Geology 492/692, Applied Geophysics
Resistivity and VLF Surveys along the V-Line Canal near Fallon, Nevada
By:
George T. Lightwood
Scott Craig
Danny Lazzareschi
Resistivity Surveys
Resistivity methods have been used for evaluation of seepage through canals and levees to evaluate the potential for internal erosion of the levee leading to failure. Usually, this method needs to be augmented wit another technique to identify likely seepage zones. Panthulu, et.al, ( 2001) used the electrical resistivity method to delineate zones favorable for seepage under a dam in India, and SP measurements were made to delineate actual seepage paths C. Chen, et. al. (2006) used the shallow seismic reflection method, surface wave dispersion techniques, and resistivity sounding to identify areas of pipe formation in embankments along the Yangtze River in China. These studies focused on indentifying active seepage zones though embankments. A study by Asch, et. al, (2008), used electrical resistivity and electromagnetic surveys to detect the thickness of sand lenses that underlie portions of American River levees in Sacramento, California. The purpose of this study was identify geological conditions that where high levels of seepage and thus piping leading to failure of the levees might occur.
On March 19, 2009, Danny Lazzareschi, George Lightwood, Laura Huebner, and Scott Craig, completed an electrical survey of 13 stations by use of a mini-res along the V-Line Canal bank outside of Fallon, NV. The first survey was done on the northwestern crest of the southeastern bank. Along that survey data was collected from seventeen stations. After the first sequence of stations, the tenth station was collected 125m southeast from station nine in what we labeled the playa. The eleventh station was east of station one at the toe of the canal bank by vegetation. The twelfth and thirteenth stations were completed along the same line as the Remi Line Two days earlier. Stations are found on Figure 1.
Station one through nine were the initial readings and after data computing, it was deduced that a distance of 4.64m showed the best anomalies for the area so eight more collection points were gathered at the original stations but with only 4.64m intervals. Along the canal, different apparent resistivity was measured where vegetation and water were evident outside of the canal bank. Figures 4 through 9 show the resistivity of low and high resistive layers as obtained from modeling the apparent resistivity verses electrode spacing in a Wenner Array.
The plots of apparent resistivity versus depth made along the canal bank at stations E-1 through E-9 all are typical of 3-layer models of the K type. This indicates the situation where there are two horizontal interfaces, with a lower resistivity layer with higher resistivity layers both above and below the low resistivity layer. Sounding E-12 was modeled using a 2 layer model with a lower resistivity surface layer and a higher resistivity lower layer. All resistivity soundings were processed using a demonstration version of the computer program, Resix. Resix has an interactive user interface that allows models to be rapidly made by interactively changing the depth and resistivity of soil layers. Since the demonstration version of Resix does not allow printing the model, the results of the model were then input into the program Resist (Burger, et. al., 2006) for plotting purposes.
Many alternative thickness and resistivity models could be developed to model the apparent resistivity; so some additional information was considered to constrain the models developed for each sounding. In this case, site observations and National Conservation Soil Service (NCRS, 2009) soil data was used to make a preliminary evaluation of likely ranges of resistivity of material types. The soil survey for the area indicates that the soil on the south side of the canal is mapped as Badlands (Ba) and the north side of the canal the soil is mapped as the Hawsley Sand (Ba). The Badlands deposits consist of approximately 53 percent clay, 30 percent silt, and 6 percent sand and are classified by the Unified Soils Classification System as high plasticity silt (MH) . The Hawsley Sand is approximately 96 percent sand, and approximately 4 percent silt and clay, and it is classified by the USCS as silty sand. Measurements made on extracts by the NRCS (2009) from soil paste made from Badlands soil indicate that the electrical conductivity of this soil is 24 millimhos per centimeter indicating a saline soil. The Hawsley Sand is less saline, and the electrical conductivity is reported to be equal to approximately 0.9 millimhos per centimeter. Based on these measurements, we would expect that there would be about an order of magnitude difference or more between the resistivity of the two soil types. Typical resistivities for soils consisting of predominantly of clay minerals containing brackish water can have are on the order of 1 to 100 Ω-m. Soils consisting of sand and silt typically have a resistivity on the order of several hundred Ohm-meters (Sharma, 1997 and Ward, 1990).
Resistivity measurements can also indicate the presence of ground water, and the resistivity of fresh water is on the order of 1 to several hundred Ohm-m, and the resistivity of brackish and saline water is typically less than 1 Ohm-m (Sharma, 1997). There are no wells at the site and the quality and depth to ground water is not known. A check of USGS Water resources data (USGS, 2009) indicate that there are two shallow wells located approximately 5 km from the site. These wells show the ground water levels are approximately 2 to 5 m below ground surface. Standing water observed in the V-line canal may also be indicative of the ground water table in this area.
Engineering Data from the United States Bureau of Reclamation (USBR, 2009) indicate that the canal was constructed between 1904 and 1905 and has a bottom width of about 7m with side slopes of 2H:1V. Water depth in the canal varies, but is typically 3.7 m; however, the canal was dry when the survey was conducted. It is likely that the canal was built on native ground by excavating material in the sandy Hawsley soils to the north and depositing them on the Badlands soils to the south. The road on the surface of the canal bank appears to have been surfaced with clayey soils to an unknown depth. Typically such roads are several feet thick and become compacted with repeated grading and compaction due to wheel loads. Typical values of resistivity of compacted clay soils typically range from approximately 10 to 160 Ohm-m and decrease with increasing compaction (McCarter, 1984). A reasonable model of the canal bank might be that it consists of three soil layers with a low resistivity layer at the surface consisting of compacted clay, a higher resistivity layer consisting of fill consisting predominantly of the Hawsley Sand placed on top of the original ground surface consisting of clay Badlands soil. Areas along the canal bank where the second layer is thicker and has higher resistivity might indicate zones of sandy soil where seepage through the canal bank is likely to occur.
Measurements made at the toe of the canal were made directly on Badlands soil and indicate an H type resistivity curve (E-11) with a low conductivity soil between two soils of higher resistivity. The resistivity measured at the toe of the canal bank, is probably representative of the resistivity of the native ground beneath the fill placed to make the canal bank. The range of values in resistivity for these soils is on the order of 13 to 24 Ω-m. The shape of the apparent resistivity curve for E-11 shows the presence of a deeper higher resistivity layer. The identification of this layer cannot be determined from resistivity measurements alone, but could be due to the presence of sandy soils or evidence of a water table at a depth of greater than 6 meters.
Constraints for the models shown were based on the assumption that the lowest layer should have a resistivity similar to that of that obtained at the toe of the embankment (Sounding E-11) and be greater than one and less than about 30 to 40 Ω-m. The resistivity of the top of the embankment where compacted clay soils at the road surface were also limited to a range between 3 to 40 Ω-m. The thickness and resistivity of the middle layer of the three layer model was then adjusted, so the resistivity was at least an order of magnitude greater than the top and bottom layers. The results of the modeling efforts are shown in the curves showing the measured and modeled apparent resistivity versus depth for stations E-1 through E-9. Figure 2 is a pseudo section made by combining the models for each sounding and shows the estimated thickness of each layer along the canal bank. A rigorous error analysis was not accomplished; but based on the interactive modeling process used with resix,. the resistivity for each layer could very by as much as 20 to 50 percent and the depth of each layer could very by as much as 0.25 to 0.5 meters and still obtain reasonable models that fit the data.
Figure 4 is the resistivity for station E-1. This graph along with figure 2 shows a more resistive layer at about 1m depth and 1.5m thick. Figure 5 shows the resistivity of E-3 which is adjacent to vegetation growth. The data on figures 5 and 2 show that a higher resistive layer is about 2.25m below the surface and is about 3m thick indicates that there is more fluid saturation below station 3 which is evident by vegetation adjacent to the station. Figure 6 represents the resistivity for E-5. The data supports a more resistive layer at greater depth, about 1.5m below the surface with a thickness of about 2.3m. Figure 7 shows stations E-7’s resistivity. E-7 is at a depth of 1.25m with a thickness of about 1.5m. Figure 8’s data indicates the thickest resistive layer at station E-8 at a depth of 1.4m with a thickness of about 4m. Figure 9 shows station E-9 with a thickness of 1m and a depth from the surface of about .75m.
The data collected from stations 1-9 are representative of what was observed from the canal. Near vegetation or areas of standing water, the high resistivity layer was thicker than areas where vegetation was not visibly evident. For example, at station E-1 the high resistive layer was thinnest. At that location the vegetation was just north and to the south vegetation was 1.5 stations away. Also, the southern vegetation was the smallest amount of bushes around the area surveyed. As the survey moved southwest the vegetation increased and standing water became evident. The thickest high resistivity layer was found by station 8. From observation the vegetation and standing water begin at station 6.5 and last until someplace between 7.5 and 8. The high resistive layer thins out after station 8 to station 9. At station 9 no vegetation was observed adjacent to the station. In general, then vegetation and standing water was observed in the field, a high resistive layer thickened indicating sandy soils where seepage through the canal bank might occur.
The first layer is a low resistive layer with ranging from 3 to 38 Ω-m. Layer two is a highly resistive layer with a resistivity ranging from 200 to 513 Ω-m. Layer three is another low resistivity layer with resistivity ranging from 6 to 24 Ω-m. Layer one is the shallowest and three is the deepest from the surface.
Station E-10 was 125m southeast of E-9 in the playa (figure 1 for location and 10 for resistivity data). There, three layers were found with the first one being a low resistive layer of 92.8 Ω-m at a depth/thickness of 1.2m. The second layer was a high resistive layer with a resistivity of 791.1 Ω-m and a thickness of 2.02m. The third layer, low resistive layer, had resistivity of 10.7 Ω-m with it thickness extending to infinity. The playa compared to stations 109 indicate that the high resistive layer is fairly thin, but if much more resistive. The increase in resistivity could be due to more saturation in the playa.
Station E-11 was collected east of the vegetation that is north of E-1, figure 1 (resistivity data on figure 11). At this station the first layer was moderate resistivity of 48.8 Ω-m and a thickness of .52m. The second layer had a lower resistivity of 12.2 Ω-m with a thickness of 6.44m. The third layer had the highest resistivity of 126.3 Ω-m at a thickness of infinity. This station has the opposite layering that what was observed from prior station and is of the H type. The high resistivity surface layer may be a thin layer of dry silt at the surface and the second lower layer may be reflective the highly saline Badland clay soils upon which the canal bank was constructed. The data also indicates the presence of a higher resistivity layer below about 7 meters and this may indicate water saturated sandier soils at depth.
Station E-12 through E-13 covered the Remi line on the opposite side of the canal that stations 1-9 were surveyed. The coverage of the Remi line indicated that the first layer is at 1.1m depth, second layer at 2.6m depth and the third goes to infinity. The data also shows that layer on had a resistivity of 18.5 Ω-m, layer two 6.0 Ω-m and layer three 66.8 Ω-m. Figure 18 shows the depths of the layers in relation to the canal bank for E-13.
VLF-EMR Survey
On March 18, 2009, Danny Lazzareschi, Scott Craig, George Lightwood, Laura Huebner, and Betsy Littlefield performed a VLF_EMR survey along the canal bank. A Geonics EM16R was used to make the measurements. Signals transmitted from the VLF station located at Jim Creek, Washington provided the source of the VLF signals. Measurements were made along the canal bank every 16 meters corresponding approximately from Station E-1 to E-9 shown on Figure 1. The apparent resistivity measured ranged from 8 to 11 Ohm-m and the measured phase angles ranged from 34 to 38 degrees. An attempt was made to develop a 2-layer model from the data (Geonics Limited). This technique requires that the resistivity be assumed for one of the layers. In this case, the resistivity of the top layer was assumed to be on the order of 10 to 30 Ohm-m, which corresponds to the Wenner Array resistivity measurements made along the canal bank at an electrode spacing of 1 meter. Using the phasor diagrams (Geonics Limited) to determine the resistivity of a second layer was not successful; that is, the depth of the interface and the resistivity of the lower layer could not be determined using this technique. This is due to the fact that the method is poorly suited to defining the boundary between a highly conductive layer overlying a layer of higher resistivity (Podder and Rathor, 1983), and the measurements are at the lower resistivity range of the instrument.