Passive-solar has been known to work well but certain things about our architecture fail the thermodynamic efficiency test and there are reasons why.
First is heat-transfer through conduction, the wood, drywall, sheathing all having rather high heat-transfer coefficients. Next is convection from a heat-riser which takes the warmth of the floor and moves it to the ceiling.
Until these are dealt with they cost a lot of energy, if one is trying to not require external heating for a building.
A first principle to recognize is that we need to insulate the OUTSIDE of a building to prevent heat-transfer from anything nailed to the studs which act like pathways for heat to be lost to "radiators", the sheathing & siding of the building. Insulation between the studs doesn't do much to prevent this, a direct insulation of exterior heat-transfer surfaces needs to be added, using insulated sheathing is the simple way to do this.
The second is if you don't collect heat you can't create an autonomous building that maintains comfort zone, followed by if you don't have enough thermal mass to heat during the day the building will cool back down before dawn.
Those are the three pieces to use in any situation but standard framing and insulation techniques need to be altered to gain this idea in a design.
The largest single key for comfort is to create and circulate warm air during the day that heats up the thermal mass of the building. If you use insulation on the outside of the structure all of its mass becomes thermal-mass to maintain comfort zone for the interior instead of being able to conduct heat to the outside world.
Then, a greenhouse wall is mandatory in passive solar, so how to get it to function as an air heater requires two panels with an air gap to warm the air well, this is needed to force a flow from floor to ceiling, but that's not good enough, the warm air from the ceiling needs to be drawn all the way down to not lose too much by it all gathering high in the room.
This implies using a slightly different way of building. First the typical plywood sheathing is used on the INSIDE of the walls opposite where the greenhouse wall is with joists between the roof-bearing wall used as ductwork so between studs is open near the ceiling with inlet vents and inside the wall open to the floor joists inside the walls & floor so the air must pass through the floor joists to get to the double greenhouse panels that heat it. If done correctly this air-flow will not require fans. The floor joists are covered by exterior insulated sheathing to open the space to this airflow.
The result is a space heater built into the home that keeps the floor warm on a daily basis, uses all the mass of the building as thermal storage by insulating outside the structure and avoiding easy heat-transfer paths. I'm working on ceiling systems that fit into this scheme, to gather heat or cold for the needs of the season but wanted to put these concepts out to people on the list to consider in their remodels & building from scratch.
Tom, could you provide a sketch of this concept? I understand keeping the structural members inside the insulated envelope, glazing over the exterior surface for the greeenhouse wall, using the interior air space between the joists for air circulation, and relying on the interior gypsum board and plywood sheathing for thermal mass, but am unclear how the warm exterior air gets passed to interior air cavity and floor joists. Is there an air opening or louver?
Jim, quick reply from work, the solar wall is double glazing, very similar to doubled panes in thermo-panes but the top & bottom are open.
This narrow space is the heat-riser and the concept is to create a lot of flow in it to be able to suck the hot air from the ceiling down under the floor to it. All the stud & joist spaces have to be caulked to keep leakage as low as possible for sure to make this happen.
Then to prevent cold air reversing everything there are going to be bi-metal louvres to shut it down at night.
Hi Tom, I could use a picture as well, but from what you have said, it sounds like you will be creating a large convective loop with the hot air created in the gap of the solar wall double glazing. When you use the term "heat-riser" I have to jump in and point out that heat does not rise. Yes, we often see it move up, but it is the colder air pushing down somewhere else in that loop that creates the movement. The importance of this, for what you are doing, will be calculating the forces that are powering that convective loop. Now, if you are way ahead of me, I would be interested in seeing your calculations, as there are many assumptions and my confidence level in what I'm using needs all the help it can get. In either case, you help me or I help you, we might be able to predict the air flow and then see if that is what is achieved.
My apology for not doing it sooner, will get drawings up tonight later, need to scan them.
Bud, totally appreciate your help however you see fit.
To explain, you have to confine the "heat riser", and I'll say that if air expands in the glazing space it's easier for it to rise than sink so that's where the term is coming from, you don't get a sink from hot air, you get expansion and if that's in a cooler mass it'll rise; otherwise, it's pretty slim to think ceiling air will get sucked back to the solar wall without a fan and that's the goal, no fans. In this case it's using the cooler air as a pressure gradient to expand upward, far as I can tell.
I'm using the standard framing as the ductwork by not using batting in those spaces and using foam sheathing outside and plywood sheathing inside to form the duct wall. All that will get caulking to seal it well as ductwork so there are few leaks to the bottom of the solar wall.
So, the expansion of the heated air gets away easiest into the room by rising, but, that creates a flow and once that flow gets going it should self-regulate to some degree, this being an experiment to gather metrics, after it's built it'll be instrumented to help tune it.
Gravity is the moving force and it is the colder air winning the battle for the bottom that displaces the warm air and pushes it up (buoyancy). If you look at your loop and consider it as two vertical columns connected by two horizontal columns, one high and one low, then the moving force would be the difference in the weight of air in one column vs the other and that weight difference is a function of temperature and the height of the column. For a first order approximation, use this equation.
P (in Pascals) = 0.0069 x h x (T1-T2) in degrees F
Since you will need air flow simultaneously out the top of the hot air column and in the bottom, h is half the height of the columns. T1 and T2 are the average temperatures of each column.
From the attic ventilation and stack effect numbers I have been working with, an eight foot ceiling will produce very modest pressures with 30° or 40° temp differences. What height and temperatures are you anticipating?
The main solar wall now uses 12ft panels, the rear wall is 8ft, so, this is a no-brainer if I get it right. My original for another site used the reverse, 8ft solar height because I've found solar walls gain more than design specs over time by heating everything any time it works and we just calculate averages is the thought, many having to have windows open all winter after five years of heating the soils at the site.
This site is a Pacific NW mountain ridge with decent solar gain but windy, cloudy & cold a lot, snow load is 4-6ft average, winds on this particular ridge are quite high but trees soak up most of it, southern Olympic Mts. so plenty of cloudy days but not dark so IR is OK most of the time.
Anyway, by adjusting the gap on the solar panes I adjust the flow velocity and that's important to sucking up so much colder air volume per hour and that heats the floor & wall mass, have used smoke to detect currents on a standard greenhouse wall and came away with using the panes to restrict the flow and thus create power to use for circulation.
I'm planning on instrumenting this once it's built, early spring.
So, thanks Bud, I get the gravity relationship though not formally so appreciate you pointing it out, and in this case you have a venturi, constriction, from the joist space [23.375"x8.625"] into the narrow & wide glazing space, and, the volume of colder air is in the floor between joists is quite a bit larger than the wall volume [2x4 vs 2x10] so that fits your concerns on the mass density of the heating using bouyancy to move air that is one-way from the ceiling. I can evaluate equations if you have formulas for this.
This will spin off so much volume of colder air per second and emit so much volume of warmer air at the top into the room.
The vents at the top of the wall that gather the warm air then are only 8ft max from the floor while the exhaust of the solar glazing is about 11ft up, consequently the delta T is smaller by some amount, the roof pitch is 4"-in-12".
I'm not finished working the roof & rafters into this, the roof is a huge gain much of the year and a cooling in summer at night so not done on how to use that space yet.
I see you prefer pencil and paper as well :). and coffee. I finally added a small graphics program so I could go directly to images I can post. Still practicing, but will soon post my first attempt at some graphics to help explain the movement of warm air.
I practiced with a 12' air column this morning at breakfast to see what pressures might be available. Just for practice I used an 85 degree difference between two columns of air. The weight difference will provide in the range of 4 Pascals of moving force to power the convection. If less delta t, then less delta p and more with more.
1. Once the hot air is moved out of the solar panel, where does it give up its energy?
2. What supply and return temperatures would you consider normal? Supply being the solar collector and return being the complementary vertical column of cooler air that powers the convection.
3. I wasn't able to visualize the 11' and 8' dimensions you mentioned and how they are connected. Essentially, they will both need to be defined as 11' columns of air to determine the pressure differences.
Tom, I posted my link onto another thread so we don't get sidetracked on this thread. To review my presentation on "hot air", go to:
I gave this some more time at breakfast this morning and here are some thoughts. To power convection, you want heavy (cold) air pushing down somewhere and hot (light) air in a position where it can be pushed up. If the column of cold/cool return air becomes warm air, the convective loop will slow down. If there is a pool of colder air at the bottom it will not be contributing to the process. Suggestion, if you could extend the heat from the collector through conduction (think metal plates) or other means down to as low a point as possible, you will be creating a volume of warmer lighter air directly below your collector supply column and the cold air adjacent to it will force it up as part of your convective loop. Something similar on the return side might help as well by extending the cold from below up into the return column to make that air heavier.
Warm and cold air are affected by gravity and the cold air wants to go to the bottom while the hot wants to go to the top and both will want to stay there until forced to move. But conductive and radiant heat transfer are omnidirectional. If you can use this property to add or subtract heat from places that will assist your convective process, the engine will run over a wider range of temperatures.
In the buoyancy model, the warm air has a high-pressure vs what's in the floor some delta-T from the warm air so more dense by some amount. This cooler air has a single outlet from the ceiling vent and that's a venturi going into the solar collector that forces it into a lower volume space and to expand, so it's expanding air and by what you say that's sitting on a pool of cooler air.
This seems to fit what you're saying, the floor will be colder, the ceiling warmer, I'll be creating a flow velocity up a slot that adds volume to the ceiling and as the wall air cools it contracts as it moves to below the floor by expanding the air in the slot.
The delta-T isn't very high because it's insulated pretty well from the outside world, yet that means all the wood is thermal-mass to add to thermal inertia to maintain comfort-zone for the daily cycle.
This also means that while conductive losses are omni-directional by insulating the outside you reduce those losses versus the standard of siding-sheathing-batting between studs-drywall which can be looked at in cross-section and seen to be radiative systems with studs or joists as conductors between inside and outdoors so the batting while helping has an upper limit due to the conductivity of the design.
By moving the insulation to the outside of the studs you isolate the inside from this effect.
Hi Tom, we are still not totally connected in our thinking, but improving. Here is another association that may add to the confusion.
When the air in your collector expands because it is heated, this does not contribute to the convective flow. This is a loop and that expansion is being countered by an equal contraction elsewhere in the loop. If the net temperature of the loop increases, there is a volume increase which will escape to the outside. Conversely, if the net temperature of the loop decreases, then there will be a volume decrease which will bring in outside air. As soon as I understand your cavity arrangement better I will draw a diagram with the related pressures, all created by the density/weight of the air.