I'm dense packing my new home's walls, as well as the cathedral attic. The attic will be unvented.
The wall assembly is Vinyl->Tyvek->.75" XPS-> Tyvek, OSB-> Cellulose (Dense)-> ?
The roof is Dimensional Shingles-> Roofing Felt -> OSB -> Cellulose (Dense) -?
I live in Milwaukee, and to my knowledge code requires a Class 1 or 2 vapor barrier on the warm side. The inspector is fine with the vapor barrier paint, or the plastic. What would you pick?

I would use plastic - most important is to prevent any interior air from flowing into the wall cavity - and plastic is much better at that.

National Fiber recommends against the use of poly vapor retarders in dense-packed assemblies, since poly reduces the drying capacity too much, extending drying times from incidental bulk water leaks by many months. Poly can even interfere with drying from moisture that got into the cavity from air leaks at electrical outlets, etc. Vapor retardent paint would be enough to sufficiently limit moisture loading of the cavity via vapor diffusion during the coldest weeks of winter, yet still allow drying during the warmer months. (Even standard latex might be enough, which is ~2-3perms, vs. ~0.5 perms for vapor retardent latex.)

Air transported moisture in a standard construction wall is usually a larger moisture transport mechanism than vapor diffusion with or without vapor retarders in place- making the interior air-tight is key. Dense-packed cellulose is nearly an air-barrier by itself (call it an "air retarder"), but not sufficient by itself. Use air-tight drywall methods (google, it there are many detailed explanations out there, eg: http://www.greatlakeshomeperformance.com/elibrary/DIY%20Airtight%20Drywall.pdf ) Using poly is not a guarantee of greater air-tightness, but if you're a fan of air-barrier sheeting, you could use permeable or semipermeable housewrap on the interior, and adjust the vapor-permeance of the interior wall with paint.

Looks to me like National Fiber recommends no vapor barrier of any kind - but code won't allow that. I agree you can do better than either of the original options presented.

From: http://www.nationalfiber.com/docs/CelluloseInsulationMoistureAndVaporBarriers0909.pdf

"By blocking the movement of moisture-laden air, cellulose will reduce moisture movement to manageable levels within the building assemblies. Any remaining moisture moving by diffusion will be further blocked by primers and paints used on the interior surfaces."

They're still counting on the paint to manage vapor diffusion, but rule out true vapor retarders like poly or foil.

Vapor redardent paint is still ~10x more permeable than 6mil poly, and qualifies as a Class-II vapor retarder. Standard latex paint is 5x more permeable than vapor retardent paint, and only tests out as a Class-III vapor retarder, but STILL might be enough in that location with the R4 XPS on the exterior. The average winter temp of the outermost cellulose will be considerably warmer than the outside of the XPS.

What works in National Fiber's backyard (Springfield, MA) works in Milwaukee, with very comparable winter temp averages in those two locations. Their recommendations might need some adjustment in Whitehorse Yukon, but there are houses in Saskatchewan insulated with cellulose 80-90 years ago with no vapor retarders, and no moisture problems. With less than an inch of XPS on the sheathing it'll dry toward the exterior somewhat in winter, and with vapor-retardent interior paint, in both directions seasonally. With poly on the interior of an air-conditioned house there could even be issues with condensation on the poly during the summer if the assembly isn't allowed to dry toward the interior through the spring.

For some reason, I would think a vapor barrier would be best placed on the outside of the framing to prevent vapor from the outside getting into the wall assembly. You can control indoor humidity, you cannot control outdoor humidity. Or am I missing something?
Or is it best to have vapor retarders (not barriers) on the outside and on the inside so the wall assembly could dry in either direction?

> but rule out true vapor retarders like poly or foil

You're coming up with things they didn't say. They recommend no vapor barrier and that is not an option here.

"In summary, we do not recommend the use of vapor barriers with cellulose insulation...".

Posted By Jere on 27 Jul 2010 10:31 PM
For some reason, I would think a vapor barrier would be best placed on the outside of the framing to prevent vapor from the outside getting into the wall assembly. You can control indoor humidity, you cannot control outdoor humidity. Or am I missing something?
Or is it best to have vapor retarders (not barriers) on the outside and on the inside so the wall assembly could dry in either direction?

yes you have trapped moisture in the exterior wall with no way to escape

Posted By jonr on 27 Jul 2010 11:48 PM
> but rule out true vapor retarders like poly or foil

You're coming up with things they didn't say. They recommend no vapor barrier and that is not an option here.

"In summary, we do not recommend the use of vapor barriers with cellulose insulation...".

I've only applied some examples to what they did say to flesh it out, since the poly vapor barrier was part of the question

In the US the common definition of vapor barrier material is " less than 0.1 US perm" in an ASTM E 1745 test:

http://en.wikipedia.org/wiki/Vapor_barrier

But there are variations on the theme in some industries b require it to be less than 0.01 perms  in an ASTM E 1745 test to qualify as a vapor barrier :

http://concreteconstruction.net/industry-news.asp?sectionID=715&articleID=350812

At  ~0.03-0.05 perms, poly is a vapor barrier, by the more common definition.  So is foil.

Vapor retardent paint is not a vapor barrier, since it typically tests at ~0.5 perms. 

Standard latex paints are also vapor retardent, but test out at ~2-3 perms.

What they DO say (and I previously quoted) was:

"Any remaining moisture moving by diffusion will be further blocked by primers and paints used on the interior surfaces."

Between those two statements the strong inference is that using the vapor retardency of the paints to control moisture is superior to or less risky than using highly vapor retardent materials on the interior.

Ergo, National Fiber's recommendation is consistent with using vapor retardent latex, and against using poly sheeting in a cellulose insulated wall.

Posted By Jere on 27 Jul 2010 10:31 PM
For some reason, I would think a vapor barrier would be best placed on the outside of the framing to prevent vapor from the outside getting into the wall assembly. You can control indoor humidity, you cannot control outdoor humidity. Or am I missing something?
Or is it best to have vapor retarders (not barriers) on the outside and on the inside so the wall assembly could dry in either direction?

In winter the dew points of outdoor air are at low-moisture extremes that aren't healthy or comfortable for humans at room temp.  If air leaks from the more humid warmer air into the cavity it will condense moisture onto the colder parts of the assembly.  When the water-vapor pressure of the interior air is high enough to diffuse water through the wall into the cavity at a rate higher than it can dry to the exterior during the winter, that too can cause condensing levels of moisture to build up.   The dew point of 30% relative humidity 68F air is ~37F- any point in the assembly that is 37F or colder will have condensation issues if exposed to even a miniscule stream of that air. Running the interior RH much lower than 30% runs into health & comfort issues, so that is often used as a lower bound in this type of analysis.

Air sealing is the most critical, but adjusting the vapor retardency counts too.  Cellulose is somewhat more forgiving than most insulation materials, since it can absorb a large fraction of it's weight in moisture without losing R value or long term degradation, and it will wick condensation away from structural matierals. But counting on this buffering capacity alone wouldn't be a good idea.

In hot humid areas, keeping summertime moisture out is the larger problem, in which case using higher exterior side than interior side vapor retardency works better.  An interior vapor barrier in FL is a recipe for condensation & stud-rot on the interior side in much the same way that an exterior side vapor barrier in WI will cause condensation & rot on sheathing & exterior edges of the studs.

Since many places in the eastern US see both the winter & summer, vapor BARRIERs can be as much of a problem as they are a solution. But if humidity is allowed to cycle in/out at a rates low enough by use of vapor RETARDERS, peak moisture levels within the structural materials can be controlled by design. The details of what works may be climate-specific. Use of cellulose as a hygric buffer can offer a margin of protection in areas with very large seasonal swings in temperature & outdoor humidity, and in very temperate climates the insulation type & vapor retardency of the inner/outer surfaced can be non-issues (as long as you avoid double vapor BARRIERS, which can form moisture traps.)

Bottom line is, whether a vapor barrier is preferable to vapor retarders is a local-climate thing, but if you can, it's usully better to design the assembly to be resilient without true vapor barriers.

Most commonly, "vapor barrier" and "vapor retarder" are used interchangeably, by the experts and in the Wikipedia article you reference. If you want to talk about vapor permeability, use "perms" instead of vague, misleading and generally unused binary distinctions.

"We have vapor retarders, we have vapor barriers, ... What do these terms mean? It depends on whom you ask" or yet another example of interchangeable usage.

Thanks Dana, that makes sense!

Posted By jonr on 28 Jul 2010 11:54 AM
Most commonly, "vapor barrier" and "vapor retarder" are used interchangeably, by the experts and in the Wikipedia article you reference. If you want to talk about vapor permeability, use "perms" instead of vague, misleading and generally unused binary distinctions.

"We have vapor retarders, we have vapor barriers, ... What do these terms mean? It depends on whom you ask" or yet another example of interchangeable usage.

The National Fiber document clearly recommends using paint for vapor diffusion control, which means they aren't including paint in their functional definition of "vapor barrier". (But they don't specify what paint, how many layers, etc.)

They didn't mention poly or foil either, but those would fit most practitioner's functional definition of "vapor barrier", and are the materials most commonly sold as "vapor barrier" in the industry.

I too would have been happier if they had specified a minimum perm rating for the interior, but something that passes about an order of magnitude more vapor than 6 mil poly (like vapor retardent paint), and but only ~1/5 that of a shot of cheap latex wouldn't necessarily fall outside of their recommendation against vapor "barrier". 

Shall I call up the engineering firm they used for their modeling behind the recommendation and give them the "WhaddayaMEAN by that?"

To be sure, vapor retardent paint reduces the capacity of the assembly to dry by diffusion toward the interior by about 80% as compared to standard latex.  But a layer of poly or foil would be about a 98% reduction.  The difference is that instead of extending the spring drying period from weeks into months, a sheet of poly would extend the relative drying time beyond  the next winter-accumulation period, which means it has to rely exclusively on exterior drying.  In Milwaukee the bulk of the potential moisture accumulation period is on the order of 6 mid-winter weeks.  If it takes 6 months instead of 6 weeks to shed the winter accumulation of moisture via diffusion in the other direction it's not a disaster, but if it takes a year or more, it is.  But since the vapor-retardency of the exterior of the assembly isn't super-low (maybe about 1 perm of sheathing + XPS)  shoulder-season drying in both directions in this stackup should be pretty good.