New Greenhouse build with thermal mass storage; air-to-water heat exchange system

Hi,

I am in the process of building a greenhouse with underground thermal mass storage(water) and I would like to get some advice.  This is my 2nd greenhouse, but 1st using thermal storage and double paned glass.  I am basing my design off of this http://www.builditsolar.com/Projects/Sunspace/GreenhouseHX/HeatPumpEnergyStorage.pdf, but there will be two 275 gallon tanks buried under the greenhouse instead of one tank above ground.  I am a mechanical engineer so it should explain why there are so many components

Goals: grow vegetables and herbs year round. Keep greenhouse 50 degrees or above at night with an average Dec-Jan low of 17 degrees. Organically grow tomatoes, squash, strawberries, leafy greens, and possibly peas/green beans.
Location: Denver, CO, zone 5b
Heat storage: water, in two 275 gallon ibc totes. tanks buried under greenhouse.
Tank Insulation: above/below tanks: 2" extruded polystyrene foam(pink Dow foam), sides of tanks: 4" polyiso insulation with 1" R-Max polystyrene adhered to outside(so only polystyrene is touching earth). All foam joints were sealed using Great Stuff expanding foam
Tank piping: 1/2" pvc piping, entering and exiting tank through top fill cap(purchased fill caps that have two 2" NPT ports each). I plan to use quick connect pvc fittings and install an 8-10" access tube to access the tank.
Greenhouse design: 9.5' x 11.5' footprint x 9'-9.5' tall. South facing glazing, 55 deg angle. Will possibly glaze the front knee wall and some of the east wall.  Construction will be pine 2x4s and 2x6s(glazing wall), osb and then horizontal cedar fence panels over the osb.  Insulation behind osb is not worked out yet. If the glazing loses too much heat I will consider an automated insulating curtain .  This is my current pick for the curtain material: https://www.reflectixinc.com/  
Greenhouse Anchor posts: qty. 6 4x4s embedded in concrete using 8" sonnet tubes. Three 4x4s along front and three along back(one on each end and one in center)  
Glazing: double paned glass that was collected free, mostly 76" x 32". (have not figured out what to do about potential hot spots yet).  Also looking for the simplest way to affix and seal them to the 2x6s. most of the glass panels have no frame.
Heat exchange system: parts are not purchased yet but current design uses a blower to force hot greenhouse air through an air-to-water heat exchanger. Pump circulates water between heat exchanger and tanks.
Growing bed: will be entire footprint of greenhouse, 15-18” deep with 4” of extruded polystyrene foam around border and 2” along bottom.

Current status:  water tanks are buried with bottom and side foam installed.

Nice design Dan, I can tell you're a mechanical engineer.  I like the idea of storing the heat in water and hiding it underground.  I'll just ramble out some thoughts in no particular order...  

If you're aiming for maximum winter light I think the angle of the glazing may be a little flat (unless you're talking 55 degrees from horizontal).  You're about at 40 degrees N so the equinox sun will be at 50 degrees and the winter solstice sun will be around 27 degrees from horizontal.  Please double check me on this because I'm going from memory a bit.  I believe many people aim for about 15 degrees more vertical from their equinox sun angle (35 degrees from vertical for you).  That way you are optimally catching sun from Nov-Jan instead of only on Dec 21.  If your write-up meant 55 from horizontal, you'd be good to go.  55 from vertical will likely give you lots of sun in the summer and less in the winter which might be the opposite of what you want.

I've heard that keeping the soil warm does more good for plants in the winter than the air.  Having the warm tanks under the beds should help.  You might want to leave access to route the hot water lines for the heat exchanger through the soil in the areas not over the tanks so they get some heat as well.

I don't know anything about hydroponics but you may be able to use your soil bed to your advantage by not sloping the bottom of it.  If you had it flat and sealed it to hold in water, the plants could draw up the water they need from the bottom of the soil bed.  You'd want a drain an inch or two off the bottom so it doesn't get too wet.  And since I don't know anything, don't just do what I'm saying here.  But that may be a way to make irrigation easier or more automatic once the plants have their roots down a ways.

How will you keep the dirt from crushing the tops of the tanks in?  It would suck to build it all, put in the dirt and then have a sunken spot over the tanks.

Here's a far out idea but since you're an engineer I figure you can handle it   Make one of the tanks a phase change battery using glycerin.  Before you can't get at it anymore, coil a whole bunch of pex or irrigation line down in the tank so you can run water through it and up to your heat exchanger.  Then fill the tank with glycerin.  It phase changes at 65 degrees which takes a ton of energy to complete.  So when it's warm you circulate water through the coiled pipes to melt the glycerin.  Then when it's cold you circulate cold water from the room down through the glycerin to heat it up.

If you can keep it above 50 degrees you can grow citrus in there too...

If you're putting in a vapor barrier (which I think is recommended) I'd put it on the interior side of the osb so the osb doesn't get moldy or water damaged.  

I don't follow the need for the anchor posts.  Aren't the existing raised bed timbers enough foundation?

Some glass has low-e coatings or other things that could help or hurt you depending on which side you face out.  If you can figure out what you have and if it matters, it may be beneficial.

You'll probably want a walkway to access all your plants.  Maybe incorporate that with the access to the tanks so you have a stacked function (access, walkway, plumbing chase, etc).  Plus it's an area for less dirt.  Oh, once you're past the tanks make it a worm composting chamber.

Good luck, looks like a fun project!

Howdy Dan , welcome to permies!  I like it. This thing is gonna get really hot in the summer so be sure to have some sort of openings to let that heat out.
I think you will find that during the growing season that you will be able to leave large areas open to allow for airflow and to let the bees in to pollinate your plants.

Looking good.

Any plans on how you're making (or what you're using) for your air/water heat exchanger?

I'm planning something similar, only I've got half the greenhouse running on a 'standard SHCS' with air tubes under the earth, and the other half with an old radiator painted black- heating up water and taking it to an insulated buried ibc container.

Mike,
Thanks for the detail review.  I am going 55 degrees from horizontal. I'm not sure what the standard reference is, horizontal or vertical.  This angle maximized solar radiation in the winter months.  I added a pic below of the solar radiation info I got from NREL at my location

I planned to slope the foam under the soil to prevent standing water, I know I may be wasting water here but I wanted to play it safe.  I also have no experience here, I just know good drainage is important.

I purchased some metal racks to put over the top of the tanks and weld or bolt to the tank cages.  I do need to do some calcs to make sure the soil weight won't bow the racks enough to make a low spot.

Glycerin sounds like a good option and the phase change temp seems perfect.  Cost is one issue though, from what I found it would be about $2700 to fill one tank.  I can always add it in the future as long as I make access tubes to the tank fill caps!

A vapor barrier sounds good for the osb.  I am concerned about moldling of the wood inside, and also about creating crevices for bugs when I insulate the inside walls.  I'm hoping I can use expanding spray foam to seal crevices.

I planned to add vertical timbers because the soil is mostly clay and things move quite a bit. I wasn't sure if the total weight of the greenhouse could be properly supported by the raised bed timbers.

Thanks for the tips on the glass. Maybe I can do light tests with the glass, to test which ones are best to use and which side should be out.  I'm thinking I could make a 2x4 fixture to hold the glass and on a sunny day put a light meter on the back side. put each panel on the fixture to test.  Just an idea.

A Walkway is good.  the tank access will be about 3' from the back wall which would be a good placement for the door.  I could align the door, walkway and access tubes 3' from the back.

I'm excited about designing and building an auto thermal curtain for the glazing with that Reflectix material. After doing the math to see how much heat is lost out of the glazing vs. the insulated walls I will lose about 4-5x more heat through the glazing! Even if the Reflectix only gives me an R value of 1, it will reduce the heat loss through the glass by 4x.

Miles,

Thanks for the tip.  My 8x10 greenhouse did get really hot in the summer and it was barely insulated! I'm planning to add a large window or vent at the peaks of each end.  I'll probably need a massive fan. My 8x10 greenhouse had a pretty powerful attic fan exhausting in the summer and it was still in the 80s inside.

Charli,

Nice to hear you've got a similar system! I was actually going to go the SHCS route but chickened out because of radon concerns (probably unwarranted considering that the air exchange out of the greenhouse is much higher than a basement) but nevertheless I went with water!

Do you have a fan blowing over the radiator? I have not chosen the heat exchanger, pump or fan yet.  I'm considering buying the heat exchanger from ebay. I see a huge selection of sizes and they seem reasonably priced.  I haven't sized it yet, but I was planning to error on the bigger side.

I was thinking of getting a 6" or 8" Can fan for moving air.  They are very reliable fans.

Dan Miano wrote:Mike,
Thanks for the detail review.  I am going 55 degrees from horizontal. I'm not sure what the standard reference is, horizontal or vertical.  This angle maximized solar radiation in the winter months.  I added a pic below of the solar radiation info I got from NREL at my location

Sweet, then you're good to go.  Yeah, I never know if people are talking from vertical or horizontal.  The sketch made it look like you were measuring from vertical.  My bad.

Dan Miano wrote:Glycerin sounds like a good option and the phase change temp seems perfect.  Cost is one issue though, from what I found it would be about $2700 to fill one tank.  I can always add it in the future as long as I make access tubes to the tank fill caps!

Ooof, that is a bit pricey.  I checked Bulk Apothecary and they're only $1800 and it comes in the tank already.  Still quite expensive.  If you know any biodiesel makers I believe glycerin is a byproduct they have to dispose of.

Dan Miano wrote:I planned to add vertical timbers because the soil is mostly clay and things move quite a bit. I wasn't sure if the total weight of the greenhouse could be properly supported by the raised bed timbers.

Gotcha.  I struggle with that myself.  How do you support a structure from a place underground that won't frost heave but then connect the structure to the ground so critters can't get in or it can't lift up your walls.  If you can insulate outside of the timbers, they hopefully won't get frost under them and they shouldn't move.  Maybe then they'd be good structural support?  Instead of going 3-4' deep with foundation insulation you can just go down a foot and then horizontally the remaining distance.

Dan Miano wrote:
I'm excited about designing and building an auto thermal curtain for the glazing with that Reflectix material. After doing the math to see how much heat is lost out of the glazing vs. the insulated walls I will lose about 4-5x more heat through the glazing! Even if the Reflectix only gives me an R value of 1, it will reduce the heat loss through the glass by 4x.

I'm excited for that as well!  If my heat plans fail and I have to go with nighttime insulation, I'll be very interested to see what you come up with.  Be sure to post the ideas here so we can optimize it  Don't forget that the R1 is only part of the benefit.  Radiant cooling to the cloudless night sky is a huge loss as well.  Reflecting that radiation back into the greenhouse will really help.

Nice! this our attempt at a similar project seems to add about 4 degrees at night on a sunny day. http://www.ecodiy.info/solar-greenhouse/

Dan,

This may be my first post ever!  Saw this thread and thought I'd forward an interesting video link (have you seen this already?).  The concept of annualized geo-solar intriques me; perhaps you can do something on a smaller scale.  Kind regards.  https://vimeo.com/168761278

Are you heating water in the summer to use for winter? Have you ever heated up a 300 gallon tank of water before?

"Before you can't get at it anymore, coil a whole bunch of pex or irrigation line down in the tank "

Dan Miano's green house project is twin of what I've been designing for the last year (28' X  40'). I have bought the same tanks. I heat them with a large Fresnel lens. This year (as soon as the snow melts!) I was going to "plumb" the pex pipe under the beds so I reacted immediately to this comment about the pex pipe. What do I need to know. As far as underground heating pipes - I chose the pex because I have been told is used in subfloor heating. Is there a better way to do this?
Thanks for the pictures.

Charli Wilson wrote:Looking good.

Any plans on how you're making (or what you're using) for your air/water heat exchanger?

I'm planning something similar, only I've got half the greenhouse running on a 'standard SHCS' with air tubes under the earth, and the other half with an old radiator painted black- heating up water and taking it to an insulated buried ibc container.

Do you add anything to the water to inhibit microbial growth, or is it not an issue?

Dan Miano wrote:
Do you add anything to the water to inhibit microbial growth, or is it not an issue?

So my water system is this:

- an old steel radiator painted black in a wooden case, with styrofoam insulation behind and a sheet of acrylic infront. It is plumbed to PEX pipe that is lagged, and the pipe goes into an ibc container, where there is a length of copper pipe as a heat exchanger (it was meant to be a spiral shape.. I never quite achieved the spiral, but it isn't a straight bit of pipe either). In the pipeline there is a pressure release valve so nothing can blow up. And a pump and a temperature monitor. When the temperature probe senses a temperature warmer than the ibc water it closes a relay and the pump comes on, so that warm (hopefully!) water from the radiator goes into the ibc, to be replaced by colder water from the rest of the plumbing. The water in the closed plumbing system is just water, nothing added. I probably should have added inhibitor or something, but I haven't! It doesn't really get cold enough here on an average year to freeze pipes.

The ibc water isn't just for a heatsink, though it is underground and so in the dark. I use the water for watering the greenhouse and refill it from rain water, so it isn't the same body of water all the time. I don't put any kind of microbial inhibitor in it. compared to my above-ground tanks (which go rather green, even though they're shaded from the light too) it seems to stay remarkably clean. Any rainwater going into it is filtered over a sponge filter to remove larger bits of debris, but that is it.

Charli

Chris Southall wrote:Nice! this our attempt at a similar project seems to add about 4 degrees at night on a sunny day. http://www.ecodiy.info/solar-greenhouse/

Chris,  That's awesome!  I like the design.  Thanks for sharing.  

kw hatten wrote:Dan,

This may be my first post ever!  Saw this thread and thought I'd forward an interesting video link (have you seen this already?).  The concept of annualized geo-solar intriques me; perhaps you can do something on a smaller scale.  Kind regards.  https://vimeo.com/168761278

I had not seen that video before, thanks. I just watched it.  I was on the verge of doing a 4' deep SHCS (Subterranean heating and cooling system) like Charli but decided it wasn't complicated enough for me
 That's very interesting to use a thermal mass large enough to utilize summer heat months later.  I'm anxious to see how my system performs with respect to storing heat for multiple days.  I think if I reduce the night heat loss through the glazing with a thermal curtain, then the 550 gallons of water may get me through 1 or 2 cloudy winter days without backup heat. We'll see!

ronie dee wrote:Are you heating water in the summer to use for winter? Have you ever heated up a 300 gallon tank of water before?

No I haven't heated a 300 gallon tank before.  Any advice is welcome!
I'm not heating in the summer for winter use.  I'm heating the water in the fall/winter/spring with the excess heat during the day and then transferring it back into the greenhouse at night.  So the heat stored one day will be used that same night.  I'm hoping the 550 gallons of water will be able to get through 1 or 2 cloudy winter days without backup heat.  
During the summer I plan to have a large fan to expel the excess heat.

What is OSB please?

It's Oriented Strand Board.  They take wood chips and glue and press it together into a cheaper version of plywood.  

Mike Jay wrote:Ooof, that is a bit pricey.  I checked Bulk Apothecary and they're only $1800 and it comes in the tank already.  Still quite expensive.  If you know any biodiesel makers I believe glycerin is a byproduct they have to dispose of.

Thanks for the tip.  I'll do some research in my area.  

It seems like you could get automotive radiators from junkyards that have electric fans built into them. Hanging from the north celing they would circulate the air in the greenhouse cooling it during the day and warming it at night. They would not have to be run at night during the summer so the tanks will start off hot. You could probably use a separate radiator for each tank only drawing heat from one tank on mild nights conserving heat for cloudy days or very cold nights.

As for the planters, wicking beds work very good in my greenhouse.  I layed down a water tight fabric with the sides and ends folded up to the height of the bed. The bottom was sloped 2" to one side and I placed a 3" slotted drain pipe on the deep side with an elbow on one end coming to the surface. I keep a plastic bottle in that as a float level.  When I water from the surface any excess drains into the slotted pipe. but once it is uniformly damp keeping the drain pipe full wicks water back into the soil keeping it uniformly damp. I only do this on the south edge of the greenhouse. the rest of the greenhouse has 3' tall barrels with the drain tube circled in the bottom. They work the same and I can put them on a hand truck and move them around and outside for the summer.

I have some experience here.  I built a similar size solar greenhouse about 5 years ago, with Twinwall polycarbonate glazing on south wall and east wall.  (This last was probably a bad idea.)  Initial heat storage was in 1 gal plastic water jugs stacked along north wall.  Not very efficient at capturing heat basically by radiation alone.  And the jugs broke down after a year or two.  Now have 2 tanks of 450 Litres, (total of about 230 US gallons), with water circulated through 2 car radiators in the peak of the greenhouse by mini 12 V. submersible pumps   On sunny days, return from radiator is about 3 degrees C above water going to the radiators.  But there is no way that I can maintain a temp of 50 degrees F through the night - it generally falls to just above freezing, despite my running the pumps and fans through the night, (essentially drawing the heat back out of the water in the tanks) (I am in zone 5b in Nova Scotia, Canada, with outside temps dropping at times as low as minus 10 F, (-22 C).  My next direction is going to be at reducing heat loss, rather than trying to increase heat storage.  (The water gets up to + 20 C after 5 days of accumulated sunshine, dropping gradually to about 4 C after 5 days of heavy overcast, snow, and generally unpleasant weather.)
Incidentally, the cheapest and easiest way to provide exhaust fans during the summer is with a hot-air furnace blower, (available from furnace repair places when they replace somebody's old furnace)

Hans Quistorff wrote:It seems like you could get automotive radiators from junkyards that have electric fans built into them. Hanging from the north celing they would circulate the air in the greenhouse cooling it during the day and warming it at night. They would not have to be run at night during the summer so the tanks will start off hot. You could probably use a separate radiator for each tank only drawing heat from one tank on mild nights conserving heat for cloudy days or very cold nights.

As for the planters, wicking beds work very good in my greenhouse.  I layed down a water tight fabric with the sides and ends folded up to the height of the bed. The bottom was sloped 2" to one side and I placed a 3" slotted drain pipe on the deep side with an elbow on one end coming to the surface. I keep a plastic bottle in that as a float level.  When I water from the surface any excess drains into the slotted pipe. but once it is uniformly damp keeping the drain pipe full wicks water back into the soil keeping it uniformly damp. I only do this on the south edge of the greenhouse. the rest of the greenhouse has 3' tall barrels with the drain tube circled in the bottom. They work the same and I can put them on a hand truck and move them around and outside for the summer.

Thanks Hans,
I like the idea of getting an inexpensive heat exchanger and fan in one piece.  I would need to do some research to see if the heat transfer rate will be sufficient though.
Glad to hear you have a successful wicking bed going.

David Maxwell wrote:I have some experience here.  I built a similar size solar greenhouse about 5 years ago, with Twinwall polycarbonate glazing on south wall and east wall.  (This last was probably a bad idea.)  Initial heat storage was in 1 gal plastic water jugs stacked along north wall.  Not very efficient at capturing heat basically by radiation alone.  And the jugs broke down after a year or two.  Now have 2 tanks of 450 Litres, (total of about 230 US gallons), with water circulated through 2 car radiators in the peak of the greenhouse by mini 12 V. submersible pumps   On sunny days, return from radiator is about 3 degrees C above water going to the radiators.  But there is no way that I can maintain a temp of 50 degrees F through the night - it generally falls to just above freezing, despite my running the pumps and fans through the night, (essentially drawing the heat back out of the water in the tanks) (I am in zone 5b in Nova Scotia, Canada, with outside temps dropping at times as low as minus 10 F, (-22 C).  My next direction is going to be at reducing heat loss, rather than trying to increase heat storage.  (The water gets up to + 20 C after 5 days of accumulated sunshine, dropping gradually to about 4 C after 5 days of heavy overcast, snow, and generally unpleasant weather.)
Incidentally, the cheapest and easiest way to provide exhaust fans during the summer is with a hot-air furnace blower, (available from furnace repair places when they replace somebody's old furnace)

David,

Thanks for sharing your system. Do you have any pictures?  What's the temperature difference inside and out in the middle of the day?  I'm curious how much heat the radiators capture.  Are the tanks in the greenhouse?  My plan up to this point is to use something similar to this magicaire exchanger http://www.magicaire.com/wp-content/uploads/2014/01/SHW-2.2-4-6-09.pdf
This is what Russell used in the project I'm following: http://www.builditsolar.com/Projects/Sunspace/GreenhouseHX/HeatPumpEnergyStorage.pdf  
I still need to look at the data to figure out the best size.
It looks like your temps are pretty close to mine, in Jan your avg is 5 degrees cooler than Denver.

I think reducing heat loss is a good idea.  Here are my calculations for heat loss and gain for my greenhouse.  You can see how much is lost through the glazing!  I have not figured out how to calculate heat gain into the greenhouse yet but I do know how much solar radiation the glazing will receive.
My "curtain" will use the reflectix material I mentioned in the original post.

Pool blankets make cheap insulators
Or take a page from the double poly green house .
Though a sheet of poly over and use a fan to inflate it.

BTW.  Glycerin is a by product of biodiesel.

Tom Robertson wrote:Pool blankets make cheap insulators
Or take a page from the double poly green house .
Though a sheet of poly over and use a fan to inflate it.

BTW.  Glycerin is a by product of biodiesel.

Thanks for the tips Tom,
Mike mentioned that about glycerin.  That would be great if I could get it free.  Part of my goal is to see how many of the materials I can gather for free.  Craigslist's search alerts have been working great for me!

I will look into the inflated barrier.  I think I could fairly easily attach and seal poly to the inside of the 2x6s that the glazing will rest on, but I'm sure there will be some pros and cons to weigh between this and an auto insulating curtain.

I am getting way beyond my competence here, but perhaps my data will be useful, and somebody who knows what they are talking about can check my reasoning.  The question was posed as to whether a car radiator, with air blown through it by a fan is an efficient heat exchanger.  (I am in Canada, so tend to use metric measures.)  I have a temperature monitor which records at 20 minute intervals the temperature of the water in my storage tank, (going to the radiator), and the temperature of the return from the radiator. On  Feb. 3, 2017, (a sunny day),  the temperature of the return flow first stated to rise above the flow to the radiator at 0945, and this differential persisted until 18:45, (coasting on the heated air in the peak of the greenhouse evn after the sun had set).  Thus we had a total of 9 hours of heat collection., or 28 time intervals..  Now we get into alligators.  I subtracted temp To from temp FROM at each interval to get the increase in temp during that interval.  (Was this legitimate?)(The gain ranged from 0.2 degrees C up to 11.35 degrees C in the 20 minute block)  I then treated this as 27 trapezoidal blocks of heat gain and calculated the area of each block using the formula 1/2(T1 +T2), where T1 and T2 are the height of the two sides of the trapezoid, and multiplied each block by 1/27 to get the amount of heat gain in that block.  (I believe this corresponds to something called Simpson's Rule (?))  Finally I added the area of each block, to get a total of 6.414 degrees C rise, essentially integrated over the 9 hour  period.  
The flow of the pump is 3.15126 L/min., giving a total volume of  1701.68 Litres of water flowing through the system during the 9 hour period.  1 kilocalorie heats 1 litre of water 1 degree C, so I captured 6.41435 X 1701.6804 = 10915.17 Kilocalories , or 43, 314.88 BTUs.  And, with two tanks, (of 450 Lites each), the total energy capture for the day was 86,629.76 BTUs.
Now, what does all this mean?  Assuming that my methodolgy sounds reasonable, (and I am distinctly uncertain about the legitimacy of, for example, my summing the areas of 27 trapezoidal blocks to get a total area under the curve, is appropriate), and comparing my total to the elegant heat loss calculations above, I am not capturing enough heat to last more than a day or two of cloud.  But the efficiency of heat transfer from air to water, across the radiators is not bad, (a gain of up to 11 degrees C during the passage through the radiator, at a flow rate of 3.15 L/min, (0.83 US gallon/minute).)
Have at me!

I love your trapezoidal math system.  I believe it's correct and should be within a few percent of the true heat gain.  One question.  How did you get from 43,315 BTU's to 86,630 BTU's?  I believe the heat gain from the radiator is a good number but it doesn't double just because the tank got twice as big.  Or do you have a second radiator to go with the second tank?

I like this method to capture heat in thermal mass.  Instead of putting your mass in the sunshine (where plants could be) you hide it anywhere convenient and move the heat to/from it.  

One question.  How did you get from 43,315 BTU's to 86,630 BTU's?  

This is the only bit I am confident of.  I have two tanks, two radiators, 4 fans, (2  in each radiator).  Just in passing, the suggestion was made earlier to simply use the installed fan on the junk radiator.  These fans draw an awful lot of power.  Large fans designed for computers are both much less energy demanding, (somewhere around 5 or 6 watts each), and, (for me more significant),  a lot quieter.  (I have to listen carefully to tell whether mine are running)

Ok, I was missing that part.  Cool, that's a neat way to store a lot of BTUs.  I'll have to consider that in my design.  I wonder if you'd get the heat back out of your system better if you routed the warm water through your soil instead of into the air up at the radiator?  The peak of the roof is a great place to collect heat but I'm guessing you'd rather return that heat closer to the ground.  I've seen studies that say heating the soil helps the plants much more than heating the air (assuming the air isn't freezing).

The other thing I like about moving the heat to a thermal mass storage tank is that the size of the tank doesn't affect the speed of the heat return to the system.  I'm making these numbers up but I'd guess that if a person used 350 1 gallon water jugs the heat gain would be great due to the added surface area.  But the heat would get all used up before the sun came back up.  Seven 50 gallon drums may accept less heat due to their reduced surface area but they may bleed heat for 24 hours.  A 350 gallon caged tote tank would accept heat much slower yet but hold onto that heat for days.

It's easier to plumb one big tank rather than interconnecting many smaller tanks.  So if you can gather the heat and move it to the tank (say with a radiator) and then distribute the heat AS NEEDED to where the plants need it (low to the ground or underground) it may be the best thing ever...

So if you can gather the heat and move it to the tank (say with a radiator) and then distribute the heat AS NEEDED to where the plants need it (low to the ground or underground) it may be the best thing ever...

Dead on!  And I have considered the idea.  Technically it is actually not that hard to effect with a couple of differential thermostats.  (I think I would simply use two extra pumps, rather than try to rig divertor valves.)  Do you have any evidence one way or the other of the importance of soil vs. air heat?  Is the limiting factor in growing through the winter the temperature (either air or soil) or light availability?

I agree, two pumps would probably be easier than a diverter valve.  At least for my electrical skills.

I don't have evidence from personal experience but I've read that soil heat is more valuable in a number of places (newer greenhouse books and the internet).   The gist I got from it was that if you have the choice of 50 degree air and 40 degree soil or 40 degree air and 50 degree soil, the plants would do better with the warmer roots.  I believe this was for "normal" greenhouse plants like greens and cool weather crops.

What you are trying to grow will dictate if the temps and/or the light is your limiting factor in the winter.  Elliot Coleman grows plenty of greens in Maine under a row cover and hoop (blocking 20+% of his limited light) and he can harvest all winter.  Those greens have limited growth in the winter but they stay perky and harvestable all winter.  If you want fruiting plants in your greenhouse (tomatoes, etc) both light and heat will be an issue.  

David Maxwell wrote:I am getting way beyond my competence here, but perhaps my data will be useful, and somebody who knows what they are talking about can check my reasoning.  The question was posed as to whether a car radiator, with air blown through it by a fan is an efficient heat exchanger.  (I am in Canada, so tend to use metric measures.)  I have a temperature monitor which records at 20 minute intervals the temperature of the water in my storage tank, (going to the radiator), and the temperature of the return from the radiator. On  Feb. 3, 2017, (a sunny day),  the temperature of the return flow first stated to rise above the flow to the radiator at 0945, and this differential persisted until 18:45, (coasting on the heated air in the peak of the greenhouse evn after the sun had set).  Thus we had a total of 9 hours of heat collection., or 28 time intervals..  Now we get into alligators.  I subtracted temp To from temp FROM at each interval to get the increase in temp during that interval.  (Was this legitimate?)(The gain ranged from 0.2 degrees C up to 11.35 degrees C in the 20 minute block)  I then treated this as 27 trapezoidal blocks of heat gain and calculated the area of each block using the formula 1/2(T1 +T2), where T1 and T2 are the height of the two sides of the trapezoid, and multiplied each block by 1/27 to get the amount of heat gain in that block.  (I believe this corresponds to something called Simpson's Rule (?))  Finally I added the area of each block, to get a total of 6.414 degrees C rise, essentially integrated over the 9 hour  period.  
The flow of the pump is 3.15126 L/min., giving a total volume of  1701.68 Litres of water flowing through the system during the 9 hour period.  1 kilocalorie heats 1 litre of water 1 degree C, so I captured 6.41435 X 1701.6804 = 10915.17 Kilocalories , or 43, 314.88 BTUs.  And, with two tanks, (of 450 Lites each), the total energy capture for the day was 86,629.76 BTUs.
Now, what does all this mean?  Assuming that my methodolgy sounds reasonable, (and I am distinctly uncertain about the legitimacy of, for example, my summing the areas of 27 trapezoidal blocks to get a total area under the curve, is appropriate), and comparing my total to the elegant heat loss calculations above, I am not capturing enough heat to last more than a day or two of cloud.  But the efficiency of heat transfer from air to water, across the radiators is not bad, (a gain of up to 11 degrees C during the passage through the radiator, at a flow rate of 3.15 L/min, (0.83 US gallon/minute).)
Have at me!

Great analysis!  Thank you.  I agree with that approach, since the amount of heat transferred decreases as the difference between the AIR IN temp and WATER IN gets smaller.  Do you know about what the AIR IN temp was,  and how many watts the water pump and fans use?  With the AIR IN temp I could compare it to the Magicaire hot water coil spec sheet.
So you replaced the radiator fans with computer fans?

I'm wondering if it would work better or worse to circulate the water through a few hundred feet of black irrigation pipe up in the peak of the roof as a direct solar collector (instead of the radiator and fan).  If the sun can strike the pipe and it's in the hottest part of the greenhouse it would pick up both radiant and conductive energy.  This would be only for heat collection, not distribution.

Does anybody have experience installing frameless double paned glass on a greenhouse?  This is one area I'm not too confident in how to approach it.  My plan was to use 2x6s to build a truss and have them resting on the 2" portion of the 2x6, probably putting a rubber strip down on the 2x6, using plastic brackets screwed to the 2x6 to hold the glass in place, leaving a 1/2" gap between 2 panes and caulking this gap.  The 2x6s will be 55 degrees from horizontal.

Does anybody have experience installing frameless double paned glass on a greenhouse?  

In my own greenhouse I acquired some frameless glass, very cheaply, (double glazed, argon-filled, low-e coated,  original customer withdrew from manufacturer).  But they are relatively small, (about 2 1/2 ft X 3 1/2 ft.).  I mounted them on the South wall, vertically.  And I simply made hinged frames for them, permitting my opening them for ventilation in hot weather.  (I seal them in winter).  That is, an alternative to a frameless installation is to frame them.  And doing so may have advantages, in making them movable.

I have an aquaponics greenhouse with 1 300 gallon IBC above ground for fish and and another 200 gallon in the ground in a similar setup as yours for a sump. Even with the 2" foam board around the sump heat does leave it to the ground water around it. I just found this guy on YouTube last night and thought I would share it with you. Might help.  
This year I plan to build a rocket mass heater and use the mass as a bed for 2-3 IBC fish tanks to heat my fish water and greenhouse. Water is a really good thermal mass. I thought about geothermal but I think this will work much better. Putting your tanks in the ground I think you're already creating a disadvantage since the ground will suck out a lot of the heat even with insulation.

Jeffrey Sullivan wrote:I have an aquaponics greenhouse with 1 300 gallon IBC above ground for fish and and another 200 gallon in the ground in a similar setup as yours for a sump. Even with the 2" foam board around the sump heat does leave it to the ground water around it. I just found this guy on YouTube last night and thought I would share it with you. Might help.  https://www.youtube.com/watch?v=6SD5HpXZc1w
This year I plan to build a rocket mass heater and use the mass as a bed for 2-3 IBC fish tanks to heat my fish water and greenhouse. Water is a really good thermal mass. I thought about geothermal but I think this will work much better. Putting your tanks in the ground I think you're already creating a disadvantage since the ground will suck out a lot of the heat even with insulation.

Thanks for the info.  So you had 2" foam around your 200 gallon buried tank? Do you know what the temp change was over a given period of time?  From my calculations I should lose around 3,500 BTUs per day to the ground, through the bottom and sides.  I hope the actual loss matches my math!

Thanks for the link. That's great info   I previously came across some of the info he presented on builditsolar here: http://www.builditsolar.com/Projects/Cooling/EarthTemperatures.htm

I have an extremely high water table and when there is a good rain the sump will actually start to float if it's not filled above the water table level. I only noticed that the water temp outside the tank is pretty close to the temp inside the tank. No way of knowing how far the dissipated heat goes out underground. Question there is does this give you a larger heat storage area or are you having to generate more heat to compensate the loss.

Jeffrey Sullivan wrote:I have an extremely high water table and when there is a good rain the sump will actually start to float if it's not filled above the water table level. I only noticed that the water temp outside the tank is pretty close to the temp inside the tank. No way of knowing how far the dissipated heat goes out underground. Question there is does this give you a larger heat storage area or are you having to generate more heat to compensate the loss.

Ok I see, so the ground water is coming in contact with the tank and drawing all the heat out.  I don't think the water table is very high in Denver, but I'm not sure where to find that info.  There is at least one company here that installs the geothermal system in the video you sent, so I'd think that system would have issues as well if the ground water was high.  It sounds like keeping the tank above ground is the only option with a high water table.