I read in some places that this might cause the roof deck to rot if water infiltrates under the slate roof and is trapped by the closed cell insulation.

I have read that, too, but always as an unproven theoretical concern. Building Science Corp., a leading independent building science research group, however, advocates unvented attics and the use of ccspf. Here's how one of their experts answers the question in part:

"Closed-cell spray foam has negligible water permeability, minimal water absorption, and excellent adhesion allowing it to act as a secondary rainwater barrier to limit damage when primary roof assembly rainwater control membranes leak. Rainwater migration is severely limited due to the low water transmission and high adhesion ('waterproofing') characteristics of the foam and damage is limited to the area immediately adjacent the hole in the primary rainwater control membrane. This tends to contain the damage, making it easier to identify the source and preventing it from spreading throughout the assembly and to interior finishes which can be costly to repair."

A number of builders at Journal of Light Construction forums have used this method successfully too.

Nevertheless, if you have doubts, there is absolutely nothing wrong with a properly vented attic. Once you have resolved the k&t issue, seal the attic floor and use a high-performing insulation such as cellulose or foam or fg capped with cellulose; and be sure there is adequate soffit to ridge or roof vent flow. You--well, someone--will get another 100 years of use out of that slate roof.

Here is a link that might be useful: Building Science Corp.: Unvented Attics for all Climates

Thank you, worthy, so helpful!
Lynn

Do I understand correctly that someone has proposed spraying closed cell foam on the underside of your slate roof to insulate it and create a so-called "cold roof"?

That sounds like a terrible solution to me!

Terrible because it goes against the way slate roofs actually work.

Terrible because it will cause damage to the slates themselves over time because slates must be fairly loose to succesfully withstand extreme temperature changes over the course of a year.

Terrible because it will make any subsequent repairs very difficult, if not impossible.

Terrible because the foam expansion could force the slates, which now lie flat against each other, apart actually creating places for water to enter where there are none now. Slates are not sealed or nailed down tightly to the roof beyond the two nails from which each hangs; their collective weight is what keeps them in place but an expanding force from underneath, like sprayed foam, would disrupt that.

Terrible because it won't really solve the problem, anyway.

Slate roofs are inherently a loose and essentially free hanging roofing material. The foam would inevitably interfere that and would result, I'm pretty sure, in considerable eventual damage when the slates were denied that movement.

Slate roofs are completely different from other roof materials in this way. The field of the underside of a slate roof is not intended to be hermetically sealed against water penetration (except around protrusions, valleys, edges, etc., where flashing is used to make a water tight seal.) but instead relies on redundant, overlapping (but loose) slates and the never-failing force of gravity to channel water away safely. That's why the ice and watershield product isn't really a recommended thing for slate. Not detrimental, just not needed.

A proper slate roof is laid over builders' paper above 5/4 wooden sheathing, but the builders' paper is not there to make the roof watertight, except during the brief period the roof is being first installed. (And used again if large areas must be later opened up all at once due to casualty or structural changes.)

Nor is the wooden sheathing intended to be water tight. Mine, which is nearly 140 years old, was simply rough cut lumber laid up as tightly as the uneven, still bark-edged slabs could be. Both the builders' paper -or what passed for that in 1870 - and the wooden sheathing are absent in about 25% of the toal area of my roofs where you can look up and see the underside of the slates themselves - all without leaks. The wooden sheathing is there for stiffening of the connections among the common rafters and to be nailers for the slate.

Ice dams can be prevented or managed by other means from raking to removing and replacing lower rows of slate with flashing in intractable cases that can't be accessed by rakes.

How high is your slate roof? Two stories, three? Is there a wing?

I live in an older house than yours, timber frame w/o a speck of insulation, and with a slate roof. I'm in northern NY.

We use roof rakes to keep the critical roof surfaces clear of ice dams, when and where they are likely to occur, which isn't on all surfaces, and not every winter, or indeed from every storm. Study of each roof and the patterns will help keep the work load at a minimum and still effective.

I also have several nearby barn buildings which are completely unheated, some of those roofs develop ice dams on occasion, so despite conventional wisdom, it's not just insulation (or lack thereof) at play in ice dam formation.

I also repair all my own slate roofs, all ten of them.

I don't think that insulating the underside of the entire roof would be either wise or effective. Ihe ice dams aren't forming on the majority of it and you don't need a totally "cold roof" you just need it above where warm air is rising through the wall assembly, which happens to be the place where the roof plane and the wall plane meet, and hence you couldn't realistically spray anything there. There are other more effective ways to block that convective and radiant channel.

With all due respect the Building Science Institute, they can't be talking about slate roofs in this quote: (from Worthy's post above)

"Closed-cell spray foam has negligible water permeability, minimal water absorption, and excellent adhesion allowing it to act as a secondary rainwater barrier to limit damage when primary roof assembly rainwater control membranes leak..." (my emphasis)

Slate roofs don't have "rainwater control membranes", they rely instead on carefully calculated, redundantly overlapping but loose layers of slates, and of course, gravity.

One final thought: slate roofs being actually thin layers of rocks are perhaps the ultimate in a "cold roof". Mine acquire a thick coat of frost inside the attic, easily visible from underneath between sheathing members. This is of no concern because the loose, unsealed nature of a slate roof also means it's extremely self-ventilating and impervious to moisture as long as nothing interferes with that, such as any man-made material (Watershield, closed foam, pink sytrofoam ventilation channels, etc.). When you introduce these foreign, unnaturally tight layers you risk changing the environmental conditions in unforeseen, and possibly detrimental ways. Old houses have survived many decades, and even centuries, because they possessed inherently successful technological systems. Old houses with different technological systems have failed and disappeared from the scene. What remains are building technologies that work. Be careful when considering adding newer systems to the old houses - you may very well be altering a critical factor.

A some good resources for you are:

John Leeke's website: www.historichomeworks.com (Many topics of old house care maintenance covered from a profoundly conservationist point of view. Particularly useful re insulation and window repair issues.)

John Jenkins webite: www.slateroofcentral.com (Site with much info on proper care, repair and installation of slate roofs.)

and, of course, the National Parks Service Preservation Briefs Series on variety of topics, including #29 on care maintenance and repair of slate roofs. See link to index of the series below.

HTH,

L

Here is a link that might be useful: Preservation Briefs Topics Indexed

liriodendron

Thank you for some excellent references and personal experience.

(BTW, the point is to create a warm roof, not a cold one. Or, put another way, an unvented attic. It has been used in more than 100,000 homes since 1995.)

Dr. Lstiburek recently specifically reaffirmed the use of foam under decking with a slate roof on top in a report by Building Science Corp. on "Deep Energy Retrofits" (See Link.)

Of course, the OP can take the conservative approach of using a traditional vented assembly, but tightening up the ceiling to attic interface and adding more appropriate insulation.

Here is a link that might be useful: Final Report on the Expert Meeting for Details for Deep Energy Retrofit

Worthy,

Thanks very much for attaching that interesting summary report. I just went through it quickly (and will do it more justice with a slower read this evening when I have more time) but at this point I remain unconvinced that spraying closed cell foam underneath the OP's slate roof would be either viable or desirable. And without significant other adaptations and accomodations to manage her house's cumulative moisture load it might turn out to be extremely damaging.

The notion of a "deep energy retrofit" is an extreme alteration to virtually all the existing technology in most 19th and early 20th c structures. The discussion in the report acknowledges this problem. (Viz: the droll reference to whether homeowners would go along with the "chainsaw approach" to getting effective moisture and infiltration barriers around the intersection bdetween the vertical and roof surfaces. It involves the wholesale removal of the wall/roof eave edges and all the trim in the area.)

I have no doubt that taken as an integrated, systemic change that deep energey retrofits can produce remarkable improvements in energy use. Where I think the concept runs into problems is where only some components are adopted, but others are not. This risks fatally disturbing the existing equilibrium.

I am of two minds about intact structures with inadequate energy conservation profiles. On one hand I deeply appreciate the personal, economic, social, and environmental need for increased energy conservation. On the other, I believe that continuing to use functional buildings even in the absence of modern energy usage profiles is quite acceptable, too. These buildings are already "paid for" in terms of global energy cost because they are so old, and whatever harvest, transportation and construction costs they needed are long since "sunk". They are by definition the most "natural" buildings since they were made long before modern materials were invented. They remain living laboratories of the old ways of building "science".

I nibble away at the edges of my own buildings' energy use. As you may already know I continue to heat only with wood, largely obtained from my own woodlot. I work on energy improvements around major infiltration points (doors, windows, various wall penetrations, and sills) and I have installed some attic insulation (with mixed results both on the improved energy conservation and on whether or not it messes up existing moisture migration patterns). Later this year year, or early next, my DH and I are hoping to install a significant, grid-tied, photovoltaic electrical generation system which will improve our household's overall energy-cost footprint. (Not really to save us money, but more to capture the, until now, unexploited solar potential to offset the lack of energy efficiency of our buildings.)

But returning to Lynn's query, I still think that unless one is prepared and willing to institute most or all of the alterations detailed in the deep energy retrofit report, simply sealing up the bottom surface of the roof below the slate would be a bad choice.

L.

Liriodendron and Worthy, thank you for this thoughtful and fascinating discussion and relevant references. I'll study up, and let you know when I have a response from our experienced roofer. The bid from the foam insulation contractor is impressive: over $8K to insulate the roof with closed cell insulation below the roof deck! This would be one way of avoiding interfering with knob and tube wiring, and may or may not make a big difference with ice dams, depending on which of the arguments outlined above is actually true. I wish there were more empirical evidence! A randomized controlled trial of old houses treated with different methods of insulation and follow-up outcomes research. Is that too much to ask? (Yes.) But in medicine, the guiding principle is "primum non nocere": first do no harm.
Lynn

Old houses have survived many decades, and even centuries, because they possessed inherently successful technological systems.

In this case, I, too, would very likely take the conservative route, maintaining the vented nature of the attic, but sealing all the leaks in the attic/interior interface as outlined above.

However, the K&T is a technology you can safely do without. Here, it is virtually impossible to get householders' insurance coverage with K&T in place.