Benchmarking: Amelunxen SGI - Agnico Eagle

Benchmarking: Amelunxen SGI Specific Energy Consumption - Agnico Eagle Laronde

Starkey, J., Robitaille, J., Cousin, P., Jordan, J. and Kosick, G., Design of the Agnico-Eagle Laronde Division SAG mill. Proceedings of SAG 2001, pages III-165 to III-178.

Survey conducted shortly after start-up

Design Criteria

Actual operating data from Esperanza

• A×b of Esperanza ore (Aug-Dec 2012); avg= 39; range is 30.7 to 45.5.
• Wi_BM average is 16.0 kWh/st, = 17.6 kWh/tonne.
• E_SAG for "normal" blasting, average = 5.5 kWh/tonne; range is 3.9 to 7.2 kWh/tonne.
• E_SAG for "high intensity" blasting typically 10% lower E_SAG.
• SAG feed size range 33.6% to 49.0 percent passing 1¼ inch (32 mm).

Design criteria for expansion cases (which may reflect Esperanza operating conditions)

• SAG circuit F₈₀ = 132 mm
• transfer T₈₀ = 3500 µm (text erroneously says "3500 mm")
• BM circuit P₈₀ = 180 µm
• SAG mill power utilization = 92.3%
• Ball mill power utilization = 95%
• Plant availability = 93%

Mill criteria

• SAG mill: 40 foot nom dia by 26 foot flange-to-flange cylinder length
• ball mill: 27 foot nom dia by 45 foot flange-to-flange cylinder length (2 units)
• SAG mill gearless motor 31,500 HP (given in Thiele & Parraguez, Relpas 2011 conference)
• Ball mill gearless motors 25,000 HP

(some references give 30,000 HP for SAG motor; this is thought to be the value excluding a 5% design margin)

Modelling

The Bond/Barratt SABC-B model is used with the default 10% E_ssbm calibration factor.

• circuit F₈₀ = 132 mm
• circuit P₈₀ = 180 µm
• The average ball mill work index given as 16 kWh/short ton = 17.6 kWh/tonne
• The relationship between A×b and Wi_RM for Andean copper porphyries can be used to determine the following rod mill work index values.
  • A×b of Esperanza ore; avg= 39; becomes Wi_RM=14.5 kWh/tonne
• No crushing work index is given. Assume a value of 12 kWh/tonne (slightly higher than typical porphyry)
• No ore density is given. Assume similar to Los Bronces, density=2.64 t/m³

SAG mill is modelled using following:

• Austin model
• assume ore density 2.64 t/m³, same as Los Bronces
• nominal diameter 40 foot, effective grinding length 26 foot (deduct 2 ft from cylinder length)
• assume 6 inch liner effective thickness, mill speed 75% of critical
• ball charge 16% v/v; total filling 26% v/v gives the desired 92.4% utilization of motor power

Ball mill is modelled using following:

• Nordberg wet overflow model
• assume ore density 2.64 t/m³, same as Los Bronces
• nominal diameter 27 foot, effective grinding length 44.5 foot (deduct 6 inches from cylinder length)
• assume 6 inch liner effective thickness, mill speed 75% of critical
• total filling 30% v/v gives the desired 95% utilization of motor power

• 

  Published A×b and Wi_RM results for Andean porphyries

• 

  Model configuration for benchmarking

• 

  Model configuration for benchmarking

• 

  Esperanza benchmarking circuit model screen grab

Results

Only one ball mill work index is provided, claimed to be the 'average'. The predictions of the min/max range are dubious because the ball mill work index should change between hard and soft ore.