D.W. Ellis & Associates Ltd. has developed over 90 simulation models with over 30 of them being large comprehensive models. During this time we have developed a structure for the whole model and modules for commonly occurring features needed for the proper modeling of operating/process features. These modules considerably reduce the development time for simulation models. As well, the modules have proven logic.
The following are descriptions of these structures and modules. Note that for a given application not all of them may be relevant.
Although all simulation languages handle the advancement of time, D.W. Ellis & Associates Ltd. has developed a logic for handling special situations. These include the periodic changing or randomizing of equipment maximum operating rate, and the 'lookaheads' needed to recognize when inventories will hit capacity, zero, or other special levels.
Any business that is affected by seasonal variations in temperatures, rain/snow fall, volumes, etc. must have these reflected in a simulation model. We have a structure that can recognize anything from a simple summer/winter difference to differences by month of the year. The recognition of different seasons can be used to vary maximum operating rates, speeds, efficiencies, the incidence of storms, demand, etc.
Outdoor operations are typically affected by storms. We have a module that handles the generation of storms with different frequencies and severities as a function of the season. As well, other logic dictates whether the storms slow down, shut down, or slow down and then shut down operations.
Maintenance scheduled by the calendar or by operating hours is an important factor in determining equipment availability and utilization as well as overall process production. We have modules for handling calendar PM with a user input of the schedule. As well, we have logic that accumulates operating time for pieces of equipment towards various maintenance activities. This logic is sophisticated enough that it determines if it is appropriate to take down a piece of equipment at any given time and to recognize the value of opportune maintenance should any unscheduled down occur close enough to the time of required PM.
Equipment failures are, by their nature, unscheduled and unpredictable. The operating performance of each piece of equipment must be known well enough that a distribution of the frequency and duration of unscheduled downs is known or can be estimated. Considerable work is usually required to organize this data. We have experience to help you with this task and a structure in our models to allow you several different distributions of down times (frequency and duration) for each piece of equipment. The handling of these unscheduled downs is one of the strongest benefits to simulation modeling.
We have experience working with many different inventory management rules. At capacity, upstream must be slowed down to downstream, and when empty, downstream must be slowed down to upstream. However, at high levels, upstream may be slowed slightly to delay hitting capacity or extra resources may be brought in to increase the downstream throughput. At low levels, downstream may be slowed slightly to delay hitting bottom. We have modules to handle all these situations. In a continuous operation it is also necessary to 'lookahead' to see when these critical inventory levels will be reached.
A histogram of the inventory levels during the case run is also produced.
A popular use of simulation modeling is facility planning or debottlenecking. In our simulation models we always produce a restriction report. It shows which pieces of equipment are responsible for times of 0 production, and for reduced operating rates. This is a very important piece of information in selecting the next case to analyse.
Most simulation languages handle the reporting of queue statistics reasonably well. We have developed many sets of special rules for items, trucks, etc to join queues.
Equipment operators usually take breaks for coffee, lunch, shift changes, equipment fueling or servicing. We have a module for handling the schedule of these breaks. As well, we have the logic to test various rules for these breaks. For example, how many operators of a particular type may be on a break at the same time, or should breaks be staggered or at the same time?
If you have questions or comments, please e-mail .
Back to D.W. Ellis & Associates Ltd.'s simulation page