The Stepped Terrace

This post is a supplement to the last post I wrote about stepped terrace buildings in 2022.

Stepped Terrace is as good a name as any and I’ll continue to use it as long as it’s not confused with terraced mat buildings that step up a slope instead of stepping up from flat land. The idea of both is to give everyone some sky overhead by horizontally displacing successive levels to give their occupants a small patch of sky above their heads and a taste of what it might feel like to live in a detached house. In most cases, paired buildings step up in opposite directions to create a void like an inverted V or Y. If this large empty space can be used for something useful then so much the better and this is sometimes the reason for stepped terrace buildings.

The first examples of stepped terrace buildings are those Antonio Sant’Elia depicted in his 1914 drawings. He didn’t tell us what he imagined happening in the spaces created beneath although, in the second image below, I’d expect at least a railway station or turbine hall. Rather than the elevators being within the building, most of Sant’Elia’s sketches have them on the outside as a design feature more expressive than the buildings they serve. The history of the elevator isn’t as short as the photograph of Elisha Otis at the 1854 World’s Fair would have us believe. Apparently the 1870 Equitable Life Building in New York was an eight-story affair and the first office building to have elevators but it was the Sant’Elia’s 1914 when the global mass production of elevators really took off and it was obvious that elevators were the vertical transportation of the future. Sant’Elia is said to have been a Futurist but, as far as elevators were concerned, he was well behind the curve.

External elevators didn’t turn out to be the future of vertical transportation but Henri Sauvage had no choice but to put them on the outside of his 1922 Social Housing and Swimming Pool at Rue des Amiraux in Paris because of the eponymous swimming pool. There’s one elevator on each side of the building, in those shorter protrusions that don’t go all the way to the top.

These two two-sided elevators are accessed directly from the street but residents on levels 1, 2 and 3 have no access to them (see below left image). These elevators go up to level 5 and residents on levels 6, 7, and 8 must then switch to the stairs using corridors overlooking the pool rooflight and passing rows of what look like storage spaces. This is bizarre. If a person living on level 3 wants to visit someone on level 8, they would need to walk down two levels, take the elevator to the fifth and then walk three levels up to 7. There’s five levels of stairs whether you choose to use the elevator or only the stairs. Those stairwells continue vertically to enable access to corner apartments until even the uppermost floors have to step back (middle below). The lightwell to the pool makes sense and the uppermost three residential levels may not have elevator access but at least they can be dual aspect. It’s the duplicated corridors behind the two galleries of changing cubicles as well as the two levels of fairly capacious storage units that make me think Sauvage had too much space he needed to fill.

Sauvage had a penchant for stepped terrace buildings. Whether the elevators were on the inside or outside must have depended on what he proposed for the space created below. The image on the right below is unusual in being a stepped terrace pyramid above a perimeter block. These never took off but it illustrates the problem with stepped terraces. For any given degree of stepping, the building geometry is fixed and the site density is proportional to the area of the site. In other words, you can’t keep stepping up and back forever.

Kenzo Tange’s 1959 Terraced Housing Project for Boston Harbour was produced when he was visiting lecturer at MIT 1959-1960 and is another of those projects history forgets because it’s not history as we’re led to like it. With Tange, it’s difficult to say whether a project actually came first or whether it was just announced first. My guess is that Tokyo Masterplan was mostly finalized before Tange went to MIT, and the Boston Harbour project was then made as a studio topic and apparent precursor to Tokyo Masterplan. Nobody bothered with environmental impact assessments back then but at least the structure looks do-able.

Kenzo Tange’s 1960 Tokyo Masterplan features stepped terrace buildings spanning transportation infrastructure but, unlike Boston, not triangulating. However, Tange did take pains to explain that automobiles need to transfer from a high-speed network to a low-speed network before they reach their destination and this is true.

Speaking of stories’ it’s difficult to tell how many storeys these Tange vision buildings have. Guessing from the size of the cars and the roads in the model, Id say the buildings vary between ten and twenty storeys. The section seems to imply the two sides are leaning against each other but there’d need to be some sort of concrete pad linking them, ideally above water level. Speaking of water, the average depth of Tokyo Bay is 40 metres. Tricky. For reference, the average depth of Venice Lagoon is 10.5 metres and the maximum depth 21.5.

Tange’s stepped terraces also have unexplained yet apparently meaningful holes. This is probably just design. What probably isn’t just design are the roofs curved to evoke some association with the curved roofs that had been part of Japanese (not to mention East and South East Asian) architecture since the 6th century. A curve may be longer than a straight line but it still has the same number of floors, only now the lower levels have more sky than the upper ones. I’ve only just noticed this, but what’s happening onshore in that lovely wood model looks like Isozaki’s 1960 version of City In The Air.

1967 Lower Manhatten Expressway, Paul Rudolph

This is another dubious project resulting from The Ford Foundation asking Paul Rudolph to produce this ‘speculative design’ that everyone hated. But the drawings are nice and well exhibited by MoMA. The project is short on details such as structure and program other than roads and how to remove leaded petrol exhaust fumes from them.

1968 Allt-Erlaa, Harry Glück, Kurt Hlawenciczka, Thomas Reinthaller and Franz Requa

The 1960s seem to have been the heyday for stepped terrace housing developments. This one illustrates the basic problem of stepped terrace housing. Henri Sauvagee increased density by putting his steps on a podium but here extra units are added to the top to make an inverted Y-shape. This shows that, if you don’t have infinite land, then in order to achieve housing or budget targets, you’re going to need steep steps and, like it or not, you’re going to have to go straight up sooner or later.

Tange’s curved blocks could be an elegant solution to this problem of a stepped terrace needing to become increasingly vertical as the limits of the site or the structure are reached. In other words, the curve prevents a building of two halves as at Allt-Erlaa, and disguises the fact that the promise of equal amounts of more sky for everyone hasn’t been met.

These steep steps at Allt-Erlaa don’t leave that much below volume to be filled. The middle image below is the closest to a photograph of it I could find – but it somehow still manages to indicate what these spaces aren’t. Information is sketchy. What the unit layouts reveal by all bathrooms and utility rooms being internal and the kitchens almost so reveal that terraces are expensive space. In fairness, this problem isn’t unique to stepped terraces.

1972 Kawaramachi Housing Complex, Sachio Otani

I’d previously noted a certain forced Japaneseness about some of Tange’s early work but, if Otani was in Kenzo Tange’s studio in 1955 then its unlikely he helped design the Hiroshima Peace Memorial Museum because 1955 was the year it opened. Yet, he started his own practice in 1960 (while Tange may or may not still have been in Boston). Otani’s four most photographed projects are the 1966 Kyoto International Conference Center (left, below), the 1969 Kanazawa Institute of Technology, and the 1970 Kawaramachi (stepped terrace) Housing project (middle) and the 1987 Okinawa Convention Centre (right).

The photogenic 1966 Kyoto International Conference Centre could well be early Tange with its mashup of Japanesque and proto-Metabolist form.

Otani’s Expo ’70 Sumitomo Pavilion (left, below) is obviously from the same stable as Isozaki’s Expo Plaza and Kurokawa’s Expo Tower and Toshiba Pavilion respectively.

Otani’s 1970 Kawaramachi Housing is mostly forgotten, but it looks like it’s been well maintained these past fifty years. Once again, we have relatively steep steps and a large amount of vertical building but this one is bizarre for creating the terraces and then building over half their area. High-level windows bring some extra light to the space behind these balcony additions and the fifth image below shows some challenging fire escape ladders. It’s difficult to see what the benefits of the stepped terraces are in this case because, for what sky there is, the balconies are used no differently from any other, and the space formed below isn’t used for anything meaningful either. The middle photo below shows what happens to pipes in buildings with inclined structures. Good to know. [Photos: Tokyo Fox]

So far, we’ve seen three different ways to deal with terrace drainage. Excuse me. Kawaramachi seems to have open channels. There’s no reason the other two couldn’t have.

Lastly, there’s Kunshan City in Kunshan, not far from Shanghai. M.A.O. architects from Japan were the designers.The development has found fame as “the pyramid building” but there are three buildings, two with adjacent sides stepped and the other two with only one and this highlights another of the problems of this typology. Stepped terrace buildings may provide occupants with a corner of real sky above their heads but illumination isn’t evenly distributed as there’s always a sunny side and a not-so-sunny side. Tange’s Tokyo Bay “masterplan” had its double-side blocks aligned NW-SE but, the lack of structural benefit from them being double-sided is why I think their curviness is intended to either evoke the oriental roof association and/or disguise to how the density was increased.

Plans and sections aren’t forthcoming and again, this is revealing. However, there is this YouTube video walkthrough. From these few screenshots below of one of the smaller buildings, we can see a deep typical floor plan (1st photo) and (2nd) the balconies stepping back to meet the elevators and create a pleasant but useless space. Inside a unit (3rd) there’s a double height void looked over by the bedroom upstairs but this is a development that’s said to have over 1,000 unit layout variations. Again, and it’s a consequence of the geometry, many adjacent units have handshake balconies. I’ve written about BIG’s King Toronto before.

So why do it? The problems are all inherent.

• The deep plan either creates very long and poorly illuminated units that are often single-aspect.
• Providing everyone with a terrace sets the site density as a function of the angle of stepping.
• Stepping terraces creates problems of vertical overlooking, while pixel geometries also introduce horizontal overlooking.

I’ll ignore the problem of the unequal access to direct daylight as this problem isn’t unique to stepped terrace buildings. The problem of the space below the terraces needing to be filled with something is unique to stepped terraces as it is a direct consequence of them. John Portman’s solution to put a hotel lobby in it was a good one if you’re designing a hotel with balconies in small steps on the other side. Rudolph’s and Tange’s solutions were to fill it with roads. I suggested just using the first two levels for one level of car parking but this accepts a height and density limit. Maybe that’s it should be? Stepped terrace residential units of the inverted V type have their occupancy determined by the depth of the site. Too small and there’s no room for terraces and too large and there’s structural complications and a surfeit of inner space.

• • • 

Pasadena Heights V3.0

This was my attempt to ventilate all rooms and for occupants to decide which spaces receive the best illumination. It borrowed heavily from Kunio Maekawa’s 1973 Pasadena Heights project. It’s proof of concept and every room has windows on two sides but, as the old saying goes, if you want windows, then you’re going to need some external walls and this building has many. It’s basically a checkerboard of internal spaces and light wells and this building is all external wall, and with no party walls. It’s an expensive way of building but, on the other hand, there aren’t any huge spans. For now, I’ll file this one under Possible Future Mud-Brick Residential Development.

Professional Development
Moneymaking Machines #6: King Toronto
The Terrace