Monitoring System Audits & Certification
CEMS & RATA Testing for Southern California Facilities
Independent reference method testing for continuous emissions monitoring systems — RATAs, cylinder gas audits, linearity checks, and certification under SCAQMD Rule 218.3 and EPA Part 60/75.
The Process
What CEMS Certification and RATA Testing Actually Involve
A continuous emissions monitoring system reports your emissions to the District every hour of every day. A Relative Accuracy Test Audit is the periodic check that those numbers are true. We bring reference method equipment to your stack, measure the same gas your analyzer is measuring, and compare the two data sets run for run.
A valid RATA needs a minimum of nine sets of paired reference method and monitor data. That number is the same under PS-2 section 8.4.4, under 40 CFR Part 75 Appendix A, and under SCAQMD Rule 218.3(f)(3)(A). You may run more than nine. If you do, up to three sets can be rejected as long as at least nine are used in the calculation — and all data, including the rejected runs, still gets reported.
Relative accuracy is not a single number you can memorize. It depends on the program, the pollutant, and how close your unit runs to its limit. Under PS-2, a monitor on a unit emitting at or above 50 percent of the standard is judged against the reference method mean and must come in at or below 20.0 percent. Below 50 percent of the standard, the calculation switches to the emission standard in the denominator and the criterion tightens to 10.0 percent. PS-4A allows CO monitors 10 percent against the reference method mean, 5 percent against the emission standard, or agreement within 5 ppmv.
SCAQMD sets its own criteria on top of that. Rule 218.3(f)(3)(E) allows 20.0 percent relative accuracy for pollutant concentrations, or a de minimis value instead: 0.5 ppm for NOx, rising to 1.0 ppm where the unit's permitted concentration limit is above 5.0 ppm; 2.0 ppm for SO2; 2.0 ppm for CO, or the permitted limit where that limit is lower. Diluent monitors are held to 10.0 percent, relaxing to 20.0 percent when measured O2 or CO2 is at or below 15 percent. Stack flow is 15.0 percent. Those de minimis alternatives exist because at very low concentrations a percentage criterion stops being physically meaningful — a well-built monitor can fail a percentage test simply because the denominator is tiny.
Certification is the larger exercise. Before a new or rebuilt CEMS can report compliance data, Rule 218.3 requires the system to run a minimum of 168 continuous hours, then pass a battery of tests: a seven-day calibration drift test built from eight calibration error tests across seven days, a linearity error check, response time, NOx converter efficiency above 90 percent, sampling system bias, and the RATA itself. Under Rule 218.3(f)(6), those certification tests have to be performed by a firm approved through South Coast AQMD's Laboratory Approval Program. We hold that approval — LAP #07-101.
For the deeper regulatory walkthrough, see our explainer on CEMS and RATA testing under SCAQMD Rules 218.2 and 218.3.
Scope
What We Audit, and Under Which Rule
Which of these applies to your unit is set by your permit and by which program your monitor was certified under. If you are not sure, send us the permit conditions and we will tell you.
| Scope | Governing Rule | Jurisdiction |
|---|---|---|
| Relative Accuracy Test Audit — annual | Rule 218.3(g)(2); 40 CFR 60 App. F §5.1.1 | SCAQMD · EPA Part 60 |
| Relative Accuracy Test Audit — semiannual | 40 CFR 75 App. B §2.3.1.1 | EPA Part 75 |
| Initial CEMS certification & recertification | Rule 218.3(f); PS-2, PS-4A | SCAQMD · EPA Part 60 |
| Cylinder Gas Audit (CGA) | Rule 218.3(g)(3); 40 CFR 60 App. F §5.1.2 | SCAQMD · EPA Part 60 |
| Relative Accuracy Audit (RAA) | 40 CFR 60 App. F §5.1.3 | EPA Part 60 |
| Seven-day calibration drift test | Rule 218.3(f)(1); PS-2 §8.3.1 | SCAQMD · EPA Part 60 |
| Linearity error check | 40 CFR 75 App. B §2.2.1 | EPA Part 75 |
| Bias test & bias adjustment factor | 40 CFR 75 App. A Eq. A-11/A-12 | EPA Part 75 |
| Low level RATA / bias factor determination | Rule 218.3 — LLR/BFD | SCAQMD |
| Stratification & cyclonic flow checks | PS-2 §8.1.3.2; Rule 218.3(f)(4) | SCAQMD · EPA Part 60 |
Where a RATA sits alongside a periodic compliance source test, we run both on one mobilization. See source testing and stack testing for that side of the work, and SCAQMD compliance for how District rule obligations stack up against the federal ones.
Reference Methods & Parameters
What We Measure Against Your Monitor
PARAMETERS NOₓ · CO · SO₂ · O₂ · CO₂ Stack flow · THC · Opacity · Moisture PERFORMANCE SPECIFICATIONS PS-2 — SO₂ and NOₓ CEMS PS-3 — O₂ and CO₂ diluent monitors PS-4A — CO CEMS, low emission standards PS-6 — Continuous emission rate monitoring QUALITY ASSURANCE FRAMEWORKS 40 CFR 60 App. F — Procedure 1, gas CEMS QA 40 CFR 75 App. A — Certification & accuracy specs 40 CFR 75 App. B — QA/QC, linearity, RATA frequency SCAQMD Rule 218.2 — CEMS general provisions SCAQMD Rule 218.3 — Certification & QA specifications
Reference methods are selected to match the monitor and the governing program. PS-2 names Methods 3B, 4, 6 and 7 — diluent, moisture, SO2 and NOx — as the defaults unless an applicable subpart says otherwise.
Scheduling
Scheduling Around Your Unit
RATA windows are hard to hit for a reason that has nothing to do with testing. The audit has to happen while the unit is running, at a representative load, inside a fixed regulatory window — and your unit runs when the business needs it to, not when the calendar says the RATA is due.
The windows are narrower than people expect. Under SCAQMD Rule 218.3(g)(2)(A), the annual RATA is due no later than the end of the calendar quarter of the previous relative accuracy test, and it must be run in the as-found operating condition. Under 40 CFR Part 75, the clock counts QA operating quarters rather than calendar quarters, and no more than eight successive calendar quarters may elapse before a RATA is performed.
Miss the window and the exposure is not just a late test. Under Appendix F, data from an out-of-control CEMS may not be used to calculate compliance, nor counted toward the minimum data availability your permit requires. You lose the data, and then you have to explain the gap.
There is relief written into the rules for units that are down. Rule 218.3(g)(2)(D) allows the RATA within 14 days after a non-operating unit restarts and resumes normal operation. Appendix F section 5.1.4 handles the parallel case federally: if the facility is off-line in the fourth calendar quarter since the last RATA, the RATA moves to the quarter in which the unit recommences operation. Part 75 provides a 720 unit operating hour grace period past the eighth quarter. Each has conditions.
Our side of it is unglamorous. We hold the crew and the reference method equipment against your window — days, evenings, weekends, turnarounds. Dispatch is from Stanton, so mobilization is short. If your unit comes up at 2 a.m. on a Saturday and that is the representative load, that is when we test.
Questions we get,
by phone, mostly.
If yours isn't here, call dispatch. Wally Moe picks up.
Get the RATA on the Calendar
Send us the permit conditions and the monitor configuration. We'll tell you which program governs, what's actually due, and when — quote back in 24–48 hours.