Selection of optimum ferronickel grade for smelting nickel laterites

CIM Bulletin, Vol. 3, No. 2, 2008

M.Y. Solar, I. Candy, and B. Wasmund

The nickel laterite industry is divided into two camps when it comes to the grade of the ferronickels produced.  One group, led by Falcondo and Cerro Matoso, markets high-grade products in the range of 35 to 40% Ni, while another, led by SLN and the Japanese smelters, favours lower grades in the range of 20 to 25%. Because the ores processed vary widely in terms of iron to nickel ratios, it is probably more accurate to classify these plants according to their iron recovery: Falcondo and Cerro Matoso recover only 15 to 30% of the iron in their ores, while SLN and the Japanese recover 45 to 65%. Lower reductions generally imply higher Ni slag losses and lower Ni and Fe recoveries.  But the advantages are lower power and reductant requirements (and thus higher throughputs in equally sized furnaces), smaller furnace off-gas volumes, and lower refining and product transportation costs.  It also eliminates the high metal superheats and the carbon boil and silicon reversion problems experienced at high reductions. Can both camps be right or is there an economically optimum ferronickel grade for a given ore? This paper proposes a methodology for determining this optimum. Complete smelter balances are developed for a range of ferronickel grades, using appropriate correlations for the variations in slag losses, ferronickel impurity levels and reductant usage.  Differential capital and operating costs are calculated for each case.  Finally, differential net present values (DNPV) are calculated, taking into account the expected ramp-up schedule and major shut-downs during the first 20 years of operation. The conclusion is that iron credits are the major factor directing the selection of the ferronickel grade.  For a medium-size plant smelting a conventional saprolite at 2.4% Ni and a Fe/Ni ratio of ~5 in a 45 MW furnace, the economically optimum ferronickel grade is around 35% for iron credits of <$300 per tonne of iron.  It is only for credits of >$400 that the economic optimum shifts to lower grades below 25% (see figure).  Because iron credits (if any) are typically significantly less than $200 per tonne, it is clear that the higher ferronickel grades should be favoured.  None of the other factors investigated (LME price, ore costs, transportation costs, reductant prices, discount rate) change this conclusion.  Only if the smelter is limited in the amount of calcine it can supply to its furnace is the inflection point moved to lower iron credits of $100 per tonne. Neither do metallurgical considerations override the economic conclusions.  For example, for the ore quoted above, production of a 35% ferronickel requires reduction of 35% of the iron.  The C, Cr and Si concentrations estimated at this reduction potential are about 0.2% each.  At these low levels, carbon boils and silicon reversions cannot occur, thus avoiding a problem that often plagues smelters that target higher reductions. Of course, every ore and every location is different.  But the authors have yet to see a project for which the conclusions are substantially different. Producers and customers should thus recognize the extra cost of producing lower ferronickel grades.