Brackish Water Reverse Osmosis: Design and Operating Considerations

The Growing Importance of Brackish Water Reverse Osmosis

Brackish water reverse osmosis (BWRO) has emerged as a critical technology for addressing freshwater scarcity in inland regions where seawater is inaccessible but brackish aquifers are abundant. An increasing number of BWRO and nanofiltration (NF) systems are being designed and operated on high-salinity brackish feed water sources ranging from 2,000 to 12,000 mg/L total dissolved solids (TDS). As documented in the seminal 2010 AMTA conference paper by Bates, Bartels, and Lai of Hydranautics, increased feed TDS dramatically impacts RO/NF system design across multiple dimensions: hydraulic configuration, feed pressure requirements, permeate quality projections, and energy economics. For purposes of design classification, high-TDS brackish waters — up to approximately 12,000 ppm — can typically be treated with brackish water NF and RO membranes at feed pressures up to 450–600 psi (31–41.4 bar). Beyond this threshold, higher-pressure seawater RO membranes rated up to 1,000–1,200 psi (69–82.7 bar) become necessary to overcome the elevated osmotic pressures generated at the concentrate end of the system.

Fundamental Design Parameters for High-TDS BWRO Systems

The design of a brackish water RO system treating elevated-salinity feed sources must address a series of interdependent parameters that collectively determine capital cost, operating cost, permeate quality, and long-term membrane reliability. Bates et al. (AMTA 2010) define the following essential steps:

• Target Permeate Quality: This is the primary driver of membrane selection. Potable water standards (typically <500 mg/L TDS) permit more flexible membrane choices than industrial process water requiring single-digit conductivity.
• System Capacity: The quantity of feed water to be processed dictates the physical scale of the installation and influences the economic viability of energy recovery devices.
• Percent Recovery: Recovery directly impacts permeate quality, system hydraulics, concentrate TDS, and osmotic pressure. Higher recovery reduces concentrate volume but increases the salinity challenge at the tail elements.
• Feed Water Source Quality: Detailed characterization — including seasonal variations — determines pretreatment requirements and appropriate flux design to mitigate fouling.
• Specific Ion Analysis: Beyond bulk TDS, the concentrations of key scaling ions (Ca²⁺, Ba²⁺, Sr²⁺, SO₄²⁻, SiO₂, F⁻) determine osmotic pressure requirements and antiscalant dosing regimes.
• Temperature Range: The warmest temperature represents the worst-case scenario for permeate salt passage; the coldest temperature dictates the maximum feed pump pressure requirement.
• System Aging Factors: Design projections must account for fouling factors (annual percent flux decline), annual percent salt passage increase, and normal membrane degradation from usage and cleaning cycles.
• Energy Cost and ERD Payback: The operating cost of energy versus the capital cost of an energy recovery device must be evaluated over the projected plant lifetime.

Hybrid RO/NF Systems: Optimizing Flux Balance in High Osmotic Pressure Gradients

One of the most significant design challenges for high-TDS brackish water systems is the hydraulic flux imbalance created by the large osmotic pressure differential from the feed end to the concentrate end of the system. Consider a representative design case from the Hydranautics study: feed TDS of 4,000 ppm at 80% recovery produces a concentrate TDS of 20,000 ppm. The osmotic pressure at the feed end is approximately 46 psi (3.2 bar), rising to 210 psi (14.5 bar) at the concentrate end — a differential of 164 psi (11.3 bar) that must be overcome entirely by the feed pump before any permeate can be produced by the tail elements. This wide osmotic pressure swing makes it challenging to achieve balanced hydraulic flow and flux between stages.

Hybrid RO/NF systems — which deploy different membrane types across successive stages or even within a single stage — have emerged as the preferred solution for these applications. The design engineer can select from several membrane performance tiers:

• Low-Pressure RO (LP-RO): High-flow elements rated at 9,000–12,000 gpd with 99.3–99.6% NaCl rejection at 150 psi/1,500 ppm NaCl standard test conditions (STC). These provide the lowest energy consumption but moderate permeate quality.
• High-Pressure Brackish RO (HP-RO): Elements rated at 11,000 gpd with 99.7% NaCl rejection at 225 psi/1,500 ppm NaCl STC. Offering an intermediate balance of rejection and energy consumption.
• Seawater RO (SW-RO): High-rejection elements rated at 12,000 gpd with 99.8% NaCl rejection at 800 psi/32,000 ppm NaCl STC. When deployed in the first stage of a hybrid BWRO system, these elements convert the energy that would otherwise be wasted in permeate throttling into superior permeate quality.
• Nanofiltration (NF): High-flow, low-rejection elements rated at 8,200 gpd with ~91% NaCl rejection at only 75 psi/1,500 ppm NaCl STC. Used as lag elements in the final stage to boost flux while maintaining acceptable blended permeate quality.

The data from the Hydranautics comparative analysis (Tables 1 and 2, AMTA 2010) is instructive. A conventional all-LP-RO two-stage design (2×1-7M array, 15 gfd system flux) at 80% recovery produced permeate TDS of 108 ppm but required 90 psi of first-stage permeate back-pressure to balance fluxes at 17.5 gfd (stage 1) and 10 gfd (stage 2) — an energy penalty that any serious designer would seek to mitigate. By contrast, a hybrid SW-RO/LP-RO design under identical operating conditions produced permeate TDS of just 75 ppm (44% better) with only 18 psi of permeate back-pressure, at equivalent specific energy consumption of 3.44 kWh/1,000 gallons.

Energy Recovery Devices in Brackish Water Systems

While energy recovery devices (ERDs) are more commonly associated with seawater RO applications, their use in high-TDS brackish water systems is increasingly justified. The Hydranautics analysis demonstrates that adding an interstage turbine-based ERD to the all-LP-RO design reduced energy consumption by 24% — from 3.44 to 2.77 kWh/1,000 gallons — while simultaneously reducing first-stage permeate back-pressure from 90 psi to 35 psi. However, the hybrid SW-RO/LP-RO design with interstage ERD achieved the best permeate quality (70 ppm) with zero permeate back-pressure and 16% energy savings, though at a slightly higher absolute energy consumption of 2.98 kWh/1,000 gallons.

The decision to incorporate an ERD in a brackish water system is fundamentally an economic one. Unlike seawater systems — where ERDs are essentially mandatory for competitive specific energy consumption — brackish water ERDs must demonstrate a suitable return on capital investment. Key considerations include the cost of repiping high-pressure concentrate lines, downsizing the main feed pump to improve operating efficiency, and the fact that ERDs are not readily retrofitted once a system is installed. The industry rule of thumb is that permeate back-pressure throttling is acceptable up to approximately 30 psi; beyond this threshold, the operating cost savings from an ERD or interstage booster pump typically justify the capital expenditure.

The Effect of High TDS on Salt Passage: Donnan Potential and Charge Shielding

One of the most important — and historically underappreciated — phenomena in high-TDS BWRO operation is the increase in salt passage that occurs at elevated feed salinities. Since 1998, Hydranautics researchers have studied and quantified this effect, developing correction factors now incorporated into industry-standard computer design projection programs.

The salt passage correction factor for brackish polyamide membranes reveals a non-linear relationship: at feed TDS between 100 and 2,000 mg/L, no significant correction is required. Below 100 mg/L TDS, salt passage increases due to reduced ionic shielding of the membrane surface charge. Above approximately 2,500 mg/L TDS, salt passage increases again — this time due to charge shielding of the Donnan potential. The Donnan potential is an electrical potential generated at the membrane surface by the repulsion of negatively charged anions away from the negatively charged polyamide membrane, coupled with the attraction and alignment of positively charged cations at the membrane surface. This potential normally enhances salt rejection by inhibiting anion passage. However, as feed TDS increases, the greater abundance of cations increasingly shields the membrane’s negative surface charge, progressively reducing the Donnan rejection potential. The correction factor rises linearly from 1.0 to approximately 1.8 across the TDS range of 2,500–5,300 mg/L, then stabilizes — indicating that the Donnan shielding effect saturates.

This phenomenon has significant practical implications. A design projection that fails to apply the high-TDS salt passage correction factor may underestimate actual permeate TDS by 40–80% at the concentrate end of the system. Importantly, this correction is not applied to seawater elements, which are factory-tested at 32,000 ppm NaCl and therefore already account for the high-salinity effect in their rated performance.

Real-World Case Study: North Miami Beach Norwood Oeffler WTP

The North Miami Beach Norwood Oeffler Water Treatment Plant in Florida provides an instructive validation of hybrid BWRO design methodology. Commissioned in April 2008, this 6-mgd (22,710 m³/day) facility operates three 2-mgd trains at 75% recovery, treating Floridan Aquifer groundwater with feed TDS ranging from 2,900–3,800 ppm at a depth of approximately 425 meters (1,400 feet). The groundwater has the advantageous characteristics of low total organic carbon (TOC), color below 1, and iron below 0.1 ppm.

The system employs a true hybrid configuration: ESPA2 elements (9,000 gpd, 99.6% rejection) in the first stage for higher rejection without permeate throttling, and ESPA1 elements (12,000 gpd, 99.3% rejection) in the second stage for higher flow and energy efficiency. An interstage turbine ERD improves energy performance and aids in flux balancing. After one year of operation (February 2009 data), the actual salt passage of 1.4% exceeded design projections by 20% — meaning the computer-projected salt passage had been conservative and the system performed better than expected. Specific ion rejection data showed good correlation between projected and actual values, with hardness and bicarbonate passage lower than projected.

The Norwood Oeffler plant also demonstrates the versatility of membrane technology: it operates alongside three NF trains treating low-TDS (400 ppm) Biscayne Aquifer water, with the blended product from RO, NF, lime softening, and filtration producing 32 mgd of finished potable water.

Concentrate Management for Inland BWRO Facilities

Unlike coastal seawater RO plants that can discharge concentrate to the ocean, inland BWRO facilities face significant concentrate disposal challenges. At 75–85% recovery — typical for high-TDS brackish applications — the concentrate stream can reach 20,000 ppm TDS or higher. Options include deep well injection, evaporation ponds, zero liquid discharge (ZLD) systems with thermal crystallizers, and — in limited cases — sewer discharge where permitted. The choice of concentrate disposal method often constrains the maximum economically viable recovery, as higher recoveries reduce concentrate volume but increase its salinity — potentially exceeding the capacity of deep well injection zones or requiring prohibitively expensive ZLD infrastructure.

Conclusion: Designing for the Full System Lifecycle

Successful high-TDS BWRO design requires a holistic approach that integrates membrane selection, hydraulic configuration, energy recovery economics, and accurate salt passage modeling. The hybrid RO/NF approach — deploying higher-rejection seawater-grade or high-pressure brackish membranes in the lead stages and higher-flow elements in the tail stages — has emerged as the industry best practice for systems where the osmotic pressure differential between feed and concentrate exceeds approximately 150 psi. When combined with appropriate pretreatment, conservative flux design (typically 12–18 gfd system average depending on fouling potential), interstage energy recovery where economically justified, and accurate salt passage projections incorporating Donnan potential effects, modern BWRO systems routinely achieve permeate quality below 100 ppm TDS at energy consumption under 3.0 kWh/1,000 gallons.

For technical guidance on BWRO system design, membrane selection, and concentrate management strategies, visit https://tiwa.co.id.

Baca juga: Untuk informasi lebih lanjut mengenai solusi pengolahan air bersih dan sistem reverse osmosis, kunjungi layanan instalasi dan perawatan sistem reverse osmosis industri.