# The differences between mass and insulation

| Posted in Green Products and Materials on

The “I installed more attic insulation and now my AC runs more” thread got me thinking about the mass effect a bit.

Let’s say I have a lightweight wall insulated to R-10. Say, 2″ XPS sheathing over an empty framed wall. On the outside, it’s 90 degrees F. On the inside, it’s 70 degrees F. According to the calculations I can do from the wall’s R-value, each square foot of wall will transmit 2 BTUs per hour to the interior. Easy peasy. It’s slowing down the heat transfer such that only 10% of the heat that wants to enter every hour is able to. If I wait 10 hours, as much heat will have passed through the wall as in a single hour if the wall was only R-1: 20 BTUs. From the perspective of the occupant, heat flow has been slowed.

But don’t massive materials like concrete and adobe also slow down heat flows through the building envelope? If I have a foot-thick concrete wall, it’s going to take hours for heat originating on one side to reach the other side, right? Maybe even 10 hours…

So that got me thinking about the actual differences between insulation and mass, because clearly they both slow down heat transfer, but in different ways. It seems kind of like insulation is akin to a funnel that reduces the size of the heat stream, and lets through a constant trickle proportional to the size of the funnel’s orifice, while mass makes the heat move through a tar pit–eventually all of it gets through all at once, even though it might take a long time, depending on how big a tar pit it is.

If we take both to their logical extremes… A super-insulated box would only admit or leak a negligible trickle of heat, allowing the occupant to heat or cool it at very little cost. By contrast, the interior of a cave or fortress with 20 foot thick stone walls would remain the same temperature year-round, because the walls would be so thick and dense that most of the heat transfer would take place within them, never fully reaching the interior or exterior. You couldn’t budge the interior temperature if it was uncomfortable.

So here’s my question: is the reason we don’t use mass as the sole thermal control material for energy-efficient construction because we would need so much of it that building the darn things would become impractical? Do we primarily use insulation because of its properties of being far cheaper, easier to build with, and more space-efficient for a given desired rate of heat slowness through the building envelope?

That is to say… if we had the magical ability to cheaply and easily build mass homes that kept the interior right at the average comfort levels year-round without requiring any HVAC equipment, then we would do that instead, right?

## Join the leading community of building science experts

### Replies

1. | | #1

I would think of mass as like a sponge that is getting wet. It is absorbing water and storing it. Slowly that moisture will reach the inside but at a slower rate and some may evaporate and never reach the inside. So when its really raining it absorbs water faster and releases more inside.

The insulation would be like a raincoat that sheds water but is not total waterproof. So it keeps most but not all the water out.

Now a cave is a different beast. You have the deep earth that maintains a rather constant temp. The thick mass above keeps the heat movement to a lower level than the earth below is absorbing the heat.

Now if you build a high mass home it would matter what climate you live in. If the temp had daily fluctuations above and below the desired temp then it would work, But a high mass home would not work in Phoenix or Alaska that stay well above or below the desired temps.

Most areas of the country stay above and/or below the desired temps for long periods of time. The only feasible method is is to use insulation,

2. GBA Editor
| | #2

Nathaniel,
If you build an underground house with at least 8 feet of dirt on top of the roof, and no windows, you will be building your house in an environment that has a temperature that is close to the annual average air temperature at that location.

Where I live, in Vermont, the soil around the house might be at about 40 degrees or 45 degrees. That's warmer than the coldest temperatures in January, of course, but it's still not very comfortable. So I would need to heat the house.

Once it's clear that I will be heating the house, I need insulation to keep the heat in. Mass alone is not enough to keep me comfortable in Vermont -- but it might be in Florida.

Even in Florida, however, a few windows are nice. So, if you are building above grade, you end up building a conventional house -- rather than your hypothetical house with 20-foot-thick stone walls and a 20-foot-thick stone ceiling, or my hypothetical underground house surrounded by 8 feet of dirt.

3. | | #3

The biggest difference between the two is mass is far less understood. I could post links to mass homes not only all around the USA but world that function well. In the NE many that never get below 65 no hvac, in SW hot desert rarely need cooling while all the neighbors with batting and drywall have high cooling bills. Download ORNLs CMASS calculator that used concrete compared to batt, it will give you expected hvac cost in all areas of the country. “The most favorable climates were in Phoenix and Miami and the worst locations were Minneapolis and Chicago”

Mass does not resist steady state temperature flow as well as batts since it is denser, its u-value is higher and a thermal bridge can occur depending on delta T, thickness, specific heat, and density. It has a parameter insulation does not have “heat capacity’ or an ability to store and release temperatures based on differentials. Some mass like adobe clay for example, have an additional benefit to regulate moisture. It is complex, a lot to know here….

Most often designs will create a thermal brake with structural foam in real hot or cold temps since the thickness, density, and specific heat is inadequate, like ICF. The best location is centered not to the interior like ICF, next best to the exterior wrap. An exterior foam wrap would provide more interior mass, the question is how much is needed since it can only store so much in a 24 hr time frame. ORNL test show 4” to the interior and a 10% hvac drop compared to 2” thick concrete but’ at some point there is no benefit like in a cave and a waste of money. If centered the inner and outer mass becomes independent storage systems, the exterior regulated by the environment it sees, the interior what it sees which can include solar passive and hvac or any temperate source. There is a lag time associated with regulations, ideally 6-12 hours. Some in hot climates like AZ have used night time radiation. The thermocouple graphs I have seen on the exterior and interior walls show the lag times that can be completely independent of one another, but never bridge.

It has taken me a long time to understand mass and I'm still learning, the chemistry and physics, I could write a book here….Looking at heating, cooling, humidity control mass benefits only dents the surface when comparing mass to layered conventional wall and roof systems which can be difficult to put cost saving's to, there are so many more benefits. My mass designs are showing a 10-30% reduction to the stick builds we do. I’d never use or consider commercial grade concrete or sheetrock as quality mass. Mass is not created equal, the highest quality comes from Mother Nature herself and everyone knows she is one woman that is difficult to understand at times :>)

4. | | #4

Nathaniel : mass does not change the rate of transfer of energy.
It does not have other effect ( in reasonable quantity ) than storage which then translates to stability.

You need the same amount of energy to keep your mass at your desired temp than it will seep from the exterior , it balances itself.

Mass provides with thermal storage that regulates if located close to the interior.
Where you might gain some energy is that you would be able to accept MORE SHG from windows during heating season because it will take more energy to bring the room temp to unconfortable level ( 24c+ ) .

But otherwise, mass will not save energy,
believe me , i live in a nearly 2 million lbs concrete house. :p

5. | | #5

The primary way mass saves energy is solar passive designs that take the need or hvac load down, solar and night time radiation cooling are a FREE renewable source of energy. When you isolate interior mass from exterior mass and do not allow it to bridge, the center wall temperate is less which produces less interior load on the hvac system.

Clemson U demoed that in this build using a 3 wyke brick design....look at the thermocouple graph...
http://hopeforarchitecture.com/blog/

Vermont building and school uses solar passive design in cold climates, many, net zero with clay as mass...
https://yestermorrow.org/courses/detail/strawbale-design-build

AZ builder been doing it 40 years now, net zero homes, check out the latest massive on in the middle of the hot desert at net zero while other superinsulated neighbors have high cooling bills...
http://www.michaelfrerking.com/gall_deserthome.html

Due to high levels of dense portland cement, stone and sand, commercial concrete has a low heat capacity and is not effective mass, it will bridge. I would put not 4" of EPS but 8 in the center to get it to perform like earth, lime, pozzolans, etc.....One bad design does not make the ones I posted or the extensive testing ORNL did a fantasy and there are many more builds, much, much more proof. I have talked to all the builders and universities, and their engineers above and many more. You have to do that and have the ability understand to get a real answer to your questions.

6. | | #6

The electronic analog of thermal mass and R-value is resistance and capacitance. A wall with fairly uniform thermal mass and R through the assembly (materials like brick, concrete or AAC) model like a lossy transmission line of distributed R/C filter. A wall with thermally massive element with low-R and less massive elements of higher-R (say, an ICF) will model fairly well as a simpler R/C filter. Either will have a characteristic time constant based on the R & C that determines how much of the temperature or heat-input "signal" is attuated over time when measured at different places within the assembly.

The inputs to the filters are on both sides- the outdoor air temp & solar gains vary the input/loss over time on one side, and the internal heat gains (including solar gains from windows) are the other side. The amount of heat stored in the assembly is finite- there is only so much thermal mass, and only so much temperature difference/heat-inputs being imparted on it. But in assemblies of higher thermal mass there is sufficient time lag that the peak loads that would need to be covered by heating/cooling systems are much lower than the steady-state R-value would indicate.

The notion that commercial concrete has "low heat capacity" would be a mischaracterization of the material. Low as compared to what phase change materials, mayhaps?. The distributed R of concrete is pretty low, and there are other thermally massive materials that will have a more favorable R/C for planets with 24 hour days, but it's possible to tune a solid concrete wall to minimize loads, but it's a much thicker wall than would usually be needed for structural reasons. It's generally better to just go high-R with other elements and keep the thermal mass inside that insulation envelope and treat it's mass as a lump, even though there is still some thermal time lag across the concrete itself. While it's possible to model that lag as a third order effect, in practical terms it really doesn't matter. In a building that is actually used are bigger "errors" in the "signals" coming from interior side heat sinks/sources than from the additional temporal lags of the massive elements in the wall when you have "real" insulation in one or more layers.

This doesn't take hard math to model, but it's not 5th grade arithmetic either.

7. GBA Editor
| | #7

Now that I see that this thread is turning into a tutorial about high-mass walls, I guess it's time to post a link: All About Thermal Mass.

8. | | #8

I've read that page a few times but came away feeling like I wasn't getting the full picture. Terry and Dana's additions to the conversation are doing a lot to fill in some of the details (thanks guys!).

9. | | #9

I just came from Eaton test lab in MI where I wrote hot box test plans, and developed designs we sold to clients around the globe. My designs where modeled by a team of experts including myself(CATIA), others including thermo, aero, stress engineers, chemist. I did preliminary analysis but I had to deal with manufacturing, cost, sustaining, design changes, test reports and plans, etc.....Most of the models much better than the ones this industry uses, I have seen are fairly accurate in organizations that have a long history of data to back calibrate the model to prototype builds, such as ORNL did and even after the initial release there are still changes that get very costly. These are companies with very deep pockets. For once Dana is making sense to me, :) I think we are on the same page for the most part. Do you have a test report and model or field build you can put what you wrote in your last post to prove? I don't and I won't even try the math or CFD modeling, my plan is by educated guess based on experience.

I would take the path ORNL did to model my spec homes and to be very honest I don't have the R&D budget nor team I had in corporate but, I am in the final design phase and would be happy to give a CAD model to Dana to import to his CFD or "math models' to determine the thickness of the walls I should build, and I will tell you the materials. Perhaps you can define my lag times too which would be useful in those areas that have a peak surcharges, size my heating, cooling, and humidity control, electrical loads, etc.. Perhaps we could put the whole design development out here along with your test plan I will conduct? The best way to learn is by doing?

Test instrumentation and design is critical, not to be taken lightly, so you will need to give it alot of thought, actually ORNL and Clemson U made some mistakes, but I know they both ran out of budget. The test we ran normally were in the millions of \$ per client and still struggled until we had a few test articles which I will also provide.

10. | | #10

Dana : WHAT?? i have re-re-read your post and i am still not sure i get half of it.
Told you you were the smartest kid on the block sifu :p
Your electrical analogy may seems simple to you, but it is not for a non educated as me.

When you think about the thermal lag impaired by high mass , you can imagine that it might not be
" timed" appropriately with the daytime/temperature ( probably variable with mass ) and should be variable with every situation thus is probably impossible to calculate precisely enough to be of any use.
Once any interior mass as reach interior temperature, it is then always working against all energy changes. It may reduce "peak" loads, but will also take longer to energize .

Then as you pointed out in the past, some heating/cooling aparatus are working more efficiently at part load than full load ( like minis i believe? )
That might add to the efficiency of a more stable home.

I still do not understand how to "predict" what effect it has when mass is used inside a wall assembly like in ICF ( vs interior mass that is not related to envelope walls )

Terry Lee : You are all over the place.
What is it exactly you do for living ?

Concrete as good thermal mass because it can be used as a cheap structural high mass system
( and can be much greener than some might say )

Would be superb to be able to use local granite blocs as load bearing inner wall ,
but most of us can't afford to do that ( and there is no point in building high efficiency 1mil\$ houses )

As far as i can tell, portland cement needs to be added in good quantity to any structural "rammed earth" elements to be accepted by eng. as structural.

Then, running tests in the mil\$ is good for keeping high wage jobs.

Pretty sure you could make a low scale test with some insulation and concrete models
and precise digi thermometers that could prove much of this.
You guys already have all the maths required to base the tests on neway.

Dana is probably basing his post on his personal knowledge and we've had much proofs
that he can be relied upon for this kind of technical stuff up to now !
( cheering for sifu Dana !! ) :p

11. | | #11

Jin wrote: Terry Lee : You are all over the place….What is it exactly you do for living ?
I have been doing this type of design, test, and build as a Design Project Engineer for over 30 years now, not that I know it all, as I said the subject is complex and I usually have a professional team.

I do not agree with Dana's statement that "this doesn't take hard math to model, but it's not 5th grade arithmetic either"....so I asked Dana to show this "simple math" and explain to the community as such in an existing mass building he designed, built, and tested. I have final reports (FTR’s) I could show of aircraft, auto, that are off topic, very similar to the hot box testing that ORNL did, and it is not "simple math" by any stretch of the imagination...ORNL built if I remember right 10 homes in different climate zones around the USA and produced 1000's of CFD (computational fluid dynamic models) from those builds and stated they ran out of funding in the FTR. You can read the final report on their website, I asked Dana to provide his. I know the design process since I have been doing it a very long time and I am spot on with the statements I made answering the complex questions at hand.

What Dana should have done is stayed on topic not change to electricity which is an entirely different animal, although your can draw an analogy to resistance and capacitance at the atomic level, and look at them as "filters" I guess if you want, a better approach, if the knowledge and "experience' truly exist, is talk to some of the formulas. The fact that you Jin are obviously confused, and others, should tell you something. The analogy to electricity I understand since I have the education and background, but it probably made things more confusing. A better electrical analogy would have been this,

"The measurement of heat flux is most often done by measuring a temperature difference over a piece of material with known thermal conductivity. This method is analogous to a standard way to measure an electric current, where one measures the voltage drop over a known resistor."

Anyone that has experience "modeling" these designs knows the output of the CFD model is heat flux or a vector load that is useful in many other analysis like structures, hvac, etc, let’s look at it that applies to all mass in a building, and try and simplify it: http://en.wikipedia.org/wiki/Heat_flux

Dana said this,

"A wall with thermally massive element with low-R and less massive elements of higher-R (say, an ICF) will model fairly well as a simpler R/C filter. Either will have a characteristic time constant based on the R & C that determines how much of the temperature or heat-input "signal" is attuated over time when measured at different places within the assembly."

Confusing the topic and basically true, that walls that have higher r-value or resistant to flux in, out, and accumulated (like ICF foam, batts-drywall) create more steady state analysis. Higher mass and lower r that creates more flux dynamic flow cannot be treated this way. There is no constant steady state resistance assumption that can be made, or in the real world every flux point on the surface of the mass cannot be measured by a thermocouple (that is the mistake Clem U and ORNL made, and they modified the production configuration with the thermocouple), but an integral or approximation by randomly selecting points on the surface can be accurate, or the start in a model or test lab setting to determine values for given materials that are imported to paramertize design models. It is done all the time, tabulate mechanical and thermal properties of materials for designing. The model will generate a time line or lag fairly accurately based on ALOT of R&D, in some areas the lag time can be designed to cut energy cost in half, San Diego, CA where I am at for one. I have asked Dana to show his math or model or I will provide a design so he can for learning purposes. I already know the outcome the effort and expense will not vary from ORNL and will be “complex" as I stated, far from “this doesn't take hard math to model”, as he stated.

Sounds like some need to do some reading here too: http://en.wikipedia.org/wiki/Heat_capacity

And, the manufacture of Portland cement carbon footprint and its atomic structure compared to certain soils. I'll leave that to those qualified to interpret the data, otherwise this thread is getting too confusing based on theory, complex analogies, wording, and personal opinion (not experiences)....the best way to demonstrate the design process and what Dana is trying to convey is by him showing the applicable accurate "simple" math or a model that produced an instrumented validated design and build?

12. | | #12

Jin, Terry- the simple math lives here http://www.electronics-tutorials.ws/filter/filter_2.html See also: http://en.wikipedia.org/wiki/RC_circuit http://www.tpub.com/neets/book7/27h.htm

If you substitute temperature for Voltage, R-value for Resistance (R), and thermal mass for Capacitance (C), the basic math is identical to RC filters. There is no thermal-equivalent to electrical inductance (L), so when looking up filter equations you can ignore any of the stuff related to LC or LCR stuff. Inductance introduces a temporal change to changes in voltage, but are essentially lossless for in the pass frequencies, unlike RC circuits.

If you ignore the modest thermal resistance of the concrete, and the modest thermal mass of the EPS, an ICF models fairly well as an RCR or "T-filter". If you want to refine it a bit you can include the much smaller thermal mass of the wallboard it's effectively RC1RC2. (A SCIP wall models as a CRC or "Pi-filter".)

When looking at the behavior of uniform materials, be it a concrete wall or a rock wool batt, the behavior is that of a gazillion pole filer of tiny Rs and tiny Cs. The ratio of the R-to-C will differ though, since the thermal mass of the concrete large compared to it's R-value at any thickness, and the R-value of the batt is large relative to it's thermal mass.

Since there IS at least some R value to the concrete there will be an error if you model it simply as a lump- as you vary the temperature on one side up and down over time the amount of temperature shift measured on the other side will not be perfectly sychronized to it- there is a temporal phase shift that occurs. But that time constant is negligible compared to the RC constant of EPS R and the concrete's C. The R-value of the EPS is roughly an order of magnitude higher than the entire R value of the concrete, which makes the concrete's resistive contribution a second-order effect.

When the temperature of the outdoors is varying roughly sinusoidally over a 24 hour period the time phase delays of the fundamental RCR in an ICF become relevant. The heat input requirements for keeping the room at temperature are not synchronized with the delta-T between the interior & exteriors, since the exterior RC of the RCR stackup keeps the temperature of the concrete within a much attenuated temperature range. In a low-mass wall the mid-point of the insulation runs about half the temperature difference between the indoors and outdoors at any one time, since the outdoor temperature swings are slow relative to the fairly short delay when there is little thermal mass in play. In an ICF the temperature in the middle of the concrete stays within a few degrees of the daily or even weekly average temperature difference, which takes a big whack off the PEAK temperature differences due to wall losses/gains that have to be supported by the HVAC systems.

But walls are only part of the whole-house heat load, and comprise a very small part of the cooling load. The largest benefits for ICF mass walls come during the shoulder seasons when the outdoor temps vary between heating & cooling, and the RCR of the wall nulls that part of the load to the HVAC systems. When the daily outdoor temps are way below 65F as a high, the benefit becomes a lot less from an average energy use point of view, but it does still reduce the peak loads to the heating system from wall losses to about half, but stretches it out in time. With a designer who is willing to do the math this means they can reduce the size of the equipment (in larger buildings by quite a bit), and if downsized to match the lower peak loads the equipment will run longer, more efficient duty cycles.

But in and ICF wall, since the "C" is isolated from the interior by the "R" of the interior EPS, things like solar gain from windows (usually much larger than that from walls) don't get nearly the same utility out of that mass. The "signal" isn't all from one direction. With a SCIP wall "Pi-filter" configuration there is substantial thermal mass from the 1-2" of concrete inside the EPS to limit how fast the indoor temperature can rise from the solar gains from windows, etc..

13. | | #13

BTW: The analogous aspects of RC to thermal resistance and thermal mass have been used to advantage for a long time. Before digital computation was ubiquitous cheap & powerful engineers would build & observe electronic circuits to solve complicated thermal design & temperature control issues. (Now those same issues are solved using digital simulations in MathCad or other tools rather than building circuits.)

It works both ways- sometimes observing a the thermal delays on a system are required to determine the phase delays of a thermal system in order to develop temperature controls for the thermal system that won't result in temperature oscillation. (Important when trying to control the center frequency of a diode laser to a tight tolerance, for instance.)

14. | | #14

Dana, you are quite entertaining! First off there is nothing simple about the math in the links you provided, and your trying to relate the use controlling electrical frequencies with resistors and capacitors, or filters is making it more complicated. Your barking up the wrong tree. If I had on my resume I use electronic filters to design mass effect I'd never get a job or be fired, lol!..so would you. Substituting electrical formulas for the ones like heat flux, heat capacitance, inertia, is nuts!

You still did not post a design analysis using this electrical filter method of yours, so I take it you have no build to show the simplicity of it all, nor can we conclude by the lack of your real live data as proof that ORNL made the testing to produce the C-MAS calculator FAR to complicated, as with the rest of the labs I have worked in, similar test procedures defined by ISO internationally and across the world invalid.

You are creative I will give you that, but diverting the readers to electronic is only confusing the subject and again, you have no design-build to show.

I'm done here, got better things to do.

15. | | #15

The electrical analogy is just that... an analogy. No need to ream Dana for it. Personally I thought it made perfect sense. Insulation LITERALLY corresponds to resistance, it's just thermal resistance instead of electrical resistance. Thermal mass is a reservoir that acts as a buffer, just like a capacitor does electrically.

If you don't understand the electrical concepts, that's fine. Just move on. Obviously analogies are never perfect.

16. | | #16

Yeah, the analogy is confusing to me as well, not having an extensive background in electronics. It also seems to me that ICFs are out of the range of the discussion here because the interior mass being insulated from the interior makes the whole wall assembly behave more like a super-insulated framed wall rather than a traditional mass wall that absorbs and releases heat.

It seems to be commonly accepted wisdom that uninsulated mass is only useful when exterior temperatures are cycling above and below the indoor comfort range. But here is an intriguing graph I found in one of Terry's links:

The temperature at the outside of this wall (12" uninsulated brick) ranges from 70 to 110 through the days, yet the interior temperature stays rooted at around 70. The range of outdoor fluctuation is almost entirely above the interior temperature. How is this explainable?

17. | | #17

It seems to be commonly accepted wisdom that uninsulated mass is only useful when exterior temperatures are cycling above and below the indoor comfort range. But here is an intriguing graph I found in one of Terry's links:

Nathaniel, give the director there a call, forgot his name, I had to make sure I understood the graph too and I explained it above...Last we talked about a year ago, they were pursuing a better brick from Europe w/o the portland cement, less dense, and more aerated...

You seem to have caught on quite well congrads! You know, the test data and graphs are a MAJOR chore in themselves to interpret sometimes needing alot of professionals. At times, the data can very drastically based on test set up, that is why test engineers like Clem U play a critical role. Anyway, one of the take always Clem U had you will read in their blogs is mass can work in any climate zone contrary to their believes of the past, same finding's as ORNL. Once they explain to you their version of why with brick, since that is all the deal with, lets keep it our little secret ;) Let everyone else confuse themselves with big words and creative writing's the tech writers are only good at, analogies that really don't apply when you get right down to it. Everything you need starts with the flux formula, substitutions, solving for what you want. There is NOTHING simple about it, once understood can drop energy bills dramatically, which for most people does matter.

18. | | #18

The aerated brick you're speaking of will start to behave more like an AAC block, right? Both storing some heat but also resisting its flow to a certain extent. I know AAC-type blocks are popular in Europe, as well as perforated, mostly-hollow clay blocks.

So it seems quite possible to turn a masonry unit into an insulator, with some creativity in trapping air. These types of products seem easier to model: you just treat them as weak insulators (R-1.5-2) and ignore their mass effects! :P I'm still interested in actually quantifying the mass effects, though. It definitely seems like there's a dearth of work on this subject compared to highly-insulated lightweight wall systems, where our models can do an almost perfect job of predicting actual heat flows and energy use patterns.

19. | | #19

Exactly, it work better in the envelope where you want insulation. ORNL did make an attempt to quantify mass effect with their DBMS (Dynamic Benefit Mass System), note the words "Dynamic" and "Mass Systems" you'll find in the software of the C-MASS calculator. BCS hot box testing quantified how r-value is trumped by air in 8 different wall assemblies is another attempt to quantify or place value on, both have a LONG way to go...

I'm using a natural plant insulation in my envelope, stick for structures, designs that will behave as an AAC big block of sorts, 2-4" plaster as mass, isolated stucco by the center core insulation that does not have a high DBMS. Internally dense packed mortar non-insulated, I want to bridge. There will be no trumping of r-value or mass effect in my designs and there will be no sealing material or labor cost other than fenestrations, that is how I get my cost below mainstream and ICF, no foam or toxins , and some other secrets :)

20. | | #20

Dana: this second time, i was able to understand it much better!

Dana and Terry : sometimes, you big brains type like we know all of your sutff, which is not at all obvious for uninitiated .

Terry : damn i have hard time following you ...

I think we all understand that everything is much more complex than it seems,
but this is a "green building" community , what is important here is to help
others understand how to build better, what is good and what is not.

The discussions on MASS and INSULATION is a very important one,
and i am sure everyone here is happy that you are participating with all your knowledge/experience.

Now could you please translate it all to a "simpler" applied text ?

Nathaniel : mass and insulation is a trade off ,
more air bubbles = less mass ...there is no secret sauce here
And beware of using a simple linked graph as a universal answer.

Terry again : could you please share in more details your " absolute" wall system ?

Dana/Nathaniel : ICF does not mean you do not have OTHER interior mass.
My house envelope walls are all ICF ( excetp for the roof exit room which is a "test" of a REMOTE type system )
But all my floors and a few interior partitions are concrete and steel .

Last one for TERRY :
why is it so complex to test the effects of mass in walls and interior of buildings?
Why can't some small scale tests be used to determine what performs better etc.. ??
I am a fan of small scale testing with some prior planning of course,
often saves plenty of calculation and more precise testing .

Then, how hard it is to "plot" some temperature exchange in an excel file
based on simple assumptions for lets say hourly states ??
Do not tell me we cannot simply model thermal exchange for simple 2d materials
such as an ICF wall ??
As Dana pointed out, 2nd order or less does not matter much in " house " type buildings.

Dana2: predicting heatloads for sizing equipment to run at the most efficient point is possible ?
how do you plan for all of possible temperatures ??
adding thermal mass delays to this is even more of a mess

21. | | #21

What is this you linked :

This is what was discussed here?

Probably all readers understand the reason to use natural materials and neutral products such as mortar/bricks , but seriously in 2014, how would one expect such a building to perform in terms of efficiency.
If it would be doubled in bricks with an insulation in between, be it foam or rockwool etc..

Even in this very neutral weather of Georgia.

Can't imagine how a "brick" building would feel during winter up north.

Please Terry explain what is to understand from this website/building ?

22. | | #22

The exact thing they're describing is a great example of why I started this thread, because I don't really understand it either.

I grew up in a structural brick house built of uninsulated double-wythe walls with interior plaster (not drywall) in the upper midwest. The house was terribly cold in the winter and boiling hot in the summer. My parents, who still live there, report monthly utility bills of \$400 a month or more. Then again, knowing what I know now about all of these sorts of things, I also know that the house is easily over-glazed by a factor of two (and the plurality are on the north face, go figure). That the windows are terrible and have been in dire need of replacement for decades (80 year-old single-pane windows that leak air like a sieve and are nearly impossible to open or close). That there is nowhere near enough attic insulation. That the basement is totally uninsulated, and often floods. That the gas furnace is decades old and operates at less than 70% efficiency. That the AC unit they had installed when I was a kid probably isn't even SEER 10 by modern measurements (Don't think they even measured SEER back then). Lots of "old house" type problems.

It's difficult for me to envision that adding a third wythe of brick to the walls would have make such a huge difference to my childhood home, but I'm willing to believe that properly addressing the other listed problems could have had a big effect. And I'm willing to accept the HopeForArchitecture folks' reports that the house they've built, which doesn't seem to have all of these problems, is a very comfortable house. Still, it would be nice to be able to mathematically quantify why it is a very comfortable house! That graph of indoor temperatures consistently below the lowest recorded outdoor wall temperature in summer still kind of has me scratching my head. Is it all just the ground-coupling effect from having an uninsulated basement? If so, won't that hurt them in the wintertime? Won't humidity in the basement be really high? I'd love to see some actual reports of energy bills, interior temperature and humidity readings, that sort of thing. The people involved in this project don't seem like the kinds of folks who wouldn't have thought a lot about such things.

23. | | #23

Sentimental and efficiency doesn't go far together.

By definition, a house should be comfortable, point.
If it is not, it failed.

Old buildings, old problems and usually expensive to fix.

This type of thread is why i come here.

and this mass/insulation relation is a swifty subject that will probably take more time than i have to be able to understand.

The important is to limit learning time to basic knowledge required to use it for the intended purpose.

But Nathaniel,
go back to basics, nothing is magic.

We want a comfy interio temp which is around 21c for everybody.
Exterior temperature is an uncontrollable factor, so you need to work against it to maintain proper interior temp.
This Insulation which slows the thermal transfer is required.

Mass can be used to change the timings or storage or whatever purpose Terry and Dana will attribute to it, but you still need insulation .

Make a 1ft cube box out of 1" steel plates.
Same box out of 1" EPS or XPS insulation
Same insulation box with a steel weight inside 2-4 lbs prob.
or a water recipient of a few liters.

measure and log hourly temps exterior and inside the box

i'll do it soon just for kicks
this winter ( since i'm interested solely on heating climate performance )

add energy measurement with thermostat and heating device to maintain 20c
and then start playing with different wall types

much much fun with the kids ( ok the kids won't probably be very interested in more than putting the boxes together :p )

24. | | #24

We were discussing this on another site a while back, and an Engineer claims and sells a “smart” mass controller that has thermocouple inputs and a program that understands the lag time for loading, unloading, mass. The design used radiant floors for mass, solar thermal and passive. The software has a “phase lead” that shuts down the HR before it got too hot or turned on before it got cold. She claimed the sensors on floors from south facing windows understood heat from solar diffused or direct reflected light, or radiant heat and stack effect in the whole house. Some large claims of quantifying mass effectively and controlling it with just surface mounted thermocouples I don’t buy but who knows maybe she is successful. When asked for proof she had none, just a fancy description and award winning paragraphs and writing skills. She said she designed the systems by some complex math she was unwilling to share, what she did share was inaccurate. I told her you need some air sensors. Mass gets much more complicated when you are phase changing and distributing it to various parts of a HR systems, now you are dealing with phase changes from water to mass, air to mass, at different rates depending on where the heat input sources are from the active system and passive solar.

Some may not like hearing the truth here, but she was right about one statement she made. She is not in business to sitting out on the internet and educating or training everyone on the Engineering principles behind this, especially when they do not have the education and experience or prereq's. Degreed professionals like her, Dana, Martin, me, etc. do not always agree on Engineering principles or methods to design. You can take ten Engineers in a room and they all can have a different approach. At my last assignment in a test lab I’d sit in telecons with 10 around the world trying to figure out why seals were failing. I asked the CFD-Structures engineer to model the seals based on the limit test we had which included heat and pressure cycles, and model a lateral torque with the heat and pressure applied to seal in a hot box. After we listen to more than an hour of BS theory and opinions, the model revealed the issues proving we did not need some complex fault tree cause and corrective actions, or any more telecons and emails---WOOHOO! The test later validated the model accurate problem solved.

When we go to production and the field, since the test and model can only simulate so much of the actual global environment, and we fail, we have data to help us understand why. That is why quantifying is important, that is why there is no simple answer. If you think it is simple and you have the education and experience to make it such, then by all means have at it, design and build a home, see what happens after lots of expense. Who knows you may get lucky.

The Engineering drawing is the medium or bridge between those that are Engineers and Scientist and those that are builders. A good Engineer will obtain the input from builders to produce the drawing. That is often difficult for Architect’s in this industry since they are not usually part of the same company, and is a requirement in most factories. I would not try and fully understand or do a skill set I am not qualified for, I leave that to the builders. Some of them have great knowledge but fail to see the complexities or other aspects of the design that have to be satisfied, like structural loads, systems integration, they know their piece of the pie and for the most part that is where it ends. Same holds true for Engineers that fail to see the build operations due to a lack of experience. A good design-build Project Engineer works out the differences.

If you want answers call and talk to designers and builder’s that have had success putting their designs methods to a build. That will vary. Mine do not fall into any of those builds, some closer than others I follow, so as I said I do not have all the answers. I will build a prototype spec home test it within my budget, then move on hopefully with lessons learned to a couple more in hopes of only needing two. That is all I can do. Mass is a risk, so is insulation just read of all the problems out here with super insulated, passive, they are not perfect nor is mass. ORNL, BCS, Clem U, etc..are working to quantify products and produce design guides, read their publications thoroughly, and if you think they are over complicating the subject and you can save them large amounts of money with your simplified plan by all means do the industry some good, call them don't discuss it out here. If you still do not get it, hire an Engineer or go pay a college to become one.

25. | | #25

Terry Lee: I never said it was simple math, just that it isn't hard math. Most people who made it through a decent high school math program can handle it, but not if you did the bare-minimum for graduation.

And the RC filter model analogies are actually quite good until you start looking at things like phase change layers or dramatically non-linear R with performance. (The non-linearities of actual resistors & actual capacitors can complicate real-world performance modeling of the electronic behavior too.) In the few instances where I've used exactly that sort of modeling to solve precision thermal control problems I wasn't fired (probably because the solution worked- just lucky, I guess ? :-) )

The signal to noise on the actual vs. modeled heat inputs renders any precision in the modeling moot for this type of application. Are we really going to get into the business of modeling north facing walls differently from those that get sun, or modeling the emissivity of the paint? Yes it makes a difference, but not a difference that actually matters.

26. | | #26

Well Dana sounds like you are the road to taking mass to next level of surviving a nuclear holocaust and e-bomb with your resistors, capacitors and filters! I'll just be happy to get to net zero, and with luck survive a tornado :)

27. | | #27

Terry: are you trying to end this discussion ? cause it feels so.

I do not care much about discussing with "high ranked" engineering at ORNL as they'll probably have not much interest to discuss with me.
But you are HERE still typing which means you are interested to share,
because the way you talk about this now is like you have no more to learn from this discussion.

There is only 1 perfect solution to everything, and we can only strive to choose the best compromises.
I believe we agree that for low cost buildings ( houses, small energy buildings )
separating the large influential factors/systems from the minute one is important
to chose where to invest money and time.

Mass and insulation are 2 key parameters.

Why do you say super insulated buildings have many problems?
What are you referring to ?

Dana: stop robbing me with the low-e paint :p
I am at the base of the learning curve and at least i'm asking questions!! :p

Terry: ah yeah ..sorry but i do not see any down side to mass except maybe its price or carbon imprint.
But with the recent info about how something like 19 of the world's largest container boats are polluting as much as all of the world's cars ... now sure i'm interested in carbon anymore that much.

28. | | #28

Hmm, so hempcrete is vapor-permeable, non-structural, and insulates to R-2.5 per inch. What's the advantage over mineral wool boards covering a 6" CMU wall? Is it really cheap or something?

29. | | #29

Jin Wrote: Terry: ah yeah ..sorry but i do not see any down side to mass except maybe its price or carbon imprint.

As I said earlier cost is less....

So lets compare some mass insulation to batt insulation for kicks and grins....How about hempcrete and typical fiberglass batt or blown in insulation, or cellulose.

Both need frame work to survive so we will call that a constant.

When batts get moist the r-value drops, it takes alot of seals, sealants, gaskets, most made of toxic materials that produce vocs, some produce carbon monoxide the worse since it prevents oxygen from getting to the blood, others produce carbon dioxide which will keep you from breathing and can create respiratory issues, out-gassing acids, etc... We all know about the toxic blowing agents and fire retardants in batts.

At a minimum a redundant seal plan is needed to equal mass, common today is peel and stick ZIP outer moisture-air barrier that when moisture-pressure-temperature cycled will eventually fail (there is no data on fatigue life data on the website, it could be 5 years and most builders will not follow up so who knows) A redundant outer insulation (foam wrap) , or sealant-seal (SPF, elastomer) is needed at the wall studs-sheathing, sills, windows, doors, etc, and perhaps some means to protect the sealants or seals from the elements and fatigue cycles to be fair. Lots of labor hours and cost for vocs and a high carbon foot print. Who knows how long the seal will last and the critical r-value.

Now look at hempcrete or other designs made with lime or calcium oxide, clay. Here are some images.

Unlike batts, hempcrete reduces racking of studs and is structural. The calcium creates a strong bond to cellulose hemp and wood framing, an automatic air seal requiring no other trades or cost. As the calcium bonds to the cellulose depending on silica content, it becomes hard as a rock, batts do not get better in time. A 12 inch thick wall yields about R-30 with some mass effect, batts have no mass effect and degrade.

If lime sees moisture it hydrates or phase changes from an oxide back to a rock as it re-cures and this will happen for the lifecyle, it need to carbonate to do this so it absorbs carbon dioxide from it's surrounds like a plant and is -CO2, so in that sense it self cures any cracks, is hydrophilic but, not as much as certain clay soils that love water and regulate it too, unlike batts. Typically, a lime stucco and plaster is used, windows and doors will bond to picture framing to create what I like to call "air fins" secondary and ternary air seals that sustain much longer than seals and sealants used in stick construction to protect batts. It would be a sin to use an air barrier, so you have to work with your AHJ that can show intent is good, depending on code. The reason it needs to breath is due to the high surface area of the calcium and/or clay that continues to regulate moisture and air as described, just as they do in nature. Toxic fire retardants? Not necessary they do not burn and meet a 1 hr burn rate with flying colors.

Joe L said it best,

"What I have been more or less able to figure out is that the 0.6 [email protected] Pa doesn’t come from any energy conservation rationale directly; it seems to be based on the need to prevent moisture problems in highly insulated building enclosures. That is the argument for the number 0.6 [email protected] Pa as I understand it. Never mind that that the number, in itself, makes no sense as you can easily design highly insulated building closures without moisture problems that are not anywhere that tight."

While some think it makes sense to spend extensive time and money sealing up every stud bay to protect batts, it makes no sense and the cost and trades are not required with certain mass. If you get into mass you will see there is nothing simple about building with some of the methods I listed, it takes years to understand soils, lime, plants, and natural superior methods to factory products. You are the factory if you choose these methods, and the Quality Control. In natural highly effective mass production the builder decides what toxins will be in his builds, takes responsibility and cannot point at a factory when things go wrong, nor some ASTM BS test with no quality control, or any lieing data sheets from a factory that is not controlled or verified by indi third parties. You'll also find to succeed you need to be very educated in order to explain the differences to inspectors, realtors, appraisers, insurance agents, clients, etc... You will need a PE, Chemist, Geologist, and lab.

That's the difference, choose your poison. :)

30. | | #30

Regrettably I missed the "hot and heavy"stage of this discussion while I was on vacation. However I find the topic both fascinating and poorly understood. My interest in the topic is arose from my desire to use a "masonry heater" in my house. In a masonry heater or rocket mass heater wood is burned in bursts (for clean and complete combustion) and the resulting heat stored in thermal mass. Fundamental to making a masonry heater useful is the ability to " slow down" heat transfer.
Dana mentioned the R C delay network but didn't elaborate very much. In my quest to "design" a "better" masonry heater I realized that materials have a "thermal time constant per unit length" that is simply not discussed, it is found by multiplying density by specific heat and dividing by conductivity (the results are hours/ inch in English units) FWIW dense firebrick is about 2 hour/in and ordinary concrete about 4 hours/in. I find it very interesting that the thermal time constant gives a pretty good explanation why masonry heaters have the wall thickness they typically have.

31. | | #31

That's exactly what I've been looking for, Jerry! What are the units you used for these things? Would you mind working out an example for us?

32. | | #32

You can mix it structurally or with a higher r-value depending on the ratio of cements (calcium, magnesium, pozzolan's) to hurd (cellulose stem of the plant) . Right now, it is expensive since Europe is the only supplier, but, some states are in their first growing season. If you are a chemist, you can do a DIY binder as I have, or hire one, to get your cost down, otherwise, shipping both hurd and binder from Europe can cost 2-3 time the wall cost of stick construction. This is where education pays off, and having an E&O insured LLC. :)

Mineral wool is one of my top choices, yes the hempcrete performs like it in it's ability to wick and condense water like a rock, but mineral wool in man made from iron oxides and a different product than a calcium oxide fiber reinforced hemp. You get compression strengths on the order of 40-100 psi, while MW is much lower not sure 10 psi. Hempcrete also has a higher tensile strength (depending of hurd properties) and reduces racking as I said, MW does not. In Europe and the USA the PEs have lowered structural compression and racking loads, it depends on the PE and how conservative they are, some may use it to create a positive margin of safety, but in general it is consider "non-structural" until the industry has test standards. Right now the USA hemp industry is in turmoil due to lack of material standards, but that will not stop me. Right on the r-value depending on the amount of hurd, more means more r-value, less means more structural value. Can you adjust batts that way? ;)

33. | | #33

Jerry, you are going to need to run some program, software, or lots of math for a masonry or rocket heater to determine the rate at which the mass absorbs heat per length of plumbing (yes as I said specific heat, density, thickness or heat capacity) will be some of the parameters. There will be a phase change from a gas to solid that will be complex to model. Most guess at some length of pipe and mass to produce the clean burn, plenty of examples out there. But if someone can demo an analysis for you, that you can verify by build, I too would LOVE to see it done?

34. | | #34

If I were to try an RMH it be made of ratios of lime to absorb CO2, clay, pozzolans, I'd never find a value for specific heat and need a lab if wanted to spend a bunch of \$.

35. | | #35

Finding the input data on common materials is a challenge then doing the units conversions. Basically I found that common materials used as insulation and thermal mass all have remarkably similar "time constants" ranging from 1 to 4 hours per inch. I found data on perlite, perlite concrete fire brick of several densities and common brick and even mineral wool, they all fall in the 1 to 4 hour/inch range. Liquid water is very different with a time constant of 1/2 second/ inch and steel is even faster at about 1/10 second/ inch. To store lots of heat economically, liquid water is far superior because of it's huge specific heat.

What I conclude is that masonry heaters and rocket mass heaters are not designed analytically but built and evaluated. There are empirical "rules of thumb" that, sort of, predict performance. A rocket burner heating water that is pumped into a large well insulated storage tank has great potential to be the winter heat source for a tight well insulated home. Masonry heaters and rocket mass heaters may work but will always have greater temperature variations during use.

36. | | #36

I’m getting ready to fly from sunny So, CA home to my next build climate zone KS tomorrow, here is a quick response, more later if you want? Input data on materials is a small part of what needs to be quantified here. Flux loads to the materials is more complex in this design. The fluid dynamics in a vessel such as a straight duct produces a diminishing turbulent flow and heat at the boundary layer due to friction, laminar high velocity cooler at the center, you would need a model or test to determine the loads along the length of pipe that will depend on the diameter and shape of the duct. A divergent duct will decrease pressure and increase velocity not ideal for the duct, convergent duct the opposite good at the heat source and duct, the reason why a rumford mass heater is so efficient but different ducting than a RMH. So you need to determine a load for starters. If you want a high transfer rate of air to the mass look at an aluminum duct and do use galvanized zinc coated steel, do use any chem heat treat such as passivated either, pure aluminum is best 1000 series. Nothing beats water in specific heat a given. Perlite is a non-structural insulation and expensive. If you want to insulate heated gas flow, not sure why, use it, plan on a smaller high pressure, longer duct to get the clean burn. You want to phase change the hot gas to solid mass and emit it to the interior space. Most will not take an analytical approach here it is too complex when considering heated air flow and flux input loads to materials, rather build a test mock up or just hope for the best, same holds true for mass walls, floors, and roofs.

37. | | #37

Nathaniel G,

If it is equations you want, take a look at this. I think it is a fairly good primer for the topic at hand.

I've worked through it making my own examples in excel up to, and including, transient heat conduction, and it is pretty straight forward. Once I started to try and model my example slab dynamically, there is a lot more to consider and it became tough sledding.

While the concept of amplitude reduction due to thermal mass is fairly easy to grasp, modelling an ever changing average temperature across a slab as a sine wave in excel is proving challenging.

The trick, as I understand it, is to predict the heat flux at the interior surface of the slab (wall, etc) as the outside temperature ramps up and down, around some mean temperature. The closer your indoor design temperature is to this mean temperature, the better, and of course, we want the outside temperature to behave like a sine wave around it (the mean temperature). Ideally, the variation, or amplitude of this wave is significant, and behaves like a sine wave over a diurnal cycle.

Oh yeah, then we get to factor in energy stored in the slab and how this affects the response time of the mass.

The article also explains why a 4" to 6" slab is the "sweet spot", and highlights a few other key concepts.

I did not see the value Jerry mentioned, though I do see some equations that may interest you.

Thermal Diffusivity, an indication of the speed at which a temperature profile moves through the mass (slab), f^/hour.

It is Conductivity / (Density x Specific Heat), or k/pc.

The other equation appears on page 16, and is explained on page 17. They define it as a measure of time for the mass (slab) to thermally respond to change.

Hopefully that is of some help.

cheers.

38. | | #38

Ahhhh, now we're getting somewhere. Thank you, Jason! What a goldmine of information!

39. | | #39

Jason,
Dynamic thermal analysis is not simple or straight forward as you have found. Such analysis simply must rely on calculus and differential equations. Transient analysis of distributed networks is taught in electrical engineering as senior year or graduate level courses.
What you found called "Thermal Diffusivity", is the inverse of what I've called the thermal time constant per unit length. If you analyze the units you'll find that temperature is not in the result but it is a simple velocity ie units of inch/hour. As Dana pointed out the electrical analog is a distributed resistor capacitor network or transmission line and as such it does have a "propagation velocity". but it also has attenuation and "filtering" properties. A step change in temperature on one "side" will result in an exponential ramp on the other "side" after the propagation delay. The initial slope of the exponential is a function of the time constant and the input magnitude. Also the final output temperature will not occur till several additional time constants have elapsed.

40. | | #40

Jerry,

So pc/k gives us a constant that is independent of unit length, or temperature - basically a material property.

Normally, I would call this a coefficient, similar to heat loss coefficient that we use to multiply by a given Delta T. However, I'm wondering if that is a mistake.

To get down to examples, using the link I provided, concrete has a thermal time constant of (145*0.19)/1.1, or 25.06, or 25.06 inches/hour.

Following that line of thought, how is temperature applied? My understanding of heat flux is that delta T is the driving factor. However the use of the term constant implies that this rate is unaffected by anything, that it just is (you used the word velocity). What, if anything, affects this velocity, and how is it applied in our above example. Is input magnitude a fancy word for Delta T? If not, what is it in this context (heat moving through concrete)?

Also, when you say that the final output temperature will not occur for several additional time constants I take that as:

0.3 L^ / (k/pc) (again from the article) is the time it takes for the output to occur, or rather to reach it's full value. They break it down into 3 smaller time constants which achieve percentages of final value (36.80%, 86.50%, 95%). Are we just talking about the same thing?

Care to shed some light, hopefully extending our concrete example?

Jerry, you are right, this wasn't in my grade 9 math class...

cheers

41. | | #41

"Following that line of thought, how is temperature applied? My understanding of heat flux is that delta T is the driving factor. However the use of the term constant implies that this rate is unaffected by anything, that it just is (you used the word velocity). What, if anything, affects this velocity, and how is it applied in our above example. Is input magnitude a fancy word for Delta T? If not, what is it in this context (heat moving through concrete)?"

I see three questions. 1. The velocity is a property of the material, it may change if there is enough temperature change, such as a phase change, obviously this really adds complexity.
2. Temperature difference is the "driving function" and analogous to electrical potential (voltage). Or yes input magnitude is tech-speak for delta t in this case. .
3 . Heat flux is the result of temperature difference across a material and analogous to electrical current.

The change in output temperature over time follows a an exponential curve. This means that it's rate of change gets ever smaller as time passes and it, theoretically, never reaches the exact final value but gets ever closer. After one time constant' worth of time it is within 36.8% of final value, then after two time constants it is at 86.5%, three time constants get to within 95% etc. So yes I agree with the paper's description.

42. | | #42

Jason,
The paper is excellent but the problem can be simplified from the three dimensional case described in the paper by the use of boundary conditions to a single dimensional case. Basically this simplification assumes an infinitely large plane with no heat flow in two directions so that all heat flow is in a single direction. I use this simplification and the result is the "diffusivity" becomes the inverse of linear velocity. Diffusivity is the inverse of a "spherical" wave velocity propagating in three dimensional space.. In applying the simplification the spherical velocity is divided by pi to convert to the one dimensional velocity. so the "time constant" for the dense concrete becomes aproximately 8"/hour .

43. | | #43

As much fun as it is to analyze this stuff to the nth degree, in practical terms anything beyond a first-order approximation is silly, given that it's a house, with many other unconstrained factors and inputs. Real walls have corners, materials are not uniform, all houses leak air, and any house that you live in has other thermal inputs.

The DOE2 tools make lots of simplifying assumptions and use really crude thermal models, yet it proves to "good enough" to hit within the error bars of real-world variations for annualized energy use. You might want to use something better than that to stabilize the optics of a large telescope to within a 1/16 of a wave of some particular color or something, but seriously, for estimating the heat loads and annualized energy use of a house it's bonkers.

44. | | #44

Jerry,

Thanks for the additional responses. I was wondering what the velocity would be until you divided it by ~3 - which makes sense to me now, if only in a general way. It would appear that concrete conducts heat quite fast, and that even a small portion of a slab being radiated upon by sunlight would quickly increase the slab's average temperature. Quicker than I would have guessed.

By the way, your explanation of the time constant concept, and how it approaches infinity, etc, is better than the one in the paper. Thanks.

Dana,

I'm starting to come to the same conclusion, given my own personal context (modelling a slab on grade home), though I would hesitate to use the word bonkers. My goal was to get a better understanding of thermal mass, the cause/effect of thermal lag, and the opportunity presented by amplitude reduction.

For me, it has been instructive to delve in to the sundry details. We see so many rules of thumb, standard assumptions, etc, it is nice to actually take a look under the hood so to speak. While I would still classify my understanding as basic, it is significantly greater than it was two weeks ago. I may still continue to plug away at it, out of self-interest.

cheers.

45. | | #45

So what is the conclusion of all this on and off discussion ??

Can we get a definite idea of the impact of mass withing wall systems ?

and
Is it really necessary to discuss the effect of mass within a building ?
Can you really have too much mass within the thermal envelope ?

46. | | #46

Terry Lee,
I you are still reading this thread: Facts! Heat transfer from a gas to a solid does NOT involve a phase change as you claim in post #33.. There is simply no "plumbing" needed in a rocket mass heater using a masonry bell.
If you are interested in rocket mass heaters you might want to visit: http://donkey32.proboards.com/
Where you'll find sufficient experimental work to assure a successful first try build.

47. | | #47

Jim,
The link Martin gave in post #7 is really a good summary of the effect of thermal mass on a building.
From it I would answer your question with a simple YES! Because too much mass means changing the interior temperature is slow and requires more energy.

48. | | #48

Jerry : why would you want the interior temperature to change ??
are you referring to mass within envelope or wall system ?

for internal mass...
It will require more energy input to change the temperature, but then it will dwell much longer before it goes down.
I thought that stable temperature what was we are all going after ?
You can accept more SHG if you have more internal mass without overheating,
which is then released longer through the evening/night than if lower mass.

as for the question ...
I was more referring as to how to plan or model for wall system with high mass effects though ..

49. | | #49

Although the horse is long since dead, here's a picture that might help:

If heat is water flow, brick/concrete takes a long time to "fill up"
of heat will keep flowing. Wood-frame and similar light construction
has less capacity and equally dismal R-value, so you'll not only
feel similarly cold but you'll do so sooner. Polyiso puts the squeeze
on heat escaping but has almost no *capacity* on its own. [Leaving
out its downside that its effective "pipe diameter" grows a bit in
really cold weather, arrgh] Then you can put these together and
conceptualize wall assemblies. Put your brick wall on the interior
and you'll have a big internal mass to keep temps stable, but still
lose less total heat over, say, several days with a cold average.

_H*

50. | | #50

Long flight yesterday, read the report again Jason posted along the way and am still reading it, been a while since I read these reports. Been trying to keep up with the thread too.

Jason wrote: Jerry, you are right, this wasn't in my grade 9 math class.

Yes it was, you were too busy pulling the girls pig tails in front of you. I ditched class that day. Dana was the only one that paid attention….hahaha!

Guess your HS program sucked like mine or you did the “bare minimum” I ditched class again.

Dana wrote: I never said it was simple math, just that it isn't hard math. Most people who made it through a decent high school math program can handle it, but not if you did the bare-minimum for graduation.

LOL!

1983, oldie but goody! Take note of the MANY assumptions made to simplify this complex subject for the "average reader" not requiring an extensive technical background, that is the audience. So if you struggle here (as I am) and obviously others are, it only gets more complicated. Most of the time when involved in a project like this there are very qualified people around, and lots of meeting’s. No one person will have the answer, but everyone on the team will have the expertise to contribute to it, even then sometimes the answer is not found or quantified, many times due to budget or schedule. One thing that I have experienced at the beginning of the programs are if the time to establish accurate loads (in this case flux) is not taken usually on a “load cell” or hot box test, than all the math and models can become inaccurate very fast. There is usually a lot of back calibrating the models to the lab and field builds too.

Here is where you can find a wealth of info too, updated 2001.

By the time you read all this if your head is not spinning at least a little you must be Albert Einstein the 2nd, or if nothing else in your own mind. I took this stuff in Thermo-3, Physics 4, Calc 4, Aero 3, Fluid Dynamics 2, in college. Appears to me by the lack of real time home designs demos on the thread (and that includes mine) I have asked for many times, ORNL has the best and has done the most research and there is NOTHING simple about it. I have admitted I do not have all the answers, I know better than to call this simple. I have participated in hot box test like this on mechanical parts, but never walls and roofs. The only way you will ever get a full understanding is to be involved in the design-build program like this and involved in the full(not water down) test report, or, design, build, or test your own. In order to have the opportunity to be in a test lab program like this you have to be hired and will need to be an Engineer, mechanic, or program manager, or perhaps volunteer if allowed.

I’ll go through the thread and make some comments if the participants don’t mind giving me the time to catch up, excuse my writing skills, although this old man is all about practical real life applications, not theory these days, I have not been a college boy for over 30 years. My build methods will be mass so I’m interested, since I like all the parameters I can design to. I do wish it was easier to quantify, and perhaps programs like Chief Architect had a workbench program better than ORNL C-Mass including more than concrete, like hemp, earth, natural materials. CMASS is the best model out for comparing batt insulation to mass.

51. | | #51

Jerry wrote:Finding the input data on common materials is a challenge then doing the units conversions. Basically I found that common materials used as insulation and thermal mass all have remarkably similar "time constants" ranging from 1 to 4 hours per inch.

My comments: From report: The thermal diffusivity is an indication of the speed at which the temperature profile moves through a wall. It has typical units of m2/s (ft2/h). An exact, analytical solution is available for this case. However, it would not give the reader much insight into the behavior of a wall and therefore is not presented. It does confirm that the quantity L2/a determines the time required to achieve steady-state conditions. The greater this quantity is, the longer the time required to attain steady-state conditions. A wall will seldom experience a step change in its surface temperature, but there are periods during which the exterior surface undergoes a ramp increase in temperature. Therefore, walls experiencing ramp temperature increases on one side and a constant temperature on the other side will be examined next. This situation is shown in part (a) of Fig. 2.4. The heat flux a t the inside wall surface is given in part (b). The heat flux calculated from the steady-state equation is also given For comparison. As can he seen in the figure, the actual heat flux lags behind that predicted by the steady-state equation. Taking an arbitrary value of heat flux, q , the time at which the steady-state equation would predict this value of heat flux, t,, is earlier than the time a t which i t actually occurs, t,. This difference in time is referred to as a “time lag.” For a single-layer, homogeneous wall, the time lag is less than or equal to L2/6a. When the transient first begins, there is no lag, but as the transient continues, the lag progressively becomes larger approaching the value of L2/6n. The greater the value of L 2 / a , the longer the time lag. The time required to reach the ultimate lag time is also of importance. The actual lag approaches L2/6a exponentially, and, therefore, would take an infinite amount of time to reach it. However, for practical purposes, the final lag time is reached quite early. To determine a time when, for all practical purposes, the lag is no longer changing, another concept is introduced. This is the concept of a “time constant.” The time constant is the time required in a transient for a value to reach 36.8% of its final value. After two time constants, it has reached 86.5% of the final value, and after three time constants, it has reached 95.0%. Therefore, after an elapsed time
equal to three time constants, a transient is essentially complete even though it theoretically continues forever.

My comments: A “time constant” is applied for practical purposes since the actual lag times varies exponentially and is infinite in reality. If you look at Fig 2.7 you should start getting a feel for the difference between steady state lag time and actual from mass and how using steady state models can yield inaccurate results that can cost a lot of money in errors.

Table 2.2 shows thermal diffusivity of some materials that varies quite a bit. Table 2.4 thickness for a 1 degree change or BTUs/FT2/hr that varies a lot.

Take note of the sinusoidal wave or frequency at the “wall surface”, it you read through this you will see there is no filtering of unwanted frequencies like an electronic device does, rather, a simplified pulsating path that is shown for the heat flux in two dimensions, x-z, at the surface, more than two dimensions are needed. The heat flux through the wall initially will be influence by these surface pules that affects lag time. The made some assumptions about steady state ramp ups. AT peak load or amplitude no more flux can be added, doing so would be counter-productive if the wall is experience and heat loss so, there is a limit on thickness and adding mass where say in the case of concrete can be an expensive loss in money. That is where a good analysis will pay off, in max effective thickness. Adding 1" to walls, roofs, floors, of concrete with no effective can cost significantly.

Jason wrote: While the concept of amplitude reduction due to thermal mass is fairly easy to grasp, modelling an ever changing average temperature across a slab as a sine wave in excel is proving challenging.

My comments: Don’t knock yourself out, use a fixed amplitude and pulse like ORNL did.

Jerry wrote: What you found called "Thermal Diffusivity", is the inverse of what I've called the thermal time constant per unit length.

My comments: Time (t) has one unit, time, I like to use secs. Thermal Diffusivity = ft2/hr, Velocity = distance/time…

I’m NOT seeing the relationship between ft2/hr and t/l in this thermal application that has a wall surface or in a “simplified” quantity the report made for Thermal Diffusivity a square area? How does taking the inverse change the units?

Jerry wrote: Dana pointed out the electrical analog is a distributed resistor capacitor network or transmission line and as such it does have a "propagation velocity". but it also has attenuation and "filtering" properties. A step change in temperature on one "side" will result in an exponential ramp on the other "side" after the propagation delay. The initial slope of the exponential is a function of the time constant and the input magnitude. Also the final output temperature will not occur till several additional time constants have elapsed.

My comments: I think I understand what you are getting at. Please make reference to this “propagation velocity” to an equation in the report. Point to the report an explanation of what is being “filtered” by pointing to the reports explanation of such as filter?

The initial slope as to do with pulses, how the mass is coupled via convection and radiation. The "time constant" noted in the report is in 3 phases and an approximation, noting more.

Jason wrote: So pc/k gives us a constant that is independent of unit length, or temperature - basically a material property.

My comments: Page 19: Steady-state conditions cannot be attained until at least an amount of energy, &, suffi- Rearranging the previous equation gives cient to give the steady-state temperature profile has been conducted into the wall. This form is used because it gives the energy storage per unit surface area of the wall, and the total area of wall need no longer be considered for now. The only properties of the wall appearing on the right-hand side of the equation are p , e , and L. (The temperature change
is not a property of the wall, but a condition imposed on it.) Thus, the peL product is ii measure of the wall’s ability to store energy.In the earlier steady-state discussion, it was shown that the wall’s U-value
(U = k / L ) is a measure of the rate a t which a wall can conduct heat. Even though the steady-state results do not apply to the transient case, the U-value is still an indication of a wall’s ability to conduct energy. The time necessary to attain steady-state conditions is related to the ratio of the wall’s ability to store energy to its ability to conduct energy. ability to store energy - pcL ability to conduct energy.
Thus, there are two properties of a wall which influence the time required to attain steady state: the wall thickness, L , and the quantity a! (a = k / p c ) , which is referred to as the thermal diffusivity. The thermal diffusivity is an indication of the speed at which the temperature profile moves through a wall. It has typical units of m2/s (ft2/h). An exact, analytical solution is available for this case. However, it would not give the reader much insight into the behavior of a wall and therefore is not presented.

Jason wrote: Following that line of thought, how is temperature applied? My understanding of heat flux is that delta T is the driving factor. However the use of the term constant implies that this rate is unaffected by anything, that it just is (you used the word velocity). What, if anything, affects this velocity, and how is it applied in our above example. Is input magnitude a fancy word for Delta T? If not, what is it in this context (heat moving through concrete)?

My comments: Temperature is applied a lot of different ways, air coupling (air-mass) or convection that is slower at ramp up, radiation from solar and thermal coupling (mass-mass), hvac, people, lights, appliances, etc….that is what it gets complicated determining a temperature at a given time on and through a wall per some area of wall or total mass in a home.

Two different things, The measurement of heat flux is most often done by measuring a temperature difference over a piece of material with known thermal conductivity. This method is analogous to a standard way to measure an electric current, where one measures the voltage drop over a known resistor.

Magnitude or better amplitude is heat flux at some capacity of storage, the way I understand it. Look at the peak curve in any of the sinusoidal curves in the report that is peak flux, max capacity, not a function of thickness because once at peak more mass will not increase peak capacity.

Jason wrote: 0.3 L^ / (k/pc) (again from the article) is the time it takes for the output to occur, or rather to reach it's full value. They break it down into 3 smaller time constants which achieve percentages of final value (36.80%, 86.50%, 95%). Are we just talking about the same thing?

Jerry wrote: The paper is excellent but the problem can be simplified from the three dimensional case described in the paper by the use of boundary conditions to a single dimensional case. Basically this simplification assumes an infinitely large plane with no heat flow in two directions so that all heat flow is in a single direction. I use this simplification and the result is the "diffusivity" becomes the inverse of linear velocity. Diffusivity is the inverse of a "spherical" wave velocity propagating in three dimensional space.. In applying the simplification the spherical velocity is divided by pi to convert to the one dimensional velocity. so the "time constant" for the dense concrete becomes aproximately 8"/hour.

My comments: Where did the paper talk about taking a “single dimension approach to 3”? Look at the temperature profile graphs, figure 2.2, put an axis on a surface surface plane (x-y-z) is used, sinusoidal and depth or thickness(L), heat flux is a multi-directional vector in reality, a vector normal (z) transient is assumed. The assumption they made is not to use the whole surface of the wall and more than three axis, which to be accurate necessary. 5-7 axis would reveal more. A sphere over complicates matter as they described.

Not sure where you are getting all this about spheres and propagated velocities? Can you provide a link and explanation to the paper Jason posted that to me is more accurate? Seems like all these analogies are confusing the readers, we would be better served to stick with the experts ORNL and discuss their reports as if they are not confusing enough. Remember guys, not everyone out here is an Engineer or Scientist, and I’m old :)

Jerry wrote: I you are still reading this thread: Facts! Heat transfer from a gas to a solid does NOT involve a phase change as you claim in post #33.. There is simply no "plumbing" needed in a rocket mass heater using a masonry bell.

If you are interested in rocket mass heaters you might want to visit:http://donkey32.proboards.com/
Where you'll find sufficient experimental work to assure a successful first try build.

My comments: Thanks for the link. I was in a hurry my bad been a while, metal is often used and a phase change does occur, I'll assume you did not know since you did not correct me with it, here is the basics, note the “issues” at bottom. I’ll probably go with a rumford if I can get it past code.
http://en.wikipedia.org/wiki/Rocket_mass_heater

An internal vertical insulated chimney, the combustion chamber, ensures an efficient high-temperature burn and creates enough draft to push exhaust gases through the rest of the system. Flue gases are cooled to a relatively low temperature within the thermal store, approximately 50 °C (122 °F), and steam within these gases condenses into liquid releasing the associated latent heat of evaporation, which further increases the efficiency in the manner of a condensing (gas) boiler.[1]

52. | | #52

Terry Lee,
Unfortunately detailed understanding heat transfer within a solid requires what most would call "advanced" math including differential equations and boundary conditions.. It simply will not happen here! Those truly interested simply must get the necessary background and educate themselves.
A similar statement can be made regarding "dynamic system" behavior or the response of a system over time to various "stimulus".
The Wikipedia definition of a Rocket_mass_heater is at odds with the site I gave.
Most successful RMH still require "chimney draft" and avoid condensation of the combustion products.

53. | | #53

Fluid dynamics!

Understanding constrained walls verses infinite walls;

My time in aeronautical engineering is flashing back;

We analyze wing airfoils with theoretically infinite length wings and with wingtips.

Walls end so often, at window, door, corners, floors, roofs...

Best to just test builds and then write the simplified formula that corroborates the data sets.

Done.

2nd year engineering lab basic calc and trig and algebraic math. Leave the diffyQ multiple postulates to the cosmic theoreticians.

54. | | #54

Fluid dynamics!

Understanding constrained walls verses infinite walls;

My time in aeronautical engineering is flashing back;

We analyze wing airfoils with theoretically infinite length wings and with wingtips.

Walls end so often, at window, door, corners, floors, roofs...

Best to just test builds and then write the simplified formula that corroborates the data sets.

Done.

Leave the diffyQ multiple postulates to the cosmic theoreticians.

55. | | #55

AJ agreed. Time for the engineers and physicist to sashay on over to the engineers and physicist forum, study it, analyse it and discuss it, then come back with an expanation that a chap that drives nails for a living can understand, then go hanggliding or go for a trail run.

56. | | #56

" We analyze wing airfoils with theoretically infinite length wings."

That's called an Earth Ship, that takes us to another galaxy.

"Best to just test builds and then write the simplified formula that corroborates the data sets."

That's called Reverse Engineering, that takes us to "Oh Sh** back to the drawing board.

"come back with an expanation that a chap that drives nails "

You can leave the nails at home.....

57. | | #57

AJ,
Unfortunately the effect of thermal mass is non exist-ant in the "steady state" . Coupled with the fact that describing ANY "system" under conditions that vary over time requires differential equations and "advanced" math which makes this thread "difficult". Martin's link in post #7 is about all that can be understood by mere mortals that drive nails for a living (thank you Debra)..

58. | | #58

And I can guarantee this having been there and done it, those ORNL Engineers and Scientist that developed the reports disagreed and argued A LOT due to a lack of misunderstanding, especially when the bean counters came around and told them to quit fighting they are over budget and under schedule! Dammit!

The simple watered down version, you know the one for the “average reader needing little tech background” the so called engineers and scientist out here either make look like high school math since they don’t even want to discuss it, or just don’t get it, resulted in a mosh pit, since making an Engineering ego simple always does.

AJ is right, a college degree and complex math is just a start, you have to build a prototype and test it. You will find the correct design-test-prototype process on the ORNL website and do not think the end result is completely accurate there are too many climate changes in the world, and did not take a lot of blood, sweat, tears, time, money, combined education and experience to get there.

Threads like this would be much more informative if a white paper review was conducted slowly. The poster having the education and experience were a mod too that could delete post that got ahead of them or did not stick with the white paper as it being discussed from top to bottom. Problem is people will destroy the thread with analogies that confuse the “average reader”, it becomes a competition of writing skills and whom has the biggest words, opinion, vs build data-fact and empirical evidence the report provides. Then people get upset they are being over modded for not following the rules of the thread. I tried it once, that is what happen on a HVAC peer review calling pro’s, the subject matter was a white paper review of HR I asked to not get ahead of me and stay on topic, no opinions just data. By the third post the subject was changed to HRVs opinions. Soon a "engineer" arrived trying to point everyone to her website for sales, far off topic. Since I am not a mod there was nothing I could do, the paper review failed. But yes, even the pros did not agree on that thread too and there were many debates.

We don't need to re-write what ORNL has already written, even if we have the education this thread does not have the lab, money for 10 or more builds in many different climate zones.

It is too bad mass is not better understood. It could save lives in hurricane and especially here in tornado alley. There is a myth it cost more than stick, batts, and nails....As I said first post the biggest difference is it is misunderstood therefore not as popular. Many builders can not afford Engineers and Scientist so they stick with what they know.

59. | | #59

This horse is DEAD! Why to we keep beating it?
Well I'll give it one more flogging with facts that require no math.
Mass and insulation are different and have different purposes in a building!
Mass is not EVER a substitute for insulation and insulation cannot do what mass can!
Insulation reduces heat flux! continuously!
The distinction between mass and insulation is one of degree as all insulation has some mass and all mass has some insulation properties.
Mass stores heat! All the heat that goes into mass will come out but it will be delayed and spread over a longer time with smaller temperature changes than those that caused the heat flux into the mass and some of the heat will come out at the same place it entered.
Insulation will reduce the energy needed to maintain a steady temperature difference

60. | | #60

http://web.ornl.gov/sci/buildings/2012/1982%20B2%20papers/006.pdf

Is this part of the research you linked Terry ?

This one is very easy to understand as conclusions go,
but i am not sure how it is applicable to a house.

Note that for heating, they concluded that a 2 layers and a 3 layers (concrete in with ext insulation and a similar to ICF ) had no difference in performance when mass is important.
(why did they use 0.2" of concrete to do tests??neway )

61. | | #61

Okay okay, let me see if I can summarize this before my brain explodes! :)

1. In cooling-only climates, where the average diurnal temperature is equal to or above the indoor comfort point, you can probably use all mass for interior temperature regulation if you really want, because even when the average temperature is above the comfort zone, nighttime radiant cooling will pull out more heat than you might expect. No amount of structural, uninsulated mass will be too much, and more will be better, but in all practicality, it will be limited by constraints of budget, available space, foundation strength, etc. A dehumidifier will do for for humidity regulation.

2. In any climate with a heating load, you need to heavily insulate your house; there is never too much. Go by typical calculations to determine how much insulation is the minimum acceptable amount, but exceed it wherever you can.

2a. Try to minimize the amount of mass outside your house's insulation, as having it there will delay desirable changes in the delta-T through the building envelope (e.g. a black shingle roof over an unconditioned attic makes the entire roof and its supports into a gigantic superhot thermal mass that doesn't cool off until hours after the sun goes down).

2b. Try to avoid using insulation that is "massy" to protect the interior against any exterior mass you can't get rid of (e.g. using cellulose on an attic floor actually seems like a bad idea to keep your house cool if your attic is very very hot, since the cellulose will saturate with heat and conduct it down through the ceiling more than a lighter material might).

3. Add as much mass as you can afford and is practical inside the house; there is never too much as long as it's insulated from the exterior

4. The more insulated mass you add to the inside of the conditioned space, the less pronounced your peak loads will be, and you smaller you can size your HVAC equipment to be compared with an identical building with the same amount of insulation but much less interior mass.

Is all of that right? I am particularly interested in knowing whether my 2a and 2b points are correct; if they are, then it could explain why the poor fellow in the thread I mentioned in my original post had higher cooling loads after adding a huge amount of cellulose to the floor of his outrageously hot attic. If so, we might want to rethink the wisdom of advocating for cellulose attic floor insulation in cooling climates/seasons. Maybe using blown fiberglass on the floor or turning it into a conditioned attic with spray foam might be better ideas in such climates.

62. | | #62

Nathaniel,
I'll certainly agree with point #1. !a is a second or third order effect and not worth any effort.. The amount of heat absorbed doesn't depend on the mass but has much more to do with the "emissivity" of the roof meaning a light colored metal roof will absorb far less solar radiation than a black asphalt shingle roof.. 1b here again there is probably no significant mass effect but be aware that fiberglass is relatively transparent to infra red radiation and is a poor choice for attic insulation. On points 2 & 3 there is negligible benefit in adding internal mass unless it is to "average" the effect of solar gains through windows to the extent necessary to prevent overheating. Adding more mass costs \$ with no ROI. The effect of added mass on peak to average load for HVAC size determination is well below the other errors in this process

63. | | #63

Jerry: the point of internal mass is to use cheap high mass materials within design parameters and or to replace existing situations.

cie : concrete floors, tile walls, concrete bearing wall, natural stone wall etc..

Most of the times, it will be more expensive than gypsum walls,
but it can be designed to add significant ambiance also.

Cost of mass if why most use concrete floors and that's it.
Still better than no mass at all.

Nathaniel: i believe you've summed the simplistic conclusions right.

64. | | #64

Not to dismiss the value of all the interesting posts on quantifying the effects of mass etc. But just for my own peace of mind, seconding what Jerry said: is there anything in any of the other posts which contradicts the advice in Martin's blog on mass?

65. | | #65

Malcolm, yes there are and there are contradictions in most of the post and Matin's blog with ORNL's concrete final test reports, and today's building methods . I could go through and quote them, spark up more opinions, but as I said long ago and Jerry confirmed, unless you have the capability to understand there is no need to beat this horse it is dead! Lots of write-ups out there on the net, if you compared them to the ORNL lab, field, models are contradictory and based on the writers level of understanding at the time. The best way to learn is to take quotes from the ORNL reports and bring them out here for review and interpretation, starting at the top. That would require one long thread. Even then, in reality, there are so MANY variables that will effect that understanding. That is why some try to to simply to steady state as Martin did in some examples. Again, ORNL makes it clear when steady state applies and when it does not. In realty steady state can cause huge errors.

Martin did a great job trying to explain or simplify to the average reader, he had to spend alot of time making analogies to real life situs people can identify with which bogged down the blog to the ones that do not need that, it is no where close to what you need as a designer, the best discussion was in reference to ORNL. ORNL did that, and to date I have not seen anyone else with that level of comprehensive theory backed empirical evidence. Clem U had some good data I posted earlier, if your into brick, they probably have updates.

66. | | #66

Terry, pretty well all your answers to every building question seem to be that you can't build any type of house without a lab, field testing crew of engineers and large amounts of pozzolans. Somehow I've managed so far and seem to get fairly good results by using general rules of thumb and advice borne out of the profession's experience.
I'm asking a pretty general question which doesn't need degree in advanced math. What is there in the the advice in Martin's blog that doesn't work?

67. | | #67

Malcom, the thread was about quantifying thermal mass to be able to build accurately in comparison to an insulation build with higher than expected cooling bills, that was not quantified completely accurate using r-value. If you have a building that has thermal mass and insulation that you have done sucessfully with analysis, or predetermined by analysis, or even guess work please post the results? I agree with you, I won't be spending alot of R&D time with it, I said long ago when I asked Dana to prove a "simple analysis" build he tested or has utility bills for, etc. I'll just use the data that ORNL has, other builders, and universities I have discussed it with, have to make an educated guess. "I do not have the budget nor personnel ORNL did" I did not see a build was illustrated on Maritn's blog, so I can not say what works or what does not. If that blog had a build with instrumentation those questions woud be answered.

If you are looking for quantitative analysis, Martin's blog fell far short, Again, look at the ORNL Final Engineering Test Report.

My offer stands to anyone that thinks they have a complete understanding of this and can prove it to me with some past build. I will be producing two fully instrumented prototype specs in the near future, you run the analysis, I will do the build, we can post it out here as a learning tool guest blog if allowed.

That is how I will quantify this, by reverse engineering....Build #2 will correct the issues, if any, and hopefully set the final configuration. Analysis or a model would help if issues arise in determining what went wrong, cause and corrective action so the same perhaps mistakes are not repeated in build #2. If "high cooling bills" resulted like the one this thread referenced in the OP, I would look at the back calibrated model not come out here for more guessing. It is no different than what ORNL did, it can be done without the math or models but, there are greater chances for repeated costly errors or starting off with a less than robust configuration. The biggest benefit in up front models are to choose between design options A or B, look at several digitally you can not build up-front. I did not see that on Martins blog. It will also aid in design changes, pre-determining impact. There are risk in this guess method that could be mitigated by Engineering. That is no different than building a home without a PE, structurally.

68. | | #68

Malcolm,
Martin's blog is a highly useful summary with out meaningful errors His last sentence is the best "advice" "Remember: the better insulated your house, the less thermal mass matters."

The one area I disagree with Martin is he didn't go far enough in "debunking" the ICF industries claims of mass enhanced "equivalent" r value. Martin mentions the idea of a mass enhanced "equivalent" r value which is kind of an average equivalent heat transfer under changing heat flux. This "equivalent" r value has been used to "hype" and sell multi layer mass enhanced JUNK. If one were to conduct the same tests with ordinary insulation materials ONLY the result would be a mass enhanced "equivalent" r value greater than the static r value! This is because of the fact I stated in post 59, that all insulation has mass. The mass of the insulation and it's effect on heat flow during changing conditions is ignored in the static r value. Stated differently the heat flux averaging of mass enhanced walls occurs in walls with real insulation and no added mass.
Data and actual tests of simply insulated walls under dynamic heat flux are rare or non existent but the properties of the materials, particularly the "diffusivity" suggest rather substantial increase of "equivalent" r value over the static value.
.

69. | | #69

Jerry, I'm not sure how far your think Martin needed to go in 'debunking' ICF claims, it is out of his control. ORNL effort was to place DBMS or give a value to products and mandates to control valse claims, similar to r-value. DBMS is not 'kind of an average r-value" or unit, it is a comparison or "equivalent r-value" with NO units denoting how much mass it takes to equal a house with conventional stick-batts and what is considered little to no mass . It is developed by test comparison of the two, test data, and a function of wall configurations, climate zone, and building type that was derived by 10 builds in 10 different climate zones and 1000s of models, you can read about here: http://web.ornl.gov/sci/roofs+walls/AWT/ExperimentalWork/index.htm

I think ORNL efforts were the most useful quantitative analysis to date and a strong effort to stop the lies. I got no idea why ICFs have grown in popularity as foam junk mass, as you said must be the false claims. ORNL also quantified ICF or ICI and compared it to CIC and showed the reductions in ICI comparatively. Martin need not say anymore that ORNL has not already proved! You will find that comparison in their c-mass calculator too.

Here is what Martin said which is accurate....

DBMS values are obtained by comparing the energy performance of a one-story ranch house built with lightweight wood frame walls to the energy performance of the same house built with exterior massive walls. The product of DBMS and steady-state R-value is called an R-value equivalent for massive systems. This R-value equivalent does not have a physical meaning. It should be understood only as an answer to the question, ‘What wall R-value should a house with wood frame walls have to obtain the same space-heating and -cooling loads as a similar house containing massive walls?’” This new metric was both useful and easy to misinterpret. Ever since ORNL researchers started writing about the concept of “an R-value equivalent,” advertising copywriters have been having a field day with the idea.

Now people need to spend more time with the ORNL reports and other credible info he made reference too, since there is obviously alot of confusion out there.

70. | | #70

Nathaniel, right! I was going to show that graph again as the true behavior of mass that does not bridge, seems like noone is readying the links except you. Martin's description never let go of a thickness that bridged and was not a sine wave. The cobb and stoners back in day realized the bridge and that is why mass got real thick (12-24"). Today a foam core Mass-insulation-Mass or CIC, ORNL shows most effective with concrete but true for any mass. SCIP (2-4" shotcrete on foam) is more effective than ICF. Now the ICF oems are stripping the inner foam off, makes no sense, high cost. As ORNL points out the initial ramp up temp is effected by the wave at the surface, later by how the heat is coupled and mass and it's capacity.

71. | | #71

I believe there are deficiencies in the linked thermal mass article that misrepresent or understate the benefits of thermal mass.

The analogies used right off the bat are largely misleading. For example, the cistern analogy seems to ignore the actual reasons why you would want a cistern: because storing water in bulk lets you avoid having to go out and buy water from the water merchants every time you're thirsty (real-world analogue: reduces intermittent equipment cycling and discomfort due to constantly changing interior temperature, especially in the summer), and also so you can supplement your water supplies from a free though intermittent stream of rainwater (real-world analogue: so you can benefit from free sunlight without overheating the house like it would in a low-mass situation).

The frying pan analogy further confuses things by implying that thermal mass is just a big liability since it irritatingly reduces response time when you're cold and turn on your heater, and might keep you too hot after your heater turns off. A frying pan is not the right analogy, since frying pans are sitting on the shelf cold 99% of the time until high heat is suddenly needed from a heat source that is turned on for the entire duration of the desired heating period. That's not the way inhabited houses are used. We want a relatively constant amount of internal heat, and we may use heating sources that are themselves intermittent (wood stove, solar gain, etc). This is exactly the use case where thermal mass is actually beneficial, because it can store and slowly release the heat from any intermittent sources, especially free ones like sunlight. And once the desired interior temperature is reached, the mass allows the heaters to turn off for many hours as the heat is slowly released into the living space.

When the article talks about how thermal mass can help you make use of free solar heat, the truck-going-70-mph analogy seems inappropriate. There are many ways to slow down or stop the solar truck: close exterior shutters or blinds; use reflective interior shades or blinds; use windows with high solar heat gain on the south side and low gain on the east and west sides; plant deciduous trees to shade the windows and house during cooling seasons; use intelligently-designed roof overhangs that mostly shade the windows and upper parts of the wall during cooling seasons, etc.

While the article mentions the time delay effect of retarding the first entry of heat, it misses the primary benefit: it's not so that you can time-shift your air conditioner loads into the night, but rather so that you can reduce them in the first place by making most of the temperature change take place in the mass of the structure rather than in the air, where you feel it. With enough mass, the heat that penetrates your insulation won't be radiating into the house in the evening; it will still be struggling to get through the mass by that point. As a result, once the sun goes down and the outside temperature drops, the flow of heat will reverse and the heat will start to get sucked out of the mass before it even gets a chance to affect the interior air.

Additionally, there is no mention of the ill-understood but seemingly important phenomenon of nighttime radiational cooling. This phenomenon results in more heat loss at night than the exterior ambient air temperature alone might indicate. Nighttime radiational cooling can drop the effective temperature at the exterior of a wall to low enough levels that it can lose heat even if the exterior air temperature is close to the interior temperature.

It is also not explained how this phenomenon is actually inhibited by insulation, because while the insulation will reduce the heat entry into the mass (good), it will also reduce mass's nighttime heat loss, too (bad). As a result, in an insulated mass house, the wall may not actually cool down to the exterior much, you will have to rely more on forced air to flush the house with cooler nighttime air, which may not work if the actual outside air temperature isn't low enough (this part is mentioned, to be fair). In other words, insulated mass loses the benefit from radiational cooling and has to rely chiefly on air temperatures cycling below the interior temperature. This important distinction is not mentioned.

By way of illustration, here is a graph of the interior and exterior wall temperatures for a house with uninsulated 12" brick walls over the course of a month in a Georgia summer:

I suspect that this unexplainably excellent passive performance may involve nighttime radiational cooling. If this wall were insulated on the outside, I suspect that the low points in that graph would rise, and the average interior temperature would, too.

While it is mentioned that if you do have insulation it's important to place the mass between it and the interior, it is not mentioned how much mass you can accidentally place between the insulation and the exterior without realizing it. For example, if you have an unconditioned attic with a shingle roof, that whole assembly represents thousands of pounds of mass outside of your house's conditioned envelope. If your attic reaches 110 degrees when the sun is shining (as mine does), then it doesn't suddenly drop as soon as the sun goes down; it takes HOURS to cool off, only reaching its coolest temperature practically as soon as the sun is about to rise again. In building science terms, the Delta-T through the ceiling is not dropping at night as much as you would expect because of all of the mass that is still hot. Ironically, this is a better instance of the frying pan analogy. I live with this personally and the effect is noticeable and measurable.

72. | | #72

Terry, what do you mean by "mass that does not bridge"? The mass wall measured in that graph was an uninsulated structural brick wall that touched both the interior and the exterior. Seems like the whole thing was a bridge, unless I'm misunderstanding you.

Do you mean "mass that is thick enough to not admit outside heat inside during a 24-hour cycle" ?

73. | | #73

Terry, your posts have me back to ...

insulation is a mass
mass is a insulation
some of it is fluffy
some of it is less fluffy
and some of it you wouldn't want to bang your head against

... and relates to why this whole topic is.

My home is standard 1980s fiberglass batts... south exposure is primary, huge cathedral ceiling. My home performs very well during the summer no real need for AC. Winter not horrible not great biggest problem is the stack effect and air leaks, but not horrible.

Anyway... I think my cathedral ceiling/roof benefits greatly from night time cooling and by having a bit of mass, enough anyway to aid a bit in keeping my home cool. My home does not overheat summers unless we get a long long spell of very warm day and night and haze etc... weather.

Mass is interesting as insulation. I build log homes... Not the best for insulation... but... definitely have wooden mass. The ORNL tab for log homes... says... stay tuned... not ready yet... Our NYS RESCHECK does take into account logs and we have to really add insulation and R-5 triple pane windows to get to 0% pass percentage. No way to pass by 50% or whatever.

I hope you are getting ready to build as often as you post Terry and post some pics of projects.

Jerry, just read your latest post... I think we both are similarly thinking... all insulation has mass and that mass matters and vice versa...

Martin, if RR was still posting this thread would be 500 posts by now with some interesting added thoughts along with at least 3 of us burned at the stake by now.
;)
aj

74. | | #74

Nathaniel, the black line is the outside wall thermo-couple, the red the inside, the graph shows a 6 or so hr lag between the inside temp and outside over time. This is wyke design with an air gap between wykes if I remember right. You can see they are independent of one another. if they bridged the two curves would be on the same temps and frequency and the thermos would show the same readings, since these thermos were placed at the same locations in space but on two different inside and outside surfaces. I confirmed this by calling the director, that is what he explained to me, unless I have alzheimer's which is very possible. They were doing hot box test on a new aerated brick from Europe when I called not allowed in the USA by placing it to the outside, and 2-3 wykes with air gaps. They are trying to use air as the insulator, no foam. Most massive cobb and rammed earth use a SIRE wall, foam wrap or core as a thermal break.

Speaking of cobb or clay, we have been discussing concrete with high density from portland cement, rock, stone, sand. Clay has an ability to store moisture, much better than concrete. Calcium oxide, CO2, moderate moisture. That is what I was referring to on the thread your referenced to humidity and cooling issues, had it been earth the issues would solve themselves. Talk about "Super Positioning" ORNL discussed in the report you linked to solve a complex heat analysis, now add moisture and CO2. I'll leave that to you college kids ;)

75. | | #75

Actually, terry, I believe it's a solid triple-wythe brick wall. No air cavity.

The European clay blocks that try to be insulators with trapped air in cavities are very interesting to me as well. Because the material itself is massive, but the many cavities full of trapped air are insulators. I almost want to model such a thing as an insulator with a number of very narrow but long thermal bridges that behave like mass. I wonder why we don't use things like that here. Maybe they are very expensive (they are from Europe, after all!).

76. | | #76

So I looked at the ORNL site, http://web.ornl.gov/sci/roofs+walls/research/detailed_papers/thermal/index.html
some more and found that adding 30#/sqft of interior mass can result in between 8% and 18% reduction in energy flow through walls of equivalent r. So, assuming the 18%, I can have the energy performance of r 30 with 30 #/sq ft if I use lite framing and r35.4. Or assuming I add mineral wool between double stud walls I'll need to add \$0.25/sq ft to achieve the benefit of 30#/sq ft of mass.

I'll consider adding mass if I find it priced at 1 cent per pound!

77. | | #77

Ok, I stand corrected, I can't remember been couple of years. I am testing and mocking up hempcrete which is aerated and works on the same principles. It has been interesting changing the mix to allow more aerating with more fiber, or less with more cement. Hempcrete is an insulation core, that gets a render permeable (mass) on both sides, around 14" thick for around R-30 mass. Again, r-value is not quantifying it's mass effect including heat, moisture, and CO2 absorption. Fortunate for me right now anyway is my area has no energy code, otherwise I would need to show an r-value (since that is all the ahjs understand or can relate to) of 20 for 2012 IRC we are thinking of adopting. I am trying to talk them out of it, since for one most builders will go out of business since their additional cost to insulate to 20 walls, and 49 roof, and 3 ACH most cant get below 5 without tons of SPF. Mass walls need no sealing that comes with the cement, these guys here will never try mass other than brick-stone and it overheats since they don't understand it.

I better take AJs advice and get back to my testing. This keeps going round and round and I think I am missing something...haha! ;)

Jerry most only know concrete, or portland cement products as mass and do not understand the other benes, they cost themselves out of the market before they get started therefore. Keep reading there is no way you got their reports full value that fast, unless you were there. There are a lot of good reviews of those reports on the internet too, check em out.

78. | | #78

Nathaniel, I could not have said, more less written, that better myself. Kudo's.

Not sure I understood this, " Building a house with thick mass walls in Vermont probably sounds laughable. "

Just to clarify, here is a Vermont strawbale builder with many net zero mass-insulation-mass homes there, he is part of a large association of natural builders that use other mass types. Guy is brilliant! and his partner, very spot on tech, great book with a video that takes you to their job sites: 18" thick bale center core insulation, straw that wicks water as a natural capillary break to condensation, 2" earth plaster as mass on both sides, they carry through to the interior walls, natural builders answer to structural and thermal SIPs :) , except both perform much better than factory wood-foam-wood products if done right.

79. | | #79

Terry, When I watch videos of high mass builds, like earthships, straw bale and rammed earth, one of the things i always notice is the presence of students or interns helping out in return for acquiring alternate building skills. Their real function is to compensate for the high labour component of these methods. A builder without access to these volunteers can't do anything but boutique projects. The same critique we level at Passive Houses applies here. I'd be interested in how mass can be integrated economically without relying on the owners or other non-market labour.

80. | | #80

I agree, Malcolm. The achilles heel of these traditional earthen building methods is that they're really best suited for huge crews of third world laborers paid a dollar a day. And they're not very automatable, either. Poured concrete à la ICF or Thermomass seems to be the most cost-effective strategy right now.

Even Terry's favorite--hempcrete--is extremely expensive. It can't replace the structural wood frame, the interior finish, or the exterior finish. So it winds up being a very expensive (if effective) insulation.

81. | | #81

Nathaniel, Thanks for the summary. For me anyway it was the most useful thing to come out of this discussion.

82. | | #82

Unfortunately, as I see it and I'd like to be wrong, the ORNL study compares the DYNAMIC performance of mass wall systems to the STATIC performance of wood stud insulated walls. Simply ignoring the dynamic characteristics of the wood stud walls (real insulation has mass and heat transfer is not instant). This serious error grows with wall thickness..

83. | | #83

Those were not volunteers, they were students...those natural builders have done well for themselves in profits despite the high labor cost myths, they wrote a book and opened a school that has alot of students.
https://yestermorrow.org/

Many of them do not believe in high debt to own a home most go in. The materials cost is reduced by far, as are the trades. I'd show a detailed cost outlay of the two, mainstream vs my natural mass, but now we are getting into my trade secrets and designs to cost. Besides, it varies on location. Soon I will need to put a non-disclosure and non-compete for everyone to sign. Haha! Yes, I have taken the cost out by substituting materials and automation, that is why I am pursuing mass, monolithic. If I did not think I could profit I would do what everyone else is doing. We can learn from other countries, labor exchange, stay out of debt. I thought of volunteers, local schools, they can slow production down, and I'd need to endorse them on my insurance. They also open up liability issues. They are not what they seem, a cost reduction. Rule that one out. The skill set required to do the mass builds is far less than the mainstream, I can hire kids at minimum wage, my mainstream employees and subs cost me 3xs that since they are skilled labor.

Jerry, thanks for the chuckles before night night, call ORNL head of the dept you are not getting it. I'm exhausted, you guys are wearing me out. Time to stay focused on builds tomorrow. Good luck all!

84. | | #84

Nathaniel,
Nigh time radiation certainly happens BUT the amount of heat lost is limited by the air temperature! Because the radiating surface will be warmed by conduction and convection. to very close to the air temperature.

85. | | #85

Okay, some more random brain percolations after re-reading a few of the ORNL links that have been posted to this thread:

In hot weather
- A thick enough mass wall in contact with the interior and exterior can entirely stop heat from entering the building during the day, and the mass can cool off sufficiently during the night via radiational cooling such that it never saturates with heat.

- Very thick insulation can almost entirely stop heat from entering, and it can prevent cooled interior air from giving up its heat. However, a highly-insulated house cannot benefit from nighttime radiational cooling and needs to take great pains to minimize solar gain and other sources of internal heat gain, as the heat cannot easily escape except by mechanical means.

Hot weather summery:
- Ideal: very thick mass wall, no need for mechanical cooling appliances! Mechanical dehumidification may be necessary in humid climates.

- Compromise: mass-insulation-mass wall, with very small mechanical cooling system (exterior mass is needed to benefit from nighttime radiational cooling, and interior mass is needed to absorb solar gains; mechanical cooling will occasionally be needed to purge heat).

In cold weather
- Very thick insulation can reduce heat loss to almost nothing. The thicker the insulation, and the sunnier and more mild the cold, the more important it is to actually reduce solar gains through the windows, which could overheat the interior.

- Adding mass inside of a heavily-insulated envelope can reduce peak loads, absorb solar gains and diffuse them internally over time, and stretch out the period of time between heating equipment use, while making the equipment run longer (more favorable duty cycle).

Cold weather summary:
- Ideal: Very thick insulation with a few inches of mass inside, not too much.

- Compromise: Very thick insulation alone, in conjunction with possibly minimizing solar gain.

So it seems like Terry is right, and that the sweet spot for a wall that works well everywhere (including cold and mixed climates), is a mass-insulation-mass wall. The exterior mass allows nighttime radiational cooling in the summer; the interior mass absorbs, time-shifts, and diffuses solar and other internal heat gains; and very thick insulation separating the two masses is crucial in winter to prevent the interior mass from losing all of its heat to the outside.

The opposite: (insulation-mass-insulation) allows neither nighttime radiational cooling nor diffusion and gradual release of internal heat gain. That explains why ICFs generally perform like slightly better insulated wood-framed walls.

I feel like I can also now better understand the perspective of those living in cold, cloudy climates. During cold weather, insulated internal mass is really not necessary as long as you take care to avoid excessive solar gains. With a super insulated envelope, all mass really does for you is permit you to reduce the size of the heater and give you a bit of forgiveness if it's often sunny and you have a bunch of windows. Building a house with thick mass walls in Vermont probably sounds laughable.

The oversight, I believe, is that most areas of the USA are not entirely cold with no cooling load. The vast majority of people living here will experience uncomfortably hot weather during the average year, many of them quite a bit (I'm on month number five for running my swamp cooler). It's gonna get really uncomfortably hot for large parts of the year in the southeast, the southwest, California, and the midwest. In these places, all the insulation you've added to deal with the cold weather can help you, but perhaps not as much as you might sometimes hope, especially not if it's very sunny and you have black asphalt shingles over an unconditioned attic (groan). That's where having a home with substantial interior mass can help a lot. And needless to say, a house with structural mass rather than structural wood also benefits from the advantages of the walls having no susceptibility to fire, flood, insects, mold, and even a lot of physical damage.

And seriously, black asphalt shingles over an unconditioned attic? What an awful idea.

86. | | #86

Enough of the JUNK science from ORNL! Spend some time reading" http://archbps1.campus.tue.nl/bpswiki/images/5/5b/H1.pdf
ORNL in creating the DBMS compares a real wall with mass elements to a simply irrelevant number (r value) completely IGNORING the fact that real insulation materials have distributed mass and delay the propagation of a temperature change.
The reference above shows that sandstone (very massive) and rock wool ( pretty good insulator) have the same temperature propagation but dramatically different amounts of heat energy are involved Under the right conditions of cyclic temperature variation no net energy will flow through the same thickness of either but when the temperature variation is not ideal much more heat will flow through the sandstone. ONLY INSULATION reduces heat flow under ALL conditions. Stated another way using mass to reduce heat flow is a FOOLS ERRAND!

87. | | #87

Jerry, I am not trying to prove that mass is superior or identical to insulation. In fact, in my summary a couple of posts back, I acknowledged that insulation always works in every climate. That's probably why it's universally recommended. But insulation is not mass, and mass is not insulation; they have different properties that are useful in different circumstances. Insulation is clearly the only thing that can prevent heat loss in cold climates. But something insulation cannot do is absorb and diffuse internal heat gains to prevent overheating, for example--which is useful not only in hot climates but also cold ones. For example, if the inside of your super-insulated box is nice and comfortable and then absorbs a lot of heat due to momentary solar gain or cooking, it might get too hot inside despite the freezing temperatures outside! In summer, you'd need to run your AC immediately to vacate that heat; in winter, you'd crack a window for a few minutes. But if your super-insulated box happened to have a lot of mass in it, the heat would be transferred to the mass and only later, once the source of the heat was gone, would it begin to release back into the air, and the rate will be gradual. It could even take place be at night, when everybody is asleep and nobody felt it. In summer, running the AC at that time would be more efficient and possibly cheaper. In winter, the extra heat would help keep the temperatures from dropping too much overnight.

One thing that is personally important to me is durability, which largely rules out lightweight wood framing sheathed with OSB and paper-covered gypsum. So if I'm going to build with masonry and have to accept its mass anyway, it behooves me to understand the properties of that mass so I can actually benefit from it thermally as well as from its greatly increased resistance to fire, water, insects, mold, and physical damage.

88. | | #88

Nathaniel, On an emotional level I share your affection for durable houses, and the desire to do that through the mass based construction methods we see in the traditional buildings around us. The present reality however tells us that the things that affect the longevity of buildings aren't primarily influenced by how durable they are. The short answer to why houses get abandoned is because people no longer want to live in them. If they don't like them because they perform poorly, are in the wrong place, are the wrong size, or simply don't provoke the affection of their owners, they won't be maintained and will decay. I'd worry about getting all these other things right before putting too many eggs in the mass based durability basket. Now if the mass you are adding in some way evokes affection in you and future occupants then putting it in probably makes sense.

89. | | #89

And this is in part what appeals to supporters of Rammed Earth. The mass of earth moderates the temperature swings of the day. R-value does not apply to rammed earth. Not sure about log homes though. I hope this thread continues.

90. | | #90

David. Wherever you may decide your views fall on the mass vs insulation argument, R-values will continue to apply to rammed earth. R-values apply to all building assemblies. They apply to logs, straw, earth, concrete, bodies you may have hidden it the wall and and cavities stuffed with the crumpled pages of physics books we have for some reason decided are obsolete and stopped reading.

91. GBA Editor
| | #91

Malcolm,
Thanks for bringing a smile to my face this morning.

92. | | #92

+1 to Malcolm :)
lol

93. | | #93

I've been willfully ignoring this thread for weeks, but had to peek when it came back from the dead (post -Halloween, no less!)

I'm not going to respond to most of it but would like to point out to Jerry Liebler that his statement here is not exactly true:

"Nigh time radiation certainly happens BUT the amount of heat lost is limited by the air temperature! Because the radiating surface will be warmed by conduction and convection. to very close to the air temperature."

The minimum temperature of the shingles can be more than 10F below the ambient air temp, since there is very little help from convection. The limiting factor to the temperature is usually the DEW POINT of the exterior air, which can be either close to the ambient air temp (if it's foggy out), or tens of degrees below the ambient air temp. Convection induced by that tempature difference brings more air to condense on the roof, but the roof is at the dew point temp, not the ambient air temp.

You can visually verify this in the AM by looking at the roof after a clear-cool night. All of that roof wetting did NOT happen at the nighttime AMBIENT air temperature.

But does it really matter from a total energy use or peak load point of view? Not much. Solar gains raising the roof deck temp to above ambient usually more than offsets the nighttime radiant loss factor, and the roof is typically a fairly small fraction of the total load.

For the rest of it, first-order approximations of the thermal mass effects are generally "good 'nuff" for predicting the peak loads and total energy use of a house. Differences in occupant behavior alone are much much larger than the inherent error in really dumb 2-D lumped mass type assembly analysis. The physicist in me knows that the model doesn't come close to reflecting exactly what is going on, but the engineer in me knows that in practical terms in most construction types being discussed here it doesn't really matter, a case of the the quest for the perfect being the enemy of the "good enough".

94. | | #94

So good enough Dana, to level the playing field I guess we could say the same things about insulation. Let's not bother understanding the r-value knock down of moisture infiltration and convective loops, and allow manufactures to produce values that are not real. BCS wasted alot of time looking at wall assemblies, and hot box testing to determine those values are not necessary, good enough without knowing, and just as designers don't have a need to understand mass systems, there is no need to determine whole wall r-values since it's "good enough" without knowing.

I'll have to disagree, we need a better understanding of mass such that people are not afraid of heating and cooling issues that can easily result from a lack of understanding and a "good enough" swag....we need better codes and design guides, more accurate software, too.

I have some rammed earth info I'll share soon, that in itself, developing thermal conductivity and mass effect from complete unknowns is a challenge. The first order is structural integrity as it should be with any wall assy, that can potentially increase the r-value and lower mass effect as the surface density increases. The soil alone has the same type of knock down tolerances as insulation, but can increase in benefit when wet, not degrade like insulation(s). Go figure! Perhaps we should just assign some r-value and call it good? Spend \$100-\$500 grand on a build, hope for the best?

95. | | #95

I been experimenting and running test on different mass systems, the latest Rammed Earth. I'm glad I could not get a skid steer into the driveway at this job site and got to experience the history of building by hand has to offer, as back breaking work as it is.It put me more in touch with it, a better understanding. I conclude the labor is a no brainier, the chemistry much more complex. I can easily see that there is a myth that claims it last for centuries, compared to light construction that dominates America. It takes years to understand this and it can easily be screwed up. I could find little help but did get some basic guidence from a Europen guru on another site, it's not like the product is sold in home depot. Lots of products in Europe. I had to search for the best materials at local quarries, rock, sand, soil (the binder)....and some stabilizers since my lime based soil is weak.

I built a retaining wall to start with around five different mixes, some stabilized with other binders in conjunction with soil. I don't have a lot of time to explain it all, so I'll explain some basics in small bits at a time.

Structures is my first concern, a lab test will prove, then thermodynamics I will share some basic test I will do within my means (equipment).

Water is a key component, soil has a high ability to manage compared to concrete due to its surface properties. I started with a jar that separates the different mass into particle size and soil type to give me an idea of what the binder looks like, there are some other basic test, lab test I will do for a home build will reveal more..A PE will require lab test to give mechanical properties, atterberg limits, plastic index, etc..... I did get some cracking to an over expansive soil.

I'm doing an interior pony wall over a footing that is thermally bridging I'll get to later....there I'll go for more of an aerated insulation mix since it is non-structural and run some test.

Here are some pics to get started....I'm experimenting with surface renders (pigments of high or lower density) since some of my surfaces are friable and my soil colors are limited.

This is an investment property I hope to have on the market soon and the chance to complete my testing. I hope th new owners allow me to monitor it over the winter. I have developed ways to cut the labor down and automate, so we are looking at building a spec home next spring with some sort of mass.

I canted the wall and added drain holes. I'm not done sculpting, the ability to create art with these mass methods is unsurpassed and very rewarding. I am the manufacture with complete control over quality and what I bring into the home walls anyway, another thing I value highly.

96. | | #96

Did you just fill the bottom of an interior wall with dirt?

97. | | #97

Malcom, shhhhsh, don't tell Dana, he'll start throwing good nuff r-values and perm rating's at it ;)

Yes sir, that is what rammed "earth" is partially made of dirt...You were expecting toxic foam, OSB? I'll also be putting water in the walls. Nuts ehhh? Amazing what mother nature has provided that works so well and cost so little if you understand it.

One of my test is to prove it in a rehab so we can move it to that side of our business. There is something to be said about it lending itself better to new construction, access. Inside I have to haul in 5 gal buckets of the wet mix, or mix inside.We'll see today I start. I have alot of mix left outside, there are other natural material choices, but I'm challenged with RE so we shall see. I have a new interior soil stabilizer I am trying.

I put some heat against one side of the outside wall to see if it would bridge and get a feel for thermal loading of mass just using my hand. The 12" wall with around a 15F temp difference between sides at the heat source showed no bridging. The surface with the heat around 75F, other side 60 with a 5-10 mph winds. It's around 25 out there right now, I'll try again let it run all night. Snow is coming this weekend they say, good! The wall seems to be taking a long time to load, heat is staying on the surface (mainly dense rock, makes sense) and not spreading far from what I can tell, or the lime wash coating I have on this side could be resisting loading, I'll try the other side next. Tomorrow I'll dig out the area after 24 hours and see whats up. Perhaps I should invest in a temp gage, if anyone knows a good one I can get a lowes or hd let me know.

So far the wall has a r-value and perm rating of 1000 for those that need some point of reference ;) It is in a class of it's own. ;) Big guy in the shadow, Frankenstein, I need no costume :)

98. | | #98

My 24 hour heat source ran all night in 14f lows, checked in 19f, surface very hot to the touch but did not spread out far. I show heat penetration to 4" which is in line with ORNL hot box and field test of much denser, higher psi concrete I think perhaps due to soil expansion and storage being higher. ORNL 6" showed slightly better dynamic benefit than 4. I imagine if Im let this run a week the penetration would be at least 6". I not sure if that makes me comfortable to not put in an insulation core. I'm thinking very unlikely I bridge. 12 inches is the norm in Europe for what its worth. Our average low is around 10F. I had a good 50+ temperature differential across 12 inches of rammed earth with a high rock content. I'll check a more dense mix next.

99. | | #99

Wow this thread is intense and it would take forever to read it all. As a contractor I've built many super insulated "boxes" with either (stucco over 6"foam-2x4-w/cellulose) or (hardi-board-3"rock wool 2x6w/cellulose)..etc... These all being custom jobs for the most part wealthy clients willing to pay the upfront costs of the labor intensive "passivhaus" approach. Overall we've had great results from a thermal comfort point of view. My personal opinion is that we seriously need to simply this process. I would also like to point out that anyone who has ever filled a 12" cavity with dense packed cellulose would have to admit there is a hell of a lot of mass in that wall! My real question is though is what about Hempcrete!

I've done a lot of research on the Tradical Hempcrete system of late. I am dying to hear what you guys think about it. With average r-values of 2/" plus high mass content -wouldn't this setup hit a sweet spot of both considerations? The Brits seem to be very gung-ho for it. And with the likely hood of industrial hemp becoming legal and available locally it is a very intriguing prospect for me. Im surprised it only gets mentioned a few times on here or am I missing something?

100. | | #100

David, Terry is also really excited about Hempcrete. I don't share his enthusiasm, because it's not strong enough to be structural. As a result, it doesn't really simplify anything, since you still need to build a frame or a structural wall anyway. And if you're going to do that, you might as well use non-structural insulation that's better-insulating, cheaper, more available, and with a base of knowledge for how to work with it.

101. | | #101

Indeed that is true and my father feels the same as you. However it does eliminate a lot of different procedures as well. Like bolting on layers of exterior insolation and the sheetrock for that matter. I personally think it looks really great as a finish.

Ive seen videos of French workers forming their own hemp blocks and rather easily stacking their walls like we do traditional cmu walls over here. Have you seen anything regarding hemp blocks formed with magnesium oxide instead of lime? Supposedly these blocks are structural. One such block called Lifecrete or greencrete looked promising however they seemed to fall off the map. Now I'm seeing a guy pushing Hemp Adobe honeycomb composite sandwich panels out west. He's touting very, very high -r-values and although i would immediately say its snake oil, after listening to him to speak in videos on youtube i admit he sounds rather credible. Hemp adobe homes is the co. name. Even if the wall is legit there definitely seems to be a lack of real explanation of how to put these things together.

• |
• |
• |
• |