University of Twente Student Theses

Login

Life cycle cost optimisation: integrating spare parts with level of repair analysis

Smit, Martijn (2009) Life cycle cost optimisation: integrating spare parts with level of repair analysis.

[img] PDF
1MB
Abstract:Research motive Thales Nederland is a major developer and manufacturer of expensive defence related systems, such as radar equipment. Radar equipment requires maintenance and support in order to achieve sufficient operational system availability. In case of expensive technical systems, such as radar systems, the life cycle costs play an important role. The objective of the department Logistic Engineering (LE) is to provide customers with a maintenance plan such that the total life cycle cost are minimal. The current life cycle cost optimisation approach that Thales uses is not optimal. Problem statement Currently, Thales LE uses two optimisation techniques in order to minimise the total life cycle cost. First, they perform a level of repair analysis (LORA) to identify efficient repair policies and subsequently they use a spare part optimisation to optimise the spare part stock levels for the given LORA solution. LORA decides whether it is efficient to repair or discard a component upon failure. If a component is repaired, LORA also decides where it must be repaired in the repair network. The repair actions and locations of all components are used as input for the spare part optimisation. The spare part optimisation determines where spare parts must be stocked in the repair network such that the total spare part costs are minimal and the requested system availability is achieved. With this approach the requested system availability is achieved and for the given LORA solution the spare part costs are minimal. However, the total life cycle costs are not minimised because LORA does not take the costs of spare parts into consideration. Alternative solutions exist in which the costs for LORA are higher but the costs for spare parts are significantly lower. In this case, the total life cycle costs can be significantly lower. The current optimisation approach uses a simplified model for LORA, it requires a lot of time and concentration, it is sensitive to mistakes and errors, and the optimisation is not reproducible since it is no formalised method. Approach In the literature, no integrated model is available that minimises the costs for LORA and spare parts in one optimisation. Therefore we use an iterative optimisation approach where we use the spare part costs that come out of a METRIC optimisation as fixed costs in the LORA model of Basten et al. (2008-b). METRIC only calculates the spare part costs for the repair decisions that are part of the LORA solution. However, it does not calculate the spare part costs for the repair decisions that are not part of the LORA solution. After every iteration, we save the spare part costs that come out of the METRIC optimisation to a database. With this approach we build up a database that contains spare part costs for every efficient repair decision that is available for components in the system. At the end of the iteration process, the optimisation approach can make better decisions because the spare part costs are known for all efficient repair decisions. This approach however, has two drawbacks. It does not account for the component interaction effect of the spare part optimisation and the spare part costs in the database can be overestimated. We offer a solution for the last drawback. The optimisation approach has been tested on three cases. The first case reflects the actual situation of a customer of Thales. The other two cases further test the optimisation approach. The case study is based on the Variant 4 radar system and the logistical data that is used is based on existing project data. Thales LE provided solutions for the first two cases. We used the results from Thales LE as a benchmark for the optimisation approach. Results The test results of the case study show that a significant cost reduction can be achieved. For Case 1, a total cost reduction of € 16 million (17%) can be achieved compared to the solution of Thales LE. For Case 2, a cost reduction of €21.6 million (24%) can be achieved. The results of the optimisation approach show that more components are repaired downstream in the repair network. With downstream repair many repair locations need test equipment and therefore the total costs for test equipment are high. However, downstream repair results in a short repair cycle time (i.e. weeks instead of months) which requires significantly less stocks of expensive spare parts. The optimisation approach that is used in this report requires a fraction of the time for the analysis compared to the original approach of Thales LE (days instead of weeks. It requires less concentration and reduces the probability of errors and mistakes. A sensitivity analysis of the input parameters shows that the MTBF and the repair probability have a large influence on the total life cycle costs. The costs for tools and test equipment, the MTBF and the repair cycle time have a significant influence on the decisions that are made by the optimisation approach. Conclusions Conclusion 1: Current life cycle cost optimisation approach of Thales is not optimal. Conclusion 2: No literature available that satisfies the requirements of Thales. Conclusion 3: A cost reduction of 17% and 24% can be achieved for Case 1 and 2. Conclusion 4: The optimisation approach gives good results but it is not very robust. Conclusion 5: Overall, the MTBF is the most sensitive parameter. Recommendations Recommendation 1: Use simplified algorithms for LORA The LORA model that is used in this report is very complex. A simplified algorithm that does not require a sophisticated solver would simplify the implementation process and increase the understanding of the employees at Thales LE. A simplified algorithm could be based on simple calculation rules that, for example, indicate per test equipment whether it is efficient to procure it or not. Recommendation 2: Improve logistical data at Thales LE The logistical data that are used at Thales LE holds a lot of uncertainty. The quality of the optimisation can be improved if the input parameters are more reliable. The sensitivity analysis shows that the MTBF is the most important parameter. The cost factor for tools and test equipment and the repair cycle times are also important. Recommendation 3: Study the factors that influence the MTBF and repair probability Both factors have a large influence on the life cycle costs. The life cycle costs can be reduced if these component characteristics can be improved. Recommendation 4: Improve the robustness of the optimisation approach The robustness of the optimisation approach can be improved by an alternative improvement technique that is discussed in the recommendations section of this report.
Item Type:Essay (Master)
Clients:
Thales Nederland
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:85 business administration, organizational science
Programme:Industrial Engineering and Management MSc (60029)
Link to this item:https://purl.utwente.nl/essays/60762
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page