Economic evaluation of ion-exchange processes for nutrient removal and recovery from municipal wastewater

Excess nutrients discharged from WWTPs cause eutrophication, prompting stringent discharge limits (e.g., <1 mg NH₄-N/L and as low as 0.5 mg PO₄-P/L). While nutrients, especially phosphorus, are valuable and should be recovered, conventional biological nutrient removal (BNR) processes dominate but can struggle to reliably achieve very low nutrient limits and face issues such as temperature sensitivity and carbon dependence. Ion exchange (IEX) has historically seen limited use due to media selectivity, bed clogging, and regeneration costs, but advances such as synthetic zeolites (mesolite) for ammonium and hybrid anion exchange resins (HAIX) for phosphate have improved performance. This study evaluates the performance and economics of scaling IEX as tertiary treatment for ammonia and phosphorus removal and recovery after either conventional activated sludge (ASP) or an anaerobic membrane bioreactor (AnMBR), comparing both with a traditional BNR + iron dosing flowsheet for a 10,000 PE municipal WWTP.

Recent work shows mesolite (synthetic zeolite) has higher NH₄-N capacity (≈4.6–4.9 meq/g) than clinoptilolite (~2.0 meq/g), linked to Si:Al ratio and high cation exchange capacity. Optimal EBCT of ~10 min balances performance and cost, with regeneration at 300–400 bed volumes (BV) for influent 10–20 mg NH₄-N/L and up to 1000 BV to a 1 mg/L breakthrough depending on influent conditions. Potassium chloride (KCl) regenerant outperforms sodium chloride (NaCl), remaining effective down to 0.1 M, with 5 BV KCl (10%) at pH ~12 effective for regeneration. For phosphorus, HAIX resins (with ferric oxide nanoparticles) can achieve very low effluent PO₄-P (<0.1 mg/L), with operational capacities ~4.9–6.2 mg P/g at 5 min EBCT (decreasing to 2.5–3.7 mg P/g after multiple runs). Regeneration with 2–4% NaOH at 10 BV is effective and regenerant can be reused ~10 times before clean-up. However, frequent regenerations and hazardous spent regenerant disposal (~£65/ton) historically reduced economic viability. Emerging approaches recover nutrients and enable regenerant reuse: HFMC with sulfuric acid to produce ammonium sulfate from KCl brine, and lime addition to NaOH brine to precipitate hydroxyapatite. Media lifetimes around 600 regeneration cycles have been suggested but need demonstration-scale confirmation.

Three flowsheets for a 10,000 PE WWTP (flow 5400 m³/d; average UK temperature 14°C) were designed to meet effluent COD <20 mg/L, NH₄-N <1 mg/L, and PO₄-P <0.5 mg/L: (1) BNR (A2O) + iron dosing + tertiary filtration; (2) ASP (carbon removal) + IEX (separate N and P columns) with regenerant clean-up and nutrient recovery; (3) AnMBR (UASB + submerged membrane) + IEX with regenerant clean-up and nutrient recovery. BNR and ASP were designed via SRT-based methods. To protect IEX from suspended solids and organics, ASP effluent received drum filtration (loading 4500 L/m²·h). IEX comprised: N column (mesolite) designed with EBCT 10 min, bed volume 37.5 m³, 7 vessels (one spare), bed depth 0.88 m, media capacity 350 BV to 1 mg/L breakthrough; regeneration with 10% KCl (5 BV, EBCT 60 min) at pH ~12 (2% NaOH), cycles 148/year, brine tanks 4×50 m³; spent KCl regenerant cleaned every 3 cycles via hollow fibre membrane contactor (Liqui-Cel, 20 m² area, ~5 m³/h, ΔP 0.41 bar) using H₂SO₄ to recover (NH₄)₂SO₄. P column (HAIX) designed with EBCT 5 min, bed volume 18.75 m³, 5 vessels (one spare), bed depth 0.88 m; regeneration with 4% NaOH (10 BV, EBCT 20 min), regenerant reuse up to 1000 BV; clean-up every 10 cycles by adding hydrated lime to precipitate hydroxyapatite. AnMBR design targeted reduced energy via intermittent dead-end biogas sparging (SGD ~0.5 m³/m²·h, energy ~0.13 kWh/m³ permeate), with degassing membrane to recover dissolved methane. Reported AnMBR performance used: COD removal ~87±1%, methane yield 0.18–0.23 Nm³ CH₄/kg COD removed, flux 10–13 L/m²·h. Economic assessment used equipment cost curves (USGC basis) and unit cost data, applying location factors and CEPCI adjustments, and a Lang factor (4.74) to estimate CAPEX. OPEX included energy, chemicals, sludge disposal, maintenance (2.5% of CAPEX), and labour. Whole-life cost (WLC) assumed 40-year life and 7% discount rate, approximated as Initial CAPEX + 14×Annual OPEX.

- All flowsheets (BNR+iron, ASP+IEX, AnMBR+IEX) met effluent targets: COD <20 mg/L, NH₄-N <1 mg/L, PO₄-P <0.5 mg/L. - Stability: BNR was sensitive to low temperature, shock loads, and carbon limitation, struggling to consistently achieve >95% nutrient removal; ferric dosing (~6438 kg/year, ~2–3 Fe:P ratio) and tertiary filtration were required. IEX processes were less sensitive to temperature and diurnal fluctuations due to controllable regeneration triggers. - ASP+IEX: ASP removed ~90% BOD/COD; HAIX provided additional 40–50% COD removal; energy dominated by aeration (53%) and drum filter (38). Nutrient recovery: HFMC converted spent KCl brine NH₄-N to ~98 t/year ammonium sulfate; lime precipitation from NaOH brine produced ~3.4 t/year hydroxyapatite. - AnMBR+IEX: AnMBR achieved ~87% COD removal; remaining COD reduction supported by HAIX; biogas production offset energy, yielding ~0.12 kWh/m³ net energy production (before other process consumption); intermittent sparging lowered energy demand. - IEX considerations: >85% nutrient recovery achievable; reduced sensitivity to temperature and shock loads; performance can be reduced by high suspended solids and competing ions (e.g., SO₄²⁻, Ca²⁺, NO₃⁻), necessitating pre-filtration. SO₄ at ~50 mg/L had limited impact on HAIX capacity; very high SO₄ may affect AnMBR via sulfate-reducing bacteria but can be mitigated. - Economics: • BNR+iron: CAPEX £3.94 M; OPEX £316k/year; WLC £8.4 M. OPEX: maintenance 31%, sludge 25%, labour 17%, energy 17%. • ASP+IEX: CAPEX £3.48 M; OPEX £282k/year (ASP ops 79%); WLC £7.4 M, ~17% lower than BNR. • AnMBR+IEX: CAPEX £3.6 M; OPEX £177k/year; WLC £6.1 M, ~27% lower than BNR. • Regenerant clean-up and reuse substantially reduced OPEX: for N IEX, regenerant reuse cut OPEX by ~50%, and nutrient recovery reduced it a further 15–65% relative to single-use regenerant. - Stringent limits: BNR OPEX increases sharply to achieve extremely low P (e.g., doubling from <1 mg P/L to 0.1 mg P/L), while IEX OPEX is less sensitive. - Products and markets: Recovered products’ potential revenue varies widely—if sold as fertilizers, ~£12,834/year; as reagents, up to ~£450,285/year (though entry to food/pharma markets is currently impractical).

Findings indicate IEX-based nutrient capture and recovery, particularly when paired with AnMBR, can reliably achieve stringent nutrient limits while reducing costs and greenhouse gas emissions. IEX transfers nutrients to regenerant brines, enabling extraction as marketable products—ammonium sulfate via HFMC and hydroxyapatite via lime precipitation—supporting circular economy goals. Although nitrogen recovery currently cannot economically compete with Haber–Bosch synthesis on a per-kg basis, it offers environmental benefits due to the high energy and CO₂ footprint of industrial ammonia. Phosphorus recovery is strategically important given finite phosphate rock reserves and market volatility. Product quality and purity are crucial for marketability; for fertilizers, specifications include high solid content, controlled particle size, absence of pathogens/heavy metals, and reliable nutrient contents/release rates. Sales channels range from direct to end users to distribution via wholesalers/brokers/dealers. The coupling of AnMBR with IEX provides additional sustainability gains: energy neutrality or positivity through biogas, reduced direct GHG emissions, and solids-free effluent that enhances IEX stability. Overall, the results support IEX as a robust, lower-cost, and lower-emission alternative or complement to BNR for meeting tight nutrient limits, especially when regenerant is cleaned and reused and nutrients are recovered.

Ion-exchange using mesolite (for NH₄-N) and HAIX (for PO₄-P) effectively removes nutrients to very low levels and enables recovery as ammonium sulfate (~98 t/year) and hydroxyapatite (~3.4 t/year) with regenerant reuse. Compared with a traditional BNR + iron dosing plant, ASP+IEX and especially AnMBR+IEX achieved lower WLC (by ~17% and ~27%, respectively), with reduced energy consumption and greenhouse gas emissions, aligning with circular economy objectives. IEX processes are less sensitive to temperature and shock loads than biological processes and can be deployed at small-to-medium WWTPs; for large plants, modular column designs allow practical operation. Future work should validate long-term media lifetimes, optimize regenerant reuse/cleanup cycles, and develop stable market pathways for recovered products to enhance economic viability.

- Economic and performance evaluations rely on design assumptions and pilot-scale data; full-scale demonstration is needed, including verification of media lifetimes (e.g., ~600 regeneration cycles) and long-term performance. - IEX performance can be reduced by high suspended solids and competing ions, necessitating pretreatment (e.g., filtration), which adds complexity and cost. - BNR performance sensitivity to temperature and influent variability may differ across climates; results assume UK-average 14°C conditions. - Market and regulatory uncertainties for recovered products (especially entry into high-value reagent/health markets) limit revenue realization; fertilizer markets yield modest revenues. - Nitrogen recovery economics remain less favorable than industrial synthesis; broader energy and carbon externalities are not priced. - Costing uses generalized cost curves, location factors, and a simplified WLC factor (14×OPEX), introducing uncertainty. - Potential impacts of sulfate management (e.g., on AnMBR via sulfate reducers) and iron dosing trade-offs require site-specific optimization.