Thermal desalination: what MSF and MED still do best

It has become fashionable to write thermal desalination's obituary. The market data half-supports it — new procurement is overwhelmingly membrane — yet a large fraction of the Gulf's installed capacity remains distillation, much of it running at availabilities membrane plants envy, and a set of duties keeps specifying it. A reviewer needs to understand the machine, not the fashion.

The two machines

Multi-stage flash drives hot brine through a cascade of chambers at successively lower pressure; a fraction flashes to vapour in each and condenses on tubes preheating the incoming feed. No boiling on hot surfaces means superb tolerance of scaling and feed quality — the property that built the Gulf fleet. Multiple-effect distillation boils seawater on the outside of tubes in a series of effects, each driven by the vapour of the last; with thermal vapour compression (TVC) an ejector recycles vapour and lifts efficiency further. MED runs cooler and converts steam to water more efficiently; MSF forgives more abuse.

Parameter	MSF	MED-TVC
Gained output ratio (kg water/kg steam)	8–10	10–14
Top brine temperature	90–112 °C	60–70 °C
Electrical parasitic	2.5–4.0 kWh/m³	1.5–2.5 kWh/m³
Product TDS	< 25 mg/L	< 25 mg/L
Scaling exposure	Low (flash evaporation)	Managed by low TBT

Where thermal still wins

• Committed steam: co-generation stations and industrial hosts with extraction steam or genuine waste heat — the energy accounting (equivalent work, not raw kilowatt-hours) can favour MED.
• Brutal feedwater: high-salinity, bloom-prone or hydrocarbon-exposed intakes where membrane pretreatment risk prices itself out.
• Product purity: boiler make-up and process duties wanting sub-25 mg/L distillate without a second RO pass.
• Brine concentrators: thermal equipment is the backbone of ZLD trains — the technology is migrating, not dying.

The honest review method is the one I use for any selection: price thermal energy as electricity forgone at the turbine, price membrane pretreatment for the worst credible month, and test both against the offtaker's penalty regime. Run that way, most new capacity goes membrane — and the exceptions that go thermal go for reasons that survive scrutiny.