In this article I present a case study demonstrating the application of manufacturing simulation for plant design. In fact, the best time to optimise a new manufacturing plant is in the design phase. There are no costly equipment modifications, no staff retraining, and no shutdowns whilst you change your plant’s layout. Everything can be done on paper where it is much cheaper to make changes.

Manufacturing simulation for complex plant design

Plant design is easier said than done. When you are designing a new manufacturing plant it is simple to optimise a small linear section of the process, but to optimise the full system it is very difficult. Even with skilled use of Excel you cannot account for all resource constraints, process variation and mix of products. A second issue is that as new information generated through process trials and design development are difficult to incorporate into your original calculation.

To create a useful model of your design you need to use manufacturing simulation (or discrete-event simulation) software, we use Simul8. Once you have the model you can verify your design to ensure that you are hitting all of your design specifications. You can de-risk the design by testing the sensitivity of your assumptions and assessing potential failure points. This information can then inform the direction of development work.

If you want to optimise your design, you can identify bottlenecks and underutilisation. Then iteratively test resizing and rebalancing of the equipment until you identify the most efficient use of capital.

If you are concerned about sustainability and operational costs, you can use the model to directly track materials, labour and energy usages and produce a detailed operational P&L. And finally, one of the biggest benefits of having a model is that you can demonstrate your design to the stakeholders. It is incredibly persuasive letting the Finance Director and Managing Director interrogate your design through the model.

Creating a manufacturing simulation model

At the start be very clear about what you are trying to achieve. It is an easy trap to try and model everything to a high level of detail. Once you understand what you are trying to achieve (with an idea or how you may grow the model in the future), then start planning what information you need. Remember the key is to only model what you need to model.

When we are looking at creating a site model, we consider four parts:

• Manufacturing process: How materials flow and are transformed by the process.
• Raw Materials: The supply chain and materials handling. Getting the raw materials to the
  manufacturing processes.
• Products distribution: How the product is distributed to either a warehouse or the customer.
• Administration: How sales orders, production orders and supply chain are managed.

You very rarely need all four parts modelling in great detail and the process can just be replaced by a black box process. For example, you can represent manufacturing as an activity that just magically spits out a product every 5 minutes, especially if you are only interested in the product distribution part of the system.

Examining each part of the site model is an article in itself which I will leave for another day. Here I wanted to talk a little more about modelling raw material call-off, before discussing an example of using manufacturing simulation in an Engineering Design study.

Supply chain flows in manufacturing simulation

Supply chains are usually complex systems. Though from a manufacturing point of view the requirement is simple. That is, the right parts should be in the right place at the right time. Typically, there is a side-line store next to the process where the parts are consumed. The side-line store is replenished by the business’s warehouse which in turn is replenished by a supplier.

If inventory management is irrelevant to the objectives of the model you can ignore it and just assume there is no constraint arising from availability of raw material. The opposite end of this is to build the model around the warehousing operations and model replenishing the bays, and collecting parts though pick lists.

Modeling side-line stores in manufacturing simulation

Most of Production Support 56’s work is focused on the manufacturing process. Nevertheless, we are quite often interested in the replenishment of side-line stores and even the business’s warehouse inventory. In these cases, we model the manufacturing line consuming stock and then triggering the replenishment of the side-line stores or business’s warehouse.

To model this, we need a Bill of Material (BoM) for the manufacturing process. When the manufacturing process is triggered, it removes inventory from the side-line stores. If the side-line store’s inventory goes below a re-order level it triggers a replenishment process. This takes a defined batch size of parts from the warehouse and adds them to the side-line stores. If we are also modelling the warehouse, we will track the part’s inventory level in the warehouse and if this drops below the re-order level then it triggers a re-order process, this is likely to have a higher re-order batch size and have a longer lead time. This example uses pull inventory management. It is just as easy to model a push inventory management, where parts are delivery to a fixed schedule.

The beauty of modelling is that you can test different inventory management protocols and adjust re-order levels and batch sizes. From a design perspective you can use the model to optimise the size of your side-line stores and inventory management protocols to ensure the smooth flow of your manufacturing processes.

New fertiliser plant design simulation case study

Here it would be helpful to talk through an example of using manufacturing simulation to support front end engineer design (FEED) studies. We were commissioned to provide a model to demonstrate a new fertiliser plant design. The model would replicate the new design but would run historic production data.

The objectives were:

• Validate and demonstrate new design to all stakeholders.
• Identify the operational requirements to prevent raw material delivery lorries from queuing on the public highway.
• Help size the storage silos to ensure there is sufficient raw materials to last for 24 hours in case of disruptions to the supply chain.

Method for conducting fertiliser plant manufacturing simulation

Production Support 56 created a dynamic computer model of the engineering design and ran it using the client’s historic production data. The model simulated four major operations.

• Administration: Setting production schedules and managing raw material inventories.
• Raw materials delivery: Delivery schedules, arrival, unloading and storage of raw materials.
• Product manufacturing: Withdrawal and transfer of raw materials, manufacture of product,
  and loading of distribution tankers.
• Product distribution: Clean down, loading, distribution and return of product tankers.

Take a look at this video to see the results.

Value resulting from the fertiliser plant manufacturing simulation

The model became an important part of the engineering design process, and produced more benefits than originally expected, here are a few:

• Validation: The new engineering design was proven to be capable of meeting production
  targets for all seasons (Spring, Summer, and Autumn), and prevent build-up of delivery truck
  queues outside of the site entrance.
• Design optimisation: The engineer team were able to remove bottlenecks and underutilised
  equipment at the design phase and hence provide better capital efficiency.
• Communication: The model was used to demonstrate and interrogate the design by all
  stakeholders.
• Growth plan: The fertiliser company were able to push the design and identify a
  development plan to increase throughput by 20%.

Conclusion: Use simulation to improve designs

Manufacturing simulation is the perfect tool to use during a front-end engineering study. At that point in the project, most of the required information is available. The model can give you real insight into your design and allow you to get the most out of your capital spend. It is also invaluable as a communication tool, to be able to show all the stakeholders the design in action and allow them to ask it questions is a great way to build confidence and buy-in.

Related articles related to manfacturing simulation

Below is a list of a some other SCDA articles related to manufacturing simulation, plant design and discrete-event simulation.

• Link: AGV simulation of part routings in AnyLogic
• Link: Simulation-based capacity planning
• Link: Discrete-event simulation procedure model
• Link: Visual Components financial KPI simulation
• Link: Simmer in R for discrete-event simulation
• Link: Machine learning and discrete-event simulation
• Link: Receival inspection simulation with simmer
• Link: Parking lot simulator with simmer in R

Dr Michael Stephenson

Co-Founder of Production Support 56. Improving operational performance with process improvement, process development and simulation.