REVERSE OSMOSIS AND NANOFILTRATION

Group FourRobert Gordon

Catherine HigginsZaman Sajid

Malcolm Wilkie

SIMILARITES AND DIFFRENCES

• Both pressure driven processes.• Boundary between the two processes is non

precise• The distinction is made because– slight difference in membrane permeability – the ability of nanofiltration to separate low

molecular weight non-ionic molecules

SIMILARITIES AND DIFFERENCES

• Both processes are used to separate low molecular weight solutes from a solvent.– Solvent is typically water– Solutes are typically salts or amino acids

• Both have applications on an industrial and domestic scale

• Nanofiltration more open polymer structure• Pure solvent flux much greater in Reverse Osmosis• Transport mechanisms are considered the same and

generally thought of as primary solution diffusion

DEVELOPMENT OF REVERSE OSMOSIS

• Originally developed by the US army to produce potable water from brackish sources.

• Research leading to commercial application began in the early 1950’s at the University of Florida.

• Advanced by Loeb and Sourirajan at University of California – first high flux asymmetric reverse osmosis membranes

• Original membranes made of cellulose acetate highly susceptible to chemical and biological degradation

• Further research funded by US government resulted in cellulose ester membranes– Higher flux and rejection of ionic species– Still susceptible to biological degradation

REVERSE OSMOSIS APPLICATIONS• Potable water from brackish sources

– Recent advances have made the production of potable water from sea water possible

• Used in conjunction with ultrafiltration and ion exchange to produce ultra pure water

• Extensively used to clean waste waters with small solutes and high BOD– Starch recovery from potato processing

• Concentration of fruit and vegetable juices– Superior flavour to those produced by heat concentration– Lower energy requirements than evaporation processes– Smaller plant cost and size– Very high dissolved salt concentration– High operating pressures– High fouling– Shortens membrane life

NANOFILTRATION APPLICATIONS

• Advantages over reverse osmosis when retention of mono – valent salts are not required– Much more permeable membrane– Higher solvent fluxes, often a factor of 10– Lower energy and capital costs

NANOFILTRATION APPLICATIONS

• Potable Water– Remove colour from water drawn from peat lands– Remove humic and fulvic acids– Conventional treatments labour and chemical intensive– Suitable for small domestic supplies in isolated communities– Can operate unattended for long periods of time– Does not remove minerals

• Sugar Concentration– Concentration of lactose from deproteinised whey stream– Concentrates lactose while allowing mono-valent ions to pass

through– Concentrated lactose stream used as feed for fermentation

NANOFILTRATION APPLICATIONS

• Dye Production– Improved by using nanofiltration– 10% NaCl reduced to 0.2%– Concentrate product by 30%

MATERIALS

• Originally derived from modified cellulose precursor

• Aromatic polyamides no used– Advantage – Significant lower tendency to

biological degradation– Disadvantage – Susceptible to damage from free

chlorine• Drawn as hollow fibres by melt or dry spinning– Typical diameter – 100 µm– Wall thickness – 20 µm

MATERIALS

• Significant advance in RO/NF technology due to composite membrane

• Casting active layer on surface of UF membrane• First membrane of this type utilised the addition

of polyethylenimine (PEI) to a polysulfone UF membrane

• Addition of toluene 2,4 diisocynate followed by heat drying improved salt rejection to > 95%

MATERIALS

• FilmTec (Dow subsidiary) developed the first composite membrane by interfacial polymerisation

• The membrane is wetted with polymer precursor and brought into contact with co-reactant

• Polymerisation takes place at the interface• Formation of the polymer prevents further

reaction taking place• m-phenylenediamine (wetting agent) & trimesoyl

chloride (non-aqueous co-reactant)

INTERFACIAL POLYMERISATION

TRANSPORT METHOD

• Both RO & NF separation achieved by application of pressure gradient

• Small solutes therefore membranes are considered non-porous

• Molecular volume of water and salts are similar therefore not a sieving process

• Exact nature of transport is under debate– Preferential sorption-pore flow– Solution diffusion

PREFERENTIAL SORPTION-PORE FLOW

• Assumes that transported species forms an adsorbed layer at membrane interface

• layer occludes the pores preventing the passage of certain molecules

SOLUTION DIFFUSION

• Assumes that molecules are not convected through pores but preferentially dissolve in the membrane

• The transported species diffuses through polymer down a chemical potential gradient

SOLUTION DIFFUSION

When 2 solutions of different concentration are separated by a semi-permeable the solvent will permeate through the membrane from the dilute to the concentrated phase (HWC to LWC)

SOLUTION DIFFUSIONIf a pressure is applied to the concentrated phase so that a pressure difference is generated across the membrane the net transport of the solvent will be reduced. The pressure difference required to reduce the net transport of solvent to zero is known as the osmotic pressure difference.

SOLUTION DIFFUSION

To relate the flux of a species to the chemical potential we need to establish the differences in chemical potential for the species across a membraneFor isothermal conditions the chemical potential of the solvent in the concentrated phase is expressed as

SOLUTION DIFFUSION

The chemical potential of the solvent in the dilute phase is expressed similarly as

The difference in chemical potential can thus be expressed as

SOLUTION DIFFUSION

For dilute solutions the activity of the solvent will be approximately equal on both sides. Therefore the potential is highly dependant on the pressure difference

SOLUTION DIFFUSION

There will be a net flow of solvent through the membrane until the chemical potentials in both phases are equal

Which gives:

If the dilute phase consists of pure solvent then the osmotic pressure of the concentrated phase can be expressed as

For dilute solutions

If the solute is denoted as component j

Therefore

SOLUTION DIFFUSION

SOLUTION DIFFUSION

Also for dilute solutions

Therefore

If the salt dissociates then the above relationship has to be modified, as the number of moles of solute will be increased. Therefore:

Osmotic Pressure of Real Solutions

• Previous expressions assume that solution behaviour is ideal

• π of solutions departs significantly from ideal behaviour at high concentrations

• It is possible to estimate π using a virial expansion

• Unfortunately coefficients are not always readily available over the range of concentrations required

Osmotic Pressure of Real Solutions

• The vapour pressure of the solution can also be used to estimate the osmotic pressure with reasonable accuracy

Trans-membrane flux of solvent

From Fick’s law

For isothermal operation the expression can be reduced to

i.e. for constant retentate conditions the flux will be linear w.r.t. the trans-membrane pressure

TRANS-MEMBRANE FLUX OF SOLVENT

TRANS-MEMBRANE FLUX OF SOLUTE

From Fick’s law:

As before this reduces to

Independent of pressure as is the rejection coefficient

TRANS-MEMBRANE FLUX OF SOLUTE

The concentration of the solute in the permeate can be written as

the rejection coefficient becomes

we can see that as the pressure drop across the membrane increases so does the rejection coefficient

REVERSE OSMOSIS SYSTEM DESIGN AND FUTURE CHALLENGES

• Membrane Fouling Control• Future $

MEMBRANE FOULING

• Membrane fouling is the main cause of permanent flux decline and loss of product quality in Reverse Osmosis systems, so fouling control determines system design and operation.

SOURCES OF FOULING

• There are four Categories1. Scale2. Silt3. Bio Fouling4. Organic Fouling

SCALE

• Cause– Precipitation of dissolved metal salts in the feed

water on the membrane surface. • Salts that form Scale– Calcium Carbonate– Calcium Sulphate– Silica Complexes– Barium Sulphate

METHOD TO DETERMINE • Scaling is determined by calculating Expected

Concentration Factor in the feed Brine solution.

Where

• Scaling is not a problem if and only if –Concentration Factor < 2 or–Recovery Rate = 50%

• Most systems are operated at Recovery Rate of 80 – 90%, the Concentration Factor Exceeds 2

CONTROL OF SCALE

• Calcium Carbonate Scale is a common problem and is controlled by the acidifying the feed.

• Others can be reduced by adding Antiscalant Chemicals such as Sodium Hexametaphosphate.

SILT• Cause– It is caused by the suspended particulates of all

types that accumulate on the membrane surface. • Typical Sources– Particulates– Iron Corrosion Products– Precipitated Iron Hydroxide– Algae– Fine Particulate Matter

METHOD TO DETERMINE SILT• SILT DENSITY INDEX (SDI) of the Feed Water– The time required to filter a fixed volume of water

through a standard 0.45µm pore microfiltration membrane.

Ti = Total Elapsed Test Time

Tf = Time required to collect 500ml Water

Tt = Total Elapsed Test Time

If:-• SDI <1 ---- RO can run several years without

colloidal fouling

• SDI< 3 ---- RO can run several months without cleaning

• SDI 3-5 ---- RO need Cleaning on regular basis

• SDI > 5 ---- Unacceptable and additional pre-treatment is required.

BIOFOULING• Cause– Growth of Bacteria on Membrane Surface.– Membrane itself becomes nutrient for bacteria

• Industrial examples are – Cellulose Acetate Membrane– Ployamide Fibres

• Solution is sterilization by heat, chlorine or chemicals.

ORGANIC FOULING• Causes

– Attachment of materials such as oil or grease onto the membrane surface.

• Such fouling may occur accidentally in municipal drinking water system, but is more common in industrial applications in which RO is used to treat a process or effluent stream.

CONTROLLING ORGANIC FOULING

• Organic Materials can be removed by

– Filtration– Carbon Adsorption

TYPICAL SEAWATER RO PLANT

Q & A SESSION

THANK YOU!