Lead and zinc: selective flotation on ores that rarely cooperate

A one-line email and one good-looking rock. It's a common way enquiries reach us, and one reason so many of them stall. Nobody can size a polymetallic circuit from a hand specimen. Here is what actually moves an enquiry to a flowsheet.

The sample, and how it was chosen

Send composites, not your best material. Split them the way your geologist already splits the orebody, by ore type, by depth, by mineralisation style, because a preferential circuit that behaves on one domain can misbehave on the next and a single blended composite hides exactly that. Sample mass follows the test. Bench flotation runs on modest charges; a semi-continuous campaign needs bulk material. Confirm the charge each test needs before the crate leaves site.

The assay suite is wider than Pb and Zn

Pb, Zn, Cu, Fe, S, Ag and Au is where you start, not where you stop. Minor elements decide the economics on both sides of the ledger. The USGS deposit model for Mississippi Valley-type lead-zinc ores records germanium, gallium and cadmium at recoverable economic levels in some deposits. A different group of trace elements matters for the opposite reason: arsenic, antimony, bismuth and mercury are the names that turn up on the penalty schedule of a smelter contract. Get both groups measured before a flowsheet exists. Rejecting a penalty element is a circuit design decision, not a commissioning adjustment.

Mineralogy, not only chemistry

An assay tells you how much silver the ore holds. It doesn't tell you where the silver sits, substituted in galena, in discrete sulphosalts, or locked inside sphalerite, and that answer decides which concentrate ends up carrying the payable metal. Iron in sphalerite works the same way: iron-rich sphalerite resists the usual depressant regime, and you'd rather learn that in a laboratory than in a cleaner bank. Gangue counts as well. USGS lists dolomite and calcite among the dominant minerals of these carbonate-hosted ores and puts their iron sulfide content generally below 5%, which is a very different pH and depression problem from a pyritic massive sulphide feed.

What the site imposes

Water source and water chemistry, grid power or on-site generation, road and port access, altitude, frost depth, seismicity. None of it changes the mineralogy. All of it changes how the plant gets built, and whether a modular plant is the sensible form. Our test work group needs the sample; the design institute needs the site data. Send both at once and you save a round trip.

Two orebodies can both be labelled lead-zinc and still need different plants. Choosing the route from the deposit's name is a recurring judgement error in tender documents, and once a tender has specified preferential flotation the flowsheet is locked before anyone tested whether the ore supports it.

The scale question people skip

USGS puts the median Mississippi Valley-type deposit at 7.0 million tonnes grading roughly 7.9% Pb plus Zn, and notes these deposits usually occur in districts of several to as many as 400 deposits rather than standing alone. Read that as a plant brief and it says something concrete. Many lead-zinc orebodies are individually small, so the capital case often rests on feeding one mill from several of them, or on a plant that can be moved.

Our own lead-zinc references sit at opposite ends of that spread: a 0.99 Mt/a lead and zinc plant in Inner Mongolia and a 500 t/d lead and zinc plant in Heilongjiang, with a 1,100 t/d copper and lead plant in Namibia somewhere between. Same pair of payable metals, plants of a different order. Crushing stages, surge capacity, the number of cleaner stages that can be justified, whether the concentrate leaves by truck or by container, the size of the maintenance crew the site can carry: all of that moves with tonnage, and none of it copies across. A route can be metallurgically sound and still fail on tonnage. That is a mine planning answer rather than a flotation answer, and it belongs in the conversation early.

The smelter is a design input, not an afterthought

Concentrate has to go somewhere, and that somewhere may not be in your country. USGS reports that nearly all United States lead concentrate has been exported since the last domestic primary lead refinery closed in 2013, while zinc ore and concentrate reaching US buyers over 2021–24 came mainly from Peru, Canada and Turkey. Long concentrate journeys are ordinary. They also mean the treatment charge, the payable metal schedule and the penalty list of a smelter you may never visit are inputs to your cleaner circuit.

Companion metals sharpen the point. USGS records domestic lead coming from five Missouri lead mines plus byproduct output at two Alaskan zinc mines and two Idaho silver mines, and counts 4 primary silver mines against 31 base- and precious-metal operations that produce silver as byproduct or coproduct. Most silver revenue rides on somebody else's concentrate. Silver most often follows the lead concentrate, but that is an outcome of where the silver sits in the ore, not a rule, and confirming it on your ore is half a flotation question and half a contract question.

What a lead-zinc handover has to define

The delivery shapes themselves are set out on EPC turnkey delivery and contract mining and operation, and choosing between them before the enquiry goes out changes what gets quoted; what follows is what a two-product polymetallic plant adds on top of whichever shape you pick.

Which ore the performance run is fed. An acceptance test has to name its feed: which domain, which grade range, which mineralogy. Where an oxidised upper zone sits over primary sulphide, a run fed from sulphide stockpiles proves nothing about the first years of the mine schedule, and a run fed from the oxide zone condemns a circuit that was never drawn for it. Tie the acceptance feed to the mining sequence, and write down in advance what happens when the plant meets a domain the test programme never saw.

Two products, two sets of numbers. Lead concentrate and zinc concentrate each carry a grade and a recovery, and the two pull against each other. Depress harder to keep lead out of the zinc product and lead units go to tailings; ease off and the zinc concentrate arrives at the smelter carrying a deduction. A single blended recovery figure hides that trade completely. Each product line needs its own acceptance criteria, and the contract needs a stated order of priority for the case where the ore will not give both at once.

Where the silver credit lands. When the payable silver rides in the lead concentrate, the party running the flotation circuit is steering a revenue line that may sit on someone else's account. Set out whether silver is assayed in both concentrates, which party the credit belongs to, and whether silver sits inside the acceptance test or outside it. A reagent regime tuned only for Pb and Zn grade can quietly move silver into the product that pays less for it.

Who owns the moisture number. A polymetallic plant ships two cargoes, and each is certified on its own. Lead concentrate and zinc concentrate are separate bulk cargo shipping names in the IMSBC Code mineral concentrates schedule, classified Group A for the liquefaction risk, so each carries its own transportable moisture limit determined on its own material. That means two filter duties rather than one, and they rarely land in the same place: a dense galena product and a fine sphalerite cleaner product do not dewater alike. The Code puts the signed TML certificate and the moisture declaration in the shipper's hands, not the plant builder's, so a scope that ends at filter cake leaves the owner certifying a number somebody else designed for. Name in the contract who determines each limit, who samples the cake and how often, and who carries it when a parcel is turned away at the load port.