The following is a study of a hypothetical water farming infrastructure for the arid city of Luanda, Angola; using fog harvesting nets with varying capabilities.

Luanda, the fastest growing city in the world, is desperately short of clean water. Only one in six Luandan households has running water, forcing most of the inhabitants of the musseques (the vast slums that constitute the majority of Luanda’s land area) to depend on contaminated water brought by truck from rivers hours north and south of the city. The price of water in the musseques can be as high as 12 cents a gallon, a huge burden on a populace which lives on an average of $2 per person per day. In 2006, the worst African cholera epidemic in a decade devastated the musseques, killing 1600, spread by contaminated drinking water as well as contact with sewage.

What if water, already inextricable from agricultural farming processes, was itself farmed? Beyond the direct benefits a renewable source of fresh, clean water would provide Luanda, farming water seeds the city with potential. By establishing an infrastructure to effect the farming of water, one may farm landscapes, societies, production: a city.

Farming water in parched terrain seems paradoxical, but an established process for doing so already exists: fog farming. Like most forms of farming, fog farming can be seen as a special instance of infrastructure and, as an infrastructure, it has both direct and indirect effects: it is directly responsible for condensing and collecting airborne moisture; but it also has many indirect effects on the growth and health of the urban system. The fundamental inquiry of our project is a unique treatment of infrastructure, considering not just how its programmed function can enable positive development of some urban condition, but how the thing in itself, the system which enables that function, can have other, non-function-related (indirect) effects — and how they can be understood and utilized.

Before exploring those indirect effects, though, a brief description of the techniques and requirements for fog farming may be helpful. This technology has been explored for several decades as a means to obtain potable water in arid environments, though until now it has been confined to rural locales, owing to problems with airborne particulate pollution. A typical apparatus consists of a nylon or polypropylene mesh, at least a meter or two square, stretched across a metal or plastic frame, with condensed moisture dripping down the mesh into a collection pipe at the bottom of the net. The basic necessary conditions for the deployment of such an apparatus are an arid environment and strong fog. Such fogs are found primarily along particular ocean coasts (the Pacific coasts of Chile and Peru, the Atlantic Coasts of Namibia and Angola, or the Indian coast of the Arabian peninsula) where certain ocean currents produce atmospheric moisture which is then confined and concentrated by mountains.

Luanda is by far the largest city in the world in which fog farming is practicable, as it is projected to jump from a current population of four million to over eight million by 2025. Though Luanda experiences something of a rainy season between March and April, for most of the year precipitation is nearly absent — averaging 0 millimeters/month from June to August — as the Benguela and Angola currents of the Atlantic Ocean combine to prevent the humid air from condensing into rain. This combination of humidity, aridity, and explosive population growth creates a situation uniquely suited to an experiment in urban fog farming.

Fog farming has the direct potential to meet this need for potable water; but an investigation of this form of farming can also reveal important things about the role of infrastructures in urbanism, if we consider its indirect effects. We believe that this move — the equal valuation of indirect effects and direct effects — is essential to the development of an alternative modernist urbanism, for several reasons:

A We argue (like many others have, particularly in the past decade or two) that the modernist planning movement has run its course; has been exposed as insufficiently flexible to react to the flux that characterizes a vital urban condition. This is because modernist urbanism was concerned primarily with controlling the city, which inevitably produced conflict between the dynamic nature of the urban system and the static nature of the designer’s reality.

B We also take as a starting point the inadequacy of some alternate proposals, particularly those such as new urbanism and parametricism, that propose an alternate set of forms to modernism while failing to account for the fact that it was the overreliance on formalism itself — on the controlled expression of the rationality of the designer — that made modernist planning unable to cope with the reality of the urban condition.

C We suggest that infrastructure is an appropriate object of design for the urbanist, the architect, and the landscape architect, as infrastructure can be embedded with some characteristics that provide definition (a means for the urbanist to have influence on the direction of change), as well as characteristics that permit appropriation by inhabitants of the urban system. In other words, an infrastructure can withstand appropriation while remaining coherent as an intervention.

Fog farming, then, can only be fully understood as an urban intervention if it is understood as performing ‘infrastructurally’ — as having effects on the urban system that extend far beyond the direct or spatial, to the alteration of streams and flows — liquid, capital, human, traffic, and so on. Thus we have sought not merely to provide a form of farming that meets a need for water, but also to understand how, in placing an infrastructure that would meet that need, we might also alter the streams and flows of the urban system for the better.

Consequently, the infrastructure we have developed is unified by its fundamental nature as a system for farming airborne moisture, but diversified in architecture, deployment, funding, distribution, and ownership, as the modification of each of those characteristics affects its interaction with the city.

Furthermore, we remain extremely sensitive to the fact that one of difficulties of working (as a designer) in a city like Luanda is that the city government cannot always be relied upon to act in the interests of the poorest inhabitants. Indeed, even we — seeking as designers to act in ways that we perceive as beneficial those who live in the musseques — are likely (if not certain) to be mistaken at times about what will benefit the musseques. And so we have varied the degree of control embedded in the infrastructure, allowing it to be more rigid in the wealthier portions of the city (where citizens have access to the levers of government) and relinquishing nearly all control in the poorest regions. Some functions, which we hope will remain uncontroversially beneficial, are embedded in the nature of the infrastructure we make available (most essentially, that it cultivates useable water from fog).

To these ends, we propose four basic forms for fog farming in Luanda (though the boundaries between forms would exist more as gradients than as hard delineations):

1 In the central districts of the city, particularly Ingombotas, a highly formal farming architecture appears, spanning the gaps between buildings and hanging off of scaffolding over streets, providing water treated with nutrients for the benefit of a system of tubes growing algae, which is in turn harvested at processing plants to make biodiesel. The profits from an industry with a fixed end-date (the oil industry) can be invested to build an infrastructure that will meet the energy needs of Luanda far past that end-date. The distribution of this water is centralized; it is a striated, non-rhizomatic network. It would be government funded, with the intention that economies of scale (resulting from the erection of this network) would drive down the cost for building future fog networks in lower income areas of the city.

2 Middle class suburbs, which ring the central districts, would also receive investment from the government and see an interface arise between the municipal water system and the fog farming infrastructure, though some of the water supply generated there would be diverted to growing personal gardens and other vegetative uses.

3 In the musseques, a combination of NGOs and microlending practices would fund the development of an infrastructure, while the deployment would be at the whim of the citizens, who stand to benefit from the income generated by new sources of water. Appropriation by the community would be nearly total, while the benefits of the infrastructure would accrue entirely to the local communities. Distribution of the water would be completely local and decentralized, disconnected from the municipal supply and its filtration. Because of this, we propose nets treated with a slowly-released store of chlorine whenever affordable, with untreated nets supplementing the supply.

4 On the arid periphery of the city, fog farming would serve to seed the future of the city. Though the practice is deplorable, it is unlikely that the government will stop clearing informal settlements in the center of the city, so the periphery will likely continue to swell with a constant mixture of slum dwellers expelled from their homes and new arrivals from the hinterlands of Angola. Expanses farmed for fog would be gradually impregnated with nutrient-rich water, cutting down on dust storms, prompting the growth of farms and eventually settlements based around this infrastructure. Through farming water, we farm landscape, farm city, farm future communities.

a statistical index of luandan fog farming:

83% the percentage of Luandan households without running water

12 cents the price per gallon of water in the musseques of Luanda

0 the average monthly precipitation in Luanda from June to August

6% the current yearly average growth of Luanda

4.8% the current average growth of the next fastest growing major city in the world, Kinshasa

4,000,000 the population of Luanda in 2007

8,200,000 the project population of Luanda in 2025

12.9 L the projected average amount of water collected by a square meter of fog net in one day

244 the number of days it would take to recoup the cost, including both materials and installation, of a square meter of fog net at the current price per gallon of water in the musseques

1,200,000 the amount of water trucked into Luanda every day from the Bengo and Cuanza rivers

$35,152,969 the estimated cost of replacing the volume of water brought using trucks from the river with water from fog farming

$100 the estimated cost of a square meter of fog net, including materials and installation

2/3 the number of Angolans living on less than $2/day

450 the number of tanker trucks that travel from the rivers to Luanda every day

1,600 the number of Luandans killed by a major cholera epidemic, due to contaminated water, in 2006