Influence of gas distribution, field velocity and power supply technique for small scale industrial

Transcription

1 Influence f gas distributin, field velcity and pwer supply technique fr small scale industrial ESP s Dminik STEINER Scheuch GmbH Austria d.steiner@scheuch.cm Wilhelm HÖFLINGER Vienna University f Technlgy, Austria wilhelm.heflinger@tuwien.ac.at Manfred LISBERGER Scheuch GmbH Austria m.lisberger@scheuch.cm 1 Abstract In this wrk the influence f gas distributin and field velcity n the electrstatic precipitatr (ESP) separatin efficiency when using either a cnventinal T/R-set r a High-Frequency pwer supply was investigated. The tests where fcused n small scale industrial ESP s fr bimass cmbustin systems in the range f 300 kw t 2 MW thermal pwer. A small scale industrial ESP which culd be equipped with different perfrated screen cmbinatins in the inlet sectin was used t investigate the influence f tw different pwer supply techniques n the separatin efficiency when using different gas distributin systems and altering gas velcity within the electrical field f an ESP. 2 Intrductin Much wrk was dne n the gas flw regime within electrstatic precipitatrs the last years; a gd verview is given in [1]. Newer papers including CFD simulatins can be fund in [2]. Basically, fr gd precipitatin results the primary crude gas flw shuld be distributed as hmgenus as pssible within the electrde system f the ESP accrding t standards like IGCI EP-7 [3] r VDI 3678 [4]. Fr sme applicatins gd results were btained with skewed gas flw due t inhmgeneus dust cncentratins acrss the inlet sectin f the ESP [5]. Anyway, crude gas flw has t be distributed via guiding- r baffle plates r perfrated screens in the inlet sectin f the ESP. When using perfrated screens the distributin behavir can be influenced via altering screen prsity. The crude gas flw (primary flw) inside the electric field gets furthermre influenced by a secndary flw referred t as electric wind. Thus actual flw regime within an industrial ESP is a cmbinatin f primary and secndary gas flw which is called electrhydrdynamic (EHD) flw. Secndary flw depends n current strength which may change when using different pwer supplies. Fr smaller ESP s als bypass flws in the tp and bttm regin f the electric field may play a significant rle. Summing up, t investigate the gas flw regime within a small industrial ESP the fllwing tests shuld be perfrmed: Altering primary gas flw velcity Altering primary gas flw distributin by mdifying gas flw distributin devices Sealing tp and bttm znes and ash hpper f the ESP Using different pwer supplies t pwer the ESP 3 Mtivatin 3.1 Influence f gas distributin When designing ESPs fr industrial applicatins e. g. pwer plant r cement industry applicatin, nwadays Cmputatinal Fluid Dynamic (CFD) simulatin is used as a state-fthe-art tl t ptimize gas flw distributin upstream and within the ESP. Simulatin f ttal particle precipitatin prcess with high accuracy, including primary and secndary flw, particle migratin behaviur within the electric field, precipitatin f particles at the cllecting electrde and reentrainment phenmena seems t be t cmplex even fr smaller industrial ESP s. Hence, in this wrk nly the primary gas flw within the ESP was simulated via CFD, separatin efficiency was determined by means f a labratry setup and measurements.

2 Mst f the previus wrk t investigate the influence f gas distributin n separatin efficiency was dne with regards t large ESP s. The influence f the tp zne between ESP rf and discharge electrdes (DE), the bttm zne between cllecting electrdes (CE) and DE, respectively, were nt taken int cnsideratin. Sme results f the influence f gas distributin fr small ESP s were already presented in [6] whereas the influence f tp and bttm zne culd nly be assumed. In this wrk als thse tp and bttm zne effects, nticeably fr smaller ESP applicatins, will be discussed. 3.2 Influence f pwer supply Using High Frequency (HF) pwer supplies instead f cnventinal T/R sets leads t higher vltage and current values fr same gas cnditins and electrde arrangements as investigated in previus wrks [7] [8]. Thus the rati f primary t secndary flw within the ESP is mdified when changing the pwer supply. 3.3 Aim f this wrk The aim f this wrk nw was t investigate the influence f perfrated screen cmbinatins with different prsities (β1 β5) n primary gas velcity distributin thse different perfrated screen cmbinatins n ESP separatin efficiency average primary gas velcity n ESP separatin efficiency influence f tp and bttm znes n ESP separatin efficiency the influence f different pwer supply techniques n ESP separatin efficiency fr the abve mentined parameter variatins 4 Experimental setup A small industrial ESP was set up at a labratry as shwn in Fig Air was circulated via a fan acrss a heating system t maintain gas temperature and furthermre thrugh a highspeed nzzle where bimass ash was redispersed. The gas then entered the ESP which was pwered either with a cnventinal T/R set r a High-Frequency pwer supply, als referred t as Switch Mde Pwer Supply (SMPS). A High-Vltage splitter switch was used t change the ESP pwer supply within secnds. Dwnstream the ESP a fabric filter was installed as a back-up filter t cllect remaining particles after the ESP. Fr all tests 150 m³/h f circulated gas was substituted with fresh air which was heated up and sucked thrugh a vaprizer t maintain cnstant gas humidity. All prcess parameters except vlume flw were kept cnstant fr all tests as shwn in Tab The specificatins f the different pwer supplies can be fund in Tab Bth pwer supplies were perated in autmatic mde; the cntrl parameters were set t sparking rates f abut 4 sparks/minute. Fig. 4-1: Simplified test setup with main cmpnents gas temperature 100 C ±4K water dewpint 30 C ±1K inlet dust cncentratin 600 mg/m³ ±15% vlume flw 2.000, 4.000, 5.000, 6.700, m³/h ±2% Tab. 4-1: Experimental prcess parameters 50Hz T/R-set HF pwer supply current rating 200 ma 250 ma vltage rating 70 kv peak /50 kv ar 120 kv switching frequency 50 Hz >20 khz vltage ripple ~30% < 3% Tab. 4-2: Pwer supply specificatins Bimass ash frm a 30 MW pwer plant ESP was used as test dust. The measured crude gas particle number cncentratin distributin is given in Fig Fig. 4-2: Crude gas particle number cncentratin The ESP used fr the tests was a standard type with 300 mm spacing, typically applied at bimass cmbustin plants with abut 1 MW thermal pwer.

3 Tw screens at the crude gas duct (screen 1 and 2) were arranged in series t distribute the incming crude gas acrss the electrde zne f the ESP. The utlet screen was nt changed because f less influence fr gas distributin upstream the screen [9]. The prsity ε f the used perfrated screens was calculated as given in Eq n... hle quantity d... hle diameter l... perfrated plate length b... perfrated plate width Eq. 4-1 T affect the gas velcity distributin within the ESP different perfrated screen cmbinatins were used at the inlet duct but were nt ptimized fr this particular case. The cmbinatins are given in Tab Lchblech 1 Lchblech 2 β 1 Ø36/0.4 Ø36/0.3 β 2 Ø41/0.5 Ø41/0.44 β 3 Ø41/0.5 Ø47/0.52 β 4 Ø41/0.5 nne β 5 Leitbleche Ø47/0.52 Tab. 4-3: Gas distributin setups; hle diameter [mm] and prsity ε In Fig. 4-3 the 3D-CFD mdel prfile with indicated psitins f the perfrated screens fr gas distributin is shwn. Als the psitin f the crss-sectin where the velcity values fr gas distributin calculatins were taken frm is indicated. T determine the influence f the tp- and bttm zne n separatin efficiency as indicated in Fig. 4-4 sealings were applied belw the cllecting electrdes (CE sealing) r abve and belw the discharge electrdes (DE sealings). Fig. 4-4: ESP fixtures and tp- and bttm zne with sealing psitins Fig. 4-3: Simplified test setup with main cmpnents S the influence f gas flw and ptential whirls between ash hpper, baffle plates and precipitatin zne gas flw (partial bypass flw) in tp and bttm zne as indicated in Fig. 4-4 culd be investigated. Ttal particle cncentratin was determined with the laser scattered light measurement system SICK FW100 which was calibrated via gravimetrical measurements accrding t VDI 2066, Part 1 [10]. Particle size distributin and cncentratin was measured with a scattered light aersl spectrmeter Palas WELAS All cncentratin measurements were taken at the clean gas side f the ESP; the high vltage-ff values were taken as crude gas cncentratins. The CFD Simulatin was perfrmed with ANSYS Fluent 12.1 (k-ε mdel with prus media fr perfrated screens) and verified with velcity measurements taken with an impeller wheel and visualized with a fgnzzle.

4 5 Results and discussin 5.1 Primary gas velcity distributin The results f primary gas velcity distributin fr perfrated screen cmbinatins β1 β5 are shwn in this sectin. A ttal unifrm gas velcity distributin already at the leading edge f the electrstatic field culd nt be achieved with the tested perfrated screen cmbinatins. Three main tendencies f the velcity prfiles at the leading edge f the field were determined: 1. Smth center/tp distributin An almst unifrm gas velcity distributin with n high velcity peaks and the main gas flw dminant in the center f the electrstatic field (β2, β3) 2. Center/tp distributin The main gas flw was in the center f the electrstatic field with high velcity peaks (β4, β5) 3. Smth bttm distributin An almst unifrm gas velcity distributin with n high velcity peaks and the main gas flw dminant in the bttm regin f the electrstatic field (β1) Accrding t the IGCI and VDI standards, the gas velcity distributin shuld be: IGCI EP-7 [3] VDI 3678 [4] 85% f velcities must be 1.15 times v average and 99% f velcities must be 1.40 times v average 75% f velcities must be 1.15 times v average and the relative standard deviatin shuld be 25% The crss-sectin fr evaluating the velcities was given with 0.9 m dwnstream the leading edge f the first field fr IGCI and befre the first field fr VDI Due t a ttal field length f ~2.5 m fr the test ESP and nly ne field the crss-sectin fr evaluating velcities were taken at 0.4 m dwnstream the leading edge as indicated in Fig A cmparisn f the CFD simulatins and referred standards are shwn in Fig The screen cmbinatins β1 and β2 achieved the mst even velcity distributins whereas n distributin prfile culd meet the standard requirements. Fr cmbinatins β3 - β5 a rather wrse velcity distributin was fund. Fig. 5-1: ESP fixtures and tp- and bttm zne with sealing psitins 5.2 Influence f perfrated screen cmbinatins n separatin efficiency The fractinal separatin efficiencies fr the 5 different primary gas velcity distributins when altering perfrated screen cmbinatins are shwn in Fig. 5-2 as average f T/R-set and HF-pwer supply measurements. Fig. 5-2: Average fractinal separatins efficiencies fr perfrated screen cmbinatins β1 - β5, v=1 m/s The efficiencies fr β 5 and β 3 are clse tgether at high level whereas efficiency fr β 1 and β 4 are the lwest; efficiency fr β 2 is in the middle. These results d nt match when nly standard deviatin f the primary gas velcity distributin is used as a parameter t assess ESP efficiency because β1 and β2 shuld have had the best ESP perfrmance.

5 5.3 Influence f tp and bttm znes n separatin efficiency In Fig. 5-3 the average fractinal separatin efficiencies fr nrmal peratin withut any sealing, when CE was sealed and tp and bttm zne was sealed (DE sealing) can be fund. smaller particles, especially in the range f µm d decrease dramatically. The average ttal separatin efficiency f bth pwer supplies fr different gas velcities is give in Fig Fig. 5-3: Average fractinal separatins efficiencies fr nrmal peratin, CE and DE sealing, β2, fixed vltages The gas flw and ptential whirls between ash hpper, baffle plates and precipitatin zne d nt have majr influence n precipitatin efficiency since bth efficiency curves are in the same range. A nticeably influence f tp and bttm zne was fund as shwn in Fig. 5-3 as efficiency did increase when the znes were sealed. 5.4 Influence f primary gas velcity n separatin efficiency Fig. 5-4 shws the average fractinal separatin efficiencies fr average gas velcities frm m/s. Fig. 5-5: Average ttal separatins efficiencies fr β1 - β5 and m/s Fr all perfrated screen cmbinatins with velcity decreasing ttal separatin efficiencies were fund. β1 and β2 again shwed the lwest separatin efficiencies whereas fr all ther cmbinatins almst similar efficiencies were investigated. 5.5 Influence f different pwer supply techniques n separatin efficiency The separatin efficiencies fr 3 different particle sizes dependent n average gas velcity fr the tw different pwer supplies is given in Fig Fig. 5-4: Average fractinal separatins efficiencies fr m/s, β2 The separatin efficiency fr particles >5 µm slightly decrease with higher average gas velcities whereas separatin efficiency fr Fig. 5-6: Separatins efficiencies fr 0.3, 2 and 5 µm, T/R-set and HF pwer supply, β2 Separatin efficiency fr bth pwer supplies shwed decreasing efficiency with increasing gas velcity. Fr smaller particle sizes and higher gas velcities bigger differences f the

6 separatin efficiency using the tw different pwer supply techniques were fund. 6 Cnclusin Fr smaller industrial ESP s like thse similar t the ne used in this wrk, meeting hmgenus gas distributin as indicated in sme standards seems nly pssible with excessive effrt. A linear cnnectin between gas velcity standard deviatin and ESP perfrmance culd nt be fund. Cncluding fractinal as well as ttal separatin efficiency measurements, mst imprtant is a prevalent central gas flw with lw velcity peaks. A majr influence n ESP efficiency was fund in partial bypass flw abve and belw discharge electrdes. This tp and bttm znes d have increasing imprtance with decreasing electrde height as shwn in Fig. 6-1 fr cnstant clearings f x spacing distance. Fig. 6-1: Share f tp and bttm zne n ttal ESP height fr varying electric field height and 3 different spacings In Fig. 6-1 the share f tp and bttm zne fr the used ESP with 300 mm spacing and 2 m field height f ~19% is indicated. It was fund that with increasing gas velcity the separatin efficiency did decrease as expected. Especially fr smaller particles in the size range f 0.4 µm the efficiency drp was dramatically. When using the HF pwer supply higher separatin efficiencies were determined fr all measurements. The greatest imprvement was achieved fr high velcities and small particle sizes. 7 Literature [1] Schmid, H. J.; Zum Partikeltransprt in Elektrischen Abscheidern; Achen, Germany: Shaker; 1999 [2] Yan, K.; Electrstatic Precipitatin - 11 th Internatinal Cnference n Electrstatic Precipitatin. Hangzhu, China: Springer/Zhejiang University Press; 2008 [3] Gas Flw Mdel Studies; Publicatin N. EP- 7; Internatinal Gas Cleaning Institute Inc., Revisin 4. Octber; 1981 [4] Elektrfilter Przessgas- und Abgasreinigung VDI 3678; Richtlinien-Entwurf, Kmmissin Reinhaltung der Luft im VDI und DIN - Nrmenausschuss KRdL; 2010 [5] Hein, A. G.; Gibsn, D.; Skewed Gas Flw Technlgy imprves precipitatr perfrmance ESKOM experience in Suth Africa; 6 th Internatinal Cnference n Electrstatic Precipitatin. Budapest, Hungary; 1996 [6] Steiner, D. et al; Influence f Gas Distributin and Field Velcity n Separatin Efficiency at ESP s with Regards t Different Pwer Supply Techniques; Prceedings f Internatinal Cnference & Exhibitin fr Filtratin and Separatin Technlgy, FILTECH 2011, Wiesbaden, Germany; Vl. II, p ; 2011 [7] Steiner, D.; Höflinger, W.; Lisberger, M.; High frequency pwer supplies fr ESP's at bimass applicatins; 6 th Eurpean Meeting n Chemical Industry and Envirnment, Mechelen, Belgium: "EMChIE 2010 Cnference Prceedings (Vl.2)" p ; 2010 [8] Steiner, D.; Höflinger, W.; Lisberger, M.; Steigerung des Abscheidegrades durch Einsatz vn Hchfrequenz-Hchspannungsgeräten bei Elektrfiltern hinter Bimasseverbrennungen; Gefahrstffe Reinhaltung der Luft, 70 (2010), 6; Seite , 2010 [9] Parker, K. R. (Ed); Applied Electrstatic Precipitatin; UK: Chapman & Hall, Lndn; 1997 [10] VDI; VDI 2066, Blatt 1: Gravimetrische Bestimmung der Staubbeladung; Verein Deutscher Ingenieure; 2006