Experiences in the Installation of Hydrocarbon Monitoring Systems for Flares and Cooling Towers

摘要

As a part of the Texas State Implementation Plan (SIP) for the Ozone Non-attainment Area including the Houston-Galveston Region, the Texas Commission on Environmental Quality (TCEQ) promulgated new regulations that require the continuous measurement of ethylene, propylene, 1,3-butadiene and isomers of butane [highly reactive volatile organic compounds (HRVOC)], from flares and cooling towers. While analytical methods exist for measuring these compounds in various process applications, they typically do not involve applications where the concentration ranges between the ppm levels up through significant weight percent levels as might be found in a flare sample. Likewise, the measurement of these compounds in a cooling water system presents its own unique problems. The regulation requires use of the "El Paso Stripper" type air stripping device or to demonstrate by actual testing an equivalent technology. When these requirements are coupled with a deadline of having such a system operational by the end of December 2005, the challenges can be significant. This paper describes some of the experiences with the installation of these systems at the Lyondell Chemical Company, Channelview Chemical Complex. This facility is installing six flare monitoring systems and eight cooling water monitoring systems to meet these requirements and adding six new analyzer shelters with all of the associated infrastructure to support them. Since the required sampling locations in most cases are not close enough to be able to use existing process analyzer shelters, the logistics of the installations were a significant part of the project. Overall the Lyondell facilities are installing 43 monitoring systems in 33 shelters located at 10 different facilities in the Houston-Galveston area. The methodology associated with these systems can best be described as equivalent to compliance stack sampling with the added requirements that it be done continuously (i.e., have a 95+ percent on-stream time and an analyzer cycle time of not more than 15 minutes). Sample location issues created challenges not envisioned when the rules were developed. For example, flares located hundreds of feet from adjacent operations and underground cooling water lines.