Engineering Conferences InternationalECI Digital Archives

Advanced Membrane Technology VII Proceedings

9-14-2016

Salinity gradient energy: Assessment of pressureretarded osmosis and osmotic heat engines forenergy generation from low-grade heat sourcesTzahi Y. CathColorado School of Mines, tcath@mines.edu

Johan VannesteColorado School of Mines

Michael B. HeeleyColorado School of Mines

Kerri L. HickenbottomHumboldt State University

Follow this and additional works at: http://dc.engconfintl.org/membrane_technology_vii

This Abstract and Presentation is brought to you for free and open access by the Proceedings at ECI Digital Archives. It has been accepted for inclusionin Advanced Membrane Technology VII by an authorized administrator of ECI Digital Archives. For more information, please contactfranco@bepress.com.

Recommended CitationTzahi Y. Cath, Johan Vanneste, Michael B. Heeley, and Kerri L. Hickenbottom, "Salinity gradient energy: Assessment of pressureretarded osmosis and osmotic heat engines for energy generation from low-grade heat sources" in "Advanced Membrane TechnologyVII", Isabel C. Escobar, Professor, University of Kentucky, USA Jamie Hestekin, Associate Professor, University of Arkansas, USA Eds,ECI Symposium Series, (2016). http://dc.engconfintl.org/membrane_technology_vii/27

Salinity gradient energy: Assessment of pressure retarded osmosis andosmotic heat engines for energy generation from low-grade heat sources

Johan Vanneste, Kerri L. Hickenbottom, Tzahi Y. Cath

Colorado School of Mines

Dept. of Civil and Environmental Engineering

ECI, Advanced Membrane Technology VII

September 14th, 2016

Cork, Ireland

2

Structure of the Presentation

Energy

PressureRetardedOsmosis

Low-grade heat

Membrane Distillation

Output:

3

Structure of the Presentation

Energy

PressureRetardedOsmosis

Low-grade heat

Membrane Distillation

Output:

Part I: Pressure

retarded osmosis

(PRO): membranes

and spacers evaluation

Part II: Osmotic heat

engine (OHE): working

fluid selection

Part III: Techno-

economic assessment

of the OHE

4

Pressure Retarded Osmosis (PRO): Utilizing Energy from the Ocean

PX

TurbineDraw SolutionHigh Concentration

High Pressure

Feed SolutionLow Concentration

Low Pressure

Applied Pressure < Osmotic Pressure

Semi-permeable,

hydrophilic membrane

Power Density (W) = JwΔP (power output per membrane area)

Water Flux (Jw) = A(Δπ-ΔP) (simplistic!)

5

Open-Loop PRO

Applications

Seawater – River water

RO concentrate –freshwater

Dead Sea and Great Salt Lake

Limitations of open loop PRO

Extensive pretreatment needed

Potential irreversible membrane fouling

Solution chemistry and temperature are fixed

Source: Statkraft

6

LC HC

Jw

Wt

Wh,pro

Closed-Loop PRO: Osmotic heat engine (OHE) Controlled solution

temperature and chemistry

High osmotic pressures

yields high power densities

Lower reverse salt flux

Can utilize low-grade heat

Potential for energy

storage

Thermal

separation

Closed loop system =

No backwashing or

chemical cleaning!

HC = high concentration (DS)

LC = low concentration (DI water)

7

Membrane Distillation (MD): Utilizing LGH for Draw Solution and Feed Stream Regeneration

High Concentration

High Temperature

Low Concentration

Low TemperatureMicroporous, hydrophobic

membrane

Water Flux (Jw) = Aw(ΔPv*) (simplistic!)

Jw

Can utilize low-grade heat to simultaneously

separate and concentrate mixed streams

8

PRO MEMBRANE AND SPACER ASSESSMENT

9

PRO: An Energy Generating Membrane Process

PRO membranes are not

yet commercially

available and FO

membranes are used

PRO membranes must:

Exhibit high power density

& low reverse solute flux

Withstand high

operating pressures

PRO membrane spacers

must provide good mixing

and adequate support

LC HC

Active

Layer

Porous

Support

Js

Jw

CF,b

CF,m

CD,m

CD,b

ICP

C

EC

PD

HC = high concentration

LC = low-concentration

CD,b = bulk draw concentration

CD,m = membrane interface draw conc.

CF,b = bulk draw concentration

CF,m = membrane interface draw conc.

ECPdil = dilutive external CP

ICPC = concentrative external CP

10

Operating Conditions: PRO

Source water

Draw solution (DS):

1, 2, & 3 M NaCl

Feed: deionized water

Bench scale testing

SCADA system controls

temperatures, and collect data to

calculate water flux, batch recovery,

and salt rejection

Flat sheet FO membranes

Hydration Technologies Innovations

(HTI) thin-film composite (TFC)

HTI cellulose triacetate (CTA)

Oasys Water TFC

X company TFC

11

PRO Bench-Scale SystemMembrane module

PROzilla

Membrane spacers

Tricot (35-ch) Tricot (20-ch) Extruded mesh

12

PRO Membrane Evaluation:

0 100 200 300 400 500

0

1

2

3

4

5

6

7

8

9

10

0

1

2

3

4

5

6

7

8

9

10

0 1 2 3 4

DS hydraulic pressure, psi

Po

wer

den

sit

y, W

/m2

DS hydraulic pressure, MPa

X HTI CTA Oasys HTI TFC

0 100 200 300 400 500

0

2

4

6

8

10

12

14

0

2

4

6

8

10

12

14

16

0 1 2 3 4

DS hydraulic pressure, psi

Sp

ec

ific

re

ve

rse

so

lute

flu

x,

g/L

DS hydraulic pressure, MPa

X HTI CTA Oasys HTI TFC

1 M NaCl draw solution (DS); 2x – 20 channel tricot on feed✓ High power densities

✓ Low specific reverse solute flux

✓ Good mechanical stability

13

Spacer Orientation

Parallel to flow

45° to flow1 M

2 M

3 M

0 200 400 600

0

5

10

15

20

25

0

5

10

15

20

25

0 1 2 3 4 5

DS hydraulic pressure, psi

Po

wer

den

sit

y, W

/m2

DS hydraulic pressure, MPa

HTI TFC 3 M HTI TFC 2 M HTI TFC 1 M

Feed spacer orientations

Corrugated heat

exchanger plate

Spacerless patterned membranes!

48 % increase!

14

Membrane Deformation

Virgin

membrane

Membrane

after 450 psi

Surface

Cross-

section

Membrane

after 600 psi

0

2

4

6

8

10

12

14

1 2 3 4 5 6 7W

ate

r fl

ux

, L

m-2

hr-

1S

pe

cif

ic r

eve

rse

so

lute

flu

x, g

L-1

Elapsed time, hr

Js/JwJw

Importance of long

term experiments!!!

700 psi

15

OHE WORKING FLUID SELECTION

16

𝐽𝑤 = 𝐴 𝜋𝐷,𝑏 exp −

𝐽𝑤

𝑘 − 𝜋𝐹,𝑏 exp 𝐽𝑤

𝑆

𝐷

1 +𝐵

𝐽𝑤 exp 𝐽𝑤

𝑆

𝐷 − exp(−

𝐽𝑤

𝑘)

− ∆𝑃

Reverse Solute Flux in PRO and Implications for OHE Process Performance

Non-idealities in PRO gives rise to

reverse solute flux (RSF, Js):

To sustain osmotic driving force, must

bleed a portion of the PRO feed

stream to the MD feed

Impacts of RSF on net power outputs

and generation costs are unknown!

HC LCLC HC

Jw

Jw

Js

Wt

MDPRO

Legend:

π = osmotic driving force

k = mass transfer coefficient (DS side)

S = membrane structural parameter

B = solute permeability coefficient

D = draw solution diffusivity

𝐽𝑠 = 𝐵 𝜋𝐷,𝑏 exp −

𝐽𝑤

𝑘 − 𝜋𝐹,𝑏 exp 𝐽𝑤

𝑆

𝐷

1 +𝐵

𝐽𝑤 exp 𝐽𝑤

𝑆

𝐷 − exp(−

𝐽𝑤

𝑘)

17

Eight salts met screening criteria (π > πNaCl & non-hazardous)

Tested at concentrations equivalent to 17.4 MPa osmotic pressure

(osmotic pressure of 3 M NaCl)

Draw solution Cost Concentration

$/kg M

CaCl2 83 1.6

HCOONa 14 4.1

KBr 42 3.2

LiBr 121 2.2

LiCl 74 2.6

MgCl2 14 1.5

Na(C2H5COO) 38 4.1

NaCl 6 3.00

4

8

12

16

20

24

0

100

200

300

400

500

600

LiC

lL

iBr

CaC

l2M

gC

l2N

a(C

2H

5C

OO

)H

CO

ON

aN

aC

lK

Br

(NH

4)2

SO

4K

Cl

C2H

3N

aO

2 N

a3C

6H

5O

7N

a2S

O4

Mg

(CH

3C

OO

)2M

gS

O4

NH

4H

CO

3K

HC

O3

NaH

CO

3

PV

P, k

Pa

Os

mo

tic

pre

ss

ure

, M

Pa

Osmotic pressure PVP (b)

Draw Solution Screening and Selection

18

Operating Conditions: PRO

Source water

Draw solution: varied, 30 ºC

Feed: deionized water, 30 ºC

Membrane: HTI TFC

Draw solution hydraulic

pressure kept constant at 2

MPa (300 psi)

Membrane

cell

Feed

Feed

loop

Draw

loop

Control

system

Draw

TT

P P

PP

CC

Dosing

~~

Flow meter

Back pressure valve

Gear Pump

Peristaltic pump

Plunger pump

Legend

Ultrasonic sensor

Conductivity probe

Pressure transducer

Temperature probe

Heat exchanger

~

P

T

C

19

Operating Conditions: MD

Source water

Feed solution: varied

Distillate: deionized water

Stream temperatures

Feed solution: 55 ºC

Distillate: 25 ºC

Flat sheet, hydrophobic,

microporous (0.2 μm)

membranes from 3M

Flow meter

Gear Pump

Peristaltic pump

Scale

Stir plate

Legend

Ultrasonic sensor

Conductivity probe

Pressure transducer

Temperature probe

Heat exchanger

~

P

T

C

0

.

0

0

0

0

0.0000

Membrane

cell

Distillate

Distillate

loop

Feed

loop

Control

system

Feed

TT

T T

PP

CC

Dosing

20

PRO Performance

Difference in power densities is because of the difference in

diffusivity and solute permeability coefficient (B)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

0

2

4

6

8

10

12

14

16

18

20

Dif

fusiv

ity,

(10

-9)

m2/s

Po

wer

den

sit

y, W

/m2

Power densityDiffusivity

21

Higher diffusivities lead to higher reverse salt fluxes

Salts with high reverse salt fluxes can be detrimental to

system costs

PRO Performance

0

1

2

3

4

0

20

40

60

80

100

120

140

160

B,

L/m

2-h

Dif

fusiv

ity,

(10

-9)

m2/s

Revers

e s

olu

te f

lux,

g/m

2-h

Solute flux

Diffusivity

B

22

Distillate conductivity decreased over time, indicating 100%

rejection of salts

Water flux decreases with decreased partial vapor pressure

difference

MD Performance

0

2

4

6

8

10

12

14

16

0

5

10

15

20

25

30

35

40

ΔP

VP, kP

a

Wate

r fl

ux,

L/m

2-h

FluxΔPVP

23

Preliminary Economic Evaluation of Draw Solutions

Co,h,pro

Qo,h,pro

HC LC LC HC

MD PRO

Jw Jw

Js

Co,h,md

Qo,h,md

Co,l,md

Qo,l,md

Ci,h,md

Qi,h,md

Ci,l,md

Qi,l,md

Ci,l,pro

Qi,l,pro

Co,l,pro

Qo,l,pro

Buffer TANK

Co,h,t

Qo,h,t

Ci,h,t

Qi,h,t

Ci,b

Qi,b

Co,b

Qo,b

Qr,l,pro

Legend Q = flow C = concentration

HC = high concentration LC = low concentration

Jw = water flux Js = salt flux

Subscripts i = inlet

o = outlet r = return l = low concentration

h = high concentration p = permeate b = bleed

t = tank T = turbine

Symbols

Heat exchanger

Gear pump

Plunger pump

Turbine-generator

Pressure exchanger

Wh,pro

Wl,pro

Wl,md

Wh,md

WT

Ci,h,pro

Qi,h,pro

MD

-LC

Ta

nk

PR

O-L

C

Ta

nk

Co,h,pro

Qo,h,pro

HC LC LC HC

MD PRO

Jw Jw

Js

Co,h,md

Qo,h,md

Co,l,md

Qo,l,md

Ci,h,md

Qi,h,md

Ci,l,md

Qi,l,md

Ci,l,pro

Qi,l,pro

Co,l,pro

Qo,l,pro

Buffer TANK

Co,h,t

Qo,h,t

Ci,h,t

Qi,h,t

Ci,b

Qi,b

Co,b

Qo,b

Qr,l,pro

Legend Q = flow C = concentration

HC = high concentration LC = low concentration

Jw = water flux Js = salt flux

Subscripts i = inlet

o = outlet r = return l = low concentration

h = high concentration p = permeate b = bleed

t = tank T = turbine

Symbols

Heat exchanger

Gear pump

Plunger pump

Turbine-generator

Pressure exchanger

Wh,pro

Wl,pro

Wl,md

Wh,md

WT

Ci,h,pro

Qi,h,pro

MD

-LC

Ta

nk

PR

O-L

C

Ta

nk

Co,h,pro

Qo,h,pro

HC LC LC HC

MD PRO

Jw Jw

Js

Co,h,md

Qo,h,md

Co,l,md

Qo,l,md

Ci,h,md

Qi,h,md

Ci,l,md

Qi,l,md

Ci,l,pro

Qi,l,pro

Co,l,pro

Qo,l,pro

Buffer TANK

Co,h,t

Qo,h,t

Ci,h,t

Qi,h,t

Ci,b

Qi,b

Co,b

Qo,b

Qr,l,pro

Legend Q = flow C = concentration

HC = high concentration LC = low concentration

Jw = water flux Js = salt flux

Subscripts i = inlet

o = outlet r = return l = low concentration

h = high concentration p = permeate b = bleed

t = tank T = turbine

Symbols

Heat exchanger

Gear pump

Plunger pump

Turbine-generator

Pressure exchanger

Wh,pro

Wl,pro

Wl,md

Wh,md

WT

Ci,h,pro

Qi,h,pro

MD

-LC

Ta

nk

PR

O-L

C

Ta

nk

Co,h,pro

Qo,h,pro

HC LC LC HC

MD PRO

Jw Jw

Js

Co,h,md

Qo,h,md

Co,l,md

Qo,l,md

Ci,h,md

Qi,h,md

Ci,l,md

Qi,l,md

Ci,l,pro

Qi,l,pro

Co,l,pro

Qo,l,pro

Buffer TANK

Co,h,t

Qo,h,t

Ci,h,t

Qi,h,t

Ci,b

Qi,b

Co,b

Qo,b

Qr,l,pro

Legend Q = flow C = concentration

HC = high concentration LC = low concentration

Jw = water flux Js = salt flux

Subscripts i = inlet

o = outlet r = return l = low concentration

h = high concentration p = permeate b = bleed

t = tank T = turbine

Symbols

Heat exchanger

Gear pump

Plunger pump

Turbine-generator

Pressure exchanger

Wh,pro

Wl,pro

Wl,md

Wh,md

WT

Ci,h,pro

Qi,h,pro

MD

-LC

Ta

nk

PR

O-L

C

Ta

nk

Experimental PRO and MD

results were used in the

model

Design constraint: PRO feed

concentration of 4 g/L

(DS specific)

Specific membrane and

module costs for PRO and

MD were referenced from

RO and MF literature,

respectively

24

Impact of Draw Solutions on System Costing

CaCl2 and MgCl2 were the salts that performed best in MD and had the

lowest specific reverse solute fluxes

Electricity generation costs are higher than expected because of low

PRO operating pressures (low power densities)

0

1

2

3

4

5

6

7

8

9

Ele

ctr

icit

y g

en

era

tio

n

co

st,

$ / k

Wh

25

MD membranes are the highest costs

Salts with high RSF result in more bleeding, subsequently

decreasing net energy and increasing MD membrane area

System Costing and Net Energy (1 MW gross)

0

100

200

300

400

500

600

0

2

4

6

8

10

12

Net

po

wer,

kW

Ca

pit

al

co

st,

$ m

illi

on

Salt costs

PRO membrane costs

MD membrane costs (bleed only)

MD membrane costs (no bleed)

Net power

26

With addition of MgCl2, high water flux of NaCl can be

maintained while RSF drops significantly

Mixed draw solutions: best of both worlds?

0

10

20

30

40

50

60

70

80

90

0

2

4

6

8

10

12

14

16

0 20 40 60 80 100

RS

F, m

mo

l/m

2-h

r

Wa

ter

flu

x, L

/m2-h

r

Percent MgCl2-OP

Water flux Solute flux

27

OHE TECHNO-ECONOMIC ASSESSMENT

28

OHE System Model and Cost Assumptions

Base case scenario

2.5 MW (net power) system

Draw solution: 3 M NaCl

PRO operating pressure:

3.4 MPa (~500 psi)

MD temperatures:

70 °C feed, 30 °C distillate

Assumptions

Low grade heat and cooling is free

Membrane costs referenced from

commercially available membranes

PRO costs referenced from RO

MD costs referenced from MF

Equipment efficienciesPump efficiency 70 %Turbine efficiency 90 %Pressure exchanger efficiency 95 %Generator efficiency 95 %Heat exchanger efficiency 60 %

Data and other assumptionsPlant life 20 yrPlant availability 90 %PRO membrane replacement 10 %/yr

MD membrane replacement 10 %/yrInterest (discount) rate, I 8 %Inflation rate, n 3 %Amortization factor 0.1Labor cost 0.03 $/m3

Specific membrane costs*

PRO membrane element cost 11 $/m2

MD membrane element cost 24 $/m2

PRO membrane housing cost 17 $/m2

MD membrane housing cost 14 $/m2

*Assumed 35% mark-down from

quoted distributor cost

29

OHE Power Generation

Power outputs

Gross Power: 4.9 MW

Net Power: 2.5 MW

About 20 % of capital costs

due to bleeding

Process efficiency

Theoretical efficiency1 4%

System efficiency 0.1%

System costs: $0.48 per kWh

Benchmark <$0.20 per kWh

1Lin et al., ES&T 2014

Hickenbottom et. al, in Prep.

Civil work36%

Generation - PRO13%

Regeneration - MD (total)26%

Hydraulic system

4%

Control system

21%

Capital Cost $57 million

MD

26 % PRO

13 %

Civil work

36 %

Control

system 21 %

Hydraulic

system 4 %

Cost of labor and

spare82%

PRO membrane replaceme

nt 4%

MD membrane replaceme

nt14%

O&M Cost $72 million

MD 14 %

PRO 4%Labor 82%

30

0

10

20

30

40

50

60

70

80

90

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0 2 4 6 8 10

Po

we

r d

en

sit

y, W

/m2

Ge

ne

rati

on

co

sts

, $

/kW

h

PRO hydraulic pressure, MPa (psi)

Cost Sensitivity: PRO Power Density and solute permeability(B)

(290) (580) (870) (1,160)

0

5

10

15

20

25

30

35

40

45

50

0

1

2

3

4

5

6

7

8

0 2 4 6 8

Po

we

r d

en

sit

y, W

/m2

Ge

ne

rati

on

co

sts

, $

/kW

h

B, L/m2-hr

Generation cost - w/bleeding

PRO power density

31

Assumptions and Model Outputs: Ideal Case OHE

25 MW (net power) system

(economy of scale)

Draw solution: 3 M NaCl

PRO operating pressure:

7.6 MPa (1,100 psi),

corresponding to a PRO

power density of 76 W/m2

MD temperatures:

85 °C feed, 15 °C

distillate

OHE Sensitivities

Base Ideal Unit

System size

(net power) 2.5 25 MW

Electricity generation cost 0.48 0.10 $/kWh

PRO operating pressures 4 7.6 MPa

PRO power density 45 76 W/m2

PRO recoveries 15 40 %

MD feed (LGH)

temperature 70 85 (95) °C

MD distillate (cooling)

temperature 30 15 (5) °C

MD recoveries 6 30 %

MD water flux 27 38 L/m2-h

Membrane replacement 10 5 %/yr

32

$0.48

$0.32

$0.20

$0.16 $0.13 $0.13 $0.12

$0.10 $0.10 $0.10 $0.10 $0.08 $0.09

$0.07

Ge

ne

rati

on

co

sts

, $

pe

r k

Wh

Costs of Competing Energy Generation Technologies

*Figure adapted from US EIA’s Annual Energy Outlook 2012;

and ElectraTherm 2013

Stationary

Engines

Organic

Rankine Cycle

(ORC)

OHE

Conventional coal-fired power ($ 0.03 per kWh)

33

Conclusions

Energy

PressureRetardedOsmosis

Low-grade heat

Membrane Distillation

Output:

Changing spacer

orientation from

parallel to 45°

increased PRO

power densities

with 48 %

HTI TFC membranes are the

best commercially available

PRO membranes,

withstanding operating

pressures up to 500 psi and

attaining power densities up

to 22 W/m2

Use of CaCl2 and MgCl2 as

working fluids, can

decrease OHE electricity

generation costs by > 46%.

Mixed draw solutions with

low RSF have the potential

to further decrease costs.

Future improvements

to PRO membranes

and power densities (>

76 W/m2) could make

the OHE a competitive

renewable energy and

energy storage

technology.

34

The future of osmotic power and the OHE?

0.0

0.5

1.0

1.5

2.0

2.5

1980 1990 2000 2010 2020

$/m

3

Year

Greenlee et al., Water Ressarce 43 (2009) 2317– 2348.

RO Desalination Costs

35

Industry partners

Hydration Technologies

3M

Academic collaborators

Yale University

Funding agencies

ARPA-E

EPA-STAR

Research Group

Ryan Holloway

Mike Veres

Tani Cath

Estefani Bustos-Dena

William Porter

Katie Shumacher

Curtis Weller

Acknowledgements

36

Thank you!