By Dr Mark Cooksey
Highly aggressive environments are synonymous with metal production, and obtaining accurate measurements of your processes can be challenging. High temperatures, corrosive materials, dust and logistics can all impact your ability to measure your plant process parameters.
Difficulty measuring key parameters can prove detrimental to your plant’s finances. There is tremendous process variation in a plant environment, and there are significant consequences from erroneously assuming parameters are constant when they are changing, or changing when they are constant. Problems can result from either no measurement or measurement error.
In our experience, most process engineers responsible for plant process feel limited by their capability to reliably measure the right parts of a process.
The old maxim, “if you can’t measure it, you can’t improve it” holds true, but in a plant environment, there is greater complexity. The reality is, if you can’t measure it you can’t prevent it from failing. And, if you can’t measure it, it’s going to cost you.
An inability to accurately measure key process variables primarily impacts a plant’s operating costs.
A common challenge in today’s processing plants is coping with changes in raw materials, as economic pressures are leading companies to have a larger number of raw materials sources. You may know that a new raw material will impact your process, but if you can’t measure key parameters quickly, it may take weeks before you learn that your efficiency has reduced significantly.
For example, an aluminium smelter modified the way it managed alumina in the plant, which resulted in a reduced particle size of the alumina. The plant was controlling the volume of alumina fed to the pots, so the smaller particle size meant that more alumina was being fed than thought. This led to the accumulation of alumina powder in the pot, and lower efficiency. It took a few weeks to detect the scale of the problem, and a couple of months to rectify it.
If the plant had recognised the key inputs that required measurement and had the capability to measure them, this incident could have been prevented and the cost to rectify the error avoided.
You can’t measure everything. There may be countless variables to measure in your plant, but this doesn’t mean you should measure them all.
Firstly, it’s worth taking the time to prioritise which parameters you should invest in measuring.
Secondly, never rely on flawed measurement. Responding to inaccurate data is likely to be worse than having no data. You will be reacting to false signals and introducing variation into the process. For example, a control system in an aluminium smelter was responding to 1°C changes in temperature, but the temperature measurement was only accurate to within ±3 °C. The control system was introducing variation, rather than reducing it.
At CSIRO, we’ve developed new measurement methods that are more capable, cheaper, and faster.
Being able to monitor the key parameters in your processes ensures you’re able to reduce operating costs through continued process optimisation.
Furthermore, good measurement is also critical for validating computer models to ensure they represent reality. In our March blog, we discussed how plants could improve their processing via predictive modelling, and a key means of validating the model is through accurate measurement.
If your plant has good measurement data that validates the modelling, you’re capable of using the model more reliably and accurately. Good data ensures your models have a robust framework, allowing you to test different processes in your plant without the risks associated with a live process.
Overall, the ability to measure your plant process parameters is key for ensuring your plant can perform optimally while minimising any financial risks.
For further information about how CSIRO can support you in measuring your plant’s process parameters contact the Responsible Metal Production and Recycling team on +61 3 9545 8865 or email Mark.Cooksey@csiro.au.