The development of a discrete-event based simulation model to improve large diameter steel water pipeline construction processes

Abstract

Water pipelines is a convenient way of transporting water over long distances. Construction processes of water pipelines consist of a series of sub-processes with tasks that must be executed in a certain way to complete a project work item. The tasks within these sub-processes include, but are not limited to, people and resources, and the tasks time is influenced by safety - and environmental factors and management restrictions. Difficulties are found when planning and implementing such a complex construction process. The main objective of the project was to develop a simulation model to identify inefficient and constraining activities and find improvements opportunities for the tie-in process within a water pipeline construction process, since the importance of timely execution there-of. The tie-in process consists out three major consecutive tasks, and consists of two crews. The first task is welding the two pipe-ends together and is executed by Crew 1. The next two tasks are coating the outside of the pipe for protection and then applying internal lining for protection. The latter two tasks are both executed by Crew 2. A discrete event simulation model was developed to incorporate all the relevant input data, experimental factors, as well as outputs. The model was used to simulate an as-is case and comparing the results to known results of a completed project. The model simulated a 1.4 joints/day, which compares well with the actual recorded 1.2 joints per day. The simulation results were scrutinised to identify the constraints in the process. It was found that the idling time of the Crew 1, caused by it waiting for the Crew 2 to access the pipe, represented the primary constraint. The model was then used to experiment to find improvement opportunities for this constraint. The introduction on access holes in the pipe, which decouples the two crews, was modelled. This resulted in a less than expected improvement, due to the Crew 1 still idling when reaching the Management Exposure Restriction m. To resolve this, the production of Crew 2 had to be improved by allowing for simultaneous execution of lining and coating (effectively adding another Crew 2), which then resulted in a balanced process where idling of Crew 1 (the chosen constraint), is minimised. The model simulated a production rate of 2.1 joints/day and a cost R17.6m (for 4,000m). The rate represents an improvement of 50% on the as-is case production rate. It is also concluded that the model may be used as a management tool, which can aid Aurecon in the planning and control of future projects.