Utilising “Little’s Law” to drive your improvements through reduced variation
Improved quality, increased throughput and better recoveries. We can have it all if we understand and know how to reduce variation.
So writes John McConnell in a post “Optimising Continuous Operations”. John is practitioner of the quality approach to management, where you apply statistical methods to understand, control and monitor process’s.
A common theme through John’s work, which in addition to the website, includes several excellent and practical books including “Safer than a Known Way”, “Analysis and Control of Variation” and “7 Tools of TQC”, is the importance of removing variability from your process and the significant benefits that come with this. Known as “Little’s Law” it states that:
all other things being equal, as variation in flow of material through a circuit is reduced, throughput rises. This increases productivity and reduces cost per tonne. In our experience, many continuous plants operate at between 5% and 10% below their true capacity, and that variation in flow of material through the circuits and over-control are chief amongst causes for this reduced performance… Little’s Law is a law, not just a good idea… if variation in flow is reduced, cycle time (or residence time) will fall and either Work-In-Progress will fall or throughput will rise, or some combination of both; not sometimes; always.
In working to control variation, it is critical to understand the difference between common cause and special cause variation. Common cause variation is a function of the system, the accumulation of all the many system factors interacting randomly to produce an output. Reacting to common cause variation (ie changing setpoints) invariable makes the output worse! Special cause variation has a non random impact from a change outside the system (eg a piece of equipment failing) so how to identify whether there is special cause or common cause variation? Use a control chart! More on this in a later edition.
In addition to John’s books, I have got a lot out of the following links on his website:
Optimising continuous operations
- There are many variables that plant personnel struggle with in any continuous operation. However, experience shows that Little’s Law works. If we can stabilise the flow of material through the process (reduce variation), output capacity rises. In addition, a circuit that is reasonably stable will operate more effectively and efficiently from a technical perspective. This makes intuitive sense. If the tonnes flowing through the process are reasonably steady, control of other aspects such as particle size and chemistry nearly always becomes less necessary and easier to execute when it is necessary.
Lessons for Manufacturing for mining and milling operations and other continuous operations
- These case studies illustrate the significant advances possible if a business focuses successfully on understanding and reducing variation, especially when the variation of inputs and to the early stages of the process is addressed.
Control Charts for Continuous Operations
- When confronted with the question: “Which type of chart should I use?” the better answer is often: “Both”. We need not be trapped in an “either — or” mindset.
- It remains common for both manual control methods and instrumentation designed to either control or reduce variation (or both) to actually increase this process variation. Before we attempt to use instrumentation to reduce or control variation, we would be wise to ensure our control systems are not injecting it.
- In many operations, scheduling a plant or a supply chain for maximum equipment utilisation is a terrible mistake. Such an approach to scheduling has the potential to actually increase costs.
- There is no trade off between quality and quantity. The laws of statistics dictate that only operations that successfully reduce variation and improve quality will achieve high output and low work in progress. The good news is that we can have it all.
