1
Expansión Antapaccay – Tintaya Dante García* & Juliano Villanueva Concentradora Antapaccay, Concentradora Antapaccay, Xstrata, Peru
ABSTRACT The construction phase of the Glencore Xstrata’s 70 000 tonnes per day copper concentrator near Cusco, Peru ended in 2012. Plant start‐up was accomplished during the last quarter of 2012. Comminution is achieved by achieved by primary crushing, SAG milling and pebble crushing, followed by followed by ball ball milling. The Concentrator is a single line configuration, comprised of one 40 ft. diameter x 22 ft. long (EGL) SAG mill driven by driven by a 24,000 kW gearless drive. The SAG mill feeds two ball two ball mills, each 26 ft. in diameter x 40 ft. long (EGL) with each driven by driven by a 16,400 kW gearless drives. This expansion represents the first Glencore Xstrata Standard Concentrator, which is the first 40 ft. SAG mill in Peru, and which features the highest power SAG mill in the world, a 6.5 km overland conveyor with gearless drive, the use of an old pit as a tailings dam and the highest torque tailings thickener in existence. This paper reviews the history, background, grinding circuit design and operations start‐up focusing on safety, start‐up strategies, main issues and improvement opportunities. The main strategies used were people recruiting and training, benchmarking training, benchmarking related to other recent start‐ups, strategic contracts with vendors and consultants, phases for start‐up and reagent strategy.
RESUMEN La etapa de construcción de la Planta Concentradora Antapaccay para minerales de cobre con capacidad de tratamiento de 70 000 TM/día, perteneciente a Glencore‐Xstrata, terminó a fines del año 2012. El inicio de operaciones se efectuó durante el último trimestre del año 2012. Las operaciones de conminución se efectúan con el chancado mediante una chancadora giratoria; luego con un molino de bolas tipo SAG; chancado de Pebbles mediante chancadoras cónicas y finalmente molienda fina con molinos de bolas de bolas convencionales. La Planta Concentradora está diseñada según configuración en línea simple. Incluye un molino SAG de 40 x 22 pies, (EGL), accionado por un motor tipo “gearless” con potencia instalada de 24,000 kW. Este molino alimenta a dos molinos de bolas de bolas , , cada uno de 26 pies de diámetro x 40 pies de longitud, (EGL), accionados cada uno por un motor tipo “gearless” de 16,400 kW. Esta expansión corresponde al primer Concentrador Estándar de la Empresa Glencore‐Xstrata, que incluye el molino SAG más grande del Perú, cuya potencia de motor corresponde al más grande del mundo. Se integra además una faja tipo “Overland” con una longitud de 6.5 km, accionada también con motores tipo “gearless”. Para el depósito de los relaves, se utiliza el tajo agotado de la Mina de Tintaya Este artículo describe la historia, antecedentes, características de diseño del circuito de molienda y las operaciones del inicio de producción. Se enfatiza en los aspectos de seguridad, estrategias de arranque de planta y oportunidades de mejora. Destacadas estrategias fueron las relacionadas a selección y entrenamiento de personal, de generación de contratos con los proveedores y consultores, fases de arranque, uso de reactivos. Todo ello se contrasta muy favorablemente con el arranque de plantas concentradoras similares.
2
1930’s: Skarn mineralization, located on Atalaya in the northwest margin of Antapaccay South, has been mined sporadically since these years. 1985: Tintaya commenced production. 1996: Tintaya was acquired by BHP Billiton following their acquisition of Magma Copper Company. 1998: The Antapaccay porphyry centres were discovered. 2006: Xstrata Copper acquired Tintaya and the Antapaccay copper‐gold prospect from BHP Billiton. 2010: Xstrata approved a US$ one and a half billion investment in Antapaccay following approval of the Environmental and Social Impact Assessment by the Peruvian Mining Ministry.
Our Antapaccay open pit mine and processing facilities are located 4,100 metres above sea level in the Yauri district of Espinar Province in southern Peru’s Cusco Region. Antapaccay is situated approximately 10 kilometres from Tintaya. Figure 1 shows Antapaccay’s location.
Figure 1: Antapaccay – Tintaya location
BACKGROUND Social management Tintaya‐Antapaccay’s Social Management involved the following voluntary dialogue and agreement processes:
3
Framework Agreement 2003: Voluntary agreement between Espinar province and Tintaya occurred in 2003. A Citizen Participation Process was conducted with 72 communities to identify and carry out sustainable development projects. Meetings with each community were held at least once a quarter to discuss / define project implementation. Development Table: A Dialogue Process was held with six communities in the direct area of influence (Espinar District) by CooperAcción; CONACAMI, OXFAM and Xstrata Tintaya. The Dialogue Process was designed to identify projects that would enhance development and coexistence with the communities in the direct area of influence of our operation. Bi‐annual participatory monitoring and at least quarterly meetings are ongoing. Cañipía Bilateral Process: A Bilateral Dialogue was held with the four communities in the Cañipía Micro basin (Espinar District) and Xstrata Tintaya. The Bilateral Process included a comprehensive programme to improve cattle for the livestock farming development of the Cañipía Micro basin. Bi‐ annual participatory monitoring and at least quarterly meetings are ongoing. Regional Mining Fund: Representatives of the Cusco region, UNSAAC and Xstrata Tintaya agreed to a Five‐year fund that contributes to the development of the Cusco region, under the Solidarity with the People Mining Programme that was signed with the central government in 2006.
Mineralisation The copper mineralization of the Antapaccay project is located mainly in intermediate intrusive rocks as dissemination, veining, and hydrothermal breccias and in contact with pre‐mineral rocks like diorites, and sedimentary rocks (limestone, chalky lutites, limonites and sandstones) that formed mineralized breccias through contact, exoskarn and “stockwork” in sedimentary rocks, with a marked domination of chalcopyrite over bornite up to 350 m deep. At greater depth, this role is inverted and is associated with a level of gypsum‐anhydrite. Two apparently isolated bodies have been identified. The south body is the larger and extends 1 300 m in a NW‐SE direction and varies in width from 250 until 430 m. The north body measures 300m NW‐SE and is 450 m wide. Contact with limestone generates the conditions for metasomatism to occurring generating irregular bodies of granite‐magnetite+/‐pyroxene exoskarn with patches of mostly chalcopyrite. In addition, zones of intense veining of grey quartz “stock‐work” with a strong content of bornite and chalcopyrite always occurs near hornfels‐intrusive contacts, reaching several meters into the hornfels. Based on the Geo‐metallurgical and Resources model, the project’s economic copper mineralization is distributed in the following form: 1.9% Cu oxide, 70.0 % Cu sulphide porphyry, 12.9% mixed porphyry, 6.5% porphyry‐ breccia‐gypsum, 5.9% mineralized breccia and 2.8% skarn, with the highest grades found in the mineralized breccias and the skarn. Figure 2 shows Antapaccay’s geology. Ore Type Sample Definition Geo‐metallurgical ore types with the prefix “UGM” were defined by the project geologists based on mineralization with seven UGMs (1‐7) initially identified for metallurgical purposes. These ore types formed the basis for comminution and flotation test work.
4
Figure 2: Antapaccay geology During the flotation variability program, an eighth ore type was identified. This one contained native copper which subsequently had a detrimental effect on copper flotation recovery. Table 1 shows the description of each ore type, its relative tonnage, distribution and copper head grade (based on geological core logs and resource model at beginning of metallurgical test program) representing the LOM Life of Mine. Table 1: Antapaccay ore type UGMs UGM Description
Ore kt
Ore %
% Cu
1
Porphyry Sulphide
328 035
65.9
0.56
2
Porphyry Mixed
55 255
11.1
0.63
3
Porphyry Breccia
31 702
6.4
0.54
4
Breccia Mineralisation
29 098
5.8
1.11
5
Skarn
13 678
2.7
0.98
6
Other
22 963
4.6
0.42
7
Oxidised
9 002
1.8
0.42
8
Porphyry Mixed with Native Cu
8 345
1.7
0.42
498 078
100.0 0.60
Total
The main ore types, UGMs one and two have very similar characteristics. In terms of competency, both could be considered as regular with to respect to SAG grinding and hard with respect to ball grinding (Dechert, 2006).
5
The majority of the ore types can be classified as being of medium hardness and low abrasiveness. Table 2 shows the average results for each ore type from the characterization program. Table 2: Summary of comminution results from characterization program Ore Type UGM 1 UGM 2 UGM 3 UGM 4 UGM 5 UGM 6
JKDWT JKDWT A*b ta 47.7 60.2 42.9 54.6 66.6 48.6
0.8 1.5 1.0 0.8 1.0 1.1
SMC Dwi (kWh/t)
SMC Axb
SPI (min)
LEIT (kWh/t)
AI
5.7 5.5 5.8 5.8 4.8 6.6
46.4 49.4 45.3 51.6 63.2 46.3
79.4 74.5 72.7 72.4 61.0 90.8
5.6 5.7 5.2 7.9 7.3 8.2
0.1 0.2 0.1 0.1 0.0 0.1
BRWI BBWI (kWh/t) (kWh/t) 13.5 13.5 11.6 12.1 10.8 14.3
17.6 16.7 15.5 14.7 10.0 14.6
Flotation test work Flotation test work consisted of a characterization program and a variability program which included one hundred and forty seven (147) samples. The most abundant ore types responded well to very simple processing parameters. At primary grind sizes of 150 to 200 μm P80, over 90% of the copper was recovered in concentrates that graded about 37% copper. Gold was 80% recovered in the final concentrate. Regrinding the rougher concentrate ahead of the cleaning stages was determined to be an important factor influencing the copper concentrate grade. However, the concentrates could be processed at regrind sizes of 30 to 50 μm P80. There were differences in metallurgical response between some of the ore types. The porphyry and breccia samples performed well with very little variation in response in individual sub samples. The skarn and other rock classifications demonstrated considerably more variability in metallurgical response. These ore types contained many non‐sulphide forms of copper that negatively impacted metallurgical response. The rock type classification proved useful in delineating metallurgical response and ore hardness for the deposit. The concentrates were essentially free of significant deleterious elements.
CONSIDERATIONS AND GRINDING CIRCUIT DESIGN The Antapaccay mine exploits a deposit of type massive porphyritic with low grade, the reserves are 520 million mineral tonnes with a copper head grade of 0.60%. The mine provided employment to more than 6 000 people during the construction peak and to about 1 460 people in the operating stage. The construction of the project was accomplished between 2010 and 2012, and the life of mine is expected to be 22 years. Progressive closing of the waste dams and the tailings dams from Tintaya mine will be done in conjunction with the Antapaccay operation. The closing of the Antapaccay facilities at the end of the operation implies remediation of the land, the waste dams and the tailings dam. Xstrata Project Development made a strategic partnership with Bechtel to act as the main engineering company for their projects, FLSmidth for mills and flotation equipment, and Siemens
6
for mills motors. Bechtel together with Graña and Montero were the main construction contractors. On the other hand Xstrata itself made strategic procurements of mill liners, grinding balls and flotation reagents.
Circuit design and equipment selection Xstrata Copper in 2008 had six potential large scale copper concentrator projects in their project execution pipeline. One of the largest capital cost items and also the longest lead time items on some of the projects are those of the equipment for the primary grinding circuit. The general engineering philosophy for the selection of such equipment has been to select the largest proven equipment available. This currently corresponds to a 12.2 m (40 ft.) diameter SAG mill and 7.9 m (26 ft.) x 12.2 m (40 ft.) ball mills. With the delivery of this type of equipment becoming critical, the decision was made to place an order to secure a production slot for either a 11.6 m (38 ft.) or 12.2 m (40 ft.) SAG Mill and two 7.9 m (26 ft.) ball mills for one of the first projects. The decision whether to select an 11.6 m (38 ft.) or a 12.2 m (40 ft.) SAG mill for the first project, initiated the recommendation that an evaluation of all of the projects be carried out by industry experts, with a focus on the Antapaccay project. The selection of grinding equipment for any project is generally a complex issue and without the constraints of extended equipment delivery cycles is still normally a critical path item in the design cycle. As such the evaluation methodology to be used in the current economic climate needs to be relatively conservative, but also carried out by experienced personnel (Vanderbeek et al., 2006). An evaluation of the available data for the six projects was carried out by consultants Steve Morrell and John Starkey who were specifically chosen due to their extensive experience in the industry, and also due to the fact that they use two fundamentally different approaches to SAG mill design. Steven Morrell was the initiator of the core model used by JKSimmet, subsequently marketed by the Julius Kruttschnitt Centre, Queensland, Australia, whilst John Starkey was the inventor of the SAG Performance Index (SPI) testing methodology now marketed by Minnovex, Toronto, Canada (SGS Lakefield). Both of these methods have had significant backing by operating companies in the minerals world. A review of available test data was also carried out on a few of the projects by M Erickson of FLSmidth. Additionally Walter Valery from Metso made a throughput forecast using modelling and simulations of the comminution circuit. The primary objective for the reviewers was to evaluate the expected throughput of the different ore types from each project through the proposed 12.2 m (40 ft.) SAG mill and the two 7.9 m (26 ft.) ball mills. The desired throughputs for each project ranged from 80 000 tonnes per day to 120 000 tonnes per day. The primary findings of the grinding evaluation were as follows: The throughputs predicted by the two grinding consultants were in reasonable agreement with each other. All of the other South American projects were SAG limited at rates lower than desired, with some significantly lower than desired. The suggested one 12.2 m (40 ft.) SAG mill and two 7.9 m (26 ft.) ball mill configuration does not fit all of the proposed projects because the SAG limitation reduces the ball mill load, and in some cases the preferred grinding configuration is one SAG and one ball mill. The decision to secure the production slot for a 12.2 m (40 ft.) SAG mill was still the correct decision, but further evaluation of expected SAG mill performance, final power selection, primary grinding circuit configuration and sizing of ball mills is recommended for future projects.
7
During the initial SAG mill performance review it was very quickly identified that desired throughputs for some projects may not be achieved. At the same time, bid adjudications on mill motor sizing and vendor selection were continuing, and the potential for increased motor rating via enhanced cooling systems was raised. The end result was the decision to upgrade the SAG mill motor size to 24 MW, via the addition of extra cooling capability, at a relatively low incremental capital cost. The rated mill motor was increased primarily as a conservative safety measure to allow for increased confidence in predicted throughput based on the preliminary nature of grinding characterization test work completed at that time. Therefore a linear increase in capacity from the data presented above based on the 21.0 MW motor to the proposed 24.0 MW motor should thus not be assumed.
Plant Overview The metallurgical plant consists of a 1 524 mm x 2 879 mm (60 in x 113 in) primary crusher located close to the proposed mine, a 6 500 m overland conveyor to the copper concentrator, a 12.2 m (40 ft.) Semi‐Autogenous Grinding (SAG) mill and two 7.9 m (26 ft.) ball mills to prepare feed for a conventional flotation circuit to recover the copper, gold and silver contained in the feed ore to the copper concentrate. The copper concentrate is thickened and filtered on site prior to being trucked to the port of Matarani for shipment to smelters. Flotation tailings are thickened and delivered to the existing Tintaya open pit for disposal. The nominal throughput of the concentrator is 70 000 tonnes per day. Figure 3 shows the flow sheet and Figure 4 shows the grinding mills.
OPERATIONS START‐UP Safety The most important issue from operations start‐up was safety. During the first two months since start‐up we had 61 days with 255 072 man‐hours without lost time incidents.
Strategies The development of the start‐up strategies began in November 2011 and involved people, strategic contracts and consulting. The people training strategies included visits, team building and lessons learned from other start‐ups with teams from Las Bambas, Alumbrera and Antamina; trainings for supervisors and operators in Alumbrera and Antamina; technical visits for superintendents and supervisors to Confluencia (Los Bronces), Esperanza, Andacollo, Chuquicamata, Valle Central, Cerro Verde, Toquepala, Cerro Corona and Conchán Refinery; training courses with suppliers (Tecsup, Sandvik, Yura, Cytec, Moly‐Cop, Mepsa, Renasa, Orica and Flomin); and participation in congresses. The objective of the start‐up phases strategy was to produce copper concentrate with the minimal possible equipment; we developed a start‐up plan with three phases taking into account the equipment priority in order to fulfill the objective. Each phase, beside the main equipment considered the secondary equipment as belt conveyors, pump with boxes and cyclone batteries (Rondestvedt, Bustos & Torres, 2011). For the reagents strategy we added two collectors and one frother in order to improve the recoveries.
8
Figure 3: Antapaccay concentrator flow sheet
9
Figure 4: Antapaccay concentrator grinding equipment
Start‐Up Benchmarks
24/10/12 Initial operation of the SAG Mill using standard ore
9
Figure 4: Antapaccay concentrator grinding equipment
Start‐Up Benchmarks
24/10/12 Initial operation of the SAG Mill using standard ore
08/11/12 End of 50% charge test and plant shutdown for readjustment
10/12/12 First tests with pebble crushers
11/12/12 First Copper Concentrate Shipment
11/01/13 Impact meter installed for SAG Mill
Figure 5 shows the grinding throughput evolution.
Figure 5: Grinding throughput evolution
Improvement Opportunities Several potential improvements to the operation are being considered:
Installation of standby secondary equipment in order to improve the availability.
Processing of crushed pebbles in balls mills using a SABC2 circuit to process more tonnage.
Precrushing of SAG mill feed in order to increase the throughput.
10
CONCLUSION The successful start‐up of the Antapaccay Concentrator was due to the implementation of strategies that were developed before the start‐up.
ACKNOWLEDGEMENTS Many people and organizations contributed to the activities described above the authors would like to thank to all of them.
REFERENCES Dechert, C. (2006) ‘Antamina – Design through to operation’, Proceedings of SAG 2006 , Vancouver, Canada, pp. 11‐26. Rondestvedt, C., Bustos, M. & Torres, L. (2011) ‘Carmen de Andacollo – start‐up of Teck’s new copper concentrator’, Proceedings of SAG 2011 , Vancouver, Canada, pp. 1‐14. Vanderbeek, J., Linde, T., Brack, W. & Marsden, J. (2006) ‘HPGR implementation at Cerro Verde’, Proceedings of SAG 2006 , Vancouver, Canada, pp. 45‐61
