Feasibility Study for CARE-Shelter First

1
Feasibility Study for CARE-Shelter First

• Apr 13, 2006

Thomas Choi, Shao Hung Goh, Marco Gutierriz,
Randy May, Zhao Ki So, Kai Wang
2
Outline

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

3
Relief Aid

• 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

4
Aid Organizations

• Subdivided into GOs and NGOs.
• Government organizations
• Within government (USAID)
• International level (United Nations)
• Non-governmental organizations
• 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)

5
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.

6
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

7
Humanitarian Aid NGOs Response Chain
8
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.

9
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.

10
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.

11
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

12
Packaging Ideal Design

• Estimated weight 200lbs.
• Size of standard pallet.
• Slots for forklifts forks.
• Disassembled into 4 color-coded, 50lbs 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.

13
Outline

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

14
Challenges

• 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

15
Outline

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

16
Distribution Network

• 3 stages
• Manufacturing sources to Warehouses
• Warehouses to Staging Area
• Staging Area to Beneficiaries

500 to 2000 miles
100 miles
Staging Area
Supplier
Warehouse
Beneficiaries (Disaster Zone)
17
Deliverables Network Strategy
Existing Timeline of Shelter Delivery by CARE
Mobilization and Procurement (21 days)
Delivery to Beneficiaries (7 days)
Proposed Timeline of Shelter Delivery by CARE
Long Haul Transit (1 day)
Last 100-Mile (1 day)
Mobilization (3 days)
18
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

19
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

20
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, SUVs, small trucks

21
Distribution Network
22
Outline

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

23
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

24
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
25
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.

26
Budget Constraint

• Projected budget for purchase of shelters,
  transportation, warehousing (excluding admin
  overhead)
• Year 1 1.2 million
• Year 2 2.4 million
• Year 3 4.8 million
• Year 4 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.

27
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

28
Model Formulation
Cost components
Budget constraint
Inventory balance
Inventory adequacy
Tpt budget adequacy
Warehouse fixity
29
Demand Estimate

• Collected historical information on the number of
  people displaced because of natural disasters
  around the world
• Although historical information of disasters
  cant predict future events, the analysis
  provides an idea of frequent geographic locations
  and expected demand
• Aggregating of expected demand across many
  locations

30
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.

31
Candidate Locations
32
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

33
Results from Model

• 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
• Setup 350K
• Operating 2.3M
• Purchasing 28.2M (4.2M in inventory)
• Transporting 7.5M

34
Sensitivity Analysis

• 2 shelters per pallet
• 11 increase in coverage (53249)
• 4 pallets per pallet
• 18 increase in coverage (56886)
• 20 increase in aircraft charter rates
• Increase in tpt costs 100K/year or 13
• 20 increase in warehousing costs
• 1.2 decrease in coverage (47563)
• 50 in setup costs
• 0.5 decrease in coverage (47877)

35
Sensitivity Analysis

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

36
Outline

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

37
Flood in Vietnam

• 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

38
Flood in Vietnam

• Affected areas

39
Flood in Vietnam

• 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

40
Flood in Vietnam

• Local Governments Reaction
• Displace the affected residents to the nearest
  higher land for sheltering
• Provide limited quantity of relief material for
  short term consumption

41
Flood in Vietnam

• NGOs Reaction to the Flood
• 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

42
Flood in Vietnam

• You cannot do an act of kindness too soon, for
  you never know how soon it will be too late

43
Flood in Vietnam

• Shelter Firsts Goal
• Shelters reach the victims who need them most
  within 3 days after the occurrence of the disaster

44
Flood in Vietnam

• Proposed response by Shelter First Initiative

Long Haul Transit (1 day)
Last 100-Mile (1 day)
Mobilization (3 days)
200 miles by air-lift or air-drop
2320 miles by UPS
Staging Area at Tan Son Nhut Airport at Ho Chi
Minh City
Warehouse in Mumbai
Disaster Zone
45
Flood in Vietnam

• Proposed reactions by Shelter First Initiative
• 18,000 homeless
• 3,600 shelters needed since 5 person/shelter
• Shelter Firsts 10 commitment of 360 shelters
• 54,000 pounds at 150 pounds each or volume of
  23000 ft3 (cube governs)
• 5 entire Airbus A300 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

Corrected!
46
Flood in Vietnam

• Distribution to beneficiaries
• CARE staff receive the shelters from trucks or
  boats and distribute them to the local
  beneficiaries
• Alternatives strategy for last 100-mile
• Local military aircraft if available

47
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

48
Thank you

• Questions?