How is the best way to reduce moisture in the attic for the house described below? Black mold is present on surfaces. Thanks for all guidance! Construction: built in 2012; open-architecture has default vertical-flow of moisture; no trees within striking distance of house in case of natural un-rooting. Location: Charlotte NC suburb; 200 linear-feet from wetland and large creek (exterior (sides and roof) collects condensation every night year-round). House: 1,200 square-foot footprint; 3-story with attic; frame construction; Hardy-Plank exterior over Tyvek building wrap and foam-board; 6-inch studs with Icynene foam in walls; Total Electric; lighting is fluorescent, CFL, and LED with only a few decorative incandescent sources. No basement; house is built on concrete pad with gravel, vapor-barrier, tile-field drain surround. Attic: contains 1 of 3 air-handlers for air-conditioner/heat-pump; air-handler drains into sewer-stack-pipe; air-handler has an overflow pan with drain into sewer stack-pipe; Icynene foam on gable-ends and shed-roof-planes; eaves-vents are(supposedly) sealed; no gable-vents; no ridge-vent. Roof: metal; light beige color; plywood decking; metal has slight separation from decking; penetrated by 3 SolaTube light-pipes, 1 VeluxUSA skylight, sewer-vent; no chimney (no fireplaces in house); externally mounted pull-fans for bathroom vents; 50% cover on half of roof by photovoltaic solar panels.

The interior surface of the Icynene should never be cold enough to condense moisture, but in Zone 4A you need at least R15 on the exterior of the roof decking to not need an interior side vapor retarder in an unvented roof, which I'm guessing is what you have? The light metal roof is also a problem, since it lowers the average wintertime roof deck temperarture, which results in even MORE moisture becoming adsorbed into the roof deck over the winter.

In the spring the warmer temps force moisture out of the roof deck, leading to very high moisture content in the attic and Icycene. For much of the shoulder seasons it can be cycling moisture into and out of the roof deck with high attic moisture peaks during the day, lower levels at night as the cool roof deck re-adsorbs the moisture.

If you had a Class-II vapor retarder such as half-perm "vapor barrier" latex on gypsum, or better yet a "smart" vapor retarder such as 2-mil nylon (eg Certainteed MemBrain) the total moisture adsorption over the winter would be minimal, and the moisture cycling leading to surface moisture high enough to support mildew on the foam would go away. That may be easy or hard to retrofit into your house depending on the particulars.

Actively condition the attic with an AC register should reduce the amount of moisture in the attic air, and eventually the total amount of moisture cycling into & out of the roof deck. A room dehumidifer in the attic spaces can accelarate the drying. once the attic is reasonably dry, then it's time to put something on the mildew to kill it, then install the interior side vapor retarders.

Unfortunately vapor barrier latex applied directly to the foam doesn't work (many have tried and failed to make that work)- it needs to be bound on the gypsum board facer to reach it's ~0.5 perm vapor permeance.

Smart vapor barrier membranes are best mounted to strucural framing, and detialed as an air barrier with overlapping & taped seams , all seams supported by framing. How easy or hard that is to pull of in your house can't be predicted without more information.

How much foam do you have on the underside of the roof deck? Current IRC code min would require about 13", but under IRC 2009 10-11" would have met code, but many installers only put 5-6" with some sort of "...that's all you really need..." theory marketing BS for skimping that much. What they really mean is "...that's the most that won't spook the client with sticker shock when the quote comes in...".

Sounds like you don't have good air movement between the house interior and the attic. Cooling (from any source - exposed ducts, interior, exterior) and no latent heat/moisture removal is a bad combination. A pair of small, matched fans (one pushing, one pulling to/from interior to attic) will bring down the relative humidity without unbalancing the pressure. Or a supply AND return. Verify with a humidity meter.

I don't know about any danger that the mold might present. Or smoke/fire.

Wait, is this an unconditioned attic with foam on the underside of the roof deck/rafters?

Yes, it is.

Unconditioned space should be well vented to the exterior. Semi-conditioned space (ie, sealed and insulated) should be turned into conditioned (at least for % RH) space.

Posted By Tom Tolbert on 20 Jul 2016 04:43 PM
Yes, it is.

Ok, and is there insulation in the joist bays/above the ceiling?

The spaces between the joists that support the ceiling of the top occupied floor and the floor of the attic have no insulation. An area of about 200 square feet of these joists is decked with plywood in the area where the air-handler is located. The rest of the attic "floor" has no decked walking-surface. The total area of the attic floor is about 1,200 square feet. The elevation at the ridge-beam is 9.75 feet above the tops of the joists. The total volume in the attic is about 5,880 cubic feet. The air-handler is located near the center of the attic (i.e., front-to-back and side-to-side). I am not an HVAC design-engineer. However, if I understand the direction that incoming advice is taking, the entire attic should have been treated as just one more level of air-conditioned space. The HVAC contractor did not do this. There is no indication that any attic supply-&-return was provided for and later forgotten during installation. If I add a supply and return for this air-handler, the load on the top-floor HVAC system would increase. Reduction of radiation from the top-floor ceiling into the top-floor should off-set the need for lower thermostat-settings in the summer; compensation for winter-time may be harder to predict. I believe that reserve cooling is available; I have no "feel" about heat-pump capacity. Would one (1), supply-duct mounted high in each gable-end of the attic provide the temperature-control and air-mixing (within the attic) to fix this moisture-condensation problem? This would be a total of two (2) supply-ducts located about 42-feet apart. I think that in-line butterfly-dampers would allow flow-control if the ducts were sized correctly. What diameter ducts would be appropriate? Would the attic need any supplemental circulation-fan to prevent air-stratification? Would I leave the thermostat at its current location in the top occupied floor? Would the filtered return need a damper to balance flow-rates between itself and the top-floor return? The top-floor return penetrated the ceiling directly under the attic air-handler.

You're right, it sounds like you have what should be a "conditioned attic" but they didn't supply conditioned air to it. I wouldn't think it would require much though. Maybe someone more expert will chime in. But you can't simply vent it to the outside.

Thank you for your comments. I gather that you are more concerned about moisture "pumping" into the attic from the roof-deck than about moisture rising vertically from the occupied space. (The "techies" did not mention infiltration from outside when the proposals were presented before construction. They were concerned with high moisture-content air from cooking, breathing, showering, etc.) In re. you last paragraph: On the roof-planes, Icynene foam fills the spaces between the rafters (Southern yellow pine on 16-inch centers); 50% of the nominal 2-inch faces show; the rest have up to 0.5-inch of over-spray. The Icynene is not skived flush; it has the rough surface-texture resulting from free-expansion. I do not know whether these rafters are 2x10 or 2x12; they are planed to dimensions (not rough sawn). On the gable-ends, Icynene foam fills and overruns the 6-inch studs. No studs show through the foam. The Icynene is not skived flush; it has the rough surface-texture resulting from free expansion.

With the insulation a the roof deck you definitely DON'T want to vent it to the exterior- if it's inside the thermal boundary of the house it is by definition conditioned space, even if you're not actively heating & cooling that space or using it for anything other than a mechanical room for the ducts & air hanlder, etc. Conditioned spaced needs to be inside the pressure boundary of the house too. But it does need to have reasonable amounts of air exchange with the rest of the house so that the humidity of the air (absolute humidity, not relative humidity) tracks with the rest of the house.

Open cell foam is still fairly permeable to water vapor even at 10-11", and needs at least a Class-III (10 < 1.0 perms) to limit the peak moisture content of the roof deck. At 10" most half-pound foam runs 3-7 perms, which may be enough to save the roof deck, but it will accumulate moisture in the outer 2" over the winter. This is the moisture that you need to purge by air conditioning/dehumidification before adding an interior side vapor retarder. If it's possible to put up gypsum board (even if it means trimming the foam a bit), a coat standard latex primer will cut the seasonal moisture accumulation in the foam & roof deck in half. If 2 mil nylon is used under the gypsum it will cut it by an order of magnitude, without significantly reducing the drying capacity, due to it's variable vapor permeance. Both the rafters and roof deck should be given this treatment.

If you can't easily install gypsum board the sheet nylon alone will be good enough, as long as you make it reasonably air tight. The foam itself is an air barrier, and even if the nylon leaks a little there won't be much of a convection drive, and it will work. (The air tightness would be far more critical if it were fiber insulation.)

If I add a supply and return for this air-handler, the load on the top-floor HVAC system would increase

You could control the vents (or fans) based on relative humidity (mold growth is NOT based on absolute humidity). Maybe $200 for two motorized dampers and a humidistat.

Only conditioning when needed would save some energy.

Below ~40F mold growth is extremely slow, independent of the RH.

Above 65F mold growth doesn't really take off exponentially until the RH is above 70%, but holding the line at 60% makes a noticeable difference with mildew.

If the foam is within the thermal envelope, is there a need for a thermal barrier on the foam? Smoke resistance?

The foam IS the thermal envelope from a heating & cooling point of view.

Foam burns, and fire codes require a fire rated timed theramal assembly to protect the home from fire spread if the space is occupied used space (even if that use is only dead storage). In space that are not used a lesser thermal barrier may be used. Increasingly intumescent paints are being allowed for the thermal barrier against ignition and fire spread, but that is not universal- check with your local building department.

This is an idiosyncrasy between codes that will probably change in years to come. For the Canadian Codes which focus on smoke and gas generation, if the foam is within the perceived thermal barrier of the fire, it must be covered with the equivalent of 1/2" drywall. Intumescent coatings are not allowed.
Other codes focus on ignition and in that case intumescent coatings would be allowed in some cases.
I suspect that in the next few years, all north American codes will tend to come into alignment one way or the other.

I find it interesting that this document http://www.icc-es.org/News/Articles/AY126ThermalBarriersSPF2011-51811.pdf seems to indicate that an attic space c/w hvac equipment is allowed to have alternative covering of the foam. If there is an attic fire, the hvac unit would push smoke throughout the house unless there is a requirement for a smoke alarm activated shutoff for the unit.

That might make a permanently installed dehumidifier an easier option.

Lots of info if you google for "open-cell-spray-foam-and-damp-roof-sheathing". My conclusion - just getting the attic %RH under control should fix the OP's issue. Do more if that proves to not be enough.

This is a hot topic of research in recent years, and the best method of dealing with it seems to be fully conditioning the attic space, while controlling the vapor permeance properly to limit the rate of moisture adsorption & release from the foam & roof deck.  See this (which JonR seems to have responded to 2+ years post-publication...)