Category Archives: spatial planning

Garden cities and forest gardens

In Ebenezer Howard’s Garden Cities of Tomorrow (1902), we can read his proposal to surround a 1,000-acre town of 30,000 residents with 5,000 acres of farm land. Since in space industry and agriculture will mainly be located outside space settlements, we could look for alternative uses for this land.

Howard pays little attention to recreation in his book. Though his design includes a central park, his garden city plan does not reserve land for sports such as field hockey and ultimate. Because sport does positively affect human wellbeing, we propose to reserve some of the land originally reserved for farms to sport fields.

In chapter 10 of his book The High Frontier (1976) O’Neill suggests that space settlements are ideal locations for gardening. Our suggestion would to combine “O’Neill” and “Howard” by reserving a substantial portion of the “farm” land for allotment gardens. These are gardens assigned for individual, non-commercial gardening.

O’Neill also advocate the concept of polyculture in his book. Polyculture is the practice of growing of multiple types of crop in the same plot of land. It’s the opposite of monoculture, which dominates modern agriculture. Scientific studies have shown that polyculture is preferable to monoculture.

There are several types of polyculture, one particular interesting concept are forest gardens. The practice of forest gardening is an ancient one, but Robert Hart has formalized it in his seven-layer outline. The first layer consists of mature fruit and nut trees, which (of course) will provide fruit and nuts. The second layer consists of smaller trees, while the third layer contains fruit bushes (currants and berries). Then the fourth layer are perennial vegetables and herbs. The fifth layer are root vegetables, the sixth are edible plants which grow horizontally. The final layer consists of vines and climbing plants.

Forest gardens is considered as a low maintenance type of gardening, and hence suitable for allotment gardens. Most people with allotment gardens, work there in their spare time. An additional benefit is that forest gardens contribute to food security by providing a wide variety of food. Potentially forest gardens could be established on rooftops.

Forest gardening is a type of agroforesty, the combination of agriculture and forestry in one system. Agroforestry does not need to limited to non-commercial allotment gardens, the utilization of non-timber forest products in forest farms can make forests management commercially interesting. We propose that forest farms should be leased to private parties, on the condition that these forests will be open to the public (see also our post on Allemansrätten).

Time zones and separation of functions

One of the major advantages of space colonization by the use of free space habitats instead of planetary “space” colonies, is the separation of functions. Gerard O’Neill already advocated that residence, agriculture and heavy industry should be separated from each other, i.e. that agriculture and heavy industry should not be done in the same structure where most residences are located.

In regard of the separation of agriculture and residency, O’Neill gives two main arguments. First, in a space settlement we have full control over both climate and day length. However, the climate preferred by most citizens is not necessarily the most optimal climate for the cultivation of crops. Second reason is pest control. If in an isolated space farm a pest will occur, it will be easy to deal with it by sterilizing the farm by increasing temperature above the limit life cannot survive. It’s quite obvious that we cannot do this, in a space habitat populated by humans.

For the separation of heavy industry and residency, the arguments are even more straightforward. Heavy industry impose a great danger to health and safety through its pollution and potential of explosion and similar disasters. By banning heavy industries from space habitats, we create a clean and save environment for people to live.

A second argument put forward by O’Neill is related to his proposal to divide space settlements over three time zones, with a 8-hour difference between each successive zone. Because heavy industry is located outside any space habitat, they can be in continuous operation. And if the industry hires shifts from different time zones, night work which is considered as unpleasant by most, will be avoided.

O’Neill imagined that space settlers employed in heavy industry, would commute each day between their home and their workplace. But technology has improved much since the mid 1970s, that nowadays much work can be automated and where people are still needed teleoperation will allow workers to run factories without leaving their space habitats or even their homes.

Besides the desire the avoid night work, there’s another reason for dividing space settlements among different time zones (which surprisingly is not mentioned by O’Neill). The principal power source of space settlements will be solar power. And since there’s no night in space (in space settlements night has to be created by covering the windows), space based solar power plants will run continuously and hence have a continuous output. But the demand for power is not continuous over the day, causing surpluses at some moments and shortages at others.

If we divide the population of three time zones with an 8-hour difference, the power demand curve will be flattened. This because if one settlement is facing a power shortage at some point, it’s likely that another settlement has a surplus since their population is experiencing another phase of the day.

Public transportation in O’Neill cylinders

In a previous post I discussed the spatial planning of the interior of O’Neill Cylinders. In a note I promised to make another post about (public) transportation inside O’Neill cylinders. For the sake of the argument, I will assume here that the chosen spatial planning is either the Broadacre cityGarden city or Colombia design. Further I want to recall that a O’Neill cylinders has a length of approximately 35 kilometers and a diameter of 6 kilometers (specific dimension may vary among different sources, however the difference is usually only a few kilometers).

A key feature of the design of the O’Neill cylinder is the alternating arrangement of “valleys” (stripes of land) and windows, three of each. It follows from the given dimension that each valley is approximately 3 kilometers wide and 35 kilometers long. Gerard O’Neill himself proposed that there would be parallel to the valley’s heartline a subterranean maglev line. This would function like most subway systems on Earth and would enable (long distance) rapid transit in an O’Neill cylinder. However this system, would not quite suitable for short distance travel, therefore a second transportation system is required.

While the maglev subway will serve as the core of the framework of intra-habitat transportation, there will be finer second network. What requirements do we look for? Ideally we would like an on-demand service, great amount of privacy and the ability to choose our destination. However do not like to waste a lot of time for searching for parking lots. Personal Rapid Transit (PRT) is a proposed idea which would combine the best of private and public transportation.

In order to show what a PRT system might look like, I have selected two YouTube videos about personal rapid transit systems. The first YouTube video (of 8.45 minutes) is about the personal rapid systems as designed by Swedish company Vectus.

This second YouTube video (5.55 min) is a promotional video of Vectus, in which they explain how their product will work.

Yes, I do realise that Vectus is a commercial company which seeks to sell its concepts. Nevertheless, I think that this “sales man videos” give a clear picture how PRT systems would operate in practice.

The prospects of personal rapid transit systems are bright. They will enable to establish the first car-free society in history without sacrificing the individual freedom of movement.

O’Neill Cylinders and spatial planning

This post was originally posted on on October 18, 2012

In an earlier post I discussed the potential of Bernal spheres and Stanford tori for city states, in this posting I will discuss several ideas for the spatial planning of O’Neill cylinders.

The ideas I will discuss here are not developed for space colonization as such, but can nevertheless be very inspiring for Space settlers. Especially for the larger space habitats spatial planning is an important topic. In this post I will discuss three proposals: O’Neill’s own idea, Ebenezer Howard’s garden cities and the ideas of Frank Lloyd Wright. The purpose of this post is not to force a certain spatial plan on to Space colonies, but rather to provide a framework for developing better societies.

Since O’Neill cylinders provide a large plot of usable land, they allows for more sophisticated spatial planning then smaller habitats. The latter will typically be highly populated and most of their usable land will be used for housing and closely related activities. Consequently the smaller habitats will lack any significant amount of nature (forests for example), while many, if not most, people will appreciate nature.

Since the days of O’Neill, the consensus among space colonization advocates (and we follow this) is that industry, agriculture and living should be separated (the first two should not be located inside space habitats), this is an important difference with terrestrial spatial planning. Combined with the practically unlimited resources in space, we are free to design the interior of an O’Neill cylinder as we like.

In his book, The high frontier, O’Neill has given an example of spatial planning. In chapter 5 he describes the build cities at the ends of each stroke of land, referred to as “valleys”, and using the land areas it self for villages, forests and parks.

It would be interesting to look at a few spatial planning concepts from the past. In the 1930s the American architect Frank Lloyd Wright  designed his famous Broadacre City. In this proposal “true” cities would disappear, while people would spread out over the country (for this reason his plan was not very popular outside the USA). One feature of this scheme was that each family receives a 4,000 square meter plot of land [1]. Which was to be developed according to wishes of the receiving family. While there many really good aspects to his vision, there is some important critique about the Broadacre City idea, which can be found here. A serious drawback of the original design is that it heavily depends on automobiles for transportation. In a space based nation, in which people are spread over many different space habitats, cars are really cumbersome to handle. As O’Neill explained the main modes of transportation in and between space habitats are space ships, maglevs, bicycles and walking [2].

My personal favorite is, however, the garden city, a concept developed by Ebenezer Howard around 1900. In short this urban design is an attempt to reconcile the city and the countryside. In Howard’s plan a garden city should require 6000 acres of land (which is approximately 25 square kilometers or 2428.2 hectare), of which 1000 acres are used for the actual city and the other 5000 acres are destined for agriculture [3]. As I have already said, most space habitat advocates favor a physical separation of agricultural and living areas. At first sight Howard’s idea seems to be outdated, and it is to some degree. Nevertheless I believe that this garden city concept is good starting point for our own spatial plans. We should look for alternative destination for these agricultural lands, a portion can be reserved for allotment gardens, while another portion is reserved for sport associations (think about field hockey clubs, rugby clubs and so on). In Howard’s original designs there is a remarkable lack of recreation areas (to be fair Howard planned a park in the center of his city, but this is one is to small for serious sport practice.)

The actual city itself, would be an annulus around this central park and would be divided into six wards, each with 5,000 inhabitants. This would give a total city population of 30,000 thousand, in addition a further 2,000 would live in the rural area of the city. Howard also thought about what to do when the city population would grow, unlike the natural course of urban growth by which new buildings are attached to the existing settlement, he foresaw to build new garden cities a few miles away of the old one. In fact he suggested to build a central city, a garden city with 55,000 inhabitants, first  and later to build six (normal) garden cities around it. The central city would serve as a regional center. This particular configuration is not feasible for a standard O’Neill Cylinder (diameter 6.4 km and length 32 km), but is we would increase these dimension s with a factor 5 (which would give an areal increase of a factor 25) then it would become an interesting option. However such O’Neill cylinders XL will not be realized in early space colonization.

A different but related concept is Columbia, Maryland. Like the conglomeration of garden cities, the different villages of Columbia are not one single area but separated by green areas (called the Tivoli garden). The city’s 100,000 residents [4] are spread among nine villages, with a land area of 82.7 square kilometers (compare this with a valley of 107.23 square kilometers). If we would organize a valley in a similar fashion as Columbia, than we would get a city with population of between 130,000 and 143,000 [5]. In order to keep this city mostly car-free, they designers envisioned a minibus-network.

As I have said at the beginning of this post, the purpose of the mentioned examples is to inspire the spatial planners of O’Neill cylinders. And I hope they will not make the same mistakes as those made by terrestrial urban planners. Space colonization is a nice occasion to experiment with innovation on spatial planning. Of course the specific spatial plans will depend on the political choices made by the owners/governments of space habitats, different political ideologies require different spatial plans. The examples I selected here, reflect my personal believes about decentralized republicanism with its preference for small non-urban communities as the framework for active citizens participation in public affairs.


[1] A valley of a typical O’Neill cylinder is 3.35 by 32 kilometers, which is 107.2 million square meters. And using Wrights 4,000 square meters per family, we can calculate that a valley provides land for 26,800 families.

[2] I will discuss transportation in space colonies more deeply in another post.

[3] Using the standard dimensions of a O’Neill Cylinder (length 32 km, diameter 6.4 km), we can calculate that each valley can host 4.4 garden cities. This gives a total population of 141,000 people for each valley (4.4×32,000).

[4] Originally (in 1966) it was estimated that Columbia would have 110,000 residents in 1980.

[5] The lower estimate is based on Columbia’s current population, the higher one the estimate from note [4].