05-2169 - Advanced Thermoelectric Power Generation...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Advanced Thermoelectric Power Generation Technology Development at JPL 3rd European Conference on Thermoelectrics 3rd September 2005 Nancy, France presented by T. Caillat J. Sakamoto, A. Jewell, J. Cheng, J. Paik, F. Gascoin, J. Snyder, R. Blair, C. -K. Huang, J. -P. Fleurial Jet Propulsion Laboratory/California Institute of Technology Outline State-of-practice Radioisotope Thermoelectric Generators 975K-300K skutterudite unicouples and multicouples development High-temperature materials and devices for operation up to 1275K Future work and conclusions Power Technology • Missions are long – Need power systems with >15 years life ~30 AU ~20 AU Neptune ~40 AU • Mass is at an absolute premium – Power systems with high specific power and scalability are needed Pluto nce ista ~6 AU D 1 W/m2 ~10 AU Uranus Saturn 2 W/m2 4 W/m2 ~1.5 AU 1 AU ~0.8 AU Earth Mars Jupiter 15 W/m2 50 W/m2 610 W/m2 S lar o rad Ir nce ia High efficiency radioisotope power sources 1373 W/m2 Venus • 3 orders of magnitude reduction in solar irradiance from Earth to Pluto • Nuclear power sources preferable 2200 W/m2 U.S. missions using radioisotopes power and/or heating sources Multi-Mission PbTe/TAGS conductively coupled RTG (MMRTG) N-Leg Fe Cold Cap PbTe P-Leg Fe Cold Cap TAGS PbSnTe Fe Cup Fe Cup Ni Hot Shoe MMRTG MMRTG couple Item/Con verter PbTe/TAGS MM RTG Hot side temperat re (K) u Cold s temperat (K) ide ure Converterefficiency (%) Systemefficiency (%)* T he rmal power (BOM)(Wth) The rmal effciency (%) i Electrcal power (BOM) (We) i 125.3 8 5.02 43.8 2.85 Numberof GPHSmodules Total PuO2 mass kg) ( Total system massestimate(kg) Specific power estimat (We/kg) e 823 483 7.6 6.4 2000 Spring-loaded TE converter General Purpose Heat Source RTG GPHS-RTG Performance Data Hot Shoe (Mo-Si) Power output-We 290 beginning of life 250 end of life 40,000 after launch 55.5 28 42.2 diameter 114 long 1270 566 PuO 2 SiGe 572 7,561 Operational life - hrs Weight-kg B-doped Si0.78Ge0.22 B-doped Si0.63Ge0.36 p-type leg P-doped Si0.78Ge0.22 P-doped Si0.63Ge0.36 n-type leg Cold Shoe Output voltage Dimensions Hot junction temperature-K Cold junction temperature-K Fuel Thermoelectric material Numbers of unicouples Mass of Pu-238-g GPHS SiGe unicouple Specific pow er - We/kg 5.1 Outline State-of-practice Radioisotope Thermoelectric Generators 975K-300K skutterudite unicouples and multicouples development High-temperature materials and devices for operation up to 1275K Future work and conclusions Segmented Thermoelectric Technology Uses improved high ZT materials Development initiated in 1991 and supported by ONR and DARPA Large T, high ZT -> high efficiency ZT 1.2 p-Bi0.2Sb 1.8Te3 n-PbTe 1.0 n-C oSb 3 0.8 p-CeFe Ru1Sb 12 3 Higher efficiency values compared to PbTe/TAGS 0.6 p-SiGe nBi2Te2.9Se0.1 Segmented unicouples development and integration into an advanced RTG Using a combination of state-of-the-art TE materials (Bi2Te3-based materials) and new, high ZT materials developed at JPL Skutterudites : CeFe3Ru1Sb12 and CoSb3 0.4 0.2 p-PbTe 0.0 200 300 400 500 600 700 800 900 1000 1100 1200 1300 Temperature (K) Current materials operation limited to ~ 1000K Higher average ZT values Higher material conversion efficiency 8.8 % for a 480-1000K temperature gradient Efficiency T -T = H C 1+ZT-1 TH T 1+ZT + C TH Skutterudite unicouples Skutterudite unicouple key technology gates Key technology gates • Developed improved low temperature skutterudite materials • TE materials synthesis and scale upprocessing • Low electrical contact resistance between TE segments and cold- and hot-shoes • Demonstrate unicouple performance though testing and modeling • Unicouple thermal-mechanical integrity • Lifetime and performance validation • Sublimation control • Stable thermoelectric properties Current Focus Demonstrating skutterudite materials and bond stability and determining degradation mechanisms in order to validate lifetime operation up to 1000K Material property measurements as a function of time and temperature Sublimation rates as a function of time, temperature, and environment Sublimation control: thin metal foil, aerogel, pressurized environment Interface diffusion studies Couple screening tests and optimization of unicouple configuration in anticipation of lifetime testing Design and fabrication of four couple modules to facilitate technology insertion into MMRTG High Power Thermoelectric Converter Technology Development for Future High Power Science Missions Segmented Thermoelectric Multicouple Converter (STMC) technology for 100 kWe class power systems Primary objective is technology development based on high performance advanced thermoelectric materials for future NASA missions 2x increase in conversion efficiency High rejection temperature (600-700K) Limit size of heat rejection system And minimize overall system mass STMC Scope focused on: JIMO 100kW Thermoelectric Space Power System Goals Power Conversion System design and Projected Performance Improvements using modeling Advanced TE Materials over SiGe Alloys used in RTGs Advanced TE materials evaluation and optimization STMC TE Technology Development Team Advanced TE Couple Array engineering • Jet Propulsion Laboratory • University of California at Davis development • Clemson University • Boeing/Rocketdyne Scale-up converter fabrication • Princeton University • Teledyne Energy Systems Planning for technology insertion • University of Michigan • Michigan State University • University of South Florida • Cornell University • University of Southern California • University of New Mexico ZT values for 1000K skutterudite baseline materials 1.2 p-Bi .2Sb1.8Te3 0 n-PbTe 1.0 n-CoSb 3 0.8 p-CeFeRu1Sb12 3 Skutterudites ZT 0.6 p-SiGe nBi2Te2.9Se0.1 0.4 0.2 p-PbTe 0.0 200 300 400 500 600 700 800 900 1000 1100 1200 1300 Temperature (K) Advanced Materials Thermoelectric Conversion Efficiency 18 17 Segmented Couple Conversion Efficiency (%) 16 15 14 STMC Baseline Materials with Bi 2 Te 3 Segment 975K-480K low temperature skutterudite STMC Baseline Materials 13 12 11 10 9 8 7 6 5 275 300 325 350 375 400 425 450 475 500 525 550 575 Cold Junction Temperature (K) MMRTG Thot = 811K Thot = 1273K LT Skutterudite Baseline Materials with Bi 2 Te 3 Segment LT Skutterudite Baseline Materials T hot = 973K GPHS-RTG T hot =1273K State-of-Practice Si 0.78 Ge 0.22 State-of-Practice PbTe/TAGS Synthesis and some properties for n-CoSb3 and Ce1Fe3Ru1Sb12 Demonstrated materials synthesis scalability to large quantities Melting (~1200C in glassy carbon crucibles) followed by ball milling in steel vials under Argon Hot pressing at temperatures between 600 and 700C, graphite dies, 20,000 psi Developed 100g batch process for n-type and p-type Overall process similar to state-of-practice thermoelectrics; powder metallurgy process easily scalable to larger quantities Properties N-type CoSb3 Uses Pd, Te (~ 1at% each) as dopants to optimize carrier concentration CTE: 9.1 x 10-6K Decomposition temperature: 878C Ce1Fe3Ru1Sb12 CTE: 12.1 x 10-6K Decomposition temperature: 830C Unicouples legs fabrication Developed uniaxial hot-pressing technique for segmented and un- segmented (skutterudite only) legs fabrication Powdered materials stacked on the top of each other Temperature optimized to achieve density close to theoretical value In graphite dies and under argon atmosphere With metallic diffusion barriers between the thermoelectric materials Metallic contacts at hot- and cold-side Low electrical resistance bonds (<5μ cm2) achieved negligible impact on overall unicouple performance n- Bi2Te2.85Se0.15 Metal contact N-type segmented leg p- Bi0.4Sb1.6Te3 Metal contact P-type segmented leg n- CoSb3 p- Ce 1Fe 3Ru1Sb12 Metal contact Metal contact Unicouple fabrication techniques developed to date Skutterudite unicouple fabricated by co-hot-pressing Skutterudite unicouples fabricated by diffusion bonding 1) Co-hot-pressing Hot-press first leg onto hot-shoe Hot-press second leg onto hotshoe to complete unicouple fabrication 2) Diffusion bonding technique Hot-press metallized legs first Diffusion bond legs to hot-shoe in a second step Encapsulating Skutterudite Unicouples in Aerogel Close contact between aerogel and the unicouples legs can be achieved using the casting process Thermoelectric module mock-up to test aerogel permeation How it was made •Block placed upside down in mold so that notches were facing down. •Liquid precursor (Sol) poured into mold. After several hours the liquid transformed into a free-standing gel •The entire assembly including the mold was placed in the autoclave and supercritically dried 500 μm aerogel Demonstrated aerogel permeation into narrow spacing down to ~ 10 μm Skutterudite unicouple performance testing Load ~ 3 lbs per leg Shielded heater Mo hot-shoe Thermocouple hole for measuring hot-side temperature Voltage measurement contact Machined aerogel sleeve Cu cold-shoe Cooling loop Thermocouple holes for measuring cold-side temperature for each leg Spring loaded unicouple test fixture for skutterudite unicouple performance testing JPL unicouple and performance life testing capabilities Up to 16 unicouples/multicouples can be simultaneously tested Tests can be conducted in a vacuum or inert atmosphere Life tests designed to monitor power output performance over time Thermal and electrical testing - Segmented unicouple 15 SEPT-03, Skutterudites Segmented Unicouple Performance Test at UNM-ISNPS Th = 973.1 + 3.9 K Tc = 301.7 + 7.1 K = 13.8% peak peak Efficiency Estimate in Test (%) 14 = 13.5% Heater Mo hot-shoe 13 N-leg P-leg 12 11 1st Sweep, 9:0 AM September 29, 2003 rcont, n = 256 μ -cm2, rcont, p = 364 μ -cm2 2nd Sweep, 6 PM September 29, 2003 2 2 rcont, n = 253 μ -cm , rcont, p = 220 μ -cm Cold-shoes Spring loaded pistons 10 6 eta-two-sweeps.epg 8 10 12 14 16 18 20 Load Current (A) Fully validated projected power performance on skutterudite and skutterudite/Bi2Te3 unicouples (corresponds to ~ 14% efficiency for 975K-300K T Independently confirmed at the University of New Mexico (for up to 1800 hrs of continuous testing) Thermoelectric Multicouple Enhancement Strategy Planned improvements to fabrication and performance of conductively coupled Thermoelectric SiGe couple stack developed for SP-100 Improved TE Materials (increase conversion efficiency up to 10%) Low contact resistance interconnects (From 35-50 to less than 25 μ .cm2 at 1275K) Refractory Aerogel for superior thermal insulation and ease of module/TCA assembly (no glass between couple legs or around legs) Module arrangement facilitates interconnection (all handled from exterior SP-100 Multicouple (8-couple series) of TCA) Thick compliant pads and graphite layers reduced/eliminated to reduce waste T from 30% to 5–10% (compliance achieved through structural engineering) TE couples arranged in modules to facilitate fabrication, assembly and ensure lifetime Detailed thermal/mechanical/fractural analysis for robust design that will survive fabrication, assembly and operation STMC Mechanical Design & Engineering • Key Structural Integrity Challenges - Coefficient of thermal expansion mismatches within TE device stack, and between stack and large heat exchangers - “Bowing” of thermoelectric legs due to large T - Surviving fabrication and assembly steps – and operation - Reviewed preliminary design of STMC and TCA - Key goal is to redistribute thermally induced stresses by selecting optimal materials combinations, element geometries 1.00 1 Mechanical Displacements (1/4 TCA model) Elastic Stress Model of STMC Stack Red Line – Heat Exchanger Held Rigidly Blue Line – Free to Expand but no Bending Purple Line – Free to expand and bend - Secondary objective is to minimize parasitic losses ( T across non-TE layers and fill factor thermal losses) • Mechanical tests of TE samples - 4-point bend data obtained on skutterudites - Testing and modeling of interface fracture toughness and development of fail-safe structures S0 j 0.5 UTS j S1 j UTS j S2 j 0 UTS j 0.5 1.00 1 0 0.78 1 2 3 4 5 t tj mm 6 7 8 9 10 11 10.17 Fraction of Ultimate Strain VS Distance in mm STMC Module STMC Module fabrication: fewer and simpler fabrication steps, scalable to mass production High Voltage Insulator Sub-Assemblies (1000V) (HVISA) and electrical interconnects Aerogel Insulation for leg thermal packaging and sublimation suppression Metallized Thermoelectric Legs Sub Assemblies (TELSAs) HVISAs Cold Side Heat Exchanger Sub-Assembly (HXSA) Hot Side Heat Exchanger Sub-Assembly (HXSA) (Refractory metal) Vaporizable Polymeric Egg-crates for alignment of all sub-assemblies prior to bonding Segmented Thermoelectric Multicouple Converter (STMC) 1000K STMC Technology Demonstration • Fabrication of several modules completed in May 2005 • Performance testing to be started in late June 2005 2 x 2 STMC aerogel filling demonstration 2 x 4 STMC Module 1000K – 425K Operation, 13W 2 x 2 STMC Module 1000K – 425K Operation, 5W Outline State-of-practice Radioisotope Thermoelectric Generators 975K-300K skutterudite unicouples and multicouples development High-temperature materials and devices for operation up to 1275K Future work and conclusions TE Device Configuration: Segmenting vs Cascading Segmented Thermoelectric Segmented constrained by constant current u = I/Qc u Constant Segmented TE Generator 975K p-CeFe4Sb12 n-CoSb3 I constant Qc constant u = I/Qc constant 675K 475K 300K p- -Zn4Sb3 p-Bi0.4 Sb1.6Te 3 n-Bi2Te2.95 Se0.05 Cascaded Thermoelectric Cascaded independent circuits for each stage Current different in each stage Heat different in each leg u optimized for each stage Load Cascaded TE Generators i1 R Load,1 Hot n n p p I different Qc different u = I/Qc adjustable i2 R Load,2 Cold I constant Qc different u = I/Qc adjustable STMC High Temperature TE Materials Effort High Temperature n-type La2Te3 published zT = 1.3 Needs to be reproduced High Temperature p-type Greatest development need Cu2Mo6Se8 zT = 0.6 Zintl Clathrates Skutterudites ph pc Half Heusler Clathrates Skutterudites Hot T i+ nh nc i+ Cold low Temperature n-type Skutterudite CoSb3 today zT = 0.8 ACo4Sb12 goal zT = 1.1 RL V Skutterudite CeFe4Sb12 today zT = 1.1 CeFe4Sb12 goal zT = 1.4 low Temperature p-type Best TE Materials to Date Advanced Materials Thermoelectric Conversion Efficiency 18 17 Segmented Couple Conversion Efficiency (%) 16 15 14 STMC Baseline Materials with Bi 2 Te 3 Segment STMC Baseline Materials 13 12 11 10 9 8 7 6 5 275 300 325 350 375 400 425 450 475 500 525 550 575 Cold Junction Temperature (K) MMRTG Thot = 811K Thot = 1273K LT Skutterudite Baseline Materials with Bi 2 Te 3 Segment LT Skutterudite Baseline Materials T hot = 973K GPHS-RTG T hot =1273K State-of-Practice Si 0.78 Ge 0.22 State-of-Practice PbTe/TAGS Lifetime performance demonstration elements THERMOELECTRIC PROPERTIES (FY04) Examples : • SiGe: dopant precipitation • Fine grained SiGe: grain growth • TAGS : compositional change Testing: • Coupons Impact: • Change in efficiency, P output Solution: • Composition, doping control MATERIALS SUBLIMATION (FY04) Examples : • SiGe: Si & Ge • TAGS, PbTe: Te, Se, Ag, Sb • Skutterudites : Sb Testing: • Coupons Impact: • A/l, porosity, contact resistance, mechanical failure Solution: • Encapsulation, coatings • More refractory materials THERMAL INSULATION (FY05-06) Examples : • Si in MFI Testing: • Unicouple Impact: • Shorting • Contamination Solution: • Processing and engineering control THERMO-MECHANICAL INTEGRITY (FY05-06) Testing: • Unicouple Impact: • Contact resistance • Mechanical failure Solution: • Device engineering/modeling LIFETIME MODEL FUEL DECAY Thermoelectric materials sublimation Sublimation is the loss of volatile species in TE materials at or near hot-junctions of unicouples Rate is temperature and material dependent (different chemical bonding) Sublimation can result in life time issues: Electrical Resistance Increase Cross-sectional area reduction over time for congruent sublimation Change in unicouple performance Dissociation of TE materials into different compounds Example CoSb3 -> CoSb2 -> CoSb as a result of Sb losses Electrical and Thermal shorting Sublimation products can condense on the cold-side of the unicouples and/or in the insulation between the legs, potentially forming electrical and thermal shorts High-temperature TE materials sublimation rates Uncoated TE material: beginning of life sublimation rate at operating temperature (g/cm2 hr) TAGS at 675K PbTe at 800K Low Temperature n-Skutterudites at 975K Low Temperature p-Skutterudites at 975K High-Temperature n-Skutterudites at 1175K High-Temperature p-Skutterudites at 1175K Chevrels (MxMo6Se8) at 1275K P-type Zintl at 1275K LaYbTex at 1275K SiGe at 1275K ~ 4.7 _ 10-2 ~ 1.1 ~ 2.15 ~ 1.40 ~ 1.62 ~ 1.05 ~ 3.66 ~ 4.04 ~ 2.11 ~ 4.80 10-1 10-2 10-3 10-2 10-2 10-3 10-3 10-4 10-5 BOL sublimation rates in dynamic vacuum do not meet requirements for any of the bare TE materials Sublimation control techniques/materials are required and have been successfully developed for state of practice TE materials for over 20 years of operation Sublimation rates for skutterudites Demonstrated BOL sublimation rates < 10-5 g/cm2hr both in vacuum (with metallic foil) and argon atmosphere (aerogel + 0.1 atm. Ar) ;comparable with other PbTe/TAGS at BOL I nitiated life testing to determine rates over long period of time Conclusions and Future Work Conclusions Developed first generation 1000K-300K skutterudite-based unicouples and multicouples with 14% efficiency demonstrated; initiated life testing Identified 1275K-1000K high-temperature materials for integration into 2nd generation of segmented devices with the potential for achieving 17% conversion efficiency Future work Continue life time studies for 1000K skutterudite materials, unicouples, and bond stability to validate lifetime operation up to 15 years Design and fabrication of four couple modules to facilitate technology insertion into advanced RTG Conduct performance testing of multicouple (Scheduled to start late June 2005) Further optimize 975K-375K skutterudite materials Continue development of high temperature materials (properties optimization, thermal stability and coatings development) Integration of high-temperature materials into high-T multicouples (up to 1300K) Acknowledgements NASA Exploration Systems and Science Missions Directorates for funding Collaborators: Caltech, Boeing/Rocketdyne, Teledyne Energy Systems, University of Michigan, Michigan State University, University of South Florida, University of California at Davis, Clemson University, Princeton, University, Cornell University, University of Southern California, University of New Mexico ...
View Full Document

This note was uploaded on 05/21/2010 for the course MS Thermoelec taught by Professor Snyder during the Spring '10 term at Caltech.

Ask a homework question - tutors are online