Traditional Roofing Magazine, a division of Joseph Jenkins Inc.

Tools, materials and supplies for slate and traditional roofing. Traditional roofing tools. Soldering bars, wire, irons, flux, and everything you need to solder for roofing or construction. Snowguards and rails for all types of roofs.

Download Entire Issue as a PDF l Download This Article as a PDF

Issue 2, Spring 2002

TR Readers Write:


Apart from the additional cost, are there technical reasons why slate should not be installed over IWS? I understand about the role of felt for temporary covering during installation, but in chronic ice damming climate (where I work) where the usual design solutions cannot be used on old construction, we apply IWS. We do not use it as a "prophylactic" for bad craftsmanship!

B.B., Grand Rapids, MI

Regarding your article in Fall 2001 "Traditional Roofing" entitled "Why Slate Roofs Don't Need Ice and Water Membrane", may I respectfully object? I lived...under a slate roof [that] leaked in several places. We were able to trace the problems to some of the valleys.[S]late roofers look[ed] at the roof, and ...were generally of the opinion that the only way to completely stop the leaks would be to remove and rebuild most of the roof slating, widening the flashing. This proved too costly, and we lived with the leaks. I have kept this experience in mind during my practice as an architect, and have always paid close attention to roofing details. I love slate roofs, and try to talk clients into them whenever possible. However, I have always been concerned that the nature of the slate material creates relatively large open channels between the slates, with the potential to allow water to be driven up under windy storm conditions. This would probably not cause leakage in the field of the roof, but at valleys, hips and ridges, where there is less overlapping coverage, it seems likely that water can be driven above the top course of slate and onto the substrate. There is also the whole concern about ice damming at eaves... Over the years, use of such membranes has virtually eliminated leakage problems with roofs on buildings I've been involved with. [Y]ou seem to be in denial about people's experience with slate roofs. Unfortunately, I have heard too many people object to my proposal to use a slate roof on their house because their experience of slate roofs is that they leak! If I want to put slate roofs on my buildings, I need to be able to assure my clients that they will not leak, because modern materials, like the membranes to which you object, will make such a thing possible. M.W.

My family has been in the slate roofing business since my great grandfather immigrated here from Ireland in the early 1880s. He was trained in England first as an apprentice and then a roofer. My grandfather started our slate roofing company in the 1920s; we have been in operation ever since.

I just spoke to my uncle that is now running our company and discussed with him ... about the use of ice and water membrane under slate and have the following comments. Ice and water membrane under slate as well as other roofing material becomes brittle and can crack. We assume this happens due to the high temperatures slate roofs achieves or the expansion of different materials. Slate as well as roofing shingles stick to the ice and water and can be difficult to replace without breakage. Ask any of your asphalt shinglers what it is like to strip a roof totally covered with ice and water with the shingles installed directly to the ice and water. The same happens to slate and the slate can be destroyed or become unsalvagable.

Some of the places we have observed ice and water membrane failures are at dormer/ roof deck intersections and chimney/ roof deck intersections. The materials expand and contract differently and can cause cracking of the ice & water at these areas. We install ice and water to our felt underlayments on shingle installations and not to the roof deck. So beware! We only use two solutions when it comes to ice damming on a slate roof. First one is a copper ice belt. The second involves installing two layers of 30 pound felt with a layer of slaters mastic in between. We have never had a failure. My uncle spoke to the manufacturer of the failed ice & water system and they told him they now make a new improved high temperature ice & water membrane for slate and metal installations. We don't utilize ice and water under slate and can not vouch for this new ice & water membrane. K.S.

Joe Jenkins replies:

I have never used a square inch of ice and water membrane on a slate roof and never intend to, basing this position on my own personal experiences over the past 34 years since I started working on roofs. You readers can decide for yourselves what to do. Here’s my side of the story:

Having examined slate roofs around the world under many different circumstances, it’s obvious to me that ice and water membrane is not a required element of a slate roof system and is instead a fad peculiar to the United States and only appearing within the past 15 or so years. Frankly, of the thousands of slate roofs I have examined, none have had ice membrane except those installed in the past decade or two in the United States by roofers whose primarily business is to install asphalt shingle roofs on plywood roof decks.

One of my aversions to ice and water membrane is simply that the manufacturing industry is trying to make it a required component of every roof, while it has no real purpose at all with permanent roofs such as slate. Its original roofing purpose was to prevent plywood from delaminating, and properly designed non-plywood roofs don't need it. I have never used ice and water membrane because I have never built a roof out of plywood; I only use board decking when building roofs. I also do not install “asphalt” roofing, which does not make a real roof at all — it’s a temporary covering for people who can’t afford or don’t really care about having a real roof. Plywood decks under temporary roof coverings are all the rage in the United States, where wasteful throw-away construction has become the norm. [Read an article about ice membrane and cancer-causing chemicals.]

In about 1984, a neighbor asked me to build a porch on the front of their 15 year-old ranch-style house. The house roof was 4:12 and the new gabled porch roof was to run on up the front roof of the house, with two valleys. I matched the existing architecture, building a 4:12 porch roof with 1/2 inch plywood and "20 year" fiberglass shingles, as my neighbor desired. This was purely a favor — I do not do this kind of work otherwise (although I have done a lot of building construction) and have not done anything like this since on a contractual basis. When I removed the shingles from the front of the house where I had to tie in the new roof, I found to my surprise that the bottom three feet of the plywood roof deck — 15 years old — was delaminated! It was shortly after that when the building supply houses started insisting on using ice and water membrane along the eaves of roof decks, and it was obvious to me why: to prevent plywood delamination. However, as I do not work on plywood roofs, I do not have the need to prevent plywood delamination, and have never taken a second look at ice and water membrane. Traditional slate roofs are built with board decking, not plywood. Boards don’t delaminate.

Ice build-up along eaves is often an insulation problem. Uninsulated roofs thaw the ice and snow from the field of the roof and the water drains down to the eaves where it freezes again, building up dangerous levels of ice. You can largely prevent this by insulating the roof. You have leakage at the eaves anyway? You think the solution is ice and water membrane? Think of it this way: If the leakage is being caused by water penetration through the slates due to ice back-up, then the water penetration should occur along the entire length of the eaves. This is often not the case. If, in fact, that is what happens, more headlap may likely solve the problem (unless the problem is due to insufficient slope or other design flaw, in which case you’re beating a dead horse).

In any case, water penetrating a slate roof anywhere, for any reason, means the roof was not installed properly or the slates or flashing simply need repaired [or we are having a bad ice-dam year]. When installing a slate roof in an area known for ice damming problems, install the bottom three or four feet of the roof along the eaves with 4 or 5 inches of headlap rather than the typical 3 inches and make sure the roof slope is 8:12 or greater. Architects must learn that slate roofs have certain design considerations that are different from other roofs — sufficient slope being one of them. The object is to create a waterproof roof, regardless of the underlayment. Underlayment material is only expected to have a relatively short life as a "weathering in" material. To depend on it for the proper functioning of a 150 year roof is only to fool yourself (or the home-owner). [Read an article about how to install ice-dam resistant eaves.]

What I have found during times of heavy ice build-up and leakage along eaves on steep slope roofs is the leakage does not occur along the length of the eaves at all. It occurs in one spot or two along the eaves, showing up inside the building in pin-pointed places. When I have examined the roof closely, I have found holes, nails, or cracks at those points on the roof. These minor roof flaws only leak under ice damming situations and once these points are repaired, the problem is solved.

Here’s an ice dam tip: the roof is leaking at the eaves during ice damming, but you can’t find where the water is penetrating. Simply cut copper bib flashings about 6 inches wide and long enough so that when you slide them under the slots (where the roof slates abut at their sides) the top of the bibs hit the slate nails. Curve the bibs lengthwise slightly so they friction fit under the slates. Use a drop of caulk under the bibs if you feel the need for extra adhesion. Install the bibs along the affected area of the eaves. You can do this in a very short time and it will likely stop the leaking. It has never failed for me and is a hell of a lot easier and more permanent than removing the slates and installing new temporary underlayment.

I don’t use any underlayment under valleys of around chimneys when I'm replacing the flashings because I know my work will not leak — and that's 100% guaranteed (although I do use 30 lb. felt when installing new roofs, sometimes half-lapped). If my work leaks one drop, something was done wrong and it needs to be redone. If it isn't going to leak a drop, then underlayment isn’t necessary. One school of thought is that if the roof leaks, then a hefty underlayment will stop the water. I don’t buy this. If any roof I build or repair leaks, I want to know it immediately and find the problem and fix it. I don’t want the problem obscured by a false sense of security so someone else can worry about it five years later. The slates and flashings keep the water out, not the underlayment. When a slate roof is installed properly, or reflashed properly, not a drop of water will penetrate the roof, underlayment or no underlayment. I have had a couple of instances over the decades when a chimney reflashing job I did leaked. It took me a while to figure out the problem, but I did, and once I figured it out, repair was simple and the mistake won’t be made again. A heavy underlayment would have hidden the problem.

I am very grateful when I can learn from my mistakes — these are the most important lessons. I think it would be a good practice for all slaters to have to install a roof without any underlayment at all on a building where the underside of the roof sheathing is exposed, in a rainy part of the world. You will either see that the roof does not leak a drop and thereby gain confidence in your work, or you will see where you’re making your mistakes and thereby improve and perfect your work. In the end, underlayment or none, light rain, heavy downpour, or flood, the roof should not leak.

Regarding the assertion that valleys and flashings on slate roofs leave gaps where wind will, by design, drive water into the roof — that’s totally bogus. The writer who mentioned leaking valleys that could not be effectively repaired was suffering from either roofing contractors who did not know what they were doing, or a faulty roof design (or both). Having replaced literally miles of valleys on slate roofs under all sorts of circumstances and never having a leak, using no underlayment ever, I know that the assertion that valleys will leak without ice and water membrane is ridiculous.

Here is one final issue with ice and water membrane: I was originally attracted to slate roofing systems because they’re natural roofs — roofs of stone and wood, biodegradable roofs, recyclable roofs, ecological roofs, environmentally friendly roofs. They’re simple, low-tech roofing systems that are fabulously successful when properly installed. Adding unnecessary asphalt materials to these roofing systems steers them away from the ecologically friendly roof systems that they now are and forces them toward the petrochemical toxic waste roofing systems so popular in America today. There’s no excuse for that. Architectural and construction planners would do well to place a high priority on choosing building materials and methods that are environmentally friendly.

I have to add that the bottom side of roofing slates will adhere to peel and stick ice membrane once the membrane gets hot. Then, when you have to remove a slate to do a repair, you will have a nasty problem.

Finally, the leading brand of ice membrane states in its Material Safety Data Sheet that the product may be 25% carcinogenic. Read more about cancer-causing underlayment.

[Read an article on how to install a copper snow apron.]


I had 3,000 sq.ft. of Spanish slate (from Galicia region of Spain) shipped to my home in NJ. It is roughly 3/16 inch in thickness. When it arrived, I was surprised to see no holes in the tiles. G.M., NJ

Joe Jenkins replies:

They don't typically "hole" the slate in Spain during quarrying like they do here, as much of the Spanish slate is shipped to Europe where they tend to install the slate with hooks rather than nails, so no nail holes are needed. When slates are produced for the American market, they are holed before shipping (usually). Looks like you got a shipment that wasn't holed. You'll probably have to hole them by hand with a slate hammer. [Read an article about nail holes in roofing slates.]

Buying slates from a foreign source can be a gamble unless you’re familiar with the quarry, the slate mill, and the work ethic of the people who produce the slate. I've heard quite a few horror stories about large, new, slate roofs that are failing after only 10 years (or less) due to “bad” slates coming from foreign sources. Some of these failing jobs have become a huge liability for the installer or for the company that provided the slate — nightmare scenarios that you do not want to experience and don’t need to if you educate youself. [Here is a source list of roofing slate suppliers.]

Not only are slates that are destined for a European market usually not holed, but they can also be very thin, maybe 1/8 inch thick. American slates are always holed and are usually 3/16 inch to 1/4 inch standard minimal thickness. Furthermore, all nail holes in standard thickness roofing slates must be punched to allow for a counter-sinking effect on the front of the slate so the nail head can sit down into the slate. If the slate is drilled, the drilled holes must also be counter sunk. Many foreign slates are being shipped to the U.S. without the counter sinking, and some domestic slates may also be lacking in the same manner. Slates that have nail holes without counter sinking will allow the nail heads to rub against the overlying slates, eventually resulting in leaking holes popping up all over the roof — a disaster.

So anytime you’re buying slates, make sure you’re getting a) slates with nail holes; b) nail holes that are counter sunk; c) 3/16 inch to 1/4 inch minimum thickness; d) slates that are high quality with a proven longevity. Color should never be the first consideration when buying roofing slates; quality of the rock should be the first consideration. You can acquire good quality slates in many colors, but do your homework first. It will be either your money, your home, or your reputation at stake.

That being said, I have to add that some excellent slate comes from Spain, including from the Galicia and Leon regions, and from other foreign sources as well. Take your time and shop around — your slate roof should last at least a century if the right slate is used and it’s properly installed, so why rush into it? Buyer beware.

Download Entire Issue as a PDF

Download This Article as a PDF

We have the largest selection of snow retention systems for roofs available on the internet!

Send us an email!


Find us at Joseph Jenkins Inc.

Slate Roofing on Facebook

Slate Roofing Contractors Association on Facebook


Snow Guard Warehouse
All types and sizes of snowguards!

Mullane Eagle Snowguard


SOLDER WAREHOUSE - Your source for bar and wire solder, soldering irons, flux, and other soldering supplies.

Propane soldering iron sold at the Solder Warehouse.

Solder Bars


Roofing Fasteners

The award-winning book!

Slate Roof Bible, 3rd edition by Joseph Jenkins

America's best source!

Stortz Genuine slate cutter



Find it here.

Find contractors now.


Slate Roof Message Board


Slate Tools

Snow Guards

New Roof Slate

Salvaged Roof Slate

Ceramic Roof Tile

Asbestos Roof Tiles

Architectural Sheet Metal

Want to find something? Search our entire website!

All material herein © 1998-2019 Joseph Jenkins Incorporated, all rights reserved.

Call us at 814-786-9085, M-F, 9-5, Eastern time, USA.