Acid Consumption Calculation: Leach & POX Acid Balance

Last updated: 2026-10-24

A disciplined acid consumption calculation turns scattered assay data into one number that decides reagent economics: on the illustrative pyrite concentrate below, 98% sulphide oxidation generates about 960 kg H2SO4 per tonne (dry basis) while gangue and precipitate sinks absorb roughly 190 kg, leaving a 770 kg/t surplus the plant must neutralize rather than purchase. This is the general methodology for a sulphuric acid balance across leaching and pressure oxidation circuits. For the oxide-specific case, see our companion guide to acid consumption in copper oxide leaching at [/resources/acid-consumption-copper-oxide/]; this article covers the transferable procedure for sulphide and POX systems.

Key Takeaways

• A structured acid consumption calculation prevents the 10–20% reagent budgeting errors common on new projects.
• The balance is generation (sulphide oxidation) minus consumption (gangue, dissolution, losses), per dry tonne.
• Oxidizing sulphide sulphur to sulphate releases 3.06 kg H2SO4 per kg of sulphide S.
• Basic iron sulphate and jarosite lock sulphate, cutting net POX acid release by up to 1.75 kg per kg of iron.
• Gangue acid consumption dominates oxide-ore demand and scales one-for-one with tonnes treated.

What Is an Acid Balance and Why Does It Control Process Economics?

An acid balance is a quantitative ledger that records every reaction generating or consuming H2SO4 in a circuit and closes on a single net figure, expressed in kg of H2SO4 per tonne of dry feed. It can be built per tonne for design and testwork, or per day for operating reconciliation; the per-tonne form is the reference. Gangue acid consumption (GAC) is a measure of the acid consumed by carbonate and reactive silicate minerals that would otherwise remain available for metal dissolution. Net acid requirement is a reconciliation figure that states the acid a circuit must import — or the neutralizing reagent it must buy downstream — once internal generation and every sink are counted. Free acidity is the operating control variable the balance predicts: residual H2SO4 in solution, in g/L.

The sign of the net figure drives real money. Sulphide concentrates treated by pressure oxidation are usually net acid generators; the capital question is not how much acid to buy but how large the limestone neutralization plant must be and how much gypsum the effluent system will produce — gypsum, in turn, sets thickener and tailings-pond loadings. Oxide copper ores are net acid consumers; there, an underestimated GAC means falling free acidity, poor extraction, and emergency acid freight. Acid is routinely the largest or second-largest reagent cost, and a 10% error in the acid consumption calculation propagates directly into reagent budgets, tankage sizing, and effluent compliance. Emergency spot purchases carry freight premiums that dwarf the nominal reagent price, which is why the ledger belongs in the feasibility study, not the post-commissioning review.

The balance also disciplines testwork interpretation. A bottle roll reporting “45 kg/t acid consumption” without stating whether neutralization reagent or washed-solids losses are included cannot be scaled, and two laboratories using different conventions will report different numbers on the same ore. Acid balance hydrometallurgy is therefore a consistent accounting convention — one basis, one set of reaction factors, a closing check against solution chemistry — applied identically from the first bottle roll to the annual reconciliation between purchased and consumed acid.

Reaction Stoichiometry Reference for Leaching and POX

Every term in an acid balance reduces to a balanced reaction with a fixed mass factor. Convention: sulphide-to-sulphate oxidation is the core generation term; iron-precipitation steps are adjustments that leave sulphate free or lock it in a solid — avoiding the classic ferric-hydrolysis double-count.

Sulphide oxidation (the generation term). Elemental sulphur to sulphate:

S + 1.5 O2 + H2O → H2SO4

i.e., 3.06 kg H2SO4 per kg of sulphide S oxidized. Written for pyrite with iron to hematite:

2 FeS2 + 7.5 O2 + 4 H2O → Fe2O3 + 4 H2SO4

and with iron to ferric hydroxide:

FeS2 + 3.75 O2 + 3.5 H2O → Fe(OH)3 + 2 H2SO4

Both routes release 3.06 kg per kg S — ferric hydrolysis (Fe2(SO4)3 + 3 H2O → Fe2O3 + 3 H2SO4) also releases acid, already embedded in this figure.

Iron precipitation as basic salts (the locking adjustments). When iron reports to FeOHSO4 or jarosite, sulphate leaves solution locked in a solid:

Fe2(SO4)3 + 2 H2O → 2 FeOHSO4 + H2SO4
3 Fe2(SO4)3 + K2SO4 + 12 H2O → 2 KFe3(SO4)2(OH)6 + 6 H2SO4

Per kg Fe to FeOHSO4, one mole of sulphate is locked: 1.75 kg H2SO4-equivalent per kg Fe; jarosite locks 2/3 mol per mol Fe, 1.17 kg. This is the largest lever on POX acid stoichiometry: with ~0.875 kg Fe per kg S, iron mineralogy swings the balance by hundreds of kg/t. Hot, acidic autoclaves favor hematite; cooler operation favors basic salts. Biooxidation precipitates iron mostly as hydroxides and jarosites — see our POX vs biooxidation comparison.

Gangue neutralization (the consumption terms). Carbonates react stoichiometrically:

CaCO3 + H2SO4 → CaSO4 + CO2 + H2O
CaMg(CO3)2 + 2 H2SO4 → CaSO4 + MgSO4 + 2 CO2 + 2 H2O
MgCO3 + H2SO4 → MgSO4 + CO2 + H2O

Reactive silicates consume acid materially over hours:

Al2Si2O5(OH)4 + 3 H2SO4 → Al2(SO4)3 + 2 SiO2 + 5 H2O (kaolinite)

Copper-bearing species. Acid-soluble oxides and silicates take one mole of acid per mole of copper:

CuO·SiO2·2 H2O + H2SO4 → CuSO4 + SiO2 + 3 H2O (chrysocolla, simplified)
Cu2S + 2.5 O2 + H2SO4 → 2 CuSO4 + H2O (chalcocite — acid converts the second copper)

Sign convention: positive factors add free acid; negative factors remove it.

How to Run a Step-By-Step Acid Consumption Calculation

The procedure applies to any leach or POX circuit; pressure oxidation autoclave hardware is covered separately.

1. Define the basis. Fix 1 tonne of dry feed as the reference; every mass refers to this basis, and moisture is a separate solution term.
2. Assay the acid-forming and acid-consuming species. You need sulphide S (by difference from total S), Fe, carbonate CO3 with a calcite/dolomite split where possible, and a read on reactive clays.
3. Fix the extents of reaction from testwork. Sulphide oxidation percentage, the iron speciation split, and metal extractions come from batch or pilot data, never from assumption.
4. Total the acid-generating terms. Multiply sulphide S oxidized by 3.06, then subtract sulphate locked in basic salts at −1.75 (BIS) or −1.17 (jarosite) per kg Fe.
5. Total the acid-consuming terms. Apply factors to GAC minerals, to each kg of metal dissolved (1.54 chrysocolla Cu, 0.77 chalcocite Cu), to neutralization reagents, and to acid in washed-solids moisture.
6. Close the balance against solution chemistry. Net generation minus consumption, divided by solution volume per tonne, predicts free acidity in g/L; if it misses the target, the balance — not the plant — is usually wrong.
7. Convert to plant rates and add margin. Multiply kg/t by dry tonnes per day, correct for commercial acid strength (98% w/w typical), and carry a 10–20% contingency.

Worked Example 1 — POX pyrite concentrate (illustrative, dry basis, per tonne). A concentrate assaying 32% sulphide S, all as pyrite, is treated in a POX autoclave; pyrite iron follows stoichiometry: 32 × (56/64) = 28.0% Fe.

Arithmetic:

• S oxidized = 320.0 × 0.98 = 313.6 kg
• Acid potential = 313.6 × 3.06 = 959.6 kg H2SO4
• Fe oxidized = 280.0 × 0.98 = 274.4 kg; Fe to BIS = 274.4 × 0.25 = 68.6 kg
• Sulphate locked in BIS = 68.6 × 1.75 = 120.1 kg H2SO4-equivalent
• Net autoclave generation = 959.6 − 120.1 = 839.5 kg
• GAC = (25.0 × 0.98) + (40.0 × 1.14) = 24.5 + 45.6 = 70.1 kg
• Net surplus = 839.5 − 70.1 = 769.4 ≈ 770 kg H2SO4/t

The acid consumption calculation concludes that this concentrate needs zero purchased acid — it is a net generator of roughly 770 kg H2SO4 per tonne, so the real acid problem is downstream neutralization. Discharge solution at 30 wt% solids carries 2.33 m³ of water per tonne; at 15 g/L that is 35.0 kg H2SO4, needing 35.7 kg CaCO3, i.e., 37.6 kg of 95% limestone — about 45 kg/t with a 20% margin for dissolved Fe, Mg, and Al.

Worked Example 2 and Sensitivity: Which Variables Move Acid Consumption Most?

An oxide ore with secondary sulphides is agitated-leached (equipment basis in our agitated leaching tank guide); recovery estimation follows the companion leaching recovery calculation.

Worked Example 2 — mixed oxide/secondary-sulphide copper ore (illustrative, dry basis, per tonne of ore).

Arithmetic:

• Chrysocolla Cu dissolved = 8.0 × 0.60 = 4.8 kg; acid = 4.8 × 1.54 = 7.4 kg
• Chalcocite Cu dissolved = 8.0 × 0.25 × 0.70 = 1.4 kg; acid = 1.4 × 0.77 = 1.1 kg
• GAC = 45.0 kg
• Solution per tonne = (1 − 0.40)/0.40 = 1.50 m³; free acid held at target = 1.50 × 8 = 12.0 kg
• Acid in washed-solids moisture = 0.20 m³ × 5 g/L = 1.0 kg
• Total = 7.4 + 1.1 + 45.0 + 12.0 + 1.0 = 66.5 kg H2SO4/t

On a once-through batch basis, consumption is 66.5 kg/t. In a continuous circuit with CCD wash liquor recycled to the leach, the 12.0 kg of free acid is working inventory rather than consumption, and steady-state make-up falls to roughly 54.5 kg/t — so the basis must always be stated.

Sensitivity analysis (Example 2 basis):

GAC moves the balance one-for-one and dominates everything else. Free-acid target times solution volume per tonne is the second lever — dense pulps are cheaper in acid as well as size. Copper grade and mineralogy barely matter at these factors (≤1.5 kg H2SO4 per kg Cu). For POX feeds the ranking inverts: 98% → 95% oxidation cuts generation by 9.6 × 3.06 = 29.4 kg/t; shifting the iron split from 25% to 50% BIS locks an additional 68.6 × 1.75 = 120.1 kg/t. Sulphide assay and iron mineralogy matter most.

Which Errors Distort an Acid Consumption Calculation Most?

The same faults recur across projects; most surface in draft spreadsheets or design reviews, and we audit for all of them in due-diligence work. Each item follows the pattern: symptom → cause → fix.

1. Wet vs dry basis confusion. Consumption looks 5–15% better than plant purchases → acid factors were applied to wet tonnes while plant throughput is reported dry, or vice versa → declare “dry basis” on every published number and convert with the actual moisture assay of each stream, not a nameplate value; on concentrates, sub-1% moisture errors are 10 kg/t errors, so reconcile moisture assays monthly.
2. Ignoring acid in washed-solids moisture. Balance predicts 55 kg/t, plant consumes 58–60 → the acid dissolved in the interstitial water leaving with tailings was never counted → add moisture fraction × residual acidity (g/L) as an explicit sink line, as done in Example 2; the loss scales with wash performance, wash-liquor strength, and ore.
3. Forgetting the neutralization reagent in the balance. Limestone plant undersized at commissioning → effluent neutralization was treated as “water treatment” rather than as a term in the sulphuric acid balance leaching ledger → close the balance over the whole flowsheet, gypsum loadings included; on a net-generating concentrate this term can exceed every other reagent.
4. Double-counting ferric hydrolysis. Net acid requirement comes out absurdly high for a pyrite feed → the hematite or ferric-hydroxide hydrolysis acid release was added on top of the full sulphide-to-sulphate term → use one convention — the sulphide-oxidation core term, or the ferric-hydrolysis reactions alone — never both; applied consistently, both give the same answer, so audit any spreadsheet for both terms.
5. Using total S instead of sulphide S. Generation overstated on concentrates carrying residual sulphate → total-S assays were fed into the 3.06 factor; float reagent sulphur and sulphate from prior processing both inflate total S → request sulphide S by sulphate-difference analysis or mineralogical S speciation on every composite.
6. Assuming complete oxidation or complete locking. Balance shows a comfortable surplus, autoclave runs acid-starved → oxidation percentage and the iron speciation split were design assumptions, not testwork results → anchor both to residue assays, and run the acid consumption calculation at 95%, 98%, and 100% oxidation and at both iron-split extremes before fixing a design number; feed-forward from daily S assays then prevents drift.
7. Ignoring acid strength and recirculating impurities. Purchased tonnage “disappears” against the balance → acid was booked at 100% instead of 98% w/w, and accumulating Mg and Al sulphates in recirculating solution consume acid invisibly — magnesium is the usual culprit on bore-water circuits → correct for assay strength and track impurity build-up monthly.

From Testwork to Plant

Laboratory numbers and plant purchases never match exactly, and the gap is diagnosable rather than mysterious. Bottle-roll and stirred tests are closed systems: fixed liquor volume, complete final wash, no recycle streams. Plants run open circuits with solution recycle, drifting ore blends, commercial acid below nominal strength, and seasonal water balances; stirred-tank tests also wash more completely than plant CCD trains, so laboratory predictions embed quiet optimism. Batch POX tests capture oxidation extent and iron speciation well but scale imperfectly into continuous discharge chemistry, where residence-time distribution changes precipitate speciation; translation issues are treated in our autoclave scaling guide. The reconciliation habit is simple: state every number on a 100% H2SO4, dry-tonne basis, book deliveries by assay rather than invoice volume, then attribute the gap term by term — strength corrections reduce apparent consumption while losses increase it, so sign errors here hide real problems. Silicate GAC also grows with residence time, so plant GAC can exceed short-test values.

An illustrative reconciliation shows the method. A bottle-roll campaign predicts 62 kg H2SO4 per tonne (100% basis) for an oxide ore. Twelve months of plant data show purchases equivalent to 71 kg/t of commercial acid. Correcting for strength: 71 × 0.98 = 69.6 kg/t on a 100% basis — a 7.6 kg/t gap. Assigning causes: washed-solids acid losses were 3.0 kg higher than assumed because the CCD wash ratio ran below design; GAC drifted up by 4.5 kg/t as the pit entered a more dolomitic ore zone; metering and assay rounding account for the remaining 0.1 kg/t. The reconciled ledger closes at 69.5 versus 69.6 kg/t; sampling frequency matters, and daily composites beat monthly averages for attribution. Every acid consumption calculation should be rebuilt this way quarterly: start from testwork, add named losses with magnitudes, and force the ledger to close against purchase records. Done consistently, this catches ore-zone drift months before it reaches the production forecast and supports defensible reagent contract negotiations; where GAC dominates, blend management is acid management. For new flowsheets, an integrated pilot campaign closes most of the uncertainty before the reagent contract is signed.

Frequently Asked Questions

What is the difference between acid consumption and acid generation?
Acid generation is acid created in-circuit, chiefly sulphide-to-sulphate oxidation at 3.06 kg H2SO4 per kg S. Acid consumption is acid removed by gangue neutralization, metal dissolution, precipitation, and moisture losses. Net acid requirement is generation minus consumption: positive means purchased acid; negative means a neutralization duty.

How is gangue acid consumption measured in the laboratory?
A ground sample is contacted with a known excess of sulphuric acid at target pulp density, held for a defined time, and the residual free acidity is titrated. The difference, per dry tonne, is GAC. Because silicates dissolve slowly, specify contact time — 2-hour and 24-hour results can differ substantially on clay-rich feeds.

Why can a pyrite concentrate need no purchased acid at all?
Oxidizing pyrite releases two moles of acid per mole of FeS2 when iron precipitates as hematite or hydroxide: a 32% sulphide-S concentrate generates roughly 960 kg H2SO4 per tonne at 98% oxidation. Gangue and basic iron salts absorb perhaps 190 kg. The reagent bill appears as limestone, not as acid purchases.

How do I convert kg/t acid consumption into daily reagent demand?
Multiply the per-tonne figure by dry tonnes leached per day, then divide by the commercial strength fraction (0.98 for 98% acid). A circuit consuming 66.5 kg/t at 5,000 dry t/d requires 332.5 t/d of 100% acid, about 339 t/d at 98% — roughly 124,000 t/year.

Does a higher free-acid target always improve extraction?
Higher free acidity accelerates oxide copper dissolution and suppresses fouling, but raises gangue consumption and washed-solids acid losses. Acid cost scales with target g/L times solution volume, so the optimum is economic: raise the target only where testwork shows recovery benefit exceeding marginal reagent cost.

Why do plant acid purchases exceed bottle-roll predictions?
Typical causes: GAC drift as ore blends change; acid leaving in tailings moisture that lab washing did not replicate; acid strength corrections; growing Mg and Al inventories; and unrecorded neutralization additions. A reconciliation ledger rebuilt quarterly usually recovers the gap to within 2–3%.

How does POX acid stoichiometry differ from biooxidation?
Both oxidize sulphide to sulphate, but POX runs at 190–230 °C, where iron precipitates as hematite or basic iron sulphate, and BIS locks up to 1.75 kg H2SO4-equivalent per kg Fe. Biooxidation runs near ambient temperature, precipitating hydroxides and jarosites, so net acid release differs for identical feeds.

Conclusion

An acid consumption calculation is not a formula but a disciplined ledger: balanced reactions, fixed mass factors, one declared basis, and a closing check against solution chemistry — the connective tissue between the laboratory and the reagent contract. The two worked examples show the full range of behavior: a pyrite concentrate that generates a 770 kg H2SO4 surplus per tonne and must be designed around neutralization, and a mixed oxide ore that consumes 66.5 kg/t with gangue acid consumption dominating every other variable. Sensitivity analysis tells you where to spend testwork money: iron speciation and sulphide assay for POX feeds, GAC and pulp density for oxide leaches. Run quarterly, the ledger becomes an operating instrument, not a design relic. If you are evaluating a new leach or POX project, our pilot testing services generate the oxidation, speciation, and GAC data these calculations require; more practical guides are in our resource library.

References

1. F. Habashi, A Textbook of Hydrometallurgy, Métallurgie Extractive Québec (1993).
2. M. L. Free, Hydrometallurgy: Fundamentals and Applications, Wiley (2013).
3. J. Marsden & I. House, The Chemistry of Gold Extraction, 2nd ed., SME (2006).
4. Perry’s Chemical Engineers’ Handbook, 8th ed., McGraw-Hill.
5. USGS, Mineral Commodity Summaries 2024 — Copper.

Changelog

2026-10-24 — First published.