Transport of Escherichia coli in 25 m columns - Transport of Multiple Escherichia Coli Transport in Saturated Porous Media - page 67

5.3.2 Power-law distribution in sticking efficiencies and determination of minimum sticking
efficiencies

We have previously shown that a power-law best described the distribution of bacteria cell
affinity for quartz grain surfaces (Lutterodt et al., 2009a). Therefore, the power-law distribution
was applied to assess the relation between F and their corresponding Α by fitting. The
coefficient of determination ( R 2 ) was used to evaluate the goodness of fit. As indicated in the
Introduction section of this paper, the minimum sticking efficiency ( mi Α ) was defined as the
sticking efficiency belonging to a bacteria fraction of 0.001% of initial bacteria mass flowing
into a column, after removal of 99.999% (5 log reduction) of the original bacteria mass has taken
place. However, within this minor fraction of bacteria cells, the sticking efficiencies are not
constant, but distributed, and within this 0.001% sub-fraction, the minimum sticking efficiency is
the highest possible sticking efficiency. The minimum sticking efficiency was extrapolated from
a power-law equation, given as:

F

= A Α (5.5)

min

where A and Β are constants obtained from fitting the experimental data.

5.4 Results

5.4.1 Porosity determination

Porosities determined for 1 m column slices ranged from 0.35 to 0.41 with a mean of 0.37, and a
standard deviation of 0.016. Porosities measured for all slices within a segment were averaged to
obtain the porosity of each segment. As a result, each column segment had a porosity of 0.37
with the exception of the second segment (6.0 to 12.15 m; porosity = 0.38). From the low
standard deviation of the mean, we concluded that the column was uniformly packed.

5.4.2 Tracer and bacteria breakthrough

Breakthrough curves ( Figs. 5.2 and 5.3 ) obtained at the sampling ports show systematic
reductions in peak relative concentrations with distance. Tests for asymmetry yielded low
positive (0.84 at 6 m and 0.26 at 12.15 m) and low negative skewness (-0.83 at 19 m and -0.31 at
25.65 m) for the first two sampling ports and the two most distant sampling ports, respectively.
Percentage mass recovery of tracer was comparatively higher at the first two sampling ports
(90% for both distances) than at the two most distant sampling ports (75% for 19 m and 70 % for
25.65 m). R 2 values obtained by fitting curves with second order polynomial were good (0.96 for
19 m and 0.98 for 25.65 m), computed tracer recovery for extrapolated curves were 77% and 90
% at 19 m and 25.65 m, respectively.