Lubricating Oil Company

Project

Computer Simulation

Lubricating OIL COMPANY

1. Problem Formulation

A manufacturing system produces Lubricating oil: P1, P2, and P3. There are four Vacuum Distillation Unit (labeled DU1 - DU4), three Extraction Unit (EU1 - EU3), two Removal Unit (RU1 - RU2), two Additive Unit (AU1 - AU2).

All products require a deterministic transit time to move from one process to the next. When the parts arrive at a Unit station, they are placed in a buffer for holding. I assume that the buffers have a sufficiently large capacity to accommodate the arriving parts. The parts are removed from the buffer as soon as a machine is available for processing. The path network and the resources characteristics of the transportation are also known. The Vacuum Distillation Unit fail as a function of the number of times they operate on parts and that the time for flushing, cleaning and waiting time for repair service are known.

The Production Engineer has been tasked to develop an animated discrete-event simulation model of the system. In specific, the modeler is tasked to compare the downtime for the Vacuum Distillation Units in several scenarios and determine the throughputs and utilization of processing locations with ultimate production rate 300 barrel per day. However, some structural improvements may be suggested as a by-product of the simulation. Processing times for the parts are deterministic and are indicated in Table 1

Table 1. Processing time in minutes for products.

Machine: P1 P2 P3

Distillation 17 - -

Extraction - 12 -

Removal - - 8

Adding - - 8

2. Model Implementation

In this system, four locations are designated; one for each of the four machine tools involved in the processing. Downtime was modeled only for the Distillation. The Distillation location has a capacity of four, so four single capacity units are depicted on the screen. The Extraction has three units, the Removal has two units, and the Additive has two units. Other locations are the InBuffer, DE Conveyor, ExR Conveyor, RA Conveyor, and the OutBuffer from which the parts exit the system. All of these locations have an infinite capacity but only one icon is depicted. Figure 1 depicts a snapshot of the system configuration. The figure shows the schematic of the entities in the manufacturing system as well as some of the parameters for machines such as DU in the system. The numbers on the right of the InBuffer location indicate a counter of number of products in the location. Products after being processed on the Distillation go to the DU conveyor. The OutBuffer is also represented as an empty pallet with an entity spot and counter. Products from the OutBuffer exit the system.

Figure 1. A snapshot of the configuration of the system simulation

InBuffer and OutBuffer are the entry and exit points, respectively, for entities in the model. No operations or delays occur at these locations. The arriving entities are depicted in the InBuffer if they are waiting for their first processing location to become idle. Otherwise, they move from the InBuffer to their first processing location

The conveyors were used to transport the products from one machine process to another in the required three minutes. The transit time for conveyor was accounted by specifying the length to be 30 feet and the speed to be 10 feet per minute. The conveyors were also used to model the machine buffers by making them the accumulating type.

3. Output Analysis

The simulated time set to be for 24 hours. I simulated the manufacturing system and executed the program several times using the real life situation as mentioned in the problem formulation. For each program execution result or replication, various performance measures were output by ProModel. These include the location utilization chart during 24 hours, single capacity location states chart and in Buffer contents. From these Output charts I see that the down time is high about %16.67 and the whole system stops for 2 hours due to the service in the distillation units. However, the production rate is lower than the target. Different scenarios need to be tested in down times issue in order to get the best result with no extra cost.