Feasibility Study for CAREShelter First Apr 13 2006

Feasibility Study for CARE-Shelter First Apr 13, 2006 Thomas Choi, Shao Hung Goh, Marco Gutierriz, Randy May, Zhao Ki So, Kai Wang 1

Outline n n Introduction CARE Shelter First Objectives Challenges Deliverables n Distribution Network for Disaster Relief Warehouse Location Model and Cost Estimates n Case Studies n n Conclusions 2

Relief Aid can be subdivided into 2 categories: • Humanitarian Aid • • • Provided on an urgent basis in response to a humanitarian crisis Objective is to save lives, alleviate suffering, and maintain human dignity. Development Aid • Address socioeconomic factors that may have led to a crisis 3

Aid Organizations • Subdivided into GOs and NGOs. • Government organizations • • • Non-governmental organizations • • • Within government (USAID) International level (United Nations) Local International (CARE, Shelter First) Funded by individuals, corporations, governments, and other organizations Some aid organizations carry out both kinds of aid (CARE) Some organizations specialize (Red Cross and Shelter First, humanitarian aid) 4

Humanitarian Aid Trends • Funding and delivery increasingly being organized at an international level. • • • Faster and more effective response for major emergencies. International NGOs becoming an increasingly effective vehicle for delivery. Standardization and improvement efforts • SPHERE minimum standards for disaster response. • Although mostly qualitative, provides guidelines for effective and efficient response. 5

Humanitarian Aid NGOs – Challenges • • • Funded in response to emergencies Little, if any, funds are dedicated to preparedness (including pre-positioning) Typically NGOs must wait until after emergency to source supplies. Response time is sometimes too slow causing additional deaths and sufferings 6

Humanitarian Aid NGOs – Response Chain 7

Shelter First Concept • • • To provide immediate short-term shelter to victims of natural disasters worldwide. Purchase and pre-stage temporary shelters at various points around the world. Start with a single warehouse. Expand over time to multiple locations. If possible, airdrop shelters directly to affected area. If not, fly shelters to closest airstrip. The ultimate goal is to airdrop shelters to destination within 24 hrs after a request. 8

Perceived Value • • • Provide a vehicle for positive publicity for the hospitality industry worldwide. Focus on temporary shelters, as opposed to the other so many needs that aid organizations face. Most importantly, reduce response time and reduce human suffering worldwide. • • Initially, reduce shelter provision response time from 2 -3 wks to 5 -7 days. Ultimately achieve 1 day response time goal. 9

Shelter Design • • • Shelter design must meet SPHERE minimum standards: adequate for climate, lighting, ventilation, places to store food, water Estimated cost of $500 per shelter. Only standard that raises an issue is size. • • Minimum of 17. 5 sqm for a 5 member family. Might make shelter kit too large and more expensive to transport. 10

Kit Contents • • • Contents should address basic needs Should last until other aid reaches the beneficiaries (3 -4 wks) Minimum contents: shelter, blankets, food rations, water purification kits, first -aid kits 11

Packaging Ideal Design • • • Estimated weight 200 lbs. Size of standard pallet. Slots forklifts’ forks. Disassembled into 4 colorcoded, 50 lbs “boxes”. Each separate box can be safely handled by a single person. Rationale: take advantage of both worlds (commercial and humanitarian), avoid need of pallet. 12

Outline n n Introduction CARE Shelter First Objectives Challenges Deliverables n Distribution Network for Disaster Relief Warehouse Location Model and Cost Estimates n Case Studies n n Conclusions 13

Challenges n n n Identification of issues in Relief logistics for Shelter First Initiative Different focus by CARE and Shelter First Feasibility of One-day response time target Data collection Pre-positioning of possible warehouse positions 14

Outline n n Introduction CARE Shelter First Objectives Challenges Deliverables n Distribution Network for Disaster Relief Warehouse Location Model and Cost Estimates n Case Studies n n Conclusions 15

Distribution Network • 3 stages: • • • Supplier Manufacturing sources to Warehouses to Staging Area to Beneficiaries 500 to 2000 miles Warehouse Staging Area ~ 100 miles Beneficiaries (Disaster Zone) 16

Deliverables: Network Strategy Existing Timeline of Shelter Delivery by CARE Mobilization and Delivery to Beneficiaries Procurement (7 days) (21 days) Proposed Timeline of Shelter Delivery by CARE Mobilization Long Haul Transit Last “ 100 -Mile” (3 days) (1 day) 17

Distribution Network • From Manufacturing sources to warehouses • • Delivery of shelters does not have to be expedited (Push to stock) Stocking at warehouses can be done through economical means such as trucks, rail, sea-freight, air, etc 18

Distribution Network • From Warehouses to Staging Area: • • • Expedited delivery by commercial express air carrier Strike of Disaster Report on damage assessment Notify commercial express air carrier Warehouses prepare shelters for shipping Staging area is the nearest airport for regular aircrafts 19

Distribution Network • • From Staging Area to Beneficiaries (“ 100 mile”): Handoff to beneficiaries within 24 hrs • • • If air-drop or air-lift is feasible, fly from Staging Area to Beneficiaries If airdrop is infeasible, use any transportation mode available for final distribution Modes of transportation • • Animal carts/animals Boats Helicopters Pick-up trucks, SUV’s, small trucks 20

Distribution Network 21

Outline n n Introduction CARE Shelter First Objectives Challenges Deliverables n n Distribution Network for Disaster Relief Warehouse Location Model and Cost Estimates Case Studies Conclusions 22

Model Rationale • • Requires fast response time to unpredictable (unscheduled) demands Basic strategy: • • Stockpile inventories “near” possible disaster zones Activation of outbound transportation at short notice 23

Warehouses: Centralized vs Decentralized Decentralization Centralization Inventory High Low Response lead time Low High Overhead cost High Low Inbound transportation cost High Low Outbound transportation cost Low High For a given response lead time and a given budget, balance the tradeoffs 24

Warehouse Location Model • • • Provide a staged plan based on projected budget growth. Objective: Maximize coverage Decision variables: • • • Number of warehouses. Location/s of warehouse/s. Inventory level/s. Constraint: budget constraint. Inputs: • • • Possible demand points, magnitudes and frequencies. Candidate locations for warehouses. Costs. 25

Budget Constraint • Projected budget for purchase of shelters, transportation, warehousing (excluding admin overhead): • • Year 1: 2: 3: 4: $1. 2 $2. 4 $4. 8 $6. 0 million Assume half of the funds after year 1 will be used for replenishment. The goal for each warehouse is to have 10, 000 shelters on hand. Estimated costs only for shelter. Cost estimates for additional kits to be determined. 26

Cost of Typical Airdrop Operation • • • C-130 - $3, 500 / hour of charter 500 shelters – 7 flights 1000 mi trip – 3 hr flight 2 hrs loading Approx $185, 500 air carrier $371 transport / shelter $0. 371 / mile / shelter Total shelter cost: $500 + $371 Grand total: $435, 500 27

Model Formulation Cost components Budget constraint Inventory balance Inventory adequacy Tpt budget adequacy Warehouse fixity 28

Demand Estimate • • • Collected historical information on the number of people displaced because of natural disasters around the world Although historical information of disasters can’t predict future events, the analysis provides an idea of frequent geographic locations and expected demand Aggregating of expected demand across many locations 29

Candidate Locations • Carried out analysis based on: • • • Connectivity (port and airport volumes) Cost of living Ease of doing business Political Stability First-cut shortlist of 80, narrowed down to 20. 30

Candidate Locations 31

Candidate Locations • Based on input from CARE, the following list is very much preferred: • • Panama Dubai (UAE) Mombassa (Kenya) Mumbai (India) Hong Kong Johannesburg (South Africa) Reasons for preference: • • Existing suppliers near these cities Established transportation network 32

Results from Model n n n First warehouse to open: Mumbai, India Year 3, open warehouse in Hong Kong Total shelters delivered over 8 years = 48122 “Steady state” inventory = annual utilization – 2000 in Mumbai & 6500 in Hong Kong Cost structure over 8 years n n Setup: $350 K Operating: $2. 3 M Purchasing: $28. 2 M ($4. 2 M in inventory) Transporting: $7. 5 M 33

Sensitivity Analysis n 2 shelters per pallet n n 4 pallets per pallet n n Increase in tpt costs: $100 K/year or ~13% 20% increase in warehousing costs n n 18% increase in coverage (56886) 20% increase in aircraft charter rates n n 11% increase in coverage (53249) 1. 2% decrease in coverage (47563) 50% in setup costs n 0. 5% decrease in coverage (47877) 34

Sensitivity Analysis n Half shelter cost, a quarter of volume n n 112% increase in coverage (101874) 10 times annual budget n Order of warehouse opening n n Year 1: 2: 3: 4: Mumbai & Hong Kong Dubai Panama & Mombasa Johannesburg 35

Outline n n Introduction CARE Shelter First Objectives Challenges Deliverables n n Distribution Network for Disaster Relief Warehouse Location Model and Cost Estimates Case Studies Conclusions 36

Flood in Vietnam n n Disaster strike on Dec 04 2005 and concluded on Dec 24 2005 Influence to the residents: • • • Affected region: 63, 090 sq km Damage: $ 27 Million Dead: 62 Displaced: 20, 000 Homeless: 18, 000 Infrastructure of houses: 171 destroyed, 335 damaged, 5300 inundated… 37

Flood in Vietnam n Affected areas: 38

Flood in Vietnam n Influence on the transportation: • 49 bridges damaged • Traffic backed up to 30 km as landslides cut off 2 sections of highway • Emergency bypasses to two sections of Highway 1 were built in Phu Yen Province 39

Flood in Vietnam n Local Government’s Reaction n n Displace the affected residents to the nearest higher land for sheltering Provide limited quantity of relief material for short term consumption 40

Flood in Vietnam n NGOs’ Reaction to the Flood n n n Assessment and report of the damage and the number of people affected Estimation of the demand of the relief material Coordination in the provision of the relief material Mobilize the logistics operation provided by the commercial service providers Usually the arrival of the first batch of relief material is 4 to 5 weeks after the disaster 41

Flood in Vietnam n You cannot do an act of kindness too soon, for you never know how soon it will be too late… 42

Flood in Vietnam n Shelter First’s Goal: Shelters reach the victims who need them most within 3 days after the occurrence of the disaster 43

Flood in Vietnam n Proposed response by Shelter First Initiative Mobilization Long Haul Transit Last “ 100 -Mile” (3 days) (1 day) 2320 miles by UPS Warehouse in Mumbai 200 miles by airlift or air-drop Staging Area at Tan Son Disaster Nhut Airport at Ho Chi Minh Zone 44 City

Flood in Vietnam n Proposed reactions by Shelter First Initiative n n n Corrected! n n n 18, 000 homeless 3, 600 shelters needed since 5 person/shelter Shelter First’s 10% commitment of 360 shelters 54, 000 pounds at 150 pounds each or volume of ~23000 ft 3 (cube governs) 5 entire Airbus A 300 payload (89, 000 pounds or ~70 pallets) Need ~ 5 chartered flights Leg 1: Transportation Cost ~$250, 000 Leg 2: Airdrop Cost ~$50, 000 Total ~$300, 000 or $0. 33 per shelter per mile 45

Flood in Vietnam n Distribution to beneficiaries n n CARE staff receive the shelters from trucks or boats and distribute them to the local beneficiaries Alternatives strategy for last “ 100 -mile” n Local military aircraft if available 46

Conclusions • • • 1 -day response near impossible Significant costs in air-dropping Costs most responsive to volume per shelter Most critical to open warehouses in Mumbai and Hong Kong Potential savings from using 2 -legged air freight Great uncertainty in last “ 100 -miles” transport modes and transit times 47

Thank you • Questions? 48