Plume Distortion and Apparent Attenuation Due to Concentration Averaging in Monitoring Wells

Abstract

Mathematical models that simulate common monitoring well sampling demonstrate the distortions that vertical concentration averaging causes during the mapping and modeling of an idealized, three‐dimensional contaminant plume emanating from a simple source of constant solute concentration. The apparent extent of the plume, mapped using simulations of a regular grid of screened monitoring wells, ranged from a worst case of 0% of the original plume area for long screens (4 m) in a low‐permeability formation to 90% for short screens (1 m) in a high‐permeability formation. When well design and purging procedure were inconsistent among wells, the mapped plume exhibited spurious directional skewing, bifurcation, zones of low concentration, intermittent sources, or multiple sources. Although the study plume was not retarded, calibrating a transport model to the monitoring well data resulted in retardation factors of up to 23. If first‐order decay was assumed, the apparent decay constant was found to be as much as 1.8 × 10‐7 sec‐1 (T1/2= 45 days). Apparent retardation or decay was inconsistent from well to well, depending on the saturated screen length, the degree of screen desaturation during purging, and the distance from the source. The study indicates that the quantitative assessment of contaminant distributions and transport processes requires discrete vertical sampling in the common situation where concentrations vary sharply with depth, even in the most ideal hydrogeologic environment. If screened monitoring wells are used, screen lengths and placements should be appropriate to the contamination situation being assessed and inherent biases must be considered. Even so, vertical concentration averaging biases and the resulting inconsistencies can result in highly misleading evaluations of ground‐water contamination problems.

Analytical Solutions of Three‐Dimensional Contaminant Transport Models with Exponential Source Decay

“…This approximated solution can be easily derived by following the approach proposed in Domenico and Robbins (1985) and Martin-Hayden and Robbins (1997). Alternatively, the proposed solution can be obtained from the instantaneous pulse solution, by following the three steps defined in rules (A), (B), and (C) and by making some simplifying hypothesis.…”

“…The numerical solutions can introduce approximation errors, can be computationally demanding, and may explain why models using the same input parameters can generate very different results. Domenico and Robbins (1985), Domenico (1987), and Martin-Hayden and Robbins (1997) proposed some 3D approximated analytical solutions in closed form. The key advantage of the Domenico and Robbins (1985) approach and its derived extensions included in U.S. EPA tools as BIOSCREEN (Newell et al 1996), BIOCHLOR (Aziz et al 2000), currently used in risk assessment procedures, is that it provides a closedform solution that can be evaluated with much less significant numerical computations because it contains only complementary error functions.…”

An Assessment of Aquifer/Well Flow Dynamics: Identification of Parameters Key to Passive Sampling and Application of Downhole Sensor Technologies

“…Many studies have identified how anomalous or otherwise unrepresentative results may be generated from traditional purge and sampling techniques (e.g. Robbins, 1989, Gibs andImbrigiotta, 1990;Reilly and Gibs, 1993;Conant et al, 1995;Church and Granato, 1996, Martin-Hayden and Robbins, 1997, Reilly and LeBlanc, 1998, Hutchins and Acree, 2000, Elci, et al, 2001. Common situations shown by these and other studies have caused investigators to question what traditional purge-and-sample ground water monitoring results represent.…”

“…Other sources of misdirection from sampling data include partial penetration (Robbins, 1989;Martin-Hayden and Robbins, 1997;and Chaing et al, 1995) and lithologic heterogeneity . Anderson and others (1992) showed how DNAPL fingers and pools result in very strong dissolved contaminant stratification in an aquifer.…”

Measuring Salinity within Shallow Piezometers:Comparison of Two Field Methods

“…There are different approaches to study saltwater intrusion in coastal aquifers, for example geophysical methods [3], borehole investigations, and numerical models [4]. Borehole data are widely used in these kinds of studies, especially for calibrating numerical models, but in recent years the validity of data collected directly from wells and piezometers is under discussion [5][6][7][8][9]. Characterizing piezometric heads and salinity within a large coastal phreatic aquifer is a complex endeavour.…”