(A conventional paper covering all three parts and offering more detail and examples is at sdl.re/SESDpaper).
Modeling systems of physical factors
Since SD’s foundations lie in engineering control theory, the method has naturally been applied to a very wide range of physical-system challenges. Project management SD models are mostly built around stocks of work-to-be-done (as shown in the small part-1 model) and work-completed (Lyneis and Ford 2007). However, such models usually include changes that occur to quantities of relevant physical items or materials themselves – units produced, tons of material consumed and so on.
Other application domains are more centred on the dynamic behaviour of physical factors themselves. Supply chains are of course made up of interconnected stocks of items and materials, between which goods and materials flow. SD models of supply chains capture their high-level dynamics – how aggregate quantities rather than individual entities move and change over time – but can add connections to non-physical factors, such as workloads and the financial value of those materials.
Models can also account for often-powerful intangible factors and their impact on the management of supply chains. Fear of stock-outs, for example, may cause over-ordering, with potentially damaging and costly implications for inventory levels and flows (Sterman, 2000: chapter 17). Figure 1 shows a business holding inventory in order to meet customer-orders and replenishing that inventory with orders placed on a supplier and received after a delivery delay. The figure shows changes occurring to orders and inventory in response to a change in customers’ order-rate, given a particular policy for setting the order quantity to place on a supplier. The model at sdl.re/SEsupplychain demonstrates the complexities of designing an ordering policy that best-meets changing customer orders while minimising the costs incurred by holding inventory.
Figure 1: Changing orders and inventory levels in a simple supply chain
Asset management is another domain in which SD models of physical factors have been applied. Such models can track populations of different types of equipment through a typical life-cycle – after a short bedding-in period, units have a long reliable life, before degenerating and becoming quite unreliable. The commitment of staffing and expenditure to maintenance, refurbishment and replacement of assets is a complex challenge that must balance system-performance aims (notably reliability) against the considerable costs of sustaining the network of physical assets. (See a demonstration model of such a challenge at sdl.re/assetpipeline).
A key SD contribution – connecting systems of physical and non-physical factors
Physical factors feature in many SD models of environmental challenges, such as the management of water resources, natural resources (crops, fish, livestock …), and climate-change impacts (Ford, 2011). The special contribution of such models is the ability to capture interactions between physical and non-physical factors in a faithful simulation of an entire situation or episode – the asset-management model includes financial factors for example.
A much larger model requiring integration of physical, non-physical and financial factors concerns a large-scale engineering project to rejuvenate water quality and wild-life in a moribund lake, explained at sdl.re/LakeModel *. The project required integrated modeling of the hydraulics and water quality, power-generation, physical construction, and the financial business case. Achieving this integrated simulation depended on capturing the knowledge of experts from several disciplines in a shared mental model, enabling all parties to see the relationships between their own part of the system and the whole. The resulting model enabled all parties to see, clearly and immediately, the impact of alternative assumptions and options for the project.
Suggested next steps
Continue to the part-3 article at sdl.re/LIPSESD3, to see how the simple model from the part-1 article can be built into a model of a larger business initiative. Part 3 also explains the wider opportunities for SEs to exploit SD business models.
* This case is courtesy of Copernicos Groep.
Ford, A., 2011. System dynamics models of environment, energy and climate change. In Extreme Environmental Events. Springer, New York, NY.
Lyneis, J.M. and Ford, D.N., 2007. System dynamics applied to project management: a survey, assessment, and directions for future research. System Dynamics Review, 23, pp. 157-189.
Sterman, J., 2000. Business Dynamics: Systems Thinking and Modeling for a Complex World. McGraw-Hill, New York.