vernacular shelter projects 2009

8/3/2019 Vernacular Shelter Projects 2009

1/138

SHELTER PROJECTS

2009


2/138

Shelter Projects 2009

i

Introduction

Foreword

SHELTER PROJECTS

2009


3/138

Foreword

ii

Introduction


Published 2010

Available online from www.disasterassessment.org

Copyright for this booklet is retained by UN-HABITAT and IFRC.

Reproduction for non-profitable objectives is encouraged.

The copyright for the photographs and images remains with the photographers whose names are indicatedon each photograph.

The maps contained in this publication are for illustrative purposes only and should not be consideredauthoritative.

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.

Front Cover: A fishermens hut in Bacoor Cavite area south of Manila, the capital city of the Philippines, March 3,2008. Fish is the primary source of protein in the Filipino diet. Manoocher Deghati.

Back cover left to right: Afghanistan, 2009, Shaun Scales. Hargeisa, Somalia, 2009, Joseph Ashmore.Dadaab, Kenya, 2009, Jake Zarins. Uganda 2007, IFRC. Bangladesh, 2007, Xavier Gnot. Italy, 2009, Dipartimento Protezione Civile. DRC, Goma, 2009, Angela Rouse. Somalia, Hargeisa, 2009, David Womble.Guatemala 1976, Ian Davis.


4/138


iii

Introduction

The attempt to build institutional memory of past disaster responses has always been a challenge withinany implementing organisation. Building this memory collectively has been an even greater challenge. Shelterprogrammes, both the strategy and adopted technical solutions, are at best captured in evaluations, final reportsor annual reviews. Most often these sit on the shelves of agency headquarters, are buried inside field managerslaptops or become anecdotal snapshots passed on by the people involved. If not properly documented, memoriesfade away, year after year, disaster after disaster.

The second edition of the Shelter Projects 2009 contributes to changing this trend. It is an institutionalcollaboration between two key organisations involved in shelter and settlements after disasters, IFRC and UN-HABITAT. It offers a concrete tool for investigation and knowledge-sharing within the sector. It aims to servethe entire community of shelter practitioners, who are called to respond to natural disasters and conflicts bothnationally and internationally year in, year out.

As a philosophy, Shelter Projects 2009 promotes the idea that each intervention, whatever its impact on thecommunity, offers us lessons learned. Highlighting both good and bad practices will serve to improve quality andaccountability of our actions.

Shelter Projects 2009 draws the attention to the following three key findings:

Firstly, several examples show that the use of cash grants, material vouchers and other non-conventionalapproaches to shelter perform well in emergencies. Not yet mainstreamed, this approach has increased andencourages agencies to move away from standard emergency relief solutions such as tents and tarpaulins, whilebalancing advantages and disadvantages of this form of assistance with tangible examples at hand.

Secondly, there is increasing evidence that the world is going urban, and so are disasters. Although Haitis andChiles 2010 earthquake responses will be captured only in next years Shelter Projects publication, other examplesin this edition address the challenges in providing shelter in cities. How are collective centres and multi-occupancybuildings planned and used in Azerbaijan to host refugees? Does this affect the affected populations ability andwillingness to return home or find more sustainable solutions? Can a cash assistance programme still work whenthere is not much to buy, as is the case in Gaza city? What are the skills needed in urban damage assessments? Dothese assessments inform both short term shelter solutions and housing policies?

Thirdly, 2009 has also witnessed disasters in the so-called developed world. The earthquake that struckthe Abruzzi region in Italy and left a whole provincial city and its historic core in ruins. The case study tacklesdisaster response and reconstruction from an angle that humanitarians are less familiar with. It also highlightsthat, although available resources are considerably higher, many of the issues and bottlenecks are similar to thoseof the developing countries: loss of human lives, property and assets, pressure to identify interim locations toresettle people, political and legal constraints, equity issues, loss of memory and identity.

On behalf of our agencies, we hope that readers find the case studies relevant to their work and feel inspired tofind out more about past projects and to apply this knowledge for future projects. Many of our readers are familiarwith the projects or have contributed directly by providing examples and lessons learned from their field work forthis edition. Their experiences are also writing the pages of the future editions. Without their contribution therewould be no story to tell.

Foreword

Esteban LeonRisk Reduction and Recovery CoordinatorShelter and Early Recovery Clusters Focal PointUN-HABITAT

Graham SaundersHeadShelter & Settlements DepartmentInternational Federation of Red Crossand Red Crescent Societies


5/138

Acknowledgements

iv

Introduction

Acknowledgements

Project coordinated by: Esteban Leon (UNHABITAT), Sandra DUrzo (IFRC) and Joseph Ashmore

Compiled and edited by: Joseph Ashmore

Authors of specific sections: Joseph Ashmore, Ian Davis, Jim Kennedy, Jon Fowler, Seki Hirano, Sandra DUrzo

Case studies have been provided from the programmes of the following organisations:

American Red Cross CARE International CHF International CRS (Catholic Relief Services) GOAL ICRC (International Committee of the Red Cross) IFRC (International Federation of the Red Cross and Red Crescent Societies) IOM (International Organisation of Migration)

Italian Red Cross NRC (Norwegian Refugee Council) Oxfam GB Predes Peru Save the Children SDC (Swiss Agency for Development and Cooperation) UNHABITAT UNHCR (Office of the United Nations High Commissioner for Refugees) UNICEF (United Nations Childrens Fund) UNJLC (United Nations Joint Logistics Center) WFP (World Food Programme)

Special thanks to the following individuals who have been essential in the compilation of this book:Alberto Monguzzi, Angela Rouse, Antonella Vitale, Bakary Doumbia, Bryan Withall, Caroline Saint-Mieux,Christopher Nixon, Corinne Treherne, Dave Hodgkin, David Womble, Ed Cook, Eddie Argenal, Elin Holmn,Ernesto Morosin, Fabio Torretta, Felix de Vries, Filiep Decorte, Ged Robinson, Glenn Costes, Harry Jeene, IyadAbu Hamam, Jake Zarins, Jama Yasin Ibrahim, John Flomo, Jonathan Puddifoot, Julia Macro, Lucia Ashmore,Mariam Betlemidze, Martijn Goddeeris, Mawanda Shaban, Melisa Tan, Michele de Tomaso, Neil Bauman, NeilBrighton, Nuno Nunes, yvind Nordlie, Pete Manfield, Peter Nicolaus, Qurat-ul-Ain Sadozai, Rita Richter, SebastienPennes, Shane Middleton, Shaun Scales, Tom Corsellis, Tim Foster, Usman Qazi, Varatharajah Ramesh, WolfgangGressmann, Xavier Genot, Zeyad Elshakra.

We would also like to thank those who contributed to Shelter Projects 2008 whose work is replicated here in the

case study summaries.

We would like to thank the Cuny Center in Washington DC for allowing access to the documents that haveprovided many of the case studies in the historical section.

Documents originally developed for the Shelter After Disaster Guidelines published by The Office of the UN DisasterRelief Coordinator (UNDRO) in 1982. (UNDRO later evolved into UNOCHA). The Shelter Guidelines were developedby an international team of consultants including: Fred Cuny and Paul Thompson of INTERTECT, Fred Krimgoldfrom Virginia Tech. and edited by Ian Davis of Oxford Polytechnic.

Photographs are reproduced by kind permission of those whose names appear next to them in the text.


6/138


v

Introduction

ContentsContentsForewordAcknowledgementsContentsIntroduction

Section AConflict and Complex Disasters

A.1 Afghanistan - 2009 - Conflict returnees Full case study 3A.2 Afghanistan - 2002 onwards - Returns Update 7A.3 Azerbaijan - 1992 - Conflict displaced Summary 8A.4 DRC, Goma - 2009 - Conflict displaced Full case study 9A.5 Eritrea - 1998 onwards - Conflict Update 12A.6 Gaza, Palestine - 2009 - Conflict Full case study 13A.7 Georgia - 2008 - Conflict Full case study 16A.8 Ingushetia - 1999 - Conflict displaced Summary 19A.9 Kenya - 2008 - Conflict displaced Summary 20

A.10 Kenya, Dadaab - 2009 - Conflict refugees Update 21A.11 Kenya, Dadaab - 2007- Conflict and floods Summary 24A.12 Liberia - 2007- IDPs, refugees Update 25A.13 Rwanda - 2008 - Conflict returnees Update 26A.14 Somalia - 2008 - Conflict Overview 27A.15 Somalia, Puntland - 2009 - Conflict Full case study 29A.16 Somalia, Somaliland - 2009 - Conflict Full case study 32A.17 Somalia - 2007 - Conflict Summary 35A.18 Sri Lanka - 2007 - Conflict returns Update 36A.19 Sudan, Darfur - 2004 (ongoing) - Conflict Update 38

Section BNatural disasters

B.1 Bangladesh - 2009 - Cyclone Aila Overview 41B.2 Bangladesh - 2007 - Cyclone Sidr Overview 43B.3 Bangladesh - 2007 - Cyclone Sidr Full case study 46B.4 China, Sichuan - 2008 - Earthquake Full case study 50B.5 D.R.C, Goma - 2002 - Volcano Summary 53B.6 Haiti - 2008 - Flooding Full case study 54B.7 India, Gujarat - 2001 - Earthquake Summary 57B.8 Indonesia, Aceh - 2004 - Tsunami, earthquake Summary 58B.9 Jogyakarta - 2006 - Earthquake Overview 59B.10 Jogyakarta - 2006 - Earthquake Summary 60B.11 Jogyakarta - 2006 - Earthquake Summary 61B.12 Italy - 2009 - Earthquake Overview 62B.13 Italy - 2009 - Earthquake Full case study 63B.14 Mozambique - 2007 - Cyclone Summary 66B.15 Myanmar - 2008 - Cyclone Full case study 67B.16 Pakistan - 2005 - Earthquake Overview 70B.17 Pakistan - 2005 - Earthquake Summary 71B.18 Pakistan - 2005 - Earthquake Summary 72B.19 Peru - 2007 - Earthquake Overview 73B.20 Peru - 2007 - Earthquake Summary 74B.21 Peru - 2007 - Earthquake Summary 75

B.22 Peru - 2007 - Earthquake Summary 76B.23 Sri Lanka - 2004 - Tsunami Overview 77B.24 Sri Lanka - 2004 - Tsunami Summary 78B.25 Uganda - 2007 - Slow onset floods Full case study 79


7/138

Contents

vi

Introduction

Section CHistorical case studies - from the archives

C.1 Historical Introduction - by Ian Davis 85C.2 Algeria - 1980 - Earthquake Summary 87C.3 Bangladesh - 1975 - Conflict displaced Summary 89C.4 Guatamala - 1976 - Earthquake Summary 90

C.5 Guatamala - 1976 - Earthquake Summary 92C.6 Haiti - 1982 - Shelter report Report 93C.7 Honduras - 1998 - Hurricane Mitch Summary 97C.8 Honduras - 1974 - Hurricane Summary 98C.9 India - 1977 - Tropical cyclone Summary 99C.10 India - 1977 - Cyclone Summary 100C.11 India - 1971 - Conflict refugees Summary 102C.12 Italy - 1976 - Earthquake Summary 103C.13 Nicaragua - 1972 - Earthquake Summary 104C.14 Nicaragua - 1972 - Earthquake Summary 105C.15 Peru - 1970 - Earthquake Summary 107C.16 Sudan - 1985 - Conflict Summary 108

C.17 Thailand - 1979 - Political conflict Summary 109C.18 Tonga - 1982 - Cyclone Summary 110C.19 Turkey, Caldiran - 1976 - Earthquake Summary 111C.20 Turkey, Lice - 1975 - Earthquake Summary 112C.21 Turkey, Gediz - 1970 - Earthquake Overview 114C.22 UK - 1945 - Post conflict Summary 116C.23 Yugoslavia (formerly) - 1963 - Earthquake Summary 119

AnnexReference

Documents . 121Websites . 122


8/138


vii

Introduction

Introduction

This book contains summariesof shelter projects that have beenimplemented in response toconflicts and complex emergencies

(Section A, page 1) and to naturaldisasters(Section B, page 39). It alsocontains summaries of historicalshelter projects (Section C, page 83)that took place before 2000.

The case studies in this bookwere implemented by manydifferent organisations, a full listof which can be found in theacknowledgements section (pageiv). In order to allow weaknessesof programmes to be openly

shared, none of the case studiesare directly attributed to individualorganisations.

This book also containssummaries of case studies thathave a fuller write up in & ShelterProjects 2008. Where thereis significant new information,updates are included with theproject summaries.

As a result of the projects

being implemented in diverseand often challenging conditions,they illustrate both good and badpractices. From every case studythere are lessons that can be learnt,and aspects that should be repeatedor avoided elsewhere.

All projects in this book wereimplemented in different contexts.The case study Georgia 2009 (A.8page 16) gives a good examplesof how many projects must be

continually adapted to meet thechanging context. None of theseprojects should be copied directly.

Selection of case studiesGiven the scale of emergency

shelter need every year, casestudies included in this book mustbe implemented on a large scale.Trials or design concepts are notincluded.

The case studies were selected

using the following criteria as aguide:

The shelter project wasimplemented in full.

A minimum of five hundredfamilies had improved shelter asa result of the project activities.

The project was implemented

largely within the first yearfollowing natural disaster. Forconflict-affected populations,chronic emergencies and returnsprocesses, longer timescaleswere considered.

Accurate project informationhad to be available fromstaff involved in the projectimplementation.

The case studies that have beenselected illustrate a diversity of

approaches to meet shelter need.In all of them, providing shelteris more than simply designingarchitecturally impressive structures.

As a result of challengingcontexts, nothing was built at allin some of the projects. In otherprojects, the advocacy and learningsupport components of the projecthad a larger impact than theorganisation was able to achievethrough construction alone.

Shelter responses in 2009In 2009, 335 reported natural

disasters, killed over 10,000 peopleand affected more than 119 millionpeople1.

In 2009 there were alsosubstantial new displacementsdue to conflict in countries suchas Pakistan, Sri Lanka and Somalia.By the end of 2009, 43.3 millionpeople worldwide had been

forcibly displaced by conflict andpersecution, the highest numbersince the mid-1990s2. This includesan estimated 27,100,000 peoplewho had been displaced withincountries, and an additional 15.2million refugees who had beendisplaced into other countries.

1) Annual Disaster Statistical Review2009. The numbers and trends. CRED,

WHO, Universit Catholique de Louvain2) 2009 Global Trends: Refugees,Asylum-seekers, Returnees, Inter-nally Displaced and Stateless Persons,UNHCR 2010

Disasters and conflict havehad different impacts in differentregions. Asian countries were mostaffected by natural disasters whilst

countries in Africa were mostaffected by conflict.

As many of the responses todisasters in 2009 are still ongoing,many of them are not written upin this book. Instead we includethe following new case studies,as well as updates to case studiespreviously documented in &Shelter Projects 2008.

New case studies inShelter Projects 20091945

UK (post conflict) - page 116

1982

Haiti - page 54

2007

Bangladesh (Cyclone Sidr) -page 116

Uganda (flooding) - page 79

2008China, (earthquake) - page 50

Georgia (conflict) - page 16

Haiti (flooding) - page 54

Myanmar (Cyclone Nargis) -page 67

Somalia, Puntland (conflict) -page 29

Somalia, Somaliland (conflict)- page 32

2009Afghanistan (conflict returns) -page 3

Bangladesh (Cyclone Aila) -page 41

DRC (conflict) - page 9

Gaza (conflict) - page 13

Italy (earthquake) - page 62

Kenya, Dadaab (conflict/flooding) - page 21


9/138

Introduction

viii

Introduction

Non-food item

distribution

Shelterconstruction

Labour

U

rban?

H

ousehold

items

S

helter

m

aterials

T

ransitional

P

ermanent

/

c

orehousing

C

ashor

v

oucher

C

ommunity

C

ontracted

D

irect

T

echnical

expertise

A.1 Afghanistan - 2009

A.2 Afghanistan - 2002

A.3 Azerbaijan - 1992 Y

A.4 DRC - 2009

A.5 Eritrea - 1998

A.6 Gaza (Palestine) -2008 Y

A.7 Georgia - 2009

A.8 Ingushetia - 1999 Y

A.9 Kenya - 2008

A.10 Kenya (Dadaab) -2009

A.11 Kenya (Dabaab) - 2007

A.12 Liberia - 2007

A.13 Rwanda - 2008v

A.15 Somalia - 2008 Y



A.18 Sri Lanka - 2007

A.19 Sudan, Darfur - 2004

B.3 Bangladesh - 2009

B.4 China - 2009

B.5 D.R.Congo - 2002 Y

B.6 Haiti - 2008 Y

B.7 India - Gujarat - 2001

B.8 Indonesia - Aceh - 2004 Y

B.10 Jogyakarta - 2006

B.11 Jogyakarta - 2006B.13 Italy - 2009 Y

B.14 Mozambique - 2007

B.15 Myanmar - 2008

B.17 Pakistan - 2005

B.18 Pakistan - 2005

B.20 Peru - 2007 Y

B.21 Peru - 2007 Y

B.22 Peru - 2007

B.23 Sri Lanka - 2004

B.24 Uganda - 2007

Explanation

ofcolumns:

Urban?-w

astheprogrammeinanurbanenvironm

ent?

Non-fooditemsdistribution-whichtypesofmate

rialsweredistributed

sheltercon

struction-weretransitionalsheltersbu

iltorwerepermanentorcorehousesb

uilt?

Labour-co

mmunitymembersbuilttheirownshelters,contractorsbuiltsheltersordirectlabour-theorganisationbuiltthe

houses

Technicale

xpertise-theshelterprojecthadatrain

ingcomponentorhadsignificantexternaltechnicalsupport


10/138


ix

Introduction

This book includes many diverseprojects, all of which aimed toimprove the shelter of the affectedpopulation.

Projects included in this bookrange from building damageassessment (A.6, Gaza, Palestine -

2008, page 13) to cash or voucherdistribution (page 9) to hiring ofcontractors to build shelters, totraining (B.25, Uganda - 2007, page79). Many of the projects, such asthe one responding to cyclone Sidrin Bangladesh have used severalapproaches (B.3, page 46) to meetthe needs as they evolve followinga disaster.

Despite the differences betweenprojects, there are many recurring

themes which we discuss below.

Support the affectedpeople

The first and main effort in allresponses is by the affected peoplethemselves. Of the case studiesin this book, the more effectiveprojects were implemented with theclose involvement of the affectees.

& Sphere standards andindicators (Annex on page 121)provides common standards onparticipation, initial assessment,monitoring and evaluation.

& Supporting the affectedpeople is the first principle outlined

in Transitional Settlement andReconstruction after NaturalDisasters as well as in Shelter AfterDisaster (Annex on page 121)

Urban shelterAs over half of the worlds

population now lives in urbanareas, and the long term impactsof many disasters is to drive peoplefurther into cities, this editionof Shelter Projects includes casestudies of shelter programmes inurban environments.

The case study of solidarityfamilies in Democratic Republic ofCongo (A.4, page 9) provides a goodexample of how finding shelter withhost families may be supported.The case study from the Haiti floodsof 2008 (B.6, page 54) includes aprogramme of supporting familiesin collective centres to find rentalproperties. The case study of Gaza(A.6, page 13) illustrates a detailed

damage assessment of multi-storeyconcrete buildings with differentapartments owned by differentfamilies.

Phases of responseResponses to disasters or conflict

are commonly split into phases:

preparedness before the disaster emergency response recovery phase durable solutions

Many of the case studies includeemergency shelter responses aimedat bridging the gap betweenemergency shelter and durablehousing solutions, whilst reducingvulnerability to future disasters. Thesummary graph below illustrates theduration of the different projects inthis book.

Housing programmes can takemany years to complete, especiallywhen implemented on a large scale.The project in Rwanda (A.13, page26), illustrates a housing project thattook two years to build 220 houses.The speed of durable shelterconstruction can leave a gap, withfamilies in emergency shelter formany years. Transitional responsesaim to bridge this gap.

B.20Peru

B.21Peru

A.1Afghanistan

B.18Pakistan

B.17Pakistan

B.15Myanmar

A.18SriLanka

B.6Haiti

B.22Peru

A.7Georgia

B.11Indonesia

B.14Mozambique

A.9Kenya

B.5DRC

B.24SriLanka

A.4DRC

B.10Indonesia

B.13Italy

B.7India

A.15Somalia

A.17Somalia

A.12Liberia

A.6

Gaza(Palastine)

A.16Somalia

A.13Rwanda

B.3Bangladesh

B.4China

A.8Ingushetia

B.8Indonesia

4 years

3 years

2 years

1 year

6 months

disaster /displacement

Graph of the duration of the different projects.

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.

Programme duration

Time between disaster and project start


11/138

Introduction

x

Introduction

500,000

400,000

300,000

200,000

100,000

0 7momths

6months

5months

4months

3months

2months

1month

0

Pakistan - 2005 - earthquake

Households provided with a tent

Total: 520,000 tents

Jogyakarta - 2006 - earthquake

Households provided with a tent or a tarpaulin

Total: 390,000 tents and tarpaulins

Jogyakarta - 2006 - earthquake

Transitional shelter construction

Total: 75,000 built in 12 months

Graph of shelter materials distribution to households against time for four major disasters.

In all disasters, additional materials were distributed. eg. in Pakistan corrugated iron, plastic sheeting and shelter repair kits

were also distributed. The information in this graph does not take into account targetting, or other support activities such as

training, advocacy or voucher distributionData is taken from shelter cluster commodity tracking lists or from OCHA sitreps and is subject to errors in reporting.

Numberoffamilie

sreached

Pakistan - 2005 - tents

Jogyakarta - 2006 - tents and tarpaulins

Jogyakarta - 2006 - transitional shelter

Sumatra - 2009 - tents or tarpaulins

Myanmar - 2008 - tarpaulins

Time after disaster

Sumatra - 2009 - earthquake

Households provided with tents or tarpaulins

Total: 170,000 tarpaulins

Myanmar - 2008 - cyclone

Households provided with two tarpaulinsTotal: 350,000 tarpaulins

Speed of ResponseCommonly, following a large-

scale natural disaster, there is asurge of media attention at theoutset, with an initial focus onsearch and rescue and latterly onaid delivery. Media interest often

focuses at some stage on the paceof relief distributions.

The graph below illustrates thespeeds of emergency distributionfrom responses in Indonesia(Jogyakarta 2006), Myanmar (2009)and Indonesia (Sumatra 2009). It isinteresting to compare these graphswith the media interest which

peaks during the first weeks. Reliefdistributions are run on longertime frames than the initial mediareporting of the emergency.

In the first month, organisationsmust often rapidly scale up staffing,establish supply pipelines and

mobilise distribution teams oragreements and support partnerorganisations. From the examplesbelow, the quickest distribution ofshelter items lasted two months.

Transitional shelterprogramming often takeseven longer to implement. TheJogyakarta transitional shelter

response (Indonesia 2006) is knownas being relatively quick, but stilltook 12 months to build 75,000transitional shelters.

Core shelter in which a basicstructure is built with the intentionthat families can upgrade later. In

Sri Lanka (A.19 , page 36) this allowedshort term shelter funding to beused in building more permanenthousing.

Permanent housing programmesoften take three or more years tocomplete on a significant scale (seeB.8, Indonesia, on page 58 for anexample).


12/138


xi

Introduction

FundingFunding of is often one of the

key determinants of the typesof shelter responses. A goodillusatration of this is the contrastbetween the responses to the2008 earthquake in Italy (B.12,

page 62) and the 2007 earthquake inPeru (B.19, page 73) or the conflictin Somalia (A.14, page 73). Thedifferent responses to cyclones Aila(B.1, page 41) and Sidr (B.2 page 46),both in Bangladesh, are partly theresult of differing funding.

The time frame of funding hasa major impact on the projectsimplemented. Often relief fundsare for less than one year, meaningthat there is time pressure to

complete construction of agreednumbers of shelters rather than toensure that shelters are occupiedand meet the needs.

Scale of programmeThe responses illustrate the

challenge of whether to implementhigh quality programmes for fewerpeople or poorer quality responsesto support more people. Thecase studies in Somalia illustratethis challenge. One project (A.15

Somalia 2009, page 29) deliveredimproved but basic shelter to over24,000 people, the other (A.17Somalia, 2008, page 35) providedimproved sites and services for 700people over two years with higherproject costs per family.

Assistance methodsThe case studies in this book

include a diversity of ways ofgetting assistance to people whohave been affected by conflict ordisaster.

A standardised shelter kit oftools, fixings and tarpaulins thatcan be stockpiled or procuredlocally and distributed followingemergencies has recently beendeveloped and deployed in manycountries. The aim of the kit is tosupport families to build strongershelters, as well as providing itemsthat will help during reconstruction.

The case study from Myanmarprovides some lessons for its use.

Climate and riskShelter programmes respond

to different hazards in differentways. In the town of Bosaso inSomalia (A.15, page 29) one of themajor threats to families living indisplacement sites was fire, which

had destroyed many peopleshomes. The best way to reduce therisk of a major fire was to work withcommunities to improve urban solidwaste disposal, to install fire breaks,and to establish committees. Thisproved more cost-effective andpractical as an approach than tryingto build fire-proof shelters.

Projects were implemented invery different climates. The casestudy from Sozma Qala camp in

Afghanistan (A.1, page 3) illustrateswinterisation of tents as wellas water supplies for a camp inAfghanistan prior to the onset ofwinter. Many of the responses,such as that to the Uganda floodsin 2007 (B.25 page 79) take intoaccount the weather, maintainingthermal comfort of buildingswhilst including hazard mitigationmeasures.

Selecting beneficiariesProject location is often selected

by a combination of where needis greatest, where organisationsor their partners have existingprogrammes, and where thereare gaps. Giving support to aselected location can create a pullfactor and increase the populationrequesting support. The case studyfrom Gonaives in Haiti, 2008, (B.6,page 54) illustrates some of thesechallenges and some measures thatwere taken in response to the pullfactors caused by the assistance.

Selecting which individuals orfamilies will benefit from a projectis often challenging, and if badlymanaged can be open to abuse.

The case study from cyclone Sidrin Bangladesh 2007 (B.3, page 54)provides an example of selectingfamilies with the agreement ofcommunity committees.

Basing selection solely onwhether a familys house is damagedcan bias responses away from thosewho rent or squat their homes.Selecting families on the basis ofother vulnerabilities such as inSomaliland, (B.16, page 32) requires

clearly agreed criteria and requiresworking closely with communitiesto be effective. Whichever selectioncriteria are used, the larger thepackage of support being offered,the more pressure there will be ongetting the selection of familiescorrect.

TrainingMany of the projects in this

book, such as those outlined forHaiti in 1982 (C.6, page 93) or

the programme implemented inUganda in response to the 2007flooding (B.25, page 79) have asignificant training and capacitybuilding component. Successfultraining programmes have asignificant disaster risk reductioncomponent, allowing the shelterassistance programmes followinga conflict or a natural disaster toreduce vulnerability in the future.

Selecting project locations and

choosing who will benefit iscritical to the success of projects.

Photo: Joseph Ashmore


13/138

Introduction

xii

Introduction

9m2

10m2

10.5m2

11.1m2

11.5m2

12m2

13.5m2

15m2

16m2

16m2

16m2

18m2

18m2

18m2

18m2

18m2

18m2

18m2

18.6m2

20m2

20.5m2

21m2

21m2

24m2

24m2

24m2

24m2

25m2

25m2

36m2

38.7m2

48m2

B.20 Peru

B.7 India

B.17 Pakistan

C.7 Honduras

A.4 DRC

B.14 Mozambique

A.17 Somalia

B.3 Bangladesh

A.5 Eritrea

A.16 Somalia

C.17 Thailand

A.9 Kenya

A.11 KenyaB.18 Pakistan

B.21 Peru

B.22 Peru

B.6 Haiti

C.7 Kenya

A.18 Sri Lanka

C.13 Nicaragua

B.24 Sri Lanka

A.2 Afghanistan

A.8 Ingueshetia

A.7 Georgia

B.5 DRC

B.10 Indonesia

B.11 Indonesia

A.11 Liberia

C.9 India

B.8 Indonesia

A.1 Afghanistan

A.13 Rwanda

B.12 Italy

7peoplex3.5m

2

6pe

oplex3.5m2

5pe

oplex3.5m2

4pe

oplex3.5m2

3pe

oplex3.5m2

2pe

oplex3.5m2

1pe

rsonx3.5m2

Size of shelterChart showing sizes of the shelters in the

case studies in comparison with allocation of

3.5m2 per person.

Note that smaller shelters are often con-

structed after assessment of local and host

population standards as well as what is

practically possible. Size of shelters is not

necessarily a good indicator of the quality ofa shelter programme.

Shelter sizeThe illustration below shows the

diversity of shelter covered areasin these case studies. These varyfrom 9m2 (B.20, Peru 2007, page74) to 74m2 (B.12, Italy, 2009, page62) this is a result of varying needs,

permanency, budgets logisticsconstraints, host standards andofficial policies.

& Sphere 2004 - Shelter andSettlement, Standard 3: coveredliving space has the followingindicator:

Duration: in the immediateaftermath of a disaster, particularlyin extreme climatic conditionswhere shelter materials are notreadily available, a covered areaof less than 3.5m2 per person maybe appropriate to save life andto provide adequate short-termshelter to the greatest number ofpeople in need. In such instances,the shelter response should bedesigned to reach 3.5m2 per personas soon as possible, as longerdurations may begin to affectthe health and well-being of thepeople accommodated. If 3.5m2per person cannot be achieved, or

is in excess of the typical space usedby the affected or neighbouringpopulation, consideration shouldbe given to the impact on dignity,health and privacy of a reducedcovered area. A decision toprovide less than 3.5m2 per personshould be highlighted, along withmeasures to mitigate against anyadverse affects on the affectedpopulation.

45-78m2


14/138


xiii

Introduction

Nero at this time was at Antium,and did not return to Rome until the fireapproached his house, which he had builtto connect the palace with the gardensof Maecenas. It could not, however,be stopped from devouring the palace,the house, and everything around it.However, to relieve the people, driven outhomeless as they were, he threw open tothem the Campus Martius and the publicbuildings of Agrippa, and even his owngardens, and raised temporary structuresto receive the destitute multitude. Suppliesof food were brought up from Ostia andthe neighbouring towns, and the price

of corn was reduced to three sestercesa peck. These acts, though popular,produced no effect, since a rumour hadgone forth everywhere that, at the verytime when the city was in flames, theemperor appeared on a private stage andsang of the destruction of Troy, comparingpresent misfortunes with the calamities ofantiquity.

Tacitus - The Annals / Book 15 - writing in

64 AD following the fire of Rome. an early

example of emergency shelter provision.


15/138

Introduction

xiv

Introduction


16/138

Conflict / Complex Shelter Projects 2009 A.1

1

A

SECTION AConflict and Complex Disasters

A.1 Afghanistan - 2009 - Conflict returnees 3

A.2 Afghanistan - 2002 onwards - Returns 7

A.3 Azerbaijan - 1992 - Conflict displaced 8

A.4 DRC, Goma - 2009 - Conflict displaced 9

A.5 Eritrea - 1998 onwards - Conflict 12

A.6 Gaza, Palestine - 2009 - Conflict 13

A.7 Georgia - 2008 - Conflict 16

A.8 Ingushetia - 1999 - Conflict displaced 19

A.9 Kenya - 2008 - Conflict displaced 20

A.10 Kenya, Dadaab - 2009 - Conflict refugees 21

A.11 Kenya, Dadaab - 2007- Conflict and floods 24

A.12 Liberia - 2007- IDPs, refugees 25

A.13 Rwanda - 2008 - Conflict returnees 26

A.14 Somalia - 2008 - Conflict 27

A.15 Somalia, Puntland - 2009 - Conflict 29

A.16 Somalia, Somaliland - 2009 - Conflict / drought 32

A.17 Somalia - 2007 - Conflict 35

A.18 Sri Lanka - 2007 - Conflict returns 36

A.19 Sudan, Darfur - 2004 onwards - Conflict 38

A.13

A.1-A.2

A.12

A.8

A.9-A.11

A.7

A.6

A.14-A.17

A.18

A.3

A.4

A.19 A.5


17/138

Conflict and Complex Disasters Conflict / ComplexA.1

2


18/138


3

Constructioncomplete

Construction start

Tents distributed

Displacement

Afghanistan - 2009 - Conflict returnees

9 These shelters allow the the original tent to continueto be used by providing additional protection againstsevere weather conditions.

9 The shelters could be constructed quickly using localsemi-skilled labour.9 Since the shelters do not use materials that maybe considered permanent they send a clear messagethat this is a temporary emergency provision. Existingagreements with local authorities and surroundingvillages regarding the occupancy of the transit campclearly stated that this location is a temporary facility.

9 The relatively large covered area of the shelter allows

for clothes drying and safe storage of belongings aswell as catering for larger family units.

9 Ongoing assistance programmes in the area alloweda degree of monitoring throughout the winter months.

9 Staff from the organisation continued to engagewith the community throughout the winter period.

8 Early expectations of the community focussed onthe provision of a permanent house rather than theextended provision of another form of temporaryshelter.

8 An extended delivery pipeline for bamboo polesstretching from Pakistan to northern Afghanistan,meant the program was vulnerable to delays caused byinsecurity, border procedures and bad weather. Overa 6 week construction period, 12 days were awaiting

delivery of bamboo.- The organisation intends to adopt a similar

deployment methodology for future fast-onsetemergencies in the region.

A.1

Strengths and weaknesses

Country:Afghanistan

Disaster:Afghanistan returns to SozmaQala camp

Disaster date:2009

Number of people displaced:2002 - 2010 - over 5million people returned toAfghanistan.

Project target population:379 families

Occupancy rate on handover:94%. 14 of the familiescompleted 1 room mud brickshelters and used the sheltersconstructed by this project forgeneral store and livestock.

Shelter size:Covered area = 38.7m2

Materials cost per shelter:$ 300

Project cost per shelter:$ 910 overall budget including

all additional winterisationworks, project staff, transport,office accommodation,administration, etc

SummaryAn emergency team rapidly winterised a temporary transit camp. The site was for 379 families for refugees

returned from Iran to their district of origin in northern Afghanistan. To improve the existing tents, a productionline was set up in the camp to build bamboo and plastic sheeting shelters, which provided additional protectionfrom severe winter weather. The structure was developed from a model implemented in Pakistan AdministeredKashmir in 2006-2007

4 months

3 months

August 2009

Project timeline

Winterisation

Afganistan

Sozma, Quala

Case study: Full case study


19/138

Afghanistan - 2009 - Conflict returnees - Winterisation Conflict / ComplexA.1

4

BackgroundAfter the closure of a camp in

Iran that had been their home for23 years, families were returned to

the Sozma Qala area of Sar i Pulprovince of Afghanistan. As a resultof the number of years of displace-ment, many of the original housesand water facilities were destroyedand there were issues regardingland ownership. For the majority ofthe returning families an immediatereturn to their village of origin wasnot possible.

Afghan authorities in SozmaQala approved the development

of a temporary transit facility onland near to villages of return.Families would be able to live onthis temporary site until durablesolutions were developed.

The original transit camp facili-ties were built to provide temporarysupport and were not intendedfor winter occupancy. Tents wereprovided as family shelters and abasic water delivery system hadbeen developed.

By October 2009 it was clearthat by the onset of winter only alimited number of families wouldhave returned to their villages.

Inadequacy of the tents hadlead to the population in the campbeing extremely vulnerable to the

coming severe winter weather. Thelargest element of the winterisationprogram was the provision of addi-tional shelter to ensure that familiesliving in tents had improved shelter.

Selection of beneficiariesA relatively late decision by

the Iranian authorities to closethe camp occupied by these 379families resulted in their spontane-ous return to their area of origin.Previous returns tended to be pre-planned. This had allowed timefor the mobilisation of resourcesto receive them and the construc-tion of durable mud-brick sheltersthat could later be extended intopermanent dwellings.

There were few opportuni-ties for temporary hosting by thelocal community until permanentdwellings could be constructedwere practically non-existent.These factors contributed to thedecision to develop a temporarytransit facility as the first step inthe return process. Family unitswere largely maintained in line with

the households established in theIranian refugee camp. There wassome negotiation related to larger

families , which was often resolvedby providing of an additional smalltent.

It is anticipated that Afghanistanwill see more of this kind of sponta-neous return throughout 2010.

Technical solutionsBamboo structures clad in plastic

sheeting were built around existingtents. Sections of the structureswere prefabricated by local car-

penters in a warehouse tent withinthe camp. They were then passedto beneficiary assembly teams forshelter construction on designatedfamily plots.

This approach to shelter wasbased on learnings from previousearthquake responses in Pakistanfor the winter of 2006-2007

The Pakistan design wasaltered to simplify the construc-tion process and allow semi-skilled

and unskilled members of the ben-eficiary community to assemble theshelters.

Sozma Qula Transit Camp before winterisation

Photo: Shaun Scales

Sozma Qula Transit Camp after winterisation

Photo: Shaun Scales


20/138


5

The shelters were known asWeather Mitigating Tent Shelters(WMTS) and were provided asa temporary shelter suitable forwinter.

The relatively large coveredarea of the shelters allowed for themaintenance of larger family units.

The assistance delivery processwas streamlined by having rawmaterials delivered directly to siteand then processed through thewarehouse tent.

ImplementationThe project was implemented by

a mixed gender team of emergencyfocal points, engineers, field as-sistants and logistics personnel.Staff were seconded from otherprograms that were managed bythe organisation elsewhere in Af-

ghanistan. 1 expatriate coordinator 2 emergency focal points 1 team leader 1 logistics assistant 6 field assistants 2 drivers

Site winterisationGround water pipelines were

dug deeper to prevent freezing andwater storage bladders positionedon platforms and protected with in-sulating enclosures.

Mitre drains were built to divertfuture snow melt into a natural

gully before it reached the camparea.

A simple gravel road networkallowed easier access for pedestri-ans and water tankers.

Additional drainage was builtwith the roads to divert rain andsnow melt from within the camp.

Four insulated bathing enclo-

sures (two male, two female) with10 separate cubicles in each havebeen constructed adjacent to win-terised tap stands.

All 379 families were issuedwith a local Bukari (solid fuel stove)and chimney kit with 90kg of coalas the first of four fuel distribu-tions. These were intended to covera 3 month period. Technical teamswere active throughout the campproviding advice on the safe fitting

of chimneys and the maintenanceof fire breaks between shelters.

Additional winter clothing kitswere also distributed to each family.

Logistics and materialsBamboo and plastic sheeting

were procured internationally andwere subject to delays due to poorweather and insecurity. All otherhardware materials and tools forthe project were purchased locallyin Mazar-I-Sharif.

Materials list for oneshelter

Quantity unit

10mbamboopoles 24pieces

Standard5mx4mplasticsheeting

7pieces

PlywoodSheet6mmx1525mmx1525mm

4.6pieces

Bolts-6mmdiameter150mmlong

84pieces

Nutsforbolts 84pieces

Washersfornuts 168pieces

Nails100mm(4) 110pieces

SmallNails50mm(forfixing

tarpaulin)

1000pieces

Washersforsmallnails(tomakewidehead)

800pieces

Bindingrope5mmcotton 60m

Rapid construction process - materials were prefabricaetd in a wrehouswe and assembled on site

Photo Shaun Scales

Finished shelter. The end wall has been covered by the occupants

Photo Shaun Scales


21/138

Afghanistan - 2009 - Conflict returnees - Winterisation Conflict / ComplexA.1

6

We came home hoping toreceive a full house with brickand windows. All we receivedwas a tent in a camp with noprivacy and not on our ownland. This organisation helpedus to build shelters to save usfor winter. The shelters arebetter than a tent but we hopeto have a full house soon.

Exterior and interior photographs of the completed shelters, showing the previously distributed tents inside.

Photo: Shaun Scales


22/138


7

115,000 shelterscompleted

90,000 shelters

completed

40,000 sheltersbuilt

Return commence

Overthrow ofTaliban

Afghanistan - 2002 onwards - Returns

Project type:Package of shelter construction

materials,self build shelters,cash grants,technical support

Disaster:Afghanistan repatriation 2002-2008

Houses damaged by disaster:5 million returnees since 2002

Project target population:1.2 million beneficiaries to date(average family size of 6).

This programme has shelteredan estimated 25% of returningpopulation

Occupancy rate on handover:Unknown

Shelter size:It was strongly recommendedto maintain a 21m2 minimumnet floor area.The shelter consists of tworooms, one corridor, and anexternal latrine.

SummaryA large scale, self-build. shelter programme implemented through partner organisations. The programme

aimed to help meet the needs of the 5 million people returning to Afghanistan since 2002 over 20 years ofconflict. Different shelter models were adopted around the country depending on local construction technology.This programme provided materials, basic technical guidance and cash for the most vulnerable people. It wasintegrated with monitoring and support for return. Escalating steel prices severely affected the programmeleaving it 5 million US dollars under budget for 2008.

Mar 2004

July 2004

July 2003

2002

Project timeline

Update - Shelter construction

Afghanistan

A.2

Case study:See Shelter Projects

2008 for more

UpdateThis programme continued in 2009 and looks set to continue for many years to come. The lead organisation

continued to develop detailed guidance for partner organisations. This included detailed paperwork requirementswhich aimed to encourage consistency between programmes accross the country. Some organisations foundthese too prescriptive and preferred to build shelters from other funding sources.

The shelters continued to be built in exisitng settlements as well as in new settlements. Accross the country,three main designs of shelter were developed. Minor design changes were made such as varying the amount oftimber used for various structural elements. In general the use of steel I-beams in the shelters proved popular.

Projects are implemented on a one year cycle with organisations being funded for construction one year ata time. This provides an incentive for rapid and large scale construction, but can make follow on projects to re-

establish communities difficult to plan, and means that shelter occupancy is not as high as it could be.


23/138

Azerbaijan - 1992 - Conflict displaced - Upgrade of collective centres Conflict / ComplexA.3

8

Project closed

Project begins

170,000 peopledisplaced to Baku

Azerbaijan - 1992 - Conflict displaced

Project type:Upgrade of collective centres

Disaster:Nagorno Karabakh conflict

Houses damaged by disaster:700,000 displaced40,915 families (169,609people) came to Baku in 1992-1993

Project target population:27, 500 in over 60 buildingsover 8 years

Occupancy rate on handover:No data. Room allocation in

the buildings is dynamic.

Shelter sizeVariable. Individual roomsoften shared by whole families

SummaryThis programme upgraded and maintained public buildings that people had moved to during the conflict in

Nagorno-Karabakh in the early 1990s. The project worked with families who, by the end of the project had beendisplaced for over ten years. The way of working evolved over time, starting with contractor led construction,evolving into direct implementation by the NGO. Although the project closed without a clear exit strategy,aspects of the project were taken up by the government in their housing policies.

13 years

5 years

First months

Conflict

Project timeline

Upgrade of collective centres

Azerbaijan

Baku

A.3


2008 for more

Photo: NRC Azerbaijan

The project worked with families who had been displaced by

conflict and were living in public buildings


24/138


9

Project completion

Distribution ofmaterials

Distribution ofsecond vouchers(food)

Distribution of firstvouchers (food andhousehold items)

Communityworkshops defineselection criteriaand design for

shelters

Project start:

New influx ofpeople into Goma

Start of ongoingconflict

DRC, Goma - 2009 - Conflict displaced

9 An alternative to camps was found, and at a lower

cost.9 Both hosting and hosted families were given a largedegree of control9 The communities themselves, as well as theauthorities and local groups and churches were veryinvolved in the project design and its implementation.

9 A significant number of the families hosted totalstrangers. In some cases the hosting family was froma different ethnic or linguistic background than thehosted family. This showed the spirit of Umoja.9 Livelihoods of the displaced families were supportedthrough the provision of more secure shelter closer to

areas of high economic activity.9 Families were able to get the supplier to substitutesome materials for a better quality at the same price.

9 Tensions between host and displaced communities

were reduced.

8 Initially, many vendors dropped out, making pricesfor food and shelter items difficult to control. This waslater resolved.

8 As this was a pilot project, high levels of monitoringand involvement by senior management staff wererequired.

8 High levels of sensitization and monitoring wererequired

8 The project was not supported by pooled funding asit did not fall into pre-defined categories such as CampManagement or Early Recovery.- Exisiting houses were smaller than 3.5m2 per

person. The shelters built by the project respected thisto reduce the risk of tensions arising.- This project was not linked to any formal urban or

regional planning.


Country:Democratic Republic of Congo

Disaster:Ongoing armed conflict

Disaster date:1994 - Conflict in easternDRC 2008 - Offensive towardsGoma

Number of houses damaged:Unknown

Number of people displaced:>100,000 for this phase of theconflict. Millions cumulativelyover the previous 16 years.

Project target population:250 solidarity families

Occupancy rate on handover:100% on project completion.

Shelter size:11.5m2 extension to existinghouses.Increase from 1.5m2 per personto 2.25m2 per person.

Materials Cost per shelter:680 USD for shelters, latrines

and labour.Project cost per shelter:

250 USD per person, inclusiveof operational / support costs.

10 months 9 months

8 months

7 months

5 months

2 months -

Oct. 2008 -

1994

Project timeline

Urban host families, vouchers

GOMA

SummaryMulti-sectoral support to Umoja (solidarity) hosting and hosted families following an influx of displaced

people into Goma. Families were provided with materials for either repair or additions / extensions to existinghousing, as well as key household items using a voucher system.

Democratic Republic ofCongo

Goma

A.4



25/138

DRC, Goma - 2009 - Conflict displaced - Urban host families, vouchers Conflict / ComplexA.4

10

Before the displacementThere have been multiple large-

scale forced displacements of popu-lation in Goma since the RwandanGenocide of 1994. There was alsolarge-scale displacement followingthe volcanic eruption in 2002.

Prior to the conflicts, the popu-lation of Goma had been estimatedat less than 50,000 people, but by

2008 the population estimates hadreached more than 800,000. Forthe two neighbourhoods in Gomachosen for the project, both werewithin the city limits. One, Kasika,had been created in a plannedmanner, whilst the other, Ndosho,was less planned. Both areas hadsuffered stresses on infrastructureand water resources before thedisaster.

After the displacementSince 1994, population dis-placement through conflict hasbeen pendular, with families oftenmoving relatively short distancesfrom their homes, and thenreturning again, once the levels ofinsecurity had fallen. However, theapproach of the rebels betweenOctober and November 2008, andthe subsequent fighting in otherclose-by areas in early 2009 meantthat many families would not be

able to return home rapidly, andthat they would need support for alonger period of time.

The humanitarian community,with the United Nations and thegovernment, were able to providespaces inside planned camps for69,000 people. This was not suf-ficient for the entire displacedpopulation. It also required muchfunding and resources. The camplocations, outside the city, meant

that the displaced families had lessaccess to livelihoods, and less like-lihood of achieving any economicindependence.

Of those who did not residein the camps, but who looked forshelter in the city, almost all foundshelter with host families. This wasarranged through relatives, throughintroductions, through church asso-ciations and through other mecha-nisms. Some families were hostedfor free, whilst others paid rent. In

the majority of cases, indoor spacefor the hosting and the hostedfamilies was greatly reduced, andstrains increased as time went on.

ImplementationA multi-sectoral approach was

chosen, to support the Umojaor solidarity of the families whowere hosting or hosted. As a pilotproject, two neighbourhoods wereselected, where a large number ofdisplaced people were living with

host families. Key needs, includingthose of shelter and non food items,were identified through consulta-tion with affected communities.

It was decided to give as muchchoice as possible to enable thefamilies to choose items that theyneeded. As a result a voucherscheme was implemented.

Selection of beneficiariesThe organisation worked with a

committee that included membersof both the hosting and the hostedfamilies. These committees createda list of vulnerabilities, and priori-tised or weighted each differentcategory in the list.

The Chef de Quartier provided alist of solidarity families, which werethen visited and weighted againstvarious vulnerability indicators. Themost vulnerable families were thenretained as beneficiaries. Lists were

displayed to allow the communityto pick out any fraudsters. Thecommittee was very involved in thewhole process.

Technical solutionsStandard designs were created

before the bill of quantities wasfinalised. These designs werecreated through the communityconsultation process, and thenshown to the selected familiesbefore construction.

However, as houses haddifferent designs, and plots varied,families were given flexibility inthe design that they built. Some

families used the materials to repairhouses, whilst others used them tobuild extensions.

For the distribution of householditems and food, a voucher schemewas used, in co-operation with anumber of selected local merchants.The merchants then returned thevouchers to the organisation forpayment. Certain items, such as.alcohol, could not be purchasedusing the vouchers, but otherwise a

wide range of items, including mat-tresses and cooking utensils, wasmade available to the beneficiaries.

At first, many of the merchantswere hesitant about the scheme,but were finally won over. However,at the same time, there were accu-sations that some of the merchantswere over-charging, above thefixed prices that had been agreedwith the organisation.

A team consisting of committeemembers and staff from the or-ganisation monitored the use ofvouchers. Families were encour-

Left to right: Hosting families used

distributed materials to improve

their houses

Photos: Angela Rouse


26/138


11

aged to barter or leave the shop ifprices were too high.

Logistics and materialsThe food and shelter items

were identified as being a priorityduring the community consultationprocess. Vouchers were then issuedfor redemption at approved andselected local merchants. A previousmarket analysis conducted by theorganisation ensured that the localmarkets would be able to provideall the items. The logistics for thehousehold items was entirely un-dertaken by the merchants them-selves.

The method of distribution ofthe shelter construction materialswas the subject of much discussionwith those receiving them. Initiallymany did not want distributiondirectly to their homes, as this mightincite jealousy from the neighbours.Additionally, the informal layout ofthe neighbourhoods, and the roughlava-rock surfaces made it difficultfor trucks to access all of the targetareas.

In the end, two distributionpoints (one in each of the two com-

munities) were selected for theconstruction materials. Most of thematerials were sourced locally, withan acknowledgement that sourcingtimber from sustainable resources isparticularly challenging in DRC.

Materials list

Material Quantity

WoodenPlank 42pieces

2"x2"woodbeam 32pieces

CGIsheetBG32 8pcs

Cement50Kg 3sacks

Sand 1.09m3

Roughsand 0.55m3

Roofingnails 1kg

10cmnails 5kg

8cmnails 6kg

6cmnails 6kg

4cmnails 0.5kg

Doorwithaccessories80/180cm 1

Windowwithaccessories60/40cm

2

Woodenplank2"x4"(50mmx100mm)

6pieces

Plasticsheet 1piecesWoodpreservativeoil 5litres

When they came withthe vouchers, we bought amattress and sheets, and thispot. We never had a mattressbefore!

Now we have building

material... We have knockedthe old house down, and areusing the old and the newmaterial to build a biggerhouse.

Materials distribution and construction for the host family support programme

in urban environemnts in Goma

Bottom: The building on the right is the extension built during the programme.

Photos: Angela Rouse

The chef de quartier had put us on a list, and aftersome months the organisation came with many

questions. In April we got vouchers to buy food,and for mattresses, blankets and pots. We nowcook in our own house.We got more food vouchers in May, and last monthwe got building material to build an annex to thehouse. This is very nice and gives us our own space.We built it together in three days, but we still haveto put the floor in.


27/138

Eritrea - 1998 onwards - Conflict - Update - Camp upgrades Conflict / ComplexA.5

12

10,000 IDPs remainin camps

Support for returnbegins

60,000 IDPS remainin camps

Returns begin

Ethiopian troopswithdraw

Peace deal Truce Hundreds ofthousands ofpeople displaced.

First distributions

War betweenEthiopia and Eritreastarts

Eritrea - 1998 onwards - Conflict

Project type:NFI distribution

Camp support programFuel efficient stove project

Disaster:IDPs in camps in Eritreafollowing Eritrea / Ethiopiaconflict

Houses damaged:Estimated 100,000 homeswere destroyed in the war.

Number of people displaced:In 2000, around 1,000,000were displaced in Eritrea.

Project target population:Varied over time.The camp population in Gash-Barka, Debub and theRed Sea states region stabilisedto 60,000 by 2001.

Occupancy rate on handover:Varied over time.

Shelter sizeTents provided 16m2 ofcovered space. Some familieshad modified their shelters to

provide up to 40m2.

SummarySupport for variable population of Eritrean IDPs following conflict with Ethiopia. The agency in this case study

was the main provider of shelter and NFI assistance, providing IDPs with tents, tarpaulins and other non-fooditems such as stoves to those living in camps in the three areas of Gash Burka, Debub and Red Sea states. Theprovision of durable shelter items was not possible due to a political interests in ensuring that the camps weretemporary. As a result IDPs often adapted the emergency shelter items they received in order to improve theirliving conditions.

10 years

9 years

6 years

3 years

2 years

2 years

May 1998

Project timeline

Update - Camp upgrades

Ethiopia

Eritrea

Gash-BarkaDebub

S. Red Sea

A.5


2008 for more

UpdateBy mid-2008 Eritrea, officially, there were no conflict displaced people

in Eritrea. The government had resettled the last 11,000 living in campsin Debub. However, United Nations Development Programme in Eritreareported in January 2009 that an unspecified number of displaced peoplewere still living with host families.

In 2007 concerns were raised over the level of services provided inreturn areas and whether settlements would support returnee livelihoodsand one agency requested donor funding to support 10,000 returnees withemergency items and basic services.

The government of Eritrea has provided cash grants and assistancewith home-building and seed/livestock buying to some returnees. Some

families received two hectares of land and the regional administration haddisbursed over three million Nakfa ($200,000) by the end of 2008.

Over 60,000 people were living in

camps 6 years after the conflict

Photo: Joseph Ashmore


28/138


13

Gaza, Palestine - 2009 - Conflict

9 Programmes were able to adapt to the changingcontext.

9 Detailed assessments of 12,000 houses wereconducted in Gaza. There is now detailed damageassessment on the basis of which future payments canbe made.

9 By assessing apartments separately from the mainstructure of a building, those renting would also besupported by future cash payments.

9 All houses were assessed, including houses occupiedthe poorest families.

8 Because much of the support early in the response

had gone to families in collective centres and campsearly, it was difficult to encourage return.

8 No housing repairs were made as a result of thisprogram. This was due to an Israeli blockade onconstruction.- Due to lack of construction materials, the project

had to be stopped after finishing the cost assessment.- The cash component of the project that was

planned, was intended for the purpose of buildingrepair and construction. As construction could nothappen, no payments could be made.

A.6


Country:Gaza, Palestine

Disaster:Conflict Operation CastLead the war on Gaza.

Disaster date:December 27, 2008 to January18, 2009,

Number of houses damaged:60,000 shelters

Project target population:Over 12,000 assessments wereconducted and 8,947 houseswere real cases.

5,039 cases were deemed tobe eligible for the grant.29,420 persons had applied forcash assistance.

Occupancy rate on handover:Not applicable as there is nohandover

Shelter size:Variable cost paid per shelter- Average of 68,000USD perhouse paid for destroyedhouses, 14,750 for damaged

houses and 1,800 for minordamage to houses.

SummaryThe organisation implementing this project advised on the allocation of grants from families whose houseshad been damaged or destroyed by the invasion of Gaza. 12,000 assessments were carried out with 5,000found to be eligible from 29,000 applications. However, the blockade on Gaza meant that materials werenot available for families to rebuild their homes.

Project completion

Assessment process,phase 1 complete

Assessments start

Project implementa-tion start date withdesk review and dataentry

Early recovery andreconstruction planat Sharm El-Sheikh

War on Gaza ends

War on Gaza starts

13 months -

12 months -

6 months -

3.5 months-

2 months-

3 weeks-

December 272008-

Project timeline

Shelter assessments

Gaza strip

Israel



29/138

Gaza, Palestine - 2009 - Conflict - Shelter assessments Conflict / ComplexA.6

14

distributed to families, although aphysical shortage of money in Gazaslowed down initial distributions.

Cash assistance was the majorelement of the response to thedisaster. The de-facto govern-ment in Gaza handed out 4,000Euro to each family who had theirhomes destroyed, and The Pales-tinian National Authority throughthe United Nations Development

Programme handed out 5,000 USDto each family with a destroyedhome and 3,000 USD to eachfamily with major damage. Peoplewith less than 3,000 USD worth ofdamage received full compensa-tion.

The same process was carriedout for the refugees through theUnited Nations Relief and WorksAgency. by the end of the conflict,over 50,000 people had found

refuge in over 50 collective centres,many more had moved in withhost families. Following the endof conflict, the number of familiesin collective centres rapidly fell aspeople moved in with host families.

After the invasion, the Palestin-ian National Authority initiated ahousing rehabilitation and recon-struction program for all residentsaffected by the war on the GazaStrip. This included both thosedisplaced and those living on theiroriginal tract of land. The fundingwould be issued to home ownersby grants through Palestinian bankswhich operate in the Gaza Strip.

Families had to apply to thebanks to receive an amount ofmoney that could be dedicatedto rebuilding homes, or to con-structing new residences on legallyowned lands.

ImplementationThe organisation in this case

study had a technical advisory role.The ultimate authority for allocationof grants was held by a committee.This committee included the Pales-tinain National Authority, the Pal-estinian Monetary Authority andthe participating banks. The projectwas planned in two phases:

Phase 1: The compensationvalue would be calculatedwhich would be issued to homeowners in the form of grantsthrough Palestinian bankswhich operate in the Gaza Strip.

Phase 2: To monitor the

distribution of cash and serveas an advisor to the banks,authorising payments tobeneficiaries. This phase didnot happen as the blockadeprevented constructionmaterials from entering theGaza strip.The organisation reviewed ap-

proximately 29,000 grant applica-tions and assessed the homes of12,000 people. Assessment formswere entered into a database with

linked GPS data, and an overall costfor required repairs was computedfor each home.

Before the conflictThe Gaza strip is very densely

populated. Its current population is1.5 million with over 4000 peopleper square kilometre. It has a highrate of unemployment and as aresult poverty is pervasive. This wasexacerbated by the blockade onGaza, which started in June 2007.This blockade prohibits many itemsincluding building materials from

entering Gaza.In 2008, over 5,000 houses

were under construction throughinternationally supported projects.Projects in the housing estatesfor refugees from 1948 were notcomplete, and an estimated 20,000new housing units were needed inGaza each year to accommodatenatural growth. Additionally therewere refugees living in unsanitaryconditions in camps.

After the conflictFor 23 days starting on 27

December 2008, the Israeli Armycarried out a major militaryoperation in the Gaza Strip whichthey called Operation Cast Lead.The military incursion led to highlevels of damage to shelter, publicservices as well as economic in-frastructure. Blockades on goods,including cement, timber, steel,glass, and other construction

materials were still in place one yearafter the military action.

The conflict damaged ordestroyed 60,188 shelters ofwhich 10% (6,000 shelters) weredestroyed or required major repair.600,000MT of rubble needed to bedealt with.

The responseThe emergency response was

to distribute relief items. These

included plastic sheeting to coverwindows and damaged walls,kitchen sets, mattresses, blanketsand hygiene items. Cash was also

Destroyed buildings

Photo credits: CHF

Where buildings had many tennants

- different apartments were assessed

separately from the building

Photo credits: CHF


30/138


15

Repair costs for each homewere calculated through an agreedand transparent method. This wasbased upon an estimate for thecost to replace or repair each typeof damaged building element (suchas column, footing, slab, floor or

even a whole building). During as-sessments, detailed informationsuch as the volume of concrete,excavations, backfilling and steelrequired was recorded according topre-agreed reference tables.

Categories of damage Category 4 - totally destroyed,

or more than 70% of the homeis damaged

Category 3 value of destructiongreater than 5,000 USD

Category 1 or 2 - minor damageand the value of the destructionis below 5,000 USD.

Damage was further catego-rised into apartment damage anddamage to the common parts of abuilding. This was to enable tenantsof multi-storey structures to qualifyfor assistance.

Selection of beneficiariesFamilies had to apply through

the banks. Eligible families included Non refugee Palestinian citizens

in Gaza Strip whose buildingswere completely destroyedor who suffered from majordamage that made the houseunsuitable for living in, and whohad a house in category 4 and 3

Palestinian refugees livingoutside the refugee camps inGaza Strip. As of June 2010,the selection of these refugees

outside the camps and thevalue of their grants neededto be discussed between thePalestinian National Authorityand the United Nations Reliefand Works Agency.

Buildings had to have beenoccupied before the war.

Damage assessmentThree different damage assess-

ment methods were identified.Each had corresponding forms andpaperwork.

Category 1: repair is not feasible.Assessment teams must collect ad-ditional data such as area of thebuilding, the number of floors,

original drawings or photos of thebuilding and type of finish.

Category 2: damage is toocomplex. A specialist team isrequired to assess the damage. Thiswas most common for multi-storybuildings where there was damageto slabs or structure in lower floors.

Category 3: partial damage orrehabilitation is feasible. Three cat-egories were established: excessive,moderate or minor damage.

StaffingTo visit all of the 29,000 homes

in 9 months, a team of over 160skilled people was assembled. Thisis summarised below

no. role years

experience

96 SiteEngineers:CivilEngineersandArchitects

5years

9 RovingSupportEngineers(Electricaland

MechanicalEngineers

7years

16 SupervisingSiteEngineers(StructuralCivilEngineers

7years

5 ChiefEngineers(CivilEngineers)

10years

10 SocialWorkers(Councillortrainingbackground)

5years

8 OfficeEngineers(Civil,Architect,Electromechanical)

7years

20 Graduateengineerswhowerepairedwithmore

experiencedstaff.

graduateengineers

1 ProgramDeputyDirector(CivilEngineer)

15years

1 ProgramManager(InternationalExpert).

Surveyor Teams were estab-lished, each one including two siteengineers with a target of assessing3 to 5 housing units each day. EverySite Supervisor was responsible for3 surveyor teams.

Each Chief Engineer had

between 3 and 5 Site Supervisorsreporting to them. This meant thatthey reviewed between 45 and 75data collection sheets per day. ChiefEngineers took a random sample of5 data collection sheets from eachSite Supervisor for review each day.

Finally the data was approvedby the Programme Manager andProgramme Deputy Director andhanded to the banks.

PaymentThe intention was that once the

payment phase of the programmehad started, the owner of eachproperty would conduct their ownreconstruction. For this, they wouldbe paid a cash grant in installments.

However, after one year, con-struction still could not take placedue to the blockade on construc-tion materials into Gaza by theIsraeli authorities.

NOTE: One year later, themoney pledged at the Sharmel-Sheikh conference for the recon-struction of the Gaza Strip had notbeen handed over to the PalestinianNational Authority. There neededto be a political resolution betweenthe two different governments inPalestine and an end to the siegeby Israel before the donors wouldhand over the pledged money.

each building was visited by a team

Photo credits: CHF

Structural assessments requiredskilled engineers

Photo credits: CHF

A blockade on construction materials

prevented houses from bein built.

Photo credits: CHF


31/138

Georgia - 2008 - Conflict - Rural shelter construction Conflict / ComplexA.7

16

Project completion

Start construction ofone warm cottagefor returnees.

Policy change, fromrepairs to onewarm cottageconstruction

Start of rubbleremoval programmefirst communitymeetings to registerbeneficiaries

Adoption of onewarm room strategy

First draft of nationalshelter strategy

First field surveys

Ceasefire

First food distribu-tions

Conflict starts

Georgia - 2008 - Conflict

9 Support allowed returnee families to stay in their

homes during the harvest season, and during thewinter.9 The number of families having to stay in collective

centres was reduced.9 Forward preparation was made for full reconstruction

after the winter.9 The NGO showed great levels of adaptability to

changing government policies.9 The project made extensive use of beneficiary

contribution and input.9 One warm cottage provided a long-term solution

for those whose homes had suffered the most damage.9 Local markets and contractors were engaged.

9 Cottages were built that would be of use to familiesevan after they had ceased living in them.

8 Constant changes in government policy forced

shelter projects to adapt continuously.

8

One warm cottage used resources which couldhave been used for permanent repairs of originalhouses.

8 One warm cottage construction not as adaptableas initial one warm room repair strategy.

8 Limited size of 'one warm cottage' was not alwaysable to provide sufficient space for extended families.

8 Need for accelerated speed in construction ofcottages reduced potential for reconstruction ofimproved houses and technical knowledge transfer.- Targeting of the most severe levels of damage

ensured that those most in need of shelter weresupported, but the increased costs of doing so meant

that fewer households could be supported, and almostnone whose houses had suffered a mid-range ofdamage were given support.


Country:Georgia

Disaster:Conflict

Disaster date:8-12 August, 2008

Number of houses damaged:1,850 families (mainly single-family houses. some multi-unitapartment buildings)

Number of people displaced:120 000130 000

Project target population:Initially 5000 households. Laterreduced to 200 households.

Occupancy rate on handover:Initial occupancy rate 65%.Later increased to over 80%.

Shelter size:The materials distributed wereto repair houses of varyingsizes. The transitional sheltercottages were 24m2.

Materials cost per shelter:Varied for building repair.3000 USD for each wintercottage.Costs were higher for the Onewarm cottage.

7 months

3 months

10 weeks

6 weeks

1 month

20 days

4 days

2 days 8 August

2008

Project timeline

Rural shelter construction

SummarySupport of families whose homes had been damaged or destroyed during the conflict, in order that theycould stay in their homes during the first winter. Building repairs and then the provision of a one warmcottage was supplemented by distributions of NFIs and firewood.

Georgia

A.7



32/138


17

Before the conflictGeorgia had a pre-existing

displaced population of approxi-mately 200,000 people. Many hadbeen living in collective centres inurban areas since the conflict of

1991-1992.

Apart from a few families livingin apartments in the centres of thelargest villages, most families livedin stand-alone farm-houses. Oftenthese were shared between manygenerations of the same family.

Most of the houses weregrouped into small villages, andstood alone inside walled gardens.Most families still relied upon ag-

ricultural produce for their liveli-hoods to some degree, and mosthouses included storage rooms inthe bottom storey.

Houses built after the 1970swere more likely to be built inbreeze blocks. Almost all of thefamilies in the affected areas wereowners of their own homes.

Since 1990 there was a dramaticdecline in the local economy. Thisadded to the vulnerability of the

housing stock to conflict damage.

Temperatures in the affectedzones fall as low as minus 200

Celsius in winter. Houses in the areawere built under the Soviet regime,when energy was virtually free tousers, and as a result many had verypoor thermal insulation. For 97%of households firewood is the mainfuel for heating and cooking. Theaverage family consumes 7m3 ofwood during a winter.

Between the 8th and 12thAugust 2008, South Ossetia wasinvaded. Russian forces continued

26 kilometres further south. At theceasefire on 12 August, a bufferzone was declaed at the perimeterof the furthest advance. This zone

was occupied until October 2008.

After the conflictDuring the first three weeks of

the ceasefire armed militia gangsroved the villages inside the bufferzone south of South Ossetia. Oncethat threat diminished, a greaternumber of families from the villagesin the buffer zone started to returnhome.

By the second week of

September in some villages, 70%of the population were eitherpermanently returning home, orspending at least part of the timeback in their homes. The returnprocess coincided with the start ofthe harvest season.

A relatively small number ofhouses (only 5% of the total) hadbeen destroyed or heavily damaged.However, up to 2483 houses in the11 most heavily-damaged villages

had suffered sufficiently lightdamage that the families could stayin the houses over the winter.

In urban areas beyond thebuffer zone, greater strains werebecoming evident in the ad-hoccollective centres for those who hadbeen displaced and who could notreturn. There were also competingclaims for support from those newlydisplaced, and the older displacedpopulation from the 1991-2

conflict, as well as those fleeingfrom South Ossetia for whomreturn was impossible.

One warm room strategyWithin one month after the

disaster, the implementing or-ganisation had developed a one

warm room strategy, based uponprevious models from the Balkansin the 1990s. The most importantelement of this strategy was that itwould support those families whowanted to return to their housesof origin, and thus relieve pressureupon the collective centres in urbanareas like Tbilisi. It aimed to providesupport to the families who wereseeking to return home in time tosalvage their agricultural harvests.

The organisation also continuedto support people that weredisplaced into urban areas throughthe distribution of firewood andnon food items.

One warm roomTrained staff would assess

the levels of damage, and thenengineers would draw up Bills ofQuantities for those houses whererepairs needed more than plastic

sheeting or other minor items. Avoucher system would be set upwith local suppliers in Gori, the pro-vincial centre just south of the bufferzone. This would support the localeconomy and ensure that as wide arange as possible of materials wasavailable.

Housing damage was assessedon a scale of 1 to 5, based on similarscales used in the Balkans. For largerhouses, there was the possibility of

providing sufficient materials toprevent further damage to the restof the house during the winter.

The housing strategy shifted from one warm room to one warm cottage. As

a result the anticipated scale of the programme was reduced

Photo: Jonathan Puddifoot


33/138

Georgia - 2008 - Conflict - Rural shelter construction Conflict / ComplexA.7

18

A cost limit per house wasimposed for each category, Thiswas to ensure equitability betweenhouseholds. Whilst this approachwould be sufficient for those whosehouses had been merely damaged,neither the time nor the budget

constraints would have permittedthe re-construction of an entirewarm room in those houses whichhad been wholly destroyed.

One warm cottageOn October 22nd, initial en-

gagement with the affected com-munities was underway, the Gov-ernment of Georgia changedpolicy: primarily destroyed houses(category 5) would be targeted.These families would be given a

24m2 cottage, constructed by theNGOs. This solved the issue of howto support those with destroyedhouses, but reduced the number ofbeneficiary households supportedby the NGO to 200.

Selection of beneficiariesInitial surveys had identified the

villages which had suffered themost damage. Village leaders wereapproached, and asked to organisea meeting between the NGO and all

members of the community whosehouses had been damaged. At themeeting, families were registered,and asked to evaluate the level ofdamage of their houses.

During the initial repair part ofthe strategy, support was offeredto families according to levels ofdamage. Extra help for both rubbleremoval and repairs was offeredto those whose vulnerabilitiesprevented them from doing thiswork themselves.

When the strategy changedtowards the construction of aone-room cottage, criteria changed.All families in the target villageswhose houses had been assessedas being Category 5, or completelydestroyed, were then included.

Technical solutionsInitially, the proposal was for

a supply of materials, based upon

individual bills of quantities writtenby staff engineers. These wouldprovide each family with at leastone warm room for the winter in

their house of origin, and would bethe start of the full reconstructionafter the winter.

For those whose houses hadsuffered minor damage (typically,broken windows or roofing tiles)there would be a direct distribu-

tion of plastic sheeting. For highercategories of damage, a voucherscheme was planned, based ona market assessment, the limitedlogistics resources for direct delivery,and traffic limits in the buffer zone.

After the change in shelterstrategy by the government, localcontractors were engaged to buildthe 200 cottages for those familieswhose houses had been totallydestroyed, or damaged beyond

repair. The cottages were builtusing breeze-blocks and timber-and geo-textile roofs. There waslittle ground insulation. Buildingshad a ceiling to improve thermalcomfort.

Cottages were sized to respectinternational standards, whilst stillhaving enough room to actuallydo the construction in the limitedspaces of beneficiaries gardens.

The government made cashtransfers of around 15,000USD tofamilies whose houses had beencompletely destroyed. However,due to lack of experience andsupport, much of this money wasnot spent on rebuilding houses.

Household energyIt was agreed to supply 3m3

of firewood to support affectedfamilies with their heating andcooking needs. The organisation

delivered around 24,500m3 offirewod to around 8,500 House-holds, over two winters.

The organisation supplied US-AID-approved fuel-efficient wood

buring stoves to all cottages. Italso supplied 5,952 cooking gascylinders and 600 electric waterheaters. Glass fibre insulation wasprovided to reduce heating costs.

Trials on woodchip briquettes asan alternative fuel found them not

to be inappropriate as they werevery sensitive to damp.

Logistics and materialsPlastic sheeting and firewood

were provided using rented trucks.

For the second, one warmcottage strategy, the contractorswere responsible for their ownmaterials supply.

To reduce the risk of causingsignificant deforestation the or-ganisation only bought woodfrom licensed suppliers, with par-ticular criteria such as diameterand species type. Unfortunately,the large scale purchases distortedthe markets. Supply licenses weresuddenly revoked by the govern-ment and only a very few supplierswere able to obtain certification.

In general, although NGO accessinto the buffer zone was limiteduntil October 2008, local Georgianswere allowed to drive trucks intothe area from a much earlier date,and after the ceasefire of 12 August2008 transport on the nationalhighways and from other countrieswas relatively unimpeded.

Materials listMaterials for one warm cottage

(excluding electrical installation)

Material Quantity

Cement 3.36MT

Gravel 6.325m3

Ironbars12mm 102.4m

Ironbars6mm 72m

Mineralwool-roll 1

Nails 8Kg

Roofingnails 200pieces

Plasticboards12.5cm 22.5m2

Plasticdoorblock 1.89m2

Plasticwindowblocks 3m2

Rooftrim 8m

Roofsheets 25pieces

Sand 4.6m3

Smallblocks20x20x40cm 665pieces

Timberbeamsandplanks 3.5m3

Woodenskirting 18.4m

One of 200 one warm cottages

built by the programme,

Photo: Jonathan Puddifoot


34/138


19

End of the project