Bibliography: Benchmarking

Bibliography: Benchmarking of models

The following list of references show data useful for benchmarking against the different models.

Benchmarking: Circuit Specific Energy Consumption - Bond/Barratt

• Magnuson, R.; Hallow, J.; Mosher, J.; Major, K., The Fort Knox Mill: Design, Commissioning and Operation. Proceedings of the SAG 2001 Conference, Vancouver, Canada.

Result for default model conditions:

	Etotal	WiO		Tonnage
Model	11.95	16.60	kWh/t	1,522	t/h
Measured	10.50	13.97	kWh/t	1,733	t/h
Difference	1.45	2.63	kWh/t	211	t/h
Difference	12.9%	17.2%		12.9%

Show details of benchmarking

Benchmarking: Circuit Specific Energy Consumption - Bond/Barratt

• Morrison, R., Current Plant Conditions at Copper Mountain. Presentation to the BC/Yukon Branch Canadian Mineral Processors, November 29, 2012; Vancouver, Canada.

Result for default model conditions:

	Tonnage
Model	1455 t/h
Measured	1600 t/h
Difference	145 t/h
Difference	9.5%

Show details of benchmarking

Benchmarking: SAG Mill Power Draw - Fimiston

• Nelson, M; Valery, W; Morrell, S, Performance Characteristics and Optimisation of the Fimiston (KCGM) SAG Mill Circuit, Page 233 - 248, SAG 1996 Conference, Vancouver, Canada.

Survey	Survey Power, kW at input	Mill speed, %critical	Ball load, %v/v	Total load, %v/v	Pulp %solids, w/w	Morrell SAG Model, kW at input / shell	Loveday/Baratt Model, kW at input / shell
Survey 1	9,255	72.5	13	21.6	65.9	9,268 / 8,764	10,372 / 9,808
Survey 2	10,374	77	13	25.2	63.3	10,481 / 9,911	11,238 / 10,636
Survey 3	8,395	75	15	16.1	70	9,193 / 8,693	10,818 / 10,230
Survey 4	8,299	78	11.5	13.5	60	7,766 / 7,344	can't do
Survey 5	10,976	80	11.5	22.2	60	9,877 / 9,340	10,322 / 9,760
Survey 6	8,616	74	11.5	13.6	63	7,394 / 6,992	can't do
Survey 7	11,684	82	14	20	75	10,782 / 10,195	11,037 / 10,437
Survey 8	11,610	78	13	28.6	75	11,415 / 10,794	11,125 / 10,520
Survey 9	11,571	78	13	25.8	75	10,990 / 10,392	11,668 / 11,033
Survey 10	9,408	80	12	19	75	9,655 / 9,130	10,309 / 9,749

• The overall average difference between the Morrell model motor input predictions and the measurements are -3.4%, and range from -15.3% to +9.1%.
• The overall average difference between the Loveday/Barratt model motor input predictions and the measurements are +4.8%, and range from -6.1% to +25.2%.

See details of benchmarking

Benchmarking: SAG Mill Power Draw - Cadia

• Radziszewski, P.; Valery, W, Cadia SAG Mill Simulated Charge Behaviour, Annual General Meeting of the Canadian Mineral Processors, Ottawa, 1999.

Survey	Survey Power, kW at shell	Mill speed, %critical	Ball load, %v/v	Total load, %v/v	Morrell SAG Model, kW at shell	Loveday/Baratt Model, kW at shell
Survey 1	11,189	79	0	28.8	11,868	12,126
Survey 2	10,321	79	0	28.5	11,787	12,039
Survey 3	10,824	78	4	25	12,762	13,390
Survey 4	14,945	78	4	40.7	15,806	15,096
Survey 5	17,586	74	12	31.6	17,351	18,216
Survey 6	17,963	78	12	26.1	17,298	18,505

• Morrell model predicts, on average, 6% high.
• Loveday/Barratt model predicts, on average, 9% high.

See details of benchmarking

Benchmarking: SAG Mill Power Draw - Cadia

• Boghey, A.; Svalbonas, V.; Jones, S.M., Supply, Installation & Commissioning of the World's Largest Grinding Mill, Annual General Meeting of the Society for Mining, Metallurgy & Exploration (SME), 2000.

Mill filling level 33% v/v. Ball charge not measured, but expected to be in 12% to 13% v/v range. Ore density not indicated, assuming 2.65 kg/L. Ball density given as 7.85 kg/L.

Survey conducted during late commissioning, Figure 8 yields:

Survey	Survey Power, kW at shell	Mill speed, RPM	Mill speed, %critical	Morrell SAG Model, kW at shell	Loveday/Barratt Model, kW at shell
23:30	19,300	8.99	73.7	17,148	18,233
23:40	19,450	9.4	77.1	18,042	19,199
23:50	19,550	9.9	81.2	19,058	19,868

Benchmarking: Ball Mill Power Draw - Fimiston

• Nelson, M; Valery, W; Morrell, S, Performance Characteristics and Optimisation of the Fimiston (KCGM) SAG Mill Circuit, Page 233 - 248, SAG 1996 Conference, Vancouver, Canada.
  • Diameter inside shell = 5.49 m (18 ft)
  • Diameter inside liners = 5.35 m (17.5 ft, 3.0 inch effective liner thickness)
  • Belly length inside liners (EGL) = 7.60 m (25 ft)
  • Centre-line length = 8.76 m
  • Top ball size = 80 mm

Table 5 presents results of a single ball mill survey. The survey measured motor input power. Drives are assumed to have an efficiency of 0.96 and gearbox+pinion efficiency of 0.970, so the model shell power draw is converted to motor input power by dividing by 0.9312. The predicted power draw of Example project circuit number 7 (Fimiston) using sample MLE, based on the KCGM paper published by Campbell, J. et al; 1998 AusIMM Annual Conference.

Survey	Survey Power, kW at input	Mill speed, %critical	Total load, %v/v	Pulp %solids, w/w	Morrell SAG Model, kW at input / shell	Nordberg Model, kW at input / shell
Survey 1	3,864	66.7 68.3 †	38.7	72.0	3,933 / 3,776	3,592 / 3,345

^† The appendix of the paper lists the mill speed as 12.5 RPM. The mill is fixed speed, so the %critical speed is only a function of mill effective diameter (as liners wear). Doing the math (neglecting the balls) gives a 68.3% critical speed.

The ball mill belly length can be achieved with a 18 degree head angle and 1.9 m trunnion diameter.

Historic & Other Interesting Benchmarking

1. Myers, J.F., Michaelson, S.D., Bond, F.C., Rod Milling—Plant and Laboratory Data, Technical Publication No. 2175, American Institute of Mining and Metallurgical Engineers, 1947. [1]