TDA Research Inc. • Wheat Ridge, CO 80033 • www.tda.com

Sorbents for Desulfurization of Natural Gas, LPG and Transportation

Fuels

Gökhan O. Alptekin, PhD

Sixth Annual SECA WorkshopPacific Grove, California

April 21, 2004

Introduction

Advances in fuel cell technologies have the potential to revolutionize the way the power is generated and distributed Fuel cells require an ample supply of high qualityfuelPipeline natural gas is the fuel of choice because of its abundance and well-developed supply infrastructureIn addition to naturally occurring H2S, chemical odorants made with sulfur-containing compounds are added to natural gas for leak detection

Common odorants include mercaptans, sulfides and thiolsThe total sulfur content of the pipeline gas averages about 4 ppmv but can be as high as 10-12 ppmv

TDAR e s e a r c h

Sulfur As an Electrocatalyst Poison

TDAR e s e a r c h

Sulfur compounds contaminate the catalysts used in fuel cell systems and degrade power generation performance

Traditionally sulfur removal is carried out with a two-step process:HDS of the organic sulfur compounds and subsequent H2S removalIt is not practical for small-scale residential units or for transportation systems

Source: Israelson, “ Results of Testing Various Natural Gas Desulfurization Sorbents”, J. of Materials

Engineering and Performance, Vol. 13 (3), June 2004

Source: De Wild, “The removal of sulfur-containing odorants from natural gas for PEMFC”, Proceedings of

Fuel Cell Seminar, p227, 2002

Project Objective

TDAR e s e a r c h

TDA Research, Inc. is developing a passive adsorbent for ambienttemperature natural gas desulfurization

Requirements of the sorbentHigh sulfur capacity (minimum replacement frequency, small size)Low costReducing total sulfur concentration to sub-ppm levels Inertness (no side reactions or chemisorption of hydrocarbons)Tolerance to possible natural gas contaminants (hydrocarbons, CO2, H2O)Ease of disposal (no toxicity, flammability, pyrophorocity)

Fuel

Water

Steam

Burner

Steam Reformer

HEX

WGS PROX

Air

PEMFC

Air

Fuel Cell Off Gas

DesulfurizerPre-

reformer

PEM System SOFC System

WaterSteam

SOFC

HEX

DesulfurizerPre-

reformerFuel

Outline

IntroductionOutlineNatural Gas Desulfurization

Experimental SetupPerformance Comparison with Other SorbentsParametric TestsSorbent RegenerationDemonstrations

LPG DesulfurizationSorbent Performance

Desulfurization of Transportation FuelsExperiments with Model FuelsBattlefield Fuels (i.e., JP-8)

Conclusions

TDAR e s e a r c h

Experimental Setup

TDAR e s e a r c h

A high pressure quartz reactor (with internal pressure stabilization) was used for bench-scale testing

All lines and system components were either quartz, Teflon or Silcosteel– to minimize sulfur interaction with system components

Con

cent

ratio

n,

ppm

v

Time, min

0.5

1.0

2.5

1.5

2.0

3.0

3.5

Sulfur Analysis

Two gas chromatographs with sulfur chemiluminescence detector (SCD) and flame photoionization detector FPD) were used to analyze organic sulfur speciesThe detection limit of the SCD and FPD were 1 and 50 ppbv, respectively

signal/noise ratio greater than 10Restek RTX-1 column was used for separation of the sulfur species

Component name

Retention time

H2S= 2.1 ppmvDMS= 1.8 ppmvTBM= 1.3 ppmvTHT= 1.6 ppmv

Odorants Tested Formula AcronymIsopropyl Mercaptan (CH3)2CHSH IPMTert-butyl Mercaptan (CH3)3CSH TBM

Dimethyl Sulfide CH3SCH3 DMSHydrogen Sulfide H2S H2S

Tetrahydrothiophene CH2CH2CH2CHS THTEthyl Mercaptan EMC2H5SH

TDAR e s e a r c h

Natural Gas Analysis

Natural gas composition during sorbent testingComponent Volume % Volume %

Sample Jan. 2004 Sample June 2004Methane 92.83 92.39Ethane 3.30 3.42Propane 0.61 0.56Butane 0.13 0.11Isobutane 0.12 0.12Pentane 0.10 0.11Isopentane 0.10 0.10Neopentane 0.10 0.10Hexane 250 ppm 280 ppmCarbon Dioxide 0.70 0.81Nitrogen 2.01 2.08

In selected tests, the natural gas contained up to 200 ppmv water vapor

TDAR e s e a r c h

Two gas chromatographs with a TCD and FID detectors were used toanalyze natural gas componentsThe FID detector could measure as low as 1 ppmv hexane in natural gas

Typical Test Profile

0

10

20

30

40

50

60

70

0 1000 2000 3000 4000 5000 6000Run Time, min

DM

S C

onco

ncen

trat

ion,

ppm

v

Feed Analysis Feed Analysis

DMS Breakthrough

T= 22oC, P= 50 psig, DMS Inlet= 60 ppmv, GHSV= 1,500 h-1

TDAR e s e a r c h

Prior Work

Prod. Name

Source: Israelson, “ Results of Testing Various Natural Gas Desulfurization Sorbents”, J. of Materials Engineering and Performance, Vol. 13 (3), June 2004

M anufacturer Active Com ponent Product Nam e

DM S (ppm v)

Sulfur (ppm v)

DM S Index

United Cata lysts, Inc. (Süd-Chem ie)

ZnO @ 350°C (No preceding CoMo catalyst bed or hydrogen addition) G -72E 1.5 6.7 668

Norit Carbon w. chrom ium & copper salts RG M-3 0.8 5.1 24000

Calgon Carbon Carbon PCB 1.5 6.2 26550

United Cata lysts, Inc. (Süd-Chem ie) Carbon w. copper oxide C8-7-01 1.8 6.7 27900

Süd-Chem ie N ickel, N ickel O xide (Unheated) C28 1.1 4.2 44992

G race Davison

Molecular S ieve; 13X (10 Å pore) zeolite-X 544HP 0.8 4.4 110376

Supplier C Unknown Proprie tary Adsorbent S 1.2 5.1 182573

Pacific Northwest National Lab (PNNL)

Copper im pregnated zeolite-Y substrate Unnam ed 0.8 4.1 2416800

Synetix Copper O xide & Z inc Oxide 100°C top, 170°C bottom

Puraspec 2084 3.4 8.4 3661800

Supplier E Unknown Proprie tary Adsorbent T 2.3 8.2 3814300

DMS Index = average ppm DMS x cubic feet of natural gas/cubic feet of adsorbent

TDAR e s e a r c h

Performance Comparison at Bench-scale

0

2

4

6

8

10

12

0 100 200 300 400 500 600 700 800 900 1000Time (min)

DM

S C

once

ntra

tion

(ppm

v)

TDA's SulfaTrap-BR3

Siemens Sample 5

Siemens Sample 4

Grace X Zeolite

Norit RGM3, Activated Carbon

TDAR e s e a r c h

T= 20oC, P= 3 psig, DMS = 12.3 ppmv, TBM = 9 ppmv, THT= 9 ppmv in Nat. Gas GHSV= 60,000 h-1

TDA’s sorbent showed the highest sulfur adsorption capacity 60% higher sulfur capacity than Siemens Sample #5

Norit RGM3 Activated Carbon

Grace Zeolite X

Siemens Sample 4

Siemens Sample 5

TDA SulfaTrap-

BR3

0.36%Grace X Zeolite0.18%Norit RGM3 Activated Carbon

0.85%Siemens Sample 41.96%Siemens Sample 53.12%TDA’s SulfaTrapTM

Pre-Breakthrough Capacity (% wt.)

Sample

0.36%Grace X Zeolite0.18%Norit RGM3 Activated Carbon

0.85%Siemens Sample 41.96%Siemens Sample 53.12%TDA’s SulfaTrapTM

Pre-Breakthrough Capacity (% wt.)

Sample

Hydrocarbon Adsorption

TDAR e s e a r c h

The sorbent does not adsorb any hydrocarbon species from the natural gas

Even the heavy hydrocarbons such as hexane were not removed

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 500 1000 1500 2000 2500 3000Run Time, min

Con

cent

ratio

n, %

vol

.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Propane Iso Butane N-butane Iso-PentaneNeo-Pentane Pentane Hexane Ethane

Etha

ne C

once

ntra

tion,

% v

ol.

Ethane--->

<---Propane

<---Hexane

<---C4s&C5s

Potential Side Reactions

T= 20oC, P= 17 psia, DMS = 17.0 ppmv, TBM= 7 ppmv, THT= 5 ppmv, GHSV= 60,000 h-1

02.2

4.46.6

8.80

120

240

360

480600

Rel

ativ

e In

tens

ity

GC Test Time (min)

Test Time ...

Feed Analysis

Breakthrough

DMS

B-Merc

Test Start

TDAR e s e a r c h

The organosulfur species do not undergo any side reactions that cause formation of complex sulfur species

TDAR e s e a r c h

Effect of Water Content in Natural GasT= 20°C, P= 3 psig, DMS Inlet= 12.4 ppm, TBM= 9 ppm, THT= 9 ppm in Natural Gas, GHSV= 60,000 h-1

0

2

4

6

8

10

12

14

0 100 200 300 400 500 600 700 800Time (min)

DM

S C

once

ntra

tion

(ppm

v)

Norit RGM3, Activated Carbon Grace Zeolite X 2%Cu 4%Fe/MFI (Seimens)

Dashed Lines - Tested in natural gas with 45 ppmv waterSolid Line - Tested in dry natural gas

SulfaTrap-R1 (tested at Siemens)

Pipeline gas may contain up to 155 ppmv of water vaporSulfaTrapTM-R1 sorbent was impacted the least by presence of water

Cyclic CapacityT= 20oC, P= 3 psig, DMS = 7 ppmv, TBM= 7 ppmv, THT= 15 ppmv, GHSV= 75,000 h-1

T Adsorption= 20oC T Regeneration = 300oC

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

0 2 4 6 8 10 12Cycle

Bre

akth

roug

h C

apac

ity

Hydrogen RegenerationNatural Gas Regeneration

TDAR e s e a r c h

Sorbent maintains its capacity for 10 adsorption/regeneration cyclesSorbent regenerates using clean natural gas as well as hydrogen

DMS Breakthrough Profiles in the 10-Cycle Test

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

0 50 100 150 200 250 300 350 400 450 500

Time (min)

DM

S C

once

ntra

tion

(ppm

v)

Cycle 1, Hydrogen RegenerationCycle 2, Natural Gas RegenerationCycle 3, Natural Gas RegenerationCycle 4, Natural Gas RegenerationCycle 5, Natural Gas RegenerationCycle 6, Natural Gas RegenerationCycle 7, Natural Gas RegenerationCycle 8, Natural Gas RegenerationCycle 9, Hydrogen RegenerationCycle 10, Hydrogen Regeneration

T= 20oC, P= 3 psig, DMS = 7 ppmv, TBM= 7 ppmv, THT= 15 ppmv, GHSV= 75,000 h-1

TDAR e s e a r c h

Temperature-Programmed Desorption

0

20

40

60

80

100

120

140

0 50 100 150 200 250

Time (min)

Con

cent

ratio

n (p

pm)

0

50

100

150

200

250

300

350

400

Tem

pera

ture

(°C

)

DMS

TMB

THTTemperature

The sorbent DMS interaction is the weakest as evident by low desorption temperature Close sulfur balance between the adsorption and regeneration indicates full regeneration potential for the sorbent TDA

R e s e a r c h

5 kWe SOFC Alpha Testing at SWPC

TDA supplied its sorbent to Siemens Westinghouse to provide natural gas desulfurization during alpha testing of their SOFCsSorbent bed was sized to provide 1 year continuous operation

2.2 L sorbent (assuming 10 ppmv sulfur content for the gas all of which is DMS)1/16” cylindrical pellets

2,700 hrs testing was completed in March 2005The same canister will be used for additional tests

TDAR e s e a r c h

Alpha Testing Results

0

2

4

6

8

10

12

14

16

18

29-Nov-04 19-Dec-04 8-Jan-05 28-Jan-05 17-Feb-05 9-Mar-05 29-Mar-05Test Duration

Fuel

Flo

w, s

lpm

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Fuel Flow Total Sulfur, ppmv

Tota

l Sul

fur C

once

ntra

tion,

ppm

v

0

2

4

6

8

10

12

14

16

18

29-Nov-04 19-Dec-04 8-Jan-05 28-Jan-05 17-Feb-05 9-Mar-05 29-Mar-05Date

Fuel

Flo

w R

ate,

slp

m

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0Fuel Flow, slpmDMS, ppmv

DM

S C

once

ntra

tion,

ppm

v

TDA’s SulfaTrapTM-R1 sorbent removed all sulfur for over 2,700 hrsThe average concentration of the sulfur species in Pittsburgh pipeline gas were as follows:

Variation in sulfur concentration Variation in DMS concentration

*Data provided by Gordon Israelson, SWPC

H2S EM DMS TBM THT NPM Total Sppmv ppmv ppmv ppmv ppmv ppmv ppmv0.10 0.07 0.45 0.59 0.61 0.52 2.39

TDAR e s e a r c h

Water Vapor in Natural Gas

Natural Gas Water Vapor

0200400600800

100012001400

17-F

eb-0

5

22-F

eb-0

5

27-F

eb-0

5

4-M

ar-0

5

9-M

ar-0

5

14-M

ar-0

5

19-M

ar-0

5

24-M

ar-0

5

29-M

ar-0

5

3-A

pr-0

5

8-A

pr-0

5

Date

Wat

er V

apor

, ppm

v

SWPC measurements showed that natural gas used during the testing contained ~500 ppmv water vapor

*Data provided by Gordon Israelson, SWPC

TDAR e s e a r c h

Field Demonstration with a 5 kWe SOFC

A 5 kWe SWPC SOFC combined with TDA’s desulfurizer will be delivered to CTC Fuel Cell Test Facility in Johnstown, Pennsylvania

Desulfurizer dimensions 3” x 21”The fuel cell will be delivered to US Army base at Fort Meade, MD

TDAR e s e a r c h

Field Demonstration with a 125 kWe SOFC

TDA is also preparing a desulfurizer unit for a 125 kWe CHP SOFC system

Sorbent size = 150 L1/8”cylindrical pellets

TDA scaled-up the production capacity two-folds

The sorbent produced with high throughput equipment (e.g., spray dryer, screw extruder) exhibited good performance

The unit will be shipped to Hanover, Germany

TDAR e s e a r c h

Desulfurization of LPG

TDAR e s e a r c h

Liquified petroleum gas (LPG) has higher power density than natural gas on volume basis LPG-fed systems are suitable fuel for portable systems and applications in remote locationsMercaptans (mostly ethyl, n-propyl and isopropyl mercaptans) are the primary sulfur species in LPG

The presence of unsaturated hydrocarbons affects the performance of the desulfurization sorbents

0 60 120 180 240 300 360 420 480

GC Time, sec

AAA PropaneIndustrial Grade Propane

Met

hane

Etha

ne

Prop

ane/

Prop

ylen

e

But

ylen

e

Iso-

But

ane

n-B

utan

e

Sorbent Performance for LPG Desulfurization

TDAR e s e a r c h

0 2000 4000 6000 8000Time (min)

Eth

yl D

isul

fide

(Rel

ativ

e In

tens

ity)

C2H5SH C4H10S2

Ethyl Mercaptan Diethyl disulfide

TDA’s SulfaTrapTM-P sorbent achieved sulfur capacity of 0.63% and 2.35% wt. at GHSV of 30,000h-1 and 5,000h-1, respectively (based on diethyl disulfide breakthrough)

GHSV = 30,000 h-1

13 ppmv EM

GHSV = 5,000 h-1

25 ppmv EM

T=20°C, P=5 psig, GHSV= 5,000-30,000 h-1, Ethyl Mercaptan Inlet = 13 ppmv–25 ppmv

Field Demonstration with a 100 We SOFC

TDA will supply MesoscopicDevices with its LPG desulfurization sorbent75-100 We portable devices with up to 10 day continuous operation6 units will be developed for propane-fed SOFC by June 2005 and shipped to the Army for field testingUltimate goal is to develop an effective sorbent for transportation fuels (gasoline and diesel) and logistics fuels (e.g. JP-8 fuel)

TDAR e s e a r c h

TDAR e s e a r c h

0

200

400

600

800

1000

1200

1400

1600

1800

2000

0 50 100 150 200 250 300 350 400Time, min

Ben

zoth

ioph

ene

Con

cent

ratio

n, p

pmv

Cycle #1 Cycle #2

Cycle # Benzothiophene Capacity Total Sulfur Capacity(% wt.) (% wt.)

1 0.35 0.622 0.33 0.61

A Regenerable Sulfur Sorbent for Liquid Fuels

Benzothiophene breakthrough earlier than 2-methyl benzothiopheneThe sorbent was regenerated in hydrogen with a mild temperature swing

T= 60oC, LHSV= 1 h-1, 1,100 ppmw Benzothiophene, 900 ppmw 2-methyl Benzothiophene in a model fuel (70% aliphatic HCs, 30% aromatics)

Mo

S

4,6-dimethyl dibenzothiophene

In the refractory sulfur molecules, the sulfur atom is sterically hindered and difficult to activate

TDAR e s e a r c h

Sulfur Removal From JP-8

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 50 100 150 200 250 300 350

Time (min)

Per

cent

Rem

aini

ng S

ulfu

r

t= 60 min

t= 90 min

JP-8 Feed Analysis

TDA’s sorbent can remove all the sulfur in the JP-8 fuelSulfur speciation and capacity calculations are underway

T=60°C, Sulfur Content= ~3,000 ppmw, LHSV= 1 h-1 (Flow = 0.12 ml/min, 3.0 g sample)

Conclusions

TDA's SulfaTrapTM Sorbent Activated CarbonOperating Temperature Ambient AmbientBed Volume 1-2 L 30-50 LHydrocarbon Adsorption Minimal SubstantialEnd-of-Life Indication Yes NoRegenerability Yes NoFlamability No YesDisposability Easy Difficult

small volume toxic, pyrophoric

TDA’s sorbent can be effectively and economically used for natural gas desulfurization

Sorbent cost is estimated ~$10-25/lb (depending on production scale)Based on the conditions of the alpha test (~14 slpm fuel flow, 2.4 ppmv sulfur), sorbent cost is estimated as $4.71/1,000 m3 natural gas or $16 to $40/year ($3-$8/kW)TDA’s SulfaTrapTM-P sorbent can achieve 2.35% wt. capacity for desulfurization of LPGTDA’s sorbent also shows promise for JP-8 desulfurization

TDAR e s e a r c h

Acknowledgements

Funding for the research is provided by DOE Phase II SBIR Grant, Contract No.DE-FG02-03ER83795 and DOD Phase I SBIR Project, Contract No. W6HZV-05-C-0121

Authors thank:Donald Collins, NETL, DOE Magda Rivera, NETL, DOEKevin Mills, US Army TACOMGordon Israelson, Siemens WestinghouseJoe Poshusta, Mesoscopic Devices, LLC

TDAR e s e a r c h