Note: FTIR data comes to us from the refinery in the form of (raw) "Single-Beam" Spectra.
Note: "Absorbance" Spectra are the spectra which we use for analysis.
Note: Each Absorbance Spectrum is derived by taking the negative log of one Single-Beam Spectrum to another (background reference) Single-Beam Spectrum.
Our (preliminary) analysis, as well as Tosco's databases, indicated that the peak level of MTBE on the South fenceline occurred just after midnight on the morning of Nov. 4. The first step in our process was to identify the spectral file which represented the peak level of MTBE during the month, and to convert this file for use as a (temporary) background reference. Absorbance files created using this "hot" background reference were then analyzed for MTBE content. (When these absorbance files are analyzed, periods of low MTBE concentration presented themselves in the form negative concentration figures.) This step was taken in an attempt to identify the periods (when MTBE concentrations were at their lowest) where we might find a Single-Beam Spectrum suitable for final use as a "clean-air" background reference spectrum for the November data set.
In the data set generated by analyzing these "hot" absorbance files it became apparent that there were several periods where MTBE levels were seen to dip (to the negative 59 - 60ppb range) but that these periods came infrequently and typically lasted only for a very short time. Because of this our "candidate pool" of potential background spectra was relatively limited. Several of the more promising candidate spectra were re-named and saved for use as background reference files. Several sets of absorbance spectra were then created using these various background candidates. The various absorbance spectra were analyzed for MTBE content, and results were compared. Through a process of elimination, Single-Beam Spectrum # Tos00538, generated on 11/6 at 9:05am, was ultimately selected for use as the background file.
While this new background reference (Tos00538) worked well for the purpose of identifying periods of elevated MTBE concentrations, it was found to be less effective for the purpose of identifying and measuring the various compounds being monitored in the "C-H Stretch" region of the spectrum. Once again, we started the background selection process described above. This time, instead of looking at MTBE concentrations, we attempted to identify those periods when methyl pentane concentrations were at their minimum. Eventually, we settled on spectral file # Tos00727, from 11/13 at 10:01am. Absorbance files created from this second background were used for the measurement of all of target compounds except MTBE which are reported in our data summary for the South fenceline in November.