RO Configuration for Makeup RO Reject: 2026 Reuse & Discharge Guide

Why Makeup RO Reject Needs Its Own Treatment Train

A makeup-water RO running at 75–85% recovery still discharges 15–25% of its feed as concentrate, and that reject stream carries 4,000–10,000 mg/L of total dissolved solids, elevated silica, hardness ions, sulfate, and any organic carryover from upstream pretreatment. Sending it to a sewer tie-in, cooling-tower blowdown line, or evaporation pond is rarely free: a single industrial RO reuse case documented 82,000 EUR in annual water and discharge savings once the reject was treated and returned to the plant (S1, MDPI 2021). The chemistry also matters — a combined softening + RO train on FGD wastewater achieved nearly complete salt rejection and a water reclamation ratio above 92% (S2, MDPI Environments 2018-06), which sets the realistic upper bound for what a properly designed second-stage train can deliver on industrial reject.

Standard BWRO elements cannot handle reject above roughly 10,000 mg/L TDS without flux collapse and accelerated scaling; above 15,000 mg/L, seawater elements become mandatory. This is why reject cannot be fed back into the first-pass RO bank at higher recovery — the osmotic ceiling and the silica/CaSO₄ scaling envelope are already exhausted at the tail of the first pass. A dedicated second-stage train, sized to the reject's actual chemistry, is the only way to push overall system recovery to 90–95% and cut concentrate volume by 50–75% before discharge or further concentration.

An industrial RO system sized for this reject stream is the first capital decision, but the configuration choice — single-pass boost, two-pass, or brine-concentrating RO — is what determines whether the reject becomes reuse-grade water or a thinner ZLD feed.

Reuse or Discharge: How the End Use Drives RO Configuration

The end use of both permeate and concentrate controls the configuration, not the other way around. A reuse path sends the second-pass permeate back as boiler makeup (target <50 mg/L TDS) or cooling-tower makeup; the concentrate stream is minimized but not eliminated and still requires a disposal route. A discharge path optimizes for concentrate volume and residual TDS to meet the local effluent limit, and permeate quality is secondary. A zero liquid discharge path is the most demanding: the RO is only the front end of a thermal train, and the configuration must deliver a concentrate at 50,000–80,000 mg/L TDS to the evaporator or crystallizer, which means high feed pressure, feed-and-bleed staging, or a brine-concentrating RO at 30–40 bar.

S2 illustrates how recovery interacts with downstream volume. Their total-recycle RO experiment ran at TMP 18 bar on an ESPA 2540 element and TMP 26 bar on an SWC-2540 element, both at 20 °C and a recovery factor of 50%, with a corresponding weight reduction factor (WRF) of 2 — meaning 50% of the feed water was forced through the membrane as permeate (S2, MDPI Environments 2018-06). On a reject stream, every additional 10% of recovery cuts the concentrate volume sent to thermal by roughly the same factor, which is why BCRO is favored for ZLD even though its specific energy consumption is two to three times higher than a brackish second pass. The decision is not which configuration is "best" — it is which trade-off the site can accept.

The Three RO Configurations That Actually Work on Reject

Engineers evaluating bids and P&IDs for reject treatment will encounter three distinct configurations, and each maps to a different operating envelope.

Configuration 1 — Single-pass high-pressure RO on reject: a brackish element bank operating at 15–25 bar, fed directly from the first-pass reject manifold. This is the lowest-CAPEX option and raises first-pass system recovery to roughly 85–90%. It is appropriate when the site has a sewer permit, an evaporation pond, or a downstream brine line that can accept 4,000–8,000 mg/L concentrate at low volume. It is not a reuse configuration because the permeate will typically land at 100–300 mg/L TDS — acceptable for cooling-tower makeup but borderline for boiler feed without polishing.

Configuration 2 — Two-pass RO with inter-stage conditioning: the standard reuse path. The first pass of the second stage operates at TMP 18–26 bar, as documented on the ESPA 2540 and SWC 2540 elements in S2, and the permeate from this stage is either returned directly to the process or polished by a second RO pass to drop TDS below 50 mg/L for high-pressure boiler makeup. This is the configuration that earns the 92% reclamation numbers cited in the FGD study and the 82,000 EUR annual savings case in S1.

Configuration 3 — Brine-concentrating RO (BCRO) at 30–40 bar in a feed-and-bleed or total-recycle arrangement. S2's UF pretreatment was run at a recovery factor of 97.5% in feed-and-bleed as the operational analogue for what BCRO staging must achieve. BCRO pushes system recovery past 95% and concentrate past 50,000 mg/L TDS, which is the right feed for a mechanical vapor recompression (MVR) crystallizer. Above 15,000 mg/L TDS on the BCRO feed, standard brackish elements must be replaced with seawater (SWC-class) elements rated for high osmotic pressure. Selecting the wrong element class on a BCRO stage is the single most common cause of premature membrane replacement in ZLD projects.

A standard industrial RO system paired with a multi-media filter for silica and turbidity reduction is the typical skid specification for Configurations 1 and 2; BCRO skids are usually quoted as a separate, higher-pressure package with seawater elements.

Pretreatment Train: What the Second RO Demands Upstream

Reject pretreatment cannot be a copy of the well-water pretreatment. The reject stream is already concentrated, so the scaling envelope closes faster and any residual organic carryover from the first-pass pretreatment is amplified. Minimum pretreatment is 5 µm cartridge filtration; a multi-media filter is added when silica scaling risk or turbidity spikes are present. Antiscalant dosing must be re-formulated for the elevated Ca²⁺, Ba²⁺, and silica, and the dose rate is typically 20–40% higher than the first-pass dose because the reject LSI and Stiff-Davis stability index trend higher.

Softening chemistry is the lever that determines whether a high-recovery stage runs for 12 months or 12 days between cleanings. S2 showed that the proper Na₂CO₃ addition was determined with respect to the Ca²⁺ ion concentration in the wastewater being treated, and that an under-dosed softening stage collapses RO performance within hours of startup. CIP regimen matters as much as the antiscalant. The S1 case study on oleochemical wastewater reuse found that NaOH-based cleaning (pH 11–12, 35–40 °C) outperformed HCl for flux recovery on organic-loaded streams, and this finding transfers directly to reject streams carrying humic, fulvic, and antifoam residues from the first pass. An automatic chemical dosing system tied to the reject feed conductivity is the most reliable way to keep the antiscalant dose in band as the first-pass recovery drifts day to day.

Sizing Example: 100 m³/h Makeup RO With 75% Recovery

The fastest way to evaluate a configuration is to run the mass balance. Assume a makeup RO fed with 100 m³/h of well water at 1,000 mg/L TDS running at 75% recovery: permeate is 75 m³/h at <50 mg/L, and reject is 25 m³/h at approximately 4,000 mg/L TDS. Send the 25 m³/h reject to a two-pass RO operating at 80% recovery: 20 m³/h returns as reuse-grade permeate (<100 mg/L TDS) and 5 m³/h of concentrate at roughly 20,000 mg/L TDS leaves the second stage. Send those 5 m³/h to a BCRO operating at 60% recovery: 2 m³/h leaves as crystallizer feed at approximately 50,000 mg/L TDS and 3 m³/h is recycled to the BCRO feed tank.

Net result: total system recovery is 97%, total concentrate volume is 2% of the original feed, and the 12.5× reduction in liquid waste is what changes the disposal cost line. A site that was paying to haul 25 m³/h of reject to a brine disposal well is now paying to haul 2 m³/h to an on-site crystallizer or off-site固化 facility. The same mass balance applied to a single-pass boost instead of two-pass + BCRO yields roughly 90–92% total recovery and 8–10 m³/h of concentrate — a 2.5–5× improvement, but well short of what ZLD economics require.

Operating Pressures, Recovery Limits, and Common Failure Modes

Set expectations on energy and recovery ceiling before signing a PO. A two-pass RO on reject runs at 15–25 bar feed pressure and should be capped at 80–85% recovery per stage to stay below silica and CaSO₄ scaling limits — pushing past 85% on a single stage typically halves the CIP interval. BCRO runs at 30–40 bar and is capped at 50–60% recovery per stage because the osmotic pressure of the concentrate climbs past 50 bar at 60,000 mg/L, and the net driving pressure across the membrane collapses to a few bar even at full feed pressure.

Two failure modes account for the majority of unplanned cleanings on reject service. First, silica scaling on the tail element of the second pass or the lead element of BCRO: this is controlled by keeping reject temperature below 35 °C, maintaining pH in the 6.8–7.2 band to minimize silica polymerization, and selecting a silica-specific antiscalant. Second, organic fouling from upstream biology, antifoam, or pretreatment polymer carryover: this is controlled by NaOH-first CIP at pH 11–12 and 35–40 °C (S1, MDPI 2021) and by keeping the SDI on the reject feed below 3, which is a stricter target than the <5 SDI typical of first-pass service. For pretreatment sizing and media selection, the multi-media filter selection guide walks through SDI reduction targets and backwash rates specific to reject service. For ammonia-driven ZLD projects where the reject is also high in NH₃-N, the ammonia-nitrogen ZLD treatment guide covers stripping and breakpoint chlorination decisions that affect the RO mass balance upstream of the second pass.

Frequently Asked Questions

What RO configuration should treat RO reject for reuse or discharge?

Two-pass RO for reuse or compliant discharge, brine-concentrating RO (BCRO) added as a third stage when the site is pursuing zero liquid discharge. A single-pass high-pressure RO is acceptable only when the site has a low-cost disposal route and does not need reuse-grade permeate.

What overall recovery can a second-stage RO train achieve on makeup reject?

Two-pass systems on reject typically achieve 90–95% overall recovery. Adding BCRO polishing raises total system recovery to 95–97%, with 2–5% of the original feed leaving as concentrated brine to the thermal or固化 step.

What feed pressure does a reject-treatment RO require?

Two-pass brackish RO on reject runs at 15–25 bar per stage. BCRO runs at 30–40 bar, which is why it is sold as a separate, higher-pressure skid with seawater elements rather than as an extension of the first-pass RO.

What pretreatment is required ahead of the second RO?

Multi-media filter for turbidity and bulk silica reduction, 5 µm cartridge polishing, and an antiscalant tailored to the reject's Ca²⁺, Ba²⁺, and silica concentrations. CIP is NaOH-first at pH 11–12 and 35–40 °C, followed by acid; NaOH was shown to outperform HCl for flux recovery on organic-loaded reject (S1, MDPI 2021).

What membrane type is used in the second pass vs BCRO?

High-rejection brackish elements (e.g., ESPA-class) for the second pass up to about 10,000 mg/L TDS feed. Above 15,000 mg/L TDS — which is normal on the BCRO feed — seawater elements (SWC-class) are required because standard BWRO elements cannot sustain the transmembrane pressure without irreversible compaction.