Converting a Tar-and-Gravel roof to metal or membrane

If you have a tar-and-gravel low-slope roof (typical of constructions in many areas in Montreal), you may consider changing the roofing material to something else.  Two common alternatives to tar-and-gravel are two-ply membrane assembly and standing-seam hidden-fastener metal panels.

Tar and Gravel

First, let’s talk about what makes a good tar-and-gravel roof work.  These roofs are also called BUR roofs, standing for Built-Up-Roofing, and this refers to the roof covering being layers of tar paper interleaved with melted tar in between.  The more plies, the better the roof.  5-ply roofs are better than 4-ply roofs, which are better than 3-ply roofs.  Once the roof BUR is completed, it is covered by a protective layer of gravel to shield the tar from the effects of the sun (mainly the UltraViolet rays or UV).  Working with hot tar can be both dangerous and messy, so fewer and fewer companies are doing it, leading to increases in prices for those who still do it.

The key weakness in BUR roofing is the junction between the roof and the various details such as drains, protrusions and parapets.  These junctions have to be waterproof, and the require detailed workmanship seal the joints properly.  If a roof leaks within a few years of installation, the most common reason is the lack of appropriate detail work.


Tar and gravel roof, showing poor waterproofing at the edge of the roof, as well as the migration of gravel from the roof peak, exposing the tar/paper layers to the sun.

The second weakness in BUR roofing is the fact that it is made up of layers of tar paper with tar in between.  There should be no air bubbles trapped between the tar and the layers of tar paper.  These bubbles are squeezed out by the roofers rolling out the tar to force any bubbles out, while the tar is still hot.  If the layers have trapped bubbles, over time these can expand and create openings between the plies.  If such an opening is close to a junction or edge, it can become the entry point for water.


Example of a bubble in the tar-and-gravel roof which has risen (which then allowed the protective gravel to move away), and cracked once the sun acted on the material.

The third weakness of BUR roofing is that over time the UV rays damage the tar and cause it to become oxidized and brittle. Such material does not handle expansion and contraction due to heat/cold, and can crack.  To prevent this, all BUR roofs are covered by a layer of gravel which shields the organic material from the sun’s rays.  However, the gravel can move on the roof (water, snow, wind), and can work its way off the higher points, leaving them exposed.  In time these exposed areas will weather, and become possible leak sources.

So given these issues, why was the material so commonly used?  This comes back to economics – tar was cheap, as was gravel, and especially, labour.  Now tar is more expensive and labour is MUCH more expensive.  Furthermore, fewer and fewer people want to work with molten tar, and the amount of expertise present is diminishing.  Given the occasional fires that have been started by roofing crews, it is now much more difficult (and expensive) to get insurance coverage for such work.

So… what are the alternatives?  There are quite a few, but for residential installation in the Montreal area, there are two that are quite common.

Two-ply Membrane

As described, this system has two layers – a base layer that is usually nailed to the deck, and a granulated top layer that forms the external part of the roof.  The two layers can be joined by heat (this is known as torch-on membrane) or by roofing cement (known as cold membrane).

Torch-on membrane is a very effective solution (when installed correctly), but because it relies on (usually) open flame, insurance companies require a very steep insurance premium to cover this type of work.  Another way that torch-on membrane can be installed is using a heat gun (high heat, but no open flame) which does not require special open-flame insurance.  The flip side is that these types of heat guns are quite expensive.

The cold membrane approach uses a bonding agent which effectively melts the membranes together.  Here the potential weakness is the requirement to roll out the membranes to ensure that no air pocket is trapped between the two plies.

To meet the code requirement for a fire-resistant surface, it is quite common to install a product called Dens-Deck (manufactured by a number of manufacturers), which is a gypsum board (1/2″ or 5/8″) reinforced in various ways and usually 4′ x 4′ or 4′ x 8′ in size.  This board is installed over the existing wood decking to provide a clean, level, fireproof, and stable base for the membranes.

The technical challenge with creating a water-proof surface is assuring that the various penetrations that exist on the roof (eg: roof drain, plumbing vents, air vents, skylights, electrical pole supports), and the roof perimeters (parapet walls, sidewalls, transitions) are fully and properly waterproofed.  This waterproofing has to also withstand the physical stress of thermal expansion and contraction, the heat and ultraviolet exposure from the sun, the thermal shock of cold rain falling on a hot roof, and the occasional foot traffic.  There are a variety of techniques used to assure proper long-term function, but they all rely on careful preparation, proper choice of materials to use, and careful workmanship.  These details can easily represent 50% or more of the labour required to install such a system.


Membrane roof, about 15 years old.  The client was looking to replace due to persistent leaks around various roof penetrations.

These systems can be installed on slopes that are essentially flat up to slopes of 4:12 (lower end of the steep-slope regime).  Esthetically, they are rather plain, and if the roof surface is easily visible, they don’t add much to the “curb-side” appeal.  Furthermore, being an organic (asphalt) product, the material does have a shorter life than metal, as can be seen in the image below:


Membrane showing deterioration after about 20 years of use.  Some patching visible at upper left, due to poor installation at the beginning.

Hidden-fastener Standing Seam metal panels

These are painted steel panels with ribs, designed to be laid vertically, with the ribs from one panel overlapping the ribs from the adjacent panel and hiding the fastening screws below that panel.  With proper technique, this type of panel provides for an essentially water-proof surface down to slopes of about 0.25:12.  The image below is of such a panel installed over a low-slope (1:12) porch.


Hidden-fastener Steel Panel on low-slope roof

An example of one such installation is shown below.  The client wanted a more durable roof than the existing tar-and-gravel, which had various patches and repairs over the years, and was rather tired of the gravel washing off her roof and into her eavestroughs and surrounding ground.  The starting point was below:


Unlike many contractors that would “save” you money by just going over the existing roof, we stripped it to the deck, for five reasons:

  1. repair the portions of the deck that were damaged,
  2. install proper waterproofing (which needs to be bonded to the deck to work properly),
  3. level the deck (the roof was quite uneven),
  4. reduce the static weight on the roof,
  5. ensure that the roof penetrations and perimeter flashings were properly waterproofed and sealed.

Roof stripped to deck, flashings removed, deck cleaned, wood repaired.

After the membrane, strapping and edge supports were installed, we had the following:


Roof deck repaired, protected by peel-and-stick waterproofing membrane, levelled (use of strapping and perimeter plywood), and ready for panel installation.

With the roof properly prepared, we then installed the perimeter flashing, and the various other flashing details that contribute to a durable waterproof roof.  Below is some of the detail work around the chimney:


Each roof penetration requires its own set of specialized flashings and waterproofing techniques.  When we do roof inspections on such roofs, the good ones have spent the proper amount of effort at providing each penetration with two or more levels of protection.  Poor ones rely on caulking or roof cement and fail once those are broken by thermal expansion/contraction and damage by the sun.

After everything is done, the roof looks great and works as it was designed.



This metal roof was installed in 2008.  In 2016 it continues to work as designed, with no issues and no requirement for maintenance.

Of course, as with most things, workmanship is still the difference between seeing good materials properly installed, and having an expensive, poorly-functioning result.

Below is a case of a standing-seam hidden-fastener panel that was leaking despite the fact that the roof itself was steep.


When we started the disassembly process, we saw the following deficiencies:

  1. The material was installed over existing roofing.  No effort was expended to repair or remediate any underlying deck deficiencies.
  2. No waterproofing membrane was installed anywhere.  The installers were relying on the existing roofing to keep the water out, despite the fact that much of it would be penetrated by nails holding the strapping.
  3. The roof perimeter was poorly supported with several 1×3 furring strips (which can flex and move) instead of solid plywood.
  4. The valley flashing itself was made in a rather “cheap” manner with a simple return along the valley edge, instead of a much stronger double-fold and additional leak protection.
  5. The panel/valley joint was “waterproofed” with caulk, which was broken in a number of places by the pressure of the ice against the joint.

The workmanship on the panel joints was poor, with the successive panels not aligned and with the paint scratched during the installation process.


There were a number of other deficiencies on this roof which spoke of a hurried and careless installation.  Unfortunately, the client did not have the money available to remove the roof and have it properly re-installed, and the best we could do for him in this instance was repair the most egregious errors, caulk what we could, and reinforce the weakest areas.

Can I use a screw-through metal panel on a low-slope roof instead of a standing-seam hidden-fastener panel?

These types of panels are usually cheaper and faster to install than the standing-seam panels.  The usual restriction on screw-through metal panels is that they cannot (or should not) be installed on slopes less than 3:12.  The installation technique relies on foam strips, sealing tape and flexible washers on the installation screws to keep the water out of such an assembly.  As well, there is a danger with every screw penetrating the material, that if the screw is overdriven, the metal will buckle, creating a miniature funnel for the water to enter, despite the flexible washers.  Over time, the flexible washers harden and do not seal as well, so from time to time it is necessary to re-install the screws.  If the water has entered via the screw-holes, the supporting strapping is often damaged, and may not hold the new screws as well. Two examples of such roofs that we were asked to evaluate are below:


Skylights installed on a low-slope screw-through panel had leaked a number of times over the years and the owner was looking to replace the whole assembly with something more reliable and durable.



Screw-through panels on a low-slope roof, showing the damage that has occured to the panels from rusting, and the poor condition of the fasteners.  Note as well that the edge of the ribs are open, and are relying on hidden foam closures to keep the water out.  Water backup from the ice in the gutters caused leaks that the foam strips did not stop.

In short, we do NOT recommend the use of such products on a roof slope below that recommended by the manufacturer, and even in those conditions, particular attention needs to be paid to preparing the roof to counter the weaknesses of such a system.

If you have a low slope roof and you are looking to change the roofing material to something more durable and better-looking, give us a call.  The roof will need to be evaluated as a system and then an appropriate solution can be considered.


4 thoughts on “Converting a Tar-and-Gravel roof to metal or membrane

  1. Thank you for the detailed descriptions & picture examples of each scenario. This information has increased my knowledge,& will be useful as I negotiate with roofers to replace the roof on my mother’s home.


    • Hi Laura,
      Thanks for the note. If you do not have the expertise to know what you need done on your mother’s roof, please consider hiring an experienced roofer for consultation. It could cost you anywhere from $500-$2,000 (depending on where you live and the number of people competing in the business), but you will have a much better idea of what you need. I strongly suggest your “consultant” do a full inspection, both on top and from the underside of the roof deck, documenting his observations with photographs. Following this, he should prepare a report for you, identifying (and prioritizing) the areas that are damaged (or have a weakness) and the recommended solution. Ideally, he should also give you a clear document identifying the work that needs to be done, the materials that should be used, and the estimated range of prices for each element of the work. This can then become a “Request for Quotation” document which you can then shop around the contractors in your area to see who can do the work. With this approach, all the quotes will be based on the same work scope, and (hopefully) the same quality of materials. A further step may be to extend the mandate of the consultant to inspect the work-in-progress and the final result to ensure that the winning contractor is adhering to the contract, and is meeting the required standard of work.


  2. I have a metal shed which rusted after 10 years. would the metal roof easy to rust as well? my house is surrounded by trees and leaf always left on my roof which will make my roof not easy to dry


    • With metal, it’s the quality of the primer and paint that will determine how long the finish protects the metal. As well, if any metal has exposed edges of perforations, then those unprotected edges will be where rusting will start. So it’s not the metal itself, but the degree of protection the metal gets that will determine how long it will resist rusting.


Would you like to share your thoughts on the topic?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s