MLS results, these were not included because they occurred in MLS-5 and MLS-7. In MLS-5 no

FNR values were able to be calculated due to insufficient tracer arrival, and at MLS-7 there were

only a few results that were based on estimated NAPL removal (ENR) described in Section 2.11.

There was a single recorded NAPL increase in MLS-7, but this value is off the scale of Figure 4-

1 as well.

 According to the previous work exploring this relationship, the MLS data distribution on

the graph indicates a high level of heterogeneity in PCE distribution. Overall, this data

corroborates the modeling and other field tests that have demonstrated aggressive mass

removal's favorable impact on mass flux.

 Because all MLS and RWs did not experience equal remedial velocity or flow rate, each

data point was weighted by the flow rate of remedial fluids sweeping through its region. Each

MLS depth FFR value was weighted by the fraction of the total estimated fluid velocity for the

respective MLS well in Figure 2-2. The weighted MLS velocity FFR for each depth was

summed to produce a mean weighted FFR for each MLS well. The RWs were weighted by the

fraction of the total extraction flow rate for the site. The flow rate weighted FFR for the RWs

were summed to provide a mean flow weighted FFR for all RWs. The fractional NAPL removal

was then compared to the resultant weighted mean FFR values and is presented in Figure 4-2.

 The MLS wells displayed highly favorable velocity-weighted flux reductions to their

respective PCE removals. Furthermore, the effect of the heterogeneous DNAPL architecture is

evident in the high flux to medium NAPL removal. The recovery well flow-weighted average

displayed less heterogeneous NAPL morphology and less favorable flux reduction to NAPL

removal, due to the integration of liquids arriving at the well.