HCl for Steel Pickling: Bath Parameters, Inhibitors and Why It Beat Sulfuric

On continuous push-pickling lines processing low-carbon hot-rolled strip, hydrochloric acid is circulated through rubber-lined or brick-lined concrete tanks at flow rates that maintain strip travel between 0.5 m/s and 1.5 m/s. The acid dissolves mill scale composed primarily of wüstite, magnetite, and hematite through the reactions FeO + 2 HCl → FeCl₂ + H₂O, Fe₂O₃ + 6 HCl → 2 FeCl₃ + 3 H₂O, and Fe₃O₄ + 8 HCl → FeCl₂ + 2 FeCl₃ + 4 H₂O. The parallel base-metal reaction Fe + 2 HCl → FeCl₂ + H₂ is the reason inhibitor addition functions as a bath parameter rather than a secondary additive. Uninhibited HCl at 80 °C attacks clean steel faster than it attacks scale, particularly at strip edges and at cut ends, producing hydrogen gas and a risk of hydrogen-induced cracking in higher-strength steel.

Bath temperature and dissolved iron are coupled variables. As ferrous chloride accumulates, free acid must be compensated by density-controlled acid injection; otherwise the bath appears correct by titration but loses pickling rate because water activity decreases and the boiling point rises. In practice, continuous lines discharge spent liquor when ferrous iron reaches 120–150 g/L, because above this range FeCl₂ begins to crystallize in low-velocity zones such as drain legs, pump suction lines, and heat-exchanger shells. Specific gravity measured at 25 °C is used as a rapid in-line proxy for iron content, but the measurement requires compensation because free HCl and dissolved ferrous chloride both increase density. A typical bath volume on a medium-width line is 30–60 m³; this volume delays concentration changes when incoming strip scale thickness varies.

What Operating Limits Prevent Base Metal Dissolution Without Slowing Scale Removal?

The working window for a continuous hydrochloric acid line is defined by free HCl, bath temperature, ferrous ion, and inhibitor concentration. A free acid drop below 30 g/L lengthens residence time and leaves incomplete scale removal on strip edges; an increase above 150 g/L raises fume load and accelerates inhibitor consumption. Bath temperature is normally held between 70 °C and 85 °C. At 60 °C, wüstite-rich scale can still be removed, but throughput drops because the dissolution rate falls. At 90 °C, some film-forming inhibitors lose surface residence and base-metal attack rises sharply. Ferrous iron is maintained at 80–150 g/L, not because ferrous chloride is insoluble, but because high iron content slows scale dissolution and increases pickling time for a given free-acid concentration.

The range shown in Table 1 applies to low-carbon hot-rolled strip with scale thickness between 5 µm and 15 µm and strip speed of 25–60 m/min. Heavier scale from slab reheat or from extended delay after rolling requires a longer immersion cell or higher temperature, not simply higher acid concentration, because concentrated HCl at low iron content can attack the substrate at exposed edges before the heavier centerline scale is removed. Pickling time for continuous strip lies commonly between 20 s and 120 s, but the exact value is scale-thickness and temperature dependent. Published data for a specific hot-rolled grade is limited, so line trials are used to establish the speed-temperature curve.

Free acid is determined by sodium hydroxide titration to the methyl orange endpoint at approximately pH 4.5. The sample is diluted with distilled water because ferric chloride hydrolysis at higher temperature precipitates hydroxide and masks the endpoint. Ferrous iron is titrated with standard potassium dichromate using diphenylamine sulfonate indicator; ferric iron is quantified by reducing with stannous chloride and taking the difference. These methods are low-cost but require a stable bath matrix. In lines with high throughput, automatic titrators are preferred over manual analysis because the acid and iron concentrations can shift by more than 10 g/L during a single shift when scale thickness varies. Make-up acid is normally 31–37 wt% hydrochloric acid, but dilution with rinse water is controlled to avoid localized boiling. The addition point is placed before the recirculation pump or through a static mixer; direct injection near the strip causes pitting due to local acid concentration.

Hydrochloric acid pickling baths are not typically filtered to remove suspended scale particles because the particles are mostly acid-soluble. However, a side-stream settler or cyclone removes sand and undissolved scale fragments from the strip entry. Accumulated sludge is drained from the bottom cones of the pickle tank every 8–24 h; failure to do so reduces effective bath volume and can block circulation nozzles. At free HCl above 150 g/L, the titration curve becomes nonlinear due to FeCl₂ hydrolysis; dilution before titration is therefore mandatory.

Inhibitor development for hydrochloric acid pickling has concentrated on nitrogen-bearing heterocyclic film formers, acetylenic alcohols, and condensation products that adsorb on the bare steel surface without blocking the oxide/acid interface. Typical continuous-bath inhibitor dosage is 0.05–0.5 vol% of bath volume. The lower limit is set by edge pitting, the upper limit by cost and foaming in high-turbulence lines. Inhibition efficiency is evaluated by coupon immersion according to ASTM G31-72 (2018) or an equivalent static mass-loss method. Acceptable proprietary inhibitors usually reduce carbon steel corrosion in 10 wt% HCl at 80 °C by more than 99% compared with the uninhibited blank. Published data for specific proprietary formulations is limited, so qualification in spent-bath liquor with the actual dissolved iron concentration is required before a new inhibitor is placed in continuous service.

Production-scale inhibitor failures are usually detected during strip inspection rather than by bath titration. A line operating at 1.0 m/s may show acceptable strip-body pickling but visible edge attack if the inhibitor film is disrupted by high ferric ion pockets or if fresh acid injection is not diffused before contacting the strip. Ferric ion above 5 g/L can oxidize organic inhibitor molecules; this is one reason spent hydrochloric baths with long dragout residence time are less inhibitor-efficient than freshly prepared baths. Foaming in the recirculation tank may indicate an overdose of surfactant in a blended inhibitor, not a hazardous condition by itself, but it reduces pump suction head and can starve the strip spray headers. Conversely, underdosing first appears as pitting at sheared edges, because film formation is weakest at high-energy sites.

Inhibitor chemistry is generally either cathodic or mixed. Cathodic ingredients such as quaternary pyridinium compounds raise the hydrogen overpotential on the steel surface, while acetylenic alcohols polymerize into a thin organic film that limits iron dissolution. Mixed inhibitors based on Mannich-base condensation products provide both functions and are preferred where the bath temperature exceeds 80 °C because single-component films can desorb. Thiourea derivatives are effective but are avoided in closed buildings because their decomposition can produce hydrogen sulfide; this is a compatibility boundary, not a performance limit.

Hydrogen Effusion, Fume Scrubbers, and Ferrous Chloride Solubility

Hydrogen evolution in inhibited HCl pickling is significantly lower than in sulfuric acid pickling at equal descaling rate because HCl attack on wüstite is chemically fast enough that less base steel is consumed before scale removal is complete. Sulfuric acid at 10–25 wt% and 80–95 °C often removes scale by hydrogen blistering and subscale attack; the ferrous sulfate product has limited solubility and may redeposit on the strip as a slimy salt layer. Hydrochloric acid at 5–20 wt% and 60–90 °C leaves a cleaner, brighter substrate because ferrous chloride is highly soluble and is carried out of the reaction zone. The practical result in continuous lines is that HCl can operate at a lower temperature for the same speed, which reduces heat input, fume load, and base metal loss.

These differences explain why HCl has replaced sulfuric acid in most new high-speed strip pickling installations where spray-roaster regeneration is permitted. The selection is not universal: sulfuric acid can still be preferred when spent acid is available as a low-cost byproduct or when an existing ferrous sulfate crystallization system must remain in use. Where hydrogen embrittlement control is critical for high-strength steel, HCl with a properly maintained inhibitor is generally preferred because the lower temperature and shorter exposure reduce hydrogen uptake. Residual hydrogen after pickling is influenced less by the total gas evolved than by the local surface coverage of the inhibitor, which is why edge attack in an uninhibited area is considered a more serious process excursion than a slow rise in total iron.

The stoichiometric acid consumption of HCl for scale removal depends on the iron oxide phase. For wüstite, 1 kg of iron in the scale consumes approximately 1.3 kg of HCl; for hematite, the same mass consumes approximately 1.9 kg of HCl. Actual consumption is higher because spent liquor retains free acid and dragout carries acid into rinse water. In a continuous line without regeneration, total HCl consumption can range from 15 kg to 35 kg of 31% acid per tonne of steel; regeneration lowers fresh acid demand to make-up losses of less than 5 kg per tonne in some installations. These numbers illustrate the economic pressure for acid recovery rather than a universal recipe.

When a Continuous Line Replaces Sulfuric with Hydrochloric and Regeneration Capacity Must Change

Conversion of a sulfuric acid pickling line to HCl involves more than replacing the acid storage tank. Sulfuric acid lines often use steam coils made of stainless steel or lead; hydrochloric acid at 70–85 °C rapidly attacks austenitic stainless steel by chloride stress corrosion cracking and is not handled in lead. Graphite, tantalum, or fluoropolymer-sheathed heat exchangers are used for heating. Pumps with stainless steel wetted parts are replaced with polypropylene or PVDF units, silicon carbide mechanical seals, or sealless magnetic-drive pumps. Fume extraction must handle hydrochloric acid vapor; a packed-bed scrubber with dilute sodium hydroxide solution is used to keep stack emissions within applicable limits. The Occupational Safety and Health Administration ceiling for HCl is 5 ppm (7 mg/m³); NIOSH uses the same ceiling value. Scrubber blowdown pH is typically controlled between 8 and 9 to prevent acid carryover without excessive caustic consumption.

Rinse water from a hydrochloric line is more corrosive because residual chloride is not easily removed by lime neutralization alone; therefore the rinse cascade must have more stages than a sulfuric line of equal throughput. However, spray roasting of spent FeCl₂ liquor recovers HCl at 500–800 °C and produces iron oxide. The roaster feed is sprayed into a hot combustion gas stream; water and HCl vaporize, ferrous chloride hydrolyzes to Fe₂O₃, and the off-gas passes through cyclones and a venturi absorber to recover 18–20 wt% hydrochloric acid. Acid recovery can exceed 99% in a well-operated roaster, which is a key reason for selecting HCl in regions with strict sulfate discharge limits. Silica and alkali contamination are operational boundaries for this regeneration route; silica forms deposits in the roaster, while sodium and potassium chlorides reduce acid recovery and can corrode the refractory lining.

Wire and fastener plants choose HCl because the pickled surface has less smut and better phosphate coating adhesion. In sulfuric pickling, carbon smut left on the surface can reduce coating adhesion. HCl removes scale without excessive production of carbon smut because the acid attack undercuts the scale rather than oxidizing the steel surface. The smoother surface also reduces drawing die wear; this is a downstream manufacturing benefit, not a chemical descriptor. Surface cleanliness after pickling is inspected visually against ASTM A380-17 or an equivalent cleanliness practice. A water-break test is used for critical surfaces; any beading indicates residual oil or oxide.

Batch wire-rod pickling with HCl operates at 40–60 °C and uses immersion times from 5 min to 20 min depending on rod diameter and scale condition. Agitation is maintained by air sparging or mechanical circulation to prevent localized ferric chloride enrichment. After pickling, the rod is rinsed in cascading water, then coated with phosphate or borax; residual chloride should be below 10 mg/m² before coating to avoid underfilm corrosion. Galvanized or tin-plated wire requires a different rinse sequence because chloride carryover into flux solutions can destabilize the flux and produce bare spots. In these batch operations, bath life is extended by daily bleed and feed rather than acid recovery, and the spent liquor is neutralized with lime to produce a hydroxide sludge that is dewatered to 30–40% solids before disposal.

Maintenance inspections on an HCl pickling line should include rubber lining spark testing, flange gasket inspection, and heat exchanger pressure testing at 1.5× design pressure. Because ferric chloride can polymerize in stagnant pockets, shutdown procedures include draining and rinsing with inhibited water. The inhibitor may be incompatible with anionic wetting agents from earlier chemical treatments; a tank change without complete cleaning can produce foaming or a sudden drop in inhibitor efficiency. For this reason, inhibitor changes are qualified in a side-stream test loop before full-bath implementation.