Over the past year, Trinity assisted a chemical company with the permitting and installation of a new 220 MMBtu/hr package boiler. The new natural gas-fired boiler replaces coal-fired units that were installed in the 1940s. Navigating through the numerous, and at times conflicting, regulatory requirements applicable to such a unit has been challenging. With careful planning and implementation, the project resulted in a successful and compliant start-up of the new unit.
For more than 60 years, steam needed for process and comfort heating, as well as electricity generation at the facility, was provided by boilers primarily fired by coal with No. 6 fuel oil as backup. Electrostatic precipitators (ESPs) were used to control particulate matter and the main boiler stack was equipped with a continuous opacity monitoring system (COMS). Additionally, the facility’s Title V permit required extensive and labor-intensive monitoring which included field verification of ESP operating parameters, sampling and analyses for verification of the sulfur content in the coal burned, operation and maintenance of a local sulfur dioxide ambient air monitoring station, and monitoring of boiler operating parameters for parametric determination of nitrogen oxides (NOx) emissions.
The new boiler is fired primarily with natural gas and is capable of burning No. 2 fuel oil as backup. The boiler is equipped with low NOx burner technology, flue gas recirculation, and Selective Catalytic Reduction (SCR) technology for the reduction of NOx emissions when burning natural gas. The SCR system uses a vanadium/titanium catalyst with anhydrous ammonia as the reductant. The reductant is injected immediately downstream of the boiler flue gas outlet where the temperature range is approximately 500 to 700o F for maximum NOx conversion. The system is designed to reduce NOx emissions by at least 90% during normal operation.
The boiler exhaust stack is equipped with a NOx Continuous Emissions Monitoring System (CEMS) to measure and record NOx emissions. This is the primary NOx CEMS. Additionally, a sample line of stack exhaust gas is directed to a small oxidizer which converts residual ammonia from the SCR to nitrogen monoxide. The converted gas sample is then analyzed in a secondary NOx CEMS and the difference between the NOx readings of the primary and secondary CEMS is calculated by the Data Acquisition and Handling System (DAHS) to determine ammonia slip. The DAHS also receives data from the facility’s distributed control system so that boiler operators can monitor other operating parameters related to the boiler’s performance on the DAHS display screen.
Regulatory Requirements and Compliance Considerations
The new boiler meets the definition of an affected facility under the NSPS for industrial,
commercial, and institutional steam generating units contained in 40 CFR 60, Subpart Db (NSPS Db). These requirements are fuel dependent and, for a unit firing natural gas and “very low sulfur fuel” (containing 0.30 weight percent sulfur or less), include limits on NOx at all times and opacity during fuel oil firing. The No. 2 fuel oil used as a backup meets the requirements of “very low sulfur fuel.” Such units are exempt from NSPS Db limitations on particulate matter and sulfur dioxide emissions but must maintain
fuel records certifying compliance with the very low sulfur fuel criteria.
NSPS Db limits opacity to 20% based on a six-minute average except for one six-minute
period per hour not to exceed 27% opacity. The new boiler is also subject to a State Implementation Plan (SIP) opacity standard of 20%, based on a three-minute average in any one hour or fifteen minutes in any 24-hour period. These similar, but not identical opacity requirements, were both considered in determining the facility’s compliance strategy.
NSPS Db allows for the implementation of a site-specific opacity monitoring plan in lieu of installing a COMS for units fired with natural gas and very low sulfur fuel oil. In developing the site-specific monitoring plan, Trinity presented two options for the client’s consideration.
Option 1: Method 22/Method 9 Tiered Frequency Approach
Option 1 was based on the alternate monitoring options added to NSPS Db in the January 2009 revision (outlined in 40 CFR §60.48b(a)). Affected facilities meeting certain requirements may choose the alternate monitoring option in lieu of installing a COMS. This approach provides for the use of Method 9 and Method 22 opacity determinations for sources that are expected to have no significant visible emissions. The required duration and frequency for these determinations is based on the results of the most recent observations. For example, if the most recent Method 9 opacity determination yielded a result of 10% or less, a 10-minute Method 22 observation may be conducted daily when the affected unit fires fuel for which an opacity standard is applicable to demonstrate that the sum of the occurrences of any visible emissions does not exceed 5% of the observation period. That is, visible emissions are observed for less than 30 seconds during a 10-minute observation period. If the sum of visible emissions occurrences exceeds the 5% threshold, then a 30-minute Method 22 observation must be conducted immediately. If the sum of the occurrences of visible emissions is again greater than the 5% threshold (i.e., visible emissions are observed for 90 seconds or more during the 30-minute observation period) then a new Method 9 determination must be performed within 24 hours. On the other hand, if no visible emissions are observed for 30 operating days during which an opacity standard is applicable, Method 22 observations may be reduced to once every seven applicable operating days. If any visible emissions are observed during the once per seven day schedule, then daily observations must be resumed.
Option 2: Method 9 Regular Frequency Approach
The second site-specific monitoring plan option was based on a review of letters previously issued by EPA with respect to acceptable alternate monitoring plans for NSPS Db opacity requirements contained in the Applicability Determination Index database. The second option requires performing a 6-minute Method 9 observation daily during periods of No. 2 fuel oil firing when the opacity standard is applicable.
Both options would require an alternate monitoring plan approved by the regulatory authority. Although the second approach requires the client to obtain Method 9 certification for several employees, it was chosen over the first option because it is a less complex approach and the facility does not anticipate burning No. 2 fuel oil very often. Further, since the Method 9 determinations are based on six-minute observations, compliance with the SIP opacity requirement based on three-minute observations can be confirmed as well.
The new boiler is subject to the NSPS Db NOx emissions limit of 0.20 lb/MMBtu based on a 30-day rolling average which applies at all times, whether burning natural gas or fuel oil, and including periods of startup, shutdown, and malfunction. The boiler’s construction permit also contains a SIP NOx limit of 5 ppmvd (corrected to 3% oxygen) based on a three-hour rolling average basis. Unlike the NSPS Db limit, the SIP limit only applies when firing natural gas. Recognizing the need for the SCR to achieve a specified minimum temperature in order to efficiently remove NOx, the SIP limit does not apply during periods of boiler cold start. As such, this requirement led to the inclusion of a definition for cold start in the permit, limiting cold start duration to a certain number of hours per year, in order to clarify compliance requirements per year.
Compliance with both NOx limits is determined based on hourly block averages calculated from the NOx values measured by the CEMS analyzer. However, the difference in averaging times and applicable operating conditions between the two limits complicated the programming of the DAHS. The data is classified or “tagged” with process and monitoring codes that indicate the specific operating conditions at the time it is obtained (e.g., data is tagged as cold startup, warm startup, normal operation, shutdown, malfunction, etc.) based on digital signals received from the plant DCS, the CEMS sample system, or manually input by boiler operators. Similarly, the fuel type being burned at the time a given data point is collected is also determined from signals received from the DCS. The DAHS is programmed to evaluate these process and monitoring codes in conjunction with the data validation criteria required by NSPS Db to determine what data should be included in the calculation of the three-hour and 30-day average NOx emissions values for comparison against the applicable limitation. Given these complexities, great care was taken to ensure that the DAHS vendor and plant operations team properly programmed the system before initial startup of the unit.
In addition to the aforementioned implementation challenges, there were a myriad of regulatory requirements for installing and operating a CEMS. The NOx CEMS is required to meet the requirements of 40 CFR Part 60 which includes the performance specifications in Appendix B and the quality assurance procedures in Appendix F. These requirements include regular calibration of the analyzers and quarterly audit testing – Relative Accuracy Test Audits (RATAs) and Cylinder Gas Audits (CGAs). Additionally, a new CEMS must be certified initially for compliance with the applicable federal performance specifications. The required timing for this initial certification, which involves a RATA and 7-day calibration drift test, is not specifically addressed in the construction permit.
Performance Specification 2 in NSPS Appendix B indicates that the NOx CEMS should be certified at the time of installationor soon after and whenever specified in the regulations. NSPS Subpart A indicates that it should be performed at the time of the required initial performance testing of the emissions unit or within 30 days of such testing. However, NSPS Db requires such performance testing be conducted within
60 days of achieving the maximum production rate at which the affected facility will be operated but not later than 180 days after initial startup of the boiler. This could potentially delay the CEMS certification to up to 210 days or seven months after the boiler begins normal operations. Since the CEMS is relied on for process control of the SCR, the project team decided that the CEMS certification should be conducted as soon after installation as possible.
NSPS Appendix F requires the development and implementation of a QA/QC plan for operating and maintaining the CEMS. Trinity assisted in the development of this plan which provides step-by-step procedures for calibrating and auditing CEMS performance, summarizes the data recording, calculations and reporting requirements, and provides preventative maintenance (PM) schedules and corrective actions for system malfunctions. Trinity recommended that the facility obtain a maintenance contract for at least the first year to have an outside contractor with specific CEMS expertise perform quarterly PM and to be on call in the event of a system breakdown. By doing so, the site maintenance personnel, who are also becoming familiar with the maintenance routines for the new boiler and SCR equipment, will have assistance and a resource for CEMS maintenance. Since NSPS requires the facility have emissions data for a minimum of 75% of the operating hours in any day that the boiler operates, or in at least 22 out of 30 operating days, it is imperative that the facility be prepared to repair a malfunctioning CEMS in a timely manner.
To monitor ammonia slip, the use of a primary and secondary NOx CEMS with an ammonia converter prior to the secondary unit was accepted by the state regulatory agency as an alternative to installing a direct ammonia analyzer. This regulatory acceptance was contingent on the system being certified, however there are no performance specifications available for such a system. Absent any federal or state performance specification, the proposed certification strategy was to confirm the accuracy of the secondary NOx CEMS via a CGA, both initially and quarterly going forward. Additionally, discrete stack testing for ammonia via EPA Conditional Test Method 027 was proposed for the initial certification of the ammonia monitoring system. Stack testing will also be performed annually to provide data to adjust the correction factor for the calculation of ammonia slip as needed.
Initial compliance with NSPS Db can be challenging, even for a boiler with a relatively clean fuel firing scenario of natural gas and No. 2 fuel oil. The new boiler is subject not only to the requirements specific to boilers in NSPS Subpart Db, but also to the general provisions of Subpart A, the performance specifications of Appendix B, the quality control requirements of Appendix F and applicable SIP regulations (including project specific construction permit requirements). However, careful planning and implementation (including assessing compliance risks during the state construction permit issuance process) paid off with a permit that provides methods for continous compliance demonstration with maximum operational flexibility and minimum compliance burdens on the facility.