Copper wire is used for electromagnetic applications. It thus has a wide range of applications and is of great importance, as well as it is offered in many different configurations. Copper wire is produced in a two phase production process. First, the wire is drawn down to required diameter. Second, the wire is enamelled in a complex process that is influenced by various factors. A simulation study can be conducted to optimize production lot sizes and transport batch sizes for overall cost efficiency, focusing on three major cost components: Scrapping costs, resulting from insufficient wire qualtiy, warehousing costs resulting from carried inventory, and labor costs resulting from having to transport each batch from processing to quality inspection and warehouse putaway areas, picking and packaging wire spools in the warehouse shipping and putaway area, and inspection work in the inspection area. The actions undertaken as a result of the simulation study were an increase in production lot sizes of each production configuration, with lot sizes depending on the demand pattern for the respective product configuration. The transport batch size was increased too.

Larger lot sizes reduce changeover costs

Each product configuration has different requirements during drawing and enamelling. Changing over from one wire type to another can take several hours. This thus results in long machine down times and high labor effort for associated technical staff.

Larger production lot sizes result in less scrap

After setting the coating process up to produce a new product configuration complex production process parameters for the coating process must be adjusted and fine-tuned, requiring laboratory inspection and final quality inspection feedback. Both of these quality checks are characterized by feedback delay, i.e. the enamelling process can only by adjusted once feedback has been received. Between inspection sampling / final product inspection and quality feedback receival the coating process prodcues wire continuously, as the production process itself is a continuous process. Initial calibration efforts after a product changeover are thus risky and may result in a lot of scrap material. Scrap is a cost, and it directly affects the profit and loss statement of the manufacturer.

After the initial phase-in of a new wire-type (i.e. after a changeover) production quality in the enamelling department is much more stabile and inspections are only conducted in the form of a finished goods inspection process. The feedback delay between final inspection and the producing enamelling process then still depends on the transport batch size, which determines the inspection interval from then on and until completion of the remaining lot size.

Smaller transport batches: Less scrap but high transport cost

While larger production lot sizes results in less changeovers and thus a lower risk of scrap production, a larger transport batch results in a higher scrap risk. Wire spools produced by the production lines are buffered in buffer inventories that are picked up by transport operators and delivered to final inspection work stations. The buffer inventory is mobile, and is transported when full or nearly-full. Larger bins result in lower transport effort and transport costs, but also longer inspection intervals. This increases the risk of scrap production.

Larger lot and batch sizes result in higher warehousing costs

Larger lot sizes in production will generally result in higher levels of finished goods inventory, since larger amounts of finished goods are produced ahead of customer demand. Larger transport batch sizes will result in higher buffer inventories, i.e. work in progress. Both buffer and finished goods inventory causes various costs, e.g. capital costs resulting from capital tied in inventories instead of other productive assets. Another example is the warehousing space and the associated rent expenses etc.

Other direct costs are e.g. labor costs for relocating products in the warehouse, e.g. as part of warehouse zoning strategies, or finished product scrap risk if customer demand runs out but there is still finished goods inventory left over in the warehouse.

Larger transport batches reduce picking and packaging costs

Once copper wire spools have been inspected and approved they are packeded. Box types differ depending on product dimensions. Afterwards, boxes are grouped and stored on pallets. First, half-full pallets from the warehouse are collected by forklifts and additional boxes are packed on them. A pallet should ideally only contain one product type. For small lot sizes two product types might be grouped onto the same pallet. Afterwards, pallets are put away in the warehouse. This means that labor effort associated with these processes is higher when transport batch sizes are smaller.

Larger transport batch sizes reduce quality inspection costs

The quality inspection process itself has changeover costs. After inspecting a product configuration the inspector must prepare the prodution line for the new product. Typical changeover and setup activities are cleaning, tooling, heating, and calibration. This is true for any industry. The time spent on these changeover operations can be measured and modelled, e.g. with distance matrices, OEE targets or measures, etc. A simulation will simulate the changeovers and the associated time and labor cost, considering the changeover time calculation method defined during model specification (= requirements engineering).

Optimal lot and batch sizing for overall cost efficiency

A simulation study can analyze the optimal production lot sizes and transport batch sizes for each product configuration, subject to the factory layout and material flow. Resulting negative and positive impacts from reducing or increasing transport and production lot sizes, can be analyzed in a safe space in this way – without impact on the real production system during experimentation.

For example, below graph illustrates the trade-offs of the various costs related to production lot sizing in any production system with inspection feedback delay, in which the feedback delay depends on the internal transport frequency.

Below graph illustrates the trade-offs of the various costs related to transport batch sizing in any production system.

A simulation model, simulating relevant material flows and all associated costs, can help in finding the overall cost optima. The simulation model can e.g. be developed using SimPy, a discrete event simulation module in Python that handles queueing and event scheduling. The simulation model accepts any arbitrary amount of product configurations and any type of material flow, as Python-based simulation models are developed completely on sourcecode basis and are thus 100% customizable to any specific need. Each product configuration can characterized by various product and production parameters.

Simulation results support management in implementing larger or smaller lot sizes and larger or smaller transport batch sizes. The simulation model furthermore serves as ongoing production support, and should be handed over to the responsible material flow and quality inspection department. This is important, since many of the parameters considered by the analysis vary over time. For example, seasonal temperatures or other weather conditions, commidity prices (e.g. relevant for scrap cost calculation), relevant product range, lead times, changeover times, equipment availability, shift models, cost parameters, etc.

Conclusion and related content

In this article I summarized how a simulation study can generate relevant results for copper wire production, or any other production system. Results can indicate optimal lot sizes and transport batch sizes. Simulation models for material flow optimization in copper wire industry, or any other industry, furthermore support changemanagement by being used throughout the lifecycle of a production system.

The following articles with related content might also be of interest to you:

• Link: Assembly line simulation and line balancing
• Link: Changeover sequencing in Python
• Link: Setup sequencing with Excel Solver
• Link: Job shop SimPy Python simulation

Linnart Felkl

Data scientist focusing on simulation, optimization and modeling in R, SQL, VBA and Python