Posted By toddm on 08 Feb 2012 08:46 AM
An ad you didn't see during the super bowl: ford expedition! 50mpg!* (downhill, in neutral) So,FBBP, what's your view of the EPA?

toddm - most people on GBT won't accept the hotbox results for fibreglass insulation but some then turn around an quote them as gospel for other products.
Can a hot box tell you how a product performs in a cold climate as compared to a tropical one? Will your aac work as well in Mexico as it will up here?
More than a few decades ago we're were already having these discussions with regards to manufactured log home. The six inch wall was only at very best R9 but probably only R7. So what? We popped in the same furnace as we would for their R20 stick framed cousins and they always performed better than their cousins. Design temp of minus 32ºc.
What about the EPA? Does your vehicle get the exact performance the EPA says it will? Both EPA and hot boxes are useful tools, unfortunately often misused.

I'm glad that FBBP feels comfortable enough to speak for "most people" regarding hot box testing, but I'm having a hard time figuring out who these "most people" are, exactly.

The ORNL's large scale climate simulator guarded hot box approach does a pretty good job at determining how different assemblies behave in different climates, far better than ASTM C 518 applied to dumbest of 2D models. http://www.ornl.gov/sci/roofs+walls/tour/LSCS.html

BAM! BAM! BAM!

That's the sound of you guys beating this poor dead horse.
Regards...

Posted By Dana1 on 08 Feb 2012 11:12 AM
I'm glad that FBBP feels comfortable enough to speak for "most people" regarding hot box testing, but I'm having a hard time figuring out who these "most people" are, exactly.

The ORNL's large scale climate simulator guarded hot box approach does a pretty good job at determining how different assemblies behave in different climates, far better than ASTM C 518 applied to dumbest of 2D models. http://www.ornl.gov/sci/roofs+walls/tour/LSCS.html

I have not got time to record every time a poster by the handle of Dana1 has inferred that fibreglass insulation does not live up to its ratings. That is the only test I referred to as "most people". You take it on yourself to answer my question as to wither a hot box or more specifically ornl's hot box can assess different climatic conditions on an assembly. And I would agree that it can but does it do it in a productive way?

"There is sufficient heating and refrigerating capacity to vary the simulated outdoor conditions in diurnal cycles, which allows tests of the dynamic response of test sections."

But do diurnal cycles show the full story? Maybe yes maybe no. How long do they run this? One day? Ten days? A full year? How do they compare air volume to assembly mass ratios? While this particular lab has outstanding features it doesn't mean we are necessarily testing for the right things in the right way. I have seen very few reports that say AAC or ICF perform so and so at this temp and so and so at this temp. I'm sure they are out there but I have not yet seen them.

Dana1 - I for one am very appreciative of the hours and hours of good information you have provided to the readers of this forum. You could be even more helpful if you would not always say that the anecdotal evidence must be wrong if it doesn't agree with the lab report. Maybe we are just testing wrong or interpreting test results wrong. Every test has some value but we need to make sure we are drawing the correct conclusions.

When you see a one legged dog running, you say "that can't be and I can prove it by this lab result". I would look at that dog and say "what did we miss in the lab report" ;-)

Bambambam. That's the sound of FBBP beating a one legged dog.

Is that the same one-legged dog I saw kickin' the dead horse?

SFAIK nobody on this forum thinks ASTM C 518 is an adequate test for mass-wall structures, since it's a steady-state 30F delta-T type of test. But it adequately describes the average performance of fiberglass batting IF it is installed without gaps or compressions AND has full air-barriers on both sides, even if it fails to describe what happens at the extremes, or in "typical" imperfect installations. But ASTM C 518 isn't a guarded hot box test, it's a steady state fixed-delta plate-test. ASTM C236 guarded hot box tests aren't usually done on fiberglass as a material, but is often used to characterize other materials and methods when the manufacturer is making the argument that it's performance is better than it's ASTM C518 results might imply. But ASTM C236 is also a steady-state test, not appropriate for mass walls.

But guarded hot box tests CAN be and ARE used to characterize both steady state and dynamic effects. ORNL has gone to considerable lengths to model dynamic performance of many assemblies using the large scale climate simulator and other hot-box tools. Pick any climate parameters, they can test the assembly under those dynamic conditions, and they have large weather databases to back up the parameters used for any particular location. This is a FAR better (and more complicated & expensive) test than plunking the material in an ASTM C518 plate. Maybe "most people" would reject those results, but I'm not one of them (quite the contrary!)

DBMS has some validity, even if labeling it at a fixed R-value other than it's ASTM C 518 value per the FTC would be a violation. DBMS values may seem squishy but they're still legal to use from a building code point of view, and well vetted by dynamic testing, despite significant variation in actual performance based on specific designs & locations. (The performance variations of f.g. batts is also dependent on design variations affecting total framing fraction, etc. too, much of which falls outside of the basic building codes.)

The IRC and similar codes give allowances for wall assemblies at lower ASTM C518 R-value rating provided there it sufficient thermal mass, but it still may be a higher R than DBMS of a specific stackup might be for the specific climate. To know how any of it (including fiberglass batts or blown cellulose) are going to work in YOUR specific design and climate, model it. DOE2 is a far better predictor of actual performance than anybody's marketing information, labeled R values, or DBMS, and DO in fact show the one-legged dogs that can run well if you tweak the design parameters to improve whole-house performance. DOE2 does a very decent job of modeling the dynamic effects of mass walls, (vetted by the measured performance of real-world structures) and much of the internals of that tool regarding mass-wall performance is based ORNL's guarded hot box testing. I readily accept this type of guarded hot box test, where I might question performance claims based on ASTM C236 in some dubious application of radiant barrier.

R-values of walls can be important, but they're sure not ALL-important or even a dominating factor in most designs. The relative air-tightness of concrete walls can be a larger factor in energy use in ICF structures than DBMS or steady state R, but unless the house is detailed and tested to a stringent air-tightness standard that advantage is too easily lost.

(OK horse, GIT goin', you can't fool me! :-) )

Posted By Dana1 on 08 Feb 2012 02:45 PM
This is a FAR better (and more complicated & expensive) test than plunking the material in an ASTM C518 plate.

Fair enough. My apologies for suggesting otherwise

If the mass effect of ICF is just based on DBMS and hence the conclusions that ICF is best suited to high diurnal cycles, would it not be fair to suggest that after two or three days of fairly even day and night time temperatures, the effect of the ICF mass would be lost?

So in my house when we had a week of minus 32ºc (also our design temp) nights with day time recover to only minus 26º, the mass effect of concrete should have abated. After the first two or three days, the R22 to 24 of the foam should have been the governing factor with regards to heat loss. Since the heating is designed "as if" the walls are R50, I should have been freezing by day three, should I not?

I don't think we have all the facts we need to properly discuss that scenario. In my neck of the woods. -26 means clear skies and lots of solar gain on our interior mass. Is that factored in?

Posted By jmagill on 08 Feb 2012 05:33 PM
I don't think we have all the facts we need to properly discuss that scenario. In my neck of the woods. -26 means clear skies and lots of solar gain on our interior mass. Is that factored in?

It certainly should be considered. Unfortunately we have not had much sun until the last two weeks. Just after the cold spell.

Posted By FBBP on 08 Feb 2012 05:11 PMSo in my house when we had a week of minus 32ºc (also our design temp) nights with day time recover to only minus 26º, the mass effect of concrete should have abated. After the first two or three days, the R22 to 24 of the foam should have been the governing factor with regards to heat loss. Since the heating is designed "as if" the walls are R50, I should have been freezing by day three, should I not?

So what did happen? How many days of -32C to -26C days did you have? How much heat did your heating system put out? What was the calculated Manual J load on your house? What kind and how much of ancillary heat sources do you have, like stove, TVs, computers, light bulbs, etc.? What was your house temp at the start and end of the cold spell?

Manual J calculations routinely result in oversizing, first. Second, ICF slows heat loss -- and gain -- so that the impact of a cold snap is spread over days rather than hours. But here is the rub. Once the ambient temp drops below 20C and stays there mass is pretty much out of the picture. Heat is only going one way -- out. Build an identical low mass house next door at R22 and heat loss is the same.
DBMS only works when 20C happens at some point during the day. In Phoenix mass can keep the AC off during 100F afternoons, and the furnace off during 50F nights. Or not. ORNL's multiplier is an annualized adjustment. My house does better than a R17 low mass house in spring and fall and worse in summer and winter. Over 365 days, my R10 high mass is effectively the same as a R 17.
More caveats. In my design you'd spend big bucks to keep the windows from bleeding heat. The walls are no big thing, as long as the mass is exposed to the interior.

dmaceld - I was referring to my earlier post on this thread, which give that info. 2nd half of first page.

My apologies, FBBP. I assumed you were deliberately misleading people until I reread the post you referenced above. What your hvac guy meant by R50 could be literally anything according to the fellow who wrote Manual J, and argues here that rampant oversizing isn't HIS fault: http://www.greenbuildingtalk.com/buildcentral/sip/article_hvac_sizing.aspx

If he meant actual performance of R50, he was wrong. Most likely he was adjusting your walls for thermal lag. Thermal lag is the considerable time it takes for heat to penetrate a mass wall, long enough for conditions to change on the other side. You'd get some buffering even in the dead of winter. Instead of a low of 0 and a high of 30, for example, the delay built into ICF might result in temps10 to 15 degrees higher. You can downsize hvac with ICF, and R50 was probably your mechanic's way of getting it done.

Your R value remains R22. While your walls don't cool as much as the identical low-mass house next door, they don't warm as much either. Average is average, and at no point is ICF manufacturing BTUs. But make the average daily temperature 70 and it is a different ballgame. A range of 67-73 is much kinder on the wallet than 55-85 -- as long as the mass is exposed to the inside and able to buffer temps on a 24-hour basis.

Researchers have done many actual tests on actual houses, including the side-by-side study in Knoxville in 2000 with ICF industry support. I know of no study that suggests that ORNL is wrong about ICF. (We know that fiberglass insulation isn't as good in the field as it is in testing because that research exists.) Anecdotal evidence won't cut it, even if it seems self evident to you.

Toddm-

In your opinion do you believe AAC is better product than ICF? If so, how are they better?

 

Posted By toddm on 09 Feb 2012 07:12 PM
My apologies, FBBP. I assumed you were deliberately misleading people until I reread the post you referenced above. What your hvac guy meant by R50 could be literally anything according to the fellow who wrote Manual J, and argues here that rampant oversizing isn't HIS fault: http://www.greenbuildingtalk.com/buildcentral/sip/article_hvac_sizing.aspx

If he meant actual performance of R50, he was wrong. Most likely he was adjusting your walls for thermal lag. Thermal lag is the considerable time it takes for heat to penetrate a mass wall, long enough for conditions to change on the other side. You'd get some buffering even in the dead of winter. Instead of a low of 0 and a high of 30, for example, the delay built into ICF might result in temps10 to 15 degrees higher. You can downsize hvac with ICF, and R50 was probably your mechanic's way of getting it done.

Your R value remains R22. While your walls don't cool as much as the identical low-mass house next door, they don't warm as much either. Average is average, and at no point is ICF manufacturing BTUs. But make the average daily temperature 70 and it is a different ballgame. A range of 67-73 is much kinder on the wallet than 55-85 -- as long as the mass is exposed to the inside and able to buffer temps on a 24-hour basis.

Researchers have done many actual tests on actual houses, including the side-by-side study in Knoxville in 2000 with ICF industry support. I know of no study that suggests that ORNL is wrong about ICF. (We know that fiberglass insulation isn't as good in the field as it is in testing because that research exists.) Anecdotal evidence won't cut it, even if it seems self evident to you.

toddm - no it could not mean anything. When the software asks for the R value of the outside walls he put in R50. If you can prove that the btu's produced by my hot water tanks is considerably more than 50,000 btu's than yes, he was wrong.

Mass effect and buffering, as per ornl, is only effective for diurnal swing. It might have an effect for sudden temperature changes. It very likely has no effect when there is a 100ºf delta T inside to outside for more than three days. The inside temperature stays at 72ºf. There is also a huge wind scouring effect taking place. All that stands between the mass and the minus 25ºf temperature is two and one half inches of foam. If this is a straight "R-value" equation than the 6" of concrete should be down to very low numbers by way of entrained heat in a very short time. Remember the heat loss from the inside to the mass is protected by the same amount of foam but with a much smaller delta t.

I have already given my thoughts on Knoxville. The air change rates are way to high for an average ICF home. This defeats the value of ICF. I think but don't know for sure that the infiltration problem is with the vented crawlspace but there could be many other circumstances involved. As I said earlier ICF probably just provided some financial input to the study. Just because my results don't match any report does not mean they are wrong!

This is not a huge house but neither is it a doll house. It has a basement of 1700 square feet with a 9' ceiling, one third of which is rated as walkout. It has a main floor of 1400 sq. ft. with a 9' ceiling and an attached 1000' garage which is separated only by a 2 x 6 wall. The garage is ICF with two 9 x10 ft garage doors. The second floor is 1400 sq. ft. with an attached 700 sq. ft. suite over the garage. Ceilings at 8'. All this is heated as described in the earlier post.

I may not be the sharpest knife in the drawer but I have made a reasonably comfortable living for 40 years building houses and commercial buildings in cold climate so I do understand enough to design my customers needs, put it on paper and get it past by the three or four jurisdiction I work in including the City of Calgary.

Wilcox remains leery of the whole concept of mass-enhanced R-value—not that the effect exists, but whether it can be used clearly with building materials. “I don’t know if there’s any way to make it a property of the material,” he told EBN, “It’s a property of the system.” There are a lot of questions to sort out, such as how many climates need to be modeled: are six enough, as Oak Ridge researchers have used, or do we need 20? Would such a system take credit for time delays in heat transfer, or just actual reductions in the amount of heat that moves through? http://www.buildinggreen.com/auth/article.cfm/1998/4/1/Thermal-Mass-and-R-value-Making-Sense-of-a-Confusing-Issue/

If thermal mass plays NO role in ICF, theoretically if one would stack just ICF forms (2.5"EPS x 2) and leave the core hollow, the home should have equal heating and cooling loads to that of a ICF form that is filled with 6" of concrete. I wonder what that study would show? 2 homes (one 6" concrete core vs. hollow core).

If I were a betting man, I would bet the 6" concrete core home would have drastically reduced heating and cooling loads over the hollow core home.

Lbear - I'm not saying that there is no mass effect for ICF wall assemblies. I am saying that the mass effect is very limited at consistently low temperatures with high delta t's. Heat goes to cold. Therefore as the assembly has to be colder than the room temp the mass cannot give any heat back to the room. With only R11 protecting the mass from the minus 32ºc exterior temp the mass should shed it retained heat quickly. So by conventional wisdom, all we should have left is the R11 of both layers of foam as the value of concrete is negligible. However the assembly in its whole is performing way better than this.

I think I'll let you prove to me that your house is the 1 in 100 with an accurate man j. If your mechanic strayed from the published data once, how do know what else he made up? If you are an ICF builder, how do you know the R 50 bit wasn't a pitch for your business? But as long as we are using reliable sources instead of research what does Larry the weather guy on channel 3 think of ICF?


Lbear there is no one perfect system. ICF fits penna's climate better than AAC. AAC is a better fit for yours. Stud walls are kinder to budgets. SIPs are a natural for timbrrframe. Find the site. Find the house that belongs there. Choose the most suitable building system.

Posted By FBBP on 08 Feb 2012 05:11 PM

Posted By Dana1 on 08 Feb 2012 02:45 PM
This is a FAR better (and more complicated & expensive) test than plunking the material in an ASTM C518 plate.

Fair enough. My apologies for suggesting otherwise

If the mass effect of ICF is just based on DBMS and hence the conclusions that ICF is best suited to high diurnal cycles, would it not be fair to suggest that after two or three days of fairly even day and night time temperatures, the effect of the ICF mass would be lost?

So in my house when we had a week of minus 32ºc (also our design temp) nights with day time recover to only minus 26º, the mass effect of concrete should have abated. After the first two or three days, the R22 to 24 of the foam should have been the governing factor with regards to heat loss. Since the heating is designed "as if" the walls are R50, I should have been freezing by day three, should I not?

The mistake in that analysis is that the presumption that your wall-R is a dominating factor in the heat load, whereas it is not.  In most homes with R20+ whole-wall values the window losses alone exceed the wall losses, and wall area only accounts for something on the order of 25% of the whole-house heat loss. (And that's why the mass effect in mass-walls in most climates is only a single-digit percentage reduction in fuel use, even if it allows you to undersize the mechanicals by double-digit percentages sometimes.)

Nor is the heating system very likely to be designed EXACTLY to balance at the the outdoor design temp. Most heating systems even if designed precisely to Manual-J with ZERO oversizing would have at least 15% margin,  and usually more like 25%.  If the heating system was designed even 10% above Manual-J (most are more oversized than that) most would be able to dwell for weeks at temps below design temp and still keep up. 

For the sake of argument let's say your outside design temp is -32C, and you treated the walls as a low-mass R50 based on some DBMS figure in your heat loss calc by which you sized the mechanicals (not recommended, but people do- there are better models for sizing mechanicals in high-mass houses), and in the calculation the walls counted for fully 30 percent of the total heat loss (that would be the exceptional case with an R20+ whole-wall house,  but let's just say it is.)  At a steady-state ASTM C518 R of R24 that means the heat loss through the walls is ~2x what you calculated for if you stagnated at -32C for a ~30% increase in the whole house heat load above what was calculated.

If your Manual-J was only a 15% overshoot, and the heating system was only 10% upsized from the Manual-J you might fall a few degrees short after a few days, but you would NOT be freezing.  You're looking at a heating system that's only putting out 1.15  x 1.10=  127% of the  calculated load, and you're stagnating at 130% , so your heating system is only putting out (1.27/1.30=) 97.7% of your calculated heating needs. You want to keep it at +21C, but the system can only maintain 19.8C after days of -32C.  Are you freezing yet? (I didn't think so.)  If a 1.2C undershoot is too much to handle, you could close the shades at night.

And if it actually got up to -26C with a daily average of -29C you'd undershoot by less than a degree with just the thermal mass of a low-mass house, and only briefly at that, but with ICF you'd probably coast through just fine.

With more typical 15%+ oversizing of the mechanicals above you could stagnate at design temp forever and still keep up.  (1.15 x 1.15= 132% of your calculated and your stagnated heat load is at only 130% of the calculated design load- you have a slim margin.)