A Solar House
A house does not have to be architecturally weird or covered in glass to be solar-friendly; if it is, it probably isn't. This article describes our own low energy house of outwardly conventional design, in which has been incorporated unobtrusive active solar elements resulting in year round comfort and very low energy demand. The house and solar heating system has been owner-designed and built using low technology materials, producing an effective end result at low cost.
Mumballup is a rural location some 50 km inland of the Western Australian coastal city of Bunbury. At latitude 33.5° South, it enjoys cool to mild winters and warm to hot summers; 40°C is fairly common in the summer as are frosts in the winter. It is far enough from the coast to have diurnal temperature ranges of up to 20°C; at least part of the day is likely to be a comfortable temperature at any time of the year.
The house site is in a 20 ha paddock which falls only gently to the south and provides good solar access. When it was built, the pine tree and a big red gum in the photo afforded some protection from westerly sun in the summer and cold winds in the winter. Unfortunately, the red gum was blown over in a storm and the pine tree succumbed to age and termites. A rural power supply line traverses the block, but there is no mains water or gas. The block is skirted by the Bibbulmun walking track, which runs from Perth to Albany.
We wanted a house large enough to accommodate our (then) teenage daughters and occasional visitors, which would be comfortable at all times and have low energy demand. The climate seemed to suggest a well insulated envelope with large thermal inertia, which we translated into a slab on ground with framed outer construction and internal mass.
Materials were chosen on the basis of cost, durability and practicality; this precluded some which may have been more environmentally friendly. For instance, we did not consider rammed earth (very popular in this part of the world) because of its poor durability and poor thermal performance, we chose a steel frame because of early experience with voracious termites, and we used mainly glass fibre insulation because of its fire retardant properties.
We did most things by the book in terms of passive solar design. The house has a northerly aspect, with living areas on the north side, sleeping areas on the south side, and a garage protecting most of the west side. The steel roof is white for maximum reflectivity, and is insulated with R1.5 foil-backed insulation; additionally the ceilings are insulated with R2.5 batts for good measure. The steel wall frames were wrapped in building foil, carefully sealed at the joins and to the window frames, and the cavity packed with R1.5 batts. The outer brickwork is painted a light colour and the brick ties do not penetrate the foil. Window frames are wooden throughout to reduce heat loss and condensation, and fixed panes to living areas are double glazed. Selective use is made of film window insulation elsewhere. Slate floors are used to capture solar gains on the northern side of the house. An internal skylight has adjustable mirrored louvres for seasonal light and heat control, and pergolas are designed to maximise heat gain in winter and shade in summer. Did we forget anything?
The modelled energy demand of the house is about 0.5 kW per Celcius degree difference between the inside and outside temperature, indicating a space conditioning load of around 50 kW-hr for an extreme day. This and about 10 kW-hr of hot water demand constitute most of the energy requirement.
Rural living has a few drawbacks which can restrict conventional options for heating and cooling; add some environmental considerations and these options can virtually disappear. We don't have natural gas, and didn't want to place too much reliance on wood or coal because of concerns over pollution, especially of rainwater collected from the roof. The shortage of good water in summer rules out evaporative cooling; in any case we considered that we would derive more comfort from our insulation. The opening of windows on summer evenings, to admit swarms of midges with a collective contempt for fly screens, even restricts conventional options for ventilation.
We chose, out of necessity more than altruism, to add some more active solar elements to the concept. Our budget and remoteness from Perth determined that the system would be characterised by recycled materials and improvisation, rather than high cost and high technology.
Large diurnal temperature ranges suggested the use of solar heat in winter, and of nocturnal cooling in summer, to provide a comfortable living environment. Consequently the design incorporates solar thermal collectors, with heat banks which are warmed by solar heated water or cooled by night air, and used to match the conditioning supply to the demand. This precludes winter overheating or summer overcooling of living space, and permits conditioning capability to be accumulated if not immediately required. The system uses water for heat transfer and storage, and incorporates capacity for domestic hot water needs. Extraneous heat generated in warmer months is used to heat a swimming pool. The whole system is operated by simple thermostatic switches in a 12V control circuit.
Solar Thermal Collectors
Twenty square metres of selectively surfaced all copper collectors are mounted under a section of polycarbonate roofing, which is angled at 53° to the horizontal so as to maximise heat gain from autumn through to spring. On a clear day (when nights are coldest) these generate about 60 kW-hr; enough to provide space heating during winter, and useable pool heating at other times. Putting the collector under the roof improves appearance, saves the cost of a separate enclosure, and provides protection from frost through heat infiltration from the ceiling (in this circuit, the collectors drain at night, but see Conclusions below).
Hot Water Preheater
Water from the collectors passes at low pressure to a 900 litre tank. This serves as a preheater for the house's conventional electric hot water service, which is fed via a copper coil immersed in the tank. Water flowing through the pre-heat coil thus has "first call" on heat flowing from the collectors. The tank was specially rotomoulded from polypropylene, to enable a high storage temperature without the cost or corrosion risk of other available vessels of this size.
A conventional wood fuelled heater is also connected to the tank, as backup for the solar collectors.
Space & Pool Heating
Water from the bottom of the tank is directed by a 40W pump via a three way valve to either of two heat exchangers, according to season.
In winter the hot water is directed to a heat exchanger consisting only of two recycled automotive radiators, from which warmed air is circulated around the heat banks by a fan controlled by a thermostat on the inlet pipe. At other times the hot water is directed to a concentric tube heat exchanger, through which pool water is circulated by another small pump, again controlled by a thermostat on the hot water inlet.
Water in the solar circuit is cooled and returns to the collectors.
Backup for space heating and cooling is provided by a 4½ star rated R/C air conditioner, and by the aforesaid wood heater (trees planted 15 years ago provide renewable fuel for this).
The house, although split level, is built on a level slab to provide 1.2 m of clearance below the minor bedrooms, creating two 18 m³ spaces which are lined with styrene foam and used as heat banks. Water is used as the storage medium because of its high heat capacity and low mass; the only significant cost is in its containment. To maximise heat transfer rates a high surface area is required, and thus a lot of small containers are better than a few big ones of the same overall volume. The container which seems ideal is the humble 2 litre "rocket bottomed" PET soft drink bottle, which is strong, practically indestructible, and far cheaper than anything else we could find (about two cents per litre, recycled). Its high surface to volume ratio is better than other bottles because, when it stands upright, even the bottom is accessible to the air.
Our family spent most of a week filling over 7250 of these bottles, much to the amusement of neighbours. The bottles are stacked upright, about 1 cm apart, in three layers, each separated by a layer of fibrous cement sheeting. The end result is 14½ tonnes of storage which could meet the most extreme daily heating or cooling load of the house with a 3°C change in temperature.
The house also features two internal double brick walls providing more passive thermal mass.
A 400W fan controls air circulation through the house's ducting system. There are two vanes, one of which controls the intake of outside or inside air, and another which directs the flow through the heat banks or directly into the house. The air circuit is designed so that ventilation of the house can proceed independently or in tandem with circulation around the space heater and heat banks. The ductwork includes an airing cupboard which can recover heat from the flue of the wood fuelled heater.
The house was owner built and construction took three years. During this time, the temperature in the sealed heat banks varied from 14.5°C in winter to 22.5°C in summer, somewhat less than the range between average temperatures for these times of the year. This was indicative of the considerable inertia of the system, and raised expectations of its performance.
The house was occupied before completion of the heat bank ventilation system or external window shading structures. During the first year it was found that diurnal internal temperatures varied by about one third of the external temperature range, with the same average in summer and about 8°C warmer inside in winter. This translated to internal temperatures from a minimum of 15°C in winter to a maximum of 28°C in summer, in ambient temperatures from 0°C to 40°C.
Measured performance of the solar collectors at slightly over 10kW, is a little dissappointing although the pump has been operated at low speed (5L/min) and they may have been overheating.
Contribution of the system to summer pool heating appears to be about 60 kW-hr per day, or slightly over one degree. This is a bit less than the overnight cooling (although this has been corrected with a pool blanket).
Subsequent commissioning of the heat banks and ventilation system was not without its tribulations. We managed to raise the temperature of the heat banks by about 1.5°C on a sunny winter's day (about 25 kW-hr) and to lower it by about 2.5°C on a cool summer's night (about 42 kW-hr). The implied rate of heat exchange is quite good at around 6 kW.
Use of the heat banks enabled internal temperatures to be maintained within the range of 20°C to 25°C without supplementary heating or cooling.
Performance of the hot water preheater is estimated from temperature measurements at the hot water service inlet during hot water use. In winter, when cold water is around 10°C, the preheated water is usually between 25°C and 35°C, depending on the temperature in the preheater tank and the rate of drawdown. A slightly greater difference in summer means that the hot water service could probably be turned off in hot weather, although under the shower is not a particularly nice place to be reminded that it needs to be turned back on again (at least it doesn't run cold). A year round margin of 20°C equates to a power saving of around 1700 kW-hr.
With the benefit of 20-20 hindsight, there are some things we could have done differently:
The hot water preheating coil creates a bit of extra friction in the pipe so there is a trade-off between heat transfer efficiency and hot water pressure. We may yet substitute a much smaller copper preheat tank in advance of the hot water service, and run the entire system directly off the cold water supply (see Figure 7). This has risks of its own; it would not allow the collectors to drain overnight, and it would necessitate changing the vehicle radiator heat exchangers for something that would sustain the higher pressure (they have already had a couple of minor leaks). This would, however, enable us to retire the polypropylene preheat tank, which has developed a crack (high up, fortunately), and may improve heat transfer to the heat exchangers.
The heat bank and air distribution system operates at very small temperature differences, and heat losses in the ceiling ductwork may be significant. We should probably have run the ductwork through the conditioned space, possibly in PVC pipes.
We may yet get around to glazing the roof collectors properly. The polycarbonate sheeting is probably not all that efficient, and is difficult to seal.
The salty pool water corroded the copper heat exchanger and it started to leak after a couple of years. This was one of those silly oversights, corrected by substitution of a (less efficient) heat exchanger made from PVC tube.
Just about everything that we built prototypes of worked as expected; things that we couldn't, didn't (a prototype collector consisting of a polypipe coil under a polycarbonate sheet worked beautifully, but disintegrated in the course of one summer).
The whole project has been a fascinating experience. We were disappointed to find a singular lack of help from some sectors of the solar energy industry, but probably shouldn't have been surprised given that for the most part we weren't buying systems off the shelf which fitted their design formulas. We should, however, record our thanks to the people at South West Recycling (PET bottles), Adro South West (rotomoulding), and Curtin University (selective collector surfacing) for their assistance, and to all those who said it couldn't be done or wouldn't work, for the incentive to persevere.
This house has been standing since 1996, and the more time that passes, the more fun we seem to have had building it. Time has also taken its toll of of some of the components written about above, mostly through corrosion (probably no worse than it would have been if we had bought things instead of making them ourselves). A major part of the active system can no longer be relied on, but the built-in passive elements of the house's design work so well that the progressive "dropping off" of the active elements has barely been noticed. The inside of the house has seldom been below 15°C or above 28°C - although we now use a small split system sparingly, to reduce the extremes. It struggled when outside temperatures dropped to -5°C on one occasion and managed 46°C on another (both broke local records).
The moral is, if you get the passive bits right, the active bits aren't going to make a huge difference, possibly not enough to justify your expense and effort (or mine in repairing them). When our house is pulled down, someone is going to make a fortune from the scrap copper.
If you want to have a go, by all means get in touch and we will try to steer you around some of our mistakes.