You are asking the wrong question, jonr. Consider this scenario: It's 5 p.m. -- sunset, except you haven't seen it for two days, and the slab is 70 degrees. Then ask the right question: How much heat do you need to add that evening so that it will still be 70 degrees at sunrise?

The question assumes that 70-degree concrete tends to the nippy side, with air temps typically 10 degrees cooler than a radiant slab. In fact, if you surround the slab with yet more mass, and lots of it, you dampen the swing in air temperature to some fraction of the slab's amplitude. (I confess that what fraction in my case remains unknown.)

If I chose your water storage approach and dialed up the thermostat to 70 in concrete-- or its equivalent in air temp -- the house would be right on mark at sunrise. But I'd miss those dancing flames and that point during the evening news when my toes are toasty warm indeed. You say the way to avoid overheating is to under heat. I say no thanks.

Yes, the other great unknown is how competent I'll be as a human Honeywell. But, if I bomb, it won't be the typical passive solar bakeout. I've thought long about it; I have a full complement of tools. For the moment I declare myself smarter than a thermostat.

Check out this home in Edmonton...about the same climate as where I live.  Tons of south facing glass, very tight, super insulated, lots of mass in his floors (2" unheated concrete overlay).  I asked him if he experienced any overheat problems in October which is usually the problem month.  He replied that everything worked great,  no heat dumping required.

The design & control problem is to get the slab to slow it's emission of heat rapidly enough as the solar gain increases, which is no mean feat to do over a wide range of heat load and solar gain conditions. To get really massive slabs to behave well when the boiler is the only significant input is problem enough, without introducing a high and fickle solar gain into the mix. With a very modest solar gain it's not too big a deal. The time constants on slabs are long- long enough that you can't get good results with a bang-bang constant hysteresis room temp sensing. GOOD thermostats have learning algorithms to set the coefficients correctly to maintain the room temp, measuring the rate of room temp change and integrating it over time, not just the offset from the desired setpoint.

It would take a quite a human Honeywell to do anywhere near as well with a slab as a Tekmar 5xx series T-stat. Whether or not any would work well with a high solar gain (even with a slab-sensor) would be a part of the design considerations. (Sure you could set up your own set of sensors & algorithms to fine tune the place if you have infinite hours and software & engineering skills, but I'm talking off-the shelf stuff.)

Air temps won't be 10F below the slab temp an R50 house they way they might be in a less-well insulated place.

Posted By Garth Sproule on 23 Dec 2010 06:10 PM
Check out this home in Edmonton...about the same climate as where I live.  Tons of south facing glass, very tight, super insulated, lots of mass in his floors (2" unheated concrete overlay).  I asked him if he experienced any overheat problems in October which is usually the problem month.  He replied that everything worked great,  no heat dumping required.

I'm quite familiar with that design- been following it since before it was built, using some of the lessons learned from the Riverdale Net Zero house.

And they simulated it quite a bit in the design process.  While 2" of concrete floor isn't a huge amount of thermal mass but probably designed to be "right sized" for the amount of solar gain they get with the (over-10%) of floor area glazing and time constants.

Sure, if you can tolerate 80F or more in October, no need to dump or block heat.

Edmonton house: living space temperature will go up to, say 24 or 25 degrees on a sunny November day, and the concrete slab will gain an extra degree or two in temperature.

Todd: note that the air temp swings much more than the slab. And that the right question is what happens in the middle of a sunny day, not in the evening or night.

To read all this, it sounds like Sask in winter has so much insolation that you are in danger of overheating your passive house, (even in Winter at -40F). If that's the case, don't you just need to reduce your level of insulation a bit, or possibly step up your air changes?

Winter is the time you want to collect that solar radiation and the slab will receive it just as well at 65 F as it will at 80 F, the difference being the rate at which the slab is radiating to the interior. The trick is to reduce the boiler heat input at a rate commensurate with the increasing insolation, holding off the sun by a combination of shading, overhangs and glazing until finally, you need to go over to a cooling system for the summer.

Good heavens, gentlemen. You have never heard of thermal lag? The sun heats the house during the day. Thermal mass carries over the heat into the night. The cycle starts over in the morning. My change is to add heat directly to the slab when the sun isn't up to the job, which is a good bit of the time in gray Pa. Think of my radiant slab as a masonry heater without the bulk. Go to builditsolar.com and read accounts of folks who have combined glass and masonry heaters and lived happily ever after. Tell them that their challenge is "to get the slab to slow it's emission of heat rapidly enough as the solar gain increases" and their reaction will be the same as mine.

Well, no. All I want is enough heat to get me through the night.

My house is off topic except for the radiant slab part. It will have R10 AAC walls and R50 ceilings. Because the walls have three times the mass of the floor, and a circulation fan will be moving heat to darker regions, the air temp will show less volatility than the slab temp. I'll have shoji screen shutters to close on overbright days that will stop 80 plus percent of the radiation and 60 plus percent of the heat. That would be the four pillars of passive solar: glass, storage, distribution and backup. Tekmar 5xx series thermostats are strictly optional.

High mass won't work in Saskatchewan, but less buffering is needed when it is 40 below outside. Suffice to say that neither house is in Tucson.

One thing I don't see on the Edmonton house is reflectors under the lower floor windows. They should increase solar gain with minimal cost and no heat loss. Or a heliostat - I have a small one and it works well on the north side of the house.

Posted By Garth Sproule on 22 Dec 2010 04:13 PM
Harold Orr is one of Saskatchewan's great super insulation pioneers. He has stated many times that if a home is very well insulated and very tightly air sealed, that a heat source can be placed anywhere in the home and that differential temps between rooms will not be observed. With this in mind, I am thinking that a single zone radiant floor concrete slab heat source could combine with passive solar rather nicely. Even better if the floor circ pump can be made to run continously when the sun is shining. I really cannot see why this would be difficult to control. Surely there will be enough "unused" floor mass in the parts of the house that do not receive direct sun, to store all the solar heat that is collected?

How does Harold Orr account for large windows and the associated heat loss? Won't rooms with large windows be more problematic on cloudy days or nights? I think try to move heat from the solar slab to the unheated slab might be an issue as you need some delta t to effect this. If the solar slab is at 80º under constant circulation you might get 70 to 74 in the fluid which might be almost enough to get 68 to 70 in the unheated slab. I'm not so sure though. You still have to make up heat loss for air exchange and as we just sent some cold air you way, I think some type of fan coil would be the answer if you have hot water.

One of those well it depends.

I am going to up the ante, 10,500 HDD where I am and it is currently -18F outside. I looked at the Passivhouse design aspects, remember it is based around Northern Europe requirements, a lot of the US is very different.

I so not have underfloor, friends do and I really like it. BUT in an area of high solar gain (that would be us) it is not ideal for a super insulated house. It is very slow to respond, not a problem if you have a long grey winter with short days, but we do not.

I also know people around here who do no have a heating system as such, solar gain top up daytime and night loss is very low. They have a supplementary source for when it gets really cold and not much sun, like today. A small wood stove in many cases.

So it all depends.

Duplicate

Posted By FBBP on 31 Dec 2010 07:00 PM

Posted By Garth Sproule on 22 Dec 2010 04:13 PM
Harold Orr is one of Saskatchewan's great super insulation pioneers. He has stated many times that if a home is very well insulated and very tightly air sealed, that a heat source can be placed anywhere in the home and that differential temps between rooms will not be observed. With this in mind, I am thinking that a single zone radiant floor concrete slab heat source could combine with passive solar rather nicely. Even better if the floor circ pump can be made to run continously when the sun is shining. I really cannot see why this would be difficult to control. Surely there will be enough "unused" floor mass in the parts of the house that do not receive direct sun, to store all the solar heat that is collected?

How does Harold Orr account for large windows and the associated heat loss? Won't rooms with large windows be more problematic on cloudy days or nights? I think try to move heat from the solar slab to the unheated slab might be an issue as you need some delta t to effect this. If the solar slab is at 80º under constant circulation you might get 70 to 74 in the fluid which might be almost enough to get 68 to 70 in the unheated slab. I'm not so sure though. You still have to make up heat loss for air exchange and as we just sent some cold air you way, I think some type of fan coil would be the answer if you have hot water.

That is all standard passive solar design, and is well accounted for in newer high-R high-mass Canadian houses (and by the PassiveHouse tools), including Orr's body of work. 

The complicating factors come when you're also using the slab as a non-passive radiator supporting the peak loads in the chilly pre-dawn hours preceding a high-gain sunny day.  I'm not saying it's not do-able, just that it's not a simple thing to model and unless you simulate the hell out of it with some pretty good tools you're likely to have less than satisfactory results. It's definitely not a no-brainer.  It's much easier (read "more reliable") to do the passive solar design simply as a passive design, and design any supplementary heat around a low(er) mass, more responsive system.  Even optimizing the passive solar design & thermal mass to the site & climate specifics is not a trivial design problem, but much easier with tools developed since 1990 than the earlier part of Orr's career. (The PassiveHouse folks credit Orr's work as a primary foundation for their design models & principles.)

It's pretty easy to put a small hydronic coil on the ventilation air stream to add the necessary BTUs if the heat load is low enough, but not as cheap & easy as it is to put in some resistance heating.  Whether/when you do one vs. the other depends on just how often/long you expect it to run every year.

That is all standard passive solar design, and is well accounted for in newer high-R high-mass Canadian houses (and by the PassiveHouse tools), including Orr's body of work. 

Dana - I've seen lots of houses modelled with Ret software but I have seen very little after the fact data which would support the modelling. I understand that MCNZ is trying to put together a good body of data, to the extent of taking out the wood stove to get cleaner data. This will be very interesting to see next year and I hope they keep tracking it for a number of years as one year does not a case study make. If you know of links to some of the Saskatchewan homes an or Drake Landing in Okotoks that show how close they came to their projections, I would really appreciate hearing about them.

As I understand it they're still heat-charging the aquifer in Drakes Landing- it may take a decade before we know how close they came to projections. I'm not sure how much time they spent on the passive-solar simulation there.

Modeling the amount of "other" energy use in houses that are actually lived in is a big error factor that makes it difficult to measure when looking at one-off structures. (The guy who spends all winter watching hockey on the 52" plasma screen can REALLY skew the numbers!) At Drakes Landing they at least have a statistically significant sample to work with.

Concerning comparing actual energy use with modeling results, I have been working on that for my new house in Colorado. I have put information giving details on the house construction and on energy use on my web site, but I have not figured out the best format to share the calculations that I have done. It is well-insulated but not super-insulated, but using limited passive solar, a small solar thermal system, and a modest sized 3.15 kW PV system, it should meet (or come close) to net-zero source energy. The web site is not complete, but it is located at http://www.residentialenergylaboratory.com/. Folks are welcome to point out details that are missing that are required for modeling.

Lee

Posted By Dana1 on 03 Jan 2011 03:23 PM
As I understand it they're still heat-charging the aquifer in Drakes Landing- it may take a decade before we know how close they came to projections. I'm not sure how much time they spent on the passive-solar simulation there.

Modeling the amount of "other" energy use in houses that are actually lived in is a big error factor that makes it difficult to measure when looking at one-off structures. (The guy who spends all winter watching hockey on the 52" plasma screen can REALLY skew the numbers!) At Drakes Landing they at least have a statistically significant sample to work with.

A few comments. 1) I may be wrong but I'm pretty sure they are using grouted bore holes as storage medium at Drake Landing. Don't know how you would heat charge an aquifer as the water would continually erode the heat. 2) I don't think they did any passive solar simulation. It was all active (flat plate to storage and reclaim.) 3) It might take 10 years. Yeah that's why I would like to know if there is any increase in the heat stored. At 42º ground temp. and no vertical containment it will be interesting what the results are. I believe in an earlier post you had little use for soil storage. 4) It might be difficult to get true readings of one off structures but if we done try we will continue to use modelling that is not proven. Personal I would see the results of units such as MCNZ to be taken as to good simply because those of us that enter into an endeavour like that tend to be more cautious in how we spend power than the average joe. Not very likely to use the 52"!!

Posted By Lee Dodge on 03 Jan 2011 03:34 PM
Concerning comparing actual energy use with modeling results, I have been working on that for my new house in Colorado. I have put information giving details on the house construction and on energy use on my web site, but I have not figured out the best format to share the calculations that I have done. It is well-insulated but not super-insulated, but using limited passive solar, a small solar thermal system, and a modest sized 3.15 kW PV system, it should meet (or come close) to net-zero source energy. The web site is not complete, but it is located at http://www.residentialenergylaboratory.com/. Folks are welcome to point out details that are missing that are required for modeling.

Lee

Lee - had a quick look at your site. Will get back and study it later. Thanks so very much for taking the time to share this info. As discussed, actual results are much more important to me than modelled results so I take my hat off to anyone who takes the time to make the results available to the greater community. So thanks again and please continue to update your energy use. Also some comments on how well the home is meeting your expectations. Bob

Don't know how you would heat charge an aquifer as the water would continually erode the heat.

apparently you can charge an aquifer. the technology exists and it is in use by the greenhouse industry in holland. hybrid projects between residential and greenhouse industry are in the works. you have to just love what super expensive fuel does.

look at the bright side - in 50 years we'll have the same stuff they have today

adi

Posted By FBBP on 03 Jan 2011 07:49 PM

Posted By Dana1 on 03 Jan 2011 03:23 PM
As I understand it they're still heat-charging the aquifer in Drakes Landing- it may take a decade before we know how close they came to projections. I'm not sure how much time they spent on the passive-solar simulation there.

Modeling the amount of "other" energy use in houses that are actually lived in is a big error factor that makes it difficult to measure when looking at one-off structures. (The guy who spends all winter watching hockey on the 52" plasma screen can REALLY skew the numbers!) At Drakes Landing they at least have a statistically significant sample to work with.

A few comments. 1) I may be wrong but I'm pretty sure they are using grouted bore holes as storage medium at Drake Landing. Don't know how you would heat charge an aquifer as the water would continually erode the heat. 2) I don't think they did any passive solar simulation. It was all active (flat plate to storage and reclaim.) 3) It might take 10 years. Yeah that's why I would like to know if there is any increase in the heat stored. At 42º ground temp. and no vertical containment it will be interesting what the results are. I believe in an earlier post you had little use for soil storage. 4) It might be difficult to get true readings of one off structures but if we done try we will continue to use modelling that is not proven. Personal I would see the results of units such as MCNZ to be taken as to good simply because those of us that enter into an endeavour like that tend to be more cautious in how we spend power than the average joe. Not very likely to use the 52"!!

Mea culpa- at Drake Landing is indeed to to store heat in  the dirt, not the aquifer. Stable non-flowing aquifers have been used as thermal storage a similar fashion elsewhere, but the hydrogeology has to be near-ideal to work. (Somehow I'd mis-remembered Drake Landing as being another aquifer storage, not dirt.)

The dirt-storage approach can work on the larger scale where it fails miserably on the micro scale-  what's uneconomic for a single house, can be for a whole development.  The ratio of lossy surface area to total volume is far more favorable on the larger scale, where it would require massively more insulation than it's worth to get there with a single-house. The storage array at Drakes is approximately round, and insulated on both the top and the edges- it as a long line of boreholes it would need 5-10x the insulation. They have 2 METERS of XPS capping it, along with the sand & XPS perimeter insulation. (edited to correct- make that 0.2 meters - I read that as 200cm first time around.  R40 makes a lot more sense than R400...)

The few images I've seen of the homes at Drake Landing appear to have passive-solar/solar-tempering features, but I've yet to see detailed discussion.  Most homes have very little E or W glazing, most have a reasonable amount of S glazing (and I've never seen a picture of the N side.)

(further edited to add) They've got a nice li'l app showing the current system performance here.  Looks like they're running ~45C storage temps with ~30C returns from the district heating with +2C outdoor temps last time I opened it up.  It'll be interesting to watch it's behavior when it's truly cold out after a stretch of cloudy weather.  Mind you, this doesn't look like a true seasonal-storage system the way aquifer storage can be,  but it may have coasting-capacity for a week or more (It's hard to tell without better specs.)  There's a LOT of active-solar collector on the district-thermal grid feeding the thing.

At least according to their hype  the PassiveHouse folks have been upgrading the model correlating it with many houses and real-world data for getting on to 20 years- it's probably better than most for high-R structures. (And way better than the rules of thumb and WAGs Orr had to start with back in the 1970s.) But the variability with individual houses within the sample sets is large (probably skewed by how the house is actually used by the inhabitants as opposed to optimal use.)  You can always leave the windows and doors open on even the best-insulated most-optimally solar-tempered house to end up with higher heating energy use, eh?