Ice Gaurd membrane

The home was built in 1974. A 2nd layer of architectural shingles were installed over the original roof.

I would assume there was no membrane installed for ice gaurd protection.

Would you report on this missing as I would believe it is required on a 4 year old roof.

i would either ask grace (mnfct of ice & water shield) for their recommended practice or a shingle manufacturer

David, If the new roofing was installed over the old roofing, the probability that ice and water shield was installed is close to none.

Most Roofing Manufacturers will knock off 5 years on their Warranty for doing it.
Since there are no Manufactures instructions for installing a double layer of roof shingle on the packages, I would have to note that it was double layered and does not comply to a new Manufacturers installation procedure.

Marcel :slight_smile: :slight_smile:

Shingle Manufacturers and Building Codes recognize “re-roofing”, defined as, adding a second layer of shingles over 1 previous existing layer, as a legitimate and warrantee-able installation technique and manufacturers indeed DO illustrate application instructions for doing just that on their product packages and product installation data sheets.

You can see Tamko’s re-roofing practices here, for example (see docment page 3 item 8 ‘re-roofing’):

Ice and water shield, on the other hand, is only effective and required for new roofing applications.

It might be all swell and dandy to be able to recognize that Code will allow two layers of roof shingles,
But, not all Manufactures will Warranty their products in this install.
They might allow the installation, but what they don’t tell you are the ramifications of the double layered roof.

Including “Tamko”, If you read the Warranty, you will find that any defect in or failure of materials used as a roof base over which the Shingles are applied, or for damage by , will not be Warrantied.

So, it is your choice of accepting the fact that Code will allow it or face the music when the shingles don’t last the 300 months the Manufacture says it is Warrantied for.

I for one as a Contractor, would never condone installing New Shingles over another layer, whether or not allowed by Code.

I have personally experienced roofing failures 10 years below the projected life expectancy of shingles too many times.


"I have personally experienced roofing failures 10 years below the projected life expectancy of shingles too many times. "

Have the roof failures been from “roofing over existing” or from poor shingle product. I’ve seen much of the the latter since 1993-5 and it has been written about a few times. When I was roofing in the 1980’s, a light (frost white, light gray) coloured shingle (Domtar 3-in-1 Truseal brand) with a 10 year warranty would last 23-25+ years if installed correctly. Just last spring I saw failure on a 13 year old bungalow!!! As well, saw similar failure on a 15 year old $500,000 home with a light brown roof the year before.

Note: We are building inspectors and, as such, should call items by their proper names. In Canada, the membrane applied to the lower section of the roof (only roofs with up to 8/12 pitch or slope require it here) for protection from water backup from ice dams is called “eave protection”. What term is used in the US?

Hope this helps.

Marcel :slight_smile:

Sorry to disagree with Marcel, but the manufacturers of shingles offer FULL warranties to second layers installed over existing roofs just as fully as they warrant the shingles if placed on a new roof deck.

A proper installation is required for BOTH practices to be warranted.

I would also disagree completely with Marcel’s anecdotal evidence that new shingles placed over existing shingles fail at a greater rate than those placed on new decking.

Such allegations are broadly bandied about, yet is there ever any solid verifiable evidence to back such claims.

I have been installing shingles over existing roofs for 30+ years and have NEVER seen premature failure of the 2nd layer based on the fact that they were installed over an existing layer.

Neither, apparantly, have shingle manufacturers, nor the Codes, or they would not wholly approve of the practice…despite Marcel’s false claims to the contrary.

Adding a second layer of shingles over and existing layer remains a manufacturer and Code approved method of installation…even in Marcel’s state.

Marcel :frowning:

I thought this addittional information might help you understand the facts.

The ROOF-OVER Professional consulting available.

"ROOF-OVER", type roofing: the poor man’s roof. “Roofing-Over” is the practice of installing, nailing or placing new roofing directly over old roofing material, to avoid the rip-off and the disposal of the old roofing.

For the customer willing to trade approximately a 10% cost savings for a roof with rough appearance, a potentially short roof life, the cost of additional venting and a doubly difficult tear off and disposal in the future; “roofing-over” is a manufacturer approved practice. This applies to “soft” roofing only, (asphalt shingles, roll roofing, bitumen, and modified bitumen “flat” type roofing). This process is most often done on sloped (shingle) roofs, but it is also common on “flat” type roofing. An increasingly common practice, this procedure became popular in the early 1970’s in the USA. With many such “roof-over” applications in place, there is now a growing body of experience and information centered on this abbreviated approach to “re-roofing”.

An increasingly common practice, this procedure became popular in the early 1970’s in the USA. [size=3]
It is generally assumed that only an aging roof in sound condition should be “roofed-over”. As long as the old roof is not literally disintegrating, some consider it a candidate. Extreme curling and cupping of the shingles in the case of “flat” roofing, deep material or very uneven materials makes “roofing-over” difficult but not impossible. The first step in “roofing-over” is to remove excess damaged parts, so as to “level” the roof. On sloped roofs, the top peaks or ridges have their “cap” shingles removed, as do any hip ridges (the top of the downward sloping intersection of adjacent “sloped” roof planes). Areas that are not severely damaged may require some simple patching with roofing material and roof cement. At this point, the installation may begin.

On sloped roofs, the actual installation of shingles begins with trimming. A starter shingle cut lengthwise down the middle leaving a 5 inch wide strip which is nailed through and over the bottommost row of shingles; followed by a full shingle with its top edge butted against the lower edge of the old row of shingles above it. After this, the shingles are nailed over the rest of the roof in the same manner until the old roof is covered. At the top, or the ridge, again new shingles are cut, or trimmed, lengthwise to run even with the ridge and a new set of “cap” shingles are cut and installed to cover the peak.

The finished “roof-over” more often than not, will have a rough appearance. This is a reflection of the physical condition of the underlying deteriorated roofing.

Of the many factors that affect the service life and durability of a “roof-over”, is the length of the nail, or other type fastener, which is used. Depending on the quality, weight and type of shingles, or roofing material, used on the original roof, the mechanical fasteners for the “roof- over” (nails or screws for some “flat” roofs) should be one half of an inch to three-fourths of an inch longer than the fasteners in the original roof. This means a one and one half of an inch to two inch roofing fastener is generally used on a “roof-over”. Since, in the case of sloped roofs, the original shingle provides little support for the portion of the nail passing through it, these longer new nails can suffer lateral, downward “pull”. “Nail-pops” caused by expansion and contraction of the underlying material can pull a nail head through the new material thereby loosening it, or push a nail head through the material above causing a hole in the shingle (a guaranteed leak).

Another factor is the continuing deterioration of the original roof. Three basic and interacting multi-factor processes are introduced to the new combination old roof/new roof lamination termed collectively as the “roof-over”.

Asphalt shingles, asphalt “flat” roofing, and modified bitumen products are manufactured from an amalgamation (a mix) of petroleum products ranging from tar-like solids and semisolids to liquids (oil-like) and volatiles (solvent-like). These substances diminish over time by evaporation, solidification (embritlement) and by a photochemical process whereby direct exposure to sun breaks down the asphaltic compounds into carbon gas (carbon dioxide, carbon monoxide, methane, etc.) and carbon solids (black, dirty stuff). Once covered, the old shingles, or “flat” roofing material continues to break down by solidification (congealing) and evaporation of the volatiles.

When the roof (of the “roof-over”) needs to be replaced, all layers of roofing must be removed and hauled to a dump. [size=3]
Trapped between the new shingles, or “flat” roofing material, and the underlying roof felt and/or roof deck (usually wood or plywood), the old shingles gather and hold moisture. During the hottest days with direct sunlight, this moisture steams off cooking the new shingles, or asphaltic “flat” roofing. In freezing weather the moisture turns to ice, swelling and tearing or breaking up the old underlying material. These combinations of actions caused by gathering and releasing moisture also involve expansion and contraction of the old roofing.

So, with the continued deterioration of the underlying the old roofing material by 1) material breakdown 2) steaming, and 3) expansion-contraction; the new roofing experiences added stress from heat, mechanical movement (dimensional change from expansion-contraction) and loss of support by deterioration/degradation of the decaying roofing material.

In the unusual case where the new asphalt shingles are installed over a slate, wood shingle, asbestos- cement shingle, or synthetic shingle, life expectancy, performance and secondary affects (side effects) are unpredictable. Such installation and their related problems are beyond the scope of this article, however, needless to say, such practices are not recommended and are rife with negative considerations.

An odd variation on the “roof-over” is the installation of plywood directly over the old shingles followed by roofing over the plywood. There is too little experience with this approach to provide knowledgeable information. It is known that moisture is trapped in the old shingles. This suggests the possibility that some of the types of moisture related problems, mentioned earlier, are likely to occur.

If you are buying a building that has a “roof-over”, the primary consideration is that when the roof (of the “roof-over”) is worn out and needs to be replaced, all layers of roofing must be removed and hauled to a dump site or disposal point. The expense saved on the “roof- over” is more than exceeded. In other words a “roof-over” transfers the cost of trash disposal to the next “re-roofing” operation. “Three-roofing” or roofing over the “roof-over” has been done and even appears in some literature; however, such practice should be relegated to cartoons and comic books. The results of actual triple roofing practice range from deceptively acceptable to obviously and immediately defective.

There are a number of special considerations when doing or taking possession of a “roof-over” roof. Because of the moisture trapped in the old roofing, and the additional heat trapped in the extra mass, a “roof-over” roof requires additional ventilation --an added expense-- to help reduce accelerated and premature aging of the new roofing. (Which includes bubbles, and lumping in “flat” type roofs.)

Another consideration is the added load (weight) of the combination of the old and new roofing. The added weight may necessitate additional structural roof bracing. (In one instance of the author’s experience involving a one-story cottage, the added weight actually snapped the horizontal cross-ties, and ceiling joists, “flattened” out the roof, bowed the front and rear exterior walls literally ripping the exterior walls away from the interior partitions – virtually destroying the cottage. In another case an 1880 vintage sloped roof (114 years old at the time) suffered lengthwise splits in its (unsupported and unbraced with no cross ties, which were not needed in the original construction) rafters. One more heavy snow load on this roof, and collapse was likely.

With “flat” type roofing, an older roof generally provides a poor surface for the new installation, especially if the newer roof requires total adhesion. Usually when roofing over a “flat” roof, a non-adhesive mechanically fastened (nails of screws) approach is far more likely to provide adequate performance.

With both “flat” roofing and sloped, shingled roofs, “roofing-over” works best with considerable additional venting of the roof cavity and/or membrane type vents for “flat” roofs. This additional venting, when done, often offsets the shortsighted financial savings in forgoing the rip-off and dump costs.

There is only one incentive for doing a “roof-over”, to save money, in the short run.

[size=3]Is that what your Codes are for?


Marcel, I agree with you and after 22 years as a contractor I will very rarely put a second layer of shingle on a roof. The roofs I sell would be a complete strip job so the proper installation of ice and water shield can be achieved as well as a good roof installation.

Beside, I don’t care what the MFG puts on the package for a warranty or their installation instruction, listen to what they tell you when there is a problem, they have a list a mile long of excuses of why they won’t warranty the shingles… been there, done that!!!


"Ventilation plays a big role on the life of a roofing shingle, because it overheats and basically cooks and you loose a lot of life."

Research has shown over the last 20 years or more that ventilation has less and less (almost nothing) to do about shingle life. Shingle colour, shingle quality or shingle orientation are the big items!!! See research by Bill Rose , research Architect at University of Illinois.

Also adding ventilation to solve attic condensation problems is the last thing you should do!! Believe it or not???

See This is from Canada’s National Housing Agency. Don’t be too :shock: :shock: :shock: ed at what is said. It fills in where the old building science ln roof ventilation left gaps. I know! When I gave a 1 night seminar to the local architect’s association, (I was teaching building science at the local architectural college then) more than a few came up to mention how their understanding was much better now. To my surprise, I was asked to do another night within 2 weeks, (usually they have monthly meetings only and none in December) just before Xmas.

I totally agree with you. And so does the largest manufacture in North America….GAF. As part of a Master Elite Roofing Contractor our company never once put a second layer of roofing on. While the manufacture says it will warrant the shingle, they will find a way to dismiss the warranty, easily. GAF’s official position with its contractor program was it would not issue the contractor a warranty for the practice of roof over’s for its contractor program, but would offer an extended warranty if the roof was striped, and was totally looked upon as a big no no. Companies that practiced roof over’s were simply not invited back into the program and would lose the status of Master Elite Contractor. While GAF says they will warrant a roof over on the package they strongly discourage the practice. Why, because GAF knows that for a contractor to maintain a great reputation within the community, they cannot practice the poor man’s tactics. They know that a roof will not last as long as a roof that has been striped.
I ask about this same subject at a GAF seminar and was told in layman’s terms: When a roof is installed directly over another roof of decaying nature, the old brittle shingles draw the oils out of the asphalt of the new shingles laid above them, thus drastically reducing the life of the new shingles. Not to mention the old layer has to be totally nailed flat for the new shingles to lie flat or they will void the warranty on this point alone. The old shingles need to have sealant installed under all of the tabs of the old shingles were the sealant is no longer properly sealing the shingle tab or the warranty is void. The old shingles need to be totally dry before any application of the new shingles or this will also void the warranty. Furthermore, most roofing systems are not designed to handle the additional weight of two layers of roofing and can cause structural failure, especially if it is in an area with high annual snow falls or if the roof sheeting and framing do not meet the manufactures recommend requirements.
As far as roof ventilation is concerned, again GAF’s official position is they will void the roof warranty of a roof system that does not have adequate ventilation. A balanced ventilation system is absolutely critical to the life of the roofing system, both for winter and summer. I read the article that Brian was referring to and it didn’t say that ventilation wasn’t necessary. It simply stated that adding additional ventilation could be bad and I think they were referring to an unbalance system, which is true for extreme winter only.
Personally I have inspected thousands of roofs and one thing always remained constant with roofs that had two layers or more and a crappy ventilation system. They simply wore out prematurely.

Peter, Brian, and Robert, Thank you for reinforceing my point, I totally agree.

Marcel:) :smiley:

Robert Cramer:

"I read the article that Brian was referring to…"

This is just the beginning of articles I will supply on these issues that I have been fighting since 1983. The first partial one was to Roy Cooke about a weeek ago concerning the origins of the 1:300 attic venting rule…it has no scientifically proven basis yet it is in codes everywhere. It was derived from a single lab experiment in 1947-8 at Penn State University on vapour barriers in flat roof assemblies. They weren’t even testing for attic venting but derived a number due to having an incomplete understanding of how these systems truly worked back then.

If we really knew the building science of moisture movement in houses then (98-99% is air movement; the rest is vapour diffusion), vapour barriers would not be installed in homes but they would be built more airtight (as is happening now). If you put in a loose, unsealed vapour barrier, you are attacking 1-2% of the problem and the other 98% defeats you at every receptacle/switch box and unsealed joint!! Make the drywall airtight and leave out the vapour barrier and now you have hit the real problem- air leakage. Same as with attic moisture from condensation- have an airtight ceiling so that house moisture does not migrate by upward air leakage and the need for venting goes way down to about what’s naturally left in the attic anyways.

Joe Lstiburek (Phd., P. Eng.) of Building Science Corporation ( promotes a basement wall system that has no vapour barrier on the warm side of the wall behind the drywall.


“A balanced ventilation system is absolutely critical to the life of the roofing system, both for winter and summer”

Just because these companies sell roofing materials, does not mean they have a fully up-to-date technical staff that has re-trained in building science!!! I have found these types are some of the hardest to talk to- they are so engrained in the old stuff they cannot believe that it was incomplete. Bill Rose had a Q&A spot in The Journal of light Construction about 10 years ago, where he slams manufacturers about using every excuse to avoid warranty claims when the real reasons are shingle quality. The short answer also talked about shingle temps in vented cathedral ceilings- yes, the shingles are cooler at the lower sections of the roof but near the top, they are hotter than if installed in a regular, unvented roof. You can always draw all kinds of arrows on diagrams but…will the air always go there or develop the velocity/volume needed for the hoped for effect- cooling? NO. The inspection field needs a lot of training in building science; all is not as it appears.

I too have read all of the posts here, in addition to every ventilation technical paper I could find over the past several years of research on this subject.

There is a significant study to support the claim that shingle color has much more to do with shingle life longevity, but the fact of the matter remains, that the manufacturers REQUIRE proper BALANCED ventilation system between the fresh air intake and the attic exhaust. The ideal scenario in most cases would be weighed in favor of a 60 % intake to 40 % exhaust ratio.

If at least the proper true 50/50 balance can not be adequately achieved, either a complete “air-tight” vapor barrier must be installed on the warm side of the attic floor insulation.

If these scenarios can not be achieved, then you must DOUBLE the amount of attic exhaust ventilation to a 1 square foot per every 150 square feet of attic floor space ratio versus the standard 1/300 ratio.

Whether some experts defy or denounce the warranty tactics used by the shingle manufacturers, it remains that those are their warranty conditions which must be met.

For those that “build or roof to CODE”, I personally feel that they are only doing the minimal amount necessary to meet the least minimum standard regulated, which is all to frequently not even inspected and verified in the first place.

Per GAF, Alcoa, and Air Vent; Each has stated in writing, that over 90 % to 95 % of all residential shingle roofs do not meet the manufacturers specifications for warranty approval. Unfortunately, there are no mandated manufacturers inspections required to pass along a useless piece of warrany information along to the homeowner.

I have been roofing for 29 years, and have owned my own roofing company for 23 of those years and real life experience has shown me that any reroof or roof over layering will reduce the useful and functional life of the newly installed roof system to between 12-15 years before becoming desirable to replace once again.

The actual 1/300 and 1/150 ratios come from a HUD archive document originally dated 1942, which I have a copy of.

The HVI and Air Vent, “Principles of Attic Ventilation” guidelines document very well, that a more reasonable ventilation calculation would be best though of as "air exchanges per hour instead of total NFVA intake and exhaust.

Edit add on:
Marcel made some excellent points regarding the mishaps and potential hazards regarding a re-roof over an existing layer(s) in post # 9, but to add to that would also be the pockets of hot trapped air, called heat sumps, contained in this encapsulated area, which further promote the premature deterioration of the newly installed roofing materials. Additionally, why is the first roof being roofed over in the first place? Because it is worn out! Does it then not follow suit, that the crumbling substrate will be less than desirable of a surface to install a new roof on?

Ed the Roofer

Ed the roofer;

I like it, good job.

Marcel :slight_smile:

Hey Ed,

I like it too.

Bobby the Roofer

You can post what ever you want out of a text book, but my 22 years experince in NE has taught me 1, always strip a roof and start from scratch ( even metal roof ) and 2 Always vent a roof.