Benchmarking: BondBarratt - Detour Lake

Benchmarking: Bond/Barratt SAG Mill Specific Energy Consumption - Detour Lake

• J. Torrealba-Vargas, J.-F. Dupont, J. McMullen, A. Allaire and R. Welyhorsky, The successful development of the detour lake grinding circuit: from testwork to production. Proceedings of the SAG 2015 Conference, September 2015, Vancouver, Canada, Paper 38.

Operating data from 2015, A×b looks similar to 50^th percentile from the October 2012 NI43-101 report.

"Current value" data, table 7 from the reference

• Throughput = 2×1,258 = 2,516 dry tonnes per hour
• SAG circuit F₈₀ = 90.7 mm (pre-crusher product)
• circuit P₈₀ = 90 µm

• SAG power draw (at motor input, per mill) = 13,424 kW (at mill shell = 12,694 kW)
• SAG mill ball charge = 17 % vol
• SAG mill total charge = 24.3 % vol
• SAG mill discharge % solids = 70% weight
• SAG mill speed 80.24% of critical
• E_SAG at mill shell = (2×12,694)/2,516 = 10.1 kWh/t

• pebble crushers typically draw 322 kW
• E_peb = (2×322)/2,516 = 0.26 kWh/t

• ball mill power draw (at motor input, per mill) = 13,982 kW (at mill shell = 13,221 kW)
• ball mill % solids = 80% weight
• ball mill total charge = 25.4
• ball mill speed = 79.61% of critical
• E_bm at mill shell = (2×13,221)/2,516 = 10.5 kWh/t

Design Criteria

• Ore A×b = 26.6, between 25^th and 50^th percentiles
• Wi_BM use 50^th percentile = 14.3 kWh/tonne.
• Wi_RM use 50^th percentile = 18.0 kWh/tonne.
• Wi_C use 50^th percentile = 14.8 kWh/tonne.
• Plant availability = 87%

Mill criteria

• SAG mills: two 36 foot nom diam by 17.25 foot effective grinding length (EGL)
• ball mill: two 25 foot nom diam by 38.0 foot EGL (flange-to-flange length 38.5)
• SAG mills, each twin variable speed motor 7000 kW (each mill)
• Ball mill, each twin (assumed variable speed) motors 7500 kW (each mill)

Modelling

SAG mill is modelled using following:

• Austin SAG model
• ore density 2.9 t/m³
• SAG liners nominal 4 inch liner thickness
• liner thickness based on JK Simulation in Table 3
• effective dimensions 10768 mm diam × 5334 mm EGL
• predicted SAG power draw at mill shell = 13,448 kW (5.9% higher than actual)

Ball mill is modelled using following:

• Nordberg ball mill wet overflow with density correction
• ore density 2.9 t/m³
• ball mill nominal 90 mm liner thickness
• liner thickness based on JK Simulation in Table 3
• effective dimensions 7439 mm diam × 11581 mm EGL
• predicted ball mill power at mill shell = 13,273 kW (0.4% higher than actual)

Results

Result for default Optimized Bond/Barratt model conditions:

	Esag	Epeb	Ebm	Etotal	t/h
Predicted	9.0	0.2	8.8	18.0	3004
2015 data	10.1	0.3	10.5	20.9	2516
Difference	1.1	0.1	1.7	2.9	508
Difference	model 12% low	model 33% low	model 16% low	model 14% low	model 20% high

Discussion

The mill is in the final stages of ramp-up, and has not reached its ultimate capacity. Figure 5 shows they are consistently exceeding 2,700 t/h in April & May of 2015, suggesting they can do 7% more than the data in Table 7.

If we claim that 7% difference between April/May and the 2015 data, then the differences become:

• Etotal : actual = 19.5 kWh/t, model 7% low
• t/h : actual = 2700 t/h, model 13% high

The ultimate capacity of this circuit hasn't been established yet, so it could very likely continue to improve and approach the model numbers.

Trying the Raw Bond/Barratt model and comparing to the April/May 2015 values, the model predicts:

• Etotal : 19.7 kWh/t, model 1% high
• t/h : 2750 t/h, model 2% high