PCR TOOL 8
PCR Tool 5: Learning from the Housing Sector, describes how housing is produced in varying contexts. The use of a professional architect to design a house to a client’s individual specifications is predominantly reserved for the rich. Most who purchase housing in the formal sector have their houses designed by a developer and built by a contractor, with limited opportunity to influence the process. These two cases refer to formal urban housing. With informal housing, both urban and rural, residents may engage local builders who use traditional designs. One example of this is in East Africa, with the “Swahili House”. This generally contains a central corridor, with rooms on each side and a veranda to the front. Where housing is built by residents themselves, or built incrementally, there may be less thought about the design. Therefore, when a disaster strikes, it is usually the low-income, informal housing that suffers the greatest damage. This is mostly attributed to poor location, poor construction and maintenance, and low-quality materials. However, the design may also be to blame. For a building to be disaster resistant it must comply with certain rules related to the location of buildings, their shape, the position of openings, and their structure. Many architects and engineers know these rules, but residents and informal builders frequently do not. Hence the involvement of architects and engineers in reconstruction is important. This is no reason to advocate a topdown reconstruction process, however. Participatory design brings together residents and professionals, to ensure that both have a say in how houses are rebuilt. This not only creates safer housing, but also ensures that the people’s own resources are mobilised and contribute effectively. Participatory design enables residents to stamp their own identity on their living environment, generating greater satisfaction and ownership.
Why is design important for building back better?
A good design can dramatically change the resilience of a building to disasters. This is evident in studies of the design and production of housing in a given location, and disaster-damage assessments.
In PCR Tool 5: Learning from the housing sector, we described three predominant housing production processes and how these influence disaster performance. Informal housing was shown to perform worst as a result of underlying factors, such as poverty and vulnerability, forcing people to build on risky sites with poor materials and designs. Some of these factors may change following a disaster, for example, there may be temporarily more resources available for housing, but many factors will remain unchanged and need to be considered when designing for disasterresistance. Fortunately, informal housing does not always perform badly. Some is designed following informal traditional rules based on previous disaster experience; architects or engineers may be able to identify these features and incorporate them in reconstruction. For example, following the 2005 earthquake in Pakistan, traditional timber frames, dhajji dewari, were used based on experience of their resilience. In PCR Tool 3: Learning from Disasters, we explained how damage assessments of housing following a disaster can help define which designs and technologies are most disaster resistant. Studies of the damage caused by earthquakes have highlighted the most common design weaknesses: poor site selection,(on slopes prone to landslides or in plains that suffer from liquefaction); irregular building shapes; openings too close to corners or intersections or irregularly distributed; heavy roofs; absence of horizontal reinforcement and lintels over openings that do not protrude sideways sufficiently; poor or absent foundations; weak bonding, particularly at corners; poor structural connections; alterations to buildings; and weak first floors, very open in plan and with overhanging higher floors. Similar studies after storms point at the following common design problems: selection of exposed sites; poor foundations not anchored strongly to the ground or well connected to the walls; open verandas or large eaves the wind can get under to lift the roof; buildings partly on stilts that can be lifted; poor structural connections and lack of bracing; low roof pitches that cause the roof to be sucked off; poor fixing of roof sheets that can send them flying; openings too near to corners; openings only on the wind side, with no outlet on the opposite site; and louvred windows.
Whilst Tool 5 helps explain why and how people design houses the way they do, Tool 3 has highlighted the strengths and weaknesses of those designs. We need to take into account both of these factors in reconstruction. The simplistic approach of providing cash grants is insufficient to ensure effective reconstruction. Used alone it will frequently result in poorly resistant housing, because people: • may not understand how and why certain house-types are more resistant to disasters than others; • are unlikely to use a knowledgeable designer; • may copy what others are doing, without being able to verify whether that is good or bad; • may build hybrid houses, a mixture of modern and traditional shapes and materials, which tend to be less resistant; • may try to build on top of partially destroyed houses, perpetuating the risks. • can be over-ambitious, starting to build a house that is too big for their budget, then leave it unfinished or finish it poorly, increasing their vulnerability. These problems can be avoided through involving of architects and experienced professionals in the planning and reconstruction process. However, architects must understand local building techniques and preferences. A study of the housing sector can help them to do so. They also need to know existing local approaches to withstanding disasters which can be understood through conducting damage assessments, observations, published documents and talks with key informants.
future housing policies and strategies. But housing agencies struggled to implement it, facing the dilemma of determining : ‘whose participation in whose decisions and whose actions’? (Turner 1976) In today’s terminology, we would probably call this the dilemma of good governance. There is ample evidence now that participation and the establishment of partnerships between various stakeholders can be effective in solving deficiencies in housing and related services, whilst at the same time building the social and human assets of those involved (see e.g. Hamdi, 1995).Yet, many humanitarian agencies involved in reconstruction are still struggling with this dilemma. They tend to work in a ‘supply mode’ when providing relief, which makes it hard to shift to a ‘support mode’, when they get to reconstruction. As a result, participation is practised in current reconstruction projects and programmes, but not in the design stage. If people are to be less vulnerable to disasters in the future, they not only need more resilient houses, but also to become more resilient themselves. The process of participation helps to empower them, to build their capabilities and social networks, and to consider livelihoods issues in reconstruction, all of which are key components of vulnerability reduction. Thus, participation needs to be ensured from an early stage in the entire reconstruction process, including the design stage.
A changing role for architects
In some countries, the traditionally elitist role of the architect who mainly works for wealthy clients, is changing. The Dutch architect Johan van Lengen, working with the people of Mexico and later Brasil, described these reoriented professionals as ‘barefoot architects’ (1982); others, like Rod Hackney in the UK (1988) call them ‘community architects’. Pioneers in the USA include Michael Pyatok and Hanno Webber. In North America and Europe, community architects are assisting low-income families and homeless people to renovate derelict inner city buildings into living spaces; others work with the inhabitants of old and poor quality neighbourhoods to upgrade or renovate housing. Also in North America and in Japan, a network of community design centres has been set up, generally in lower-income urban areas, where local residents can obtain advice and information, get drawings prepared for buildings or renovations and get in touch with builders who have been vetted on the quality of their work. Rodolfo Livingston is well-known for his work in Argentina and Cuba. In the latter country, the Programme of the Architect of the Community is now well established, in which architects work with communities to develop housing designs that they or organised building brigades can use for construction. It is from such pioneer architects,
Disaster-resistant design principles
From decades of disaster damage observations and experience designing reconstruction projects and programmes, we now have a good knowledge of disaster-resistant design principles. Those applicable to small buildings such as houses are listed in the table on the opposite page for the most common disasters: storms; earthquakes; floods and landslides. A lot more detail can be found in the literature; see, e.g. Coburn et al. (1995) in the Resources section.
The importance of participation
It was recognised as early as the 1970s by authors such as John Turner (1976). that residents make most of the decisions in low-income housing. He argued that the process of producing housing is more important than the actual end product, since it builds people’s capacities and empowers them. In 1976, the first Habitat Conference in Vancouver made people’s participation a central element of
Some principles of designing for safety
Designing for wind resistance
• Select a sheltered site; avoid long and narrow (<6 m) streets; position houses in a staggered way rather than in rows; create wind breaks by planting trees, hedges etc.; Make buildings heavy, so it is more difficult for the wind to blow them away; Use a compact shape, with low walls, to present minimum obstruction to winds; a hipped roof, pitched at 30-45º, with small eaves to prevent uplift; avoid gables, as they may be pushed inwards; If a veranda is required, separate veranda frame and covering from the main roof; Tie roofing sheets well to the roof frame; flying sheets can be lethal; in the case of gci sheet roofing, provide overlaps of 2.5 corrugation, and closer spaced ‘U’ bolts along ridges and external walls; Reinforce structural connections with ‘hurricane straps’; Make solid foundations, well anchored to the ground; Provide strong structural joints and fixings, especially between walls and foundations, and walls and roof; use diagonal bracing; Give walls a rough finish to reduce wind suction; Position openings centrally and away from corners and intersections; provide openings on both sides of rooms, so that the wind can eventually pass through, rather than lift the roof; Ensure all windows can be closed; avoid louvres - if they are essential, provide storm shutters or board them up before storms.
Designing for earthquake resistance
• • • • Select a solid site; avoid landfills, flood plains and steep slopes; Make buildings light to reduce the horizontal forces caused by earthquakes; Make roofs light to avoid them pushing walls sideways and falling-in on people; Design compact buildings with a symmetrical shape and closely spaced walls in both directions. If that cannot be done, design them in separate blocks; Separate adjacent small buildings by at least 75 mm; Avoid gables, they may fall inwards; If buildings have more than one floor, opt for similar floor shapes and designs; Position the foundations on rock or firm soil, avoid stepped foundations; Provide strong joints between structural components; use a ring beam and a plinth beam where possible; use bracing at corners; If masonry walls are used, create good bond especially at corners and intersections; If concrete pillars are used, lap vertical reinforcements mid way between floors and not just above floors; Keep openings to a minimum, well distributed over the building and within walls; keep them centrally positioned, at least 60 cm away from the inside of corners and intersections and from the nearest other opening
• • •
• • • • •
• • •
Designing to cope with floods
• • • • Avoid sites close to rivers and other waterways that are known to flood; Provide for good site drainage and good waste management, as waste may block waterways; Plan for measures, such as small dams or gabions that can reduce the speed of water; Plan any new infrastructure very carefully. Some, such as road or railway embankments, may have devastating effects by re-directing flood waters; Lift buildings onto stilts or raised platforms - where the latter is used, a larger platform for a cluster of houses is preferable over single platforms, to reduce the effect of erosion; Provide deep foundations that keep buildings in place even in strong currents, eventually include a ring beam at plinth level. The minimum depth should be 600 mm in solid soils - if stones are used, select angular, not round ones; Avoid the use of soil in foundations or walls that may be reached by flood waters. These lower sections of walls should be made of more durable materials that can resist the shocks of debris floating in water; Protect organic materials such as timber and bamboo from the effects of humidity.
Designing to cope with landslides
• Avoid building on steep slopes do not make steep cuts in slopes to make space for infrastructure or housing; keep any cuts shallow, as steep cuts may become unstable; Drain slopes well, as they can become unstable and lose bearing capacity when soaked. For the same reason, avoid the use of soakaways, e.g. for sanitary systems or used household water, on slopes. Use stepped drains to reduce the speed of downward flow of water; Avoid blocking natural drainage ways with buildings or infrastructure; Avoid stepped buildings where possible; create terraces for small buildings, but avoid deep cuts and fills; keep any infill at the lower end to a minimum, and stabilise this well; Foresee retaining walls to retain the slope above terraces, and any infill at the lower end; Reduce erosion by planting appropriate vegetation on slopes.
that we can learn a lot about participatory design. Essentially, in participatory design, architects and residents jointly design a dwelling that is culturally and climatically appropriate. The architects, although giving up their traditional lead role and professional responsibility, assist and technically guide residents, CBOs and their local builders, on disaster-resistant design.
Where and when to use participatory design?
Participatory design can be used to develop plans for individual households, but this approach is not necessarily as effective in meeting the needs of large target groups. In non-disaster situations, participatory design is generally used by architects working with communities, e.g. in a street, apartment block, or an organised group of people who want to build anew or turn an existing building into houses, but the key is in striking a balance between these individual needs and plans, and the needs of the community. Often the result is the development of a number of standard house plans, from which community members then can choose. This method could be applied to reconstruction projects. It is possible, however, to incorporate a degree of flexibility in order to respond to individual needs. One solution is to jointly decide on the shape, structure and essential internal divisions of a house, but then leave it to the inhabitants to decide on some of the infills and finishes (anything that has no essential structural role, including resisting disasters). The idea of separating “supports” (structure) from “infill” (internal completion) in housing was first suggested by the Dutch architect N. John Habraken (1972), as a way of giving inhabitants a meaningful participatory role in design. Another option is to allow a certain degree of modification of standard house plans, through collaboration of a designer with individual households. The advance of computer-aided design (CAD) has now made this a lot easier, as illustrated by Practical Action’s post-tsunami reconstruction in Sri Lanka. CAD systems, however, are still relatively expensive. Wikipedia provides a comparison of different types of software; see the Resources section. Participatory design can also be used to design other community buildings, such as community centres, schools; health centres; markets or commercial areas; workshops; communal water and sanitation blocks, etc. This is closely linked with participatory planning (see PCR Tool 7: Planning with the People), and is best done immediately after, or even during, the planning process. Whilst planning is mainly concerned with settlement layout, the provision of infrastructure, and the position of house plots, participatory design will
Participatory design in Bangladesh
then specify how these plots are to be filled in. The discussion with communities on design needs to address house plans as well as specifications for materials and components. The design group will also need to think about how the houses will be built - by themselves, by local builders or others - as the availability and skills of those builders determine which construction technologies are feasible. Furthermore, they will need to consider whether materials can be salvaged, produced and supplied locally, or have to come from elsewhere. It is important to consider at this stage that using local materials and builders can also serve to rebuild local livelihoods. Design with communities has limitations. It may be difficult: if a group of disaster-affected people is not really a community, e.g. if they are households prioritised from an official waiting list where it may then be very time consuming for them to build relations and generate sufficient trust for a communal design process to work. Similarly, it can at times be difficult in urban contexts, particularly if these contain much more heterogeneous categories of people. There can also be problems if there is a mixture of tenants and owners and tenure issues cannot be easily resolved or landlords suppress tenants’ right to involvement. These issues may be resolved by dividing such groups into more homogeneous smaller groups for which individual plans can be designed.
Who needs to be involved?
A participatory development process tends to involve different parties that can be divided into the community to be housed and outsiders who
photo © IFRC
support the development process. In the case of participatory design, the latter will always include an architect or otherwise suitably qualified designer. This person may be supported by an engineer who can advise on making buildings disaster resistant and on services required, such as water. The outsiders may also include a representative of the agency funding the building(s), and possibly a representative of the local authority. One ‘outsider’ people should facilitate the design process, and ensure every participant can have a say. It is important not to have too many outsiders, however, as they may dominate discussions. On behalf of the community, it is important that all sections are represented, but it is not a requirement that every household takes part. It becomes difficult to run a design workshop if there are more than 30 participants, and it is often possible, in relatively homogeneous communities, to develop a series of standard house plans with the participation of less. It is preferable for a community to select its own representatives, unless there are good reasons not to, e.g. if the selection process could be biased. There needs to be sufficient women representation as women’s design priorities often differ to those of men. It is also useful to have some older participants who may know more about the history, culture and tradition of housing in the location, and have a memory of previous disasters as well as young participants who have more formal education. Local builders have valuable knowledge of vernacular technologies and designs. It is important to remember that, following a disaster, the proportion of households with missing, ill or disabled members can be much higher than usual; those households may have special design or construction requirements, and therefore need proper representation. If there are a number of very different sections in a target group – a heterogeneous community a participatory design process can become very difficult, as it may be hard to reach agreement. One solution for such cases is to divide the target group into smaller sub-groups and have a participatory design exercise with each of those. Sometimes, communities decide themselves to form smaller groups; this quite often happens when they decide to organise the construction of clusters of houses themselves, in mutual aid. Such groups tend to accommodate a limited number of households, frequently between about 10 and 30, as the management of larger groups in mutually aided construction is quite complex. In such cases, it may also be preferable to have a participatory design exercise with each of those groups.
The participatory design process
A participatory process for designing houses in a reconstruction programme needs a good facilitator.
This is not necessarily the architect or engineer, but could be another member of a support agency with the right skills. The process may involve the following steps: 1. Determine design requirements. Organise meetings with small groups of people and ask them how they design and build their houses, and use the space in them. If, in a separate needs assessment (see PCR Tool 4, Assessment of Reconstruction Needs and Resources), people have listed economic activities, seasonal calendars, or daily activities, it is useful to bring these into the discussion too, as these can highlight livelihood and other activities that may have to be accommodated in or around the house. Distinguish the activities and needs of men and women. 2. Discuss disaster resistance. Ask the groups what they normally do to protect their houses from natural hazards. Have some houses received less damage than others during the recent disaster? Why do people think this is so? Can some of the local designs and construction technologies be retained for reconstruction? Would they have to be improved? (see PCR Tool 3, Learning from Disasters). 3. People produce sketches. Ask the groups to produce some outlines of how their houses might look like using pens and papers (or even lines in the soil, they could be photographed to retain the sketches). Ask them to think about what they would like to retain from their traditional house types. Aim to get approximate floor plans for a ground floor and any upper storeys (if there is a need for those). If some people can draw well, they could be asked to produce elevations too. It can be helpful to give names to the rooms, and write in them what activities take place there. The facilitator should keep in mind that producing a vast number of drawings would create a lot of work and time; experience tells us that five or six type designs are sufficient for most locations, and that people rarely select more than three out of those. This may mean that compromises have to be reached between groups over some details. 4. Groups produce models. Groups use the sketches to produce housing models, using wood, hardboard, plastic, canvas etc. Models of individual houses should not be smaller than one twentieth of the real size. People can make some changes to the sketch plans on the models, as the purpose of the modelling is to create a clearer picture and encourage further discussion. Any changes made on the models should be marked on the sketches too. The facilitator should remind people to include measures to mitigate disasters in the models, as well as the sketches, as discussed in step 2; if there are no appropriate local disaster-resistant
Let us know which of the options below best describes you and we'll direct you to the most relevant content.
Practical Action uses technology to challenge poverty, working with poor women and men around the world.
Explore our work by Country
Explore our work by Technology
+44 (0)1926 634400 email@example.com
© Practical Action