Yeah, except that I can't find anyone that can help with a standing column well. The idea is to draw water from the well for the heat pump, extract the heat (in winter), and return to the top of the same well. Natural convection within the well would mix the returning cold water on the top with the other water in the well. However, I can't find an answer on how deep the well has to be.

I've never heard of an example of a 2 ton standing column well, so don't know if the general 50-100' well depth per ton holds for such a small heat requirement and a corresponding shallow well. There might be a "minimum threshold depth" necessary.

As far as the HRV goes, it is required for new home construction. The house must make 0.3 air changes per hour. You could just go with bathroom fans, range hoods, and open windows, but in the winter you won't be using these for fresh air, so although it may pass building code, air quality would be terrible. So the HRV will definitely be going in... With option #1 or #2, the HRV would be integrated with the air handler and there would be no additional duct work, which would be great.

Jesse, to give you an idea of the building envelope:
-2x6 walls
-cathedral ceiling throughout 2nd floor (18x42' footprint) with 12:6 slope
-2nd floor ceiling is low, with walls along front and back of house ~5 1/2' high, and vaulting to 9 or 10' high
-roof 16" thick
-dense pack cellulose insulation in walls and roof (with roof cavity completely filled, with no ventilation channel above insulation, as required for batt insulation)
-exterior sheathed in OSB, with fiber cement siding

That said, I wonder if going from 2 to 4" of under slab foam is overkill if I there are other weaker components to the envelope. I didn't know people are insulating their slab more than this, so am quite interested. Are the edges of the slab also insulated this thick? That would look awkward for slab on grade construction.

As a percentage, yes. As you fix one thing, the others increase as a % of loss. But on a $ basis, nothing shifts around, it just become less and less attractive as you keep reducing the loss in a particular area.

I'd be shocked if 8' of soil ('ceptin' maybe peat moss) really behaved like a simple insulation against outside air since it has a signficant & interactive thermal mass. The thermal mass of it would mitigate against peak flows, but it's R-value would be averaged over weekly or monthy average temps, never daily/hourly. Insulating against sub-soil temps, not exterior temps is the larger factor in basement slabs, and those temps are in the ~45F-ish range & lower around Lake Superior. 

The heat loss out of the bottom of a heated uninsulated slab is huge- typical soil interface U-factors are in the U0.75-1.5 range, but the thermal mass interacts with the heat source over the season as the near-soil heats up and humidity shifts, etc. so the model gets very complicated on a seasonal basis (ask any earth-coupled geothermal designer).

I'm sure going with  6" of XPS instead of 2" or 4" is cheaper than performing all the necessary testing to provide sufficiently accurate data into the PassiveHouse model to see if 6" is really cost-effective, eh? ;-)  It wouldn't surprise me if 4" was a no-brainer level for a 10year NPV analysis in many, if not most cases in that region. 

But the fact that his off-peak electricity prices are substantially less expensive per btu-delivered-to-the-slab than most people's natural gas (burned at 85% efficiency) may cut into the financial-return aspects a bit, just as it does to the case for geothermal.  (In my neighborhood retail-residential power costs 5x what it costs him for off-peak, making the argument for geothermal more viable, but even here superinsulation is often more cost effective than geothermal in new construction.)

I would not count the portion of a standing well that is above the water level. With that adjustment, I believe normal sizing (around 60'/ton) would apply (after corrections for climate). Ok, making the well deeper than needed for just water is some extra cost.

You don't have to guess about soil characteristics in Thunder Bay: the Canadian government's Basecalc modeling software has a specific profile for the area. http://canmetenergy-canmetenergie.nrcan-rncan.gc.ca/eng/software_tools/basecalc.html

That is the good news. The bad news is that its slab on grade function models a simple pad rather than footings and stem walls that get you below frost.

Some thoughts from my play with Basecalc: Putting a thermal break between the slab and the stem wall is surely more important than 2" vs 4" under the slab. Even then, the edges of the slab are more important than the middle of the slab, because of the straight ahead nature of conduction. The edges of footings lose heat twice as fast, and corners three times as fast, as the bottom of the slab.

While you are constrained by SOP for foundations in your area, which also depends on soil characteristics, the best approach is to pour the slab inside ICF stem walls. If that will work, you can model a pad or basement at the depth of your footings as a crude approximation of what you gain by adding inches of XPS under the slab. As Dana writes, soil has no real R value, but heating four or five feet of it takes a long, long, long time.

I am pretty sure the Canadians set insulation requirements for radiant slabs, which not only gives you a starting point but tells you to err on the side of too much rather than too little.

Thanks for the tip on the modeling software. I've tried it out and will need to get more information from my designer and/or contractor.

As far as stem walls go, since it will be a slab poured on a raised section of bedrock, I didn't think there would be stem walls. I'll have to ask my contractor about this too.

Jonr, thanks very much for the 60' per ton depth below waterline information! With that in mind I'll keep geothermal as an option and will now get quotes.

I suspect bedrock has a very different set of thermal characteristics than dirt. (Just a guess... but maybe and educated guess? ;-) )

Don't accept any simple rules of thumb on geothermal wells & heat exchangers for region/sites different from yours. Seasonal & peak capacity will vary with groundwater temps & flows, horizontal separation distances, etc. It's even less rule-of-thumb worthy when you're dealing with buried loop & soil factors instead of groundwater-proper (which is what I'd referred to earlier in reference to thermal modeling of dirt.) Bedrock isn't usually the best fit for a geothermal solution either. (Swamps can be pretty good! :-) ) I'm not sure where 60'/ton is "normal sizing" either- closed loop well designs run ~150'/ton in my neighborhood, and our groundwater is probably ~4-5C warmer than in Thunder Bay. (Once you've coated the heat exchanger with groundwater-ice, the thermal transfer characteristics change, and not for the better.)

You can spend a heluva lot on insulation & time-of-day controls to achieve the same operational costs of going geothermal and still be below the cost of a geothermal system. Get some quotes, then see if superinsulation is really more expensive than geo. In many cases in new construction, going with double-studwall Larsen Truss type walls or PassiveHouse approach will be a better investment. At 4.2cents/kwh energy pricing both high-efficiency envelopes OR high efficiency heating systems are likely to look somewhat extravagant. But if you're making the investment anyway, bear in mind that superinsulation is lower maintenance than geothermal, continues to work in a power outage, etc. When all else is equal from a financial outlay & performance POV, reduced loads from a highly efficient envelope should trump increased mechanical-systems efficiency.

If you make the envelope efficient enough, the size & cost of the systems necessary to support the reduced loads fade, and the relative efficiency of those systems become somewhat irrelevant (further reducing the system costs.) That's how the Urban IL PassiveHouse can get away with using a 1kwh resistance heating element in the HRV, as it's "heating system", and STILL use less than half the overall power (all uses,not just heating) of the US national average. If it can work for the experimental Saskatchewan House in the 1970s (they had to remove the active solar, since the place needed more cooling than heating), it can work for you. With benefit of better thermal design modeling than was available then the costs can be better bounded too. A 25kbtu/h heat load cottage can become an 8kbtu/h heat load cottage and it may be less up front cost than an inappropriately complex & expensive high-efficiency heating system. The financial "sweet spot" is probably somewhere in between.

60ft/ton is reported to be common even in Maine but definitely adjust any average ft/ton figures for your colder than average climate and for any differences in bleed rate. What that adjustment is, I don't know.

You are about 44F ground water and 10400 heating degree days.

Oops. Missed the part about bedrock, which does have different thermal characteristics. But that isn't your problem. Your problem is that surface or near surface rock will track ambient temperatures absent deep snow. If it is monolithic, it will conduct cold under your slab. You'll need insulation in all directions except up, particularly at the edges.

Can't imagine you wouldn't need perimeter walls to level the slab, create room for XPS and meet code requirements for raising sill plates above grade. Pouring a slab inside ICF walls absolutely will work on bedrock, and would be easier to build on a hard irregular surface than other framing methods, it seems to me. Then again, I know as much about building on the Canadian shield as most people here. (Zip in other words.)

Having absolutely know idea how one builds directly on rock, I'm going to guess that one might drill some holes for rebar, level the area with gravel and then pour the slab on that. Then build the walls on the slab.

Have you considered electric radiant ceiling heat?  (Disclosure: I work for a manufacturer that makes these systems.)

At 4.2 cents and a reasonable shoulder rate of what? 7.x cents, you should be able to heat the space almost entirely during the off-peak period with maybe some supplemental during the shoulder rate period on very cold days.

Remember that radiant heat warms all the objects in the room and all that mass has the ability to store heat.  The net effect is that, depending on house design, insulation levels, etc. it can take a long time for the various rooms to cool down.  Our typical experience is a 1*F temperature drop in a 4 hour span in Canadian climates.  This small drop is normally not within noticeable range for human comfort (therefore no sacrifice in comfort) and you can easily avoid the on-peak rates.  A decent programmable thermostat that supports enough periods per day is all that's needed to take advantage of the off-peak and shoulder rates.

My own personal experience is that we had a power failure in the winter of roughly 13 hours and the temperature in the house dropped about 2*C.  Now it wasn't the coldest day of winter but still.  And I have a LOT of glass.

FWIW, I heat my home, using no setback strategy (other than keeping my garage and storage room at 12*C) for less than $1000 per year.  ~3800 sq. ft., slab-on-grade.  7.x cents per kwh, ~7800 Heating Degree Days.

Great post and comments! I faced a similiar problem in the last house we just finished and I am still not sure we made the right decisions. Here is how the numbers worked on this 2,000 sq/ft house. We had a 22,000 Btu heat loss at design temps here in Southeastern MA. 0 degreee design temp and 6,000 heating degree days. Insulated slab on grade with 3" of geofoam EPS below, up the exterior and extending 4' beyond the perimeter with a 15 R-value. Walls were constructed with SIP's 5-1/2" core EXS with a true wall insulation value of R-23 after all of the components were calculated. All of the windows were at maximum of U-30, better ones didin't justify the energy savings. Passive solar design with maximum solar levels achieved though design, orientation, and even roof angles designed for solar efficiency at this location. We installed radiant heating within the slab and supply this through a waterfurnace geothermal heat pump, 3 ton. The exchange was a horizontal slinky loop consisting of 3-600 loops all running directly back into the mechanical room. a turbomax exchange tank is used to transfer the heat pump heat into the radiant system and a 26 gallon tank was used. Controlling the slab has been another issue and the controls I would like are not readily available. We are planning on cooling this house through the radiant slab as well. I know kind of risky but precautions were put in place like a whole house dehumidification system and an HRV system as well.
The wild card in the house are our prototype thermal mass windows. we have a total of 200 ft of south facing glazing, 150 sq/ft are made up of windows that contain about 3,000 pounds of water. We are data logging the windows and the house to see how they contribute to the total energy costs, so far so good. I have been documenting the whole project and posted 15 or 16 videos of the project on my youtube account, if your interested in what I did check it out.
http://www.youtube.com/user/eebuilder

Controlling the geothermal system takes some time and a little bit of tweeking to get it right and I 'm not sure after the first 2 months of operation we have achieved maximum performance. I used a couple of data logging devices on the goethermal and radiant system to ensure proper cycling, flow rates and delta T's were working as designed. It took quite a bit of manipulating to get it running at the proper cycling of the heat pump. If you get it wrong the pump will cycle wildly and not perform as it should. I was surprised that verl slight changes in the flow rates could make such big changes in how the system ran. We fired up the heating system on Feb 15th 2009 and during the first month of operation the whole house used $157 worth of electricity for all of the energy needs, bring a cold slab up to room temperature took about a week and then again the heat pump was running almost constantly. Our electric rates are 17 cents a Kwhr not 4.2 . The slab has great potential for storing energy and if you can use the energy at it's lowest rate then I would definetly suggest putting in the radiant slab heating. How you provide that energy is another question? Does an electric boiler make sense, maybe given the low upfront costs and the inexpensive electric rates. The HVAC system for this house including the geothermal exchange, heat pump, distribution system(radiant slab), HRV, Whole house and dehumidification system, cost a total of $24,000. The geothermal system including the exchange field, and heat pump cost around $17,000. The thing that made sense for me was that geothermal used less total electricity to provied the same amount of heating and cooling as with any of the other options. Making this house ready for conversion to a true zero energy homes. After the first full year of occupancy the total energy needs for the house will be calculated and a PV system will be designed and estimated. I doubt that you could justify a PV system with electric rates at 4.2 cents per Kwhr.

I would first make every attempt at reducing the heat loss not only under the slab, but everwhere with the house and try to get the loads down to less than 10Btu's/sqft at design temp. Reduce the loads as much as possible and you will find that geothermal no longer makes sense. At 20,000 Btu heat loss the cost effectiveness is borderline. Reduce your loads by even another 25% and it wouldn't make sense to go any other route than an electric boiler and maybe some solar thermal to buffer the loads as well. Turbomax a Canadian company makes a boiler specifically for this purpose, I accually used one for the MIT solar decthalon house I worked on in 2007. That would be my suggestion, the initial costs would be less than going geothermal and even if you installed a couple of solar collectors to back up the system you would find this to be a good mix, with such low electric rates, I'm jealous. If I had your electric rates the 2,000 sq/ft home I just built would have an annual total energy bill of $375, if you can use the energy when the rates are low. Is the reductions in the rate time specific, like night to day or seasonal, like summer to winter?

Keep this post going it would be good to get feedback on the choices made and how well they perform after up and running.

Tom Pittsley
[email protected]
www.eebt.org

Hey Tom,

I'm not sure how geo is at all cost effective in your scenario. Even if you run a COP of 5 which would be a penultimate achievement with today's tech, that's like 3.44 cents per kwh for heating. for easy math let's call that 1000 BTUs per penny which is a little bit generous.

For natural gas at a buck a therm (just to guess) that's 1000 BTUs/penny as well, with buffer and mod/con at 95% efficiency that figure doesn't change much. and you save more than $10k on initial costs over the figure you gave for the geothermal.. for boiler only very much more like $12k+

You probably need about 50 million BTUs/winter or so with the heat load you spec'ed at a zero design temp and 6500 degree days/year. so any fluctuations in your electrical rate, COP, or boiler efficiency can be easily used to modify those nice round figures, and multiple x 50,000 "1000 BTU units" to calculate simple payback.

I wager it's not there. for a 10 year payback.. simple payback only... you'd need to save at least $1000/year. Your total heating bill with natural gas here would be $500 with no solar gain at a buck a therm. your solar gain further hurts this comparison.

You do have that benefit of making PV MORE attractive... but there still isn't anything remotely cost effective about this. Achievable, cool... I have no problems with that... but certainly not economically appropriate for wide scale deployment.

Offpeak electric could achieve this same kind of cost comparison (if available) at half the cost of geo. Something like a steffes ETS unit.

I am wildly interested in seeing any data you get on those mass windows though. I'm also interested in what you are doing to keep the water clear?

cool video too tom: how are you keeping those bad boy windows from freezing?

Hey Tom,
Nice project. Your u tube site is great, good doc. I am curious as Rob whats with the windows, How are you keeping the units so clear, How long do you see these lasting, repair, so on..... But all in all excellent system out side of the box.
Dan

Insulation is the natural first step. I would use a FPSF system.

C:\Documents and Settings\o\Desktop\DIY radiant floor heating\FPSF\Frost Protected Shallow Foundations - Page 2.mht


Most soil does have resistance to heat flux and it is quite predictable in certain areas. This leaves the perimeter (expose to higher Delta Ts) the logical point for more insulation. When I specify insulation for radiant floors, I use a simple and practical rule of thumb here in MN (not Yellow Knife but still chilly). R20 at to the frost line (out, down or in as the cold doesn't care) and R10 under the slab (for residential as commercial buildings have much lower ROI for middle slab insulation).

Shallow foundation technology has been around for a long time but follow culture more than science (popular in the south and not in the north).

As for any kind of electric heat. Our gas has gone from $.98/therm to .52. The Canadians make their own gas. Geo-thermal and off-peak (as long as it lasts) have to work hard to compete.

Well the windows aren't my invention but they are pretty cool. We didn't have the heat up and running in this house until mid Feb and the lowest the window temp got was 48F, that was only after 4 days of no sun, lows below 0 and highs around 20F. I know that some additives were put into the water to stop algea growth and so far so good.  The exterior glazing unit that covers the water blocks is one of the best solar windows you can get. It has a .29 U-value and a .76 SHGC, so even if the water blocks fail over time we still have the passive windows that will work. We did have a couple of the blocks leak slightly and replacing them is a simple matter of removing the gaskets between them and removing 4 screws.

     As far as freeze protection in much colder climates I don't know, want to build one and find out? We considered alcohol but the codes are pretty specific about flamable materials in buildings and I didn't want to assume the risk personally. They have however reinvented these already and are begining some beta testing. They are also concerned about the long term storage of the water and the install was definetly not an easy job. Their latest version contains some other kind material that I am not at liberty to talk about. I have this non disclosure agreement with the inventor and can't tell people about the great work he's doing. I couldn't even tell people about these windows until after they were installed.

As far as the geothermal system not being a practical approach it was one that was dicussed at length with the home owner before we made the decision. The total cost of this house was about $150 sq/ft not much if any more expensive than any well built home in this area. A good air source heat pump like the arcadia might be a good choice to save money but I would like to see some data on the systems that are installed before commiting to it. does anyone have any experience with the arcadia heat pumps?

Ultimatley the choice was made so that in the future the home can be easily conveted to zero energy. The cost to do this right now is about another $25-30,000 for the PV system.

The idea is to have a home that can provide all of the energy needed to live comfortably without being subject to influxes in the energy market. What better way to retire than knowing that once your income has been fixed, your homes costs have been fixed as well. I have seen to many older people in recent years having to make tough choices about what bills to pay when the cost of energy spike during the coldest time of the year, do I buy food, medicine or oil this week, tough choices that should not and do not have to be make if you live in a zero energy home. This is the way of the future, let's all get off the fossil fuel train.

Tom Pittsley
[email protected]
www.eebt.org

Here's an update. Construction starts in about two weeks. I'm going geothermal with a water-to-water heat pump, but don't know if I should connect the hot water storage tank to just a water coil, or also a radiant floor. The PEX will go in the slab, but it might not be hooked up.

I'm going with a forced air system because I would need the ductwork anyways for the HRV, so the air handler (with a hydronic coil for heating and cooling, and electric heat strip as back-up) isn't much of an additional cost. I could do away with the radiant floor altogether, then wouldn't even need to isolate the heating water into its own closed loop...

Any thoughts?