Integrating planning, scheduling, and simulation in multi-product batch plants

Production planning in multi-product batch plants poses a number of difficult challenges: balancing forecasted demand and product pricing against available plant capacity; deciding which products to make within a given time frame, and in what batch sizes to make them. Once these planning decisions are made, there is still the issue of scheduling production: when to make the desired product batches, and in what order, within the known equipment and processing constraints. More often than not, the planning and scheduling decisions are made independently of one another, without an accurate estimate of the consequences of implementing a particular schedule. When things don't quite go according to plan, what usually results is lost revenue caused by shortfalls between production and demand.

Day & Zimmermann International, a leading engineering consulting firm with a strong presence in the bio-pharmaceutical industry, has found a unique way to help their customers with this type of problem. According to T.S. Ramesh, director of the Knowledge Systems group at D&Z, the solution lies in combining two advanced technologies, simulation and optimization, in a single framework, using a common underlying model of the process. "The power of this approach is its simplicity, " he says, "because linking planning, scheduling and simulation capabilities in a single environment ensures that the plan is profitable, the schedule implements the plan, and the schedule is verified to work within the plant constraints. Making planning and scheduling decisions in this integrated manner allows plants to make dramatic improvements in their bottom line. For example, one customer was able to realize a 19% improvement in gross profit margin through the use of this approach." Industry watchers echo the importance of these benefits. A recent international Delphi study issued by the Process Plant Equipment Association stated that "simulation technologies will swing open the doors to new dimensions of plant planning. Experts expect productivity increases of up to 30% by 2010."

D&Z's modeling tool, called Batch Design Simulator (BDS), is based on Gensym's G2 software, and provides object-oriented, time-staged simulation capabilities. The recently released version 5.2 of BDS includes a proprietary optimization algorithm that solves the planning and scheduling problems using the underlying model of the process. The same process model is then simulated to verify that the schedule is workable in practice within the process constraints.

Ramesh says that the approach is both flexible and scalable. "Our optimization algorithm can maximize or minimize a desired objective function, or simply find a feasible solution that satisfies the constraints. We've also structured it in such a way that it effectively manages combinatorics, allowing us to tackle large problems." According to Ramesh, Day & Zimmermann has successfully used BDS in over 40 consulting projects, for conceptual design and capacity analysis of new facilities as well as improvement of existing facilities, with a focus on capital cost savings, capacity improvement, cycle time reduction, and better resource utilization. Day & Zimmermann also licenses BDS to customers at their request.

For more information contact: T.S. Ramesh, Director of Knowledge Systems, Day & Zimmermann International Inc., 1818 Market St., Philadelphia, PA 19103. Tel: 215-299-8393.

Edited by Angelo DePalma
Managing Editor, Pharmaceutical Online