Maintaining feedwater pH within the alkaline range is a core requirement in industrial steam systems because dissolved carbon dioxide returning with condensate forms carbonic acid, which depresses pH and accelerates carbon steel attack. Regenerant alkalinity derived from ion-exchange systems—principally the residual sodium hydroxide and sodium carbonate discharged during strong-base anion resin regeneration—can be collected, filtered, and metered into demineralized water or reverse-osmosis permeate to neutralize dissolved carbon dioxide and buffer the feedwater. The primary neutralization reaction converts aqueous carbon dioxide to bicarbonate, CO₂(aq) + OH⁻ → HCO₃⁻, and a secondary hydroxide dose converts bicarbonate to carbonate, HCO₃⁻ + OH⁻ → CO₃²⁻ + H₂O. The dissociation constants of carbonic acid, pKa1 = 6.35 and pKa2 = 10.33 at 25 °C, establish that a pH between 8.3 and 10.0 stabilizes predominantly bicarbonate-carbonate buffer pairs while minimizing free CO₂. In low-pressure industrial firetube and watertube boilers, published ASME feedwater guidance lists a pH of 8.3 to 10.0, total hardness below 1.0 mg/L as CaCO₃, and dissolved oxygen below 7 µg/L. Recovered caustic regenerant is not a chemical-grade substitute without treatment; it requires particulate filtration to 5 µm, metals precipitation, and organic removal before injection into any feedwater train. Measured pH should follow ASTM D1293-18, carbonate alkalinity should follow ASTM D1067-16, and specific conductance should follow ASTM D1125-14.
Recovered caustic regenerant composition varies with the regeneration protocol: a strong-base anion unit regenerated with 4–6 wt% sodium hydroxide at 2–4 bed volumes per hour typically discharges an alkaline waste containing 1–3 wt% free sodium hydroxide, 0.5–2 wt% sodium carbonate, 0.1–0.6 wt% sodium bicarbonate, and residual sulfate, chloride, silica, and dissolved organic carbon. Published data for this specific configuration remains limited because regenerant alkalinity composition depends on resin age, rinse protocol, and raw water anion load; therefore batch characterization by the listed methods is required before feedwater reuse. The free hydroxide concentration is sufficient to neutralize acidic cation-regeneration waste when blended under controlled conditions. In field blending, the acid waste pH may remain below 2.0 and the alkaline waste pH near 13.0; the neutralization endpoint is commonly controlled between 6.0 and 9.0 to meet discharge compliance, but for reuse as feedwater conditioning a narrower endpoint of 8.5 to 9.5 is maintained before final polishing. The buffer capacity derived from this stream can offset the pH depression caused by residual CO₂ after vacuum degasifiers, which typically leave 5–15 mg/L dissolved CO₂ depending on tray efficiency and hydraulic loading.
| Parameter | Observed or design range | Analytical method or equipment basis |
|---|---|---|
| Free sodium hydroxide in spent regenerant | 1–3 wt% | ASTM D1067-16 |
| Sodium carbonate in spent regenerant | 0.5–2 wt% | ASTM D1067-16 |
| Spent regenerant pH | 12.5–13.5 | ASTM D1293-18 |
| Blended reuse pH endpoint | 8.5–9.5 | Two-stage neutralization skid with pH trim loop |
| Feedwater pH, low-pressure boiler | 8.3–10.0 | ASME feedwater guidance |
| Cation conductivity after polisher | <0.2 µS/cm | EPRI cycle chemistry guidance |
| Filtrate particulate retention | 5 µm nominal | Backwashable cartridge or bag filter |
At pressures above 900 psig, the role of regenerant alkalinity changes because carbonate/bicarbonate buffers partially decompose under high-temperature boiler water conditions, releasing carbon dioxide to the steam phase and leaving hydroxide in the liquid phase. The carbon dioxide partitions into the steam and depresses condensate pH, causing carbonic acid attack in return lines that a nonvolatile regenerant alkali cannot reach. Therefore, recovered caustic regenerant is used primarily to neutralize the bulk feedwater ahead of the deaerator, while volatile neutralizing amines or ammonia are still required for two-phase condensate protection. Published high-pressure cycle chemistry guidance recommends cation conductivity below 0.2 µS/cm for systems above 900 psig; recovered caustic addition must not introduce chloride or sulfate above this anion-equivalent threshold. Free hydroxide carryover into high-heat-flux waterwall zones is limited by maintaining boiler water chemistry on congruent phosphate rather than free caustic treatment. The feedwater pH target is commonly narrowed to 8.8–9.6 for high-pressure industrial drum boilers, with dissolved oxygen held below 7 µg/L by mechanical deaeration and chemical oxygen scavenging. Regenerant alkalinity is therefore appropriate for the deaerator influent and after condensate polishing, but it cannot replace volatile alkalizing agents where steam-phase transport of CO₂ controls the corrosivity of the condensate network.
Silica and charged organic matter present in recovered caustic regenerant pose a direct steam-purity risk. Strong-base anion resin releases trace trimethylamine and organic sulfonates if the resin has been fouled by humic substances; these contaminants partition into low-pressure turbine deposits or can form stable foams in the boiler drum. Regenerant reuse should therefore include ultrafiltration with a molecular weight cut-off below 10,000 Da or granular activated carbon contactors sized for 20 min empty-bed contact time. Residual silica in the neutralized blend should be limited to the boiler water silica threshold corresponding to the operating pressure, and the feedwater silica concentration should be verified by ASTM D859-16. Iron and copper released during neutralization of mixed acid streams should be removed by backwashable multimedia filtration because colloidal iron at 50 µg/L is sufficient to deposit on deaerator trays and feedwater pump wear rings. The use of regenerant alkalinity in systems with aluminum heat exchangers is not recommended because alkaline pH above 8.5 can exceed the amphoteric stability region of aluminum and produce localized alkaline attack.
Cation exchangers regenerated with 5–10 wt% hydrochloric acid or 1–4 wt% sulfuric acid produce rinse waters with pH below 2.0 and total dissolved solids above 10,000 mg/L. Blending this acid stream with alkaline anion-regeneration effluent neutralizes free acidity and precipitates hardness ions if the blend is allowed to overshoot above 9.0. In a continuous neutralization system, a two-stage reactor with 15 min average residence time per stage and pH trim loops using 30 % sodium hydroxide or recovered caustic gives a discharge pH between 6.0 and 9.0, which is consistent with typical industrial sewer discharge permits. However, when the neutralized stream is subsequently reused for boiler feedwater conditioning, the first-stage pH endpoint is held at 8.5–9.2, the slurry is passed through a lamella clarifier to remove precipitated metal hydroxides, and the clarified overflow is filtered to 5 µm before entering the feedwater tank. Sulfuric acid regenerant produces calcium sulfate precipitation risk; if the sulfate concentration exceeds 1,500 mg/L, sulfate removal or seeded crystallization may be required to avoid scaling of the recovery piping. Hydrochloric acid regenerant yields soluble calcium chloride and is preferred where brine recovery and feedwater alkalinity reuse are planned.
In neutralization tanks, stratification is a recurring failure mode when recovered caustic is introduced through a submerged distributor at low velocity. The top alkaline layer can reach pH 12.5 while the lower acid layer remains below pH 3.0, producing false endpoint readings and allowing corrosive slugs to leave the tank. A recirculation loop designed for 3 tank volumes per hour, a static mixer with 10–15 pipe diameters downstream before the pH sensor, and a vertical turbine mixer with 0.75 kW motor per 1,000 gal are common design provisions. The pH analyzer should sample from a side-stream after cooling to 25 °C as specified in ASTM D1293-18, with an automatic temperature compensator and dual-junction reference electrodes resistant to sodium ion error. Conductivity monitoring after cation exchange using ASTM D1125-14 is more sensitive than pH for detecting anion breakthrough from spent regenerant; a cation conductivity rise of 0.05 µS/cm above baseline indicates chloride or sulfate ingress and should automatically divert the recovered alkalinity stream to waste.
In plants with deep-bed condensate polishers, the polished effluent is often low in alkalinity and aggressive to carbon steel because all ionic species, including bicarbonate, have been removed. Regenerant alkalinity can be introduced after the polisher effluent line and before the deaerator to restore the pH buffering capacity lost during polishing. Because the polished effluent has a specific conductance below 0.1 µS/cm, the addition of even 10–20 mg/L of sodium bicarbonate or 5–10 mg/L of recovered caustic increases pH from below 6.5 to 8.8–9.2. The dosing point must be upstream of the deaerator by at least 30 seconds of pipe transit time to avoid localized high pH at the deaerator inlet nozzle. Injecting recovered caustic directly into the condensate pump suction can cause cavitation damage because the caustic plume is insufficiently dispersed and the pump inlet eye sees alternating high-pH and low-pH slugs. Field installations with polished condensate and recovered caustic have used an injection quill located at the center of a 24-inch line with 0.5 m/s minimum velocity and a static mixer immediately downstream; under these conditions the pH sensor downstream of the mixer tracks the dose within 20 seconds.
The deaerator removes dissolved gases but does not remove nonvolatile alkalinity; therefore any overdose of recovered caustic before the deaerator carries directly to the boiler as free hydroxide. In low-pressure boilers, free hydroxide in the boiler water can embrittle riveted seams or promote caustic gouging at tube bends and backing rings. For this reason, the feedwater pH measured at the deaerator outlet should be limited to 9.5 in systems with copper alloy feedwater heaters, because copper oxide dissolution accelerates above pH 9.2 in ammonia-bearing water. If copper alloys are absent, a pH up to 10.0 may be tolerated for low-pressure firetube boilers. Feedwater pump wear-ring clearances are also sensitive to abrasive particulates carried with recovered caustic; the total suspended solids should be kept below 1 mg/L and the particle diameter below 5 µm. Oxygen scavenger feed and caustic feed should be separated by at least 10 pipe diameters because localized high pH can reduce the effectiveness of catalyzed sodium sulfite.
| Control parameter | Limit or report value | Standard or reference |
|---|---|---|
| Feedwater pH measurement | 8.3–10.0 | ASTM D1293-18, ASME guidance |
| Carbonate alkalinity | Report as mg/L as CaCO₃ | ASTM D1067-16, ISO 9963-1:1994 |
| Specific conductance | Report at 25 °C | ASTM D1125-14 |
| Silica | Boiler pressure-dependent | ASTM D859-16, EN 12952-12:2003 |
| Sodium | Mass balance check | ASTM D4191-15 |
| Cation conductivity | <0.2 µS/cm above 900 psig | EPRI cycle chemistry guidance |
| Particulate filtration | 5 µm nominal | Cartridge filter manufacturer specification |
Recovered caustic regenerant adds sodium ions along with alkalinity, and the resulting sodium-to-alkalinity mass ratio influences boiler water cycles of concentration. In a softener-based feedwater system, the sodium-to-total-hardness ratio after blending should be checked to avoid exceeding the solubility limit of calcium carbonate in the deaerator. The use of sodium carbonate from spent caustic introduces carbonate alkalinity that can react with hardness leakage from a softener upset and form calcium carbonate deposits on feedwater control valves. Consequently, the feedwater hardness after final alkalinity addition should remain below 1.0 mg/L as CaCO₃, and the total alkalinity in the feedwater should not exceed 200 mg/L as CaCO₃ for low-pressure firetube boilers to avoid excessive carryover. Silica in recovered caustic is another critical boundary; at 600 psig, boiler water silica should be limited to 100 mg/L and at 1,000 psig approximately 40 mg/L, with lower values for turbines. Recovered caustic batches should be screened for silica by ASTM D859-16, and reuse should be interrupted if the calculated feedwater silica exceeds 0.02 mg/L for high-pressure systems. The sodium-to-alkalinity ratio is measured by ion chromatography or by calculation from sodium and total alkalinity determined with ASTM D4191-15 and ASTM D1067-16; this ratio is used to estimate the caustic demand before the deaerator.
For operational boundaries, pre-filtration of recovered caustic to 5 µm at relative humidity above 60 % prevents atmospheric carbon dioxide absorption in storage tanks; a nitrogen blanket or floating cover is specified because sodium hydroxide absorbs CO₂ and converts to sodium carbonate, reducing its neutralizing strength and increasing the solids inventory. Recovered caustic should not be mixed with ammonia or neutralizing amines in the same static mixer because localized precipitation of ammonium bicarbonate can block the injection quill. When the required feedwater pH is above 9.2 and the condensate contains dissolved CO₂ above 10 mg/L, a two-point alkalinity feed strategy—recovered caustic before the deaerator and volatile amine after the deaerator—is used rather than a single regenerative alkali dose. The regenerant alkalinity inventory should be batch-tested for total organic carbon; a TOC above 10 mg/L indicates fouled anion resin and requires diversion to wastewater treatment instead of feedwater reuse. Iron pickup during storage in carbon steel tanks should be limited by internal epoxy lining or by maintaining pH above 10.0 and using sealed storage with minimal air contact. These constraints define the safe operating envelope for using regenerant alkalinity in industrial boiler feed conditioning.