Magnetic characterization of ferrosilicon dense media recovery

Ferrosilicon remains the benchmark dense medium for gravity separation, particularly in Australian coal preparation plants, where clean coal and reject must be separated by density. In a Hunter Valley wash plant, the dense medium circuit may appear routine, yet its performance rests on fine particles that change between shifts, clean-ups and campaigns.

In dense media separation (DMS), the medium’s value cannot be reduced to nominal iron and silicon content. Particle size, shape, relative density, oxidation, magnetic susceptibility and magnetisation behaviour affect how it flows through a dense medium vessel, reports to the magnetic separation circuit and separates cleanly from coal or mineral surfaces.

For projects in Kalgoorlie-Boulder, the Pilbara and Mount Isa, verified medium specifications help engineers convert laboratory results into plant decisions for coal and selected gold and base-metal applications. That work connects ore testing with mine design, EPC, commissioning and operational support.

Why media properties drive recovery

A ferrosilicon lot can meet its chemical specification and still perform poorly if its particles are too fine, too angular or unevenly distributed. Relative density establishes the achievable cut point, while size and shape influence dispersion, terminal velocity, drainage and entrainment.

Those properties interact with the dense medium vessel, pumps, drainage screens and magnetic recovery system. They also shape maintenance and consumable costs. This makes media characterisation relevant to the broader processing equipment range, rather than a stand-alone laboratory exercise.

Sampling the dense media circuit

Results are only as dependable as the sample. A useful programme combines fresh reagent, medium returned from the plant and material recovered during a clean-up. Sampling by timed grabs can exaggerate one condition; stratified cross-cuts or a properly sized composite give a more defensible picture of the working circuit.

Document moisture, sampling location, date and campaign conditions, then seal and chain each sample. Non-magnetic tools reduce contamination. At remote FIFO sites, dispatch delays and temperature changes make duplicate samples, retained splits and clear custody records especially valuable.

Measuring magnetic and physical characteristics

Magnetic testing should measure volumetric or mass susceptibility and, where the data justify it, a full magnetisation curve with saturation magnetisation, remanence and coercivity. A susceptibility meter provides a rapid comparative screen, while specialist magnetic measurement instruments support calibration and deeper investigation of a medium lot.

Physical testing should pair pycnometer or gas-pycnometer density results with dry and wet particle size distributions, microscopy, shape observations and chemical assay for iron, silicon and loss on ignition. These measurements distinguish dilution, oxidation, fractured particles and contamination that a single Fe/Si assay can miss.

Quality systems need calibrated instruments, reference materials, replicate readings and stated uncertainty. Record sample temperature and any washing or drying performed before magnetic testing, because preparation can alter the very condition an operating plant experiences.

Linking characterization to separation performance

Laboratory data become useful when paired with washability tests and a density partition curve. The test should establish cut-point density, misplaced material, medium-to-feed ratio and product response under controlled conditions. Coal operators can relate this to ash and yield; mineral projects can track valuable metal in reject and gangue in product.

For a two-stage circuit, engineers should validate each density cut independently rather than infer one result from the other. A short double-barrel explanation can help non-specialist stakeholders understand staged choices, but it is only a communication analogy and must not be treated as process data.

Ferrosilicon carried into flotation can stabilise froth and contaminate concentrate, while magnetite and FeSi particles report to medium recovery. This is a familiar issue in coal washing. Operators should therefore track iron in product, medium losses, froth condition and wash-plant ash together rather than diagnosing a separation problem from ash alone.

Detecting degradation and handling damage

In-service ferrosilicon changes through oxidation, surface coating, attritional fracture and the production of fines. Those changes can alter density and magnetic response while increasing losses through drainage screens. Comparing fresh and aged media from several operating periods makes gradual degradation visible before recovery falls away.

Scratches on screens, mixers and handling equipment can also indicate poor medium circulation or damaged liners. Automated scratch inspection can complement routine visual checks, provided it is calibrated for wet, dusty conditions and used to show trends in surface damage rather than to replace thickness and hardness measurements.

Translating results into plant design

Characterisation should feed directly into equipment selection and process design. The dense medium plant needs suitable mixing pumps, vessels, drainage and washing screens, magnetic recovery, sampling points and access for clean-up. Lozova can align media data with these requirements through an integrated plant configuration.

Control recipes should define medium density, medium-to-feed ratio, cut point, operating temperature and sampling frequency. Engineers should set warning limits from baseline and aged-medium data, not from a generic supplier certificate. Dense media performance then becomes a managed variable across crushing, grinding and downstream separation rather than an unexplained swing in final product grade.

Perth-based project teams must also plan for imported equipment, long-lead spares and exposure between AUD and USD costs. For Pilbara and Mount Isa operations, commissioning with operating crews is valuable because site conditions and local terminology often expose assumptions that a desk study misses. Australian project reviewers often use plain language here: a result is only “good on paper” until it survives a plant trial.

Recommendations for a defensible test program

A compact, traceable test matrix gives the project a stronger basis for acceptance and operating decisions.

• Establish a lot-specific baseline for density, magnetic response, particle size and chemical composition before the medium enters production.
• Sample fresh, returned and recovered media across locations and campaigns so normal variation and ageing are both represented.
• Use duplicate and retained splits with calibrated equipment to separate genuine lot differences from sampling or laboratory error.
• Pair every characterisation result with washability, partition-curve and medium-recovery observations from the relevant circuit.
• Track medium losses, froth effects and handling-surface wear over time using consistent operating and maintenance records.
• State measurement uncertainty, preparation methods and acceptance criteria so later procurement or replacement decisions remain comparable.

Document exceptions, protect the chain of custody and review the baseline whenever suppliers, operating conditions or plant configuration change. The result should be a usable operating envelope, not a folder of disconnected laboratory sheets.

Reliable dense media recovery begins with treating ferrosilicon as a living process material whose condition must be known and managed. Characterisation gains value when engineers connect magnetic and physical properties to separation performance, equipment design and operator controls.

Bring media samples, ore composites and operating data to the engineering project team to define a practical test programme, equipment scope and commissioning plan for an Australian project.