Modeling the Production Capacity of a Continuous Plant Using Discrete Event Simulation

Thanks to its simplicity and cost-effectiveness, discrete event simulation is preferred for construction simulation and the issue of how to model the production capacity of a continuous plant by using discrete event simulation is addressed. A plant of continuous nature relies on a material-handling mechanism (such as conveyor or pipeline) to continuously convey material delivered in truckloads to a designated activity location at the site. In contrast with discrete resources commonly encountered and matched in construction (such as a truck, an excavator, and a crane), a buffer is the hallmark of a continuous plant (such as unloading container); and one or multiple feeder resources (trucks) can be simultaneously processed subject to the production capacity of the plant. With a concrete pump example, we discuss the potential pitfall of simplifying a continuous plant as one discrete resource entity. Then, we formalize an approximate method for representing a continuous plant with N discrete resource entities in simulation of a predominantly discrete operations system. We also describe two practical applications: (1) modeling the production capacity of a magnetic separation plant in iron mining operations; (2) modeling the passing capacity of an urban road section in microtunneling and pipe-jacking operations. In conclusion, the proposed method adds to the usefulness and flexibility of a discrete simulation methodology in modeling complicated construction systems. INTRODUCTION Simulation modeling builds a logical model on the computer medium as a valid, adequate representation of a complicated problem in reality, aiming at achieving a better understanding of the problem and hence resolving the problem. (Law and Kelton 2000) With respect to the mechanism by which the state of the system changes over time, simulation methodologies can be broadly categorized into discrete event simulation and continuous simulation (Prisker and O’Reilly 1999). In discrete event simulation, the modeler is concerned with how to describe the logical conditions for triggering the occurrence of events that change the system state only at discrete points in time. The majority of simulation applications in construction engineering fall into the discrete class for its simplicity and effectiveness