8.2 Hazardous Air Pollutants, CEMS, and EPA Reference Methods

Hazardous Air Pollutants, CEMS, and EPA Reference Methods

NAAQS govern six ubiquitous pollutants at ambient monitors. A separate part of the Clean Air Act governs the far larger list of hazardous air pollutants at the stack, which is where continuous emission monitoring and the EPA reference test methods come in.

1. Hazardous Air Pollutants (HAPs) and NESHAP (Section 112)

Unlike criteria pollutants regulated by ambient concentration limits, toxic air contaminants are controlled under CAA Section 112 through source-specific emission standards.

Statutory List of HAPs

The 1990 CAA Amendments established a statutory list of 189 hazardous air pollutants (subsequently revised to 187 HAPs, including benzene, asbestos, vinyl chloride, mercury, polychlorinated biphenyls, methylene chloride, and formaldehydes). HAPs are substances known or suspected to cause cancer, genetic mutations, neurotoxicity, or reproductive harm.

Maximum Achievable Control Technology (MACT)

EPA issues National Emission Standards for Hazardous Air Pollutants (NESHAP) based on technology performance:

• Major Source Threshold: Any stationary source that emits or has the potential to emit ≥ 10 tons/year of any single HAP or ≥ 25 tons/year of any combination of HAPs.
• MACT Floor for Existing Sources: The average emission limitation achieved by the top 12% of best-performing existing sources in that industrial category.
• MACT Floor for New Sources: The emission control achieved in practice by the single best-controlled similar source.
• Risk and Technology Review (RTR): Eight years after MACT promulgation, EPA conducts a statutory residual risk assessment. If residual cancer risk to the most exposed individual exceeds 1 in 10⁶ (10⁻⁶), EPA must implement more stringent health-based standards.

---

2. Stationary Source Continuous Emission Monitoring Systems (CEMS)

Continuous Emission Monitoring Systems (CEMS) provide real-time, uninterrupted measurement of stack gas effluent concentrations (e.g., SO2, NOx, CO, CO2, O2, total VOCs) and volumetric flow rates.

CEMS Architecture

1. Sample Extraction / Conditioning Interface: Heated sample probe, heated Teflon sample line (> 120°C to prevent acid gas moisture condensation), chiller/dryer to remove water vapor, particulate filters.
2. Analytical Gas Analyzers:
   • SO2 & NOx: Non-Dispersive Ultraviolet (NDUV) or Chemiluminescence (for NOx via reaction with ozone: NO + O3 → NO2(*) + hν).
   • CO & CO2: Non-Dispersive Infrared (NDIR) spectroscopy based on Beer-Lambert infrared light absorption.
   • O2 Diluent: Zirconium oxide (ZrO2) electrochemical cell or paramagnetic cell.
3. Data Acquisition and Handling System (DAHS): Logs raw data, performs temperature/pressure/moisture corrections, and computes 1-hour and rolling compliance averages.

Quality Assurance and Performance Specifications (40 CFR Part 60 / Part 75)

• Daily Calibration Drift (CD) Test: Zero and span calibration gases injected directly at the probe tip daily. Drift must not exceed ≤ 2.5% of span value.
• Relative Accuracy Test Audit (RATA): An annual or semi-annual comprehensive audit comparing CEMS output against concurrent manual EPA Reference Method testing across a minimum of 9 test runs:

$\text{RA} = \frac{|\bar{d}| + |\text{CC}|}{\overline{\text{RM}}} \times 100\%$

Where |d̄| is the mean difference between reference method and CEMS, |CC| is the 95% confidence coefficient, and RM-bar is the average reference method value. Regulatory compliance requires RA ≤ 10.0% (or ≤ 20.0% under Part 60 general applications).

---

3. EPA Stationary Source Reference Methods (Methods 1–5)

EPA 40 CFR Part 60 Appendix A details manual reference methods for stack testing:

  Method 1: Traverse Point Selection ──► 8 duct diameters downstream / 2 upstream from flow disturbances
                                             │
  Method 2: Stack Gas Velocity & Flow ─► Type S Pitot Tube + differential manometer (Δp)
                                             │
  Method 3/3A: Molecular Weight & O2/CO2► Orsat chemical absorption or NDIR/paramagnetic instrumental
                                             │
  Method 4: Stack Gas Moisture Content ─► Chilled impinger condensation train (volumetric/gravimetric)
                                             │
  Method 5: Isokinetic Particulate ────► Heated probe + filter (248°F) + isokinetic nozzle control

EPA Method 1: Sample and Velocity Traverses for Stationary Sources

• Selects sampling traverse point locations across the stack cross-section.
• Ideal Criteria: Sampling location at least 8 duct diameters downstream and 2 duct diameters upstream from any flow disturbance (bend, expansion, contraction, fan, or damper).
• Minimum Acceptable Criteria: At least 2 diameters downstream and 0.5 diameters upstream (requires an increased number of traverse points, up to 48 points for circular stacks).

EPA Method 2: Determination of Stack Gas Velocity and Volumetric Flow Rate

Measures gas velocity pressure using a calibrated Type S Pitot Tube (baseline coefficient Cp ≈ 0.84) and inclined manometer:

$v_s = K_p \cdot C_p \cdot \sqrt{\Delta p} \cdot \sqrt{\frac{T_s}{P_s \cdot M_s}}$

Where:

• vs = Stack gas velocity (m/s or ft/sec)
• Kp = Velocity equation constant (34.97 m/s · g/(g-mole·K) · mm Hg/mm H2O or 85.49 ft/s in English units)
• Cp = Pitot tube coefficient (dimensionless)
• Δ p = Velocity head differential pressure (in. H2O or mm H2O)
• Ts = Absolute stack temperature (K or °R)
• Ps = Absolute stack pressure (mm Hg or in. Hg)
• Ms = Wet molecular weight of stack gas (g/g-mole or lb/lb-mole)

EPA Method 3 / 3A: Gas Analysis for Molecular Weight

Measures concentrations of O2, CO2, and CO to compute dry stack molecular weight (Md):

$M_d = (0.440 \times \%\text{CO}_2) + (0.320 \times \%\text{O}_2) + [0.280 \times (\%\text{N}_2 + \%\text{CO})]$

Wet molecular weight (Ms) adjusts for stack moisture content (Bws):

$M_s = M_d \cdot (1 - B_{ws}) + (18.0 \cdot B_{ws})$

EPA Method 4: Stack Gas Moisture Content

Gas is extracted through a chilled condenser/impinger train. Moisture content (Bws) is calculated from condensed water volume (Vlc) and dry metered gas volume (V(m(std))):

$B_{ws} = \frac{V_{w(std)}}{V_{m(std)} + V_{w(std)}}$

EPA Method 5: Determination of Particulate Matter Emissions (Isokinetic Sampling)

Method 5 extracts particulate matter gravimetrically through a precision nozzle, heated probe, and heated filter maintained at 120 ± 14°C (248 ± 25°F), followed by a chilled impinger train.

                                  ┌─────────────────────────────┐
                                  │ Isokinetic Sampling Physics │
                                  └──────────────┬──────────────┘
                                                 │
       ┌─────────────────────────────────────────┼─────────────────────────────────────────┐
       ▼                                         ▼                                         ▼
┌──────────────────────────────┐ ┌──────────────────────────────┐ ┌──────────────────────────────┐
│     Isokinetic (Vn = Vs)     │ │    Under-Isokinetic (Vn < Vs)│ │    Over-Isokinetic (Vn > Vs) │
├──────────────────────────────┤ ├──────────────────────────────┤ ├──────────────────────────────┤
│ • Streamlines enter straight │ │ • Gas streamlines diverge    │ │ • Gas streamlines converge   │
│ • True particulate mass      │ │ • Large particles over-enter │ │ • Large particles under-enter│
│ • 100% accurate measurement  │ │ • Biased HIGH (False High)   │ │ • Biased LOW (False Low)     │
└──────────────────────────────┘ └──────────────────────────────┘ └──────────────────────────────┘

The Isokinetic Sampling Principle

• Isokinetic Condition (vn = vs): The linear gas velocity entering the nozzle probe (vn) equals the stack gas velocity approaching the nozzle (vs). Particulates of all sizes enter proportionally without streamline deflection.
• Under-Isokinetic (vn < vs, Sampling Too Slowly, I < 90%): Gas streamlines diverge around the nozzle opening. Heavy particles with high aerodynamic inertia cross the divergent streamlines and enter the nozzle, resulting in an artificially high (falsely elevated) measured mass concentration.
• Over-Isokinetic (vn > vs, Sampling Too Fast, I > 110%): Gas streamlines converge into the nozzle opening. Heavy particles resist curvature due to inertia and travel past the nozzle rim, resulting in an artificially low (falsely depressed) measured mass concentration.
• Acceptable Regulatory Window: The Isokinetic Variation Percentage (I) must fall within 90% ≤ I ≤ 110% (100 ± 10%) for a test run to be legally valid.

Isokinetic Variation Equation

$I = \frac{100 \cdot T_s \cdot [K_3 \cdot V_{lc} + (V_m \cdot Y / T_m) \cdot (P_{bar} + \Delta H / 13.6)]}{\theta \cdot v_s \cdot P_s \cdot A_n}$

Where:

• I = Percent isokinetic variation (must be 90-110%)
• Vlc = Total volume of liquid condensed in impingers (mL)
• Vm = Dry gas meter volume recorded (dcf or dcm)
• Y = Dry gas meter calibration factor
• Tm = Absolute temperature at dry gas meter (°R or K)
• Pbar = Barometric pressure (in. Hg or mm Hg)
• Δ H = Average pressure differential across meter orifice (in. H2O)
• θ = Total elapsed sampling duration (minutes)
• An = Cross-sectional area of sampling nozzle (ft² or m²)

---

4. Worked Step-by-Step Calculation Examples

Worked Example 7.1: EPA Method 5 Isokinetic Variation (I%) Calculation

Problem: An environmental compliance audit tests a coal-fired boiler stack using EPA Method 5. The field test run yields the following data:

• Stack gas velocity (vs): 50.0 ft/s
• Sampling duration (θ): 60.0 minutes
• Stack temperature (Ts): 240°F = 700 R
• Stack pressure (Ps): 29.80 in. Hg
• Standard dry gas volume sampled (V(m(std))): 45.00 dscf
• Stack moisture content fraction (Bws): 0.0400 (4.0% moisture)
• Sampling nozzle inner diameter (Dn): 0.250 inches
• Standard conditions: Tstd = 528 R, Pstd = 29.92 in. Hg

Verify whether this sampling run meets the EPA Method 5 isokinetic acceptance criterion (90% ≤ I ≤ 110%).

Solution Steps:

1. Calculate Nozzle Cross-Sectional Area (An): $D_n = \frac{0.250\text{ in}}{12\text{ in/ft}} = 0.020833\text{ ft}$ $A_n = \frac{\pi}{4} \cdot D_n^2 = \frac{\pi}{4} \cdot (0.020833\text{ ft})^2 = 3.4088 \times 10^{-4}\text{ ft}^2$

2. Apply the Standard Isokinetic Formula: $I = \frac{100 \cdot V_{m(std)} \cdot T_s \cdot P_{std}}{T_{std} \cdot P_s \cdot v_s \cdot A_n \cdot \theta \cdot 60 \cdot (1 - B_{ws})}$

3. Compute Numerator and Denominator: $\text{Numerator} = 100 \times 45.00\text{ dscf} \times 700\text{ R} \times 29.92\text{ in. Hg} = 94,248,000$ $\text{Denominator} = 528 \times 29.80 \times 50.0 \times (3.4088 \times 10^{-4}) \times 60 \times 60 \times (1 - 0.0400)$ $\text{Denominator} = 528 \times 29.80 \times 50.0 \times 0.00034088 \times 3600 \times 0.9600 = 92,624,357$

4. Calculate Percent Isokinetic (I): $I = \frac{94,248,000}{92,624,357} = 101.75\% \approx 101.8\%$

Result: The isokinetic percentage is 101.8%, which falls well within the required regulatory window of 90% ≤ I ≤ 110%. The test run is valid.

---

Worked Example 7.2: Particulate Concentration and Mass Emission Rate

Problem: From the Method 5 test in Example 7.1, the total gravimetric particulate catch recovered from the probe wash and filter is mn = 45.0 mg (0.0450 g). The stack has a circular diameter of 6.0 ft with a volumetric stack gas flow rate of Qstd = 22,000 dscfm (dry standard cubic feet per minute).

1. Calculate the particulate concentration in grains per dry standard cubic foot (gr/dscf) and in milligrams per dry standard cubic meter (mg/dscm). (1 lb = 7,000 grains; 1 grain = 64.7989 mg; 1 dscm = 35.3147 dscf).
2. Calculate the particulate mass emission rate (E) in pounds per hour (lb/hr).

Solution Steps:

1. Calculate Particulate Concentration (cs):

   • Convert mass in mg to grains: $\text{Mass in grains} = \frac{45.0\text{ mg}}{64.7989\text{ mg/gr}} = 0.69446\text{ grains}$
   • In gr/dscf: $c_s = \frac{0.69446\text{ grains}}{45.00\text{ dscf}} = 0.01543\text{ gr/dscf}$
   • In mg/dscm: $V_{m(std)} (\text{m}^3) = \frac{45.00\text{ dscf}}{35.3147\text{ dscf/m}^3} = 1.27426\text{ dscm}$ $c_s = \frac{45.0\text{ mg}}{1.27426\text{ dscm}} = 35.315\text{ mg/dscm} \approx 35.3\text{ mg/dscm}$

2. Calculate Mass Emission Rate (E): $E = c_s (\text{gr/dscf}) \times Q_{std} (\text{dscfm}) \times \frac{60\text{ min/hr}}{7,000\text{ gr/lb}}$ $E = 0.015432\text{ gr/dscf} \times 22,000\text{ dscfm} \times \frac{60}{7,000}$ $E = \frac{20,370.24}{7,000} = 2.910\text{ lb/hr} \approx 2.91\text{ lb/hr}$

Result: The particulate concentration is 0.0154 gr/dscf (35.3 mg/dscm), generating a mass emission rate of 2.91 lb/hr.