HVAC load calc ICF house with Hambro joists, foam insulation

[watercop]

Subject ICF house under construction near Jax Florida has a flat roof of 3" of poured concrete held up by Hambro metal joists whose top element is embedded in the concrete.

We'll spray 3-5" (haven't decide on precise thickness yet) of closed cell spray foam on the bottom of the concrete roof. That has an aged R-value of about 6 per inch.

However, I don't think I dare feed the full R-6 per inch into my load software because I expect substantial thermal bridging via the steel joists. The upper element of the bar joist sits in poured concrete, which, though covered in a white poolside-type cool deck material, will still get quite hot on a summer afternoon.

About every linear foot or so a 1/2" tube of steel welded to the top chord will emerge through whatever thickness of sprayfoam I specify. There are a total of 400+ such penetrations, and intuition suggests that on a hot day they'll conduct heat enough to feel warm to the touch.

I need a number for that load, either a way to correctly derate the insulation or as an added sensible load on a design day.

I surfed then called Hambro - they expressed curiosity and asked that I call them back should I reach a conclusion, in other words the guy I talked to hadn't the foggiest notion, and no suggestions as to whom I might call.

Anyone here ever run into anything like this?

 

[EnergyProfessional]

either you use a software (like DesignBuilder) where you can specify the thermal bridges as a % of the area. Or you just calculate the effective R-value (jsut a resistor calcualtion, the WI building code also has some calcualtion method for stud walls, that might work too).
You also could use 95 ft² of concrete + foam and 5 ft² of concrete + steel if 5% is the % of steel penetrations.
thermal bridging has quite some effect, much more than you think.
you also could wrap the remainder of the steel in some insulation if feasible.

 

[watercop]

Fortunately the % of area works out to about 0.5% of ceiling area. I agree that 5% would cause a significant heat transfer.

My load software doesn't provide for a bridging percentage. I like the idea of calculating the effective R value but would need to calculate the conductivity longitudinally down a steel tube set at 45 degree angle.

 

[AbbyNormal]

I built a four plex in the Caribbean. The units have a sloping concrete roof.

I put pressure treated 2x4s on the flat on top of the sloping decks to be a nailer for white standing seam roof.

I put 1.5" thick low end expanded polystyrene a little more than R4 per inch in between the 2x4s. I covered it over with tar paper and then had the white standing seam on top.

Thermal mass, with air films etc maybe an R value of 7.

The attic is sealed, I have gable walls in the attic that face east and west. I have tar paper up against the concrete block gable walls and then the same 1.5" of foam on the inside.

I would say I have less attic heat gain than a home with R30 on the ceiling plane in a vented attic space.

With a flat roof, why not build it up with insulation and go with a white membrane, it is better to keep the heat out in the first place.

Insulating the underside of the roof is the second best way but what you will be doing is allowing the sun to charge up that thermal mass to the maximum each day and then rely on your insulation from slowing down how much trickles into your house.

Back to my own sealed attics. I deliberately avoided conditioning the spaces. I really kept an eye on the unit I lived in, had hobo data loggers up there for the first six month period. I found the attic averaged 80F and 50% RH.

My wife now stores some heirlooms up there so in the one unit I added about 30 CFM of supply air up in the attic. The attic space now averages about 79F and humidity in the low 40s.

I have yet to condition the attics in the three remaining units but in three years I have not had any kind of issues.

One thing I found interesting is the attics are the warmest about an hour or so after the sun sets and I attribute this to two causes

1) Stored up solar heat is dissipating from the thermal mass of the deck

2) Because the sun is down I turn on lights. The light fixtures are surfaced mounted on the underside of the sheet rock ceiling that divides my attic from the living spce below. The attic traps the heat from the light fixtures.

The way we build has a far greater impact on our comfort, energy consumption and IAQ, than any HVAC system we install

 

[AbbyNormal]

built up insulation on top of the deck solves the short circutting through the bar joists as well.

The way we build has a far greater impact on our comfort, energy consumption and IAQ, than any HVAC system we install

 

[watercop]

I agree that a built up roof is the way to avoid the bridging. However, homeowner intends to be able to use the roof as an upper deck - it will have a walking surface similar to poolside Cooldeck.

I did some back-of-envelope calculations based on conductivity of steel, expected cross sectional area of the penetrations and a SWAG at roof deck temperature and injected an additional sensible load of 1 btuh / SF.

I've since learned that the ordered windows and doors have significantly better U-factor and SHGC than I originally assumed, so I have some headroom to rely on.

Installing ductwork into the itty-bitty Hambro joist openings is a whole other kettle of fish...