Simulation uses a model to describe product flow – product flow is defined by rate and quality. Activities within the product flow are used to alter rate and quality.
In the process environment Activities are affected by ‘decisions’ and ‘effectiveness’. Decisions are a function of knowledge and experience (ability to interpret conditions) whereas effectiveness / efficiency is a function of process, equipment and systems.
Knowledge and experience are domain specific - furthermore the level at which decision-making is applied requires different sets of knowledge and experience. The consequences of decision-making ripples up and downward in the decision hierarchy.
Decisions at the top of the decision hierarchy include setting operational objectives (aimed at realizing strategy). This requires lower levels in the decision hierarchy to formulate actions to ensure compliance with objectives i.e. operational decision-making at execution level needs to be aligned with objectives. Sub-optimal or poor decisions at the lower end of the decision hierarchy in turn ripple upwards in the hierarchy requiring decisions to counter disturbances.
Data, ability to interpret data and consistency of interpretation impact operational risk exposure. Operational risk is defined as the potential loss consequential to the failure of systems, processes, control and / or incorrect decisions. Complexity and variability, synonymous with processing plants, complicate decision-making.
In summary, to increase overall performance organizations need to improve decision-making throughout the decision hierarchy and decision-making needs to be aligned. Quality decision-making at the lower end of the hierarchy delivers higher predictability / stability and presents a great opportunity to increase efficiency and effectiveness without the need of capital investment.
Performance advances have been constrained by the way performance is managed within processing plants. It is common to see shift (daily, monthly) production targets. Reporting a totalised value has been ‘good enough’ in the past to facilitate decisions higher up in the decision hierarchy. However, aggregated values mask inefficiencies within the operation.
Effective decision-making at operational level presents major financial benefits. Substantial improvements are attainable almost immediately when understanding ‘how’ aggregated performance has been achieved. Within a processing plant aggregated performance is made up of intervals within the reporting period of exceptional good, average and poor performance!
Worthwhile to note is that operations go through intervals of good, average and poor performance while applying the same process, systems, controls and people. Although feed material characteristics can affect performance, our experience show variable performance even when processing material with similar characteristics. Analysing data at a granular level reveals the source of variable performance to be decision-making and actions (or lack of action) i.e. a mismatch between decisions and conditions observed within the process, systems and equipment.
When assessing processing plants, results typically confirm opportunities to realize recovery improvements of more than 1%-point, production throughput improvement of more than 5%, energy efficiency improvement of more than 5%, quality and availability improvement. These opportunities fall into the category of efficiency and effectiveness improvement, not requiring capital – virtually all value realized by this type of improvement reports to the bottom line. Realization of these opportunities are in most cases foiled by an inability to provide actionable information to the operational team to manage process, system and equipment conditions pro-actively.
Positioning Time-in-State® for Production Decision-Making
Operational objectives are only as good as the ability to comply with them. Objectives are typically set for production rate (throughput), quality, availability (and others). Several systems, processes and equipment make up the production process. To this point, operations have been heavily reliant on the knowledge and experience of operational personnel (production and engineering) to manage process, equipment and systems to deliver the required results. Evidence of the latter is found when analysing performance of different production teams – it is common to find shift teams preferring certain set points and / or to find that one team performing better than others.
Variable performance can be traced back to proper ‘translation’ of operational objectives into appropriate actions applied within production. In other words, aligning management and control of each system, process and equipment with operational objectives. This is particularly challenging in an environment where it is difficult to relate cause and effect (a common problem in processing plants). This highlights the need for a solution and performance metric that address the following:
Time-in-State® has been developed to address the above listed requirements. Time-in-State® consists of a MESA endorsed methodology and suite of products (software) facilitating the implementation of each aspect of the methodology.
Positioning Time-in-State® for Management Decision-Making
As mentioned earlier, simulation describes product flow in terms of rate and quality. Variability is introduced into the simulation by configuring some distribution into a variable that affects the rate and quality of the product at one or more point in the model – this is perfect for off-line simulation. Time-in-State® is the platform that will transfer the simulated environment into a model that provides real-time information required for decision-making.
To illustrate the real-time application, it is best to refer to an actual example – the example is a mining and mineral beneficiation process. Three systems represent the production chain – see Figure 1.
A selection of parameters, derived through analysis, are used to describe system behaviour. Evaluating historical data generates a definition of the Optimum Operating Envelope (OOE) for each system. Having derived the OOE for each system it is possible to generate the following information:
Using results from 1(a) and 1(b) one can derive that System B and C were able to counter disturbances introduced by System A. Irrespective of the latter, System A requires attention – this information justifies where to focus attention / invest time and money. (In this case Time-in-State® is applied to perform a detailed analysis of System A)
This illustration confirms the Time-in-State® platform’s versatility. Models can with ease also monitor aspects such as:
Simulation and real-time modelling can be applied within processing plants with great affect. Both applications emphasize the importance of having the right data and knowledge to execute interpretation. From a decision-making point of view, it is only possible to arrive at the same conclusion and actions if data is interpreted consistently.