Coal Preparation » Process Control
The objective of this project was to develop, implement and demonstrate a system that determines the optimal DMC cut‐point at which a target product ash and/or a given incremental ash can be achieved. The operation of dense medium cyclone (DMC) at the optimal cut‐point would reduce the amount of product with out‐of-specification ash and maximise the yield of a plant with multiple parallel beneficiation units.
The system combines online measurements of medium densities and DMC product yield with predictive models for on‐line determinations of the partition curve, the feed coal washability and the product ash and/or incremental ash. The partition curve and the plant feed washability are determined using methods developed in two previous projects (C13058, C20049). The system will help the control system to get the most out of limited measurements by determining the current cut‐point and predicting the optimal cut‐point.
Methods for predicting instantaneous ash curve for known and unknown coal sources were developed. A versatile empirical equation was developed for fitting any type of instantaneous ash curve. It was found that the averaged instantaneous ash curve from bore‐hole logs with weighted coefficients related to the seam thickness in each hole for a given seam or a seam mixture with a constant blending ratio has a sufficient accuracy for predictions of the product ash and the incremental ash, and this is valid even for coal feeds with intra‐seam or out‐of‐seam dilutions. It was also found that the washability curve from the bore‐hole logs is still useful for predicting the product ash in cases where the coal feed has a significant intra‐seam dilution from shale bands as long as the cut‐point is lower than 1.6RD.
The effects of errors in instantaneous ash curve on the predictions of the product ash and incremental ash have been thoroughly investigated using simulated cases. A ±1% error (absolute error) in the ash value at the density of 1.27RD on the instantaneous ash curve can cause about ±0.8% error (absolute error) in the predicted product ash value. Errors at the density of 1.65RD on the instantaneous ash curve have an insignificant impact on the prediction.
Results from plant‐based experiments indicated that the predicted product ash values were close to those from sampling and laboratory analysis for coal feeds from a mixture of seams or a single seam. The plant-based experimental results demonstrated that the method was effective in the simultaneous determination of the plant feed washability and the instantaneous ash curve through the perturbation of the correct medium density and accompanied by product sampling and rapid ash analysis. The method for determining the optimal cut‐point for a given product ash or incremental ash was successfully demonstrated in plant experimental cases.
The methods developed in this project are useful not only for online applications, but also for optimisation through sampling. Data from Marcy density gauge and rapid ash analysis are collected at three correct medium densities: i.e. current set‐point, +0.05RD higher and ‐0.05RD lower, and then used to estimate the instantaneous ash and the washability curves of unknown coal feed sources. Finally the cut‐point for the targeted product ash and/or incremental ash would be found.
The system and methods have considerable potential to add value to the Australian coal preparation industry by maximising the plant yield. Applications of the system and methods at mine sites require further improvements and integration of the system with existing plant systems. Development of an Excel Spreadsheet based program is also desirable for the optimisation through sampling.