Evaluation of potential markers of bleed air contamination

Abstract

Gas turbine engine oil and other fluids can contaminate the bleed air of an aircraft and, consequently, infiltrate cabin air. This thesis examines potential markers of bleed air contamination that can be used to detect the presence of these fluids in bleed air. Additionally, common instrumentation devices and techniques that are used to measure and identify background levels, contamination rates, particles, and chemicals in bleed air and cabin air quality on an aircraft are briefly discussed. Next, the thesis discusses the current ASHRAE 1830-RP project, which focuses on engine oil, hydraulic fluid, and deicing fluid contamination of bleed air and the resulting particulates that result as markers of this contamination. Furthermore, two major studies known as VIPR and Cranfield are addressed in this thesis and specifically deal with bleed air and cabin air on aircraft for comparison purposes. The VIPR project provides detailed information on substances, chemicals, and particulates found in bleed air from engine oil, but it is somewhat limited in scope. While a sizeable number of substances in the bleed air were evaluated the results apply to a limited range of operating conditions and a single contaminant (Mobil Jet Oil II). The Cranfield project was identified as a relevant study. It was somewhat narrow in measurement scope, but it was extensive in that it conducted measurements on 100 flights. Thus, it adds useful information and is included in this thesis. Both VIPR and Cranfield have relatively useful particulate data, and therefore will be compared to the ASHRAE 1830-RP project data. Based on the results from VIPR and ASHRAE 1830-RP, it can be inferred that there indeed can be uses of particulates as a marker for bleed air contamination, where markers are able to be measured and reliably associated with the contaminant. The results indicate that ultrafine particles (<100nm) tend to show large responses for engine oils. While ultrafine particles remain a prime candidate for detection of oil contamination, sensing needs to be upstream of the cabin in the supply air stream to ensure particles are detected from bleed air and not from other sources for future studies. Fine particles (0.3-2μm) show clear response for engine oils. Additionally, formaldehyde and acetaldehyde may be promising markers for bleed air contamination and their location in the cabin has not been ruled out. Also, carbon monoxide (CO) is likely not a good marker for bleed air contamination, but total volatile organic compounds (TVOCs) are still possible indicators of oil contamination and both markers probably cannot be used to detect oil contamination by measurements of the cabin air. Carbon dioxide (CO2) appears to be a poor marker for bleed air contamination but is a useful discriminator on determining whether particles arise from bleed air contamination or exhaust ingestion. Fine particles appear to be a good marker for hydraulic fluid. Ultrafine particles are however probably not a reliable marker for hydraulic fluid. Lastly, particles of any size are very poor markers for deicing fluid based on the results from ASHRAE 1830- RP.