废气中颗粒物检测方案

energy&fuelsARTICLEpubs.acs.org/EF Energy & FuelsARTICLE Emission Characterization and Efficiency Measurements ofHigh-Efficiency Wood Boilers Sriraam R. Chandrasekaran， James R. Laing， Thomas M. Holsen， Suresh Raja， and Philip K. Hopke* Center for Air Resources Engineering and Science， Clarkson University， Post Office Box 5708， Potsdam， New York 13699， United StatesSupporting Information ABSTRACT： Detailed gaseous and particle emissions along with thermal efficiency measurements were made on three mid-sizedhigh-efficiency wood boilers with thermal output capacities of 150 kW (514000 Btuh)(n=2)and 500 kW(1.7mmBtuh)(n=1).Wood chips and commercial wood pellets were used as fuel. Continuous emissions ofCO，NO，SO， fine particle mass (PM.s)， andultrafine particle number distributions were determined using a dilution tunnel sampling system. PM.s and semivolatile organiccompound characterization was performed. Low concentrations of CO， organic carbon (OC)， and elemental carbon (EC) duringsteady-state boiler operation indicated good combustion conditions. Fine particle mass from wood pellets was predominantly K"and SO4， with <8% OC and <2%EC. Inorganic emissions (PM2s，NO，and SO2) were found to depend upon fuel quality， whichindicates the need for wood pellet fuel standards in the U.S. Cd， Pb， Ti， Rb， and Zn were found to be enriched in PM2.s， which is ofconcern for human health. Levoglucosan was a predominant organic compound found for all fuels， ranging from 38 to 82 ug/MJ.Total particle and semivolatile polycyclic aromatic hydrocarbon (PAH) emissions were relatively low (19.4-92.8 ug/MJ). Thethermal efficiencies of the wood pellet boilers determined using the provisional American Society of Heating， Refrigeration and AirConditioning Engineers (ASHRAE) standard 155p ranged between 70 and 86% for the 150 kWboiler tested and between 75 and91% for the 500 kW boiler. The use of ASHRAE 155p allowed for the determination of the linear relationship between the energyinput and the energy output over the range of energy outputs rather than only determining the efficiency at minimum and full load asis the current standard practice. Boiler capacity had no significant effect on efficiency； however， the operating conditions， such as fuelfeed rate， outlet water temperature， and building demand， did affect the results. INTRODUCTION There has been increasing interest in developing sustainableenergy sources， especially biomass-based fuel， such as woodpellets and wood chips， for space-heating purposes because ofconcerns about fossil fuel depletion and global climate change.Biomass can be renewable and a CO，-neutral energy source. Itcan be produced domestically and stimulate local economies. Ithas been shown to be a promising source of alternate energy forcommercial or industrial hot air， hot water， steam， and electricity，especially in the northeastern U.S. In the commercial buildingsector， space heating ranks second in terms of energy use. If an increase in the use of biomass energy in the U.S. is to beachieved， health and environmental considerations must beconsidered. Conventional wood stoves， which make up themajority of the fleet in the U.S.， have relatively low efficiency.and significant emissions of CO， soot， particulate matter， andother harmful pollutants， which are a concern from the perspec-tives of global climate change and human health. Many Eur-opean countries have emission， efficiency， and fuel standards forbiomass combustion devices. These standards have driven thedevelopment of many advanced combustion systems that pro-vide substantially higher thermal efficiency and lower emissionsthan conventional systems. Currently， there is little informationavailable on the emissions and efficiency of mid-sized biomassboilers (100-500 kW). A majority of the combustion devicesinvestigated previously have either been residential burners andboilers (2-15 kW)7-10 or large district heating boilers (>1MW).11-14 Studies that have looked at mid-sized boilers have generally only measured fine particle mass (PM.s). It is im-portant to fully characterize the emissions from these boilers ifthey are going to replace existing oil and natural gas boilers. The objective of the present study was to characterize theemissions and efficiency of three high-efficiency boilers withthermal capacities of 150 kW (514 000 Btu h-) (n=2) and500 kW (1.7 mmBtuh-)(n=1). In this study， the AmericanSociety of Heating， Refrigeration and Air Conditioning Engineers(ASHRAE) standard 155p new provisional protocol was usedto determine the full load， partial load， and seasonal efficiencyThe standard can be applied to determine the space heatingperformance and is applicable to all boilers with energy inputvalues ranging from 300 000 to 12 500 000 Btu h. Previousperformance standards do not account for heat losses in the heatexchanger or only account for thermal efficiency at full-loadconditions.5 ASHRAE 155p measures the efficiency over therange of energy output values rather than only making measure-ments at minimum and fuel boiler loads . MATERIALS AND METHODS Boiler Descriptions. Measurements were carried out on threecommercial high-efficiency boilers： a 150 kW boiler installed at ClarksonUniversity’s Walker Center in Potsdam， NY (WAC)， an identical 150 kW ( Received： August 1 9， 2011 ) ( Revised： October 12， 2011 ) ( Published： O ctober 17， 2011 ) boiler at ACT Bioenergy’s facility in Schenectady， NY that wassubsequently installed at the Cayuga Nature Center in Ithaca， NY(CNC)， and a 500 kW boiler integrated with a solar hot-water systeminstalled at the Wild Center Museum in Tupper Lake， NY (WIC). Thesmaller boilers are 150 kW Hamont CATfire wood boilers imported byACT Bioenergy (Advance Climate Technologies， LLC)， and the WildCenter boiler is a 500 kW Hamont CATfire manufactured by ACTBioenergy. All boilers were of the same general design (pictures of the WAC andWIC boilers can be found in Figures S1 and S2 of the SupportingInformation). The boilers employ fully automated bottom fed fuelfeeding systems and a triple air-staging combustion process. Air stagingis accomplished by heating the fuel bed in an oxygen-deficient environ-ment and tangentially injecting secondary and tertiary air at a high air/fuel ratio to burn pyrolysis gases. Thorough mixing of combustion airwith pyrolysis gases allows the boiler to operate at low excess air levels，increasing the temperature in the combustion zone and， thereby，increasing the combustion efficiency. Low temperatures and the presenceof oxygen in the fuel bed have been observed to reduce NOand inorganicparticle emissions.The boilers were equipped with a cyclone to removecoarse particles from the flue gas. Prior to the measurements， the boilerswere run for at least 1 h to bring them to the operating temperature. The WAC boiler was not American Society of Mechanical Engineer(ASME)-certified and， thus， had to be operated unpressurized. It had anexternal heat exchanger to transfer heat from the boiler water loop to thebuilding water loop (see Figure S3 of the Supporting Information). TheCNC and WIC boilers (see Figure S4 of the Supporting Information)were ASME-certified and did not have an external heat exchanger. Fuels. Calorific value， moisture， ash， nitrogen， and sulfur contents，and elemental compositions of the fuels were determined (Table 1 andTable S1 of the Supporting Information). Most of the elements weredetermined by inductively coupled plasma-mass spectrometry(ICP-MS). Nitrogen analysis was performed according to AmericanSociety for Testing and Materials (ASTM) D5291. Sulfur and chlorinecontent were determined by ion chromatography (IC) analysis of the Table 1. Fuel Characteristics pellets A wood chips pellets B gross calorific value (Btu/lbs) 8180 6371 8060 moisture (%) 4.6 26 5.1 ash (%，dw) 0.71 1.79 0.47 N (mg/kg， dw) 1300 3700 1424 S (mg/kg， dw) 74.1 175 63.6 Cl (mg/kg， dw) 38.8 nm nm oxygen bomb washings according to ASTM E775 and United StatesEnvironmental Protection Agency (U.S.EPA) SWP-846 method 5050，respectively. Commercially available wood pellets were used at the WAC(pellets A) and the WIC (pellets B). The CNC used wood chipsconsisting of waste wood and forest residue. The wood pellets had lowmoisture (4.61 and 5.10%) and ash [0.71 and 0.47% dry weight (dw)]contents compared to the wood chips (26.4% moisture and 1.79% dwash). Nitrogen， sulfur， and trace element concentrations were alsosignificantly higher for the wood chips. Emission Measurements. A dilution sampling system with a2.5 um in-stack cyclone conforming to CTM-039 of the U.S. EPA wasused for emission measurements (Figure 1). Dilution sampling simulatesambient conditions by rapidly mixing hot flue gas with high-efficiencyparticulate air (HEPA)-filtered ambient air， enhancing secondary aero-sol formation.Flow rates of sample and dilution air were calculatedusing the differential pressure offactory-calibrated venturi flow elementsand the air temperature. Dilution ratios between 20 and 40 were used.CO， NOx， and SO2 were measured continuously using ambient gasmonitors (models 42i， 43i， and 48i， Thermo Scientific). Continuousmeasurements of the PMs mass concentration were made using a FilterDynamics Measurement System (FDMS)， consisting of a TaperedElement Oscillating Microbalance (TEOM) particle monitor with anFDMS kit (models 8500b and 1400ab， Thermo Scientific). Ultrafineparticle number concentration and particle size distributions in the rangefrom 5.6 to S60 nm were measured using a fast mobility particle sizerspectrometer (FMPS， model 3091， TSI， Inc.). All emission values arereported at dry standard state conditions (283.34 K and 101.325 kPa).Continuous measurements were made for 17.2， 7.2， and 6.0 h for theWAC， CNC， and WIC，respectively. Emission factors were converted tonominal emissions (1MJ) per fuel energy input because the flow ratein the stack and fuel energy input rate were known. Quartz fiber filters were analyzed for organic carbon (OC) andelemental carbon (EC) by the National Institute for Occupational Safetyand Health (NIOSH) 5040 method using a Sunset Laboratoriesanalyzer. Prior to use， the filters were baked in a muffle furnace at550 °C for 16 h to remove organics. OC was artifact-corrected using twoparallel sample lines： the first with a quartz filter and the second with aTeflon filter followed by a quartz filter. The quartz filter downstream ofthe Teflon filter provided an estimate of the gas-phase OC.A 1.5 cm’2punch from the quartz fiber filters was sonicated in 10 mL of Milli-Qwater for 60 min and analyzed for major anions and cations by IC.2 TheTeflon filters were microwave-digested (MARS Express， CEM Corp.) in10 mL of 50% HNO and analyzed for trace metals by ICP-MS usingEPA Method IO-3.5. Semivolatile and particle organic compounds were collected onquartz filters followed by polyurethane foam plugs (PUFs) and analyzed by gas chromatography-mass spectrometry (GC-MS). A total of 133compounds were included in the analysis. Prior to use， the PUF glasslinears were baked in a muffle furnace at 500°C for 1 h and the quartzfilters were baked at 550°C for 16 h. PUFs were extracted in 100 mL ofdichloromethane， dried with nitrogen， wrapped in aluminum foil， andsealed before use.After collection， the filters were extracted with a 100 mLsolution of dichloromethane and methanol (4：1， v/v) and the PUFs wereextracted in 100 mL of dichloromethane using an ASE 300 acceleratedsolvent extraction system (Dionex Corp.， Sunnyvale， CA). After theextract was concentrated by evaporation with ultrapure nitrogen， thesample was split into two fractions. One aliquot was analyzed immediatelyand another was reacted with 99% bis(trimethylsilyl)trifluoroacetamide(BSTFA) and 1% trimethylchlorosilane (TMCS) to convert the polarcompounds into their trimethylsilyl derivatives before being analyzed.GC-MS analysis was performed using a DSQ II single-quadrupoleGC-MS system (Thermo Scientific， Inc.， Franklin， MA) equipped witha 30 m length ×0.25 mm inner diameter × 0.25 um film thickness SGEforte GC capillary column (Restek Co.， Ballefonte， PA). Quartz filters and PUFs collected at the WAC were also analyzed forpolychlorinated dibenzodioxins and polychlorinated dibenzofurans(PCDD/Fs). Samples were extracted with toluene， and analysis was per-formed at State University of New York at Fredonia (SUNY Fredonia)."A total of 7 polychlorinated dibenzodioxins were analyzed for， including 2，3，7，8-TCDD， 1，2，3，7，8-PeCDD， 1，2，3，4，7，8-HxCDD， 1，2，3，6，7，8-HxCDD，1，2，3，7，8，9-HxCDD， 1，2，3，4，6，7，8-HpCDD， and OCDD， along with 10polychlorinated dibenzofurans， including 2，3，7，8-TCDF， 1，2，3，7，8-PeCDF， 2，3，4，7，8-eCDF， 1，2，3，4，7，8-HxCDF， 1，2，3，6，7，8-HxCDF，2，3，4，6，7，8-HxCDF， 1，2，3，7，8，9-HxCDF， 1，2，3，4，6，7，8-HpCDF， 1，2，3，4，7，8，9-HpCDF，and OCDF. Thermal Efficiency Measurement Methods. Boiler efficiencywas determined using the direct method of dividing the useful heatoutput of the boiler by the energy input of the fuel (eq1). Heat input was calculated from the gross calorific value of the fuel. Heatoutput from the boiler was determined by the temperature differencesand flow rates in the output and return water pipe. Temperatures at theWAC were measured with thermocouples connected to a portablehandheld data logger (Omega DAQPRO-5300). A handheld ultrasonicflow meter with type M1 transducers (Shenitch STUF-200H) was usedto measure the water flow rate. The WIC boiler had an automated reporting system for the input andoutput energies. On the basis of real-time measurements of the fuel feedrate via the boiler control system and the higher heating value of thepellets， the input energy was calculated to determine the boiler efficiencybased on boiler loads. The fuel feed rate for CNC and WAC wasmeasured by hand feeding a known quantity of pellets into the feedingbin of the boiler and recording the time of consumption. RESULTS AND DISCUSSION Gaseous and Particulate Emissions. During continuousmeasurement periods， the WAC and CNC boilers ran at steadystate and operated at thermal inputs of 114 and 90 kW，respectively. The WIC boiler did not run at steady state through-out the continuous emissions testing because of an insufficientheat demand in the building. Thermal inputs for the WIC boilerranged from 378 to 621 kW throughout testing. Testing at theWIC boiler was performed soon after boiler installation and fuelauger rates were manually chosen. During one portion of thetesting， the fuel feed rate， calculated from the fraction of time thatthe fuel auger ran and the weight of pellets delivered per unit time Table 2. Emission Factors (mg/MJ) with 95% ConfidenceIntervals boiler Walker Center Cayuga Wild Center fuel pellets A wood chips pellets B CO 107±2.51 125±4.29 734±125 NO. 17.3±0.17 127±1.67 35.9±1.34 SO， 0.21±0.02 1.42±0.02 0.44±0.02 PM2.5 26.0±0.47 41.1±1.40 26.3±3.53 NFMPs (×10"， number/MJ) 134±1.15 360±6.03 230±14.3 GMDFMPS 91.4±0.31 80.0±0.26 70.3±0.75 OC 1.90±0.20 0.70±0.15 0.58±0.50 EC 0.11±0.01 0.02±0.04 0.24±0.19 Figure 2. Steady-state ultrafine particle number size distributions forWAC， CNC， and WIC boilers. when the auger ran， was determined to be higher than the ratedcapacity of the boiler.All filter and PUF measurements were takenduring steady-state periods. Carbon monoxide emissions， which indicate the completenessof combustion， were significantly higher (p value <0.001) for theWIC boiler (734 mg/MJ) than for the WAC and CNC boilers(107 and 125 mg/MJ) (Table 2). The CO concentration at theWIC boiler was highly variable， characterized by intermittentlarge peaks of high concentration. Testing was performed soonafter installation， and the boiler was not running at steady state，which may account for these results. Previous work has shownthat modern pellet burners and boilers (5-20 kW) operatingwith a nominal feed rate have CO emissions ranging from 21 to194 mg/MJ， and when operating intermittently， the emissionfactors range from 155 to 1100 mg/Mj.7-10 13Sippula et al.observed CO emission factors of 6.60-117 mg/MJ for fourdistrict heating boilers (5-15 MW) operating at 心+etansteady-state. Nitrogen oxide (NO，) and sulfur dioxide (SO2) emissionswere significantly higher (p value <0.001) from wood chipscompared to wood pellets. NO， emissions from biomass burningare all derived from nitrogen in the fuel because temperatures inthe wood boilers are not high enough to form significant amountsof thermal or prompt NO.. The higher NO and SO2emissionsfrom wood chips are due to higher nitrogen and sulfur in thewood chips compared to wood pellets. This relationship has beenobserved in other studies.8-10 A relatively low percentage offuelbound nitrogen was emitted as NO. (14，25， and 21% for pelletsA， wood chips， and pellets B， respectively)， possibly because ofthe air-staging technology employed by the boilers. Figure 3. Ultrafine particle number size distributions for startup，steady-state， and shutdown conditions. Table 3. Chemical Composition of PM2.s from Wood Pellets Walker Center (WAC) Wild Center (WIC) species pellets A pellets B "na= not avaiable.

确定