Improving the blood supply chain: simulation and optimisation models to support collection, production and location-allocation decisions

This thesis introduces and studies different problems in the blood supply chain. The problems are focused on aspects less frequently studied in the literature such as the exploitation of the different collection and production alternatives, consideration of multiple products and uncertainty in demand and supply. These important features can be found in different decision levels, including daily collections, annual planning and at the strategic level when the blood supply chain is designed. For each problem presented, a suitable solution strategy is proposed. Different methods such as discrete event simulation, Monte Carlo simulation, optimisation, stochastic optimisation and multi-objective optimisation have been used to provide solutions to the problems studied. A simulation-optimisation model to support collection and production decisions in the blood supply chain is first presented. A model which integrated discrete event simulation and integer linear programming was designed to solve this problem. The model is tested using data from a blood centre in Colombia. Results show that key performance indicators such as total cost, number of donors, shortage and outdated units are improved by using the approach proposed. In addition, a stochastic multi-objective optimisation model to study the trade-off between cost and number of donors required is also included in this thesis. This model supports the decision of number of donors required by using whole blood and aphaeresis collection processes as well as considering the different blood groups and two main objectives: minimisation of cost and donors. The problem is solved using a combination of the augmented epsilon-constraint algorithm and the sample average approximation technique. A Pareto front considering stochastic demand is obtained by applying the proposed method. The final model included studies the optimal design of a blood supply chain as well as a discussion about the main motivations for centralised and decentralized systems. A stochastic mixed integer linear programming model is proposed and a solution method based on the sample average approximation technique is designed to address the problem. The complete approach is applied to a case study and several scenarios are generated to evaluate different travel time policies as well as the impact of using aphaeresis processes