Integral Logistics Management — Operations Management and Supply Chain Management Within and Across Companies

8.3 Processor-Oriented Resource Management

Intended learning outcomes: Explain campaign planning. Differentiate between processor-dominated Scheduling and material-dominated scheduling. Describe a nonlinear usage quantity and a product structure with loops.

8.3.1 Campaign Planning

Section 8.1.3 described large lots as a consequence of setup costs due to stopping, cleaning, and restarting processes. The changeover pro­cesses for transporting flow resources are of lesser significance. In processor-oriented resource management, the objects concerning capacity management and production control are not equivalent to materials management objects.

  • For control, the planning unit is the machine or facility, such as the re­actor, which thus also becomes the actual planning object. The tech­nically feasible batch size is calculated by the quantity of goods that should ideally be processed by this facility. The batch thus produced is also used for accounting, stockkeeping, and archiv­ing information for the subsequent lot traceability, for instance.
  • From the materials management viewpoint, the emphasis is placed on demand. For technical reasons, a production lot can only be a multiple of a production batch. “Optimum” batch sizes, whether calculated using stochastic or deterministic methods (see Sections 11.3 and 12.3), often have to be rounded up considerably due to the high setup costs and the required utilization of capacity. Such hidden formation of batch sizes results increasingly in block demand for, and thus a decidedly quasi-deterministic form of, materials management.
A campaign is an integer multiple of production batches of a certain item, the batches being produced one after another. 

A campaign cycle is a sequence of campaigns during which all the important products are produced up to a certain capacity and in the quantity required by demand.

The sequence of campaigns is used to reduce setup costs. As soon as the optimum batch size from the materials management viewpoint consists of several batches, it is then combined to form a campaign. Under certain circumstances, it is then advisable to produce a batch of a different product immediately afterward, if this will avoid the need for a cleaning process, for example. The formation of campaigns in this way is a characteristic feature of processor-oriented resource management. This means that the entire campaign must be considered, rather than just the individual batches, when scheduling capacity. A campaign can, of course, be split back into its constituent batches if necessary.

Campaign planning aims to create optimum campaign cycles.

Campaign planning is one type of sequencing, or the combination of optimum sequences. Optimization can target various areas: production costs, manufacturing time, or product quality. Figure shows the example introduced in Figure, with the addition of a packaging process. The example is taken from [TaBo00], p. 18 ff.


Fig.        Example of a process chain in chemical production (see Figure

The grades A, B, and C produced at a plant are packed into two drum sizes (4 liters and 20 liters) in the subsequent packaging process. The demand is for the 6 end products (3 grades times 2 packaging sizes). To simplify the example, the minimum batch is assumed to be one day’s production. The demand for an end product is specified in relation to the overall demand: A4, 30%; B4, 20%; C4, 10%; A20, 20%; B20, 10%; and C20, 10%. Bill-of-material explosion results in the proportionate demand for the intermediate products obtained from the reactor: 50% A, 30% B, and 20% C.

Assuming that the long time required to set up the packaging process arises when the packaging size is changed and that the reactor setup costs can be minimized by the sequence A, B, and C, as well as the specification of a minimum campaign of one day’s production, the campaign cycles shown in Figure result.


Fig.        Campaign cycles for the example in Figure (see Figure and a minimum campaign of one day’s production.

The reactor rhythm is determined by the minimum pro­portion of the demand, 20% for C. The campaign cycle thus lasts 5 days. The packaging rhythm is determined by the minimum propor­tion of the demand of 10% for C4 or C20. This campaign cycle thus lasts 10 days.

The ideas behind processor-oriented resource management thus correspond in some respects to those of the lean / just-in-time concept (see Section 6.2), in which the optimum sequence of operations is important with a view to reduce the setup times (see also Figure The reduced setup times should result in small lots and, therefore, continuous demand. Only then will it be possible to totally separate the processes that make up the various production structure levels, which will allow the use of the Kanban technique in the process industry.

If continuous demand cannot be achieved, then quasi-deterministic techniques will still be required. In this case, the response to a net demand will be to schedule at least one campaign for production, rather than just a batch. A batch also results in by-products. Both of these contradict the simple pull logistics of the Kanban technique, since production is determined by the technical process and savings in terms of setup time, rather than in response to consumption. The dominating factor is capacity management.

The conventional MRP II / ERP concept of resource management does not incor­po­rate the processor-oriented techniques, such as manufacture of by-products and campaigns, making them less suitable for the process industry. Cam­paign planning enables demand to be synchronized in terms of quantities with the goods to be produced at all production structure levels, particu­lar­ly with respect to end products. Where synchronization is not possible, buffers must be kept to absorb any shortfall. The aim of campaign plan­ning is thus to mini­mize the inventories that have to be kept in the inter­mediate stores by synchronizing the va­ri­ous (process) stages as accurately as possible. Figure shows how the two (pro­cess) stages (or production structure levels) could be synchronized for the above example.

Fig.        Campaign planning: how the (process) stages could be synchronized.

The diagram shows the start and end of the overall campaign for each pro­duct, but not the individual batches. It can be used to calculate the resulting stock curves for the end and intermediate product stores for given quanti­ties. The inventory curves are of the type discussed in detail in Chap­ter 11 for determining the available stock. They are used as the basis for trouble­shooting, particularly for determining the buffers that will be needed.

The campaign planning technique described here is modified finite capacity planning (see also Section 14.3) that requires continuous intervention by the scheduler. The planning diagrams are similar to the Gantt charts or planning boards used in finite capacity planning (see the illustrations in Sections 14.3 and 15.2.2). The only difference is that they include — as well as individual batches — entire campaigns or even campaign cycles.

Exercise: Campaign cycle planning
Try to plan the campaign cycle of a two-step production (reactor and packaging) for four different products, each with two packaging options. You are the production manager. Before beginning with the planning task, take note of the constraints and hints offered.

8.3.2 Processor-Dominated Scheduling versus Material-Dominated Scheduling

Processor-dominated scheduling (PDS) is a technique that schedules equipment or capacity (processor) before materials. This technique facilitates scheduling equipment in economic run lengths and the use of low-cost campaign cycles ([APIC16]).

See also [TaBo00], p. 30 ff. The campaign principle outlined in Section 8.3.1 is an example of processor-dominated scheduling. Indeed, capacity management has priority over materials management for scheduling. Finite loading is used as the scheduling principle. Materials are planned according to the results of finite loading.

Processor-dominated scheduling is characteristic of the processor-oriented con­cept. It is typically used to schedule manufacturing steps within a pro­cess stage. However, the process industry does not use it in every situation.

Material-dominated scheduling (MDS) is a technique that schedules materials before processors (equipment or capacity). This technique facilitates the efficient use of materials ([APIC16]).

Material-dominated scheduling can be used to schedule each stage within a process train. Typically, the MRP II  /  ERP concept as well as the lean / just-in-time concept use material-dominated scheduling logic. In the process industry, they have their significance as well.

The problem in the process industry is to identify the point at which the processor-oriented concept replaces the other concepts. Figure provides a simplified rule of thumb. This line of reasoning is similar to that followed in [TaBo00]. In addition, see also Figure MRP II  /  ERP concept or the lean / just-in-time concept may be used if

  • Materials are expensive related to manufacturing costs.
  • There is over­capacity.
  • Setup times and costs tend to be negligible.
  • There is job shop production rather than line or flow shop.

The processor-oriented concept may be used if

  • Capacity is expensive related to costs of goods manufactured.
  • There are capacity bottlenecks.
  • The one-off costs for each lot produced are relatively high.

Fig.        Use of the MRP II /  ERP concept or of the lean / just-in-time concept compared to the processor-oriented concept.

8.3.3 Consideration of a Nonlinear Usage Quantity and of a Product Structure with Loops

In the process industry, the quantity per, or usage quantity, corresponds to the selective use of starting materials to produce intermediate, end, or by-products.

The operation/process yield is the relationship of usable output from a process, process stage, or operation to the input quantity (compare [APIC16]).

Operation/process yield can often be expressed by a ratio, usually as a percentage. However, chemical and biological processes are subject to conditions that cannot always be predicted accurately (for example, external influences like the weather). In addition, the technologies and production processes used, as well as variations in the quality of the raw materials, have an effect on the consumption of resources that is not quantifiable in every respect. For example, excessive use may be made of certain materials in the startup phase of a process or in the course of the process — namely, as the produced quantity increases. In such cases, the usage quantity ceases to be a linear function of the quantity produced.

A nonlinear usage quantity is an operation/process yield that cannot be expressed by a linear function of the quantity produced.

Just as with the usage quantity, the duration of the process is no longer proportional to the quantity produced. Thus, the effective consumption could change, as shown in Figure See also [Hofm95], p. 74 ff.


Fig.        Quantity of a manufactured product P as a nonlinear function of the usage quantity of a resource R.

In some cases, the nonlinear function for calculating the nonlinear usage quantities may be known in advance. The problem is then solved using an appropriate formula as a parameter, rather than as a constant value for the usage quantity or quantity per attribute. In the event of a transition from the production structure to an order structure, the formula is evaluated using the parameter values associated with the order (e.g., the batch size), and the appropriate demand for the resource is thus determined. This procedure is exactly the same as described in Section 7.3 for one-of-a-kind production of products with many variants. There, formulas are linked with various attributes, and not just with the usage quantity and the operation load.

Most products with a product structure with loops are those that can be returned to the production process. These may be by-products (such as broken chocolate or energy in the form of steam or heat) or processing aids (catalysts, for example) that can be used for further production. It thus follows that the by-products or waste products are not subject to external demand, and their use can therefore be optimized internally. There are, however, certain quantity or time-related marginal conditions concerning usability (spoilage, deterioration) or storability and shelf life. Most of the soft­ware packages based on the conventional MRP technique, as described in Section 12.3, do not allow loops. This is because the technique deals with the individual items in the order of their low-level code. In a production structure with loops, the low-level code would be regarded as “infinite.”

One possible solution to this problem is to identify such items (by-products or waste products) and then to omit them from the structure-level code calculation or to allocate to them a maximum structure-level code. The MRP technique should then be used to schedule such by-products or waste products only at the end. At this time, all the demand is already known, as are also all the planned receipts in response to planned orders. Any net requirements for such by-products or waste products would then have to be produced or procured. Consequently, an additional production structure without further by-products should be allocated to each of these products. This is then converted into an order structure.

For the manufacture of chemical products, the features and qualities are determined by a multitude of parameters. Among others, there are the mixture proportions of the final product, temperature, holding time in the apparatus, etc. This often makes it difficult to plan the need for resources precisely and results in losses in production and energy during the commissioning phase.

Exercise: Process Train
In the following exercise, try to commission a very simplified continuous process, a process train of two manufacturing steps, and convert it to a steady operating state. As is often the case in practice, the criterion for product quality is the color of the products. Variation parameters are: the demand ratio to the raw materials and the holding time in the processors, which can be influenced by the demand quantity and the level. The goal of this task is to keep the amount of waste as small as possible.

Course sections and their intended learning outcomes

  • Course 8 – The Concept for the Process Industry

    Intended learning outcomes: Produce characteristics of the process industry. Disclose processor-oriented master and order data management. Explain in detail processor-oriented resource management. Describe special features of long-term planning.

  • 8.1 Characteristics of the Process Industry

    Intended learning outcomes: Explain divergent product structures and by-products. Describe high-volume line production, flow resources and inflexible facilities. Produce an overview on large batches, lot traceability, and loops in the order structure.

  • 8.2 Processor-Oriented Master and Order Data Management

    Intended learning outcomes: Produce an overview on processes, technology, and resources. Present the process train: a processor-oriented production structure. Disclose lot control in inventory management.

  • 8.3 Processor-Oriented Resource Management

    Intended learning outcomes: Explain campaign planning. Differentiate between processor-dominated Scheduling and material-dominated scheduling. Describe a nonlinear usage quantity and a product structure with loops.

  • 8.4 Special Features of Long-Term Planning

    Intended learning outcomes: Disclose the determination of the degree of detail of the master production schedule. Describe pipeline planning across several independent locations.

  • 8.5 Summary


  • 8.6 Keywords


  • 8.7 Scenarios and Exercises

    Intended learning outcomes: Differentiate between batch production and continuous production. Calculate an example of manufacture of by-products. Elaborate an example of production planning in process industries.

Print Top Down Previous Next