QBARS - v33n2 A New Greenhouse, Cool but Comfortable

A New Greenhouse, Cool But Comfortable
Frank Doleshy, Seattle, Washington

No rhododendron grower is ever quite satisfied with the plants that do well in his garden. People living near the Great Lakes would like to grow the standard British hybrids, Long Island people would like to grow the plants they see thriving in Seattle, and, everywhere, more and more people would like to grow the Malesian rhododendrons from New Guinea, Borneo, and other tropical places. The result is a great demand for plant shelters, particularly greenhouses.

Mrs. Doleshy and I bought a 6- by 10-foot greenhouse about twenty years ago. And, although we originally thought of this as a place to propagate rhododendrons from seeds and cuttings, it was soon providing winter shelter for tender rhododendrons. The constant warmth and moisture needed for propagation, however, did not suit the Malesian group. While cuttings from these were easy to root, the older plants died back and seldom produced flowers. They appeared to need a different environment, and the only practical way to provide this was to build a second greenhouse. Then, with the original greenhouse available for propagation, the new one could be a small conservatory, meeting the needs of the Malesians and serving as a place where we and our friends could enjoy their spectacular flowers, often produced during the winter.

The greenhouses from several manufacturers seemed excellent but did not include the features we wanted. Therefore we designed our own, with main components as follows:

A wood frame supports the large panels of tempered safety glass used as the covering.

Vents are at the roof ridge, controlled by automatic lifters, and the sliding doors at each end serve as air intakes.

Benches are entirely absent, the plants living in the ground or in ground-level containers, under ceiling-level lamps strong enough to give them the amount of light found in open shade.

The heart of the heating system is slow return of sun-warmed and lamp-warmed air from the ceiling to the floor. To supplement this when necessary, a thermostat turns on buried, lead-sheathed heating cables.

This article is an explanation of design decisions and their effects; it does not deal with construction techniques - which can be learned from numerous do-it-yourself publications. Any reader who likes the design is welcome to build a similar greenhouse. First, though, he should decide whether this design will meet his needs, fit his environment, and come within his expense limits - or whether he should change some features. As a partial basis for estimating expenses, I am including the prices of some materials and supplies, as quoted in Seattle on November 25, 1978. But construction cost depends on local conditions and the amount of work the owner does, so I am not including any figures for this.


The first thing to select is the covering material: fiberglass, some other plastic, or glass itself. This choice affects heating expense by increasing or decreasing the amount of "greenhouse effect", which is the conversion of sunlight to trapped heat in a structure with a more or less transparent covering. The conversion occurs when sunlight, after passing through the covering, strikes and heats an opaque object. The trapping occurs when radiant heat from the object fails to escape through the covering and, instead, is reflected or reradiated back to the interior, causing a temperature rise. Physicists explain this as the selective effect of a covering material which is transparent to the short wavelengths of visible light but opaque to the longer wavelengths of infrared (heat) radiation. Although this heat is to be had without expenditure of fuel or electricity, some authors of greenhouse books do not discuss it, and others provide only a brief, vague description. We have found no fully satisfactory comparison of glass and plastics as heat-trapping materials, but glass seems generally accepted as the most efficient.

Along with this heating efficiency, we considered other advantages and disadvantages of the various covering materials, including their availability in large panels. While plastic products can be purchased in almost any size, the greenhouse glass used until recent years was most often in pieces two feet square or smaller - installed with overlaps and joints which leaked air or filled with green slime. We had noticed, though, that newer designs included larger pieces, and this seemed desirable unless too expensive. Actually, we found, one dealer offered imperfect pieces of tempered safety glass, un-tinted, 76 inches long by 28 or 34 inches wide, at $12 each, while another charged only $7 when he had any in stock. These prices (which have remained unchanged) were about equal to those for some grades of fiber glass, twice as much as those for others, and, of course, higher than those for thin plastic sheeting. But glass has advantages other than its presumably greater efficiency in solar heating; these include its rigidity and resistance to weathering and abrasion as well as its attractive, appropriate appearance in a greenhouse designed as a small conservatory. Therefore we decided to use glass and ordered 20 of the 76- by 28- inch panels for the roof and sides; also, ordinary 12-foot sliding door assemblies for the ends ($225-305 each), and inexpensive, hinged panels for the roof vents. We had to use this glass in the manufactured sizes only, since it cannot be cut. (If this is tried, the whole piece breaks into granules like those in a broken windshield.) Although a design constraint, this was a minor one, easily overcome by changing the roof pitch.

Framing, we decided, should be wood instead of metal. This saves some heat, reduces shock hazards, and permits easy attachment of wires and ducts. Either western red cedar or redwood is suitable; we elected to use cedar with copper naphthenate treatment, since experience had shown us that this will resist decay for twenty years or more. (Caution: Wood treated, instead, with creosote or pentachlorophenol is deadly to nearby plants and can kill everything in a greenhouse.) The rafters and studs, or "mullions", are made from 1 X 4 boards, treated on all surfaces and nailed together as three-board sandwiches with enough offset to provide shoulders for glazing. Including some larger-dimension pieces, lumber expense for our 150square-foot structure would now be $240.

The foundation is reinforced concrete, extending 16 inches below grade, and the floor is a conventional, porous mixture of ⅔ gravel, ⅓ soil, about 10 inches thick and merging gradually into the unmodified soil beneath.


In the Seattle area, Mrs. Pendleton Miller has pioneered the low-temperature cultivation of Malesian rhododendrons. They thrive in her cool house, where she permits temperatures to go as low as 30° - 31° F, instead of the conventional 40° - 45° minimum. On the other hand, our own outdoor plantings have not survived a winter; they seem to stay alive only as long as temperatures rise to 45° or higher every day or two.

Both of these results are consistent with observations on Mt. Kinabalu, Borneo, when Mrs. Doleshy and I were there during September-October, 1976. Our main headquarters was a but at 10,800 feet, surrounded by rhododendrons, and we found the tropical-looking R. lowii growing to 11,300 feet, other kinds to 13,400 feet. Outside the hut, 7 a.m. temperatures were 41° - 45°, and the minimums at other times of the year are doubtless 10° lower; possibly even 20° lower. Our afternoon highs were 48° - 55°, and we suspect that the highs fluctuate more than the lows.

These pieces of information led us to design our new greenhouse for a 30° minimum, and most Malesian' rhododendrons have responded well, R. stenophyllum apparently being an exception which requires more heat. (For the present, it lives in the old greenhouse.) Also, although these temperatures seem adequate for such companion plants as a Begonia from Mt. Kinabalu, Vaccinium species from Mt. Kinabalu and from New Guinea, and a widely-cultivated Tibouchina , they are too low for the ordinary, garden forms of Impatiens that we have tried.

As mentioned, we expected some warming from the sun and some from the 1720 watts of lighting, turned on morning and evening for a total of 7½  to 12 hours per day. Also, we expected a contribution from the ground during winter. Yet, during cold weather, these sources might only be sufficient to produce a pocket of above - 30° air under the ceiling (mixed with water vapor, which also rises). Therefore we decided to cycle the ceiling air down to the floor. For this, we had 4-inch ducts installed in the northwest and southeast corners, each one supplied by a fan at the ceiling and ending in a Y which splits the stream of air and sends it along the lower walls.

The only difficulty was to find fans small enough to do this at a low speed, necessary because rapid circulation of air past the cold walls would waste heat. After several inquiries, we found a dealer who recommended the Pamotor 8500C, an instrument-cooling fan from West Germany which moves about 40 cubic feet per minute when attached to a short duct. At this rate, a pair of fans requires about two minutes to displace a one-foot-thick layer of air in our 150-square-foot greenhouse - which seems nearly ideal. These fans cost $24.40 each, and the power consumption is 13 watts (about one-tenth the requirement of a reading lamp). If operated continuously at 77° F, the expected life is 11.4 years; at 131° it is 2.3 years.

We are not the only greenhouse owners with this kind of homemade system, as we learned while attending the International Rhododendron Conference at the New York Botanical Garden. One evening before dinner, the cocktail party was in the Garden's propagating greenhouses, and Mrs. Doleshy noticed some hanging lengths of light-weight duct material, each with a slow current of slightly warm air issuing from the lower end. She explained what she thought these were, but a very famous rhododendron authority assured her that the Garden did not engage in such simple-minded inventing. Soon though, a member of the professional staff came by, and, when asked about the ducts said, "You see we knew our warm air and moisture were up at the ceiling..."

When the warmth from sun, lamps, and soil is insufficient our next source is the underground heating cables. Living in a wooded area close to salt water, we get by with 270 feet of cable supplying 1800 watts of heat, and worth about $240. One mile inland or in any colder region) a similar 150square-foot greenhouse would probably require two to four times as much cable. Good control of this heat is essential as we had learned from experience with runaway thermostats and we recommend two safe-guards: first, use thoroughly dependable thermostats such as the White Rodgers 23022, $50.90); second install a pilot light to show when each thermostat is on.

For still more heat, we turn on the main lighting during late-night hours when it is normally off - a step taken about ten nights per winter. Beyond this we'd have to use convection heaters or else apply temporary insulation. Also, we could try to store surplus heat by means of a major structural addition as outlined in the concluding paragraphs.


Although sunlight contains the whole visible spectrum of colors, plants are selective, using chiefly the blue and red portions for manufacture of food. This leads to a dilemma in selecting greenhouse lamps. For efficiency, the owner will probably want fluorescents. The question is, though, should he buy special plant lamps which produce the often-seen combination of blue and red? Or should he buy high-intensity white lamps with the thought in mind that their output consists partly of blue and red light? A dedicated investigator could become lost analyzing prices, power usage, effective hours of lamp life, total light output, and percentage of blue and red in the total output. Based partly on manufacturers' specifications and recommendations, partly on local lore, our decision was to use lamps of two colors, half-and-half: cool white and the pinkish-cream Sylvania Gro-Lux Wide-Spectrum.

Perhaps more important, we decided to use the 8-foot-long 215-watt "VHO"" (very high output) size, with the optional feature of internal reflectors to direct most of the light in one direction. These lamps are only moderately more expensive than conventional kinds (about $15 each, the cool whites being slightly higher than the Wide-Spectrums), and they do away with any need to place the lighting fixtures within a few inches of the plants. From eight such lamps, mounted 7 feet above the ground, our ground level plants receive an approximation of bright-day open-shade illumination. We are free, therefore, to do without any benches and grow our plants on the ground as in a garden or meadow - somewhat crowded but with a few rocks showing. We enjoy this and also enjoy the brightness that makes an evening visit seem like a step into daylight. (This is not quite artificial daylight, though. Because of deficiencies in parts of the spectrum, some colors are not right to the eye or to color film.)

Hours of operation are a compromise, meeting the needs for security, neighborhood acceptance, and heat and light. We avoid illumination at school bus times, home-from-the-party times, and also the late-night hours when a great blaze of light in the woods might disturb our neighbors. These limitations are not troublesome, since we use the lighting to extend and intensify natural daylight. The timer turns it on at 8:30 a.m., off at a time depending on the season, between 9:30 and 11:45 a.m., on again between 3:00 and 5:00 p.m., and off for the night at 11: 30. As mentioned, we turn it on at other times if necessary for supplemental heat, but this is infrequent and doesn't seem to disturb the plants or the neighbors.


The fans run all the time, regardless of temperature, and stir the air as they cycle the heat and moisture. Excessively warm air escapes through the vents in the roof ridge; automatic lifters open and close these as the temperature rises or falls through the 68°-75° range. (These lifters are non-electric devices, dependable and well worth the price of $ 25 or $ 50 each.) The sliding doors at each end provide the rest of the ventilation, and the openings can be as much as four feet wide.


As completion approached we wondered about several things, such as the adequacy of the ventilation and the quality of light produced by VHO lamps. These doubts disappeared with the first test, which was to let the weeds grow. We have many kinds, and they thrived. Then, when we moved the first rhododendrons inside, these changed rapidly. Leaves were greener within two weeks, and the bare stems of some plants soon produced buds. Now, more than a year later, plants formerly in the perpetual invalid group have recovered, some producing flowers for the first time.

Adaptability, Possible Changes And Designing Your Own

Although I've continually mentioned Malesian rhododendrons, this greenhouse seems equally healthful for other tender or borderline species. One such inhabitant is a San Francisco favorite, R. 'Forsterianum'. Nearly killed by a sudden chilling, our old plant produced no new foliage during an entire year. Yet, after being cut back to a bare trunk and given a few months' rest in the new greenhouse, it resumed growth and is now very much alive.

I'll be glad to supply a copy of our plans for the $3 cost of reproduction and mailing but must warn that I drew these myself, although not a licensed engineer or architect. As earlier mentioned, I suggest that the user change the design in any way that suits his own needs and circumstances. Among the many possible alterations would be the addition of a heat storage device, such as containers of water or a bin of rocks. By circulating air through this device, excess heat could be drawn off during the day for release at night. But efficiency would be low in our greenhouse, since nearby trees and salt water reduce the amount of day-tonight temperature fluctuation. Also, Malesian rhododendrons seem to thrive on the present amount of fluctuation and may actually require this. Therefore, we decided against the heat storage.

To people in other areas, this feature might be a money-saver, and further study could result in improved designs. Likewise, a person studying any other feature of his greenhouse design is likely to find opportunities for innovation, leading to monetary saving or additional high-quality space for the amount spent. I urge the reader to accept this challenge.