Let’s first take a look at the mother of all microflats, Kisho Kurokawa’s Nakagin Capsule Tower. At architecturalmoleskine there’s an excellent post which will fill you in if you’ve never seen or heard of it.

In this post however, I’ll only be concerned with how little space it takes for one person to live. At Nakagin, it’s 2.5 m x 4m = 10 sqm, including the bathroom.
(FYI ¥60,000 = US$620/month)

I get the feeling that Japanese people are going to be better at these microflats than us but the point is that somebody thought about everything a person might need in order to call a place a home. For a first attempt, it’s not bad. The year was 1970.

Here’s a one-room furnished apartment currently for rent in Tokyo. The layout is typical but this one’s unusual in having a side window as well. Most won’t.

This next one’s a variation. The entrance and shoe cupboard are cleverly positioned. The character 洋 means “western” (style room) and usually means a laminate floor instead of tatami. But if it were to have tatami, it would have eight of them @ 0.88m x 1.76m = 12.4 sqm. The balcony is for drying laundry. Shall we say this one’s 25 sqm?

This next one has a side entrance. By the front door is a space for the washing machine.

Unlike the Nakagin Capsule Tower capsules, these two plans are more like single rooms with service spaces attached for, once in the 8-mat space, you could in theory, feel like you’re in an 8-mat room anywhere – if you didn’t have much stuff, that is. I’ve mentioned before that people seem to think that Japanese live without much stuff. This is not true. Maybe, in the plan above, a futon could be rolled up and shoved in the closest, but I doubt it would be. Not with a laminate floor. It makes more sense to get a bed, use it as a sofa, and to put more stuff under it. I can easily imagine how people actually live in the SANAA Okurayama apartments. You’ve seen the plans elsewhere.

The lady in this photo is nervously waiting for the movers to arrive with her banana holder, piglet- and puppy-face rice bowls, fish-shaped chopstick rests, My Little Pony lunchbox, pantie dryer, Hello Kitty futon set and other stuff.

Over in Britain, Piercy Connor Architects is the king of microflats and probably even invented the word. A full-scale version of his microflat was installed in Selfridges Oxford Street window and lived in for two weeks to the general interest of media and amusement of passers-by.

His is interesting in what it says about the British Way Of Life. First of all, there’s nothing new about it – even its size isn’t that unusual for some converted apartments. The only major difference is that an apartment this size was being presented for the first time as an architectural solution for an as-yet unexploited market. Once even the hope of owning such a reduced product has become unrealistic, then it’s safe for architecture to shift its market focus downwards a notch.

The microflat has a separate bedroom with a wardrobe. It has a bathroom off a corridor. I has a kitchen with a full-size cooker with four hobs and an oven and – of all things – a full-size stainless steel “extractor” hood which, to me, doesn’t seem like a wise use of one cubic metre. Such is the power of aspiration. The microflat also has the two other aspirational symbols of balcony and flatscreen TV, both compromised but nonetheless present, competing for wallspace.

The idea was to build these in inner-city areas but land values have no doubt prevented that. These places lose their value when they’re sited on land that makes them feasible to build. It’d be great if small living spaces could be located in supporting and/or lively neighbourhoods but such neighbourhoods tend to add more value to upmarket apartments. The only way out of this is to locate microflats in areas that are distinctly unattractive to other markets. Families are easy to discourage, businesses less so.

With a micro-flat, the bed seems to continually be a problem. Having a separate bedroom like the Piercy Connor apartment means that one third of the space is not fully used or appreciated 24 hours a day. This is not clever but may be a difficult habit to break. Having said that, it’s not a particularly Japanese thing to sleep in a bed but, in Nakagin, it made sense to double-up on the seating and the bedspace instead of doing something clever with either or both of them. There’s just the one and only one nice place to be. The Japanese would think of it as a floor and sit on it and when it came to sleep, sleep on it. Us, we’d think of it as a day bed that can be used for sitting or sleeping.

Here’s a variation with a bed mezzanine – un petit unité d’habitation. It’s by Julyan Wickham of Wickham van Eyck. It’s not bad. Placing the bed above the bathroom is a good idea, but dramatically increases the volume of the flat. Ceilings are 4m perhaps? This added volume is no doubt pleasant for the same reasons that having ‘space to waste’ usually is but this doesn’t seem like the right place to have building volume not fully utilised. Also, it weighs in at 31.5 sqm not counting the balcony.

The access corridors seem unnecessarily generous, probably because of the unconvincing entrance lobbies – is it really a choice between them and a larger mezzanine window? And what’s going on upstairs? The wardrobe seems unnecessarily large, the line of its doors defined by the width of the kitchen in order to force some blue Rietveldt-esque blue solid. But Wickham has some good things to say in this article.

More than 70% of the worlds population live in an urban environment of some kind and 99% of them are not rich and do not have much mobility yet where they live and work is the real substance of any town or city and without them there is no market or urbanism to fuel the capitalist dream or provide occupation for architects and planners.

Do I think this is advancing architecture? No I don’t, and that’s what I like about it. It advances the cause of useful building serving an important social need in ways that don’t waste resources yet try to extract maximum value from every square metre of space. If that’s not what architecture’s supposed to be doing, then architecture is not what we’re really wanting. To be honest, when I first saw the above plan, I thought it was one of the Unité d’Habitations variations.

But then I saw that the non-‘architecture’ plan has the elevators and service risers sensibly placed in the middle of the building so as to minimise average travel distances for people and utilities. The non-‘architecture’ plan also has (what I imagine/hope are) fire escapes sensibly positioned at the end of the long corridors so that all people have a choice of two directions of escape  – unlike the Corbusier building where fire escapes take up space that could be better used for more habitations. Moreover, they will actually escape towards the light that comes in at those ends which appear to also be communal balcony area. I can see that whoever designed this building did the best they could with a tricky brief.

The idea of a cell for human habitation is not a new one, of course. This is another excuse for me to insert this image of Megaton City, The First City, from Superstudio’s Twelve Ideal Cities project.

Superstudio’s Twelve Ideal Cities project is a wry comment on twentieth-century modernist utopias, and it supposedly represents “the supreme achievement of twenty thousand years of civilization.” In the First City, or 2,000-Ton City, shown here, cubic cells stacked atop one another form a continuous building that stretches across a green, undulating landscape. Each cell is equipped with technology capable of accommodating all human desires and physiological needs.

Although I’m curious about the technology capable of accommodating all human desires and physiological needs, I’m only interested here in the size of the cells and how much space that technology takes up. Unfortunately, we don’t have that information. What MegatonCity does have in common with other microflats is a repetitive configuration.

Multi-coloured balcony separators don’t really do much to alleviate this. The Piercy Connor microflats tried alternating the direction of the balconies and cladding in timber but to little effect. The capsules at Nakagin were fully kitted out when they were lifted into place and attached with only four bolts. In 45 years, none were ever upgraded. At least with Nakagin, there was the potential for the capsules to be replaced if anyone had wanted to.

This next building only has the illusion of capsules. This is New Sky Building #3 (1972) in Higashi-Shinjuku, Tokyo, by Yoji Watanabe.

Semi-abandoned until last year, New Sky Building #3 has been fully restored for use as shared housing, becoming Tokyo’s only historic capsule apartment building with a future to speak of. The restoration was possible thanks to the building’s all-steel construction and distribution of square footage, an ultra-flexible layout for which Watanabe holds the patent.

I’d be interested to see a copy of this patent but, in the meantime, we can understand the structure from this plan of the building before its refurbishment. Judging solely by the sizes of the stairwell and doors, the individual pseudo-capsules seem like the 20 sqm we’re looking for. However, what we really have here is only two apartments – the paired corridors you see here, are internal corridors belonging to each of the two apartments. Communal corridors are in the lighter grey.

And here’s a floor plan after refurbishment. [Thanks blog binturong.] It shows how the individual capsules were combined to make smaller apartments, not larger ones.

The Nagakin Capsule Tower was the built manifestation of one such concept – the unicore typology. While Yoji Watanabe was not part of the metabolist group, his New Sky Building shares a similar design parti with the Nagakin Tower – modular capsules are affixed to a main circulation core. While the core remained a permanent structure, the continual replacement of capsules was, at the very least, an implicit idea.

However there are notable differences between the two buildings, which may have contributed to their eventual adaptability or lack thereof. First, the basic living unit in the Nagakin tower was a single capsule. In comparision, several capsules make up a living unit in the Blue Sky building – 4 [2!] such apartments occupy every level. With its small interior size, each capsule in the Nagakin Tower was limited to use as a single person residence (the prototypical salaryman). The larger sized units in the New Sky Building though, could accomodate different modes of living and other programs, thus broadening its appeal to a wider range of people.

This quote creates the impression that New Sky Building #2 has capsules when it doesn’t – it has the idea of capsules. Anyway. Another problem that all built and unbuilt Metabolist structures had and which New Sky Building #3 doesn’t have, concerns services. With a building such as Nakagin, the capsules are serviced from the structural core and this means all services have to pass through that structure, restricting their maintenance and upgrading. To compare, here’s a plan of Nakagin. There’s a few places where vertical risers could go, but everything must pass through the structure to get to them. Just like the services, people have to pass through the structure to get to the things that move them around the building. Moreover, apartments contrived to have side entrances have unnecessarily long service runs. A genuine metabolism would not do this.

I think we’ve seen enough now to generate some rules for microflat design.

RULES FOR MICROFLATS

1. Consider the structure and spatial enclosure as functionally separate entities but not necessarily as physically separate entities. This is a fundamental problem with Metabolist architecture. It reduces flexibility and complicates services.  
2. Do not think of the basic unit of human habitation as a capsule. This reduces adaptability over time. It is probably a basic human instinct to have some sort of physical entity (such as a capsule) to call their own, but this is an unnecessary luxury and not really part of the problem that should be being considered. Designing a capsule and then clustering it does not seem the way forward.
3. Try to consider the services as functionally separate entities but not necessarily as physically separate entities. Service risers need access. New services such as cable TV or fibre-optic broadband mean that wiring and cabling may need to be completely replaced within a decade or two.
4. Embrace repetition as default. Instead of forced or gratuitous variation, accessories such as sunscreening or balconies can (and should) be used to generate meaningful variations from site and environmental factors.
5. Do not build-in cultural inertia or social aspiration. Remember that these are microflats. Every cubic metre must make the best possible use of that space. 
6. Design for replacement, build-in as little as possible. These days, not many people really want to watch television on a 1970s television or listen to tracks on their reel-to-reel. Technology gets replaced. Today’s bathroom taps and windows are much more efficient and higher performance than old ones. Incandescent bulbs are no longer in production.
7. Don’t be too clever. Objects that fold away or disappear into walls still occupy space. The presence of each object has to be justified. A room that has no objects in it is a room that cannot be used.
8. Start from scratch. Do not contrive a design to allude to grander accommodations.

Here’s a design I did in 1995 as a competition entry. The structure is a concrete matrix, with access stairs and services at its periphery, and two apartment entrances at each access point.

Here’s what happens at the access points. At each landing are two front doors, and two ‘back’ doors. The gap between the stairs is an open service riser.  

Party walls and floors define the volume of the apartment between the columns and beams of the matrix. Units of this volume can be traded between vertically and horizontally adjacent neighbours. A minimum apartment would consist of a service module as standard equipment plus one other – approx. 30 sqm. There is no maximum. Party walls can be replaced or upgraded to better performing ones as technology or need dictates.

External walls are also chosen and/or upgraded for the same reasons or to reflect the aspirations and ‘individuality’ of the owner. Elevations generate themselves.