How To Keep Soil on Your Roof: Living Roof Edge Detail

by ziggy on October 18, 2013 -- 12 comments -- Follow

Living Roof Edge Board

Our living roof edge detail getting installed

The hardest part of building a living roof, other than figuring out how to get all that monstrously heavy soil or sod up there, is designing a good edge detail that will contain the soil at the eaves and gable ends. A good living roof edge detail should contain the soil on the roof, while simultaneously allowing the excess rainwater to drain off, all without puncturing your liner, and allowing for some kind of gutter system. And it should look decent. That’s a lot of design details to weigh, but I think we developed a good set of plans. Check it out.

Plans for a Good Living Roof Edge Detail

The main components for our edge detail are the following:

  • metal flashing, 8″ wide with a 1.25″ bend
  • 5″ corner braces
  • 5.5″ wide white oak boards (1″ thick)
  • roll of galvanized steel flashing
  • 1″ zinc-coated SPAX screws
  • 3/4″ screws of your choosing
  • 1″ galvanized roofing nails
  • gravel

As I mentioned above, you need to allow excess rainwater to easily leave your roof, all while holding the soil back at the edge. What we devised was a soil retaining board, in this case a 1″ thick white oak board, wrapped in steel, fastened to the roof deck about 1/2″ off the surface, with a gravel dam on the soil side to filter water. The retaining board is fastened in such a way that there are no punctures in the liner, and water should find its way into a gutter easily.

Note: the theory here is that the retaining board does not actually do that much work. Gravity does the overwhelming majority of work, holding the soil in position on the roof — the retaining board mostly keeps stuff from falling off at the very edge. In no way is this edge board responsible for holding all of the weight of the soil on the entire roof surface. (And for the record: the roof on Gobcobatron has little (if any) edge board to speak of, and the sod stays put quite well, thanks to gravity alone.)

Here’s a stunning technical drawing that I made up in Photoshop to demonstrate the different layers and components:

Living Roof Edge Detail Plans

A quick technical drawing, color-coded for your convenience

And here’s another look at the real life version:

Living Roof Edge Detail Design

The edge detail getting installed, with all elements visible.. except the soil.

And here’s a bit more of an explanation of the details:

  1. The 5.5″ inch wide white oak board (white oak is highly rot-resistant, hence our choice for that particular wood) is wrapped in metal. We used a roll of 8″ wide galvanized steel flashing (it’s more durable than aluminum). You could get pre-bent trim coil for the purpose, but we went the hard way. (A metal break would have went a long way here, too.) We fastened the metal with 1″ galvanized roofing nails (had to pre-drill for every one), spaced 3″ apart.
  2. We bought a pile of 5″ corner braces, and installed those to the now-wrapped oak board every 24″. We used a 1/2″ shim between the board and the corner braces to ensure there would be a gap. This gap is how water leaves the roof surface once the assembly is installed on the roof. We chose zinc-coated SPAX screws to install the corner braces – they look neat, and drive cleanly.
  3. We cut the EPDM pond liner along the edge of our roof decking with a pair of scissors.
  4. Next, we folded back the pond liner, and installed the 8″ wide flashing (with 1.25″ bend). We used 3/4″ screws to secure the metal to the decking. We positioned the metal so the 1.25″ bend was about an 1″ (or finger’s width) from the surface of the fascia. The gutter should tuck right behind the metal, so water can run right off the flashing, into the gutter, and into a cistern.
  5. We lifted the edge board into position, set back a couple inches from the edge of the decking, and screwed it into position.
  6. We tucked the pond liner over the corner braces, and under the 1/2 gap between decking and retaining board, and cut out the EPDM that interfered with the corner braces.
  7. Finally, we spread a layer of gravel against the retaining board to act as a filtration layer. (Some of the smaller gravel will fall through the 1/2″ gap, but that’s fine.)
Green Roof Edge Board

The edge board, wrapped in steel flashing, awaiting braces and installation

Make sense? I hope so. The goal was to, at no point, allow water to go anywhere but straight down, into the gutter, without making contact with wood. The gravel should help keep soil out of the gutter, as well.

Living Roof Termination Detail

Finished living roof edge detail, awaiting a gutter

As for the gable ends… those are a bit easier. We used the following items to devise an edge:

  • 8.5″ wide white oak boards, with a 30 degree taper cut at the bottom edge
  • 5″ wide flashing, with a 1″ bend
  • 3″ screws
  • 3/4″ screws

In this case, the steps were fewer. We cut an angle at either end of the board to match the pitch of our roof. A 30 degree rip cut along the bottom edge adds a nice bit of flair to the design, as well as a little custom cut at the bottom ends (see image below to understand better).

Gable End - Living Roof Design Details

Note the decorative cut at the end, and the 30 degree rip cut along the bottom of the board.

We put a chalk line about 2″ from the top of the rafter, and installed the white oak board with 3″ screws, being sure that we didn’t pinch the EPDM pond liner between board and rafter. Next, we cut the pond liner, flashing it up the board, but not over. Finally, we used the 5″ wide flashing to trap the liner against the wood, and used 3/4″ screws to keep the metal snug against the wood. Very simple, and the metal + EPDM creates a double barrier against moisture here.

At the top of the gable, we took the opportunity to add a nice bit of extra flair, too.

Green Roof Design Details

The peak at the gable end features a decorative touch… which also helps hide the seam between different boards

And there you have it! That concludes this living roof edge detail design…. something that has consumed me for a long, long time. Information and specific details are very scant out there. This page has a few termination details of its own, but otherwise, there is little detail out there.

I think this is a fairly slick solution.


Send Me More Updates Like This!

  • c-dub

    That’s a dangerous detail. Under a heavy rain, the growing media is going to get very heavy — those corner braces could easily straighten.

  • The theory here is that the retaining board does not actually do that much work. Gravity does the majority of holding the soil in position on the roof — the retaining board mostly keeps stuff from falling off. This I think is true, especially when the soil is rooted.

    The roof on Gobcobatron has little (if any) edge board to speak of, and the sod stays put quite well.

  • Donna

    I like the idea of a living roof. My question is : How do you compensate for the weight on the Rafters? Are they closer together than the normal 16″ or 24″ inches or do you double up on them? Or do you use a lighter material for the dirt such as a pearlite mixture?
    Looking forward to your reply.


  • I’ve not heard of anyone mixing vermiculite or perlite in their soil, but there are different methods for creating a growing medium, whether it is some depth of topsoil, sod, or basically “mulching” the roof with piles of dead leaves or straw bales. In any of these cases, you will have to engineer for the load — the maximum load, too, not just dry soil, but fully saturated soil, and soil with snow (if that is a concern in your region.)

    Every house plan is different, but some will compensate with larger rafters, or more closely spaced rafters, or both. It depends on how much soil you plan on putting up there. Our frame is designed to support a relatively shallow soil depth, about 4-5″. Our rafters are very substantial oak 3x10s, spaced two feet apart, and decked with full 1″ material. It’s a very heavy frame, and loads will not be a problem.

    Some of the info. from the American Wood Council here may be a good starting point to determine different load capacities:

  • Jim

    Is the water coming off the roof fine for rainwater collection? I had been planning on a metal roof for my cabin because we want to harvest the rainwater. Maybe I need to consider a living roof instead?

  • If you want to harvest as much rainwater as possible, metal is still the way to go. You simply won’t collect as much with a living roof — the soil will end up absorbing a lot of the rainfall. They may look a lot better than metal, but for practical rainwater catchment, metal wins out.

  • Jim

    Metal is is, then. Thanks for the helpful info!

    Hmmm… Maybe I’ll compromise and put the living roof on the remote artist studio that will eventually be built down the hill. 🙂

  • c-dub

    I largely agree with your points about the roof edge and brackets, but I think a few caveats are in order. Much depends on the slope of the roof and the particulars of the assembly. Once the root structure is developed, you’re right: even a steep roof should be stable. Until that happens, though, a hard or sustained rain could lead to a catastrophe.

    I helped repair a green roof a few years ago that was originally built with an identical edge detail. A few months after the roof was planted, a hard rain sent a 12′ x 12′ section of roof sliding onto the ground: the angle brackets flattened and more than two tons of material came off the roof in a moment. Luckily, no one was hurt, but it could have killed someone.

    For anything but the flattest roofs, either include some sort of integral erosion control in the growing medium, or build a full, continuous curb.

  • Yikes, that is dramatic. I wonder what you mean by “full, continuous curb”? What does that look like to you? Curious.

    Before we put soil on our roof, we devised a sort of “retaining ladder” system… basically, 2x4s lashed together with rope (no nails or screws over the liner!), draped over the peak. These helped contain soil at every 3 foot intervals. I think those have definitely helped the erosion situation.

  • c-dub

    By a full, continuous curb, I mean something like a stack of 2x’s laid on the flat, screwed into the roof framing below. The growing media would probably surmount the curb in a really catastrophic erosion, but it would still impede the movement and reduce the damage. Even a 2×12 on edge, screwed into the fascia blocking, would be preferable to the brackets; they are just too easily bent.

    Now that you’ve told me about your erosion control “ladder,” though, I’m less concerned. I’ve included similar structures in steep roofs, and I think they work well. As the roots establish, the ladder just rots away.

    I’ve been working with green roofs for years, and I think they’re brilliant. But caution is critical, because they can be dangerous if not installed correctly.

  • While we’re on the subject, I’d love to have the opportunity to discuss a few other points about living roofs. The issue I’m interested in is insulation. What can you say about how well living roofs insulate against heat and cold?

    It seems to me that they are less effective in cold seasons/climates, but there must be a reason they were traditionally popular in Scandinavia… or was it just the availability of the materials?

    Do living roofs, with their grasses and plants help to reflect heat gain? Are there numbers on the thermal properties of living roofs?

    Honestly, I sometimes don’t know why to recommend living roofs to folks, other than their aesthetic value, and potential longevity (if the pond liner is well cared for, at least, and not exposed). I like to think they help (thermally) in the summer, but I have no numbers to offer. Would love to hear your thoughts on this, based on your experience!

    (Very glad you chimed in here, too!)

  • c-dub

    It’s very difficult to make broad generalizations about the insulating properties of green roofs because they’re so variable: their performance is dependent on factors like the thickness of the medium, the amount of water in the system, the species and even the health of the plants. It’s an unfortunate result of working with a living system, since the lack of testable standards has inhibited their acceptance.

    Generally, though, you’re right: the bigger thermal benefit occurs during the summer, when the plants shade and cool the roof via evapotranspiration. This I’d particularly true on large commercial roofs with roof-mounted air handlers: on a conventional black roof, the air handlers will draw in air that’s as much as 90 degrees hotter than the ambient temperature; that air is then cooled to about 55 degrees for use inside the building. A green roof, though, is only a few degrees warmer than the ambient temperature, so the energy savings are very significant.

    There’s a small benefit in the winter too, just because of the additional R-value in the overburden, but a roof that freezes wet won’t insulate as well as one that freezes dry. Perhaps the most important thermal consideration is to insulate adequately between the building interior and the green roof assembly: winter heat loss can bring plants out of dormancy and leave them vulnerable to frost kill.

    Green roofs have perhaps the largest range of benefits of any sustainable building technology. They’re beneficial in a broad range of climates, and the reasons for using them largely depends on local conditions. In addition to the cooling benefits, they also extend the life of the waterproofing membrane, improve building acoustics, and allow the otherwise unused roofscape to be used for habitat, agriculture, or recreation. Here in Minnesota, their use is largely driven by water quality: green roofs mimic the natural landscape’s ability to clean and cool stormwater and mitigate the runoff “bounce” that happens after big storms.

    But the fact that they’re beautiful is enough for me.

Previous post:

Next post: