Let me say ahead of time that I’m trying to get heat loss calculations run to size my boiler. However, I’m wondering if there is any data out showing some rough heat loss numbers per sf for ICF homes. I’m running into plumbing supply companies who seem to have a desire to oversize the boiler. I’m planning a 3800 sf Amvic plus ICF Home in Northern Vt Hot water baseboard heat throughout with one small area of radiant floor in kitchen ( transfer panels). Thanks

The problem is, your heat loss calculations depend on more than simple wall composition and R value.

Can you explain Dil? Obviously it needs to include attic insulation. Wouldn't a blower test confirm what they need? HVAC folks have 'always' oversized(more money in their pockets) if not waaaay oversized the system to cover their a%$#$#$ Probably 99.9% of people have 'no' idea what they need and the HVAC folks take advantage of that.

Posted By predgw on 10 Feb 2018 02:29 AM
Let me say ahead of time that I’m trying to get heat loss calculations run to size my boiler. However, I’m wondering if there is any data out showing some rough heat loss numbers per sf for ICF homes. I’m running into plumbing supply companies who seem to have a desire to oversize the boiler. I’m planning a 3800 sf Amvic plus ICF Home in Northern Vt Hot water baseboard heat throughout with one small area of radiant floor in kitchen ( transfer panels). Thanks

There's no such thing as rough heat loss numbers for any kind of building. The last thing you want to do is design your HVAC system by guess and by gosh. That sounds like what you're doing. You'll regret it later. Engage an HVAC contractor who will do a proper heating/cooling load calculation and system design in accordance with Manual J. There are way too many parameters to consider.

I bought a $50 program, and it proved to be very accurate in real-world results. I had a 'pro' tell me (over the phone, no less) that my house needed 2-1/2 tons, minimum. Our actual cooling load worked out to 9,000 BTU cooling (3/4 ton), and 12K BTU heating (one ton). A 2-1/2 ton system would have been woefully inefficient, and would not have functioned well. It's important to be honest and accurate with all of the data you input, and there's a lot of it. Things such as floor insulation, attic insulation, windows, lot orientation, even roof overhang size all factor in, not just ICF walls.

What program did you get for $50.00? Thanks for all the replies. I have researched this subject enough to know that I’m not trying to design around rule of thumb. I’m just looking for any information. First heat loss came in at 65k, but the boiler recommended was a WM GV 90-6. Pretty big boiler. ICF is not normal construction and everyone likes to tell me such.

There was a recent independent study completed on this. See this link from the ICFMA website. http://icf-ma.org/independent-testing-reveals-thermal-mass-effect-energy-efficienty/

Just for reference. I can heat my 3700 SqFt ICF house (plus walkout basement 1/2 of it out of the ground) with 75,000 hot water heater. All floors are heated concrete. Ann Arbor Mi. Had a couple nights to -7 degrees this year. Unit was running 90 percent of the time when super cold.

I also have a geo furnace and geo water heater but it cost $9 to heat with geo and $7 per day to heat with Natural Gas in my area. So I only run the geo if I ever needed a backup. I have ran the house in both configurations. Geo furnace will be used for AC.

Is this a new house? If not you can use your heating bills to get the best estimate. We can help if you give the numbers. Good Luck...

Rules of thumb based on square footage of conditioned space, all suck, even for code-min houses. Heat loss is a function of the exterior surface area and the properties of the materials of the sub-components of the exterior surface assemblies. (Windows are lossier per square foot of window area than walls are of wall area, etc.) Even the ratio of exterior surface area to floor area can vary by more than a factor of two, depending on the shape of the house. (A 3-story cube has a lot less exterior surface per square foot than a sprawling 1 story ranch will multiple ells.)

With ICFs you have the additional time lag between the peak indoor to outdoor temperature difference to when the interior surface of the wall is coldest, lowing and time-shifting when the peak load occurs.

That said, an IRC code-min houses would typically come in at about 12-14 BTU per square foot at -10F outside (the 99% outside design temp in Barre is -11F), and 70F indoors, with MANY exceptions (in both directions.)
An ICF house should come in under 10 BTU/ft, but it's pretty easy to design one that doesn't. Bigger houses have lower load per square foot ratios due to a lower exterior surface to floor area ratio, and at 3800' you're a 1.5x the national average, which probably means a lower than typical load per square foot ratio.

A fire-tube type boiler with a 10:1 turn down ratio like a Lochinvar KHN085 or HTP UFT-080W can cover a huge range of heat loads (and individual zone loads) without short-cycling, with HUGE margin. Even at a ludicrous (for ICF house with all other aspects at code-minimum) 20 BTU/square foot (76,000 BTU/hr) either of those boilers would make it, and even at 5 BTU per square foot (19,000 BTU/hr) it won't short-cycle unless you micro-zone the bejeezuz out of it.

But any radiant floor zone really SCREAMS for an aggressive Manual-J load calculation before the radiant designer is allowed to jump in.

My heat pump was down from mid-Dec to mid-Jan due to a coil leak. During that time we heated with our backup heat source, a pellet stove. Over a period of about 35 to 40 days we used approximately 40 bags of pellets, or roughly one bag per day. Using 8200 Btu/lb for heating value that works out to be an average heat load about 15,000 Btuh. This is for a house of 2000 sq ft inside the walls with outdoor temp mostly in the twenties and thirties.

Heat load isn't an AVERAGE- it's the rate of heat loss at a particular set of outdoor & indoor temperatures. The temperatures normally used are (the code-minimum) 68F indoors, and the 99th percentile historical temperature bin at that location.

https://articles.extension.org/sites/default/files/7.%20Outdoor_Design_Conditions_508.pdf

With the average heat load over a particular time period based on fuel use (and heating appliance efficiency) it's possible to infer the 99% design heat load (or at any other arbitrary outdoor temperature) with a linear approximation, as explained here:

http://www.greenbuildingadvisor.com/blogs/dept/guest-blogs/out-old-new

If the average load at an average outdoor temp 30F is 15,000 BTU/hr and an indoor temp of 70F (a temperature difference of 40F) , it'll be about twice that, or 30,000 BTU/hr @ -10F (a temperature difference of 80F).

Window losses are very non-linear with temperature, with a portion being a function of the fourth power of the absolute temperature (above absolute zero), which can screw up linear approximations when windows are the primary heat loss, as in some very high performance houses. The labeled U-factors for windows the US is the heat loss per square foot at 0F outdoors, 70F indoors, whereas in Europe window manufacturers specify the window performance at 0C outdoors, 20C in doors. For what's probably too much (or not enough) information on that topic, see:

http://obec.on.ca/sites/default/uploads/files/members/CCBST-Oct-2014/A5-1-b.pdf

Posted By Dana1 on 21 Feb 2018 10:27 PM
Heat load isn't an AVERAGE- it's the rate of heat loss at a particular set of outdoor & indoor temperatures. The temperatures normally used are (the code-minimum) 68F indoors, and the 99th percentile historical temperature bin at that location.

True, but it is an indicator! An R-10 wall will have twice the average heat loss of an R-20 wall. I forgot to mention we like a warm house so this was with an indoor temp ranging from about 72 on the far bedroom to 76-78 in the living room where the stove is. The pellet consumption confirms what the calculated heat loss. My pellet stove has three heating ranges, low, med, high and is thermostat controlled. It ran on low probably 1/2 of the time and medium about 1/4 of the time and was off about 1/4 of the time. Only occasionally did I have it on high for a few minutes. It's tough to measure the heat output of an air to air heat pump. The engineering data for my 3 ton Daikin shows a heat output around 27,000 Btuh at an outdoor temp of 5° and an indoor of 75°. It keeps the house warm at 72+ when it's -10 outside.

So what I'm saying is that the pellet stove performance confirms my estimate of the heat pump performance and both indicate the max heat load of the house at our usual winter temps of -10 to +10 is around 25,000 to 30,000 Btuh, right in line with the Manual J calculation.

An R10 wall would only have twice the loss of an R20 wall in a steady-state simulation, or dynamically only if there is very minimum thermal mass. The thermal mass of the concrete in an ICF can reduce both the peak & average heat loss numbers fairly significantly compared to a low-mass framed wall of similar steady-state R. This is the source of often misleading "R50 equivalent" claims by ICF vendors, and why it's better to simulate the house with something like BeOpt using local weather and site factors than using simple models.

Manual-J makes a nod to the thermal mass issues, and is good enough for specifying the equipment, but doesn't fully reflect the energy use performance of mass-walls with insulation on both sides of the thermal mass. It's WAY better than a "lessee, ya got 3800 feet times xx BTU per foot gives ya a load of yy,yyy BTU" type rule of thumb approach though. There are too many ways for dumb rules of thumb to deliver numbers that totally wrong, often by 2x or more, even worse with mass walls like ICFs.

Sounds like in an ICF house it might be more inefficient if you lower the thermostat at night?

I have had two high mass houses and both I have always found it is best to kept the thermostat constant because house takes to long to react and that can lead to overshoot when trying to raise the temp backup.

Overnight setbacks won't lower the net efficiency of an ICF house. The thermal mass is isolated from the continuous interior by ~R10 with most ICFs, which is about the same performance of a 2x4/R13 type wall which is plenty for allowing reasonable recovery ramp rates, and the concrete is at a much more moderate temperature than "the great outdoors".

High mass houses where the mass is all on the conditioned space side of the insulation are less responsive, with longer ramp times.

- but much linger periods of stable temperature. If you are trying to raise the temp in a room by heating a large mass directly - like a radiant slab - there is a longer ramp time but if there is abundant mass that's insulated from the outside but exposed to the room and you are only heating the air in the room like with a heat pump, furnace, wood or pellet stove, etc., that mass can moderate temperature swings in a nice way - absorbing excess heat and releasing it back during cooler periods like night time. The combination of direct and indirectly heated mass is nice. The more exposed mass, the less of an issue temp setbacks are as turning the thermostat down has much less effect short term.

When we first moved into this house I tried the set back approach for heating for only a few weeks. The interior mass of the house just by itself plus the concrete rat slab in the crawl took too long to warm up in the morning. We keep the thermo at 74° all winter. But I do use set back in the summer, but opposite of the usual approach. I drop the temp down to 70 at night and the mass of the house helps keep it cool during the day. We are on time of day metering so I have the thermo set to 76 or 78 from 1 pm to 9 pm. The heat pump seldom kicks in for cooling during those hours.

Posted By dmaceld on 24 Feb 2018 07:10 AM
When we first moved into this house I tried the set back approach for heating for only a few weeks. The interior mass of the house just by itself plus the concrete rat slab in the crawl took too long to warm up in the morning. We keep the thermo at 74° all winter. But I do use set back in the summer, but opposite of the usual approach. I drop the temp down to 70 at night and the mass of the house helps keep it cool during the day. We are on time of day metering so I have the thermo set to 76 or 78 from 1 pm to 9 pm. The heat pump seldom kicks in for cooling during those hours.

To get reasonable recovery times from overnight setback with a heat pump requires oversizing the heat pump, or using resistance electricity heat strips, either of which takes a toll on efficiency. This is true for all houses heated with heat pumps, not just ICF or other high mass houses. Oversizing a hot air furnace has almost no impact on efficiency, making overnight setback a winning strategy. With heat pumps the impact of oversizing on seasonal HSPF efficiency is pretty dramatic, with a lot of compressor spin-up energy wasted when cycling on/off during periods of lower load.

Pre-cooling the house during the hours when electricity is cheap is definitely the smart strategy, and works pretty well in all higher-performance houses. Large amounts of unshaded west facing window area can undermine the effectiveness of that approach though, even in high-mass, high-performance houses.

We are using small mini-splits in our ICF house, and they do a good job, but recovery is very slow. They are so cheap to run that we just leave them on at a steady temperature during the season. Here in SC, we often leave them off for months at a time. We've had them off for about a month, and with a little luck, may not need them again until May.