SWRO Energy Recovery Explained: How Modern Plants Cut Power Use by 60%

The Energy Problem in SWRO

Seawater reverse osmosis is, at its core, a brute-force solution to a thermodynamic problem. Standard ocean seawater at 35,000 mg/L total dissolved solids exhibits an osmotic pressure of approximately 25 bar (370 psi) at 25 °C. To push freshwater through a polyamide thin-film composite membrane against that osmotic gradient, an SWRO plant must apply 55–85 bar of hydraulic pressure on the feed side — significantly more than the osmotic minimum, because the driving force for permeate flux is the difference between applied pressure and osmotic pressure.

Here is the catch. Only a fraction of the feed becomes permeate. Typical first-pass SWRO recovery is 40–50%, which means more than half of the feed flow exits the membrane housings as a concentrated brine stream — still at roughly 55–82 bar, having lost only the 2–3 bar pressure drop across the membrane array. Without an energy recovery device, that brine stream gets throttled to atmospheric pressure through a control valve, and all of its hydraulic energy is converted to heat and noise. For a 35% recovery plant, that wasted brine carries something on the order of 60–65% of the high-pressure pump's shaft energy out the back door.

This is why first-generation SWRO plants in the 1970s and 80s consumed 8–12 kWh per cubic meter of permeate. The economics simply did not work outside Saudi Arabia and a few water-stressed islands. The story of how SWRO became cost-competitive with surface water treatment in many coastal regions is, in large part, the story of energy recovery devices.

Without an ERD: Where the Energy Goes

Consider a 100 m³/h feed SWRO train at 60 bar membrane pressure and 40% recovery, with no ERD. The high-pressure pump delivers about 200 kW of shaft power (100 m³/h × 60 bar / 36 / pump efficiency 0.82 ~ 203 kW). The 60 m³/h brine exits the array at 58 bar, throttled to zero through the brine control valve. That throttling event dissipates roughly 97 kW — nearly half the pump's input — as low-grade heat in the discharge pipe. The plant produces 40 m³/h of permeate. Specific energy consumption: 200 kW / 40 m³/h = 5.0 kWh/m³, with another 1–2 kWh/m³ on top for intake, pretreatment, and product transfer. Total: 6–7 kWh/m³.

Add a modern ERD that recovers 90+% of that brine energy, and the high-pressure pump no longer has to do the full lift. It only has to make up the pressure differential. The same plant now requires roughly 110 kW for the main HP pump, drops to 2.7–3.0 kWh/m³ total, and the operating cost falls by 40–55%.

ERD Categories: Isobaric vs Centrifugal

All modern SWRO ERDs fall into two architectural families:

1. Isobaric (positive-displacement) pressure exchangers. A rotating or reciprocating chamber alternately fills with low-pressure feed and high-pressure brine. The brine directly pressurizes the feed by displacement, with no intermediate energy conversion. Examples: ERI PX-Q series, Flowserve DWEER, Danfoss iSave.
2. Centrifugal turbochargers. A single-shaft turbine-pump combination where the brine expands through a turbine wheel and the recovered shaft work compresses the feed stream on the same shaft. Examples: FEDCO HPB series, Pump Engineering Turbocharger.

The isobaric devices win on peak efficiency (96–98%) but introduce a small amount of feed/brine mixing and require an auxiliary booster pump. The centrifugal devices give up some efficiency (80–85%) but eliminate the booster pump, the mixing, and most of the auxiliary plumbing.

Pressure Exchangers: ERI PX

The ERI PX family is the dominant isobaric ERD in large SWRO plants worldwide. The mechanism is elegant: a high-precision ceramic rotor with axial cylinders spins between two ceramic end plates. As the rotor turns, each cylinder is alternately exposed to a low-pressure feed port and a high-pressure brine port. The brine pushes feed out at full brine pressure, then the next half-rotation refills the cylinder with low-pressure feed. The result is a near-perfect transfer of pressure with only a small amount of axial mixing at the brine/feed interface inside the cylinder — typically 1–3% by volume.

The PX models scale by parallel rotor count, sharing a common motor-less architecture:

The PX requires a booster pump on its high-pressure feed outlet, because the rotor only transfers pressure — it cannot make up the small differential between the brine pressure and the membrane operating pressure (typically 1–3 bar of array pressure drop plus any margin). That booster is usually a vertical multistage pump such as a Grundfos CR or a Danfoss APP axial-piston pump, sized for the full permeate flow at the makeup differential.

Turbochargers: FEDCO HPB

Centrifugal energy recovery turbochargers compress the entire ERD function into one rotating component. The high-pressure brine enters the turbine side of a single-piece rotor, expands across the turbine vanes, and the shaft work immediately drives a pump impeller on the other end of the same rotor — pressurizing the feed flow on its way to the membrane housings. No motor, no booster pump, no external lubrication system, no electrical connection of any kind. The rotor floats on a hydrodynamic film of process water.

The FEDCO HPB family hits transfer efficiencies above 80% and, in the case of the HPB Ultra variants, above 85%. The trade-off versus an isobaric device is roughly 12–15 efficiency points in exchange for: zero feed/brine mixing, zero auxiliary equipment, dramatically smaller skid footprint, and the simplest possible startup and shutdown sequence. For mobile and containerized plants, the trade-off is usually decisive in favor of the turbocharger.

Booster Pump Considerations

If you specify a pressure exchanger, you must specify a booster pump alongside it. The booster handles the pressure differential between the PX brine inlet and the array feed pressure — typically 1.5–3.5 bar at full permeate flow. Two common selections:

• Danfoss APP axial-piston pumps. Positive-displacement, exceptionally efficient at the high pressures used in SWRO, available in stainless and Super Duplex options. See Danfoss high-pressure pumps.
• Grundfos CR vertical multistage pumps. Centrifugal, lower capital cost, easier to service. Suitable when the booster differential is moderate and consistent. See Grundfos pumps.

The booster's shaft power adds to the energy budget — typically 5–10% of the high-pressure pump's draw — but the overall system efficiency still favors the isobaric architecture when the ERD operates near its design point.

System Integration: A Walking Tour of the P&ID

A pressure-exchanger SWRO train looks like this in process flow terms: pretreated feed splits into two streams. One stream goes to the high-pressure pump suction and is pressurized to roughly the membrane pressure. The other (larger) stream goes to the low-pressure side of the PX, where it is pressurized to brine pressure (slightly below membrane pressure). The PX-pressurized feed then passes through the small booster pump, which raises it the last 2–3 bar to match the HP pump discharge. The two feed streams merge at a tee and enter the membrane housings together. Brine exits the housings and feeds the high-pressure side of the PX, completing the loop.

A turbocharger SWRO train is simpler. Pretreated feed goes through the HPB pump-side inlet, gets boosted by the recovered brine energy, then is topped up by a smaller high-pressure pump to reach membrane pressure. Brine exits the array and powers the HPB turbine side directly. There is no flow split, no booster pump, no tee for stream merging.

Real-World Specific Energy Targets

These figures are HP-pump + ERD specific energy only. Add 0.5–1.5 kWh/m³ for intake, pretreatment, post-treatment, and product delivery, depending on site conditions.

Economic Case: 100,000 GPD SWRO Plant

Consider a 100,000 GPD (16 m³/h) permeate plant operating 24/7/350 days, with a 40% recovery train (40 m³/h feed, 24 m³/h brine). At $0.15/kWh industrial electricity:

• No ERD: ~6.5 kWh/m³ × 16 m³/h × 8,400 h/yr = 873,600 kWh/yr = $131,040/yr power cost.
• FEDCO HPB-60: ~3.5 kWh/m³ × same = 470,400 kWh/yr = $70,560/yr. Annual savings: ~$60,000.
• ERI PX Q220 + booster: ~2.8 kWh/m³ × same = 376,320 kWh/yr = $56,448/yr. Annual savings: ~$74,500.

The ERD typically pays back in 12–24 months. Over a 20-year plant life, the difference between "no ERD" and a modern device runs well over $1 million for a plant this size.

Selection Matrix: PX vs HPB by Capacity

Putting It Together

Energy recovery has transformed SWRO from a high-cost specialty technology into a viable mainstream water source. The choice between a pressure exchanger and a centrifugal turbocharger is rarely about peak efficiency alone — it is about plant size, footprint, maintenance philosophy, and the cost of the supporting equipment around the ERD itself. For mid-size mobile and containerized plants, a FEDCO HPB turbocharger typically wins. For larger fixed-site plants, an ERI PX with booster typically wins. ForeverPure supplies both. For sizing assistance, see our pages on ERI pressure exchangers, FEDCO energy recovery, and complete SWRO systems.