SWRO vs Thermal Desalination: Cost and Energy Comparison

The choice between seawater reverse osmosis and thermal desalination is a 25-year energy bet, and most of the cost lives in the operating column, not the build. A 50,000 m3/day plant consumes $4 to $12 million a year in energy alone, and the gap between the two technologies on that line item can exceed $5 million annually. Pick the wrong process for your energy context and you lock a site into a structural cost disadvantage that no operational tweak can recover.

The industry default has settled hard toward SWRO over the last two decades, and for most greenfield plants that default is correct. But the default is not universal. Where waste heat is effectively free, where feedwater is extreme, or where boron and product-purity constraints bite, thermal processes still win on total cost. Treating SWRO as the automatic answer without testing it against the site's energy and feedwater reality is how plants end up over-engineered or wrongly sited.

This article gives capital projects leads, plant managers, and sustainability directors a decision-grade comparison: how each technology consumes energy, what each costs to build and run, where each wins, and the feedwater and integration factors that flip the answer.

Quick Navigation

• How each technology works
• The energy comparison that decides it
• CAPEX and OPEX side by side
• Where SWRO wins
• Where thermal still wins
• Feedwater, boron, and product purity
• Where desalination decisions go wrong
• The CFO Hook
• Related Articles
• FAQ

How each technology works

Seawater reverse osmosis (SWRO) pushes seawater through semi-permeable membranes at high pressure (55 to 80 bar) so that water passes and salts are rejected. It is a pressure-driven, membrane-based separation. The energy goes into the high-pressure pumps, and the defining innovation of modern SWRO is the energy recovery device (ERD), which captures the pressure energy in the reject brine and returns it to the feed, cutting net energy by 40 to 60%.

Thermal desalination evaporates seawater and condenses the vapour as fresh water, leaving salts behind. The two dominant variants are multi-stage flash (MSF), which flashes heated brine across successive low-pressure stages, and multi-effect distillation (MED), which evaporates across a series of effects at descending pressure. Both are heat-driven, and their economics live or die on the cost of that heat. Where the heat is a waste stream from an adjacent power or industrial process, thermal can be remarkably cheap to run; where the heat must be raised purely for desalination, it is expensive.

The fundamental divide is this: SWRO trades a higher membrane and pre-treatment burden for far lower energy, while thermal trades simplicity and feedwater tolerance for high heat demand. Everything downstream of that trade-off, including the full desalination energy consumption picture, flows from it.

The energy comparison that decides it

Energy is the decision. SWRO with modern energy recovery consumes roughly 3 to 4 kWh per m3 of product water for typical seawater (35,000 mg/L TDS). Thermal processes consume far more total energy, but most of it is thermal (heat) rather than electrical: MED runs around 1.5 to 2.5 kWh/m3 electrical plus a large thermal load equivalent to a further 6 to 12 kWh/m3, and MSF is higher still.

The catch is the quality of the energy. A kWh of electricity is worth far more than a kWh of low-grade waste heat. So the honest comparison is not total kWh but cost of energy per m3. Where electricity costs $0.08 to $0.15/kWh and there is no free heat, SWRO's 3 to 4 kWh/m3 translates to $0.25 to $0.60/m3 of energy cost, and thermal cannot compete because its heat must be bought. Where a thermal plant can tap genuinely free or near-free waste heat from an adjacent power station, its energy cost can fall below SWRO's.

According to the International Renewable Energy Agency's analysis of desalination energy, SWRO is the lowest-energy mature desalination technology for standard seawater, which is why it dominates new capacity. Thermal retains a niche where it is co-located with a heat source that would otherwise be wasted, turning a liability (reject heat) into a desalination asset.

CAPEX and OPEX side by side

The capital and operating profiles of the two technologies differ enough that the total cost of water can swing 30 to 50% depending on site conditions. The table below normalises both per unit of capacity for a like-for-like comparison.

The CAPEX gap favours SWRO, and the OPEX gap usually does too, unless free heat is available. The total-cost-of-water decision therefore reduces to one question more than any other: is there a genuine, reliable, near-free heat source on this site? If yes, thermal deserves a full evaluation. If no, SWRO is almost certainly the lower lifecycle cost.

Where SWRO wins

SWRO wins for the large majority of greenfield seawater plants, and the reasons are structural. Its energy advantage on standard seawater is decisive wherever heat is not free, its modular membrane-train design scales cleanly from small to mega-plant, and its CAPEX per unit is lower. The technology has also matured: energy recovery devices, improved membranes, and better pre-treatment have driven SWRO's energy down by more than half over two decades. The International Desalination Association reports that SWRO accounts for the overwhelming majority of new desalination capacity, reflecting its decisive energy advantage on standard seawater.

SWRO is the right answer when: the feedwater is standard seawater (30,000 to 40,000 mg/L TDS, moderate fouling), electricity is available at industrial rates or from solar, there is no large free heat source, and the product-purity target is potable or industrial-process grade rather than ultra-pure. This covers most coastal municipal and industrial desalination.

The discipline SWRO demands is pre-treatment. The membrane train is only as reliable as the water fed to it, and membrane fouling prevention is the single largest determinant of SWRO operating cost. A plant that under-invests in pre-treatment pays for it in membrane replacement and CIP frequency for the life of the plant.

Where thermal still wins

Thermal desalination is not obsolete; it is niche. It wins decisively in three situations, and each is about either energy context or feedwater extremity.

First, co-location with waste heat. A thermal plant bolted onto a power station or a heavy industrial process that vents large quantities of low-grade heat turns that heat from a cost into a feedstock. Here thermal's energy cost can undercut SWRO, and the higher CAPEX is justified by the operating saving over 25 years. Much of the Gulf's installed thermal capacity exists for exactly this reason.

Second, extreme feedwater. Where the source water is very high salinity (above 45,000 mg/L), very warm, or heavily fouling (high algae, high organics), SWRO's pre-treatment burden and reduced membrane recovery erode its advantage. Thermal processes are largely indifferent to feedwater quality, because evaporation does not care about fouling potential the way membranes do.

Third, ultra-pure product requirement. Thermal distillate is exceptionally pure (under 25 mg/L TDS, near-zero boron), so where the product feeds a process that cannot tolerate the 200 to 500 mg/L and residual boron of single-pass SWRO, thermal can be simpler than SWRO plus a second pass.

The right answer for a specific site depends on the actual feedwater analysis and the energy context. Post your project and qualified desalination specialists will model both technologies against your real numbers rather than generic ranges.

Feedwater, boron, and product purity

Boron deserves a dedicated note because it routinely catches SWRO projects off guard. Seawater contains 4 to 6 mg/L boron, and single-pass SWRO rejects it poorly, leaving 1 to 2 mg/L in the product. For potable use the WHO guideline is 2.4 mg/L, but for agricultural irrigation and some industrial processes the limit is far tighter (0.3 to 0.5 mg/L), which forces a second RO pass at elevated pH. That second pass adds 10 to 20% to energy and CAPEX, and a plant that did not budget for it discovers the gap during commissioning. The WHO drinking-water guidelines set the boron limit that, combined with tighter agricultural thresholds, determines whether a second RO pass is required.

Thermal processes sidestep boron entirely because distillation leaves it behind with the salts. So a site with a strict boron limit for irrigation reuse faces a real choice: two-pass SWRO with pH adjustment, or thermal if free heat is available. This is exactly the kind of constraint that should be characterised before technology selection, not discovered after. Where the desalinated water feeds agricultural reuse, the boron and salinity limits in our agricultural water reuse regulations guide become the binding design constraint.

The reject brine is the other shared challenge: both technologies produce a concentrated brine that must be disposed of responsibly, a problem covered in depth in our brine disposal from desalination guide.

Where desalination decisions go wrong

Failure 1: assuming free heat is reliable. A thermal plant justified on waste heat from an adjacent process is only as economic as that heat is available. When the host process runs at reduced load or shuts for maintenance, the thermal plant either stops or burns expensive fuel to make up the heat. A plant that modelled free heat at 95% availability but gets 70% sees its energy cost double. The fix is to stress-test the heat-availability assumption before committing to thermal.

Failure 2: under-specifying SWRO pre-treatment. The most common SWRO failure. A plant builds minimal pre-treatment to save CAPEX, then the membranes foul on the first algal bloom or turbidity spike. CIP frequency rises, membrane life halves, and the replacement cost of $0.5 to $2 million per major change-out arrives years early. The correct decision is to size pre-treatment for the worst-case feedwater, not the average.

Failure 3: ignoring boron until commissioning. A single-pass SWRO plant designed for a potable boron limit, then asked to supply agricultural reuse, fails the tighter boron spec and needs an unbudgeted second pass. Cost of the retrofit: 10 to 20% of plant CAPEX, plus the schedule slip.

The way to avoid all three is to model both technologies against the real feedwater, energy context, and product spec before selecting. Nepti models your feedwater matrix and energy scenario and produces a ranked technology comparison with cost projections, so the SWRO-versus-thermal decision is made on your numbers, not the industry default. Start at Nepti.

The CFO Hook

If you select SWRO over thermal for a 50,000 m3/day plant where electricity is $0.10/kWh and no free heat exists, you save roughly $0.40 to $0.90 per m3 of energy cost, which is $7 to $16 million a year at full output, dwarfing the modest CAPEX difference over the plant's 25-year life. The biggest cost-of-doing-nothing is selecting thermal on a waste-heat assumption that turns out to be 70% available instead of 95%, which doubles your energy cost and strands a higher-CAPEX asset on a structural cost disadvantage you cannot operate your way out of.

Related Articles

• Desalination Energy Consumption: The Numbers That Matter
• Brine Disposal from Desalination: Regulatory and Technical Options
• How to Evaluate a Desalination Plant Provider
• Membrane Fouling Prevention: Protecting the RO Train
• Reverse Osmosis Systems: Industrial Guide

FAQ

Is SWRO always cheaper than thermal desalination?

For standard seawater with no free heat source, yes, almost always, because SWRO's energy use is far lower. Thermal only competes on cost where it can tap genuinely free or near-free waste heat.

Why is thermal desalination still used in the Gulf?

Largely because of co-location with power stations that produce abundant waste heat, and because the warm, high-salinity, high-fouling Gulf seawater is harder on SWRO membranes. The economics flip where free heat is available.

How much energy does SWRO use per cubic metre?

Modern SWRO with energy recovery uses roughly 3 to 4 kWh/m3 of product water for standard seawater. Older plants without energy recovery used 6 to 8 kWh/m3.

What is the boron problem in SWRO?

Single-pass SWRO rejects boron poorly, leaving 1 to 2 mg/L in the product. For agricultural reuse with a 0.3 to 0.5 mg/L limit this forces a second RO pass at high pH, adding 10 to 20% to energy and CAPEX.

Which technology produces purer water?

Thermal distillation, which yields under 25 mg/L TDS and near-zero boron. Single-pass SWRO yields 200 to 500 mg/L TDS, which is fine for most uses but may need a second pass for ultra-pure or strict-boron applications.

Does feedwater quality affect the choice?

Strongly. SWRO is sensitive to fouling and high salinity, needing heavy pre-treatment. Thermal is largely indifferent to feedwater quality, which is why it wins on very high-salinity or heavily fouling sources.

What is the biggest risk in choosing thermal?

Over-estimating the availability of the free heat source. If the host process runs at reduced load, the thermal plant's energy cost can double, stranding a high-CAPEX asset.