Home

DRI Model 2001 Thermal/Optical Carbon Analysis

1. Hydrogen for the FID flame regulated to 15 psi with a metal diaphragm regulator Hydrocarbon free air to supply O2 to the FID regulated to 15 psi by a metal diaphragm regulator from a zero air generator Compressed air for pneumatic activation regulated to 25 psi DRI STANDARD OPERATING PROCEDURE Page 25 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 At least one backup cylinder per gas type should be kept on hand at all times The 90 He 10 O mixture are typically replaced every four to six weeks H and He are replaced once a week All gases are replaced when the cylinder pressure drops below 500 psi Check the O2 scrubber and follow the manufacturer s recommendations for scheduling its replacement The flow settings on the flow meters rotameters are based on an input of 15 psi for He 90 He 10 O H2 and FID air The pneumatic drivers for the breech should have a pressure of 25 psi to operate effectively sealing the opening 2 1 7 Forms and Paperwork All samples are logged in upon receipt at the laboratory A sample analysis list will be prepared by the laboratory supervisor or designated technician indicating which samples will be analyzed plus any special instructions Figure 2 6 provides an example of the sample analysis run list As individual samples are analyzed entries are made in the Carbon
2. FID CFD6 FIDS gm e Start a run by clicking on Run After the computer states Please load gas syringe flush the gas syringe with the calibration gas at least three times and then load it with the calibration gas The computer will then state the Time remaining until load XX seconds and then Inject calibration gas Follow the verbal instructions to inject the calibration gas through the septum DRI STANDARD OPERATING PROCEDURE Page 52 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 e Calibration gas injections should be in the following ranges for 100 ul gas Manual Injection Lower Allowable Limit Upper Allowable Limit CH 20 36 ug carbon 22 50 ug carbon CO 20 28 ug carbon 22 41 ug carbon Final Calibration Peak 20 000 Calculated in a real laboratory environment For a 5 12 CH standard at 646 mm Hg at 24 C actual mass of methane is 21 43 ug carbon Calculated in a real laboratory environment For a 5 10 CO standard at 646 mm Hg at 24 C actual mass of carbon dioxide is 21 34 ug carbon Lower Allowable Limit equals to 5 lower than the actual mass Upper Allowable Limit equals to 5 higher than the actual mass Limits should be adjusted according to the real laboratory environment e Note Each time the MnO or Ni catalyst is replaced the instrument cali
3. Calibration Injection The injection of calibration gases either CO or CH 4 into the DRI STANDARD OPERATING PROCEDURE Page 12 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Calibration Peak FID Split Time Organic Carbon Elemental Carbon High Temperature OC High Temperature EC Laser Split Lower Split Time Pyrolysis sample stream at the beginning and end of each work day to check instrument performance The FID peak resulting from the automatic injection of methane calibration gas CH4 He at the end of each analysis run for each sample All integrated peak areas are divided by the calibration peak area and multiplied by an instrument specific calibration factor to obtain ug carbon per sample punch The time at which the laser split occurs plus the transit time required for thermally evolved carbon to travel from the sample punch to the FID Carbon evolved from the filter punch in a He only gt 99 999 atmosphere at 140 280 480 and 580 C plus pyrolyzed organic carbon This is the same as Volatile Organic Carbon VOC plus high temperature OC Carbon evolved from the filter punch in a 98 He 2 Oz atmosphere at 580 740 and 840 C minus any pyrolyzed OC Carbon evolved from the filter punch in a He only atmosphere at 280 480 and 580 C plus pyrolyzed organic carbon This is
4. Carbon Analysis Programi cao C eee ee 38 Figure 3 4 Tabular Printout from DRI Model 2001 Thermal Optical Carbon Analysis Program m aT RE aal NRT ar DR wn A 39 Figure 3 5 Graphical output from DRI Model 2001 Thermal Optical Analysis Program 40 Figure 4 1 Calibration thermogram from the cmdAutoCalibCheck command of the DRI Model 2001 Thermal Optical Carbon Analyzer esse 49 Figure 4 2 Carbon analyzer Analysis Setup screen ai ee eee stereo deir age denser 51 Figure 4 3 Example of calibration summary worksheet used to determine calibration slope 55 Figure 4 4 Example DRI Carbon Analyzer Calibration Curves eene 56 Figure 5 1 Temperature ramping with a Tempilaq G temperature indicator rated at 184 C Figure 5 2 Figure 6 1 Also shown are reflectance and transmittance of the temperature indicator if available The vertical dashed line indicates the achievement of the rated STUN SE AUN esha atc ee a a eM act ee 64 Reflectance and transmittance measurements and their first and second derivatives over time with a Tempilaq G temperature indicator rated at 184 C Figure 5 1 The vertical dashed line indicates the achievement of the rated target temperature 65 Tabular Output from the DRI Thermal Optical Carbon Analyzer 71 DRI STANDARD OPERATING PROCEDURE Page 7 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of
5. 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 The thermograms are checked for the following Figure 3 5 e The initial FID baseline is flat indicating that the analyzer has been thoroughly purged before analysis began e The final FID baseline prior to the calibration peak is within three milliVolts of the initial FID baseline excessive drift is cause for reanalysis e The laser reflectance signal during the first two minutes prior to sample introduction to the Analysis position appears near the bottom of the graph and shifts position once the sample is in the Analysis position The transmittance signal will be out of scale until the sample is in the Analysis position e The laser signal should dip below the initial laser line until O is introduced at which point the signal should rise steeply For most samples charring does occur High temperature soot samples may not show this characteristic e The temperature readings reflect stable and smooth temperatures at each level and quick transitions between levels e Problems or deviations from normal should be circled in red If the sample punch taped to the thermogram is not white it is also circled If examination of the tabular and thermogram printouts results in a decision that a sample should be reanalyzed write Rerun in red on the printouts and prepare a re analysis list This list should be posted immediately after the valida
6. only the monitors are turned off when the analyzers are not in use Confirm that the date and time on the computer are correct Wipe the sample tweezers flat glass plate and punching tool with clean KIMTECH wiper taking care not to contact the cleaned surfaces with fingers or other dirty items Check to make sure that no fibers from the KIMTECH wiper are left on the surfaces Begin the daily entry in the Carbon Analyzer Logbook Entries should follow the format in Figure 2 7 Make sure that the printer has enough paper for the day and that the toner cartridge is producing legible text and graphics NOTE When using the carbon analysis software clicking on the Exit button closes the program Exiting in the midst of an analysis is not advisable as the analyzer will revert to the default settings see settings under Options Manual in such a case a hot thermocouple will retract possibly damaging the Teflon seal If for some reason the program freezes allow the sample oven to cool to below 200 C and then close the program and restart DRICarb exe 3 1 2 Leak Checks Perform leak checks daily to detect leakage in the sample oven 3 1 2 1 Manual Leak Check To begin the manual leak check go to C CarbonNet then double click the DRICarb exe program icon to begin the carbon program or double click the DRICarb exe shortcut on the computer desktop The DRI welcome screen appears as depicted in Figure 3 1 Verify that the correct v
7. should have EC less than or equal to 3 8 OC should be less than or equal to 18 Circle any records that require further investigation e All operator generated flags must be either converted to standard analysis flags Table 6 1 or removed The flags in Table 6 2 are temporary flags only and are not recognized as legitimate analysis flags at DRI DRI STANDARD OPERATING PROCEDURE Title DRI Model 2001 Thermal Optical Carbon Analysis TOR TOT of Aerosol Filter Samples Method IMPROVE A e After all thermograms have been reviewed and all possible reruns have been identified post the rerun list in the carbon room and have the reruns done as soon as possible e Review the data from the reruns looking for inconsistencies Confirm that the reasons for the rerun have been addressed Mark the printout with the new values for manual insertion into the MS Excel file Previous runs must be flagged as invalid or the reruns flagged as replicates e Finally all comments flags insertions and other changes made to the printout are entered into the MS Excel file After all changes are made generate a new printout Label the new printout with the file name and printout date Forward a copy of the Page Date Number Revision 76 of 79 July 15 2005 2 216 2 3 printout and the MS Excel file on disk to the person putting the final report together Table 6 2 DRI Carbon Analysis Temporary Flags Flag Description EI Error in sample I
8. 1 Shipping and Mailing Procedures DRI SOP 6 009 1 Field and Laboratory Safety Procedures Any SOPs dealing with filter handling and shipping in conjunction with the specific sampling method used DRI SOP 2 106 1 Pre Firing of Quartz Filters for Carbon Analysis The DRI Model 2001 Thermal Optical Carbon Analyzer Owner s Manual revised 3 2004 Atmoslytic Calabasas CA The DRI Model 2001 Thermal Optical Carbon Analyzer Installation Operation and Troubleshooting Manual revised 12 2004 Atmoslytic Calabasas CA 2 APPARATUS INSTRUMENTATION REAGENTS AND FORMS 2 Apparatus and Instrumentation 2 1 1 Description The components of the DRI Model 2001 Thermal Optical Carbon Analyzer are depicted in Figures 2 1 through 2 3 the complete gas flow schematic is shown in Figure 2 4 Other details of the configuration of the DRI Model 2001 Thermal Optical Carbon Analyzer are referred to in the owner s manual The programmable combustion oven is the heart of the carbon analyzer and includes loading combustion and oxidation zones in a single quartz oven as depicted in Figure 2 5 In addition to the DRI Model 2001 thermal optical analyzer which is connected to a Pentium 4x compatible computer the following items are needed for routine carbon analysis e Stainless steel punching tool 5 16 inch diameter 0 5 cm nominal area for removing small sample punches from quartz filters This punching tool must be kept clean and sharp If the pun
9. Aerosol Filter Samples Method IMPROVE A Revision 3 1 GENERAL DISCUSSION 1 1 Purpose of Procedure This standard operating procedure is intended to e Provide a basic understanding of the principles of carbon analysis and carbon analyzer operation e Describe routine determination of organic elemental and carbonate carbon from ambient and source filter samples using the DRI Model 2001 Thermal Optical Carbon Analyzer and e Detail the concerns and procedures which will ensure a state of the art carbon analysis measurement process This procedure will be followed by all analysts at the Environmental Analysis Facility of the Division of Atmospheric Sciences of the Desert Research Institute Reno Nevada USA 1 2 Measurement Principle The operation of the DRI Model 2001 Thermal Optical Carbon Analyzer is based on the preferential oxidation of organic carbon OC compounds and elemental carbon EC at different temperatures Its function relies on the fact that organic compounds can be volatilized from the sample deposit in a non oxidizing helium He atmosphere while elemental carbon must be combusted by an oxidizer The analyzer operates by 1 liberating carbon compounds under different temperature and oxidation environments from a small sample punch taken from a quartz fiber filter 2 converting these compounds to carbon dioxide CO2 by passing the volatilized compounds through an oxidizer heated manganese dioxide MnO 3 red
10. Analyzer Logbook as shown in Figure 2 7 DRI STANDARD OPERATING PROCEDURE Page 26 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Figure 2 6 DRI Carbon Sample Analysis Run List Date From To Analysis IMPROVE NPS B04 June 2004 Batch G8 Quartz 68 26 64 Account 6300 683 6081 D Crow Carbon Lab OC EC by TOR 296 samples data in IMOETGSI DBF Sample Overview This analysis list covers samples from the NPS IMPROVE project These are 290 PM2 5 samples on 25 mm Quartz filters including no lab blanks and no fieid blanks These samples were collected with an Improve sampler Analysis Overview Sample deposit area 3 53 cm Analysis start date When Ready Analysis deadline Sample Analysis DO NOT area Carbon location Carbon Lab Details REPLICATE BLANKS OR SECONDARIES DO NOT FLAG BLANKS Q numbers have 3 53 cm2 M numbers have area 3 860 cm2 analysis data will be stored in the D MIMPROVE B 4MLOTG8 directory DRI STANDARD OPERATING PROCEDURE Title DRI Model 2001 Thermal Optical Carbon Analysis TOR TOT of Aerosol Filter Samples Method IMPROVE A Figure 2 7 DRI Carbon Analysis Logbook Format ASIA say Neme SS Y HPN Bre Hoe EA 200 LAA V Poy lt Q ER Ww ripe Bro VEI LIA BAT e Wy ze I ME E BH tales s ans aes IUE NE 6 LE Gy su
11. Bakeu tifus li anaouwnes 32 3 1 4 OC BC Analysis A A leben abo pss 34 Sv T Analysis Preparation cys sal sic eios use aeo b ipe to De barato da acid epe Li io 34 3 1 4 2 Loading the Filter Punch tei a cet eerie 35 LS PA d 36 3 1 5 Special Analysis T 41 3215 47 System Blank penna e En H a ea a 41 21 22 APIO ALS A alysiS ennie tii 41 3 1 6 Analyzer Shut DOWU nesine judas 42 3 2 Abbreviated Procedure Senor A EA dep Ui aao amat 43 3 2 1 Analyzer Stat BT T 43 DRI STANDARD OPERATING PROCEDURE Page 3 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 3 2 2 Leak Checks and Oven Baking A rodea dpa tiep oc bu EN M 44 32 3 OCTEC Analysis deest dr teer eer dte Ute Ga te alle re e ogee eee 45 3 222 0 Analysis Prepatatton uu ede e batter e aei Mate MR E UE UA ed 45 3 2 3 2 Meade the Filter Punch uui ee terree ee a set re odes 45 323 3 Post Analys Sena ste cadet esas ro dla eos ac ue o ds 45 3 2 4 O 46 Ll System Bo M eM ted Gates 46 3 2 o A depo A A a EAS 46 32 5 Analyzer Shu DOWN i cese daedewcins Maree ches dee oe aeons 46 4 QUANTIFICATION Si 47 4 1 ASTUCIA ss 47 4 1 1 End of Run Calibrations Description tdt 47 4 1 2 Routine Calibrations Description eene 48 4 1 2 1 Automated Routine Calibration Description amp Instructions 48 4 1 2 2 Manual Routine Calibration Instruction
12. Chang M C and Paredes Miranda G 2005 Comparison of the DRI OGC and Model 2001 Thermal Optical carbon analyzers Prepared for the IMPROVE Steering Committee Fort Collins CO by Desert Research Institute Reno NV Desert Research Institute 2004 Pre firing of quartz filters for carbon analysis SOP 2 106r4 Desert Research Institute Reno NV Fung K K 1986 Personal communication Huntzicker J J 1986 Personal communication Johnson R L 1981 Development and evaluation of a thermal optical method for the analysis of carbonaceous aerosol M S Thesis Oregon Graduate Center Beaverton OR DRI STANDARD OPERATING PROCEDURE Page 79 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Johnson R L Shah J J Cary R A and Huntzicker J J 1981 An automated thermal optical method for the analysis of carbonaceous aerosol In Atmospheric Aerosol Source Air Quality Relationships E S Macias and P K Hopke Eds American Chemical Society Washington DC pp 223 233 Lin C and Friedlander S K 1988 Soot oxidation in fibrous filters 2 Effects of temperature oxygen partial pressure and sodium additives Langmuir 4 4 898 903 Novakov T 1981 Microchemical characterization of aerosols In Nature Air and Methods of Microchemistry H Malissa M Grasserbauer and R Belcher Eds Springer V
13. The punch area is correct errors in entry require that the calculated carbon concentrations be recalculated e The deposit area is correct errors in entry require that the calculated carbon concentrations be recalculated by hand e The calibration peak area is in the correct range Section 4 1 e The initial and the final FID baseline readings are within three counts of each other excessive FID baseline drift is a cause for re analysis NOTE Some very heavily loaded filters will have an FID baseline drift greater than three counts no matter which carbon analyzer the sample is run on typically a FID baseline drift greater than three counts signals either a problem with the run or with the carbon analyzer e The lower laser split time and the upper laser split time are within 10 seconds of each other If the times differ by more than 10 seconds check that the lower split OC and upper split OC differ by no more than 5 OC values which differ by more than 5 unless due to a small change in laser signal resulting from an extremely clean or very dark sample requires re analysis Calculated carbon values for calibration injection runs are within 10 of the current mean value for the injected gas type on that analyzer DRI STANDARD OPERATING PROCEDURE Page 70 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Acceptance runs
14. The vertical dashed line indicates the achievement of the rated target temperature 2800 2400 Reflectance 2000 Transmittance 1st Reflectance Derivative 1200 2 d Reflectance Derivative 800 2nd Transmittance Derivative Laser Reflectance Transmittance mV 0 200 400 600 800 1000 1200 1400 1600 1800 Time sec Tempilaq s melting point 184 2 9C 35 25 1600 1st Transmittance Derivative 5 25 35 2000 First Second Derivative 6 QUALITY CONTROL 6 1 Performance Testing System blanks are performed at the beginning of each day to confirm the system is not introducing bias in the carbon results and to confirm that the laser signal is not temperature dependent Contamination is potentially due to e Operator practices such as improper cleaning of tweezers and punch DRI STANDARD OPERATING PROCEDURE Page 66 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 e Teflon particles on the push rod getting into the heated zone of the quartz oven e Sample boat contamination e Contamination of the carrier gas e Fibers left on the punch tool or on the flat glass plate during cleaning e Contamination from field operator e Contamination from normal use of analyzer A temperature dependent laser signal is potentially due to e Physical coupling of the push rod to the b
15. dries quickly to a dull opaque film when applied to a surface As the surface is heated to the designated temperature the film liquefies and is accompanied by a change of appearance that can be optically monitored to determine sample temperature 5 1 1 Standard Preparation Temperature calibration requires two pre fired quartz fiber filter 2500 QAT UP Pall Life Sciences Ann Arbor MI punches 0 3 cm and 0 5 cm for DRI analyzer and a clean matching sized quartz disk Continental Glass Engineering Burbank CA Quartz fiber filter punches are sliced in half with a filter slicing device Fung et al 2004 A thin layer of Tempilaq G 25 uL is uniformly applied to the quartz disk surface with a 0 1 ml Eppendorf graduated Combitip Brinkman Instruments Inc Westbury NY and before drying is immediately covered with a sliced filter punch For cost savings a glass instead of quartz disk can be used for Temiplaq G at temperatures less than 520 C Higher temperatures require the quartz disc to prevent melting The disc sandwich i e temperature standard is then loaded on a sample holder for analysis The mass of applied Tempilaq G is determined gravimetrically to ensure its mass is 10 5 1 2 Temperature Program After insertion of the temperature standard into the analyzer the temperature is slowly 2 C min ramped across a 50 C range containing the specified Tempilaq G melting point This slow ramping creates a quasi equi
16. for pre fired quartz filters result in lt 1 5 ug cm OC 0 5 ug cm EC and lt 2 0 ug cm TC for IMPROVE A thermal protocol Filters which exceed these levels must be re fired Items which are found to be okay are underlined in red Items which have problems are circled in red DRI STANDARD OPERATING PROCEDURE Title DRI Model 2001 Thermal Optical Carbon Analysis TOR TOT of Aerosol Filter Samples Method IMPROVE A Page Date Number Revision 71 of 79 July 15 2005 2 216 2 3 Figure 6 1 Tabular Output from the DRI Thermal Optical Carbon Analyzer CARBON ANALYSIS RESULTS Analyzer 6 Technician LAC Analysis ID Q17289 2 0EC Sample ID Q17289 Punch area 0 506 cm2 Deposit area 3 530 cm2 Analysis Start 12 22 03 14 52 Calculation 10 21 04 11 59 Analysis Stop 12 22 03 15 09 Anal program ver P6 0 02 20 03 Parm file ver v0302 Calib slope 20 27 ug C peak ratio Baseline time 11 sec Calib intercept 00 00 ug C Reflectance unc 010 counts Transmission unc 010 counts Sample transit 18 sec Calibration peak area 25078 millivolt seconds Initial FID baseline 100 millivolts Final FID baseline 101 millivolts Laser reflectance initial baseline 1291 millivolts Laser reflectance minimum 884 millivolts at 192 sec Laser reflectance final baseline 1589 millivolts Laser transmittance initial baseline 315 millivolts Laser transmittance minimum 60 millivolts at 201 sec Laser transmittance final baseline 427 millivolts Reflect Spl
17. for quartz oven tube inlet Teflon ferrules Parker or Swagelok style 0 64 cm front and back ferrule for the quartz oven tube outlet connections Swagelock T 400 SET Solon OH Refer to instrument user manual for specific ferrule sizes Teflon ferrules 0 32 cm to 0 64 cm Alltech Associates RF 400 200 T Deerfield IL for the thermocouple rod at the inlet breech Refer to instrument user manual for specific ferrule sizes style varies by location on analyzer High temperature silicone seals for quartz light rods and connector prepared by Alltech Associates 15427 and 15429 Deerfield IL Heating element for oven Watlow VC401A06A 0000R 90 bend Columbia MO Printer paper and toner cartridge Computer CD for backup if not on server backup Reagents The following chemicals should be reagent grade or better Potassium hydrogen phthalate KHP for calibration Fisher Scientific cat P 243 CAS 877 24 7 Fairlawn NJ Sucrose for calibration use EM Science SX1075 1 Gibbstown NJ Manganese dioxide MnO crystalline as an oxidizer in the oxidation oven Nurnberg Scientific C5162 Portland OR Aldrich Chemicals 24344 2 St Louis MO or equivalent Nickelous nitrate Ni NO3 2 6H2O crystalline used to prepare the nickel catalyst in the methanator Fisher Scientific cat ZN62 5000 CAS 13478 00 70 Fairlawn NJ DRI STANDARD OPERATING PROCEDURE Page 24 of 79 Date July 15 2005 Title DRI
18. have identical peak areas if the catalysts are in good condition and the calibration factor holds See Figure 4 1 Use the following steps to perform this automated calibration e From the DRICarb exe Welcome screen select Analysis from the Main submenu e Set Type to SAMPLE and select cmdAutoCalibCheck from the drop down menu in the Command table field e The project name should be Calib the Batch should be 0 and the sample ID should be in the format CxxYYYYMMDD where xx is the analyzer number e g C0720050710 for analyzer number 7 run on July 10 2005 e Set the Run 1 for first calibration of the day and 2 for second calibration of the day etc Enter 1 in the Punch area and Deposit area fields Click on OK and then click Run e Review the thermogram and record these values in the logbook as shown in Figure 2 7 The three calibration peak counts OC3 EC1 LtPyMid should be above 20 000 and should be almost identical in area and within 10 of the Calibration Peak Area value show on the tabular printouts Check the average C value for the calibration gas against those posted on each carbon analyzer DRI STANDARD OPERATING PROCEDURE Page 49 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 e Whenever the MnO or Ni catalyst is
19. replaced an automated routine calibration should be run to confirm that the previous calibration curve holds Figure 4 1 Calibration thermogram from the cmdAutoCalibCheck command of the DRI Model 2001 Thermal Optical Carbon Analyzer a Recall data jaj x m mum mem en eport xit Sued o mer Te S 0517 1 05 17 05 07 17 y 1000 LT ET ET EF ET LT ET ET 10 LaserT D 2500 FID 8 67 400 LaserR oO 2500 900 4 Lo FID_6 0 50 8004 be e 700 4 y N ee 600 4 FP L6 A O a K o m u u s o u a 6 y 5004 NC bs 3 ie 3 L 4004 ESA e E Du E 3004 E 1 N 1 tt zu Tae ee pea 4 1 2 2 Manual Routine Calibration Instructions e From the DRICarb exe Welcome screen select Analysis from the Main submenu e Set Type to CALIB and select cmdCalib HeO for example from the drop down menu in the Command table field DRI STANDARD OPERATING PROCEDURE Page 50 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 e A project name is not required The sample ID should be in the format MIxY Y Y YMMDD for CH injection or CIxY Y Y YMMDD for CO injection where x is the analyzer number e g MI720050710 for a CH injection on analyzer number 7 run on July 10 2005 e Set the Run 1 for first calibr
20. run after the autocalibration Section 4 1 2 1 has been completed and shows an acceptable range The following items should be checked and recorded on the Daily Analyzer Checklist shown in Figure 3 2 These values can be obtained by choosing the manual option from the main welcome screen o Reflectance and Transmittance must be measured with a clean blank filter in the analyze position Reflectance range should be between 1400 and 2000 and consistent with previous days values Transmittance range should be between 800 and 1300 and consistent with previous days values o System blank values Total carbon must be less than 0 2 ug cm o Calibration values Specific to analyzer and must be consistent with previous days values DRI STANDARD OPERATING PROCEDURE Page 33 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Figure 3 2 Daily Analyzer Checklist Daily Analyzer Check List CA 7 Month Aug Us y 2005 stem r Eri a reat Des hoa T 24102 25999124172 D CB Ez NN TETTE FN TEES TERES ET M i emh 241851250931 24112 120171 js a P A o fouls 2574 121 2017 gia Pm ais 2511u 2303o 248 0 z 3zu2 lalib lor saul 12 35 Lao lol loot zaz bu2301 2028203011 1314 eul GHA 24455 2u501 24479 v4S als AMIS 2 22 ud ios 0 1 125998 25578 257977 254907 512 pul Qi 24101 2
21. sample in the remarks field Table 6 1 Common DRI Analysis Flags Flag bl b2 b3 b6 fl f2 il 13 14 m2 j rl r2 r3 r5 V Description Field blank Lab blank Blank of undetermined type Transport blank Filter damaged outside of analysis area Filter damaged inside analysis area Inhomogeneous filter deposit Deposit falling off usually occurs on heavily loaded samples Abnormal deposit area possible air leakage during sampling Non white sample punch after analysis Possible air leakage during sampling First replicate on same analyzer Second replicate on same analyzer Third replicate on same analyzer Replicate on different analyzer Sample void DRI STANDARD OPERATING PROCEDURE Page 69 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE_A Revision 3 6 5 Data Validation and Feedback 6 5 1 Daily Validation Level I validation is performed by manually checking the tabular and thermogram printouts the day after the analysis is performed The laboratory supervisor or a designated technician is responsible for checking the data The following items are checked on the tabular data Figure 6 1 e The filter ID is correct and Punch e For calibration runs the tabular and thermogram printouts are checked to make sure the catalysts are operating at required level e The analysis date and time is correct e
22. temperature rating of 1 calibrations based on the two optical methods were considered to be equivalent therefore their means were used Among temperature indicators that achieve an adequate signal noise ratio temperature indicators of 140 280 480 580 740 and 840 C were chosen for IMPROVE A protocol temperature calibration Chow et al 2005 DRI STANDARD OPERATING PROCEDURE Page 64 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Figure 5 1 Temperature ramping with a Tempilaq G temperature indicator rated at 184 C Also shown are reflectance and transmittance of the temperature indicator if available The vertical dashed line indicates the achievement of the rated temperature 200 2 S Transmittance O gt 2400 2 3 an o c 3 2000 5 E di 150 2 S E c Temperature o om E 100 o o S o 2 s O 3 s e x 40 Reflectance E o cc L c D 0 0 F e 0 200 40 600 amp 1000 1200 140 160 180 a0 E Time sec DRI STANDARD OPERATING PROCEDURE Page Date Title DRI Model 2001 Thermal Optical Carbon Analysis Number TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 65 of 79 July 15 2005 2 216 2 3 Figure 5 2 Reflectance and transmittance measurements and their first and second derivatives over time with a Tempilaq G temperature indicator rated at 184 C Figure 5 1
23. thermogram Figure 3 5 for proper laser response temperature profiles realistic carbon peaks and the presence of the calibration peak at the end of the analysis e Examine the laser signal at the end of the run e Ifa problem is found register it in the analyzer log book and run list and notify the lab supervisor immediately e If all aspects of the analysis appear correct select the appropriate analysis flag from the post analysis form that appears at the end of the run e Mark the analysis date on the sample analysis run list e Using clean tweezers remove the punch from the boat and tape it to the thermogram with transparent tape ensuring that the punch is deposit side up Repeat the above steps for additional analysis runs DRI STANDARD OPERATING PROCEDURE Page 37 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Figure 3 3a Setup Screen for the DRI Model 2001 Thermal Optical Carbon Analysis Program Setup ini x r Graph 1 Type s ibi T minimax fo foo Polarity GUE Unipolar Bipolar LR min max fo 2500 al LT minimax f 2500 Hho Bale Sample y F8 min max 67 eoo Batch example Sub batch F6 minimax fo fo E TimeMax po 000 Command table cmdlmproveA E Sample ID FiDAutoScale IV on slit sl Run Flags Punch area 501 Comment
24. until load will be stated and then Inject calibration gas Inject the gas through the septum Hold the plunger down with needle still inside septum for 10 seconds or until peak appears The CO and CH calibrations are run using the Calibration options from the main menu The following volumes are injected e 100 ul CO gas use 1000 ul syringe e 250 ul CO gas use 1000 pl syringe e 500 ul CO gas use 1000 ul syringe DRI STANDARD OPERATING PROCEDURE Page 60 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 e 1000 ul CO gas do once with 1000 ul syringe and once with 2500 ul syringe e 1500 ul CO gas do with 2500 ul syringe 4 2 2 e Repeat for CH4 Record these calibration values in the logbook as in Figure 2 7 The integrated peak counts are extracted manually from the tabular printouts and entered into the spreadsheet which is used to determine the final calibration The total peak is calculated by adding the peak area from OC1 OC2 OC3 and OCA as well as ECI EC2 and EC3 Pyrolysis counts are not included in the total Full Sucrose and KHP Calibrations To perform the full calibration select Analysis from the Main menu of the DRICarb exe program Welcome screen Choose Calib under the Type drop down menu Complete the information about the sample including Project Na
25. v flag power surge or power failure DRI STANDARD OPERATING PROCEDURE Page 78 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 7 REFERENCES Chen L W A Chow J C Watson J G Moosmiiller H and Arnott W P 2004 Modeling reflectance and transmittance of quartz fiber filter samples containing elemental carbon particles Implications for thermal optical analysis J Aerosol Sci 35 6 765 780 Chow J C Watson J G Pritchett L C Pierson W R Frazier C A and Purcell R G 1993 The DRI Thermal Optical Reflectance carbon analysis system Description evaluation and applications in U S air quality studies Atmos Environ 27A 8 1185 1201 Chow J C Watson J G Crow D Lowenthal D H and Merrifield T M 2001 Comparison of IMPROVE and NIOSH carbon measurements Aerosol Sci Technol 34 1 23 34 Chow J C and Watson J G 2002 PM carbonate concentrations at regionally representative Interagency Monitoring of Protected Visual Environment sites J Geophys Res 107 D21 ICC 6 1 ICC 6 9 Chow J C Watson J G Chen L W A Amott W P Moosmiiller H and Fung K K 2004 Equivalence of elemental carbon by Thermal Optical Reflectance and Transmittance with different temperature protocols Environ Sci Technol 38 16 4414 4422 Chow J C Watson J G Chen L W A
26. 001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 LIST OF TABLES Table 6 1 Common DRI Analysis Fags ese ata 68 Table 6 2 DRI Carbon Analysis Temporary Flags sss see 76 Table 6 3 DRI Carbon Analysis Validation Comments essere 77 DRI STANDARD OPERATING PROCEDURE Page 6 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 LIST OF FIGURES Figure 2 1 DRI Model 2001 Thermal Optical Carbon Analyzer sees 15 Figure 2 2 DRI Model 2001 Thermal Optical Carbon Analyzer Schematic Diagram 16 Figure 2 3 DRI Model 2001 Thermal Optical Carbon Analyzer Sealing Diagram 17 Figure 2 4 DRI Model 2001 Thermal Optical Carbon Analyzer Gas Flow Schematic 18 Figure 2 5 DRI Model 2001 Thermal Optical Carbon Analyzer Combustion Oven 19 Figure 2 6 DRI Carbon Sample Analysis Run List sese 26 Figure 2 7 DRI Carbon Analysis Logbook Format sss sees eee 27 Figures DRI Welcome rss 30 Figure 3 2 Daily Analyzer Checklist eee e nere tete rele Ree Re de dais 33 Figure 3 3a Setup Screen for the DRI Model 2001 Thermal Optical Carbon Analysis Program o Me bod SM uc d ed MN nde sats asd be DOE 37 Figure 3 3b The Analysis Screen during Sample Analysis for DRI Model 2001 Thermal Optical
27. 16 28 Anal program ver P6 0 05 31 05 Parm file ver V0416 Calib slope 21 07 ug C peak ratio Baseline time 10 sec Calib intercept 00 00 ug C Reflectance unc 010 counts Transmission unc 010 counts Sample transit 14 sec Calibration peak area 25485 millivolt seconds Initial FID baseline 106 millivolts Final FID baseline 107 millivolts Laser reflectance initial baseline 1768 millivolts Laser reflectance minimum 0 millivolts at 0 sec Laser reflectance final baseline 1894 millivolts Laser transmittance initial baseline 1241 millivolts Laser transmittance minimum 0 millivolts at 0 sec Laser transmittance final baseline 1259 millivolts Reflect Split Time Laser FID Split Time Lower split 800 sec 1605 millivolts 814 sec Regular split 804 sec 1606 millivolts 818 sec Upper split 812 sec 1604 millivolts 826 sec Transmit Split Time Laser FID Split Time Lower split 895 sec 687 millivolts 909 sec Regular split 902 sec 688 millivoits 916 sec Upper split 908 sec 690 millivolts 922 sec Peak Area Carbon oci oc 342 mv secs 0 56 ug C cm2 1 99 ug C filter DC 2 oc 881 mv secs 1 45 ug C cm2 SES ug C filter oc3 oc 871 mv secs 1 44 ug C cm2 5 07 ug C filter OC4 oc 529 mv secs 0 87 ug C cm2 3 08 ug C filter EC1 EC 753 mv secs 1 24 ug C cm2 4 39 ug C filter EC2 EC 178 mv secs 0 29 ug C cm2 1 04 ug C filter EC3 EC 0 mv secs 0 00 ug C cm2 00 ug C filter LRPyMin Py 579 mv secs 96 ug C cm2 8 37 ug C filter LRPyMid Py 595 mv secs 98 ug C cm2 3 47 ug C filte
28. Carbon Analyzer Combustion Oven Laser reflectance detector Laser 632 nm Quartz light pipe Heater Pushrod Deposit side of 0 5 cm quartz filter Flame X C CO CO gt CH J N R Ionization N Oxidizer Reducer Detector Oxidation Oven Methanator FID Gas flow direction Combustion at 100 He or 98 He 2 O atmosphere Quartz light pipe Sample holder Thermocouple Laser transmittance detector DRI STANDARD OPERATING PROCEDURE Page 20 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 e Syringes Hamilton Gas Tight 1000 and 2500 ul syringes for calibration injections 25 ul syringe for carbonate analysis and for analyzer calibration e Quartz filters Pallflex 2500 QAT UP Pall Life Sciences Ann Arbor MI quartz fiber filter or equivalent e Flat tip tweezers e Flat glass plate e Logbook notebook e Transparent tape e KIMTECH Pure CLA Critical Task Wipers and large KimWipes EX L e Small Styrofoam cooler or refrigerator e Blue ice if using Styrofoam cooler e Butane or piezoelectric lighter e A copy of DRICarb exe the analysis program Carbon par the analysis parameter file and Microsoft Access to run CarbonNetWork mdb 2 1 2 Instrument Characterization The DRI Model 2001 Thermal Optical Car
29. D EA Error in sample deposit area ST Suspect temperature profile SF Suspect FID signal SL Suspect laser signal Mi Miscellaneous problem m2 Non white sample punch after analysis V Invalid run r Replicate b Blank i Inhomogeneous f Filter media damaged g Sample deposit damaged d Sample dropped n Foreign substance on filter W Sample wet DRI STANDARD OPERATING PROCEDURE Title DRI Model 2001 Thermal Optical Carbon Analysis TOR TOT of Aerosol Filter Samples Method IMPROVE A Page 77 of 79 Date July 15 2005 Number 2 216 2 Revision 3 Table 6 3 DRI Carbon Analysis Validation Comments Comments Anomalous laser Operator error Analyzer malfunction Poor replicate Poor initial laser Potential contamination Power failure Description Despite good initial laser laser signal drifted above initial laser signal before dropping typical of auto emissions Used with v flag operator exited program unexpectedly Used with v flag analyzer malfunction or problem beyond the control of the operator such as plugged FID broken oven heater etc Replicate is outside the normal criteria but no reason can be found for the discrepancy Used with v flag severe coupling or boat not pushed in time for calculation of initial laser signal Used with v flag rerun of sample yields lower values or different peaks Typically used with blanks or reruns of replicates Used with
30. DRI STANDARD OPERATING PROCEDURE DRI Model 2001 Thermal Optical Carbon Analysis TOR TOT of Aerosol Filter Samples Method IMPROVE A DRI SOP 2 216 1 Revised November 2005 Prepared by Desert Research Institute Division of Atmospheric Sciences 2215 Raggio Parkway Reno NV 89506 TABLE OF CONTENTS GENERAL DISCUSSION seas a A 7 1 1 AA A A rre DA RI DN 7 1 2 Measurement Principle ud oer ed tt pr ax d det atu c duse tact tuts 7 1 3 Measurement Interferences and Their Nmimizanion sss sees 8 LA Ranges and Typical Values of Measuremients id 9 1 5 Typical Lower Quantifiable Limits Precision and Accuracy sese 10 1 6 PBersonnelResponsibilitieSuu siii natalidad 11 1 7 Definitions for IMPROVE_A Thermal Protocol for Carbon Analysis 11 1 8 Ss A TE 14 APPARATUS INSTRUMENTATION REAGENTS AND FORMS 14 2 1 Apparatus and Tnstr t entatiotis uto A A AA 14 2 1 1 IS ETS 14 2 1 2 Instrument CharacterizaniOn s see ee eee 20 2 1 3 Maintenant ae en paie a e a SR oe 21 2 1 4 SDare Ta TTT 22 21 5 Reagents RM 25 2 1 6 erc DUE E mV a EM RE PED D TR 24 2 1 7 Forms and Paperwork orii dra eret EEEE EE d TENA Sareea 25 PROCEDURES a e T 28 3 1 Detailed Proeedutes ceca e a a DUE 28 3 1 1 Analyzer Start UP A onte O A sical 28 3 1 2 Leak Checks AA 29 3 12 1 Manu l Leak Glieek c A AAN 29 3 1 3 A afecto fua 31 Sd Mir POB et o ae te ee 31 31 925 Automatio Oven
31. Depost area 3 53 Tech initials TLB FID CrFD6 FD 8 OK Exit DRI STANDARD OPERATING PROCEDURE Page 38 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Figure 3 3b The Analysis Screen during Sample Analysis for DRI Model 2001 Thermal Optical Carbon Analysis Program Access database Project name Batch Sub batch Command table Sample ID Run Punch area Dep area Tech A 0705 1 0 501 3 53 CACarbonNet amp CCESS ProductioniCarbonN IMPROVE cmalmprove amp Q71633 TLB FID Run type FID 8 Sample LaserT LaserR FID 8 FID amp Oven Temp Oven Prs Slope B Y EN r Mr E Samp Switch Samp Position Interlock Event Time Status Data on Redraw Exit 1000 10 LaserT o 2500 FID 8 66 400 900 4 Lg LaserR 0 2500 FID 6 0 S 800 4 7004 600 4 500 4 Temperature Degrees C 1200 DRI STANDARD OPERATING PROCEDURE Title DRI Model 2001 Thermal Optical Carbon Analysis TOR TOT of Aerosol Filter Samples Method IMPROVE A Page Date Number Revision 39 of 79 July 15 2005 2 216 2 3 Figure 3 4 Tabular Printout from DRI Model 2001 Thermal Optical Carbon Analysis Program Analysis ID Q080292 1 0EC Sample ID Q80292 Punch area 0 501 cm2 Deposit area 3 530 cm2 Analysis Start 07 19 05 15 57 Calculation 07 19 05 16 28 Analysis Stop 07 19 05
32. Description The end of run calibration consists of a set quantity of He CH4 calibration gas which is automatically injected by the carbon program All FID readings during the analysis run are normalized to this peak to minimize the effects of FID performance and electronic drift over time The end of run calibration occurs automatically at the end of each analysis run and DRI STANDARD OPERATING PROCEDURE Page 48 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 requires no operator intervention The integrated calibration peak counts should be checked by the operator immediately after each run to confirm that the analyzer is operating satisfactorily Calibration peak area counts should be greater than 20 000 Check daily records to compare and determine analyzer performance and stability 4 1 2 Routine Calibrations Description Routine calibrations must be performed at the beginning and end of each day either manually or by using the automated routine calibration command cmdAutoCalibCheck in the CarbonNetWork database Command table 4 1 2 1 Automated Routine Calibration Description amp Instructions The automated calibration uses the Carle valve to inject the methane standard once in a He only atmosphere once in a He O atmosphere and finally the normal calibration peak at the end of analysis The three peaks should
33. I Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Figure 2 3 DRI Model 2001 Thermal Optical Carbon Analyzer Sealing Diagram B Current Sealing of Oven C5 High Temperature Silicone Seal Upper Arm Red Septum Injection Pont Oring Fitting MnO Oxidation Oven Thermocouple Teflon Fitting Push Rod Heater Loading Zone pressure tight A Teflon Fitting Carrier Gas Note In the breech there is a Teflon reducing ferrule to seal the pushrod thermocouple plus one O ring to seal the breech against the inlet coupler connector DRI STANDARD OPERATING PROCEDURE Title DRI Model 2001 Thermal Optical Carbon Analysis TOR TOT of Aerosol Filter Samples Method IMPROVE A Page 18 of 79 Date July 15 2005 Number 2 216 2 Revision 3 Figure 2 4 DRI Model 2001 Thermal Optical Carbon Analyzer Gas Flow Schematic Carle Valve can BOY ATR L He H O Scrubber FID gt Meth Oven Back Valves Sample Oven Symbols Toggle valve X py Needle valve Rotameter with O needle valve 3 Way valve DRI STANDARD OPERATING PROCEDURE Page Date Title DRI Model 2001 Thermal Optical Carbon Analysis Number TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 19 of 79 July 15 2005 2 216 2 3 Figure 2 5 DRI Model 2001 Thermal Optical
34. Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Chromosorb A 20 30 mesh used as a support for nickel catalyst in the methanator from any chromatography supplier such as Alltech cat 2 0165 Both nickelous nitrate and Chromosorb A are for preparing the reduction catalyst in the methanator Hydrochloric acid HCl 0 4 molar solution for use in cleaning punch and quartz ovens and for use in carbonate analysis Fisher Scientific cat A508 212 CAS 7647 01 0 Fairlawn NJ Hydrofluoric acid HF diluted to 15 for removing the white deposits from devitrification white deposits of SiO2 on the quartz oven parts Fisher Scientific cat A147 1LB CAS 7664 39 3 Fairlawn NJ Nanopure water used as described in Section 4 1 3 1 2 1 6 Gases The following compressed gases should be ultra high purity UHP grade or better He for a carrier gas regulated to 15 40 psi with a metal diaphragm regulator The higher pressure is required due to the pressure drop across the Supelco oxygen scrubber 5 CH by volume in He for calibration injections and calibration peaks regulated to 10 psi by a metal diaphragm regulator 5 CO by volume in He for calibration injections regulated to 10 psi by a metal diaphragm regulator 10 O5 by volume in He as a carrier gas regulated to 15 psi by a metal diaphragm regulator In addition the following gases are required
35. OC minus the first OC peak OC1 Carbon evolved from the filter punch in a 98 He 2 On atmosphere at 740 and 840 C minus any pyrolyzed organic carbon present in these two peaks This is EC minus the first EC peak EC1 The separation between OC and EC which depends on the laser measured reflectance and or transmittance of the filter punch returning to its initial value At this point all pyrolyzed OC has been removed and EC is beginning to evolve The time at which the laser measured reflectance and or transmittance of the filter punch reaches its initial value minus the precision of the laser signal currently defined as 10 counts The conversion of OC compounds to EC due to thermal DRI STANDARD OPERATING PROCEDURE Page 13 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 OCI OC2 OC3 OC4 OP ECI EC2 EC3 Regular Split Time Total Carbon TC Upper Split Time decomposition this may be envisioned as charring during the organic portion of the analysis Carbon evolved from the filter punch in a He only gt 99 999 atmosphere from ambient 25 C to 140 C Carbon evolved from the filter punch in a He only gt 99 999 atmosphere from 140 to 280 C Carbon evolved from the filter punch in a He only gt 99 999 atmosphere from 280 to 480 C Carbon evolved fr
36. arbon is calculated by uctus usines zem of KHP used in ug I no of carbon in KHP x 2 vol of solution prep in pl MW of KHP _ weight of KHP used Y 8x12 10 ml Y 10 ug 100 ml 204 23 ul g DRI STANDARD OPERATING PROCEDURE Page 54 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 The nominal 1800 ppm sucrose solution is prepared by transferring 0 428 ug of sucrose into a glass 100 ml volumetric flask Dilute to volume with acidified Nanopure water see blank solution preparation instructions below Mix the sucrose thoroughly Store this solution in a refrigerator until it is used for calibration purposes This solution is good for 40 days Label the flask with the chemical name the date of preparation the name of the chemist preparing the solution and the exact concentration The concentration is calculated by ActualugC per uL e of sucrose used in ug I no of carbon in sucrose x 2 vol of solution prep in ul MW of sucrose pem of sucrose used in ug E x 2 10 ml 10 ug 100 ml 34231 u g To prepare a blank solution add 0 4 ml of concentrated HCI to a glass 100 ml volumetric flask and dilute to volume with Nanopure water This acidified Nanopure water is made fresh each time a 1800 ppm KHP stock solution is prepared No primary standards NIST traceable currently exist for carbon analysis Ideal
37. ating from DRI are pre inspected for defects such as pinholes or tears They are then pre fired for a minimum of four hours at 900 C and 2 of these are acceptance tested for blank levels before use in the IMPROVE network Batches containing filters that fail to pass the preset acceptance levels 1 5 ug OC 0 5 ug EC and 2 0 ug TC per cm are not used for sample collection Average pre fired blank levels are 0 15 0 15 ug OC cm 0 00 0 02 ug EC cm and 0 15 0 15 ug TC cm Because pre fired filters can adsorb organic vapors during shipping and storage the LQL of analysis on a particular set of filters depends on the number of field blanks analyzed and the variability in the results from those blanks To reduce the risk of contamination during shipping and storage samples are vacuum sealed and stored in a freezer The minimum detection limit MDL of the DRI Model 2001 carbon analyzers used for IMPROVE A protocol is based on the analyses of 214 blank quartz fiber filters The MDL is defined as three times the standard deviation of their measured results They are total OC 0 45 ug cm total EC 0 06 ug cm TC 0 45 ug cm Acid evolved carbonate levels in pre fired quartz fiber filters have been shown to be quite variable 0 0 1 0 ug cm over time The reaction of ambient CO with alkaline sites on the quartz fibers may be the cause of such variable blank levels Acceptance testing for carbonate is only performed for special projects tha
38. ation could potentially invalidate the entire day s runs any deviation from the accepted ranges must be noted and the cause identified Notify the lab supervisor Leave the DRICarb exe software open If desired He 1 Cal Gas He 2 and Air may be turned off with the toggle valves to conserve gases However all other gases should be left on as long as the oxygenator and methanator are heated DRI STANDARD OPERATING PROCEDURE Page 43 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 e Place all of the day s printouts including calibration data in a file folder labeled with the date and analyzer number Place on the lab supervisor s desk for Level I validation Section 6 5 e Leave the computers and analyzers on overnight unless the potential for power outages or surges exists Turn off the monitors overnight e Make a final check of the gas cylinder pressures to ensure that gas flow especially the compressed air will continue until someone will be available to check them again e Move the samples and blue ice in the Styrofoam cooler or refrigerator back into the sample storage freezer and verify that the freezer is completely closed e If the 25 or 50 ul syringe was used for carbonate analysis thoroughly rinse the syringe with distilled water and tightly cap all solutions Store solutions in the refrigerator Free
39. ation of the day and 2 for second calibration of the day etc Enter 1 in the Punch area and Deposit area fields Click on OK and then click Run e Select the atmosphere for calibration under the Cal Gas either CH4 or CO2 menu and select the proper Carrier Gas either HeO or He HeO for beginning of day and HE only for end of day These should be alternated with the calibration gas Verify the command table matches the options selected e Choose OK to proceed with the analysis or Exit to leave the program The analysis will start with the following screen as shown in Figure 4 2 You will note that the top portion contains all the information in the analysis Setup screen The bottom half will display the thermogram when the run is initiated DRI STANDARD OPERATING PROCEDURE Page 51 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Figure 4 2 Carbon analyzer Analysis Setup screen Setup al x Graph Type Sample 1000 Polarity Unipolar Bipolar T minimax LR minimax Min 0 0 Project N LT minimax b o Ban AE cal SYSBLK Batch amp 08 F8 minimax 0 400 0 Sub batch 50 ha l TimeMax 1200 Command table emdimprove FiDAutoScale Iv On Sample ID ho Run Flags hn JA nae so Comment Deposit area i Tech initials TLB F6 minimax
40. bination particle gaseous sampler Dichotomous sampler for PM 5 PMooarse and PM 19 Gaseous High volume sampler IMPROVE NPS sampler MiniVol Sampler Audit samples Resuspension chamber Sequential filter sampler SFS Wet Deposition Unknown Y sampler DRI source sampler The final MS Excel file name is specified on the analysis list posted in the carbon room DRI STANDARD OPERATING PROCEDURE Page 74 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 e Begin validation by matching the filters listed on the analysis list with the filters listed on the MS Excel printout There must be at least one entry on the printout for every filter listed on the analysis list e Flag field and lab blanks while the list is being reviewed by placing b1 b2 b3 or b6 in the second column of the printout Because the MS Excel printout is sorted by ID number replicates and reruns will be grouped together e Indicate missing data by writing the missing filter ID in the margin with an arrow drawn to the appropriate place of insertion Scan the printout for unusual IDs which may have been mistyped or misread by the scanner during analysis Generally these will appear at the beginning or end of the printout due to the sorting process Make sure that all samples listed on a rerun list appear on the printout e Resolve all mis
41. bon Analyzer is program driven and data is stored automatically to the hard drive via a PC compatible computer processor board Response times and signal lag times are built into the parameter file that is loaded when the analysis program begins The program is driven by the thermal protocol For example when using the IMPROVE A protocol the program will advance to the next temperature or carrier gas mixture once the FID signal returns to its baseline 1 e after a minimum of 150 seconds at one analysis condition A maximum time limit 580 seconds per analysis condition is also established to prevent a slight baseline drift from holding the analyzer in one condition indefinitely For the Speciation Trends Network STN thermal protocol the program advances from one specified temperature plateau to the next temperature or carrier gas mixture when the specified analysis time is reached Both methods require at least one 0 5 cm punch per filter and do not require sample pre treatment The sample punch is destroyed by both methods DRI STANDARD OPERATING PROCEDURE Page 21 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Operator concerns for correct routine operation of the instrument include the following refer to Section 4 for more details e Verify sample oven pressure reading and specified flow range in the front panel flow me
42. bration should be checked to confirm that the previous calibration curve holds This can be done by running two points from the six CHy He calibration volumes and two points from the six CO He volumes used in the full calibration 4 1 3 Full Calibration Description Full instrument calibration performed semi annually establishes the calibration slope used in converting counts to ug of carbon as explained in the next section Instrument calibration involves spiking pre fired quartz punches with 5 0 to 20 0 ul of the 1800 ppm KHP and sucrose solutions Section 4 1 and injecting 100 to 1500 ul of the CO and CH gases Four types of standards are used to calibrate the carbon analyzers 5 nominal CH in He 5 nominal CO in He KHP and sucrose Only the calibration gases are used on a daily basis as analyzer performance monitors KHP and sucrose are used in conjunction with CH4 and CO semiannually to establish the calibration curve of each analyzer 4 1 3 1 Preparation Ranges and Traceability of Standards The calibration is done by injection of a known volume of the standard to yield a calibration curve of peak area ratio of injected carbon CH internal standard versus ug of carbon injected Internal Standard Calibration Method For the best accuracy the temperature and pressure at DRI STANDARD OPERATING PROCEDURE Page 53 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT
43. ching tool is resharpened the punch area must be re verified Verification is performed by removing 10 punches from a 47 mm quartz fiber filter 17 35 cm then calculating the punch area 17 35 cm x average weight of 10 punches original filter weight Further verification can be done by taking a precise measurement of the punching tool DRI STANDARD OPERATING PROCEDURE Page 15 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Figure 2 1 DRI Model 2001 Thermal Optical Carbon Analyzer Temperatures Oxygenator Methanator Sample Oven FID Temperature Controller Oven Pressure Flame Injection Detector FID Hel He 2 Flow He O Flow CalGas H FID Air Flow Flow He 3 Flow Flow DRI STANDARD OPERATING PROCEDURE Page 16 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Figure 2 2 DRI Model 2001 Thermal Optical Carbon Analyzer Schematic Diagram Laser Photo Detector Valve Network Oxidation Methanator Oven Control FID Signal Reflectance Gas Stream Control Microcomputer Disk Printer DRI STANDARD OPERATING PROCEDURE Page 17 of 79 Date July 15 2005 Title DR
44. cording the date Notify the lab supervisor of any problems and record them in the log book and on the run list e Repeat the above steps for additional samples 3 2 4 Special Analysis 3 2 4 1 System Blanks e Go through all the steps for a normal analysis but do not remove the punch from the previous analysis Proceed with the routine analysis e Use project name Sysblk Batch mm for the month and Sub batch dd for the day Punch area and Deposit area should be 1 e Calculated carbon concentrations from the system blank should not be more that 0 2 ug carbon Values greater than this warrant an additional system blank or oven bake 3 2 4 2 Carbonate Analysis e Follow the steps under Routine OC EC Analysis until the sample punch is loaded into the boat e Enter the Sample ID Run Punch area and Filter area e When prompted for injection eject 20 ul 0 4 M HCI onto the filter punch e Flush the syringe with Nanopure water between samples e Continue the normal OC EC analysis when the carbonate cycle is complete 3 2 5 Analyzer Shut Down e Leave the last analyzed punch in the boat with the boat positioned in the Calibrate position This punch will be used as the system blank the following morning and then taped to the corresponding thermogram DRI STANDARD OPERATING PROCEDURE Page 47 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aer
45. e deposit or any other conditions which might affect analysis results 3 2 3 2 Loading the Filter Punch Begin the analysis by clicking on Analysis in the Welcome window and inputting the Sample ID Run Punch area and Deposit area After the boat has cooled to 50 C or less remove the previously analyzed sample punch and load the current sample punch Clean the tweezers flat glass plate and punching tool with a dry KIMTECH wiper Replace the Petri slide or Petri dish containing the filter into the Styrofoam cooler or refrigerator Upon analysis completion use a small piece of transparent tape to attach the sample punch to its thermogram ensuring that the deposit side is facing up At the end of the analysis the push rod will automatically be pulled back to the Calibrate position to begin cooling 3 2 3 3 Post Analysis Examine the thermogram for proper laser response temperature profiles realistic carbon peaks and the presence of the calibration peak at the end of the analysis Examine the tabular printout to confirm that the calibration peak counts are within specifications see Section 4 1 Finally examine the laser signal at the end of the run DRI STANDARD OPERATING PROCEDURE Page 46 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Indicate successful analyses on the sample analysis list by re
46. ed above these changes are reversible and temperature dependent Some colored organic compounds can affect the laser correction causing increased reflectance or decreased transmittance as these compounds are removed This effect is ascertained by examining the laser response during the organic portion of the analysis The split between OC and EC should be examined manually if the effect is large The presence of certain elements Na K Pb Mn V Cu Ni Co and Cr existing either as contaminants on the filters e g glass fiber filters or as part of the deposit material has been shown to catalyze the removal of EC at lower temperatures Lin and Friedlander 1988 Such catalysis would affect the distribution of carbon peaks during the analysis If present in sufficient levels water vapor either contained in the deposit or remaining after acidification of the sample punch can shift the FID baseline To eliminate this effect allow the sample punch to dry in the analyzer by passing carrier gases over it before starting the analysis 14 Ranges and Typical Values of Measurements Source dominated or heavily polluted environments which would normally have carbon concentrations above the working range of the carbon analyzer may be sampled and analyzed within the range of the carbon analyzer by increasing the filter deposit area or by decreasing the sampling flow in the field equipment Deposits that are very black such that the initial ref
47. edure SOP including routine system calibration actual analysis and immediate review of the data as it is produced in order to correct system problems The responsibilities of the laboratory manager or supervisor are to ensure that the carbon analyses procedures are properly followed to examine and document all replicate standard and blank performance test data to designate samples for reanalysis to arrange for maintenance and repair to maintain the supplies and gases necessary to ensure uninterrupted analysis and to deliver the analysis results in database format to the project manager within the specified time period The quality assurance QA officer of DRI s Division of Atmospheric Sciences is responsible for determining the extent and methods of quality assurance to be applied to each project to estimate the level of effort involved in this quality assurance to update this procedure periodically and to ascertain that these tasks are budgeted and carried out as part of the performance on each contract 1 7 Definitions for IMPROVE_A Thermal Protocol for Carbon Analysis The following terms are used in this document IMPROVE A A thermal protocol is used in carbon analyzers to quantify carbon Thermal Protocol fractions evolved at different temperature plateaus The IMPROVE A thermal protocol derives from the Interagency Monitoring of Protected Visual Environments IMPROVE thermal protocol initiated in 1987 Chow et al 2005
48. edure assumes carbonate will not be measured For carbonate analysis refer to Section 3 1 5 2 Special Analysis Always execute the cmdBakeOven command to bake the oven before beginning analysis each day refer to section 3 1 3 This will ensure the system is clean 1 5 ug OC 0 5 ug EC and 2 0 ug TC per cm Run a system blank with the IMPROVE A protocol 3 1 4 1 Analysis Preparation Verify the computer date and time is correct Verify sample oven pressure reading and specified flow ranges in the front panel flow meters Wipe the flat glass plate tweezers and punching tool thoroughly with a dry KIMTECH wiper Based on the analysis list remove the sample to be analyzed from the Styrofoam cooler or refrigerator Record the filter ID in the analyzer log book Figure 2 7 Open DRICarb exe from the c CarbonNet folder or use the desktop shortcut to DRICarb exe Verify correct software version number on the Welcome screen Select Analysis from the Main submenu of the Welcome form This will initiate the analysis protocol as shown in Figure 3 3 You can also select FID ID typically FID 8 to determine the FID peak sensitivity In the analysis Setup form enter Sample for the Type Polarity should default to Unipolar DRI STANDARD OPERATING PROCEDURE Page 35 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Sa
49. eposit area should be 1 Use an ID number derived from the current date e g SB071805 for July 18 2005 Calculated carbon concentrations should not be more than 0 2 ug carbon Values greater than this warrant additional system blanks Samples may not be analyzed until the system blank is 0 2 ug carbon 3 1 5 2 Carbonate Analysis Enter the Sample ID Run Punch area and Deposit area Select cmdImproveA_Carbonate from the Command Table drop down field and start the analysis program Follow the steps under Section 3 1 3 until the sample punch is loaded into the boat Load sample and click OK When asked if you want to delay or continue analysis click OK After 90 seconds the punch automatically centers under the acid injection port The computer will prompt you to inject the HCl and then will state Load syringe and XX seconds to acid injection Inject 20 ul of 0 4 M HCI through the septum port to the sample Prior to acidification approximately 90 seconds elapsed analysis time flush the 25 ul syringe with 0 4 M hydrochloric acid HCl into a waste beaker When prompted to inject HCl eject 20 ul HCI onto the filter punch ensuring that the needle bevel is turned toward the punch and that the needle tip is touching the top of the punch DRI STANDARD OPERATING PROCEDURE Page 42 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filte
50. erlag Wein Graz Austria pp 141 165 Novakov T 1982 Soot in the atmosphere In Particulate Carbon Atmospheric Life Cycle G T Wolff and R L Klimisch Eds Plenum Press New York NY pp 19 41 Rau J A 1986 Residential wood combustion aerosol characterization as a function of size and source apportionment using chemical mass balance modeling Ph D Dissertation Oregon Graduate Center Beaverton OR Rosen H Hansen A D A Dod R L Gundel L A and Novakov T 1982 Graphitic carbon in urban environments and the Arctic In Particulate Carbon Atmospheric Life Cycle G T Wolff and R L Klimisch Eds Plenum Press New York NY pp 273 294
51. ersion of the software and database are being used For normal DRI STANDARD OPERATING PROCEDURE Page 30 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 analysis the database should be CarbonNetWorkXX mdb where xx is the analyzer number Figure 3 1 DRI Welcome Screen HE Welcome MME desert Em EE f research The Environmental Analysis Facility g b institute DR rhermalOptical Carbon Analyzer Model 2001 university and community Program DriCarb Version 6 0 Revision date 20050421 Main gt Utilities Backup Analysis Empty Tables Manual Recall Data ra eng xit Help e With the breech closed flip the oven outlet toggle on the side of the analyzer down off and let the sample oven pressure reach 5 psi Most systems are working in the 2 5 psi range but a leak is easier to detect when the pressure is at 5 psi Close the oven inlet toggle on the front of the analyzer and watch for a decline in the sample oven pressure A leak free condition is indicated by a steady pressure reading or a decline of 0 01 psi or less per second 0 01 ml sec e Ifthe pressure is stable flip the outlet oven toggle and then the inlet oven toggle back to the on position up This is to avoid pressurizing the oven if the inlet toggle is flipped first e If the pressure is not s
52. he thermogram is performed by the computer at the end of the analysis program For reference purposes the calculation is peak ug C punch integrated peak counts above baseline calibration slope internal calibration n counts For IMPROVE A thermal protocol the peaks reported are four organic peaks OC1 OC2 OC3 and OC4 corresponding to 140 280 480 and 580 C in He atmosphere respectively three elemental carbon peaks EC1 EC2 and EC3 corresponding to 580 C after the introduction of O2 740 and 840 C respectively and three pyrolyzed organic carbon peaks Lower Regular and Upper Splits by reflectance and transmittance corresponding to the peaks after the introduction of O and before the Lower Split Time Regular Split Time and the Upper Split Time respectively for the reflectance and transmittance optical charring correction see Section 1 7 and Figure 3 4 The EC reported includes pyrolyzed carbon Carbon values per punch are converted to ug C cm by ugC punch Com i puncharea Finally carbon values are converted to ug C filter by ugC filter ug C cm tilterdepositarea DRI STANDARD OPERATING PROCEDURE Page 59 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Precision calculations are performed as the data is merged into the final database Section 5 5 4 2 Instrumen
53. iffusion into the system These O levels are determined quarterly using a gas chromatography mass spectrometry GC MS instrument on the analyzer When a fresh He cylinder is installed the O2 level should be checked in one of the analyzers hooked up to the gas line to assure the quality of the gas supply and the condition of the O2 scrubber If the cmdAutoCalibCheck command is used for calibration the condition of the catalysts will be indicated and appropriate action can be taken such as catalyst replacement All calibrations repairs and checks must be recorded in the Carbon Analyzer Logbook Figure 2 7 Flow rates of all operating gases should be checked DRI STANDARD OPERATING PROCEDURE Page 22 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 and adjusted if needed whenever a new quartz oven or methanator is installed or serviced Additionally a flow balance should be performed as well 2 1 4 Spare Parts The following spare parts must be kept on hand to ensure minimal interruptions in carbon analysis e Quartz furnace tube available from the manufacturer Atmoslytic Calabasas CA e Quartz rods 3 mm nominal diameter optical quality Atmoslytic Calabasas CA polished for optical clarity with 104 mm upper arm and 119 mm lower arm lengths The version of the carbon analyzer manufactured in 2001 uses a 98 mm
54. ined for the temperature calibration to hold The analyzer must not be used if the sample boat shifts position or becomes loose in its holder see Figure 2 3 Carbonate carbon may bias carbon concentrations if it constitutes more than 5 of total carbon in the ambient or source sample Carbonate carbon may be measured as either OC or EC depending on the chemical nature of the carbonates and their thermal decomposition temperatures Acid pretreatment of filter samples can eliminate the carbonate interference Novakov 1981 1982 Rosen et al 1982 Carbonate carbon has been found at only a few IMPROVE monitoring sites and then at levels that do not appreciably bias OC and EC concentrations Chow and Watson 2002 The presence of certain minerals in some soils can affect the laser correction for pyrolysis These minerals change color as the sample punch is heated generally resulting in a darker DRI STANDARD OPERATING PROCEDURE Page 9 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 sample For samples which contain large fractions of resuspended soils the split between OC and EC should be examined manually Some minerals again predominantly in soil samples or soil dominated samples may affect the laser correction by temporarily changing color or changing the surface texture of the deposit residue Unlike the effect describ
55. ision 3 o If the analysis has some anomaly but still appears to be legitimate either flag or add notes to the comments field as appropriate o Analysis flags are defined in Section 6 4 o All samples flagged as invalid must have an entry in the comments field to describe the reason that the sample is invalid Typical notes and comments are presented in Table 6 3 e Scan the OC and EC columns looking for unusually high or low values At this time make sure that the field blanks and or lab blanks are all close to one another Circle any possible outliers for further investigation e Compare replicates against original run The values should meet the following criteria Range Criteria OC EC amp TC lt 10 ug cm lt 1 0 ug cn OC and TC gt 10 ug cm 10 of average of the 2 values EC gt 10 ug cm 20 of average of the 2 values e Check the OC TC ratio Typically it should not be less than 0 65 Circle any possible outliers for further investigation e Scan for records where EC is greater than OC These may require additional investigation depending on loading and sample source Circle records for further investigation e Scan blanks for OC being greater than 3 95 deposit area and for EC greater than the deposit area Rerun any unusually high blanks e Compare primary and secondary filters for validity Secondary filters should have OC and EC measurements less than the corresponding primary filter Secondary filters
56. it Time Laser FID Split Time Lower split 277 sec 1681 millivolts 295 sec Regular split 278 sec 1681 millivolts 296 sec Upper split 279 sec 1683 millivolts 297 sec Transmit Split Time Laser FID Split Time Lower split 311 sec 501 millivolts 329 sec Regular split 315 sec 500 millivolts 333 sec Upper split 320 sec 501 millivolts 338 sec Peak Area Carbon ocl oc 2184 mv secs 3 49 ug C cm2 12 29 ug C filter OC2 oc 2156 mv secs 3 44 ug C cm2 125 L5 ug C filter 0c3 oc 1808 mv secs 2 89 ug C cm2 10 19 ug C filter oc4 oc 2704 mv secs 4 32 ug C cm2 15 25 ug C filter ocs oc 20 mv secs 0 03 ug C cm2 LE ug C filter EC1 EC 112 mv secs 0 18 ug C cm2 63 ug C filter EC2 EC 90 mv secs 0 14 ug C cm2 Rut ug C filter EC3 EC 115 mv secs 0 18 ug C cm2 65 ug C filter EC4 EC 94 mv secs 0 15 ug C cm2 253 ug C filter ECS EC 0 mv secs 0 00 ug C cm2 00 ug C filter LRPyMin Py 759 mv secs 1 21 ug C cm2 4 28 ug C filter LRPyMid Py 732 mv secs 1 17 ug C cm2 4 13 ug C filter LRPyMax Py 707 mv secs 1 13 ug C cm2 3 98 ug C filter LTPyMin Py 182 mv secs 29 ug C cm2 1 02 ug C filter LTPyMid Py 151 mv secs 24 ug C cm2 85 ug C filter LTPyMax Py 123 mv secs 20 ug C cm2 5 69 ug C filter de de kc e he te He ede eoe de eee de ehe de e ke he e che ee e he de e dede ee e ke ke ke e oe e e e he IT e e e ke he e e e e e e e e e e IKK e e DRI STANDARD OPERATING PROCEDURE Page 72 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number
57. l KHP and sucrose solution do twice 20 ul KHP and sucrose solution no injection as a system blank see Section 3 1 5 1 20 ul acidified Nanopure water only check of background level of Nanopure water Flush the syringe at least three times with the calibration solution before taking up the volume for injection Pump the syringe plunger to remove any trapped bubbles Slowly spike the solution in the center of quartz punch and wash the syringe with Nanopure after use If the solution is spiked too quickly it will bead up and run off the punch Click OK on the analysis Setup screen The boat will load to the calibration position e The computer will ask Would you like to proceed or would you like to delay analysis Enter the length of time in seconds you wish to delay the beginning of the analysis in the Delay box This is used to purge dry a filter disc that has been deposited with an aliquot of KHP or sucrose standard solution or when the sample is acidified for carbonate removal In general allow 1 minute of purge time for every ul of solution deposited e g 5 wl 300s 10u1 600s 15u1 900s and 20411 12005 Click OK and analysis will begin Allow the punch to dry thoroughly the punch will turn from translucent to opaque as it dries The punch must be dry to avoid water vapor effects on the FID and the laser reflectance and transmittance signals Select the cmdImproveA option from the analysis menu to star
58. lectance is close to zero provide a less precise OC EC split because additional blackening due to OC charring is not quantified by the reflected light The carbon analyzer can effectively measure between 0 05 and 750 ug carbon cm for a typical punch size of 0 5 cm The upper limit depends on the particular compounds on the filter and the temperatures at which they evolve This upper limit may be extended by reducing the punch size or extending analysis times at lower temperature plateaus to avoid an over range FID signal DRI STANDARD OPERATING PROCEDURE Page 10 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Typical carbon values range between 10 and 100 ug carbon cm for 24 hour ambient samples The distribution between OC and EC depends on the particulate source types ranging from negligible levels of EC e g secondary sulfate to 80 or more EC e g diesel exhaust 15 Typical Lower Quantifiable Limits Precision and Accuracy The lower quantifiable limits LQLs of thermal carbon methods depend on the variable carbon content of the blank quartz filters as well as the analysis method For better LQLs the unexposed filters should be pre fired in an oven at high temperatures for several hours to remove any residual carbon contamination Fung 1986 Huntzicker 1986 Rau 1986 DRI 2004 All quartz fiber filters origin
59. librium condition that allows the phase transition point to be resolved When the specified temperature is reached the Tempilaq G liquefies causing a sharp change in reflectance and transmittance DRI STANDARD OPERATING PROCEDURE Page 63 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Figure 5 1 demonstrates the thermocouple temperature reflectance and transmittance as a function of thermal analysis time In the example provided the reflectance and transmittance remained relatively flat until the temperature approaches its specified value of 184 C Figure 5 2 compares the time series of reflectance transmittance and their respective first and second order derivatives The second order derivative change in the slope recorded the inflection point of reflectance or transmittance that provided the best indication of the attainment of the designated temperature Thermocouple temperature at this critical point was recorded as measured temperature The temperature deviation AT between the sample and the thermocouple temperatures is determined by comparing the rated Tempilaq G temperature with this measured value Chow et al 2005 In the Model 2001 the reflectance based method generally gave a lower liquefying temperature than the transmittance based method within 2 C Given the uncertainty in the Tempilaq G
60. ly such standards should include a range of organic compounds from low to high molecular weights and with varying degrees of susceptibility to pyrolysis as well as EC and carbonate compounds Currently KHP sucrose and the two calibration gases are used at DRI for calibration and system audit purposes 4 1 3 2 Calculating Calibration Slope The calibration slopes derived from the two gases and the KHP and sucrose spiked filter punches are averaged together to yield a single calibration slope for a given analyzer This slope represents the response of the entire analyzer to generic carbon compounds and includes the efficiencies of the oxidation and methanator zones and the sensitivity of the FID Note that the current calibration procedure is based only on TC as no routine procedure exists to check the accuracy of the OC EC split An example of the spreadsheet is shown in Figure 4 3 DRI STANDARD OPERATING PROCEDURE Page 55 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE_A Revision 3 Figure 4 3 Example of calibration summary worksheet used to determine calibration slope AA eel oe Standard Run uL m Peak Counts Peak Counts Calibration Slope MIO0719 3 MIO0719 4 MI0719 5 MIO0719 6 MI0719 7 MIO0719 11 Slope 21 05 Scatter 0 1619 Correlation 0 99857 4 1 3 3 Typical Accuracy of Calibration Standards The accu
61. m the Main submenu From the Manual screen select Control on the drop down menu list Note analyzer must be in Auto mode for the manual control to work This will bring up the Control screen Change the Oven Temperature field to 1000 Change the Sample Position field to Analyze Click Go This will heat the oven to approximately 950 C depending on the instrument s calibration Exercise caution when working around hot surfaces of the analyzer Repeat until the system is clean Sample runs or calibrations may then begin System blanks section 3 1 4 1 are run after the oven bake DRI STANDARD OPERATING PROCEDURE Page 32 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 3 1 3 2 Automatic Oven Bake From the main welcome screen select Analysis Set Type to Sample and select cmdBakeOven from the drop down menu in the Command table field The project name should be SAMPLE and the sample ID should be in the format Cxx Y Y Y Y MMDD where xx is the analyzer number e g C0720050715 for analyzer number 7 on July 15 2005 Set the Run Punch area and Deposit area fields to 1 Click OK then Run Repeat until the system is clean Sample runs or calibrations may then begin System blanks Section 3 1 5 1 are
62. me Batch and Sub batch The Project Name should be CALIB and the Batch should be SUKHPCAL A clean blank quartz punch is baked in the analyzer oven at 900 C for 10 minutes using cmdBakeOven from the Command Table After baking the quartz punch change to cmdImproveA under Command Table Perform system blank before running KHP or sucrose Enter the Sample ID number or place your mouse cursor in the field The sample ID should be in the format SUxY Y Y Ymmdd zzz where SU is for sucrose spiking use KHP for KHP spiking x is the analyzer number and zzz is the volume 05 10 15 or 20 ul You can also select FID ID typically FID 8 to determine the FID peak area and make comments and flag the analysis from this screen before the analysis starts Enter the Run the Punch area and Deposit area should be 1 for the filter being analyzed Enter technician initials in the Tech initials field DRI STANDARD OPERATING PROCEDURE Page 61 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 e After the punch has cooled to less than 50 C the KHP or sucrose solution prepared as described in Section 4 1 3 1 and kept at room temperature is injected onto the punch using a 25 ul syringe The following volumes are used e 5 ul KHP and sucrose solution 10 ul KHP and sucrose solution 15 u
63. mples Method IMPROVE_A Revision 3 e Fill out the information about the sample including Project Name Batch and Sub batch e Under Command Table select cmdImproveA e Enter the Sample ID number or place your mouse cursor in the field and use a barcode scanner to read the barcode on the Petri dish e Enter the Run Punch area and Deposit area for the filter being analyzed e Enter technician initials in the Tech initials field e Select any pre analysis flags from the drop down menu in the Flags field A list of valid choices is presented on the screen e Visually examine the filter and note any non uniformity or unusual deposit Remove it from the Petri slide or Petri dish with tweezers handling the filter only by the edge Place the filter on the flat glass plate and remove a sample punch by pushing down gently on the punching tool Rocking the punching tool slightly will ensure that the punch is completely severed Try to remove the punch from the edge of the deposit to avoid wasting the filter while trying to avoid areas of non uniform deposits e Leaving the sample punch in the punching tool place the punching tool on a clean KIMTECH wiper Return the filter to the Petri slide or dish being careful to handle only the filter with the tweezers e If this is the first run of the day or if the analyzer has been cooled down the analyzer will verbally prompt you to load the punch Please load filter anal
64. ns are not necessarily the same owing to charring of organic vapors adsorbed within the quartz fiber filter Chow et al 2004 Chen et al 2004 OC and EC determined by both methods are reported Carbonate carbon can be determined by measuring the CO evolved upon acidification of the sample punch before the normal carbon analysis procedure Seven temperature fractions as well as the TOR and TOT charring correction are individually quantified and reported when the IMPROVE Chow et al 1993 2001 temperature protocol is applied Values routinely reported include total OC total EC total carbon TC sum of total OC and total EC and pyrolized carbon monitored by both reflectance OPR and transmittance OPT Depending on the thermal optical protocol applied for quantification thermally derived subfractions of OC and EC and of carbonate carbon are reported 1 3 Measurement Interferences and Their Minimization Precision of thermal optical carbon analysis depends on the sample temperature in the analysis Therefore the correlation between sample temperature and thermocouple temperature should be established and calibrated semiannually so that the thermal protocol can truly reflect the sample temperature during the analysis Chow et al 2005 The thermocouple s position in relation to the sample as well as the different heating properties of the thermocouple and the sample govern the temperature offset This relationship must be mainta
65. oat during the run e Boat movement due to loose boat holder e A quartz rod laser light pipe ready for replacement As quartz is heated to high temperatures devitrification white deposits of SiO2 occurs that leads to a decrease in the laser intensity The end surface becomes frosty The bottom light pipe also receives droppings of quartz particles from filter discs during analysis Thus the bottom light pipe will deteriorate faster than the upper light pipe Microscopic cracks in the quartz rod will increase internal reflectance of the laser light as the number of these cracks multiply the effect of temperature on these cracks and thus on the reflectance becomes an interference in the laser signal As described in Section 4 1 the calibration peak at the end of each analysis run serves as a regular standard the integrated area under the calibration peak serves as a measure of analyzer performance In addition the daily injections of two calibration gases further serve as standards Primary standards in the form of NIST traceable spiked filter punches do not exist The cmdAutoCalibCheck command check allows the condition of the catalysts to be monitored and verified DRI STANDARD OPERATING PROCEDURE Page 67 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 6 2 Reproducibility Testing Replicates of analyzed samples are
66. of Aerosol Filter Samples Method IMPROVE A Revision 3 the time of analysis need to be taken into account For a 100 CH4 or CO standard at 760 mm Hg at 20 C each microliter 0 499 ug carbon For a 5 standard it will be 0 02495 ug carbon ul at standard temperature and pressure STP 20 C 760 mm Hg The Ideal Gas Law should be used to correct for the temperature and pressure of the laboratory Acualug C per wl a Jeteor cal gas x 12 760 T 273 15 A 0 08206 where Pa is pressure in mmHg T is ambient temperature The calibration gases are traceable to NIST standards The calibration gases are assayed for exact concentrations by the gas supplier the assay value is obtained from the tag on the cylinders and is typically determined by gas chromatography GC To prepare an 1800 ppm standard the KHP is dried at 110 C for two hours before dispensing Transfer 0 3826 g of KHP into a glass 100 ml volumetric flask after the KHP has come to room temperature inside a desiccator The weight of KHP used must be recorded Dilute to volume with 0 4 M concentrated hydrochloric acid HCl and 99 6 ml Nanopure water Mix the KHP thoroughly Store this solution in a refrigerator until it is used for calibration purposes This solution is good for 40 days Label the flask with the chemical name the date of preparation the name of the chemist preparing the solution and the exact concentration The concentration nominally 1800 ppm c
67. om the filter punch in a He only gt 99 999 atmosphere from 480 to 580 C The carbon evolved from the time that the carrier gas flow is changed from He to 98 He 2 O at 580 C to the time that the laser measured filter reflectance OPR or transmittance OPT reaches its initial value A negative sign is assigned if the laser split occurs before the introduction of O Carbon evolved from the filter punch in a 98 He 2 O atmosphere at 580 C Carbon evolved from the filter punch in a 98 He 2 Oz atmosphere from 580 to 740 C Carbon evolved from the filter punch in a 98 He 2 On atmosphere from 740 to 840 C The time at which the laser measured reflectance and or transmittance of the filter punch reaches its initial value All carbon evolved from the filter punch between ambient and 840 C under He and 98 He 2 O atmospheres The time at which the laser measured reflectance and or transmittance of the filter punch reaches its initial value plus the precision of the laser signal currently defined as 10 counts DRI STANDARD OPERATING PROCEDURE Page 14 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 1 8 Related Procedures Standard Operating Procedures SOPs related carbon analysis activities and other manuals that should be reviewed in conjunction with this document are DRI SOP 6 001
68. osition 6 3 Control Charts and Procedures Control charts are updated at the beginning of each month These charts include a month of calibration data and are posted in the carbon room until the end of the month after which they are filed with the raw analysis results The control chart gives a plot of calibration peak counts as percent deviation from a historical mean versus date Instances where the calibration peak area deviates by more than 10 from the historical mean must be investigated and the cause must be corrected The historical mean covers results from the previous three months and is updated either quarterly when the CH4 DRI STANDARD OPERATING PROCEDURE Page 68 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 calibration gas is changed when the catalysts are renewed or when extensive repairs are performed 6 4 Analysis Flags During Level I validation see Section 6 5 unusual conditions of the deposit or analysis problems are noted on the analysis printouts Errors in pre analysis data entry e g in filter ID punch size deposit area are corrected Flags are applied to the Access file created from the analysis results see Section 6 5 The analysis flags commonly used are presented in Table 6 1 Note that all results flagged with v must include a description of the reason for invalidating the
69. osol Filter Samples Method IMPROVE A Revision 3 e Execute the cmdAutoCalibCheck command to verify the analyzer s performance e When the analysis is complete record the calibration peak counts and calculated injection calibration in the logbook Any values outside the ranges defined in Section 4 1 should be investigated and rerun e Backup the day s data files to disk if not automatically backed up on a server e Remove the printouts and attach them to a file folder labeled with the date and analyzer number Place on the lab supervisor s desk e Turn off the computer monitors e Make a final check of the gas cylinder pressures e Move the samples and blue ice in the Styrofoam cooler or refrigerator back into the sample storage freezer and verify that the freezer is completely closed e If the 25 ul syringe was used for carbonate analysis thoroughly rinse the syringe with Nanopure water and tightly cap all solutions Store at 4 C e Lock the carbon analysis room 4 QUANTIFICATION 4 1 Instrument Calibration The calibration procedures for the carbon analyzers are of three types 1 the end of run calibration peak 2 the routine beginning and end of day calibration injections of He CHy4 and He CO you can also use the auto calibration check using the cmdAutoCalibCheck command and 3 full instrument calibration performed every six months using KHP sucrose and the two calibration gases 4 1 1 End of Run Calibrations
70. performed at the rate of one per group of ten samples The replicate is selected randomly and run immediately after a group of ten is completed It is run on a randomly selected analyzer it can be the same or a separate analyzer as the original This practice provides a better indication of potential differences if samples are analyzed by different laboratories The ug cm values for OC EC and TC are compared with the original run The values should fall into the following criteria Range Criteria OC EC amp TC lt 10 ug cm 41 0 ug cm OC and TC gt 10 ug cm lt 10 of average of the 2 values EC gt 10 ug cm 20 of average of the 2 values Notice that the criteria converge at 10 ug cm Replicates which do not fall within the above criteria must be investigated for analyzer or sample anomalies Analyzer anomalies include poor response as reflected in the calibration peak areas or poor laser signals affecting the splits between OC and EC Typical sample anomalies include inhomogeneous deposits or contamination during analysis or from the field sampling location Inconsistent replicates for which a reason cannot be found must be rerun again When samples are analyzed with an automated sample loader the sample chamber tray will be set up such that the loader number six location will be a replicate of the first sample the 12th location a standard spike of sucrose or KHP and the 18th location a replicate of the sample in loader p
71. r LRPyMax Py 624 mv secs 1 03 ug C cm2 3 63 ug C filter LTPyMin Py 788 mv secs 1 30 ug C cm2 4 59 ug C filter LTPyMid Py 811 mv secs 1 34 ug C cm2 4 72 ug C filter LTPyMax Py 835 mv secs 1 38 ug C cm2 4 86 ug C filter FIR RIT I TTT NARA RARARAAAARA RARA wv DRI STANDARD OPERATING PROCEDURE Page 40 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Figure 3 5 Graphical output from DRI Model 2001 Thermal Optical Analysis Program Sample ID Q80292 Analysis ID Q80292 1 0EC Analysis Date 07 19 05 Tech SMS C A 9 10 LaserT O 2500 FID_8 TE 400 9 LaserR O 2500 FID 6 0 50 1050 1200 1350 Time seconds DRI STANDARD OPERATING PROCEDURE Page 41 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 3 1 5 Special Analysis 3 1 5 1 System Blanks System blanks are run at the beginning of each day Follow the steps outlined in Section 3 1 3 on oven baking with the following exceptions Go through all the steps for a routine analysis Section 3 1 4 but when prompted to load filter punch remove the filter from the previous day and leave the boat empty for the analysis Use project name Sysblk Batch mm for the month and Sub batch dd for the day Punch area and D
72. r Samples Method IMPROVE A Revision 3 When the analysis is underway flush the syringe with Nanopure water to prevent corrosion of the syringe plunger After analysis the program will delay any further analysis for 900 seconds to allow the punch to dry After the carbonate analysis is completed a tabular summary and a copy of the graph will be printed similar in format to Figures 3 4 and 3 5 Select cmdImproveA from the Command Table drop down field and click OK Click Run on the analysis Setup screen The program will automatically cycle into the normal OC EC analysis using the same Sample ID Heat from the oxidation oven will dry the sample in this position for approximately 15 minutes without prematurely baking carbon from the sample the sample temperature should not exceed 42 C When the punch is dry proceed with normal OC EC analysis 3 1 6 Analyzer Shut Down After the final sample for the day is analyzed shut down the analyzers using the following procedures Leave the last analyzed punch in the boat with the boat positioned in the Calibrate position This punch will be used as the system blank the following morning and then taped to the corresponding thermogram Perform end of the day calibration gas injection routine or use cmdAutoCalibCheck command and record the calibration peak counts Any values outside the expected ranges should be investigated and rerun Because low values from the end of day calibr
73. racy of the calibration standards is primarily limited by the accuracy of the calibration gas assays the accuracy of the preparation of the KHP and sucrose solutions and the technician s injection technique The calibration slopes determined by these four compounds historically differ by less than 5 on a given analyzer if sufficient care is taken during the calibration procedure Section 4 1 Figure 4 4 shows an example of plotted calibration curves DRI STANDARD OPERATING PROCEDURE Page Date Title DRI Model 2001 Thermal Optical Carbon Analysis Number TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision Figure 4 4 Example DRI Carbon Analyzer Calibration Curves CA 7 Calibration 56 of 79 July 15 2005 2 216 2 3 40 y 21 055x 35 30 m a Calculated Carbon ug a 3 0 00 0 20 0 40 0 60 0 80 1 00 1 20 1 40 Peak Area Calibration Peak Area no units 4 1 3 4 Data Treatment for Calibration Data Calibration values are plotted as actual ug carbon vs the ratio of the integrated sample peak counts to the calibration peak counts Figure 4 1 Obvious outliers are identified and rerun Linear regression is performed on each set of calibration data separate calculations for KHP sucrose CH4 He and CO He The slope m is calculated from x O x Y 6 1 80 2 00 DRI STANDARD OPERATING PROCEDURE Page Date Title DRI Model 2001 Thermal Op
74. rod for the upper arm Measure the old rods for reference e Catalyst ovens Watlow 15 24 cm length 2 54 cm tube diameter element from the analyzer supplier e Quartz boats Atmoslytic Calabasas CA e Thermocouple rods 24 13 cm length by 0 32 cm outside diameter OD Type K ground isolated with Inconel sheath Omega Part TJ36 CAIN 18E 9 5 Stamford CT Remove 1 cm of the sheath with a file to obtain the longer tip needed in this application e FID flame tips for Gow Mac 12 800 FIDs Gow Mac 132 117 Bethlehem PA e Septa Standard 0 32 cm or 0 64 cm cylindrical Alltech Associates 6524 Deerfield IL for injection ports Silicon septa 0 25 cm and 1 11 cm for oven seals Alltech 15427 and 15429 Deerfield IL e 25 ul syringes e Replacement needles for syringes Alltech 7729 06 Deerfield IL e Replacement oxygen moisture trap R amp D Separations Model OT3 2 Rancho Cordova CA e Replacement hydrocarbon trap R amp D Separations Model HT200 4 Rancho Cordova CA DRI STANDARD OPERATING PROCEDURE Page 23 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE_A Revision 3 2 1 5 Replacement indicating oxygen trap Chromatography Research Supplies Model 202223 Louisville KY Quartz wool For repacking the oxidation oven Alltech Associates 4033 Deerfield IL Viton O rings Size 013 Two needed
75. ry pressures are correct Hydrogen H2 15 psi Helium He 15 40 psi check label on regulator for current setting Compressed air 15 psi for FID 25 psi for breech actuation O He mix 15 psi CH4 He mix 10 psi COy He mix 10 psi e Check that the FID is lit by holding a pair of tweezers over the FID exhaust stack and watching for condensation If the FID is not lit as immediately after the Hz or compressed air cylinders are changed relight the flame by turning the H rotameter to the upper limit as posted on the flow meter and holding a butane lighter or match over the FID stack A soft pop indicates that the flame has been lighted Verify that the flame remains lit by the tweezers test Often the flame will not stay lit the first try especially after the H cylinder is changed and air gets into the gas lines Return the rotameter to the operation setting after the flame is lit DRI STANDARD OPERATING PROCEDURE Page 29 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 Check and readjust if necessary all gas flows at the analyzer The correct readings are posted on each rotameter Read through the center of the ball If drastic adjustments are required on one analyzer recheck that flows on the other analyzers have not been affected Turn on the computer monitor Note the computers are generally left on at all times
76. s sss sese sese eee 49 4 1 3 Full Calibration Description it aie e eate a see tee ug 52 4 1 3 1 Preparation Ranges and Traceability of Standards sss 32 4 1 3 2 Calculating Calibration Slope cre ee RE Atene sq ct ei cet ORE 54 4 1 3 3 Typical Accuracy of Calibration Standards sese eee 55 4 1 3 4 Data Treatment for Calibration Data sss sese 56 41 3 5 Calculations de ii Rap sit ase tuo tea exi ln ae is nd fut cc a Ped aati 58 4 2 Instrument Calibration Instructions sss sese ee eee eee 59 4 2 1 Full Gas Cabra ta 59 4 2 2 Full Sucrose and KHP Calibrations divas 60 5 TEMPERATURE CALIBRATIONS iate tea coe dioe eu s n 62 5 1 Temperature Indicators A O a tst EO Na a EOM EU aE 62 5 1 1 Statidard ETT 62 5 1 2 Temperature PLO GCA Dirt 62 6 QUALITY CONDREO I 2 s echt eet eli E a A E a t aset aa 65 6 1 Performance Testing m Re M i a EE E EA RT ROEE 65 6 2 Reproducibility Testing sates oe sese eee eee eee 67 6 3 Control Charts and Procedure 67 GA Analysis TTT 68 6 5 Data Validation and Feedback 69 DRI STANDARD OPERATING PROCEDURE Page 4 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 6 5 1 Daily VAG at OM cese A A cds 69 6 5 1 1 Validation of Final Data File ee li o ene nia 72 7 REFERENCES ua daa 78 DRI STANDARD OPERATING PROCEDURE Page 5 of 79 Date July 15 2005 Title DRI Model 2
77. sai QN BACH BaN SH oos e a op 17 Raters 133 IC 13 E pps 2 JBI tm 13 BARI IZ Pas ZOD El 4 G 6012 1 IYd2R 6G6dO7 l BEEF ET 56050512 Page Date Number Revision LS IN uorttecUcous SAmpte DEPOSIT 27 of 79 July 15 2005 2 216 2 3 DRI STANDARD OPERATING PROCEDURE Page 28 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 3 PROCEDURES 3 1 Detailed Procedures 3 1 1 Analyzer Start Up If the analyzer is started up for the first time or after an extended period of non operation it will take a period of conditioning to reach a stable system background At the start allow all the gases to purge through the system for 30 minutes before heating the various zones in a stepwise manner Allow the FID and Line Heaters to reach operating temperatures of 120 C and 105 C respectively before heating up the oxygenator and methanator Heat both catalysts at 120 C for about half an hour then in 100 C increments with 30 min hold time until the final temperatures of 912 C and 420 C are reached for the oxygenator and methanator respectively The following steps outline analyzer start up e Check all gas cylinder pressures cylinders with gas pressures less than 500 psi should be replaced before beginning the day s analysis e Check that all gas delive
78. set temperature to 900 C and let analyzer reach this temperature With breech closed flip oven outlet toggle to the off position and let sample oven pressure reach 5 psi Flip the oven inlet toggle to the off position and watch for a decline in the sample oven pressure If system is leak free the pressure will remain stable If the system is not stable use a He leak detector to locate the leak disassemble and reassemble the port fitting if necessary and check the O ring for correct placement and pressure When system is stable flip both the inlet and outlet oven toggles back to the on position An oven bake can also be executed without a leak check by selecting cmdBakeOven from the drop down menu in the Command table Update the Daily Analyzer Checklist and verify values are within expected range DRI STANDARD OPERATING PROCEDURE Page 45 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 3 2 3 OC EC Analysis 3 2 3 1 Analysis Preparation e Clean the tweezers flat glass plate and punching tool with dry KIMTECH wiper e Based on the analysis list remove the sample to be analyzed from the Styrofoam cooler or refrigerator Verify the sample ID against the analysis list e Remove a sample punch from the filter e Record the Sample ID in the analyzer logbook along with any comments on the condition of th
79. sing data Scan the deposit area column for incorrect entries Circle the incorrect entries to ensure that corrected values replace those currently in the database e Scan the filter IDs for multiple entries of ID numbers Under normal conditions the only times multiple entries should occur are reruns and replicates All multiple entries must be flagged to indicate the reason for their existence e Scan for missing runs The most common example is the first run being aborted or lost for some reason and the only entry in the MS Excel file is the second run An entry for the first run must be inserted flagged as invalid and labeled as to the reason it was invalid All punches taken from the filters MUST be accounted for and documented in the file e Pull the analysis folders and go through the analysis summaries and thermograms one by one At this time resolve all circled items and all missing flags o Determine if analyses flagged by the operator are legitimate These flags are determined by the operator at the end of the analysis run Section 6 4 and are defined in Table 6 2 o If the temporary flag is not warranted draw a line through the flag to indicate that it should be removed o Ifthe sample should be rerun add it to a rerun list DRI STANDARD OPERATING PROCEDURE Page 75 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE_A Rev
80. t The integrated peak counts for all seven temperature fractions for the sample and calibration peaks are recorded The total peak is calculated by adding the peak area from OC1 OC2 OC3 and OC4 as well as EC1 EC2 and EC3 Pyrolysis counts are not included in the total DRI STANDARD OPERATING PROCEDURE Page 62 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 5 TEMPERATURE CALIBRATIONS Temperature calibrations are performed semiannually on all instruments to verify that the sample temperature is as accurate as possible 5 1 Temperature Indicators Since it is not possible to sense the temperature of the sample directly materials were sought that 1 could be placed where the sample would normally be located and 2 would cause sharp reactions when known temperatures were achieved Quick drying temperature indicating liquids of different melting points Tempilaq G Tempil Inc South Plainfield NJ USA were used as temperature indicators in muffle furnaces A Tempilaq G set contains long chain hydrocarbons suspended in an organic solvent which change their appearance at 44 specific temperatures spanning 80 100 C The accuracy of Tempilaq G is certified within 1 of its designated temperature and is traceable to the National Institute of Standards and Technology NIST Tempilaq G is bottled in liquid form and
81. t Calibration Instructions 4 2 1 Full Gas Calibration To perform the full calibration select Analysis from the Main menu of the DRICarb exe program Welcome screen Choose CALIB under the Type drop down menu In the Command table drop down select cmdCalib He for example Fill out the Sample ID Run and Tech Initials fields The sample ID should be in the format MIxYYYYMMDD zzz where MI is for CH injection use CI for CO injection x is for the analyzer number and zzz is the volume of gas injected You can also select FID ID typically FID_8 to determine the FID peak area and make comments and flag the analysis from this screen before the analysis starts Select the atmosphere for calibration under the Cal Gas menu and select the proper Carrier Gas Verify the command table matches the options selected Enter the technician initials in the Tech initials field Choose OK to run the analysis or Exit to leave the program The analysis will start with the screen as shown in Figure 3 3 You will note that the top portion contains all the information in the Setup screen The bottom half will display the thermogram when the run is initiated Start a run by clicking on the Run command button After the computer states Please load gas syringe flush the gas syringe with the calibration gas at least three times and then load it with the calibration gas Time remaining
82. t require carbonate analysis The precision of carbon analysis has been reported to range from 2 4 Johnson 1981 For analysis of actual ambient and source filters homogeneity of the deposit is most important for reproducible results For homogeneous deposits containing gt 5 g cm 10 times MDL of TC precision is generally 10 or better for inhomogeneous deposits replicates may deviate by as much as 30 The precision of carbonate analysis results is 10 DRI STANDARD OPERATING PROCEDURE Page 11 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 The precision of the laser dependent split between OC and EC fractions depends upon how rapidly the laser is increasing at the time of the split and whether or not the split falls in the middle of a large carbon peak Typically relative laser split times are reproducible within 10 seconds and deviations in calculated splits are lt 5 of the total measured carbon The accuracy of the thermal optical reflectance method for TC determined by analyzing a known amount of carbon is between 2 6 Rau 1986 Precision of the OC EC split is between 5 and 10 This precision is also influenced by the filter loading and source type 1 6 Personnel Responsibilities Before performing carbon analysis all analysts in the laboratory should read and understand the entire Standard Operating Proc
83. table use a He leak detector Alltech Deerfield IL to locate the leak Check the following items and correct accordingly o All ferrules fittings and seals DRI STANDARD OPERATING PROCEDURE Page 31 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 o Quartz oven o All tubing o Thermocouple o Breech O ring If the system still leaks wipe all threads and ferrules with a dry KIMTECH wiper reassemble and retry Also check the breech O ring to ensure that it sits squarely in the groove and that there is sufficient pressure to close the breech Refer to the carbon analyzer s Troubleshooting Manual for additional tips and procedures Once the system passes the leak test make sure that the analyzer s multi function switch at the left of the front panel is set at Auto in order to continue with routine analysis Allow the system pressure to return to its original value and record this value on the Daily Analyzer Checklist shown in Figure 3 2 The pressure should be consistent with previous day s values 3 1 3 Oven Bake A daily oven bake is performed to ensure the system is clean before beginning analysis The oven bake can be performed manually or by using an automated command from the command table Use the following procedures to do a manual oven bake 3 1 3 1 Manual Oven Bake Select Manual fro
84. ters e DO NOT leave the room until the analysis begins e Check the graphical printout after each analysis run to ensure that the 1 FID Flame Ionization Detector 2 temperature and 3 laser signals are behaving as expected Section 4 1 Report any anomalies to the lab supervisor immediately e The quartz oven is susceptible to breakage Care should be taken when handling and cleaning e Be careful that no fiber from the KIMTECH wiper is left on the sample punch tweezers and or glass plate 2 1 3 Maintenance Regular maintenance for the analyzer involves daily checking of compressed gas supplies cleaning the punching tool and tweezers between each sample with dry KIMTECH ensuring that the lab is clean and backing up data files to disc on a daily basis unless files are automatically backed up to server Temperature calibrations for the six temperature plateaus 140 280 480 580 740 and 840 C need to be performed semiannually see details in Section 5 Checks of laser adjustments physical and electrical and leaks are made at least monthly The procedure for leak checks can be found in Section 3 1 2 Additional leak tests are performed with a He leak detector each time a part is replaced and whenever the analyzer fails the leak check during the daily routine The system should show no He leaks at the various connections of the quartz cross oven Since He has high diffusivity freedom from He leaks will safeguard against O d
85. tical Carbon Analysis Number TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision The standard deviation s is calculated by 1 6 mx Y n i Nx Oo where injected carbon peak area calibration peak area and y calculated carbon in spiked filter or manual injection ug e Note that this is a special form of the regression formula which ensures that the curve passes through the origin e The resulting slope is compared to previous calibration results New values should be no more than 10 different than previous calibrations if no major analyzer changes 57 of 79 July 15 2005 2 216 2 3 have been made If variation is gt 10 calibration must be redone to verify values e The new slope for each analyzer derived from combined CH4 KHP and sucrose data is placed into the Carbon par file for each analyzer this file contains analyzer parameters which are read into the Carbon program when it is first started The date and version number in the Carbon par file is also updated Calibration data and plots are retained in file folders in the file cabinet with raw analysis data DRI STANDARD OPERATING PROCEDURE Page 58 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 4 1 3 5 Calculations The conversion of integrated peak counts to ug of carbon for each peak in t
86. tion is complete and those samples should be rerun as soon as they can be conveniently fit into the current day s analyses Evidence of persistent analyzer problems must be resolved either by physically examining the analyzer or reviewing the problems with the analyzer operator 6 5 1 1 Validation of Final Data File The following steps are followed to create an Excel file containing carbon data and to perform Level I validation on it DRI STANDARD OPERATING PROCEDURE Title DRI Model 2001 Thermal Optical Carbon Analysis TOR TOT of Aerosol Filter Samples Method IMPROVE A Page 73 of 79 Date July 15 2005 Number 2 216 2 Revision 3 e Each analyzer will have an Access database containing all of the raw carbon data e A query is used to generate the project data in ug cm or ug filter and a validation report is then generated from this query e The output of the Access query is saved as an Excel file for data validation and processing The MS Excel file naming convention calls for a name in the following format xxOETnnt xls where XX OET nn is the two character project identifier stands for organic elemental carbon is the two digit batch number generally used to distinguish between different projects for the same client or between sampling quarters for an extended project stands for the sample type based on sampler technology Kx ZOO TAO UO gt Agricultural burn emissions dilution sampler Com
87. ucing CO to methane CH4 by passing the flow through a methanator hydrogen enriched nickel catalyst and 4 quantifying CH equivalents with a flame ionization detector FID The principal function of the optical laser reflectance and transmittance component of the analyzer is to correct for pyrolysis charring of OC compounds into EC Without this correction the OC fraction of the sample might be underestimated and the EC fraction might include some pyrolyzed OC The correction for pyrolysis is made by continuously monitoring the filter reflectance and or transmittance via a helium neon laser and a photodetector throughout an analysis cycle The reflectance and transmittance largely dominated by the presence of light absorbing EC decrease as pyrolysis takes place and increase as light absorbing carbon is liberated during the latter part of the analysis By monitoring the reflectance and transmittance DRI STANDARD OPERATING PROCEDURE Page 8 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 the portion of the EC peak corresponding to pyrolyzed OC can be accurately assigned to the OC fraction The correction for the charring conversion of OC to EC is essential for a less biased measurement of carbon fractions Johnson et al 1981 The Thermal Optical Reflectance TOR and Thermal Optical Transmittance TOT charring correctio
88. yosa 201 zuro 18 vam 1220 11534 11020 0 03 24232 2037u 243331 24455 EN out 8 8 em GH 125200 25870 25338125901 sa am nh 2 31 1158 102 0 00 24 47 puiz3 2ujqu lauzsz 6 2 pul HH 1258435779 l2580501258838 ETT A burg 7957127773 sig em ANS 2607 20 73 24210 Haal 106 2 51 115721 1003 10 12 20J0 2uzo 24228 12198 ECTETUR 20 28 2012x 20242 2uz2x d 7 CTT T 0 alz PmlC HH 125535125501 292 22522 aa leer aaa leas la la lao CON dlrs pul AMO 2540 20555 zu 580 ls aap ip am TE 2 3 11775 1001 002 l25dul25812 125388 23221 Ail 12541 l25342 5912 2597X Leaun la l aaa el liz enad 20 85 24j42 zunpz ug ETF ERA Sas 270712782 125407 Lento 3 Six em AHA 12 23 2uI8F 2e 42 232 caes lees el tare err eer beis m ho slg Pm SMS 2400 pax 25925 25905 Y glo am Uh 2 30 COA 0 06 J24055Puo35 24117 lore 1277721 25440125408 25984 Test on a blank filter is placed on Analyze position DRI STANDARD OPERATING PROCEDURE Page 34 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 3 1 4 OC EC Analysis Based on the analysis list for the day retrieve the samples to be analyzed from the sample freezer and place in a Styrofoam cooler with blue ice or in the analysis room refrigerator Routine analysis proc
89. ysis If the analyzer was previously used it will cool to 100 C then pull the boat back to the calibration position continue cooling to 50 C and pull the boat back to the load position for the next analysis 3 1 4 2 Loading the Filter Punch e Use tweezers to remove punch from punch tool and place in analyzer boat e Click OK on the analysis Setup screen The boat will load to the calibration position and the computer will ask Would you like to proceed or would you like to delay DRI STANDARD OPERATING PROCEDURE Page 36 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 analysis Check to make sure a 90 second delay is in the delay box Click OK and analysis will begin e Wipe the tweezers flat glass plate and punching tool with a clean KIMTECH wiper 3 1 4 3 Post Analysis At the end of each analysis data is saved to the database split times are calculated carbon peaks are integrated and tabular and graphical printouts are produced The sample boat will retract to the calibration position when it is sufficiently cooled by the fan to gt 100 C and will continue to cool until it reaches less than 50 C e Examine the tabular printout Figure 3 4 to confirm that the calibration peak counts are within specifications typically 25 000 counts see Section 4 1 e Examine the
90. zer storage may cause crystallization e Lock the carbon analysis room 3 2 Abbreviated Procedures 3 2 1 Analyzer Start Up Check pressures and delivery pressures in all gas cylinders Check that all FIDs are lit by holding a pair of tweezers over the FID exhaust stack and watching for condensation Relight if necessary Check all gas flows at the analyzer readjust if necessary Turn on the computer monitor Confirm that the date and time on the computer are correct Execute the DRICarb exe either from the shortcut on the computer desktop or from C CarbonNet Confirm that the printers have enough paper for the day and that the toner cartridge is producing legible text and graphics DRI STANDARD OPERATING PROCEDURE Page 44 of 79 Date July 15 2005 Title DRI Model 2001 Thermal Optical Carbon Analysis Number 2 216 2 TOR TOT of Aerosol Filter Samples Method IMPROVE A Revision 3 3 2 2 Wipe the sample tweezers flat glass plate and punching tool with a clean KIMTECH wiper Begin the daily entry in the Carbon Analyzer Logbook Bake the oven for 10 minutes or bake during leak test Perform a leak test involving isolating the oven and operating the Carle valve Execute the cmdAutoCalibCheck command to verify the analyzer s performance Retrieve the samples to be analyzed from the sample freezer Complete Daily Analyzer Check Sheet posted on each analyzer Leak Checks and Oven Baking Use manual option to

DRI Model 2001 Thermal/Optical Carbon Analysis

Contents

Download Pdf Manuals

Related Search

Related Contents