A&B Mylec has developed a unique way of generating reliable washability data from crushed borecore data. Such data was previously unsuitable to use for generating practical wash plant yield and product quality information.
The process has been developed and utilised over dozens of resources spanning most major coal exporting basins across the world.
This presentation steps through a case study of the unification process being employed on a recent deposit. Completing this exercise allows for much improved washability data and as a result, significantly increased confidence in modelling outcomes.
Simulation comparison between Unified / Non-unified
Crushed washability data has two distinct issues for use in practical yield simulations:
- A size distribution that does not represent the mass proportions feeding each of the CPP circuits;
- A single density distribution is used in both the DMC and RC circuits which is different from the naturally liberated density distribution;
A&B Mylec has developed a technique to repair crushed washability (non-unified) data using properly pretreated reference data so that the repaired (unified) data represents the size and density distributions fed to a wash plant.
As a comparative exercise, three datapoints from seams with distinctly different washability profiles were selected for comparison:
- D141: Good washability performance;
- H16: Average washability performance;
- H15: Poorer washability performance;
Data Unification Processes
The following processes were utilised in order to unify the washability database:
- Float/sink density interpolations:
- The interpolation and extrapolation of each washability series so that all series have a common 12 density fractions (F1.30, F1.35, F1.40, F1.45, F1.50, F1.55, F1.60, F1.65, F1.70, F1.80, F2.00 & S2.00);
- Unnecessary density fractions were combined;
- Combination of pre-treated washability fraction to achieve a common series of size fractions representative of the CHPP, namely:
- DMC (+1.0mmww);
- Reflux classifier (1.0ww x 0.250mm);
- Flotation (-0.250mm);
- Development of Liberation models.
- Development of Circuit Segregation models.
Converts crushed washability data into pseudo-pretreated washability data:
Circuit Segregation Models:
Distributes pretreated washability data into designated size fractions:
Simulate the reparied data set:
Product ash/yield trends - D141 (target 10.5%):
A sample with good washability performance will not be as affected by the difference between unified and non-unified data.
The offset between pretreated (unified) data and crushed (non-unified) data is not consistent within seams or across seams.
Using the repaired dataset further analysis can be undertaken on the practical yields and ash for blending opportunities to increase total yield at the same product ash.
Spatial analysis can be completed for different mining areas.
PRESS RELEASE - ACPS 15TH BIENNIAL CONFERENCE & EXHIBITION
3 APRIL 2014
The Australian Coal Preparation Society's (ACPS) 15th National Conference "Continuing Excellence in Coal Preparation" will be held at the Gold Coast Convention & Exhibition... more