SHELTER PROJECTS EAST & HORN OF AFRICA: 14 Case Studies
CASE STUDIES OF HUMANITARIAN SHELTER AND SETTLEMENT RESPONSES RELEVANT TO EAST AFRICA
VERSION 1 -- 18 June 2018
Global Shelter Cluster ShelterCluster.org Coordinating Humanitarian Shelter Shelter Projects - East & Horn of Africa: 14 Case Studies
Released in June 2018 by the International Organization for Migration (IOM), on behalf of the Global Shelter Cluster.
All case studies are available online from www.shelterprojects.org
Copyright for this book is retained by IFRC, IOM, UNHCR, and UNHABITAT. Reproduction for non-profitable objectives is encouraged.
The copyright for the photographs and images remains with the photographers or entities whose names appear on each picture or in the caption. The Global Shelter Cluster and its members may use the pictures, if appropriately credited.
This booklet has been compiled and printed thanks to the support of USAID/OFDA, DRC and IOM.
Case studies in the Shelter Projects publication have been provided by organizations and individuals listed online at: http://www.shelterprojects.org/partners.html
DISCLAIMER
The maps contained in this publication are for illustrative purposes only and should not be considered authoritative. Whilst every effort has been made to ensure the accuracy and completeness of the content of this booklet, no liability can be accepted for any errors or omissions contained within it.
The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the Global Shelter Cluster concerning the legal status of any country, territory, city or area, or of its authorities, or concerning delimitation of its frontiers or boundaries, or regarding its economic system or degree of development.
Approximate prices are given in US Dollars (USD), based on exchange rates around the time of the project.
Copyright for front cover photo: © Rikka Tupaz, IOM South Sudan.
Copyright for back cover photos (top to bottom, left to right): © UNHCR, © IFRC, © World Vision Somalia, © Jama Yasin Ibra- him. TABLE OF CONTENTS
1 / A.7 / Democratic Republic of Congo / 2002 / Volcano 1 2 / A.9 / Ethiopia / 2012 / Conflict and drought 4 3 / A.8 / Ethiopia / 2011 / Sudanese conflict 8 4 / A.4 / KENYA / 2007-2008 / Election violence 11 5 / A.3 / KENYA ( Dadaab) / 2007 / Flooding 14 6 / A.22 / SOMALIA / 2011-2013 / Drought and conflict 17 7 / A.28 / SOMALIA / 2011 / Famine and conflict 22 8 / A.15 / SOMALIA / 2009 / Conflict 26 9 / A.8 / SOMALIA / 2007 / Civil conflict 29 10 / A.25 / SOUTH SUDAN / 2014-2016 / Complex 32 11 / A.29 / SOUTH SUDAN / 2011 / Conflict 37 12 / A.9 / SUDAN (Darfur) / 2004 / Conflict 40 13 / D.9 / SUDAN / 1985 / Conflict 43 14 / B.25 / UGANDA / 2007 / Slow onset floods 46 © Muse Mohammed © Chiara Jasna Vaccaro * Note that the original case study codes from past Shelter South Sudanese refugees in Tierkidi camp in Gambella, Ethiopia, plastering Projects editions are retained on each case study. the walls of the shelters with mud sourced locally by themselves, in areas agreed with authorities and host communities.
INTRODUCTION
This booklet is a compilation of case studies of humanitarian The publication is intended to support learning by highlighting shelter responses relevant to the East Africa region, compiled the strengths, weaknesses and some of the lessons that from the six past editions of the interagency publication Shelter can be learned from different projects, which try to maximize Projects. The series of publications, initially led by IFRC, UN- emergency funds to safeguard the health, security and dignity HCR and UN-Habitat, is now a Global Shelter Cluster product of affected people, whilst – wherever possible – supporting and includes contributions from over 300 shelter practitioners longer-term shelter needs and sustainable recovery. from across the world, from over 50 organizations and over The target audience is humanitarian managers and shelter 70 countries, including host governments’ shelter responses. programme staff from local, national and international The projects described in the case studies and overviews organizations at all levels of experience. Shelter Projects is contained in this booklet represent responses to conflict, also a useful resource for advocacy purposes, showcasing natural disasters and complex crises, demonstrating some of the work done by the sector, as well as for research and the implementation and response options available within the capacity-building activities. East African context. These include urban site upgrades (see All case studies and overviews contained in this booklet, as n. 7 and 8), distribution of shelter meterials/kits and NFIs (see well as from all editions of Shelter Projects, can be found n. 4 and 14), shelter construction (see n. 1, 2 and 3), cash-for- online at: work, technical assistance, site planning, resettlement and multi-sectoral programmes (see 6 and 9). www.shelterprojects.org © Muse Mohammed
Bentiu Protection of Civilians site in South Sudan. This site for displaced people fleeing conflict has grown to over 120,000 people at its peak and has been upgrad- ed over two years to improve living conditions, reduce health and safety hazards, and address the high increase in population. 1 / A.7 Excerpt from: Shelter Projects 2011-2012 Volcano
A.7 Democratic Republic of Congo – 2002 – Volcano / 1 Keywords: Returns, Urban neighbourhoods, Household NFIs, Construction materials, Transitional Update: shelter / T-shelter, Community engagement, Mass communication.
Country: Project timeline Democratic Republic of Congo Project location: Goma 10 years – ––Long term assess- Disaster: ment of impacts. Goma volcano eruption in 2002 No. of houses damaged: 15,000 houses destroyed (20 per cent of Goma’s housing stock) ––Periodic monitoring Number of people displaced: 300,000 people displaced 9 months – Project outputs: ––5000 families as- 5,000 families supported with sisted with transi- shelter and latrine packages tional shelter Shelter cost: US$ 250 average cost: Shelter 3 months – Goma and latrine (materials and ––First transitional shelters erected labour) Democratic Republic of Congo January 2002 – ––Volcano erupts
Project description This case study summarises an assessment by a major donor of the transitional shelter and recovery programming that it funded in Goma following the volcanic eruption in 2002. The assessment was conducted ten years after the initial response. The assessment found that transitional shelter did help to facilitate the transition to permanent housing, and became a base for many livelihood activities. It also found lasting impacts from both the settlements approach taken and from the supporting activities to help people in Goma to “live with risk”.
Strengths and weaknesses + Transitional shelter really can facilitate the transition to and concentrated the economic benefits associated with permanent housing. As intended, nearly all of the original investments made by the humanitarian community. In turn, 5,000 "t-shelters" have been improved in some way as part residents have had better access to jobs and public services of making it a permanent home. A site visit 2012 noted that in an urban context than in a remote camp, contributing most beneficiary families continue to live in their transformed further to the recovery of their city. Disaster Risk Reduction transitional shelters. After ten years, some families are still measures were incorporated into the reconstruction of making improvements leading to permanence, suggesting road and service networks, to enhance both evacuation that the process of incremental housing development is options as well as access to land and housing markets. both evident and likely to continue in the foreseeable future. The city-focused approach orientated humanitarian + Transitional shelters have become “shophouses”. As assistance towards settlement planning and also reflected intended, many project beneficiaries have expanded their beneficiaries' wishes to return to their own neighbourhoods. shelters to create space for livelihood activities of all kinds, thereby either restoring livelihoods lost in the disaster, or - In Shelter Projects 2008, the implementing organisation creating new economic activity using the shelter as a much- noted: needed platform for production. This has contributed to - For families with eight or more people, shelters were both community and regional economic recovery since the initially not big enough. volcanic eruption. - Some people felt that plastic walls compromised their + A deliberate focus on “Shelter and Settlements” is a privacy and security. critically needed approach to humanitarian assistance in urban areas. Longer-term recovery was dependent upon • The project was one of the first-ever attempts by the donor regenerating its urban economy. Providing transitional shelter to promote an explicit shelter and settlements approach to in the city, based on the “city-focused” approach, maximised shelter activities.
1 SHELTER PROJECTS EAST AND HORN OF AFRICA Democratic Republic of Congo Excerpt from: Shelter Projects 2011-2012 1 / A.7
Left: Recovery work began as lava cooled. Right: Road work was linked to evacuation planning. Photos: USAID/OFDA
The volcano 9 seismologists declared that the option aimed at rehabilitating Goma (See case study A.1 in Shelter eruption was over. itself, allowing as many people as Projects 2008) Since early 2002, Goma has possible to remain. These options Nyiragongo, a volcano located subsided by nearly 50cm. Minor sub- were discussed at length among rep- approximately 16 kilometers (ten sidences have periodically occurred resentatives of all key stakeholders. miles) north of Goma, the major town as a result of on-going tectonic activity. The perceived and real security and in Eastern Democratic Republic of political conditions in the immediate Congo (DRC), began erupting on 17th Response region affected decisions in shelter January 2002. Lava flowed from the With thousands of jobs lost, assistance by constraining relocation southern flank of the volcano, heading and the urban and regional economy options to the east, north, and west of towards Goma. devastated, national and interna- Goma. The city is also located on the This eruptive activity triggered an tional organisations mounted a rapid northern shore of Lake Kivu, making exodus of Goma, a city of approxi- response, with the international large-scale southern movement of the mately 450,000 people. Of which an community contributing a total of US$ displaced impractical. estimated 300,000 people fled briefly 40 million in assistance. There was also the local security to Rwanda, while others fled to set- In this case-study, the donor’s consideration that many people tlements to the west of Goma as well share of the contribution was nearly wished to remain close to their former as elsewhere within the DRC. Most US$ 5 million. This included US$ 2.6 houses to prevent appropriation or people returned to the city within three million in emergency relief: water, looting. months. food, health, and non-food assistance Following consultations with The lava flows and subsequent (including blankets, household goods, affected communities and authorities, fires caused severe damage in Goma. and plastic sheeting); and a US$ 2.3 the donor devised a two-pronged An estimated 13 per cent of the city’s million programme featuring a transi- strategy that would bring new life 35km2 land area was covered by lava. tional shelter project and disaster risk to Goma and reduce the impacts of It heavily inundated the central part reduction activities. future disasters. of the city, destroying up to 15,000 The response featured the dwellings (20 per cent of the city’s design and implementation of one of Transitional Shelter estimated housing stock). In inun- the donor’s first transitional shelter Due to the security, safety and dating the most developed portion of projects. economic concerns of the affected the city, arguably the most developed population, the first element of the portion of eastern DRC, the lava flows Recovery programme was to support a city-fo- destroyed numerous economic enter- After critical needs had been cused transitional shelter program, prises and community structures, and addressed, the humanitarian devoting 80-85 per cent of program thus thousands of livelihoods. community began to develop strat- funds to the Goma urban area. The An estimated 90,000-105,000 egies for helping residents of Goma donor and its partners determined that people, many of whom were already rebuild their lives and livelihoods. there was sufficient space in Goma to vulnerable because of conflict-in- Shelter quickly emerged as the most shelter residents there, and that the duced insecurity and limited economic pressing need for affected families. existing social and economic infra- opportunities, lost their homes and People displaced by the volcano structure, even post-eruption, made other assets, and were in need of needed a place to call “home”. it easier to assist people in the city shelter. Options for meeting this need rather than elsewhere. Although eruptive activity ended included moving the entire city to a The donor funded a single organ- within 24 hours, seismic activity new site, dispersing people to different isation to provide shelter in Goma to related to the volcano continued until regions of the country, moving people assist 5,000 households. All of the early February 2002. On February into camps, and a “city-focused” households were assisted within nine
SHELTER PROJECTS EAST AND HORN OF AFRICA 2 1 / A.7 Volcano
“The central business district, buried under rock, is re-emerging; there is even a new Volcano Internet Café on the edge of the destruction. The camps set up for displaced residents are now mostly shuttered, and Goma is experiencing something of a housing boom.”
The New York Times (emphasis Transitional shelter (left) has evolved into permanent housing (right) for added), November 10, 2002 thousands of families. Photo: USAID/OFDA
months of the eruption. Other donors officials and representatives. This was response, and the explicit shelter and saw the efficacy of this strategy and intended to increase the number of settlements approach adopted were provided a combined total of 8,000 evacuation routes. aimed at strengthening the resilience additional households with transitional of Goma’s citizens by promoting shelter. A further 2,000 households Outcomes recovery and lessening the impact of received other assistance from a Despite the considerable changes future disasters. variety of other organisations. in Goma during the 2002-2012 period, The 2012 assessment by this Assistance was used to expand including recent conflict in and near donor found that the activities that it or supplement host family homes, or the city, several outcomes of the supported have contributed to a tran- build on under-used or vacant private donor-supported post-eruption activi- sition to recovery and reconstruction. residential parcels of land. Shelter ties have become visible over time: This outcome is notable, for it demon- supplies were sufficient to create 21m2 strated the utility of using shelter of covered living space for an average • In addition to providing much- as a means of promoting economic beneficiary household of up to six needed shelter, the city-focused recovery and linking humanitarian people. The supplies included plastic programme had a significant community shelter activities to the impact on Goma's economy. sheeting, zinc roof sheeting, wood process of longer-term permanent Beneficiary families supported framing, and concrete screed flooring. housing development. Furthermore, nearly 45,000 person-days of A modest latrine was also provided. labour to transform their transitional shelter activity was deliberately con- Three-quarters of households shelters into permanent homes. centrated in neighbourhoods, where were assisted on land occupied by This generated nearly 3,600 new people wanted to resume their lives host families (relatives or friends); jobs, and helped to jump-start and livelihoods. It enabled people to many of these beneficiaries have economic recovery in Goma. learn to live with risk, supporting them remained on hosted land. • Volcano monitoring is ongoing, with risk reduction activities. with most of the equipment Living with risk provided still functional, though upgrades are needed. The second element of the • The Goma Volcano Observatory strategy was rooted in the basic continues to operate many message of learning to live with risk: a community-based education Disaster Risk Reduction program me activities, although updating was designed to promote improve- is required. Activities include ments in volcano hazard monitoring providing volcano activity (provision of equipment, staff support, reports to radio stations, sharing and technical assistance to the Goma information at a local volcano information center, and updating Volcano Observatory). alert levels in public areas. The donor also sponsored a • Over time, nearly all beneficiary two-year, community-based Disaster families transformed their Risk Reduction program linked to transitional shelter into permanent the Goma Volcano Observatory to housing, resulting in the re- enhance early warning systems, establishment of local markets upgrade evacuation routes, and and communities, contributing to improve community awareness of overall recovery. what to do and where to go when eruptions and earthquakes occur. The rapid response to the 2002 The road network was expanded volcanic eruption, the incorporation following discussions with local of Disaster Risk Reduction into the
3 www.shelterprojects.org SHELTER PROJECTS EAST AND HORN OF AFRICA Ethiopia Excerpt from: Shelter Projects 2011–2012 2 / A.9
2 / A.9 Ethiopia – 2012 – Conflict and Drought Keywords: Planned and managed camps / relocation sites, Transitional shelter / T-shelter, Case Study: Site planning
Country: Project timeline Ethiopia Project location: Dollo Ado 18 months – ––Construction Conflict / Disaster: ongoing Conflict and drought in Somalia 18 months – ––Transitional shelter strategy reviewed Conflict date: Conflict since 1992 Number of people displaced: Over 1 million registered Somali refugees By the end of 2012, 177,000 14 months – ––First transitional refugees were registered in the shelter strategy five Dollo Ado refugee camps Project target population: 9,000 families (2011-2012) 13 months – ––Shelter prototype Project outputs: evaluation 7 months – ––Criteria for Tran- 7,127 shelters by end of 2012 sitional Shelter Occupancy rate on handover: adopted High Shelter size: 6m x 3,5m (21m2) Ethiopia Materials cost per shelter: February US$ 525 including transport 2010 – ––First camps opened at Dollo Ado Project cost per shelter: Dollo-Ado US$ 800 excluding overheads
1992 – ––Conflict starts
Project description Four organisations built semi-permanent shelters for Somali refugees living in the camps at Dollo Ado. Each organisation set up production lines in the camps to prefabricate the components. The projects worked within the constraints of challenging logistics and very different social environments between camps. The shelter design was selected following a consultative process during which different options were shared with camp residents.
Strengths and weaknesses + The design process was coordinated between was not possible to implement. Negotiations with the organisations to avoid conflict between refugees over host communities over the use of mud slowed progress. different shelter standards. + The process to select the shelter design was designed to • Different organisations have had very different completion be transparent and include all stakeholders, including camp rates as a result of different budgets, management residents. structures, logistics, supply and relations with camp + Shelter construction provided refugees and the host residents and host communities. community with paid work. It is estimated that the shelter projects contributed US$ 16,000 per month to the economy • Most materials were not available in Dollo Ado markets of each camp. and were imported from other regions. Price fluctuations led - Joint procurement of supply of materials was attempted to a 16 per cent increase in the total shelter cost. but did not prove successful. • The shelter strategy was developed based on the - Significantly fewer shelters have been built than initially assumption that it would acheive 100 per cent shelter anticipated. The strategy supported less than 20 per cent of coverage. Production and delivery remains short of these the population of the camps by the end of 2012. targets. - The original design used mud render, but this required a significant amount of water and transportation, and
SHELTER PROJECTS EAST AND HORN OF AFRICA 4 2 / A.9 Conflict and Drought
Completed shelters in Buramino camp at Dollo Ado. It was initially intended that the shelters would be plastered in mud but this proved difficult to implement. Photo: Joseph Ashmore
Camps at Dollo Ado two camps. Camps were established were staffed by a mixture of skilled within host communities or in isolated carpenters and daily labourers. Following a resurgence of the locations. Some had been estab- One organisation found particu- conflict and drought in Somalia, a lished for months whilst others had lar challenges with the splitting of series of five camps were established existed for years. As a result each site bamboo, facing a 50 per cent shortfall in 2010 and 2011 within 100 km of the presented very different challenges. at the time that the bamboo had to be Ethiopian Border town of Dollo Ado. There was some variation in fixed. It turned out that this was due By the end of 2011, the five beneficiary selection: One organisa- to many bamboo poles being split into camps of Bokolmayo, Melkadida, tion targeted blocks in each camp two pieces by the daily labourers as Kobe, Hilaweyn and Buramino hosted according to agreed criteria. Within opposed to four or six. 34,000 Somali families, the largest each block the most vulnerable refugee presence in Ethiopia. The households were prioritized, and all Training and supervision refugee population increased during shelters in each block were completed Training was provided for skilled 2012, and by the end of the year, before moving on to the next. Another labourers who were responsible for 180,389 individuals refugees were organisation prioritized individual the on-site construction. On-site works registered in the camps. As these households across the camp rather included digging holes for foundations, camps became more established and than prioritising individual blocks. erecting the frames, fitting the roofing, the numbers of registered refugees In addition to building shelters for covering the walls with bamboo continued to increase, it seemed likely camp residents, 120 shelters were slats and fixing windows and doors. that the camps would remain open for built for host community households Training in mud rendering for walls some years. in 2012. Additional targets were set for was given where mud was available. Being close to the equator and 2013. All organisations directly hired at low altitude, Dollo Ado is subject both skilled staff and daily labourers. to harsh weather conditions with Implementation To select carpenters for on-site works, high temperatures, strong winds and Each implementing organisation candidates were asked questions on seasonal heavy rains. started at different times with different minimum foundation depths and how The people living in the camps total budgets and in different sites. to best nail a joint. They were then mainly come from rural parts of The most effective projects estab- assigned one shelter to prove their Somalia. A significant proportion are lished strong supply routes, prefabri- skills. Staff monitored the construction. nomadic pastoralists, accustomed to cation facilities and clear procedures The ground at the different sites lightweight and movable shelters. for managing supply and construction. varied. In some sites it was relatively In 2011, shelter was identified as straightforward to dig 60cm deep holes an urgent need in the refugee camps. Workshops by hand, in other sites the ground was The number of new arrivals peaked Each organisation established hard and concrete was required in the with an average of 168 persons per a workshop and materials storage foundations. day in June 2011. They were provided area close to construction sites. In with tents. However, the life span of the workshops, timber was precut, Shelter selection the tents proved to be around 6 to bamboo was split, and doors, windows In September 2011, the organi- 8 months, menaing that alternative and roof trusses were prefabricated. sations agreed to develop common solutions for the shelter in the camps A well-organised workshop with shelter standards and build shelter were required. effective quality-control mechanisms prototypes for review. was necessary to maximise production At this time, the three organ- efficiency. The minimum workshop and isations involved in the shelter Beneficiary selection storage area for efficient production programme were invited to produce Shelters were built by four organi- was 1 hectare (10,000m2). Workshops prototypes based on the shelters that sations and each was allocated one or they had been building. Each of the
5 SHELTER PROJECTS EAST AND HORN OF AFRICA Ethiopia Excerpt from: Shelter Projects 2011–2012 2 / A.9
Three different prototype shelters were built and a group of refugee representatives, the government and the key organisa- tions agreed on a common design Image: NRC three shelters was built to the same was adapted using closer spacings Materials list design brief. between the bamboo strips for walls Materials Quantity Each shelter was evaluated by a instead of rendering it with mud. Corrugated galvanised 24 gender balanced group of refugee For bamboo-only walls, plastic iron sheet sheets roofing pieces representatives, the government and sheets or fabric were often placed (2m x 0,90m) the key organisations. inside to provide protection from Eucalyptus poles (8cm 32 The model selected had a corru- the rain and the wind. The resulting diameter) pieces gated iron sheet roof, a eucalyptus structure was relatively well venti- Bamboo (6cm diamter, 62 post-and-beam structure and split lated in the hot climate and provided min. 6.5m, dry, straight) pieces bamboo wall cladding. The intention adequate protection from the rain. was to plaster the walls with mud. Households were relatively satisfied Nails (#9, #8, #6, #4) 10.5kg The shelter had an internal with these shelters. Roofing nails 3kg partition, two lockable windows, and Tighter construction quality a door that could be locked both from controls were required for bam- Metal straps (2cm wide; 10m 1.5-2mm thick) the inside and the outside for improved boo-only walls to ensure that no large security. Corrugated iron sheet was gaps were left between the lathes. Wire mesh (1.8m x 30m; 1 piece chosen for roofing on account of its The use of local wood for shelter 2cm opening) durability and fire safety. construction was a major issue for Hinges (T hinge 4 cm long 6pieces the host population, and as a result, sides) Mud plastering timber was brought in to the area. Lock system 4 pieces The shelter was originally designed However, each camp resident burned to have wattle and daub walls using a significantly greater volume of wood Black wire (10 kg rolls) 0.1roll. local mud. Bamboo laths would be when cooking than each shelter used covered with chicken wire and the in its construction. shelter would be rendered with mud. Workshop tools This was initially considered to be Materials Quantity a low cost and sustainable walling Logistics and supply Electric Radial arm saw 2 pieces solution. Although highland Ethiopia has Unfortunately, mud of suitable significant plantations and production Hammer 5 pieces quality was only available from certain of both eucalyptus timber and bamboo, Tape measurer 4 pieces locations in river beds and these were the nearest eucalyptus and bamboo Cutting table 2 pieces owned by the host communities. Each plantations are at least a day’s drive Assembling table 3 pieces shelter required slightly more than from Dollo Ado. The suppliers who can Oil barrel for treating 1 piece 2m3 of soil to render it with mud, as produce the paper work required for timber well as a significant volume of water. large procurements are further away, This worked out at over 2,000 truck mostly based in Addis Abbaba. loads for the 10,000 shelters that were The transport requirements proved planned in the first year. Up to 1m3 of demanding. One truck only carried mud would additionally be required enough materials for 15 shelters. each year for repairs after the rains. Building 10,000 shelters would require By the end of 2012 there were over 600 trucks. sufficient resources available to Over the year, the biggest cost implement mud walling for 60 per cent increases were with bamboo and of the shelters in Dollo Ado, and the transportation. This led to a 16 per decision was made to discontinue the cent increase in the cost of a shelter. use of mud in the future. The design
SHELTER PROJECTS EAST AND HORN OF AFRICA 6 Top: View of a camp before construction. 7 Each organisation established a large workshop in each camp to SHELTERstore materi PROJECTS als and EAST prefabricate AND HORN OFcomponents. AFRICA Images: Joseph Ashmore Ethiopia Excerpt from: Shelter Projects 2011–2012 3 / A.8
A.8 Ethiopia – 2011 – Sudanese Conflict / 3 Keywords: Planned and managed camps, T-shelter, Site planning, Training, Tools, Construction Case Study: materials, Infrastructure.
Country: Project timeline Ethiopia Project location: Bambasi camp, Assosa ––Second partner Conflict: begins shelter work 12 months - ––First partner begins Sudan and South Sudan shelter work conflict 11 months - ––Work on site begins Conflict date: September 2011 10 months - Number of people displaced: 40,000 refugees by end of ––Site master plan- 2012 ning Project outputs: 6 months - Camp for 12,000 people (3600 households) ––Need for new camp 2,175 shelters built (two identified organisations, 70 percent built ––54,000 Sudanese Sudan One month - refugees in Ethi- by one organisation) opia Shelter size: <2 people: 10m2 2 September 3-4 people: 14m ––Conflict restarts in 2 2011 – 4-6 people: 21m Ethiopia Sudan Cost per shelter: Assosa US$ 640 - 10.5m2 2006 - ––Returns begin 2 South US$ 800 - 14m ––Previous 2 Sudan 1990s - US$ 920 - 21m 1987 - history 1983 - of conflict and 1969 - refugee influxes
Project description The organisation planned and built a camp for Sudanese refugees. Semi-permanent shelters were constructed by refugees, with two partner organisations providing materials, carpenters and training. Refugees were able to chose their own plot configuration and the shelters were constructed with locally procured materials.
Strengths and weaknesses + The shelters followed local housing designs to make them - Initial estimates of construction time were too low, and cool in the day and warm at night. additional carpenters and masons were required. Fewer + The shelters were cost-effective, and were durable shelters were built than initially anticipated. alternatives to tents. - Many refugees did not recieve a shelter. Of those who did, + Materials were procured locally, reducing transport costs many received a tent whilst waiting. and injecting cash into the local economy. This provided some economic compensation to the host community. • Bamboo is grown extensively in the area, the eucalyptus + Shelter dimensions were tailored according to family size. was sourced from a neighbouring state owned forest. + Each plot was provided with a fence, a latrine and a • There was a very strong input from the government of shelter. Ethiopia in all issues relating to the camps. + Refugees contributed labour to build the shelters. This • Many families were separated when the first families helped to foster a sense of ownership. arrived. The rehousing of refugees was undertaken in Inital plans to for the organisation to build the shelters itself - parallel with replanning the camps and a family reunification were dropped as other organisations had management exercise. systems better suited to implementation. - Technical staffing capacity was a constant challenge. • Sudanese refugees brought large numbers of livestock - Difficulties in sourcing and transporting mud for the walls with them. Space for animals in the camp had to be were not foreseen. allocated (See B.3).
SHELTER PROJECTS EAST AND HORN OF AFRICA 8 3 / A.8 Sudanese conflict
• Average Plot Size per Household: 15mx10m Site development Despite delays, by the summer of 2012 plot demarcation had begun and the road was upgraded in order for it to be functional during the rainy season. Once Bambasi camp was estab- lished, water was provided from eight shallow wells (up to 60m deep). Later three boreholes were developed and a system of 34 tap stands was estab- lished.
Shelter construction The organisation initially planned Refugees were supported with materials, carpenters and to build all of the shelters itself, and masons to build tukuls based on the designs used by the host community. The refugees contributed labour. built some sample shelters. However, Photo: UNHCR it became clear, that the organisation lacked the management systems Before the influx communities in border areas, and required to build the numbers of Sudanese refugees have sought a transit centre was established at shelters required. As a result an alter- safety in Ethiopia since 1969, first Ad-Damazin. With the camps at full native implementation process was settling in the Gambella region. Addi- capacity, this transit site became more chosen, using partner organisations. tional refugee influxes 1983, 1987 and permanent. Given the scale of the Two organisations were identified the early 1990s led to the creation influx of refugees, new camps were to implement the shelter programme. of five refugee camps in western needed. Implementation by the partner Ethiopia. Three were established in organisations began in August/ the Gambella Regional State and Site selection September 2012. two in the regional state of Benis- Negotiations began with the A fixed design of shelter (a hangul-Gumuz in the area surround- national government’s refugee agency tukul) was built. It was based on the ing the town of Asossa. and the local government to identify shelters built and lived in by the host Following a peace agreement sites. community, differing from the shelters between north and south Sudan, A 450 hectare site owned by the that the refugees were accustomed refugees began returning home from adjacent village was identified at to building. As a result construction March 2006 onwards and three of Bambasi, 50km from the border. It training was required. the camps could be closed. 23,000 had with suitable drainage and access Carpenters and materials were refugees remained in one camp and was around 600km or a two day provided and managed by the in Gambella and a further 4,000 drive from Addis Ababa, the capital of implementing organisations, while remained in one of the Assosa camps. Ethiopia. families had to provide the labour. The refugee population included The host community and the Most families were able to provide the several hundred refugees from the refugee population had a similar tribal labour, but in the case of the most vul- Democratic Republic of Congo. Those heritage which, once some initial dif- nerable households, some support remaining in the camps included ferences were resolved, led to a good was required. several hundred Congolese people. relationship between the two commu- The implementing organisations nities. both provided a site engineer to lead Displacement in 2011 the project and a site foreman to Fighting recommenced in Site planning manage the teams of carpenters and September 2011 in the Blue Nile State From March 2012 the prepara- masons in the camp. Both organisa- of Sudan, displacing more refugees tion of the master plan began. The tions required significant logistics into Ethiopia. plan took four months to develop and support. By mid October 2011 there were an agree. The process was significantly In the project plans, a carpenter estimated 54,000 Sudanese refugees delayed by complications in awarding and a mason, working with families and asylum seekers in Ethiopia. About the contract to build the access road. would be able to build ten shelters in 34,000 were registered and accom- The camp was designed to be no fifteen days. In practice, only half the modated in three refugee camps: closer than 500m to the village. The number of shelters could be built. This Sherkole and Tongo near Assosa and camp was divided as follows: was due to an underestimation of the Pugnido near Gambella. • Number of Zones: 3 training required by those constructing Most of the new arrivals from • Number of Blocks: 40 the shelters, and an underestimation the Blue Nile State stayed with host • Number of Communities: 265 • Number of Family Plots: 5,240
9 SHELTER PROJECTS EAST AND HORN OF AFRICA Ethiopia Excerpt from: Shelter Projects 2011–2012 3 / A.8
Timber and bamboo frames shelters were built with thatched roofs as a more durable shelter solution than tents. It proved challenging to source mud to plaster the walls as originally intended. Photos: Left: UNHCR, Right: Demissew Bizuwerk / IOM Ethiopia.
of the number of households who Each shelter required significant would require additional assistance. volumes of grass for thatching the roof Materials list and for strengthening the mud walls. Below is a materials list for different Selection of beneficiaries The grass could only be harvested shelter sizes. Refugees were brought to seasonally with the main harvest Bambasi camp from the transit site being in March. This did not coincide near the border at Ad-Damazin. The with the construction, which needed Material Quantity / refugee population had continued to to continue all year round to meet the shelter size rise while the camp was being built, needs. Shelter (small) 10m2 14m2 21m2 and many refugees had settled near The sourcing of sufficient quan- Eucalyptus poles 10 11 14 the border. tities of mud also proved more chal- 5m x 10cm Shelters were allocated according lenging than anticipated. Initially mud Eucalyptus/ 27 33 40 bamboo 5m x <8cm to family size. Each family was came from digging the latrine pits Eucalyptus pole 1 1 1 allocated a 10m x 15m plot. For but this was insufficient for the initial 5.8 m x 12cm families with seven or more people, shelter needs, and for re-mudding Bamboo 5m length 37 39 90 two plots were allocated. after the rains. By the end of 2012, the Mud with grass (m3) 2.45 4.37 4.89 organisation was still trying to identify Bamboo (roof) 57 66 80 Logistics sources for mud and to organise suf- 20cm All of the shelters were built using ficient trucking for the large volumes Bamboo reeds 25 45 50 50cm locally available materials: bamboo, required. Grass (bunch) 15 22 30 grass (for a 15cm thick roof), rope The camp water supply was suffi- Rope and strings 100m 150m 200m and mud. This approach was much cient to cope with the volume needed Door with frame, 1 1 1 cheaper than sourcing materials in to mix with the soil. hinges and lock the capital, also cutting transportation Window 0.6m x 1 1 1 costs. Situation at the end of 2012 0.6m By the end of 2012, there were Nails 2. 5’’ (kg) 0.5 0.5 1 over 86,000 Sudanese refugees living Nails 4’’ (kg) 1 1 2 in Ethiopia. Used motor oil 3litres 3litres 5litres Approximately 3,700 refugees CIS Nails 0.25 2.5 0.25 formerly registered in Ad-Damazin still Fence Eucalyptus 5 remain in the local community after 5m x 10cm opting out of the formal relocation Bamboo (1m 80 process to Bambasi camp in June spacing) and July 2012. A few dozen refugees Tools moved spontaneously to Bambasi in Claw Hammer 1 September. In October 2012, 2,000 Bow saw 1 refugees were relocated to the camp Shovel 1 by local officials and were accompa- Meter rule 1 nied by around 8,000 livestock. Pick axe 1 Axe 1
A carpenter and a mason worked with each family to build shelter. Photo: Demissew Bizuwerk / IOM Ethiopia.
SHELTER PROJECTS EAST AND HORN OF AFRICA www.shelterprojects.org 10 Africa4 / A.4 Shelter Projects 2008 Election violenceA.4
A.4 Kenya4 -/ 2007-2008A.4 Kenya - -2007-2008- Election violenceElection violence Transitional shelter kits Project type: Pilot project providing transitional shelter kits Technical support for building Full construction for vulnerable households Emergency: Kenyan election crisis, 2007-2008 No. of people displaced: 125,000 - 250,000 IDPs found shelter in camps and similar settlements during the violence. An estimated 300,000 moved in with relatives or friends and around 12,000 fled to Uganda. Project target population: 481 transitional shelter kits provided as a pilot Kenya project (226 erected by the agency, 255 self-built) Occupancy rate on handover: Nairobi 86% - Those not occupying shelters wanted to wait until the shelter had been upgraded with stronger walls or until other family members returned. Both reasons related to ongoing feelings of insecurity. Shelter size 18 m2 (extendable, modular construction)
Summary Provision of transitional shelter kits as a pilot project in the Rift Valley of Kenya, before upscaling to a national response. Shelters were designed to be adapted by beneficiaries into permanent homes and, except in the case of vulnerable households, were erected by the beneficiaries themselves.
Project timeline
Dec 2007 1 month 3 months 4 months 6 months 7 months 8 months I I I I I I I
spark conflictPresidential electionKibaki resultsreached and RailaPower-sharing between Odinga MwaiShelter agreement final Strategy Shelterdraft of ClusterTransitional prepares transitional Design shelter and testingcal model craftsmenselection of Beneficiary and and labourers training Pilotselection of construction lo and beginsdistributionPilot completed with final
-
Strengths and weaknesses 9 Consideration was given to how the shelters could 9 Only viable project sites were selected, based on the be upgraded in the future to permanent homes. This security guarantees of the local administration, existence of maximised the impact of the financial investment. peacebuilding initiatives and willingness of IDPs to return. 9 Use of robust building components meant the shelters 9 Because it used local building technologies and local could be relocated. Some beneficiaries used plastic spacers craftsmen’s knowledge, the design was readily accepted by when nailing the roof to make disassembly easier. the beneficiaries and easily built. 9 Close involvement of the community and local 9 Having construction teams of mixed ethnicity administration in beneficiary selection meant that contributed to the peacebuilding process in an unplanned distributions ran smoothly and disputes were resolved. but positive way. 9 Linking the project with livelihoods interventions
11 11 SHELTER PROJECTS EAST AND HORN OF AFRICA A.4 Kenya - 2007-2008 - Election violence - Transitional shelter kits Africa Kenya Excerpt from: Shelter Projects 2008 4 / A.4 Strengths and weaknesses (continued) promoted sustainable return. roofs and doors had been looted. 8 Occupancy was not as high as hoped for, with some 8 The kit included spare sheets and plastic sheeting for IDPs not ready to move back. the construction of latrines. These materials were often 8 Not all of the materials are available locally in sufficient used to extend the roof instead. quantities. Sourcing of materials needs to be reconsidered 8 Some beneficiaries stated that they would have before the project can be upscaled. preferred to have been given the cash value of the plastic 8 Only those whose houses had been completely so that they could buy local materials themselves to build destroyed received the kit. Further attention needs to be the walls (cash grants are being considered for the post- given to those whose houses are partly damaged, as many pilot phase). Photo: Mark Lawler
Transitional shelter built on the family’s own land Situation before emergency Selection of beneficiaries A number of the tensions related to The Shelter Cluster agreed that An ad hoc beneficiary selection the ethnic nature of political affiliation 481 transitional shelter kits would be committee was established by the in Kenya, unresolved land issues, in- distributed as a pilot project to test the local administration, with appropriate equality of wealth distribution, high un- design of the shelter and the response representation of women and IDPs, employment and conflict over natural of beneficiaries. to select the final beneficiaries. This resources led to violence following the committee was monitored by the im- It was important that the site December 2007 election. plementing agency. chosen should be one where security The majority of those displaced was good, IDPs were willing to return The degree of vulnerability of the from the Rift Valley province had lived to and the community they were households was also assessed and was in small timber pole-framed houses returning to was ready to accept them. intended to be used as another filter with timber or adobe wall cladding, Mtaragon, in the Kipkelion District, fit in beneficiary selection. But as the thatch or iron-sheet roofs and the requirements. number of shelters to be provided compacted soil floors, strengthened almost matched the number of houses The local administration had a with dung or cement. completely destroyed, vulnerability record of all IDPs. Their assessment of criteria was used to determine the level the impact of the violence, correlated After the emergency of construction assistance a household with the agency’s own assessment, The election crisis was compound- required, rather than to select the ben- showed that around 500 houses had ed in April by food security problems, eficiaries themselves. flooding in some areas and drought in been completely destroyed. the north. The pattern of displacement To qualify for construction assist- The following criteria were used was complex. People were displaced ance, the household had to be headed to decide which of the 500 house- from many different parts of the by a single parent or a child or have holds who had no shelter to return to country as one ethnic group escaped members who were elderly, disabled would be chosen to receive a kit. The the threat of violence from another. or had special health requirements. selected beneficiaries: Around half of IDPs found shelter The criteria for the upscaled • were registered as an IDP by the in camps. The rest sought refuge with project was modified from the Shelter local administration; friends or relatives and some moved Cluster’s Transitional Shelter Strategy • were willing and ready to return; back to their ‘ancestral’ land where developed in March 2008, following • had proof of land ownership. support services were limited. feedback from the pilot project. A response plan was developed Proof of land ownership was only Implementation through the Cluster System, which required for this pilot project. It A prototype of the shelter was would provide non-food items and was anticipated that an appropriate tested for structural quality and tents to meet the need for emergency response would later be developed by reviewed by IDPs for its suitability. At shelter while a transitional shelter the Shelter Cluster to deal with those the same time as the final selection of design was developed to bridge the without formal titles to their property beneficiaries was being made, a second emergency and permanent shelter or whose houses were only partially prototype was built in a prominent phases. damaged. 12 12 SHELTER PROJECTS EAST AND HORN OF AFRICA 12 Africa Shelter Projects 2008 A.4 4 / A.4 Election violence
location in Mtaragon to sensitize bene- ground at a depth of around 60cm. The Materials Quantity poles supported a timber ring beam, ficiaries as to what was being provided Walls and to get feedback on the design. which in turn supported the timber rafters onto which an iron sheet was Cedar posts 14 units 9', 4" diameter Local craftsmen and unskilled nailed. labourers were recruited into ten Walling-polythene sheeting- 45 m2 teams and trained. Although not Walls were clad in plastic sheeting 1000g planned, the teams were a 50-50 mix and floors were compressed earth. The Cypres timber 2x3", 20 m from the ethnic group that had fled and doors were flaps in the plastic sheeting 6 x 2m, 3 x 2m the ethnic group that they felt threat- and weighted with timber battens. Ordinary nails 4" 2 kg ened by. This side effect of the project The design was based on the ver- Roof had a positive impact on peacebuild- nacular housing typically lived in by Cypess timber 2x3", 40 m ing. The donor organisation directly IDPs prior to their displacement. 2 x 10m, 3 x 3m, 1 x 8m procured the materials within Kenya This enabled IDPs to upgrade their Cypes timber 2x2" 41 m and delivered them to the implement- shelters incrementally using materials 6.5 x 6m ing agency’s warehouse in Nakuru. and methods that they were already CGI ridge covers-30g -1.5 m 4 units The implementing agency then familiar with. The walls could be clad CGI sheets-30g 20 units distributed the materials at three with timber, adobe or even brick and 2 x 0.9m locations. Beneficiaries collected them cement. Cement could be used to Ordinary nails, 2kg 4", 2kg 4.5 kg and took them to their plots up to increase the durability of the floor. 3", ½kg 2" three kilometres away, using their own The use of plastic sheeting allowed Roofing nails 4 kg transportation (either by hand, by shelters to be built and occupied Iron hoop 1 kg donkey, or by tractor and trailer). very quickly, though some beneficiar- Tools The kits also included the basic ies replaced the plastic sheeting walls Stanely claw hammer 1 unit tools necessary to build the shelter. immediately with adobe or reclaimed building parts, such as doors or timber. Stanley woodcutting saw 1 unit ‘I’m over 60 and unable The plastic sheeting could then be sold Panga knife 1 unit to get the materials to or used for temporary house exten- Hoe and handle 1 unit build on my own. Despite sions, and provided waterproof storage Manaila thread 30m (roll) 1 unit for seeds and fertilisers. what happened, I have Measuring tape 1 unit to continue staying here. The use of regular frame and Being my land I cannot roof sections made the construction run away. If everybody modular – it could be easily extended can be assisted in the or adapted. The choice of materials ‘The prototypes built by way I was, that would meant that there was no part of the local craftsmen in each be great. Plastic sheeting building that could not be fixed or project location enabled is OK, but I would have replaced locally. structures to be tested and preferred timber, as it’s important feedback from stronger and can’t be Most beneficiaries erected their builders and beneficiaries blown away’.- Beneficiary shelters on exactly the same site as their previous homes had been, so to be incorporated into Guidance was given by the local little site clearance or ground levelling the final design.’– craftsmen on how to put the shelter was required. Engineering coordinator together. The beneficiaries provided Logistics and materials the labour themselves and the houses Materials were sourced in Kenya, were normally completed within one and chosen for their familiarity, durabil- or two days. ity and low cost. Timber was supplied Over 45% of the beneficiaries met by private forestries who were only the vulnerability criteria and qualified considered if they had government- to have their houses built by the con- approved replanting projects in place. struction teams. Plastic sheeting was made from recycled Technical solutions plastic. The total cost of materials and labour for one transitional shelter was The structure had a covered space US$ 350, not including transport and of 18m2 (6m x 3m), was split into two agency administrative costs. rooms, and had good clearance above head height. Photo: Dyfed Aubrey The frame was made up of 10cm diameter cedar poles, dug into the Transportation
13
13 www.shelterprojects.org SHELTER PROJECTS EAST AND HORN OF AFRICA A.3 Kenya Kenya - 2007-Excerpt Flooding from: - ShelterShelter Projects and disaster 2008 mitigation 5 / A.3Africa
A.3 Kenya - 2007-5 / A.3 Flooding Kenya - 2007- Flooding Shelter and disaster mitigation Project type: Construction of self-build new shelters for refugees Community mobilisation, disaster mitigation Disaster: Ifo refugee camp flood response, Dadaab, Kenya, 2007 No. of people displaced: Approximately 6,000 households displaced, mostly from the Ifo camp Project target population: 500 households in the Ifo camp kenya Occupancy rate on handover: 100% (based on visual assessment) Shelter size Somalia 2 18m (6m x 3m) NairobiDadaab
Summary Through a combination of upgrading and emergency response funding, 500 families were assisted in making bricks and building shelters through a community-based construction programme following flooding in a large refugee camp.
Project timeline
Jan 2007 1991 Nov 2006 Dec 2007 I I I
upgradingLimited projects funds for small Project begins Project completed Kenyaopen aroundSomali refugeeDadaab, camps campscutting100,000 roadFloods access camp affectto inhabitants, nearly
soil from outside the camp required negotiations with the Strengths and weaknesses host community to avoid conflict. 9 Strong community participation through the training of 8 Water consumption was high. Water meant for domestic beneficiaries to construct their own shelters meant project consumption was used in brick production. Rainwater costs were low and construction standards were high. catchment systems will help to avoid this in the future. 9 A sense of ownership and pride in their shelters was 8 Though foundations increase the structure’s strength, demonstrated by the wide variety of self-implemented they can still degrade through contact with water. modifications, raising living conditions. Stabilising the soil with cement will help to make them 9 Mud brick production has become a major income- stronger. generating activity even though the project has finished. 8 The inclusion of people from minority groups, such as 9 Deforestation in the Dadaab area was reduced by the disabled, was not fully realised. replacing stick walls with mud bricks. - The agency needs to use the refugee initiatives that 9 The use of a thick foundation and lower wall reduces emerged from this project to help redesign its strategy. the possibility of collapse in heavy rains. Supporting livelihood activities may accelerate the 9 Broken bricks were recycled to demarcate plots, build construction pace and decrease costs. furniture or were remixed with water to be remoulded. - Opportunities for income-generation activities and 8 Soil quality was variable outside of the camp, so many broad environmental concerns require joint agency used soil from their own plots. This created hazardous solutions. This kind of shelter project requires coordination holes that may create mosquito breeding grounds. Sourcing among agencies working in different sectors. 8 8 SHELTER PROJECTS EAST AND HORN OF AFRICA 14 Africa5 / A.3 Shelter Projects 2008 Flooding A.3 Photo: Joana Cameira
Village constructed through community-based project Photo: Joana Cameira
Brick production Situation before emergency After the emergency It was decided that the agency would Three refugee camps (Ifo, Hagadera The severe flooding in the Ifo camp follow the idea of previous shelter pro- and Dagahaley) sheltering mainly destroyed over 2,000 shelters and left grammes in building mud-brick houses, Somali refugees were established close more than 10,000 people homeless. but would improve the durability of the to the town of Dadaab, in Northern This meant that many refugees had to design, increase the involvement of the Kenya, in 1991 and 1992. By 2007 they move to a new camp neighbourhood, communities and reduce the need to had a population of around 173,000 ‘Section N’. pay beneficiaries for construction. people. Section N was not a popular choice The aims of the programme and Dadaab is an area with little veg- for many refugees. Although the the implementation of the strategy etation and refugees’ access to natural ground was higher and less affected by were explained to camp leaders who resources (including building materials) floods, the site was further away from disseminated the information. As well, is limited. The government of Kenya the market and its lack of trees meant community mobilisers (agency staff does not encourage activities that are little natural shade. who were based in the blocks for eight ‘permanent’, so refugees rely on aid hours per day) ensured that the right agency support rather than self-suf- Selection of beneficiaries information was reaching everybody. ficiency through agriculture or other Beneficiaries had been preselected A public demonstration of ‘brick livelihoods. by a UN agency, following standard vulnerability criteria that was verified throwing’ to test the strength of bricks The camps are highly congested, through door-to-door checks. made from different soils ignited the creating sanitation problems and fire interest of potential beneficiaries and safety issues. The majority of shelters Implementation addressed the fears of mud-brick in the camp are of two types, both The agency faced two main chal- houses being weak. The agency con- employing highly flammable roofing lenges: convincing refugees that structed some prototype shelters that materials: traditional tukuls – 3.5m Section N could become a nice place were then used as classrooms for the diameter dome structures made of to live and that improved mud-brick construction trainees. wooden sticks, covered in fabric; and constructions would be stronger than The agency then provided a ‘training adobe huts – 6m x 3m shelters using a the previous buildings that the refugees of trainers’ to a small group of refugees large number of sticks for walls with a had seen washed away. on construction techniques and brick- roof made of local vegetation. making. Efforts were made to ensure
15 SHELTER PROJECTS EAST AND HORN OF AFRICA9 A.3 Kenya Kenya - 2007-Excerpt Flooding from: Shelter - Shelter Projects and disaster2008 mitigation 5 / A.3Africa
that training teams included women and the elderly. Each trainer super- vised around four families per month, assisting them with layout, foundations, walling and plastering. Carpenters were deployed to give technical support on ‘I used to live in a bush roof and latrine construction. house. It was not really a ‘It was my first job! It al- house. It is better here’. –
lowed me to support my Elderly refugee Photo: Joana Cameira family’. – Female refugee construction trainer Completed house mortar or cow dung and the roof was next stages of the implementation Soil-sourcing sites, both within and covered with iron sheeting. Improve- strategy, to increase the supply at lower outside of the camp, were identified by ments were made to ventilation to costs and in a shorter timeframe. the agency, which also supplied brick decrease the high internal temperature moulds, pangas (knives), wheelbarrows Logistics and materials of previous designs. and plastic sheeting to cover completed Families originally used soil from bricks during the rain. Tools were A change in the position of the planned and unplanned areas within shared among the community groups house on the plot improved sanita- the camp. A project to dig new garbage and returned to the agency when not tion. Latrines were moved to the front pits outside the camp presented an op- in use. Water storage was provided of the plot next to the street and the portunity for a new soil source. near the soil-sourcing sites. house was positioned at the back of the To reduce the water consumption plot. This left space for more construc- Agency staff maintained quality- necessary for brick production, ‘spilled tion inside the plot and prevented the control checks on all the constructions water’ from tap stands was collected. problems of a dirty backyard blocked to ensure the safety of the houses, par- The rest of the water was supplied by by wastewater runoff. ticularly as previous mud brick failures truck and stored in oil drums distribut- had been mostly due to poor con- Beneficiary modifications ed around Section N or in water tanks struction rather than design. Beneficiaries made a number of if the bricks were being produced Upon completion of the mud-brick modifications to the new structures. outside the camp. These included: structures, the agency supplied the Roofing and door materials were construction materials that the benefi- • Aesthetic: Painting and decorating. procured in the capital with support ciaries could not produce or purchase • Windows: The size was adjusted. from a UN agency, while other themselves, such as roofing sheets and Sometimes they were partially closed materials were procured in the nearest doors. with other bricks or sticks to increase large town. security and reduce sunlight but The combination of a team of The total cost of materials, including maintain ventilation. trainers able to transfer skills to the transport, was around US$ 440 if the • Furniture: Some families community and beneficiaries willing soil was sourced within the camp, constructed beds and tables out of the to participate in the construction of rising to US$ 480 if soil was sourced mud bricks, which helped to demarcate their own shelter at no cost led to outside the camp. Labour costs for the internal living space. full engagement of the community and each shelter were around US$ 30. guaranteed that people would maintain • Plot boundary: Small walls to define their properties themselves. the extent of a plot were often built Quantity Unit with spare or broken bricks. Iron sheets (2.5m length) 20 pieces Technical solutions • Plastering: Some families plastered Timber - cypress (2mx2m) 120 m The 6m x 3m houses required their house with cement mix, making 1,700 bricks, considerably more than the walls impermeable. Plain sheet (2.4m x 1.2m) 1 piece previous designs implemented in • Gutters were made out of waste Nails 4" 4 kg the camp. While disaster mitigation tin sheet and tin cans. Nails 3" 1 kg was primarily achieved by relocat- • Livelihoods: Market stalls were Nails 1" 0.5 kg ing refugees to the higher ground of built as extensions onto or between Section N, extra bricks were necessary houses, increasing the income of the Roofing nails 5 kg to build a thick foundation and lower families and providing more options Butt hinges 4" 3 pieces wall to improve the structure’s per- for other residents to shop locally. Padbolt 6" 1 piece formance in heavy rains. About 30% of the beneficiaries Tower bolt 1 piece employed other refugees at some stage Eight pillars provided support for GI Ridges (1.8m length) 4 pieces of the construction. This increased the walls and roof trusses, increas- Binding wire 5 kg ing the stability of the roof itself. the income generated in the housing Wood preservative 8 l Mud-brick walls were plastered with industry in the camp. Such initiatives inspired the agency to look into the
10 SHELTER10 PROJECTS EAST AND HORN OF AFRICA www.shelterprojects.org 16 6 / A.22 Drought and conflict
CASE STUDY 6 / A.22 SOMALIA 2011-2013 / DROUGHT + CONFLICT
KEYWORDS: Permanent housing, Resettlement, Advocacy, Infrastructure, Community participation, Land tenure
(July 2011 - June 2012) GULF OF Complex: Drought DJIBOUTI CRISIS and armed conflict. The project started at the peak ADEN of the drought in the Horn of Africa. PUNTLAND 3.7 million people affected by drought and TOTAL PEOPLE famine crisis (Source: OCHA, 2011). AFFECTED GAROWE 1.4 million internally displaced people (ibid.). BURTINLE PROJECT LOCATIONS Garowe and Burtinle, Puntland region, Somalia. ETHIOPIA
BENEFICIARIES 1,200 households (8,400 individuals).
1,200 permanent shelters built KENYA 1,800 individuals benefitting from cash PROJECT for work (masons, unskilled labourers and carpenters). OUTPUTS Other outputs include: access road, one health centre and one borehole in Garowe, water systems in both INDIAN sites, child-friendly space, public area and police post OCEAN in Garowe, 14 sex segregated toilet blocks.
SHELTER SIZE 16m2 (4x4m) one room shelters (10x10m plot).
RIO NAPO 2 SHELTER DENSITY 2.7m / person (average household size of 5.9 persons). PROJECT SITES
MATERIALS COST PROJECT COST USD 3,493 including site works, WASH USD 1,693 including labour. PER SHELTER PER HOUSEHOLD facilities and organizational overheads.
OUTCOME INDICATORS 100% occupancy rate of shelters in both sites. Secure land tenure obtained in both sites.
PROJECT SUMMARY This was a two-year, multidonor, multisectoral, project aimed at providing a sustainable shelter solution by building 1,200 permanent houses for IDP households in two relocation sites. The shelter programme was linked to Livelihoods, WASH, Health, and Education. The project adopted holistic settlement as well as community-led construction approaches. The organization managed to secure the land and receive additional funding for complementary activities, including infra- structure, facilities and common spaces.
MAY OCT 2010 2011 DROUGHT 2012 2013 2013 AND FAMINE 1 2 3 4 5 6 7 8
PLANNING PHASE IMPLEMENTATION PHASE 1 IMPLEMENTATION PHASE 2 HANDOVER PHASE TIMELINE JUL SEP NOV JAN APR MAY JUN JUL OCT JAN MAY JUN SEP
1 Jan 2012: Beneficiary selection and verification.
2 Feb 2012: Recruitment of staff completed. STRENGTHS + Achievement of tenure security and establishment of durable sites. Apr 2012: 3 Typology design agreed and start of land titles negotiations. + Continuous engagement of all stakeholders. + The selection criteria were established and agreed upon by all. 4 Apr 2012: 45 pilot houses completed in Burtinle. + Owner-driven approach, transparent and accountable systems. + Settlement approach, linkages with vocational training and sav- 5 Jun 2012: Land acquisition granted from local administration and Ministry of Interior. ings groups.
6 Oct 2012: 80 pilot houses completed in Garowe (delayed by resolving WEAKNESSES land issues). - Staff turnover and lack of flexibility of internal systems. - Limited female participation and lack of gender analysis. 7 May 2013: All 1,200 shelters and related facilities constructed, land - The project provided only one-room shelters, that were too small titles processing completed. to meet cultural needs. - Beneficiaries had to be incentivized to participate in the construction. 8 Jun 2013: Commissioning of settlements and start of handover of houses and land titles.
17 SHELTER PROJECTS EAST AND HORN OF AFRICA Somalia Excerpt from: Shelter Projects 2015–2016 6 / A.22 © World Vision Somalia Vision © World
The project built 1,200 permanent houses with accompanying infrastructure across two new sites (here, the site in Garowe).
CONTEXT a minimum kit, that would be locally procured and stockpiled Food security in Somalia had been deteriorating since 2010, by Cluster partners in strategic points in Somalia and Kenya. with almost all southern regions being affected; famine was Transitional shelter was provided to stabilized IDP settlements declared in the Bay region, for a total of 6.4 million affected (in Puntland and Somaliland). Interventions ranged from shel- people (more than half of the Somali population). Due to this, ter kits, to houses with corrugated iron roof sheets. The third and the instability and fighting within the country, the number pillar supported voluntary relocation, or return to the place of IDPs in Somalia was estimated to be 1.4 million1. The pro- of origin. Due to the presence of returnees from Yemen and ject areas were hosting the majority of IDPs in the Puntland Kenya, the Cluster adopted an equality approach, wherein region, which is primarily inhabited by people from the Somali IDPs, returnees and urban poor groups could be integrated. ethnic group (and of Muslim faith). Despite its relative stability, Although this project was initially conceived to fit under the the region had also endured armed conflict. second pillar, it ended up providing permanent shelters with secured land tenure, due to its longer engagement process. SITUATION DURING THE CRISIS PROJECT IMPLEMENTATION In the wake of a severe drought and the resulting famine in 2011, the population density in Puntland further swelled, due The project was implemented by a contracted team (Finance to the influx of IDPs who were fleeing violence in South-Cen- officer, Accountability officer, Supply chain officer, Project en- tral Somalia, concentrating around Garowe and Burtinle, and gineers) and with additional staff, brought in on a need basis, some of the long-term IDPs who had settled in Garowe. Dis- including shelter engineers and humanitarian accountability placed people were searching for life-saving assistance, due facilitators. A community-based construction approach was to limited access to water, food, health care services, and ad- adopted, whereby beneficiaries received construction materials equate shelter. The influx of IDPs led to increased tensions and technical support to build their houses. They were likewise between the host community and the new arrivals, as they responsible for identifying the skilled labour and providing the competed for limited employment, access to state services unskilled labour. Each shelter unit was constructed by an aver- and scarce resources. In Garowe, there was insufficient or age of five labourers (two masons and three unskilled workers). substandard shelter to meet their needs1. Additionally, the IDP settlements were unplanned and congested, due to the in- BENEFICIARY SELECTION flux caused by the ongoing drought. In Burtinle, all respond- The project aimed to provide shelter to people displaced from ents from a rapid assessment (conducted in two IDP camps) their homes due to conflict and drought, as well as the urban reported that the shelters were inadequate to protect from the homeless from host communities. The organization engaged weather. The houses were primarily buuls (huts made from all stakeholders (regional government, elders, religious lead- sticks, cardboard, old rugs and tents), offering little security. ers, community members) in the selection of beneficiaries. The Accountability Officer invited committee representatives SHELTER CLUSTER STRATEGY from more than 15 IDP settlements in Garowe, explaining the The Cluster response strategy in 2012 contained three pillars: shelter and vulnerability criteria, as well as the selection pro- 1) Emergency response, 2) Transitional shelter, and 3) Durable cess. The local authorities were tasked to work with settlement solutions. For the emergency response, the Cluster designed leaders in identifying the most vulnerable residents, based on agreed-upon criteria. Leaflets and posters were distributed in 1 Humanitarian gaps assessment coordinated by OCHA, 2011
SHELTER PROJECTS EAST AND HORN OF AFRICA 18 6 / A.22 Drought and conflict © World Vision Somalia Vision © World Somalia Vision © World
The project established two relocation sites (here Jilab Village, Garowe). Settlement planning included public spaces.
the settlement, wherever possible, to inform the inhabitants on BENEFICIARY ENGAGEMENT these criteria, which included: During construction, the beneficiaries were responsible for • People displaced by the insurgency within the target areas. ensuring that the houses were built according to their expecta- • Drought-affected people who had lost their livestock and tions, as well as for receiving and taking care of the construction had no shelter. materials. The community was also able to provide feedback • Rural self-settled: those outside the urban or peri-urban through suggestion boxes in each site. Regular monthly areas and those settled individually in small family groups meetings were held with the government and beneficiaries to on unoccupied land. discuss project progress, achievements, challenges, areas of • Households hosting and supporting displaced people improvement, as well as follow on the feedback received. with housing challenges. On top of these, the vulnerability criteria included age, dis- LAND TENURE SECURITY abilities, homeless widows, female-headed households, large The organization advocated from the beginning of the families, diseases, and no access to livelihoods. project to secure land tenure for IDPs, as a precondition for building the shelters. One of the challenges was that the bene- The organization carried out an independent verification ficiaries in most cases were from different clans than the land exercise once the beneficiary lists were submitted. Although owners. It was decided that these households should be pro- most beneficiaries were accepted, a few cases had tobe tected and have access to secure land tenure. Government changed in order to include the most vulnerable households. officials agreed to provide titles, as long as the organization The verified families were issued with a beneficiary ID card, would cover the registration costs. The organization publicized containing the information about their households. the contents of land documents to all stakeholders and further worked with the media to create public awareness, that the COORDINATION AND PARTICIPATION OF shelter units provided under this project were not for rent or DIFFERENT STAKEHOLDERS sale. The Ministry of Interior reposed any shelter unit that was The organization held a series of meetings with all stakehold- being sold or rented and re-allocated them to other people ers, to explain the implementation process. Firstly, awareness still living in the IDP camp. This aimed at discouraging people meetings were conducted with government officials. A design from infiltrating the system with the aim of making profit. workshop was then initiated to share information with govern- ment officials from the Ministry of Interior, Regional Governor However, the process to obtain tenure security was and Local Districts, IDP representatives, landowners and clan lengthy and delayed the project, especially in Garowe. There- elders. Different shelter design options were presented, ad- fore, the team decided to separate the issues of Burtinle and vantages and disadvantages were analysed, and the groups Garowe, in order to not delay the whole project. were requested to make recommendations to improve each In Garowe, the government was forced to stop the con- design. The coordination throughout the project avoided struction of houses after the organization indicated that per- unnecessary conflicts with the communities and other stake- manent houses could not be implemented on land with unsecure holders, such as clan elders, local authorities, and NGOs. tenure. The government was then requested to secure freehold The coordination with cluster members contributed to en- land for the IDPs, if these houses were to be implemented as sure that basic standards were maintained, based on cluster per the agreed design. A meeting was held and broadcast on guidelines. Effective coordination and information sharing television, with different sectors of the government, humani- with other sectors, particularly the WASH Cluster, enabled tarians, elders, and influential businessmen in the town, during the organization to learn from partners’ experiences and which the government pleaded to allocate special land for the achieve project goals successfully. resettlement of IDPs. This resulted into a piece of land meas-
19 SHELTER PROJECTS EAST AND HORN OF AFRICA Somalia Excerpt from: Shelter Projects 2015–2016 6 / A.22 © World Vision Somalia Vision © World
Households received construction materials and technical support, had to identify skilled labour and provide unskilled labour themselves. uring 1,000m by 150m being allocated to the organization for While the local market in Garowe was able to accommodate the shelter project, which was further subdivided into plots of the higher demand, the project in Burtinle was partially de- 10m by 10m for each household. layed due to lack of materials. The project team held meet- ings with suppliers and government officials, in order to have In Burtinle, the process was smoother, as the organization was the neighbouring businesses to assist, even though this was permitted to build on three existing sites that were identified initially objected. for the upgrading of makeshift shelters into permanent houses. Due to the high demand, the price of materials rose. Meet- Ultimately, the project’s ambitious goal was achieved in both ings were held with the settlement leaders and the govern- locations, with land allocated without time limitations and ment officials, to explain that the project budget was fixed by signed by the Ministry of Interior and relevant legal titles, the donor, thus higher prices would mean less beneficiaries. issued to each beneficiary, as part of the handover process. Additionally, in order to reduce the costs, the organization In particular, the titles were legal documents recognized by suggested to order goods directly from manufacturers. In the society and the sharia courts, and MoUs were signed with the end, both suppliers and government officials agreed to the organization. Notably, the project also included women as keep the prices stable, unless it was demonstrated that the household title holders. increase was due to external factors.
MAIN CHALLENGES WIDER IMPACTS OF THE PROJECT Apart from general security and access constraints for in- The long-term engagement with the regional government ternational staff, one of the major challenges was related to served not only to build the capacity of the government but as it was hard to recruit local engineers. The organ- staffing, also to legitimize its efforts and the goals of the overall project, ization therefore suggested to hire engineers from Somali- particularly on land tenure issues. The organization helped to land, but faced stiff oppositions from the Ministry of Labour. establish the government as a credible voice and partner in This delayed the employment process, though ultimately lo- the well-being of Puntland residents. Reciprocally, the govern- cal engineers were identified. ment formally recognized the site in Garowe as “Jillab Village”. Another challenge was to uphold humanitarian account- One of the most striking discoveries in the impact evaluation ability principles, given that the government tended to as- was the dramatic reduction of crime from the IDP camps sume they would take the lead in communicating with the to the resettlement sites. In both sites, village elders re- communities, instead of the organization. More advocacy on ported only a handful of petty crimes within memory. Wom- the importance of accountability to all stakeholders should en, men and youth unanimously reported feeling safe in all have been factored in from the start. parts of the compounds. Additionally, the evaluation indicat- ed a reduction in gender-based violence, according to the MATERIALS AND SUPPLY elders and settlement leaders, to which they credited locka- All the construction materials were procured locally. The sup- ble windows and doors in the new shelters. pliers were provided with information on the beneficiaries, including the resettlement site and plot number. The mate- Finally, the lessons learned from this project were applied rials were then distributed to and received directly by the in another shelter project that the organization started in beneficiaries, using supplier’s vehicles. Dolow.
SHELTER PROJECTS EAST AND HORN OF AFRICA 20 6 / A.22 Drought and conflict
STRENGTHS, WEAKNESSES AND LESSONS LEARNED
MATERIALS LIST FOR ONE HOUSE (ESTIMATED BEFORE COMPLETION) Description Unit Quantity Rate (USD) Tot. Cost (USD) Stones m3 4.4 6.25 27.50 Aggregates for concrete (gravel) m3 0.5 10.50 5.25 Sand m3 1.3 6.25 8.13 Portland cement (50kg) Bags 16 8.00 128.00 Blocks (40x15x20cm) made from 1:7 mix cement-sand No 700 0.65 455.00 Stirrups, 6mm mild steel diameter, 6m long No 8 2.50 20.00 Steel reinforcement 10mm diameter, 12m long No 6 12.00 72.00 Roofing Nails Kgs 3 2.50 7.50 Wire nails - assorted Kgs 6 2.20 13.20 Galvanized iron sheets, 2.4m long of 28 gauge thickness No 18 9.00 162.00 Galvanized iron ridge caps of 30 gauge thickness No 1 9.00 9.00 Roofing timber, 2”x2”, 3.9m long No 9 5.00 45.00 Roofing timber, 2”x3”, 3.9m long No 15 6.00 90.00 Formwork timber, 1”x6”, 3.9m long No 8 6.50 52.00 Steel door complete with frame, hinges, locking system – 1x2.2m No 1 60.00 60.00 Steel window complete with frame, hinges, locking system - 0.85x1m No 2 30.00 60.00 Formwork timber 1”x2”, 3.9m No 2 2.50 5.00 Labour costs Lump sum 343.00
WEAKNESSES STRENGTHS - Staff turnover and lack of flexibility of internal systems + Achievement of tenure security and establishment of a and processes impacted the project timeline. For example, community in the targeted areas. Positioning the Housing, the regional accountant and his deputy resigned during the Land and Property focal point to coordinate with the govern- implementing period and no replacement was found for long. ment contributed to the strategic engagement and capacity- This affected the timely processing of financial reports and building of the authorities. delayed the procurement approval process, as some deci- + Continuous engagement of all stakeholders to explain sions had to be referred to Nairobi. the beneficiary selection and the implementation process. - Limited female participation. Gender inclusion in Somalia This was found to have significantly contributed to managing is bound by cultural and religious considerations, which af- the expectations of suppliers and local authorities, as well fect the ability to engage female staff and beneficiaries to the as reduce rumours of theft and misappropriation of project same extent as males. Programmatic gender analysis is assets and materials. necessary and should be built into monitoring systems, in + The selection criteria were established and agreed order to tease out power relations and influence biases, flag upon by all stakeholders. Beneficiaries were able to under- the level of women participation in the project, and inform stand and explain the reasons why they qualified for assis- actions to improve equitable participation. tance; the same was true for those who were not selected. This - Although in Somalia the common practice is to build several shows how effectively the information was shared amongst single room shelters (tukuls) for one household, which offer the community, and how transparent the system was. privacy for parents, children, relatives and can accommodate + Continued engagement of beneficiaries and owner-driven large families, this project provided only one-room shelters. approach to construction. This included the transparent and These could not meet these family needs, however, houses accountable systems that were established for the benefi- could be further expanded on the plot allocated to each family. ciaries, to be in control of the materials received and accept- ed. For instance, beneficiaries refused to accept the supply of blocks when these did not meet the agreed upon standards. LEARNINGS • Beneficiaries had to be incentivized to participate in owner-driven construction. This required a good understanding of the local context and skilled community motivators. As the interest in participating in manual construction work was low, the project team advised beneficiaries that they would be given priority if they provided labour. • Being clear and consistent from the beginning on the mandate of the project, and sharing the objectives with relevant authorities, forced them to identify a suitable piece of land. • The settlement-based approach allowed the team to consider the root causes of vulnerability in this region and to avoid the “bandaid after bandaid after bandaid” situation. The organization has embedded disaster risk reduction and resilience building into its development and humanitarian practice, ever since. • Linking the programme to vocational training and saving groups helped people to build new skills and earn money. Offering options is important, as it allows community members an opportunity to exercise choice, helping to ensure that they are more than passive actors in the process and can thus find solutions tailored to their needs.
21 www.shelterprojects.org SHELTER PROJECTS EAST AND HORN OF AFRICA Somalia Excerpt from: Shelter Projects 2011–2012 7 / A.28
7 / A.28 Somalia – 2011 – Famine / Conflict Keywords: Urban neighbourhoods, Household NFIs, Construction materials, Transitional Update: shelter, Site planning, Infrastructure, Coordination.
Country: Project timeline Somalia Project location: Mogadishu ––Construction phase 14 months – Conflict / disaster: July 2011 Famine and ––Mapping of Zona K Continuing conflict 12 months – ––Dedicated Tri-Clus- 11 months – ter Coordinator Number of people displaced: arrives 200,000 IDPs in Mogadishu ––Selection of 16 Project target population: 10 months – projects to be Approximately 36,000 implemented by 14 agencies Project outputs: 3,645 housing units complete ––Strategy formed WASH and health facilities 8 months – and dedicated co- ordinator agreed Occupancy rate on handover: ––First funds allo- 100 per cent - November 2012 7 months – cated Shelter size: 15.8m2 (3.6m x 4.4m) ––Non-food items and shelter kits Materials cost per household: distribution US$ 420 July 2011 – ––Famine declared Somalia
Early 2011 – ––Displaced people start to arrive Mogadishu
Project description The Tri-Cluster project is a coordinated group of 16 projects implemented by 14 partners across the sectors of shelter, WASH and health. Zona K in Magadishu was chosen as the target area as it had the densest concentration of IDPs and was the least likely IDP settlement to be evicted once Mogadishu stabilised and developed. The project goal was to improve the protection for displaced people living in Zona K through improved settlement planning and the provision of integrated services from multiple sectors.
Strengths and weaknesses - A weak community structure combined with the fact + Regular coordination meetings achieved a that many people were already settled within the common understanding of aims and objectives settlement meant that it was not always possible to amongst all partners. follow site plans and meet minimum standards. + By integrating services the project was able to act - Communal spaces have been eroded by an more efficiently to provide shelter, access to water increase in the numbers of people living in Zona K. and sanitation and basic health serives. + Settlement planning has enabled organisations to • As the sectors work at different levels (shelter have better access and the beneficiaries have an with households, WASH with groups of five families enhanced sense of community. Displaced people per latrine and health with the whole community) were involved in the development of context- synchronising activities required complex work specific planning standards which helped manage plans. expectations. • Mapping all the stakeholders in the process was - Underestimation of the impact of other projects difficult, and their influence changed over time. funded through other sources active in the same • The project had a high profile, putting implementing project area. partners under pressure to produce results quickly, - Although eviction is unlikely in the short-term, compromising planning and construction quality. there is no clear ownership of land and so displaced • The Tri-Cluster coordinator took on many of the people are vulnerable to the Somali ‘gatekeepers’. camp management and camp coordindation duties.
SHELTER PROJECTS EAST AND HORN OF AFRICA 22 An urban area of Mogadishu was re-planned and many organisations worked together working in three sectors of interven- tion. These shelters have been upgraded by inhabitants who have built their own external shaded and cooking areas. Photo: Richard Evans
Before the displacement with an estimated 70,000 IDPs living very successful. Instead, the Cluster Mogadishu has hosted displaced in make-shift shelters called buuls advocated that organisations should people from conflicts since 1991. (traditional Somali thatched shelter). provide humanitarian assistance to However, as drought worsened in These were constructed by the IDPs the locations where IDPs had self- late 2010 and famine approached in themselves from scavenged materials settled. This has been the approach early 2011, more and more Somalis and items received from humanitarian in Somaliland and Puntland where were driven away from rural areas to organisations. the conditions and access are more Mogadishu looking for assistance and Any attempt to coordinate favourable. safety. settlements in Mogadishu would The mechanics that control the Displacement was compounded have directly interfered with the creation of new camps were deemed by the ongoing conflict in Somalia. economic relationship between the too complex and unpredictable to host population and the IDPs. As a encourage new sites. After the displacement result, no formal camp coordination Upon arrival in Mogadishu, the mechanism was established. Implementation Internally Displaced Persons (IDPs) As a response to the influx of Under the umbrella of the Tri- settled on any unoccupied land. This IDPs into Mogadishu, a three-phase Cluster there were five shelter process of self-settlement meant that strategy was developed in July 2011: projects, with a total value of US$ 4 there was no site planning. Services million. such as water and sanitation, and • Provide all displaced people with a The first project focused on access to the 100 or so settlements non-food item packages mapping the existing settlement, were sporadic. As the number of sites • Provide transitional shelter producing settlement plans, and solutions closer to the centre of town reduced creating access roads and storm • Provide site planning to improve and as Al-Shabaab’s influence drainage. living conditions and access to lessened, many IDPs settled into the other basic services such as This mapping was followed by area which became known as Zona K. WASH and health. consultations with the beneficiary Zona K’s mixed ownership, community and landowners to ensure between the government, the The shelter coordination did that people would not be evicted once university and some private individuals, not advocate the creation of new work was completed. meant that it was one of the least likely settlements for the IDPs. This strategy One organisation chose sites to be evicted. By the end of 2012, was attempted in Puntland (see A.8 in to work through long-standing the site covered an area of over 3km2 Shelter Projects 2008) but was not partner organisations while the
23 SHELTER PROJECTS EAST AND HORN OF AFRICA Somalia Excerpt from: Shelter Projects 2011–2012 7 / A.28
other contracted the work to local construction companies. Where possible the implementing organisations followed the site plans, but they were often forced to deviate from them. Reasons for this included the need to accommodate new demands from stakeholders, the construction of new permanent structures that had been built after the initial mapping, and the need to accommodate a larger population. Once the shelters were completed, two local organisations provided non-food items, including blankets, kitchen sets, jerrycans and fuel- efficient stoves. Beneficiary lists were provided by the main shelter The project integrated shelter WASH, health and site planning. partners,and distributions were Photo: Richard Evans undertaken once the shelters were handed over. could be carried out at the same time existing two partners, and 3,000-4,000 as the mapping. more shelters were planned. Selection of beneficiaries Coordination and communication Once the framework and common The whole area was sub-divided was needed with the local authorities understanding on coordination was into 25 zones, and settlement ensured that they were aware of created, it became feasible to add planning was based on the displaced the project and its implications, and additional sectors and projects. population at the time of mapping. that they approved the temporary The Tri-Cluster approach came The two main organisations started in development plans. As the final about because the Humanitarian different zones and completed all the shelter solution was semi-permanent Coordinator considered that shelter, construction before moving on to the (5 to 10 year lifespan), the urban WASH and health were the most next. Every IDP that was registered planning undertaken as part of the Tri- pressing needs for the IDPs. At the received a shelter and non -food item Cluster, will influence the development time there was surprise that other kits. The other Tri-Cluster partners of this part of the city. Access roads sectors were not also included in a provided sanitation and water points created now, will be the main roads for multi-sectorial approach. However, in the locations identified during the years to come. the coordination of just three sectors planning process. was difficult enough, and in retrospect Technical solutions the presence of additional partners Coordination The shelter actors worked with and targets may have reduced the Effective coordination was crucial the main partners to identify a unified effectiveness of the entire intervention. for success, as there were 16 projects shelter typology. Initially, US$ 80 Generally, once an organisation operating in a very concentrated area. shelter kits were planned as the secured funding, the focus was In addition, there were many actors land tenure was not known. Later, a immediately on implementing as who were already working in Zona ‘hybrid’ between plastic sheeting and quickly as possible in order to meet K. Therefore, a dedicated Tri-Cluster corrugated galvanised iron (CGI) was project targets. To combat this “tunnel coordinator was brought in to act as a adopted during the planning stage. vision” amongst organisations, the focal point for the 16 projects. This provided a better quality shelter successful multi-agency approach Initially there was reluctance from while also keeping a light footprint. invested heavily in communication some of the implementing partners The design was developed further just and consultation. This always takes to work under the same umbrella. before the construction phase into a time. The WASH and health partners did full corrugated iron model, partly due Starting with just 3 sectors not want to wait for the mapping to donors and partly due to protection enabled a culture of coordination to be process to be completed, and wanted concerns. ingrained. Only once the coordination to implement projects immediately, was working with a few key partners regardless of the output from the Future was it possible to expand to the full planning phase. The Tri-Cluster project was array of humanitarian services. Over a series of meetings, the expanded for 2013 to include importance of coordinating activities education and protection focused was emphasised and a plan was projects. It was planned for an developed where some activities additional shelter agency to join the
SHELTER PROJECTS EAST AND HORN OF AFRICA 24 Site planning for the urban areas of Mogadishu. Different potential plans were shared with focal groups. In the end, row planning was chosen because people could understand it better and could clearly mark the extent of their ‘land’. This would make it easier for people to know what 25 SHELTERbelonged PROJECTS to EAST them AND and HORN help OF to AFRICAavoid conflicts. ConflictSomalia / Complex Excerpt from:Shelter Shelter Projects Projects 2009 2009 8 / A.15A.15
A.15 8Somalia, / A.15 Somalia Puntland - 2009 - 2009 - Puntland, - Conflict Bosaso Case study: Urban shelter upgrade Full case study
Disaster: Project timeline Somalia conflict. Displacement sites in Bosasso and Galkayo. 12 months – – Final report Disaster date: Somalia conflict 1991 onwards. Project implementation 2008 10 months – – Hygeine promotion onwards programmes Population: complete 300,000 people (Bosaso) 200,000 people (Galkayo) Project target population: 7 months – – Settlement planning 4000 households (24,000 and distribution persons), including; 1,450 tents, 1000 shelter kits in IDP settlements (4 in Galkayo and 2 in Bosaso). 500 fire guard steel drums 4 months – – Partner identification Bosaso and local staff distributed in Bosaso training Occupancy rate on handover: Somalia 100% Project cost per shelter: 1 month – – Procurement starts 350 USD per tent 30 USD per shelter kit (wooden poles and ropes) January 2009 – – Project start date – 1992 – – – Conflict starts
Summary To meet the shelter needs of displaced people living in urban temporary settlements in the cities of Galkayo and Bosasso in Somalia, multiple approaches to shelter were used. To reduce risk of fire, fire breaks were made, sites were cleaned up, safe cooking areas were established and stoves were distributed. To meet shelter needs tents were designed and distributed. Additional support was provided in sanitation, hygiene promotion, and the construction of latrines.
Strengths and weaknesses 9 Flexible approaches to shelter were adopted to site layout and water and sanitation programmes. meet local needs. 8 Long term maintenance and support is required to 9 Installing fire-breaks or re-planning sites, supporting ensure that fire breaks remain and sites remain clear of communities to clear refuse in urban settlements, and flammable debris. providing oil drums for cooking in urban settlements 8 Solutions and activities in displacement sites remain proved to be the most effective way securing shelters temporary fixes. against fire. 8 Sanitation remained a significant concern after the 9 Programmes were closely coordinated with other programmes. organisations operating in the cities. - Limited funding availability and challenging security 9 Shelter programmes were closely integrated with made project implementation challenging.
A congested site in Bosaso Photo: Øyvind Nordlie
29 SHELTER PROJECTS EAST AND HORN OF AFRICA 26 A.158 / A.15 Somalia, Puntland - 2009 - Conflict - Urban shelter upgrade Conflict / ComplexConflict
Aftermath of a shelter Fire in Bosaso Photo: Jama Yasin Ibrahim
Puntland context The major populations of There was a lack of garbage other emergencies in Bosaso and displaced people in the Puntland disposal systems. Vector controls, Galkaiyo. Stocks were released after region of Somalia are centred in dustbins, garbage collection points a joint assessment by organisatons dense urban settlements in the city and landfills are almost nonexistent working on shelter programmes. of Galkaiyo and in the port city of in Bosaso and Galkaiyo. Bosaso. Outside these settlements, Settlement Selection many people have also settled in the Technical solutions Criteria for selecting which compounds or on the land of host Tents, plastic sheeting and tra- community to work in were: ditional shelters are not fireproof; families dispersed across the city. • settlements that had received there were insufficient funds to The number of displaced people no assistance before the project, build more solid shelters on a large had been increasing over previous • in Galkaiyo, items were scale, so multiple activities were years and the capacity of agencies distributed to newly displaced required to reduce fire risk. These to provide adequate social services families receiving shelter. activities included: is stretched. Humanitarian access is Sanitation activities focussed on: hindered by insecurity. • clearing sites of flammable • settlements with little or no In Bosaso, the climate is excep- refuse, • establishing fire breaks within sanitation facilities, tionally hot and dry (with annual • settlements where protection rainfall under 100mm), and there sites, and improving planning, • removing the most flammable violations had taken place due are strong winds. Galkaiyo is less to lack of sanitation facilities, windy but still hot and dry. Climate of shelters and replacing them with tents, • settlements where land was combined with overcrowding, available for the construction of poor sanitation and social tensions • establishing fire points, • distributing stoves and cooking latrines at a safe distance from means that large groups of shelters water sources, were frequently destroyed by fire. shields to reduce the risk of fire spreading, • settlements where the Most of the settlements were • when shelters had been community was willing to controlled by gatekeepers who burned, emergency shelter kits participate in the construction insist that the assistance is first containing sticks, ropes and and maintenance of latrines. provided to them, promising that plastic sheeting were distributed Water scarcity in the settlements they will then undertake the ac- in emergencies posed a challenge for improving tivities themselves. There were also A stock of shelter kits was hygiene in the targeted communi- issues with people taking control of built as an immediately available ties. Through the shelter cluster, assets once services were provided. responce to fire outbreaks and the organisation advocated though
30 27 SHELTER PROJECTS EAST AND HORN OF AFRICA ConflictSomalia / Complex Excerpt from:Shelter Shelter Projects Projects 2009 2009 8 / A.15A.15
the WASH Cluster in Nairobi for the interventions and jointly share in- All materials for the erection of provision of water for IDP settle- formation. Lists of targeted benefi- the latrines (cement, timber and ments in Galkaiyo. ciaries and locations were discussed iron sheets for the walling and to ensure that duplication was roofing, used oil drums, paints and Consulting and involving IDP avoided and the maximum number plastic pipes) were procured locally committees from the onset was of people were reached. in Bosaso and Galkaiyo by the local prioritized to improve participation. cooperating partners. They were in- Although this resulted in delays, ex- Local partners were funded to structed to use procurement proce- perience and understanding were erect tents and construct latrines. dures, approved by the internation- gained that sped up the implemen- al organisation that was funding tation as a whole. Logistics and materials The organisation directly the project. Protection concerns were procured community cleaning Shelter kit - Galkaiyo included in all of the stages of kits and distributed them during Kit for reinforcing existing planning and the implementation cleaning and hygiene promotion shelters: of the project. 93% of beneficiaries campaigns. were female-headed households. Material Quantity In total, of the 830 people who Kits of household items and Plastic sheeting 4m x 5m 1 benefited from training, 41% were tents were procured through in- Timber poles (different sizes) 10 female. ternationally advertised tenders, as there were limited stocks available Rope 50m Selection of beneficiaries in local and regional markets. Host communities living within Shelter kit - Bosaso A combination of very poor For famlies whose shelters have the IDP settlements in the same quality materials and strong winds recently been destroyed by fire. living conditions were included in meant that previous tents had very the programme. Vested interests short lifetime. For this reason a Material Quantity from the local and federal authori- frame tent was carefully developed Wooden poles 32 (3- 3.5m ties proved to be challenging as long each) with suppliers. During the process of a result of the huge need of the this tent development, two batches Rope 100m population compared to available of samples were requested, and the Plastic sheet (5m x 4m) 2 resources. final model was signed of during a Sleeping mat 1 Both displaced people and host final visit to the manufacturers in Household items kit 1 family members were included in China. The final detailed specifica- Community cleaning kit the projects. The project targeted: tion was subsequently shared with Each kit for 10 households, 175 other organisations. • disadvantaged anf marginalised distributed in total. displaced people Delays in procurement and • newly arrived displaced people delivery of the tents was a major Materials Quantity (from 2006 and beyond), challenge. Rake 1 • people who had been displaced Wheelbarrow 1 During the programme, a many times within the Spade 1 conflict broke out between the temporary settlements, Garbage disposal drums 1 two administrations in Galkaiyo in • families whose homes are badly Bill boards for public as required December resulting in suspension information or totally destroyed, of activities. • vulnerable members of the local host community living in the periphery of the IDP settlements. Implementation Seven shelter staff were engaged to implement the project and to provide technical support to local partners. There were regular meetings with agencies in Puntland to discuss
Left: Shelter materials distribution. Right: cleaning up a shelter site Photos: Jama Yasin Ibrahim
SHELTER PROJECTS EAST AND HORN OF AFRICA www.shelterprojects.org 31 28 Africa9 / A.8 Shelter Projects 2008 Civil conflictA.8
A.8 Somalia9 /- A.8 2007 Somalia - Civil - 2007conflict - Civil Conflict Resettlement Project type: Resettlement project Support to local authorities in sourcing private land Security of tenure to IDPs and urban poor Provision of extendable one-room shelter Service provision to family plots Emergency: Somalia civil conflict – 1991 onwards (chronic emergency) No. of people displaced: 400,000 IDPs in Somalia before 2007; 1 million in 2008 25,000 IDPs estimated to be in Bossaso Project target population:
140 families; 80% IDPs and 20% urban poor Bossasso Occupancy rate on handover: Puntland 100% of resettled IDPs (112 families); 25% of urban poor (7 of 28 families) Somalia Shelter size 13.5m2 extendable shelter on a 7.5m x 15m plot (including shower and toilet)
Summary A resettlement project in Puntland, Somalia, preceded by in-depth discussions on the concepts of access to land for IDPs and related negotiations on land rights. A consortium of agencies built a serviced community settlement supporting beneficiaries in the construction of extendable single- room houses and providing them with temporary shelters on their new plot.
Project timeline
1991 1998 Dec 2004 June 2005 Sept 2005 Dec 2005 2006 2007 Sept 2007 Dec 2007 I I I I I I I I I
1993 isUN withdrawn peaceleads mission tolapses in mass 1995. Somaliandbegun displacement. civil governmentin conflict autonomy Puntland col Bossaso declaresproposedFirst relocation assessmentassessment site ofShelter in reportBossaso Clusteragreement on producesIDPsPresentation on strategyagreed and forinitialtion with of Roadstakeholders Bossaso map strategy forcontrol implementa remove of Islamicment Mogadishu andSomaliCourts Ethiopian Transitionalfrom armyselection GovernBeginning selection ofFinalisation beneficiary ofcompleted beneficiaryConstruction
- - - Strengths and weaknesses a coordinated system was necessary to limit manipulation 9 Beneficiary involvement in construction increased a of the process by powerful groups. sense of ownership and sometimes meant a higher quality - Although slow, the beneficiary selection process used of workmanship at lower cost compared to contractors. simple and verifiable criteria that ensured that the target Contractors were necessary for some of the infrastructure group was assisted. works. - Donated land does not always guarantee sufficient 9 Announcing the outcomes of meetings publicly was one quantity or quality of land. As a result, an integrated urban way of avoiding a breakdown of communication with local development plan can be hard to develop. Assessment of authorities and ensuring transparency. land suitability and direct discussions with private landlords - In a place without clear land policies, laws or record to clarify donation conditions are necessary before land is systems, land issues were complicated and sensitive and formally donated to the municipality. required careful investigations, localised responses and 8 Working with the municipality was difficult, due to public awareness-raising through mass media and meetings. its low capacity, high turnover of staff and overlapping - Joint planning and implementation by agencies through responsibilities with the clan system. Documenting
23 29 SHELTER PROJECTS EAST AND HORN OF AFRICA A.8 Somalia SomaliaExcerpt - 2007 from: - Civil Shelter conflict Projects - Resettlement 2008 9 / A.8Africa Strengths and weaknesses (continued) decisions and agreements made was of little help due to 8 At times not all the humanitarian agencies involved literacy issues. communicated the same messages. This meant that 8 Selection of beneficiaries took much longer than agreements sometimes had to be renegotiated. expected, so some construction work began before knowing who the final beneficiaries would be. This limited participation. Photos: Ombretta Tempra
Sites and services: the project focused on negotiating land and providing access, secure compound walls, water and sanitation. Background Bossaso is a coastal town in the The project indirectly tackled gov- agency to improve livelihood oppor- Puntland region of northern Somalia. ernance-related issues relating to land, tunities for beneficiaries and promote Puntland, with 2.8 million mostly and broader urban development and peace between the IDPs and the host nomadic/pastoralist inhabitants, has city planning issues. Approaching these population. been semi-autonomous since 1998. issues directly, without a clearly visible A committee was established to Due to its relative stability, it has project, would have been difficult. identify land within the current urban become an attractive area for IDPs Implementation growth areas. During Ramadan, calls fleeing conflict in South Central Negotiations with authorities for were made for land donations. Five of Somalia. accepting the permanent resettlement the offered sites were selected and of- Bossaso has a significant popula- of IDPs within Bossaso, and the relative ficially handed over to the municipality. tion of IDPs, many of whom had been provision of suitable land, began in The land transfer was endorsed by the present for over ten years. The liveli- 2004. These negotiations were linked sharia court in December 2005. hood opportunities created by the fast- to a joint UN strategy for IDPs in With no clear legal framework developing port of Bossaso is a strong Somalia published in 2005. in Puntland, customised ‘letters of pull factor, along with some IDPs’ clan Once a strategy for Bossaso had allotment’ had to be developed to sub- connections in the area. been agreed upon between humanitar- stitute for an ‘ownership title’. Benefici- There is no land administration and ian agencies and Bossaso authorities, aries received the right of occupation, there are few documentary records, so the project began in 2006. It was imple- use and inheritance for the first 15 customary law, secular law and sharia mented by a consortium of agencies, years. After this, each family would also law all overlap. all of whom were represented in the acquire the right of disposal (selling the property for profit). For the document Aim of the project newly established Somalia Shelter Cluster. to provide the strongest protection for The idea of the permanent reset- IDP tenure, it was signed by the benefi- tlement project was to substantially Construction was completed by the ciary, the mayor, the minister of local improve IDP protection, security of end of 2007. The duration of the project government and the magistrate of the tenure, access to basic services and in- was longer than initially envisaged, due sharia court. frastructure (especially water and sani- to difficulties in obtaining land, a long tation), and to provide a solid base for beneficiary selection process and the Selection of beneficiaries income-generating activities (renting challenges of maintaining consensus Beneficiary selection took longer out rooms, space for shops or pro- with a relatively unstable and inexperi- than planned. Some 80% of plots ductive activities), in addition to the enced local government system. were to be allocated to IDPs and provision of a better shelter. 20% to poor families from the host Land issues community. This approach limited Freeing IDPs from paying rent for The original site proposed by the the interest of powerful members of inadequate shelter meant that they local authorities was rejected on the the host community from exerting could use resources for basic services, grounds that it was too far from the too much influence in the selection such as education and health. Con- town and limited economic integration of IDP beneficiaries. Post-occupancy struction training would provide ben- of the IDPs with the host community. assessments found that few of the eficiaries with new skills. This was a key requirement by the urban poor beneficiaries in the project 24
24SHELTER PROJECTS EAST AND HORN OF AFRICA 30 Africa Shelter Projects 2008 A.8 9 / A.8 Civil conflict Photo: J. Ashmore Many of the sites initially offered were rejected because they were far from Bossaso and possible livelihoods. The five selected sites were donated following requests for land made during Ramadan. occupied their site, preferring instead Contractors were used to trace The cost for the second phase to rent out the new accommodation roads for the new settlement and was US$ 580 per housing unit. This or leave the house empty, while the connect it to the municipal water excluded agency staff costs and food- occupancy rate on project completion supply. This also benefited those living for-work contributions but included from IDP families was 100%. along the route of the new water all other logistics, administrative and pipes. An ongoing solar-powered material costs. The selection process, managed street lighting project was also started by the multi-representative Bossaso Logistics and materials towards the end of the project. selection committee, began in Materials were procured locally, September 2006. IDP beneficiar- A plot was provided within the set- with contractors responsible for their ies were selected by April 2007, but tlement for each family to construct own procurement. agreement on urban poor beneficiaries their own house, with support from Bill of quantities was not reached until November 2007. the consortium. The following table shows the bill Before selecting individual families, Two different agencies implement- of quantities for Phase 2 of the project, the IDP settlements with the worst ed the construction of the 140 housing averaged for a single unit (some units shelter conditions were identified. units in two phases using contractors. were corner units rather than free- Selection committees were formed in The first phase took five months and standing). each of these settlements and were involved the construction of foun- tasked with putting forward individual dations, boundary walls, sanitation Materials Quantity households who had lived in Bossaso (shower, toilets and septic pits) and a Hollow concrete blocks 281 pieces for more than six years, with no fixed 4.5m x 3m floor slab. Phase 1 cost US$ (150 mm x 390 mm x 180 mm) assets and at least three children. More 1,850 per housing unit. detailed ‘vulnerability’ criteria were Cement for mortar and 5 bags The second phase began after concrete ring beam rejected due to the complexity of beneficiary selection was complete Somali family structure and the lack of Sand for mortar and concrete 1 tonne and took three months, finishing in ring beam identification documents. December 2007. The beneficiary Aggregate / ballast for ring 0 tonnes Selection lists were made public families moved onto their plot, living in beam concrete to allow time for complaints to be a temporary tent-like shelter provided Y8 bars (12m long) for ring 4 pieces investigated (one of the settlements by another agency until the work was beam produced a list that excluded an ethnic completed. The temporary shelters R6 rings (6m long) for ring 2 pieces minority). The final selection of the were later used as additional rooms or beam 112 IDP families was made through for storage. 6x1 white wood for form 12 metres a lottery broadcast on local TV and work Food for work for a maximum of radio, which was deemed a fair method 28-gauge galvanised 14 pieces 30 days was provided to beneficiar- corrugated iron sheets by beneficiaries. ies for the construction period, along Structural grade 150 x 50 18 m ‘Compared to the shelter with US$ 30 to hire a mason (families (6' x 2') timber roof rafters sometimes did masonry work them- Structural Grade 75 x 50 27 m I had before, I can now (3' x 2') timber roof purlins say that my life has im- selves, with technical support, and kept proved 100 percent. The the money). The main agency provided Roofing nails 1 kg technical support in the form of cash Ordinary wire nails 1 kg resettlement programme for skilled labour and employment of a was completely transpar- Steel single doors 1 set ent and well done’. foreman for supervision. (0.8m x 2m) Giving the families the oppor- Double leaf-steel window 1 set tunity to select their own mason (1m x 1m ) Technical solutions (rather than following the wishes of White wash 4 bags This project provided the infra- the local authorities who wanted the Brushes for whitewashing 2 structure for a serviced community whole construction process contract- Bamboo/rope for ceiling mats As settlement, well integrated with the ed out) meant that they had greater required host population, and support to IDPs quality control over the work done for the building of individual dwellings and allowed the agency to avoid the within the settlement. problems of a tendering process.
25 31 www.shelterprojects.org SHELTER PROJECTS EAST AND HORN OF AFRICA South Sudan Excerpt from: Shelter Projects 2015–2016 10 / A.25
CASE STUDY 10 / A.25 SOUTH SUDAN 2014-2016 / COMPLEX
KEYWORDS: Emergency shelter, Site planning, Phased construction, Infrastructure, Planned camps
South Sudan Civil War, CRISIS Dec 2013 - ongoing. Complex crisis SUDAN BENTIU 6.1 million in need of humanitarian TOTAL PEOPLE assistance and 1.66 million internally displaced, AFFECTED 1 ABYEI UPPER as of December 2015 . NILE For more updated figures, see overview A.23. NORTHERN BAHR EL GHAZAL UNITY Bentiu, Protection of Civilians (PoC) WARRAP PROJECT LOCATIONS ETHIOPIA site, Unity State. WESTERN BAHR EL JONGLEI GHAZAL 105,786 people (47% male; 53% female; CENTRAL LAKES BENEFICIARIES AFRICAN with 47% under five years old), relocated across REPUBLIC communal shelters, at 45 people per shelter. WESTERN EASTERN EQUATORIA EQUATORIA CENTRAL PROJECT OUTPUTS 11,778 robust shelters. EQUATORIA KENYA 2 DEMOCRATIC REPUBLIC 84m (4.5x21m communal shelters, with OF THE CONGO UGANDA SHELTER SIZE partitions to accommodate between 35 and 55 PROJECT SITE people in groups of 7 to 11 individuals).
1.5m2 at peak. Shelter occupancy has been PROJECT SUMMARY S AN SHELTER DENSITY variable due to space constraints, with huge 198,448 The project constructed 11,778 shelters in the Protec- Upper Nile
! ! ! ! ! ! ! ! ! ! ! ! ! ! !
! influx in PoC caused by repeated insecurity. !
!
!
!
!
! ! Abyei region
!
! !
tion of Civilians site in Bentiu. The project was closely ! 101,203 134,435 !
!
!
! Malakal
! ! !
! !. 299,084
! !
! ! ! ! ! ! ! ! ! 9,403 RIO NAPO ! Unity linked with the phasing of a broader USD 18 million !. MATERIALS COST USD 837 Bentiu Aweil !. 541,395 PER SHELTER project of site works, which converted a camp that Northern Bahr (Materials: USD 687, Labour: USD 150 approx.). el Ghazal ETHIO IA seasonably flooded into a habitable site. !. Kuajok Wau Jonglei 1 225,321 South Sudan Humanitarian Needs Overview 2016, http://bit.ly/2d3Y2tB. !. 1,924 Western Baher el Ghazal CENTRA ! ! AFRICAN 502,209 2,339 24,817 !. Rumbek RE IC Lakes
129,964 !. Bor !
Western Equatoria 15 DEC JUL 93,276 2013 2014 2015 2016 2017 Eastern Equatoria Yambio 9,232 Central !^JUBA !. Equatoria 4,654 !. Torit 1 2 1’ 2’ 3’ 3 4 4’ 5 5’ 6 18,492 53,415 E OCRATIC EN A RE IC OF THE CON O 172,200 48,921 FLOODS CONFLICT AN A PROJECT (OVERALL)
TIMELINE PLANNING IMPLEMENTATION PHASE 1 IMPLEMENTATION PHASE 2
JAN SEP OCT JAN FEB MAR APR JAN APR
1 May 2014: Population in Bentiu PoC: 8,000 individuals. 2’ Feb 2015: Site redevelopment begins to reduce overcrowding and provide adequate drainage, addressing the flooding risk. 2 Sep 2014: Population in Bentiu PoC: 46,000 individuals. Apr 2015: Implementation phase begins with a two-months delay 3 3’ Jul 2015: Population in Bentiu PoC: 87,000 individuals. (due to negotiations with UNMISS regarding usage of the space), and as a result of community resistance to being relocated to the 4 Oct 2015: Population in Bentiu PoC reaches 120,000 individuals. new site within the PoC.
5 Jul 2016: Population in Bentiu PoC: 102,000 individuals. 4’ Jun 2016: Site development gradually completed in a phased ap- proach, with sectors/blocks handed over to the partner NGO as the 6 Dec 2016: Population in Bentiu PoC: 120,000 individuals. site works ended.
Jan 2015: Aug 2016: 1’ Robust emergency shelter design agreed upon, and ap- 5’ Phase 2 of shelter construction completed (though on- proved by the community. going, as new arrivals continue and reinforcement is done).
WEAKNESSES STRENGTHS - Delays due to logistics and weather constraints. + Provided shelter secure from violence and localized flooding. - Assistance was provided only within the site, causing disparities + Effective coordination between all actors. with the populations outside. + Strong forward-planning for procurement and implementation. - Overcrowding in shelters. + Use of local materials where possible. - Issues in timber procurement and poor market analysis. + Enhanced cladding with grass to improve comfort and durability. - Lack of partitions in the initial design.
SHELTER PROJECTS EAST AND HORN OF AFRICA 32 10 / A.25 Complex ©Muse Mohammed Although it is widely recognized that camps are an option of last resort, for tens of thousands of residents in Bentiu PoC, conflict meant that there was no other option. However, the site was too small and would flood every year. This required massive expansion and infrastructural works.
BACKGROUND For more information on the context and the shelter-NFI re- sponse in South Sudan, see overview A.23. Before the outbreak of conflict in 2013, the bases of peacekeep- ing forces – United Nations Mission in South Sudan (UNMISS) – had hosted small populations seeking protection for short peri- ©IOM south sudan ods, with limited humanitarian response. Following the outbreak of conflict, tens of thousands of people fled to – and stayed in – Protection of Civilians (PoC) sites far longer than expected. Over the course of the conflict, multiple waves of violence af- fected the town of Bentiu, leading to 120,000 people seeking shelter in the PoC site. Bentiu is extremely hard to access, with a small airstrip of limited capacity, and is inaccessible by road during the rainy season. During the dry season, it is regularly cut off, due to poor security. As a result, all logistics and sup- plies had to be planned in advance of the wet season, and plans needed to be flexible, to allow for this variable security context. Satellite view of Bentiu PoC and some planned expansions. Humanitarians arrived in Bentiu in January 2014, to provide essential, life-saving, services to the population residing there. In March 2014, the PoC site in Bentiu hosted 11,000 IDPs, with the population rapidly rising to 43,718 by December 2014 as a result of escalated conflict in Unity State. The huge influx- es created overcrowding and difficulties in service provision. lished, based on the Dutch “polder” system. Major works (with 74 pieces of heavy machinery) led to the establishment of a In the rainy season of 2014 the site flooded for several 4m tall berm (mainly for security purposes) and 24m2 section months, leaving the camp population trapped, with many drainage ditch around the site. This was to prevent surface run- parts of the site deep in water. By mid-2014, living space was 2 off from the surrounding land. Additionally, a series of drainage limited to 9m per person across the site. Overcrowding was ditches and water retention basins were dug. These had large compounded by stagnant water, which worsened living condi- capacity pumps, to remove rainfall from inside the berm. tions and exacerbated the risk of water-borne diseases, such as cholera. The site itself remained highly insecure, with reg- The site works were achieved through contractors and a care- ular violence outside the PoC – and at times inside, due to fully phased construction plan. This plan allowed for addition- ethnic conflict – leading to fatalities throughout the project. al timing for contingencies and monitored the volumes of soil moved, as well as the length of drainage ditches and berms. SITE WORKS Beyond the major site works, the site development project in- To respond to the growing site population and address the is- cluded shelter construction, establishment of water, sanitation sues of localized flooding, during 2015 and 2016, the Bentiu and hygiene systems, health and education facilities, alongside PoC was expanded and rehabilitated over 1.68 million m2 (168 other services. Given that the site was already occupied, agen- hectares). To create better living conditions for people seeking cies needed to work together to ensure carefully phased re- shelter in the site, a massive drainage network was estab- location. Shelters, latrines and other structures could not be
33 SHELTER PROJECTS EAST AND HORN OF AFRICA South Sudan Excerpt from: Shelter Projects 2015–2016 10 / A.25 © Muse Mohammed The expansion plan included land that was already spontaneously occupied by camp residents. Careful phasing of major upgrading works was required, as the whole site needed to be upgraded. Phase 0: survey and demining Phase 1
Phase 0: survey and de-mining. Phase 1: Access to site works Phase 2: Internal access and Phase 3: block development Phase 4: block development, and perimeter. drainage. and relocations. relocation and completion. erected until ground works were ready and, if they were built COMMUNITY ENGAGEMENT before people were relocated, they risked falling into disre- Close engagement with the community leadership was crit- pair, or being looted. ical for maintaining the ability to operate safely in the camp. It was also essential to enable safe and phased relocations GROWING SITE POPULATION within the site, as new shelters were built. The site was designed for 50,000 people with a contingency of up to 75,000 people. As the site population continued to rise, PHASING AND COORDINATION reaching over 87,000 people by July 2015, revisions to site As people were already occupying the site, a phased reloca- and shelter plans were necessary. In the first phase, there was tion process allowed site works to continue, according to an significant community resistance to the programme, as the overarching project plan. The site was split into sectors and population influx meant that the number of people per shelter each sector was moved as the ground works were finished had to be increased from five to eight. In 2016, this increased and shelter frames erected. further to 11, as the population increased to over 120,000. Relocation could only take place once plots for families and IMPLEMENTING TEAM STRUCTURE communities had been established, shelter materials had been distributed and construction was completed. Given the The lead organization for the site sub-granted to a part- limited space, some sectors had to be moved to newly ren- ner NGO for the shelter activities. The implementing NGO ovated plots before all of the land could be worked on. This had a Shelter Programme Manager and a Shelter Advisor, made the timing of different activities for the entire site recon- and was supported by the lead organization by two deploy- struction project interdependent and highly time critical. ments of Shelter Cluster rapid response officers. The project also included an implementation and management team with On 21 May 2015, the camp management agency coordinat- functions such as quality control, cross-sectoral coordination ed 160 humanitarian workers in a population verification ex- and information management. In addition to project staff, the ercise, recording biometric details and assigning addresses project implementation team included around 200 camp res- within new areas. Verification was an important first step and idents, who were chosen by the community leadership and helped in demarcating plots and defining movement plans. trained by the organization on shelter design and construc- Overall, UNMISS, peacekeepers, humanitarians and the au- tion. The construction of shelters was phased employing six thorities had to negotiate between each other and coordinate different teams (including plot demarcation, digging, erecting closely in a very complex military environment and in in- skeletons and spraying walls). credibly harsh conditions, including shrinking humanitarian Technical supervisors and contractors were recruited by the access and a protracted conflict situation. partner NGO within the PoC sites, with each of the contractors further recruiting a team of labourers to build shelter frames.
SHELTER PROJECTS EAST AND HORN OF AFRICA 34 10 / A.25 Complex © Muse Mohammed The shelter project built communal shelters due to lack of land and nationally limited resources. These shelters allowed to maximize the use of limited space and impacted shelter strategies throughout the country.
CONSTRUCTION PROCESS Shelter frames were built by contractors and guards were hired to protect the shelter frames from theft, until they were allocated to a household. Once households had been allocated a shelter plot by the organization (in coordination with camp management agencies), they collected a shelter kit from the implementing partner NGO to complete their shelter. Demonstration shelters were built as prototypes and the partner NGO provided techni- cal supervision to households to ensure that the materials were used effectively. For example, care was taken to ensure that plastic sheets were attached correctly. Individuals with identi- © Brendan Bannon fied vulnerabilities, such as disabled persons, pregnant women Shelter frames were built by contractors. and the elderly, were provided additional assistance. A timber workshop was set up at the logistics base in the UNMISS site THE SITE IN THE LONGER TERM with outdoor storage for 3,000m3 of timber. At the workshop, Relative stability in the first half of 2016 and the expansion teams prepared the timber for the structures of the shelters, of humanitarian services to wider Unity State led to a net re- including treating them with anti-termite solution. duction in the number of people in the PoC site. However, a resumption in hostilities following the July 2016 crisis led to a SHELTER DESIGN population increase in Bentiu PoC (as of 31 December 2016, The shelter design was discussed with the Technical Working the population was 119,853 individuals). The sustainability of Group in Bentiu and the national Shelter-NFI Cluster before be- this and other PoC sites has been object of debate, due to ing presented to communities. Local adaptations included the the limited resources, the protracted nature of the crisis and use of elephant grass, which could be harvested by women the need of displaced populations for long-term assistance. residing in the site. The windows and doors were also revised to be based on traditional local designs. The shelter design had WIDER IMPACTS OF THE PROJECT an estimated life-span of one year, providing displaced house- Humanitarians have been running similar sets of projects in holds with a solution that is significantly more sustainable than other PoC sites, such as in Malakal, where the organiza- standard emergency shelters built in the country by humani- tion has been redeveloping and rehabilitating the PoC site tarians. The design was inspired by the local summer housing throughout 2015 and 2016. The shelter partner in that site solution known as Rakuba. has applied the communal shelter design and aimed to en- In 2016, concerns were raised by the community about security sure the continued provision of essential emergency shelter in the site and the security of shelters. As a result, the partner services through distributing shelter kits, repairing damaged NGO started the process of providing doors to shelters which communal shelters when required and providing assistance did not have one, starting with the most vulnerable, as identified to people with special needs to construct shelters. by protection partners. The implementation of activities across the country has been To protect from water coming in, it was initially planned to use in line with the Shelter-NFI Cluster objectives and humanitar- sand to raise the floors of the shelters, but this proved impossi- ian best practices, including lessons learned in Bentiu. ble to procure. Households were therefore encouraged to use Through regular monitoring and technical guidance, human- white soil to raise their floors instead. itarian shelter teams have been working to help residents construct their shelters in more durable ways.
35 SHELTER PROJECTS EAST AND HORN OF AFRICA South Sudan Excerpt from: Shelter Projects 2015–2016 10 / A.25
STRENGTHS, WEAKNESSES AND LESSONS LEARNED
STRENGTHS + The project provided (relatively) secure shelter from vio- lence and localized flooding. + Coordination between all actors was key to the success of such a large-scale programme, which required careful phasing within many constraints. + Strong, forward-planning regarding required supplies helped the project team mitigate extreme weather variability © Muse Mohammed and the lack of transport infrastructure. This enabled over 1,000 The site works were based on a Dutch “polder” system. They included 28m2 units to be constructed per week, at the height of the reloca- section drainage ditches, berms, water retention basins, and large volume tion process. pumps to evacuate water.
+ Wherever possible, local materials were used. 84,000 MATERIALS LIST FOR ONE COMMUNAL SHELTER bundles of elephant grass, bamboo and garang rope were pro- Unit cost Total cost Material Unit Quantity cured. The local elephant grass was procured from women (USD) (USD) over a period of two weeks, through a large community-mo- Plastic sheet Piece 15 8 120 bilization campaign. (4x5) + The plastic sheet cladding was enhanced with grass to Rubber binding Bundle 5 20 100 improve insulation and extend the lifespan of plastic sheets. rope Bundle Bamboo poles 5.5 10 55 of 10 WEAKNESSES Timbers 3 x 2" Piece 4 28 112 - Activities were delayed by approximately eight weeks x 3m compared to the proposed work plan. This was primarily due Timbers 2 x 2 Piece 6 10.25 61.5 to logistics and weather constraints. x 4m Timbers 2 x 2 The site became the only significant location where as- Piece 4.5 12 54 - x 3m sistance at scale could be provided in the state. This caused Timber 4 x 2 disparities between the assistance provided to those living Piece 11 4.25 46.75 in the PoC and those outside and was one of the causes of x 5m Timber 3 x 1" x population growth of the site. Piece 3.5 8 28 3m bracings - The site became very crowded and shelters were rela- Nylon Rope 30m Roll 8 5 40 tively small. Although the reasons for the lack of space were unavoidable (both political and financial), the overall density Nails 4" Kg 2 2.5 5 was higher than desirable. Nails 3" Kg 2 5 10 Nails (roofing) Kg 3 1.5 4.5 Anti-termite and Piece 10 5 50 wood borer LEARNINGS • The project demonstrated the value of early collaboration and planning, particularly in such a complex and chal- lenging environment. While shelter activities in 2014 were constrained significantly as a result of a lack of dry space and logistical challenges, the convening of stakeholders and the establishment of a technical working group to plan the redevelopment project in September 2014, as well as the relatively timely procurement of materials during the dry season logistical window, ultimately ensured the success of the project. • Shelter designs that are meant to accommodate households beyond an acute emergency phase should take into account privacy considerations and install partitions. The communal shelters were initially built without partitions, as the shelter approach was based on individuals-per-shelter (and not households). This was mainly a result of limited space available and the increasing population in the camp. • For such large projects, it is important to have a proper market analysis and adopt a design that suits locally available materials. Not enough consideration went into the procurement of timber, nor its potential environmental impact. With a non-functional timber market, non-standardized sizes and right species available, it was difficult for the supplier to keep up the demand; compounded by its limited understanding of the requirements, as well as access to appropriate tools and workshops to provide desired sizes.
SHELTER PROJECTS EAST AND HORN OF AFRICA www.shelterprojects.org 36 11 / A.29 Conflict
11 / A.29 Republic of South Sudan – 2011 – Conflict Keywords: Returns, Resettlement, Construction materials, Core housing construction, Site Update: planning, Infrastructure, Training.
Country: Project timeline Republic of South Sudan Conflict: 17 months – ––Project completion Post-war reconstruction 16 months – ––8,300 shelters Conflict date: completed 1983 to 2005 Number of people displaced: 2,000,000 Project target population: 13 months – ––Compressed mud 70,000 (includes beneficiaries of blocks introduced quick impact projects) 12 months– ––6,800 shelters complete Project outputs: 8,300 shelters 2,200: Compressed mud blocks 6,100: Bamboo / wattle and daub 8 months – ––First construction phase Occupancy rate on handover: 95 per cent Shelter size: 2 5 months – ––Procurement and 16 m - up to four people delivery of shelter 2 24 m - five people or more Republic of building material Materials cost per shelter: RepublicSouth Sudan of 3 months – ––Agreement with partners US$ 400 - 600: poles and bamboo South Sudan US$ 800 - 1100: compressed mud March 2011 – ––Project start blocks Labour: US$ 260 Average: US$ 1,100 Project cost per shelter: US$ 600-1,200 2005 – ––Peace agreement
Project description The project supported reintegration of returnees. It constructed 8,300 shelters on new land plots provided by the government. Basic urban services such as school buildings and boreholes, were constructed through parallel programmes. Two shelter designs were employed: bamboo and thatched-roof shelters (6,800) that could be built quickly to respond to large-scale returns and compressed mud block shelters with CGI sheet roofs (1,500) to provide more durable structures.
Strengths and weaknesses + Communities participated in the selection of vulnerable highly-skilled lead builder. In some early cases, skills were households and in designing shelters. lacking and build quality was poor. + Good coordination prevented returnees from being sited - Due to unexpectedly slow block production, the number in areas too far from transport or services. of mud block shelters was cut by 800. + Shelter construction was linked to projects to deliver - Plans to use drainage activities to supply the mud basic services and livelihood opportunities. required for blocks failed due to the lack of organisation at + The project was able to respond to input from authorities the community level. and change the shelter design. - The project was too big and created unsustainable + Training of affected populations improved their demands for materials, leading to concerns over the construction skills. destruction of national forests. + Partners were required to submit phased progress • Compressed mud-blocks needed to be plastered with reports for each household to keep the project on burnt oil, sandy soil and Arabic gum. schedule. • As the compressed mud-block technique was new to - Communities demanded incentives for their involvement some areas, its performance over time remains untested. in the construction phase. - The target number of shelters was reduced by 35 per cent due to rising costs and delays in block production. - Construction using compressed mud blocks required a
37 SHELTER PROJECTS EAST AND HORN OF AFRICA South Sudan Excerpt from: Shelter Projects 2011–2012 11 / A.29
Approximately 25 per cent of the shelters were built using compressed soil blocks. The technique represented a financial and environmental improvement, but was slower, requiring significant efforts to introduce as a new technology. Photo: Fernando Murillo
Before the conflict technical and two administrative per cent of shelters were constructed Sudan became independent from staff), and implemented by partner for families from the host community. the Republic of the Sudan in 2011 NGOs and community organisations. Associated projects such as following two civil wars (1955-1972 Construction teams were made up borehole and school construction and 1983-2005). The Comprehensive of nine people, including engineers, benefitted both groups. Land alloca- Peace Agreement in 2005 signalled construction supervisors, masons and tion was made through a government the beginning of a return process. carpenters. lottery process. In 2011 Sudan, (north and south Materials were procured by the Households with special needs combined) had an Human Devel- main agency on behalf of the partners. had their veranda, kitchen or oven opment index of 0.408 placing it The materials were distributed as built for them. in the “low human development” self-construction kits. Experienced group. South Sudan is relatively less masons and carpenters were identified Coordination developed than the north and faces to provide “on-the-job” construction Coordination was critical since considerable challenges in terms training for young people from both so many actors were involved. The of infrastructure development and the returnee and host communities. coordinating agency not only had to poverty reduction, with many people Construction progress was ensure coordination within the project unable to access social services or monitored by giving each shelter one in terms of working with implementing education. of four statuses: partners but also had to work closely • To be done: Beneficiaries not yet with national and state authorities After the conflict identified who were developing their planning The conflict between The Republic • In progress: Beneficiaries and building regulations from scratch. identified and land title received of Sudan and South Sudan stunted Despite many delays the land allo- • Under construction: Structure and development in the South and most cation was completed in time for the roofing complete returnees had no shelter or land to • Finished: Beneficiaries have shelters to be constructed. return to. moved in. Beneficiaries and host communi- 2011 marked the peak in return ties were also involved in prioritising as it coincided with the deadline the type of quick-impact projects to be for southern Sudanese to leave Selection of beneficiaries implemented. Khartoum, where the majority of IDPs Project areas were determined by had fled to during the war. There was the agency in collaboration with the Hazards also a significant return of the diaspora Ministry of Humanitarian Affairs. The There were a number of site in neighbouring countries, Europe and shelters were distributed according hazards, including severe flooding, the USA. to the proportion of returnees in each that prevented access to some areas. county. Introduction of significantly stronger Implementation Individual beneficiaries were compressed mud block foundations The project built 8,300 shelters selected jointly by the implementing helped to mitigate the flood risk in (6,800 in 2011 and 1,500 in 2012) and partner agencies and the government. shelters. Beneficiaries with technical more than 42 community buildings Criteria included households that were supervision, voluntarily dug site (mostly schools) across the 10 states headed by children or women, house- drainage channels to reduce flooding of South Sudan. Land was allocated holds with individuals with disabilities risks. by the Ministry of Housing and and those who had no visible means Physical Planning. to support the construction of their Technical solutions The project also implemented own shelter. Beneficiary lists were Shelters had a single slope for quick-impact projects and livelihood then verified by the main agency’s the roof to improve water harvesting. schemes. field staff. This design was replicated by other The project was coordinated by The beneficiaries came mostly returnees who were not beneficiar- an international agency (with two from the returnee community but 10 ies of the programme. A small water
SHELTER PROJECTS EAST AND HORN OF AFRICA 38 11 / A.29 Conflict
”Return back home is easy when someone supports you to build your shelter.”
Beneficiary, Central Equatoria State.
New settlement, in Central Equatoria state, showing a bamboo, wattle and daub shelter (far left), and compressed mud blocks (right). Photo: Fernando Murillo
tank, that could later be upgraded by were too expensive for domestic In the first year of using com- homeowners, was provided with every purposes. pressed blocks, 500 fewer shelters shelter. Using the same press, and mostly than planned were built, and the The shelter could be expanded black cotton soil, it was possible to project had to return to the bamboo with a veranda and an external kitchen make compressed mud blocks without design instead. to reduce the health risks of smoke a cement stabiliser. from cooking indoors. It was possible to produce 400 Logistics Sample shelters were built for the compressed blocks a day. While the Bamboo and compressed mud community to examine and comment technique is slower than traditional blocks were procured or produced on. Following feedback, shelters were mud brick production (1,000 per day) locally. Plastic sheeting and ironmon- plastered with burnt oil, Arabic gum it used much less water. gery were imported. and sandy soil. The government was positive and Different foundation designs were felt that the technique created a new developed for different ground condi- type of industry. Materials list tions. In poor soil areas, wider founda- Mud-blocks were less prone Materials Quantity tions were built on top of large stones. to attack by insects compared to CSB (mud) blocks 414 bamboo, and enabled construction of (foundation) 15m Bamboo model strong, load-bearing walls. They were Polythene sheet (1m wide) 1034 Initially, shelters were built using cool by day and warm by night, and CSB (mud) blocks-walls/ 8 pieces poles and bamboo wattle and daub did not have to be transported over columns Corrguated iron 4.2 pieces walls. These were relatively quick to long distances. sheets x 4m Timber 125mm x 2 pieces build but required significant procure- The project also demonstrated to 50mm x 4m 4 pieces ment of timber and bamboo. each community how blocks could be Timber 100mm x 50mm x 4m 11 pieces Bamboo-based structures used for energy efficient ovens. Timber 100mm x 50mm x 4m 4 pieces required “mudding” to complete and The introduction of compressed Timber 75x50mm x 4m 2.5 pieces seal the walls. In a number of cases mud-blocks in 2012 resulted in Timber 50x50mm x 4m 20m beneficiaries used plastic sheeting for different reactions from communities. Galvanized drainage zinc 2m 2kg walling instead. In some areas, people already Hoop Iron (50m roll) 2kg Shelter costs rose during con- built using dried mud-blocks. In other Nails 4" 1kg struction due to rising bamboo prices areas the technique was new. In some Nails 3" 2kg and unplanned transport costs of soil cases there was resistance to the use Nails 2.5" 5+1 and water for mudding. of the blocks, as production involved Galvinized spiral roofing nails pieces Due to the local environmental considerable heavy labour. The intro- 3" Hinges and bolts 1 piece impacts of using timber, and new duction of the block presses and the Chicken wire 2 Bags conditions set by the government to realisation that mud-blocks were a Cement (plastering) (1/6 protect timber sources, it was decided relatively efficient material in terms of cement/soil) 1m3 to switch away from these materials. water use, led to a more positive view Soil/sand for mortar 2 litres of the mud-blocks. Anti termite treatment Compressed mud blocks The holes left behind by the pro- Government representatives were duction of mud blocks were an issue aware of a project in the Republic in some areas, and more effort could of Sudan which used stabilised soil have been made to combine drainage blocks (SSB) and expressed an digging with mud block production to interest in this alternative. SSBs had facilitate a more efficient use of both been used for public buildings but labour and soil.
39 www.shelterprojects.org SHELTER PROJECTS EAST AND HORN OF AFRICA A.9 Sudan Darfur - 2004Excerpt (ongoing) from: -Shelter Conflict Projects - Materials 2008 distribution 12 / A.9Africa
A.9 Darfur12 -/ A.92004 Darfur (ongoing) - 2004 -(ongoing) Conflict - Conflict Materials distribution
Project type: Darfur shelter materials pipeline Multi-agency common logistics system Distribution of shelter materials and non-food items Emergency: Response to displacement caused by violent conflict in Darfur, Sudan, 2004 (ongoing) No. of people displaced: Over 1 million people affected (May 2004 estimates); 700,000 people internally displaced Project target population: 1 million people initially (167,000 families) Sudan Increased to 1.4 million people in September 2004
Occupancy rate on handover: Darfur 80% of target population reached by December 2004 A further 8% of beneficiaries reached by those not operating within the pipeline Shelter size Maximum of 20m2 of covered space for a family of six people, provided by a 4m x 5m plastic sheet. Actual covered space would have been less, due to the need to fix and fold the sheeting.
Summary A joint distribution mechanism, which would later include joint procurement, was set up by a consortium of NGOs and UN agencies to standardise the procurement and distribution of basic shelter materials to those displaced by the conflict.
Project timeline
Apr 2003 Sept 2003 Dec 2003 Apr 2004 May 2004 Aug 2004 Sept 2004 Dec 2004 I I I I I I
DarfurChad toRefugees escape beginviolenceholds arriving andfor in shortSudanese inCeasefire time government betweendown SLA Darfur completely ceasefireties allocated approvedbreaksPipeline and responsibiliconceptthe pipeline paperFirst distributionis donorsarrives Appeals in Nyala throughupdated ProcessProject Consolidated (CAP) is included to distribution in thereceived 70% the of firstbeneficiaries peopleroundincreased of hadTarget to 1,400,000populationfirst round aries had 80%of distributionreceived of benefici the-
-
Strengths and weaknesses easily obtained by the UN than NGOs, so collective 9 More effective use of overall emergency funds due to purchase reduced customs obstacles. economies of scale. 9 Common purchasing meant greater ability to direct 9 Reducing logistics overheads for individual humanitarian resources based on a broad overview of need in the region. agencies in the field meant they could concentrate on 9 Centralising stock in a warehouse outside of the conflict service delivery to beneficiaries. area meant that goods were not tied up in warehouses 9 Reduced competition for transport and warehousing in the ‘wrong’ areas, so interruptions to distribution by among agencies limited price inflation. security problems were minimised. 9 Customs clearance from the government was more 9 Early donor commitment of funds and air cargo allowed 26
SHELTER26 PROJECTS EAST AND HORN OF AFRICA 40 Africa12 / A.9 Shelter Projects 2008 Conflict A.9
Strengths and weaknesses (continued) the project to move quickly. Donor coordination meant 8 Without the distribution of a frame, the distribution funding went through a central channel, avoiding duplication of plastic sheeting did not constitute a complete ‘shelter of projects. solution’. If not provided by an NGO working in the field, 9 Individual agencies in the field augmented the provision IDPs had to provide these items themselves, which carried of shelter items with their own locally procured materials risks in terms of collecting material from unsafe areas or (such as poles) to provide shelter solutions. having to buy local materials at high prices. 8 Roles could have been decided more quickly at the 8 Shelter issues were seen as being dealt with by the beginning, as slow-moving discussions delayed initial provision of basic materials and the project had limited implementation. technical support to help consider what other shelter 8 Some key items (poles and cooking fuel) were issues might need to be addressed. problematic to procure in bulk and the poles were not 8 Access was severely restricted due to the conflict itself supplied. and limitations set by the government. Photo: Joseph Ashmore Queue for distribution of materials Situation before emergency Situation at the initial jerry cans and soap. In September the Before 2004, Darfur was one of response stage number of beneficiaries in need was raised to 1.4 million and by December the poorer areas of North Sudan. At the end of 2003, humanitarian 2004 around 80% of this target popu- Although there are no exact figures, agencies were able to access less than lation had received assistance. This its mostly rural population had limited 5% of IDPs due to travel restrictions equated to almost half a million blankets access to safe drinking water and had imposed by the Government of Sudan. and other material by the onset of poor child nutrition. The region’s in- This made accurate needs assessment winter. An additional estimated 8% creasingly scarce natural resources of difficult. The May 2004 revision of the was covered by other agencies not grazing land and water were one of the Consolidated Appeals Process for using the common pipeline. factors behind the conflict. Sudan estimated a US$ 22.5 million The situation in Darfur became need for shelter and non-food items Technical solutions increasingly volatile throughout 2003, for the remainder of the year. With the shelter materials supplied, with refugee movements beginning The severe lack of shelter available beneficiaries were assumed to be able as early as April. The crisis began to IDPs was a major health threat, to construct basic shelters using poles to escalate at the end of the year. with exposure to heat and dust during as frames, plastic sheeting as a roof and Response to the crisis in Darfur was the day and very cold temperatures at rope for fixings. hampered by the Sudanese govern- night. By May, exposure to rains led to However, procuring poles in bulk ment’s restriction of access to the an increase in communicable diseases. at a national level proved too difficult affected areas. Selection of beneficiaries and these were left out of the package. This meant that no complete shelter The project aimed to reach all of solution was provided. those displaced by fighting in Darfur, which by April 2004 was estimated to Many agencies in the field planned be a total of 1 million people. Benefi- a shelter response where they ciaries were then divided into catego- augmented the materials received ries of full or partial assistance so that through the pipeline with items the most vulnerable would receive a they procured locally, such as poles. full package of items. Those receiving However, the lack of a clear idea of how a partial package were assumed to be beneficiaries would use items to create able to supply missing materials them- shelters meant that some beneficiaries selves, which may not have always would have had to source construction been the case. materials themselves. These may have
Joseph Ashmore been expensive, or, in the collection By August, 70% of the beneficiar- of poles, have entailed the same risks ies were reached with ‘first tier’ NFIs, Unloading a lorry of relief items to personal security that many women comprising plastic sheeting, blankets, faced when collecting wood for fuel.
27 41 SHELTER PROJECTS EAST AND HORN OF AFRICA A.9 Darfur - 2004 (ongoing) - Conflict - Materials distribution Africa Sudan Excerpt from: Shelter Projects 2008 12 / A.9
The limited consideration of started later in the year). An NGO The centralising of procurement technical shelter issues was unsurpris- was responsible for the ongoing trans- and the management of an efficient and ing considering that the programme’s portation and storage of these items cost-effective supply chain reduced the main focus was on the logistical chal- to sub-level warehouses and the UN logistics overheads for agencies on the lenges of mass distribution. agency that made the original pipeline ground, freeing them up to concentrate proposal was made responsible for co- on direct assistance to beneficiaries. Implementation ordinating the supply chain. Meetings to discuss the idea of a The coordinating agencies were pipeline were held in March 2004, with In addition, a third UN agency able to keep an up-to-date broad a concept paper for funding included agreed on 19 April to act as a ‘consignee’ overview of the requirements in in the updated Consolidated Appeals to officially receive imported goods, Darfur, which helped to direct assist- Process released at the beginning of an extension of their role in receiving ance to those areas most in need and May 2004. food items. This was crucial, as individu- prevented a duplication of response. al NGOs were unable to clear customs With such a large affected popula- Any agency wishing to receive as quickly as the well-established UN tion and the government restricting items from the pipeline had to fill in agency. both access and imports, the main a request form and a basic needs as- aim of the pipeline was to get enough On the programmatic side, coordi- sessment form. The procurement non-food items and shelter items into nation of needs analysis, gap identifica- agency delivered items to El Obeid the hands of beneficiaries as quickly as tion and interaction with humanitarian and/or Nyala. From there the NGO possible to cover basic needs. partners in the field was carried out responsible for distribution arranged by the UN coordination agency and for deliveries to their warehouses in During April 2004 a structure was the agency that initiated the whole the different state capitals. The individ- agreed upon for the management and project. ual NGOs responsible for making the initial request would then make the final distribution to beneficiaries. Information-sharing on needs gaps, current stocks and supply chain updates was achieved through regular meetings and access to an open website. Logistics and materials A basket of non-food items was agreed on by participating agencies. This basket included plans for some sanitary items, clothing and kitchen sets. Initial plans for additional shelter items – poles and rope – were dropped after it proved too difficult to provide them. Plastic sheeting was mostly donated from abroad or imported. Other non-food items were purchased in Khartoum or Nyala. Shelter items in the NFI basket
Joseph Ashmore Note: Other items, such as sanita- tion items, were also supplied in the implementation of the pipeline. The Emergency shelters non-food item basket, but are not process was supported by several listed here. ‘Coordination with all of Materials Quantity donors and the final allocation of re- the agencies was key. We sponsibility was undertaken by the UN held weekly meetings in Blankets 2 Country Team. The organisation of the Khartoum and the field Plastic sheet 1 pipeline was split into two main parts: (4m x 5m) a supply part and a programme part. and set up information- sharing systems such as a Rope 20 m For the supply part of the website. No one had an Poles 6 planned, but programme, one UN agency was excuse for not knowing not distributed nominated for procurement, which what was going on!’ Sleeping mat 2 would be distributed as far as a central- - Coordinator ised warehouse (though this process
28 28
SHELTER PROJECTS EAST AND HORN OF AFRICA www.shelterprojects.org 42 Historic13 / D.9 Shelter Projects 2008 Conflict D.9
D.9 Sudan - 198513 / D.9- Conflict Sudan - 1985 - Conflict Case study: Planned camps Project type: Planned camps Disaster: Civil war and famine in Ethiopia (Eritrea and Tigray) 1983-1984 No. of people displaced: Hundreds of thousands Project target population: 232,000 across 15 camp complexes (June 1985) Eritrea Camp capacity designed for up to 640,000 Tigray
Occupancy rate on handover: Ethiopia Unknown Sudan Shelter size: Various
Summary Relocating refugees from smaller camps gave time to create better sites and facilities in the larger camps built as part of the second stage. Building camps using a hierarchy of shelter groupings (cluster-block-sector) helped the humanitarian actors ensure support for the cycle of repatriation.
Project timeline
1983 1984 September1984 November1984 December1984 January 1985 April 1985 June 1985 I I I I I I I I
from TigrayFirst large influx of refugeesLarge-scale migration starts Measles reported in camps per 1000Death people rates per in daycamps rise to 15than 5Death per 1000 rates people reduced per to day less55,000 Voluntary people repatriation of Failure of harvests in Ethiopia
Strengths and weaknesses
9 Working with local relief agencies allowed camp but were insufficiently known among many implementing planners to understand village and community structures, organisations. and to adapt camp layouts to those structures accordingly. 8 Unplanned camps not only had problems with water 9 Having clearly demarcated sections and blocks in a supply, but some then had health-threatening problems camp facilitated both repatriation and phased reuse of the with drainage once the rains arrived. camp for newcomers. 8 Relocation to new camps, while unavoidable, had large 9 Decentralisation of services in the camp allowed for programme costs. easier training of village health workers in preparation for 8 Not even advanced camp layouts can solve the grave repatriation. issues of malnutrition or communicable disease. 8 Multi-sectoral guidelines on camp planning and camp management had been available for a number of years,
Case study credits: Cuny Center 105 43 SHELTER PROJECTS EAST AND HORN OF AFRICA D.9 Sudan SudanExcerpt - 1985 from: - ShelterConflict Projects - Planned 2008 camps 13 / HistoricD.9 Before the influx There had been ongoing conflict between the Ethiopian government and rebel groups fighting for inde- pendence for the provinces of Eritrea and Tigray since the 1970s. Many refugees from the conflict moved to Sudan. During 1983-1984, the conflict combined with drought across many countries in Africa to create a major famine. There were no early warning programmes or adequate stockpiles until after September 1984. Before 1984, sufficient food had been supplied into Tigray from Sudan. By mid-1984 the Relief Society of Tigray, a national civil relief organisa- tion, stated that the famine had reached crisis levels and that they would lead Tigrayans out of Tigray and into Sudan, where they could receive aid. Cuny Center/UNHCR Initial camps in Sudan were sometimes located adjacent to the Block plan sites of older permanent refugee set- arrivals from each village. tlements. In early December 1984, Selection of beneficiaries it was realised that there were not There was no selection per se. As This cluster, block and sector enough water resources for these the refugees arrived in the camps in hierarchy was derived from the camps. A decision was taken to look more or less intact village groups, it Handbook for Emergencies, which had for sites that would support larger was possible to work with the village been made available two years before numbers of refugees. Even then, not leaders and social structures to identify the crisis. A Sudan-specific version of all camps had adequate clean water for vulnerable members. the handbook specific was created. many months. Waterborne disease, Land rights / ownership As the main emphasis was placed alongside measles and malnutrition in There were no permanent land on water supply, sanitation and the new arrivals, became the chief cause of rights given to refugees. In fact, the logistics of food and medicine, the death in the camps. government of Sudan insisted that basic shelter was often a traditional Although the Sudanese had new refugees would not be granted tukul tent made out of branches, welcomed hundreds of thousands of permanent residency. although there were some distribu- refugees for resettlement from Ethiopia Technical solutions tions of other shelter materials. The over the previous two decades, the Once decisions had been made to government’s insistence that the scale of the new influxes, and the transfer some of the refugees from in- camps were to be short term often fact that Sudan itself was suffering a adequate camps, the new camps were prevented the use of any more durable drought, caused a reversal of policy in set up following a hierarchy of blocks shelter materials, even if the resources the Sudanese government. Even when of buildings. This started with a cluster had been available. this decision was overturned, the gov- of shelters based on the size of each Implementation ernment indicated that they did not extended family. These clusters could The Relief Society of Tigray would expect the refugees to remain in the be grouped together to form a block often lead the Tigrayans into Sudan in long term. that would follow the size of a single entire village groups. In some cases, After the first influx village. A number of blocks would form the society would also participate in NGOs began searching for suitable a sector of a camp. the transfer of groups from one of sites for new camps. Between April Importantly, the number of clusters the first camps to a second camp with and June 1985, 55,000 refugees were in a block was not predetermined, but better facilities. able to return to Ethiopia. But this still was dependent upon the number of Materials left 258,000 new Ethiopian refugees in extended families coming from each Pressure from the Govern- eastern Sudan, in addition to 120,000 village in Tigray. To the extent possible, ment of Sudan meant that use of any Chadian refugees in the west of the services such as health units and sup- ‘permanent’ materials was avoided. country, 700,000 ‘old’ Ethiopian plementary feeding centres were de- Although there were distributions of refugees and increasing numbers of in- centralised throughout the camps. plastic sheeting, many of the refugees ternally displaced Sudanese. Space was left in each block for late
106 106 SHELTER PROJECTS EAST AND HORN OF AFRICA 44 Historic Shelter Projects 2008 D.9 13 / D.9 Conflict
lived in self-built tukul tents, made from tree branches, grass thatch and cloth. Logistics Access to the camp helped with logistics. The most important paved highway in Sudan, connecting Port Sudan with Khartoum, ran through the camps areas. A major train line also ran adjacent to the highway for part of the time, and airports capable of handling large jets or C-130s were available at towns used as logistics hubs. Most materials had to be imported using UN mechanisms, apart from in- dividual shelter materials scavenged by the refugees. During the emergency, there were some severe delays in the provision of materials, but these were caused by poor pre-planning, lack of stockpiling and internal organisational issues, as much as by lack of physical infrastructure. Cuny Center\UNHCR
Sector plan
‘[The design] had several major advantages. First, it enabled the relief agencies to train a cadre of health workers from each village. In the event that people decided to return to Tigray (which many of them did several months after arriving in the camp), the skills and training the workers had acquired would be taken back to the village with them. Second, it provided camp administra- tors with a simple way to reunite families. When anyone entered Sudan, they simply had to tell the relief authorities what Ti- grayan village they were from; they could be transferred to the camp where the people from that village were located. Family reunification could then be han- dled on a self-help basis. Finally, camp administrators were pre- sented with an intact community organization with which to work, facilitating activities which re- quired notification or organiza- tion of the refugees.’- Fred Cuny Cuny Center/UNHCR Camp plan
107
45 www.shelterprojects.org SHELTER PROJECTS EAST AND HORN OF AFRICA NaturalUganda disasters Excerpt from:Shelter Shelter Projects Projects 200 20099 14 / B.25B.25
B.25 Uganda14 / B.25 -Uganda 2007 - Slow2007 - onsetSlow onset floods floods Case study: Materials and public information Full case study
Country: Project timeline Uganda - Katakwi and Amuria districts Disaster: Floods Disaster date: Between July and mid September 2007 No of houses damaged: More than 20,000 households were severely affected No of people displaced: 58,000 people Project target population: UGANDA 100.000 families located in 96 villages Occupancy rate on handover: Katakwi district 7458 shelter completed Amurai district Shelter size: Traditional round hut 12m2 with veranda July 2007 – – Flooding starts
Summary 10,000 plastic sheets were distributed during the relief phase. These were for temporary roofing materials in the absence of grass, and were also used to prevent rain from destroying walls and moulded bricks. To ensure that communities rebuilt more flood resistant shelters, both communal and individual tool kits were distributed. These were combined with a large scale public information program on building back safer. As the traditional building season was three months after the floods, during the dry season, the project had components of water, sanitation and agriculture. The approach taken was to work through community mobilisation.
Strengths and Weaknesses 9 The programme worked in many sectors including 8 The recovery process was slow due to bad weather. distribution, water and sanitation and health activities. 8 The government had already started housing The assessment included a multi-sectoral team programs (concrete blocks and iron sheeting) which 9 A shelter specialist was rapidly deployed to support were often too expensive for the affected population. programmes. 8 There was some resistance towards earth and 9 The emergency Items arrived within 2-3 weeks of thatch buildings. the floods. This was possible because there was an 8 The national partner organisation had a lack of existing emergency stockpile in Nairobi. experience in shelter projects. 9 The project used large scale public information and participation to empower communities. 9 A simple technical solution was used, based on simple improvements to a traditional construction. 9 Different organisations operated in different geographical areas. This helped to avoid duplication. 9 A combination of communal kits and individual kits helped the organisation to target more families. 9 The international organisation worked with a national partner that was strong in community mobilisation. A boy walks past a flooded house Photo: Jacob Dall
79 SHELTER PROJECTS EAST AND HORN OF AFRICA 46 Natural disasters Shelter Projects 2009 B.25 B.2514 / B.25 Uganda - 2007 - Slow onset floods - Materials and public information NaturalSlow onset disasters floods
Before the disaster For 20 years, Katakwi and View of a village after the floods showing the traditional circular shelters. Amuria districts of Eastern Uganda Photo: IFRC were controlled by the Lords Re- sistance Army and affected by Ka- tion and technical awareness. The showed improved earth construc- ramajong raids from the North. project combined interventions tion, and illustrated the following Although security had improved in many different sectors such as details to protect the house from as a result of presence of the army camp planning and water and sani- flooding or termites: tation. and police forces, many people • The house and foundations remained afraid, preferring to sleep The programme paid less should be elevated. at night in clusters in camps rather attention to individual needs. It • Foundations should be built than returning to their family plots. focussed instead on information with a large plinth beam of The traditional local shelter sharing and the distribution of wattle and daub. This would design is a round mud hut with a communal tools. The tools could be need to be repaired by house thatched roof. used for shelter, road works, agri- owners after each small flood culture, and other uses. • A water proof barrier should be The majority of the families are put the foundations to protect engaged in farming and other small Implementation the walls and floors which are businesses. The main crops are • 10,000 tarpaulins and 2000 made of adobe blocks. sorghum and cassava, but the crops communal kits were distributed • Walls made of sun dried mud had not been harvested before the • Technical awareness posters blocks should be conical in floods struck. were distributed shape • Prototypes shelters were erected • Proper materials to build more After the disaster with the community resistant earth blocks should Heavy rains in the East of The project trained sixteen be identified. Examples are clay Uganda caused slow-onset floods members of the partner organisa- from termite hills, and using that damaged houses in the camps tion to support 224 community mud mixed with cow dung to and destroyed crops in the fields. volunteers. These volunteers were protect against termites. Programme overview active within camps. • Wood in direct contact with the earth should be trated to To reduce the impact of floods in Affected families themselves protect it from termites. the region, the program focused on built the shelters whilst volunteers improved prevention and prepar- monitored the construction. Material lists edness. It also used local building The communal kit contained: a knowledge to improve the houses. Technical solutions wheelbarrow, a hammer, a wood In the initial emergency phase, saw, a claw hammer, a machete, The supported shelter pro- plastic sheeting was distributed a hoe, an axe, a pick axe, a sharp- grammes improved awareness on along with other materials. ening tool, a tape measure, a spirit how to rebuild more safely as well level, a dumpy level and a first aid as providing tools and grants. A list of necessary but lacking tools was drawn up with the kit. Selection of beneficiaries community. These would be The household kit contained: Through coordination meetings, required to help families to re- a sickle, brick making moulds, the area was split geographically construct their traditional earth damp proof membrane (polythene between organisations. dwellings. sheeting), anti termite treatment for The shelter project focused Information, education and wood, sisal roll, nails, a 20 litre Jerry on twenty camps and promoted can, a medium trowel, a window House under construction with improved plinth communication materials such Photo: IFRC community awareness, participa- as posters were produced. These shutter, a door shutter, and wire. 81 47 80 SHELTER PROJECTS EAST AND HORN OF AFRICA NaturalUganda disasters Shelter Projects 2009 B.25 B.25 Uganda - 2007 - Slow onset floods - Materials and public information Natural disasters Excerpt from: Shelter Projects 2009 14 / B.25
Before the disaster For 20 years, Katakwi and View of a village after the floods showing the traditional circular shelters. Amuria districts of Eastern Uganda Photo: IFRC were controlled by the Lords Re- sistance Army and affected by Ka- tion and technical awareness. The showed improved earth construc- ramajong raids from the North. project combined interventions tion, and illustrated the following Although security had improved in many different sectors such as details to protect the house from as a result of presence of the army camp planning and water and sani- flooding or termites: tation. and police forces, many people • The house and foundations remained afraid, preferring to sleep The programme paid less should be elevated. at night in clusters in camps rather attention to individual needs. It • Foundations should be built than returning to their family plots. focussed instead on information with a large plinth beam of The traditional local shelter sharing and the distribution of wattle and daub. This would design is a round mud hut with a communal tools. The tools could be need to be repaired by house thatched roof. used for shelter, road works, agri- owners after each small flood culture, and other uses. • A water proof barrier should be The majority of the families are put the foundations to protect engaged in farming and other small Implementation the walls and floors which are businesses. The main crops are • 10,000 tarpaulins and 2000 made of adobe blocks. sorghum and cassava, but the crops communal kits were distributed • Walls made of sun dried mud had not been harvested before the • Technical awareness posters blocks should be conical in floods struck. were distributed shape • Prototypes shelters were erected • Proper materials to build more After the disaster with the community resistant earth blocks should Heavy rains in the East of The project trained sixteen be identified. Examples are clay Uganda caused slow-onset floods members of the partner organisa- from termite hills, and using that damaged houses in the camps tion to support 224 community mud mixed with cow dung to and destroyed crops in the fields. volunteers. These volunteers were protect against termites. Programme overview active within camps. • Wood in direct contact with the earth should be trated to To reduce the impact of floods in Affected families themselves protect it from termites. the region, the program focused on built the shelters whilst volunteers improved prevention and prepar- monitored the construction. Material lists edness. It also used local building The communal kit contained: a knowledge to improve the houses. Technical solutions wheelbarrow, a hammer, a wood In the initial emergency phase, saw, a claw hammer, a machete, The supported shelter pro- plastic sheeting was distributed a hoe, an axe, a pick axe, a sharp- grammes improved awareness on along with other materials. ening tool, a tape measure, a spirit how to rebuild more safely as well level, a dumpy level and a first aid as providing tools and grants. A list of necessary but lacking tools was drawn up with the kit. Selection of beneficiaries community. These would be The household kit contained: Through coordination meetings, required to help families to re- a sickle, brick making moulds, the area was split geographically construct their traditional earth damp proof membrane (polythene between organisations. dwellings. sheeting), anti termite treatment for The shelter project focused Information, education and wood, sisal roll, nails, a 20 litre Jerry on twenty camps and promoted can, a medium trowel, a window House under construction with improved plinth communication materials such Photo: IFRC community awareness, participa- as posters were produced. These shutter, a door shutter, and wire. 81 80 SHELTER PROJECTS EAST AND HORN OF AFRICA 48 B.25 Uganda - 2007 - Slow onset floods - Materials and public information Natural disasters 14 / B.25 Slow onset floods
Public information images on proper site planning with space between buildings Image credit: IFRC
(1) (2)
(3) (4)
Images from public information posters on building a flood resistant structure: (1) elevate the plinth and put a plastic sheet under the floor (2) fold the plastic sheet over the ground level ring beam (3) build conical walls (4) thatch the roof, render the walls with mud and elevate the area around the house to protect it from flooding Image credit: IFRC
82 49 www.shelterprojects.org SHELTER PROJECTS EAST AND HORN OF AFRICA
NOTES
This booklet is a compilation of case studies of humanitarian shelter responses relevant to East Africa, compiled across the six past editions of the interagency publication Shelter Projects.
The projects described in the case studies and overviews contained in this booklet represent responses to conflict, natural disasters and complex crises, implemented by national and international organizations, as well as host governments, and demonstrating some of the implementation and response options available.
The publication is intended to support learning by highlighting the strengths, weaknesses and some of the lessons that can be learned from different projects, which try to maximize emergency funds to safeguard the health, security and dignity of affected people, whilst – wherever possible – supporting longer-term shelter needs and sustainable recovery.
The target audience is humanitarian managers and shelter programme staff from local, national and international organizations at all levels of experience. Shelter Projects is also a useful resource for advocacy purposes, showcasing the work done by the sector, as well as for research and capacity-building activities.
All case studies and overviews contained in this booklet, as well as from all editions of Shelter Projects, can be found online at:
www.shelterprojects.org
Global Shelter Cluster ShelterCluster.org Coordinating Humanitarian Shelter