time the partitioning tracer will spend in the subsurface. Tracers should be chosen to provide

adequate retardation, but not so that the test is unreasonably long. The recommended retardation

factor is between 1.2 and 4 [Jin et al., 1995].

 The mean arrival times are calculated from the method of moments [Jin et al., 1995;

Annable et al., 1998; Jawitz et al., 2003; Jawitz, 2004]. The injection pulse is maintained at

constant concentration, and the MLS or RWs are sampled frequently. The resultant break

through curves (BTCs) are then be plotted and analyzed by the method of moments. The Nth

absolute moment (MN) of a distribution is described by:


 MN = ftC(t)dt (1-6)
 0

The first normalized moment (jp) is determined by dividing MI by Mo.


 f tC(t)dt
 = = o (1-7)
 f C(t)dt
 0

One half of the tracer pulse duration (to) is subtracted from the first normalized moment to get

the ith tracer mean arrival time, z,.


 r, to (1-8)
 2

The average NAPL saturation (SN) in the swept volume is then calculated from the difference in

the mean arrival times of the non-partitioning and partitioning tracers.


 SN = t'(KPW )] (1-9)


The effective pore volume (Ve) for each RW is determined from the mean arrival time of the

non-partitioning tracer and the RW extraction rate, Q,.