(QA Handbook Vol I, Section 11.0, Revision No: 0 Date: 05/13 Page | of 7 11.0 Instrument Equipment Testing, Inspection and Maintenance Implementing an ambient air monitoring network, with the various types of equipment needed, is no easy task. Through appropriate testing, inspection and maintenance programs, monitoring organizations can be assured that equipment is capable of operating at acceptable performance levels. Every piece of equipment has an expected life span, and its use should be discontinued if its performance quality ceases to meet appropriate standards. For amortization purposes, EPA estimates a 7 year lifespan for most monitoring instruments and a somewhat longer lifespan for more permanent types of equipment (instrument racks, monitoring shelters etc.). This schedule means that funds for replacing capital equipment are provided in resource allocations and monitoring organizations should make the best use of equipment replacement resources. Monitoring organizations may be able to prolong the life of equipment but in doing so they may run the risk of additional downtime, more upkeep and a greater chance of data invalidation, while losing out on newer technologies, better sensitivity/stability and the opportunities for better information management technologie Due to the many types of equipment that can be used in an ambient air monitoring program, this section provides general guidance on testing, inspection and maintenance procedures for broad categories of equipment only. In most cases, equipment manufacturers include inspection and maintenance information in the operating manuals. The role of monitoring organizations, in developing a quality system, is to address the scheduling and documentation of routine testing, inspection, and maintenance. Detailed maintenance documents should be available for each monitoring site. Elements incorporated into testing, inspection and maintenance documents include: equipment lists - by organization and station; spare equipment/parts lists - by equipment, including suppliers; inspection/maintenance frequency - by equipment; testing frequency and source of the test concentrations or equipment; ‘+ equipment replacement schedules; ‘+ sources of repair - by equipment; ‘+ service agreements that are in place; and ‘+ monthly check sheets and entry forms for documenting testing, inspections and maintenance performed. 11.1 Instrumentation 11.1.1 Analyzers and Samplers Aside from the specific exceptions described in Appendix C of Part 58', monitoring methods used for SLAMS monitoring must be a reference or equivalent method, designated as such by 40 CFR Part 53° and will be labeled as such’, Reference or equivalent methods also must be used at NCore monitoring sites, intended for comparison with any NAAQS. Among reference and equivalent methods, a variety of analyzer designs and features are available. For certain pollutants, analyzers employing different measurement principles ate available, Some analyzer models only meet the minimum performance specifications (see Table 7-6), while others provide a higher level of performance, Section 7 provides information on what aspects to consider when selecting a particular monitoring instrument/analyzer. " 40 CER Part $8, Appendix C htip:/iwww.ceft,gov/egi-bintext-idx?ipl-/ecfrbrowse/Title40/40tab_ 02.tpl 240 CER Part 33, °40 CFR Part Part $3.9(4)(QA Handbook Vol I, Section 11.0, Revision No: 0 Date: 05/13 Page 2 of 7 Upon receiving the new analyzer, the user should carefully read the instructions or operating manual provided by the manufacturer. Information or instructions concerning the following should be found in the manufacturer's manual: ‘unpacking and verifying that all component parts were delivered; checking for damage during shipment; checking for loose fittings and electrical connections; assembling the analyzer; installing the analyzer, calibrating the analyzer; operating the analyzer, electrical and plumbing diagrams; preventive mai troubleshooting: and a list of expendable parts. chedule and procedures, NOTE: Many vendors have specific time periods when the initial checks for damage in transit need to be made so it may be important to perform an initial check/verification of the equipment as soon as possible. The monitor should be up according to the instructions in the manufacturer’s manual Initial Set-up and Acceptance Testing It may be important to do this initial set-up and testing at the main office or laboratory facility (see Section 11.1.3) before taking the equipment to the site, Following analyzer set-up, and allowance for the instrument to reach required operating conditions, an initial verification of performance characteristics such as power flow, noise, response time and a multi-point verification should be performed to determine if the analyzer is operating properly. These guidelines assume that the instrument was previously calibrated. If the instrument was disassembled after calibration, or no calibration of the instrument had previously been performed, the monitor must have a multi-point verification/calibration to ensure itis within acceptable calibration requirements prior to use. Zero/span drift and precision should be checked during the initial calibration or measured using abbreviated forms of the test procedures provided in 40 CER Part 53. Acceptance of the analyzer should be based on results from these performance tests. If the analyzer does not perform to stated specifications, document the testing procedures and data and contact the manufacturer for corrective action, Once accepted, reference and equivalent analyzers are guaranteed by the manufacturer to operate within the required performance specifications for one year’, unless major repairs are performed or parts are replaced. In such instances, the analyzers must be recalibrated before use. 11.1.2 Support Instrumentation Experiences of monitoring organization staff; preventive maintenance requirements, ease of maintenance and general reliability play crucial roles in the selection of support equipment. ‘The following examples depict general categories of support equipment and typical features to look for when selecting thi equipment. This list is meant to guide agencies in the selection of equipment and does not represent required specifications. * 40 CER Part 53.9 (¢)(QA Handbook Vol I, Section 11.0, Revision No: 0 Date: 05/13 Page 3 of 7 ‘© Calibration Standards: Calibration standards fall into several categories: = mass flow controlled (MFC) devices; = standards that meet the 2012 Traceability Protocol for Gaseous Calibration Standards* = permeation devices; = voltage standards for equipment testing; = photometers; = flow measurement devices; = barometric pressure measurement devices; and = temperature measurement devices. It is recommended that the devices be 110 VAC, be compatible with data acquisition systems for automated calibrations, and have digital compatibility or true transistor-transistor logic (TTL). The most common standards are MFC devices and permeation devices. Both use dilution air to obtain the needed output pollutant concentration. ‘+ Data Acquisition Systems (DAS): DAS should have at least 32-bit logic for improved performance (DAS with at least 16-bit logic can still be used); have modem and internet capabilities; allow remote access and control; allow for digital input; and be able to initiate automated calibrations and polling. It is also recommended that DAS have software compatible with AQS and AQI reporting and editing. Both data loggers and analog chart recorders may be used for recording data; however, the storage, communicability, and flexibility of DAS coupled with data loggers makes the DAS systems the preferred option. More information on DAS is found in Section 14. ‘+ Instrument Racks: Instrument racks should be constructed of steel and be able to accept sliding trays or rails. Open racks help to keep instrument temperatures down and allow air to circulate freely. ‘+ Instrument Benches: Instrument benches should be of sufficient space to allow adequate room, for multiple instruments with room to work and be capable of supporting a fair amount of weight (© 100 tbs), Slate or other hard, water-proof materials (e.g, steel) are recommended. © Zero Air Systems and Standards: Zero air systems should be able to deliver 10 liters/min of air that is free of ozone, NO, NO;, and SO to 0.001 ppm and CO and non-methane hydrocarbons to 0.1 ppm or below the instruments method detection limit (whichever is lower). With NCore monitoring and the use of trace gas monitors, there may be a need to audit and calibrate at lower level. Therefore monitoring organization may need to acquire zero air systems capable to delivering zero air at 20 to 30 liters/min. There are many commercially available systems; however, simple designs can be obtained by using a series of canisters. In addition, the 2012 Traceability Protocol for Gaseous Calibration Standards includes a discussion of zero gas standards which are commercially available. Although not required for use under protocol gasses, the standards can be used as a check on zero air systems. S EPA Traceability Protocol for Assay and Certification of Gaseous Calibration Standards (EPA-600/R-23/531) http://www. epa.govinrmrlapped/mmadidb-traceability-protocol him](QA Handbook Vol I, Section 11.0, Revision No: 0 Date: 05/13 Page 4 of 7 11.1.3 Laboratory Support While it is not required, monitoring organizations should employ full laboratory facilities. These facilities should be equipped to test, repair, troubleshoot, and calibrate all analyzers and support equipment necessary to operate the ambient air monitoring network. In cases where individual laboratories are not feasible, a monitoring organization may be able to find a central laboratory (PQAO) where these activities can be performed. Itis recommended that the laboratory be designed to accommodate the air quality laboratory/shop and PM) and PM); filter rooms’, as well as enforcement instrumentation support activities. The air quality portion consists of several benches flanked by instrument racks. One bench and rack are dedicated to ozone traceability. The other instrument racks are designated for calibration and repair. A room should be set aside to house spare parts and extra analyzers. A manifold/sample cane should be mounted behind the bench. If possible, a sample cane that passes through the roof to allow analyzers that are being tested to sample outside air should be mounted to the bench. This configuration also allows any excess calibration gas to be exhausted to the atmosphere. It is recommended that the pump room be external to the building to eliminate noise. Each bench area should have an instrument rack attached to the bench, The instrument rack should be equipped with sliding trays or rails that allow easy installation of instruments. If instrumentation needs to be repaired and then calibrated, these activities can be performed on the bench top or within the rack. Analyzers then can be allowed to warm up and be calibrated by a calibration unit. Instruments that are to be tested are connected to the sample manifold and allowed to sample air in the same manner as if the analyzer were being operated within a monitoring station. The analyzer is connected to an acquisition system (e.g., DAS, data logger, chart recorder, ete.) and allowed to operate. Any intermittent problems, that occur can be observed on the data logger/chart recorder. The analyzer can be allowed to operate over several days to see if anomalies or problems reoccur; if they do, there is a record of them. Ifthe instrument rack has a DAS and calibrator, nightly auto QC checks can be performed to see how the analyzer reacts to known gas concentrations. In addition, the ozone recertification bench and rack should be attached to a work bench. The rack should house the local ozone level 2 standard’ and the ozone transfer standards (level 3 and greater) that are being checked for recertification. Zero air is plumbed into, this rack for the calibration and testing of ozone analyzers and transfer standards. During FRM/FEM testing EPA tries to ensure that monitoring equipment manufacturers test instruments a varying environmental extremes, However within the period of testing some extremes that exist in some monitoring areas may not be achieved. Monitoring organizations that have large regions with varying extremes of temperature, humidity and pressure may want to invest in an environmental chamber that can be used to test monitoring instruments against the manufactures advertized performance standards. 11.2. Preventive Maintenance Every monitoring organization should develop a preventive maintenance program. Preventive maintenance is what its name implies; maintaining the equipment within a network to prevent downtime and costly repairs and data loss. Preventive maintenance is an ongoing element of quality control and is, * Guidance on filter room requirements can be found in methods 2.10 and 2.11 for PMio and 2.12 for PMas http:/www.epa,govitin/amtifiles/ambient/gage/OzoneTransferStandardGuidance.paf(QA Handbook Vol I, Section 11.0, Revision No: 0 Date: 05/13 Page $ of 7 typically enveloped into the daily routine. In addition to the daily routine, scheduled activities must be performed monthly, quarterly, semi-annually and annually, Often the standard operating procedures and/or operation manuals will provide preventative maintenance activities for the particular instrument/method. It is suggested that these sections could be assembled into a preventative maintenance document that could be kept at each site and accessed electronically so that maintenance can be implemented and documented in a consistent manner. Preventive maintenance is the responsibility of the station operators and the supervisory staff. Itis important that the supervisor review the preventive maintenance work and continually check the schedule. The supervisor is responsible for making sure that preventive maintenance is being accomplished in a timely manner. Preventive maintenance is not a static process; procedures must be updated for many reasons, including, but not limited to, new models or types of instruments and new or updated methods. The preventive maintenance schedule is changed whenever an activity is completed or performed at an alternate time. For instance, if'a multipoint calibration is performed in February instead of on the scheduled date in March, then the subsequent six-month calibration date moves from September to August. On a regular basis, the supervisor should review the preventive maintenance schedule with the station operators. Following all repairs, the instruments must be verified (multi-point) or calibrated. Lists can facilitate the organization and tracking of tasks and improve the efficiency of preventive maintenance operations. A checklist of regular maintenance activities (e.g., periodic zero-span checks, daily routine checks, data dump/collection, calibrations, ete.) is recommended. A spare parts list, including relevant catalog numbers, is also recommended, as it facilitates the ordering of replacement parts. Such a list should be readily accessible and should include the types and quantities of spare parts already on-hand. 11.2.1 Station Maintenance Station maintenance is an element of preventive maintenance that does not occur on a routine basis; rather, these tasks usually occur on an “as needed” basis. Station maintenance items are checked monthly or whenever an agency knows that the maintenance needs to be performed. Examples of station maintenance items include: ‘+ floor cleaning; ‘+ shelter inspection; ‘+ security inspection fencing, locks, surveillance cameras, lightin; ‘+ visual inspection of probes, and met gear + air-conditioner repair; + AC filter replacement ‘+ weed abatement and grass cutting; roof repair; general cleaning; inlet and manifold cleaning; manifold exhaust blower lube; desiccant replacement; and ladder, safety rails, safety inspection, if applicable. ‘Simple documentation of these activities, whether in station logs or electroni evidence of continuous attention to data quality. c logs, helps provide(QA Handbook Vol I, Section 11.0, Revision No: 0 Date: 05/13 Page 6 of 7 11.2.2 Routine Operations Routine operations are the checks that occur at specified periods of time during a monitoring station visit. These duties must be performed and documented in order to operate a monitoring network at optimal levels. Examples of typical routine operations are detailed in Table 11-1. ‘Table 11-1 Routine Operation Checks, hem Each V Weekiyihfonthh imam ‘Observation of unusual X conlionsievents Review Data x Mark chants, where applicable x, ‘Chesk/Oil Exhaust Blower x. ‘Cheok Exterior x ‘Check Change Deiceant x ‘Manifold Leak Test x Tspect tubing x Replace Tubing naval Tnspeet manifold and cane x ‘Clean manifold and cane very 6 months or as needed ‘Check HVAC systems x ‘Check electrical connections x Field site supply invent x Residence time caleulation Tmanifold and lets altered Trtubing is used externally a an let device W may need tobe replaced every 6 months or more frequently depending upon site specific issues. In addition to these items, the exterior of the building, sample cane, meteorological instruments and tower, entry door, electrical cables, and any other items deemed necessary to check, should be inspected for wear, corrosion, and weathering. Costly repairs can be avoided in this manner. 11.2.3 Instrument Logs and Site Logs Each instrument and piece of support equipment (with the exception of the instrument racks and benches) should have an Instrumentation Repair Log (either paper or electronic). The log should contain the repair and calibration history of that particular instrument, Whenever multipoint verification/calibration, instrument maintenance, repair, or relocation occurs, detailed notes are written in the instrumentation log. The log contains the most recent multipoint verification/calibration report, a preventive maintenance sheet, and the acceptance testing information or reference to the location of this information, If an instrument is malfunctioning and a decision is made to relocate that instrument, the log travels with that device. The log can be reviewed by staff for possible clues to the reasons behind the instrument malfunction, In addition, if the instrument is shipped to the manufacturer for repairs, itis recommended that a copy of the log be sent with the instrument. This information helps non-agency repair personne! with troubleshooting instrument problems. Improper recording of instrument maintenance can complicate future repair and maintenance procedures. The instrument log should be detailed enough to determine easily and definitively which instrument was at which site(s) over any given time period. If'a problem is found with a specific instrument, the monitoring staff should be able to track the problem to the date it initially surfaced and invalidate data even if the instrument was used at multiple sites A site log should be kept documenting maintenance of a specific monitoring site and the auxiliary monitoring equipment located there. Information that could be recorded includes the activities listed in the Station Maintenance and Routine Operations sections (Sections 11.2.1 and 11.2.2).(QA Handbook Vol I, Section 11.0, Revision No: 0 Date: 05/13 Page 7 of 7 The site log is a chronology of the events that occur at the monitoring station, The log is an important part of station maintenance because it contains the narrative of past problems and solutions to those problems. Site log notes should be written in the form of a narrative, rather than shorthand notes oF bulleted lists. Examples of items that should be recorded in the site log are: ‘+ the date, time, and initials of the person(s) who have arrived at the site; ‘+ brief description of the weather (e.g., clear, breezy, sunny, raining); ‘+ brief description of exterior of the site, Any changes that might affect the data should be recorded for instance, if someone is parking a truck or tractor near the site, this note may explain high NO, values; ‘+ any unusual noises, vibrations, or anything out of the ordinary; ‘+ records of any station maintenance or routine operations performed; ‘© description of the work accomplished at the site (e.g, calibrated instruments, repaired analyzer); and ‘+ detailed information about the instruments that may be needed for repairs or troubleshooting, It is not required that the instrument and site logs be completely independent of each other. However, there is an advantage to having separate instrument logs. If instruments go in for repair, they may eventually be sent to another site. Having a separate instrument log allows the log to “travel” with the instrument, Keeping electronic instrument and station maintenance logs at stations and at centralized facilities (see LIMS discussion Section 8) also has record keeping advantages, but there needs to be a way that these records can be considered official and not be tampered with or falsified. Newer electronic signature technologies are helping ensure that electronic records can be considered official. It is important, however, that all of the required information for each instrument and site be properly recorded using a method that is comprehensive and easily understood. Many monitoring organizations have developed standard station maintenance forms that contain all the items to be checked and the frequency of those checks. It then becomes a very simple procedure to use this form to check off and initial the activities that were performed,(QA Handbook Vol I, Section 12.0 Revision No: 0 Date: 05/13 Page fof 10 12.0 Calibrations Calibration is defined as: the comparison of a measurement standard, instrument, or item with a standard or instrument of higher accuracy to detect and quantify inaccuracies and to report or eliminate those inaccuracies by adjustment Prior to the implementation of any ambient air monitoring activities, the sampling and analysis equipment must be checked to assure itis within calibration tolerances, and if it fails these tolerances, must be appropriately calibrated. This function is most routinely carried out at the field monitoring location Calibration of an analyzer or instrument establishes the quantitative relationship between the actual value ofa standard, be it a pollutant concentration, a temperature, or a mass value (in ppm, °C or ng, ete.) and the analyzer's response (chart recorder reading, output volts, digital output, etc.). This relationship is used to convert subsequent analyzer response values to corresponding concentrations. Once an instrument's calibration relationship is established, itis checked at reasonable frequencies to verify that it remains in calibration, Verification Versus Calibration Since the term calibration is associated with an adjustment in either the instrument or software, these adjustments should be minimized as much as possible. Sometimes performing frequent adjustments to provide the “most accurate data possible” can be self-defeating and be the cause of additional measurement uncertainty. For example, adjusting an instrument based upon a standard that might be degrading or contaminated may actually cause data to be farther from the true concentration, Therefore, quality control procedures that include measurements (e.g., I-point QC, flow rate verifications, etc.) and multi-point verifications are considered “checks without correction” and are used to ensure the instruments are within the calibration tolerances. Usually these tolerances have been developed so that as Jong as the instrument is within these tolerances, adjustments do not need to be made. However, verifications should be implemented at reasonable frequencies to avoid invalidating significant amounts of data, NOTE: When the term “calibration” is used in the remainder of this section, it is assumed that a multi-point verification is initially performed and the operator has concluded that calibration (adjustment) is necessary. EPA does not recommend post-processing of data to “correct” for data failing one point or multi-point verifications. For example, ifafter failure of a one-point QC check a subsequent verification and calibration found that data was biased high by 15% the previons rontine data up until the last aeceptable I-point QC check is not adjnsted down by 15% and reported. Based upon validation criteria, the data is either reported as initially measured or invalidated. " American National Standard Quality Systems for Environmental Data and Technology Programs ANSI/ASQ E4 haps /Avww.asq.org(QA Handbook Vol I, Section 12.0 Revision No: 0 Date: 05/13 Page 2 of 10 Each analyzer should be calibrated as directed by the analyzer’s operation or instruction manual and in accordance with the general guidance provided here. For reference methods for CO, NO2, SO2 and Os, detuiled calibration procedures may also be found in the appropriate reference method Appendix in 40 CER Part 50° and the method guidance and technical assistance documents listed in the fact sheets in Appendix A. Calibrations should be carried out at the field monitoring site by allowing the analyzer to sample test atmospheres containing known pollutant concentrations. In the case of PM and Pb monitors where concentration standards are not available and impractical, calibrations take place on flow, temperature and pressure devices as best as possible. At times this may need to be accomplished in laboratory settings rather than the field. The analyzer to be calibrated should be in operation for at least several hours (preferably overnight) prior to the calibration so that itis fully warmed up and its operation has stabilized. During the calibration, the analyzer should be operating in its normal sampling mode, and it should sample the test atmosphere through all filters, scrubbers, conditioners, and other components used during normal ambient sampling and through as much of the ambient air inlet system as is practicable All operational adjustments to the analyzer should be completed prior to the calibration (see section 12.7). Some analyzers can be operated on more than one range. For sites requiring the use of FRM or FEMs (NAAQS sites), the appropriate ranges are identified in the Designated Reference and Equivalent ‘Method List found on AMTIC*. Analyzers that will be used on more than one range or that have auto- ranging capability should be calibrated separately on each applicable range. Calibration documentation should be maintained with each analyzer and also in a central backup file. Documentation should be readily available for review and should include calibration data, calibration equation(s) (and curve, if prepared), analyzer identification, calibration date, analyzer location, calibration standards used and their traceability, identification of calibration equipment used, and the person conducting the calibration. Full Se: ration Seal vs. Many older documents and some of the CFR reference methods refer to calibration at “full scale”. The interpretation of this meant that monitoring organizations would calibrate to full scale of one of the FRM/FEM approved operating range(s) of the instrument. For example, ozone instruments are approved at 0-500 ppb or 0-1000 ppb, Many monitoring organization calibrate the instrument by evenly spacing four upscale points up to around 500 ppb. In this scenario, with most sites reading less than 80 ppb, the majority of the upscale calibration points would be at levels not measured in ambient conditions. EPA suggests monitoring organization calibrate using points that are more applicable to the concentrations found in their networks while still be protective of concentrations exceeding the NAAQS. Using this procedure more points can be used to calibrate the instruments at these lower concentration levels and better inform monitoring organizations of stability. For convenience, EPA will use the term “calibration scale” to refer to the concentration range used for calibrating the monitoring instruments. Section 10 provides more details on this process. * http:/www access. gpo.gov/nara/efricfi-table-search,html ® hup:lvww.epa.gov/ttn/amtic/erteria, him!QA Handbook Volume Il, Seetion 12.0 Revision No: 0 Date: 05/13 Page 3 of 10 12.1 Calibration Standards and Reagents Calibration standards are: © Reagents of high grade * Gaseous standards of known concentrations that are certified as EPA protocol gasses Instruments and or standards of high sensitivity and repeatability. 12.1.1 Reagents In some cases, reagents are prepared prior to sampling. Some of these reagents will be used to calibrate the equipment, while others will become an integral part of the sample itself. In any case, their integrity ‘must be carefully maintained from preparation through analysis. If there are any doubts about the method by which the reagents for a particular test were prepared or about the competence of the laboratory technician preparing them, the credibility of the ambient air samples and the test results will be diminished. It is essential that a careful record be kept listing the dates the reagents were prepared, by whom, and their locations at all times from preparation until actual use. Prior to the test, one individual should be given the responsibility of monitoring the handling and the use of the reagents. Each use of the reagents should be recorded in a field or lab notebook. Chemical reagents, solvents, and gases are available in various grades. Reagents can be categorized into the following six grades" L. Primary standard - Each lot is analyzed, and the percentage of purity is certified. 2. Analyzed reagents- Can fall into 2 classes: (a) each lot is analyzed and the percentages of impurities are reported; and (b) conformity with specified tolerances is claimed, or the maximum percentages of impurities are listed. 3. USP and NF Grade - These are chemical reference standards where identity and strength analysis are ensured 4, “Pure,” “e.p..” “chemically pure,” “highest purity” - These are qualitative statements for chemicals without numerical meaning. 5. “Pure,” “purified,” “praetieal grades” - These are usually intended as starting substances for laboratory syntheses. 6. Technical or commercial grades - These are chemicals of widely varying purity The reference and equivalent methods define the grades and purities needed for the reagents and gases required in the Ambient Air Quality Monitoring Program. All reagent containers should be properly labeled either with the original label or, at a minimum, the reagent, date prepared, expiration date, strength, preparer, and storage conditions. Leftover reagents used during preparation or analysis should never be returned to bottles. * Quality Assurance Principles for Analytical Laboratories, 3rd Edition. By Frederick M. Garfield, Eugene Klesta, and lorry Hirsch AQAC Intematianal (1000) hitp-iwww aoae ong!QA Handbook Volume Il, Seetion 12.0 Revision No: 0 Date: 05/13 Page 4 of 10 12.1.2 Gaseous Standards In general, ambient monitoring instruments should be calibrated by allowing the instrument to sample and analyze test atmospheres of known concentrations of the appropriate pollutant in air. The following is an excerpt from 50 CFR Part 58, Appendix A Section 2.6.1 “Gaseous pollutant concentration standards (permeation devices or eylinders of compressed gas) used to obtain test concentrations for carbon monoxide (CO), sulfur dioxide (SO2), nitrogen oxide (NO), and nitrogen dioxide (NO2) must be traceable to either a National Institute of Standards and Technology (NIST) Traceable Reference Material (NTRM) or a NIST-certified Gas Manufacturer's Internal Standard (GMIS), certified in accordance with one of the procedures given in reference 4 of this appendix. Vendors advertising certification with the procedures provided in reference 4 of this appendix and distributing gasses as “EPA Protocol Gas” must participate in the EPA Protocol Gas Verification Program or not use “EPA””in any form of advertising.” "Traceable" is defined in 40 CFR Parts 50 and 58 as meaning that a local standard has been compared and certified, either directly or via not more than one intermediate standard, to a primary standard such as a National Institute of Standards and Technology Standard Reference Material (NIST SRM) or a USEPA/NIST-approved Certified Reference Material (CRM)". Normally, the working standard should be certified directly to the SRM or CRM, with an intermediate standard used only when necessary. Direct use of a CRM as a working standard is acceptable, but direct use of an NIST SRM as a working standard is discouraged because of the limited supply and expense of SRM's. At a minimum, the certification procedure for a working standard should: ‘establish the concentration of the working standard relative to the primary standard; ‘© certify that the primary standard (and hence the working standard) is traceable to a NIST primary standard; include a test of the stability of the working standard over several days; and ‘specify a recertification interval for the working standard, Certification of the working standard may he established by either the supplier or the user of the standard. As describe in CFR, gas supplier advertising “EPA Protocol Gas” will be required to participate in the EPA Protocol Gas Verification Program. Information on this program, including the «gas supplier participating in the program, can be found on AMTIC’. EPA has developed procedures for the establishment of protocol gasses in the document: EPA Traceability Protocol for Assay and Certification of Gaseous Calibration Standards*. Table 2-3 in the Traceability Document provides the ‘maximum certification periods for verification and calibration standards used in the ambient air program. Since these periods sometimes change the table is not presented here. ‘Test concentrations of ozone must be traceable to a primary standard (see discussion of primary standards below) UV photometer as described in 40 CFR Part 50, Appendix D and the guidance document: Transfer Standards for the Calibration of Ambient Air Monitoring Analyzers for Ozone’. For ambient air monitoring activities zero concentrations can be acquired through zero air generation devices or purchased as standards. Although zero concentrations are not required to be traceable to a * http:/www.epa.govittwamti ° hutp://www.epa,gow/nrmeliappedmmddb-traceability-protocol, html 7 huyp:/www.epa,gov/ttn/amtie/files/ambient gage/OzoneTransferStandardGuidance,pdf(QA Handbook Volume Il, Section 12.0 Revision No: 0 Date: 05/13 Page 5 of 10 primary standard, care should be exercised to ensure that zero device or standards used are adequately free of all substances likely to cause a detectable response from the analyzer and at a minimum, below the lower detectable limit of the criteria pollutants being measured, Periodically, several different and independent sources of zero should be compared. The one that yields the lowest response can usually (but not always) be assumed to be the “best zero device/standard.” If several independent zero device/standards produce exactly the same response, iti likely that all the standards are adequate. Certification periods decrease for concentrations below the applicable concentration ranges provide in Table 12-1. For example the certification period for SO. standards between 13-40 ppm is 6 months, Also, tank size may affect stability in low level standards. Some gas manufacturers claim that standards supplied in smaller tanks are stable for longer periods of time then the same concentration in larger tanks. Although this claim has not been verified, if true it may be helpful in making purchasing decisions. Primary Reference Standards A primary reference standard can be a defined measurement standard designated for the calibration of other measurement standards for quantities of a given kind in a given organization’. NIST’s standard reference material (SRM’s) are examples of primary reference standards. NIST also describes a Primary Reference Standard as a standard that is designated or widely acknowledged as having the highest metrological qualities and whose value is accepted without reference to other standards of the same quality. For example, the NIST-FI Atomic Clock” is recognized as a primary standard for time and frequency. A true primary standard like NIST-FI establishes maximum levels for the frequency shifts caused by environmental factors. By summing or combining the effects of these frequency shifts, it is possible to estimate the uncertainty of a primary standard without comparing it to other standards. NIST maintains a catalog of SRMs that can be accessed through the Internet”. Primary reference standards are usually quite expensive and are often used to calibrate, develop, or assay working or secondary standards. In order to establish and maintain NIST traceability the policies posted at the NIST Website" should be observed. It is important that primary reference standards are maintained, stored, and handled in a manner that maintains their integrity. These standards should be kept under secure conditions and records should be ‘maintained that document chain of custody information. 12.1.3 Instruments The accuracy of various measurement devices in sampling and continuous instruments is very important to data quality. For example, in order to produce the correct flow rate to establish an accurate PM2s cut point, the temperature and barometric pressure sensors, as well as the flow rate device, must be producing, accurate measurements. Table 12-1 provides some of the more prevalent instruments that need to be calibrated at a minimum annually or when shown through various verification checks to be out of acceptable tolerances. In addition, the audit standards used to implement the checks and calibrations should be certified annually in order to establish their accuracy and traceability to higher standards * definition of reference measurement standard from International vocabulary of metrology ~ Basie and general ‘concepts and associated terms (VIM) hitp:/www.bipm.org/enpublications/guides/vim,html ° hup://www.nist.govipml/div6R8grp50/primary-frequency-standards.cfin "tp: //www.nist-gov ° opus. nist.gov/traceabilityQA Handbook Volume Il, Section 12.0 ‘Table 12-1 Instruments and Devices Requiring Calibration and Certifications. Revision No: 0 Date: 05/13 Page 6 of 10 Criteria Acceptable Range 4crR Reference Barometric Pressure Temperature Flow Rate Design Flow Rate Adjustment Clockitimer Verification Mirobalance Calibration Lab Temperature Lab Humidity Mirobalance Calibration Standard Reference Photomerer (SRP) SRP recertification to local Verification/Calibraion of devices sampler 210mm Hg, £2°C of standard of transfer standard Yo of design flow rate | minima Readability 1p Repeatability 1 yg 2c Readability 1 ng Repeatability 1 yg 44% or +4 ppl (whichever greater) RSD of sx slopes < 3.7% Sid. Dev. of 6 intercepts 1S iyzer/laboratory against an authoritative standard Part 50, App.L, See 9.3 Part 50, App.L, See 9.3 Part 50, App.L, See 9.2 Part 50, App.L, Sec 9.2.6 Part 50, App.L, See 74 Pat 50, App.L, See 8.1 Verification/Calibraion of devices in shelter or lab against an authoritative standard not deseribed not deserbed Part 50, App.L, See 8.1 Verificaion/ealibraion standards requiring certification annually not described not described Primary standard New slope = + 0.05% of previous Flow rate 2% of NIST —Traceable Standard Part 50, App L Sec 9.2 Pressure “4 1mm Hg resolution, 1 mm Hig not deseribed accuracy Temperature 0.1° C of standard resolution, + 0.5° C 1 not deseribed mm Hg accuracy Gravimetric Standards 0.025 me not deseribed 12.2 Mult point Verifications/Calibrat ns Multi-point calibrations consist of a zero and 4 upscale points, the highest being a concentration above the NAAQS (for SLAMS criteria pollutants) and higher than any routine values one might expect at the site. This is defined as the calibration scale for the instrument and is different then what has been traditionally defined as the “full scale” operating range defined in the FRM/FEM approval documentation. Multi- point calibrations are used to establish or verify the linearity of analyzers upon initial installation, after major repairs and at specified frequencies. Most modern analyzers have a linear or very nearly linear response with concentration. If @ non-linear analyzer is being calibrated, additional calibration points should be included to adequately define the calibration relationship, which should be a smooth curve. Calibration points should be plotted or evaluated statistically as they are obtained so that any deviant points can be investigated or repeated immediately. Most analyzers have zero and span adjustment controls, which should be adjusted based on the zero and highest test concentrations, respectively, to provide the desired scale range within the analyzer's(QA Handbook Volume Il, Section 12.0 Revision No: 0 Date: 05/13 Page 7 of 10 specifications (see section 12.5). For analyzers in routine operation, unadjusted ("as is") analyzer zero and span response readings should be obtained prior to making any zero or span adjustments. NO/NO,/NO, analyzers may not have individual zero and span controls for each channel; the analyzer's operation/instruction manual should be consulted for the proper zero and span adjustment procedure. Zero and span controls often interact with each other, so the adjustments may have to be repeated several times to obtain the desired final adjustments. After the zero and span adjustments have been completed and the analyzer has been allowed to stabilize on the new zero and span settings, all calibration test concentrations should be introduced into the analyzer for the final calibration, The final, post-adjusted analyzer response readings should be obtained from the same device (chart recorder, data acquisition system, etc.) that will be used for subsequent ambient measurements, The analyzer readings are plotted against the respective test concentrations, and the best linear (or nonlinear if appropriate) curve to fit the points is determined. Ideally, least squares regression analysis (with an appropriate transformation of the data for non-linear analyzers) should be used to determine the slope and intercept for the best fit calibration line of the form, y = mx +b, where y represents the analyzer response, x represents the pollutant concentration, m is the slope, and b is the x- axis intercept of the best fit calibration line. When this calibration relationship is subsequently used to compute concentration measurements (x) from analyzer response readings (y), the formula is transposed to the form, x eee For the gaseous pollutants, Sapcemiamanamatnttee | Be rite tales, is considered acceptable if all calibration points fall aT aE] Sh esac aetoumetan Within 2% (or an absolute ine ae wnnmecguecaneneamese — difference) of the calibration scale, best fit eta ‘tim Cc Cin straight line. Which = acceptance criteria (percent or absolute difference) is used depends on the concentration of the calibration points. EPA has developed a Data Assessment Statistical Calculator (DASC)'? tool that automates this process a Calibration information for data evaluation (See oa ~ Fig. 12.1), For manual “ me samplers the flow rate, temperature, pressure devices are checked at different settings. Acceptance criteria for these devices can be found Figur 12.0886 catvaton Tot " up://www.epa,gov/ttn/amticigareport.himlQA Handbook Volume Il, Section 12.0 Revision No: 0 Date: 05/13 Page § of 10 in the MQO Tables in Appendix D. As a quality control check on calibrations, the standard error or correlation coefficient ean be calculated along with the regression calculations. A control chart of the standard error or correlation coefficient could then be maintained to monitor the degree of scatter in the calibration points and, if desired, limits of acceptability can be established, 12.3 Frequency of Calibration and Analyzer Adjustment An analyzer should be calibrated (or recalibrated): upon initial installation, following physical relocation, after any repairs or service that might affect its calibration, following an interruption in operation of more than a few days, ‘upon any indication of analyzer malfunction or change in calibration, and at some routine interval (multi-point verification, see below). ‘When calibration relationships are applied to analyzer responses to determine actual concentrations, it is suggested that the analyzer undergo multi-point verification/calibration periodically to maintain close agreement, ‘The frequency of this routine periodic recalibration is a matter of judgment and is a tradeoff among several considerations, including: the inherent stability of the analyzer under the prevailing conditions of temperature, pressure, line voltage, etc, at the monitoring site; the cost and inconvenience of carrying out the calibrations; the quality of the ambient measurements needed; the number of ambient measurements lost during the calibrations; and the risk of collecting invalid data because of a malfunction or response problem with the analyzer that wouldn't be discovered until a calibration is carried out When a new monitoring instrument is first installed, zero/span and one point QC checks should be very frequent, perhaps daily or 3 times per week, because little or no information is available on the drift performance of the analyzer. With the advancement in data acquisition system technology, many monitoring organizations are running these QC checks daily. However, the QC checks are required to be implemented every two weeks. Information on another unit of the same model analyzer may be useful; however, individual units of the same model may perform quite differently. After enough information on the drift performance of the analyzer has been accumulated, the calibration frequency can be adjusted to provide a suitable compromise among the various considerations mentioned above. To facilitate the process of determining calibration frequency, itis strongly recommended that control charts be used to monitor the zero/span and one-point QC drift performance of each analyzer. Control charts can be constructed in different ways, but the important points are to visually represent and statistically monitor drift, and to be alerted if the drift becomes excessive so that corrective action can be taken, Such control charts make important use of the unadjusted zero and span response readings, 12.4 Adjustments to Analyzers Ideally, all ambient measurements obtained from an analyzer should be calculated on the basis of the most current multipoint calibration or on the basis of both the previous and subsequent calibrations (see Section 12.5), Some acceptable level of drift (ie., deviation from an original or nominal response curve)‘QA Handbook Vol I, Section 12.0, Revision No: 0 Date: 2/13 Page 9 of 10 can be allowed before physical adjustments (i.e., calibration) must be made because the calibration curve used to calculate the ambient measurements is kept in close agreement with the actual analyzer response. The chief limitations are the amount of change in the effective scale range of the analyzer that can be tolerated and possible loss of linearity in the analyzer's response due to excessive deviation from the design range. Cumulative drifts of up to 15 percent of full scale from the original or nominal zero and span values may not be unreasonable, subject to the limitations mentioned above. Due to the advancement in monitoring technologies, ambient air monitors are much more stable and adjustments not as necessary, Earlier versions of this Handbook included sections for zero/span calibrations as well as physical zero/span adjustments. Precise adjustment of the zero and span controls, may not be possible because of: (1) limited resolution of the controls, (2) interaction between the zero and span controls, and (3) possible delayed reaction to adjustment or a substantial stabilization period after adjustments are made. Precise adjustments may not be necessary because calibration of the analyzer following zero and span adjustments will define the precise response characteristic (calibration curve). EPA feels that frequent adjustments of instruments should not be necessary and may in fact lead to more data quality uncertainty. EPA does not recommend span adjustments be made between multi-point calibrations but zero adjustments are appropriate. EPA is no longer including guidance suggesting that the calibration equation be updated after each zero/span check and suggests the ambient readings be calculated from the most recent multipoint ibration curve or from a fixed nominal or “universal” calibration curve (Section 12.5). In this ease, the zero and span checks scrve only to measure or monitor the deviation (drift error) between the actual analyzer response curve and the calibration curve used to calculate the ambient measurements, Automatic Self-Adjusting Analyzers Some air monitoring analyzers are capable of periodically carrying out automatic zero and span ibrations and making their own zero and span self adjustments to predetermined readings. EPA discourages the use of automatic span adjustments but considers automatic zero adjustments reasonable when: 1) the automatic zero standards pass through the sample probe inlet and sample conditioning system, 2) the zero test is performed every day, and 3) both the adjusted and unadjusted zero response readings can be obtained from the data recording device. EPA does not suggest zero adjustments on checks that occur every two weeks. However an adjustment does not mean a post processing correction on zero (adjusting the previous 24 hours routine data based on the difference between the current zero reading and the previous 24-hour reading). In fact, the automated zero does not correct routine data. The zero is “reset” every 24 hours. EPA does not recommend making automatic or manual adjustments (corrections) to the span. It is expected that the difference between the unadjusted and adjusted zero response is negligible and not greater than the zero drift acceptance criteria listed in the validation template (App D). Data invalidation and corrective action should occur if the differences between the 24-hour unadjusted and adjusted zero drift is greater than the validation template acceptance criteria, Data loggers should be programmed to provide flags or warnings of this occurrence.‘QA Handbook Vol I, Section 12.0, Revision No: 0 Date: 2/13 Page 10 of 10 12.5 Data Reduction Using Calibration Information As noted previously, an analyzer's response calibration curve relates the analyzer response to actual concentration units of measure, and the response of most analyzers tends to change (drift) unpredictably with passing time. These two conditions must be addressed in the mechanism that is used to process the raw analyzer readings into final concentration measurements. Two practical methods are described below. They are listed in order of preference, 1) "Universal" Calibration-A fixed, "universal" calibration is established for the analyzer and used to calculate all ambient readings. All verifications and checks are used to measure the deviation of the current analyzer response from the universal calibration, Whenever this deviation exceeds the established zero and span adjustment limits, the analyzer is recalibrated, 2) Major Calibration Update~In this method, the calibration slope and intercept used to calculate ambient measurements are updated only for "major" calibration (i.e., semi-annual or annual multi-point verification/calibrations). All ambient measurements are calculated from the most recent major calibration. Between major calibrations, periodic zero and span verifications are used to measure the difference between the most recent major calibration and the current instrument response. Physical or automated adjustments of the zero may be appropriate however span adjustment to restore a match between the current analyzer response and the most recent major calibration is not suggested. Whenever this deviation exceeds the established zero and span adjustment limits, the analyzer is recalibrated. 12.6 Validation of Ambient Data Based on Calibration Information When zero or span drift validation limits (see Figure 12.1) are exceeded, ambient measurements should be invalidated back to the most recent acceptable zero/span/one-point QC check where such measurements are known to be valid. Also, data following an analyzer malfunction or period of non- operation should be regarded as invalid until the next subsequent calibration unless unadjusted zero and span readings at that calibration can support its validity Documentation All data and calculations involved in these calibration activities should be recorded in the instrument log book described in Section 11