28 - TU Dresden · 28 page I GASIFICATION-POST ... continuous stirred real reaktor plug flow reactor reactor Pe=O O

28

page I

GASIFICATION-POST-COMBUSTION OF WASTE MATERIALS;

INFLUENCING PARAMETERS AND ON-UNE MONITORING OF

ORGANIC SUBSTANCES

Bcckmann, M.I) and Zimmermann, R2)

I}Clall~(hakr Ulllwl.'irtt.'chnik-lnstiltH Gmbll (CUTEC), ClaUSlhal.Zclterfeld. Gerll1an~

2) In~ti(ll! für Ö~ologisch(' Chellli~1GSF.Forschungszr:ntnllll für UJllwch und Gölllldheit, Ob('r5chlt.~ißh:?il1l. GCfmany

I\bstract

Gasilication (Ir W;)511..' materials Oll grate systems with inclcpcndcnd post-combustion is ;:1 !lewi)' ck'\'C'lorec1rroccssing 1'01' thcrrnnl \\'<1stc lrentmclH. Thc proposccl paper sho\\'~ iJl"cstigatiolls of innucncing parametersconccrning dccoll1position of organic trace componcl1ts in the post-colllbuslion proccss at a pilot plant inconjullctiOIl wilh thc ::lpplication oran on-line analysis mCllioc\. The applicd an:1lysis Illc!hocl is n combination cfIascr indllccd Resonancc-[nlianccd ~llllii Photon loni~alion (RE~IPl) and '1 imc-of-Flighl "lass Spcclromclry(TOf.\1S) ",hieh rCpröCl11:-:: Clu:-i~rul 1001 rür the c!ercctioll ortracc qU31ltilics ofürganic subs!anccs.

Contents

1 IntrorIuction. .......................................................•.................................................................................. 2

2 Gasi ficat iOIl- I'ost-( ombustioll Process ...........................................•.

2.1 GCl1eml i\SpcCls .

2.2 Pilol Plant..

. 2

. 2

. 3

3 Occomposilioll 01' Organic Trace COlllpolinds in thc Post-ComiJustiOIl Proccss 5

3.] Variation of InOucllcing Parameters... . 5

3.2 Resldls and DisC1I55iol1 5

4 Hesonance-Enhanced 'lultipholon Jonization Timc-of-Flighl "ass Specrromeicr 6

4.1 Experimental ~lelhod 6

4.2 Resldls and Di5cussion .

5 S u nl Tl13 ry a 11d Fu rt her I n"estiga [ion s 8

6 Ac" n 0" Ied g e m c n IS 9

7 \' om eu cla r u re ............................................................•......................................................................................... 9

8 R e fe re n ces " , 9

page 1

GASIFICATION-POST-COMBUSTION OF WASTE MATERIALS;

INFLUENCING PARAMETERS AND ON-UNE MONITORING OF

ORGANIC SUBSTANCES

Beckmann, M.I) and Zimmermann, R.2)

I) Clausthaler Umwelttechnik-lnstilUt GmbH (CUTEC), Clausthal-Zellerfeld, Gennany

2) Institut rur Ökologische Chemie, GSF-Forschungszentrum für Umwelt und Gesundheit, Oberschleißheim, Gennany

Abstract

Gasifieation of waste materials on grate systems with independend post-eombustion is a newly developedproeessing for thermal waste treatment. The proposed paper shows investigations of influeneing parameterseoneeming deeomposition of organie traee eomponents in the post-eombustion proeess at a pilot plant ineonjunetion with the applieation of an on-line analysis method. The applied analysis method is a eombination oflaser indueed Resonanee-Enhaneed Multi Photon lonization (REMPI) and Time-of-Flighl Mass Speetrometry(TOFMS) whieh represents a useful 1001 for the deteetion of trace quantities of organie substanees.

Contents

I In trodu etio n .........•.............................................................................................................................................. 2

2 G asi !ication - Post -Com b ustion Proeess 2

2.1 General Aspeets.................................................................................. .2

2.2 Pilot Plant . 3

3 Deeomposition of Organie Traee Compounds in the

3.1 Variation oflnflueneing Parameters .

3.2 Results and Diseussion .

Post-Combustion Proeess .......•..•...•..•....•.................. 5

.. 5

. 5

4 Resonanee-Enhaneed 'Iultiphoton Ionization Time-of-Flight Mass Speetrometer 6

4.1 Experimental Method..... .6

4.2 Results and Diseussion...... . 8

5 S u m m a ry an d Fu rt h er In vestiga tio ns 8

6 Aeknowledgemen ts 9

7 omen cl a tu re 9

8 Ref erenees .........................................•.•.•............................................................................................................. 9

page 2

1 IntroductionPlants for thermal waste treatment may basically be subdivided into two parts. The first part is the so-ca liedthermal main process, with stoker firing systems being frequently applied. The second part includes alladditional treatment steps as for example purification of exhaust gases, post-treatment of ash or generation ofelectrical power. With regard to the reduction of pollution concentrations, in the past the emphasis was on thedevelopment of an efficient purification of tlue gas and the post-treatment of residuals (i.e. secondarymeasures). Currently equipped plants now generally comply to the legally required limits conceming thedischarge or the disposal of noxious malter in air, water and soi!.Further reduction of stack emissions is expected by optimizing thc thennal main process by primary measures.Among investigations at research facilities, pilot plants etc. and mathematical modeling, on-line detection ofinorganic as weil as organic pollutants is an important prerequisite. While the on-line detection of NO, SO, orCO belonged for a long time to the state of the art the on-line detection of organic trace components has still tobe developed. As far as speed and chemical preparation of the sampie are concemed the conventional GC-MShas reached its limits of performance.The proposed paper shows the application of an on-line analysis method for organic trace components inconjunction with the development and optimization of a multi-staged process which consists of

• gasification with air on a grate followed by• an independent post-combustion ofthe gases generated [1,2].

The applied analysis method is a combination of laser induced Resonance-Enhanced Multi Photon Jonization(REMPI) [3] and Time-of-Flight Mass Spectrometry (TOFMS) which represents a useful tool for the detectionof trace quantities of organic substances [4-6].The application as an environmental analysis method will be discussed with reference to test results at a pilotplant (thermal capacity 0.5 MW) using the above-mentioned multi-staged gasification/combustion process.Special emphasis lays here on the investigation of the independent post-combustion of the gases generated inthe grate process conceming organic trace compounds. The pattem of organic substances in the tlue gasdepending on the combustion conditions (temperature, residence time and mixing conditions) can be detected bythe REMPI- TOFMS. So this leads to a better understanding of pollutant fomlation and decompositionmechanisms and beyond this an on-line optimization of the process will be possible.

2 Gasification-Post-Combustion Process

Figure 1. Maininfluencialparameters forthermaltreatment[21.

I additional substances I :~::::~~al(~~~torboonding of poIlutants. controlling the melting behaVlOUr)

bed (e.g. f1uidized bed. solid bed, cifculallng bed. bindlng matrix. ash feedback)

Sir, oxygen. (nitrogen), carbon dioxide, steam. recirculated flue gas. ete.

low temperature until high temperature } drylng,~ '"600 ·e ~> 1ooo·e ':::: oe :,~:cation,

low press ure atmospheric pressure high pressure eva~bOr'I,pcO.1MPa p.O,1MPa p»Q,lMPa sublimallOn

continuous stirred real reaktor plug flow reactorreactor Pe=O O<Pe<CXl Pe-+oo

feed momentum, swirt, atomization, disperse, etc.eg f()(waSlllmatenals,reactlooga"addlllonalsubstanoll'S

composition (problematic substances)

combustion",

very longhours up to days

gaseous, liquid, pasty, solid (dustyorcoarse)

gasification1< 1

longminules up to hours

short,,,,,,m",

physical prOllerties

chemical properties

thennolysls.1.=0

pressure

temperature

reaction gas

oxygen supply

residence time

waste materials

reactor behaviour

I method of material input I

2.1 General AspectsA concept of thermaltreatment of waste shouldguarantee a low level ofpollutants. Before secondarymeasures are expandedfurther, the main thermalprocesses must be optimizedabove all by primarymeasures. The processconditions are chietlydetermined by the level ofmain intluencing parametersC1l1Ll). The distribution ofthese parameters along thereaction path is importantlogether with the differentreaction steps.

page 3

From the teehnieal point of view of the eombustion proeess, eoarse waste materials are more diffieult to treatthan regular gaseous, liquid and powdered fuels. For very different sueeessive tasks, like:

• eonversion of solid matter and• eombustion of gases generated,

the separation into different proeess units provides opportunities for individually optimizing eaeh task.Depending on the supply of oxygen or air the eonversion of solid matter ean be run as a pyrolysis, gasifieationor eombustion. The so called classical process of waste treatment is the combustion-post-combustion proeesswith grate systems. With regard to optimization an additional leeway arises if the first process is operated as agasification. A almost complete conversion of organic solid matter is possible both at gasification and atcombustion (avoiding of pyrolysis eoke). The gasification leads among others 10 a combustible gas whichmakes an independent post-eombustion without additional fuel possible in comparison to the combustion.Furthennore primary measures to reduce pollutants can be applied [e.g. 7-10]. The way of conducting theproeesses

• gasifieation ofsolid matter on the grate wirh air (1..8"'°.4 10 0.8),• post-combustion ofthe gases almost stoichiometrically (1..'0<'" 1.2 to 1.4) and• heat exchange

separately, is shown schematically in fig. 2. There are many possibilities available with grate systems forcontrolling the main acting variables [e.g. I]. In contrast to shaft fumaces, rotary kilns and tluidized bedreactors, grate systems enable a eontrol of reaetion steps (drying, degassing, gasifieation, bum-out) over thereaetion path. Gaseous, liquid and powdered fuels are mostly bumed in eombustion ehambers.This gasification-post-eombustion eoneept, currently examined on a test-size scale, appears promising, as, ineomparison to the eonventional ineineration proeessing in grate systems,

• the tlue gas mass tlows are signifieantly redueed,• eombustible gases whieh enable an independent post-eombustion process are generated.• the post-eombustion proeess itself ean be optimized regardless of the process on the grate with the

help of familiar primary measures for redueing the NO,·emissions and at the same time achieving highbum-out results,

• emission loads can be redueed eonsiderably [2, 11, 12].As mentioned before here the optimization of the post-combustion, especially the minimization of organiceompounds and the on-line monitoring ofthis trace compounds are ofspecial interest.

Figure2. Blockflowdiagramwithindependentunitsof thermaltreatmentforcoarse waste (1).

2.2 Pilot Plant

The tlow diagram of thepilot plant, including thescheme of thecorresponding processmeasuring and controlequipment, is presentedin .!l&J.. The plantconsists of the maincomponents• 5-zone reverse acting

grate or, altematively,3-zone advancinggrate,

• combustion chambersystem,

• heat exchanger and• tlue gas purification

unit.

additionaltuel

(llneeessaryl

wastematerialsadditive

(ifneccessary)

reactiongas (air)

1) of'flake of solid rE!Sldues not shown.2) coarse and pasty waste.3) gaseous, liqid and finely diVlded wasle

exhaustgas

page 4

compres-sed air

natural gas

model tuel

air

nitrogenloxygen

airairair

Figure 3. Measuring and control flow chart of the pilot plant (reverse-acting grate) [2).

dust

air

flue gas

remnants

Within the framework set here, the essential aspects are briefly outlined. Complete details on plant technology,data logging system and analysis are reported elsewhere [11]'

The model combustible solid or waste material is supplied in intervals lO the first grate zone via a feeding ramlocated at the lower end of the feeding hopper. The reaction gas (usually air; recycling of flue gas or oxygenenrichment is possible) is supplied to the underside ofthe grate.

The combustion gas generated in a sub-stoichiometrical operation of the grate process (gasification) is fed lO thecombustion chamber unit for an independent, muhi-staged post-combustion. Depending on the primarymeasures to be examined in the area of the combustion chamber uni:. astaging of the air and. if necessary, ofthe combustible can be provided for the sening of the required conditions nf temperature, oxygen-concentrationand residence time along the reaction path.

The scheme ofthe process measuring and control equipment in the flow diagram in figure 3 gives an outline ofthe quantities measured during the test operation. The mass flow and composition of the input material, the massflow and the composition of the fuel gas and the ex haust gas: CO2, O2, CO, H2, CH., NO, S02, HC1,temperatures in the grate and the combustion chamber are important conceming balancing and evaluation of theexperimental investigations and belong lO the standard program. If indicated not differently, all gasconcentrations are referred lO IjI02= I I vol-% (i.s.s. dry). The sampling of the organic trace compounds(polychlorinated dibenzo-p-dioxines and polychlorinated dibenzofuranes (PCDD/F), polycyclic aromatichydrocarbons (PAH) , polychlorinated biphenyles (PC8), polychlorinated benzenes (PC8z) etc.) was carriedout first of all discontinuously recording VDI-guideline 3499, page 3. For absorption a condenser followed by 2series of gas bubbiers filled with methoxyethanole as absorptive solution were used. The compounds weremeasured by conventional GC-MS after corresponding clean up of the sampie. The PCDDIF-concentrations areindicaled as lOxicity equivalents (TEQ) referred to the sucked gas volume at ljIo, = 11 vol-% (i.s.s. dry). Aboutthe on-line measurements with REMPI-TOFMS will be reported below separately.

page 5

3 Decomposition of Organic Trace Compounds in thePost-Combustion Process

3.1 Variation of Influencing Parameters

mass flow rale of 'uel rh F,g [kg/hJ 60 60 60 80 80supply (grale)

mass flow rate 01 air supplyfiJA,PC,I Ikglh) 110 153 129 202 118(posl-combustlon)

mass 'Iow rate 01 nitrogen rh Nz.pc.1 IkglhJ 0 0 146 0 146supply (post-<:ombustlon)

mass .Iow rate of air supply rh A,pc,2 (kglhl 134 61 54 314 279(posl-combustlon)

mixture rallo ii 0,61 0,85 1,63 1,00 1,33

First of all there was the aim to indicate trends on the decomposition of organic trace compounds depending onthe process conditions in the post-combustion process. In the context of first grope tests five clear differentadjustments of the main influence parameters temperature, oxygen concentration and residence timerespectively residence time behavior were I 2 I s Iprovided (table 1)[13). leslnumber(n

Striving temperatures of :1'" 1200 °C atthe first stage of the combustion chamberthe first trial (T I) was carried out at nearstoichiometric conditions (A.pe.!'" 1,0). Toimprove Ibe mixing conditions in the firststage of the combustion chamber the airsupply was increased in a second test run(T2). A further improvement of themixing conditions should be achieved by Table 1. Parameter variation of the tests Tl to T5 [13J.additional supply of nitrogen (simulation of flue gas recycling) (TI). At the !Wo last lest runs a decrease of thetemperature was intended. Therefore the mass flow of air was increased (T4) and nitrogen was suppliedadditionally (T5).

The gasification process put in front was operated with crushed waste wood (mixture of eoated woods, pallets,plastie materials, window wood ete.) and remained unchanged conceming the air ratio )'g'" 0.6. The mass flowrate ofinput material was eonstant in the test runs TI 10 T3 and T4 and T5.

3.2 Results and Discussion

temperature in the 09 ",,' ["CI 1200 1050 1050 1050 750

post-combustion 09 "'~ ["CI 950 1000 950 750 750chamber

[} pc.3 ["CI 900 900 900 700 750

resldence Urne T oe [51 2,8 3,1 2,2 2 2,4

mixlure ratio ii I·J 0,615 0,855 1,627 1,005 1,333

01-concentatJon '11 02"" [vol"%1 4,34 4,31 2,76 10,2 6,47

CO-concentratlon l/I CO,pe Img/m'] <10 <10 <10 <10 <10

N01-concentration VI N02.pe [mg1m'] 265 357 254 409 297

PCDD-concentralion l/I PCDD.pe [pg/m'] 224 201 204 1046 89

PCDF-concentratlon VI PCDF,pc Ipg/m'l 333 180 605 2322 2774

PCDDfF-concentraUon VI PCoaF,pc [pglm'] 556 381 810 3368 2862

PCOO/F (TEQ) l/I PCOOF,pc [pglm'] 21 10 20 59 32

The results ofthe above mentioned parameter variations are summarized in table 2 [13). The temperatures weremeasured at the first stage of the eombustion chamber (Spe.!), at the second stage behind the seeondary airsupply (Spe.]), and at the outlet of the eombustion chamber (Spe,,). The gas eoneemrations have been deteeted atthe out let of the eombustion ehamber. To eharacterize the mixing conditions at the first stage of the combustionchamber a simplified mixture ratio 8 is introdueed here: the sum ofthe primary air mass flow rate ",.", and thenitrogen mass flow rate '"N,.", ('",.;".1 : '",." .•+",,,.,,.,) is referred to the mass flow rate ofthe fuel gas from thegasification Ii'a.,'With regard to the PCDD/F-concentrations three summarized aspects seem to be important:

• All TEQ-eoncentrations are below Ilesl number (n I 1 I 20.1 nglmJ (TE) (limit value [14]).

• The sum concentrations for the testruns TI 10 T3 are on the same level inthe context of analysis precision.

• The sum concentrations for the testruns T4 and T5 are on Ihe same leveltoo, however in comparison with TIto T3 around a order often higher.

This allows the following eonclusions.High temperatures alone like in the firsttest Tl of about S", 1200 °C do notrepresent any mandatory prerequisite fora high deeomposilion of organic traceeompounds. Also at lower temperaturesof about S",950°C aecording 10 beller Table 2. Selecled results of the tests Tl to T5 [13J.

page 6

mixing conditions (T2, T3) a high decomposition is possible. The PCDD/F-concentrations of test runs T4 andT5 are comparatively to the test runs Tl 10 T3 about one order of ten higher, which can be referred to the clearlowering of the temperature level below 3< 800 oe. evenheless the PCDD/F-concentrations are also at thislow temperature level below 0.1 ng/m' TEQ. This is remarkable, if one takes into account, that even atgasification of waste wood several organic compounds are produced. This will be shown by the REMPI-TOFMS spectrum below.A further important task, also in connection with the development of the on-line analysis, is the investigation ofso called indicator compounds for PCDD/F. Previous investigations have shown that the chlorinated benzenesand phenols are good dioxin indicators [e.g. 15-18]. Also at the here mentioned investigations a good correlationbetween chlorinated benzenes and PCDD/F was recocnized e.ti&..i) [13]. However polychlorinated biphenylesshows not such a behavior. The PCB varies in a range ofabout 'I'Pes'" 1000 to 3000ng/m'. A direct dependenceon process conditions was not detectable.Next 10 the organic pollutants consideration should be given 10 CO- and 'O-concentrations. In all cases the CO-concentrations are below 'I'co<IOmg/m'. As reponed elsewhere [e.g. 19] the results point out that CO is notany indicator for more stable organic compounds Iike PCBz, PCDD/F etc.The staged processing in test run T3 with extemal flue gas recycling (simulated by nitrogen) leads to low NO-concentrations. At the tests mentioned here no minimization of NO was scheduled. However, if the process isoptimized as reponed in [2, 20], results lower than 'I'N01'" 200 mg/m' can be obtained.The resuhs altogetber lead to the conclusions, that also at gasification-post-combustion processing ofcontaminated waste wood the temperature, residence time and residence time behavior in the post-combustionprocess contribute all together to a decomposition of organic trace substances. A temperature level of about3",950°C accompanied by a high mixing intensity in a first sub-stoichiometrical stage (continuollsly stirredtank reactor) followed by near plug flow conditions in a second overstoichiometrical stage is weil sllited toachieve a high decomposition rate both of organic trace substances and NO.Optimization tests with regard to organic sllbstances require a high expendirure of time and money if thedetection of these compounds is performed o!f-Iine-as reponed until now. An on-line monitoring of organictrace substances would open the possibility of an on-line optimization too. So a new set of process parameterscan be evaluated immediately after transition into a new steady state. The following attempts with a on-linemeasurement method are of special interest both conceming on-line optimization and a better understanding ofpollution formation and decomposition mechanisms.

4 Resonance-Enhanced Multiphoton lonization Time-of-Flight MassSpectrometer

oo 1000 2000 3000 4000 SOOO

concentration PCDD/F (pglm~ I.s.s.dry (11 val.·" 02)

Figure 4. Concentration of chlorobenzene referring toconcentrationof PCDD/F [13).

4.1 Experimental Method

While the on-line detection of NO, SO, or CObelonged to the state of the art for a long time theon-line detection of organic trace components hasstill to be developed. The conventional GC-MS hasreached its limits of performance as far as speedand chemical preparation of the sampie areconcemed. The resllhs are availabJe hours or evendays after the sampIe has been taken.On-line measurements of trace compollnds inexhaust gases of combustion processes requiremethods that combine selectivity with sensitivity.Laser spectroscopic techniqlles, like e.g. theabsorption spectroscopy, the laser inducedflllorescence (L1F), the matrix assisted laserdesorption and ionization (MALDI) or the light

'000

'00 +--+"

page 7

Laser torResonance-EnhancedMultiphoton lonization

(REM Pt)

d;bile REMPI·TOFMS Ins'rumen'

On-line couplingfor process analysis

Molecular Ion

Molecular Ground State

Molecular Excited State,J, ,.. (UV-Spectrum)

>-~Cl)l:W

Figure 5. lonization process.

Resonance-Enhanced Laser detection and ranging (LlDAR), find more frequently practicalMass Speclromelry (REMPI-TOFMS) application. A promising possibility for a species selective on-line

detection is given by a newly developed, mobile laser massspectrometer (Resonance-Enhanced Muhiphoton lonization Time-of-Flight Mass Spectrometer, REMPI-TOFMS). REMPI-TOFMScombines the optical selectivity of Resonance-EnhancedMultiphoton Ionization with a Time-of Flight Mass Analysis to atwo-dimensional analytical method.

For the REMPI detection of aromalic compounds, usually a onephoton resonante /two photon ionization process is used Ul&.2 ).The first photon is ahsorbed and the moleeule is in an exited state.This step is similar to gas phase UV absorption spectroscopy. In themeasurements described helow a second photon is absorbed,ionizing the molecules of interest. For strong transitions in the firstexcitation step the ionization yield can be in the range of severalper cent of the illuminated molecules.

Fig. 6 shows the REMPI-TOFMS equipment schematically. A veryimponant pan ofthe instrumental design is the setup ofthe probingand sampie inIet system for real-time on-line measurements. Theexhaust gas of waste incineration processes consists of corrosiveinorganic suhstances e.g. HCI, NO, SO" a variety of organic

substances of different volatility up to low volatile compounds (e.g. tar) and is loaded with dust panicles. Therequirements for direct probing of PAHs from the combustion flue gas are:

• avoiding of condensation of low volatile compounds,• minimization of memory effects for the semi-volatile compounds.• reduced catalytic activity ofthe surface and• a rugged and reliable design.

The sampie system inIet is huilt with quanz glass surface from the sampling point to the MS. It consists of aquartz gl ass tube, a quanz gl ass wool filter paper and a capillary. The whole system is heatable up to 200°C toavoid memory effects for larger PAHs. The flue gas is sucked with a flow rate of about 100 l/h and the residencetime in the system is ahout 3 sec [6].

The applied KrF excimer laser(248 nm) supplying laser pulseswith IOns duration, up to50 Hz repetition rate und up to10 mJ pulse energy (MINEX,Lambda Physics Inc.,Gerrnany). The laserwavelength of 248 nm is weilsuited for detection of benzeneand its alkylated derivates asweil as for many PAHs andtheirs alkylated derivates. Theionization yield is proponionalto the laser intensity, howeveran adcquate calibrationprocedure is necessary foranalytical application. Thecalibration and adjustment ofthe laser mass spectrometerwere perforrned with ppb

Figure 6. Schematic presentation of Ihe REMPI-TOFMS instrument

page 8

standards prepared dynamically with permeation and diffusion devices. Details of the in-house designed micro-devices will be published elsewhere. Naphthalene can be quantified directly using the external calibrationstandard. Quantitative results are given in reference. The calibration factors for other PAHs have been adaptedin combination with naphthalene calibration from other experiments [6]. For naphtalene@248nm using the KrFexcimer laser a detection limit of 86 ppt was achieved.

4.2 Results and Discussion

350300

N

"'N

~~.8

I

N

N

oN ..,

r f

Because of its unique selectivity, combined with high sensitivity, the REMPI technique is ideally suited for real-time applications. Simultaneous to the above mentioned investigations of the process conditions of the post-combustion process on the decomposition of organic trace substances first on-line measurements with theREMPI-TOFMS were carried out atthe gasification-post-combustionpilot plant. The results discussed ;::.here were obtained at the outlet of '00the combustion chamber. cThe selectivity and sensitivity of ~this method analyzing complex flue ctlgases from a waste incineration c

~~~ces~t~t~~n~t b~es~val~~~~~ti~~t:.~Fig. 7 shows a REMPI@248nm ~mass spectrum recorded after partly Nextinguished post-combustion due @)to reduced air supply. The 0:::measurement time was 500 ~milliseconds, corresponding to 25 Waveraged mass spectra with a laser 0::repetition rate of 50 Hz. The massspectra shows a large number ofmonocyclic and polycyclicaromatics and their alkylated 50 100 150 200 250derivates. A tentative assignment of mass ratio m/zthe peaks is given in fig.7. It is Figure7. REMPI@248TOFMSmass spectrum[5).interesting to note the very lowintensity of toluene (rn/z =92) and higher alkylated benzene derivcte~ in the mass spectrum. Benzene.naphthalene, its methylated derivates, anthracene and pyrene are typical products of incomplete combustion.The spectrum also shows phenol (rn/z =94), a typical product of wood gasification. The REMPl-sensitivity forphenolic compounds is lower compared with benzene, since the introduction of the hydroxyl group causes alarge red shift ofthe first UV transition.The relatively prominent peak at m/z =212 may be due to hexamethylated naphthalene. The rn/z =252 peakcorrcsponds to the ensemble ofpentacyclic PAH isomers (perylene, benzoflouranthenes and benzypyrenes). Thehighly carcinogenic benz[a]pyrene (BAP) molecule represents an indicator compound for the toxicity ofcomplex PAH mixtures. A selective on-line BAP monitoring is possible with a two-Iaser REMPI scheme and asupersonic inlet valve for sampie cooling [21].

5 Summary and Further InvestigationsThe first part of the paper deals with the optimization of a newly developed gasification-post-combustionprocess. Special emphasis is given to the decomposition of organic trace compounds depending on theconditions at the post-combustion process. Thereby the main influential parameters temperature, oxygenconcentration, residencc time and residence time behavior have to be considered coherently. Not only a hightemperature of i}", 1200°C is the prerequisite of a good burn-out, even if high concel1lrations of aromatic

page 9

substances and pre-cursors for PCDDIF occurs in front of the post-combustion. A temperarure level of about& '" 950°C accompanied by a high mixing intensity in a first sub-stoichiometrical stage (continuously stirredreactor) followed by near plug flow conditions in a second overstoichiometrical stage is weil suited to achieve ahigh decomposition rate both of organic trace substances and NO.

However, optimization of the post-combustion conceming organic decomposition is expensive due to the higheffort of discontinuously sampling and analyzing. A promising possibility for a species selective on-linedeteclion is given by a newly developed, mobile laser mass spectromeler (Resonance-Enhanced Multiphotonlonization Time-of-Flight Mass SpeclTometer, REMPI-TOFMS). REMPI-TOFMS combines the opticalselectivity of Resonance-Enhanced Multiphoton lonization with a Time-of Flight Mass Analysis to a lWO-dimensional analytical method. Thc results presented here prove that a REMPI- TOFMS device is applicableunder rough environmental conditions (dust, temperature changes, vibration etc.). A highly time-resolved, on-line analysis of organic compounds in flue gases of industrial incinerators (waste incinerators, power plants,industrial therrnal processes) is possible. It allows on-line detection of PAHs with molecular mass of 252 andhigher with sensitivities in the low ppb region and bettel. The direct monitoring of process-caused concentrationflucruations in the post-combustion chamber at high temperatures opens additional leeway for on-lineoptimization ofthe process.

One future aim is the on-line detection of indicator compounds for PCDDIF. Previous investigations haveshown that the chlorobenzenes and phenols are good dioxin indicators. Next to the funher improvement of theREMPI- TOFMS there is the task of on-hne optimization of the post-combustion process. With regard to furtherminimization of the experimental effort the optimization should be accompanied by mathematical modeling.

6 AcknowledgementsHere we would like to give special acknowledgement to the "Deutsche Bundesstiftung Umwelt", Osnabrück,Gerrnany. With the financial help of this foundation lhe pilot plant was erected, the majority of the testspresented here were carried out and the REMPI-TOFMS device was developed.

7 NomenclatureSymbols

9 temperarureA. air ratio, stoichiometric ratio'IJ concentration (volume related)Ö mixing index

residence timem massrn/z mass ratio

8 References

CSR continuously stirred tank reactorPFR plug flow reaclorPAH polycyclic aromatic hydrocarbonesPCR polychlorinated biphenylesPCRz polychlorinated benzenesPCDDIF polychlorinated dibenzo-p-dioxines

and poychlorinated dibenzofuranesTEQ toxicityequivalent

Indices

1,2,3 staging indexA airF fuelg grateG combustible gasi.s.s. in standard statepe poslcomustion ehamberRG reaction gas (air)101 total

Oux

[I] Scholz, R.; Beckmann, M.: Möglichkeiten der Verbrennungsfiihrung bei Restmüll in Rostfeuerungen.VDI-Berichte Nl. 895, VDI- Verlag GmbH, Düsseldorf, 1991.

[2] Beckmann, M.; Scholz, R.; Wiese, c.; Davidovic, M.: Optimization ofGasificalion ofWaste Materials inGrate Systems. 1997 International Conference on Incineration & Thernlal Treatment Technologies,San Fransisco-Oakland Bay, California, 12.-16. May, 1997.

[3] Boesel, U.; eusser, H.J.; Schlag, E.W.: R. Naturforsch. 33a (1978)1546.

[4] Williams, B.A.; Tanada, T .. ; Cool, T.A.: 24'h Symposium (International) on Combustion,The Combustion Institute, Pittsburgh(1992), 1587.

page 10

[5] Zimmermann, R.; Heger, H.J.; Kettrup, A.; Boesel, U.: A Mobile Resonance-enhanced MultiphotonIonization Time-of-flight Mass Spectrometry Device for On-line Analysis of Aromatic Pollutants inWaste Incinerator Flue Gases: First Results. Rapid. Communic. Mass Spectrom. I I (I 997) I095.

[6] Heger, H.l.; Zimmermann, R.; Dorfner, R.; Beckmann, M.; Griebei, H.; Kettrup, A.; Boesel, U.: On-lineEmission Analysis ofPolycyclic Aromatic Hydrocarbons at pptv Concentration Levels in the Flue Gas ofan Incineration Pilot Plant with a Mobile Resonance-Enhanced Time-of-Flight Mass Spectrometer.Anal. Chem. (In press.).

[7] Kolb, T.; Sybon, G.; Leuckel, W.: Reduzierung der NOx- Bildung aus brennstoffgebundenem Stickstoffdurch gestufte Verbrennungsflihrung. 4. TECFLAM- Seminar, Heidelberg, 1990.

[8] Kolb, T.; Leuckel, W.: NOx- Minderung durch 3-stufige Verbrennung -Einfluß von Stöchiometrie undMischung in der Reaktionszone. 2. TECFLAM- Seminar, Stuttgart, 1988.

[9] Kremer, H.; Schulz, W.: Reduzierung der Ox- Emissionen von Kohlenstaubflammen durchStufenverbrennung. VDI-Berichte NI. 574, VDI-Verlag GmbH, Düsseldorf, 1985.

[10] Klöppner, G.: Zur Kinetik der NO-Bildungsmechanismen in verschiedenen Reaktortypen am Beispielder technischen Feuerung. Dissertation, TU Clausthal, 1991.

[11] Beckmann, M.: Mathematische Modellierung und Versuche zur Prozeßführung bei der Verbrennung undVergasung in Rostsystemen zur thermischen Rückstandsbehandlung.CUTEC-Schriftenreihe, 1995, ISB 3-931443-28-0.

[12] Beckmann, M.; Scholz, R.; Vergasung von Abfallen in Rostsystemen. In: Born, M.; Berghoff, R. (Hrsg.):Vergasungsverfahren für die Entsorgung von Abfallen.Springer-VDJ-Verlag GmbH, Düsseldorf, I 998, ISBN 3-18-990035-3.

[13] Beckmann, M.; Griebel, H.; Scholz, R.: Einfluß von Temperatur, Durchmischung und Verweilzeit aufden Abbau organischer Spurenstoffe bei der thermischen Behandlung von Abfallholz.DGMK Tagungsbericht 9802, 1998, ISB 3-931850-40-4.

[14] Siebzehnte Verordnung zur Durchführung des Bundes-Immissionsschutzgesetzes(Verordnung über Verbrennungsanlagen für Abfalle und ähnliche brennbare Stoffe - 17. BlmSchV)vom 23. November 1990 (BGBl.! S. 2545, berichtigt S. 2832).

[15] Öberg, T.; Bergström, J.; Chemosphere 14 (1985) 1081.

[16] Kaune, A.; Lenoir, D.; Nikolai, U.; Kettrup, A.; Chernosphere 29 (1994) 2083.

[17] Blumenstock, M.; Zimmermann, R.; Schramm, K.-W.; Henkelmann, B.; Kenrup, A.: Presence ofPolychlorinated Dibenzo-p-Dioxines (PCDD), Dibenzofuranes (PCDF), Biphenyles (PCB), ChlorinatedBenzenes (PCBz) and Polycyclic Aromatic Hydrocarhons (PAH) emder Various Combustion Conditionsin a Post Cornbustion ChambeI. Organohalogen Compounds Vol 36,Edited by the Swedish Environrnental Protection Agency (1998) 59, ISBN 91-89192-05-2.

[18] Blumenstock, M.; Zimmermann, R.; Lehnhardt, R.; Schramm, K.-W.; Kaune, A.; Nikolai, U.; Lenoir, D.;Kettrup, A.: Estimating I-TEQ Emissions ofPolychlorinated Dibenzo-para-Dioxines and Dibenzofuranesfrom Lower Chlorinated PCDDfF and Benzenes at Hazarded Waste IncineratoI. OrganohalogenCompounds Vol36, Edited by the Swedish Environmental Protection Agency (1998) 59,ISB 91-89192-05-2.

[19] Hasberg, W.; Römer, R.: Organische Spurenschadstoffe in Brennräumen zur thermischen Entsorgung.Chem.-Ing.-Techn. 60 (1988) 6.

[20] Beckmann, M.; Davidovic, M.; Scholz, R.; Weichert, c.: Vergasung und Verbrennung von Abfallholz inRostsystemen. VDI-Berichte, VDJ Verlag GmbH, Düsseldorf, 1998, ISBN 3-18-091387-8.

[21] Zimmermann, R.; Lenoir, D.; Kettrup, A.; Nagel, H.; Boesel, U.: On-Line Emission Control ofCombustion Processes by Laser lnduced Resonance Enhanced Multiphoton lonization Combined with aTime of Flight Mass AnalyzeI. 26'h Symposium (International) on Combustion, The CombustionInstitute, Pittsburgh(1996), 2859.