V. A Sustainable Greenhouse

Here's an economical zero-fuel, all-winter greenhouse. It has been running since late 2018 in East Greenwich, RI. It puts 150% of our relatively weak winter sun onto the plants and it solves humidity issues well. It can work in harsh winter climates. We used ordinary glass windows, not sheet plastic or a plastic/fiberglass mixture. All greenhouses and all buildings would benefit by storing solar heat.

This chapter explains, item by item, the design choices that were made.

Raising the reflector wall of the greenhouse.

V1. Preamble: parabolic dishes and linear troughs:

This parabolic satellite dish has been aimed by the dish's installer straight at a particular orbiting satellite. The satellite orbits the earth exactly once every 24 hours, so that the satellite appears to always stay at exactly the same spot in the sky as seen from earth. The dish bounces the satellite's weak broadcasting signal onto a small target, greatly magnifying the signal's strength.

A linear trough focuses sunlight onto a central pipe in order to heat water flowing inside the pipe. The linear trough is aligned in the east-west direction, so that when the sun moves across the sky from due east to due west, each square inch of mirror stays focused on at least some part of the pipe. A parabolic linear trough can concentrate up to 30 times of reflected sunlight.

A heliostat mirror with a long-range focus, such as a focus on a power tower 200 meters away, has an almost flat but not perfectly flat mirror with a barely perceptible parabolic arc. A great mirror would put an exact image of the sun onto the power tower's black heat-absorbing surface. In practice a somewhat cheaper imperfect mirror will more or less get the job done. A flat mirror might work for shorter range reflection.

A few cautions: no one should ever stare at the sun directly because it can cause eye damage. I've heard one recommendation that even 6 times normal solar concentration is getting into a danger zone because heat buildup can lead to melting and fire hazards. As a practical safety measure I purchase my eyeglasses with a protective UV coating, and I would recommend this precaution for any solar experimenters or builders.

Caution: Linear troughs are unpopular because the overconcentration of sunlight sometimes starts fires. A friend built a linear trough and a metal pipe filled with water on a plywood backing. First he achieved a boiling water geyser out of the heat absorber pipe. Then the plywood backing caught fire. The Vdara Hotel in Las Vegas was built roughly in the shape of a parabola, with many reflecting glass windows, and it had a problem. It kept setting its poolside patrons' hair on fire.

To the right is the not-quite-parabolic reflecting wall of my Rhode Island greenhouse. The south-facing reflecting wall uses rows of flat 12” x 12” mirrors tipped at five angles, where the middle row is 18 inches high and the other four rows are 12 inches high. By building five flat reflecting rows, I know that no square inch of the greenhouse's row of target windows will receive more than five reflected suns. It usually takes somewhat more than 10x reflected suns to start a fire. As a matter of fire and overheating safety, 5x is about my personal limit.

As the sun moves from due east to due west across the sky, the linear trough continually reflects sunlight onto some part of a line of target windows on the greenhouse's north wall. The solar reflection slides from west to east across the window as the sun moves from east to west.

V2. The Greenhouse's History

My first tiny parabolic reflector in 2007 used a piece of plywood warped with straps into a curve. Technically, warping plywood creates a smooth section of circle, an arc, which isn't quite a parabola but it's close. Then I glued on glass mirror pieces that I had cut with a glass-cutting tool.

At a certain distance away from the target, all eight rows of three inch high glass mirrors focused rather precisely through the model's cutout window area, or in the picture, onto a piece of cardboard placed in front of the target window area. My wife and I sat in the front yard and watched the horizontal motion of the bright solar reflections across the cardboard as the sun moved across the sky.

Invention is a matter of solving side issues one by one. Early on, the Wright Brothers had to find an airplane motor with a high ratio of power to weight. Then they had to get the curvature of the plane's wings correct by testing various wings in a simple wind tunnel. Then they had to think about steering the airplane.

Getting eight times normal sunlight into a greenhouse through a small window might be a great idea because the rest of the greenhouse can be well-insulated and inexpensive, but will the greenhouse maintain its heat all night? My second prototype greenhouse was 4 feet high and well-insulated on top and on its sides. I put a large plastic water barrel inside the greenhouse for thermal heat storage, and I placed the greenhouse on a bare lawn in a friend's backyard. I may have overspent on the greenhouse's nice shingled roof, but I wanted it to look nice.

For the next reflector wall I cut a curve into the sides of wooden 2” x 6” boards so that the 4' x 8' plywood sheet would permanently hold its curvature. I also cut two air holes in the middle of the plywood because various people were cutting wind holes in political protest banners made out of bedsheets. Their goal was to reduce the banner's overpressure in high wind gusts.

Results were mixed. The system was weatherproof. I only had to change the reflector's focus once a week in February and not at all in December, because the sun rises to nearly the same elevation each week in December. Not that I expected any problems, but I observed no fire or heat melting problems all winter, so having eight rows of flat reflectors wasn't unworkable. When I stood in front of the concentrating reflector wall my hand felt nice and warm in winter temperatures, but not hot.

The greenhouse had a major heat holding problem. Most of its captured heat sank down into the moist topsoil, then migrated sideways below the insulated walls, then came up and was gone. The plastic barrel of water had some ice freezing inside the barrel but it never completely froze, so that the measured temperature inside the greenhouse might remain at 30 to 32 degrees for a week. The grass inside the greenhouse loved the indirect sun off the back wall, the relative warmth and the lack of snow cover all winter, except the grass underneath the barrel received zero sunlight. The prototype was too small for me to think about how the incoming sunlight would actually get to the plants.

I tried insulating the ground just outside the greenhouse with dry bales of straw. It didn't help much. Most of the greenhouse's heat still leaked out. My main problem had to do with the greenhouse's size. I've seen large solar greenhouses in Wisconsin that stored heat well in midwinter because their huge solar inputs were far greater than their heat leakage through the topsoil around the greenhouse edges, and the heat had to migrate through many feet of insulating soil to get outside.

V3. The Attleboro Prototype

A year later I found someone who wanted an algae greenhouse with which to try algae growing research, and so we built the 2010 Attleboro greenhouse prototype. For algae, my goal this time was to bring the concentrated sunlight in at about waist level and then use 45 degree mirrors just behind the windows to further reflect the concentrated sunlight down into metal algae growing tanks. I designed the Attleboro building to be all algae tank on the north side, on the reflecting greenhouse's business side, with 16 east-west feet of sunlight-gathering window and with just a bit of building depth to house any auxiliary equipment behind the tanks. This particular prototype building had a triangular shape on the south side so that we could save on construction costs, but with full access to the algae tanks located just under the windows.

As you can see, when we reflected 10 times normal sunlight into the Attleboro prototype's window, the bright incoming reflections would spread out on the back wall.

The reader might observe that in the picture above, a bit of the incoming sunlight is hitting just above the target window and just below the target window. My optical calculations for a certain parabolic reflector were excellent, but we installed one of the heavy parabolic reflector frames upside down. Remember that a parabolic reflector, a circular reflector and an upside-down parabolic reflector all look somewhat the same.

My partner in Attleboro decided on his own to build this particular greenhouse two feet off the ground and on wooden piers. This would make sense for a normal cabin in a wet field, but now we couldn't use the ground to store heat and we would have to insulate the underside of the building. This illustrates an important principle that all inventors need to recognize: again and again, carpenters, partners and other craftspeople think that they know better than the mechanically clumsy inventor and so they usually build whatever they feel like building to get the job finished. They don't tend to ask the inventor for any opinions. I sometimes just have to roll with the punches.

In the Attleboro building the target windows start six feet off the ground on the outside, so that the eyes of people and animals standing on the outside would be underneath the 10x sunny reflection zone at the target windows. Always err on the side of solar safety.

Early on, my partner lost all of his interest in growing algae. My first two-dollar primitive algae tank design didn't impress him, perhaps I should have built a better tank in the first place, and then he never wanted to observe a proper tank in action. In a nutshell my goal is to multiply the algae tank's surface area for growing algae under concentrated sunlight by getting the sunlight deep into the water tank. I'll return to algae in a later chapter.

In any case we tried to grow greenhouse plants. We gave up on focusing a tall vertical rack of full-power 10x concentrated sunlight in favor of 5x reflected sunlight. We furnished the inside of the greenhouse with a layer of bricks on the floor to hold heat and with thin metal water barrels stacked two-high, on top of each other.

We had some success. My partner's house was at 68 degrees and the house was rather dark. The prototype greenhouse on a sunny Fegruary afternoon might get well into the 70s and was wonderfully sunny. So, my friend would walk through the snow out to the greenhouse and read his newspaper out there. It was great! This might indicate that everybody needs a greenhouse for better personal happiness. Why freeze in a cold house? The sunshine fights seasonal affective disorder and it keeps Vitamin D levels up.

We discovered new side issues. On one February afternoon I measured water barrel temperatures. The top half of the topmost water barrel was bathed in 5x sunlight, and that half of the barrel was recorded at 80 degrees F., respectably warm. Within the same water barrel, the bottom half was 50 degrees F. Apparently the hot water at the top of the barrel wasn't mixing with the cold water at the bottom of the barrel. The bottom barrel was 40 degrees F. The layer of bricks covering the floor was also cold. So, we had a heat storage issue. We weren't getting full use out of our heat storage medium.

The picture on the right looks like a great success. I'm holding a freshly rolled snowball from outside and the inside thermometer temperature says 78 degrees Fahrenheit. In fact, I would have preferred better heat storage with a somewhat lower inside afternoon temperature. We also had an electronic temperature recording device.

Our concentrated sunlight came blasting in at waist to chest level. I planted a good number of tomato and broccoli seeds in little pots. I discovered that when I put the seedlings into anything approaching direct 500% of normal sunlight the sprouts soon cooked and died, or else their little root balls dried out and then the plants died. White plant pots fared slightly better than black plant pots. Seedlings on the floor of the greenhouse got no direct sun and a small amount of indirect sun, so they didn't grow quickly but they didn't die quickly. We had a sunlight strength issue.

The Attleboro parabolic reflector frames worked well. We cut shims to specific lengths and nailed them onto the sides of 2” x 6” boards to create our parabolic curves, then we nailed straight 2” x 3” horizontal boards on top of the shims. We mounted the mirrors top and bottom on these horizontal boards. We put hinges on the centers of our reflector frames so that we could rotate them.

I left one inch air gaps between adjacent 12” x 12” glass mirrors on the reflector frames. These air gaps did a good job at spilling wind gusts when Hurricane Irene hit the area. The gaps may have given away 10% of our collection ability per square foot of collector frame, but it was a good trade-off. Perhaps leaking a bit of sunlight through the frames will allow some grass to grow behind the reflector frames.

These reflector panels were photographed from inside our Attleboro, Massachusetts prototype in 2010. Notice how images of the same clouds appear on successive vertical rows of reflector mirrors.

After three years of exposure to wind gusts the individual glass mirrors began to show problems. These frames supported the glass panels on their tops and on their bottoms only. Years of strong wind gusts could worry the slightly flexible glass pieces back and forth until a crack appeared. The wind constantly pushing back and forth on each glass mirror piece tended to cause a crack to spread through the glass piece's middle until the glass mirror snapped into two pieces. We lost about four 12” x 12” mirrors in three years. Also, our vinyl plastic mirror holder strips couldn't hold the mirror pieces on securely – a couple of mirrors weren't perfectly secured and they popped off.

V4. The Greene School Prototype, West Greenwich, RI, 2018

In 2012 I shifted to the relatively unrelated problem of properly storing solar heat for heating commercial buildings in winter. I didn't get back to building greenhouse prototypes until 2018.

This time I calculated to locate my reflector wall as close to the greenhouse as possible while still catching sunlight coming over the greenhouse roof at a 10 degree elevation above the horizion. My goal was to make the distance between the top of the windows and the top of the roof as small as possible, so that the sunlight could clear the roof, hit the bottom edge of the reflector panels and reflect into the greenhouse. The closer I could move the reflector panels, the better the aim.

I wanted the sun to be well above eye height, 6 feet, both inside and outside of the greenhouse. After the sun entered by the target windows, If the reflector wall was 8 feet away from the target windows, the target windows would be 8 feet away from the plants for a perfect deconcentration back to the 100% sunshine level on plant leaves.

I wanted the sun to spread back out on the plants at ergonomically comfortable levels for gardeners. To focus the reflected sunlight between 1 and 6 feet above the greenhouse floor when the target window would be 6 to 8 feet above the ground, the reflector wall would need to be located 8 to 13 feet above the ground. A tall reflector wall looks unusual, but this design delivers sunshine to plants at knee level up to shoulder level. We could build supermarket-style plant shelves on the far wall, on the south wall, for easy plant grower ergonomics. The greenhouse was designed to be ADA-wheelchair accessible. Also, the reflecting mirrors would all be above people's reach outside, and furthermore, the reflector panels would be reaching up and over the greenhouse for the best possible sun, relatively less obstructed by rather tall local trees.

I used a spreadsheet to calculate my mirror angles. The chart to the right is lifted from my spreadsheet. Looking from west to east, I calculated a number of (x,y) points for the reflecting mirrors and targets. Then I created a chart that drew a number of lines from one point to another. The two points on the far right of the chart are the top and bottom points of the optimum target window on the greenhouse, 12 inches apart, 90 inches and 102 inches up the wall of the greenhouse. The six points on the left are top and bottom points of an array of five 12” high rows of mirrors. Five of the long diagonal lines are tracings of sunlight rays from the bottom edges of five mirrors to the bottom edge of the target. The long diagonal line on top is a sunlight ray tracing to the top of the target from the top edge of the topmost mirror. The spreadsheet chart didn't include every line, but I could manually connect the dots to get a rough picture of the five reflector mirror positions and the target window. On top of this graph I put another reflecting mirror calculation that I didn't use.

It was useful to graph spreadsheet results to see if the results matched my idea of how the reflector should look. This chart showed that the spreadsheet's numeric results would at least form the right shape. On my earlier calculation try I could see that the end result looked completely wrong. I found my calculation error and fixed the spreadsheet/chart.

The Greene School reflecting panels were mounted at the center points of their backs onto the frame timbers with hinges. They can point up about 35 degrees above the horizon to send June 21 sunlight into the windows, or they can rotate down to point about 8 degrees above the horizon to send December 21 sunlight to the windows. Its best that the panels have no more range of motion than needed. The rather heavy panels will normally stay in their December position, but pulling on the chains will raise them to any needed intermediate position.

To reinforce all of the array's 12” x 12” glass mirrors I cut many 12” x 12” plywood backing panels and screwed them onto the framing timbers. When wind gusts blow from the north the plywood blocks the added air pressure of the wind gusts. When wind gusts blow from the south the glass mirrors are leaning against the plywood. As a result the somewhat malleable glass mirrors don't get pushed back and forth by any wind gusts.

On the Greene School reflectors, mirrors are held in place with screws above and below the mirror pieces and with 1” steel washers on the screws. This system worked well, although in four instances I over-tightened the screws and the extra pressure on the mirrors plus wind gust pressures eventually snapped four mirrors.

The greenhouse solar heat storage issue was solved by a number of separate improvements. I had been working with rock beds, active solar air systems and heat storage during 2012-2016.

Active solar air heat storage using a fan and a rockbed traditionally has had a show-stopping issue: rocks would occasionally get cold and damp and then the damp rocks grow toxic mold. People can't stand toxic mold, and so the rockbed would be thrown out. I know of someone who has a gravel driveway because their neighbor threw out his rockbed.

One solution is to never, ever let the rocks get cold or damp. Inside the Greene School greenhouse we installed an attic gable vent fan that comes packaged with its own photovoltaic panel. Whenever the sun shines on the PV panel, summer or winter, the fan turns and a stream of greenhouse air blows through the rockbed's rocks. In the heat of the day this system transfers that heat into the rocks. At night, the rocks are much too warm for condensation and the fan isn't running.

The fan's noise inside the air chase has been barely audible. I removed the vent fan system's electric thermostat with a screwdriver so that the fan will always push air into the rocks as long as sunlight is shining on the PV panel, and when the sun shines in winter the greenhouse air is always relatively warm. Normally the attic gable vent fan won't cool an attic whenever the temperature is below 80 degrees F.

Here's a 3-d sketch of the building's fan and air chase, along with the 2”x 4” framing timbers in the north and east walls. The fan is tilted at a 45 degree angle. It pulls air down an air chase from near the building's ceiling. It blows that warm air somewhat down a horizontal ground-level air chase. Here, I used 3-d graphing software to get an accurate picture of every board length and to make sure that no board in the greenhouse would interfere with some other board. Measure twice, cut once.

Here's what visitors can see of the fan housing and the air chase. We made the top of the low air chase look like a bench running along the north wall, below the windows, by building a back rest above the top of the chase. The Greene School sometimes needs extra seating for classes.

V5. Ground-level heat storage details

Sand can be useful as insulation because it doesn't compress underneath heavy loads. We put s five ton truckload of 1 1/2” rocks (rocks from a nearby gravel pit, small enough to pass through a 2” screen) under the center of the greenhouse's floorboards and preferably more than two feet from any wall of the greenhouse. Below the rockbed is 1” of sand. Heat doesn't tend to travel downward through the small air pockets in the sand, and then it takes forever for heat to migrate several feet laterally through the dry ground beneath the greenhouse toward the outside in winter. Sand has been piled high near the dry-laid concrete blocks under the greenhouse's four walls, which act as the greenhouse's building foundation. The sand insulates the concrete blocks which are exposed to the outside air. Near the blocks, this sand is covered by housewrap to always keep most of the sand dry from water spills inside the greenhouse and to reduce air infiltration in cracks between the concrete blocks. The housewrap above the sand is tilted to conduct any running water away from the dry sand at the edge. We piled a bit more sand on top of the housewrap to hide the housewrap. We also tossed sand into the centers of the concrete blocks and into any cracks between the blocks.

The ground level horizontal air chase has a sand bottom, with extra sand piled against the outer concrete blocks for insulation. This horizontal air chase runs almost the entire length of the greenhouse.

An interior line of concrete blocks runs the length of the greenhouse, 20 feet from east to west. It serves as a foundation for the inner side of the horizontal air chase. These blocks are laid on their sides. Warm air flows from the horizontal air chase through the hollow centers of the concrete blocks to the rocks under the floorboards. The inner side of the air chase, part of the bench above, is supported by this line of concrete blocks.

Plywood flooring backed with 2” x 3” boards for stiffness sits on the rocks. The flooring forces air from the air chase to travel all the way through the rock bed, from the north window side of the greenhouse to the south plant rack side. The rocks sit on a two inch layer of sand, and the sand sits on the undisturbed natural topsoil complete with a few dried-out natural grass blades. My sense is that the warm rocks slowly transfer some heat downward into the sand and into the rather dry topsoil, so that we have several extra tons of functional weekly to monthly heat storage in the topsoil. The 1 1/2” rocks have, in aggregate, an enormous surface area, so that they absorb heat quickly from the air flowing through the rockbed. We need quick heat absorption because when solar heat comes to the greenhouse in November it arrives rather quickly and it may not last that long.

Because the rocks are below the floorboards and below the plant racks, at night when the greenhouse needs heat, hot air naturally rises into the greenhouse from the rocks. The plants are on the warm side of the greenhouse away from the windows so that they will stay warmer and frost-free. At night, coldness from the north window flows downward to the floor and stays low. The rocks directly under the plant shelves can release heat directly upward at night.

Rockbeds sometimes have an air circulation issue – a fan wouldn't be strong enough to push air through the rockbed. The correct solution is to spread out the airstream and push air through the spread-out rockbed at an imperceptably slow speed. The Greene School prototype has a rockbed 6 inches thick by 5 feet wide by 18 feet long. Air is pulled downward from the top of the greenhouse where the hottest daytime air will be found, run through the fan, pushed through an 18 foot horizontal ground-level air chase, then small parts of the airstream are drifted quite slowly through 5 feet of rocks, giving the air time to transfer heat to the rocks. Air friction is related to air velocity to the fourth power, so slowing the air down within the rock bed can drop total air friction toward insignificance.

V6. Weeping Windows

Single pane greenhouse windows will weep water in winter. When humid greenhouse air comes in contact with single pane greenhouse glass on a cold winter's night, water condenses on the glass. We don't have to care how much heat we lose by this weeping process because we have so much solar heat to store, and because 90% of the wooden greenhouse is reasonably well-insulated. We have gotten away with inexpensive R-10 insulation in the walls and roof, and we have been maintaining about 25 degrees F. higher than the average weekly outside temperature. We could increase our insulation to achieve much better performance except we're already happy right now.

The Greene School greenhouse has special aluminum gutters on the insides of its windows. With warm, humid air inside and snow outside these windows don't weep, they bawl their eyes out. The water runs down aluminum drains into quart buckets. Students regularly pour the distilled water onto the greenhouse's seedlings. This weeping window design means, first, that the plants never suffer from leaf mold, and second, that the building's internal wooden structure probably will never rot away.

The interior window gutters leave about a 1-inch air gap at the bottoms of the windows. This allows the windows to cook out and air out any residual moisture below the windows. Airflow below the gutters is limited by the gap's size, so that moisture condensation on the windows at night is also limited.

V7. White window trim and heat issues

I didn't use vinyl window frames. Vinyl might melt under 500% of normal sunlight. We created white-painted wood frames.

The frames were designed to have only 2”x6” wood framing timbers between adjacent windows. We wanted to capture as high a percentage of incoming sunlight as possible, but then the wood framing timbers have to be strong enough to support the roof.

Anything that's black and close to a window is going to heat up, and heat can crack a window. In the Attleboro prototype we installed an edge of a storm window hard against white housewrap that had some black lettering. Where the black lettering touched the glass, the glass overheated and cracked.

For the Greene School greenhouse, a subcontractor installed ordinary window glass for us without asking me, as opposed to installing heat-tempered safety glass. So far the regular window glass has worked, except apparently some students tried melting a snowball that was pressed against a target window. The enhanced temperature difference made a snowball-shaped round crack in that particular window. Oops!

It's useful to wash 5x solar target windows regularly. Dirt can eliminate 30% of incoming solar heat, and the plants might miss that heat on a cold winter's night.

We have already had an accidental water spill in the greenhouse. I filled a 50 gallon water barrel with a hose and we spilled 5 or 10 gallons after the barrel filled. The water ran down a gravel drain pre-dug under the rockbed and sand at the topsoil level. The water drained down a slight slope to the outside of the greenhouse. I was worried during construction about adding french drains to properly drain the soil around the greenhouse to keep ground water levels low, but we haven't in practice had a drainage problem.

We never disturbed the original topsoil in building this greenhouse. There aren't any pilings because ground frost has become impossible near this heat-storing greenhouse. The greenhouse legally qualifies as a “temporary” greenhouse because the building has no fixed foundation, it sits on the dried-out lawn and it has no wiring for grid electricity. I use temporary in quotes because the greenhouse is likely to be a pretty permanent and sustainable fixture.

At some point I want to remove a few of the rocks below the floorboards and plant a tomato vine into the natural topsoil. The vine might grow to become 25 feet long after 24 months of growth. We have yet to try that experiment.

We added reflective film to the inside south wall. Sun-loving plants can get roughly 100% of weak winter sun on their north-facing leaves and maybe an additional 50% of doubly-reflected sun on their south-facing leaves, for better photosynthesis. Our sun-loving plants don't look spindly; as the above picture shows, the plants look properly stocky and happy despite winter sunshine. This particular greenhouse view looks downward from a plant rack to the bare rockbed. The white plywood flooring starts on the right side of the picture. Most of the wood inside is painted white to slightly increase indirect light reaching the plant leaves.

The picture on the right illustrates a neighboring grower's main problem with plastic sheeting greenhouses. If they don't throw away their plastic poly sheeting every few years, bad things tend to happen. A blizzard threw a large tree branch right through a sheet of plastic, causing a freezeout of the entire crop. A crop freezeout can cause a regular customer to not trust them the next time that they promise to deliver a crop in midwinter. Growers insurance doesn't reimburse growers for the expensive loss of regular customers.

The Greene School greenhouse is supported by 2”x6” roof rafters under roofing plywood, so that it won't collapse and freeze out the crops even with 6” of ice and snow buildup on the roof. Branches aren't going to be driven through plywood walls or through the roof. Branches flying on a powerful north wind are going to first be blocked by the greenhouse's sturdy reflector frame that protects the windows and the mirrors. The Greene School greenhouse recently came through Hurricane Henri unscathed. The reader may notice that Rhode Island gets enough hurricanes.

On July afternoons it can make sense to open a greenhouse's vents. We never installed vents on the Greene School greenhouse because the school is closed during the summer. Otherwise, I had designed a system whereby the greenhouse's fan would exhaust hot air from the greenhouse in July.

On an extremely hot July afternoon, this type of greenhouse's reflector panels should be pointed upwards to reflect sunlight back into space. Selectively turning off all of the greenhouse's heating/lighting in July plus thermal mass means that greenhouse crops that don't prefer heat can be saved from July's heat.

We specified an extra-wide greenhouse door to accommodate wheelchairs and garden carts. Supermarket-style multiple plant shelves work well for any wheelchair-bound gardeners who might have a limited reach.

V8. Summary

Self-heating midwinter greenhouses can be easy for carpenters to build. They can be inexpensive. They can be zero-heating, zero-lighting and off-grid. They inherently control their own humidity. Because of their great sunlight control, I expect them to grow more crop per square foot of space than most mid-winter greenhouses. They are puncture-resistant and collapse-resistant. The growers aren't breathing volatile plasticizer chemicals from freshly replaced plastic sheeting. Greenhouse crops can be marketed as sustainable. There's no fundamental reason why a person couldn't move into one of these greenhouses.

V9. Future Greenhouses

I want to build a double-action greenhouse with a line of target windows on the top of the north wall and another line on the top of the south wall, with a different reflector panel design on the south side of the greenhouse. Twice as much reflected sunlight comes into the greenhouse. Both the north and south walls have plant racks. Twice as much crop can be grown with a cart path down the middle of the greenhouse and a jungle of green growing up both walls. Glazing over the entire south wall of the greenhouse might possibly work except on a New England winter night a glass south wall would lose far too much heat. The double-action greenhouse will get better performance for less cost.

A commercial greenhouse for growing organic crops will want an adjoining clean room where growers can check themselves for insects that might hitchhike their way inside.

A one-acre sustainable greenhouse needs to bring down 100% of the available sunlight falling on its roof, and a bit more. The greenhouse roof would have alternating rows of east-west rooftop reflectors and sawtooth-like glass target windows to bring the sunlight down. It should outclass a greenhouse made with throwaway plastic/fiberglass roof panels. See elsewhere on this site: K8. Industrial-sized vegetable and flower greenhouses

There's nothing special about building this greenhouse's walls and roof out of lumber. Somebody will want to build future greenhouses out of straw bales. Two igloo-shaped ends for the greenhouse might work.

I want more accurate reflector frames. I can see some slight aiming inaccuracies in the Greene School greenhouse reflectors because some of the reflector's plywood sheets became a bit warped. I would prefer constructing long-lasting aluminum reflector frames that won't rot away or need restaining.

I want snap-in non-glass mirror modules for easy reflector wall assembly and for occasional mirror replacement. If I'm building my own reflector mirrors, I can reduce the distance between the reflector's wind-spilling holes to improve the reflector wall's performance in hurricane-force wind gusts, and I might shrink the sizes of the air holes.

I want a reflector frame with a greenhouse motor so that I can track the sun's curved path across the sky while making the target window narrower, reducing heat losses, and no one would ever have to worry about whether the reflectors were on target. In some cases I might want to install double pane target windows in response to Manitoba-like winter weather conditions, leaving one small single-pane weeping window for inherent dehumidification.

I want to take a straight-line power sander to the greenhouse's line of windows before installation. Straight-line or belt sanding in a vertical direction would horizontally diffuse the light coming down to the plant racks, which could be easier on the eyes and on the plant leaves than ordinary sunlight. A plant's back leaves might get more diffuse light from more directions so that the back leaves wouldn't be 100% shaded out by leaves in front.

I want a more standardized interior water gutter system for the greenhouse's weeping windows.

V10. Linear trough reflectors aren't perfect

Our reflector panels are mounted on hinges so that they can tilt upward to focus June sun into the windows and downward to get December sun. Chains are attached to each reflector panel section. Ninth grade Greene School students pull on the chains to adjust the angle of sunlight weekly between December and June, so that the concentrated sun is aimed on the windows. They point the solar focus, an easy-to-see sunny solar multiple reflection that shines on the windows on sunny days, at the windows.

In December the sun travels in a curved line across the sky. It rises on the southeast horizon, not in the east, at about 8:00 a.m., it reaches an elevation of 24.5 degrees by noon in Rhode Island and it sets on the southwest horizon at 4:00 p.m. The curve means that a fixed linear trough can't track the sun perfectly from sunup to sundown. If the sun is high in the sky, the reflection is low on the windows. If the sun is low on the horizon, the reflection is high. This pronounced curvature in the path of the sun somewhat disappears by February.

We solve some of this focusing problem by positioning the reflector panels as close to the target windows as possible. We want as few inches as possible between the top of the line of windows and the top of the roof, so that sunlight coming over the roof at perhaps 10 degrees above the horizon will just barely touch the bottoms of the reflector panels.

Our line of target windows is 28 inches high and each of our lines of mirrors is only 12 inches wide. This gives us 14 inches of extra play on the windows. Our focusing rule of thumb is to point the 12 inch high solar focus at the bottom of the line of target windows at noon but at the top of the line of target windows around 9:30 a.m. and 2:30 p.m.

After 3:00 p.m. the reflected solar focus unfortunately starts drifting above the top of the target window, and by 4:00 the focus has drifted completely above the windows. However, the sun is close to the horizon at 4:00 and the school it gets somewhat shaded by nearby trees around 3:00, also by low clouds, so that sunlight on the horizon isn't that valuable to us at 4:00 p.m.

I somewhat arbitrarily assigned the perfect angle of my incoming sunlight at solar noon on February 21, one quarter of the way between the lowest solar elevation on December 21 and the highest solar elevation on June 21. The school has its summer vacation in June-August and plants can grow outdoors all summer, so December 21 sunlight will be relatively more important to the school. Around June 21, the top of the tilted back reflector is a bit too far away from the target windows and the bottom is a bit too close. Not all of the reflected sunlight hits the target window. The reflection from the highest row of mirrors hits a bit too low on the greenhouse windows and the lowest row also hits too low. We're willing to put up with the June imperfections because we don't need any extra solar heat in July.

A linear trough shape isn't equipped for spring/summer growing in another way. It completely stops reflecting light into the greenhouse after roughly 4:30 pm and before 7:30 a.m. When the incoming sun is almost parallel to the line of mirrors, so that the sun doesn't shine at all on the mirrors after 6:00 p.m.

In the future I may want to use a reflector with two separate aiming motors, called a heliostat, to bring concentrated sunlight into a greenhouse. With a heliostat we might be able to apply as much as 14 hours a day of full sunlight onto a sun-loving crop in the months of April and May. We went further into heliostats in a previous chapter.