24-MMP-A3 Mineral Processing · May 2013
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
EGBC National Exam — Mining and Mineral Processing Engineering, 09-MMP-A3 Mineral Processing, 2013-May. 3 hours duration, closed book; only an approved Casio or Sharp calculator permitted. Question 3 allows a choice of any six (6) of nine sub-terms. Question 5 and its bonus (page 5) were to be handed in with the exam booklet; Question 4's log-log plot (page 6) is reproduced here as a computed inline figure.
Reference texts: Wills & Finch, Wills' Mineral Processing Technology, 8th ed. (flotation circuit design and metallurgical balances – Ch. 12; comminution, Bond's law and circulating load – Ch. 3 & 6; particle size analysis – Ch. 4; sampling theory, Gy's equation – Ch. 3; gravity concentration, dense medium separation, magnetic/electrostatic separation – Ch. 10, 11 & 13); Taggart, Handbook of Mineral Dressing (classical economic-recovery/economic-efficiency formula used in Question 1(vi)).
Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.
All nine pairs are answered below (the exam permits choosing any six of nine; every pair is worked so this set doubles as a complete study reference).
Similarity. Both are tailings/process-water clarification techniques that aggregate fine, slow-settling suspended particles into larger units so they settle faster in a thickener or tailings pond, and both work by overcoming the natural electrostatic (zeta-potential) repulsion between fine particles.
Difference. Coagulation uses an inorganic multivalent-ion reagent (e.g. lime, alum, ferric salts) that compresses the particles' electrical double layer, letting van der Waals attraction pull them into small, dense, tightly-bound coagula. Flocculation uses a high-molecular-weight organic polymer (anionic, cationic or non-ionic polyacrylamide) that physically bridges across many particles at once, producing much larger, looser, faster-settling flocs. In practice coagulation is often the first step (destabilising the fines) and flocculation the second (bridging the destabilised fines into a strong floc) in the same thickener feed-well.
Similarity. Both are gravity concentration devices that separate particles by specific gravity difference rather than by size or surface chemistry, and both are effective on relatively coarse (roughly 0.1–10 mm) liberated particles.
Difference. A jig separates by vertical hindered-settling/stratification: a pulsating water column (via a diaphragm or air pulsator) repeatedly dilates a static bed of particles on a screen, and denser grains work down through the bed (or through the screen into the hutch) while lighter gangue stays on top and reports over the weir. A shaking table separates by a combination of horizontal transport (an asymmetric reciprocating stroke moves the whole bed slowly along the deck) and cross-flowing wash water (which preferentially strips lighter, less-riffle-trapped grains down-slope), producing a continuous fan of products across the deck rather than a single stratified bed.
Similarity. Both attach hydrophobic (collector-conditioned) particles to air bubbles that rise to form a mineral-laden froth removed as concentrate, while hydrophilic gangue remains in the pulp and is discharged as tailings.
Difference. A mechanical cell uses a motor-driven impeller/stator to both generate bubbles and keep coarse solids in suspension in a relatively short, turbulent, well-mixed vessel – good throughput but with some entrainment of fine gangue into the froth. A column cell has no moving parts: it is a tall, quiescent vessel with bubbles generated by a static sparger at the base rising counter-current to the descending pulp, and (uniquely) wash water applied at the top of the froth washes entrained gangue back down, giving columns a markedly higher selectivity/lower entrainment – the reason Gibraltar's regrind circuit uses columns as the primary cleaning stage ahead of a Denver mechanical-cell scavenger.
Similarity. Both are flotation reagents essential to forming a selective, stable mineral-laden froth, and both act at an interface (frother at the air–water interface, collector at the mineral–water interface).
Difference. A collector (e.g. sodium isopropyl xanthate, at Gibraltar) selectively adsorbs on the target mineral's surface, rendering it hydrophobic so it can attach to an air bubble; it acts on the solid. A frother (e.g. Orform F-2, MIBC) is a surface-active organic compound that lowers water's surface tension, produces small, stable bubbles, and builds a froth layer strong enough to carry mineralised bubbles to the lip without bursting; it acts on the air–water interface and does not render minerals hydrophobic itself.
Similarity. Both crush by the same eccentric-gyration mechanism – an inner mantle gyrates inside an outer concave/bowl, alternately closing and opening the crushing gap around the circumference – and both give a continuous (not intermittent, unlike a jaw crusher) discharge.
Difference. A gyratory crusher has a steep, near-vertical mantle and a large feed opening, sized for continuous, high-tonnage primary crushing of run-of-mine ore straight from the pit. A cone crusher has a much shallower mantle/bowl angle and gyrates faster with a smaller throw, giving a finer, more uniform product with a lower reduction ratio – the standard choice for secondary/tertiary reduction ahead of grinding, typically fed material already reduced by a primary gyratory or jaw crusher.
Similarity. Both are hydrocyclones of essentially the same mechanical design (tangential feed inlet, vortex finder, apex/spigot) that use centrifugal force in a swirling vortex to split a feed stream into two products.
Difference. A classification cyclone splits a water-only slurry purely by particle size/settling rate: fines report to the overflow (via the vortex finder), coarse to the underflow (apex) – the separation is size-based only, with no true density selectivity beyond its influence on settling rate. A dense (heavy) medium cyclone is fed a slurry of ore suspended in a dense medium (typically fine magnetite/ferrosilicon in water, adjusted to a target SG between the two minerals to be split) and separates almost purely by particle density, near-independent of size over a wide range – sink (denser) reports to the underflow, float (lighter) to the overflow, making it a true gravity/density preconcentration device rather than a classifier.
Similarity. Both are staged embankment-raise methods that build a tailings dam progressively higher over the life of a mine using cycloned tailings sand as (part of) the embankment fill, rather than constructing the full final height at once.
Difference. An upstream raise places each new dyke crest behind (upstream of, over the tailings beach of) the previous one, so the embankment's centreline migrates toward the pond – cheap and fast, but each raise is founded partly on previously deposited (potentially loose, saturated) tailings, making upstream dams the most susceptible of the three classical raise methods to static/seismic liquefaction failure. A downstream raise places each new dyke in front of (downstream of) the previous crest, so the centreline migrates away from the pond and each raise is founded on the original, engineered starter embankment/foundation – structurally the most stable method, but requiring substantially more fill volume and footprint for the same final height.
Similarity. Both are dry, physical (non-chemical) separators that exploit an intrinsic mineralogical property difference to split a dry, sized feed on a moving belt or rotating drum, and both are commonly used together in mineral-sand (e.g. ilmenite/rutile/zircon) separation trains.
Difference. A magnetic separator exploits differences in magnetic susceptibility: ferro-/paramagnetic grains (e.g. magnetite, ilmenite) are deflected/held by a magnetic field (permanent or electro-magnet) while diamagnetic gangue is not. A high-tension (electrostatic) separator exploits differences in electrical conductivity: as grains pass through a high-voltage corona field on a rotating drum, conductive grains (e.g. rutile, ilmenite) lose their induced charge quickly and fly off the drum on a ballistic path, while non-conductive grains (e.g. zircon, silica) retain their charge, cling electrostatically to the grounded drum, and are removed further round by a brush.
Similarity. Both describe the particle size that has an equal (50%) chance of reporting to either product of a real classifier (e.g. the hydrocyclone of Question 2), read from the same experimental partition (Tromp) curve.
Difference. $d_{50}$ (the "actual" or "analytical" cut size) is read directly off the raw partition curve at 50% recovery-to-underflow, without correction. $d_{50c}$ (the "corrected" cut size) first removes the effect of water/fines by-pass to the underflow – the fraction of very fine (near-zero size) material that reports to underflow purely by entrainment in the underflow water split, not by true classification – and reads the 50% point off that corrected curve; it is the better estimate of the size at which the cyclone's true classifying action is centred. The vertical spread of the corrected curve around $d_{50c}$ is summarised by the sharpness index, the Imperfection $I=(d_{75}-d_{25})/(2d_{50c})$ – a small $I$ (steep curve) means sharp, efficient classification; a large $I$ (shallow curve) means a poorly selective split.
