Ref. No. [UMCES]CBL 2013-015
ACT VS12-02
The time series response of the UviLux-CDOM and UviLux-HC fluorometers to the
stepwise addition of crude oil and dispersant are plotted in figures 6 and 7, respectively. Each
lettered panel represents a day-long test of specific source oil and dispersant ratio at seven
different concentrations including ambient background. The highest concentration was not
tested on day 1, but this whole experiment was repeated on day 3 during which the highest
concentration level was included. The background fluorescence of the incoming seawater varied
with each trial, averaging between 400 – 800 mV for both sensors. The elevated ambient
background fluorescence is also evident in the non-zero EEM
QSE
values at the start of each trial
series (Fig. 10 & 11, panel B). Representative EEM maps from reference samples collected after
the fourth oil addition (mass added ca. 85 grams; concentration ca. 3 ppm) are presented in
figures 8 and 9. The optical windows used for estimating the integrated fluorescent intensities of
the UviLux-CDOM and UviLux-HC units are overlaid on the maps in figure 8 and 9,
respectively. The overlap of the optical window of the UV-HC unit to the region of maximum
fluorescence of the oil mixtures was clearly greater than that for the UV-CDOM unit.
The response of the UviLux CDOM unit to the oil additions was minimal but linear only
in trials conducted with the chemical dispersant Corexit 9500 (Fig 10, panel A). In all cases the
background reading in seawater was not subtracted from the instrument response during oil
additions. The response of the UviLux-HC unit was significantly greater (Fig 11 panel A) than
the UviLux-CDOM unit for all oil exposures, and this response difference was greatest for the
three tests in which chemical dispersant was used in addition to physical wave dispersion..
Chemical dispersion resulted in an approximate doubling of the fluorescence detected by the
UviLux instrument. The UviLux-CDOM also reported a similar relative signal enhancement but
greatly dampened absolute signal response.
Cross plots of instrument response versus oil concentration and estimated EEM
QSE
intensity clearly reveal differences in the detection capabilities of these two UViLux
configurations (Fig 10, CDOM unit and Fig. 11, HC unit). The response of the CDOM unit to
increased oil concentration was much dampened and the overall range of EEM intensity within
the optical window was greatly compressed (Fig. 10). The response of the UviLux-HC unit was
much greater but was highly dependent on both the type of oil and most importantly to whether
or not chemical dispersant was added to the crude oil (Fig. 11). The instrument response was
roughly 3-5 times greater when the crude oil had been dispersed by the Corexit. The instrument
response of the HC unit to oil additions was linear through 3000 ppb added oil but showed a
significant response above background to even the lowest oil addition of
ca
300 ppb TPH in the
presence of Corexit 9500. There was good agreement between instrument response versus oil
concentration for independent trials with the same source crude oil (Fig. 11A; ALC, DOR 1:25).
Figure 12 summarizes various water quality parameters over the course of the five tests.
Concentrations of chlorophyll, CDOM, and turbidity were conducted on discrete reference
samples, while particle concentration estimates were generated in situ with a LISST. Although
levels of chlorophyll, CDOM and turbidity varied at the start of each day, their effect on the
initial background fluorescence of the seawater was relatively small. Changes in chlorophyll and
CDOM concentrations during the step-up oil additions were relatively small. Turbidity
increased almost linearly when dispersant was present with the oil, but showed little change to
increasing oil concentrations above 1.5 ppm without dispersant. Similarly, the increase in mean
particle concentrations was much greater in the presence of dispersant than without, indicating a
physical repacking of the oil is also taking place, which would likely account for much of the
differences in fluorescent response of the test mixtures.
20
