Testing the In‐Well Horizontal Laminar Flow Assumption with a Sand‐Tank Well Model

Abstract

The assumption of horizontal laminar flow within a monitoring well is a commonly cited basis for interval sampling using low‐flow or no‐purge sampling techniques. A few studies have shown horizontal flow over short distances within the well for short periods of time. Others have demonstrated specific circumstances under which the assumption fails. But surprisingly, little focus has been given to confirming the underlying concept—that under “normal” conditions (i.e., no vertical hydraulic gradient) water enters one side of a well and exits the other side of the well at the same elevation. To test the horizontal flow assumption, a physical sand‐tank model was constructed to observe flow through in a simulated monitoring well. The well, filter pack, and aquifer largely mimic real world conditions of a submerged well in a moderately high‐permeability sand. To observe flow behavior in the simulated well, a dye “stringer” was introduced into an injection port upgradient of the simulated well. In all tests, regardless of flow rate or small density differences, the dye stringer eventually mixed throughout the model monitoring well. Since the model approximates a section of an at‐scale well subjected to real world bulk flow rates, mixing appears to be the rule rather than the exception for near–neutrally buoyant contaminant stringers in homogeneous flow fields. Despite additional heterogeneities introduced by field conditions, there are several clear and important implications of this study: (1) some degree of in‐well mixing and flow‐weighted concentration averaging may occur in a well before any purge or sampling efforts are made; (2) in‐well mixing may mask low to moderate contaminant stratification in an aquifer; (3) contaminant stratification, if present inside a well, implies strong contaminant stratification outside the well; (4) contaminant stratification inside a well may not correspond to stratification at the same interval outside the well; and (5) vertical stratification within an aquifer may not be accurately monitored by sampling multiple intervals within an open well.

Effects of groundwater withdrawal on borehole flow and salinity measured in deep monitor wells in Hawai'i: implications for groundwater management

“…In long-screened wells in a sandy aquifer in France, sampling for groundwater age at discrete depths was biased by ambient vertical flow (Corcho Alvarado and others, 2009). A sand-tank model study indicated that a long-screened well can create an artifact even in the absence of vertical ambient flow and that water-quality stratification in a borehole may not correspond to stratification in the aquifer (Britt, 2005). Sukop (2000) used a simple solute mass-balance and water-balance method to compute the vertical distribution of solute concentration in the aquifer using water-quality information from a long-screened monitor well.…”

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

“…In this objective we expanded on the work done by (Britt 2005) through a combination of 2D and 3D experiments that:…”

Evaluation of the Snap Sampler for Sampling Ground Water Monitoring Wells for VOCs and Explosives

“…Convective circulation within wells also results when there is a density gradient in the well water. A density gradient can result from differences in the concentrations of constituents in the water or from differences in the temperature of the water within the well (Diment 1967, Gretner 1967, Samuel 1968, Britt 2005, Vroblesky et al 2006. In instances where there is convective flow, oxygen can be transported from shallower parts of the well to deeper parts of the well, including the screened portion of the well (Vroblesky et al 2006).…”