By: Richard J. Wozmak, P.E., P.H., LSP, LEP

A large number of disposal sites in New England contain Light Non-Aqueous Phase Liquids (LNAPL) such as oil and gasoline. The presence of LNAPL adds complexity in site characterization as well as developing remedial alternatives to achieve specific cleanup goals. Because these liquids have densities less than water, they will float on the capillary fringe and the water table as a separate, immiscible phase (if a large enough release occurs) and migrate in the general direction of ground water flow. Also, water table fluctuations caused by seasonal variations in precipitation-evapotranspiration and tidal influences will affect LNAPL thickness in a well, often increasing with declining water table elevations.

Estimating the thickness of LNAPL in site characterization studies is critical because many regulations contain thickness thresholds for release notification and for response action triggers. For example, the Massachusetts Department of Environmental Protection (MADEP) has notification and risk thresholds for LNAPL. For LNAPL observed in a monitoring well between one-eighth and one-half inch in thickness, a 120-day reporting condition exists, and for NAPL observed in a monitoring well greater than one-half inch in thickness, a 72-hour reporting condition exists. The MADEP has also established an Upper Concentration Limit (UCL) of one-half inch in thickness within a formation for characterizing future risk of harm to public welfare and the environment.

In addition, understanding the volume and distribution of LNAPL is necessary to evaluate appropriate active or passive remedial alternatives, as well as to estimate the time to achieve a remedial standard. There are predictive hydrogeologic models now available that incorporate LNAPL dissolution coupled with dissolved-phase contaminant transport designed specifically to estimate times of remediation. Important parameters required to successfully apply these models include the volume and distribution of LNAPL in the subsurface formation.

Coated measuring tapes, oil/water interface probes, and clear bailers can be used to measure LNAPL thickness in monitoring wells. However, the thickness of LNAPL in a monitoring well typically exceeds the thickness of LNAPL in the subsurface formation by a factor estimated to range between 2 and 10. Due to this difference, the LNAPL thickness measured in a monitoring well is commonly referred to as the “apparent thickness” and is not an accurate measurement of the LNAPL thickness in the subsurface formation.

The difference in LNAPL thickness is primarily caused by LNAPL floating on the capillary fringe (water held above the water table in soil pore spaces), which is not present in the monitoring well, and the weight of the LNAPL (see figure). The absence of the capillary fringe in the monitoring well causes the well to act as a low point into which LNAPL will drain. When LNAPL accumulates in the well, the weight of the LNAPL will depress the water table in the well resulting in additional LNAPL drainage into the well. The difference in thickness between the formation and the well will be greater for heavier oils such as No. 4 or 6 oils, due to the higher densities of these oils further depressing the water table in the well. In addition, the difference in LNAPL thickness increases with decreasing formation grain size due to the presence of a higher capillary fringe in the smaller grain size formation. Therefore, silty soils will produce a greater thickness difference than coarse-grained sand.

Many studies have been performed to correlate LNAPL thickness in a monitoring well to actual LNAPL thickness in the subsurface formation.  These studies have produced empirical relationships that correlate actual LNAPL thickness in a formation to apparent LNAPL thickness measured in a well. These methods typically involve relationships that require an understanding of hydrogeologic properties such as the height of the capillary fringe and the physical properties of the LNAPL fluid. However, these correlations may not be accurate under a variety of field conditions and typically produce only order of magnitude estimates.

Other methods for estimating LNAPL include boring programs and the use of a cone penetrometer. Multiple borings with continuous sampling can be used to define the thickness of LNAPL from visual observations. The cone penetrometer is a geotechnical tool that is capable of rapidly producing information regarding subsurface stratigraphy and LNAPL distribution. A cone penetrometer typically consists of a truck-mounted hydraulic ram that pushes a cone-shaped instrument through the soil. A signal is sent back to the drilling rig that is translated into thickness of various materials. Some cone penetrometers are now equipped with devices to sample ground water, LNAPL, and soil during cone advancement. However, this method is more expensive than conventional drilling methods and, also, does not accurately distinguish between mobile (free phase) and immobile (residual phase) LNAPL trapped in the soil pore space.

The method selected to estimate LNAPL thickness is generally dependent upon the degree of certainty needed to characterize LNAPL distribution. The simplest approach is to assume that the observed thickness in a monitoring well is the actual thickness in the formation. This approach is required for many regulations, including the MADEP notification requirements for LNAPL. However, use of this approach is conservative and will overestimate the actual LNAPL thickness in the formation.  Order-of-magnitude estimates obtained from empirical relationships that correlate LNAPL thickness in monitoring wells to the actual thickness in the formation may provide a better estimate of the true LNAPL thickness and are generally acceptable for evaluating remedial approaches and for use in models designed to estimate times of remediation to achieve specific remedial goals.

Another approach is to sequentially apply more aggressive methods as needed to characterize LNAPL distribution and thickness. This approach is particularly useful when demonstrating compliance with a regulatory standard.  The MADEP has suggested this approach to determine compliance with the LNAPL UCL thickness of one-half inch. They suggest first collecting LNAPL thickness data from wells within a contiguous LNAPL plume to evaluate LNAPL thicknesses and conservatively compare the largest thickness results (the average of the thickest value measured in each well within a contiguous LNAPL plume) to the standard.  If the average does not exceed the UCL standard, compliance has been achieved. If the result is greater than or equal to the standard then additional characterization or calculations such as the methods presented above can be performed to evaluate regulatory compliance.

In summary, the distribution and behavior of LNAPL in the subsurface is complex. There are a number of different methods that can be used to characterize LNAPL distribution and migration that vary in complexity, accuracy, and cost. Selection of the appropriate method(s) depends primarily on the degree of certainty needed in characterizing LNAPL thickness. Other factors to consider include regulatory acceptance of the selected methods, and the cost of the method versus the potential benefit obtained from improved accuracy in the result.

Published in “Environmental Law and Policy” a periodic publication of Morrison, Mahoney & Miller, LLP Environmental Practice Group, July 2003