AAOM Handbook

the measure. For example, if we are measuring the utilisation of a process, the time boundaries should be set to enclose the period when the process is performing its Purpose, and to exclude the period when the process is not performing its Purpose. The boundary conditions for the theory and purpose measures should be matched. When processes interconnect, as they always do, people often get confused trying to decide whether a particular measure is of a lagging or leading nature. This is understandable because the nature of a measure is relative. The output parameters for one process are typically elements of the theory measures (input in this case) for the following process. For example if you are looking at the grind quality from a mill, it is a lagging measure of the mill operating performance. For the leach circuit described in the example above, the grind quality will be a leading measure of the process performance. The output performance of a process will always exhibit some variation. Variation can arise from two types of cause – common causes and special causes. Common causes are defined as those that regularly recur when the process is behaving in a consistent way (this does not mean that the process is meeting specification – it might consistently not be meeting specification). The characteristics of consistent performance are that over time the average and the variation around that average are predictable. Special causes are defined as those instances where the process did not behave in a consistent way. Statistically, special causes account for less than 1% of the variation that occurs in stable process. You can readily identify the common and special cause variation in performance using a control chart to present the measures of the process purpose and theory. Each special cause should be investigated as soon as it becomes evident in the Control Chart, because it indicates that the process is probably changing (unstable) and hence may require an immediate intervention. Common causes do not need to be investigated in each instance. If the process is stable and meeting specifications, or if it is stable but not meeting specifications and you do not intend to take any immediate action (you have bigger fish to fry), then it is not necessary to collect the reasons for common causes. Importantly you only need to collect a representative sample of common causes, since in a stable process the same common causes will continually recur. See the BPF paper Quantify Common Causes for more detail. Purpose To create the data necessary to indicate when an intervention in the Process may be required. Quantity The Purpose of a Process shall be measured using the following parameters:  Effectiveness (output units on spec per time period)  Efficiency (Recovery, Unit Cost, ROI etc)