ÿWPC$  , gN~iµ[4KCÜîN±,¹7Y?wU0”çIîÇ8—>¬™SNð‹Kù)µìÉâ¤ì¶y#áÔ3Ÿë£?!øÌOW{²Ùø-·ì`Èž7ÌqP5æ* ¯9"Ct°?l׫Þf °¤Âûõ±<ÑÝ)ÞS¯ÚmO®¯K²?Hâ/B%mwNÚ#¼+&'Îl™ëÕZŽ^kSnBv+·kÞo0»m™ágš£ËO¬”„€9öäJdßÍÅÆiQ4Wr½Ò,˜%² ĺΉPÆ8ºŸ³'µ” ßÕ»yüÇŒ^Mɹü1®áªkŒÂ–t„÷Ë܉Š/ÍE p°ùEH±”{HûÎõûÀ%Èô5_‡H3¡2­ô8=Ž¢Ì¾’º=J ß÷Ô9Ʀ;LµÁӽа¾I ëhþy>BKmI\á“[¸EÙe~DF\°l{-1еõ++‡tBŸdqxü»4þ†19!®ÂO¡’•…Ü꾉˜ÄË¿BŒ¼Å%Ì9è‹={ÑgXÆg%ØY'wW½—ù:[lTð±œ@ÚèÑ@£"îËì¡øV}_M––ÿÎ"ÅùhQBíF|“§hªy8CbWI)aŽ'œ±n:(°Œh&"U:î %(Œ. 0[ºE î 0[`n­p 0@ B ]zðzj wú 4 <þ6X9`("Courier 10cpiX ‚Ý ƒ«¢'ÝÓ  ÓÓ  ÓÝ  ÝÒ„°ÒÒX°Òß?€ ' °°xdE„¯ü!?ßÐ ° Ðò òEMC€GD„034à Œ àà ä àà < àEMC€GUIDELINE€DOCUMENTà ô àà L àà ¤ àPAGE€Ú  Ú1Ú  ÚÐ ¿ Ðß?€ ' °°xdE„žü!?ßó ó(32«¢$§§Ý ƒ øÝ!ÝÓ  ÓÓ  ÓÝ  Ý äÝ ƒ«¢)ÝÓ  ÓÓ  ÓÝ  ÝÒ„°ÒÒX°Òß?€ ' °°xdE„¯ü!?ßÐ ° ÐPrepared€by€ò òEmission€Measurement€Centeró óà D àà œ àà ô àò òEMC€€GD„034ó óÐ ¿ ÐEMAD,€OAQPS,€EPAà ä àà < àà ” àà ì àà D àà œ àà ô àSeptember€28,€1994Ð Ÿï Ðß?€ ' °°xdE„rü!?ß(32«¢$©©Ý ƒ øÝ!ÝÓ  ÓÓ  ÓÝ  Ý3|x«<þ6X9`("Courier 10cpiXôxþ6X@ÉÇ;X@(øÝ$¡¡Ó  ÓÓ  Ó d æÝ ƒ«¢)ÝÓ  ÓÓ  ÓÝ  ÝÒ„°ÒÒX°Òß?€ ' °°xdE„¯ü!?ßÐ ° ÐPrepared€by€ò òEmission€Measurement€Centeró óà D àà œ àà ô àò òEMTIC€€GD„034ó óÐ ¿ ÐEMAD,€OAQPS,€EPAà ä àà < àà ” àà ì àà D àà œ àà ô àSeptember€28,€1994Ð Ÿï Ðß?€ ' °°xdE„rü!?ß †Ý ƒ«¢'ÝÓ  ÓÓ  ÓÝ  ÝÒ„°ÒÒX°Òß?€ ' °°xdE„¯ü!?ßÐ ° Ðò òEMTIC€GD„034à Œ àà ä àà < àEMTIC€GUIDELINE€DOCUMENTà ô àà L àà ¤ àPAGE€Ú  Ú1Ú  ÚÐ ¿ Ðß?€ ' °°xdE„žü!?ßó óô{‡$Ý ƒ øÝ!ÝÓ  ÓÓ  ÓÝ  ÝÑ,°ÑÒ„°ÒÒX°ÒññÖ€ÿÿÖññññÖ€ÿÿÖññÑ  ÑññÖ€ÿÿÖññññÖ€ ÿÿÖññß?€ ' °°$xdE„¯$ü!?ßò òÐ ° ÐÓ  ÓEMISSION€MEASUREMENTññ€TECHNICAL€INFORMATIONññ€CENTERÌà@@‚..JàGUIDELINE€DOCUMENTÌß?€ ' °°$xdE„’$ü!?ßÐ “ã ÐÌà@@‚..àEMTIC€WORKSHOP€BACKGROUND€PAPER€FOR€METHOD€26€AND€26Aó óÐ –æ ÐÌÌò òÓ  ÓDESCRIPTION€OF€METHODÐ  n Ðó óÌMethod€26€is€a€proportional€sampling€procedure€that€uses€midget€(Method€6Ïtype)€impingers€whereas€Method€26A€uses€large€(Method€5€type)€im‚pingers€andÏthe€sample€is€collected€isokinetic‚al‚ly.€Note€that€it€is€necessary€to€useÏMethod€26A€when€water€droplets€are€present,€such€as€after€a€scrubber,€andÏone€wants€to€account€for€the€halides€in€the€water.ÌÌBoth€sampling€trains€use€a€heated€teflon€matte€filter€to€remove€theÏparticulate€salts€prior€to€the€impingers.€€Both€proce‚dures€have€provisionsÏfor€using€alternative€filter€types€in€some€cases.€€In€the€halogen€andÏhalide€analysis€scheme€everything€up€to€and€including€the€filter€isÏdiscarded€and€not€part€of€the€sample€although€it€may€be€recovered€as€aÏMethod€5€particulate€sample€in€some€cases€for€Method€26A.ÌÌThe€first€set€of€impingers€in€both€trains€contain€a€dilute€0.1€N€Hòò2óóSOòò4Ð –æ Ðóósolution€and€the€back€set€of€impingers€contain€a€dilute€0.1€N€NaOHÐ jº Ðsolution.€€In€Method€26A€there€are€options€to€add€a€cyclone€when€heavyÏparticulate€matter€loadings€are€expected€or€to€keep€the€moisture€dropletsÏoff€the€filter,€and€an€addition‚al€impinger€if€a€lot€of€moisture€condensa…¼tion€is€anticipated.€€ÌÌA€pre„purge€of€the€probe€and€filter€with€stack€gas€is€re‚quired€in€MethodÏ26€to€avoid€a€significant€bias,€because€of€the€smaller€volumes€sampled,Ïresulting€from€sampling€clean€gas€initial‚ly.€€For€Method€26A,€if€theÏoptional€cyclone€is€used,€or€if€mois‚ture€conden‚sation€is€observed,€then€itÏis€necessary€to€conduct€a€post„test€purge€to€evaporate€the€halogens,Ïhalides,€and€the€condensed€moisture€before€the€filter€so€it€can€be€capturedÏand€recovered€in€the€impinger€solutions.€€Failure€to€do€this€dryingÏcompletely€results€in€significant€negative€bias.€ÌÌThe€separation€of€the€halides€from€the€halogens€is€accom‚plished€physi‚callyÏin€the€impingers.€€Halogens€(Clòò2óó€and€Bròò2óó)€have€a€low€solubility€in€anÐ ª$ú( Ðacidified€solution€and€hence€tend€to€not€be€captured€to€a€large€extent€inÏthe€first€set€of€impingers€containing€0.1€N€sulfuric€acid,€whereas€they€areÏcap‚tured€effec‚tively€in€the€0.1€N€NaOH€solution.€€Meanwhile,€the€halidesÏ(HCl,€HF,€and€HBr)€are€effec‚tively€cap‚tured€in€the€0.1N€Hòò2óóSOòò4óó€and€henceÐ ú'J#, Ðremoved€before€they€can€get€to€the€NaOH€solution.€€Analysis€is€accomplishedÏin€both€sets€of€solutions€by€ion€chromatography€looking€for€the€halide€ionsÏ(Clòò„óó,€Fòò„óó,€or€Bròò„óó)€using€standards€pre‚pared€in€the€appropriate€solutionÐ v*Æ%/ Ðmatrix.Ìâ âÐ ,n'1‘ Ðò òÔ& T Ôâ âDISCUSSION€OF€IMPLEMENTATION€PROBLEMS€OR€ISSUESÐ v ÐÌHClÌÌó óIt€is€important,€but€difficult€to€confirm,€that€any€con‚densed€moistureÔ'Tv¬ ÔÐ ö € Ðpresent€during€sampling€be€removed€completely€during€a€post€purge€of€theÏMethod€26A€train€to€avoid€a€signifi‚cant€negative€bias.€€As€little€as€oneÏdrop€moisture€is€sufficient€to€cause€a€significant€negative€bias€becauseÏof€the€solubility€of€halides€in€water.€ÌÌFrom€an€analysis€of€data€available€today,€it€is€also€likely€important€thatÏthe€probe€and€filter€temper‚ature€be€maintained€at€greater€than€250€ð$ðF€toÐ ÂL  Ðassure€transmission€of€at€least€HCl€(and€likely€HF)€through€the€probe€andÏfilter€in€the€gas€streams€with€moisture€as€little€as€2%€by€volume€to€avoidÏa€negative€bias.€Additional€data€are€being€collected€to€confirm€thisÏobserva‚tion.€€However,€if€volatile€halide€salts€such€as€ammonium€chlorideÏare€present,€there€is€an€increased€risk€of€a€significant€fraction€ofÏammonium€chloride€being€in€the€vapor€state,€passing€through€the€filter,€andÏbeing€captured€in€the€0.1€N€Hòò2óóSOòò4óó€solution.€€When€this€happens,€the€ammoniumÐ Ž Ðchloride€disassociates€in€the€solu‚tion,€the€Clòò„óó€is€measured,€and€reportedÐ bì Ðas€HCl.ÌÌAmmonium€chloride€(NHòò4óóCl)€in€the€vapor€state€also€disassoci‚ates€into€HClÐ Þh Ðand€NHòò3óó€at€higher€temperatures.€€The€exact€tempera‚ture€dependency€is€notÐ ²< Ðclear€at€this€point.€€Hence€one€theory€exists€that€choice€of€samplingÏlocation€can€become€critical€if€the€measurement€location€stack€temper‚atureÏis€much€hotter€than€the€stack€temperature€just€prior€to€exiting€the€stack.€ÏThe€theory€is€that€some€of€the€HCl€would€form€NHòò4óóCl€before€it€leaves€theÐ Œ Ðstack€and€hence€sampling€in€the€"wrong"€location€results€in€overestimatingÏemissions€of€HCl€from€the€process.€€ÌÌIf€the€filter€and€probe€temperature€is€much€greater€than€the€stackÏtemperature,€this€same€logic€indicates€that€a€positive€bias€can€beÏintroduced€in€this€case€too.€ÌÌConversely,€if€the€probe€and€filter€temperature€results€in€cooling€of€theÏgas€stream,€it€is€possible€to€introduce€a€nega‚tive€bias€too.ÌÌUse€of€glass€fiber€filters€is€not€allowed€especially€at€low€concentra‚tionsÏof€HCl€because€of€the€potential€for€reac‚tion€with€the€filter.€€In€thoseÏcases€where€higher€stack€tempera‚tures€preclude€the€use€of€teflon€filtersÏbecause€the€teflon€would€be€unstable,€quartz€filters€are€recommendedÏalthough€the€exact€effect€of€using€these€filters€is€not€clear€at€present.€Ï€Ìò òHFó óÐ + $, ÐÌHydrogen€fluoride€(HF)€likely€reacts€with€the€glass€probe€liner€to€formÏsilicon€tetrafluoride€(SiFòò4óó)€but€because€of€SiFòò4óó€vapor€pressure€at€250€ð$ðFÐ ž-('/ Ðwe€think€it€will€pass€through€the€filter€and€be€captured€in€the€0.1N€H2SO4Ð r.ü'0 Ðsolution€and€correctly€be€reported€as€HF.€€If,€however€SiF4€is€present€inÏthe€sample€matrix,€it€will€likely€be€a€positive€interference€by€passingÏthrough€the€filter€and€being€captured€in€the€impinger€solution.€€The€dataÏwe€have€at€present€is€too€limited€to€confirm€that€this€is€quantita‚tive;Ïhowever,€we€are€presently€collect‚ing€data€that€may€confirm€that€if€thereÏis€a€HF€reaction€with€the€probe,€€SiF4€will€still€be€cap‚tured€and€report‚ed.€ÏWe€are€not€col‚lecting€data€at€present€to€confirm€that€the€reac‚tion€actuallyÏtakes€place.ÌÌBoth€methods€allow€use€of€teflon€probe€liner,€filter€holder,€and€impingers.€ÏIt€is€possible€that€a€positive€HF€bias€can€be€introduced€from€the€heatingÏof€the€probe€and€filter.€€Some€degassing€of€HF€was€observed€from€the€heatedÏteflon€sample€line€in€another€study€currently€underway€to€evaluate€HFÏmonitors.€€Likely€this€can€be€minimized€by€precon‚ditioning€the€teflon€atÏthe€temperature€the€train€is€anticipated€to€see€during€the€test,€but€weÏhave€limited€data€to€define€how€effective€this€preconditioning€may€be.€€€ÌÌò òSimultaneous€Presence€of€HBr€and€Clòò2óóó óÐ Š ÐÌIf€HBr€and€Clòò2óó€are€present€at€the€same€time,€it€appears€that€the€HBr€andÐ >È ÐClòò2óó€values€will€be€under„reported€and€the€correspond‚ing€HCl€and€Bròò2óó€valuesÐ œ Ðwill€be€over„reported.€ÌÌò òAnalysis€€ó óÐ Ž ÐÌThere€was€a€recent€change€made€to€both€methods€requiring€that€sodium€thioƒ¼sulfate€be€added€to€the€0.1N€NaOH€solution€after€sample€recovery.€TheÏpurpose€of€this€requirement€is€to€assure€that€the€hypohalous€acid€ion€isÏreacted€to€form€a€second€Clòò„óó€(or€Bròò„óó)€for€each€Clòò2óó€(or€Bròò2óó)€molecule.€€InÐ ¾H Ðearlier€versions€of€the€methods,€the€assump‚tion€was€made€that€one€Clòò„óó€(orÐ ’ ÐBròò„óó)€and€one€HClOòò„óó€(or€HBrOòò„óó)€was€formed€for€each€Clòò2óó€(or€Bròò2óó)€present€in€theÐ fð Ðsample€matrix.€€This€assumption€was€not€always€precisely€correct.€€TheÏequations€were€therefore€changed€in€the€latest€version€of€the€methods€toÏavoid€duplicating€this€correc‚tion.€€