Designing a PSD PM2.5 Monitoring Program: Challenges and Considerations

On January 22, 2013, the D.C. Court of Appeals vacated the PM2.5 Significant Monitoring Concentrations (SMCs) citing the SMCs were inconsistent with the requirements of Section 165(e)(2) of the Clean Air Act. As a result, all Prevention of Significant Deterioration (PSD) permit applications must now include ambient air monitoring data for the preceding 12-month period for the permit application to be considered complete. If existing data are not determined to be representative by the permitting authority, the applicant must conduct its own monitoring program. A thorough understanding of quality assurance requirements under PSD and a familiarity with proper sampling equipment are key components to designing a successful monitoring program to meet PSD permitting requirements.

Monitoring Photo 1

EPA Designated Sampling Methods

Monitoring for a PSD permit application must be conducted using a reference, equivalent, or EPA-approved method. EPA has developed design and performance based criteria defining three levels of method designations for ambient monitoring samplers: Federal Reference Method (FRM), Federal Equivalent Method (FEM), or an approved method accepted by the permitting authority.

FRM monitors are defined as the principal standard for collecting particulate samples. Samplers designated as FRM have met a combination of design and performance based criteria as defined in Title 40, Part 53 of the Code of Federal Regulations (40 CFR Part 53) and 40 CFR Part 50 Appendix L for PM2.5. FRM samplers collect particulates using a filter-based system by pumping ambient air through a 46.2 mm diameter polytetrafluoroethylene filter for 24-hours at a low volume (1 m3/hr). Samples are typically collected on a predefined three or six day schedule (updated annually by EPA) and are processed at a laboratory using gravimetric analysis to determine the 24-hour PM2.5 concentrations. EPA maintains a list on the Ambient Monitoring Technology Information Center (AMTIC) webpage of samplers that have been designated reference and equivalent methods in accordance with 40 CFR Part 53. Currently there are 17 models designated as reference methods for sampling PM2.5.

Federal Equivalent Method samplers meet performance criteria defined in 40 CFR Part 53 and are separated into three classifications (Class I, Class II, and Class III). Class I FEM samplers have minor design modifications from an FRM, but do not meet all the requirements to be designated an FRM. Many Class I FEMs (BGI PQ200-VSCC or PQ200A-VSCC, Thermo Scientific/Rupprecht & Patachnick Partisol Model 2000, Thermo Electron RAAS2.5 (100, 200, or 300), and Thermo Scientific Partisol 2025 Sequential Sampler) have a dual reference and equivalent method designation. Minor modifications to the sampler or quality assurance policies may change the actual designation for these samplers. Like FRMs and Class I samplers, FEM Class II samplers collect by filtration, require filter conditioning and gravimetric analysis, but have substantial design differences from FRMs. An example of a Class II equivalent sampler would be the Thermo Scientific Partisol-D 2000 and Partisol 2025-D dichotomous samplers which measure PM2.5, PM10, and PM10-2.5.

On October 17, 2006, EPA published test criteria (FR 71, 61236) for continuous PM2.5 monitors to qualify as a Class III FEM. The time resolution of Class III FEMs are on the order of minutes to an hour providing PM2.5 data in real-time or near real-time on a daily basis. FEM Class III Beta attenuation monitors such as Met One BAM 1020 and 1022, Teledyne 602 Beta, and Environmental S.A. MP101M, collect samples onto a filter tape each hour and analyze the amount of particulate through beta ray attenuation. Other methods for continuous monitoring include light scattering detection (Grimm EDM 180) and mass gain using a tapered element oscillating microbalance (TEOM) coupled with a Filter Dynamics Measurement System (FDMS) to account for semi-volatile materials. Automated continuous monitors do not require filter collection, conditioning, and gravimetric analysis at a laboratory.

Background Determination

Background values used in cumulative impact analyses are typically obtained from representative 24-hour and annual PM2.5 concentrations. Depending on the type of monitor and sampling schedule selected, background concentrations can be skewed. High concentrations may be missed when daily 24-hour concentrations are not available and statistics may be affected by the number of samples collected.

The annual PM2.5 design value is the three-year average of the annual average PM2.5 and the three-year average of the annual 98th-percentile 24-hour average concentrations. Filter-based systems (FRM and FEM) sample on an EPA defined 3-day or 6-day schedule resulting in approximately 122 or 61 samples in a 12-month period, respectively. The design value for the 24-hour samples would be the third highest concentration for a 3-day sampling schedule and the second highest concentration on a 6-day schedule.

Recently, EPA allowed continuous FEM monitors to be assigned as the primary sampler at a monitoring site. The result of this decision allows 24-hour concentration data for each day to count toward data completeness. Using daily 24-hour concentrations would allow for background determination to be based on the eighth highest concentration in a year.

EPA’s PM2.5 modeling guidance allows for the use of seasonal background concentrations using the maximum value for each season after the design value has been determined. Additional analysis of the continuous monitor may be helpful in determining where particulate originated when paired with meteorological data, it can be used to determine if emissions from an existing nearby source are being measured or “double-counted.”

Quality Control Monitoring

The PSD and SLAMS monitoring quality assurance programs require the reporting of precision, bias, and accuracy. To meet these criteria, a collocated monitor must be installed at the site per 40 CFR Part 58, Appendix A. Each pair of samplers must have a designated primary and audit sampler. Concentrations from the primary sampler will be reported as the PM2.5 concentrations measured at the site while concentrations from the audit sampler will be used in determining precision, accuracy, and bias. Although both PSD and SLAMS monitoring projects allow sampling with a primary monitor that is FRM or FEM designed, under Appendix A, the first collocated monitor in a network must have an FRM designation.

Agreement between measurements of collocated samplers may be relatively poor for low concentrations. According to Appendix A, PM2.5 data quality assessment calculations for precision and bias should be made for sample pairs where both the primary and secondary analyzer measure 24-hour concentrations greater than 3 µg/m3. Appendix A also recommends that approximately 25 of these acceptable sample pairs be collected in a year which can be challenging particularly on a 6-day sampling schedule where only 60 samples are taken annually. Sampling at a higher frequency with a continuous monitor (daily 24-hour concentrations) or on a 3-day sample schedule increases the likelihood of meeting the recommended number of sample pairs.

Comparability of Collocated Continuous FEM and FRM Samplers

Using two different sampling methods can introduce challenges for data analysis. Differences in operation and data collection can lead to concentration discrepancies resulting in precision and bias criteria not being met. Continuous monitors provide near real-time concentrations on an hourly basis while filter-based samples are analyzed gravimetrically in a laboratory days after the sample was taken.

In 2011, EPA’s office of Air Quality Planning and Standards (OAQPS) conducted an assessment1 of continuous PM2.5 FEMs with collocated FRMs throughout the country. The study comprised of 61 Met One beta attenuation monitors (BAM 1020) and 17 Thermo Scientific 8500 filter dynamic measurement system (FDMS) each monitor reported to EPA’s Air Quality System (AQS). Valid 24-hour concentrations were compared to data quality objectives (DQOs) that are defined in 40 CFR Part 53 for FEM designation. Figure 1 presents the slope and intercept comparison for each monitor type (61 BAM units and 17 FDMS units). The orange box identifies the acceptable slope and intercept criteria defined in 40 CFR Part 53, and the red arrow points to the average.

Monitoring Figure 1

The study found that a majority of collocated FEMs produced data meeting DQOs while a third of the monitors had unacceptable slopes and/or intercepts. The study also found that most stations that were reporting within DQOs also had a positive bias leading to higher overall concentrations being measured by the continuous monitor. EPA recommended that state agencies running continuous PM2.5 FEMs that did not meet DQOs use the FRM as the primary sampler for reporting concentrations while working toward improving the FEM data. While SLAMS networks include monitors that are run indefinitely, PSD projects are for a finite term and any data loss due to poor sampler performance can delay the submission of the permit application and/or granting of a permit.

In conclusion, for PSD projects where pre-construction monitoring is required, several factors are key when choosing the appropriate method for the monitoring program. High resolution data from continuous FEM monitors provide a more complete annual dataset that can be analyzed for particulate origination. However, 40 CFR 53 requires the use of an FRM designated sampler for the first collocated sampler in a network. Comparability between FEM and FRM sampler data can be difficult and presents a challenge for meeting the established quality assurance criteria. A study conducted by EPA-OAQPS comparing continuous FEM monitors to FRM samplers found that a good portion of sites reporting to the AQS system lacked acceptable performance data questioning the use of continuous FEMs. Meteorological Solutions Inc., a Trinity Consultants company, has a thorough understanding of the quality assurance requirements of the PSD monitoring program and can design, build and implement a quality/robust monitoring program that will meet a sources monitoring goals. Defensibility of the data is key and using a high resolution, well performing method is essential for PSD monitoring.

“Purpose and Goals of PM Continuous Program and FEM Status”, Tim Hanley, EPA OAQPS – AAMG, November 6, 2006

“Finally, a Continuous FEM for PM2.5”, Gobeli, Meyer, Schloesser, Pottberg, AWMA EM February 2009

“Guidance for Using Continuous Monitors in PM2.5 Monitoring Networks,” EPA-454/R-98-012, May 1998

EPA Memo to PM NAAQS Review Docket, “Assessment of PM2.5 FEMs Compared to Collocated FRMs”, (EPA-HQ-OAR-2007-0492), dated April 7, 2011

National Ambient Air Quality Standards for Particulate Matter. Final Rule, Federal Register 2006, 71, 61236

EPA Guidance Memo, “Guidance for PM2.5 Permit Modeling”, (EPA-454/B-14-001), dated May 20, 2014

AMTIC – Sampling Schedule Calendar

AMTIC - Air Monitoring Methods- Criteria Pollutants Designated EPA Reference and Equivalent Methods


1 EPA Memo to PM NAAQS Review Docket, “Assessment of PM2.5 FEMs Compared to Collocated FRMs”, (EPA-HQ-OAR-2007-0492), dated April 7, 2011