Powdered NaOH pH Setpoints and Selectivity in Refinery Gas Scrubbing

Powdered sodium hydroxide is introduced into refinery fuel gas scrubber recirculation loops as a dry alkalinity source when site logistics favor solid handling over delivery of 50 wt% railcar caustic or when water balance constraints require that additional dilution be minimized. In acid gas contactors, the control of pH is not a simple alkalinity inventory problem because the two principal sour gas constituents, hydrogen sulfide and carbon dioxide, respond to hydroxide ion concentration through fundamentally different reaction regimes. Hydrogen sulfide undergoes proton transfer to hydroxide at a rate that is effectively diffusion-limited in the liquid film, whereas carbon dioxide is consumed by a finite second-order reaction with hydroxide exhibiting a reported rate constant near 8.5×10³ L mol⁻¹ s⁻¹ at 25 °C. The resulting separation is therefore controlled by the liquid-film enhancement factor for carbon dioxide, which remains below 0.5 at pH values near 10.0 but exceeds 1.2 at pH 11.0 and 3.7 at pH 12.0 under standard kinetic assumptions. This pH-dependent acceleration of carbon dioxide fixation is the central reason why high-pH operation increases caustic consumption and generates bicarbonate-carbonate scale, while selective operation at a lower pH retains hydrogen sulfide absorption because the H₂S–OH⁻ reaction remains well within the fast pseudo-first-order regime. pH setpoint selection for a given scrubber must also account for total dissolved sulfide concentration, the gas-side mass transfer coefficient, the liquid residence time in the sump, the carbonate alkalinity already present in the blowdown, and the sodium ion error of the pH electrode in high-alkalinity service.

What pH control band preserves hydrogen sulfide selectivity against carbon dioxide in recirculated caustic scrubbing?

Maintaining a pH control band between 9.5 and 10.5 is generally required for selective H₂S scrubbing when the scrubber feed contains both H₂S and CO₂, because the dimensionless Hatta number for CO₂ remains below approximately 0.4 while the corresponding Hatta number for H₂S exceeds 200. The Hatta number, defined as the square root of the product of the second-order rate constant, the hydroxide concentration, and the dissolved acid-gas diffusivity divided by the physical liquid-film mass transfer coefficient, determines whether a reaction is confined to the interface or occurs throughout the bulk liquid. At pH 10.0 and 25 °C, the hydroxide ion concentration is 1.0×10⁻⁴ mol/L, the CO₂–OH⁻ rate constant is 8.5×10³ L mol⁻¹ s⁻¹, the diffusivity of CO₂ in water is approximately 1.6×10⁻⁹ m²/s, and a representative liquid-film coefficient in a packed tower is 1.0×10⁻⁴ m/s, yielding a Hatta number of 0.37. At pH 11.0, the same inputs yield 1.17, and at pH 12.0 the value becomes 3.69. H₂S, by contrast, reacts with hydroxide through a proton transfer pathway whose apparent second-order rate constant may be taken as 1.0×10¹⁰ L mol⁻¹ s⁻¹; even at pH 8.0 the Hatta number exceeds 40, so H₂S absorption remains gas-film controlled and is relatively insensitive to pH within the range 8.0–11.0. This asymmetric kinetic response means that reducing pH from 12.0 to 10.0 suppresses CO₂ uptake by roughly an order of magnitude while leaving H₂S removal efficiency governed chiefly by gas-side resistance and liquid-to-gas ratio. The lower end of the selective band is bounded by H₂S retention: the first pK_a of H₂S is 7.0, so at pH 8.0 only 91% of dissolved sulfide is present as HS⁻, whereas at pH 9.0 the HS⁻ molar fraction exceeds 0.99. The upper end is bounded by the second pK_a of the carbonate system at 10.33, above which the carbonate ion fraction rises rapidly and calcium carbonate scaling becomes kinetically favourable in hard waters.

Stoichiometric alkalinity consumption differs significantly between selective and nonselective operation. Absorption of 1 kg of H₂S consumes 1.17 kg of NaOH if the product is NaHS, or 2.35 kg of NaOH if the product is Na₂S. Absorption of 1 kg of CO₂ consumes 0.91 kg of NaOH if NaHCO₃ is formed, or 1.82 kg of NaOH if Na₂CO₃ is formed. In a refinery fuel gas containing 1,000 ppmv H₂S and 2.0 mol% CO₂, the molar CO₂-to-H₂S ratio is 20:1; complete CO₂ absorption as Na₂CO₃ would consume 20 times as much caustic as complete H₂S absorption as Na₂S under nonselective conditions. Selective scrubber operation therefore has both an economic limit and a scaling limit, and the pH setpoint is used to reduce CO₂ absorption efficiency to less than 10–20% of the CO₂ feed. Because hydrogen sulfide absorption is gas-film controlled, liquid-side mixing and pH have only secondary effects on H₂S removal; carbon dioxide absorption is liquid-side or kinetically controlled and is therefore sensitive to residence time and liquid hold-up. Selective caustic scrubbers consequently tend to use low liquid hold-up, controlled irrigation density, and moderate pH, while nonselective high-pH towers use higher hold-up and higher free caustic concentration. Published data for a particular refinery fuel gas composition may differ because hydrogen partial pressure, operation at elevated pressure, and the presence of olefins in fluid catalytic cracking off-gas alter gas-liquid equilibrium and the effective liquid-film mass transfer coefficient.

When Powdered Sodium Hydroxide Replaces 50 wt% Liquid Caustic in Refinery Fuel Gas Polishing Service

Dry powdered sodium hydroxide substitution alters the dynamics of pH control because the heat of solution and the dissolution rate must be managed before the material reaches the scrubber sump. Anhydrous NaOH releases approximately 1,112 kJ/kg when dissolved to a finite concentration, and a 10 wt% solution made with 20 °C dilution water can exceed 50 °C adiabatically; a 15 wt% solution can approach 65 °C depending on mixing heat losses and the specific heat of the final solution. Consequently, the powder handling system should be configured with a 316L stainless steel dissolver, an eductor-driven wetting cone, a hinged bag dump station with dust extraction, and a positive-displacement metering pump that transfers the resulting 10–15 wt% caustic solution to the scrubber circulation line downstream of the recirculating pump. Direct addition of dry NaOH to a scrubber sump is not recommended because localized high pH zones promote carbonate precipitation on the pH electrode and mineral scale on pump impellers. The dissolution skid should include a 2.0–5.0 m³ tank with a 0.75–1.5 kW pitched-blade mixer, a 2 mm basket strainer, and a low-level interlock to prevent the transfer pump from running dry. Published vendor data for a specific powder-dissolution skid configuration is limited; the values quoted here are typical industrially accepted ranges and must be confirmed against the selected equipment supplier’s heat and mass balance. Because powdered NaOH is deliquescent, hopper blanketing and dust extraction should use instrument air with a dew point below -40 °C when ambient relative humidity exceeds 60%.

Because sulfidic liquors poison silver-silver chloride reference junctions through Ag₂S precipitation and carbonate deposits coat the measuring surface, the pH loop requires a double-junction electrode with potassium nitrate reference electrolyte, a flat or guarded glass membrane, and a retractable insertion housing that permits removal under system pressure. The sample stream should be extracted from the circulating line after the injection point but ahead of the scrubber spray header, passed through a 0.5 mm strainer and a heat exchanger to bring the sample to 25–35 °C, and returned to the sump to avoid a steady blowdown loss. pH calibration should follow ASTM D1293-18 and ISO 10523:2008, using two buffers bracketing the setpoint and a third buffer to verify slope; a drift of more than 0.2 pH units over an eight-hour shift typically indicates coating or sulfide poisoning. The controller should use a proportional-integral algorithm with a proportional band no greater than 1.0 pH unit and an integral time of 60–180 seconds, with the pump stroke limited to 20% of full scale per minute to avoid overshoot. Because pH measurement alone cannot distinguish carbonate alkalinity from hydroxide alkalinity at high pH, a conductometric or total alkalinity titration should be run at least once per day on the scrubber sump sample. Above pH 12.5, sodium ion error of standard glass electrodes becomes measurable; electrodes with lithium glass formulations and high-alkalinity correction are required for reliable service in concentrated caustic or high free-hydroxide liquors.

Spent Caustic Oxidation Alkalinity Demand, Carbonate Scaling Boundaries, and Sulfur Speciation

When the recirculated scrubber liquor is directed to spent caustic oxidation after sulfide loading rises, the pH setpoint shifts from selective acid gas rejection to sulfur species retention and prevention of volatile H₂S release. Spent caustic oxidation reactors require pH above 11.5 to keep dissolved sulfide as HS⁻ and prevent H₂S release from the feed tank. Powdered NaOH is often added to the spent-caustic feed conditioning tank to maintain pH between 11.5 and 12.5 before oxidation; at this pH, sulfide is predominantly HS⁻ and the CO₂ absorbed earlier in the scrubber exists mostly as carbonate. During wet air oxidation, sulfide is oxidized to thiosulfate and sulfate, and sulfate accumulation decreases pH; therefore supplemental NaOH must be added continuously to the oxidation reactor to hold the pH above 11.0 if complete sulfide destruction is required. The pH control point in oxidation service differs from the selective scrubber setpoint because the goal is no longer CO₂ rejection but rather retention of sulfur species in the aqueous phase and prevention of volatile H₂S breakthrough. At pH 11.5, the equilibrium H₂S partial pressure over a sulfide-containing liquor is negligible for dissolved sulfide concentrations up to several thousand milligrams per litre; below pH 8.5 the same solution can release significant un-ionized H₂S, especially at elevated temperature.

Carbonate scaling is governed by the second pK_a of the carbonate system at 10.33. At pH 10.0, the carbonate ion fraction is approximately 0.32; at pH 10.5, it rises to approximately 0.60; and at pH 11.0, it exceeds 0.82. If calcium or magnesium hardness enters through dilution water or tower drift, carbonate scale can blind packing and reduce the gas-liquid interface, so the selective pH band is bounded on the upper end by scaling potential and on the lower end by H₂S retention. Softened process water with total hardness below 1 mg/L as CaCO₃ is required for make-up if the scrubber is operated above pH 10.5 for extended periods. Operational boundaries for powdered NaOH pH control in refinery gas scrubbing include: keep the scrubber sump pH between 9.5 and 10.5 for selective H₂S rejection of CO₂; avoid pH below 8.5 unless the vessel is gas-blanketed and designed for acid gas release; avoid pH above 12.0 unless carbonate scaling and caustic consumption are explicitly accepted; pre-dry the powder handling system and purge hopper headspace with instrument air at a dew point below -40 °C when ambient relative humidity exceeds 60%; use 316L stainless steel only below 50 °C for 10–20 wt% NaOH; and do not combine powdered NaOH make-up with amine-based H₂S scavengers in the same closed recirculation because the high pH can liberate amine and generate mixed carbonate/carbamate precipitates that foul exchangers. Analytical verification should include ASTM D4084-07 for gaseous H₂S, ASTM D1945-14 for CO₂ and fuel gas composition, ASTM D1293-18 and ISO 10523:2008 for pH, and ion chromatography per ASTM D4327-13 for sulfate, thiosulfate, and carbonate in the spent caustic blowdown. For installations where the refinery fuel gas H₂S concentration exceeds 1,000 ppmv or the CO₂-to-H₂S molar ratio exceeds 50:1, published data for this specific configuration is limited and the pH setpoint must be validated with a pilot column using the actual hydrocarbon matrix, liquid-to-gas ratio, and scrubber temperature.