You are on page 1of 134

SHELTER PROJECTS

2010
By the end of 2010, over 43 million people worldwide had
been forcibly displaced due to conflict and persecution.
In addition, during 2010, over 400 natural disasters
were reported killing over 300,000 people and affecting
more than 300 million people. This made it the second
worst year since 1980. The corresponding scale of global
shelter need has required a diversity of approaches that
go beyond simple design solutions.

Spanning humanitarian responses from over 100 years,


Shelter Projects 2010 is the third annual compilation of
shelter programmes. The project summaries included
aim to illustrate some of the project options available
to organisations working in both post disaster and post
conflict situations, as well as to support learning from
the strengths and weaknesses of different projects. The
focus of this book is on projects that maximise emergency
response funds to support sustainable recovery.

This document is targeted at humanitarian managers


and shelter programme staff from local, national and
international organisations at all levels of experience.

HS Number: HS/010/12E
ISBN Number(Series): 978-92-1-131943-9
ISBN Number(Volume): 978-92-1-132426-6

UN-HABITAT
United Nations Settlements
Programme
P.O. Box 30030, GPO Nairobi, 0100, Kenya
Telephone +254 20 762 3120
Fax: +254 20 762 3477
The international Federation of Red Cross infohabitat@unhabitat.org
and Red Crescent Societies promotes the www.unhabitat.org
humaitarian activities of National Societies
among vulnerable people.

By coordinating international disaster relief


and encouraging development support
it seeks to prevent and alleviate human
suffering.

The international Federation, the National


Societies and the International Commitee
of the Red Cross togetheer constitute the
International Red Cross and Red Crescent
Movement.

International Federation of Red Cross and UNHCR


Red Crescent Societies, United Nations High Commissioner for Refugees
P.O. Box 372, P.O. Box 2500
CH-1211 Geneva 19, 1211 Geneva 2
Switzerland, Switzerland
Telephone: +41 22 730 4222,
E-mail: secretariat@ifrc.org, www.unhcr.org
www.ifrc.org
We believe 1 refugee without hope is too many.
Saving lives, changing minds. Tell the world you do too: http://do1thing.unhcr.org
Introduction Shelter Projects 2010

Foreword

SHELTER PROJECTS
2010

www.ShelterCaseStudies.org i
Foreword Foreword Introduction

Shelter Projects 2010

Published 2012

Available online from www.ShelterCaseStudies.org

Copyright for this book is retained by IFRC, UN-HABITAT and UNHCR.


Reproduction for non-profitable objectives is encouraged.

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

Back cover left to right:


© IFRC, Usman Ghani, Joseph Ashmore, UN-Habitat Veronica Wijaya, Jorge Romo, Habitat for Humanity
Tajikistan, David Sacca, Julien Goldstein.

ii
Introduction Shelter Projects 2010 Foreword

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.

Esteban Leon Graham Saunders


Risk Reduction and Recovery Coordinator Head
Shelter Cluster Focal Point Shelter & Settlements Department
Early Recovery Focal Point International Federation of Red Cross
UN-HABITAT and Red Crescent Societies

Monica Noro
Chief of UNHCR Shelter and Settlement Section
United Nations High Commissioner for Refugees

www.ShelterCaseStudies.org iii
Acknowledgements Acknowledgements Acknowledgements Introduction

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
Introduction Shelter Projects 2010 Contents

Contents
Contents
Introduction
Foreword iii
Acknowledgements iv
Contents v
Introduction vii

Section A - Case studies


A.1 Afghanistan - 2009 - Conflict returnees 3
A.2 Chile - 2010 - Earthquake 6
A.3 Grenada - 2004 - Hurricanes Ivan and Emily 9
A.4 Haiti - 2010 - Earthquake - Overview 12
A.4 Haiti - 2010 - Earthquake - Overview continued 14
A.5 Haiti - 2010 - Earthquake 17
A.6 Haiti - 2010 - Earthquake 20
A.7 Haiti - 2010 - Earthquake 23
A.8 Haiti - 2010 - Earthquake 26
A.9 Haiti - 2010 - Earthquake 29
A.10 Haiti - 2010 - Earthquake 32
A.11 Haiti - 2010 - Earthquake 35
A.12 Indonesia - Sumatra - 2009 - Overview 38
A.13 Indonesia - Sumatra - 2009 - Earthquake 40
A.14 Indonesia - Sumatra - 2009 - Earthquake 44
A.15 Indonesia - Sumatra - 2009 - Earthquake 47
A.16 Kyrgyzstan - 2010 - Conflict 51
A.17 Malawi - 2009 - Earthquake 54
A.18 Mozambique - 2007 - Cyclone Fávio 57
A.19 Myanmar - 2008 - Cyclone Nargis 61
A.20 Myanmar - 2008 - Cyclone Nargis 64
A.21 Pakistan - 2009 - Conflict 67
A.22 Pakistan - 2010 - Floods - Overview 70
A.23 Pakistan - 2010 - Floods 74
A.24 Pakistan - 2010 - Floods 77
A.25 Pakistan - 2010 - Floods 80
A.26 Philippines - 2010 - Typhoon Megi 83
A.27 Romania - 2010 - Floods 86
A.28 Sri Lanka - 2009 - Conflict Returns 89
A.29 Tajikistan - 2010 - Earthquake 92
A.30 Tonga - 2010 - Tsunami 95
A.31 Vietnam - 2009 - Typhoons Ketsana and Mirinae 98

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
Contents Contents Contents Introduction

vi
Introduction Shelter Projects 2010 Introduction

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.

Selection of case studies


The case studies were selected
using the following criteria:
• The shelter project was wholly In just two responses, Pakistan and Haiti, over 2 million tarpaulins were distrib-
uted in 2010.
or largely complete, or solid
Photo: IOM Sukkur
learnings from the project could

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
Introduction Introduction Introduction Introduction

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
Introduction Shelter Projects 2010 Introduction

Table illustrating which types of response took place in each case study

Distribution Shelter type Support methods


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
Introduction Introduction Introduction Introduction

In addition to the use of cash,


the case study from Pedang
(Indonesia 2009, A.13), illustrates
an Emergency Market Mapping
Analysis assessment (EMMA).
This aimed to better understand
markets before procurement for re-
construction began.

Needs and approaches


Most of the projects in this
book respond to a need, providing
assistance through construction,
materials, cash or training. In
addition to responding to shelter Carpenter training in Myanmar. Training and communications projects are a key
needs, some projects show inno- component of many shelter projects.
vative approaches. Examples of Photo: Veronica Wijaya
this are handling cash with credit
cards (Chile 2010, A.2) or develop- provide additional funding from In this book we use the
ing construction technologies for their own resources or from remit- terminology that was used in country
risk reduction (Mozambique 2007, tances. A case study from Pakistan for each response. Practically in each
A.18). For some of these projects (A.24) provides a good example of response, national understandings
the major impact is not in terms this - providing 300 USD in cash are developed surrounding the
of number of shelters built but the per shelter. If all of the organi- use of these terms. In some cases,
impacts on policies and behavioural sation’s recommendations were flexibility in terminology has helped
change of the at risk population. followed, each house would cost projects to take place sooner.
500 - 1000 USD - depending on the
Remittances type of construction used. The risk Interpret and contribute
According to the World Bank , 1 of such an approach is that some In reading this book, or
remittances to developing countries of the more vulnerable families can browsing relevant case studies, it is
account for three times the amount benefit less. hoped that readers will be able to
of official development assistance. draw their own lessons and identify
Additional cash flows also occur Terminology useful techniques and approaches.
within countries. In many, if not There has been a lot of academic
and practical debate surrounding Readers are encouraged to send
most disasters, remittances (money in their own projects for future
sent home from overseas) and terminology used in shelter. Addi-
tional confusions have been added editions. In this way, the humanitar-
other fund transfers outside the ian community can compile good
international relief system, are the by language translation issues.
and bad practices and hopefully
major part of responses. Issues of the definition of words implement increasingly effective
It has been difficult to find have been particularly great sur- shelter projects in the future.
examples of humanitarian shelter rounding the language used for
different phases of assistance. As Contribute at:
projects which directly engage on
the issues of remittances and cash an example the terms “transitional www.ShelterCaseStudies.org
flows within a country. shelter"2, "T-shelter", "temporary
shelter", "semi-permanent shelter"
Some of the projects which give and "incremental shelter" have all
cash assistance provide slightly less been used in responses to define
than is required to build a shelter, both types of shelters and the
implicitly expecting families to processes used.

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

x
Introduction Shelter Projects 2010 Introduction

Graph of the duration of the different case studies

A.1 Chile 2010


A.2 Grenada 2004
A.4 Haiti 2010
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.12 Indonesia 2009
A.13 Indonesia 2009 Programme duration
A.14 Indonesia 2009
Time between disaster / end
A.15 Kyrgyzstan 2010 of conflict and project start

A.16 Malawi 2009


A.17 Mozambique 2007
A.18 Myanmar 2008
A.19 Myanmar 2008
A.20 Pakistan 2009
A.22 Pakistan 2010
A.23 Pakistan 2010
A.24 Pakistan 2010
A.25 Philippines 2010
A.26 Romania 2010
A.27 Sri Lanka 2009
A.28 Tajikistan 2010
A.29 Tonga 2010
A.30 Vietnam 2009

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
Introduction Introduction Introduction Introduction

Graph of the distribution against time for major international shelter responses

1,000,000
Number of families provided with

Pakistan flood 2010


800,000
basic shelter

600,000

400,000
Haiti earthquake 2010

200,000 Myanmar cyclone Nargis 2008

Indonesia, Sumatra, earthquake 2009

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

Haiti earthquake 2010


80,000
Indonesia Jogyakarta
built

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
Introduction Shelter Projects 2010 Introduction

“The poor inhabitants were dispersed about St. George’s


Fields, and Moorfields, as far as Highgate, and several miles
in circle, some under tents, some under miserable huts
and hovels, many without a rag or any necessary utensils,
bed or board, who from delicateness, riches, and easy
accommodations in stately and well furnished houses, were
now reduced to extremest misery and poverty.”
September 5th, 1666.

“His Majesty and Council indeede took all imaginable care


for their reliefe by proclamation for the country to come
in and refresh them with provisions... they now began to
repaire into the suburbs about the Citty, where such as had
friends or opportunity got shelter for the present, to which
his Majesty’s Proclamation also invited them.”
September 7th, 1666.

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.

“Following the 1827 Great Fire of Turku, the tsar opened


warehouses of the Russian garrison to supply flour and
grain, exempted the city from taxation and the obligation to
accommodate soldiers and established a construction fund
to give interest-free loans...”

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
Introduction Introduction Introduction Introduction

xiv
Case Studies Shelter Projects 2010 Section A

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
Section A   Case Studies

2
Conflict / Complex Shelter Projects 2010 A.1

A.1 Afghanistan - 2009 - Conflict returnees


Update:

Country: Project timeline


Afghanistan
Disaster:
Afghanistan returns to Sozma
Qala camp
Returns date: Two years - –– 320 families still
present on the site
August 2009
Number of people displaced:
2002 - 2010 - over 5 million
people returned to Afghanistan
Project target population:
379 families 2009
320 families 2011
Occupancy rate on handover:
94%
Shelter size:
Covered area 38.7m2

Sozma Qala
4 months – –– Construction
complete
3 months – –– Construction start
Afganistan

–– Tents distributed

August 2009 – –– Displacement

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.

Strengths and weaknesses


99 The shelter design and fabrication proved to be hard to factor into planning for support packages.
more durable than expected. -- Although the camp was established as a short term
88 During this extended period of displacement, fix, it may well become a permanent settlement. Initial
increases in family size further complicated the land allocations and site layout will therefore have a
inevitable land disputes upon their return. This was long term impact upon the families at the site.

A Weather Mitigating Tent Structure. Left: in 2009. Right: two years later.
Photos: Richard Hamilton and Shaun Scales

www.ShelterCaseStudies.org 3
A.1   Conflict / Complex

Sozma Qula transit camp after winterisation in 2009.


Photo: Shaun Scales

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

4
Conflict / Complex Shelter Projects 2010 A.1

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
A.2   Natural disaster

A.2 Chile - 2010 - Earthquake


Case study:

Country: Project timeline


Chile
Disaster:
17 months - –– Project completion
Earthquake
Disaster date:
February 27th 2010
No. of houses damaged or 15 months - –– Final delivery of
destroyed: cards
More than 200,000 houses 13 months - –– Number of ben-
eficiary households
Project target population: increased
10,000 households
Shelter size:
Variable
Materials Cost per household: 6 months - –– First delivery of
375 USD value per household cards
on the card
5 months - –– Partnership agree-
ment with support-
ing company signed

3 months - –– Start
Chile
2 months - –– Project coordinator
hired

1 month - –– Shelter specialist ar-


rives to support the
February country team
27th 2010 - –– Earthquake

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.

Strengths and weaknesses


99 In general the project was well received by 88 More time should be given for the use of the card
beneficiaries giving them flexibility to spend resources or it should have been distributed earlier than it was.
as they saw fit. 88 The project did not provide technical support on
99 The project team invested time to explain the safer and more earthquake resistant construction. It
project to the beneficiaries. Suppliers were also able did not build on the experiences of recent programmes
to explain the process well to beneficiaries. in neighbouring Peru.
99 Community members were encouraged to group -- The prices of a basket of selected materials at
their purchases together to receive free or reduced various hardware stores should have been monitored
price delivery of their materials from the merchants. over the course of the project. At the start, a baseline
88 The process of choosing beneficiaries was not as price survey could have been conducted to check that
clear as it should have been. Many affected people the project had not lead to price escalaiton. However
felt that many of those who received assistance in a mid-term evaluation, 80% of the targeted families
didn’t suffer major damage to their homes. Others found the prices in the stores acceptable, and there was
noted that the project excluded some families who little evidence of price escalation due to the project.
they thought should have been eligible to receive the
assistance. This led to some jealousy and resentment
from community members who did not receive cards.

6
Natural disaster Shelter Projects 2010 A.2

The earthquake caused a tsunami.


Photo: Sebastián Klarén

After the earthquake Emergency response Implementation


On 27 February 2010 an earth- During the emergency phase of
quake of magnitude 8.8 struck operations, the organisation distrib- It was decided to implement a
Chile. The epicentre was located uted the following non-food items: voucher scheme using a plastic card
60km southeast of the nearest city with a magnetic strip.
• Tents (1,587families)
in the Maule region (400km south The voucher scheme com-
• Tarpaulins (20,650 families)
of Santiago). plemented the delivery of the
• Blankets (44,740 families)
The earthquake generated • Hygiene kits (11,290 families) emergency items, as it allowed
a tsunami, affecting 500km of • Kitchen sets (11,175 families) for the improvement of housing
coastline. The earthquake and suc- • Buckets (22,370 families) solutions through the purchase of
cessive tsunami caused hundreds different household items, as well
However this first phase of the as material for the reconstruction
of deaths and serious damage to response was slow and did not
homes and other infrastructure, of damaged homes. The monetary
meet all of the needs. As a result value of the card (equivalent to 375
primarily in the Maule and Bío Bío other approaches were developed.
regions. USD) was decided in line with the
legal minimum wage at the time.
The survivors of the earth-
quake had to survive the remaining
months of the winter without ap-
propriate shelter.
According to the Chilean
Ministry of Planning, in the worst
affected region (Maule) nearly one
in five people had a damaged or
destroyed house. The earthquake
affected 5 cities with over 100,000
inhabitants, 45 other cities with
over 5,000 inhabitants each, and
more than 900 villages. It affected
both rural and coastal communities.
The project provided cards which could be redeemed for construction materials.
Photo: Mirna Suárez

www.ShelterCaseStudies.org 7
A.2   Natural disaster

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.

8
Natural disaster Shelter Projects 2010 A.3

A.3 Grenada - 2004 - Hurricanes Ivan and Emily


Case study:

Country: Project timeline


Grenada
Disaster:
Hurricane Ivan (Cat. 4) 2 years - –– Over 750 roofs
& Hurricane Emily (Cat. 1) rebuilt and houses
Disaster date: built
–– 2,000 families
September 7th 2004 installed hurricane
and July 13th 2005 straps
No. of houses damaged or –– 128 carpenters
trained
destroyed:
14,000
No. of people affected:
About 61,000 people; 1 year - –– Start of the second
project
50% of the population was left
homeless
Project target population:
- 750 families received a new
roof or a house 9 months - –– Hurricane Emily
- 2,000 families received
hurricane straps 9 months - –– Over 250 roofs
- 128 carpenters trained rebuilt
Occupancy rate on handover: 3 months - –– Construction train-
100% (estimated) ing of the first teams
Shelter size: of carpenters and
11m2 -70m2 Grenada first roofs rebuilt
Materials Cost per shelter: 1 month - –– Project start
Average cost per shelter repaired
2,500 USD
September 7th
2004 –– Hurricane Ivan

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.

Strengths and weaknesses


99 Capacities in hurricane-resistant construction 99 Community members in 32 communities received
techniques were increased through training of men training on safer roof techniques.
and women. 88 The project did not meet the needs of many of the
99 Those trained during the project received a most vulnerable. The weakest houses could not get a roof
certification in carpentry. At the end of the project because they needed too much retrofitting.
most of them were able to find a related job. 88 More houses should have been built to replace the
99 The project was integrated with an island-wide destroyed homes.
project of disaster preparedness. 88 The trainees who received materials did not get the
99 Over 2,000 houses were strengthened with community help anticipated. Carpenter teams had to
hurricane straps as a risk reduction and risk mitigation be deployed to help.
project. 88 The project focused on the needs of homeowners
99 Fact sheets were distributed through newspapers and did not support tenents.
and with materials. They promoted safer construction 88 Larger houses received a higher financial value of
techniques. support as their houses were built from more materials.

www.ShelterCaseStudies.org 9
A.3   Natural disaster

The project focussed on building safer roofs.


Photo: Emeline Decoray

Before the hurricane Implementation than twenty teams of 4 people


Before the disaster, Grenada had Initially, the project focused on working island- wide.
not been hit by a hurricane since re-roofing 100 homes. Six team A disaster preparedness project
1955. People had forgotten about leaders were trained in hurricane-re- was also implemented in 32 com-
the hurricane-resistant techniques sistant techniques which had been munities. On weekends, some public
which had previously been applied used by carpenters 50 years earlier. awareness activities were held to
by carpenters. Trainees were selected and assigned train some community members on
to each team leader. different topics including rebuild-
The houses in Grenada are con-
structed in two types; either from In total 128 men and women ing better roofs. As a mitigation
wood or from concrete. Wooden were trained. They received a project, 2000 vulnerable homes
houses have timber frames and one-day theoretical course followed received hurricane straps which
are clad in timber and have corru- by hands-on training. At the end were installed by trained community
gated iron roofs. Concrete houses of the course, the most capable members.
are commonly made from concrete became assistant carpenters. After
blocks and have a corrugated iron gaining more experience some of Selection of beneficiaries
roof. them became team leaders. The beneficiaries were selected
by the organisation according to
Many wooden houses were The trainers who qualified, criteria defined by the government
resting on concrete or wooden received a certificate in Carpentry and the agencies involved in the
pillars, their structures had no and Masonry from the Technical relief emergency operations. Two
braces, not enough studs and the College (T.A. Marryshow). They types of criteria were used: social
roofs were flat with long eaves. were evaluated after rebuilding 5 to (vulnerable people affected by the
6 roofs with a team composed of a hurricane) and technical (house
After the hurricane team leader, an assistant carpenter damaged or destroyed by the
The hurricane damaged 90% of and 2 trainees. hurricane).
the housing on the island. Concrete
structures partially or entirely After this the trainees could All of the houses were techni-
lost their roofs. Wooden houses receive material to rebuild their own cally assessed before the beneficiar-
were severely affected or totally destroyed roofs. ies were selected. This allowed the
destroyed. The agricultural sector The project ultimately had more organisation to decide on the type
was also severely affected. of assistance the beneficiary would

10
Natural disaster Shelter Projects 2010 A.3

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.

Logistics and supply


All materials were purchased
locally, through local suppliers, even
if it was imported material.
After an assessment of each
damaged house, a bill of quantity of
the material needed to rebuild the
roof was drawn up. This was calcu-
lated by putting the size of the house
into a standardised spreadsheet.
The material was delivered on
site before the start of the work. At The project retrofitted 2000 houses with hurricane straps.
the beginning, all the logistics was Photo: Emeline Decoray

www.ShelterCaseStudies.org 11
A.4   Natural Disaster

A.4 Haiti - 2010 - Earthquake - Overview

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

Dealing with the rubble has been a central issue to recovery.


Photo: Joseph Ashmore

12
Natural Disaster Shelter Projects 2010 A.4

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)

A humanitarian response to urban context: Two-story tran-


sitional shelters, part of a project to apply a “neighbour-
hood approach” in central Port-au-Prince.
Photo: USAID/OFDA.

www.ShelterCaseStudies.org 13
A.4   Natural Disaster

A.4 Haiti - 2010 - Earthquake - Overview continued


Overview:
Summary
The earthquake of 12 January 2010 resulted in
over 222,000 deaths and over 300,000 people injured.
Over 180,000 homes could no longer be occupied, the
majority in densely populated informal settlements,
generating a large scale challenge in terms of debris and
increased pressure on space. Spontaneous and planned
camps were established throughout the affected area,
accommodating at peak 1.5 million people.
The international response was large scale and well
funded. It used a wide range of actors, with varying
degrees of experience of humanitarian response, urban
crises and coordination.
The shelter sector recovery strategies evolved from
meeting emergency needs to addressing a range of
shelter solutions including T-shelter and housing repairs.
The Shelter, Camp Coordination Camp Management,
and Early Recovery Clusters were mobilised to address
these needs.

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
Natural Disaster Shelter Projects 2010 A.4

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

Initial strategies also made


provision for host family support,
but in general projects were not
able to scale up to quickly meet
these needs on any scale. Two years
later over 6,000 households had
received rental subsidies.

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
A.4
SHELTER IN HAITI   Natural Disaster

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.

TRANSITIONAL SHELTERS HEAVY DUTY TARPAULINS

Today, shelter cluster agencies are


increasingly focusing on transitional 633,052 = 1,000 TAUPAULINS 93,287
shelters. These are simple structures DISTRIBUTED ON THE WAY
that provide beer protection than
tents or tarps but take longer to build.

= 1,000 TRANSITIONAL
SHELTERS

OF TOTAL

125,000
PLANNED
45,722
IN STOCK

3,264
COMPLETED

12,175
IN COUNTRY

27,214
IN PIPELINE

HOST FAMILIES TENTS HOUSEHOLD NFI’S COVERAGE AND GAP

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.

THE WAY FORWARD


The provision of transitional shelters is gaining momentum,
particularly in rural areas where more land is available. It is essential
that the identification of additional, safe relocation sites, debris
removal and the required planning processes are urgently addressed
by the authorities to enable the large scale construction of transitional
shelters and ultimately the provision of permanent housing solutions.
EMERGENCY SHELTER TRANSITIONAL SHELTER PERMANENT SHELTER

SOURCE: IASC HAITI SHELTER CLUSTER, 2010 CC BY-ND GRAPHIC BY STANFORD KAY STUDIO.COM
16
Natural disaster Shelter Projects 2010 A.5

A.5 Haiti - 2010 - Earthquake


Case study: See A.4 “Haiti - 2010 - Earthquake - Overview”, p12 for background.

Country: Project timeline


Haiti
Disaster:
18 months - –– 2,550 transitional
Earthquake shelters installed,
Disaster date: and 1,126 tool kits
January 12th 2010 distributed
No. of houses damaged or
destroyed: 12 months - –– 20,000 reinforce-
180,000 ment kits and 500
rural repair kits
Project target population: distributed
10,000 emergency shelter kits
distributed
20,000 reinforcement kits 6 months - –– 5 years habitat
strategy
distributed.
2,550 T-shelters installed
Materials Cost per shelter: 4 months - –– 1 year shelter strat-
egy revised
T-shelter: USD 1,700 per unit
Project cost per shelter: 3 months - –– 1 year shelter strat-
T-shelter: USD 2,800 per unit egy agreed
HA

(materials and project costs)


IT

2 months - –– 10,000 emergency


I

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.

Strengths and weaknesses


99 Multiple approaches were taken to shelter 88 The quality of non-food items and tents procured
provision, allowing projects to match the evolving and imported by the organisation was variable.
context. -- Immediately after the earthquake, there was an
99 The organisation was able to deploy several apparent “equality of vulnerability” as everyone has
experienced shelter team members, who were able to lost their home. However, it quickly became apparent
influence national strategy and programmes beyond that who, prior to the disaster, had the power, identity,
the organisation. connections and resources – in particular housing,
99 The organisation carried out extensive advocacy land and property assets – were able to reassert these
on land rights and access to land. networks and recover more quickly;
88 Procurement and logistics caused significant -- A given neighbourhood was likely to need an array
delays to the transitional shelter projects. Recognising of services and it was not always clear whether it is
that logistics capacity within the organisation was more efficient for a single, non-specialist agency to
weak, attempts were made to establish partnerships deliver all services or for specialist agencies to provide a
for supply with other organisations. These were not single, specialist service across several neighbourhoods
all successful, and three months were lost trying to or indeed the whole city.
establish a working partnership.

www.ShelterCaseStudies.org 17
A.5   Natural disaster

Hillside showing transitional shelters built on small plots of land.


Photo: Mildred Beliard, CARE

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
Natural disaster Shelter Projects 2010 A.5

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
A.6   Natural disaster

A.6 Haiti - 2010 - Earthquake


Case study: See “A.4 Haiti - 2010 - Earthquake - Overview”, p.12 for background.

Country: Project timeline


Haiti
Disaster:
Earthquake
Disaster date:
January 12th 2010 13 months - –– Project completion
No. of houses damaged or
destroyed:
180,000
Project target population:
Repair - 14,000 households
Structural assessment - 400,000
structures
Occupancy rate on handover:
Once a building had received a 11 months - –– 1,500 houses were
green-tag, occupancy jumped repaired
from 50% to 80%
Shelter size:
1-floor earthquake damaged
structure (1 – 3 rooms): average
HA

of 15 - 35 m²
IT
I

3 months - –– Project start


Materials Cost per house:
Repairs: average 2,000 USD per Port au Prince
structure

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.

Strengths and weaknesses


99 The project used a community based approach organisations who were conducting repairs. This
and maintained open channels of communications caused some confusion.
with the relevant government ministries and the -- Initially, owners were suspicious of the engineers.
population at large. As the project became better known, owners began
99 A repair and rehabilitation project was developed. asking the engineers to assess their homes.
This considered the types of housing, differing -- The repaired houses are stronger than they were
neighbourhoods, government guidelines and the when the earthquake struck, but they look virtually
local community. identical to how they looked before the earthquake.
99 Local builders learned cost efficient but safe -- The assessment showed that nearly every
techniques for rebuilding. neighbourhood of Port au Prince contained a mixture
99 Public awareness campaigns assisted displaced of levels of damage.
community members to return to homes which were -- An analysis of the damage showed that residential
structurally safe. buildings, schools, and churches were the hardest hit
88 A shortage of local companies, combined with while commercial buildings fared best.
presidential elections and security issues lead to a -- Although all the houses repaired were more
delay in the start of the public information campaign. resistant to earthquakes than they had been before, it
88 The public information campaign suffered is not possible to guarantee that the repaired houses
from poor messages and overlapped with other would be able to withstand another major earthquake.

20
Natural disaster Shelter Projects 2010 A.6

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
A.6   Natural disaster

The assessment highlighted how The involvement of the


widespread the damage was. Rather community facilitated the setting
than having a core area of red up of meetings with the inhabit-
tagged houses surrounded by rings ants, and municipal authorities. It
of yellow tagged and then green has also facilitated the design of a
tagged houses, nearly every neigh- public awareness and information
bourhood is a mixture of green, campaign.
yellow, and red tagged buildings.
Workshops with local popu-
2) House repairs lations and existing community
Once a house had been assessed, projects helped to identify the key
the next challenge was to repair it. people to meet and to accompany
The cost of rebuilding yellow tagged and support the teams on the
buildings was relatively inexpensive ground.
compared to the cost of new con- To build back safer, three key
struction or comparable transitional changes were made to the way that
shelters. However, it was also clear the masons built walls:
that the reason that most buildings Team of engineers assessing buildings.
• High quality materials: rather Photo: Chiara Jasna Vaccaro
had collapsed was that they were
poorly built. than allowing the masons to Different repair specifications
make their own blocks using were developed for walls with and
Based on the information gained river sand, stronger blocks were without windows, cracked ring
during the damage assessment, made in factories. They were beams, walls that had separated
twelve different types of repairs made with clean materials and from the roof, and for minor
were identified. were vibrated after casting. cracking in walls and columns.
The most common repair was of Masons were required to use
The specific steps to repair each
an X-shaped crack in masonry wall. clean sand for the mortar.
type of damage were detailed in a
The specific steps to repair each • A thinner layer of stronger
separate illustrated guideline.
type of damage were detailed in a mortar: the masons used a 3:1
separate guideline accompanied by sand : cement ratio rather than 4) Training
clear illustrations. the traditional 6:1 ratio. The The following people were
masons were instructed to use trained:
To ensure that builders continued
only a thin layer of this mortar.
to use the better techniques, the • engineers (who had been vetted
This helps to compensate for
organisation, working with an in- by the government) - to conduct
the higher cost of the mortar.
ternational contractor, conducted damage assessments, to use
• Steel reinforcement bars in
inspections of the work on site. PDAs and to how complete the
the wall: the masons were
instructed to add two steel bars required forms,
3) Public communication
between every four courses of • masons - on repair techniques,
& training manuals
blocks and vertically every three • contractors - on repair
Four areas were chosen, for a
blocks. The horizontal steel techniques,
public communications project. In
bars are tied into the vertical • international NGOs and their
each area, a community based or-
columns and the vertical bars technicians.
ganisation was contacted.
are tied into the ring beam. The focus was on how to build
more safely. Since the changes were
The repair process minor, the masons and contractors
1. The damage assessment database was used to identify the number of could be trained in just three days.
houses that can be repaired.
2. Project engineers visit the neighbourhood to verify that the houses are Trained on conducting Damage
not in high risk areas, nor in rights of way. evaluations
3. Community animators meet with local leaders to identify the house 270 engineers for Damage evaluations:
owners. The owners sign a repair agreement. 105 during the 1st Phase
4. Local engineers assess each house. The engineer fills in a form on the 165 during the 2nd Phase
PDA and writes the details of the repair required on the house. Trained on conducting Repair evaluations:
5. A contractor is assigned to repair a group of houses. 32 engineers
6. As each repair is completed, the supervision engineer certifies that the Trained on conducting repairs:
repairs are complete and the contractor is paid. 11 sub-contractors
• Contractors work on groups of three to six houses at a time. Trained to support subcontractors on
• Only masons and contractors who had successfully completed the conducting repairs:
training on the improved construction techniques were allowed to 30 engineers
work on the repairs. 210 masons

22
Natural disaster Shelter Projects 2010 A.7

A.7 Haiti - 2010 - Earthquake


Case study: See “A.4 Haiti - 2010 - Earthquake - Overview”, p.12 for background.

Country: Project timeline


Haiti
Disaster:
Earthquake 13 months- –– Project completion
Disaster date:
January 12th 2010
No. of houses damaged or
destroyed:
180,000
Project target population:
5,690 households or 34,140
individuals
Shelter size:
18 m2
Materials Cost per shelter:
Wood framed shelter 878 USD
Steel framed shelter 1,800 USD
Host family grant 800 USD
HA

Project Cost per shelter:


IT

1 month - –– Project start


I

Wood framed shelter 1,060 USD


Steel framed shelter 2,500 USD Port au Prince

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.

Strengths and weaknesses


99 The projects provided an economic benefit to both 88 Steel framed shelter components were delayed in
shelter recipients and through supporting activities shipment and customs.
such as paid labour for rubble removal. In total, the 88 Effective sanitation for shelters was delayed.
projects injected 750,000 USD into the local economy 88 Relatively low capacity of local builders required
in paid wages. extensive capacity building and oversight.
99 The projects trained and / or employed nearly 400 88 Complications with land tenure and land verification
local masons and builders. Many of whom went on processes slowed shelter provision and created an
to secure formal employment for the first time. unexpected staffing and administrative burden.
99 The projects successfully prevented over 5,000 88 Procurement of some shelter components was
households from going to settlements. delayed, leaving some incomplete shelters.
99 Many households converted parts of their new -- Challenges with coordination often resulted in
homes into shops, salons or cafes, leading to a more duplication and a wide variation in shelter assistance.
rapid recovery. -- Removal of debris was a key factor in the ability to
88 The projects were delayed. This was primarily construct transitional shelters.
due to unavoidable circumstances such as domestic -- Limited local leadership from the local or national
shortages of key construction materials, severe governments, which varied from location to location.
weather conditions, disease outbreaks (cholera), and -- Assembly lines and serial production were largely
post-election tensions. newly introduced concepts and required a lot of
advocacy, training, and oversight.

www.ShelterCaseStudies.org 23
A.7   Natural disaster

Before the earthquake Plot identification Steel shelter


(See “A.4 Haiti - 2010 - Earth- Individual shelter plots were The organisation built 2000
quake - Overview”, p.12.). identified through written state- light gauge steel shelters in areas
ments by community members and outside of Port au Prince, Leogane
Target groups local leadership. and Petit-Goave. These were more
The projects aimed to encour- resistant to hurricanes and heavy
age affected families to stay in their Upon finalising the location of
rain, being designed to resist winds
communities of origin to depres- the plot, shelter construction teams
up to 120-140 miles an hour. These
surise formal or informal camps. To coordinated with cash for work
18m2 shelters were anchored into
achieve this aim, mobilisation teams teams to assure that all rubble and
concrete floor slabs.
worked with settlement leaders to dangerous material was removed
identify households who wished to from the construction site, and The shelter components were
settle near to their properties. from access paths. shipped pre-cut from USA, from an
American design firm in 40 contain-
In most cases, the organisation Engineers worked closely with
ers of 50 shelters per container.
worked with ‘spontaneous settle- shelter construction teams to assure
ments’ that were no more than a that placement of the shelter would Different teams off-loaded the
cluster of households squatting provide the safest possible space for containers, assembled the parts,
on private land or in the streets or the beneficiary household. loaded and off-loaded prefabri-
public spaces next to their property. cated structures and installed the
Wooden shelter shelters on site.
Selection of beneficiaries The transitional wooden shelter
had an area of 18m2 and was Approximately 200 male and
Beneficiary criteria were devel-
intended for a family of five. The female workers were trained to use
oped with community leadership
structure was composed of almost drills in the assembly of metal parts.
structures in neighbourhoods and
50 pieces of timber, ten corrugated Additionally, 8-10 other drivers and
informal settlements, and through
galvanized iron sheets of 12 feet loader crews were used to deliver
local authorities.
(4m) and a concrete floor. the assemblies to the construction
The starting point for the ben- sites.
eficiary selection process was the The structure was strengthened
with hurricane straps. The main Once the assembly mecha-
Mayor’s office in any given location.
bearing wooden columns were nism was fully operational, each
Identification of informal settle-
anchored to the soil using cast-in- facility prefabricated about 45 steel
ments in this way was highly de-
place concrete piers. The walls were shelters each day and installed or
pendent on the support and activity
clad with plastic tarpaulin. The life “completed” approximately 17
level provided by each Mayor.
expectancy of this structure was 24 shelters per day on individual plots.
To triangulate vulnerability as- to 36 months. Only a short training time of 4-5
sessments, project staff also con-
Once materials were delivered days for each assembly team was
sulted with other local organisations
to site, a team of one skilled car- required to start producing roofs,
and community leadership. Shelter
penter and two unskilled labour- sides and front walls.
assistance was prioritised for single
female-led households, the elderly, ers built two shelters a day. On Once the shelters were built, an
and households with more than four average, the project completed 15 additional 6-8 three man crews of
family members. wooden shelters per day. masons installed the cement floors.
Previous homeowners rather The organisation hired ap-
Owner contribution
than renters were targeted as a proximately 120 carpenters in five
The beneficiaries made a floor
result of the added complexity of communes of Port-au-Prince and
fill from broken rubble so that the
determining viable rental agree- installed wooden shelters in various
concrete floor would use minimum
ments and entitlements. areas of the capital.

A steel framed shelter converted into a shop. Family in a timber framed shelter.
Photo: CHF International Photo: CHF International

24
Natural disaster Shelter Projects 2010 A.7

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
A.8   Natural disaster

A.8 Haiti - 2010 - Earthquake


Case study: See “A.4 Haiti - 2010 - Earthquake - Overview”, p.12 for background.

Country: Project timeline


Haiti
Disaster:
Earthquake 26 months - –– Project completion -
EXPECTED
Disaster date:
January 12th 2010
No. of houses damaged or
destroyed:
180,000
Project target population:
3,960 households
Occupancy rate on handover:
One year after the beginning of
the project, the occupancy rate
was 89%
Some households did not 5 months - –– Construction of
occupy shelters still covered with shelters started
tarpaulin for fear of theft
HA

Shelter size:
IT
I

1-5 people 18m²


6-10 people 36m² Port au Prince
11-15 people 54m² 1 month - –– Assessments started
Materials Cost per shelter:
2,400 USD (18m² module) January 12th
Project cost per shelter: 2010 –– Earthquake
4,700 USD (18m² module)

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.

Strengths and weaknesses


99 Support was provided irrespective of land tenure. some construction materials were purchased locally.
99 Modular design allowed for living space to be However the local market could not provide these
varied according to family size. materials easily. This resulted in construction delays.
99 All construction materials, except the steel frames 88 The project was still ongoing two years after the
and a part of the roofs, were purchased locally, disaster, and water and sanitation solutions were not
promoting the local economy. complete.
99 The project included safer construction awareness 88 Few resources are being allocated to follow up and
activities for all families and safer construction monitoring of incidents (occupation, evictions, etc.).
trainings for construction workers. -- Some of the land where the beneficiaries were living
99 As a part of the integrated programme, the access was very close to a river. All the shelters have a raised
to water and sanitation was improved. floor to prevent flood damage. In areas with higher
88 Beneficiary participation in the construction is low flood risk, a deeper foundation would be built as an
as rapid construction was prioritised. additional measure.
88 Power tools were needed to assemble the shelters -- The traditional Haitian house has several exterior
and as a result generators were required. This had doors. Many beneficiaries added doors to their shelter.
logistical and financial implications.
88 Due to lack of understanding of the market,

26
Natural disaster Shelter Projects 2010 A.8

Before the earthquake


See “A.4 Haiti - 2010 - Earth-
quake - Overview”, p.12.

After the earthquake


The town of Leogane’s popula-
tion was estimated at more than
134,000 people. The earthquake
is estimated to have destroyed
32,000 buildings (around 80%
of Leogane’s buildings). After the
disaster there were around 300
camps in the area, with more than
60,000 people living in them.
The construction of shelter was
the beginning of a programme that
provided support to affected house-
Shelters had a steel frame and were modular. They could be personalised to
holds. The support also included meed household needs.
water and sanitation, hygiene pro- Photo: Beti Egea
motion, health, disaster prepared-
ness and livelihoods projects.
during the earthquake, it was ini- Up to ninety shelters were built
Land issues tially unclear whether the shelter per week, but delays with material
The shelters were allocated on would be the property of the ben- supply slowed production.
land where the beneficiaries lived eficiaries who fulfilled the selec-
Beneficiary participation in con-
before the earthquake, promoting tion criteria, or whether the shelter
struction was low. Rapid construc-
the return of displaced people to would be the property of the house
tion was prioritised, leaving little
their places of origin. owners.
time to mobilise, train and incorpo-
Land ownership was difficult It was decided that shelters rate beneficiaries into the work.
to verify. Many beneficiaries did would always be the property of
The shelters were adapted ac-
not have personal identification the beneficiaries. A document was
cording to the number of people
documents, and there were many signed between the beneficiary
in the family. The basic module
difficulties in obtaining legal and and the owners, where the owners
is 18m². Families with up to 5
official land property records. There authorise the beneficiaries to build
members received one module,
were many owners or heirs that did their shelters on their land. This
families over 5 members received
not have documents to prove that document was valid for five years.
two modules and families with
the land belonged to them. Rental If the owner did not respect this
over 10 members received three
agreements with the land owners agreement the beneficiary could
modules.
were made verbally in most cases. move the shelter.
The construction of the progres-
To meet shelter needs of all the If families were landless, the
sive shelter is implemented in two
people living in the communities, community networks were encour-
phases: a first emergency response
solutions for all households who aged to help them to find some
shelter (structure covered with
fulfilled the selection criteria were land. There were also negotiations
tarpaulin) and a second durable
developed, whatever their tenure with local authorities to find a
solution (permanent housing
situation. Intensive community mo- solution for beneficiaries who had
with cement cladding). Different
bilisation was undertaken, and local lived in squatter settlements. Finally
cladding materials were tested for
authorities were involved. authorities let these shelters be con-
the permanent housing.
structed.
In the case of owners or heirs A prototype was erected to
without official identification or Implementation compare the practicality of instal-
land ownership documentation, After the validation and signing lation and the acceptance by the
validation meetings were organised of the documentation, construction target population. The beneficiar-
where the community certifies their materials were distributed. ies chose cement cladding as they
identity and their land ownership. A found it more durable, safer and
The construction team had 4
document was signed by the ben- very similar to the construction
shelter specialists, 4 local coordina-
eficiary, a neighbour, community technique they traditionally used.
tors and 15 local engineers. Each
representatives and local authori-
engineer led a team of workers The project included safer con-
ties.
from the communities, and each struction awareness activities for all
In the case of tenants who lived team built 6 shelters per week. the families and safer construction
in houses that were destroyed trainings for construction workers.

www.ShelterCaseStudies.org 27
A.8   Natural disaster

Two-module shelter with a door added by the family


Shelter made from two modules and later upgraded (standard two-module shelter has two doors, one on the
by family. front and one at the back).
Photo: Betisa Egea Photo: Sandra Tapia

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
Natural disaster Shelter Projects 2010 A.9

A.9 Haiti - 2010 - Earthquake


Case study: See “A.4 Haiti - 2010 - Earthquake - Overview” p.12 for background.

Country: Project timeline


Haiti
Disaster:
Earthquake 18 months - –– All families have a
Disaster Date: transitional shelter
January 12th 2010 –– Ongoing provision
of services required
No. of houses damaged or
destroyed:
180,000
No. of people displaced:
Approximately 1.5 million
Project target population:
1,356 families
Occupancy rate on handover:
105% 3.5 months - –– Full occupancy with
tents
Site density:
30m2 / person
Materials Cost per shelter:
HA H

Tent 300 USD (excluding


IT AIT
I I

3 months - –– Relocation starts


transport)
Transitional Shelter 1,600 USD Port au Prince
Corail
Project cost per shelter: Port au Prince
Unknown
6 weeks - –– Decision taken to
open site
January 12th
2010 –– Earthquake

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.

Strengths and weaknesses


99 Key actors worked together to prepare the site from the sun and wind.
within an extremely limited timeframe. 88 The site does not represent a durable solution
99 Strong coordination greatly assisted with the for the relocating families and remains one of 802
logistics of the relocation through information occupied camps for displaced families in Haiti.
campaigns and consultation with the affected 88 Rapid site preparation required significant
population. investment at a time when financial resources for the
88 The urgency of the relocation initially left little provision of basic services were limited.
opportunity for activities beyond the provision of -- The impact of having a camp in any location has
shelter, water, sanitation, food, education and health to be carefully considered since it might end up as a
services. permanent settlement.
88 Greater emphasis on ensuring access to existing -- The decision to relocate the people was based on an
or developing livelihood activities would have been engineering assessment of the risk of flash floods (high
beneficial had time allowed and the site was far from volume, fast moving water) at several spontaneous
existing livelihoods. IDP locations. The identified population faced life
88 There was a significant delay in the follow up threatening risk in their current location. In addition,
construction of transitional shelters, meaning people there was an urgent need to decongest the camp to
had to stay in tents in an area with little natural shade allow the introduction of basic services.

www.ShelterCaseStudies.org 29
A.9   Natural disaster

Left: an aerial photo of a typical spontaneous settlement in


Port au Prince.
Right: An aerial photo of Corail shortly after construction.
Photo: Shaun Scales / NRC

Background Planning the site Why tents?


See “A.4 Haiti - 2010 - Earthquake The new site was based on a firm Allowing relocating families to
- Overview” p.12. plan. Site assessments identified four bring their existing shelter materials
separate ‘sectors’ for development with them was not seen as a sensible
Identification of families
with ‘Sector 4’ selected as the first to approach as they were generally of
Given the large population in
be prepared and occupied by the re- too poor a quality to re-use and it
camps within Port au Prince, weeks
locating population from Delmas 48. was too logistically challenging.
after the disaster, assessment teams
identified specific areas at risk from The outline of the site was deter- It was recognised that the
flash flooding. They also assessed mined by existing natural drainage. commonly adopted emergency
which engineering works could This was upgraded to protect plots shelter strategy focused on the
mitigate identified threats to life. from surface water from above the provision of plastic sheeting, but
site and to allow the development of given the circumstances tents were
The assessment was conducted
an internal drainage network. provided as they were the best
in spontaneous settlements within
emergency shelter solution.
Port-au-Prince. Amongst others, The camp was planned for
it identified the Delmas 48 site as occupancy as a transitional site Relocation
being at risk from flash floods and with defined individual family plots, The Camp management agency
landslides during the approaching internal road networks and space with support from the Camp Coor-
seasonal rains. The site had over for education, health, recreation and dination and Camp Management
25,000 people living in high densities distribution facilities. The plan was lead organisation initiated a settle-
on a steep hillside. strictly followed so that future devel- ment wide information campaign to
opment with longer term infrastruc- identify families willing to relocate to
The engineering team developed
ture could be possible. Although the a new planned camp.
a mitigation plan that included the
site was officially temporary, the site
diversion of surface water and land The relocation of 1,356 families
planners took account of the possi-
stabilisation works. To complete was completed in stages over a ten
bility that it might not close soon.
these works, an estimated 7,500 day period with transport provided
people would be required to move Pending the development of by the United Nations mission. A plot
from their current high risk plots. durable solutions for the significant identification system allowed each
displaced population within Haiti, the arriving family to be allocated an in-
The area of the settlement that
maintenance of essential services to dividual plot which was recorded as
needed to be vacated was marked.
all camps, including Corail, remains a part of the registration process and
The high density population left little
prolonged and significant challenge. assisted with the future delivery of
room for internal relocation and re-
organisation. services.
Site construction
Land clearance and the develop- Transitional shelters and
Selecting the site
ment of a gravel road network were other structures
State land is limited in Haiti and
completed within two weeks. Con- The delivery of transitional shelter
the power of the government to
struction progress was accelerated was significantly delayed. However
claim land for public emergency use
by foreign military forces, some who by mid 2011, each family plot had an
is even more limited. Identifying al-
were due to depart imminently. 18m2 transitional shelter on it.
ternative land close to neighbour-
hoods of origin was problematic as Land clearance allowed plots to Each shelter included a raised
most potential sites were already be marked for shelter and infrastruc- cement finished plinth and a small
occupied. The only immediately ture. Tents were then erected and veranda area covered by an extended
available land of sufficient size was temporary water and sanitation facil- truss roof.
16km away. This did restrict oppor- ities provided. Fire breaks were built
tunities for relocating families whilst and a population density of 30m2 per Education and health facili-
maintaining access to livelihoods. person was maintained. ties were formalised with semi

30
Natural disaster Shelter Projects 2010 A.9

permanent or permanent structures


of wood and brick construction. The
original temporary latrines were also
replaced with blocks built of bricks.
Eighteen months after the occu-
pation of the site, kitchen gardens
and a market selling foodstuffs,
household items and handicrafts had
been established. Small businesses,
including restaurants, carpentry
workshops and an art gallery were A typical street in Corail with transitional shelters.
also established, although the Photo: Shaun Scales / NRC
primary source of income comes
from work off site. However services were falling back solutions for the displaced popu-
as funds fell and organisations began lation. However a lack of recon-
The school was adopted as a
to close projects. It was recognised struction continues to hinder this
government institution with ministry
that camp based services could con- process, and Corail was not likely to
of education providing salaries for
tribute to the sustained presence in close soon.
teachers.
camps however an acute shortage of
Corail was less densely
Following the occupation of return solutions for the majority of
populated than many spontaneous
Sector 4, further development of the displaced population of former
sites in Haiti. Transitional shelters
adjacent sites continued to allow tenants, remained the primary factor
were built, and this caused some
for further relocations including 178 hindering camp closure. This may
confusion regarding the ‘status’ of
families affected by Hurricane Tomas have contributed to the sustained
the site. The future closure of Corail
in November 2010. presence of camps.
would require the same efforts as
The longer term Two years after the earth- other emergency and transitional
Almost two years after the earth- quake, the future for camp based settlements. It also became sur-
quake, people in camps in Port au populations across Haiti remained rounded by thousands of Haitians
Prince continued to receive limited unclear. The exit strategy for Corail who had built their own shelters
free services in water, education, was always the closure of the and houses.
health, and other assistance. camp following delivery of durable

Tents provided initial shelter at the site. This was later upgraded to transtional shelters.
Photo: Shaun Scales / NRC

A Market area in “Corail Sector 4”.


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
A.10   Natural disaster

A.10 Haiti - 2010 - Earthquake


Case study: See “Haiti - 2010 - Earthquake - Overview”, p.12 for background.

Country: Project timeline


Haiti
Disaster: 22 months - –– Construction
Earthquake complete
Disaster Date:
January 12th 2010
No. of houses damaged or
destroyed:
180,000
Project target population:
Families with disabled persons
Shelter size: 8 months - –– Project scaled up
12m2, 18m2 or 24m2 with a 6m2
porch dependent upon family
size and land.
6 months - –– Pilot 50 shelters
start

5 months - –– Supply chain and


workshop estab-
lished
HA
IT
I

2 months - –– Participatory work


Port au Prince and pilot shelter
5 weeks - –– Project start

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.

Strengths and weaknesses


99 The construction technique of clissade is well and resource consuming to receive beneficiaries referred
known by the local population as it has been in this way. This increased the logistical challenges as
traditionally used in rural Haiti. As a result it is easy beneficiaries were identified as the project progressed
and affordable to maintain and upgrade. and were not identified from the start.
99 The shelter was designed in panels. Each panel has 88 If the beneficiaries do not upgrade their shelter by
the same width as a door, allowing beneficiaries to covering their panels, water could enter and it could
create new openings in their shelter. be cold.
99 The project paid particular attention to 88 Logistics were demanding and slow as rural
beneficiaries with disabilities. Each individual shelter locations meant that some families could not always
and its sanitation facility was adapted to the type be reached by vehicles.
of disability. It was accompanied by a rehabilitation 88 The project and the design was very labour intensive.
program for people with disabilities, to increase their -- The shelter was prefabricated in pieces in the
mobility and build capacities in the use and access to central workshop and sent to the field for assembly by
the latrine and the shelter. beneficiaries themselves. The concept was that shelters
99 The project worked with students from a youth could later be moved if required.
vocational training centre. It aimed to increase their
capacity to join the labour market.
88 Beneficiary selection depended on a referral
system from other organisations. It proved very time

32
Natural disaster Shelter Projects 2010 A.10

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.10   Natural disaster

A traditional shelter that survived the earthquake. A completed shelter, based on vernacular styles.
Photo: Joseph Ashmore Photo: David Sacca

Technical solutions Materials list


The T-shelter was made from The shelter was designed and Materials Quantity
pressure treated pine wood. Panels tested by structural engineers to be Timber 2"x2"x14' 4 pieces
were prefabricated in the workshop resistant to hurricane, earthquake (50x50mmx4.3m)
and were then transported to and floods. It was also designed to Pine 2"x4"x14 89 pieces
the field. Once on site, the pieces ventilate naturally. (50x100mmx4.3m)
were bolted together. All the nails Pine 1"x4"x14 23 pieces
and screws (the panels were fixed Logistics and materials (25x100mmx4.3m)
Once the shelters had been Pine 1"x6"x14 3 pieces
with nuts and bolts, not nails) were (25x150mmx4.3m)
double hot dip galvanized. prefabricated in the workshop, it
proved challenging to get the com- Plywood 1/2" (13mm) 3 pieces
For roofing, corrugated bitumi- ponents to remote locations in the Plastic mosquito net 48" (1.2m) 20' (6m)
nous sheets were selected. They mountains of southern Haiti. Wood Glue 0.5l
were selected due to their 15 year Corrugated fastener 1"x5" unit
guarantee, their thermal properties Many of the raw materials had Corrugated roof sheets (Onduline) 19 pieces
and their strength. to be imported to Haiti. For example Ridge (Onduline) 9 pieces
the timber used was pressure Twisted roofing nails for wood
The site for each shelter was treated pine that was not available 2 1/2"x9" (60x230mm)
prepared by a team who were in Haiti. Most were shipped in and Threaded rod 3/8" 80" (10mm) 23' (7m)
tasked with taking into considera- then trucked into the workshop in Nails: 1 ½"-5" (30mm-125mm)
tion possible risks, such as land- Petit Goave. In the workshop, the Coiled strap (Hurricane strap) 15 m
slides, of each plot. The field teams whole shelter was pre-fabricated in Hinge 4"x4" (100mmx100mm) 1
were expected to conduct work to panels and trusses. The pre-assem- Hinge 3"x3" (75mmx75mm) 2
mitigate the risks. bled components were then trans- Bolt 4", 3"(100mm, 75mm) 2
Each shelter is raised by between ported to the site, by truck or by Wood screw 3½"x10
30 and 50 cm from the level of the hand in difficult to access areas.
Cement 18 bags
ground preventing water entry in Sand 6 m3
case of floods. Gravel 5/25 4 m3
Cement blocks 70 pieces

Some areas were difficult to access and materials needed to be transported by hand.
Photo: Olivier Dorighel

34
Natural disaster Shelter Projects 2010 A.11

A.11 Haiti - 2010 - Earthquake


Case study: See “A.4 Haiti - 2010 - Earthquake - Overview”, p. 12 for background.

Country: Project timeline


Haiti 36 months- –– (Anticipated) - On-
Disaster: going monitoring
Earthquake
24 months - –– 8450 households
Disaster Date: supported
January 12th 2010
No. of houses severely 20 months - –– Decongestion of
camps: Delmas 75
damaged or destroyed:
185,000 18 months - –– Decongestion of
Project target population: camps: Croix de
8,450 households after 24 Bouquests
months 13 months - –– Decongestion of
T-Shelter size: camps: Simon Pele
Aim for 18m2 minimum
Less considered when
11 months - –– Decongestion of
insufficient space camps: Annex de la
Materials Cost per household: Mairie
T shelter: 2,800 USD
500 USD livelihoods grant 10 months - –– Decongestion of
camps: Sint Luis de
Project cost per household:
HA

Ganzague
IT

T-shelter projects: 4,500 USD


I

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.

Strengths and weaknesses


99 The project took a broad approach to shelter, 88 The process was very labour intensive and required
looking at the overall settlement issues. constant monitoring and support.
99 Households were involved in identifying a shelter 88 The process for cash transfers was cumbersome
solution with which they felt comfortable. and needed to be shortened.
99 Families were able to quickly pick up some threads 88 Technical support for some construction aspects
of normality with the cash support to develop income has been limited. In particular, viewing the land
generating activities. and identifying the work that was required before
99 Physical security for people was improved once construction could begin.
they were out of the camps. 88 Camp committees were difficult to manage as they
99 Cash gave people a greater degree of choice believed that they should be receiving a salary.
and permitted them to spend money according to -- Some people did not want to leave the camps as
their own priorities. This in turn helped to maintain they believed that they would continue to receive
people’s dignity. goods if they remained there.
99 Cash had potential benefits for local markets and -- Some households split across multiple sites to
trade. receive a greater total amount of assistance.

www.ShelterCaseStudies.org 35
A.11   Natural disaster

Background
See “A.4 Haiti - 2010 - Earth-
quake - Overview”, p. 12.

After the earthquake


Up to eighty percent of the
population in Port-au-Prince rented
either the house or the land. In other
urban centres such as Leogane, up
to seventy percent of the popula-
tion rented.
Reconstructing houses would
restore the assets of the landlords,
but would not ensure the availabil-
The project also had a significant information component and the organisation
ity of this accommodation to the actively promoted public health messaging.
former tenants who are currently Photo: Julien Goldstein
shelter-affected.
Different options offered they could rent. If the accommo-
Residential reconstruction activi- Different options were provided dation was deemed to be secure,
ties therefore included measures to depending upon the context that had water and sanitation facilities
ensure that former tenants received the family found itself in: and was seen as a safe dwelling,
benefits in kind through agreed the family received up to 500 US
rent-free tenancies for a defined 1) Own land dollars to resettle. This sum covered
timeframe, separate cash grants Some people had the option to a year´s rent.
linked to rental accommodation, or move back to where their house
shared usage rights. was or to a piece of land to which Often, people moved towards
they could show ownership. They the areas they lived in previously as
Settlement approach received a T-shelter on their land they were familiar with the area.
The organisation implement- and received a 150 USD grant. 72% of families in the project
ed projects using a ‘settlement chose this option.
approach’. Communities and infra- 8% of families received this
structure were supported, integrat- form of assistance. 5) Resettlement in the
ing other sectors such as water and 2) Access to land provinces
education. Many of the projects Some people knew someone 19% families chose to return
had strong economic and social who had a plot of land who agreed to their provinces of origin. These
‘livelihoods’ components. that they would be able to reside families received a resettlement
on the plot for two years. They grant.
Shelter was seen as including
support to all of the settlement had to produce a signed document Additional support
options chosen by affected popula- stating that they can live on the All Families additionally received:
tions, including host families, rental land for two years, and a copy of
accommodation and, where neces- the ownership documents and their • A livelihoods grant of 500 USD
sary, camps. In choosing between identification divided into two distributions
options, families and groups can of 250 USD. The first was one
They received a T shelter built on month after having left the
make best use of their coping strat- the land and a 150 USD grant.
egies. camp and the second was after
3) Repairable houses three months.
Five months after the earth- People who had houses clas- • A training was provided on
quake, the shelter team began sified as green (having minor managing finances and business
registering people in four camps in damage) were offered cash or a opportunities of their choice.
an area of Port au Prince. A variety voucher to access the needed ma- • Families were supported with
of solutions to support households terials, an unconditional business health insurance for one year.
were identified. grant, and training on earthquake The health insurance was
The interventions were based on resistant construction. provided by a local organisation.
assessments and discussions with The insurance was 1 USD
In the first two years of the monthly per person, and entitled
families. Three areas of support project, no families chose this
were identified: them to free consultation at
support option. clinics run by the organisation.
• an improved shelter solution, It also limited their payments for
4) Resettlement in Port au
• support for livelihoods, medicines to a maximum of 150
Prince
• an option to help their children USD. They could also have low
Families identified accommo-
return to school. cost medical investigations.
dation within Port-au-Prince that

36
Natural disaster Shelter Projects 2010 A.11

The project included support for livelihoods , support getting children back to school and access to improved health care.
Photo: Julien Goldstein

The small minority of families


who did not take up any of the
support offered signed a document
to show that they had refused the
offered support and would remain
in the camps. Once families moved
out of the camps, sometimes other
families might settle in space made.
It was the responsibility of the
Haitian authorities to deal with
these cases.

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
A.12   Natural disaster

A.12 Indonesia - Sumatra - 2009 - Overview


Case study:
Summary
On 30th September 2009 a series of earthquakes struck West
Sumatra, not far from the provincial capital of Padang. 13 out of
the 19 districts in West Sumatra province were affected. Between
earthquakes and landslides nearly 250,000 houses were destroyed
or heavily damaged.
The Government of Indonesia responded rapidly, with the
assistance of the national and international humanitarian community.
Whilst non-government agencies focused on emergency shelter,
distributing an average of 2 tarpaulins per family, the government
focused on rebuilding provincial government capacity, search and
rescue and emergency relief. The emergency phase was declared
over within 8 weeks.
The Government of Indonesia committed to providing affected
families with a community based economic stimulus package
for permanent housing reconstruction, leaving the provision of
emergency and transitional shelter to the humanitarian community,
many of whom also focused on Disaster Risk Reduction based
construction skills training.

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
Natural disaster Shelter Projects 2010 A.12

200 USD “transitional shelter” of dissapointing quality built by


an international organisation. Rural self help shelter built by earthquake affected family.
Photo: Dave Hodgkin Photo: Dave Hodgkin

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
A.13   Natural disaster

A.13 Indonesia - Sumatra - 2009 - Earthquake


Case study: See “A.12 - Indonesia - Sumatra - 2009 - Overview” p.38, for background

Country: Project timeline


Indonesia, Sumatra, Padang
Disaster:
Earthquake
Disaster date:
September 30th 2009 5 weeks - –– Final report
No. of houses damaged:
115,000 destroyed houses
135,000 damaged houses

This was a market assessment


into brick production and so
did not directly lead to the
construction of shelters
3 weeks - –– Surveys were con-
ducted over 3 days

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.

Strengths and weaknesses


99 The assessment was conducted with team 88 Surveys looked at the use of bricks but not the use
members from nine different organisations. This of timber to make the bricks.
process increased buy-in to the findings of the 88 The survey did not address issues of the living and
assessment report, and helped to form consensus on working conditions for those in the brick kilns.
the issues surrounding markets in the response. 88 The survey used human resources, meeting time
99 The bricks survey findings were used to advocate and vehicles that could otherwise have been used in
for a cash based response, and for a move away from implementing the response.
solid masonry buildings which potentially carried a -- It is difficult to accurately measure the impacts
greater risk of causing injury in an earthquake. of this survey. Whilst it used human resources and
99 The survey came at an opportune moment after absorbed time during an emergency response, there is
the earthquake. The timing of the survey needed some evidence that it helped to inform the strategies
to be long enough after the earthquake that team and programmes adopted.
members could be identified, access was possible and -- There are many markets that could have been
those working at brick kilns could easily be found. surveyed. Bricks were chosen following experiences in
Had it been any later it would not have been able to Aceh (2004) and Yogyakarta (2005).
inform the strategy.

40
Natural disaster Shelter Projects 2010 A.13

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

Much of the capital for small scale manufacturers


Many of the bricks were made by hand. was the bricks in their kilns.
Photo: Unknown Photo: Unknown

www.ShelterCaseStudies.org 41
A.13   Natural disaster

were 1,800 small scale brick manu-


facturers, who produce an average
of 15,000 bricks per month. These
were the most severely affected
of all brick manufacturers. The
financial capital of these producers
was often tied up in the number of
bricks they had in their kiln, making
it difficult to restart manufacture.
Medium scale manufactur-
ers (45,000 bricks per month) also
suffered production losses due to The survey used teams from nine different organisations working together.
the earthquake, but their stronger Photo: Unknown
financial position meant that they Brick demand brick production capacity
were better able to resume produc- 60% of all households inter- would likely lead to higher
tion. It was estimated that it would viewed indicated that they would brick prices and delays in rural
take 6–8 weeks for these manufac- re-use as many bricks as possible. A housing reconstruction. Large
turers to bring new bricks to the rough estimate suggested that many brick manufacturers were likely
market. households would be able to salvage to reach previous production
Most of the larger scale brick 800-1200 bricks from the rubble. As capacity within two months.
manufacturers were located up to an average size brick masonry house Resulting transportation cost
90km North East of Padang. Some of 10m X 12m used approximately increases could lead to a price
large brick manufacturers reported 10,000 bricks, approximately 10% increase of between 100% and
losing 35% of their brick produc- of this demand would come from 150% per brick.
tion in the earthquake, while others recycled materials. • Small - scale brick manufacturers
did not report significant losses. would be slow to resume pre-
Although 67% of all households
earthquake production levels
Brick prices and financing interviewed said they lived in a brick
without financial assistance or
Pre-earthquake brick prices masonry house before the earth-
favourable credit terms. Their
ranged considerably according to quake, 54% of the brick masonry
ability to resume production
quality, seasonality and transport households indicated they would
was restricted due to capital
costs. prefer to rebuild timber and brick
shortages, or favourable credit
houses. Safety concerns were most
Following the earthquake brick arrangements.
often cited as the reason for this
prices from suppliers for mid range • The demand for timber and
preference, followed by cost consid-
quality bricks increased by between bricks was high, and was likely
erations.
25% and 50%. The assessment to increase. Over 60% of
found that these prices were likely There was some concern raised earthquake affected households
to continue to rise to 150% of their that recycled bricks would not interviewed in this survey
pre-earthquake cost. perform so well as new bricks indicated that they planned
because as cement mortar cannot to rebuild (or would prefer)
Two years after the survey, brick bind to them so well. timber frame houses with brick
prices in Pedang were between masonry infill walls over full
60% and 100% higher. Gender issues masonry construction. Concerns
Women made up 40 - 60% of the over seismic safety, speed of
Both small and medium scale
labour force of small and medium construction, and lower costs
brick manufacturers used informal
scale brick manufacturers. They were the main reasons for this
credit and selling arrangements
were typically paid on a piecework change in preference.
with their customers and distribu-
basis for each brick they made. Male
tors. Local supply stores typically Impacts of the survey
brick labourers are likely to receive a
paid small-scale manufacturers for Because the survey was
daily wage for their work.
bricks once they had sold them. conducted by teams from many or-
As current brick production ganisations, it helped to get support
All brick manufacturers, but es-
for many small-scale producers is for the findings. Although not all of
pecially small and medium scale
affected, the ability of brick making the recommendations were imple-
producers, had limited storage and
women to earn wages was tempo- mented, it did help organisations
warehousing space. These space
rarily disrupted. and coordination teams to form
limitations forced manufacturers to
move their bricks to market quickly. Possible scenarios an advocacy position away from
It encouraged large suppliers and The analysis suggested that: building full masonry structures,
distributors to increase their prices instead promoting semi-timbered
to meet speculative market demand. • Earthquake damage to regional structures with support provided in
cash.
42
Market-system Map: The Brick Marker – Padang –W est Sumatra Symbol Key
BUILDING PRICE INFORMAL
The market environment: ENVIRONMENTAL PREFERENCES TRENDS CONTRACTS Critical issue !
FACTORS
Natural disaster

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

RAW MATERIALS BUFFALO TRANSPORT AND LABOUR CASH Colour key


! DELIVERY INCOME
Target groups
Key infrastructure, inputs CREDIT AND WAREHOUSING ROADS &
and market-support services LOANS Other type 1
! BRIDGES
Other type 2

41
This map is for the brick market in Pedang following the earthquake. The black arrows show how bricks reached homeowners
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
A.13
A.14   Natural disaster

A.14 Indonesia - Sumatra - 2009 - Earthquake


Case study: See A.12, “Indonesia - Sumatra - 2009 - overview”, p.45 for background.
Country:
Indonesia, Sumatra, Padang Project timeline
Disaster:
Earthquake 7 months- –– Project completion
and evaluation
Disaster date:
September 30th 2009
No. of houses damaged: 6 months- –– Cash distributions
115,000 destroyed houses
135,000 damaged houses 5.5 months- –– Market surveys
No. of people affected:
Approximately 1,250,000 people 5 months- –– Registration
affected through total or partial
loss of shelter and livelihoods 4 months- –– Project assessment
Project target population: –– Project start
Shelters for 750 families
Household items to 30,000
3 months - –– Shelter kits, tool
families kits, household and
Occupancy rate on handover: hygiene items distri-
Unknown bution complete
Shelter size: Padang
Variable
Materials cost per household:
Indonesia 4 - days –– Non-food items
275 USD
were distributed
from pre-positioned
stock
September 30th
2009- –– Earthquake

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.

Strengths and weaknesses


99 Cash grants helped people buy what they needed and not have encouraged families to build safely.
for construction. People had flexibility to build what 88 There was some resentment from those who
they wanted. did not receive cash grants. There were sometimes
99 The injection of cash into the markets boosted the very slight difference between recipients and non-
local economy and has assisted the self-recovery of recipients circumstances, which made it hard for some
other community members, who are also starting to to understand why they had not received support.
rebuild their homes. 88 Transitional shelter support should have arrived
99 Despite the amount of money being insufficient earlier. After three months of living in inadequate
to complete all work required, it gave people a shelter, many households were ready to build semi-
strong starting point to begin recovery. Many people permanent structures.
became motivated to begin construction. 88 The half day of training provided to beneficiaries
99 Existing relationships between project staff and was insufficient. House improvements were not
communities helped trainings and cash distributions covered in trainings.
run smoothly, even though there was some unrest -- There are strict rules that limit logging locally. Many
from those who had not received support. beneficiaries only used trees from their own land.
88 The sum of money was too small for all -- The local cost of materials did not increase. However,
construction. there was a reported increase in the cost of skilled
88 Project timeframes may have rushed construction labour, which was in low supply and high demand.

44
Natural disaster Shelter Projects 2010 A.14

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
A.14   Natural disaster

A family who used the cash grant to purchase


House for 9 people under construction in the foreground. timber beams and concrete.
Photo: Save the Children Photo: Save the Children

semi-permanent structures or reno- a note in a confidential complaints • A minimum of 3.5m2 covered


vations to damaged homes which box. During February the monitoring living area per person.
the majority of beneficiaries had and evaluation team also enabled a • A minimum of 2m from the
opted for. “complaints hotline” for all sectors, ground to the eaves.
so people could call or send their • The roof should provide
2) Training beneficiaries
comments using text messages. adequate strength and have a
Project staff held 11 two hour pitch of at least 250.
workshops in the villages, to dissem- Monitoring • There should be adequate
inate technical information about During the monitoring phase, ventilation.
construction standards and methods the team used guidance and an • The shelter should provide
among selected beneficiaries. At the agreed format to check the com- protection from rain.
end of their training, beneficiaries pliance with the following cluster- • There should be at least one
received vouchers to be exchanged agreed minimum standards: internal division for privacy.
by cash. • Materials and construction • Building should use safe
should allow for 24 months of construction techniques to
Complaints response minimize the impact of further
use.
mechanism (CRM) natural hazards.
1) At the targeting level
The committees posted the final
list of names on community notice
boards. At the same time, boxes
were installed to collect complaints
from those who had not been
selected, so they had an opportu-
nity to make their case. Three days
later, boxes were collected. After
analysing the messages and com-
plaints, meetings were to be held
with committees. If those who had
complained qualified, they would be
added to the final beneficiary list.
2) At the implementation level
The community would be able
to file complaints and give feedback
throughout the entire duration of
the project, not only during the
selection phase. The communities
would have the opportunity to meet
A “renovation”: the roof and foundations were solid - the owner used materials
directly with staff during their visits, bought with the grant to repair the shelter.
approach shelter committees or drop Photo: Save the Children

46
Natural disaster Shelter Projects 2010 A.15

A.15 Indonesia - Sumatra - 2009 - Earthquake


Case study: See “A.12 - Indonesia - Sumatra - 2009 - Overview” p.38 for background.

Country: Project timeline


Indonesia, Sumatra, Padang
Disaster: 12 months- –– Project completion
Earthquake with no-cost
Disaster date: extension
September 30th 2009
No. of houses damaged:
115,000 destroyed houses 9½ months- –– Planned project
135,000 damaged houses completion
8½ months- –– Second cash dis-
(approx. 70,000 in Padang bursement
city)
Project target population:
3,400 households (3% of
overall houses destroyed) 6½ months- –– First cash
Occupancy rate on handover: disbursement
66% of all shelters occupied 6 months- –– MoU signing with
12 months after the beneficiaries begins
earthquake.
Shelter size: Padang 5 months- –– Training for field
Variable teams
Materials Cost per household:
Cash grants for T-shelter: Indonesia
4 months- –– First beneficiary
330 USD per unit training
Government estimates for 3½ months- –– Funding proposal
reconstruction of a destroyed submitted
houses: 1,600 USD

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.

Strengths and weaknesses


99 Each family was able to build according to their 88 People without land or with damaged housing did
needs and wishes. This improved ownership. not get cash or any technical assistance and often
99 Families built shelters that they felt were rebuilt dangerous brick structures.
permanent. Families invested and built quickly. -- Standard designs would have made quality control
99 A transparent complaints mechanism helped with much easier. However this would have curtailed the
the perception that beneficiary selection was fair. freedom of the beneficiaries to build according to their
99 The project worked in remote rural remote areas needs.
because people had space, owned that space and -- Donors had some concerns that permanent housing
owned non-productive coconut trees. had been built with emergency funding.
88 A disaster risk reduction opportunity was missed -- The houses built might have been “safer”, but It
for people with damaged housing. is a mistake to refer to them as earthquake or hazard
88 The 120 field monitors and community volunteers resistant.
had only a few days technical training. It was not
realistic to expect them to check the construction
quality of 3,400 unique houses.

www.ShelterCaseStudies.org 47
A.15   Natural disaster

Background To be included in the project, The draft lists were posted in


See “A.12 - Indonesia - Sumatra - beneficiaries had to have land for the communities along with posters
2009 - Overview” p.38. a shelter and a destroyed house. explaining the selection criteria,
Selected families were in a good detailed definitions of the project,
Before the earthquake
position to complete their shelters an outline of a step-by-step im-
In West Sumatra, most families
as: plementation plan, and a hotline
owned their houses before the
telephone number to call or SMS
earthquake. The region has a mat- • They were in less urbanised
feedback, complaints or requests
rilineal system with women owning areas and had previously lived
for information.
and inheriting land and housing. in single storey buildings.
On marriage, the new husband will • They had access to timber and Senior project managers
move on to the land of his wife’s experience of using it. operated the phone and were avail-
family. Housing has symbolic and • They saw the transitional shelter able for office visits and had after
social importance. as a permanent home, worth hour telephone numbers posted on
finishing and worth investing in. the office door. Each and every case
Family houses are built bit-by-
More than 9000 households was followed up on an individual
bit. In rural areas people usually
were surveyed and given a vulner- basis with village government and
paid local builders to build or some-
ability and eligibility score. Selection community committees.
times built houses for their own
families. criteria included female and senior
Implementation
headed households, low-income
Houses are not purely a finan- Assessments and existing expe-
families, pregnant women and
cial investment. Remittances are riences showed that communities
children under 5.
a major source of housing finance had the capacity, access to materi-
and cash incomes are irregular and Feedback and complaints als, labour and community cohesion
seasonal. The community feedback and to manage cash to build transitional
complaints mechanisms were es- shelters. A cash approach was also
Organisational capacity sential to the running of the project. promoted by the Shelter Cluster.
Before the 2009 earthquake, the This system built on lessons learned Beneficiaries built according to their
organisation had significant practi- from the 2005 tsunami response needs, wishes and resources. This
cal emergency experience. Both the and Jogyakarta / Central Java earth- encouraged fast construction and a
international organisation and its quake response programmes. sense of ownership leading to high
partner understood the need for ex-
perienced staff and sufficient time
for community engagement.
“Lots of people got jobs
The organisation also had ex- as masons [because
perienced senior managers and of the project]. New
partners who knew the community masons were called
and spoke the local languages. The ‘toukonggumpa’
local partner organisation addition- [‘earthquake masons’].”
ally had good and long term rela-
tionships with the affected com-
munities. This reduced the need for Rural community leader in Pariaman
lengthy formal assessments.

After the disaster


The earthquake of September
2009 destroyed 115,000 houses,
and damaged 135,000 houses. In
Padang the government responded
with assessments and the promise
of compensation. Many households
affected by the 2007 earthquake
were only just receiving compensa-
tion at the time of the 2009 earth-
quake so families did not expect
compensation to arrive quickly.

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
Natural disaster Shelter Projects 2010 A.15

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
A.15   Natural disaster

Guidance used for a feedback and complaints


handling mechanism
• Ensure that simple complaints and feedback mechanisms are written
into project strategy and budget.
• Ensure that ‘complaints handling’ is written into job descriptions of staff
at all levels of the organization, and that staff are adequately prepared
and trained in handling complaints.
• Consult communities and select context appropriate means of
communication and technology to receive feedback and complaints
and provide a response (e.g. phone or email systems, visiting hours,
feedback boxes).
• Define the process for complaints handling including timeframes,
appeal process and explain the complaints you can and cannot handle.
• Ensure the mechanisms are safe, non-threatening and accessible to all.
• Inform communities about the complaints process, explain it is a right
and encourage communities to use it.
• As much as possible, involve local community members, leaders and
authorities in the handling of registered complaints.
• Provide communities with relevant and timely information about
project criteria and parameters to use the feedback and complaints
mechanisms, and of improvements and changes made to the project
(or why changes are not possible).
• Ensure sufficient time and flexibility of implementation to respond to
complaints.
• Keep records of incoming feedback and complaints, and evidence
of follow-up to allow senior management supervision and external
evaluation.
• Ensure mechanisms are in place for serious complaints, like allegations
of sexual abuse, fraud or other sensitive issues.

Various types of structure were built during the project.


48 Photos: Bill Flinn
Conflict / Complex Shelter Projects 2010 A.16

A.16 Kyrgyzstan - 2010 - Conflict


Case study:

Country: Project timeline


Kyrgyzstan
6 months - –– Project completion
Disaster/conflict:
Civil disturbances
Disaster/ conflict date:
June 10th–11th 2010
5 months - –– All construction
No. of houses damaged: complete
2,000 compounds damaged –– Floor/windows
1,690 completely destroyed complete
4 months - –– Walls, ceiling, roof
No. of people displaced: complete
300,000 people in Kyrgyzstan
75,000 refugees in Uzbekistan
Project target population: 3 months - –– Foundations com-
plete
1,668 family shelters
(13,400 people)
Shelter size: 2.5 months- –– Reconstruction can
28 m2 covered living area begin
(2 rooms of 14 m2) with an 2 months- –– Clearing Debris and
additional verandah of 16m2 demolition
Materials Cost per household:
Up to 5,100 USD per shelter Krygyzstan 6 weeks - –– Emergency shelter
(materials) depending on the Uzbekistan strategy approved
level of damage –– 120 tons of aid
Project cost per household: airlifted
5,900 USD per shelter excluding 3 weeks - –– Project start
operating costs. 2 weeks- –– Offices opened
People building their own –– Rapid joint shelter
houses received 800 USD assessment
June 10th-11th
2010 - –– Disaster/ conflict

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.

Strengths and weaknesses


99 Shelters were designed with the beneficiaries. site engineers to oversee the work and provide advice.
Families were allowed to make modifications. 88 Materials were difficult to procure in volume.
99 The homes were built using better material than Transportation costs were high, flooding of quarries
previous dwellings. stopped sand production, timber was delayed at the
99 Homes were insulated and seismically resistant in border, and some suppliers withdrew from contracts.
line with national codes and international standards. 88 Small access roads, and lack of security on site meant
99 People built their own shelters and were assisted materials could only be delivered in small volumes.
in their work through contribution towards labour 88 Removal of debris was slowed by lack of heavy
costs. machinery and heavy traffic.
99 Homes were built in existing compounds. This 88 Lack of proper documents prevented payment
allowed the families to monitor the construction. through the bank. Security made other means of
99 The project was structured so that implementing payment challenging.
organisations shared responsibilities. -- By building their own houses, work was delayed
99 The implementing partners had good numbers of and quality reduced, but the process acted as training.

www.ShelterCaseStudies.org 51
A.16   Conflict / Complex

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
Conflict / Complex Shelter Projects 2010 A.16

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
A.17   Natural disaster

A.17 Malawi - 2009 - Earthquake


Case study:

Country: Project timeline


Malawi
Disaster:
Earthquake
Disaster date:
December 6th and 20th 2009
18 months - –– Project completion
No. of houses damaged:
6,000
No. of people affected:
24,000
Project target population:
2,400 people (rural and urban)
Government construction
guidelines also developed.
Shelter size:
New build houses - 45m2
House repair - 20m2 to 40m2
Materials Cost per shelter:
House construction (including Malawi 3 months - –– Project start
labour) approx 2,400 USD
Repair grants were 310 USD /
household

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.

Strengths and weaknesses


99 Strong links with communities, government, and 88 The projects had to be implemented within a
other organisations enabled access to the affected short time to coincide with the dry season, to meet
communities. donor requirements, and to meet the expectations
99 Communities and local government were of the community. This created a constraint in
responsible for developing the selection process that terms of time available for staff development and
was managed and implemented by the communities. training, maintaining quality assurance, and the timely
99 International links provided access to technical resourcing of the programme, such as the purchase of
support and specific assistance, especially during the equipment and access to funds.
first phase of the emergency. -- The recovery programme was able to engage
88 The national organisation lacked the technical with other initiatives that were running prior to the
experience to implement large scale shelter programmes earthquake, such as housing and urban planning
and had to rely on external support, especially during the projects, and disaster risk reduction planning and
first phase of the programme. preparedness. Through the support of an international
88 The organisation was the main, and often the only, agency, partnerships were formed with government
provider of assistance after the earthquake. Deploying and other stakeholders to develop a disaster risk
and sharing the resources of other ongoing programmes reduction strategy to assist reconstruction.
was a challenge in terms of personnel, vehicles, office
space, and finance and administration systems.

54
Natural disaster Shelter Projects 2010 A.17

Before the earthquake


Malawi is one of the lowest
income countries in the world,
with many households having cash
income below one US dollar per
day. The population is manly rural,
living in scattered communities
focused on agricultural activities.
Land is allocated through traditional
authorities. In urban areas property
and land is bought or rented.
Traditional houses are built from
wattle and daub with thatched
roofs. While lacking durability, these
dwellings were largely undamaged
by the earthquake, provide good
thermal comfort, and are con- The project lead to the establishment of national guidelines on safer house con-
structed using local materials. sustruction and had a strong training component.
All other buildings are of brick and Photo: Jamie Richardson
block construction.
The number of deaths and • Construction of new houses
At the low-income end of injuries were relatively low but, as • House repairs
brick construction, houses are this part of Malawi had never expe- • Construction of latrines
built with un-burnt brick and mud rienced an earthquake, the popula- • Training of hygiene promoters
mortar. Where there are sufficient tion was traumatised by the event • Training of artisans
resources, bricks are burnt using and was fearful of going back to • Beneficiary dissemination
locally sourced firewood. their homes. workshops
To economise on the use of • Guidelines for safer house
On the 20th December there was
bricks, walls a single brick thick construction
another earthquake.
were often built. These walls are not Two project officers were
earthquake resistant. Additionally, In some areas near the town appointed to manage the urban
the position and size of doors and of Karonga, the land is lower than projects and the rural projects and
windows and the type of un-braced Lake Malawi, and there was the both shared resources and staff.
roof construction, added to the natural fear that the earthquake
structural failing of the buildings. may cause fissures that would lead Guidelines for safer house
to flooding. This community moved construction
Houses are generally construct- to an area of high ground where During the emergency phase
ed over a period of time, as families the government and other agencies of the response, an internation-
gain the resources to purchase the established a temporary camp. al shelter specialist assisted. This
required materials. expertise led to the organisation
In the first months after the
In rural areas most households taking a national lead in shelter
emergency most households slept
owned the houses they lived in. In and allowed an alliance to develop
in temporary shelters outside their
the urban areas many were tenants with government, other agencies
houses. There was a limited dis-
and had less opportunity to build or and non-government organisations
tribution of tents, but for most,
repair their homes. working in housing and shelter.
temporary shelters were construct-
This group was given the responsi-
Issues of public health were ed using local materials such as
bility by the government to produce
greater in the urban area. Latrines in timber and thatch.
guidelines on house construction
the rural areas were generally con- to assist the recovery process. The
Implementation
structed using local materials, whilst guidelines were produced as a
One of the guiding principles
in urban areas many were built of manual and as a series of posters.
for the project was that household-
brick, but were no longer usable.
ers, communities, and government It was recognised that infor-
After the earthquake were responsible for providing safe mation should be made available
The first earthquake on the 6th and adequate housing. The or- nationally to reduce the risk of all
December destroyed and damaged ganisation would provide support hazards, including earthquakes.
thousands of houses, hundreds of were there were gaps in skills, The guidelines would be the start of
schools and public buildings, cracks knowledge, and resources. a process to create national guide-
appeared in the ground and the The following parallel activities lines and standards for construc-
levels of the earth altered in some were implemented: tion.
locations.

www.ShelterCaseStudies.org 55
A.17   Natural disaster

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
Natural disaster Shelter Projects 2010 A.18

A.18 Mozambique - 2007 - Cyclone Fávio


Case study:

Country: Project timeline


Mozambique
Disaster:
Cyclone Fávio 34 months - –– Project completion
Disaster date:
February 22nd 2007 32 months - –– Technical manual
No. of houses damaged: printed
6,500
No. of people affected: 28 months - –– Completion of
160,000 people the first phase of
Project target population: building houses
10 communal cyclone shelters
22 months - –– Project start
Shelter size:
35 m2
Materials Cost per shelter:
3,500 USD
Project cost per shelter: 13 months - –– Assessment ends
5,000 USD
10 months - –– Assessment and
design starts

Mozambique 1 month - –– Initial damage as-


Vilankulo sessment

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.

Strengths and weaknesses


99 A construction handbook was developed and 88 High cost of ferro-cement for the context prevents
accepted by local governments as a building code. This large scale uptake.
was translated and disseminated in other countries. 88 During the first phase, kitchens, latrines and water
99 The project focused on mitigating the impacts of wells were not included in the project.
future events. -- There was a process of experimentation and
99 Continuous technical support provided by “learning by doing”.
architects and civil engineers. Training of local master -- Some architecture students and other municipalities
builders. that have the same risk were interested by the project.
99 Construction techniques were accepted by local -- After the project, some residents constructed
population, local master builders and the municipality. houses using the ferro-cement roofing technique, and
99 The city council remained the owner of the shelters. the council decided to replicate the vaulted structure
They were freely provided to the most vulnerable in a school.
households in the city. -- Some humanitarian institutions were interested in
99 The houses serve as community cyclone shelters replicating the construction techniques in new schools
for the neighborhood sheltering about 1000 people. and emergency infrastructure located in disaster prone
88 The design did not allow the houses to be modified areas.
or extended.

www.ShelterCaseStudies.org 57
A.18   Natural disaster

The project built 10 cyclone shelters, but used the project to advocate nationally
for safer construction.
Photo: Arianna Francioni

Before the cyclone winds. This was due to the construc-


Mozambique has a large coastline tion techniques and quality.
exposed to the Indian ocean leading The study reported that it was
to the threat of cyclones. Addition- more sustainable to reconstruct in
ally Mozambique is prone to floods, a resistant manner than to spend
droughts and earthquakes. The money every two or three years re-
number of events has dramatically building after cyclones.
increased this century.
Coastal areas of Mozambique Catalogue of housing
are very prone to cyclones. However, Following the cyclone, a
at country and local level, there catalogue of different housing
was neither interest nor technical solutions adapted to urban contexts
knowledge on how to build resistant in Mozambique was produced. It
houses and / or community shelters. included some pre-industrialised
techniques and materials. The
Vilankulo municipality is a geo- catalogue targeted master builders
graphically exposed coastal town and technical staff.
that has been hit several times by
cyclones and strong winds. Prototype testing
To support the introduction of A construction manual was
In the urban context of Vilankulo cyclone-resistant construction tech- developed, adopted nationally and
municipality, infrastructure and niques in the country, and to dem- shared regionally.
houses were very vulnerable to onstrate the real feasibility of the
strong winds. technical proposals in the manuals,
After the cyclone different prototypes using different
materials were tested by being built The organisation worked with
In 2007, Cyclone Fávio hit some
in the field. the municipality at local level, and
coastal areas of the central region
the National Institute of Disaster
of Mozambique, generating tor- In building the prototypes ways Management at country level. This
rential rains and wind speeds up to improve them would be iden- allowed different government insti-
to 220 Km/hour. It hit an area that tified. It was also hoped that by tutions to be involved and allowed
had already been flooded the month building them replication would also for advocacy at different levels.
before. High wind speeds caused the be encouraged, as craftsmen would
majority of damage. “learn by doing”. Implementation
Some of the techniques tested
Field assessment The site where the shelters were
were already available in Mozam-
One month after cyclone Fávio, built was located in a council owned
bique and others were imported
a technical team conducted a field land in a suburb area, home to some
from different countries around the
assessment in order to determine of the most vulnerable people in the
world. Architects, engineers, and
damage on houses and public fa- city. The original houses that were
students of the national university
cilities, and realised that most of the built on the site with local materials
participated in the process of design
buildings were not resistant to strong were destroyed by the cyclone.
and building.

58
Natural disaster Shelter Projects 2010 A.18

“The purpose of demo


projects is to ultimately
influence national
and local policies,
so that if proper
building techniques are
integrated in the codes
and regulations, then
the scale of the impact
is both national/local.”

As part of the project a manual, posters and other materials were developed.
Credit: Eduardo Feurhake Project technical director

In the town, local master builders


and municipality technical staff were
very involved in the construction of
the shelters. They received special
trainings on the new techniques and
practiced by building the shelters.
Inhabitants of the suburb could
see the site and visit the houses. This
involvement allowed local people to
become aware of the importance of
building resistant houses, and which
techniques to use.
In the municipality some families
are now building their new houses
replicating the techniques. The mu-
nicipality is also building new facili-
ties according to the building recom- In building the cyclone shelters, local builders “learnt by doing”.
mendations and new techniques. Photo: Fernando Ferreiro

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
A.18   Natural disaster

60
As part of the project a game was developed to share learning surrounding the threat of cyclones and flooding.
Credit: Eduardo Feurhake
Natural disaster Shelter Projects 2010 A.19

A.19 Myanmar - 2008 - Cyclone Nargis


Case study:

Country: Project timeline


Myanmar
Disaster:
Cyclone Nargis
Disaster date: 43 months - –– Originally planned
project completion
May 2nd 2008
No. of houses damaged /
destroyed:
39 months - –– Project completed
42,194 in Dedaye Township
(172,000 in all Nargis affected
areas)
No. of people affected:
160,000 in Dedaye Township
(2,433,300 in all Nargis affected 25 months - –– Project start
areas)
Project target population:
1,658 households (8,250
people)
96 carpenters employed Myanmar
Shelter size:
15.6m2 covered space per family
Project cost per shelter:
650 USD
Dedaye

May 2nd 2008- –– Cyclone Nargis

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.

Strengths and weaknesses


99 Local communities were at the centre of the process 88 The project met the needs of less than 4% of the
of decision-making and all activities performed at the affected population.
local level were recognised and owned by them. This 88 In one village, beneficiary selection became highly
led to the project concluding four months before the contentious because nearly everyone in the village had
originally planned completion date. suffered great losses as a result of the cyclone.
99 Communities benefitted from complementary 88 Some timber on shelters scheduled for retrofitting,
water and sanitation activities such as reservoir turned out to be rotten on the inside requiring
ponds, tube wells, water tanks and school latrines. additional work and materials.
88 The project did not start until 25 months after the 88 While some of the target villages were located in
cyclone. remote areas of the township, the project was less
88 Buildings made from toddy palm timber can successful at reaching individual households or clusters
withstand strong winds, but are not as strong as of households that were far from village centres.
buildings made from hardwood timber. Hardwood -- It is hoped that villagers who are not direct
timber was too expensive for the available budgets. beneficiaries of this program will take note of the
88 The shelters will not be sufficient to withstand Disaster risk reduction components of the project.
another event of the magnitude of Cyclone Nargis.

www.ShelterCaseStudies.org 61
A.19   Natural disaster

Construction was managed by Village Reconstruction Committees who


handled all of the funds required.
Photos: UN-Habitat Veronica Wijaya

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
Natural disaster Shelter Projects 2010 A.19

The allocated funds were dis-


bursed in two instalments; 80% of
funds were released at the incep-
tion and the remaining 20% were
given once a benchmark of works
stipulated by the Community
Contract was completed.
The Village Reconstruction Com-
mittees were responsible for paying
the carpenters, other artisans and
labourers, and for disbursing funds
for the purchase of materials. In
the interest of transparency, the
amount given to each committee
and then to each group of benefi-
ciaries, was publicly posted so that Community meeting. People organised themselves to identify and prioritise their
needs and together take decisions on their recovery.
it could be reviewed by anyone in
Photo: UN-Habitat Veronica Wijaya
the community.

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
A.20   Natural disaster

A.20 Myanmar - 2008 - Cyclone Nargis


Case study:

Country: Project timeline


Myanmar
23 months - –– Project completion
Disaster: and evaluation
Cyclone Nargis
Disaster date: 22 months - –– Construction
finishes
2nd May 2008
No. of houses damaged:
172,000
No. of people affected: 16 months - –– Start of
2,433,300 construction
Project target population:
533 households 15 months - –– Training of local
partner
Shelter size: –– Assessment,
20m2 consultation,
Materials cost per household: selection of
beneficiaries
600 USD
Project cost per household: 13 months - –– Evaluation of local
970 USD approximately partner
12 months - –– Second periodic
Myanmar review

7 months - –– First periodic


Yangon review

May 2nd 2008- –– Cyclone Nargis

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
Natural disaster Shelter Projects 2010 A.20

“It is not difficult


to build a decent
house, but it is
hard to get good
materials. “

The local partner organisation

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
A.20   Natural disaster

Training focussed on key messages such as making Households were later able to upgrade their shelters.
good connections.
Photo: Tina Salsbury

winning the contest in 31 groups


A table from an end of project evaluation assessing the quality of shelters and the
spread through the four villages.
shelter design
13 teams of 4 carpenters were 1: Choose location Poor adaptation to local site : some sites
trained and helped to build the full to avoid force of flooding at high tide ; some on rock required
wind different foundations.
scale model house in Yangon at the
start of the project, so that they were
2: Use simple Good.
familiar from the outset about the regular shape
ten principles of cyclone resistant
construction and about the different
techniques being proposed to make 3: Keep roof angle Good.
the houses more storm resistant. above 30°
Posters were distributed. These
showed ten key principles of safe 4: Separate roof, No lean to structures were planned, and only
construction and details about safe avoid large roof at the end of the project have families started
bamboo and frame construction. overhang to add on to their house. Most know about
having a separate roof and respecting the key
They were put up on nearly all the principles.
houses and in the villages. 5: Good Yes, quite good; families have difficulty to find
Most groups of families could connections the same fishing line, and suggest using nylon
fishing string, which would be ok; people like
remember many of the ten key the use of nuts and bolts.
points, and in several cases this was
6: Diagonal Yes, well integrated.
done with considerable animation bracing
and mime. Non-beneficiary families
also knew some of the principles.
7: Fix roof down Yes, with bamboo trellis frame over the
In a project evaluation, carpen- thatching panels.
ters knew the construction princi-
ples, but could not always articulate
this verbally. They said that they did 8: Opposing Yes.
not know how to convince clients to openings
spend money on greater safety.

Tools 9: Window/door Yes.


leaves shut
The teams did not get any tool
kits. Each house required about
110 holes to be drilled. The holes 10: Plant trees as Many sites so far have nothing on them, and
for bolts were made with an auger, wind breaks planting may be difficult because of terrain in
which was laborious. The carpen- 89 cases on rock.
ters said that the work would have
been easier if each team had been Strong enough
adequately supplied with good tools.
Acceptable quality, needs to be improved

Poor, needs more attention in future

66
Conflict / Complex Shelter Projects 2010 A.21

A.21 Pakistan - 2009 - Conflict


Case study:

Country: Project timeline


North Pakistan, FATA and KPK
Conflict: 18 months- –– Project completion
IDP (Internally displaced people)
crisis
Conflict date:
July 2009
14 months - –– T-shelter project
No. of houses damaged: resumes as flood
More than 30,000 houses waters recede
No. of people affected /
12 months - –– 9,585 transitional
displaced: shelters built
2.7m people internally displaced –– Project suspended
Project target population: due to flooding
16,260 households
(approximately 115,000 people)
Occupancy rate on handover:
93% at time of floods (July
2010)
Shelter size:
Approximately 25m2
Materials Cost per shelter:
682 USD (including kitchen, Pakistan
latrine, purdha wall and tools) –– Project start
2 months-
Additional costs per shelter: –– Returns begin
44 USD (partner costs)
132 USD (design and
monitoring) –– Displacement
July 2009 -
25 USD approx. (cash for work)
Staffing / admin costs excluded

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.

Strengths and weaknesses


99 The project considered cultural needs; with separate 88 Landless families depended upon land being
kitchens and latrines for men and women, and purdha allocated to them.
walls for the women's privacy. 88 Procurement delays, material transport problems,
99 Community organisations (Jirga), consisting of a security concerns and flooding in 2010 led to the
diverse range of community members, were established project being delayed and extended by 8 months.
to be involved in the selection of beneficiaries. 88 Initially partner organisations collected construction
99 The project used unskilled labour from within the materials from the central warehouse, leading to delays.
community for construction work. This contributed to It took many months to readjust the project to allow
good relations between families that received shelter partners to procure their own materials. Occasionally,
and those that did not. the wrong materials were delivered.
99 Each household received 2,000 PKR (22 USD) upon 88 Timber in particular was not procured quickly
completion of the shelter. This ensured the timely enough to meet the planned programme.
completion of the shelter and reduced the later sale of 88 During the first phase of the project the vulnerability
shelter materials. assessments were not carried out. As a result some
88 The depletion of local material resources and skilled families never occupied their shelters.
labour led to varying costs of shelters, as well as 88 Limitations on movement and access often created
delaying the delivery of shelters. problems for material delivery which lead to delays.

www.ShelterCaseStudies.org 67
A.21   Conflict / Complex

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

Before the conflict On return, most returnees Selection of beneficiaries


People lived in three main types found their houses either partially Shelter assistance was priori-
of houses prior to the conflict: or completely destroyed. A damage tised for vulnerable people who
assessment was made, using the had non-repairable homes. People
• Pucca houses with a reinforced
following classifications: whose houses were repairable but
cement concrete frame and
not habitable were also eligible for
foundation. The walls are made 1) Repairable assistance, but other vulnerability
out of burnt bricks, blocks or • Pucca houses: the reinforced were considered. These included:
stone, with sand and cement concrete structure remained in
mortar. The roof is made from good condition, even if walls • health status (chronically sick
tiles, slates, reinforced concrete, were completely destroyed. people, people with disabilities,
cement sheets, or metal • Semi-pucca houses: the etc),
sheeting. structure was in good condition • financial assets (regular income
• Semi-pucca houses are similar to and damage to the walls or roof or resources family size),
pucca houses but do not have a was away from the corners. • social assets (position within
reinforced concrete frame. They • Katcha houses: damage was the community, social network,
have good quality masonry walls limited to a small section of etc),
cement mortar and a reinforced the wall or roof away from the • natural resources (ownership of
cement concrete or corrugated corners and not affecting the land, forests, orchards, water
galvanised iron (CGI) sheet roof. structural integrity. resources, etc),
• Katcha are houses with mud or In all cases, any surface damage, • number of dependents (children
dry stone masonry walls and a such as bullet holes, cracked plaster, under 18 and women),
mud roof with wooden panels, broken windows and doors, etc, • family situation (widowed,
wooden beams or CGI sheets. could be repaired. orphaned, elderly without
Most Katcha houses have a family, etc).
timber frame. 2) Non-repairable
• Houses that were completely
Community organisations
After the conflict destroyed as well as those with
Community organisations,
Following the military opera- known as Jirga, were formed. Each
significant structural damage.
tions in FATA and KPK provinces contained at least 12 members
• All Katcha houses with any
in July 2009, more than 2.7 million from different tribes, and at least
structural damage.
people were internally displaced. one school teacher, imam, khan
• Pucca and semi-pucca houses,
When the area was considered and nazim (village councillor). It
those with more than 40%
safe, the Government of Pakistan was encouraged that there should
structural damage.
started a returns process. be women in the Jirga.
The Jirga were involved in the
selection of beneficiaries along
with the six implementing partner
organisations. A survey form was
completed for each beneficiary to
verify their eligibility for support,
and the information was stored in
According to local tradition, privacy a database along with GPS data to
screens (purdah walls) were built
record the locations prior to the dis-
around shelters.
Photo: Arif tribution of materials.

68
Conflict / Complex Shelter Projects 2010 A.21

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
A.22  Overview Natural disaster

A.22 Pakistan - 2010 - Floods - Overview


Overview
Summary
The 2010 monsoon season caused the worst flooding in
Pakistan’s history, and one of the larger humanitarian crises
of this century. The floods affected every province, over
half of the districts in Pakistan, and one-tenth of Pakistan’s
population. They damaged or destroyed 1.8 million homes,
from the mountainous north where winters are cold, to the
south where flooding caused the most damage. The scale
was vast, but the funds did not meet the needs.
For the first months, the government of Pakistan and
many organisations working in the affected areas distributed
tarpaulins, tents and other non-food items. The government
also made cash payments to registered flood affectees using
a “WATAN Card”.
Following the emergency response, a “one room shelter”
approach was adopted, by which organisations supported
families to build a permanent shelter, which families could
later extend. However the scale of the floods was such, that
less than 10% of those who lost a house received such a
shelter.

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
Over 1 million tarpaulins 380,000 tents were distributed. Additionally there was also a very large scale, though less coordi-
nated, response from Pakistani civil society.
Photo: Joseph Ashmore

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
A.22  Overview Natural disaster

BEFORE THE FLOOD

DURING THE FLOOD - People are displaced to higher ground.

Spontaneous Collective Host families Camps


sites centres / rental only where
absolutely
necessary
Note: Spontaneus sites includes dispersed shelter where one or two familes settle on elevated land near their houses.
Camps require significant resources and can increase the challenges in return and recovery.

RETURNS BEGIN

Public information individuals


Advocacy
Livelihoods support Support for return
Damaged Destroyed Spontaneous Collective Host families
houses houses sites centres / rental

Targetting of vulnerable Displacement tracking and needs


Tents not required for houses that can easily be repaired, nor for collective centres.
Tool kits and Community clean up kits for damaged houses.
Collective centres rehabilitated.
Cash and vouchers should be considered where local markets are able to support the needs.

RETURNS CONTINUE - Some people have lost their land and are unable to return.
Coordination

Damage assessments Analysis of gender differences

Damaged Destroyed Spontaneous Host families


houses houses sites / rental
Monitoring and review

Longer term solutions


Tarpaulins & rope,
should be sought
Construction materials Tool kits,
Tool kits, Household kits
Household kits cash
Community clean up kits

As returns continue:
- Most people move to the land where their damaged or destroyed house was.
- Collective centres and many spontaneous camps close. sites and buildings are rehabilitated.
- Organisations consider transitional shelter as support.

An illustrated version of the initial Shelter Strategy for Pakistan floods - 20th Aug 2010

72
IASC PAKISTAN FLOODS
SHELTER CLUSTER
Natural disaster
Inter-Agency Standing Committee
Shelter Projects 2010
www.shelterpakistan.org
A.22

PAKISTAN FLOODS RESPONSE AFTER 8 MONTHS


Although the disasters compared in this document are very different in nature, human impacts, and challenges, this document makes
some numeric comparisons based on data collected from the shelter cluster for different responses.

Area affected People who lost their houses


Pakistan 2010 floods 11 million
Pakistan floods 2010
100,000km2
1.5 million
Haiti 2010 earthquake
The ‘slow tsunami’ that hit Pakistan in 2010 damaged or destroyed an
estimated 1.7 million houses, leaving at least 11 million people home-
less.
In Punjab alone, twice as many houses were damaged destroyed by the
floods than by the 2010 Haiti earthquake.

Haiti 2010 1,000,000

Damaged / Destroyed
<5,000Km2

Gilgit Baltistan
750,000

Punjab

Haiti 2010
Sindh
Houses

Balochistan
(drawn to scale)
500,000

KPK
PAK
250,000

Emergency shelter GAP map - 9th November 2010 0

EMERGENCY SHELTER RESPONSE


The cluster estimated that 70% of those with damaged and
destroyed houses will need emergency shelter support.
381,000 tents 2.5 million
delivered
blankets
delivered or on
Emergency shelter has been delivered to over 1 million their way
1.36 million
households. Despite the scale of response, only 67% of the
emergency shelter needs have been met. tarpaulins
delivered 1.5 million
more required
further gap filling still required for the
Other Non-Food Items distributed are 438,600 bedding sets, remaining need
603,200 kitchen sets and 94,500 tool kits. estimated 30% of self recovery, with
host-family support or in collective centre

SHELTER CLUSTER FUNDING 48% funded $168 million unmet requirement


US$ 322 million requested

EARLY RECOVERY SHELTER RESPONSE

Emergency: “One room” shelters As time progress families upgrade


Tents or tarpaulins Technical guidance one room shelters to form a home

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
 
A.23 Natural disaster

A.23 Pakistan - 2010 - Floods


Case study: See A.22 “Pakistan - 2010 - Floods - Overview”, p70 for background.

Country: Project timeline


Pakistan, Sindh.
Disaster: 18 months - –– 21,700 shelters
Floods constructed.
Disaster date:
July 2010
No. of houses damaged:
About 1.8 million in 77 of 139
districts across Pakistan
No. of people affected:
More than 20 million
Project target population:
Pilot project 175 households 8 months- –– Project completion
Occupancy rate on handover:
100% as of 10 March 2011
Shelter size:
25m2
Materials Cost per shelter: 5 months - –– Project start/flood
USD 740 Pakistan waters recede

2 months - –– Flooding reaches


Sindh northern Sindh
–– Relief operations
start

July 2010 - –– Floods start

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.

Strengths and weaknesses


99 Conducting a pilot project allowed issues with the Although it was extended to 6 weeks, the actual time
project process to be identified before a large scale taken to build the 175 shelters was 11 weeks. This was
project was implemented. due to time taken to mobilise the community and rain
99 The returning families were supported to return to interrupting work.
their original locations. 88 Lack of planning a delivery schedule and use of a
99 The community were consulted and involved single supplier led to the late supply of construction
throughout the project, including the needs materials.
assessments process, shelter design, materials 88 Lack of detailed specifications led to poor quality of
sourcing and shelter construction. materials. This was compounded by lack of ownership
99 Skilled labourers from nearby villages also by the flood affectees leading to additional damage of
benefitted from the opportunity to work on the materials during transport and unloading.
shelter construction. 88 There was a problem accommodating labourers
99 The projects were flexible for the different needs from other villages, so tents had to be provided.
of each village, depending on the social situation and 88 While the affected population were involved in the
the preferred balance of shared facilities and privacy. project they could not focus on normal livelihoods
88 Once the project was started, the planned project activities, so the provision of food became an issue.
duration of 4 weeks proved to be insufficient.

74
Natural disaster Shelter Projects 2010 A.23

The project was a pilot for a larger construction programme.


Photo: Kpakpo

Before the floods original design: formal Non Objection Certificate


See Background: A.22 “Pakistan (NOC) was obtained. In one of the
• A high water table meant that
- 2010 - Floods - Overview”, p70. three villages, all families owned
the height of the foundation
a piece of land, but in the other
Selection of beneficiaries wall was increased from 150mm
two villages the land belonged to
A village assessment was carried to 450mm.
a landlord, who provided the NOC.
out to select three villages in one • 10% cement was added to
district in Sindh for the pilot project. the mud mortar to increase its Implementation
durability. The project was designed to
A family assessment was then • At the request of the ensure a degree of participation by
used to identify eligible benefi- beneficiaries, the pitch of the those receiving the shelters. Each
ciaries. Each selected village had roof was reduced and the door family provided unskilled labour,
between 40 and 80 families who dimensions were changed to and was responsible for plastering
were eligible for the project. 1.2m x 2m. inside the shelter. The community
• As the sites were dense, the based organisations located and
Community mobilisation
distances between shelters contracted skilled labour from
Community meetings were
were reduced and sometimes nearby villages.
organised to discuss the needs of
shelters joined.
the returning families and participa- In the pilot project beneficiaries
• To meet individual community
tory rapid assessments were carried were not paid. Their contribution to
requirements, one village,
out. Community based organisa- the project was to provide unskilled
consisting of one extended
tions were established to manage labour and salvage materials.
family, built communal toilets
the projects. Members were trained
and washing facilities. In other Logistics
on mobilising their communities.
villages where families wanted A single supplier was identi-
The projects were coordinated more privacy, houses and fied following a tendering process
and monitored through a district individual toilets were built in that included taking out advertise-
coordinator and senior engineer, long rows. ments in the newspaper. However,
regularly reporting to the District
Coordination Office and the District Land allocation the contract did not stipulate the
Before construction could delivery schedule. As a result no
Disaster Management Authority. In
begin, field teams verified that materials were delivered in the first
addition, each village had its own
there were no land disputes and a 2 weeks of the project.
site engineer and logistics assistant.

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
A.23 Natural disaster

During the procurement,


material specifications only
indicated the dimensions of the
products and as a result, the quality
of material varied. Furthermore
there was lack of clarity over who
was responsible for the materials
once they had arrived on site. As
a result, a lot of bricks and roofing
materials were damaged during
off-loading and moving.
Families were expected to
provide around 10% of the
bricks that were required through
salvaging materials.

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
Natural disaster Shelter Projects 2010 A.24

A.24 Pakistan - 2010 - Floods


Case study: See “Pakistan - 2010 - Floods - Overview”, p. 70 for background.

Country: Project timeline


Pakistan
Disaster:
2010 Floods
Disaster date:
July – September 2010 16 months- –– 38,500 shelters
No. of houses damaged: complete
1.8 million houses damaged or
destroyed
Project target population:
38,500 households 12.5 months - –– 20,000 shelters
completed
Estimated 217,617 beneficiaries
targeted 11 months - –– 10,466 shelters
Shelter size: completed
Shelter sizes vary. 225 square
10 months - –– 3000 shelters com-
feet (70m2) was recommended plete
For mud structures, this was the
suggested maximum
Materials cost per shelter:
300 USD cash per shelter
provided
Pakistan
If DRR recommendations are 5 months - –– Project start
followed cost to beneficiaries
500 USD for a mud house, 4 months - –– Most people able to
return
1,000 USD for a fired brick
house July to
September
2010 - –– Floods

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.

Strengths and weaknesses


99 Cash transfer allowed households to use money in 88 Because households were free to choose the
the way they saw fit. construction materials they wanted, giving out disaster
99 Transferring cash instead of materials meant that risk reduction advice to each household was difficult.
materials were purchased locally. 88 Not all of the implementing partners had the shelter
99 Using community focal points to distribute cash experience or the staff capacity to cope with the project
proved to be overwhelmingly reliable. requirements.
99 25 households built shelters as a group, supporting 88 Some organisations working in nearby sites
each other in order to receive the next payment. provided different amount of money, leading to initial
99 Disaster Risk Reduction trainings and messages to dissatisfaction among recipients and some drop outs.
communities resulted in safer houses. 88 Internal requirements on financial accountability
99 A monitoring and evaluation and an information led to a significant amount of paperwork, requiring
management system ensured that the programme 59,064 separate signed documents (various forms,
was carefully tracked. MoUs, approvals, receipts, checklists, etc.).
88 The banking system in Pakistan lead to cash -- Identifying the most vulnerable households required
transfers often being delayed. major efforts from the implementing partners and
extra verification from the organisation.

www.ShelterCaseStudies.org 77
A.24   Natural disaster

Before the floods Beneficiary Selection Distribution in 3 tranches


See Background: “Pakistan - 2010 The village committee was re- Each household received 100
- Floods - Overview”, p. 70. sponsible for selecting the most USD as an advance for digging
vulnerable households using the the foundations and constructing
Implementation overview
following criteria: the shelter up to the plinth level.
The project built One Room
Once the implementing partner
Shelters (ORS) through a cash 1) The house must have been
had verified that all plinths had
transfer system. One room shelters destroyed or heavily damaged as
been completed, a request for the
were traditional houses that could a result of the 2010 floods.
2nd tranche of 100 USD was made
be extended later. They were 2) In addition, one or more of the
and funds were subsequently dis-
generally built with brick or mud following criteria were met: no
tributed.
walls. adult male in the family, disabled
family member, medically unfit Once the Implementing Partner
The project allocated a total
family member, elderly family had verified that all walls had been
of 300 USD to build each shelter.
member, family taking care of completed, a request for the 3rd
Households were then able to use
orphans, large family, or poor tranche of 100 USD was made and
the money to procure materials
family. a final cash distribution was made
and/or labour as they saw fit.
The implementing partners for the construction of the roof.
Technical advice was provided verified that the targeting had
to help families to improve their re- Cash transfer focal points
been done using the criteria before
silience to future disasters. In each village, a representative
people joined the programme.
village committee was established.
The 300 USD that was provided Monitoring staff also verified com-
This was responsible for choosing
to each household was nationally pliance with targeting criteria.
a highly dependable and respected
agreed between organisations in person from the community who
Groups of 25 households
November 2010. Some projects would act as the focal point for the
The organisation insisted that
that started later, or that worked in project.
every household participating in
areas with seismic risks provided a
the project worked as a group and The focal point brought money
larger amount per shelter. The or-
completed each stage of construc- paid via his/her personal bank to
ganisation continued with 300 USD
tion before any household would the village and distributed it to each
per household because it allowed
receive the next tranche of funds. group of 25 beneficiary households.
more people to benefit.
The group of no more than 25 When the focal point accepted
Many households added their
beneficiary households had to work the responsibility, he/she was given
own resources to build their houses,
together as a unit. It was stressed a Memorandum of Understanding
in some cases selling assets such as
that none of the beneficiaries to sign and told that bank charges
livestock. Many beneficiaries also
would succeed unless all of them related to the programme activities
used emergency funds provided
succeeded. would be covered. Upon the suc-
by the government through the
“WATAN” card system. It was understood that they cessful conclusion of the project,
had to help the most vulnerable he/she would also receive 175 USD.
Most households participating in
people to complete their shelter as It was explained that when ben-
the project built shelters that were
a precondition for getting money to eficiaries signed the Memorandum
significantly better than the house
construct their own shelter. of Understanding to participate
they had been living in before the
floods. in the programme, they were also

The project provided cash in three tranches


so that people could purchase materials and
hire labour.
Photo: Usman Ghani, IOM Pakistan
78
Natural disaster Shelter Projects 2010 A.24

accepting the focal point to receive


money on their behalf. Once the
focal point had received the money,
it was no longer the responsibility
of the implementing organisations.
The focal points brought three
separate tranches totalling up to
2,500 USD each to the community.
In only a handful of cases were
funds not delivered to communi-
ties according to plan. This rela-
tively small number of cases should
be seen in the context that the
programme worked with approxi-
mately 1,600 focal points.
Housholder in front of her completed house. This was one of 11 photographs
taken per household as part of the required project documentation.
The 11 required photographs Photo: IOM Pakistan
were:
1. Head of household standing Land rights is a major issue in In an effort promote transpar-
in front of existing shelter Pakistan, especially in Punjab and ency and answer any questions
2. Head of household holding Sindh provinces which have large relating to the programme, posters
CNIC (National Identity tracts of land under the control of in local languages were printed and
Card) landlords. In this project, no dis- posted in the communities. These
3. Close up of CNIC card tinction was made between those posters had the phone number of
4. Head of household standing owning and those renting land. a call centre where people could
in front of the empty plot obtain information and make com-
Monitoring plaints.
where they are going to Implementing partners were
build required to fill in forms that es- Programme management
5. Head of household receiving tablished the vulnerability of the The programme operated in 3
first tranche (thumbprint on household, tracked the progress provinces. The programme head-
documentation) of the construction and tracked quarters were in Islamabad but the
6. Plinth level completion the distribution of tranches to ben- day-to-day management responsi-
7. Head of household receiving eficiary heads of households. In bilities were devolved to four hub
second tranche (thumbprint addition, the organisation had its offices.
on documentation) own team that monitored around
8. Wall level completion In Northern Pakistan, the or-
7% of the households to verify ganisation directly implemented the
9. Head of household receiving targeting and to ensure that the
third tranche (thumbprint construction of shelters, whereas in
construction progress reported by Punjab and Sindh provinces the or-
on documentation) the implementing partners was
10. Roof level completion ganisation worked with 44 Imple-
being accurately described. menting Partners, of which all but
11. Household occupying the
shelter A great deal of monitoring both 4 were local agencies.
by implementing partners and the
Technical aspects organisation ensured that house-
After testing techniques during holds were meeting the construc-
the pilot phase, multiple, highly tion thresholds.
practical trainings for the im- GPS coordinates and 11 photo-
plementing partners were held. graphs taken during the course of
Different messages were developed cash distribution and construction
for different types of construction. were required for each beneficiary
Many field visits were made household so that construction
to ensure that the messages were progress could be comprehensively
being disseminated to communi- tracked throughout the process.
ties and used in the construction. All of the information and pho-
Three posters were produced that tographic evidence was uploaded
showed the three main construc- into a large database that was Focal Points distributed money paid
tion typologies and techniques that managed by a team of information via their personal bank account to
could make shelters stronger and managers. each group of 25 households.
more flood resistant. Photo: Carlo Gherardi

www.ShelterCaseStudies.org 79
A.25   Natural disaster

A.25 Pakistan - 2010 - Floods


Case study: See “A.22 Pakistan - 2010 - Floods - Overview”, p.70 for background

Country: Project timeline


Pakistan
Disaster/conflict:
Floods
Disaster/ conflict date:
July – September 2010
No. of houses damaged: 12 months- –– Project completion
1.8 million houses damaged or
destroyed
No. of people affected:
More than 20 million people
Project target population:
300 families
Occupancy rate on handover: 7 months- –– Project start
Unknown
Materials cost per shelter:
316 USD
The overall programme cost was
significantly higher

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.

Strengths and Weaknesses


99 Thick mud walls and roof keeps the inside of the 88 Material costs increased by 38% during the
shelter cool during summer. implementation period.
99 Mud was sourced from the immediate vicinity, 88 Demands for skilled masons exceeded local supply.
reducing logistical delays and environmental impacts. 88 Harvesting and planting seasons interrupted
99 Beneficiary led material procurement and quality construction.
monitoring resulted in high quality materials. 88 Some local leaders tried to influence community
99 Vouchers redeemable for cash to make labour committees and suppliers for political reasons.
payments allowed people to use the cash payments 88 Budget did not allow sanitation facilities to be built.
for other requirements. 88 Variable security hindered monitoring activities.
99 This shelter design was acceptable to landowners. 88 Recruitment of qualified field staff was extremely
The roof is the majority of the investment and difficult given the competition and scale of
belonged to the families. reconstruction in the area.
99 The shelters were designed so that during flooding, 88 Structural integrity of the shelter highly dependent
the mud walls could be washed away leaving the roof on good quality foundations. This was difficult to
intact. monitor.
99 The shelter could be easily extended. Another 2 88 The project was unable to provide guaranteed
columns and 1 girder would allow the shelter size to security of tenure for the recipients due to the immense
be increased in size by 50%. power of local landlords and the entrenched feudal
88 Relatively small scale project given size of disaster. systems of landownership.

80
Natural disaster Shelter Projects 2010 A.25

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
A.25   Natural disaster

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

Technical solutions impossible to negotiate secure ciaries had undertaken work to


The design was based upon tenure. the required quality, and reached
brick and cement mortar columns. Despite negotiation, the land- the next stage in the construction
Each column was built on an indi- owners refused construction of process.
vidual foundation. These columns any form of full masonry (Pukka) Pre-selected vendors would only
were aligned to support 3 steel structures. Under customary law receive payment once all vouchers
girders and a traditional style flat the landlord automatically owns for a community had been signed
roof of bamboo, reeds (Khick), any part of a structure sunk into the off by agency staff. Brick factories
plastic sheet and mud plaster. ground. The roof which is the major engaged in the project were
Beneficiaries chose whether to part of the value of the shelter monitored to ensure that no child
build walls with either traditional would still belong to the families. labour was used.
compacted mud or with sun dried The expensive roofing girders could
mud bricks. in theory be removed and taken Implementation
away if the family were ever evicted. Construction was completed in
The shelter was finished with a comparatively short timeframe
either a mud or a stabilised mud MoUs were agreed with land- once project preparations had been
plaster on both the interior and owners prior to construction to finalised. The use of vouchers was
exterior walls to provide protec- ensure beneficiary rights were extremely successful in ensuring
tion from slow rising flood water or secured as much as practically timely delivery of good quality
heavy rain. possible and would not be evicted materials. The communities policed
from their homes to make way for the process vigorously and did not
In the event of serious flooding
other workers or families. hesitate to reject any materials they
and fast flowing waters the walls
between the columns would Vouchers considered to be of poor quality.
dissolve leaving the roof intact. To increase participation, and in Any reloading and additional trans-
consideration of a highly variable portation costs were the responsi-
The approach used materials bility of the suppliers and they rarely
and techniques that are familiar to security environment, a voucher
system was used which also reduced attempted to supply poor materials
the targeted communities. The thick more than once.
mud walls help to keep the interior both the logistical burden and some
of the Shelter cool even during the of the quality control responsibility At some stages in the project,
extreme summer heat when tem- of the implementing agency. seasonal cultivation activities
peratures rise above 500 C. Beneficiaries were trained in reduced the availability of labour.
minimum quality requirements During the 4 months of project
Land tenure
of the materials and then issued implementation, inflation increased
The majority of the families in
with a booklet containing phased material prices by 38%. The use of
the project were tenant farmers or
vouchers for both materials pro- vouchers meant that the agency
indentured labour who had for the
curement and labour payments. could negotiate directly with
most part occupied these areas for
Vouchers could only be vendors for any adjustments in the
generations. Due to the complex
redeemed following sign off from value of the redeemed vouchers
feudal system of land ownership
field staff who ensured benefi- without slowing construction.
that dominates the region it was

82
Natural disaster Shelter Projects 2010 A.26

A.26 Philippines - 2010 - Typhoon Megi


Case study:

Country: Project timeline


Philippines
9 months- –– Project Completion
Disaster:
Typhoon Megi
Disaster date: Philippines
October 18th 2010
No. of houses destroyed: 7 months- –– Reassessment
30,048 (destroyed)
118,174 (damaged)
Project target population:
49,765 people (9,953
households) in Cagayan, Isabela, 4 Months- –– Ban on harvesting
Kalinga and La Union timber - new
approach to
Materials Cost per household:
vouchers
160 USD for damaged houses, 3 months- –– Suppliers and ben-
340 USD for destroyed houses eficiaries identified
through cash vouchers

1 month- –– Project start

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.

Strengths and weaknesses


99 The cash voucher approach ensured that to agree fixed prices with the suppliers and guarantee
beneficiaries played a bigger role in their own quality.
recovery. 88 Initial attempts to restrict which materials could
99 According to a project evaluation people assisted be used failed due to supply shortages following a
felt that orientation and information sessions enabled government ban on harvesting timber.
them to understand what they were entitled to 88 Some dishonest suppliers could cheat beneficiaries
receive. of some items and claim them in invoices. Financial
99 Recommending several hardware stores allowed controls aiming to prevent this required a very large
people to shop around, but also allowed them to amount of documentation and massively increased the
choose the most convenient stores. workload for project and finance staff.
99 Vouchers allowed people to identify and prioritise 88 A minority of beneficiaries colluded with suppliers
their own needs. and used their cash vouchers for other unintended
99 The value of the vouchers was sufficient to meet purposes. In part this was due to shelter not being
the immediate shelter needs. However many people seen by all of them as the highest priority.
added their own resources to repair their houses. 88 Not all households adopted improved typhoon-
99 The majority of people supported by the project resilient construction techniques. The project could
preferred vouchers to direct cash. Their main reason have better promoted and trained in safer construction
was that vouchers enabled them to avoid spending techniques.
cash on other needs. It also allowed the organisation

www.ShelterCaseStudies.org 83
A.26   Natural disaster

Families are not given a pre-


defined list of materials. Instead,
the organisation conducted price
surveys and recommended several
shops from which beneficiaries
could obtain shelter materials.
Families repaired or rebuilt
shelters through bayanihan. This
is a tradition common in Philip-
pine rural areas, where community
members help each other. Through
bayanihan, those households who
Vouchers were provided that could be used to purchase materials up to a given are physically unable to build [older
cash value. people, people with disabilities,
Photo: Hajime Matsunaga/IFRC households headed by women and
households headed by children]
Before the typhoon voucher worth 7,000 PHP (150 are supported by their fellow
The Philippines has a history USD) to obtain the same materials community members.
of storms. In late 2009 Typhoons and tools as in Category I shelter The period during which
Ketsana and Parma caused con- repair kits. Under this category vouchers could be redeemed was
siderable damage. Three of the the families would also receive limited to a fixed period. This
districts hit in 2009 were also hit by the following materials to enable amount of time depended upon the
typhoon Megi in 2010. them to place poles in reinforced capacity of the shops and number
concrete footings: of beneficiaries per shop. Selected
After the typhoon
Typhoon Megi caused signifi- • three bags of cement, shops were required to display fixed
cant damage to houses, livelihoods • six timber posts - 6”x6” prices of main shelter materials
and infrastructure. The damage (150x150mm) or 4”x4” throughout the time.
was mainly due to the powerful (100x100mm), Each voucher could only be
category 5 winds when the typhoon • eight x 6m, 10mm diameter redeemed in one shop. However,
made landfall. The damage was steel bars, beneficiaries of Category II shelter
largely focused on five provinces. • four x 6m, 8mm diameter bars. repair kits received two vouchers of
Two weeks after Typhoon Revised implementation USD 150 and were able to redeem
Megi, heavy rains caused further In February 2011 a government each voucher at separate shops.
damage. The typhoon and the ban on harvesting timber was es-
tablished. This lead to a new meth- Selection of beneficiaries
rains combined further stretched As relief operations progressed,
community coping capacities. odology being established. In this
approach, people were provided the organisation reverified the ben-
Implementation with cash vouchers, which they eficiary lists. Details were initially
The shelter interventions had then use to purchase their choice of provided in lists by the government.
two components: shelter materials. During reverification, the sites of
all damaged or destroyed homes
• Category I - shelter repair kits
for families whose homes were
damaged.
• Category II - shelter repair kits
for families whose homes were
destroyed.
Initial plan
For Category I shelter repair
kits, families were provided 7,000
PHP (150 USD) .They could collect
any combination of materials and
tools in a predetermined list from a
shop of their choosing, as long as
the total cost did not exceed the
allocated amount.
For Category II shelter repair kits,
each beneficiary family would also Families rebuilt the shelters through community self-help.
receive an additional commodity Photo: IFRC

84
Natural disaster Shelter Projects 2010 A.26

The organisation monitored the shops.


Photo: Hajime Matsunaga/IFRC

A typical house rebuilt using the grants.


Photo: IFRC

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
A.27   Natural disaster

A.27 Romania - 2010 - Floods


Case study:

Country: Project timeline


Romania
12 months - –– Project completion
Disaster:
Heavy rain and flooding in 24 11 months - –– Finished construc-
counties of Romania tion of 30 houses
Disaster date:
8 months - –– Finished the renova-
June, July 2010 tion of 50 houses
No. of houses damaged: –– Finished the
Over 800 houses completely construction of the
destroyed school
6 months - –– Second resource
More than 7,000 houses centre opened
damaged
No. of people displaced:
15,000 2½ –– Construction starts
months- –– First resource centre
Project target population: opened
400 households and 3 school
Occupancy rate on handover: 2 months - –– Project team formed
and communities
100% identified, family
Shelter size: selection done
60 m2
1 month - –– Partnership with the
Materials Cost per house: Romania government signed
Average 2,500 USD
Project cost per house: 2 weeks - –– Project start
Average 3,000 USD –– First impact and
damage assessment
visits

June 2010 –– Heavy rainfall

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.

Strengths and weaknesses


99 Beneficiaries contributed their time and labour 88 Government delays in delivering on promises for
towards construction of their new homes. materials.
99 Business partners were fast in reacting and 88 Logistics not adapted for acting in multiple locations
providing much needed materials and financial (10 different locations at the same time).
resources. 88 Difficulty in engaging the beneficiaries selected for
99 Partners mobilised their workers to volunteer at relocation.
the construction site. 88 Delay in developing the infrastructure for the plot
99 The municipality was involved from the start which of land selected for relocation of the new homes.
helped the project to proceed quickly. 88 The weak involvement of local volunteers interested
99 The government endorsement of the campaign in renovation activities.
helped generate needed resources and partnerships. 88 Severe weather conditions in winter and spring
99 Assistance was offered not only in reconstructing (-10C in April) delayed construction and prevented
homes, but also in supplying and distributing more local volunteers from taking part in the project
construction materials through the resource centres. as planned.
This allowed families whose houses were not severely 88 Lack of sufficient local capacity in terms of skilled
affected to quickly renovate them with their own people available for running more resource centres.
resources.

86
Natural disaster Shelter Projects 2010 A.27

Floods destroyed 800 over homes.


Photos: Ovidiu Micsik and Mihai Grigorean, Habitat for Humanity Romania

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
A.27   Natural disaster

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
Conflict / Complex Shelter Projects 2010 A.28

A.28 Sri Lanka - 2009 - Conflict Returns


Case study:

Country: Project timeline


Sri Lanka
Conflict:
Population movement due to
civil war
Conflict date: 48 months - –– ANTICIPATED
Conflict ended in May 2009 Project completion
No. of houses damaged:
160,000
No. of people affected /
displaced:
300,000
Project target population:
2,511 families
(Estimated 12,555 people) –– Project starts
11 months -
669 houses built by
January 2012 India
Shelter size:
7 months - –– Returns begin
46m2
Materials Cost per house:
4,600 USD
Additional costs per house:
3,100 USD May 2009 - –– End of conflict

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.

Strengths and weaknesses


99 The owner-driven approach with direct cash grants another. As a result experienced staff were required to
to the beneficiary bank accounts created a sense of ensure that the project was effective.
ownership. 88 Because the project started with beneficiary
99 The project was able to build upon experiences and lists from the local authorities, it was open to some
use staff from the post-tsunami recovery programme. politicisation.
99 The shelter projects were part of a larger 88 Each house required a minimum of eight documents
programme including support in water, sanitation, and five separate transactions. This led to delays when
livelihood, disaster risk reduction, community combined with the processing of funds requests and
infrastructure and efforts to rebuild civil society. bank transfers.
99 The project aimed to empower civil society, 88 The project aimed for high quality but as a result
and provoke them to take initiatives on behalf of is relatively small scale, aiming to meet 1.5% of the
themselves. housing needs. In total all organisations together aim
88 The project was delayed by lack of legal documents. to meet a total of 20% of the overall housing need.
88 Often beneficiaries could not produce evidence -- The organisation was able to establish good
of land ownership, such as land permits, birth/death/ relationships with the government and military allowing
marriage certificates etc. as they were lost during the improved access to difficult to work in areas. However
conflict. given the context this required significant efforts to be
88 Construction skills differ from one family to seen to remain impartial.

www.ShelterCaseStudies.org 89
A.28   Conflict / Complex

Houses were rebuilt using cash grants.


Photo: Silvester Kueenseger, IFRC

Background Of those houses that needed with a list of beneficiaries by the


Conflict between forces of the to be reconstructed, 74% needed local authorities (a list from the
government of Sri Lanka and Lib- to be constructed anew, and the District Secretariat, approved by the
eration Tigers of Tamil Eelam (LTTE) rest required repairs. Considering Government Agent), and given an
began in 1983 and continued until the size, complexity, and evolving opportunity to verify beneficiaries
2009. nature of the situation, it had and communities.
been a challenge to gain access to
Between 2006 and February The final selection was done
isolated or inaccessible areas and to
2009, over 281,000 people became by the by the organisation after
assess the needs of the most-affect-
internally displaced. This was in conducting interviews. Each ben-
ed people.
addition to over 214,000 people eficiary in the given list provided
who had been displaced before Implementation the following documents at the
2006, meaning that over half a The national organisation with interview:
million people had been displaced support from its international coun- • family details,
by the conflict. terparts received approval from the • copy of the National Identity
The total population that lived government to assist 2,511 house- Card,
in the Northern Province of Sri holds. • copy of the bank pass book/
Lanka prior to May 2009 is yet to Of these, 2,181 are new build bank details,
be assessed. houses for fully damaged houses, • copy of the deeds,
and 330 are repairs for the partially • consent letter by land owner
The conflict left hundreds of
damaged houses. 669 were approved by the assistant
thousands of people in transit,
completed by the end of 2011 with government agent if the land is
displaced, and seeking refuge with
1,294 ongoing. not owned by beneficiary,
host families or in government-
• plan of the site.
run camps or centres for inter- Selection of beneficiaries
nally displaced people. Two entire After the selection of the
The organisation was provided families, each community was given
districts were fully deserted and
three other districts had partial
displacements as a result of the
conflict.
Hundreds of thousands of
people lost almost everything and
suffered without shelter, water,
sanitation, health care, livelihoods
or other basic facilities.
The government and humani-
tarian actors estimated that ap-
proximately 160,000 houses in the
north of Sri Lanka were in need of
reconstruction. This figure excludes
more than 100,000 families who
were scheduled for return from Rebuilt house in Northern Sri Lanka.
Photo: Silvester Kueenseger, IFRC
India and other countries.

90
Conflict / Complex Shelter Projects 2010 A.28

a chance to object when the lists


were publicly displayed.
Noting that families are re-build-
ing their houses on their own land,
a site investigation was carried out
following the beneficiary selection.
The site audit was done by a
project technical officer and a field
engineer to satisfy that the house
had stood on the site before the Grants were phased, and each grant was conditional upon the previous
war and was completely damaged level of construction being attained.
Photo: Ganga Kariyawasam, IFRC
and to ensure suitability for recon-
struction. • a baseline survey form, The standard house under the
• documents such as affidavits, project is a permanent house of
Beneficiary files certificates etc. in the absence minimum size of 500ft2 (46m2) floor
A beneficiary file was built of required documents, area with two rooms. The following
up for each beneficiary with the • any other documents relevant elements are required for the house
file cover clearly marked with the to the project. to be considered as complete:
project name and number, benefi-
ciary name and beneficiary address. Housing construction • The house is bounded by brick
The beneficiary file consists of : A cash grant amount of 2,900 or block wall and covered by a
USD (LKR 325,000) was paid to tiled roof.
• a family details form, each beneficiary through the bank • There is one internal lockable
• a copy of the beneficiary's in five instalments. An additional room with internal plastering,
National Identity Card, grant of 267 USD (LKR 30,000) was flooring, doors & windows and
• a copy of the beneficiary Bank provided to construct the toilet and sashes.
pass book indicating name and for the water supply. • There is one internal or external
account number, kitchen with internal plastering
• copies of documents indicating Each stage was to be completed
and flooring.
ownership of land (copy of the by the beneficiary within three
• There is one internal or external
deed or consent letter by the weeks of receiving the instalment.
toilet with adequate effluent
land owner approved by the On completion of each stage,
disposal.
assistant government agents the next grant should have been
received within one week. Each beneficiary may adjust the
if the land is not own by the
standard design to suit their indi-
beneficiary), The technical advisors and vidual needs, within the param-
• a plan of the site, support is given by the technical eters of the budget and minimum
• the site audit report conducted officers and the field engineers, standards.
by the organisation, who supervised 100 families and
• a copy of the house plan, 500 families respectively.

Housing construction with each instalment


Instalment Amount Work to be completed Technical details
1 440 USD Foundation (house and toilet) Laying of 3” screed concrete, rubble masonry foundation in 1:5
- within 4 weeks cement mortar along with damp proof course plastering and
applying of damp proof course tar.
2 790 USD Construction of super structure (house and The brick or block work of the walls up to roof level including the
toilet*) with brick or block walls up to roof level lintel tie beam above the window level using two 10mm diameter
- within 6 weeks tor steel bars in 1:3:4 concrete mixture and toilet super structure.
3 790 USD Roof (house and toilet*) House - roof work using clay roof tiles
- within 4 weeks Toilet - roof work using clay roof tiles or concrete slab
4 615 USD Internal plastering and floor concreting of one The internal plastering of one bed room and kitchen in 1:5 cement
bed room, kitchen and toilet. Fixing of all door sand mortar along with fixing of all door and window frames. Also
and window frames must concrete the floor areas of one bedroom and kitchen with
- within 6 weeks 1:3:4 concrete mixture.
5 220 USD Flooring of one bedroom, kitchen and toilet. The fixing of door and window sashes in one bedroom. Each
Complete one lockable room (fixing of doors and beneficiary must have at least one lockable room. Also lockable
windows and sashes in one bedroom) room floor and kitchen floor needs to be plastered and finished
- within 2 weeks with cement.
Total 2,900 USD Time taken to complete the construction work
is 20 weeks. Additional time required for funds
request and transfer.
*Water supply can be provided instead of a toilet

www.ShelterCaseStudies.org 91
A.29   Natural disaster

A.29 Tajikistan - 2010 - Earthquake


Case study:

Country: Project timeline


Tajikistan 26 months - –– Phase 2 - survey of
Disaster: construction prac-
Earthquake tices,120 houses
retrofitted
Disaster date:
July 29th 2006
No. of houses destroyed:
1,000 destroyed 10 months - –– Phase 2 - Survey of
1,500 damaged construction prac-
No. of people affected: tices, 120 houses
16,000 people lost their homes started
Project target population:
83 new build houses
120 houses retrofitted 7 months- –– Phase 1– 83 new
earthquake-resistant
Occupancy rate on handover houses finished
100%
Shelter size:
New house - 45 m2
5 months - –– Project start
Reinforcement of two rooms in
the house – 32 m2
Materials Cost per house:
New houses – USD 6,405
3 months - –– Phase 1– 83 new
Reinforcement of two rooms – earthquake-resistant
USD 895 Tajikistan houses started
Loans average 800 USD
Project cost per house:
July 29th
New house – USD 7,945 2006 - –– Earthquake
Reinforcement of two rooms –
USD 984

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.

Strengths and weaknesses


99 The project used local materials to reduce costs 88 Affected villages were located far from each other
and avoid delays in transportation. leading to logistical challenges.
99 The low technology reinforcement and 88 The technology was viewed positively by the local
construction technology was simple and was 30 population. However, many families who were not
percent cheaper than the standard reinforcement affected by earthquakes soon forgot or dismissed the
techniques. risks, as they had other priorities and limited resources.
99 The project promoted owner-driven home 88 Structural elements were more easily applied in
reconstruction and involved the whole community in new construction than in retrofitting after the disaster.
the procurement and construction process. Reinforcement involved the destruction of wall finishes.
99 Families were offered loans that they could repay 88 Loans require that families have sufficient income
into a revolving fund. This was used to finance later to be able to repay them. The poorest people may
projects. therefore be excluded.
88 The project met the housing needs of 10% of -- This project involved a number of partners.
the affected population. However, there was limited
response by other organisations.

92
Natural disaster Shelter Projects 2010 A.29

Before the earthquake


Tajikistan lies in a mountainous
region. Annually, it experiences over
5,000 tremors and earthquakes. In
total, from 1980 to 2010, more
than 6 million people in Tajikistan
suffered from the effects of natural
disasters.
The main causes of destruc-
tion during earthquakes are poorly
constructed homes which lack any
earthquake resistant, seismic tech-
nology. This is also coupled with
lack of knowledge, and limited
financial resources to reinforce or
even maintain homes.
Around 75% of the population
in Tajikistan lives in remote rural
areas in the mountains. These are
the families that are most vulnera-
ble to disasters. In addition to that,
almost half of the population lives The project rebuilt about 10% of the damaged and destroyed houses using food
on less than 2 dollars a day. for work, loans and materials distribution.
Photo: Habitat for Humanity Tajikistan
After the earthquake
On 29 July 2006, two earth- development, “food for work” to the project ben-
quakes in separate locations hit Ta- • training, eficiaries. Families received the food
jikistan. They fully damaged 1,000 • construction / retrofitting. based on the amount of working
houses and partially damaged Starting from May 2007 con- days they invested in the construc-
1,500 houses. There were concerns struction team members and an tion of their houses (wheat flour 2
for the winter ahead. expert at the national seismic kg/person/working day, vegetable
institute conducted a joint survey of oil 75g/person/working day, salt
1st Phase: 83 new houses newly constructed houses in three 25g/person/working day, pulses
In the initial phase, the organi- 200g/person/working day.)
target communities of Kumsangir.
sation, in partnership with interna-
tional non-government organisa- Based on the results of the The families were selected by the
tions, the government of Tajikistan survey, the design, technical inter- Committee of Emergency Situation
and the affected communities, vention and the scale of the project of Tajikistan, a United Nations
supported 83 families to build new were identified. The institute agency and the organisation itself.
earthquake-resistant houses. initiated technical research on The criteria for family selection
the effectiveness of using local were based on three pillars:
Beneficiary families were
materials to reinforce mud walls.
involved in the whole process of • housing need,
planning and building. They were Training workshops on house re- • ability to repay,
trained in construction skills, earth- inforcement using locally available • willingness to partner and
quake resistant construction and materials were conducted in target contribute hours of labour.
house maintenance. The families communities. During the workshop
The organisation aimed to
provided all necessary unskilled participants learned about the
ensure that it serves the most vul-
labour. methods and practiced by reinforc-
nerable first. However, as the or-
ing and retrofitting one house.
2nd Phase: 120 houses ganisation encouraged self-driven
reinforced Both women and youths were reconstruction and ownership of
In the second phase of the actively involved in the production homes, families were required to
project, the organisation worked and mounting of grids made from have a minimal stable income and
with the UN to support the retro- mulberry branches to reinforce the be willing to work on the construc-
fitting of 120 homes in earthquake houses. A permanent project team, tion or reconstruction of their own
affected districts. made up of engineers and supervi- house and, if needed, neighbouring
sors assisted the families by giving homes.
This phase of the project was them training and assistance.
implemented in three stages:
An external organisation
• research and technology provided support by providing

www.ShelterCaseStudies.org 93
A.29   Natural disaster

“Now my house is solid


and I am not in fear
of earthquakes as I
was in the past. It has
reinforcement frames
made of wood and
mulberry branches and
I am sure that it can
protect my family from
earthquakes in the future”

Shamsov Sharofidin, 58-year-old head of


the family Houses were retrofitted with a mesh of mulberry branches to increase seismic
resistance developed by the Tajikistan institute for seismology.
Photo: Habitat for Humanity Tajikistan

Technical solutions the wooden structure of the wall. community.


A mesh of mulberry branches This approach of adding a grid One year after the project com-
was used to make rural homes of branches could also be built into pletion, the organisation returned
more seismically resistant. This was the construction of new homes to support 206 families with non-
a new technology to the area. The or retrofitted into existing homes, profit loans to retrofit their houses
aim of this was to reduce the risk specifically those already suffering to improve their resistance to earth-
to life due to building collapse as from some earthquake damage. quakes.
well as to reduce the chances of re-
inforced houses being damaged in Loans Achievements
small or middle magnitude tremors. The organisation offered loans Perhaps the most important
The external and internal averaging 800 USD. These were achievement of the project was
damaged walls were channelled designed to cover the construction that at the end of it, families and
out to take the structural timber and reinforcing costs. Homeowners organised community groups
frame, to where the floor and walls could repay their loans with 18 USD developed vital skills of safer house
joined and the walls and ceiling monthly instalments over a period construction. These skills could be
joined. Filling this, mulberry twigs of 3 years. These repayments went then applied without additional as-
woven into grids, were attached into a revolving fund. Later, this sistance, as well as transferred to
to the timber framing, and then fund was used to offer more loans other communities.
plastered with an adobe render/ to families for retrofitting houses in
plaster, which was fortified with the disaster prone areas. Families
straw and wool. paid monthly instalments at the
nearest bank in their respective
Mulberry branches were tested Materials list
for sufficient strength as required
Materials Quantity
by Tajik national building codes
Timber 16
and norms. According to the data (150mm x 100mm x 2.7m)
from the Institute of Seismology in Timber 8
Tajikistan, if the walls are properly (50mm x 100mm x 3.2m)
plastered, mulberry branches Timber 4
could serve a minimum of 25 years (50mm x 100mm x 5.2m)
without losing their strength and Mulberry mesh 150mm x 96m2
150mm
without getting brittle.
Timber 4
The approach was cheap to (50mm x 150mm x 3.2m)
build as mulberry branches are free Timber 4
to people living in rural Tajikistan (50mm x 150mm x 5.2m)
and the total reinforcement process Timber (for doors/windows) 16
(100mm x 50mm x 3.2m)
costs 33 percent less than the steel
Timber (for doors/windows) 4
reinforcement bar (rebar) alterna- (100mm x 50mm x4.1m)
tive. Binding wire 4kg
To make the grid, branches are Tightening wire 3kg
bound to each other with steel wire. Wire 10kg
The wire is also used to fasten the A wall being retrofitted. Total timber 1.5m3
mesh element to the walls and to Photo: Habitat for Humanity Tajikistan

94
Natural disaster Shelter Projects 2010 A.30

A.30 Tonga - 2010 - Tsunami


Case study:

Country: Project timeline


Tonga
Disaster:
Tsunami 15 months - –– Project completion
(known as the Samoa Tsunami)
–– Construction com-
Disaster date:
pleted and gov-
September 30th 2009 ernment building
No. of houses damaged: certification received
79 destroyed, 30 with major
10 months - –– Construction of
damage
shelters and latrines
No. of people affected / on Niuatoputapu
displaced: begins
9 months - –– Shelter pre-fabri-
465
cation completed,
Project target population: prototype latrine
74 households built.
8 months - –– Construction of
Occupancy rate on handover:
footings on Ni-
Estimated 90% at handover uatoputapu begins
Shelter size: 7 months - –– Shelter pre-fabricat-
18m2 ed design tendered
Niuatoputapu and contract
2.4 m tall
awarded
Materials Cost per shelter: 6 months - –– Shelter prototypes
4,350 USD constructed on
Niuatoputapu
Project cost per shelter:
8,900 USD
5 months - –– Community assess-
ments
–– Project start

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.

Strengths and weaknesses


99 The project successfully addressed the significant 88 Community consultation could have been stronger
needs of a remote population. For the first months at critical points of the process of shelter design and
after the disaster no other non-government latrine construction.
organisation worked on Niuatoputapu. 88 Construction of latrines was not completed by
99 Interviews with beneficiaries as part of a project a number of households in one village. Follow up
evaluation indicated the shelter had met, and in many was required to understand the issues and ensure
cases exceeded, their expectations. completion where feasible.
99 Excellent logistical organisation with the support -- Initial assistance was requested in water supply only
of a well-established local implementing partner but the deployment of an engineer quickly identified
helped to keep the project on time. other needs including shelter.
99 Rainwater harvesting was included with the -- As this project had a low number of beneficiaries,
shelters to supplement drinking water sources. and high costs, the project team could not be large. It
88 A formal handover of the shelters to beneficiaries was difficult to provide a range of skills with the limited
did not take place during the project leading to some number of personnel.
uncertainty about ownership.

www.ShelterCaseStudies.org 95
A.30   Natural disaster

“The house is important


to me, especially the
water tank as this is my
only source of drinking
water. With this house I
can manage ok.”

Maka Holi
Project beneficiary

Over half of the people on the island lost their houses.


Photo: Kathleen Walsh

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
Natural disaster Shelter Projects 2010 A.30

“If the transitional


house hadn’t been
given then we would
still be in the tent
and the small shack.
Without the house we
wouldn’t have water
and would have to find
it from somewhere
else.”

Neomai Osika
Project beneficiary

The project illustrated the challenges of running small projects on remote islands
with a small project team.
Photo: Paul Davenport

A government building assessor Technical solutions to be of standard according to


resident on the island provided The technical and resource Tongan building regulations by the
both interim and final certification capacity of Niuatoputapu is very government building inspector on
for the buildings based on govern- limited. Therefore it was decided Niuatoputapu.
ment standards for cyclone resistant to fabricate transitional shelter kits
shelter. The design of the shelters
in the capital. These could then
ensured they were simple enough
be flat packed and shipped to the
Selection of beneficiaries to be built in a remote location
island. The erection of the prefab-
Beneficiaries were those families and that they could be dismantled
ricated elements was undertaken
identified by the local implementing and re-built as the government
by local trades people supported
partner in an initial damage assess- was offering land further from the
by the householders where appro-
ment. To qualify for a shelter, their coast to encourage people to move
priate.
home and assets had to have been for their future safety.
completely destroyed or lost in the Shelters were designed to be
tsunami. cyclone resistant and were certified

The project built shelters using contracted teams from Niuatoputapu, 600 km away.
Photo: Kathleen Walsh

www.ShelterCaseStudies.org 97
A.31   Natural disaster

A.31 Vietnam - 2009 - Typhoons Ketsana and Mirinae


Case study:

Country: Project timeline


Socialist Republic of Vietnam
15 months - –– Internal review
Disaster:
Typhoon Ketsana and Typhoon
Mirinae
13 months - –– 650 houses
Disaster date: completed
September 29th 2009 (Ketsana)
November 2nd 2009 (Mirinae)
No. of houses destroyed:
10 months - –– Beginning of
23,500 construction
No. of people evacuated:
356,790 people evacuated 9 months - –– Shelter advisor
present (4 months)
Project target population: –– Trainings, selection
Around 2,730 people (650 of beneficiaries and
households) in seven provinces house design
Occupancy rate on handover:
100% (estimate)
Shelter size:
26 m2 average 4 months - –– Project start
–– Shelter needs
Materials Cost per shelter: assessment
1,650 USD cash grant
1,300 USD average spend on 3 months - –– Household items
material only distributed to
Vietnam 60,286 people

2 months - –– Typhoon Mirinae

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.

Strengths and weaknesses of ownership of their own homes. Some members of


99 Houses were built according to traditional design ethnic minority groups expressed their appreciation for
with necessary reinforcement. Daily construction their houses being reinforced.
work was closely supervised by local engineers. 88 The organisation was slow to start the project. In part
99 Families decided on the house design and were this was due to not getting the right people in place in
able to adjust the home according to their individual time to start recovery planning.
needs. 88 Water and sanitation (both hardware and software
99 Many families made additional contributions as components) should have been included in the shelter
they considered it a lifetime investment. programme as part of the house package.
99 The conditional cash grant enabled families to 88 The houses were not all culturally acceptable to ethnic
select local suppliers and builders whom they trusted, minorities. More detailed needs assessments should have
while benefitting from technical advice. been conducted.
99 Technical training helped families to follow each 88 More attention should have been given to the
step of the construction work while being supported disparities between provinces regarding the availability of
by project engineers. local labour and prices for material and transport.
99 A participatory approach helped to provide a sense

98
Natural disaster Shelter Projects 2010 A.31

Households were allowed to choose from certain given


Training was conducted on safe construction techniques. designs and encouraged to adapt them to meet their needs.
Photo: DWF Photo: DWF

Before the typhoon After the typhoon 1. The organisation conducted


The Socialist Republic of Vietnam Houses were destroyed because field surveys to assess needs
is a single-party state. The Govern- they were in vulnerable locations, and local conditions for
ment at local level is represented by were poorly constructed, materials construction, paying special
the People’s Committee, in every were used poorly and lacked rein- attention to ethnic minority
province, district and commune. forcement. Houses were destroyed needs and customs.
both by the winds and by flooding. 2. Based on information gained,
Vietnam had been rapidly in-
The poor quality of construction house designs were prepared
dustrialising and there had been
was compounded by a lack of in line with Vietnamese
a significant improvement in
financial resources and awareness. national and local government
people’s living standards. However
standards, taking into account
there remained wide disparities in For the response the organisa-
culture, geography and
income and living standards across tion provided support with food,
exposure to hazards. Three
the country. The seven provinces safe water and support for liveli-
standard house designs were
covered by this shelter project are hoods. It also distributed basic
developed for each province,
among these poorest provinces of household items to 60,286 people
and later adapted for each
Vietnam. within the first three months.
household beneficiary.
Vietnam has a tropical climate Implementation 3. The organisation approved
with a hot summer and colder The project started with trainings final beneficiary lists and cross-
winter (especially in the north). The in each province to cover the spe- checked information. Working
storm / typhoon season mainly takes cificities of the shelter programme, with the partner organisation,
place from August to November. beneficiary selection criteria, cash each family was consulted on
grant distribution process and the design, family contributions,
Houses are mostly based upon
related guidelines. The trainings availability of materials and
traditional styles, but using different
were targeted at members of the skilled local labour.
materials (brick, cement blocks,
organisation, People’s Committee 4. Trainings were conducted on
concrete, corrugated Iron sheet)
(representatives of the Vietnamese safe construction techniques.
instead of wood and clay tiles used
government) representatives from These targeted local builders,
in the past.
the province, district and commune project staff and beneficiaries.
When Typhoon Ketsana struck levels. 5. Construction then began.
the central and highland areas of Beneficiaries received the
Vietnam at the end of September This training was followed first allocation of the cash
2009, the government evacuated by community meetings in each grants following the laying of
over 100,000 households. commune to select beneficiaries foundations by local builders.
following agreed criteria. Grants were paid in cash, as
Five weeks later Typhoon
An international partner organi- are all other transactions at this
Mirinae hit central Vietnam, causing
sation was identified to provide level in Vietnam. Payment was
floods that swept away nearly
technical support and oversight. also made to material suppliers
2,400 houses, and hitting the same
The houses were constructed at this time. The organisation
people who were recovering from
according to the following process: and its partner monitored all
Ketsana.
stages of construction.

www.ShelterCaseStudies.org 99
  Conflict / Complex

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

Families were given cash to build houses according to given designs.


Photo: DWF

100
Conflict / Complex Shelter Projects 2010

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
  Conflict / Complex

102
Historical Shelter Projects 2010 B

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
B   Historical

104
Historical Shelter Projects 2010 B.1

B.1 Sphere Project - 2011 - Global standards


Case study:

Country: Project timeline


Global
Disaster: 2011 - –– Third edition of
Sphere hand book covers both sphere handbook
conflict and natural disaster

2004 - –– Second edition of


sphere handbook

2000 - –– First published


edition of sphere
handbook

Early 1998- –– Working group


established to draft
the Humanitarian
Charter

1997- –– Sphere Project


launched

1994- –– Rwanda refugee


crisis

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
99 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.
99 The shelter and settlements chapter of the 2010
handbook is based on consensus surrounding
shelter provision. It was developed during a broad
consultation process.
99 The document can be very useful starting point
for establishing advocacy positions with donors,
governments and within organisations.
99 When used well, it can lead to better programming,
and as a basis for training in humanitarian shelter
provision.
88 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
B.1   Historical

Sphere project formation


During the 1990s, humanitarian
response was going through many
changes. New actors including the
military and development agencies
were actively engaging in humani-
tarian operations.
Following the 1994 Rwanda
genocide where 800,000 people
were killed, a multi-donor evalua-
tion concluded that if humanitar-
ian agencies had done a better job,
then more lives would have been
saved.
This led to a group of humanitar-
ian non-governmental organisations
and the Red Cross / Red Crescent
Movement establishing the Sphere
Project in 1997.
At the core of the Sphere
Handbook is a summary of common
principles and values, known as the
humanitarian charter, as well as life
saving universal minimum standards
in key sectors.
Sphere is based on the humani-
tarian imperative - that there is the
right to give and receive humani-
tarian assistance wherever needed
to prevent and alleviate suffering,
protect life and ensure respect for
all human beings without discrimi-
Structure of the shelter, settlement and non-food items chapter
nation.
Sphere has not been adopted by been attained. Planning”. It contained six sections
all organisations. However most of • Guidance notes are points to including one on household items.
the major organisations working in consider when applying the
In the second and third editions
shelter do aim to adhere to it. minimum standards, key actions
of the handbook, the shelter
and key indicators in different
Standards, chapter was renamed the “shelter,
situations.
settlement and non-food items”
key actions, In addition to the sector specific chapter. This chapter now contains
key indicators, chapters, the Sphere handbook has five standards on shelter and settle-
and guidance notes core standards in: ment issues and five standards on
The “shelter” chapter in the • people-centred humanitarian non-food items and their distribu-
Sphere handbook now contains 10 response, tion (see illustration of the chapter
standards. Each standard has ac- • coordination and collaboration, structure above).
companying key actions, key indica- • assessment,
tors and guidance notes: 2011 edition
• analysis and design,
The shelter chapter of the 2011
• Minimum standards are • performance, transparency and
edition of the sphere handbook,
qualitative in nature and specify learning,
recognises some developments in
the minimum levels to be • aid worker performance.
humanitarian shelter, including:
attained in humanitarian the Shelter settlement and
provision of shelter. • The need to consider transitional
non-food items and longer-term recovery/
• Key actions are necessary Since the inception of sphere,
activities and inputs to be taken reconstruction issues during the
shelter has been seen one of the core
in order to meet the minimum initial or emergency response
areas of humanitarian response. The
standards. phase.
first edition of the sphere handbook
• Key indicators are ‘signals’ that • Changes in interagency
had a chapter on “Shelter and Site
show whether a standard has

106
Historical Shelter Projects 2010 B.1

coordination. and reprioritisation of themes. A and not recovery/early recovery. In


• The integration of risk and number of former key indicators the 2011 edition, enabling early
vulnerability reduction in shelter have now been incorporated into recovery is an explicit theme, with
programmes. the guidance notes. supporting guidance to match.
• The “embedding” of cross- The thematic issues incorporat- While Sphere is still primarily
cutting issues in relation to ed in the key actions and indicators intended for the humanitarian
individual responses. reflect developments in both the response phase, the guidance given
• Access to non-food items or sector and in overall humanitar- in the 2011 edition can inform
relief assistance is more than ian action. New content primarily preparation for and recovery from
distribution. There is now consists of new or expanded disaster. Without significantly
greater emphasis on the use of guidance on cross-cutting issues. affecting content, this orientation
cash, vouchers and access to has influenced the language and
local markets. In the 2004 edition, promotion
guidance on use of the content.
of livelihoods was less of a priority,
New in 2011 with the focus primarily on response
Much of the content of the 2004
edition has been revised, edited and
updated to reflect evolving practice
M inimum S t andar ds in S hel ter, S e t t lemen t and n on- f o od I tems
since 2004. Some of the text in the
2004 edition has been cut.
The sphere shelter chapter Shelter and settlement options and response scenarios
now contains a diagram outlining
shelter and settlement options, as PRE-DISASTER
they relate to non-displaced and SETTLED POPULATION
Temporary or permanent shelter
displaced populations.
Although in practical terms the Settlement typologies
shelter and non-food items needs
may be the same for both popula- House Apartment
tions after a disaster, the settlement owner–occupier owner–occupier
options are very different and will in
turn impact on the type of shelter
assistance to be provided. House tenant Apartment tenant

2011- changed standards


“Physical planning” has become Land tenant Occupancy with
“settlement planning” to reflect no legal status
its focus on space planning issues
rather than the strategic issues.
Disaster
The former “design” standard
has been merged within standards
on “covered area” and “construc- POST-DISASTER NON-
POST-DISASTER DISPLACED
DISPLACED POPULATION
tion”. POPULATION
Temporary or transitional shelter;
Temporary or transitional shelter
repair or reconstruction
A new “non-food items”
standard has been introduced to
provide an over arching standard SETTLEMENT TyPOLOGIES SETTLEMENT TyPOLOGIES
on ensuring access to relief items,
including the provision of cash House Apartment Dispersed self- Short-term
and vouchers and access to local owner–occupier owner–occupier settlement with land, house or
no legal status apartment tenant
markets.
The non-food items standard on House Apartment tenant Hosting by Collective
personal hygiene has been moved tenant families centres
to the “water supply, sanitation
and hygiene promotion” chapter. Land Occupancy with Self-settled, Planned and
tenant no legal status unplanned managed
2011- key actions and key camps camps
indicators
The use of key actions as well DURABLE SOLUTIONS
Reconstruction
as a limited number of key indica- Resettlement
tors has enabled a review, revision Reintegration
245

www.ShelterCaseStudies.org 107
B.1   Historical

The Sphere shelter, settlement and non-food items standards

Shelter and settlement standard 1: Strategic planning


Shelter and settlement strategies contribute to the security, safety, health and
well-being of both displaced and non-displaced affected populations and promote
recovery and reconstruction where possible.
Shelter and settlement standard 2: Settlement planning
The planning of return, host or temporary communal settlements enables the safe
and secure use of accommodation and essential services by the affected population.
Shelter and settlement standard 3: Covered living space
People have sufficient covered living space providing thermal comfort, fresh air and
protection from the climate ensuring their privacy, safety and health and enabling
essential household and livelihood activities to be undertaken
Shelter and settlement standard 4: Construction
Local safe building practices, materials, expertise and capacities are used where
appropriate, maximising the involvement of the affected population and local
livelihood opportunities.
Shelter and settlement standard 5: Environmental impact
Shelter and settlement solutions and the material sourcing and construction
techniques used minimise adverse impact on the local natural environment.
Non-food items standard 1: Individual, general household and shelter support
items
The affected population has sufficient individual, general household and shelter
support items to ensure their health, dignity, safety and well-being.
Non-food items standard 2: Clothing and bedding
The disaster-affected population has sufficient clothing, blankets and bedding to
ensure their personal comfort, dignity, health and well-being.
Non-food items standard 3: Cooking and eating utensils
The disaster-affected population has access to culturally appropriate items for
preparing and storing food, and for cooking, eating and drinking.
Non-food items standard 4: Stoves, fuel and lighting
The disaster-affected population has access to a safe, fuel-efficient stove and an
accessible supply of fuel or domestic energy, or to communal cooking facilities. Each
household also has access to appropriate means of providing sustainable artificial
lighting to ensure personal safety.
Non-food items standard 5: Tools and fixings
The affected population, when responsible for the construction or maintenance
of their shelter or for debris removal, has access to the necessary tools, fixings and
complementary training.

108
Historical Shelter Projects 2010 B.2

B.2 USA - San Francisco - 1906 - Earthquake


Case study:

Country: Project timeline


USA
Disaster:
San Francisco earthquake June 30th
and fire 1908 - –– Moving of
cottages complete
Disaster date:
April 18th 1908 June 2nd
No. of houses damaged: 1908 - –– Last camp closed
28,000 buildings and 500 city
blocks – one quarter of the city
of San Francisco. July, 1907 - –– Moving of cot-
No. of people homeless: tages begins.
225,000
Shelters built:
5,610 timber cottages Aug 1st
1,709 housing grants 1906 - –– Shelter item distri-
9,064 housing furniture grants bution ends
–– Camps moves
1573 loans from army to
civilian authorities
June 1st
San Francisco 1906 - –– 34,000 people
in tents, 5,000 in
shacks

April 20th - –– Timber for tent


floors, latrines and
shacks delivered
April 18th
1906 - –– Earthquake

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.

Strengths and weaknesses


99 Immediately after the earthquake, railway and caused more damage than the earthquake itself.
ferry companies provided free transport for those 88 Barracks were often dense and lead to crowded
wishing to temporarily leave the city to find shelter conditions with limited sanitation and privacy.
elsewhere. 88 A "Shoot to kill policy" established by mayor to deal
99 Commitees were established to identify host with looters. Many victims of this policy were salvaging
families outside of the city. materials from their own houses.
99 5,610 timber cottages were rapidly built and 88 Some minority groups such as chinese-americans
rented on a lease-to-buy deal. were evicted from land that they had before the
99 Multiple approaches to support recovery were earthquake and were moved from camp to camp.
established including cash and loans. -- Although significant, the number of schemes for
99 The military established a warehousing system for cottages, barracks, loans and grants were not on the
relief goods where elected civilian chairmen could scale of the number of houses damaged.
put in requests for their communities. Cash was given -- The quality of the response was dependent upon the
for those whose requests could not be met. decision making of individuals more than pre-exisitng
99 Tents were provided and were easy to move and systems. There are documented cases of both positive
provided adequate shelter until the rainy season. and negative behaviour by public officials.
88 There was a lack of preparedness and planning to -- All formal camps were closed within two years of
reduce disaster risks. A fire caused by the earthquake the earthquake.
www.ShelterCaseStudies.org 109
B.2   Historical

Before the earthquake


Before the earthquake, San
Francisco had a population of
450,000 people.
With the exception of public
buildings, and the houses of
wealthier city residents, buildings
were built from wood.
The city contained significant
minorities of immigrants from other
countries.

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
Historical Shelter Projects 2010 B.2

• in the official camps, Reconstruction


• in shacks and tents outside Reconstruction was relatively
official camps, swift, largely along the existing
• in the city who were living in grid plan of the city. Most of the
cellars or similar places, those reconstruction was completed by
who were receiving shelter from the Panama-Pacific Exposition, nine
Large numbers of people left the friends, years after the earthquake and fire.
city, free transport was provided.
Photo: The U.S. National Archives
• citizens living outside the city. However there were criticisms that
and Records Administration Some people whose houses the rebuilding was not to the seismic
survived the fires but needed better safety that it should have been.
by the army and by the American housing received cottages and moved
National Red Cross. The majority of the funds for
them to plots for permanent use. reconstruction came from private
Tents were seen as a practical capital and insurance payments.
shelter solution for the emergency
Charges for cottages
A nominal rental (2 USD per Grants and loans
phase, more effective than barracks.
month) was charged for the A committee administered
They were easy to move and
cottages. This was to avoid a culture grants and loans for reconstruc-
provided adequate shelter until the
of dependency and distortion of the tion. It tailored funding decisions to
rainy season.
economic conditions. the needs of individual applicants.
Camps Applicants were required to Grants were offered in various cat-
The army gradually assumed sign a lease agreement with the egories including tools, re-establish-
control of 21 camps. These camps “San Francisco Relief and Red ing houses, business enterprise and
were known as “permanent Cross Funds” (a corporation) before transportation.
camps.” occupying the cottages. This was a • Where applicants planned and
To live in a “permanent camp”, purchase contract that stipulated built their own houses, the
residents had to abide by rules of that the tenant would: committee set a maximum cost
decency, order, and cleanliness. • own the cottage if rent was of each house to be erected,
When a person was ejected paid until August 1, 1907, with the applicant paying the
from one camp all other camps • pay rent and gas rates, majority.
were notified so that he could not • abide by the camp regulations, • Where the committee planned
relocate. 488 people were ejected • not sublet without written and directed the construction of
from camps for reasons ranging permission, the house, the grant:
from drunkenness to disorderly • vacate the house at the • covered the entire cost of the
conduct. Discrimination forced expiration of his lease unless all house,
Chinese–Americans to be shuffled payments had been made, • supplemented the grant with
from camp to camp. • on acquiring ownership the a loan to be repaid by the
tenant would remove the house applicant,
Camps were generally located from the camp at his/her own • supplemented the grant with
in parks and squares. One camp expense before August, 1907. a cash payment from the
had nineteen two-story tenement applicant.
buildings and a one-story bath- When through ill-health a
person was not able to pay rent, 1,709 housing grants and 9,064
house and laundry building.
the Rehabilitation Committee was housing furniture grants were
Cottages / shacks informed. provided. An additional 450 housing
Three and a half months after relief grants were issued, averaging
708 cottages were purchased at 644 USD per household.
the earthquake, the city corporation on a lease-to-buy deal. The cottage
launched its plan to build timber could be bought for 50 USD. These Of 2,098 applications for the
cottages, and established contracts could be moved to plots where combined grant and loan plan, as-
for their construction. were rented by the authorities at sistance was given in 1,572 cases.
Building began six months after 3-15 USD per month. Loans ranged from 37 - 595 USD,
the earthquake. It took a further Whenever a person could prove
three months, before considerable that he had purchased or leased
numbers were available. By the end a lot in the city or county, he was
of the project, 5,610 cottages, 667 permitted at his own expense to
“patent flush closets”, over six miles move his house.
of gas and water pipe and over five
miles of sewer pipe were built. In just over two years the Burning of earthquake refugee
cottages were all removed and the shacks during “Nine Years After”
The cottages were assigned camps closed. celebration at the Panama-Pacific
according to the following priorities International Exposition.
- people who were: Photo: San Francisco Public Library

www.ShelterCaseStudies.org 111
B.2   Historical

“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

A distribution queue, early in the response.


Photo: Records of the United States Senate, National Archives

A military camp, four days after the earthquake.


Photo: Records of the United States Senate, National Archives

112
Annexes Shelter Projects 2009 Country Index

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
Country Index Annex 1 - Index - by country Annex 1 - Index - by country Annexes

Annex 1 - Index - by country

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

Afghanistan - 2002 onwards - Conflict Update Shelter Projects 2009


Afghanistan - 2002 onwards - Conflict Case Study Shelter Projects 2008
Afghanistan - 2009 - Conflict Case study Shelter Projects 2009
Algeria - 1980 - Earthquake From “Shelter After Disaster” Shelter Projects 2009
Azerbaijan - 1992 - Conflict Case study Shelter Projects 2008
Bangladesh - 2009 - Cyclone Aila Response Overview Shelter Projects 2009
Bangladesh - 2007 - Cyclone Sidr Response Overview Shelter Projects 2009
Bangladesh - 2007 - Cyclone Case study Shelter Projects 2009
Bangladesh - 1975 - Conflict Shelter Upgrades Shelter Projects 2008
Chile - 2010 - Earthquake Case Study Shelter Projects 2010
China, Sichuan - 2008 - Earthquake Case Study Shelter Projects 2009
D.R. Congo, Goma - 2002 - Volcano Case Study Shelter Projects 2008
Darfur - 2004 onwards - Conflict Case Study Shelter Projects 2008
DRC, Goma - 2009 - Conflict Case Study Shelter Projects 2009
Eritrea - 1998 onwards - Conflict Update Shelter Projects 2009
Eritrea - 1998 onwards- Conflict Case Study Shelter Projects 2008
Gaza, Palestine - 2009 - Conflict Case Study Shelter Projects 2009
Georgia - 2008 - Conflict Case Study Shelter Projects 2009
Grenada - 2010 - Hurricane Case Study Shelter Projects 2010
Guatemala - 1976 - Earthquake From Cuny Archive Shelter Projects 2008
Haiti - 2010 - Earthquake Response Overview Shelter Projects 2010
Haiti - 2010 - Earthquake 7 Case Studies Shelter Projects 2010
Haiti - 2008 - Flooding Case Study Shelter Projects 2009
Haiti - 1982 - Hurricane From Cuny Archive Shelter Projects 2009
Honduras - 1998 - Hurricane Mitch Case Study Shelter Projects 2008
Honduras - 1974 - Hurricane From “Shelter After Disaster” Shelter Projects 2009
India - 1977 - Cyclone From “Shelter After Disaster” Shelter Projects 2009
India - 1971 - Conflict From Cuny Archive Shelter Projects 2008
India, Andhra Pradesh -1977 - Cyclone - From Cuny Archive Shelter Projects 2008
India, Gujarat - 2001 - Earthquake Case Study Shelter Projects 2008
Indonesia - 2009 - Earthquake Response Overview Shelter Projects 2010
Indonesia - 2009 - Earthquake 3 Case Studies Shelter Projects 2010
Indonesia, Aceh - 2004 - Tsunami & Earthquake Case Study Shelter Projects 2008
Indonesia, Jogyakarta - 2006 - Earthquake - Response Overview Shelter Projects 2008
Indonesia, Jogyakarta - 2006 - Earthquake 2 Case Studies Shelter Projects 2008
Ingushetia - 1999 - Conflict Case Study Shelter Projects 2008
Italy - 2009 - Earthquake Response Overview Shelter Projects 2009
Italy - 2009 - Earthquake Case Study Shelter Projects 2009
Kenya - 2008 - Conflict Case Study Shelter Projects 2008
Kenya, Dadaab - 2009 - Conflict Update Shelter Projects 2009
Kenya, Dadaab - 2007 - Conflict / Flooding Case Study Shelter Projects 2008
Kyrgyzstan - 2010 - Conflict Case Study Shelter Projects 2010

114
Annexes Shelter Projects 2009 Country Index

Liberia - 2007 - Conflict Update Shelter Projects 2009


Liberia - 2007 - Conflict Case Study Shelter Projects 2008
Malawi - 2009 - Earthquake Case Study Shelter Projects 2010
Mozambique - 2007 - Cyclone Case Study Shelter Projects 2008
Myanmar - 2008 - Cyclone Case Study Shelter Projects 2009
Myanmar - 2008 - Cyclone 2 Case Studies Shelter Projects 2010
Nicaragua - 1973 - Earthquake From Cuny Archive Shelter Projects 2008
Nicaragua - 1972 - Earthquake From “Shelter After Disaster” Shelter Projects 2009
Pakistan - 2005 - Earthquake Overview of the response Shelter Projects 2008
Pakistan - 2005 - Earthquake 2 Case Studies Shelter Projects 2008
Pakistan - 2009 - Conflict Case Study Shelter Projects 2010
Pakistan - 2010 - Floods Overview Shelter Projects 2010
Pakistan - 2010 - Floods 3 Case Studies Shelter Projects 2010
Peru - 2007 - Earthquake Overview of the response Shelter Projects 2008
Peru - 2007 - Earthquake 3 Case Studies Shelter Projects 2008
Philippines - 2010 - Cyclone Case Study Shelter Projects 2010
Romania - 2010 - Floods Case Study Shelter Projects 2010
Rwanda - 2008 - Conflict returns Case Study Shelter Projects 2008
Rwanda - 2008 - Conflict returns Update Shelter Projects 2009
Somalia - 2008 - Conflict Overview Shelter Projects 2009
Somalia - 2007 - Conflict Resettlement Case Study Shelter Projects 2008
Somalia, Puntland - 2009 - Conflict Case Study Shelter Projects 2009
Somalia, Somaliland - 2009 - Conflict Case Study Shelter Projects 2009
Sri Lanka - 2007 - Conflict Update Shelter Projects 2009
Sri Lanka - 2007 - Conflict Case Study Shelter Projects 2008
Sri Lanka - 2009 - Conflict Case Study Shelter Projects 2010
Sri Lanka - 2004 - Tsunami Overview of the response Shelter Projects 2008
Sri Lanka - 2004 - Tsunami Case Study Shelter Projects 2008
Sudan - 1985 - Conflict From Cuny Archive Shelter Projects 2008
Sudan, Darfur - 2004 onwards - Conflict Case Study Shelter Projects 2009
Tajikistan - 2010 - Earthquake Case Study Shelter Projects 2010
Thailand - 1979-1980 - Conflict From Cuny Archive Shelter Projects 2008
Tonga - 2010 - Tsunami Case Study Shelter Projects 2010
Tonga - 1982 - Cyclone ‘Isaac’ From Cuny Archive Shelter Projects 2008
Turkey, Caldiran - 1976 - Earthquake From “Shelter After Disaster” Shelter Projects 2009
Turkey, Gediz - 1970 - Earthquake From “Shelter After Disaster” Shelter Projects 2009
Turkey, Lice - 1975 - Earthquake From “Shelter After Disaster” Shelter Projects 2009
Uganda - 2007 - Slow onset floods - Case Study Shelter Projects 2009
UK - 1945 - Post Conflict historical Case Study Shelter Projects 2009
USA, San Francisco - 1906 - Earthquake historical Case Study Shelter Projects 2010
Vietnam - 2009 - Typhoon Case Study Shelter Projects 2010
Yugoslavia (formerly) - 1963 - Earthquake From “Shelter After Disaster” Shelter Projects 2009

115
Country Index Annex 2 - Further reading Annex 2 - Further reading Annexes

Annex 2 - Further reading

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
Annexes Shelter Projects 2009 Further reading

IFRC, Oxfam GB, Plastic sheeting, 2007


Available from: www.plastic-sheeting.org
A guide to the use and specification of plastic sheeting in humanitarian relief, 2007. An illustrated booklet on
when and how to use plastic sheeting most effectively in emergencies.
IFRC, The IFRC shelter kit, 2010
Available from: www.shelterlibrary.org
A guide on the IFRC shelter kit and how to use it.
IFRC, Transitional Shelter: Eight Designs, 2011
A review of risks in shelter construction and detailed structural analysis of eight different transitional shelters
designs that have been used in the field in large scale projects.
Mike Albu, The Emergency Market Mapping and Analysis Toolkit,
Available from: http://emma-toolkit.org
A toolkit designed for generalists, as well as specialist staff on how to conduct an emergency market mapping
analysis.
NRC, Shelter Centre, Urban Shelter Guidelines
Available from: www.shelterlibrary.org
general guidance for urban humanitarian response.
Shelter Centre, UN, DfID, Shelter after disaster - Strategies for transitional settlement and reconstruction, 2010
Available from: www.shelterlibrary.org
A book containing information and guidance on how to agree strategies for reconstruction after natural disasters.
contains description of the types of shelter programmes that organisations can implement.
Sphere Project, Sphere - Humanitarian charter and minimum standards in humanitarian response, 2011
Available from: www.sphereproject.org
Contains consensus standards agreed among major humanitarian organisations for key sectors including shelter
and settlement. It also contains actions, indicators and guidance notes as to whether standards have been achieved.
Sultan Barakat, HPN Network paper 043, Housing reconstruction after conflict and disaster, ODI, 2003
Available from: www.odihpn.org/documents/networkpaper043.pdf
Review of housing reconstruction experiences and approaches.
UNDRO, (now UNOCHA), Davis, I., Shelter After Disaster, Guidelines for Assistance, 1982
Available from: www.shelterlibrary.org
Guidelines and description of shelter provision in all aspects of natural disasters (from preparedness to reconstruction).
UNHABITAT, IFRC, Shelter Projects 2009
Available from: www.ShelterCaseStudies.org
Case studies of shelter projects implemented between 1945 and 2009. Includes many different types of response.
UNHABITAT, IFRC, UNHCR, Shelter Projects 2008
Available from: www.ShelterCaseStudies.org
More case studies. Stee also UNHABITAT, IFRC, Shelter Projects 2009.
UNHABITAT, Land and Natural Disasters - Guidance for Practitioners, 2010
Available from: www.disasterassessment.org or from unhabitat.org
A book containing guidance on land issues following natural disasters.
UNHCR, Handbook for Emergencies, Third Edition, 2007
Available from: http://www.unhcr.org
A book containing guidance on the management and all the key sectors in refugeee emergencies.
UNICEF, Compendium of Temporary Learning Spaces (TLS): Design and practice in emergencies 2011
Designs and guidance on the construction of temporary learning spaces.
UN/OCHA, Tents - A guide to the use and logistics of tents in humanitarian relief, 2004
Available from: www.shelterlibrary.org
A booklet describing when and how to use tents as well how to support those living in them to best adapt them
to meet their needs.
UN/OCHA / IFRC / CARE International, Timber as a construction material in humanitarian operations, 2009
Available from: www.humanitariantimber.org
An illustrated booklet on how to source and use timber for the construction of basic structures.

117
Further reading Annex 2 - Further reading Annex 2 - Further reading Annexes

General statistics and case study related information

CRED, EM-DAT disaster database.


www.emdat.be.

Construir com os ventos: guiäo de construcäo para zonas de risco de ciclone, Government of Mozambique
UN Habitat.
Available from www.sheltercentre.org

IFRC, World Disasters Report, 2011 - Focus on hunger and malnutrition


IFRC, World Disasters Report, 2010 - Urban Risk
Available from www.ifrc.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

UNHCR, Global Trends 2010


Available from www.unhcr.org

118

You might also like