potential (including selection of collector technologies)
- Structural requirements
- Selection of roofing materials
- Aesthetics of Massing
Based on these three considerations, the first decision we made was that the roof would be composed of a relatively few number of flat planes, not a curvilinear surface or a tessilated solid form. Certainly these are options, but they would complicate the framing of the roof structure and limit the range of options for cladding materials and solar collectors. These forms would tend to drive the cost of construction up and would also be out of character with the neighborhood. Some people really dig Frank Gehry and Steven Holl’s work, but we decided this just wasn’t the place. There was no compelling reason to do it.
On the other hand, we DO wan the form of the building to suggest its function, including it’s ability to harness the rays of the Sun. If the roof form were to telegraph that relationship and still reflect an elegant and practical solution in terms of construction, all the better. After a lot of tinkering with forms, we arrived at the following four basic options. In each of the figures the ARROW points SOUTH.
A variation on the hip roof shape where we have cut the corner to create a large triangular area of roof pointing approximately south.
The cut creates a vertical (or even slightly overhanging) surface on the north side of the cut, which could become a clerestory window for daylighting.
In this scheme, the ridge of the gable runs along the diagonal of the southern bar of the dwelling’s footprint. The orientation of the roof surfaces depends on the aspect ratio of the bar. In this case, it is about 29 degrees off the N-S axis (instead of 45 degrees if the ridge were parallel to to the bar axis).
The orientation of the roof toward the south is clearly visible on the facade of the building
This is basically a gabled roof with a hipped corner. Simple, not fancy. The ridge of the hip does ‘point’ approximately south.
Not much to say here. The roof surface is virtually horizontal, sloped overall or in sections to direct water to downspouts. This form would be particularly well suited to green roof or roof-deck applications.
Solar Power Potential
In order to assess the potential performance of the different roof forms, independent of collector efficiency and panel sizes, we can start by looking at the solar insolation values in the Washington DC area. No, I didn’t just mispell insulation. InSOLation is the amount of solar energy that will fall on a unit area in a given year. It depends on the orientation of the surface (what direction the section of roof faces) and on the tilt of the roof. In theory, given the average weather data for a given location, you could even take into account the amount of cloud cover common for different seasons of year. Such insolation maps exist, but often for very specific locations (not necessarily where you live).
I found the maps below on-line. The first is a 3D map of the insolation values in London (from the SolStat website). The second is a color contour plot showing similar information but using color to represent the ‘height’ dimension. This second map is for Baltimore-Washington area, and was downloaded from Solmetric’s site. Solmetric has on-line maps for sites all over the U.S. and a nifty little app for interpolating the data on the map – you enter the orientation and tilt, and it tells you the isolation value in kWh per square meter (per year).
So, by breaking the roof shapes down into various surfaces and entering corresponding tilt and orientation on Solmetric’s web-app, I got the insolation values below and the potential energy that each roof form can collect in a year (assuming perfect collector efficiency AND that every sq. ft. of the roof is covered in solar panels.
So, looking at these numbers, it appears that Cutting Corners has the highest average isolation efficiency (energy per area covered in solar cells), while the Old Standard has the highest total energy potential. Note that we are assuming that all surfaces on the roof (not actually shaded by other parts of the roof) are fair game for solar collectors. In the “old days”, solar panels were so expensive that no one would ever think of putting them on a roof that was pointing slightly north. As solar panels get ever cheaper, it starts to make sense to put them on less efficient orientations.
In this day and age of computer aided manufacturing, it is possible to fabricate all sorts of non-standard and non-traditional building forms; none of the options cited here are impossible.
Both the Cutting Corners and the Diagonal Gable forms could be created using custom shaped trusses bearign on the exterior walls. Cutting Corners would also lend itself to using a few ‘collector beams’ bearing on a few concentrated points. These will add to the cost and the complexity (risk) of the project, but are solvable if these really are the preferred forms selected by the client.
The Flat and Old Standard forms are quite conventional. The Flat roof option requires deeper members than the Old Standard, but the spans are not extraordinary.
Realistically, aesthetics can only really be evaluated in the context of the rest of the building. Part of the process that got us to these four finalists involved trying these ‘hats’ on different massings for the rest of the house and thinking about how each would drive the form and the performance of the rest of the house. Ideally, the roof form would suggest its role as a solar collector and personalizing the house to its site. On the other hand, the intent here (which is certainly NOT the case with many architectural designs) is to have the house fit into the residential character of the neighborhood as much as possible (and not create the equivalent of a shark protruding from the roof as seen below).
The short list for materials here, based on sustainability considerations, include shingles, standing-seam metal, or something like an EPDM or TPO membrane (tough but generally not toxic). As long as we are not talking about asphalt shingles here, all of these options are viable for rainwater collection and for solar array mounting. We can even consider using vegetated roofs here, depending on the solar collector technology selected.
In the case of the Flat roof, there is really only one suitable material here, and that is a membrane with solar panels mounted to a rack anchored through the membrane. Shingles and standing seam metal panels generally are NOT suitable for low-pitch roof surfaces. Folded-seam metal roofing is an option, albeit an expensive one. The regular rectangular footprint of the house is well suited to the use of conventional solar panels.
The other roof forms cited here could all be done using standing seam metal or shingle roofing. The more challenging issue is the selection of solar collectors. The photovoltaic panel choices generally available today are: conventional rigid solar panels (typically 2′ x 5′ in size or thereabouts), thin-film, and solar shingles. Solar hot water collectors generally come as an array of glass tubes or as flat-plate panels (there are other options that include cylindrical or dish parabolic concentrators, but those are not generally mounted on sloped roofs (they would be fine for the Flat roof options, but would not exactly blend in).
Rigid solar panels are by far the most common type used today. They are relatively inexpensive and the costs are coming down. Their efficiency and durability are both relatively good (about 19%). The challenge in all of the sloped roof forms is the nonrectangular shapes of the individual roof surfaces. Even the Old Standard roof poses a challenge in the south hipped corner. Using panels like these would probably mean a lot of lost area for solar collection, especially in the Cutting Corners scheme.
(Photo by Jeff Gipson – UMD WaterShed House – 2011 U.S. Solar Decathlon)
Thin film collectors come in strips and are generally designed to be mounted on standing seam metal roofs (between the seams). The strips can be cut to different lengths, so it would be relatively easy to cover almost all the roof surfaces with little loss of space. The efficiency and durability of the thin film panels is not as good as for the rigid panel units, but the cost per kWh is not too different. Unfortunately, if we hoping to achieve net-zero energy, we need power, not just value. Ultimately, then, the decision will rest on how efficient we can make the rest of the house (which would also allow us to use rigid panels if we don’t need as many).
(Photo from solarpowerrocks.com website)
Finally, there is the option of using solar shingles. These are essentially just smaller solar panels designed for attachment directly to the roof deck. They also serve the function of regular singles, to protect the roof. These things are really cool, but are not in widespread use. They are comparatively expensive and they do tend to complicate the wiring a bit.
In the case of the curved tiles shown here, we are basically dealing with thin film solar panels factory-applied to the tiles. For flat shingles, it is possible to use small rigid polycrystaline PV cells.
Neither the thin form nor the shingles are very effective at shading the roof to reduce heat loads. They are generally darker in color and so may actually add to the heat load. We might be able to ventilate the roof on the interior, especially if we are including an unconditioned attic space.
Overall, the variation in total solar energy potential between candidate roof forms is only about 7%, and the variations in ‘efficiency’ (average insolation) is about 4%. Not necessarily compelling data for declaring a winner based on solar performance.
As noted above, though, it is too early to pick a winner. We will keep our options open and move forward with the rest of the massing investigation.