NI 43-101 Preliminary Economic Assessment: BK11 Mine

This report presents the proposed design, historical operating data, and testwork for the BK11 diamond mine processing facilities in Botswana.

Report context

This NI 43-101 Preliminary Economic Assessment for the BK11 Mine reports on a project that involved open pit mining of kimberlite material with planned processing plant upgrades in three phases. The report documents both the design basis for the processing facilities and actual operational data from July 2010 to February 2012, when mining operations ceased at the end of January 2012. Firestone’s planning assumed the majority of the first 120 vertical metres of kimberlite material would be variably weathered and diluted between 10% and 40% with clasts, cobbles and boulders of country rock, mainly basalt.

Processing route

Design basis and phased upgrades

The process plant was designed to be upgraded in units of 100 t/h across three phases: 100 t/h, 200 t/h and 300 t/h. Equipment from another mine was to be transported and built progressively in streams. An old alluvial diamond processing plant was dismantled and the equipment was re-conditioned and built at BK11. Most of the design and infrastructure was based upon the plant achieving 200 t/h or a yearly target of 1.5 Mt for a period of 10 years.

Phase 1 design and operation (July 2010 to July 2011)

Phase 1 was designed as a simple flowsheet with primary crushing, scrubbing, screening, dense media separation (DMS) and final recovery plant. All oversize material (+25 mm) was stockpiled for re-treatment at a later stage. This phase allowed for early project implementation and production at up to 100 t/h. Phase 1 ran from July 2010 to July 2011.

Phase 2 design and operation

The Phase 2 plant design doubled the capacity to process at 200 t/h head feed with 125 t/h DMS capacity, comprising a 75 t/h DMS and a 50 t/h DMS. However, the process did not incorporate secondary crushing and +25 mm oversize material was discarded until December 2011. The secondary section was not fully commissioned and therefore did not operate up to when the mine was stopped at the end of January 2012.

Actual throughput performance

The process plant started in July 2010 but was run at relatively low throughput of approximately 38 ktpm and intermittently until December 2010. When the plant was in stable operation from about April 2011 to when it stopped at the end of January 2012, it averaged approximately 70 ktpm or approximately 120 t/h. This was lower than the design of 200 t/h. Daily data showed there were times when the plant operated at 150 t/h but this was less than 10% of the time. The reduced throughput versus plan was due to a number of factors but mainly due to the stickiness of the ore, which the scrubbers were not able to cope with. Sticky ore also caused downstream problems in the DMS as clay fines affected the Ferrous Silicon (FeSi) viscosity.

Process flow description

The flowsheet included two streams consisting of two primary crushers and two scrubbers. ROM material was loaded by truck into two ROM receiving bins, each fitted with a 400 mm tipping grizzly. A rock breaker was provided to reduce any oversize in the ROM feed. Material was withdrawn at a controlled rate from the receiving bin by a belt feeder on one stream and an apron feeder on the other, then screened on a vibrating grizzly. The oversize was fed to the primary jaw crusher with the crusher product combined with the grizzly undersize. A self-cleaning magnet was provided over the main feed conveyor to remove any tramp material. The jaw crushers were operated at a closed side setting of approximately 70 mm to ensure a feed size of minus 110 mm to the scrubbers.

Primary crusher products were conveyed to the scrubber circuit where they combined with pulping water from the process water dam and fed into the scrubber. Scrubber products were then screened and combined into three fractions: +25-110 mm oversize, -25+8 mm DMS feed and -1.6 mm slimes. Above the DMS bins a single deck screen separated material into -25+8 mm and -8+1.6 mm material. The coarse -25+8 mm material was fed to the 50 t/h DMS and the 8+1.6 mm material was fed to the 75 t/h DMS. The undersize (-1.6 mm) from the screen was pumped to the slimes dam.

In the last few months of operation tests were performed to assess the crushing of +25 mm oversize material to -15 mm. The crusher product was recycled back to one of the two scrubbers, which reduced the amount of material that could be fed through the primary crusher to the scrubber.

Dense media separation circuit

Material was withdrawn at a controlled rate from the DMS feed bin and fed to the DMS mixing box where the gravel was mixed with FeSi medium at the correct density. The cyclone feed pump transported the gravel and medium mixture from the mixing box up to the dense medium cyclones, where separation was effected. Higher density particles moved down the wall of the cyclone and discharged through the spigot. Less dense particles moved towards the central axis of the cyclone, where they were caught in the vortex and discharged into the overflow chamber.

The tailings (floats) were discharged from the DMS cyclones to the static drainage panel where most of the FeSi medium was removed. The gravel and remaining medium was carried onto the drainage section of the floats screen for draining and washing. The drained medium was returned to the circulating medium tank and the washings to the dilute medium tank. The concentrate was discharged onto the sink screen where the medium was drained and the concentrate was washed to remove any adhering medium. The concentrate was then conveyed to the recovery feed surge bin.

Final recovery section

Material was drawn from the DMS concentration holding bins via a jet pump and coarse and fines were pumped separately approximately 120 m to the recovery plant and then onto a sizing screen. The sizing screen separated the material into two size fractions: -16+8 mm (coarse) and -8+1.6 mm (fines). Each size fraction went into a separate surge bin, with the coarse processed through a single stage Flow Sort machine and the fines through a double stage Flow Sort machine. The tailings from the X-ray machines were supposed to be conveyed to a grease scavenger section, however this section was never built. The concentrate from the X-ray machines was dried using an infrared drier and then transferred to a holding bin above the X-ray concentrates glove box for hand sorting.

Water circuits

All -1.6 mm material was pumped to the slimes dams with penstock overflows returning slimes water to the thickener. Flocculent was added to assist with settling. The thickener overflow was collected in the process water reservoir. Make-up water from the raw water reservoir was added to the process water reservoir to maintain a constant level. Process water consumption volumes were actually 50 Ml per month when 70 ktpm was processed, which is about 0.7 m³ per tonne ROM feed. This was higher than the design mainly because only one thickener was commissioned out of two planned. Water was pumped from six pit dewatering boreholes. The raw water dam was sized at 10,000 m³ to hold approximately 2 days worth of process water. The dam was HDPE lined with a 1.5 mm membrane.

Proposed plant upgrade – autogenous milling

The report describes proposed autogenous milling as an appropriate technology for the BK11 Project, noting that autogenous milling was not used in Botswana prior to 2012. The proposed upgrade would involve a 5.5 m diameter by 3.1 m length autogenous mill with a 750 kW motor, which would be retrofitted into the current circuit. A comprehensive lump sum turnkey quote was received from a Johannesburg based engineering company.

The report states that autogenous milling would be the preferred comminution and diamond liberation unit process at BK11 for several reasons including: undertaking the same diamond liberation role as attrition scrubber, secondary crusher, tertiary crusher and high pressure grinding rolls (HPGR) unit processes; treating heavily weathered kimberlite in sticky areas of the open pit thereby resolving materials handling problems relating to clay; reducing required DMS capacities; removing the need for internal water recovery circuits; reducing capital expenditure and operating costs; reducing operating manpower requirements; and allowing for a flexible flowsheet.

Rock hardness and mill sizing

Samples taken at BK11 in August 2015 by Gemet (Pty) Ltd were processed by internationally accepted SMC tests. Results were interpreted and reported by the Julius Kruttschnitt Mineral Research Centre (JKMRC) in Brisbane, Australia. BK11 ores were considered to be relatively soft such that the autogenous mill would operate with higher throughput than originally estimated.

Key reported parameters

Technical qualifications

This report presents historical operating data from July 2010 to February 2012 for a plant that is no longer operating, with mining operations having stopped at the end of January 2012. The proposed autogenous milling upgrade is based on testwork conducted in August 2015 using SMC tests interpreted by JKMRC, but no operational results or economic performance data for the proposed upgrade are provided. The report notes that the secondary crushing section was not fully commissioned and did not operate. The grease scavenger section intended for the final recovery plant was never built. The thickener circuit had only one thickener commissioned out of two planned. The process water consumption was higher than design due to this limitation. The report does not present any current operational status, ownership information, economic analysis results, or links to project websites.

—

Source: NI 43-101 Preliminary Economic Assessment: BK11 Mine, relevant sections including Recovery Methods, Process Description and Equipment Selection, and proposed autogenous milling upgrade.