USING SIMULATION TO EVALUATE A BLOOD SUPPLY NETWORK IN ATLANTIC CANADA
John Blake
Department of Industrial Engineering
Dalhousie University
& Canadian Blood Services
john.blake@dal.ca
Matthew Hardy
Department of Industrial Engineering
Dalhousie University
& Canadian Blood Services
Keywords: Blood system management, simulation, logistics, inventory modelling
Background
Canadian Blood Services (CBS) has developed a consolidation plan for blood component production, distribution and support functions for the Maritime Region of Atlantic Canada to be located in Dartmouth, Nova Scotia. Once consolidated, blood products for New Brunswick hospitals will be distributed from the new Dartmouth production facility either directly or via a stock holding unit (SHU) located in Saint John, New Brunswick. A number of stakeholders have expressed concern about the timeliness of deliveries from Dartmouth to either Saint John or to hospitals in the north-east and north-west of New Brunswick. However, CBS has committed to a standard of service under the new network that is at least as good as that of the existing network.
The current logistics distribution network for blood and blood products to New Brunswick hospitals uses ground distribution from the Saint John facility for regularly scheduled runs, as well as ad hoc and emergency deliveries to hospitals. The proposed logistics infrastructure, once consolidation of production activities to the new Dartmouth facility has occurred, will utilize existing modes of ground transportation as well as air deliveries for hospitals in north-east and north-west New Brunswick. CBS will service some New Brunswick hospitals from the Dartmouth facility or the Saint John stock holding unit. Transport of raw and finished products between Dartmouth and Saint John will be provided by a dedicated CBS ground transport service.
Objectives
This papers describes the results of a simulation experiment to evaluate and compare expected service levels for customers supported through Saint John under the “current” and “to be” distribution networks.
Approach
There is a rich literature describing the application of simulation to blood centres and hospital blood banks (see for example Katsaliaki and Brailsford (2007)). However, we believe that this is the first time simulation has been specifically used to evaluate the impact of network logistics on customer service levels in a blood supply chain.
A simulation model of the Saint John facility was constructed in Rockwell Arena Version 13. The simulation was designed to model the “current” situation where collections are processed at Saint John as well as the “to be” situation where collections are processed at the proposed Dartmouth facility and delivered to Saint John via ground transport. The simulation model is able to reproduce processes of inventory management: collections, testing, end-labelling, and distribution and to represent order arrivals, order backlogs, and order completion and dispatch. An MS-Access database is used to simulate the CBS production database and to track the physical location and status of all units in the simulation model.
Results
System performance metrics (order fill time and outdate rates) were recorded for both the current and proposed distribution networks using the simulation model as a test platform. A comparison of means tests (α = 0.05) indicates that the proposed network has performance metrics that meet, or exceed, those of the existing network.
Conclusions
Levels of customer service experienced by NB hospitals will not be degraded by consolidation of production and support facilities in Dartmouth.
References
- Katsaliaki K, Brailsford S. Using simulation to improve the blood supply chain. Journal of the Operational Research Society 2007;58(2):219-227.