At the same time, most of the planning tools in use today still function on deterministic inputs – producing a single result for each activity. Thinking of the system holistically, understanding a process means understanding its capacity, constraints, and variability, which in turn allows mining companies to frame goals as ranges rather than fixed points.
This raises an important question for long-term and tactical planning: if performance is inherently variable, should schedules reflect that variation as well?
Mining operations continuously collect production data for reporting. Planning engineers can use that data to build probability distributions for availability, utilization, and task durations, creating an opportunity to improve our understanding and communication of risk. Before diving into the specifics of building these distributions, it’s worth asking – why plan for variation at all?
Planning variation into schedules allows us to make informed decisions about risk and opportunity. Running hundreds of probabilistic scenarios can reveal to us which tasks pose the greatest risk, where the schedule is most sensitive, and where we can confidently invest extra time.
Analyzing production data presents an opportunity to distinguish between meaningful signals, and background noise. Instead of reacting to fluctuations in daily performance, we can make decisions with statistical confidence, ensuring we stay focused on long‑term goals rather than overcorrecting for normal, expected variation.
Engineers can build probability distributions to inform key schedule variables: jumbo production rate, equipment availabilities, utilizations, and production rates. This shift in rates and durations allows us to model a range of possible outcomes, shifting from assuming a single deterministic path to understanding a spread of likely results.
Once distributions are established, they can be input into Deswik Planning’s Variations tool to start evaluating scenarios for our mine plans. Hundreds or thousands of scenarios can be run in Variations, and tools such as Microsoft Power BI can be used to aggregate the results and convert raw outputs into planning insights.
In figure 2, we can see that 2026-2028 have relatively tight maximum/ minimums, while 2029 and 2030 are much more sensitive to utilization. Tools like Power BI allow us to isolate these scenarios and take a deeper look.
Evaluating the scenarios generated by Deswik Variations helps identify which tasks consistently appear on the critical path. This information is invaluable to guide the planning priorities. It enables engineers to focus resources on the activities that are most critical to the plan’s success. In contrast, deterministic planning provides only a single point of reference for our critical path – one which can shift under normal production variability.
Aggregated scenarios can be evaluated to track key tasks within each scenario to provide a confidence range. The above chart shows a wide gap in start date between P25 (the point where 25% of outcomes finish earlier) and P80 (the point where 80% of outcomes finish earlier), highlighting the risk present in our scenarios. This allows the planner to evaluate the chain of tasks leading to the ore bin, and plan contingencies and trade-off to narrow the gap.
Embracing inherent variation in mine planning improves stakeholder communication and supports more resilient planning and decision making, helping teams build schedules that better reflect the realities of mining.
Do you incorporate probability in your mine plans? Comment below to explain how you’re applying it, and if you’re leveraging Deswik’s Variations tool, included with Deswik Planning.
Understanding Variation: The Key to Managing Chaos by Donald J. Wheeler is an excellent resource in the world of variation and provides excellent insight into the use of variations to manage unpredictability and risk. It offers practical examples of how to analyse and interpret variation across different business contexts.