Suppose we take seriously Le Corbusier's idea that a house is a "dwelling machine"; what would it look like? I propose the Subur-bean or Suburbean, a sort of immobile or minimally mobile alternative to a submarine, designed to give the humans maximal control over their living situations. Previous notes on this subject include Thermodynamic systems in housing.
I have been careful in this note not to rely on any novel technology that the humans do not have working already; it's all just a straightforward application of known, working processes to the problem.
The moment in my life when my house was the most convenient was the three months in 2006 when I lived in a Volkswagen Vanagon with Beatrice, which we dubbed the Magic Bus. This is sort of counterintuitive: you would think that living in such a tiny space would be very inconvenient, and indeed accessing some kinds of things was somewhat inconvenient, requiring disassembly of the bed for example; but the day-to-day tasks in it were quite convenient. At first we had a terrible time finding things, but after a few days we got into the habit of always putting things away as soon as possible, and it became easier to find things than in a larger house, which allows you to get into bad habits.
I say it was a tiny space because it was a little under two meters wide, about four meters long, and about a meter and a half tall; I couldn't stand up inside unless we cranked up the pop-top, which raised the roof to a height of over two meters in the center portion of the vehicle (say, 1.5 meters by 1 meter), also providing access to another bunk up there that was two meters by 1.5 meters long, about 0.8 meters tall. (This two-meter length included the 1-meter hatch from below, which you would close with the second half of the bed if you were going to sleep up there, which we never did.)
Many camper Vanagons had the front "bucket" seats mounted to swivel around toward the back, so that when it was parked, you could turn them to face the center of the interior. Ours was an aftermarket conversion that didn't have that, so the front driving area was nearly partitioned off from the main interior area, making it even smaller than it would be otherwise. So the only seating in the main living area was the vinyl-padded bench seat that the bed folded down into, which could seat about three people.
As a vehicle, the Vanagon was a terrible piece of shit, but as a house it was wonderful. Unfortunately we drove it a lot in kind of an abusive way and thus had to spend a lot of time fixing it.
It had a vinyl awning or canopy that could be rolled out from one side of it to provide a shaded area to set our chairs in.
One thing it couldn't provide internally was sewage facilities; for this we relied on external bathrooms or, occasionally, Gatorade bottles. And of course the external scenery was usually gorgeous and changed often.
So the inner sanctum of the Magic Bus was about 12 cubic meters, much of which was storage cabinets (and the refrigerator, sink, stove, and batteries), and there were another couple of cubic meters of machinery mounted below, including the hundred-kilowatt air-cooled boxer engine, wimpier than many modern motorcycle engines; and the additional bunk was about 2.4 cubic meters more. The canopy occasionally enclosed another 10 cubic meters or so, and another 12 cubic meters or so were needed for access to the back lift gate.
And this was very comfortable and convenient for two people nearly 24 hours a day for several months, with occasional breaks. Without, it should be emphasized, any drugs other than the occasional ibuprofen or beer for one of us, or any particular level of spiritual enlightenment, since of course drugs or enlightenment can make you comfortable under almost any circumstances.
There was something very empowering about having everything you need within arm's reach, and it also had very nice acoustics and a seriously nice sound system.
Back when I had cars, I often enjoyed sitting in them listening to music. Bucket seats in cars are more comfortable than many chairs, though not all, and acoustics inside of cars are generally much better than in houses, offices, or classrooms, largely due to the necessity of damping engine and road noise; although false-ceiling acoustic tiles help significantly, it's very expensive to buy and install the tens or hundreds of square meters of acoustic foam needed to give really good acoustic characteristics to a house, office, or classroom. If the car engine is running and you have an air conditioner, you have tens of kilowatts of power available for heating or cooling the interior, so within a minute or two you can get the temperature to a comfortable level, whatever your preference is.
When larger spaces go out of control and become unlivable, it happens more slowly than in smaller spaces; and bringing them back under control takes longer.
For example, consider CO₂ homeostasis. Mina's bedroom is 36 cubic meters, three times the size of the Vanagon, and in Reducing nighttime bedroom CO₂ levels I calculated that if Mina and I are in there for 12 hours without any airflow, we could raise the CO₂ levels from their default 400 ppm to 14000 ppm; it would take us 20 minutes to double the CO₂ level. Her whole apartment is 14 m × 3 m × 4 m, or 168 m³, so if it were sealed, it would take us 90 minutes to double the CO₂ levels. By contrast, in the Vanagon, it would take only 6 minutes. So if you manage to seal the Vanagon up too tightly, you'll notice within minutes that it's getting uncomfortably stuffy, and you'll be able to tell within a few minutes if improved ventilation measures are successful.
Similarly, if the air conditioner in Mina's apartment breaks, it may take a few hours for temperatures to reach the level where you realize this has happened; and if the house has gotten too hot while you were out, it can take half an hour or so for the air conditioner to bring the temperature down to a livable level.
However, larger spaces generally require more energy and more material to remain under control, although as the scale increases, many of the things you might want to control become more efficient to control.
There is an example above of acoustic tiles: a single-person listening booth of 0.8 m × 0.8 m × 1.5 m has a total surface area of 6.1 m², which works out to 6.1 m² of expensive acoustic materials per person. By contrast, a 5 m × 5 m × 3 m listening room has a total surface area of 110 m² and can hold about 35 people, so it needs only about π m² of expensive acoustic materals per person.
A more extreme example has to do with climate control.
A minimal private bunk, like the upper level of the Magic Bus, might be about 2 m × 1 m × 0.8 m, with a surface area of 8.8 m². (This is a bit cramped; this body is 0.95 m high, seated; but we can imagine that there's a bit of ceiling height variation to accommodate such things.) If you're trying to maintain an insulated tent of this size at 22° at a time that the outside temperature is 0°, and your insulation is 100 mm of 0.04 W/m/K insulation (typical for insulation materials; see Air conditioning), you have roughly 10 m² of effective surface area, so you need 88 W and 1 m³ of insulation, both per person.
If, by contrast, you have a sort of barracks with two stories, each 2.4 m tall, containing a 5-m-long corridor whose walls are covered with three levels of 0.9 m × 0.7 m doors leading into such capsules --- sort of like a morgue or mausoleum, but for the living, like a Japanese capsule hotel --- the overall building is 5 m × 5 m × 5 m, not counting the 100 mm of insulation around the outside. Its surface area is about 160 m², so under the same conditions it loses about 1400 W and needs 16 m³ of insulation. But it holds 48 people (assuming the last meter of corridor is used for travel between the levels and whatnot) so this 29 W and 0.33 m³ of insulation per person, one third of the individual tent. In fact, if occupancy is over about 30%, it won't need active heating; it will need cooling.
There is no limit to this kind of increase in efficiency from scale, assuming the whole interior can legitimately be at the same temperature.
Another example has to do with lighting. As described in Illuminating yourself with 10 kilolux of LEDs to combat seasonal affective disorder, a few years ago you could supply artificial daylight (10 kilolux) to an individual human in a sort of tanning booth or SAD-treatment pod at a cost of about 200 watts and US$150 of LEDs. But obviously illuminating a normal-sized living room with 200 watts will not come close to daylight. (More information about lighting costs is in Illumination cost, including an option that uses many times more energy but uses only US$3.20 of quartz-halogen bulbs; notes on using lightpipes instead are in Can artificially-lit vertical farming compete with greenhouses?, Subterranean glazing, and Notes and calculations on building luxury underground arcologies for whoever wants them.)
Indoor pot growers often use portable closets illuminated in such a way to daylight levels, sometimes with low-pressure sodium bulbs; modern high-efficiency LEDs are also apparently considerably better than the ones I mentioned above. (See Can artificially-lit vertical farming compete with greenhouses?.)
The Suburbean is a self-contained living space capable of months-long autonomous habitability for three people under a wide variety of external conditions, much like a nuclear submarine, but with the objective of bringing habitability to urban rather than marine environments. It is built into a standard-width refrigerated "high-cube" TEU shipping container, 6.1 m × 2.44 m × 2.90 m, so that it can be easily loaded onto a truck, train, or ship for moving, although it's better not to do this while people are inside of it. The refrigeration system is used to maintain the interior at a comfortable temperature chosen by the inhabitants, rather than a temperature suitable for preserving food. The Suburbean has very limited autonomous mobility --- it's more of an autoimmobile than an automobile.
Unlike a nuclear submarine, the Suburbean does need access to external air, at least to exhaust its waste products into and usually also to burn its diesel fuel with. When no external power is available, its refrigeration unit and other internal energy consumption is powered by two 1-kW onboard diesel engines, providing some 2 kW of mechanical or electrical power.
The whole Suburbean weighs some 20 tonnes, well short of the 30-tonne limit for a loaded TEU. This means that when lifting itself vertically with its six built-in winches with its 2-kW-output generator engine running at 100% duty cycle, it can rise only about 10 mm/s, although when running off batteries or external power, it can manage about 130 mm/s, limited only by the 25 kW of the winches. In horizontal or downward movement, when it need only overcome friction, whether suspended in the air by its winch cables or running on wheels, it may be able to move several meters per second over limited distances under its own power. This is what I mean by "very limited autonomous mobility".
How realistic is this?
Amazon lists a 6 horsepower (=4500 W) winch for US$320. eBay lists a bunch of 1kW handheld portable generators for around US$200 to US$300, but they all run on gasoline; diesel generators seem to start around US$1200 and 7kW. A plain TEU-sized shipping container seems to cost around US$2000, but refrigerated ones seem to cost around US$6000, or up to twice that new.
Like some refrigerated shipping containers, the interior of the Suburbean is superinsulated with vacuum insulated panels (< 8 mW/m/K) to reduce the power necessary for refrigeration; it uses 50 mm of them in two staggered layers, reducing the interior dimensions by 100 mm in addition to the 200 mm or so consumed by corrugation, to about 5.7 m × 2.04 m × 2.5 m (32.3 m³, of which 20 m³ is air). Over the effective surface area of 70 m² or so, this works out to about 11.2 W/K, so over the 0° to 44° temperature range, only a maximum of 250 W of forced heat flow is needed at steady state, or about 85 W electric to power the heat pump with its CoP of 3, which works out to about 210 W of diesel fuel. If three people are actually present inside, this adds about 300 W of forced heat flow when the outside temperature is too hot rather than too cold, increasing these numbers to 550 W of heat flow, 180 W electric, and 460 W fuel.
The usual wooden flooring of a shipping container is not present, since the vacuum insulation layer includes the floor as well, requiring a rigid protective support surface to prevent damage to it.
The air conditioner's CoP of 3 is achieved by liquid-cooling its condenser coils with antifreeze that is circulated through pipes that heat the metal outer walls of its entire 35-m² lateral area, providing a large surface for radiative and convective cooling. When the air conditioner is operated in heating mode as a heat pump from outside to inside, this operates as liquid-heating the evaporator coils with the same antifreeze circulated through the same pipes, where they absorb heat from the environment.
When the Suburbean is being heated rather than cooled, the diesel generators' exhaust is routed through a heat exchanger to transfer the heat they produce to the interior as well through a closed-circuit heat-transfer fluid, the same way that car heating systems typically work. This reduces the energy demand of the heat pump, possibly to zero.
The Container Handbuch says that typical power consumption for cooling a ThermoKing Smart Reefer TEU is around 3.6 kW, which is about 43 times what I've calculated above; EPRI's "Electric Refrigerated Container Racks: Technical Analysis" says that diesel reefers have 2-liter engines with 30 to 40 horsepower, or in modern units, 20 to 30 kW. Container Handbuch section 3.1.1.2, "Container design and types, Part 2", suggests perhaps unintentionally that typical mechanically powered refrigerated containers are not insulated, and that insulated containers typically use 50-100 mm of polyurethane foam, reducing heat conductance to 0.4 W/m²/K (type code H5) or 0.7 W/m²/K (type code H6). Instead, the 50-mm thickness of VIPs described above would have heat conductance of some 0.16 W/m²/K. This explains a discrepancy of about 3.4, leaving another factor of 13 or so from my calculation above, to be explained by people using uninsulated containers (!?) and/or the much lower temperatures.
It seems that self-powered refrigeration units are (or recently were) dominant in land transport, but plug-in electric refrigeration units (three-phase 400VAC) instead are dominant for sea transport. Carrier and ThermoKing make all the refrigerated container machinery for sea transport.
I found a "Technical Specification for Refrigerated Container Model No. SS1WN1" of half a TEU from Shanghai Reeferco. It claims they use a Carrier refrigeration unit and their insulation is 63-80 mm in thickness everywhere except on the floor where it ranges up to 135, and that the heat transfer rate Umax is 20 W/K at 20°C. Its surface area is about 43 m² so that works out to 0.47 W/m²/K, which is in the range of the Container Handbuch numbers above. Yet the Carrier 69NT40-541-300 it ships with can do 3.2 to 10 kilowatts of cooling with inside-outside temperature differences of 67° (3.2 kW) to 36° (10 kW). You would think that with a temperature difference of only 36° you would only need 720 W, not 10 kW. Maybe it's shipped with this huge air conditioner for the initial pulldown?
The Suburbean carries half a tonne of diesel fuel, which weighs 0.832 kg/ℓ and provides 43.1 MJ/kg, so this occupies 0.6 m³ of the 32.3 m³ available and stores 21.6 GJ. At the outdoor high-temperature extreme of 44°, when it needs to use 460 W of diesel fuel to maintain thermal homeostasis, this provides almost 18 months of autonomous operation before needing to refuel, as long as it has access to oxygen from air and somewhere to release exhaust.
The Suburbean also includes half a tonne of lithium-ion batteries, which at some 500 kJ/kg amounts to 500 MJ, about 2% of the energy content of its diesel fuel tank. Still, at the 180 W electric needed to maintain thermal homeostasis at extreme temperatures, this amounts to a bit over a month of autonomy without access to air; for example, during a flood or while buried under rubble, although assumptions about external temperature and insulation may be called into question in such circumstances. Lithium-ion batteries have a self-discharge rate on the order of 2% per month, which is insignificant in this context.
The batteries are rated for a discharge rate of 5C (12 minutes), which is lower than the 15C discharge rate (4 minutes) used for quadcopters and the like, but higher than the lowest-end batteries. At a conservative discharge rate of 3C (20 minutes), the Suburbean can muster a peak output power of some 400 kW, a bit over 500 horsepower; for example, for arc welding.
To recharge the batteries without consuming diesel fuel when sunlight is available, 6 m² of the 15-m² roof of the Suburbean contains 22%-efficient SunPower Maxeon Gen II monocrystalline solar panels under an openable protective cover. This provides nominally 1300 W peak of solar power, providing 260 W as a 24/7 average, at a typical 20% capacity factor. Moreover, the back of the cover is mirrored, and it is positionable under motor control, permitting higher capacity factors than are possible with statically positioned panels, and potentially even providing over 1300 W at times.
See also House scrubber and Notes on a possible household air filter.
Of course, when air is available from the outside, the Suburbean will use it, after appropriate filtration and temperature control. But above, it is explained that the batteries can provide a month's worth of climate control without air for the diesel engine. So, what about air for the humans inside in such a situation?
Remaining human-habitable without access to air requires very roughly 600 g of oxygen per person per day, which is the amount in about 2 m³ of air (2.4 kg) or 0.4 m³ of oxygen, at 1.429 g/ℓ. You'd think you could generate oxygen on demand in such situations through electrolysis of water, and although that does work, it turns out to be very expensive, about 250 MJ/kg O2 (see Why you can't run a diesel engine on water and diesel fuel with electrolysis), which at 600 g/day/human works out to about 1700 W per person, a large energy drain which also adds to the cooling load when the situation is hot. The 20 m³ of air contains enough oxygen to last only 10 person-days, or 3 days with 3 people, and storing compressed air at ordinary pressures such as 15 atmospheres (1.5 MPa) would only give you 15 normal m³ of air per m³ of compressed-air storage.
Instead, the Suburbean stores oxygen as relatively nontoxic sodium chlorate, NaClO3. Aircraft decompression masks are supplied from sodium-chlorate chemical oxygen generators; a 63-mm-diameter, 223-mm-long canister (0.0007 m³) generates enough oxygen for two humans for about 15 minutes, which suggests you'd need 1 m³ per person per month, which is closer but still too bulky. A simpler chemistry is used in chlorate candles, which are mostly sodium chlorate (2.5 g/cc) but with iron powder to produce heat, and provide 6.5 person-hours of oxygen per kg, which works out to 110 kg per person-month or 330 kg per 3 person-months.
However, the Suburbean's sodium chlorate is not mixed with a fuel in this way, so there is no risk of an oxygen-candle explosion. Instead, heating the sodium chlorate to the requisite 300° is done with an electrical heating element. This produces NaCl and 48 daltons of oxygen per 106.4 daltons of NaClO3, so the 55 kg of oxygen required for three person-months of autonomy require only some 122 kg of NaClO3, occuping 49 ℓ.
(I think it should be possible to run the diesel engines from stored sodium chlorate and diesel fuel, but the amount of sodium chlorate required is rather large, and the reduction to NaCl consumes some of the energy; 106.4 daltons of NaClO3 yield 48 daltons of oxygen, which can oxidize only 14 daltons of diesel, so you need 7.6 kg of sodium chlorate for every kg of diesel, so you only get 5 MJ/kg from the total mixture (minus whatever the endothermic chlorate decomposition energy is), and only 1.5 to 2 MJ/kg of exergy. Still, that's three or four times the exergy density of the batteries, and it might provide a viable approach to multi-month underwater autonomy. See Underwater energy autonomy for more.)
The Suburbean generates its own sodium chlorate, when water and energy are abundant, from electrolysis of an aqueous solution of NaCl, with some HCl to lower the pH, at 90°. Because this reaction also produces some hydrogen, the reaction chamber is outside the interior space, so that if there is a hydrogen leak it dissipates into the environment rather than making the internal atmosphere explosive. Normally the hydrogen is fed into the diesel engines through a secondary injector to burn it, or burned in external air with a spark igniter if the diesel engines remain off for too long. If there is no external air, the hydrogen is just released into the environment.
It isn't enough just to add oxygen to the atmosphere, though; CO₂ must also be removed. I reviewed the possibilities in House scrubber; the Suburbean's CO₂ scrubber uses six small beds of high-surface-area caustic magnesium oxide to absorb the CO₂, which it regenerates by heating them to 450° by recirculating a stream of CO₂ through the bed and through a heating element. One person-hour of CO₂ is 25 grams, if the figures in Reducing nighttime bedroom CO₂ levels are correct; MgO weighs 40.3 daltons; MgCO3 weighs 84.3 daltons; CO2 weighs 44 daltons, which is the difference. So each bed is sized to be able to hold 75 g of CO2 during its active hour, which makes the conservatively 150 g of MgCO3 (2.96 g/cc, so 50 cc) which becomes 71.7 g of MgO after heating. Excess CO2 is vented to the environment.
(Hmm, I'm not sure this is actually the right solution; I can't find any notes about anyone using MgO as a CO2 scrubber, and Wikipedia's CO2 scrubber article says the Space Shuttle's metal-oxide-adsorbent scrubber was regenerated with 10 hours of 200° air through the "Metox Canister", which it turns out actually used silver oxide, perhaps in a thin layer on the surface of some kind of inert support material that comes with a high surface area. More recent design proposals use solid amine adsorbents and Fe2O3 but I don't know if they've flown or if they're suitable for the Suburbean.)
As a backup in case all of this fails, the Suburbean is also equipped with lithium-hydroxide curtains which provide a few days of autonomy before needing to get access to air.
As mentioned in House scrubber, there are other contaminants that must also be removed, and a variety of means to remove them.
The reason there are six separate CO2 scrubbers is that the Suburbean is divided into two compartments with a hermetically-sealed bulkhead between them, connected through an airlock; experience with the ISS has shown the importance of being able to isolate the effects of accidents such as fires and refrigerant leaks to only part of the environment.
If the air quality in one compartment is becoming bad because of a broken or inadequate pollution control system rather than because of a polluting accident, the inhabitants can open the airlock to allow air to flow.
Of course, the Suburbean has the usual big doors that are the entire rear end of the shipping container. But it has some other features as well.
The Suburbean is stackable; each of its two independent compartments has a hermetically-sealed hatch in one corner of the top surface and the bottom surface, making it possible to connecting multiple Suburbeans into a single larger dwelling-machine by stacking them on top of one another and twistlocking them together, then opening the hatches and sealing them together. ISO 1496(1) requires that TEUs be tested for stacking nine containers high, but that's only safe when the doors are closed! The Suburbean has extra bracing just inside those rear doors, part of which is the hatch shaft itself, so that it can safely support that amount of weight even with the doors open.
This allows you to construct an ad-hoc autonomous minimally-mobile hermetically-sealed apartment building with space for up to 26 people to live comfortably, assuming you use the rear compartment of the bottom Suburbean as a common entry and exit.
When no other Suburbean is atop yours, these hatches can also serve as entry and exit, if you either have a way to get on and off the roof, or space underneath. Unlike the rear doors, the hatches cannot be locked by people outside, only by the Suburbean's computer systems, ensuring that escape is possible in an emergency (even if the rear compartment is on fire) and preventing outsiders from locking the doors as a way to apply coercion to the inhabitants.
The shaft connecting the top and bottom hatches is also hermetically sealed from the compartment it runs through, with a door providing access to the compartment. A positive-displacement ventilation pump runs only when air-quality sensors report that the air in both the shaft and the compartment are not contaminated. This means that air-contamination incidents in compartments do not discourage you from traveling vertically past those compartments in a stack of three or more Suburbeans. It also makes it possible for the inhabitants of a compartment to lock it against access from their neighbors, while still allowing those neighbors to pass through.
If there are stable high points to attach the winches to, no separate crane is needed to stack the Suburbeans; once some winches are attached, the upper Suburbeans can move themselves into position. The winch cables can also guy the stack of Suburbeans to points on the ground or elsewhere to reduce the risk of falling in wind. (See Bootstrapping rope bridges and other tensile structures with UHMWPE-bearing drones for one way to bootstrap the winch rope attachment.)
If your Suburbean is hanging from its winches out of reach above you, you can command it to descend to within reach and open a bottom hatch so that you can enter; then, once you have entered, you may want to command it to ascend again.
Both of the hatch shafts are on the right side of the Suburbean if you're facing its nose; this means that if you're facing the rear door, the rear hatch shaft is on your left. This means that if one Suburbean is yawed 90° on top of another, one twistlock point and adjacent hatch shaft can dock, but it's the rear hatch of one connecting to the front hatch of the other. With some additional support beyond that provided by the twistlock points, this design permits the assembly of much larger and more stable ad-hoc hermetically-sealed interconnected Suburbean Voltrons.
Because the Suburbean is hermetically sealed and has a volume of 43 cubic meters, but a gross weight of only 20 tonnes, it floats if it falls into water as long as it doesn't leak, like other shipping containers.
Humans who feel that they're in a small, cramped space for a long period of time will be unhappy. So within the Suburbean there are areas with "infinity mirror" illusion ceilings and parallel mirrors on the walls to create the illusion of a large horizontal space. Extra acoustic foam on the other walls and behind holes in the mirrors provides the acoustic sensation of being outdoors; fans silenced by baffles followed by laminar ductwork provide a breeze.
These are potentially a big issue, as they are in space travel and nuclear submarines.