How to Plan for a Relative Accuracy Test Audit (RATA)
Learn about what a Relative Accuracy Test Audit (RATA) is and how to plan for a successful RATA Test.
What is a Relative Accuracy Test Audit (RATA), and Why Does it Matter?
According to the US Environmental Protection Agency (EPA), a Relative Accuracy Test Audit (RATA) must be conducted once every four calendar quarters. Conduct the RATA as described for the Relative Accuracy test procedure, which is as follows:
“The absolute mean difference between the gas concentration or emission rate determined by the CEMS and the value determined by the RM’s plus the 2.5 percent error confidence coefficient of a series of tests divided by the mean of the RM tests or the applicable emission limit.”
In addition, analyze the appropriate performance audit samples received from EPA as described in the applicable sampling methods.
What are Pre-RATA Activities That My Facility Should Participate in?
The first step in preparing for an upcoming RATA Test is to conclude when the RATA’s are due. This includes finding and reviewing the in-house RATA procedure designed for the Continuous Emission Monitoring Systems (CEMS) in your facility. Once you have located and accomplished the in-house procedure, generating a copy of the “QA Test Status Report” is the next step. This report will provide information about your RATA history and any other relevant information regarding your RATA Test.
Once you’ve generated a copy of the “QA Test Status Report,” verify in StackVision™ when the last RATA was completed for all of your Part 75 systems and when the next RATA needs to be completed. For Part 75, some systems might require the next RATA to be completed within two (2) calendar quarters or within four (4) calendars. You must check the results from the last RATA for each system.
After you have checked the last RATA for each system, review the operating load records recorded in StackVision for not less than four calendar quarters. Each quarter must contain at least 168 operating hours to qualify as a “QA operating quarter.” Use the “Operating Load Analysis Report” (Reports ⇒ Part 75 ⇒ Operating Load Analysis) starting with the first day following the completion of the previous RATA for each system. Gas RATAs are not necessarily performed all on the same day, so review your previous RATA records carefully.
Every five (5) years, the flow RATA must be performed in each of the three operating load ranges. Review the existing RATA record to determine when the last three load flow RATA were performed. During flow RATAs, their loads need to be separated by no less than 25% of the normal operating range. The load analysis results need to be reported or updated in the “Load” records of ECMPS.
The RATA for gas systems is conducted a single load and must occur either of the two most frequently operated load ranges identified in the Load Analysis Report. Prepare for the RATA by selecting the performance date and scheduling the RATA with various parties such as your boss, operating staff, etc. In many cases, RATA dates should be scheduled at least several weeks to months in advance.
Most state and local air quality agencies require advanced written notice stating when performance evaluation tests (RATAs) will be conducted. Submit this notice under the signature of the Designated Official. Notification is also required to the regional office of the EPA at least 21 calendar days in advance of the scheduled test date. It must fall under certain recertification provisions to avoid this notice requirement, reviewed in section §75.61(a).
What Test Activities Should My Facility Complete the Day of Our RATA Test?
Preparation is key to successfully passing a RATA test. To prepare on the day of your RATA test, review your in-house RATA procedures (again) so you can answer questions from those you work with or interface with. Be sure to check the testing schedule for that day with the control room operator and the supervisory operating staff. then, give an estimate of how long the RATA test will take to complete the day’s testing schedule.
Once you have given an estimation, discuss the overall strategy of how the RATA will be conducted with your stack testing lead. Decide with your lead whether the gas or flow RATAs will be performed first, at what load, and how long each run will last. You should compare the RATA results generated by StackVision and the results calculated by your stack testers. The two sets of results should match.
Each year, the stack testers should perform a stratification test before beginning the gas RATA to determine the number of sample points needed for the stack testers to get a representative set of reading for each run for the gas RATA. This requirement is identified in Appendix A, §6.5.5., §6.5.6.1. Stratification test results are retained on-site as part of the RATA supplementary records.
Before beginning the flow RATA, the reference method traverse points must be selected using the Performance Specifications listed in Appendix A, §6.5.6. and following reference method 1 of Part 60 Appendix B. After entering data sets from about six or seven RATA runs, review the projected relative accuracy in StackVision, and decide if you need to run more than ten runs to allow the opportunity to drive the RA lower if close to 7.0%. Your final relative accuracy results need to be <7.5%.
You must report the gas cylinders used by the stack testers to perform their calibration checks in the QA EDR file. Then, get copies of the certification sheets for gas cylinders that the stack testers used in the QA check on their monitors and enter them into the “Cal Gas Manager” section of StackVision. Once the certificates are entered in the Gas Manager, they can be linked in the gas RATA records in the RATA Editor for each gas RATA performed. Be sure your stack testing company identifies who the “Qualified Individual” is for this RATA. To link to Part 75 RATA results through the RATA Editor, you need to have the required information about this individual in StackVision. This can be found in the “AETB” tab under “QA and Certificates” in StackVision.
Exchange the averages recorded for each run for each monitor/system between yourself and the stack testers after every two to three runs. After recording the results for seven or eight runs, compare the relative accuracy results calculated by StackVision and the results calculated by the stack testers. The two sets of results should match. If the results do not match, go back and review the results that both parties have recorded in their data systems and correct any discrepancies. Part 75 requires that at least nine sets of RATA runs be performed and recorded. ESC Spectrum recommends performing at least ten runs.
Acceptable RATA Relative Accuracy Results
Acceptable RATA Relative Accuracy (RA) results are:
- P75 gas and flow must not be >10.0%
- If RA is > 7.5 and < 10.0%, then the next RATA must be performed within the next four (4) calendar quarters. Many CEM operators strive to achieve a relative accuracy of < 5%
Repeating a RATA is possible and undoubtedly advisable, but remember, to keep the stack testing team on-site costs about $3,000 to $5,000. Bringing the stack testing team back on-site will cost even more due to additional mobilization costs.
Once the RATA runs begin, you have 168 hours to complete all necessary runs, although most single load RATAs can be completed within one (1) to six (6) hours. In the RATA editor, there is an option to “Optimize” the RATA results. If this option is selected, StackVision will evaluate all the data for your RATA runs and determine which set of nine pairs will give you the lowest relative accuracy (RA) results. StackVision will select the nine runs that gave the best relative accuracy (RA) results if ten runs are recorded.
Other than reviewing the relative accuracy results and the duration for when the next RATA must be performed (two or four calendar quarters), make sure to observe the Bias Adjustment Factors (BAFs) derived for each of the RATA results. The BAF calculated by the stack testers and in StackVision must match. BAFs are not used or applied by either CO2 or O2 hourly data.
Wrapping Up and Moving On From Your RATA Test
You should receive the final RATA report from your stack testers within two-to-three weeks. Once received, review and compare each of the averages recorded for each run, the relative accuracy results, and the other results shown on their summary page versus what you have recorded in the RATA Editor. If the two don’t agree, you must contact your testing contractor and work out those differences.
After finishing the steps mentioned above, review your in-house RATA procedure for updating hourly data based on the RATA results. Suppose during the final review; you changed any data in StackVision. In that case, you must reprocess the hourly data because a change in BAF could trigger the need to accumulate daily, monthly, quarterly, or yearly (YTD) totals or averages. Be sure to consult and follow your existing data processing procedures. The reprocessing begins with the first full hour following the last RATA run of the affected monitoring system.
When your RATA records are finalized in StackVision, generate a QA EDR file and run it through ECMPS. If the evaluation report shows any critical errors, they must be resolved in StackVision. Including a new file generated and re-evaluation in ECMPS. You must resolve any errors at this step. When the RATA is completed, make sure the data and results are saved. It is also essential to review the hourly data for hours surrounding when the RATA was conducted.
Final Thoughts:
- Does your RATA procedure need to be updated?
- Are there any lessons learned or more efficient steps you could follow uncovered during this round of testing?
- Consider locking down your RATA database so that it doesn’t get changed or altered. Use the “Data Locking Wizard” found under “Tools” in the Main Menu of StackVision.
How Can I Streamline my RATA Testing Process?
It is no secret that Relative Accuracy Test Audits (RATA) are time-consuming and complicated. ESC Spectrum offers services to streamline your RATA Testing Process with our RATAView™ software. Developed by stack testers for stack testers, RATAView was purpose-built to help meet the challenges of stack testing for an easier and faster testing experience while improving accuracy.
