How to Evaluate a Desalination Plant Provider

A desalination plant is a 25-year commitment, and the provider you select carries more of that risk than the technology you choose. A membrane is a commodity; an engineering, procurement, and construction partner who under-scopes the pre-treatment, walks away after commissioning, or whose proposal hides the real lifecycle cost can turn a sound technical concept into a plant that never hits its design output. The difference between a strong and a weak provider on a $40 million plant is routinely $5 to $10 million in lifetime cost and a year of schedule.

The common failure in provider selection is treating it as a price comparison. The lowest bid on a desalination plant is frequently the most expensive over its life, because it wins by under-specifying the parts that do not show up in a headline price: pre-treatment robustness, energy recovery quality, instrumentation, and the operating support that keeps the plant at design output. Apples-to-apples comparison is the entire game, and most RFPs are not structured to enforce it.

This article gives procurement leads, capital projects teams, and plant managers a structured evaluation framework: the qualification criteria that separate credible providers from optimistic ones, the proposal elements that must be normalised before comparison, the contract terms that allocate risk correctly, and the red flags that predict trouble.

Quick Navigation

• Why provider selection outweighs technology selection
• Qualification: what to screen before the RFP
• Normalising proposals for apples-to-apples comparison
• The contract terms that allocate risk
• Red flags that predict trouble
• Scoring framework
• Where provider evaluation goes wrong
• The CFO Hook
• Related Articles
• FAQ

Why provider selection outweighs technology selection

Desalination technology is mature and largely commoditised. The major membrane, pump, and energy-recovery suppliers are known, their performance is documented, and the SWRO versus thermal choice is usually settled by the feedwater and energy context before any provider is engaged. What is not commoditised is the integration: how the pre-treatment is sized for your worst-case feedwater, how the energy recovery is configured, how the controls and instrumentation are specified, and whether the provider stays engaged to get the plant to design output. The International Desalination Association's resources document how integration quality, not core technology, drives the spread in delivered plant performance across otherwise similar projects.

A pattern that recurs across desalination projects is that two providers offering the identical membrane technology deliver plants with a 20 to 30% difference in lifetime cost, entirely from integration quality. The strong provider sizes pre-treatment for the algal bloom that happens once every three years; the weak one sizes for the average and the plant trips the first time the feedwater turbidity spikes. The membrane is the same. The outcome is not.

This is why provider evaluation must assess engineering judgement, not just price and reference list. The question is not "can this provider build a desalination plant" but "will this provider's specific design hold up against my feedwater for 25 years, and will they be accountable if it does not." That is a harder question to answer from a glossy proposal, which is why the evaluation framework below is built to surface it.

Qualification: what to screen before the RFP

Qualification screens out providers who cannot credibly deliver before you spend effort comparing detailed proposals. The screen is about track record on plants like yours, not just plants in general.

Screen on these five criteria:

• Reference plants of comparable scale and feedwater. A provider with ten 5,000 m3/day brackish plants is not qualified for a 50,000 m3/day seawater plant. Demand references at your scale, on your feedwater type, that have been operating for at least three years (long enough for the real operating cost to surface).
• Verified operating performance, not commissioning performance. Ask references for their actual three-year output, energy consumption, and membrane replacement history versus the original design. Plants hit design at commissioning and drift after; the drift is the truth.
• In-house pre-treatment engineering. Pre-treatment is where desalination plants fail. A provider who subcontracts pre-treatment design wholesale has outsourced the part that matters most.
• Financial stability for the warranty and support period. A 25-year plant needs a provider who will exist in year 5 to honour the performance warranty. Screen the balance sheet.
• Local operating and support presence. A plant supported from another continent loses days to every fault. Local or regional service capability is a hard requirement for a critical-supply plant.

These criteria mirror the general principles in our water treatment provider evaluation guide, sharpened for the specific risk profile of desalination, where pre-treatment and long-horizon support dominate.

Normalising proposals for apples-to-apples comparison

The core discipline of evaluation is normalisation: forcing every proposal onto the same basis so that price differences reflect value differences, not scope differences. Unnormalised, the cheapest proposal usually wins by quietly omitting scope, and the omission surfaces as a variation order or an operating shortfall later.

Build the comparison on total cost of water ($/m3 over 25 years), not CAPEX. A proposal that is 15% cheaper to build but 20% more expensive to run loses badly over the life of the plant. The total-cost-of-water lens is the only one that exposes the under-scoped bid, and structuring the RFP for a water treatment project to demand this basis is what makes the comparison honest.

The contract terms that allocate risk

The contract is where risk is allocated, and on a desalination plant the risks worth allocating explicitly are performance, feedwater variability, and energy. A contract that leaves these vague defaults the risk to whichever party has the weaker negotiating position after a problem appears, which is usually the owner.

Performance guarantee with liquidated damages. The provider must guarantee the plant's output, product quality, and energy consumption, with defined liquidated damages if it underperforms over a test period. A guarantee without teeth is marketing. The test period must be long enough to capture real operating conditions, not a perfect commissioning week.

Feedwater risk allocation. Define the design feedwater envelope in the contract. Within the envelope, the provider owns performance; outside it, the owner owns the consequences. Without this boundary, every performance shortfall becomes a dispute over whether the feedwater was worse than expected. This is the most-litigated clause in desalination, so it must be explicit. According to the US Bureau of Reclamation's desalination research, feedwater-variability disputes are among the most common and costly sources of conflict in desalination delivery, which is why an explicit feedwater envelope matters.

Energy and membrane warranties. The energy guarantee must specify the ERD and the conditions under which it holds. The membrane warranty must define expected life and the replacement-cost basis, so a short-life membrane choice cannot hide behind a low headline price.

Where the desalinated water serves a critical or expanding operation, the broader technical due diligence on water treatment framework provides the independent-review layer that catches a weak contract before signature.

Red flags that predict trouble

Some signals reliably predict a difficult project. None is disqualifying alone, but a cluster of them should drop a provider down the ranking sharply.

• A headline price well below the field with no scope explanation. The cheapest bid by a wide margin has almost always under-scoped something, usually pre-treatment or operating support.
• Reluctance to share three-year reference operating data. A confident provider volunteers it. A reluctant one is hiding drift between design and reality.
• A vague or absent energy guarantee. Energy is the dominant lifetime cost. A provider who will not commit to a firm kWh/m3 is leaving you exposed on the biggest line item.
• Pre-treatment designed to the average feedwater. Ask explicitly what feedwater event the pre-treatment is sized for. "Average conditions" is the answer that precedes a fouling failure.
• No local support presence and no firm response-time commitment. A critical plant supported remotely will lose production to every fault.

Scoring framework

Score providers on a weighted matrix that reflects the actual risk, not a generic procurement template. For desalination, lifetime cost and integration quality should dominate; headline CAPEX should be a minority of the weight. The World Bank's guidance on desalination procurement stresses total-cost-of-water evaluation over headline CAPEX as the single most important discipline in provider selection.

Suggested weighting: total cost of water over 25 years (35%), guaranteed performance and the strength of the liquidated-damages regime (20%), pre-treatment engineering robustness (15%), reference operating track record at comparable scale (15%), operating and support capability (10%), CAPEX (5%). The deliberately low CAPEX weight is the point: on a 25-year asset, the build cost is a small fraction of the total cost of water, and over-weighting it selects for the under-scoped bid.

The right weighting depends on whether your plant is critical-supply or has a fallback, and on your feedwater risk. Post your project and qualified providers will respond against a normalised specification, giving you the apples-to-apples basis the scoring framework needs to work.

Where provider evaluation goes wrong

Failure 1: selecting on CAPEX. A team picks the lowest build price, the provider under-scoped pre-treatment to win, and the plant fouls and underperforms from year one. The lifetime cost overrun of $3 to $8 million on a mid-size plant dwarfs the few hundred thousand saved on the build. The fix is the total-cost-of-water lens with normalised scope.

Failure 2: accepting a vague feedwater basis. The contract does not define the design feedwater envelope, the source water turns out harsher than assumed, and every performance shortfall becomes a dispute the owner loses. The fix is an explicit feedwater envelope with risk allocated inside and outside it.

Failure 3: no post-commissioning support in scope. The plant is built, commissioned to design, and the provider departs. Six months later the output has drifted and there is no one accountable. The fix is a defined operating-support period in the contract, scored as part of the evaluation.

To run an evaluation that survives these failure modes, start from a defined, normalised specification rather than a vague brief. Nepti characterises your feedwater and models the plant requirement, producing the design basis and total-cost-of-water framework that lets you compare providers on identical terms. Start at Nepti.

The CFO Hook

If you select a desalination provider on total cost of water over 25 years with normalised scope rather than on headline CAPEX, you typically save $3 to $10 million in lifetime cost on a $40 million plant by avoiding the under-scoped pre-treatment and weak energy guarantee that the cheapest bid hides. The biggest cost-of-doing-nothing is awarding on price to a provider who designed pre-treatment for average feedwater, then watching the plant foul on the first algal bloom and limp below design output for the rest of its life, a structural shortfall no operating effort can fully recover.

Related Articles

• SWRO vs Thermal Desalination: Cost and Energy Comparison
• Water Treatment Provider Evaluation: A Buyer Framework
• How to Structure an RFP for a Water Treatment Project
• Technical Due Diligence on Water Treatment Assets
• Brine Disposal from Desalination: Regulatory and Technical Options

FAQ

What is the single most important criterion when choosing a desalination provider?

The robustness of the pre-treatment design against your worst-case feedwater, because pre-treatment failure is the most common cause of a plant missing its design output. Price is a poor proxy for this.

Why is the lowest bid often the most expensive?

Because it usually wins by under-scoping the parts that do not show in the headline price: pre-treatment, energy recovery quality, and operating support. Those omissions surface as variation orders or operating shortfalls over the plant's life.

How long should reference plants have been operating?

At least three years, because plants hit design at commissioning and drift afterward. Three-year operating data reveals the real energy use, membrane life, and output the proposal promised.

What contract terms matter most for a desalination plant?

A performance guarantee with meaningful liquidated damages, an explicit design-feedwater envelope that allocates feedwater risk, and firm energy and membrane warranties. These three carry most of the lifetime risk.

Should I weight CAPEX heavily in the scoring?

No. On a 25-year asset the build cost is a small fraction of total cost of water. Over-weighting CAPEX selects for the under-scoped bid. Weight total cost of water and guaranteed performance instead.

How do I compare proposals fairly?

Normalise every proposal to the same design feedwater, guaranteed output, recovery, energy basis, and membrane-life assumption, then compare on total cost of water over the plant life rather than CAPEX.

What is the biggest red flag in a desalination proposal?

A headline price well below the field with no scope explanation, almost always paired with pre-treatment sized to average feedwater and a vague energy guarantee.