CCGT’s wastewater reckoning: lime for phosphate, high‑pH for zinc

The wastewater coming off combined‑cycle gas‑turbine (CCGT) units isn’t just warm water. It’s cooling‑tower blowdown and process effluent loaded with nutrients (notably phosphate and ammonia) and metals such as zinc (Zn) and copper (Cu) from corrosion‑control programs. Regulators are moving fast: U.S. EPA proposals target cooling‑water Zn (≤1 mg/L) and chromium (Cr, ≤0.2 mg/L) in effluents (power-eng.com), while many new permits demand essentially zero phosphate discharge (power-eng.com).

The ecological logic is straightforward: heavy metals like zinc and copper are highly toxic to aquatic life (power-eng.com), and phosphates fuel [harmful algal blooms](https://beta.co.id/en/blog/red-tide-real-risk-a-desalination-contingency-plan-for-harmful-algal-blooms). By contrast, older cooling regimes can leave system water at 5–15 mg/L phosphate and roughly 0.5–2.5 mg/L Zn (chemengonline.com)—so removal is essential.

In Indonesia and elsewhere, environmental standards (e.g., PermenLH rules) similarly require low‑nutrient, low‑metal effluent; while specific limits vary, industry best practice aims for single‑digit mg/L total phosphorus (P) and sub‑ppm heavy metals. Note: mg/L is milligrams per liter; “ppm” is parts per million and often interchangeable with mg/L in water.

Phosphate precipitation using lime

Orthophosphate (PO4³⁻) is typically removed by converting it to insoluble calcium phosphate. Lime (calcium hydroxide, Ca(OH)2) supplies Ca²⁺ and elevates pH (a measure of acidity/alkalinity), driving the reaction 3Ca(OH)2 + 2 PO4³⁻ → Ca3(PO4)2(s) + 6 OH⁻ (mdpi.com). Studies show that raising pH to around 9.5 with lime—about 150–200 mg as CaCO3 per liter—typically precipitates ≈80% of phosphate (for example, cutting 20 mg/L influent to ~4 mg/L P) (nepis.epa.gov).

Going higher can push performance further. At pH ≥10.5, removal often exceeds 90% (frequently leaving residuals below 1 mg/L P), albeit with heavier chemical use and highly alkaline sludge (nepis.epa.gov). Typical lime dosing to reach pH ≈9.5 is 150–200 mg/L as CaCO3—roughly 100–130 mg/L Ca(OH)2 (nepis.epa.gov). Metering lime accurately is normally done with a dosing pump, and solids are settled in a clarifier.

In practice, >80–90% removal often needs sequential settling or polymers. Experiments found simple lime dosing without flocculation gave ~80% removal (nepis.epa.gov), whereas >90% required pH >10 or additional flocculants (nepis.epa.gov). Plants commonly add flocculants to improve settling.

Ferric and alum coagulant polishing

Ferric chloride (FeCl3) and aluminum sulfate (Al2(SO4)3) also precipitate phosphate as FePO4 or AlPO4 and can achieve >90–95% removal down to sub‑mg/L levels (mdpi.com). This chemistry often works near neutral pH and is widely used to meet ~0.1–1.0 mg/L P permits (mdpi.com). The trade‑off is sludge: organic sludge weight can increase by ~40–100% (mdpi.com). As Figure 23 shows, Ca(OH)2 precipitation is most effective at high pH (8–12) (mdpi.com), whereas Fe/Al salts work via charge‑driven flocculation. Many facilities use lime first, then a small ferric dose for polish. Where chemical purchasing matters, specifying the right coagulants can tighten compliance margins.

Performance example: in a typical CCGT blowdown (moderate alkalinity), lab work showed lime dosing to pH 9.5 (200 mg/L as CaCO3) reduced total P by >80% to under 2 mg/L (nepis.epa.gov). If controls require <0.5 mg/L, that same plant would need higher pH or ferric polishing. A design check is simple: if influent P = 10 mg/L, 80% removal leaves ~2 mg/L; meeting 0.5 mg/L typically means pushing pH >10 (by ~80–100 mg more lime per liter) or adding ~5–10 mg/L FeCl3.

Zinc precipitation by pH adjustment

Zinc is commonly removed as zinc hydroxide via pH control. Solubility of Zn(OH)2 falls rapidly above pH ~8.5, so raising pH to ~9–10 with caustic (e.g., NaOH) or lime can precipitate Zn(OH)2. The core reaction is Zn²⁺ + 2 OH⁻ → Zn(OH)2(s), with a theoretical solubility product Ksp ≈ 4×10^−17, so thermodynamically most Zn precipitates by pH ~10. Bench data back the practice: adjusting a ZnO‑industry wastewater to pH 9.5 reduced Zn from 79 mg/L to ~3.7 mg/L—95.3% removal (researchgate.net).

Lower pH values (8–9) only partially remove Zn; pH 9–10 is optimal. In real cooling systems with lower influent (say 5–10 mg/L Zn), that performance implies final Zn below ~0.5 mg/L. Achieving pH ≈9.5 in moderately buffered water may require 100–200 mg/L NaOH (or equivalent lime) addition; downstream neutralization (with CO2 or acid) often returns effluent to a discharge pH of 6–9. After hydroxide precipitation, solids are separated by clarification/filtration—in many plants, a clarifier handles the bulk solids.

Raising pH to this range can co‑precipitate other metals (e.g., Cu, Ni) and remove some co‑occurring phosphate as calcium phosphate (mdpi.com). Sulfide addition (Na2S) can precipitate ZnS even at neutral pH, but sulfides are hazardous and generate toxic sludge—generally reserved for very low residual goals in niche cases.

Alternative technologies and reuse routes

Beyond precipitation, ion exchange and adsorption can polish P and Zn to very low levels. Strong‑base anion or chelating resins are common, and zirconium/lanthanum‑based adsorbents remove P to <0.1 mg/L in drinking‑water studies. Natural zeolites or activated carbon also adsorb Zn. Given capacity and regeneration costs, these approaches fit best at low flows or as a final step. For resin‑based polish, packaged ion‑exchange systems are often integrated near discharge; for Zn adsorption tasks, activated carbon is a familiar option.

Membrane filtration—nanofiltration (NF) or reverse osmosis (RO)—can strip nearly all dissolved species, including P and metals. Cooling‑water reuse studies show RO concentrating blowdown to near‑seawater salinity, with permeate meeting strict limits. Downsides include higher CAPEX/OPEX, pretreatment needs to avoid scaling, and a brine stream. Utilities that go this route commonly evaluate integrated RO/NF/UF membrane systems sized for industrial duty.

Passive systems such as constructed wetlands or biofixation can sequester P and Zn via biological uptake or mineral formation; some experiments report 50–90% P removal in slag‑based wetland filters. These installations require land and generally suit municipal/industrial effluents more than power‑plant blowdown.

Volume reduction is another lever. Raising cycles‑of‑concentration (COC, the ratio of dissolved solids in circulating water to makeup) from 3 to 6 halves blowdown flow (and pollutant load) (power-eng.com). This does not remove contaminants, but it slashes the volume to be treated and can pair with evaporation/crystallization toward partial or full Zero Liquid Discharge (ZLD). One study even advocates ZLD design for future plants, given tightening limits (power-eng.com).

Design implications and compliance math

The practical playbook at most CCGT facilities is chemical precipitation plus clarification, because it is simple and effective. A plant aiming for ~1 mg/L P can precipitate ~80–90% with lime, then polish with a small alum/ferric dose. Zn‑rich blowdown is typically treated with caustic to pH ~9–10. Bench work—jar tests or pilots—tunes dosages and sizing of reactors and settlers.

Two data points guide design: “95.3% Zn removal at pH 9.5” (from 79 mg/L to ~3.7 mg/L; researchgate.net) and “~80% phosphate removal with 200 mg/L as CaCO3” (nepis.epa.gov). From there, it is arithmetic: dosing ~150 mg/L lime to blowdown with 10 mg/L P would cut it to ~2 mg/L; hitting 0.5 mg/L P often means either 80–100 mg more lime per liter to push pH >10 or ~5–10 mg/L FeCl3. Solids load then drives clarifier design, with polymers and clarifier detention time tuned for capture.

Sources for these figures span power‑industry handbooks and environmental studies (power-eng.com; power-eng.com), lime‑phosphate precipitation benchmarks (nepis.epa.gov; mdpi.com), coagulation performance and sludge impacts (mdpi.com; mdpi.com), typical cooling‑system concentrations (chemengonline.com), and zinc hydroxide trials (researchgate.net).