Microsoft Word - numero 15 articolo 4 D H Ni Un nzu AB (PB hu per wa rel the aci gas tem ma KE IN atte low obs into dev Th sin bet is c Am pro car A Design a High Tem icola Zulia niversity of Tri uliani@units. BSTRACT. H BI) polymer midification rformance. T ater managem ate to the d e main desig id membran s mixture si mperature di anagement. T EYWORDS. H NTRODUCTIO fuel-ce density sources ention becaus w temperatur stacles hinde olerance to C velopment of here are sever gle phase of w tween the fue considerably i mongst all typ omising. They rbon monoxid A and expe mperatu ani ieste, Departm it High Temper r and phosp n is not requ This is what ment proble development gn features a nes. Perform mulating a istribution a This could s High Tempe ON ell power-sou y, ergonomics s. Amongst a se of their hig e (< 100 °C) er their wides CO usually co f PEM fuel c ral advantages water need to el cell stack an increased ther pes of HT PE y can be oper de (CO) cont eriment ure PEM ment of Mechan rature Proto phoric acid, uired and CO t makes HT ems are avo t and charac nd the perfo mance curves typical steam analysis sugg simplify stack rature PEM urce system s and enviro all types of f gh power den ) and are co spread comm ontained in t ells that can s in operating o be consider nd the coolan reby allowing EM, the ones rated at temp tent up to 2% N. Zuli tal chara M fuel c nical Engineer on Exchange can be ope O content up T PEM very oided. Till n cterization o ormance cur s refer to the m reformer gest that cat k design and M Fuel Cells; can generate onmental com fuel cells Pro nsity (> 1 kW mmonly calle mercialization the reformate be operated g at higher te red; (ii) the co nt; (iii) waste h g fuel cells to based on pol perature betwe % in fuel can b iani, Frattura ed acteriza cell stac ring , via A. V e Membrane erated at tem p to 2% in attractive, a nowadays, fr f high temp rves of a 25 e stack oper output. The thode air cou d system BO Experiment e electrical p mpliance (low oton Exchang W/kg), simple ed Low Tem n. These inclu es fuels. To o above 100 °C emperatures: ooling system heat can be r use lower qu lybenzimidazo een 120 °C a be tolerated, a d Integrità Struttu ation of ck Valerio 10 34 e (HT PEM) mperature be the fuel can as low qualit rom experim perature stac cells HT PE rating with t e stack volt uld be used OP, thus incr tal analysis; P power with s w or no em ge Membran design and q mperature PE ude a compl overcome the C (High Tem (i) water man m is simplified ecovered as a uality reforme ole (PBI) and nd 180 °C, re affecting only urale, 15 (2011) f a 350 W 4127 Trieste ) fuel cell ba etween 120 n be tolerated ty reformed mental point cks. The aim EM stack bas two type of age distribu as coolant reasing syste PBI; Hydrog significantly missions) than ne (PEM) fue quick start. U M (LT PEM lex water an ese problems mperature PEM nagement can d due to the in a practical ene d hydrogen. d phosphoric eactants hum y marginally th ) 29-34; DOI: 10 W ased on pol °C and 180 d, affecting hydrogen c t of view, o m of this wo sed on PBI a fuels: pure h ution analysi leading to a m performa gen. improved: e n other con el cells have sually, PEM M) fuel cells. d heat mana s research has M fuel cells, n be simplifie ncreased temp ergy source; ( acid membra midification is he performan .3221/IGF-ESIS.1 lybenzimidaz 0 °C. Reacta only margin can be used nly few stud rk is to pres and phosph hydrogen an s and the st a better ther ance. fficiency, en nventional po drawn the m fuel cells wor Several techn agement and s focused on HT PEM) [1 ed because on perature grad (iv) CO tolera anes are the m not required nce. 15.04 29 zole ants nally and dies sent oric nd a tack rmal ergy ower most rk at nical the n the 1, 2]. nly a dient ance most and http://www.gruppofrattura.it mailto: nzuliani@units.it http://dx.medra.org/10.3221/IGF-ESIS.15.04&auth=true N. 30 In tem PB of ref sys exp the cell HT stac ST (se and the gra pat cha (se tor TE con for ope har exp T T Zuliani, Frattur literature the mperature me BI and phosph carbon mono formates cont stem [8]. Sev perimental an e single fuel c l, a non unif T PEM fuel c ck based on P TACK DESIGN he stack experim Cell Cel e Fig. 1). The d has a phosp e cathode and aphite plates, ttern and the annels. The c e Fig. 2). Be rque of 7 Nm Figure 1: E EST BED he stack transmit thermoc ntrolled using r the hydroge erated in con rdware, was perimental da T T ra ed Integrità St major resear embrane or th horic acid bas oxide. They fo taining up to eral literature nalysis of the cell system. T form tempera cell stack, thi PBI and phos N k design relie mental activity ltec P-1000 M e MEAs have phoric acid co d 1 mg/cm2 that can with e air cooling cells are stack lleville spring m. Exploded view k experimenta tters are used couples fitted g two mass fl en circuit. Lo nstant current used to cont ata. trutturale, 15 (20 rch effort on he single fuel sed single cell found that, fo o 2% of CO. e papers refe HT PEM fu This could be ature distribu s work prese sphoric acid m es on some f y [8]. The stac MEA, two bip e an active are ontent of mor on the anod hstand a max channels. Th ked and assem g washers we w of the single H ation was carr d to measure d on the lowe low controlle oad is contro t mode. A me trol the stack 011) 29-34; DO this kind of f cell system [ l operated wi or temperature Similar resul er to membra uel cell stack [ related to se tion, an incre ent the main membranes, f features that w ck, has 25 ce polar plates an ea of 50 cm2 re than 95 wt de [15]. The p ximum opera he reactant flo mbled betwee ere used at o HT PEM cell. ried out on a the reactants er and on th ers: a Sierra S olled by a Th easurement an k main opera OI: 10.3221/IGF-E fuel cells relat [3-7]. For exa ith pure hydro es above 160 lts were obta ane degradat [13, 14]. Stack everal factors eased electric design featur focusing on v were already ells fed in par nd the gasket and an averag t% in a PBI m plates were d ting temperat ow field has en two steel s one side of th Figu test bench, s s inlet pressur e upper part martTrack M hurlby Thand nd control sy ating paramet ESIS.15.04 te to the stud ample Korsga ogen and five °C, the cell c ained by the a tion issues [9 ks usually sho as: a non un cal resistivity. res and the p voltage and te tested with rallel. Each c ts which seal ge thickness o matrix. The pl designed and ture of 180 ° a multiple se square end pl he tie rod an ure 2: Picture o Mechanical D schematically res. The stac of the stack M100 for the a dar Instrume ystem, based o ters (reactant dy, developme aard et al. [7] e types of ref could operate authors of th 9-12], but on ow lower per niform distrib With the aim performance c emperature di good results ell contains a the active ar of 860 µm. T latinum cataly built by the °C [16]. Each erpentine pat lates, fixed w nd the stack w of the 25 cells Department of presented in k temperatur . Reactants fl air circuit and ents TTi LD on the Nation ts flows, tem ent and mode studied the p formates, con e with good p his paper for ly some stud rformance wh ution of reac m of experim curves of a 2 istribution. during a pre a commercial rea and the re The membran yst loading is authors, usin h plate contai ttern with fiv with eight all t was tightened HT PEM stac f the University Fig. 3. Two re is measured flow rates we d a Bronkhor 300 electron nal Instrumen mperature) an elling of the h performance ntaining up to performance w a single fuel dies relate to hen compare ctants inside e mental analyze 25 cells HT P evious single l PBI BASF F eactant manif ne is about 60 0.75 mg/cm ng Sigracet B ins the flow f ve square sec threaded tie r d to a maxim ck designed at t y of Trieste. Baumer pres d using 6 K- re measured rst El-Flow F nic load mod nts CompactR nd to acquire high of a o 5% with cell the d to each e an PEM cell Fuel folds 0 µm m2 on PP4 field ction rods mum the sure type and F201 dule, RIO the http://www.gruppofrattura.it http://dx.medra.org/10.3221/IGF-ESIS.15.04&auth=true gau Th con inp usi a A EX (O con cur typ unt Tem infr A Figure 3: Sche uge, 6-valve, 7- he CompactR nfigured with put, analog ou ng a NI cRIO Agema Therm XPERIMENT start u degrada °C usin CV) and purg ntrolled, using rves were coll pical steam re til the lowest mperature di frared camera Figure 4: A ematic represen -solenoid valve RIO hardware h modules de utput for mas O 9201 and N movison 570 in TAL PROCED up and shutd ation. During ng an externa ged with nitro g an on/off a lected operat eforming refo t stack cell v stribution wa was pointing : Polarization c ntation of the t e, 8-hydrogen c e, consisting edicated to sp ss flow contro NI cRIO 920 nfrared came DURE down proced g the start-up al hot air blow ogen, on both algorithm. Th ting the stack ormate: H2 56 voltage reach as observed o g the top side curves and pow Stack a [V ] N. Zuli test bed (1-air cylinder, 9-pre of the NI cR pecific tasks: ols and digita 6 analog inpu era. dure were im procedure th wer. During t h anode and c he control sys at 160°C wit 6.35%, CO 0. h about 0.3 V operating the e of the stack. wer density var and single cell 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 0 200 [ ] Single fuel 25 cells sta Single fuel 25 cells sta iani, Frattura ed filter, 2-air com ssure regulator RIO 9004 rea analog input al output for t ut modules. S mplemented in he fuel cell sta the shut-down cathode sides stem acted on th two differe .5%, CO2 43. V. Single cel stack with pu riations with lo mean tempera 400 [mA/cm2] cell - voltage ack - mean cell voltage cell - power density ack - mean cell voltage d Integrità Struttu mpressor, 3-flo r, 10-nitrogen al time contr t for voltage temperature c Stack tempera n order to a ack is heated n procedure, s. At normal o n the hot air ent fuels: pur .15 %. Stack lls voltages w ure hydrogen oad for the 25 ature: 160°C. F 600 800 0 50 10 15 20 25 30 35 40 urale, 15 (2011) owmeter, 4-pre cylinder, 11-th roller and the and current control. Singl ature distribu avoid water c up from roo the stack wa operation, the blower or on re hydrogen a voltage was were measure n at 400 mA/ cells stack and Fuel: H2. 0 00 50 00 50 00 50 00 [m W /c m 2 ] ) 29-34; DOI: 10 essure transmit hermocouple, 1 e NI cRIO 9 measurement le cell voltage tion has been condensation om temperatu as set in Open e stack mean n the cooling and a gas mix measured sta ed at 200 an cm2 current d d for a single fu .3221/IGF-ESIS.1 tter, 5- pressur 12-electronic lo 9104 chassis, ts, thermocou es were measu n measured u n and membr ure to at least n Circuit Vol temperature fan. Polariza xture simulatin arting from O nd 400 mA/c density load. uel cell [8]. 15.04 31 re oad). was uple ured using rane 120 ltage was ation ng a OCV cm2. The http://www.gruppofrattura.it http://dx.medra.org/10.3221/IGF-ESIS.15.04&auth=true N. 32 ST cell 0.4 dec mW the Th the 2], imp ST HY to inv An per obs     l F F Zuliani, Frattur TACK PERFO ig. 4 sho hydrogen fuel cell l voltage, cor 46 V mean ce creases as loa W/cm2 for th e stack (14 % his behaviour e end plates c cell voltages proving perfo TACK PERFO YDROGEN AN ig. 5 sho hydrogen of fuels a some extend vestigated. Fu n important as rformance of servation can stack voltag stack voltag Fig. 8; the ends cel at 400 mA/ Figure 5: Mean oad. Stack mea reformat 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 [V ] F F ra ed Integrità St RMANCE OP ows the stac n at 160°C. I [8] at the sam rresponding t ell voltage. A ad increases: f he stack (27% less). could be exp auses a mark s are lower ormance. RMANCE AN ND REFORM ows the mea n and the sim are 5% at 200 d lower than urthermore, du spect regardin f the entire s n be made: ge distribution ge distribution lls exhibit the /cm2, the cell n cell voltage a an temperature te (H2 56.35%, 200 [m H2 - mean cell volt Sim. reformate - m H2 - power density Sim reformate - po trutturale, 15 (20 PERATING W ck polarizatio n the same fi me operating to a power de s expected st for example o % less). At 0.5 plained referri ked stack temp at stack end ND VOLTAGE MATE FUELS an cell voltag mulated reform 0 mA/cm2 an n that with h ue to the CO ng the stack o tack. Fig. 6 s n patterns are n follows a p e lowest volta number 1 rea and power den e 160°C. Fuels , CO 0.5%, 43 400 600 mA/cm2] tage mean cell voltage y ower density 011) 29-34; DO WITH HYDR on curve and igure the pola conditions is ensity of 292 tack perform operating at 0 5 V power de ing to the stac perature non ds. Increasing E DISTRIBU S ge and mean mate at 160°C nd 9% at 400 hydrogen as O poisoning ef operation is th shows the sta e similar both parabolic patt ages; ach a very low nsity variation w s: H2 and simul .15 % CO2). 800 0 50 100 150 200 250 300 350 400 [m W /c m 2 ] OI: 10.3221/IGF-E OGEN the stack p arization curv s included. Th 2 mW/cm2. T mance is lower 0.6 V power ensity is 338 ck temperatur uniformity (s g stack load UTION OPER n power den C. Referring mA/cm2. Sta previously o ffect, perform he voltage no ack voltage d h for the hydro ern that is sim w voltage of 0 with lated Figu S ESIS.15.04 ower density ve and the po he stack has The maximum r than single density is 221 mW/cm2 for re distribution see Fig. 8). A tends to de RATING WITH nsity variation to Tab. 1 per ack performan observed in mance differen on-uniformity distribution a ogen and refo milar to that 0.326 V when ure 6: Voltage d Stack mean tem reformate 0.3 0.4 0.5 0.6 0.7 1 V y variation w wer density v a nominal po m reached sta cell perform 1 mW/cm2 fo r the single fu n: at low load s performanc ecrease temp H n with load rformance di nce related to [7, 8] where nce increases y, since the lo t 200 mA/cm ormate case; of the stack n the stack is distribution at 2 mperature 160 (H2 56.35%, C 5 9 1 Cell n H2 - 200 m Sim. reform H2 - 400 m Sim. reform with load ope variation with ower of 365 W ack power ha mance. Perform for the single fuel cell and 2 ds the heat di ces depend on perature non for the stack fferences bet o reformate o e a single fu as load incre owest voltage m2 and 400 m temperature fuelled with r 200 mA/cm2 a °C. Fuels: H2 a CO 0.5%, 43.15 13 17 21 number mA/cm2 mate - 200 mA/cm2 mA/cm2 mate - 400 mA/cm2 erating with p h load for a si W at 0.5 V m as been 433 W mance differe fuel cell and 292 mW/cm2 spersion thro n temperatur uniformity t k operating w tween the 2 ty operation are o uel cell has b ease [17]. cell will limit mA/cm2. Sev profile show reformate. and 400 mA/c and simulated 5 % CO2). 25 pure ingle mean W at ence 162 2 for ough e [1, thus with ypes only been t the veral wn in cm2. http://www.gruppofrattura.it http://dx.medra.org/10.3221/IGF-ESIS.15.04&auth=true Th to t cell Th effe low wit ST tem nam the cat C F Current densit [mA/cm2] 200 400 580 Table 1: P he voltage dist the temperatu ls could be re he voltage non fect is strongl west voltage d th reformate f TACK TEMPE ig. 7 and mean te central a mperature beh med T1, T2 a e stack middl thode air cool Figure 7: Th Current density F ty Me volt 0 0 0 Performance d tribution follo ure dependen elated to the d n-uniformity ly associated due to the co fuels stack tem ERATURE DI d Fig. 8 show emperature an and end sta haviour. It is and T3 (Fig. 7 le reaching 2 ling effect. In hermal image o y 400 mA/cm2 fuel: pur F ean cell tage [V] 0.63 0.56 0.50 differences at 2 ows a parabo nt performanc different reac becomes eve to cell tempe ombined effe mperature dis ISTRIBUTION w the temper nd 400 mA/ ack parts (ab s also possibl 7 and Fig. 9). 25°C between n fact, reactan of the 25 cells 2, mean stack te re hydrogen. Figure 9: Simpl N. Zuli hydrogen Mean cell pow density [mW/cm2] 126 223 292 200 and 400 mA olic pattern th ce of the sing tant distribut en more notic erature [17]. I ct of CO con stribution is c N rature distribu /cm2 current bout 20°C). le to observe . T1 temperat n T1 and T2 nt air enters th HTPEM stack emperature 16 lification schem iani, Frattura ed wer ] Stac pow [W 157 279 36 A/cm2 for the hat is similar t gle PBI/H3PO tion along the ceable when f In this case th ntent in the f crucial in orde ution along th density load The heat di e the temper ture is the low 2. This tempe he stack from k. 60°C, me of the air p °C d Integrità Struttu ck wer W] Me volt 7 0 9 0 5 stack operatin to the stack t O4 based fuel e stack. fuels containi he first stack fuel and low er to obtain b he stack ope d. There is a ispersion thr rature differen west compare erature behav m the T1 sectio Figure 8: Sta 400 mA/ path through th 60 80 100 120 140 160 180 200 0 C urale, 15 (2011) ean cell tage [V] 0.60 0.51 - ng with hydrog emperature p l cell. Voltage ing CO are us k cell (see red cell temperat best performa rating with p a marked tem rough the en nces amongs ed to T3 and viour could b on and leave ack temperatur /cm2, mean sta fuel: pure h he fuel cell stac 50 100 ) 29-34; DOI: 10 reformate Mean cell pow density [mW/cm2] 119 204 - gen and reform profile (see Fi e differences b sed. In fact th d arrow in Fig ture. In the c ance. pure hydrogen mperature dif nd plates co st the 3 secti T2. Differen be explained from the T3 re profile. Curr ck temperature hydrogen. ck. 0 150 mm .3221/IGF-ESIS.1 wer ] Stac pow [W 14 25 - mate as fuel. ig. 8). This is between adja he CO poison g. 6) presents case of opera n at 160°C s fference betw ould explain ions of the s nces are highe referring to (Fig. 9). rent density e 160 °C, 200 T1 T2 T3 15.04 33 ck wer W] 9 5 due cent ning s the ating tack ween this tack er at the http://www.gruppofrattura.it http://dx.medra.org/10.3221/IGF-ESIS.15.04&auth=true N. 34 Th tem sin sim com CO of Th rea stac des pur RE [1] [2] [3] [4] [5] [6] [7] [8] [9] [10 [11 [12 [13 [14 [15 [16 [17 I Zuliani, Frattur his simple inv mperature and gle blower c mplification co mplex balanc ONCLUSION n this pape The stack with reform a single fuel c he stack volta ach the best p ck temperatu sign improve rpose. EFERENCES J. Zhang, Z J. of Power Q. Li, J. O. D. F. Chedd K. Scott, S. O. Shamard M.K. Daleto A. R. Korsg R. Taccani, Application H. Tang, Z. 0] J. Hu, H. Zh ] Z. Qi, S. Bu 2] A.D. Modes 3] J. Scholta, M 4] S. J. Andrea 5] T. J. Schmid 6] Sigracet BPP 7] Q. F. Li, R. I ra ed Integrità St vestigation s d reach a bet could be used ould not easy e of plant. NS er the main d can deliver a mate stack pe cell system. A age and temp performance ure distributio ement could S Z. Xie, Y. Tan Sources, 160 Jensena, R. F die, N. D.H. M Pilditch, M. M dina, A. Chert ou, N. Gourd gaard, R. Refs , R. Radu, Z ns, EFC 09-17 . Qi, M. Rama hanga, Y. Zh uelte, J. of Pow stov, M.R. Ta M. Messersch asen, S. K. Kæ dt, J. Baurmei P Datasheet, H. He, J. A. trutturale, 15 (20 uggests the ter temperatu d for cathodi y apply to a lo design charact about 350 W erformance de As expected st perature distri particularly on noticeably lead to a sim ng, C. Song, T 0 (2006) 872. F. Savinell, N. Munroe, J. of Mamlouk, J. o tovich, A.A. K doupi, J.K. K shauge, M. P. Z. Nicola, A 7056 (2009). ani, J. F. Elte haia, G. Liua, B wer Sources, arasevich1, V midt, L. Jöris ær, Int. J. of H ister, J. of Po Eisenhuth G Gao, J. O. Je 011) 29-34; DO opportunity ure distributio ic air supply ow temperatu teristics and a at 0.5 V mea ecreases only tack performa ibution analy when operat suggesting th mpler BOP a T. Navessin, . J. Bjerrum, P f Power Sour of Appl. Elec Kulikovsky, A Kallitsis, J. of M Nielsen, M. A. Damnjano r, J. of Power B. Yia, Int. J. 161 (2006) 1 V.Ya. Filimono ssen, Ch. Har Hydrogen En ower Sources, GmbH, (2004) nsen, N. J. B OI: 10.3221/IGF-E of using ca on. Additiona y and coolant ure PEM fuel an experimen an cell voltag y slightly. This ance is lower ysis reveal tha ting with refo he opportuni as only a sin Z. Shi, D. So Progress in P rces, 160 (200 ctrochem., 37 A.R. Khokhlo Membrane Sc Banga, S. K. ovic, In: Pro r Sources, 158 of Hydrogen 126. ov, N.M. Zag rtnig, J. of Pow nergy, 33 (200 176 (2008) 4 ). jerrum. J. Ele ESIS.15.04 athodic air as ally, the BOP t purpose. It cell based sys ntal analysis o ge when oper s performanc than single c at temperatur ormates. It h ty to use it fo ngle blower c ong, H. Wang Polymer Scien 06) 215. 7 (2007) 1245. ov, Int. J. of H cience, 252 (2 Kæra, J. of P oceedings of 8 (2006) 1306 n Energy, 31 gudaeva, Elec wer Sources, 08) 4655. 428. ectrochem. So s coolant in P of the syste t is significan stem where h of a 25 cells H rated with pu ce behaviour h cell performan re manageme has been obse or temperatur could be use g, D. P. Wilki nce, 34 (2009) . Hydrogen En 005) 115. Power Sources the Europe 6. (2006) 1855. ctrochimica A 190 (2009) 8 oc., 150 (2003 order to m em can be sim nt to highligh humidification HT PEM stac ure hydrogen has been foun nce. ent has a key erved that ca re manageme d for air sup inson, Z. S. L 449. nergy, 30 (200 s, 162 (2006) an Fuel Cell Acta, 54 (2009 3. 3) 1600. manage the s mplified as on ht that this B n requires a m ck are presen . When oper nd similar to y role in orde athode air aff ent purpose. T pply and coo Liu, S. Holdcr 09) 1. 239. l Technology 9) 7121. tack nly a BOP more nted. rated that er to fects This oling roft, y & http://www.gruppofrattura.it http://dx.medra.org/10.3221/IGF-ESIS.15.04&auth=true