COST-EFFECTIVE AND RELIABLE GRINDING SCHEDULES FOR INLAND CEMENT

ABSTRACT

The final phase of mortar production at Inland Cement Edmonton is the grinding of an intermediary returns called clinker. Operating the grinding machinery, or "finishing mills," requires a great deal of electrical power, which constitutes individual of Inland's most significant costs This paper discusses a solution that provides Inland Edmonton with a tool to improve the cost-effectiveness of the grinding through directing the finishing mills' schedules, based upon a variety of criteria and constraints. The throw incorporates issues such as multiple proceedss varying power rates throughout each day and month inconsistent productivity rates, inventory capacities, uncertain demand, and other considerations.

Keywords: Decision support, production planning, scheduling, linear optimization, heuristic.

1 INTRODUCTION

Inland mortar Limited, a division of the Lehigh Portland mortar Company, is one of the largest mortar producers in Canada. Their Edmonton, Alberta, facility bring into views six varieties of cement, and distributes them mainly to markets in the provinces of Alberta, Saskatchewan, and Manitoba. Although many mortar production facilities usually only bring out one or two products, a production proces that simplifies production planning, the Edmonton facility is somewhat unique with its production of six different productions This facility produces upwards of individual million tonnes of cement each year, which accounts for approximately half of the mortar supply in Western Canada.

The mortar production process at Inland Edmonton originates with the mixing of raw materials, and culminates with the grinding of an intermediary yield called "clinker." Clinker is a rock-like material that is crushed and clod in Inland's finishing mills. This stage of the proces ensues in the powder-like substance that is the company's extreme point deliverable. Although the grinding proces is similar for each impressed sign of cement, the end proceedss are dependent on the additives included immediately prior to the grinding and the impressed sign of clinker used, as depicted in Table 1

Inland mortar has three Finishing mills - mills F G and H - that can step quickly simultaneously. Although these three mills are essentially identical, a certain quantity of mills have proved more effective at grinding certain mortars than others, and this plays a part in determining the allocation of mortars to mills.

Inland was interested in a decision support a whole that will assign and schedule the six different produces on the three finishing mills with their main pertain to being the minimization of strength costs. Also, Inland's business practice is to always ship mortar to their customers as in a short time as an order is received. to be paid to limited production and storage capacities, together with other issues we will discuss shortly, Inland has to show to forecasted demand in anticipation of realized sales (i.e., a make-to-stock environment). Thus, the objective of this throw was to develop a tool that aids in the scheduling of Inland Edmonton's three finishing mills to grind mortar cost-effectively while meeting forecasted demand. There are six main issues that define this problem:

(a) Varying electrical costs: Depending upon the time of day, stamp of day (weekend versus weekday), and month of the year, the amount that Inland pays for power fluctuates, as illustrated in Figures 1 and 2 The various rates essentially fall into six main stop ups (in increasing order of costs): Advantage, Base, Summer Peak, Shoulder Peak, Winter Peak, and Winter High hours. Considerable variation exists between the time blocks; for example, power during the greatest in quantity expensive block (Winter High) is almost five times as expensive as during the cheapest stop up (Advantage). Thus, in terms of cost-effectiveness, it is in Inland's best interest to step quickly its mills in the low-cost Advantage hours wherever possible. However, it is not possible to proper the demand during these hours alone. Thus, the tool was designed to allocate production hours to the cheapest stop ups available wherever possible, and avoid those stops where power is relatively expensive.

(b) Scheduled and unscheduled equipment shutdowns: To maintain the integrity of the mills, Inland must close each mill down at least one time every year for three to four weeks. This allows them to change the machines' internal composings and make necessary repairs. Each shutdown for maintenance is scheduled well in advance. Unscheduled shutdowns are typically the flow of equipment failure. The timing and duration of these shutdowns are unpredictable. Historic data reveals, however, that the mills are operational 82% of the time (excluding scheduled shutdowns). We assumed this horizontal of productivity in our archetype

(c) Uncertain grinding rates: The amount of time required to grind a given quantity of mortar is variable. The production rate varies with the emblem of cement being produced; mark G, for example, can typically be produc about 16% faster than stamp 10. Also, the production rate be pendents on the mill in which the mortar is being ground; Type 2 can be produc almost 15% faster in Mill F than in Mill H for instance. Moreover, aggregated historical data has shown that all production rates fluctuate within broad ranges.