Professional Documents
Culture Documents
Foreword
SHELTER PROJECTS
2010
www.ShelterCaseStudies.org i
Published 2012
The copyright for the photographs and images remains with the photographers whose names are indicated on
each photograph.
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 Secretariat of the United Nations 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. The analysis, conclusions and recommendations
of the report do not necessarily reflect the views of the United Nations Human Settlements Programme, the
Governing Council of the United Nations Human Settlements Programme or its Member States.
Approximate prices are given in US Dollars (USD), based on exchange rates at the time of the project.
Front Cover:
© Mildred Beliard, CARE
ii
Foreword
In this third annual Shelter Projects publication, containing summaries of a range of programming experiences
in post crisis situations, we learn several essential lessons that should become principles for wider discussion and
adoption.
One of the issues that triggers some reflection is that in every single shelter response, there is a need and obli-
gation to involve and strengthen local capacities to enable sustainable solutions and proper housing reconstruction
for the affected population. Once again, we need to emphasise the importance of putting survivors of these crises
at the centre of the sheltering process, supporting their role in re-building their own dwellings and the training and
awareness raising of local builders in safe building design and construction.
Another key lesson, clearly reflected in the Haiti 2010 earthquake response, is that since settlements provide
the context for any shelter intervention, the focus on the provision of shelter “products” alone is too limited.
Instead, a larger settlement response is required - without immediate strategic planning covering many areas (land
use, tenure, livelihoods, essential services, housing reconstruction, etc) shelter response plans will always be limited
in impact and at risk of failure due to the lack of integration with these other critical issues.
Identifying and addressing shelter and settlement related vulnerabilities through the reconstruction process will
also enhance the resilience of the disaster affected population at risk to future such events. Disaster risk reduction
must pass from messaging to explicit actions.
The rapid meeting of post disaster shelter and settlement needs, whilst enabling the rapid transition to more
durable solutions by the affected populations themselves, requires informed support and engagement. Given the
typical disparity between the scale of need and the availability of resources, involving and strengthening local ca-
pacities and supporting integral shelter and settlements responses that consider future risks is key.
This publication is an appeal to all those involved in responding to post disaster sheltering needs – affected
governments, local, national and international response actors, and the affected populations themselves - to
learn from and apply the practical lessons from the relevant, recent experiences. Initial response activities have a
significant impact on the approaches to longer term reconstruction. Is therefore imperative that donors and imple-
menting agencies work with Governments and affected communities to plan from first response to full recovery,
maximise available resources and expertise, and utilise emerging better practices from the field.
On behalf of our institutions, and in appreciation of the many and varied contributions from shelter sector
agencies, we are pleased to present this Shelter Projects 2010 publication. We encourage all to learn from this
review of current practices in post disaster shelter and settlement.
Monica Noro
Chief of UNHCR Shelter and Settlement Section
United Nations High Commissioner for Refugees
www.ShelterCaseStudies.org iii
Acknowledgements
Project coordinated by: Esteban Leon (UN–HABITAT), Mark Hopley (UNHCR), Manoucher Lolachi (UNHCR), Sandra
D’Urzo (IFRC) and Joseph Ashmore.
Compiled and edited by: Joseph Ashmore, with additional editorial support from Carlo Gherardi, Jim Kennedy and
Wan Sophonpanich.
Case studies have been provided from the programmes of the following organisations:
• American Red Cross
• Australian Red Cross
• British Red Cross Society
• CARE International UK
• CARE Haiti
• CARE Indonesia
• CHF International
• Croix Rouge Française
• Development Workshop France
• Federation of Handicap International, DAU
• Grenada Red Cross
• Habitat for Humanity Tajikistan
• Habitat for Humanity Romania
• Haiti Red Cross Society
• IFRC - International Federation of the Red Cross and Red Crescent Societies
• International Organisation for Migration (IOM)
• Malawi Red Cross Society
• Norwegian Refugee Council (NRC)
• Oxfam GB
• Pan American Development Foundation (PADF)
• P3SD
• Save the Children UK
• Spanish Red Cross (Cruz Roja Española)
• Sri Lanka Red Cross Society (SLRCS)
• Tonga Red Cross
• UNHCR - United Nations High Commissioner for Refugees
• UN-Habitat - United Nations Human Settlements Programme
The editors would like to express their gratitude to the following individuals who wrote and reviewed the case
studies in this book:
Alberto de Castro, Amanda Ball, Ann Lee, Arshad Rashid, Ascension Martinez, Behruz Dadoboev, Bill Flinn,
Brian Kelly, Britt Christiaens, Carmen Ferrer, Carolina Cordero-Scales, Charles Setchell, Charlie Mason, Chiara
Jasna Vaccaro, Clementine Ludford, Corinne Treherne, Dan O’Neil, David Sacca, Eddie Argenal, Edgar Scrase, Ela
Serdaroglu, Elizabeth Babister, Emma Feeny, Emeline Decoray, Ganga Kariyawasam, Gregg Mcdonald, Guillaume
Chantry, Helen Seeger, Hugh Earp, Ingvild Solvang, Iñigo Vila, Irantzu Serra, Jake Zarins, Jabborov Mardon, James
Bellamy, Jamie Richardson, Jan-Willem Wegdam, Jonathan Brass, Javier Cidón, John Norton, Julia Macro, Julien
Mulliez, Kate Crawford, Katerina Bezgachina, Kathleen Miner, Le The Thin, Lee Malany, Maggie Stephenson,
Mahesh Gunasekara, Manuel Moniz Pereira, Meredith Sisa, Michele Detomaso, Mihai Grigorean, Milton Funes,
Muhammad Abbas Khan, Nadia Tithi, Natalia Prokopchuk, Neil Brighton, Pascal Bourcher, Peter Clark, Pilar Bravo,
Pornphun Sutthiprapha, Richard Hamilton, Rick Bauer, Robin Black, Samba Sibide, Sarah Davies, Shaun Scales,
Stephen Robert Hutcheson, Terry Charles, Thomas Carey, Vanessa Larsen, Waqas Hanif, Xavier Genot.
We would also like to thank those who contributed to Shelter Projects 2008, and Shelter Projects 2009 whose
work is reflected in this document.
We would like to thank the San Francisco Public Library for allowing access to the documents and images used for
the San Francisco 1908 case study.
Photographs are reproduced by kind permission of those whose names appear next to them in the text.
iv
Contents
Introduction
Foreword iii
Acknowledgements iv
Contents v
Introduction vii
Section B - Historical
B.1 Sphere Project - 2011 - Global standards 105
B.2 USA - San Francisco - 1906 - Earthquake 109
Annexes
Annex 1 - Index - by country 114
Annex 2 - Further reading 116
www.ShelterCaseStudies.org v
vi
Introduction
This book contains summaries be gained. insecurity2,3. Over half of the world’s
of shelter projects that have been • Given the scale of emergency disaster affected people were in the
implemented in response to shelter need every year, case People’s Democratic Republic of
conflicts, complex emergencies, studies must have had large China2,3. However, to be disaster
and to natural disasters (Section scale impacts to be included. affected does not necessarily mean
A). It also contains a section on Discontinued trials or design homeless or in need of shelter
historical shelter projects (Section concepts are not included. support.
B) including the sphere handbook • The majority of the project
The two single largest interna-
which was updated in 2011. must be implemented within
tional emergency shelter responses
A full list of case studies, the first year following a
in 2010 were to the Haiti Earth-
organised by country, in Shelter natural disaster. For conflict-
quake and the Pakistan Floods.
Projects 2008, Shelter Projects 2009 affected populations, chronic
and this book, (Shelter Projects emergencies and returns Conflicts in 2010
2010) can be found in Annex 1. processes, longer time scales By the end of 2010, there were
were considered. 43.74 million forcibly displaced
The case studies in this book • Accurate project information people worldwide, a rise of 400,000
were implemented by many was available from staff involved people since 2009. This figure
different organisations, a full list of in the project implementation. includes refugees (15.4 million),
which can be found in the acknowl- • The case studies illustrate a asylum seekers (837,500) and inter-
edgements section (page iv). In diversity of approaches to meet nally displaced people - IDPs (27.5
order to allow weaknesses of pro- shelter need. Providing shelter million5). Many of these displaced
grammes to be openly shared, the is more than simply designing people have been displaced for
case studies are not directly attrib- architecturally impressive struc- many years - for example 7.2
uted to individual organisations. tures. million refugees are in situations of
Host government projects are not protracted displacement.
included. Natural disasters in 2010
In 2010 over 304,000 people It is not known which propor-
As a result of the projects being were killed by natural disasters2,3, tion of these conflict affected
implemented in diverse and often nearly three quarters of them in people required or received support
challenging conditions, they illus- Haiti. This was the highest number with shelter.
trate both good and bad practices. of fatalities in one year over the
From every case study there are decade. A further 304 million It is estimated that 3.1 million
lessons that can be learned, and people were affected by natural people who were displaced due to
aspects that may be repeated or disasters, the vast majority by conflict were able to return during
avoided. floods, droughts and ensuing food 20104. To illustrate some of the
Warning
Each project must take into
consideration the local contexts
and needs of the affected popu-
lation, which will differ in every
case. Projects should not be directly
copied or there will inevitably be
programmatic weaknesses and
failures.
1
UN-Habitat, UNHCR, IFRC Shelter Projects 2008, UN-Habitat, IFRC Shelter Projects 2009. from www.ShelterCaseStudies.org
2
CRED’s EM-DAT disaster database: www.emdat.be
3
See also IFRC, World Disasters Report, 2011
4
UNHCR, Global Trends 2010
5
IDMC/ NRC Internal Displacement Global Overview of Trends and Developments in 2010
www.ShelterCaseStudies.org vii
The earthquake in Haiti had a high media profile and brought challenges of working in complex urban environments.
Photo: Joseph Ashmore
issues relating to returns, we include studies in this book, the more Cash, vouchers, markets
an update on Sozma Qala camp for effective projects were set up with During 2010 there was a sig-
returnees in Afghanistan (A.1), a a clear understanding of the needs nificant research focus on market
project for conflict returnees in Kyr- of beneficiaries and with clear two based assistance, led by projects
gyzstan (A.16), a transitional shelter way communication between the such as ‘The Cash Learning Partner-
programme for returning IDPs in implementing organisations and ship’ (CALP) (www.cashlearning.
Pakistan (A.21), and a case study the affectees. org).
on housing in Sri Lanka (A.28). Ad-
To encourage projects to address Shelter cash interventions range
ditional case studies on shelter for
the needs, Sphere standards1 and from small amounts of cash to
returnees can be found in Shelter
indicators (Section B.1) provide support families whilst they build
Projects 2008 and Shelter Projects
common standards on participa- shelters (North Pakistan, 2009,
2009.
tion, initial assessment, monitoring A.21) to voucher schemes using
Shelter responses in 2010 and evaluation. plastic credit type cards (Chile
International shelter responses 2010, A.2) to phased cash dis-
Scale bursements of several thousand US
in 2010 were dominated by an
After the emergency response dollars (Sri Lanka 2009, A.28).
earthquake in Haiti and floods in
phase, the largest projects outlined
Pakistan. Haiti (Case studies A.4 There are however very few
in this book were a structural
- A.11) brought exceptional chal- examples of cash based shelter
damage assessment (Haiti, 2010,
lenges of working in highly complex projects that were operational
A.6) which assessed 400,000
urban environment, whilst the within the first weeks of a response.
earthquake affected houses, and
Pakistan floods (Case studies A.22 -
the construction of 38,500 one Cash assistance is not a new
A.25) brought challenges of how to
room houses by providing groups form of response. The historical
respond on an exceptionally large
of households with cash (Pakistan case study (USA 1906, B.2) illus-
scale with limited funding to meet
2010, A.24). Even projects on this trates the use of cash and loans,
the needs.
scale are not able to meet the in addition to other forms of assis-
This book also includes three majority of the recovery needs tance such as loans, following the
case studies from the response for “mega disasters” such as the 1906 San Francisco earthquake and
to the earthquake in, Pedang Pakistan floods in 2010 (overview fires.
(Indonesia 2009, A.12 - A.15). A.21) which officially damaged 1.8
and two recovery projects from million homes. The case study from Tajikistan
Myanmar (A.19-A.20). (2010, A.29) provides an example
Urban environments of the use of loans.
There were many shelter The diversity of responses
projects that were implemented in The table on page ix opposite
in Haiti (A.4-A.10) shows how
2010 that are not included in this shows 16 projects or responses that
multiple approaches are required
book. These include those from use cash/vouchers or some sort of
when working in complex urban
disasters and conflicts in many market assistance to meet shelter-
environments, where traditional
other countries including ongoing ing needs.
concepts of “community” develop-
or ”forgotten emergencies“. ed in rural contexts may be harder
to apply.
Affected people are the
first responders Following the Haiti earthquake,
The first and main response Urban Shelter Guidelines were
after all disasters is by the affected launched in late 20102. Many case studies feature the use of
people themselves. Of the case cash and vouchers in shelter projects.
Photo: Carlo Gherardi
1
Sphere Project, Sphere - Humanitarian charter and minimum standards in humanitarian response, 2011
2
NRC, Shelter Centre, Urban Shelter Guidelines launched in 2010
viii
Table illustrating which types of response took place in each case study
Transitional shelter
Emergency shelter
Advocacy / legal
training materials
Cash / vouchers
housing / repair
Guidelines and
Host families /
General items
Core housing
Site planning
Construction
assessment
Permanent
/ T-shelter
Structural
materials
Training
Rental
Loans
A.1 Afghanistan 2009 update
A.2 Chile 2010
A.3 Grenada 2004
A.4 Haiti 2010 (overview)
A.5 Haiti 2010
A.6 Haiti 2010
A.7 Haiti 2010
A.8 Haiti 2010
A.9 Haiti 2010
A.10 Haiti 2010
A.11 Haiti 2010
A.12 Indonesia 2009 (overview)
A.13 Indonesia 2009
A.14 Indonesia 2009
A.15 Indonesia 2009
A.16 Kyrgyzstan 2010
A.17 Malawi 2009
A.18 Mozambique 2007
A.19 Myanmar 2008
A.20 Myanmar 2008
A.21 Pakistan (North) 2009
A.22 Pakistan 2010 (overview)
A.23 Pakistan 2010
A.24 Pakistan 2010
A.25 Pakistan 2010
A.26 Philippines 2010
A.26 Romania 2010
A,28 Sri Lanka 2009
A.29 Tajikistan 2010
A.30 Tonga 2010
A.31 Vietnam 2009
B.1 Sphere
B.2 USA 1906
Explanation of columns:
covering materials) to meet emergency shelter needs.
Distribution:
• Transitonal shelter, T-shelter, Temporary shelter, semi-perma-
• General items - tents / blankets and other non-food items
nent shelter. Terminology is used according to the wording used in
were distrubuted
the response.
• Construction materials - were provided to build the shelters /
• Host families / rental support
houses.
• Core housing (basic component of a house designed to be
Shelter type:
upgraded at a later date)
The project aimed to provide one or more of the following types
• Permanent housing / repair. (durable housing or the repair /
of shelter.
upgrade of exsiting pre-disaster housing)
• Emergency shelter - shelter (often non-food items and
Note definitions may change between disasters.
www.ShelterCaseStudies.org ix
This book contains only a small sample of projects implemented around the world. There were many disasters and conflicts,
some ongoing for many years that have not been documented here. This photo is following the 2010 floods in Benin.
Photo: Joseph Ashmore
1
World Bank website: http://www.worldbank.org/prospects/migrationandremittances. Jan 2012
2
IFRC Transitional Shelter: Eight Designs, 2011
0 6 12 18 24 30 36 42 48
Time in months
Some of the projects in support of long term displacements have not been included due to their long timelines. The majority of
projects were implemented under funding cycles of less than one year.
www.ShelterCaseStudies.org xi
Graph of the distribution against time for major international shelter responses
1,000,000
Number of families provided with
600,000
400,000
Haiti earthquake 2010
1 2 3 4 5 6 7
Time in months
Note: “Families provided with basic shelter” means the number of families (or households) who have received tents or plastic
tarpaulins. The number of tarpaulins received per household varies between response and occasionally changes as responses
continue. For most responses two tarpaulins per household were provided.
In the case of the Pakistan 2010 floods, the start of the response was phased over 6 weeks as the flood waters moved through
the country, and the full extent fo the floods became known. As a result the curve may appear to be shifted to the right.
These graphs display the data that was reported to shelter coordination teams during the responses.
Graph of number of early recovery shelters* built against time for some major
international shelter responses
120,000
Pakistan flood 2010
100,000
Number of early recovery sheltetrs
earthquake 2006
60,000
40,000
200,000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Time in months
* Early recovery shelters include “T-shelters, one room shelters or any other shelters to designed support the recovery process.
xii
Excerpts from John Evelyn’s diary following the great fire of london in 1666, illustrating
how families moved in with host families whilst others settled in spontaneous camps.
Following the fire the navy provided tents and canvas for some of the affected people.
Source Jacob F. Field (2011). Charitable giving and its distribution to Londoners after the
Great Fire, 1666–1676. Urban History, 38 , pp 3-23
www.ShelterCaseStudies.org xiii
xiv
SECTION A
Case Studies
This section contains case studies of projects from both conflicts and natural
disasters. It also contains one update from a project (A.1 Afghanistan) that was
included in Shelter Projects 2009. See “Annex 1 - Index - by country” for case
studies that are in previous editions.
www.ShelterCaseStudies.org 1
Sozma Qala
4 months – – Construction
complete
3 months – – Construction start
Afganistan
– Tents distributed
Project description
An emergency team rapidly winterised a temporary transit camp. The site was for 379 families of refugees
returning after 23 years. Two years later 320 families remained at the site with dwindling funding for external
support. To improve the existing tents, weather mitigating tent structures (WMTS) were built from bamboo and
plastic sheeting. They lasted for more than two years - longer than expected.
A Weather Mitigating Tent Structure. Left: in 2009. Right: two years later.
Photos: Richard Hamilton and Shaun Scales
www.ShelterCaseStudies.org 3
Background - 2009 own land have returned to these insulated WASH facilities were also
(See Shelter Projects 2009 for plots and reconstructed their homes provided at the site.
the original case study.) whilst the remaining families have
remained on the Sozma Qala site. Technical solutions 2011
The Sozma Qala Transit Camp in The original weather mitigat-
the Sar i Pul district of Afghanistan As of 2011, it was planned
that 150 families would return ing tent structures have lasted well
was established in late 2009 to ac- beyond their initially planned life
commodate 379 families returning to their place of origin, a village a
few kilometres away, and would span and will remain standing for a
to their area of origin after the set- fair amount of time to come.
tlement in Iran that had been their be supported to build their own
home for the previous 23 years was houses. In some cases the shelters were
closed by Iranian authorities. Of the families who have left upgraded with mud blocks and new
the site: sheeting by the remaining families.
Local authorities had initially At this site, plastic sheeting lasted a
approved the development of the • 34 families have received a maximum of 24 months. As a result,
temporary transit facility to house winterised mitigating tent the plastic sheeting distributed was
the families whilst long abandoned structure (known as WMTS) at used to patch up the shelters.
houses and infrastructure were their place of origin. They have
rebuilt and issues over land then rebuilt their own homes. The design of the weather miti-
ownership were resolved. • 25 families have moved to a plot gating tent structures has been
of land (owned by one of them) adopted, and many were built two
However during the 2 decades years later for conflict IDPs from
of displacement, family numbers at the bottom of the camp and
have built permanent houses. Sayad district in and around Sar i
had grown significantly, and their Pul district centre. The design was
original land was no longer suf- Two years later, the site has a also adapted for northern Pakistan
ficient to accommodate all of the mosque on the opposite side of the in response to the 2009 IDP crisis.
returnees. main road, and a school is being
built in the camp. It also had three The two remaining agencies
Update: site issues - 2010 wells. Six camp residents were paid who continue to work in the camp
Of the 379 families who moved and armed as Afghan Local Police. have provided limited support
to the temporary site, 320 remained of heating fuel (350kg coal per
two years later. It is highly likely that a number family), blankets (3 per household)
Representatives of the returnee of families will remain at the site and other non-food items.
families then began a long process and it will become a small village in
The latrines and washing areas
to try and be granted the area on the future.
fell apart early in 2011 as did the
which the site had been developed. Technical solutions 2009 garbage collection process.
This was opposed by local com- The original transit camp facili-
munities who had remained in the The design of the weather
ties were built to provide temporary mitigating tent structures used
area during the previous decades of support and were not intended
conflict. To increase the pressure on in this project was informed by
for winter occupancy. Tents were shelters built following the 2005
the local government many families provided as family shelters and a
with access to their original plots earthquake in Pakistan.
basic water delivery system had
refused assistance to rebuild their been developed.
original homes in a show of solidar-
ity with their community members As winter approached the
and extended families. tented site was winterised through
the provision of bamboo framed
As time has passed and the plastic sheet structures over the
land issue has remained unresolved tents. Drainage, gravel roads and
many of those families with their
Views of Sozma Qula transit camp in 2011, two years after it was established
Although intended as a temporary transit site, the majority of the families remained on the site.
www.ShelterCaseStudies.org Photos: Richard Hamilton, Jake Zarins
5
3 months - –– Start
Chile
2 months - –– Project coordinator
hired
Project description
Following a non-food item distribution to 10,000 households, plastic cards with magnetic strips were given
to earthquake affected households. These cards were valid for 30 days from manufacture and could be
redeemed in 40 pre-designated hardware stores located in the affected regions.
www.ShelterCaseStudies.org 7
The organisation trained both would be a second distribution in signed for future collaborations
staff from the participating suppliers the community at a later date. between the organisation and
and beneficiaries on the use of the the relevant government Ministry.
There was also the risk that
cards. In a project evaluation, recip- During an interim project review,
segments of the affected popula-
ients of the cards generally found approximately 80% of respondents
tion were not included because
the staff at participating hardware stated that they had the knowledge
they did not have good relations
stores knowledgeable about the to make their own repairs with the
with the community leaders or
project. materials purchased with the card,
they lived in sites between targeted
17% paid for someone else to do
Staff from the hardware stores communities.
them, while 4% stated that they
travelled to communities with
Technical solutions did not have the knowledge and
product catalogues. This assisted
Once the users received their would have liked to have been
affectees who had limited access to
cards, they had one month to use trained in how to make the repairs
transportation.
it. Partial purchases were allowed, themselves.
A partnership was established meaning that they could buy several
with the Corporate Social Respon- Project conclusion
times during the month in smaller
sibility programme of a Chilean Initially the project targeted
volumes. One other way to use the
company. The company verified ben- 8,400 households, but this was later
funds on the card was to make a
eficiary data, printed relevant docu- increased to 10,000 families. The
bulk purchase for the total value of
mentation and opened a permanent project was implemented in one
the card.
call centre to answer any questions year – from May 2010 to May 2011.
about using the card. At a later stage, initial home It took a little longer to close the
repair guidelines were delivered at project as some transactions could
A call centre was also estab- the same time as the cards. These only take place once all invoices had
lished to allow beneficiaries to verify were in line with an agreement been received.
the amount of funds remaining on
their card along with the location of
participating stores.
The validity of the card was set
on the magnetic strip by the manu-
facturer, but an expiration date was
also printed on each card. Due to
the time needed to distribute the
card, some beneficiaries had less
time to purchase material.
Selection of beneficiaries
The criteria for selecting ben-
eficiaries was very broad, and took
into account which families had
received relief kits. No detailed
damage and needs assessment was
conducted. In practice, the project
relied on beneficiary lists that were The project required significant amount of paperwork.
Photo: Jorge Romo
provided by local authorities and
community leaders along with lists
provided by project staff. These lists
were developed during the distribu-
tion of relief supplies in March and
April 2010, some months before
the distribution of cards.
In some cases the data in the
lists wasn’t accurate, leading to
the misprinting and subsequent
voiding of the cards at the distri-
bution sites. During distributions
there were families at the distribu-
Team members,
tion sites who claimed that they a phone line and
should be included in the project. posters explained
In these cases, they were added to how the project
a waiting list and told that there worked.
Project description
Over 2 years, the roofs of over 650 houses were repaired and 100 homes were built from scratch. 128 people
were trained and certified as carpenters, over 2,000 houses were strengthened with hurricane straps and 32
communities were better prepared to face the next disaster.
www.ShelterCaseStudies.org 9
10
The project used “old time” techniques, learning from the past and which buildings had survived the hurricane and why.
Photo: Emeline Decoray
obtain. Able-bodied beneficiaries done from the project warehouse: were sent to the supplier 3 days
were invited to become trainees storage of the material, loading prior to the delivery date. This way,
and receive the material to rebuild of the truck and delivery on site. most of the logistics issues were
their roofs. If the beneficiary was Because the project grew rapidly, transferred to the supplier. As a
elderly, or was unable to undertake the supplier was asked to manage consequence of this, the organisa-
construction himself or herself, a a part of the logistics from his tion had to coordinate closely with
carpentry team was sent to reinforce warehouse. The bills of quantity the supplier.
and re-roof the house.
Technical solutions
The techniques applied to
rebuild the roof and to strengthen
the house before building the roof
were “old time” techniques, which
had resisted Hurricane Ivan. The
“old time” wooden houses resisted
the wind forces better than the
newly built houses, even concrete
houses.
The houses received some rein-
forcement, such as doubling studs
in the corners, around doors and
windows, bracing the corners in both
directions, attaching the flooring
beams to the pillars, and attaching
them to the foundations with metal A spreadsheet was developed that calculated the materials required given the
straps. The smallest houses received dimensions of each house.
a gable roof with a 30° slope and
25cm eaves, while the largest one
received a hip roof.
www.ShelterCaseStudies.org 11
Sheltering in Haiti: While the case studies that follow reflect extraordinary
Looking forward while looking back and laudable effort, they also at least suggest that
In August 2010, seven months after the devastat- the questions remain only partially answered, to the
ing Magnitude 7.0 earthquake near Port-au-Prince, a detriment of those living in - and out of - camps.
think tank made the following key shelter-related rec- As central as the rubble issue has been to recovery,
ommendation1: the more important issue, and underlying rubble both
“The Haitian government, together with the donor literally and figuratively, is the land that was the locale of
community, should accelerate removal of rubble. This the homes, shops, schools, neighbourhoods, and other
is the single most important step toward reconstruction features of a primarily densely populated urban area
of housing and infrastructure that the Haitian govern- affected by the earthquake. The rubble and broken
ment and donors can take.” buildings littering settlements after the earthquake ef-
fectively decreased the size of those settlements, and
The study went further:
thus the supply of land available for sheltering people
“For housing to be reconstructed, sites have to and recovering economic, educational, governance,
be cleared… Unless rubble is cleared expeditiously, and other activities. The land and housing markets
hundreds of thousands of Haitians will still be in tent in those settlements, constrained by myriad tenure,
camps during the 2011 hurricane season.” infrastructure, service, and hazard risk issues prior to
the earthquake, were exacerbated significantly by its
That hundreds of thousands of Haitians still face
impacts, making it extremely challenging to respond
the very real prospect of remaining in camps during
to widespread shelter needs, while also affecting the
the upcoming 2012 hurricane season, and perhaps
longer-term process of recovery.
beyond, speaks volumes about the challenges of de-
livering humanitarian shelter assistance and housing Shelter and land issues in urban areas pose par-
reconstruction in Haiti - and elsewhere. ticular challenges to humanitarian organisations, many
of which have their genesis, institutional memories,
The difficult, dangerous, and generally thankless
protocols, and expertise in rural areas. Confronting
task of clearing rubble is viewed largely as a means to
rubble, land, and related issues in dense urban areas
the end of enabling the recovery of lives, communi-
anywhere would thus be a challenge to even the most
ties, and societies in the wake of disasters. Clearing
experienced humanitarian organisations. All the more
rubble, then, is a critical precursor to recovery; it can’t
so in Haiti, where extreme poverty, environmental
be overlooked or sidestepped. Perhaps more so than
degradation, and a host of hazards, coupled with the
any previous natural disaster since the adoption of the
limited capacities of a complex network of regulatory,
UN cluster system in 2005, the Haiti earthquake chal-
political, community, and market actors, combined to
lenged that system significantly with the profound
create the highly vulnerable settlements that sustained
issue of ownership: which cluster would take the lead
such overwhelming destruction, and making it all the
in addressing clearance of the enormous rubble pile
more difficult to respond to needs generated by the
generated by the earthquake? Which donors would
earthquake.
fund the planning and clearance of rubble? Which
organisations would actually do the clearance work?
1
RAND Corporation. Building a More Resilient Haitian State, 2010. Available from http://www.rand.org
12
Many of the case studies that follow contended are fundamental to any effort to address shelter needs.
directly with land and related settlements issues, No less important than these foundational elements of
bringing both reaffirmation of and new meaning to sector strategy is the communication of strategy, for
the phrase “shelter and settlements” (S&S) sector that even the best of strategies are less than effective if not
has been used increasingly by humanitarian actors in understood widely, adopted by key actors, and imple-
recent years to reflect a recognition that sector activities mented expeditiously. The strategic communications
entail not just the four walls and roof of a shelter, but outputs of humanitarian actors in urban areas must
also its contextual setting. A focus on the settlements be disseminated early and repeated often in order to
side of the sector will likely remain a feature of continu- inform and guide response activities. Messaging also
ing efforts in Haiti, as well as future sector responses needs to be creative, visible, and pervasive to compete
elsewhere, particularly those in urban areas. To do with the multiple and voluminous messages received
otherwise would only further increase the vulnerability daily by those living in urban areas. Although this was
of populations in hazard-prone settlements. and remains a challenge in Haiti, as it is anywhere, the
rapid emergence of numerous forms of social media
Perhaps the zenith of shelter and settlements sector
enabled not only delivery of strategic messages, and
programming in Haiti has been the “neighbourhood
much needed feedback, but also actual implementa-
approach” adopted by several actors to plan and
tion of shelter programmes, with “mobile money”
integrate multi-sector, area-based programming, often
initiatives to pay for rent and other necessities a good
in collaboration with other humanitarian agencies, civil
example.
society organisations, the private sector, and local and
national government offices. This settlements-based Finally, the following case studies reflect consider-
approach to shelter provision was identified early on able innovation and flexibility by humanitarian actors in
after the earthquake as a means of both working in response to numerous constraints, an awareness that
rubble-strewn areas to provide humanitarian assistance risk reduction is paramount to “Building Back Better”
and establishing a platform for subsequent reconstruc- and a recognition that “one-size-fits-all” approaches, if
tion. Although initial results of the neighbourhood they ever were effective in rural settings, are most defi-
approach are promising, there are still more earthquake- nitely inappropriate in urban settings. Moving ahead, a
affected neighbourhoods than actors to work in them. focus on the neighbourhood approach will likely remain
Further, a macro-level, city-wide complement to the a feature of continuing efforts in Haiti, as well as future
neighbourhood approach, which could link currently Shelter and Settlement sector responses elsewhere,
disparate and distant efforts, is still very much a work particularly in urban areas. In Haiti, the range of inter-
in progress in Haiti, despite the intensive and concerted ventions will have to expand, as impoverished families
efforts of UN-HABITAT and others. Finally, it must not in camps, limited land supplies, complex land tenure
be overlooked that the neighbourhood approach, if issues, and limited resources will likely conspire to
adopted and implemented early in the response effort, produce not just more transitional shelters and more
is an effective means of promoting inter-cluster coor- repairs of damaged housing, but also greater resort to
dination, lending critically important on-the-ground hosting support, rental housing production, and rental
support to the cluster approach, which is, after all, the subsidies. It is hoped that the effort going forward will
primary means of guiding humanitarian action. feature the continuing quest for clarity on the seminal
issues that confound and define the sector, perhaps the
One very large “lesson learned” of the Haiti earth-
largest alluded to in the study quoted above: what is
quake is that both the neighbourhood approach and its
shelter, what is housing, and what is meant by “toward
macro-level complement, an emergency master plan,
reconstruction”?
Charles A. Setchell
Charles A. Setchell is the Senior Shelter, Settlements, and Hazard
Mitigation Advisor, USAID Office of US Foreign Disaster Assistance
(OFDA)
www.ShelterCaseStudies.org 13
Background uncertainty, annual risks due to rain In the first four months,
Prior to the earthquake, Haiti and hurricanes and an outbreak of 560,000 tarpaulins, 62,000 tents
was the least developed country in cholera at the end of 2010. and 130,000 kits containing tools
the region, ranking 145th of 169 and fixings were distributed by 80
countries in the United Nations Emergency Response organisations.
Human Development Index. More During the first three months,
many affected families moved As per the initial plans, distri-
than 70% of the population lived bution data showed that 100%
on less than 2 USD per day. from damaged neighbourhoods
onto available spaces, establish- of households received emergency
In the cities people lived in ing spontaneous camps. Some of shelter items by 1st May 2010.
crowded neighbourhoods with poor these were subsequently formalised
infrastructure and without access to T-Shelter and early
and serviced by various supporting
basic services. Living space in Port- recovery
agencies. In less damaged areas,
au-Prince’s permanent housing was Many donors and agencies
many stayed with host families. For
reported at just 1.98m2 per person developed projects to provide tran-
the first months, many people slept
before the earthquake. sitional shelters (also referred to
outside damaged houses afraid to
as T-Shelters) to agreed standards.
The urban context, with high go back in.
Given the need for large scale
proportions of tenants, needs for An estimated 500,000 people material imports, pressure for land
urban planning and challenges left the earthquake affected area and other challenges, it took two
of engagement with the govern- in the first month but the majority years to build over 100,000 planned
ment contributed to the complex returned by mid 2010. shelters, missing the initial planning
operating environment. target of 18 months - the start of
The initial response provided
After the earthquake, thousands the hurricane season of 2011.
emergency shelter support through
of non-government organisations provision of basic materials, tar- Repairs to damaged houses
with varying levels of experience paulins, fixings and other non-food were slow to start but accelerated
appeared in Haiti. At times this items to a maximum number of from the end of 2010 to almost
undermined an already weak gov- people. This was to supplement and 14,000 houses repaired by agencies
ernment sector that had lost infra- weather proof the large number by the end of 2011. This figure does
structure and personnel. Recovery of self-made shelters built from not include the houses repaired by
was further challenged by political salvaged materials. people themselves without support.
Many earthquake affectees found themselves living in temporary settlements through the rains.
14 Photo: Joseph Ashmore
Over 630,000 plastic tarpaulins were distributed, allowing people to protect themselves from the sun and rain. However
there was a risk that many of the spontaneous settlements would become the slums of the future.
Photo: Joseph Ashmore
Housing and
neighbourhoods
A strategy was developed
during 2010 to promote support
in the areas of origin to accelerate
return from camps and reconstruc-
tion in rehabilitation. This was not There were major shortages of land - in this settlement, families built in the
adopted until the beginning of central reservation of a major road.
2011 and formed the basis of the Photo: Joseph Ashmore
majority of neighbourhood based
recovery programmes.
At the end of 2011 there were
still over 500,000 people in camps.
This included both people directly
affected by the earthquake but also
reflected a pre-existing housing
deficit and urban poverty.
Official permanent reconstruc-
tion assistance shows limited
progress with approximately 5,200
houses built within two years, and
limited support for host families.
However, the rate of self recovery
and formation of spontaneous
new settlements by Haitian families
themselves is significantly higher.
Support programmes including in-
formation and training have been
limited, and much of the rubble has Many families built their own temporary shelters using reclaimed materials.
Photo: Joseph Ashmore
yet to be cleared.
www.ShelterCaseStudies.org 15
EMERGENCY SHELTER consists primarily of TRANSITIONAL SHELTERS are simple timber or steel
188,383 DESTROYED
OR SERIOUSLY
tarpaulins and fixings such as ropes, nails, a hammer
etc. Tents can also be used for emergency shelter
frame structures that provide beer protection, more
privacy and more space. Transitional shelters will oen
DAMAGED HOUSES IN HAITI
but, because they are less versatile than tarps, their have a concrete foundation and can last years. Once
1.5 MILLION PEOPLE NEED use is limited. Emergency shelter can be distributed people have found permanent homes, transitional shelters
SHELTER ASSISTANCE quickly but offers only limited protection against can be be put to other uses. They take longer to build but
heavy rains. can be dismantled and moved if necessary.
THE ACHIEVEMENTS AS OF
6/25/10
MEMBERS OF THE SHELTER AND NON-FOOD-ITEMS CLUSTER have delivered vital aid to the estimated 1.5 million people who
were directly affected by the earthquake. Despite a destroyed port, a severely damaged airport and a lack of infrastructure, cluster
members reached an average of 100,000 people per week in the first four months of the response operation. Each family received
two tarpaulins or one tent.
= 1,000 TRANSITIONAL
SHELTERS
OF TOTAL
125,000
PLANNED
45,722
IN STOCK
3,264
COMPLETED
12,175
IN COUNTRY
27,214
IN PIPELINE
70,279
DISTRIBUTED
= 1,000
TENTS
PERCENT
140
NEEDS MET
120
OR EXCEEDED*
Tents are less 100
versatile, need
more space and 80
do not last as
500,000-600,000
long as tarpau- 60
lins. For that 40
PEOPLE LIVING WITH reason fewer
A HOST FAMILY tents than tarps 20
were distrib-
uted. 0
TOTAL DISPLACED Blankets Buckets/ Hygiene Kitchen Mats Mosquito Ropes
PEOPLE Jerry cans kits sets nets
PERCENT OF
30%
Distributed In country On the way Remaining need
DISPLACED PEOPLE
LIVING WITH A HOST
FAMILY *Needs are oen exceeded because items are lost or destroyed by weather or used up.
THE CHALLENGES Because most people were renters or squaers and don’t own land, all aspects of shelter are very
complicated. All steps have to be agreed with the tenant and the land owner.
KEEP OUT
PORT-AU-PRINCE
1 2 3 4
OWNERSHIP OF LAND IS OFTEN SITES ARE BLOCKED BY DEBRIS. EVEN MANY ROADS ARE TOO NARROW FOR HURRICANE SEASON: EMERGENCY
UNCLEAR BUT OWNERS HAVE TO WITH HEAVY EQUIPMENT IT WILL HEAVY EQUIPMENT. MULTI-FAMILY SHELTERS CAN BE DESTROYED BY
GIVE PERMISSION BEFORE ANY TAKE YEARS TO REMOVE IT. BUILDINGS CANNOT BE EASILY HEAVY WIND AND RAIN.
WORK CAN BE DONE. REPLACED.
SOURCE: IASC HAITI SHELTER CLUSTER, 2010 CC BY-ND GRAPHIC BY STANFORD KAY STUDIO.COM
kits distributed
Port au Prince
1 month - –– Draft 1 year strategy
January 12th
2010- –– Earthquake
Project description
This project provided different forms of support for people with differing needs. In the emergency phase the
organisation distributed 10,000 emergency shelter kits. It went on to provide 2,550 transitional shelter kits,
20,000 reinforcement kits for those did not have land to build upon, 500 rural repair kits and over 1,000 tool
kits. These kits were accompanied by trainings and posters on staying safe during hurricanes. The organisation
also actively supported inter-agency coordination and had a strong advocacy role.
www.ShelterCaseStudies.org 17
Before the earthquake • Spontaneous camps in highly Transitional shelter kits required
(See A.4 “Haiti - 2010 - Earth- damaged zones close to the that people had access to a space
quake - Overview”, p12.) epicentre of the earthquake to build a shelter. These were not
(Leogane). necessarily the most vulnerable
Before the earthquake the or-
• Dense spontaneous settlements families.
ganisation in Haiti had concen-
along roads to Leogane, that
trated in poor rural areas and on Reinforcement kits targeted
were likely to be neglected by
smaller scale projects. The organisa- families who were unlikely to
other agencies (Carrefour).
tion was not focused on shelter or receive a transitional shelter kit
• Spontaneous settlements close
construction. and who would remain in self-built
to the office and warehouse
(Port-au-Prince). shelter during the hurricane season.
Many of the organisation’s
Training sessions were held on how
experienced staff were directly
Neighbourhoods to use the kits and printed fliers
affected by the earthquake. The
Following the emergency dis- were distributed. Trained carpen-
country office had very few staff,
tributions, the organisation shifted ters also supported families to re-
no partners and little experience in
target to neighbourhoods rather inforce their makeshift emergency
areas directly affected by the earth-
than camps. The main reason for shelters.
quake. Scaling up the capacity of
this was to push to more durable
the country office was also difficult Toolkits were given to agencies
shelter solutions than could be
because many non-government or- that were training technicians, but
found in camps.
ganisations arrived – all trying to who had limited resources.
recruit locally. Although massive shelter needs
remained, the organisation decided Land tenure
Emergency shelter kits not to continue providing shelter The organisation’s approach to
The organisation initially re- assistance in spontaneous settle- tenure was to:
sponded by distributing emergency ments in Port-au-Prince. This was • Record reported tenure status
shelter kits. These contained plastic due to the large number of other during registration.
sheeting, mattresses, hygiene sets actors working there, and also to • Develop a Memorandum of
and kitchen sets. These materials allow them to focus activities. Understanding (MoU) with
were delivered to affected people
All families with destroyed beneficiaries in coordination
within the first three months after
housing in the most vulnerable with other agencies. This
the earthquake and before the
neighbourhoods were targeted. highlighted that beneficiaries
major rains arrived.
will own the shelter but that
It was difficult for any agency Transitional Shelter Kits tenants must take responsibility
to identify the neediest geograph- Kits were developed to protect for seeking the consent of their
ic areas in terms of the highest people from the imminent rains landlord to erect a transitional
number of the most vulnerable and hurricanes. 2,550 transitional shelter for 3 years.
people, highest levels of damage, shelter kits (6 million USD of mate- • Engage the municipality in
and zones most likely to be ne- rials), 20,000 reinforcement kits (3 a similar agreement which
glected by responding agencies in million USD of materials) and 500 outlines the approach and puts
the first 3-6 months. The organisa- repair kits for timber-frame houses the onus on municipalities to
tion decided to deliver emergency were distributed. Half of these tran- resolve disputes.
shelter kits to: sitional shelters were built in part-
nership with another organisation.
18
Shelter Design
The following are the seven key
stages in the transitional shelter
programme:
• Assessment and beneficiary
selection: visit dwelling and
complete assessment form.
• 1st verification: visit destroyed
house, and plot. Check with
neighbours. Fill in verification
form.
• 2nd verification: visit proposed
plot to check that it is ready.
• Explanation and 1st MoU
signature: explain and sign
the MoU to clarify that the Different approaches were used to procurement - some items
beneficiary has consent to use were prefabricated off site.
Photo: Mildred Beliard, CARE
the plot and that the roles and
responsibilities are understood. publishing accounts or account- 20,000 Reinforcement kits
• Delivery and 2nd MoU signature: able to identifiable shareholders. Item Quantity
sign MoU to confirm that the This made it difficult for the or- Plastic sheet ( 4m X 5m) 1
shelter has been received. ganisation to monitor problems Timber 2" x 4" (50x100mm) 24m
• Installation: teams install the with labour rights, health and Hurricane strap 6m
shelter (2 carpenters, 5 helpers safety, environmental regulation or Roofing nails 1Kg
from the beneficiary’s side, check that materials – particularly Nails - 1inch (25mm) 2Kg
supervised by a technician). imported timber – were from sus-
• Final handover and 3rd MoU Nails - 4 inch (100mm) 1Kg
tainable sources. Metal corner spikes 50cm 6
signature: sign the MoU to
confirm that the shelter has Emergency staff were unable to 8 mm nylon rope 25m
been installed. build sufficient capacity for efficient Bag for ironmongery 1
procurement. As a result the project Plastic box 1
Kits and the accompanying used multiple approaches for pro-
information campaign were de- curement. These were: 500 Rural Repair kits
veloped in partnership with other Item Quantity
agencies using a commonly agreed • A partner organisation and local
private contractors purchased Timber 2" x 4" (50x100mm) 48m
transitional shelter brief. Shelter
the timber and all other Hurricane strap 10m
designs were checked by qualified
components and delivered Nails - 1" and 4" (25, 100mm) 4Kg
structural engineers from partner
them to site. Plastic sheet 4m x 5m 2
organisations both in Haiti and
• The organisation itself Corrugated iron 2m2
Europe, who offered their services
purchased and delivered plastic Roofing nails 1Kg
to check the designs.
sheeting, hurricane strapping Cement 42.5kg 2 bags
Daily labour on construction and cement. It also provided
sites was supervised by technicians 1,126 Tool kits
truck rental for later deliveries.
who had been trained by engineers. Item Quantity
• The beneficiaries themselves
The organisation itself directly provided gravel and sand. Bucket - 20l with cover 1
monitored implementation of the • Local private sector Rope - polypropylene 10mm 15m
project and quality. manufacturers assembled Iron wire gauge 12 or 14 15m
roof trusses and frames. This Hammer carpenters 0.5kg 1
Logistics and supply allowed quality to be controlled Mallet - 1.3kg 1
Haitian companies were not before kits arrived on-site. Crowbar 45cm 1
necessarily registered, paying tax, Cold chisel 20cm 1
Wire cutters 20cm 1
Dust masks 2
Gloves 1
hacksaw 30cm 1
Hacksaw blades 30cm 4
Roofing nails 25mm 50
Wood saw 50cm 1
Extension built by a family to upgrade a transitional shelter. Chisel 3cm 1
Photo: Mildred Beliard, CARE Nails - 1 inch (25mm) 2Kg
www.ShelterCaseStudies.org 19
of 15 - 35 m²
IT
I
January 12th
2010 –– Earthquake
Project description
The programme provided safe and improved housing which helped people to leave the camps and allowed
them to restart the recovery process. The programme included: 1) damage assessment, 2) house repairs 3)
public communication and training manuals 4) training.
20
Many buildings that were tagged yellow could be repaired at a lower cost than
building a new transitional shelter.
Photos: Chiara Jasna Vaccaro
Before the earthquake afraid that it was. Large numbers of Haitian engineers were trained
(See “A.4 Haiti - 2010 - Earth- people would leave camps and tents to conduct the evaluation. They
quake - Overview”, p.12.) and return to their homes if they were then sent in groups to assess
could be sure that their houses were the structures in a neighbourhood.
Prior to the earthquake, there
safe. The engineer would use a PDA to
were no enforceable building codes
photograph each building and take
and no inspections. As a result Implementation its GPS coordinates.
homeowners could build as cheaply The programme was divided in
and therefore insecurely, as they four separate components. They then inspected every room
chose. The same was true for urban of the building, and completed a
planning and zoning. Houses were 1) Damage assessments short questionnaire on the PDA.
regularly built into existing roads, on Damage assessments were im- At the end of the inspection, each
steep, unstable slopes, or in ravines plemented working closely with the building was spray-painted with a
prone to flash floods. Ministry of Public Works (known highly visible red amber or green
by its french acronym MTPTC). The tag. Each engineer was able to
Most structures were built in survey was conducted by teams of inspect an average of 10 structures
stages as and when money was engineers. Each team had between a day. At the end of each day, the
available. Additional floors and one and fifteen engineers. During data was downloaded directly into
rooms were often added without the project there were up to 18 the central database and used to
checking the original foundations or teams at any one time; a total of create a map.
structures. Entire neighbourhoods 270 Haitian engineers.
were built and developed without To standardise assessments, the
planning. The assessment tagged buildings ATC20 form was modified for use
according to the damage using the in Haiti. The ATC20 is the standard
The main problem with construc- following “traffic light” system: form used in California to rapidly
tion in Haiti is that the structures are
assess earthquake damage.
too brittle. Almost all the structures • green - safe for use,
are built out of masonry blocks with • yellow - damaged, but stable During the assessment, over
reinforced concrete columns and (needing minor repairs to be 400,000 structures were tagged;
beams. made useable), this was nearly every building in the
• red - unstable, either major Port-au-Prince metropolitan area
After the earthquake repairs or demolition and that was impacted by the earth-
An international seismic engi- rebuilding required. quake.
neering company was brought to
Haiti a week after the earthquake to
help the organisation with the early
response. Initially the focus was on
the main government buildings as
well as the main hotels and factories.
Many people were sleeping under
Buildings were sprayed with green, yellow or red markings according to the
tarpaulins not because their house level of damage sustained.
was unsafe, but because they were Photos: Joseph Ashmore
www.ShelterCaseStudies.org 21
22
January 12th
2010 - –– Earthquake
Project description
This organisation ran several projects focused on supporting economic, social, and political recovery. Shelter
assistance was delivered through a variety of “shelter solutions”, including traditional wooden framed
transitional shelter construction, steel framed transitional shelter construction, supporting host families
through a livelihoods-based incentive system, and the removal of rubble. The projects targeted those who
decided to stay in or around their homes of origin.
www.ShelterCaseStudies.org 23
A steel framed shelter converted into a shop. Family in a timber framed shelter.
Photo: CHF International Photo: CHF International
24
Projects were implemented with the common goal of encouraging affected families to stay in their com-
munities of origin to depressurise formal or informal camps.
Photo: CHF International
cement. The families were also tion and expenditure of household Local teams were responsible
expected to help clear rubble in livelihoods grants. for managing and tracking shelter
preparation for the arrival of the components from the assembly fa-
shelter. Logistics cilities. In many cases, steel frame
Existing relationships with shelter components were trans-
Since the project was only brokers and familiarity with customs ported to individual building sites
funded to provide a metal sheet systems built over the previous years by groups of labourers.
roof and a tarpaulin as side helped more rapid procurement of
covering, it was left to the families materials required for the wooden
to build more durable walls. This shelter. Local vendors sourced
Materials list
lead to some issues between the timber in bulk from the USA and For 1500 Wooden Shelters
organisation and the beneficiaries. the Dominican Republic, and deliv- Timbers 2”x4”x12’ yellow 9,000
ered directly to warehouses. pine (50 x 100mm x 3.7m)
Host family Timbers 2”x4”x14’ yellow 11,500
Rather than distinguish between Shelter managers submitted pine (50 x 100mm x 4.3m)
the displaced and the host families, order forms for each project site for Timbers 2”x2”x12 yellow 10,500
the project viewed the combined remaining materials such as nails, pine (50 x 50mm x 3.7m)
households as one household unit cement, and iron sheeting. Corrugated iron roof 5,000
so that the economic assistance sheeting, 28 gauge. 12'
Shelter mobilisers and team lengths (3.7m).
would be tailored to the needs of
leaders organised the delivery of Portland cement ( 42,5 Kg) 2,500 bags
both families and agreed upon by
specific material quantities to con- Hinges 4" (100mm) 3,000 pairs
both the displaced and hosting
struction sites on a daily or weekly Sliding lock 1,500
heads of household.
basis, to reduce the possibility of Nails 3" (75mm) 900 kg
Each household unit was offered graft and wastage. Nails 4"(100mm) 900 kg
a choice of vouchers that could be Roofing nails ( Umbrella 900 kg
Customs delays resulted in
spent on a variety of needs, includ- Type)
some interruptions in the supply
ing: tuition, household supplies Doors and windows 1,500
chain, and other materials such
and groceries, medicines, and small Staples (boxes of 1000 1,000 boxes
as sand and plastic sheeting were staples)
business re-stocking.
also delayed due to high demand Mosquito nets metalic type 50 Rolls
Project staff worked with each among non-government organi-
household to select the vouchers sations and slow-moving customs
needed to support the joint family processing. Host Family Livelihoods Grant Options
unit. Both families signed tri-partite Small business Through a selection
agreements with the organisation grants process with a committee
with beneficiaries
and a local government representa- submitting business plans
tive to document their cooperation, Household Buckets, cleaning supplies,
agreement, and intent to mitigate supplies cooking supplies
any arguments with local officials. Fees for tuition Direct payment to schools
through vouchers
Each household unit received School supplies school books, pens, paper,
800 USD to support the host family etc.
arrangement for a minimum of four Work tools hammers, drills, nails,
months. In most cases, the arrange- Wood framed shelters under con- paint, brushes, etc.
ment lasted long past the distribu- struction.
Photo: CHF International
www.ShelterCaseStudies.org 25
Shelter size:
IT
I
Project description
This project built progressive shelter in two phases: a first emergency response (structure covered with
tarpaulin) and a second durable solution (permanent housing with cement cladding). The project included
safer construction awareness activities and safer construction trainings. The shelter project was the beginning
of an integrated programme that also included water and sanitation, hygiene promotion, health, disaster
preparedness and livelihoods projects.
26
www.ShelterCaseStudies.org 27
Shelter construction was part • no water and sanitation. Other materials were sourced locally
of an integrated programme to • Vulnerability criteria: and transported by trucks to site.
support affected households and • number of dependants,
Due to lack of understanding
communities, access to water and elderly, or handicapped
of the local construction materi-
sanitation was later improved. people or children,
als market it was decided to locally
There were plans to drill bore holes, • single-parent families,
purchase some materials that the
to provide 70 litres of water per • no monthly income.
local market could not provide
person per day.
Technical solutions easily. This resulted in construction
Selection of beneficiaries The shelter had a galvanised delays.
Coordinated project assess- steel frame with a mono-pitch roof
and a raised floor. The shelter was Materials list
ments started one month after
3 x 6m on plan and had 6 columns Materials Quantity
the disaster intervention in areas
agreed through coordinating with spaced on a 3m grid, fixed to rec- Cement (42.5kg bags) 3 bags
other organisations. 3,960 families tangular reinforced concrete foun- Sand 0.38 m3
living in rural and semi-urban areas dations using a base plate and four Gravel (20mm aggregate) 0.38 m3
of Leogane were targeted. ordinary bolts per base. The shelter Iron bars 12 mm 36 m
could be demounted and founda- Column base plate (300mm 6 pieces
All of the families of the in- tion bolts cut to reuse the frame. x300mm x6mm plate)
tervention areas were surveyed. Steel 2mm (80mm x80mm) 27.65m
Since this was an integrated pro- The main structure was made
Floor beams 2mm (40mm x 100.9m
gramme, shelter support was not from three primary frames spanning 40mm)
only provided to families directly in the transverse direction with rec- Window and door framing 9.9m
affected by the earthquake, but tangular hollow section columns. (32.5mmx100mm)
also to families whose houses did The roof cladding was corrugated Plywood door (1.94m x 0.7m) 1 piece
not achieve a certain minimum steel sheeting nailed to steel sec- Plywood flooring (21.8thk) 18 m2
habitability criteria. The aim was to ondary roof members spanning Steel sheeting (0.75m x 1.83m) 18 pieces
avoid creating inequalities within between the three primary frames. Plastic sheeting (6m x 4m) 4 pieces
the communities. Timber studs are screwed to the Mosquito net 8 m2
steel members and the tarpaulin Bolts, nuts + washers (20, 10, 200 pieces
Selection criteria 6.25 d.)
(emergency response) or the per-
The following selection criteria Brackets (35wide, 70+20legs, 70 pieces
forated metal sheet of the cement 2thk)
were used:
cladding (durable solution) attached Hurricane straps – angles 36 pieces
• Families whose main residence to them. Additional timber sub- (75x75)
became uninhabitable because framing is used to form windows Self tapping screws 75 pieces
of the earthquake. and doors. Nails (10, 8, 4 d.) 22.7 kg
• Families whose house does Hinges 3 pieces
The intention was that the struc-
not achieve a certain minimal Door latch + padlock 1 piece
ture could be used in a modular
condition of habitability, even if Cement cladding:
manner, putting two side by side to
it has not been affected directly Perforated metal sheet 27 pieces
form a double pitched roof struc-
by the earthquake. These
ture of 36m2. Cement (42.5kg bags) 16 bags
included:
Sand 1.25 m3
• lack of space in relation to Logistics and supply Natural fibre 0.34 m3
the number of people who Steel frames were procured in-
live there, ternationally and shipped to Haiti.
28
Project description
Families were relocated from a spontaneous settlement in the Haitian capital to a new planned camp in an
area called Corail 20km away. The initial establishment of the camp was according to a carefully considered
plan and relocation took place within a month. As with many sites in Haiti, two years after the earthquake,
the future for the camp based population remained unclear.
www.ShelterCaseStudies.org 29
30
Tents provided initial shelter at the site. This was later upgraded to transtional shelters.
Photo: Shaun Scales / NRC
Thousands of people spontaneously moved into the land surrounding the planned sites at Corail, many building durable
houses. This spontaneous settlement was not planned.
Photo: Michelle Dupont
www.ShelterCaseStudies.org 31
January 12th
2010 –– Earthquake
Project description
The project targeted displaced disabled people in rural locations in the south of Haiti. The project used a
participatory approach to build durable shelters. The project re-engineered a well known traditional technique
known as clissade making it more durable, suitable for mass assembly and later upgrade by beneficiaries.
32
Before the earthquake other non-governmental organisa- staff of over 150 people working in
See “Haiti - 2010 - Earthquake - tions. Once designed, the next three the workshop, on site, in logistics
Overview”, p.12. months were spent negotiating and as social mobilisers.
with donors, tendering, organising
Before the earthquake, the Day Stage Worker days
logistics and preparing workshops.
majority of Haitian families who 1 Ground 2 x technical
The workshop was designed and
lived in rural areas lived in self- preparation advisor,
organised with a chain of produc- 6 x beneficiaries
built houses. Many were built
tion producing around 30 shelters 2 Digging 6 x beneficiaries
using clissade, a Haitian technique
per week with almost 45 persons foundatrions
of weaving bars of palm wood to
working inside. 3 Bolting and fixing 1 x chief carpenter
make walls. These walls were later columns 1 x chief mason
covered by mud and cement. The The programme included a sani- 6 x beneficiaries
roof was covered with corrugated tation component providing with 6 x labourers
zinc. access to latrines or an adapted san- 4 Embankments 6x workers
itation solution. Both the shelters 5 Installation 1 x chief carpenter
After the earthquake and the sanitation component were of panels and 6 x beneficiaries
carpentry 3 x workers
In general, the clissade houses adapted to the disability of the ben- 6 Paving and 1 x chief mason
resisted the earthquake much better eficiaries of the shelter. drainage 6 x beneficiaries
than the concrete houses. Where 3 x workers
they were damaged in the earth- To build the shelters, 60 USD
7 Fixing roof 1 x chief carpenter
quake, the injuries to the occupants was given to the beneficiaries to windows and 6 x beneficiaries
were not as severe as those caused pay local workers. The organisation doors 3 x workers
by collapsing concrete houses. provided skilled workers to lead the
construction. Selection of beneficiaries
Pilot shelter The project targeted vulnerable
Less than 40% of the families
The project began with a par- families affected by the earthquake,
owned their land. For these families,
ticipatory process that lasted 10 including people with disabilities.
a multi-party document was signed
days. During this time, community A survey form was prepared to
to keep the beneficiary on the land
groups were organised in a remote select the most vulnerable people
for free for at least for 3 years. This
village. The focus was on under- amongst those who were referred
was signed by the beneficiary, the
standing the daily activities of each to the organisation. A social officer
landowner, the community leader,
member of the family, including worked in close collaboration with
the mayor and the organisation.
working, cooking and sleeping. the organisations field office, with
After 3 years, the beneficiary will
This process lead to a shelter design other non-governmental organisa-
remain the owner of the shelter and
being developed that could be used tions referring families with disabil-
the owner will keep the latrine.
for a pilot shelter. ity cases and with local organisa-
At its peak, the project had a tions and associations.
A location for building the pilot
shelter had to be negotiated with
the local authority. It was intended
that the pilot shelter would be
useful for the community. In the
end it became a treatment centre
for disabled people.
Once a site was identified, it took
another 10 days to organise teams
and materials to build. The pilot
shelter allowed different technical
solutions to be tested. Different
technical and design corrections
were made to the pilot in order to
improve it and to fit it in the budget.
The shelter was assessed by struc-
tural engineers offered by another
organisation. Specific changes
including additional bracing and
hurricane straps were required to
ensure that it could withstand 100
mph (161 Km/h) peak wind speed.
The shelter was later adopted by
The shelters were built using a traditional technology known as clissade.
the local authority and by several
Photo: David Sacca
www.ShelterCaseStudies.org 33
A traditional shelter that survived the earthquake. A completed shelter, based on vernacular styles.
Photo: Joseph Ashmore Photo: David Sacca
Some areas were difficult to access and materials needed to be transported by hand.
Photo: Olivier Dorighel
34
Ganzague
IT
8 months - –– Decongestion of
camps: Carradeux
Port au Prince
6 months - –– Project start
January 12th
2010 –– Earthquake
Project description
The project supported people to leave overcrowded camps and encouraged them to lead their own
recovery process. It provided transitional shelters for those with land, cash for those who needed to rent,
and relocation grants for those who moved to different areas. It also subsidised health care and provided
livelihoods grants which were used to help re-establish businesses, or to support children going to school.
Camp decongestion required at least one year of monitoring and support after families had relocated.
www.ShelterCaseStudies.org 35
Background
See “A.4 Haiti - 2010 - Earth-
quake - Overview”, p. 12.
36
The project included support for livelihoods , support getting children back to school and access to improved health care.
Photo: Julien Goldstein
Monitoring and
evaluation
The organisation was asked to
intervene in the camps that it is
The organisation provided transitional shelters for those with land to build
working in either by the govern- on. It provided cash grants to help people other rent or resettle elsewhere.
ment, local organisations that were Photo: Julien Goldstein
involved there or by the commu-
nities themselves. In some cases
camps under threat of eviction
asked the organisation to help.
All families in the camps were
eligible for one of the support
options above. The focus was on
people without a land title. After
registration, people were respon-
sible for organising their preferred
accommodation.
Camp decongestion did not
end with finding shelter solutions
and moving families out of the
camp. At least one year of monitor-
ing with support in livelihoods and
vocational training followed. All families were provided with cash grants and training to allow them to
establish livelihoods.
Photo: Julien Goldstein
www.ShelterCaseStudies.org 37
Earthquake damage to a former 3 story government Emergency distributions of two tarpaulines per household
building in Padang. were made by reponding organisations.
Photo: Dave Hodgkin Photo: Dave Hodgkin
Before the earthquake “Permanent” (masonry) houses, for Whilst rural housing was
West Sumatra is located at the “Semi Permanent” (part masonry), commonly self-built, urban housing
convergence point of four tectonic and for “Non-permanent” (timber was more commonly commercially
plates and is highly prone to earth- or bamboo) houses, however, constructed with a mixture of rental
quakes. A recent earthquake in limited certification (15%) along and non-rental housing.
2007 had damaged or destroyed with poor compliance and enforce-
over 43,000 houses. ment had resulted in a low quality After the earthquake
of general construction. The disaster caused an estimated
As a result of numerous disasters, 2.3 billion USD damage to infra-
both the provincial and national In West Sumatra, most homes structure and housing. Over 30% of
government had significant experi- were privately owned particularly housing stock in the affected areas
ence. The recently formed National in rural areas, with most inherited was destroyed, making shelter a
Disaster Management Agency through matrilineal ownership priority.
deployed a Technical Advisory Team systems. They were constructed in-
to assist in the immediate response crementally often with the support Initially rural and semi-urban
and assist in the formation of it’s of remittances from male family areas were prioritised. In these
provincial equivalent. members working in the “Padang” areas, many families were living
restaurants across Indonesia and in inadequate, unsafe makeshift
Although established national shelters, under tarpaulins within
Malaysia that the area is famous
building codes, including seismic their plots of land, or staying in
for.
resistant construction guidelines for other people’s homes or gardens.
38
Concerns over the approaching An international coordination included advice on permanent re-
rainy season added to the sense of team arrived within four days of the construction. 63,000 transitional
urgency. earthquake to assist the Indonesian shelter packages were provided
government in coordinating over with a cost varying from 200 USD
Previous experiences within to 500 USD per household.
200 national and international re-
Indonesia indicated that public
sponding agencies. Later assessment highlighted a
outreach programmes on earth-
quake resistant construction were The initial shelter strategy was lack of assistance to urban areas,
important to ensure safe recon- agreed eight days after the earth- with a range of agencies then
struction. quake. The strategy focused on the running clean operations in these
distribution of tarpaulins and tents areas. Delays in material supplies
Response capacity for the emergency phase, whilst and limited capacity saw transi-
The first few weeks saw intense identifying the need for transitional tional shelter projects continuing
international media attention and shelter and disaster risk reduction for over 9 months after the earth-
an ensuing influx of internation- activities in the recovery phase. quake, overlapping significantly
al and national funds. Over 200 with the arrival of permanent re-
agencies both national and inter- Despite an overwhelming initial construction funds.
national responded rapidly. Many response to the disaster there
had prior experience in Yogya- remained a shortfall in funding, Government response
karta earthquake and/or remnant particularly in shelter and liveli- The government of Indonesia
capacity in nearby Aceh and Nias hoods. A total of 170,000 families provided grants of approximately
Island from post tsunami and earth- were supplied with emergency 1,500 USD for heavily damaged
quake projects. shelter within the first two and a houses, 1,000 USD for medium
half months. damage (from the State Budget)
However many organisations, and 100 USD for lightly damaged
including the newly formed provin- Recovery shelter houses.
cial disaster management agencies The Early recovery phase saw
quickly found themselves over- the government focusing on the Two years after the earthquake,
stretched. Many were still respond- development of permanent shelter not all funds had been released,
ing to an equivalent scale earth- assistance programs, whilst non though much of the community
quake in West Java less than one government agencies focused on had self funded reconstruction.
month before. Many of the interna- transitional shelter needs through The 2010 earthquake in the West
tional agencies soon had to relocate a range of shelter packages. Most Sumatra district of Mentawai
capacity to the Haiti earthquake. assistance was in the form of cash Islands, further stretched and
grants or material supply, to small expanded provincial response
Emergency response community groups in line with gov- capacity.
Extensive collapse of commercial ernment proposed methodology The initial government decision
and government building in Padang for community built reconstruction. to focus only on permanent shelter
resulted in an initial focus on search
Transitional shelters commonly was later reviewed in light of out-
and rescue with 21 teams of various
had timber frames. They were standing transitional shelter needs,
sizes being deployed.
mainly clad with corrugated iron or with funds then allocated to transi-
The Indonesian Government tarpaulins for roofs and tarpaulins, tional shelter in West Sumatra, and
announced an end to the search plywood or timber for walls. Shelter again in Mentawai Island and other
and rescue phase within weeks, packages commonly included a later responses.
and allocated an initial 10 million technical advice component. Many
USD to emergency relief.
www.ShelterCaseStudies.org 39
Padang
September 30th
Indonesia 2009- –– Earthquake
Project description
This project surveyed brick production and anticipated supply and demand. It was conducted one month after
the earthquake. The survey was conducted as a trial of the EMMA (Emergency Market Mapping and Analysis)
methodology. The survey findings were used to inform the adopted strategy of using cash to support the
construction of shelters that used both timber and bricks.
40
Poorly built brick-masonry buildings were a significant cause of the damage to housing.
Photo: Unknown
Background Mapping and Analysis). EMMA is 3.Air drying: The bricks are laid to
See “A.12 - Indonesia - Sumatra a tool designed to analyse markets dry in the sun for 5 days. Bricks
- 2009 - Overview” p.38. following a disaster. EMMA uses are then stacked and air dried for
background research, interviews, 30-60 days, depending upon the
After the earthquake and graphic representations of weather.
The earthquake in September market systems to help inform 4.Kiln drying: The dry bricks
2009 destroyed or damaged over humanitarian response options. are loosely stacked in open air
200,000 houses in West Sumatra. EMMA defines a market system kilns without chimneys. These
Poorly built brick based masonry as “a web of people, businesses, kilns are rectangular or circular
caused many of these buildings to structures and rules that take part in shapes. Mud is plastered around
collapse. producing, trading and consuming the outside of the brick kilns to
The Indonesian Building Code a product or service.” trap the heat from the fire, with
specifies that a "Permanent House" space for smoke to escape and
For more information on the
means masonry, “Semi Permanent” oxygen to enter. The average
EMMA methodology, download
means masonry sub walls and height of a brick kiln is 2m tall.
the EMMA Toolkit from: http://em-
timber above, whilst “non-perma- Bricks are typically kiln dried for
ma-toolkit.org
nent” means timber or bamboo. 10 – 14 days.
Brick making in Sumatra 5.Distribution: Manufacturers
Experience from previous sell their bricks directly to
Brick making involves five steps
disasters in Aceh (2005) and Yo- masons, home owners, brick
and is labour intensive.
gyakarta (2006) showed that the distributors, and / or building
demand for bricks for housing re- 1.Mixing: Clay, sand and water supply stores. Transportation
construction quickly outstrips the are mixed together in open charges are typically 30 - 60%
available supply. This often led to an pits by foot, shovels or water of the total brick price.
increase in the price of bricks, and buffalos. Larger manufacturers
/ or periodic supply shortages that use mechanical mixers. Damage to supply
delay reconstruction progress. 2.Shaping: The mix is The survey suggested that over
compressed in wooden 50 million bricks were damaged in
What is EMMA? frames. On average, a skilled the earthquake.
This research was conducted labourer can produce 1,000– The majority of the supply was
to trial EMMA (Emergency Market 1,500 bricks per day. through small scale suppliers. There
www.ShelterCaseStudies.org 41
institutions, rules,
! !
norms & trends Major disruption
SOCIAL LAND GOVERNMENT
BRICK RENTAL & NGO Partial disruption
RELATIONSHIPS
QUALITY !
!
The market chain:
market actors & their linkages
RURAL
SMALL BRICK PRODUCER HOUSEHOLDS
Volume = < 10,000 LOCAL SUPPLY
!
Number = 1000 – 2000 families STORE
Price = 350 – 600 rph
CONTRACTORS
LARGE BRICK PRODUCER DISTRIBUTOR
Volume = >160,000/month
Number = 5 -1 0
Price = 500– 600 rph
Shelter Projects 2010
LARGE SUPPLY
STORE
MEDIUM BRICK PRODUCER
www.ShelterCaseStudies.org
Volume = 25,000 –40,000/month
! URBAN
Number = 40 – 80 HOUSEHOLDS
Price = 500– 600 r p h
43
This map is for the brick market in Pedang following the earthquake. The black arrows show how bricks reached homeowners
A.13
from the different scale suppliers, and the red lines show which supply routes were interrupted.
the earthquake.
A main tool in EMMA is the Market-System Map. This helps to visualise the difference between the markets before and after
07/03/2012 14:04:43
A.14 Natural disaster
Project description
Cash was distributed to allow 750 families to build transitional shelters. It built on the initial emergency
shelter response in West Sumatra in which a package of shelter materials, toolkits, common household
supplies and basic hygiene items had been supplied to 30,000 families. Each beneficiary household received
approximately 275 USD and technical training on safe construction and minimum standards for shelter. A
partner organisation provided technical advice on construction.
44
Background ganisation requested that these were After consulting the communi-
See “A.12 - Indonesia - Sumatra - gender balanced and representative ties and other organisations working
2009 - Overview” p.38. of different age and social groups. in the sector, the Indonesian postal
service (Pos Indonesia) was selected
Distributions The committee’s role was entirely
as the best way to distribute the
The organisation initially voluntary and a Memorandum of
cash grant.
responded with non-food items. Understanding was signed with each
This started 4 days after the initial committee to lay out clearly their A mobile post office distributed
disaster. Rapid response was made roles and responsibilities. the cash grants directly to each ben-
possible by pre-positioned stocks eficiary in their village. Other organi-
Each local committee was asked
in Indonesia, held in the cities of sations had already used this system
to produce an initial list of ben-
Medan, Jogjakarta and Ambon. and its feedback was very positive.
eficiary households, whom they
Since cash grants would be distribut-
From October to December 2009, believed matched the targeting
ed directly to each beneficiary, there
shelter kits, tool kits, household and criteria. These lists were then posted
was no need to establish beneficiary
hygiene items were distributed to publicly.
groups and train their members to
30,000 families. Project staff verified each manage the funds.
Transitional shelter household recommended by the
committee and selected 620 names Market analysis
In January 2010 the organisation
for the final beneficiary lists giving In order to monitor the impact
shifted its focus to transitional shelter
priority to the most vulnerable and of the cash injection into the local
through cash programming. This was
needy, taking into account the economy; market surveys were
aimed to complement the organisa-
targeting criteria. carried out at 3 project intervals.
tion’s previous work and give earth-
A baseline market survey was
quake affected people the flexibility Implementation conducted prior to cash distribution,
to purchase materials and construct The organisation distributed cash in order to establish the local avail-
homes that met their needs. grants in two instalments. ability and cost of materials. This
The approach of providing cash was followed by two further market
An initial cash grant of 80% was
to enable self build was encouraged surveys after the disbursement of
followed by house by house moni-
by the government, as it comple- the first and second instalments of
toring to assess whether cash was
mented its own program to distrib- the cash grant.
being used for shelter and the com-
ute larger cash grants to facilitate pliance with minimum standards.
permanent construction. Technical solutions
A second grant of 20% was Technical support was provided
Selection of beneficiaries distributed. For both payments, through two different kinds of
The selection of the community vouchers were given that were later trainings:
was based on the organisation’s exchanged for cash by the mobile 1) Training facilitators
existing knowledge from its initial post office.
response and consideration for the Project staff received training
need to have a close liaison with local Delivery mechanism: from an international organisa-
authorities and key stakeholders. The organisation initially con- tion. While the training provided
sidered using a bank to distribute on T-Shelter gave staff sufficient
In each community, the organi- funds, but not all beneficiaries had grounding in good T-shelter con-
sation presented the information a bank account or could go to the struction both for community
in meetings. The communities then nearest town to collect the funds. training and monitoring, they were
elected local committees. The or-
not sufficiently equipped to assess
Many materials could be salvaged. Cash grants allowed Temporary shelter built whilst owner was awaiting labour
people to pay for materials and labour according to their to complete his house.
needs. Photo: Save the Children
Photo: Save the Children
www.ShelterCaseStudies.org 45
46
September
30th 2009- –– Earthquake
Project description
An international non-government organisation working through a local partner provided cash grants for
shelter. Conditional cash grants were given to 3,400 families in two instalments. The local partner used six
mobilisers to give technical support. Beneficiaries paid for materials and labour to build timber homes. Most
shelters took 10 weeks to build. 77% of the shelters were completed within 12 months of the earthquake.
www.ShelterCaseStudies.org 47
Beneficiary selection
The communities were selected
because the partner organisation
knew them well. The project provided cash to allow families to build what they needed.
Photo: Bill Flinn
48
Monitoring safety of the structures was very challenging given that each
family had the freedom to build according to their needs.
Photos: Bill Flinn
completion rates and additional trainings on construction and how Participants interviewed during
investment by beneficiaries. This to use salvage materials. Better con- the final evaluation stated that they
was despite their low and irregular struction was promoted through had spent between 500 USD and
incomes. minimum construction standards; 1,000 USD of their private funds in
training for field staff, beneficiaries completing the shelters, and that
The amount of cash was agreed
and masons; production of posters the grant served as an “injection of
with other agencies. It was enough
and pictures; and weekly technical motivation to a traumatised popu-
to build a shelter if supplemented
monitoring visits for all recipients of lation”.This resulted in variations in
by salvage and available resources.
the cash. final shelters with many exceeding
The cash was given in two instal-
the minimum quality standards.
ments (3 million rupiah or 330 Logistics and materials
USD). People could only get the Outsourcing material procure- It is difficult to evaluate impacts
second amount if they built a safer ment and cash distributions was on a local economy (especially
house. decided to be more effective than without baseline data) but new jobs
using the organisation’s internal as “earthquake masons” and as
Grants were delivered via the In-
and limited capacity. “chainsaw masons” were created
donesian post office in two stages.
by the project. The injection of cash
First the participants received 75% Good roads for material supplies and short time frame for building
of the funds to complete 85% of and spare local capacity for labour- briefly inflated the prices of some
the construction. In the second ers and suppliers to start up helped labour and some materials. Cash
phase, the remaining 25% of the the project. also appeared to have pushed
grant was disbursed.
It was possible that more some new businesses to open (e.g.
At the outset of the project, remote communities might have a hardware store).
families had to sign a Memoran- to pay higher prices for transport Completed homes were likely
dum of Understanding that com- and labour. However, it turned out to be “safer” than the construc-
mitted them to spend the money on that people further from roads paid tion practices that have become
timber framed transitional shelter only slightly higher prices. The fixed prevalent over the past 30 years but
and not on a permanent house or cash grant for all families was seen cannot be described as earthquake
repairing an original house. as fair. or hazard resistant. The freedom
Technical which was a strength also lead to a
Impact
Four models of shelter were wide variation in quality and diver-
Twelve months after the earth-
designed, but beneficiaries were gence from design principles.
quake: 77% (2,603) of the transi-
free to build according to minimum tional shelters were complete, 11%
standards. (369) of the shelters were incom-
A 60-strong team of mobilisers plete but in progress, 8% (265)
was established to motivate benefi- of the shelters were incomplete
ciaries to build to an agreed quality and without sufficient progress to
and on time, over 10 weeks. receive the second cash instalment,
and less than 5% (163) had not
Participants received technical been built.
www.ShelterCaseStudies.org 49
Project description
Working through international partner organisations, the lead agency was able to build 1,668 seismically
resistant winterised homes in time for winter. Homes were rebuilt using locally procured materials on the
foundations of destroyed properties. Teams of engineers, foremen, community mobilisers were hired to
ensure that all families received the material and technical expertise needed.
www.ShelterCaseStudies.org 51
Needs assessment
A house-to-house survey was
“We were born here, we conducted, assessing every recently
worked our entire life in damaged residential structure. In-
Jalalabad, we built a house, formation on structural damage, as
invested all our savings and well as pre and post conflict data
everything went in one about the household was collected.
day. We lost everything...
This house gives me hope The survey started with a pilot
to see my grandchildren phase on 3rd July 2010. The survey
again” was completed for Osh city on 10th
Project beneficiary July. The Jalalabad component
was carried out from 11–13th July
2010. Surveys included staff from
different agencies.
A man stands in front of a destroyed building in the
Preliminary results gave an indi-
neighbourhood of Kizil Kishtak in Osh, Kyrgyzstan.
Photo: Rodrigo Ordonez cation of the damage: 770 houses
in Osh city, which is an estimated
Before the conflict Selection of beneficiaries 38% of the expected overall total of
In Krygyzstan, families tend to The project was for displaced houses that were damaged in Osh.
live in compounds, containing an families. The agreed selection
Technical solutions:
average of 2 families (15 persons). criteria for beneficiaries was:
The “emergency transitional
Households are defined as ‘one or
• Displaced people living outside shelter strategy” was developed
more nuclear families related by
(homeless) or in collective by the Shelter Cluster participants
blood or law who share the same
centres. with the Ministry of Emergen-
compound. Most compounds
• People returning to (refugees cies and the State Directorate for
contain 2-4 small houses.
and IDPs) their damaged homes. Reconstruction. Technical issues
The main type of house is a • Displaced people (including such as selection of the building
“Private One Storey”, and each separated family members) materials for the emergency transi-
compound has around 300m2 who were unable to return to tional shelter was developed by the
of covered living space. The vast their homes due to damage. In Shelter Cluster Technical Working
majority of homes have plastered particular, where five or more Group which included representa-
walls and timber floors. Over 80% displaced people are living with tion of the government engineering
of the houses have a slate roof. a host family. team.
• Very vulnerable individuals, and
Nearly every house had access to The actual design of, and
their displaced or returning
water before the crisis through the support for, the emergency transi-
family, including, but not
municipal tap network. A minority tional shelter was based on the level
necessarily limited to, single
has access to a private well. Some of damage to the existing structures
parented headed households
neighbourhoods had collective and the living space requirements
and families supporting disabled
wells. of the house owners / users. The
or chronically sick people.
main building in the shelter assess-
After the conflict • Households who lost family
ment was assigned a damage cat-
The inter-ethnic violence of members in the fighting.
egorization consisting of four levels,
10-11 June 2010 prompted a large Emergency response with corresponding entitlement to
scale displacement of mostly ethnic Tents and non-food items were support:
Uzbeks from within the Kyrgyz pop- initially distributed as an emergency
ulation. The displacement occurred • Category 1) minor damage. Up
measure.
rapidly within 3-4 days. to 500 USD of materials.
A planning figure of 2,000 • Category 2) moderate damage.
After the violence of June 2010, was used for damaged / destroyed Up to 1,500 USD of materials.
more than 1,500 families were shelters. This initial figure was • Category 3) major damage. Up
without basic shelter or supplies in arrived at through analysis of to 3,000 USD of materials.
the south of Kyrgyzstan. satellite imagery, and was based on • Category 4) Full reconstruction
The damage led to large amounts two areas, Osh (1,500 households) required. Up to 5,100 USD of
of rubble and debris, including and Jalalabad (500 households). materials.
asbestos. Winter was approaching 75% of housing units assessed
and temperatures would fall signifi- were fully destroyed (category 4).
cantly below zero.
52
Left: Construction workers pour concrete for the base foundation of a home. Right: A construction worker lays bricks for a
new home.
Photo: Rodrigo Ordonez
The following prioritisation prin- theless, the basis of the project struction provided cash transfers
ciples were applied: was modular, with two-room units to affected families, after which
providing 28 m2 covered area. the affected families purchased
• A minimum of two rooms
supplies directly from government
of 14m2 each per family. If This modular approach allowed
construction suppliers. All efforts
there is more than one family for flexibility. In situations where
were made ensure that all organisa-
per compound, support for a full structure could not be
tions worked to similar approaches
additional rooms was provided. completed, the superstructure
and specifications.
• All damaged houses or could be made of panels instead of
compounds were provided with brick and mortar. All other struc- Logistics and materials
materials under a controlled tural details remained the same for The government was committed
monitoring regime, technical the sake of equity. to support the temporary / tran-
support and some support sitional housing scheme and
The majority of homes did not
for labour. This was to ensure offered tax exemptions for building
have indoor latrines prior to the
appropriate and warm living materials, warehousing and
conflict. Damage done to the
space for the family during the transport.
outside sanitary facilities was not as
harsh winter. The estimated
extensive as to the main buildings. Implementing agencies had to
requirements were limited per
However, latrines were restored, as quickly source and purchase large
level of damage.
needed, to meet the toilet needs amounts of construction material,
• Families were engaged in the
through the winter. including sand, cement, bricks and
reconstruction / repair process.
They had to provide labour Given the scale of the challenges timber. Each day, 300,000 bricks
through a self help programme. and the cost of the project, a third had to be sourced, procured and
Family composition and capacity party neutral monitoring scheme delivered, as well as 800m3 of sand,
were taken into account. was established to provide objective 600 cubic meters of gravel, 750m3
• Priority was given to the most information with regular feedback of aggregate, and many more
vulnerable households. on what was working, and what materials. In total, the programme
• All partially damaged houses needed to be improved. used around 10 million bricks as
were provided with support that well as 7,350 metric tons of cement.
The State Directorate for Recon-
ensures that their homes were
repaired to the same standard Level of assessed Number of houses Estimated cost Total amount (USD)
and quality. damage** (USD)
Shelters were designed on a
Category 1 92 500 46,000
case-by-case basis, taking into
Category 2 94 1,500 141,000
account the time frame, struc-
Category 3 271 3,000 813,000
tural integrity of the existing
Category 4 1,419 5,100* 7,236,900
foundation, availability of skilled
communal labour (masons in par- Unconfirmed category 6 unknown Up to 30,600
ticular for brick laying), bricks, and Total 1,876
the amount of living space required * Assuming that there are on average two families per compound
for the household members. None- ** As per preliminary results of shelter assessment
www.ShelterCaseStudies.org 53
December
6th and
20th 2009 - –– Earthquake
Project description
The project provided materials, cash grants and training to build and repair houses. The project led to national
guidelines on safer house construction that were adopted by the government. The project also provided
psychological support, hygiene promotion, sanitation facilities for households and schools, and disseminated
better building practice.
54
www.ShelterCaseStudies.org 55
Construction of houses
The Government of Malawi had
already produced designs for rural
housing and these designs were
adopted and modified to improve
structural performance.
Every beneficiary was given a
range of designs to choose from.
They were given the possibility to
make further modifications so long
as these met the design guidelines.
Both householders and artisans
were provided training to ensure
Houses were built through cash grants. Cash was transferred by mobile phone.
that important construction details Families were encouraged to purchase work in groups to obtain lower prices.
and methods were implemented. Photo: Jamie Richardson
The organisation provided con-
struction supervisors to monitor process the application, and to have Logistics and supply
and assist the construction process. officers from government and the Beneficiaries were encouraged
Cash grants were provided to the organisation verify the applications. and supported to buy materials
householder to purchase materials An appeals process was established in groups. The suppliers could
and pay for labour. Payments were to allow for the review of an appli- then maximise efficiency and
made in tranches aligned with the cation. minimise rates charged by deliver-
phases of construction. The house- This method of selection ing in quantity. The organisation
holder was responsible for the con- empowered the communities, provided four wheel drive vehicles
struction. allowed government to have re- for transport when required.
The houses were constructed sponsibility for the administration of Local suppliers were used for all
using locally made burnt brick, mud the recovery process, and enabled materials other than the doors and
and cement mortar, timber for the the organisation to provide support windows, which went out to tender
roof structure and joinery, and iron and monitoring of the process. and were purchased in the capital.
sheeting for roofs. Technical solutions Materials list
House repairs There were many constraints Materials Quantity
The construction supervisors, in terms of available materials, Timber 1"x8" (25x200mm) 17
with the householder, surveyed the financial resources, skill level, and Bricks (230 x115x 75mm) 8400
houses to identify the repairs and cultural aspiration. This led to the Wire mesh 25 m2
produce a prioritised schedule of choice of brick construction. Chlorodine (Anti-termite 1l
treatment)
work and an approximate budget. Proper brick bonding, the use
Cement 22 bag
The householder was paid a of lintels to brick openings, the
Reinforcement bars 12mm x 12m 17
grant in two phases to carry out the bracing of roofs and methods in
connecting the brickwork, were not Supporting plain bars 6mm x 6m 5
work. Repairs focused on strength- Solignum (timber treatment) 1l
ening each element of the structure. previously applied. The position and
size of door and window openings Damp proof course 3
Hygiene promotion and was addressed, as was the design Quarry stones for the ring beam 3 Tonnes
sanitation of unsupported masonry such as Timber 2" x 3" (50x75mm) 16
The urban part of the recovery gables and internal partition walls. Timber 2" x 4"(50x100mm) 3
programme identified a need for Timbers 2" x 6" (50x150mm) 25
better sanitation and hygiene Cash transfers Wire nails 2" (50mm) 5Kg
practice. In addition to house The transfer of funds provided a Wire nails 3"& 4" (75,100mm) 7Kg
repairs, 250 household latrines and challenge in the rural areas, and was Wire nails 5" & 6" (125,150mm) 15Kg
school sanitation facilities were implemented through a partnership Roofing nails 18Kg
constructed. with a mobile phone company. Galvanized ridges 5
Beneficiaries were given a Iron Sheets 28 gauge x14' (4.3m) 28
Selection of beneficiaries phone and funds as credits were Sand for pointing, plastering, 8 Tonnes
The project targeted the most transferred to their phone. These flooring
vulnerable within the communities. credits could be exchanged for Cement for pointing, plastering, 26 bags
Vulnerability criteria were collabo- flooring
cash through the phone company’s
ratively identified. outlets, or exchanged for materials
Community groups were estab- at specified hardware stores.
lished to identify beneficiaries, to
56
February 22nd
2007- –– Cyclone Fávio
Project description
The project identified and tested innovative small-scale mitigation interventions for cyclones. It used
participatory approaches and focused on local capacity building in vulnerable pilot areas. The major focus of
the project was to disseminate the initiative and prepare the conditions for future replication. It also built 10
cyclone shelters.
www.ShelterCaseStudies.org 57
The project built 10 cyclone shelters, but used the project to advocate nationally
for safer construction.
Photo: Arianna Francioni
58
As part of the project a manual, posters and other materials were developed.
Credit: Eduardo Feurhake Project technical director
At national level, the building online version of the manual is now Walls were made from concrete
process was periodically presented to available and spread worldwide in blocks (first phase) and compressed
national government institutions and different languages. earth blocks (second phase).
other stakeholders (including hu-
manitarian institutions, universities Different institutions from the Different solutions for covering
and private sector), which were also government and humanitarian were tried:
invited to visit the construction site. agencies are interested in the ex-
• 3cm thick ferro-cement vaults
perience and want to replicate the
(0.70 x 6m) manufactured on
Manual developed building techniques for public fa-
the ground and then raised.
A first manual with simple recom- cilities such as schools and health
• 8 cm thick concrete vaults (3
mendations on how to improve local centres.
x 6m) using a metal formwork
construction techniques had been
The purpose of developing on the beams for easy assembly
developed before the cyclone.
technical manuals and implement- and disassembly.
With the experiences gained ing pilot projects was to ultimately • Self-supporting dome made
during construction of the proto- influence national and local policies, with compressed earth blocks.
types, the existing manual “Building so that proper building techniques The houses were built by local
with Winds” was reviewed and and be integrated in the codes and masons (the beneficiaries are very
reprinted. It was targeted at tech- regulations. old or handicapped people so they
nicians and local communities. It could not participate in the con-
included simple graphic designs and Technical solutions
struction).
explanatory texts. Its content was The cyclone shelters were built
used for developing posters that can with reinforced concrete structure
be put up in the communities. The (foundations, columns and beams).
www.ShelterCaseStudies.org 59
60
As part of the project a game was developed to share learning surrounding the threat of cyclones and flooding.
Credit: Eduardo Feurhake
Project description
850 shelters were built and 800 shelters were retrofitted. All 1,650 shelters were provided with a latrine and
a ceramic jar for water collection. The project aimed to address multiple issues of security, shelter recovery,
livelihoods and future disaster resilience to provide a sustainable and holistic solution for the affected
population. The project was implemented through the “People’s Process” where people organise themselves
to identify and prioritise their needs and together take decisions on their recovery.
www.ShelterCaseStudies.org 61
After the disaster to represent them on the Village Priority was given to people cur-
Cyclone Nargis hit Myanmar in Recovery Committees. The commit- rently living in structurally unsafe
May 2008 damaging or destroy- tees worked directly with the imple- dwellings such as tents, camps or
ing an estimated 800,000 houses. menting agency during the project. makeshift huts precariously con-
450,000 of these were totally de- structed from weak, low quality
The committees were generally
stroyed. Damage was caused by a and/or temporary materials like tar-
comprised of 10 to 12 members,
combination of high winds and a paulin roofing. All of these families
of which 4 members occupied the
storm surge up to 4m tall in coastal and individuals had faced acute
leadership positions of Chairman,
areas. water and sanitation problems.
Secretary, Treasurer, and Assistant
Village selection Treasurer. Of the 287 members Training of carpenters
The 50 worst affected villages in of the 32 committees, 46% were Selection of carpenters began
Dedaye were selected for commu- women, and 42% of members as soon as villages were selected.
nity-wide interventions. Of these 50 in management positions were Training began during the third
villages, 32 were selected. Selection women. week of August 2010. The training
was based on damage assessments, Training was provided to guide emphasised cylone-resistant build-
perceived vulnerability to future members in best practices for com- ing techniques, consistent with the
cyclones and flooding. The selec- mittees, such as ensuring repre- goal of “building back safer”.
tion was based on the experience sentation of all village inhabitants, The basic criteria for selection
of Nargis and other more recent training on quality control, procure- of carpenters, as identified by the
storms. ment, finance and bookkeeping. committees, included that the can-
The villages selected were To ensure fairness of the procure- didates come from the beneficiary
located in relatively inaccessible ment and certification process, lists village, maintain a strong sense of
areas and had benefitted the least of materials and local labour wages community spirit and service, and
from aid and recovery efforts by and charges were obtained from practice carpentry or a similar trade
other humanitarian organisations township and village authorities as a livelihood activity.
during the two years following and upheld during the implementa-
tion process. A total of 96 carpenters were
Nargis. trained, and each trainee received a
Village recovery Selection of beneficiaries tool kit containing 21 tools.
Within villages, the community
committee Community contracts
members were responsible for se-
Community mobilisers visited Once designs for house con-
lecting the individual beneficiar-
the affected areas to establish a struction / retrofitting were agreed
ies. The basic selection criteria was
rapport within the communities and upon, 32 Community Contracts
that the families and individuals
to help to organise mass meetings were signed with the 32 commit-
were not capable of repairing or
during which residents were en- tees. These specified the work to
rebuilding their own homes. This
couraged to understand the need be performed, its duration and the
included, for example, female-
to organise themselves. schedule of payments.
headed household, widows, the
At these meetings, the com- elderly and persons with disabilities
munities nominated the individuals that had no family support.
62
Women’s participation the use of a system of settling and waste goes directly into a septic
The project gave equal attention ponds. tank before it can attract flies or
to involvement of local women in other pests. Very little maintenance
Crude oil was used as a wood
target areas. Out of 287 members is required for these units. They can
preservative only for key structural
of the Village Reconstruction Com- be flushed with water.
components of the shelter. Only
mittees, 46% were women. the exact amount of crude oil Hygiene education had previ-
Women community facilitators needed was bought. ously been given to all communi-
played key roles in empowering ties.
Complementary activities
and involving local women in activi- Communities were also engaged
The programme had house-
ties of the programme in the field. in upgrading village roads and foot-
hold water and sanitation facilities
Some committees had actively mo- paths, upgrading or constructing
built in to the budget, so that every
bilised women in procuring, super- village flood protection dykes and
household receiving shelter support
vising and monitoring the retrofit- embankments, upgrading and con-
also received a water storage jar
ting and construction of shelters in struction of small bridges and pond
and a latrine.
their villages. renovations.
The latrines provided are called
Women participating in pur-
“Fly-proof Latrines” because the
chasing and transportation of con-
toilet is covered with a wooden lid
struction materials, land cleaning
and levelling, construction, supervi-
sion and monitoring of works and
management of funding, gained
confidence and benefited from
learning programme implementa-
tion activities.
In all village reconstruction
committees, the treasurers were
women.
Environmental mitigation
Materials used such as toddy
palm and bamboo are natural
products and are sustainable
sources of timber (growing locally
and quickly). While concrete was
only used for the footings of the
shelter, the mixing of concrete can
contaminate water sources if care is
not taken. Carpenters and masons
were trained to avoid this through Shelters were built using locally available materials including toddy palm and
bamboo.
Photo: UN-Habitat Veronica Wijaya
www.ShelterCaseStudies.org 63
Project description
The project constructed 533 shelters by providing materials and carpenters, and was in response to a review
one year after the cyclone which found many families remaining in poor shelter. The project had a significant
training component, but had significant issues with procurement of materials of suitable quality.
Strengths and weaknesses quality and timeliness of materials. Using local suppliers
99 The beneficiaries who received support were later in the project reduced these issues.
pleased with their new houses. 88 The bill of quantities should have been better
99 The training of the carpenters was efficient and the defined.
work was well organised. This is particularly in evidence 88 There were missed opportunities to engage the
in the consistent good standard of construction. beneficiaries in making the bamboo mats for walls and
99 The houses are much stronger than contemporary floors and in preparing the thatching panels.
houses built by families on their own. 88 The project only provided shelters for families who
99 The beneficiary families were familiar with the had land to build on.
key principles of safer construction, and were able 88 The beneficiaries think the house will last 4 to 5
to explain the majority of the points. However it years, but some components will have to be changed
was not clear how many non-beneficiaries learnt the before that time.
techniques. -- Families said that the size of the house is fine for
88 Some families were not entirely happy about the a quite small family, but for a large family it is a bit
beneficiary selection process. It would have benefitted cramped and they wished to add on extensions. By the
from more transparency and community participation. end of the project, many families were already adding
88 Construction materials supplies and quality are the a small extension to the rear of the house.
weakest point of the project. Yangon based suppliers
were initially used, and there were problems with
64
A model house built as part of the project to illustrate improved cyclone resistant techniques.
Photo: Tina Salsbury
Before the cyclone In May 2009, a review showed full scale house was then built in
The four villages in the project that the majority of the households Yangon over four days so that all
area were home to 4,213 house- that reported severe and complete the details could be worked out.
holds. The region is largely flat and damage to their house could • Training on fraud awareness, on
low lying, with salt flats and paddy not undertake repairs due to the accountability and humanitarian
fields, and is divided by streams and absence of cash or materials. accountability partnership
a few navigable waterways. Many principles. Guidelines were
Very few of the houses built after
houses were in sites that were provided for activity and financial
the cyclone incorporated significant
exposed to the wind. reporting.
disaster risks reduction features.
The main livelihood activi- There was a lack of bracing, con- There were requirements for
ties were fishing, fish drying, salt nections were not good, and many monthly reporting, but in practice
production, coconuts, rice, stone roofs had too flat a pitch. this was not very detailed. This made
cutting and stove production, and it difficult to clarify questions relating
some vegetable production. The in- Implementation to the selection of beneficiaries that
habitants were poor and had a low The project initially targeted 569 arose later.
capacity to improve their homes households, focusing on the most
vulnerable families, to assist with The international organisation
without support.
the provision of materials and the had a full time engineer to oversee
Most housing had a framed construction of shelters that are the project. It also conducted
structure, bamboo secondary struc- disaster resilient. Subsequently, support missions for technical and
tures with thatched roofing and the number of households was administrative control.
thatched walls. Some houses had adjusted to 533, taking account
sawn timber frames and plank walls Training
of revised construction costs at the Through seven workshops, of
with corrugated galvanised iron start of the project.
(CGI) roofing. There were a few which two in Deedukone and the
masonry or stone block houses. Institutional setup rest in five other villages, a total of
The international organisation 607 people were trained (carpen-
Houses did not incorporate ters, beneficiaries, local authorities
would partner with a local community
any features designed to resist the and leaders). 46 village leaders were
based organisation which had been
impact of high winds. They relied given information about the prin-
working on the island in support of
on vertical posts for strength, ciples of safe construction at the
local families.
but many of these snapped off at beginning of the project.
ground level. At the beginning of the project,
the international organisation trained The project reached 2,607 people
After the cyclone the implementing organisation in: through the awareness raising activi-
One year after the cyclone, ties. 83% of these were non-benefi-
120,000 families were still living in • Safe construction: this covered ciaries of the project.
inadequate shelter that was neither the technical issues related to
safe houses – which resulted in 1,148 people participated in a
sufficient to protect families against
making some changes to the competition about the safer con-
the current monsoon, nor able to
proposed design of the house. A struction principles, with 115 people
resist any future cyclones.
www.ShelterCaseStudies.org 65
Training focussed on key messages such as making Households were later able to upgrade their shelters.
good connections.
Photo: Tina Salsbury
66
Project description
The lead organisation worked with six partners and established community committees (jirga) to provide
shelter for people returning to damaged or destroyed houses. Kits for constructing transitional shelters,
including a kitchen and latrine, were distributed. Households were given cash towards the construction cost
on completion of their house.
www.ShelterCaseStudies.org 67
Transitional shelters were built for people returning after being displaced by conflict. Material and cash were provided and
skilled workers were hired by partner organisations.
Photo: Schellenberg
68
Land allocation
To receive shelter assistance,
a family required a plot of land.
Implementing agencies had no
responsibility for the provision of
land so families without land relied
on the Jirga to negotiate with the
community, allocate land and
resolve any community conflicts.
Material distribution
A group of transitional shelters showing external purdah walls.
The implementing organisations Photo: Arif
were responsible for transporting
construction materials from the • Monitoring of project progress. Initially these flood affected
central warehouse to the villages, • Issuing completion certificates. families were provided with
and on to the site where the shelter • Coordinating with local emergency non-food items and
would be built. Once delivered to authorities, the army and were sheltered in public buildings,
the site, security of the materials implementing partners. such as schools, or in tents and
became the responsibility of the make-shift shelter on higher
beneficiary. Where people were unable to
ground.
provide the labour, the implement-
“It may not be a mansion but ing organisation was responsible KPK was the first province
our shelter certainly provides for the construction of the shelter. where the flood waters receded,
and from September 2010 to
some relief to the returning Skilled labourers, such as masons
January 2011, a total of 6,675 tran-
population, whose houses and carpenters, were organised by
sitional shelters were constructed
have been partially damaged the Jirga and paid by implementing
for people affected by both the
organisations. Cash was only paid
or even completely destroyed. floods and the conflict.
to the beneficiary once the shelter
We try our best to come was complete. In total, 16,260 transitional
up with a design that suits shelters were constructed for
the environment and local Shelter design conflict and flood affected people
Each transitional shelter was in the 16 months from when the
traditions.”
designed to accommodate a conflict returns began.
Implementing partner
family of seven, with space for a
living area, a kitchen and a latrine. Materials list
Implementation Families with more than seven Materials Quantity
The construction of the shelters members were entitled to addi- Timber pole 2" (50mm) dia 100m
was the responsibility of the benefi- tional materials in order to meet Timber pole 3" (75mm) dia 69m
ciaries. Cash was paid for unskilled minimum standards. Coverings:
labour. Technical support was Plastic sheet (5 sheets) 120m2
provided by the implementing Within one year of the conflict, Bamboo mat 61m2
agencies and social support came a total of 9,585 transitional shelters Nails 6” (150mm) (number 2) 6kg
from the local Jirga. were constructed. 5” nails (125mm) 5kg
2” nails (50mm) 3kg
nail caps 1.5”dia (40mm) 3kg
One partner organisation was After the flooding Rubber washers 1.5”dia (40mm) 150
responsible for overall technical as- In July 2010, floods inundated Hinge, hold fast iron size 4” 2
sistance of the project. This included many areas of Pakistan displacing (100mm)
the following activities: Nails 1.5” (40mm) 1
millions of people. This led to the
Doors (using poplar timber)
• Preparation of shelter standards construction of shelters for people Timber 3”x1” (75 x 25mm) 26m
and bill of quantities. displaced by the conflict being Timber 3” x 1.5” (75 x 40mm) 6.5m
put on hold until the flood waters Timber 2” x 2”(50x50mm) 4.4m
• Quality control of materials in Bow handle 2
the central warehouse. receded. Door lock 2
• Training and orientation for field Locka (Plant matter used for 80 bundles
staff of implementing partners. insulation)
• Support for initial beneficiary Iron strip for reinforcing joints
30 gauge 12” x 1” 60
selection and re-verification of (300x2.5mm)
beneficiary lists. GI wire 16 gauge 34m
• Mobilisation of communities Corrugated Iron 26 gauge 14
and technical assistance to 10’ x 3’ (3x0.9m)
Iron sheet for ridge 24 gauge 5.5m
implementing partners. This katcha house was one of 18' x 2.5’ (5.5x0.75m)
• Construction of model shelters. those repaired.
Rope (cotton 1 head, 13mm) 55m
Photo: Arif
www.ShelterCaseStudies.org 69
Before the floods inundated. Before the end of July, flood-affected individuals stayed
Pakistan has a strong and over half a million people had with host families, 13% in col-
recent experience of dealing with been affected and the emergency lective centres, 19% in planned
humanitarian emergencies, from response began. camps, 10% in spontaneous set-
conflict displacements (including tlements and 40% returned to, or
It was another six weeks before
the Afghan refugee crisis and the remained in, their place of origin
the full extent of the floods became
2009 IDP crisis) to natural disasters by September 2010. Many schools
known. The initial United Nations
(with major earthquakes in 2005 were used as collective centres.
floods appeal was launched as
and 2008 and floods in 2007). As a waters were still rising in Punjab and Of the flood-affected areas in
result there was significant experi- Sindh in the south of the country. Pakistan, Sindh province was the
ence in dealing with the aftermath worst affected, with more than
of disasters. However much of this By mid-September 2010, the
80% of affected houses either
was focused in the north of the National Disaster Management
heavily damaged or completely
country. Authority (NDMA) estimated that
destroyed, while in Punjab province
the floods of 2010 had damaged
There were also significant stock- 65% of affected houses were
or destroyed 1.8 million households
piles of relief items some of which heavily damaged or completely
in Pakistan and that approximately
got flooded. Additionally there was destroyed.
75% of the flood devastation was
a manufacturing industry, being concentrated in Punjab and Sindh
one of the world largest manufac- Response capacity
provinces. The disaster management ca-
turers of humanitarian tents and
other key relief items. The floods led to wide scale pability of each affected province
displacement. Some people were was quite different. In Khyber Pa-
After the floods displaced for days. In other areas khtunkhwa province, where earth-
The floods began in the north flood waters took six months or quakes and other natural disasters
of Pakistan in late July 2010. longer to recede. occur more frequently, the Provin-
Heavy rains lead to flash flooding, cial Disaster Management Authority
landslides and areas becoming Surveys indicated that 9% of (PDMA) was relatively well-pre-
The floods caused significant damage to infrastructure and made 11 million people homeless.
Photo: Joseph Ashmore
70
pared to coordinate and manage Recovery shelter In total less than 150,000 one
the flood response. However, in The focus of the recovery room shelters and transitional
Punjab and Sindh provinces which strategy was on the construction shelters were built. Although this is
were the most affected, the ca- of one room shelters for those able one of the largest shelter responses
pacities were much smaller because to return to their place of origin in history, it met only a small
they had not previously managed a and transitional shelters for those fraction of the total needs.
natural disaster of this scale. people who remained displaced,
those with limited access to land, WATAN cards
As temperatures in the north
and for seasonal migrants. To support families during the
were due to fall below zero a few
relief and reconstruction phases,
months after the floods, shelter One room shelter (ORS) was the Government of Pakistan estab-
would become a lifesaving priority. defined as “a more durable solution lished a WATAN card scheme. The
However, particular focus was built at place of origin with indig- WATAN Card is an ATM card that
needed in the south as delayed enous materials and techniques.” the government was able to make
flooding meant that the response The envisaged lifespan of the one payments to. It was distributed to
was four to six week behind the room shelter was 3 to 5 years, which people in affected villages.
north and there was an urgent can be extended upon upgrading of
need to encourage an increase in the shelter. In the 12 months up to August
capacity. 2011, the government of Pakistan
Transitional shelter was defined issued WATAN Cards to 1.6 million
Although strong national shelter as “a transitional solution that households, through which they
coordination was established, the responds to temporary needs, e.g. received grants of 225 USD each.
scale of the disaster was so large for those facing extended displace-
that coordination efforts were chal- ment or those living in frequent A planned second phase would
lenged, especially at district level. flooded areas”. Transitional shelters support 1.1 million households
should have had a lifespan of at with cash grants of 450 USD per
Emergency response household.
least 1 year, and a design that
The Emergency response was
allowed for reuse of materials.
relatively swift and on a very large
scale, especially when taken in
“The floods of 2010 are being termed as SUPER FLOODS
comparison with other disasters.
for Pakistan due to their large scale devastation in the country.
Within the first six weeks of the
They have laid new benchmarks for the country both in terms of
response over 300,000 families
preparedness and response. The floods affected 21% of cultivable
had been supplied with emergency
land of the country and uprooted 20 million people from their
shelter items.
homes and lands. Shelter, being the private asset, was once
By the end of the emergency again the focus of attention for the government as well as the
response, over 1 million households humanitarian community. It was important to help people get
had been provided with a tent or back to their homes quickly to avert another food disaster in the
two tarpaulins. country...”
Despite the scale of this Waqas Hanif - Advisor National Disaster Management Authority (NDMA)
response, it only amounted to 67% and focal person for Shelter Cluster
of the total estimated need. These
shortcomings were a result of the “Nobody was prepared for the magnitude. We were trying
massive scale of the disaster, the to think big, but that was not enough. We went to provincial
shortage of funds and shortage of coordination, to hub coordination, to district coordination; this
experienced implementing partners had never happened before...”
in the south of Pakistan. Arshad Rashid - Shelter cluster coordinator Pakistan floods
PA
BEFORE THE FLOOD Al
so
A
P
1
DURING THE FLOOD - People are displaced to higher ground.
RETURNS BEGIN
RETURNS CONTINUE - Some people have lost their land and are unable to return.
Coordination
As returns continue: On
- Most people move to the land where their damaged or destroyed house was. fro
- Collective centres and many spontaneous camps close. sites and buildings are rehabilitated. wh
- Organisations consider transitional shelter as support.
Tr
be
ne
Te
inc
An illustrated version of the initial Shelter Strategy for Pakistan floods - 20th Aug 2010 loc
Dat
Som
72 Jos
31/
Damaged / Destroyed
<5,000Km2
Gilgit Baltistan
750,000
Punjab
Haiti 2010
Sindh
Houses
Balochistan
(drawn to scale)
500,000
KPK
PAK
250,000
One Room Shelters are simple traditional structure made To date, Shelter Cluster members have constructed over 40,500 one room
from mud or brick, that allow families to upgrade and extend and transitional shelters and are committed to supporting the construc-
when they have the means to do so. tion of over 247,000 in total. - a rate similar to that following other major
disasters.
Transitional Shelters are lightweight structures that can
Current commitments will, however, only meet 31% of the total need.
be relocated. They are for those who cannot return to perma-
nent land.
Large numbers of families have started rebuilding on their own the Shelter
Technical guidance for the ongoing self recovery process, Cluster’s priority is to support communities’ capacity for self-help. Other
includes outreach messages and training of skilled labour and
forms of support such as training centres are urgently required to
local engineers.
provide technical assistance and help families to rebuild more safely.
Data sources: cluster websites for Haiti and Pakistan, FTS. The designations employed and the presentation of material on maps do not imply the expression of any
Some graphics reproduced courtesy of Stanford Kay Studio.com and opinion whatsoever on the part of the Secretariat of the United Nations concerning the legal status of
JosephAshmore.org.
www.ShelterCaseStudies.org 73
any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or
boundaries.
31/03/2011
Project description
This pilot project built 175 one room shelters for flood affected families in South Pakistan. It was later
followed by a much larger scale project (building thousands of shelters over 18 months). Working through
partners, the agency provided the construction materials and paid for skilled labour. Each shelter was built
from burnt brick and had an accompanying kitchen and latrine.
74
Shelter design
The one room shelter was
designed to the following brief:
• minimum floor area of 25m2,
• separate latrine and kitchen,
• durable foundations,
• brick/ concrete block
construction with cement
mortar.
Due to site conditions and con-
sultations with the beneficiaries, Shelters were built from brick with the aim of being more disaster resistant.
modifications were made to the Photo: Kpakpo
www.ShelterCaseStudies.org 75
Design modifications
Following the pilot project, it
was agreed with the local authori-
ties that future projects would
include two structural improve-
ments:
• Walls should be strengthened
by specifying cement mortar for
the full height of the walls and
not just the bottom 0.5m.
• There should be greater
resistance to earth tremors
through the addition of a
reinforced concrete ring beam
at the top of the walls.
The modifications increased the
unit cost from around 740 USD to
over 1,100 USD.
In the 10 months after the
pilot project, significant numbers
(thousands) of one room shelters
had been completed for the floods
response.
Materials list
Materials Quantity
Excavation 8m3
Foundation: 4.6m3
3% cement & soil mixture
Brick masonry (Plinth) 1:4 4.6m3
Damp proof course 0.52m3
Brick masonry (Wall) 1:4 16.13m3
Lintel (reinforced concrete) 1.7m
(long)
Wooden girder (roof) 12.3m
Bamboo 83.4m
Mat (2.4m x 6m) 3
Plastic sheet 6mx4m 1.5
20mm thick mud plastering 32m2
Door (wooden) 1
Woven mats 12 Different stages of the construction process. Houses were built
with a bathroom and a kitchen.
Photo: Kpakpo
76
Project description
This large scale project provided cash to provide households with the means to buy materials and hire
labour. Each household received the cash in 3 tranches. Each payment was made when a group of up to 25
households constructed to an agreed level. Payments were made via an agreed focal point for each group of
households. The project was managed by 44 Implementing Partners spread over 3 provinces, most of them
local agencies.
www.ShelterCaseStudies.org 77
www.ShelterCaseStudies.org 79
Pakistan
August
2010 - –– Displacement in
Sindh
Project description
Provision of ‘One Room Core Shelter’ for flood affected vulnerable families in Jacobabad, Sindh Province,
Pakistan. This project used a staged voucher system for beneficiaries to source all materials and to pay labour.
This reduced logistical delays and greatly increased beneficiary participation. The design incorporated some
disaster risk reduction considerations whilst still using predominantly local materials and practices.
80
The one room shelters were built using vouchers to pay for both materials and labour.
Photo: Jake Zarins
Before the flood Communities were prioritised the shelters were to be built. They
See “A.22 Pakistan - 2010 - where the majority of buildings also led on shelter maintenance ini-
Floods - Overview”, p.70. prior to the flooding had been con- tiatives, the planting of trees (flood
structed using traditional materials break/plinth binding) and other
Jacobabad district has tradition-
(mud or ‘kacha’) and had been aspects of community safety and
ally not been considered a high risk
completely destroyed. improvement (e.g. danger of illegal
area for flooding; it was only due
electricity connections).
to the exceptional flood levels in Additionally those villages with
the 2010 floods (and again in 2011) higher than normal proportions Committees were also respon-
that the area was submerged and of extremely vulnerable people sible for overseeing the selection
population affected. and groups (single parent families, of vendors for the supply of bricks
persons with disabilities, the elderly, and for the monitoring of delivered
There were extremely high levels
and those with no secure land materials. This group-led pro-
of poverty before the flood with
tenure or rights) were given priority. curement allowed communities
both bonded labour and a lack of
as a whole to reject poor quality
land ownership for the majority. Selection of beneficiaries materials and negotiate timely
This greatly increased the affected The implementing organisation delivery. This empowered commu-
population’s vulnerability. used the following guidance on pri- nities and reduced the need for the
oritising vulnerable groups for as- implementing agency to be present
After the flood
sistance: during each delivery and ensured
Jacobabad district was one of
the hardest hit areas during the • Poor families with three or more unscrupulous vendors could not
flooding, with almost the entire children. take advantage.
district submerged. An estimated • Women headed households.
Beneficiary agreements
160,000 houses were destroyed • Households supporting orphans
Prior to construction each ben-
and as many as 1.4 million people or disabled families and
eficiary signed an agreement clearly
left without adequate shelter in chronically ill family members.
stating the roles and responsibilities
Jacobabad district alone. Under these criteria all benefi- of the beneficiary, the community,
The bowl shaped topogra- ciaries selected for shelter assistance and the implementing organisation.
phy prevented flood waters from were considered to be vulnerable. It highlighted that any deviation
receding and much of the land from the prescribed process or
Community committees
remained submerged many months design would be solely their respon-
Committees were formed con-
after the initial flooding. sibility.
sisting of beneficiary family repre-
Selection of villages sentatives, village leaders and local The community committees
Villages were selected through decision makers. were also responsible to ensure the
close coordination with other hu- conditions of the MoU were rein-
These committees, under
manitarian organisations, govern- forced and to assist project staff in
guidance from the community mo-
ment authorities and relevant local dealing with any dissatisfaction or
bilisation staff assisted in promoting
actors to prevent duplication, and complaints.
disaster risk reduction activities such
also with the organisation’s projects as the raising of plinths on which
in other sectors.
www.ShelterCaseStudies.org 81
If the owners had to move they would be able to keep the materials in the roof. These were the majority of the cost of the
structure.
Photo: Jake Zarins
82
October 18th
2010- –– Typhoon Megi
Project description
Vouchers were distributed to provide materials for the repair of 9,953 shelters. Two types of vouchers were
tried. Initially people could choose from a given list of materials. Due to supply issues the project was adjusted
so that people could choose the materials that they wanted up to a given value and from an approved list of
suppliers. Families also received information on how to reinforce their homes against typhoons.
www.ShelterCaseStudies.org 83
84
were visited, to assess the extent niques. During the sessions, ben- which people could obtain shelter
of damage, and check that families eficiaries were provided with posters materials.
met agreed beneficiary selection showing how to construct typhoon-
These visits ensured that shops
criteria. This was to ensure that the resistant shelters to encourage them
applied fixed pricing for basic
most vulnerable were supported to construct houses with steady
shelter items as agreed prior to dis-
and that they had not received as- foundations, and to place poles in
tribution. This helped to eliminate
sistance from other actors. concrete footings with reinforce-
the possibility of shops inflating
ment.
Shelter assistance targeted prices or overcharging beneficiaries.
families that lacked the capacity In the initial approach of
People in the project were also
to repair or rebuild their homes. commodity vouchers, carpenters
encouraged to conduct their own
In addition to this, the beneficiary were part of the project team and
independent comparison of prices,
selection criteria prioritised families participated in beneficiary orienta-
to bargain for better prices with the
headed by women without income, tion sessions. Their role extended to
shops, and to decide independently
families headed by children, persons assisting beneficiaries in selecting
from which of the recommended
with disabilities, families with young materials and guiding them when
shops to redeem their vouchers.
children or elderly family members, repairing or rebuilding their houses.
families from ethnic minorities and Though prices varied slightly
In the new approach of providing
other socially excluded groups. from shop to shop, monitoring
cash vouchers, carpenters were
showed that beneficiaries were able
Team members undertook con- no longer a part of project teams.
to select shops from which they
tinuous reverification to ensure Instead, beneficiaries were encour-
got most competitive prices and
that only deserving beneficiaries aged to engage the services of car-
therefore more materials from the
received shelter assistance. This penters independently. This was
fixed voucher amount. The shops
took into account the reality that because beneficiaries purchased
saw an opportunity to make profit
other actors could have served their choice of materials according
from larger sales volume rather
some of the targeted beneficiaries to their respective, unique needs.
than per item.
in between the initial reverification
and the period they were scheduled Logistics supply
to receive shelter materials. Throughout provision of shelter
assistance using the cash voucher
Technical solutions system, team members monitored
Before the beneficiaries received the market prices and visited des-
the materials, they attended orien- ignated shops on a regular basis
tation sessions organised by project to observe how families were
teams composed of carpenters, and obtaining shelter materials. Through
project staff. The orientation sessions this monitoring, the team was able
highlighted basic building tech- to recommend several shops from
www.ShelterCaseStudies.org 85
Project description
This project mobilised 497 volunteers to help build and repair half of the homes damaged by the floods. It
also built or repaired three schools. It managed to use donated materials and supplied families with materials
and technical assistance to support self-help home repairs and renovations.
86
Before the floods The project also mobilised meetings were held in the camp
Heavy rainfalls in June 2010 led local volunteers from the business for flood victims in August 2010 to
to flooding in Romania. Over 800 community. 34 teams consisting explain the housing project, eligibil-
homes were completely destroyed, of 497 volunteers were hosted by ity criteria, conditions for participa-
more than 7,000 damaged, and the community. These volunteers tion and future obligations towards
over 15,000 people were displaced. worked more than 3,200 hours over the project.
54 days.
Many of the people most affected At the same time, social surveys
by the floods were from low income Implementation were conducted for each family on
households, mainly old people or The city hall allocated a new plot the official victims list of the mu-
families with young children from of land for the construction of new nicipality. Through this process,
rural communities. Two-thirds of houses along with the needed infra- 43 families were initially selected
Romania’s poor live in rural areas. structure: running water, electricity, for the renovation project and 17
In times of disasters, these are the sewage and access roads. All 30 al- families for the construction of new
most vulnerable groups, as many of lotments were in one area. homes during the first phase of the
them find it hard to recover without project. The 340 other beneficiaries
additional support after losing all A warehouse was set up close were selected in the second phase
possessions and homes. to the main construction site to of the project after the project staff
receive and store in-kind donations team was strengthened and was
After the floods and materials, later distributed able to perform an initial needs as-
The first stage of the campaign throughout the resource centre. It sessment.
“Now, more than ever!” focused on also disbursed materials for the con-
fund raising. At least 60 companies struction of new houses. Two local General selection criteria took
and over 20,000 individuals raised companies were subcontracted to into account the vulnerability of the
650,000 USD in cash and 290,000 perform core/specialised work both family, the urgency of the housing
USD in construction materials. with new constructions and renova- need, willingness to relocate and
tions. volunteer. For the renovations,
Later stages of the response dealt single parent families and families
with logistics and implementation. A local project team was with at least one minor child were
The first resource centre assisted 80 formed to manage the project. It prioritised.
families to rebuild their houses. included a public relations special-
ist, a volunteer coordinator, a family The major problem for the
Later, the second resource centre selection process was that the
support officer, and a construction
was established, and it focused on number of beneficiaries kept
site manager.
the distribution of materials in the changing, especially for the new
area. In total, it assisted 240 families Selection of beneficiaries builds project. This was mainly
with timber for roof, aerated A selection committee was caused by that fact that some of the
thermal blocks, sand, cement and composed of two representatives families finally refused to relocate.
plasticiser for external and internal from the organization, a town hall Other families did not perform the
walls, polystyrene stucco plaster for representative, and a community required number of working hours
insulation and external finishes. representative. Public information on the construction which was one
www.ShelterCaseStudies.org 87
Interior of a repaired house. The project mobilised volunteers to build and repair flood damaged houses.
Photo: Ovidiu Micsik and Mihai Grigorean, Habitat for Humanity Romania
of the main conditions for partici- The houses were finished Construction workers were
pating in the project. This problem using standard quality materials deployed to provide families with
was solved when the town hall for interiors, including drywall, design and technical support in con-
approved a final official list that laminate parquetry, stoves or tiles. struction, and the use of tools and
could not be amended. The bathroom and kitchen were equipment, house to house in order
equipped with basic amenities; sinks to ensure the quality of construction
To encourage families to work
and showers. and health and safety requirements.
equally on all new constructions,
the beneficiaries were not assigned For the renovations of the
a home until the very end of the partially affected houses, the
project. solution was selected on a case by Materials list for 400 houses
case basis. The bulk of work was on
Technical solutions Materials Quantity
restoring walls, insulation, reinforc-
For the new builds, the aerated Aerated thermal 9,800 m³
ing foundations, replacing flooring blocks
thermal blocks were chosen as and internal finishes. Timber 380 m³
a solution due to availability and
climate conditions, the type of Technical assistance was Windows 120
in-kind donations and related con- provided to a large number of Cement 250,000 kg
struction costs. All 30 new houses families via the resource centres. Iron 30,900 kg
were built on the same plot of land, Families received materials for Parquetry (flooring) 1,800 m2
making it easier to develop the their own work. Transport from Doors 210
logistics and organise the produc- the warehouse to each location Polystyrene 875 m³
tion and supply chain management. was organised and offered to each
family.
Using volunteers, the project rebuilt half of the houses destroyed by the rain and flooding in Romania. Many of the volun-
teers were from outside Romania.
Photo: Ovidiu Micsik and Mihai Grigorean, Habitat for Humanity Romania
88
Sri Lanka
1983 - –– Conflict start
Project description
This owner-driven programme provided cash to support people to build houses damaged or destroyed by the
conflict. The project aimed to contribute to the sustainable rehabilitation and reconstruction in the north of Sri
Lanka. It primarily supported people who have been displaced who were resettling after the conflict.
www.ShelterCaseStudies.org 89
90
www.ShelterCaseStudies.org 91
Project description
This project helped to rebuild communities affected by earthquakes in the Kumsangir district. It also aimed
to help prepare remote rural communities against further earthquakes and mud slides. The project used
alternative and affordable construction technologies and provided loans to help families to rebuild or repair
their homes.
92
www.ShelterCaseStudies.org 93
94
September
30th 2009 - –– Tsunami
Project description
This project provided cyclone resistant transitional shelter, water supply and sanitation to 74 families who
lost their homes and elected to remain on Niuatoputapu, while waiting for assistance to re-build permanent
housing. The tsunami had destroyed the houses of more than half the island’s population. The shelter
materials and construction teams were imported from an island 600km away.
www.ShelterCaseStudies.org 95
Maka Holi
Project beneficiary
Before the tsunami The initial assessment indicated the island with the project manager
Niuatoputapu lies at the that 79 homes were destroyed and arranging the shipping.
northern edge of the Kingdom of 30 had major damage. The total
Tonga and although small, is the A contract team of carpenters
number of people affected at that
main island among the Niua group. was set up on the island with a local
point was 465. These families were
Niuatoputapu, occupied by approx- overseer. The local implementing
initially housed in tents, often on
imately 850 people, is extremely partner had a representative over-
the land of other families, or shared
remote and highly vulnerable to seeing the process, supported by
housing.
natural hazards. It has very limited the project manager. This contract
transport and communications, Implementation team accessed materials from the
and just three settlements on its The locations of houses was depot of the implementing partner
northern shore. discussed with each family sepa- via consultation with the repre-
rately to ensure that the land was sentative. They constructed the
After the tsunami either their own or that they had footings in phases, leaving time for
An earthquake measuring 8.3 consent to locate a house and toilet the concrete to cure, and then in
on the Richter scale, 300 Km north on the site. Each household signed stages, constructed the sub-frames
east of Niuatoputapu caused three an agreement that this was the and erected the shelters. There were
tsunami waves up to six metres in case. For the land closest to the sea up to three teams working on the
height. Nine lives were claimed in the town of Hihifo, there were island at one time.
and four people were left critically strong government sensitivities The project manager, local rep-
injured. to re-constructing housing in this resentative and the contractor all
The townships of Hihifo and area, and finally it was decided not had some responsibility for moni-
Falehau were severely damaged, to build on this land. toring progress and quality.
and all government houses and As there was a lack of resources To encourage householders to
offices in Hihifo were totally on Niuatoputapu, a contract for contribute, the toilet superstructure
destroyed. All essential services prefabrication of shelters, toilets was only provided once pits had
including the local hospital, airport, and water tanks was offered to been dug by households. Water
communication offices, ground a company based in the capital tanks were only provided once
and surface water were seriously Nuku’alofa, six hundred kilometres platforms had been constructed.
damaged. away. They were responsible for This was only partly successful.
shipping materials to the dock on
96
Neomai Osika
Project beneficiary
The project illustrated the challenges of running small projects on remote islands
with a small project team.
Photo: Paul Davenport
The project built shelters using contracted teams from Niuatoputapu, 600 km away.
Photo: Kathleen Walsh
www.ShelterCaseStudies.org 97
September
29th 2009 - –– Typhoon Ketsana
Project description
This permanent shelter project was implemented as part of the recovery phase of the typhoon Ketsana
response. 650 households who had lost their homes were supported through cash grants to rebuild storm/
flood resistant houses. A technical consultant was hired to support a national organisation to organise
trainings on safe housing, develop house designs and supervise the construction of houses.
98
www.ShelterCaseStudies.org 99
6.Within two months, most of the community meetings to select ben- Logistics and supply
650 houses were completed. eficiaries. The number of beneficiar- Households living in highland
Some delay was experienced ies was defined based on the criteria provinces faced problems regarding
due to heavy rain and lack of and on the allocated amount of the availability of qualified labour
access to certain communes. cash grants. force and transport of material. One
Eleven months after typhoon local company was often building
The list of beneficiaries was then
Ketsana, all houses were all houses for a selected commune.
reviewed. All beneficiaries were
completed.
verified on site and finalised by all In all other areas, families could
7.In the last month of the project
levels of the organisation in coor- easily select the builders and buy
an awareness campaign was
dination with local authorities and building materials in the commune
conducted on “safe housing”.
other community based organisa- shops with credit. Payment was
This was implemented by the
tions representatives. made after receiving the cash
organisation with the technical
10% of the beneficiaries were grants.
support of the partner. This
included the printing of 1,000 later checked through field visits. Generally speaking, all materials
calendars displaying the storm/ Once approved, the lists were were available in the localities.
flood-resistant house designs, a issued and publicly posted in each
In two provinces, due to lack of
children’s play emphasising the Commune’s People’s Committee
capacity, the material supply and
basic principles of safe housing, office.
construction was done by small
posters of the newly constructed
Technical solutions local companies paid for directly by
houses in each commune, and
The following technical issues the families. In the other provinces
the preparation of an atlas
were standardised to make the where more material and local
displaying typical houses from
houses flood/storm resistant: builders were available, the families
the seven provinces .
paid the material supplier and the
8.In December 2010, the shelter • reinforcement of the local builder directly
project was externally reviewed. foundations,
Selection of beneficiaries • reinforcement of the structure, Materials list
The organisation established the with reinforced concrete Example for a house built in Kon
selection criteria that households: columns (example: 4 steel bars Tum province:
instead of the traditional 3
Materials Quantity
• were listed on the poverty list, bars), ring beams,
• had lost their means of Gravel 3m3
• reinforcement of the links
generating income as a result of Gravel 3.7 m3
between roof structure and
the disaster, Cement 3,300 Kg
walls, and roof covering,
• had no labour force (elders, • protection of tiled roof with Sand 12 m3
family with young children (0-5 concrete ribs and of corrugated Sand 4 m3
years), pregnant and lactating iron sheets, with steel bars in Brick 6,000.00
women, disabled people, single coastal areas (with high risks of Steel bar 6mm diameter 55 Kg
female headed households), strong winds), Steel bar 8mm diameter 75 Kg
• had no significant support • doors and windows which can Steel bar 10mm diameter 120 Kg
received from other sources. be securely closed, Corrugated iron sheet 28 m3
Village chiefs and members • there should be an attic above Door 2 opening 2.46 m3
of the organisation chaired the the flood levels. Door 1.64 m3
Window 2.4 m3
Window frame 3
Lime 52 Kg
Tool 1 Kg
Steel wire 10 Kg
Paint 7 Kg
Nail 1.5 Kg
Tiles edge 54
Timber 5mmx10mm 0.36 m3
100
The project allowed families to adapt basic models of shelter to suit their needs (top).
It also provided technical guidance on safer construction (drawings and computer rendered image below).
www.ShelterCaseStudies.org 101
Photos: DWF
102
SECTION B
Historic
This section contains case studies from projects that started before the year 2000. In this edi-
tion we have an update on the Sphere project (B.1) which was launched in 1997, and the San
Fancisco earthquake from 1906 (B.2). See “Annex 1 - Index - by country” for more historic
case studies.
www.ShelterCaseStudies.org 103
104
Project description
The third revision of the Sphere Handbook was released in 2011. It built upon the previous two editions and
contained a section on “Minimum Standards in Shelter, Settlement and Non-Food Items”. This chapter is the
closest there is to consensus in humanitarian sheltering practices, and is available for download free of charge
from www.sphereproject.org.
Strengths and weaknesses
The shelter and settlements chapter of the Sphere
Hhandbook is the best place to get an overview of a
commonly accepted understanding of shelter issues.
The shelter and settlements chapter of the 2010
handbook is based on consensus surrounding
shelter provision. It was developed during a broad
consultation process.
The document can be very useful starting point
for establishing advocacy positions with donors,
governments and within organisations.
When used well, it can lead to better programming,
and as a basis for training in humanitarian shelter
provision.
The Sphere handbook is often quoted but seldom
read in detail. The shelter chapter is less than 50
pages long.
- The figure of 3.5m2 per person is often quoted as
being the sphere standard for shelter, however, it is
only a “key indicator” under the broader standard of
covered living space, and has accompanying guidance
notes.
- It can be hard to meet Sphere standards where
there are resource constraints or pre-disaster shelter is
below Sphere standards. This is now discussed in the
introduction to the handbook under "In cases where
the standards cannot be met...".
Front cover of the Sphere project 2011 handbook.
www.ShelterCaseStudies.org 105
106
www.ShelterCaseStudies.org 107
108
Project description
Following the San Francisco earthquake of 1906, shelter was initially provided in tents and with the
distribution of household items. Formal camps were established and cottages built, which people living in
them were allowed to rent and purchase at a subsidised rate. Reconstruction for some households was
supported through a system of grants and loans.
After the earthquake 5610 Cottages were built. They could be rented at relatively low cost.
The earthquake hit at 5:15 am Photo: San Francisco Public Library
on April 18. Immediately after the
earthquake, fire broke out. The fires In the first month, it handled a number of people who lost their
lasted for three days and destroyed daily average of twenty truckloads. houses.
28,000 houses, making 200,000 Goods included towels, sheets,
pillows, pillow cases, blankets, mat- Committees helped to identify
people homeless. This was nearly
tresses, stoves, cooking utensils, communities who were willing to
half of the city’s population. After
cutlery, dishes, brooms, wash tubs, accommodate those affected by the
the fire, rents rose, leading to an
washboards, boilers, irons, clothes earthquake and fires.
increase in homelessness.
lines, axes, chairs, tables, and Barracks
Very limited amounts of con- sewing machines.
struction materials could be Barracks were built by the
salvaged. It was estimated that 500 To distribute the materials, “committee on housing the
million dollars worth of property requests came from the cities’ homeless”.
was lost (more than 8 billion USD elected civilian chairmen. The army These were arranged in camps.
in current value). About 40% of handled the goods and the Red One of these camps had 18
this was insured. On average each Cross verified aid entitlements, buildings with 16 two-room apart-
citizen lost around 650 USD of following an initial registration. ments in each, separated by an 8ft
property. When the distributions ended, (2.6m) partition. The rooms were
On the day of the earthquake, those who had not received items 100 ft2 (9m2) in plan and had a front
the mayor of San Francisco issued a were given the cash value of the room with a window and a door
proclamation that authorised police articles that they had requested. and a rear room.
“to kill any and all persons found The first barracks were occupied
engaged in looting or in the com-
Relocation / host families
Immediately after the fires, large one month after the earthquake.
mission of any other crime.” The last ones were closed just over
numbers of people left the city.
From the first day of the fire, free one year later.
Distribution
The first three and a half months transport by boat and train was The barracks were often dense
of the response were led by the army. provided across the bay, down the and lead to crowded conditions
They rapidly established supply lines. peninsula, and to inland locations. with limited sanitation and privacy.
Ten days after the earthquake, they The Southern Pacific railroad
established a warehouse for second Tents
transported 300,684 free pas- Tents were provided from
hand clothing, and set it up on the sengers mainly around the San
model of a department store. the first days of the response by
Francisco bay and to elsewhere in voluntary agencies, by the sub-com-
California. This was more than the mittee on housing the homeless,
The relief operations lasted for 3 and a half months. Many people were
provided with tents. 21 formal camps were established.
Photo: San Francisco Public Library Photo: San Francisco Public Library
110
www.ShelterCaseStudies.org 111
“Souvenir hunters”; the mayor of San Francisco authorised police “to kill any and all persons found engaged in looting or
in the commission of any other crime”.
Photo: RG 46, Records of the United States Senate, National Archives
112
Annexes
The Appendices contain a list of case studies in the three editions of this report to date (“An-
nex 1 - Index - by country”, p.114), and further reading, including some of the public docu-
mentation on which this report was based (“Annex 2 - Further reading”, p.116).
113
This index is to help readers find case studies of shelter projects from Shelter Projects 2008, Shelter Projects
2009 and Shelter Projects 2010. it is sorted by country and by date. Projects are colour coded as follows:
Conflicts, complex emergencies or post conflict returns conflicts
Natural disasters
Projects implemented before the year 2000
114
115
Websites
www.disasterassessment.org
A site where members of the disaster management community can meet to exchange tools and case studies
related to disaster risk assessment.
www.oneresponse.info
The home pages of the project to establish clusters as a coordination mechanism. Includes links to the shelter
cluster and the Early Recovery Cluster. Contains further reading and links to current documents for major responses.
IFRC/ICRC Emergency relief items catalogue - website
http://procurement.ifrc.org/catalogue/
Detailed specifications of all items commonly used by IFRC and ICRC
IFRC Shelter video channel
bit.ly/ifrcshelter
Red Cross Red Crescent videos related to emergency shelter
www.reliefweb.int
Up to date information on complex emergencies and natural disasters as well as an archive of information, field
reports and situation reports from emergencies since 1996. OCHA situation reports (sitreps) and IFRC appeal
documents and operations updates have been of particular use in compiling these case studies.
SDC - Cash transfer Projects
http://www.sdc-cashprojects.ch/en/Home/Experiences/SDC_Cash_Transfer_Projects
Compilation of cash projects by SDC. Includes shelter case studies.
www.shelterlibrary.org
A library of free documents relating to transitional settlement and reconstruction.
www.sheltercluster.org
Home page of the global shelter cluster - the coordination mechanism for shelter responses. Contains links to
individual responses.
www.sphereproject.org
Download the sphere handbook, find information on trainings and other activities from the Sphere Project. The
Sphere Project aims to improve the quality of humanitarian assistance and the accountability of humanitarian
actors to their constituents, donors and affected populations.
Background Documents
Camp management project, Camp Management Toolkit 2008
Available from: www.nrc.no/camp
A comprehensive field manual for camp management organisations and stakeholders involved in camp operations.
Corsellis and Vitale, Transitional Settlement: Displaced Populations, Oxfam publishing, 2005
Available from: www.shelterlibrary.org
Guidelines for the strategic planning and implemention of settlement responses for displaced populations.
IASC, Shelter Centre, Selecting NFIs for shelter - 2008.
Available from: www.shelterlibrary.org
Provides information, case studies and guidance on how to choose the best items to distribute to those affected
by natural disaster or conflict.
ICRC/IFRC Guidelines for cash transfer programming - 2007
Available from: www.ifrc.org/
Provides information on when and how to distribute cash in disaster response.
IFRC, Guidelines for assessment in emergencies - 2008.
Available from: www.ifrc.org
Practical information and guidance on how to conduct assessments in emergencies.
IFRC Owner Driven Housing Reconstruction Guidelines (ODHR), 2010
Available from: www.ifrc.org
Guidance on the planning and implementation of assisted self help reconstruction projects.
116
117
Construir com os ventos: guiäo de construcäo para zonas de risco de ciclone, Government of Mozambique
UN Habitat.
Available from www.sheltercentre.org
IDMC/ NRC, Internal Displacement Global Overview of Trends and Developments in 2010
Available from www.internal-displacement.org
Rick Bauer & Jonathan Brass, Oxfam GB, Brick Markets in West Sumatra after the 30 September 2009 Earthquake,
an inter-agency assessment of issues in brick supply & demand 16 –19 October 2009 Pariaman & Padang Districts
- West Sumatra.
Available from www.cashlearning.org
Sacca David, Abri Transitionnel (T-shelter), Cahier des Charges, Handicap International, Département Action
d’Urgence, Haïti 2010-11.
Available from www.sheltercentre.org
Scott DiPretoro, American Red Cross, RED Card Program (Repair and Development) Chile Earthquake 2010
Beneficiary Satisfaction & Program Evaluation, 2011
Available from www.cashlearning.org
118