Sie sind auf Seite 1von 19

Forward osmosis under pressure:

Performances and challenges of


pressure assisted osmosis (PAO)
School of Chemical Engineering

Gaetan Blandin Chuyang Tang


Arne Verliefde Darli Myat
Amy Childress Pierre Le-Clech

now Hong Kong Uni now USC


Outlook
FO-RO hybrid concept in desalination pre-dilution
Assessment of three commercial membranes performances in FO
and PAO operation:
Membrane characteristics and deformation
Water and solute mass transfer
Fouling
Trace organic contaminants rejection (not discussed today)

Energy, capital cost model


Practical challenges (module and pressure drop)
New concept of PAO-NF

2
FO-RO hybrid: Seawater pre-dilution

First study: Cath et al. (1,2)


Lower energy for seawater desalination thanks to the osmotic dilution
Beneficial reuse of wastewater, i.e. water recycling
Multi-barrier protection to increase consumer confidence in water recycling
Reduction in reverse osmosis membrane fouling

Promising but need for flux improvement


1 Cath, T. Y. et al. (2009). "A Novel Hybrid Forward Osmosis Process for Drinking Water
Augmentation using Impaired Water and Saline Water Sources." WERC.
2 Cath, T. Y. et al. (2010) Journal of Membrane Science 362(12): 417-426.
Osmotic processes

PRESSURE ASSISTED
OSMOSIS (PAO)

4
Membrane characteristics under pressure
Water permeability: effect of the SPACER (in RO cross flow setup)

3 3
RO feed spacer as membrane
Permeate spacer as membrane support
support

2 2
A (LMH.bar-1)

A (LMH.bar-1)
CTA
CTA
TFC1
TFC1
TFC2
1 1 TFC2

0 0
0 2 4 6 0 2 4 6
Pressure (bar) Pressure (bar)

Improved permeability with TFC membranes


FO membrane deformation with pressure when RO feed spacer is used!

5 Blandin et al. (2013) Journal of Membrane Science


Membrane characteristics
Salt permeability (RO cross flow setup)
B/A
A (L.m-2.h-1.bar-1) B (10-7.m.s-1) S0
(0bar)

4 bar / 4 bar / RO 4 bar / 4 bar / RO


0 bar permeate feed 0 bar permeate feed m
spacer spacer spacer spacer

CTA, HTI 0.7 0.9 1.0 1.5 2.1 3.1 2.1 663

TFC1, HTI 1.3 1.7 2.1 0.3 0.6 0.7 0.23 1227

TFC2, Porifera 2.1 2.2 2.2 1.2 1.4 1.4 0.57 344

Membrane deformation also increases Salt permeability (B)

Improved selectivity with TFC membranes

High variation on S values!

6
Experimental vs. modelled water flux
DI as feed, 35g.L-1 red sea salt as draw
30
CTA TFC1 TFC2
FO: 2 times higher flux with TFC2!
(lower S value)
20
Jw (L.m-2.h-1)

PAO: Pressure increases flux in all


cases
10

Deformation further improved flux


0
0 bar 4bar / permeate 4bar / RO feed
spacer spacer

Both novel membrane development and hydraulic


pressure (PAO) help to improve water flux

7
Reverse salt diffusion
DI as feed, 35g.L-1 red sea salt as draw
0.7

0.6
PAO: decrease of RSD for all membranes
0.5 Despite deformation/ stretching
Thanks to more severe ICP resulting of
Js/Jw (g.L-1)

0.4 pressure-enhanced water flux!


0.3
PAO operation allows to tackle the
0.2 CTA water/salt permeability trade off of FO!!
0.1 TFC1 Higher water permeability
TFC2 Lower reverse salt diffusion
0.0
0 bar 4bar / permeate 4bar / RO feed
spacer spacer

Potential interest in using more permeable membrane in PAO even with lower salt rejection!

8
Fouling behaviour
Foulant mix: 1.2g/L Sea salts, 200mg/L CaCl2, 200mg/L Alginate, 200mg/L Humic acid
Five (5) consecutive batches at constant recovery =30%
20 20
FO CTA PAO CTA
TFC1 TFC1
15 15
TFC2 TFC2
Jw (L.m-2.h-1)

Jw (L.m-2.h-1)
10 10

5 5

0 0
batch 1 batch 2 batch 3 batch 4 batch 5 batch 1 batch 2 batch 3 batch 4 batch 5

Lower flux membranes: no flux decline Performance decrease observed with all 3
More severe flux decline with high flux Porifera P membranes in PAO mode!

Is fouling due to higher operating flux? Is fouling due to applied pressure or higher
operating flux?
Fouling behaviour
Impact of hydraulic pressure
CTA, AL-FS Sparse and smooth cake deposition (FO70)
60%
FO70 PAO

50%
Batch 1 Batch 3 Batch 5
Recovery

Batch 1 Batch 3 Batch 5


40%

30%
batch 1 batch 2 batch 3 batch 4 batch 5
More severe loss of recovery when hydraulic Formation of a layer on the membrane
pressure is implemented! and then accumulation (PAO)

Confirms higher fouling behaviour with PAO


Different foulant cake structure (compaction!)

More on: Is FO really a low fouling process 1


Poster IDA-51730 0
Cleaning strategy for high flux fouling

FO, TFC2, AL-FS


100%

80%
Initial flux recovery

60%

40%

20%
Possible to recover initial
0% flux with extended time of
No cleaning 15 min 1h 1h, high CFV osmotic backwashing at
high CFV
Osmotic cleaning also adapted to operation of FO with high
permeation flux!

11
Impact of FO recovery on RO energy consumption
How to use the osmotic dilution?
Scenario A: Constant RO feed flow and RO recovery Scenario B: Constant RO feed flow (diluted by pre-FO)
and increased RO recovery
2.5 4.0

RO energy consumption (kWh.m-3)


RO energy consumption (kWh.m-3)

3.5
2.0
3.0

2.5
1.5

2.0
Constant RO feed flow (4000m3.h-1)
1.0
1.5
0% 50% 100% 45% 55% 65% 75%
% FO recovery %RO recovery

Possible to reduce by 30% RO energy needs


or to increase RO recovery up to 60-65% for similar energy consumption
RO energy consumption considering different scenario of FO-RO hybrid
implementations using ROSA software.
Blandin et al. (2015). Desalination 363: 26-36
SWRO desalination economics
FO/PAO-RO hybrid can allow for lower energy needs (30%) and possible lower
maintenance costs. But
Average water production cost: 0.76$.m-3 If similar flux in FO/PAO and RO!

Energy FO-RO hybrid


Capex
32%
38%
Other CAPEX costs

RO filtration unit
Mainten
ance/ RO
FO filtration unit
labour
30%
Desaldata.com
0 0.1 0.2 0.3 0.4 0.5
Capital investment costs ($.m-3)

Important to consider investment cost in FO/PAO-RO hybrid economics!


FO vs. PAO
CAPEX vs. energy?

Model based on HTI CTA membrane


(FO/PAO recovery=50%)
0.3
Filtration investment costs Pumping energy costs

PAO is beneficial vs. FO thanks to a


0.2 significant decrease of filtration costs
$.m-3

Optimum of operation
0.1 (6bar for HTI CTA membrane, and economics
assumptions considered)

0.0
FO PAO 2bar PAO 4bar PAO 6bar PAO 8bar PAO
10bar

Despite higher energy need, PAO could be more financially attractive than FO

14
FO Module operation
Development/commercialisation: Spiral wound, plate and frame and hollow fibre configurations

Difference with RO:


Need to circulate both feed and draw solutions

Challenges in design

Packing Optimised Membrane


density hydraulics support

Pressurisation will be needed for feed/draw


transfer (PAO operation could not be avoided)

Further work needed to understand feed/draw


hydraulics optimisation and interconnections!

http://www.globalwaterintel.com/news/2013/45/toray-boosts-membrane-market-share-woongjin-buy.html
http://porifera.com/products/pfo-modules-systems/
http://www.toyobo-global.com/news/pdf/2014/05/press20140521.pdf
Hydrodynamics in 8 module (from initial pilot testing)
Typical FO operation

42L.min-1 46L.min-1
Feed

Draw
10L.min-1 14L.min-1

Volume Increase by 40% in one pass in the module!!


Slight pressure build up in draw channel!

16
Reverse salt diffusion
DI as feed, 35g.L-1 red sea salt as draw
0.7

0.6
PAO: decrease of RSD for all membranes
0.5 Despite deformation/ stretching
Thanks to more severe ICP resulting of
Js/Jw (g.L-1)

0.4 pressure-enhanced water flux!


0.3
PAO operation allows to tackle the
0.2 CTA water/salt permeability trade off of FO!!
0.1 TFC1 Higher water permeability
TFC2 Lower reverse salt diffusion
0.0
0 bar 4bar / permeate 4bar / RO feed
spacer spacer

Potential interest in using more permeable membrane in PAO even with lower salt rejection!

17
The concept of PAO-NF
50 1.0
(here at 2bar)
Jw RST
40 0.8
Jw (L.m-2.h-1)

30 0.6

RST (g.L-1)
20 0.4

70
10 0.2 60
Jw

Jw (L.m-2.h-1)
50
40
0 0.0 30
Jw, p
CTA FO Dow90 Dow270 Hydra200 Hydra1000 GE 20
10
0
Jw,
Small contribution of osmotic pressure 0 2 4 6
Higher fouling rate, but easily reversible Hydraulic Pressure, bar
But option to consider when high flux is needed

More on that Wednesday 2pm


VALIDATION OF PAO-NF CONCEPT
IDA #51503
Thanks!

PAO Questions ?
HTI TFC

100%
90%
80%
70%

Recovery (%)
60%
50%
40%
Porifera 30%
20%
10%
0%
0 50 100 150 200
membrane surface area (m2.m-3.h)

Das könnte Ihnen auch gefallen