PAPER from 10th International Colloquium Fuels Conventional and Future Energy for automobiles  January 20‐22 2015 Technische Akademie Esslingen in Stuttgart/Ostfildern.                      Permission is given for storage of one copy in electronic means for reference purposes. Further  reproduction of any material is prohibited without prior written consent of Innospec Inc.  ©INNOSPEC INC 2015  All Rights reserved.   

Injectoor and Fuel F Sysstem Deposiits

Jim Barker Innospec Ltdd, Ellesmere Port, P UK Jacqueline R Reid Innospec Ltdd, Ellesmere Port, P UK

with w for example reduced sullphur whilst blending b nonreefinery produccts such as FA AME or ethano ol. The ou utcome of thiss is a rise of deeposits throug ghout the fuel sy ystem. In this paper p we willl be concerned d with the naarrow field of diesel and gaasoline system ms. Th he causes of diesel d fuel systtem deposits are a described in n Figure 1 and continue to bbe pertinent. One O recent ex xample has beeen vehicle prooblems as a reesult of diesel beeing adulterateed with vegetaable oil in Arg gentina [1]. Fu urther with dieesel injectors now being at pressures of 27 700 to 3000 baar the opportuunities for form mation of deeposits will increase. he purpose off this paper is tto add to the knowledge k Th reeported by ourrselves [2] andd others [3,4] in the

Abstract Early studiess of deposits inn injectors and d fuel systemss were limited to injector noozzle orifice deposits. The majority of thhese deposits were the resu ult of fuel decompositioon during the combustion process p and weere controlled byy deposit contrrol additives (DCA). ( Modeern diesel fuel syystems with hiigh pressure common c rail architecture aand the severee conditions itt applies to thee fuel, coupledd with the introoduction of hiighly refined fuels such ass Ultra Low Suulphur Diesel (ULSD) withh its different ssolubilising caapacity, has seeen an upsurgee in deposit forrmation at varrious points in n the fuel system. Furthher, historical nozzle depossits were organnic in nature wheereas the latesst deposits hav ve been shownn to be not onlyy complex inoorganic and orrganic mixturees but also, in thhe case of inteernal diesel inj njector depositts (IDID), layerred in nature. It is also the case c that deposits havee been found throughout t the system, especially onn the fuel filterrs. The resultss of such deposits are m manifested in field problem ms such as severe issuess with drivabillity, fuel consu umption, emissions, annd engine failuure. Problemss in injector systems are nnow also being observed wiith direct injection gasoline systemss. Although co onventional DCAs have ttheir place in combating c dep posits there arre also new genneration DCAss that have been developed to combat thesee deposits. Thiis paper discu usses the varieety and range off deposits founnd, their origin ns and the effectivenesss of some of thhe new DCA chemistries c inn treating thesee deposits.

Figure 1. Sources off Diesel Deposits.

stu udy of fuel deeposits and theeir causes. Diesel Depositss

n Introduction

Diesel deposits have been annd remain the subject of nu umerous studies to characteerise them and d understand th heir origin [2-5 5, 12-15,16-200]. The diesel fuel system in n a modern com mmon-rail enggine has seen deposits in vaarious parts off the injector ssystem and in filters [1 19,21]. A typiccal common ra rail schematic is shown in Fiigure 2.

The worldwiide legislative drive to reduce emissions continues with for example Tier 3 being g introduced inn d the Bharat the United Sttates, China V in China and regulations inn India. Theree are also regu ulations proposed or ccoming into effect e regardin ng marine and locomotive trransport. Thee result of this has seen fuell injector manuufacturers goiing to ever-hig gher pressuress and temperattures to attain these targets whilst the fueel manufacturerrs have had too re-engineer their t product

Th he majority off work carriedd out has centrred on the in njectors. The Industry I Standdard engine teest procedure (C CEC F-98-08) is designed too assess foulin ng of injector tip ps. Internal diesel injector ddeposits (IDID D) have been th he focus of reccent work and are regarded as the most im mportant probllem in the indu dustry at present. The types off IDID observed can be cateegorised as:  Fuel ageing deposits ts: tacky brown or black in colourr.

  

Carbbonaceous: bllack particulattes containingg arom matic graphitee/ene precurso ors. Mettal Carboxylattes: off white carboxylate saltss of sodium. Am mide lacquer deeposits: thin fiilms of deposiits, polyymeric in natuure linked to poor p quality loow mollecular weightt PIBSI speciees.

Th hese novel dep posit control aadditives havee also been sh hown to be efffective in clean aning up existiing deposits reesulting in a raapid restoratioon of power on n addition of th he additive afteer 32 hours. T This is shown in Figure 4.

Deposits are also found onn fuel filters an nd recent worrk has investigaated the insidee of the common-rail itself.

Figuree 2. A Typical Common-Rail C Schematic S

Diesel Injecttor Tip Deposits The introducction of the Inddustry Standarrd test procedure (C CEC F-98-08) to assess the injector i foulinng tendency of ffuels dosed with different deposit d controll additives hass led to the inttroduction of new n chemistriies to combat theese deposits effectively. e

Figure 3. Observed power data from DW10B keep cleann engine tests oon fuel containiing Zn neodecaanoate plus noveel deposit control additive.

Engine tests conducted usiing the CEC F-98-08 F DW10B engiine using RF-06-03 (European certificatiion test fuel) havve shown the effectiveness e of o these novell additives as sshown in Figuure 3. As per the CEC F-98808 test proceedure the fuel was w adulterateed with zinc neodecanoatee to give 1mg/kg of zinc in the fuel. A brief outline of the proceduure is given here. h The testt used the 1997cm3, 4-cylinder, turbo-chaarged, engine of the DI type w with a high preessure commo on-rail fuel system and µ µ-sac six-hole injectors. Th he engine was operated on tthe engine test bench accord ding to a test cycle consistting of 12 steaady state conditions to give a total cycle tim me of 3600 seeconds. The cy ycle was repeated to ggive a total test time of 32 hours.

Fiigure 4. Observ ved power data ffrom DW10B clean c up enginee tests on fuel con ntaining Zn neoodecanoate pluss novel deposit control addditive.

nternal Diesell Injector Depposits – Polym meric Amidee In Lacquers Po olymeric amid de lacquer depposits and theiir link to low molecular m weig ght polyisobute tene succinimiides (LMW PIIBSI) has been n investigatedd in a number of pu ublications [13 3-20]. In many ny cases these have in nvestigated an ill characterissed LMW PIB BSI of un nknown proveenance. Recent work [2] on these depposits saw eng gine testing caarried out on a well characteerised, lab preepared, nonco ommercial low w molecular w weight polyiso obutylene su uccinimide sam mple, (LMW PPIBSI (non-co ommercial)). Th he paper descrribed a non-coommercial LM MW PIBSI which w in the CE EC F-98-08 Peeugeot DW10 0B engine caaused injector sticking. A coommercial PIB BSI deposit co ontrol additivee (DCA) was ffound not to cause c sticking in n the same testt and also prevvented stickin ng when used in n addition to th he low molecuular weight material. The jaammed injecto or needles from m the test werre subject to a nu umber of analy ytical tests. Thhe presence of amide on th he needle surfaace was confirrmed by infra-red sp pectroscopy (F FTIR). Tiime of Flight Single Ion Maass Spectromeetry (TofSIIMS) showed,, by spectral ccomparison wiith the starting materiaal, that the LM MW PIBSI (no onco ommercial) waas present in a lower layer of o the in njector depositt. This is show wn in Figure 5 using the io on C4H2O2N- which w is comm mon to both th he LMW PIIBSI (non-com mmercial) andd the injector deposit. d

Figure 7. Infraared spectra froom varoius poin nts along the JFTOT tubee deposit. Figure 5. Toff SIMS Depth Profile P Study off „stuck“ injectoor from use of LMW PIBSI P (non-com mmercial).

vailable for No Industry Standard test is currently av IDID. CEN (Committee European E de Normalisation) N ) TC19/WG244 Injector Depposit Task Forcce and CEC (Coordinatinng European Council) C TDFG G-110 engine test committeees are workinng to develop an engine tesst for IDID cauused by LMW PIBSI. Although an engine test will w be the final arbiter in anyy studies there is also a needd for a bench test t to assist inn the understannding of deposits. The Jet Fuel F Thermal Oxidation Teester (JFTOT)) is the Industrry Standard qualification test for the thhermal stabilitty of aviation fuel. This tecchnique has beeen extended to t diesel fuel bby a number of groups [13-155 ] including ourselves o as ann investigativee bench test for IDID. Using mapping infra-red techhniques we haave extended the t degree thaat the test can innform regardiing the chemisstry of depositt formation. Thhe recent publication [2] deetailing enginee testing with a LMW PIBSI (non-commeercial) also noted work oon the lab preppared sample using u the JFTOT and a reference fueel. This resultted in deposition onn the JFTOT tube. t This sam mple will be used to illusttrate the use of infra-red miccroscopy as a characterisatiion tool for saamples generaated by the JFTOT benchh test. The inffra-red map off the deposit iss shown in Figgure 6.

Figuree 6. Infra-red Microscope Map of JFTOT Tube Deposit.

he spectra werre collected allong the tube and a Th traansition can be b seen in Figuure 7. At the beginning b of th he tube, amide is predominaant with a strong vibration att ~1670cm-1. The T further aloong the tube one o travels thee im mide of the PIB BSI comes to the fore with a band at ~1 1705cm-1. Thee amide is form med from the acids in the fu uel. It may be speculated thaat the two matterials deeposit on diffeerent parts of tthe tube becau use of diifferences in solubilty in thee fuel. Thus a bench test sh howed both an n imide and ann amide speciees exist in a LM MW PIBSI (n non-commerciial) deposit. This T agrees with w what was found f previouusly on the eng gine test geenerated needlle [2]. nternal Diesell Injector Depposits – Meta al In Carboxylates In nternal diesel injector i depossits (IDID) fro om sodium caarboxylate sou urces have alsoo been the sub bject of inveestigation. As for the polym meric amide laacquers, there is currently no Industry I Standdard engine teest method. In n the US, CRC C committee suub panel on ID DID (CRC DIESEL Perforrmance Groupp - Deposit Panel Bench/Rig/Inveestigation sub panel), and in n Europe, CE EN TC19/WG G24 Injector D Deposit Task Force, F and th he CEC TDFG G-110 panel, ar are working to develop a standard enginee test using a ““fuel soluble”” sodium salt an nd dodecenylssuccinic acid ((DDSA). Teests were cond ducted using tthe CEC F-98-08 DW10B en ngine to assesss the injector ssticking tendeency of fuels do osed with sodiium salts. The he tests were carried out ussing RF-06-03 3 (European ceertification test fuel) as reeceived and the fuel was nott adulterated with w zinc neeodecanoate. The engine teest procedure was w carried ou ut as described d above with aadditional mo onitoring of th he exhaust gas temperaturess. When the engine was started the exhaaust gas tempeeratures for eaach cylinder were w recorded during d 15 minnutes. The eng gine then co ompleted 8 tesst cycles follow wed by a 4 ho our soak peeriod after which the enginee was re-starteed and the ex xhaust gas tem mperatures forr each cylinderr recorded. Th his is referred to as 8 hour ddata. The eng gine then ran fo or a further 8 hours h of test cy cycles followed by a 4 hour so oak period beffore being re-sstarted and thee exhaust temperatures reecorded to givve 16 hour dataa. This co ontinued until 32 hours of teest cycles had d been

completed orr until the enggine failed to start, s which signified seriious injector sticking. s Previous worrk has shown the use of fueel soluble sodium 2-ethhylhexanoate in i conjunction n with dodecenylsucccinic acid (D DDSA) to cau use injector sticking afterr 8 cycles. Soodium 2-ethylh hexanoate alonne did not causee injector stickking. To aid solubility s in thhe fuel sodium 22-ethylhexanooate was added as a 10 %w//w solution in 2--ethylhexanoll. Further worrk was carriedd out to determ mine if commeercial deposit control c additives werre effective inn preventing th he formation oof sodium carbooxylate deposits within the injectors.

sticking in the presence p of soodium species added to the fu uel. A novel DCA D was addeed at 1X treat rate r to the fu uel containing 0.5 mg/kg Naa (as Na 2-eth hylhexanoate) an nd 10 mg/kg DDSA. D The teest ran for the full 32 hours without w signs of injector stickking. This is shown in Fiigure 10.

The change iin observed poower for each test is shown in Figure 8. Thhe sodium 2-etthylhexanoatee/DDSA test failed to startt following the first soak peeriod after 8 hours of runnning.

Figure 10. Exhaust gas temperratures from DW W10B engine ontaining Na 2-eethylhexanoate, DDSA and test on fuel co novel n deposit coontrol additive.

on-Rail Depossits Diesel Commo Th he work above has been lim mited to the injjector part of th he common-rail diesel fuel iinjection systeem. Prreliminary wo ork on analysiss of a diesel common-rail will w now be desscribed. . Figure 8. Obsserved power daata from DW10 0B engine tests on fuels conntaining Na 2-eethylhexanoate/DDSA, plus conventtional and novel deposit contro ol additives.

A commerciaally available conventional deposit contro rol additive com mmonly found in diesel fuel globally was added at 2X treat rate to thhe fuel contain ning 0.5 mg/kg kg Na (as Na 2-ethylhexanoatte) and 10 mg g/kg DDSA. The test ran ffor the full 322 hours withou ut signs of injector stickking. This is shown s in Figu ure 9.

In n a diesel fuel injection systtem one of thee most im mportant parts is the commoon-rail, a presssure accumulaator where the fuel is stored at high pressu ure which feeeds fuel to thee injector systtem. Though injectors i have beeen investigateed [10] until nnow the rail itself has not beeen studied. Here H we presennt preliminary y data deescribing the corrosion c that has occurred in a co ommon-rail in n field use. A common-rail from a mediuum duty dieseel engine which w had suffeered a failed innjector was reecovered from the field in n the USA. Itt had been sub bjected to use with w ULSD unttil injector faillure. The com mmon-rail was cu ut up and one side of the pippe removed to o investigate th he inner surfacce of the rail. T This was then n mounted on reesin and subjecct to SEM anaalysis as show wn in Figures 11 1, 12 and 13. Th he surface sho owed general ccorrosion acro oss the inner su urface of the pipe p with rust ddeposits seen across the su urface. There was also pittin ing observed to t a depth of 3 to 25 micronss.

Figure 9. Exhhaust gas temperratures from DW W10B engine teest on fuel coontaining Na 2--ethylhexanoatee, DDSA and coonventional depposit control add ditive.

Next generattion novel depposit control additives have also been shoown to be effeective in preveenting injectorr

fo ormed are activ ve sites for fuuel degradation n in the fuel sy ystem [11]. Th he corrosion m may be the ressult of water in ngress, biodiessel componentts [12] or disaarmament of th he corrosion in nhibitor additiv ive. Fu urther investig gations will bee reported in a future publiccation. Diesel Fuel Filter Deposits

Figurre 11. Inner Surrface Common-rail Pipe.

Fu uel filters are designed to reemove particlees from fuel an nd as such can n accumulate ssimilar deposiits to those ob bserved in the injectors or pprecursors to those t deeposits. Filterrs removed aft fter DW10B teesting of fuel do oped with zincc neodecanoatte showed a build up of bllack deposits. Filters taken from tests conducted with th he addition of novel depositt control additiives showed a reduction in th hese deposits as shown in Figure F 14. Clearly these no ovel deposit ccontrol additiv ves are venting not onnly injector deeposits but efffective in prev deeposits throughout the fuel ssystem.

25 microns

Figure 12. SE EM Micrographh of Common-rrail Inner Surfacce with pits 255 microns deep..

Figure 14. Fuel filters from D DW10B enginee test on fuel containing c Zn neodecanoate wiith and withoutt novel deposit control addditive.

Inside surface e of common rail

8 microns 10 microns

Figure 13. SE EM Micrographh of Common-rrail Inner Surfacce with pits 8-110 microns deep p.

The presencee of such corroosion is of interest from twoo aspects. One is that it show ws the corrosiv ve effect of ULSD on a m metal fuel systtem surface bu ut also the pitss

Th he application n of infra-red sspectroscopy to the an nalysis of diessel filter samplles has been carried c out forr many m years [22]. The adventt of diamond ATR A technology has made possiblle direct acquiistion of data from the filter surface s routinee. An examplle of a filter from a Europeaan diesel filterr spectrum is shown s in Fiigure 15. In n addition to bands associateed with diesell, the infrareed shows a dom minating bandd at 1565 cm-11 assigned to assymmetric CO OO stretch andd in conjunctio on with a COO symmetric stretch at 14446 cm-1, sugg gests the prresence of carb boxylate saltss. Although a very v useful technique, the caveat c that onlly the surface of the depos-it is analysed sh hould always bbe borne in mind. m

nalysis (EDAX X) did not shoow significantt amounts of an metals, m just tracces of molybddenum and callcium. The viisual morpholo ogy of the depposits would suggest s seeveral differen nt chemical sppecies being prresent.

Figure 15. Euuropean Field Diesel D Filter Dirrect Surface AT TR Speectrum.

The work onn diesel has shoown that altho ough injector deposits are iimportant, undderstanding an nd characterissing the depossits formed is analytically challenging. c Itt should also bbe noted that there t are otherr parts of the common-raill diesel system m which requirre investigatioon. Hence our prrevious work on o diesel filters [22], and thhe preliminary w work on the coommon-rail described d hereiin. Gasoline Injjector Deposiits It would be nnaïve to believve that depositts are restricteed to one fuel syystem alone annd there have been a numbeer of recent repoorts regardingg direct injectiion gasoline injector depoosits [5, 6]. Thhe drivers of vehicle v fuel economy connsumer satisfaaction and emiissions are toward directt pushing gasooline engine manufacturers m injection gasoline engine technology, t with w estimates oof market sharee of this type of o powertrain to t be around 20% of the w world market by b 2020. The number of papers and patents on depoosit formation n or its prevention haas also increased [5-8]. The majority of these have foocussed on injector design or o fuels, and studies of thee chemistry off the deposits at the tip and in the body of tthe injector. Here H we report initial work oon a gasoline dirrect injection injector which h had failed.

Figuree 17. SEM Imagge of Injector Hole. H

he injector holes where all bblocked to a certain c degree Th an nd an examplee is shown in FFigure 17. Ag gain, EDAX an nalyses showeed a lack of meetals being present.

Figure F 18. Injeector Needle. he injector needle, Figure 118, had deposiit present Th accross the ball and a shaft. In th this case, tracee amounts of lu ube oil constitu uent metals (ffor example molybdenum) m were w found. Deeposits were allso noted on the t injector sp pring but again n no significannt amounts off metals were prresent. In sum mmary, the depposit on the neeedle appears to o be different in i nature to deeposits found elsewhere, with w a more cry ystalline naturre being obserrved. The deeposits are maainly organic. In nitial data from m SEM and Innfra-red micro oscopy studiess haas shown the deposit d to be ccomplex, with h both fuel gu um, intermediates and carboonaceous depo osits present. Th his study will be fully descrribed in a futu ure pu ublication. Conclusions

Figure 16. SE EM Image of Gasoline G Direct Injection I Injecttor T Tip.

The SEM of the injector tip, t Figure 16, showed deposit arounnd the tip and the injector holes. h The anlaysis of thhe deposit by X-ray X fluoresccence elementtal

Th his work has shown s that thee bench JFTO OT test co oupled with in nfra-red spectrroscopic mapp ping techniques can be regarded aas a useful testt for un nderstanding IDID I formatioon.

The challenge of removing injector tip, IDID and fuel filter deposits has been met by the development of a new generation of DCAs for which successful engine test data is described. Initial investigations of a diesel common-rail has shown corrosion levels to be high with concommitant pitting of the surface producing active sites which have potential to act as fuel degradation sites. The current factors causing deposits in diesel systems are also to be found in other fuel systems such as gasoline. The analytical techniques developed for diesel systems are useful to characterise and understand these deposits. Further it would be naive to think that deposits in the future will be restricted to these two fuel systems. The challenges remain to the DCA producers to keep injectors clean, and to the analytical community to apply techniques to understand the deposits formed. Regarding the latter, work on the application of Focussed Ion-Beam Scanning Electron Microscopy (FIB-SEM), Transmission Electron Microscopy (TEM), Atomic Force Microscopy (AFM) and Raman Spectroscopy, to diesel deposits will be the subject of a future publication.

[11] Venkataraman, R. and Esar, S., Chemistry Central Journal 2008, 2:25 doi:10.1186/1752-153X-2-2 [12] Farzal, M.A., Haseeb, S.M.A., and Masjuki, H.H., Fuel Process Technology 2010, 91,1308-15. [13] Reid, J. and Barker, J., SAE Technical Paper 2013-01-2682, doi:10 4271/2013-01-2682. [14] Ullmann, J. and Stutzenberger, H., TAE Fuels 9th International Coloquium, January 2013. [15] Bohenke, H., Gaiol, H., Benoist, G., Guillo, S. et al., JSAE 20145121, 2014. [16] Ullmann, J., Geduldig, M., Stutzenberger, H., Caprotti, R. et al., Diesel Injector Deposits, TAE 7th International Colloquium Fuels, Esslingen, 2009. [17] Schwab, S., Bennett, J., Dell, S., Galante-Fox, J. et al., SAE Int. J. Fuels Lubr. 3(2):865-878, 2010, doi:10.4271/2010-01-2242. [18] Lacey, P., Gail, S., Kientz, J., Milovanovic, N. et al.,SAE Int. J. Fuels Lubr. 5(1):132-145, 2012, doi:10.4271/2011-01-1925.

References [1] Rzeznik, M., ASTM Indianapolis 2014. [2] Barker, J., Reid, J., Snape, C., Scurr, D. et al., SAE Technical Paper 2014, doi:10.4271/2014-01-2720. [3] Dallanegra, R. and Caprotti, R., SAE Technical Paper / Journal Article, 2014, doi:10.4271/2014-012728. [4] Barbour, R., Quigley, R., and Panesar, A., SAE Technical Paper, 2014, doi:10.4271/2014-01-2721. [5] Dearn, K., Xu, J., Ding, H., Xu, H. et al. SAE Int. J. Fuels Lubr. 7(3):2014, doi:10.4271/2014-01-2722. [6] Smith, S. and Imoehl, W., SAE Technical Paper 2013-01-2616, 2013, doi:10.4271/2013-01-2616. [7] Kinoshita, M., Saito, A., Matsushita, S., Shibata, H. et al., SAE Technical Paper 1999-01-3656, 1999, doi:10.4271/1999-01-3656. [8] Rogerson, J. S. and Dawson, J.E., EP2145940A1P. [9] Graupner, O., Mundt, M., Schutze, A., Jacobus, J. et al. US78901470B2. [10] Karamangil, M.I. and Taflan, R.A., Journal of The Energy Institute 2012, 85, 4, 209-219.

[19] Lacey, P., Gail, S., Kientz, J., Benoist, G. et al., SAE Int. J. Fuels Lubr. 5(3):1187-1198, 2012, doi:10.4271/2012-01-1693. [20] Barbour, R., Quigley, R., Panesar, A., Payne J. et al., TAE 9th International Colloquium Fuels, Esslingen,2013. [21] Barker, J., Richards, P., Goodwin, M., Wooler, J., SAE Technical Paper 2009-10-1877.