Simulation Model - X-CD System Conference Management

Report
Planning Tools for Swine Health
Emergency Response
David Cornejo and Haleh Byrne
Ph.D. Candidate, North Carolina State University
NCSU Industrial Extension Service
Team members
• Presentation Co-Authors:
• David Cornejo- NCSU Operations Research
• Haleh Byrne-NCSU Industrial Extension Service
• Animal Science/Engineering Collaborators:
• William Morrow –NCSU Animal Science
• Larry Stikeleather- NCSU Bio & Ag Engineering
• Craig Baird –NCSU Bio & Ag Engineering
• Robert Meyer – Mississippi State, College of Veterinary Medicine
• Darrel Styles – USDA
Funding Acknowledgement:
USDA APHIS cooperative agreement and a Department of
Homeland Security Science and Technology Division Interagency
Agreement (System to Administer Inhaled Gases for Mass
Depopulation of Swine in a National Emergency, Agency Reference
Number: 09-9137-1280-CA)
Horia Varlan - Creative Commons, Flickr
2
Problem Statement
• Major outbreaks of swine diseases are not
common, but are economically devastating.
• Containment requires mass euthanasia.
• Effective, humane euthanasia practices can
be developed and tested on a small scale.
• Simulation can estimate the resources
required to scale-up process to disaster
response proportions.
3
Outline
• Background on swine health epidemics.
• Summarize new, humane euthanization
method developed.
• Present simulation model developed to
scale new techniques.
• Demonstrate how model was utilized to
develop strategic and tactical decision
aid.
4
Swine Health
• Swine diseases
• Foot and Mouth, Swine Flu
• Aersolisable=Spread though the air
• While pigs are alive generating more
aerosolized vectors.
• Mass Depopulation allows containment.
• Timely depopulation reduces total
morbidly and economic cost.
5
Effective Depopulation
• Primary Goal: Minimize Time
• Minimize
• Resources required
• Animal handling
• Ensure proper disposal
6
New Depop Procedure
1.
2.
3.
Swine loaded onto trucks.
CO2 gas pumped into containment.
Burial
• Process tested and evaluated measured on
small scale at research farm.
Reference: Robert E. Meyer, W.E. Morgan Morrow, Larry F. Stikeleather, Craig L. Baird, J. Mark
Rice, Haleh Byrne, Burt V. Halbert, Darrel K. Styles. Administration of carbon dioxide for onsite mass depopulation of swine in response to animal health emergencies. Journal of the
American Veterinary Medical Association 2014; 244(8): 924-933
• Create Simulation Model to determine
resource requirements on disaster scale.
7
Simulation Model
Jobs, Innovation, Growth, Stability
www.ies.ncsu.edu 8
Mass Depop Process
• 4 Distinct Stages
• Moving pigs from house to loading slot
• Loading pigs onto available farm truck
• Preparing trucks and gasification
• Disposal of carcasses
• METRIC: Time to complete process.
9
Moving Pigs from house to truck
• Influenced by following user inputs
• Pigs: Number, Weight class
• Houses: Number, Size, Loading Slot length
• Workers: Number available
• Variables affect Travel Time of run.
• Calculated travel time parameterized travel
time distribution(stochastic travel time).
10
Using Time-in-Motion for Data
Collection
• Walk speed data was collected using a
Time-in-motion study.
• Timer Pro was utilized to collect data
from videos of truck loading operations.
• Multiple observations were used to
develop simulation distribution
parameters.
11
Data that supports calculations
• Weight classes
Class Name
Nursing
Weaning
Finishing
Breeding Boar/Sows
Weight Range (lbs)
≤70
70-100
≥100
≥200
• Run Size/Walk Speed
Class Name
Run Size
1 Handler Rate
(ft/min)
Nursing
Weening
Finishing
Breeding
Boar/Sows
21
9
5
1
107.59
209.975
240.385
625.734
12
Load Truck
• Pigs Loaded onto truck
• Capacity of trucks from pig weight class
Class Name
Nursing
Weaning
Finishing
Breeding Boar/Sows
Space Requirement(sq. ft./pig)
1.09
2.635
3.48
5.55
• Assumed 15 seconds to load each pig
• Details
• One or two loading slots available
• Handler will wait with run until truck arrives
13
Preparing Trucks for Gas
• Before each application of gas some
preparation (check tarp, attach hose etc.)
• Assumed to take between 5 and 10 mins
with a mode of 10 mins
• May physically take place at gas location.
14
Administer Gas
• Applied for minimum of 5 minutes. May
take longer.
• Applying gas consumes available stock.
• Process halted when no more gas
• Gas can be delivered on intervals.
• Sublimation a way to extend gas stock
15
Dispose of Carcasses
• “Dwell Time” of 10 minutes included in
minimum travel time to Disposal Site.
• Assumes can travel immediately after gas
• No particular time allocated to disposal
• Trucks take time to return to loading slot
16
Statistics Monitored
• Time To Terminate: How long to cycle all pigs
• Average Walk Time: Average Time spent with
Handler(until loaded on truck)
• Accounts for walk time and queuing for truck
• Average Time in Gas Process: From starting
to load until disposal.
• Accounts for queuing at constrained gas resources
17
Parameters and Control Variables
• Farm Parameters
• Size of House
• Size of Pin
• Number/Length of Load Slots
• Truck Controls
• Truck Size
• Number of Trucks
• Handler Controls
• Number of Handler
• Handler Efficiency
• Special Process Parameters
• Dry Ice recovery process option
18
Planning Tools using Simulation
Jobs, Innovation, Growth, Stability
www.ies.ncsu.edu 19
Simulation Model to Practical Tool
• Simulation Model can optimize handler
worker combination given number of trucks
for a specific scenario.
• Disaster planners want to get estimates for a
wide array of planning scenarios or on the fly
“what-if” scenarios.
• Most interested in obtaining resource
estimates to complete the operation in a
specified time window.
• Farmers do not want to interact with the
simulation model.
Web App Solution
• Best resource solution form a large array of
scenarios batch run in simulation model.
• Results stored in database.
• Web application exposes relevant controls to
decsionmaker and calculates resource
requirements based on stored simulation
results.
• Web app runs fast (seconds) and requires no
specialized software.
21
Web App Screen Shot
22
Web App Output Page
23
Conclusion
• Swine health disasters deserve advance
planning.
• Simulation provided a viable method for
estimating the requirements when scaling up a
new process.
• Data was collected through careful studies of
portions of current processes.
• Results of simulation model were encapsulated
in web application.
• Web application requires no special software or
training to use.
24
Questions?
Jobs, Innovation, Growth, Stability
www.ies.ncsu.edu 25

similar documents