The power-plant RO playbook: filtration, antiscalants, and cleanings that prevent costly shutdowns

High‑purity boiler and cooling water in power plants commonly pairs RO (reverse osmosis) with polishing via ion exchange. In practice, that means RO plus ion exchange. Because RO membranes are extremely fouling‑sensitive, robust pretreatment—multiple filtration stages to remove particulates and colloids, chemical conditioning (antiscalants, pH adjustment), and disinfection (biocides, chlorination/dechlorination)—is mandatory. Done right, these steps prevent scale and biofilm on membranes and can extend membrane life from the typical ~2–4 years up toward a decade (www.intechopen.com).

The payback is visible in the field. One RO retrofit in a steam plant showed that proper pretreatment let the membranes run ~12–18 months between major cleanings (shanghai-cm.com). And scale isn’t the only threat: worldwide, biofouling alone is estimated to cost ~$15 billion/year (www.sciencedirect.com). Rigorous pretreatment thus has tangible payback in reduced downtime, energy and regeneration costs.

Filtration and particulate control

First principles: remove all visible solids and colloids before RO. Conventional trains often deploy a clarifier followed by multimedia filters, typically sand and anthracite. A well‑designed clarifier‑filter can routinely produce effluent turbidity ≤0.5 NTU (nephelometric turbidity units) (www.watertechonline.com).

Media beds using sand/silica and, in a multi‑layer configuration, anthracite handle the heavy lifting, while modern lines polish with submicron barriers, e.g., 5–10 μm cartridge filters. Two critical metrics at the RO feed are turbidity and the SDI15 (silt density index measured with a 0.45 μm membrane for 15 minutes). The typical target is SDI15 <5 (ideally <3), which roughly correlates to turbidity <0.1–0.2 NTU (www.watertechonline.com), and the SDI test is a standard quality‑control tool (www.watertechonline.com). If SDI exceeds spec, additional pretreatment is required.

Membrane pretreatment raises the bar. Plants replacing a clarifier/media pretreatment with microfiltration (MF) serving a supercritical steam generator reported several months of operation with no maintenance other than a quarterly offline chemical clean (www.watertechonline.com). Full‑scale trials using UF (ultrafiltration) in Saudi and Gulf plants produced permeate with SDI ≈2.2 (www.researchgate.net), and in 6‑month trials no RO cleaning was required at all (www.researchgate.net) (www.researchgate.net). Adopting ultrafiltration as pretreatment delivers consistent low turbidity/SDI and supports high RO recovery (up to ~95%) (www.sciencedirect.com) (www.watertechonline.com). It also intercepts microscopic colloids (0.01–1 μm) that can pass a 0.45 μm SDI test yet foul RO; UF removes down to ~0.01 μm (www.moruiwater.com).

Design targets are straightforward: remove all suspended solids so the RO lead elements see essentially “clear” water; influent turbidity <0.5 NTU, SDI15 <3–5 (www.watertechonline.com) (www.watertechonline.com). Plants often place 5–10 μm cartridge filters immediately before the RO pressure vessels for final polishing, with continuous monitoring of SDI and turbidity. One operating rule of thumb: “no particles visible, SDI<3, chlorine free” at the RO inlet.

Antiscalants and hardness control

RO concentrates salts 2–10×, quickly pushing even moderate hardness or silica past solubility. Common scale‑forming salts include CaCO₃, CaSO₄, BaSO₄, CaF₂, and iron/metal oxides, all known to deposit on the membrane surface and block flow (www.intechopen.com). The first line of defense is continuous antiscalant dosing. Modern antiscalants (typically phosphonate/polymer blends) act at low dose—e.g., 1–5 mg/L—to prevent crystal nucleation and agglomeration via threshold inhibition (www.intechopen.com) (www.moruiwater.com). Plants often meter these with a dedicated dosing pump just ahead of the final filters, and program selection falls under membrane antiscalants.

The effect is material. Designs relying on antiscalant routinely tolerate higher saturation ratios than acidification alone. One design table shows CaCO₃ inhibition allowing Langelier Saturation Index (LSI) ≈+3.3, versus ≈+2.5 with only acidification (www.intechopen.com). Antiscalants can also hold sulfates and fluorides at tens to hundreds of times saturation (www.intechopen.com). In practice, proper dosing can often double productivity by enabling higher recoveries (70–80%+).

If raw water hardness is extreme (hundreds of mg/L as CaCO₃) or dominated by non‑carbonate hardness, a softening step may pre‑remove Ca²⁺/Mg²⁺ before RO; operators commonly deploy a softener. When iron or manganese are present, pre‑oxidation (Cl₂ or KMnO₄) plus filtration is advised (shock doses of oxidant to precipitate metals, then media filters). Many plants ensure Fe/Mn <0.1 ppm by aeration or greensand filters; a dedicated media such as manganese greensand is typical for this duty. Without these steps, antiscalants alone may not stop Fe‑oxide colloids from fouling the membrane surface.

Silica is a special case. Dissolved silica (SiO₂) can form colloidal gels; specialized silica inhibitors (a type of antiscalant) or diatomaceous–ultrafiltration pretreatment are used if silica >50 mg/L. Some designs limit silica to <25–30 mg/L in feed when targeting high recovery; above that, silica fouling can necessitate lower recoveries or frequent cleaning.

On dosage, municipal brackish RO plants often feed 1–3 mg/L of active antiscalant to handle ~200–300 mg/L hardness at 75% recovery, an application space that aligns with brackish‑water RO. A pragmatic goal often cited: no scale spot visible after a 500+ hour RO run.

Biological disinfection and dechlorination

Any residual organics or bacteria can rapidly form biofilms on RO. Bacteria can double every 20–30 minutes under warm conditions (www.intechopen.com), secreting sticky polysaccharides that bind particulates and block flow (www.intechopen.com). Even a few viable cells can colonize the RO’s large surface area, so upstream disinfection is standard.

Chlorination is common at 0.5–2 mg/L NaOCl to kill plankton/algae, but standard polyamide membranes cannot tolerate free chlorine; any chlorination step must be followed by full dechlorination. Plants use sodium bisulfite (NaHSO₃/metabisulfite) or activated carbon to reduce Cl₂ to effectively zero (e.g., <0.1 mg/L) by the RO inlet (hydrochemie.co.uk). In practice, 3× the stoichiometric SO₃ dose is often applied to ensure complete quench. These dechlorination tasks map to dechlorination agents or to carbon beds such as activated carbon.

Many facilities opt for “shock” treatment rather than continuous disinfection: switching to free chlorine for 6–8 hours each week (to ~1 mg/L residual), then dechlorinating afterward. Studies report >99% kill by such weekly shocks, with minimal membrane exposure (link.springer.com). High‑dose treatments with other oxidizers (ozone, peracetic acid) or sanitizers (500 ppm H₂O₂ or glutaraldehyde soak for 30 minutes) are also used periodically. The key is avoiding a continuous chlorine residual in the RO feed.

Non‑oxidizing, RO‑compatible biocides—such as DBNPA (2,2‑dibromo‑3‑nitrilopropionamide) or isothiazolinones—are dosed either inline or batch at a few ppm and can be used on‑line (in non‑potable systems) or off‑line to eliminate and control microbiological fouling (hydrochemie.co.uk). Plants slot these under biocides. When fouling is severe, operators may add a circulated soak of peracetic acid or sodium hypochlorite as a final cleaning step, with immediate bisulfite neutralization (shanghai-cm.com).

Effective biological pretreatment keeps bacterial counts extremely low (e.g., <200 CFU/ml) entering the RO. Routine microbiological monitoring or ATP assays should confirm near‑zero active biomass. The ROI is clear: even a thin biofilm can double membrane ΔP in days; controlling it saves frequent cleaning. In practice, careful application of biocides and dechlorination has enabled many plants to run months between RO cleanings (e.g., the UF‑RO trial above ran 6 months with no biofilm issues: www.researchgate.net) (www.researchgate.net).

RO membrane cleaning (CIP) steps

Despite best‑in‑class pretreatment, some fouling accumulates. Clean‑in‑Place (CIP) is standard. Typical triggers: a normalized permeate flow drop of ~10–25% or a feed‑brine ΔP rise of ~15–50% over baseline (www.wwdmag.com) (shanghai-cm.com). Economic analyses (e.g., Orange County’s groundwater reuse RO) suggest optimal CIP intervals for similar trains on the order of 5–10 months (www.wwdmag.com).

Operators typically follow a standard sequence, using proprietary blends under membrane cleaners as needed:

• Pre‑flush: Isolate permeate outlet and pin down concentrate. Flush with RO permeate (at ~50% design flow) for ~30 minutes to wash out loose particulates (shanghai-cm.com).
• Alkaline clean (organics/bio removal): Circulate alkaline cleaning solution (~0.1–0.5% caustic with a chelating agent) through all elements for ~60–90 minutes; heat to ~35–40°C if possible; maintain low transmembrane pressure (TMP) <0.2 MPa; then soak for ~4–6 hours at ≥30°C (shanghai-cm.com) (shanghai-cm.com) (shanghai-cm.com).
• Intermediate rinse: Drain and rinse with permeate until effluent turbidity is low (e.g., NTU <5) (shanghai-cm.com).
• Acid clean (scale removal): Circulate acidic cleaner (e.g., 1–2% citric acid or dilute HCl, often with a corrosion inhibitor) for ~45–60 minutes; keep pH ~2.5–3.5 and temperature 25–30°C; stop when conductivity change is <5%/hr (shanghai-cm.com) (shanghai-cm.com).
• Disinfection (if needed): For severe biofouling, add an oxidizing flush or soak (e.g., 0.1% peracetic acid circulated 30 minutes, or 1% NaHSO₃ soak 2 hours) (shanghai-cm.com).
• Final flush: Alternate forward and reverse flushing with RO permeate until effluent pH is ~6.5–7.5 and conductivity returns to within ~10% of baseline (shanghai-cm.com).

After cleaning, compare performance to pre‑fouling baselines. Industry standards aim for ≥90–95% recovery of original flux and full recovery of salt rejection (shanghai-cm.com). If cleaning recovers <80%, it may indicate irreversible fouling or membrane damage; element replacement should be considered. One guideline targets permeate flow recovery ≥95% and salt rejection improvement ≥3% after CIP (shanghai-cm.com).

Well‑executed CIP extends membrane life and saves cost. A report estimates 8–12% energy savings after a thorough CIP due to lower pump pressure (shanghai-cm.com). Delaying cleaning increases daily energy and risks membrane injury. An economic study of 15 RO trains found that halving the cleaning frequency (to ~5–8 months) roughly halved total annual operating cost compared to very frequent cleaning (www.wwdmag.com).

Key performance metrics

• Feed quality (post‑pretreatment): Turbidity <0.1–0.5 NTU, SDI15 <3–5 (www.watertechonline.com) (www.watertechonline.com), and essentially no visible particles. All biologically active chlorine must be removed (e.g., ≤0.1 mg/L free Cl₂) before RO (hydrochemie.co.uk). These are the conditions systems under membrane systems are designed to receive.
• Antiscalant dosing: A few mg/L active often suffices for mg/L‑range hardness and silica; design saturation ratios—e.g., CaCO₃ LSI up to +3—should be validated by lab tests (www.intechopen.com). Coagulant (FeCl₃/Al₂(SO₄)₃) is often added ahead of media filters when raw water contains colloids or iron.
• Biocide practice: If chlorine is used, apply 100% removal agents (bisulfite) before RO (hydrochemie.co.uk). Shock biocide cycles (weekly chlorination or peroxide) are recommended for SWRO contexts (link.springer.com), paralleling practices in sea‑water RO. Non‑oxidizing agents (DBNPA, isothiazolinone) can be fed continuously at a few ppm (hydrochemie.co.uk).
• Cleaning schedule: CIP at a normalized flux drop of 10–25% or ΔP rise of ~15–50% (www.wwdmag.com). Many plants clean every 3–9 months; one case saw an economic optimum ~5–8 months (www.wwdmag.com).
• CIP effectiveness: Well‑executed CIP should restore >90% of initial flux (shanghai-cm.com). Failure to achieve ~80–90% indicates fouling not fully removed or irreversible damage.

By implementing this pretreatment guide—effective solids filtration, proper scaling and biofouling inhibitors, and systematic CIP—a power plant can keep its RO membranes operating at or near design capacity. Quantifiable outcomes include longer run times between cleanings (e.g., running many months instead of weeks), fewer membrane replacements, and measurable energy savings on the order of 8–12% (shanghai-cm.com).

Sources: Peer‑reviewed and industry references underpin the above guidelines. Key sources include Abrahamse et al. Desalination (2015) and Ning IntechOpen (2012) on fouling mechanisms and pretreatment, Shariff & Basri Applied Water Sci. (2017) on biofouling, Owens & Patel W&WD (2018) on cleaning economics, and technical industry guidelines from ChemTreat/HydroChemie and Shanghai CM (2025). Full citations: Abrahamse A. J., Reungoat J., Bastviken D., Joss A., Comparison of pre‑treatment technologies towards improving RO desalination of cooling tower blowdown, Desalination 357 (2015) 140–149. DOI:10.1016/j.desal.2014.11.018 (www.sciencedirect.com) (www.sciencedirect.com); Ning R. Y., Chemistry in the Operation and Maintenance of Reverse Osmosis Systems, In: Advancing Desalination, InTech (2012) (www.intechopen.com) (www.intechopen.com); Goh P., Lau W. J., Othman M. H. D., Membrane fouling in desalination and its mitigation strategies, Desalination 425 (2018) 130–155. DOI:10.1016/j.desal.2017.10.018 (www.sciencedirect.com); Shariff H. F., Basri I. I., Biofouling in RO: phenomena, monitoring, controlling and remediation, Applied Water Science 7 (2017) 2637–2651. DOI:10.1007/s13201-016-0493-1 (link.springer.com) (link.springer.com); Owens E., Patel M., “RO Cleaning Frequency: A Balance of Costs,” Water & Wastes Digest (Apr. 2018) (www.wwdmag.com) (www.wwdmag.com); Buecker B. (ChemTreat), “High‑Purity Water Treatment: Using Membranes to Protect Membranes,” Water Technology, June 21, 2022 (www.watertechonline.com) (www.watertechonline.com); HydroChemie Ltd. (UK), Reverse Osmosis Chemical Treatment (product literature, “Membrane Cleaning Chemicals/Biocides”) (hydrochemie.co.uk) (hydrochemie.co.uk); Shanghai‑China Membrane Tech. Co., “CIP Chemical Cleaning Technical Guidelines for RO Systems” (2025 blog) (shanghai-cm.com) (shanghai-cm.com).