Engineering Limits of Forward Osmosis and Membrane Distillation: Forward Osmosis Has Group-Standard Values, Membrane Distillation Still Lacks a Design Code

1. Two Routes at Very Different Levels of Standardisation

Discussing membrane distillation and forward osmosis in the same breath hides an important fact: their standardisation status is not comparable. Forward osmosis has a drafted group standard with feed water thresholds, design calculations, cleaning triggers and case-study data all present. Membrane distillation, at national and industry level, has only a terminology definition and no engineering design code. Mixing the two at scheme stage, and applying FO logic to MD, will produce a wrong capital estimate.

2. Forward Osmosis: A Complete Set of Values Exists

On 13 March 2023 the Shanghai Society of Environmental Sciences released the draft group standard Technical Guidelines of Forward Osmosis Membrane Process for Wastewater Treatment for public comment, with feedback due by 12 April 2023, drafted by Donghua University, Tongji University and others. Clause 3.3 defines forward osmosis as a process driven by the osmotic pressure difference across a heterogeneous selectively permeable membrane, in which the solvent passes through the membrane into the high osmotic pressure side while other components such as salts and organics are retained on the low osmotic pressure side and leave with the concentrated stream.

Feed water thresholds (clause 7.4.2, Table 1)

• pH: CTA 3 to 7, TFC 2 to 11, AQP 3 to 10.
• Temperature: CTA 5 to 50 degrees C, TFC 5 to 45 degrees C, AQP 5 to 40 degrees C.
• SS 1 000 mg/L or less, COD 50 000 mg/L or less, total hardness as CaCO3 15 000 mg/L or less.
• Turbidity: CTA 1 000 NTU or less, TFC 1 NTU or less, AQP 1 NTU or less.
• Residual chlorine: CTA 2 mg/L or less, TFC 0.1 mg/L or less, AQP 0.1 mg/L or less.
• Silica 50.0 mg/L or less, oil 20.0 mg/L or less.
• Silt density index: no requirement for CTA, 5.0 or less for TFC and AQP.

The most useful insight from this table is the trade-off it exposes: CTA tolerates far higher turbidity and residual chlorine than TFC and AQP, by three orders of magnitude on turbidity and a factor of twenty on chlorine, and pays for it with a much narrower pH window. Membrane selection is essentially a choice between feed tolerance and chemical robustness.

Design calculations and cleaning triggers

Clause 7.3 supplies five formula groups: product water flow corrected by an assembly coefficient of 0.90 to 0.96 and a stability coefficient of 0.6 to 0.8; membrane flux as the volume permeating to the draw solution per unit time divided by effective membrane area; reverse salt flux from the change in feed volume and salt concentration; rejection by the dilution factor method; and the number of elements as design product flow divided by the stable output of one element. Clause 7.1.4 sets the cleaning triggers at 15 percent loss of flux, 15 percent loss of salt rejection and 15 percent rise in differential pressure.

Clause 7.1.2 describes three membrane materials: cellulose acetate (CTA) with good hydrophilicity, strong oxidation resistance and high mechanical strength; thin-film composite polyamide (TFC) with abrasion resistance, heat resistance, chemical resistance and high water flux; and aquaporin (AQP) composite membranes that exploit the highly specific selectivity of aquaporin for water molecules, giving high recovery and good permeate quality. Clause 7.1.3 requires equipment selection and control systems to comply with HJ 579, HJ 2010, GB/T 39717, GB/T 19249 and HJ/T 270.

Two case datasets

• Incineration plant leachate: system recovery above 70 percent, with raw CODCr at 28 000 mg/L, SS 2 000 to 6 000 mg/L, chloride 4 000 to 5 000 mg/L and TDS up to 25 000 mg/L, producing permeate at 5.6 mg/L CODCr and 98.5 microS/cm.
• Power plant desulfurization wastewater: feed with mean TDS of 16 471 to 32 820 mg/L was concentrated to a mean TDS above 120 000 mg/L, a concentration factor of 4 to 7.5, with 90 percent salt removal, FO stage recovery of 75 to 90 percent and an hourly average throughput of 0.7 to 0.85 m3/h.

Clause 8.1.2 addresses the concentrate: where salt recovery is wanted, membrane processes combined with evaporative or freeze crystallization may be used to separate and recover salts. Clause 8.2 classifies draw solutions into inorganic salts, organic solutes and responsive draw solutes, and sets regeneration requirements. Forward osmosis therefore does not remove concentration energy; it relocates it to draw solution regeneration.

3. Membrane Distillation: A Definition, Not a Design Code

A review of national standards, industry standards and the ecological and environmental standards system confirms that membrane distillation has no national or industry engineering design code. The only current document that can be cited is at terminology level: HY/T 203.2-2016, Terms of Seawater Utilization, Part 2: Seawater Desalination Technology, approved by the State Oceanic Administration on 5 May 2016 and effective from 1 August 2016. Among its membrane desalination terms it defines membrane distillation as the process in which, driven by the vapour pressure difference across the membrane, vapour on the hot side passes through the membrane pores and condenses on the cold side; the same standard also defines forward osmosis.

At group standard level, membrane distillation only saw its first dedicated project initiation in 2025: the group standard Chemical Precipitation and Membrane Distillation System for Phosphate Rock Wastewater Reuse was formally launched by the China National Federation of Urban Industrial Products Trade Centres with a project notice on 16 June 2025, with a draft issued for public comment in July 2025. Standardisation of membrane distillation is thus still at the starting line, and the first such standard targets a single industry sector.

The conclusion must be stated plainly: design parameters for membrane distillation, including flux, hot-side and cold-side temperatures, wetting criteria and concentration factor, have no national or industry standard values to cite and must be determined by bench or pilot testing, or by reference to comparable projects.

4. A Benchmark: Pressure-Driven Membranes Are Fully Covered

As a yardstick, HJ 579-2010 shows how complete coverage looks for pressure-driven membranes such as NF and RO: feed limits in Table 3, membrane suitability in Table 4, pretreatment in clause 5.3, design calculations and instrumentation in clause 6.3, commissioning parameters in 7.2.2 and cleaning triggers in Appendix C. The forward osmosis draft deliberately mirrors this structure, since clause 7.1.3 requires compliance with HJ 579 and others. No document for membrane distillation currently reaches this level of completeness.

5. Practical Conclusions on the Engineering Boundary

1. What can be fixed in a scheme for FO: feed thresholds from Table 1, membrane material chosen on the pH, temperature, chlorine and SDI windows, product flow corrected by the 0.90 to 0.96 and 0.6 to 0.8 coefficients, and cleaning triggered at 15 percent.
2. What must be tested for MD: flux, temperature difference, concentration factor, wetting behaviour and membrane life, all determined by testing or by reference to comparable projects, with no reliance on circulating industry figures lacking a source.
3. Shared technical role: both are concentration and volume-reduction tools, not energy- eliminating tools. FO shifts energy into draw solution regeneration under clause 8.2; MD carries it in hot-side heating and latent heat recovery. Either way the tail end still needs evaporative or freeze crystallization, the route given in clause 8.1.2.
4. Document status: the forward osmosis guideline and the phosphate rock membrane distillation standard are both drafts for comment, not formally published versions. Values should be treated as design reference only, and contracts should state that they are to be checked against the final published texts.