In late 2008, T3, a division of Trinity Consultants, developed an environmental management information system (EMIS) to manage “near real-time” emissions data at a petroleum refinery. The system is designed to support Consent Decree requirements for compliance demonstrations with air quality regulations at the facility. Functionality delivered includes advanced data gathering, validation, rule-base correction, and averaging requirements for newly installed continuous emissions monitoring systems (CEMS). The system is designed to perform hourly emission calculations, determine compliance with reportable quantity and other emission limits, and generate notifications and reports to facilitate additional engineering assessments.
As a result of a Consent Decree, a refinery was required to install additional emission control devices. Two tail gas clean-up units (TGCU) were installed to reduce sulfur dioxide (SO2) emissions from the Sulfur Recovery Units (SRUs). A Selective Catalytic Reduction (SCR) unit was installed to reduce emissions of nitrogen oxides (NOx) from the Fluid Catalytic Cracking Unit (FCCU) flue gas, while a wet gas scrubber (WGS) was installed downstream of the SCR to control emissions of SO2 and particulate matter (PM). Finally, modifications were made at the Benzene Recovery Unit (BRU) to reduce hydrogen sulfide (H2S) emissions.
Besides the new control equipment installed at the refinery, additional limits and recordkeeping requirements were imposed. For the FCCU/WGS, compliance with new reduced SO2 and NOx emissions limits must be demonstrated on 7- and 365-day rolling averages. Hourly emissions limits of PM and carbon monoxide (CO) must also be met with. For all regulated pollutants, hourly emissions of startup, shutdown, and malfunction (SSM) periods must be determined and documented. For the TGCUs, 12-hour rolling SO2 limits must be met in addition to continuously determining 24-hour rolling total SO2 emissions to ensure the emissions do not exceed the reportable quantity. At the BRU, the facility must demonstrate compliance with the 3-hour rolling limit on H2S emissions.
In order to demonstrate compliance with the limits mentioned above, CEMS were installed in support of each of the control devices. Besides measuring concentrations of pollutants and oxygen, status flags from the analyzers are used to determine the quality of the data and whether the analyzer is in calibration. CEMS data are stored in the plant process data historian along with other process variables (flow rates, temperatures, etc.) required to perform calculations necessary to determine compliance with all requirements. The CEMS data quality must comply with the standards specified under the New Source Performance Standards (NSPS).
The refinery selected T3 to assist in implementing an EMIS that would streamline the facility’s rigorous compliance requirements. The implementation proceeded according to a well-documented system requirements analysis and design specification. The goal of the EMIS project was to collect, process, and calculate results from the stream of continuously monitored data and document results to meet the requirements of the Consent Decree. To meet these objectives, the EMIS was designed to perform three distinct functions:
- Data averaging and validation
- Emissions calculations
- Reporting/data retention
Data Averaging and Validation
Data from the process historian are collected on a minute-by-minute frequency. These data are assessed for quality and other factors to determine whether to include the value for averaging, including analysis of up to six analyzer status values on each minute. Per NSPS performance standards, a minimum of one data value per 15-minute block is required to constitute a good hour’s data from the analyzer. After assessing each minute value for a 15-minute block and excluding any values that do not meet the data quality requirements, the resulting record set is averaged into a 15-minute block. The process is repeated for the rest of the hour. Assuming that four such 15-minute blocks are available for the hour, the average of those 15-minute blocks constitutes the hourly average. In the event that a 15-minute block does not meet the minimum data requirements, additional logic was developed to automatically substitute conservative values from previous “good” hours.
For this particular refinery, approximately 100 variables are collected, validated, and averaged for every minute of operations. On an annualized basis, this represents over 52 million records analyzed by the EMIS.
Beyond the determination of hourly validated averages for each of the required process variables, more than 200 equations must be executed to meet all new compliance obligations at the refinery. These calculations span a wide variety of equation types, such as correcting concentrations read from the analyzer for excess oxygen, correcting flue gas volumetric flow rate to dry basis and standard temperature, converting concentrations into mass-based emissions for the hour, and other such process-specific calculations. Additionally, the calculations include “IF … THEN … ELSE” conditional logic to allow for operating conditions to determine the path and result of certain equations.
One of the finer points of the application is that the equation inputs may consist of either hourly averaged process variables discussed previously or intermediate results of other calculations (e.g., calculating a stack flow rate from continuously monitored process data and then using the resulting stack flow rate to determine the mass emission rate). Further, these calculations must be manageable by the refinery engineers as they enhance and improve their knowledge of the new systems. As a result, the history of calculation methodologies must be saved and retained for audit purposes. Another requirement of the calculation engine is to allow for “data catch-up” and “data substitution.” Since compliance obligations are on a continuous hourly basis and the EMIS itself is continuously operating, any interruption in data availability could result in a lag between current time and the results calculated in the system. Therefore, data catch-up is necessary to minimize the duration of these gaps and ensure the EMIS can return to normal operations shortly after any downtime or periods of re-calculation. Additionally, during any analyzer failures, results pulled from the process historian and the dependent calculations would likely be an inaccurate representation of actual operations. In these circumstances, refinery engineering personnel must be able to import substituted, corrected results into the EMIS for compliance determination and documentation purposes. Finally, additional logic was required to process multi-hour outputs, such as the rolling hour averages, rolling 24-hour totals, and rolling multi-day averages required to demonstrate compliance.
To meet the dynamic and complex calculation requirements of the refinery, T3 developed a “calculation engine” and interface. All of the required functionality was delivered as part of this engine and interface; however, the refinery engineers also have the option of modifying and updating calculations within the system as needed to address operational changes. Furthermore, the refinery can add emission sources and build whole new data collection points and calculations to manage their compliance requirements. In this way, the EMIS can be maintained and extended beyond the immediate project scope.
Reporting and Data Retention
Due to the high volume of data analyzed, validated, and calculated within the EMIS, a limit was needed on how much data would be available and maintained. The project team determined that the minute and 15-minute data sets were duplicative of data readily available in the facility’s plant data historian. Therefore, only the hourly averages and calculation results are stored and maintained in the system. In addition to data retained within the EMIS, the hourly values calculated within the system are written back to the process data historian each hour, providing an avenue for communicating these results to unit operations and process engineers.
Within the application, several reports were developed to assist with managing and reviewing compliance status. Examples of these reports include:
- Excess emission and continuous monitoring system downtime reports
- Exceedance and emission limit violation details on both hourly and rolling-period bases
- Emissions during periods of startup, shutdown, and malfunction
- Details on reportable quantity events and duration
Each of the reports was customized to meet the refinery’s specified output and provided within the EMIS interface so that plant personnel can generate these reports as needed. All exceedances of corresponding regulatory limits included within these reports are also identified and sent to selected personnel via email to trigger additional investigation and follow-up within the system. Additional administrative and support outputs were created related to CEMS data quality, audit history, and ad-hoc data querying capability.
Complying with a Consent Decree can be challenging. Besides the control device equipment requirements and lower emission standard obligations, the refinery was faced with onerous data gathering, validation, calculations tracking, and reporting demands. Further, the challenge of managing this high volume of data on a near real-time basis exacerbated the need for advanced processing applications. By efficiently implementing an EMIS solution, T3 provided valuable assistance with meeting requirements and automating the ongoing efforts associated with these compliance demonstration challenges.