Natural Attenuation - Dealing with Contaminated Sites: From Theory Towards Practical Application - page 994

• Emission is almost independent of the LNAPL mass, but depends on the source

zone extension. Thus a reduced LNAPL saturation, e.g. as a result of phase
extraction, will not in the short term lead to significantly reduced mass flow
rates out of the source zone, neither to reduced concentrations in the plume nor
a reduction in plume length. This is because saturation concentrations of a spe-
cific contaminant in the groundwater as a result of NAPL dissolution depend
on the compound's molar fraction and its solubility, and not on NAPL satura-
tion. If, however a source zone remediation measure achieves not only a reduced
LNAPL saturation, but also a reduction in source zone extension (e.g. partial
source removal by Dig and Dump), then a reduced mass flow rate out of the
remaining LNAPL will occur. In any case, whether the partial source removal
leads only to decreased NAPL saturation or indeed to a reduced source zone
extent, a reduced LNAPL mass translates directly into a shortened source lifetime
and consequently also shortened plume lifetime.

With respect to the contaminant plume in the groundwater , the following general
conclusions can be drawn:

• Generally, petroleum hydrocarbons are readily degradable, i.e. sites contaminated

by petroleum hydrocarbons generally may come into consideration for MNA.
However, also less degradable contaminants such as MTBE or trimethylbenzenes,
forming long plumes and possibly hazardous metabolites, might occur within
petroleum hydrocarbon mixtures, posing a challenge for MNA at those sites.

• For plume delineation, a high number of data with some uncertainty is better

than a small number of highly precise data, improving interpolation or calibration
results. Thus, low-cost methods to investigate the plume, such as Direct-Push
methods, are recommended for a first detailed plume investigation.

• Although plume stationarity is one of the essential prerequisites for accepting

and implementing MNA, this stationarity is hard to prove. Meta studies have
shown that steady state plumes seem to be prevailing at about one third of
271 investigated petroleum hydrocarbon contaminated sites (Rice et al. 1995 ).
However, most processes that influence plume development seem to be transient,
e.g., source emission (see above), hydraulic conditions (e.g., changing ground-
water flow velocities and/or directions, groundwater recharge), biodegradation
due to changing geochemical conditions (e.g., due to depletion of immobile
electron acceptors like Fe(III) or Mn(IV)), sorption due to the kinetic nature of
sorption processes and/or due to backfilling of sorption places and potentially
subsequent desorption. Transience of these processes occurs at time scales that
might be shorter than the monitoring frequency (e.g., changing hydraulic condi-
tions) or longer than the monitoring frequency (e.g., changing source emission,
biodegradation, sorption). Thus, it can be expected that a transient plume devel-
opment might not be detected due to too long monitoring frequencies or due
to too short monitoring periods relative to the expected time scale of process
instationarity.