Sulfuric acid remains the principal acidulation reagent for converting fluorapatite-bearing phosphate rock to wet-process phosphoric acid and phosphogypsum. Phosphate fertilizer production consumes approximately 60% of global sulfuric acid output by mass. In the dihydrate route, fluorapatite Ca₅(PO₄)₃F reacts with 5 mol H₂SO₄ and 10 mol H₂O to yield 3 mol H₃PO₄, 5 mol CaSO₄·2H₂O, and 1 mol HF. Most merchant-grade phosphate plants are designed around sulfur-burn acid with a concentration of 93–98 wt% H₂SO₄, a low trace-metal burden, and predictable filtration behaviour. A by-product smelter acid stream can be introduced into this process if the impurity profile, supply logistics, and quality assurance system are formally qualified. This article addresses Case Study: Qualifying a Smelter Acid Supplier for a Phosphate Plant, with emphasis on wet-process filtration risk, corrosion metallurgy, and gypsum disposal constraints.

As a dense oily liquid, sulfuric acid has a specific gravity of 1.84 at 93–98 wt% H₂SO₄. Freezing points vary sharply with concentration: 93 wt% acid freezes near −32 °C, while 98 wt% acid freezes near 3 °C. Railcar deliveries of 98 wt% smelter acid therefore require heating during cold weather. At 20 °C, the viscosity of 96–98 wt% H₂SO₄ is approximately 21–26 mPa·s. The product is classified under CLP as Skin Corr. 1A, H314, and is transported as UN 1830. Standard acidimetry per ASTM E223-16 is used to verify total acidity; concentration is reported as wt% H₂SO₄, and density or Baumé gravity is used for transfer-volume correction.

What Impurity Profile Distinguishes Smelter Acid from Virgin Sulfur-Burn Acid?

Smelter acid originates from SO₂-containing off-gas in copper, lead, zinc, or molybdenum roasting and smelting. After gas cleaning, the sulfur dioxide is converted in a contact acid plant. Because upstream gas cleaning cannot remove all volatile metals and halides, smelter acid typically carries higher arsenic, chloride, fluoride, mercury, selenium, lead, and fine particulates than sulfur-burn acid. The acid strength is not the primary qualification issue: a well-operated smelter acid plant can deliver H₂SO₄ at 93–98 wt%, within the same range as a virgin sulfur-burn acid plant. The procurement risk is the trace impurity spectrum and its effect on filter cake permeability, evaporator pitting, and gypsum leachate.

The comparison above is a screening framework, not a universal specification. Site acceptance limits must be derived from the specific phosphate rock mineralogy, plant metallurgy, and gypsum disposal permit. A smelter acid lot that is acceptable for a dihydrate plant processing igneous fluorapatite may be unacceptable for a plant processing high-organic sedimentary rock, because humic substances interact with metallic impurities and change filter cake rheology. Therefore, qualification is always site-specific and must include rock-acid interaction testing rather than relying on supplier certificates alone.

Sampling of by-product smelter acid is more demanding than sampling of sulfur-burn acid because fine particulates can settle in railcars and road tankers. A representative sample must be drawn from top, middle, and bottom zones after recirculation or after road tanker pressure-offloading has agitated the heel. Sampling devices are constructed of PTFE or polypropylene; glass sample bottles are avoided because trace fluoride can leach borosilicate glass. Samples for mercury are preserved with dichromate in nitric acid and analyzed within 28 days. Without this sampling discipline, a supplier certificate may underestimate the particulate load delivered to the phosphate plant.

Storage of by-product smelter acid at the phosphate plant requires carbon steel tanks with a design corrosion allowance of 3–6 mm, provided the acid concentration is maintained above 93 wt% and the temperature remains below 40 °C. If the tank headspace is allowed to ingest humid air, sulfuric acid absorbs water and weakens at the surface, accelerating iron corrosion and generating hydrogen gas. Vent desiccants or dry-air pads are therefore used. Pump materials are usually carbon steel or Alloy 20 for tank transfer, but high-shear pumps handling acid with particulate loads above 100 mg/kg require hardened trim to limit erosion-corrosion.

Three-Stage Qualification of Smelter Acid for Dihydrate Phosphoric Acid Service

The qualification sequence in the case study follows three stages: supplier capability audit, bench-scale acidulation, and a limited plant trial. In the supplier audit, the supplier is required to demonstrate ISO/IEC 17025:2017 laboratory accreditation, statistically stable lot-to-lot impurity data over no fewer than 12 months, and REACH registration for the acid stream. The supplier is also required to disclose the upstream smelting route, gas-cleaning sequence, and any blending operations that may affect trace impurity speciation. If the smelter acid originates from a copper smelter with a wet electrostatic precipitator, the particulate carryover profile may differ from that of a lead smelter with a gas-phase mercury condenser. These differences affect the phosphate plant risk assessment.

A bench-scale acidulation rig typically comprises a 2 L jacketed reactor with PTFE baffles, a razor-blade impeller, and a vacuum leaf filter. The phosphate rock is ground to 90% passing 150 µm, then mixed with recycled weak phosphoric acid and smelter acid. The reaction is held at 75–80 °C for 2–4 h to approximate a dihydrate plant. The resulting slurry is vacuum-filtered at 200–250 mbar absolute, and the filtration rate is measured as kg dry gypsum m⁻² h⁻¹. A candidate smelter acid can be considered equivalent only if the filtration rate is statistically not less than the sulfur-burn control at 95% confidence; if the rate drops by more than 5%, the lot is referred for impurity-specific precipitation testing. Concurrent analyses include water-soluble P₂O₅ loss in the filter cake and particle size distribution by laser diffraction according to ISO 13320:2020.

In dihydrate phosphoric acid, the free sulfate concentration is controlled between 1.5 and 3.0 wt% SO₃. If insufficient sulfate is present, gypsum crystal growth is inhibited and the filter cake becomes slimy; if excess sulfate is present, calcium sulfate may precipitate rapidly on the rock particles and reduce P₂O₅ recovery. Smelter acid with high metal loads can interact with this free sulfate window by forming sulfate complexes with iron and aluminium, shifting the measured free sulfate. The plant trial therefore includes frequent sulfate titrations by ion chromatography or precipitation titration; the target is re-established with acid feed blend changes, not assumed constant.

The plant trial uses a segregated acid storage tank and a controlled feed rate to the reactor. Typically 500–1,000 t of smelter acid is blended with the normal sulfur-burn acid feed over 7–14 days. During this window, the free sulfate in the filter acid is held at 1.5–3.0 wt% SO₃, and the filter cloth pressure drop is monitored continuously. A rise in pressure drop greater than the site-specific baseline upper control limit at constant solids loading indicates blinding, which may be caused by metallic gels or fine crystals; the trial is then stopped and the filter cake is examined by scanning electron microscopy for crystal aspect ratio changes. Corrosion coupons of 316L are installed in the evaporator recirculation loop before the trial and removed at the end for mass-loss and pitting evaluation according to ASTM G1-03 and ASTM G48.

The table should not be read as a substitute for a formal management-of-change review. In plants that produce feed-grade phosphoric acid, trace-metal limits are often lower than fertilizer-grade limits, and the smelter acid blend fraction may be restricted to a single truck per batch until long-term data are available.

Trace-metal analysis in concentrated sulfuric acid is subject to matrix suppression. Direct ICP-OES analysis of undiluted 98 wt% acid is not practical because of viscosity, nebulizer blockage, and spectral background. The standard practice is a 1+9 dilution by mass into high-purity water, which raises the detection limit by a factor of 10. For arsenic and selenium, hydride-generation ICP-AES or ICP-MS may be required to achieve the lower fertilizer-grade thresholds. The supplier and the phosphate plant laboratory should agree on a referee method before the first lot is shipped.

When Chloride and Arsenic Approach Phosphate Plant Tolerance Bands

Chloride is the most consequential corrosion species in smelter acid for phosphate service. In the wet-process phosphoric acid evaporator, free chloride concentrates with recycled acid and can initiate pitting on 316L stainless steel at temperatures above 60 °C. The threshold depends on chloride concentration, sulfate activity, and pH. If the smelter acid contributes more than 10–20 mg/kg Cl to the reactor, the resulting recirculation loop should be evaluated by electrochemical polarization. Plants with 316L evaporators may require dilution with low-chloride sulfur-burn acid or upgrade to Alloy 20, 904L, or super-duplex materials in chloride-rich zones. Arsenic is less corrosive but enters the product acid. The acidulation reaction does not remove arsenic; it partitions between phosphoric acid and gypsum depending on sulfate and iron content. High arsenic smelter acid can push merchant-grade acid outside fertilizer or feed-grade specifications and require sulfide precipitation or solvent extraction.

Operational boundaries therefore include a maximum smelter acid blend fraction rather than a fixed impurity concentration. A conservative first-qualification position is to limit smelter acid to 10–20% of the total acid feed for the first 30 days and to maintain the flash cooler scrubber pH at 2–4 to absorb stripped SO₂. If the supplier lot has detectable nitrosylsulfuric acid from incomplete NOx scrubbing, the acid should not be stored in carbon steel tanks because of hydrogen evolution and stress-corrosion risk; PTFE-lined or 316L storage is used until the concentration of reducible nitrogen species is below the plant limit. Avoid combining high-NOx smelter acid with organic flocculants that are susceptible to oxidation, because the resulting redox reactions can reduce scale-inhibitor performance in the filter feed.

In the dihydrate process, filter rate is controlled by gypsum crystal size and aspect ratio. Well-formed gypsum crystals of 80–120 µm with a compact aspect ratio give high filtration rates. Foreign ions, particularly Fe³⁺ and Al³⁺, can poison crystal growth, producing needle-like or tabular crystals that blind the filter cloth. Smelter acid may introduce these ions in milligram-per-kilogram quantities, but the effect is magnified by recycle loops. A side-stream pilot filter should be run for at least 72 h to detect crystal habit drift; scanning electron microscopy is used to measure the thickness-to-length ratio of gypsum crystals.

Published data on arsenic partitioning for specific rock-smelter acid combinations is limited. Bench and pilot tests are therefore necessary whenever the smelter acid arsenic load exceeds the historical sulfur-burn acid baseline by more than 2 mg/kg in the diluted acid feed. The measurement endpoint is not total arsenic alone; it is the combined response of product acid purity, filter-cloth life, and gypsum leachability.

Mercury, Selenium, and Gypsum Disposal Constraints

Mercury and selenium are priority pollutants in phosphogypsum stack leachate. In the dihydrate process, mercury may partition to the product acid or to the gypsum depending on chloride and redox conditions. Selenium can be leached from phosphogypsum under alkaline conditions. A smelter acid qualification therefore includes a leachate risk review using the phosphate plant's phosphogypsum permit limits. The filter cake is subject to Toxicity Characteristic Leaching Procedure EPA 1311; mercury is determined by EPA 7470A, and selenium by EPA 6010D after acid digestion. If leachable mercury exceeds the permitted land-disposal threshold, the smelter acid blend ratio must be reduced or the acid must be treated for mercury removal before acidulation. These constraints are site-specific and may be more restrictive than corrosion or filtration limits.

Gypsum from smelter acid campaigns may show slight discolouration from iron, lead, or selenium. Discolouration alone is not a specification failure unless the receiving environment or by-product gypsum user imposes a whiteness index. The relevant measurements are leachable concentration, free moisture, and residual water-soluble P₂O₅. For by-product gypsum sold to wallboard producers, additional parameters such as total heavy metals and alpha-emitting radionuclides may apply.

Post-qualification monitoring for a qualified smelter acid stream consists of weekly trace-metal composites from the acid storage tank, monthly filter cloth inspection, and corrosion coupon retrieval at the evaporator recirculation loop. If any parameter drifts outside the baseline control limits established during the plant trial, the supplier lot is quarantined and the blend fraction is reduced to the previous qualified level.