Introduction

In 1994 Vector Engineering, Inc. was hired to conduct geophysical and engineering studies for assessing the relation between the Bell Rapids Irrigation District canal system and the perched aquifer discharge on the hillside. Vector hypothesized if the canal water was the recharge source then an electrical continuity between the canal water and perched aquifer discharge water would be evident. The Vector study employed electrical techniques to determine if there is a correlation between the aquifer discharge and the canal water.

Pelton (1995) then evaluated these surveys for accuracy and acceptable methodology. According to Pelton, Vector Engineering does not provide complete information on the data acquisition and reduction procedures for any of these electrical studies in its report. Pelton (1995) states that despite the lack of detail it appears the Wenner and Schlumberger data were acquired and reduced using procedures acceptable to most professional geophysicists although the data post-processing for the Schlumberger study using a computer model produced skeptical results.

Description/Purpose

In 1994 Vector Engineering conducted a Misa-a-la-masse study near the Fossil Gulch Canal with the goals of collecting data for identifying ground water flow geometry and establishing an electrical connection between the perched aquifer discharge zones at the hillside and the canal. Figure 23 illustrates the locations of each Misa-a-la-masse array station marked with a triangle where voltage potential was measured lateral to the canal while the canal water was charged with an electrical current.

Results/Conclusion

The seeps at the bluff face exhibit high electrical potential indicating the perched aquifer discharge zones are hydraulically connected to the canal. Perched aquifer discharge water always exhibited higher voltage potentials than other areas at the bluff face, which had no water discharging and were apparently "dry" (Vector Engineering, 1994). Pelton (1995) states the current electrode placed at 3,000 to 5,500 feet from the canal may not have been far enough away due to zones of elevated potential difference at the margins of the survey area. Also, the procedure used by the Vector Misa-a-la-masse study differs with the procedure described by Parasnis 1973 in two fundamental ways.

1. The access point of the current source to the conductive mass was not fixed.
2. A roving electrical potential difference measurement was made rather than the electrical potential measured relative to a fixed base.

Additionally, the current electrode located at the canal was not fixed and so it is incorrect to combine all of the Misa-a-la-masse data together on a single map. The experiment design is not described completely and the exact locations of the current electrodes is unknown but, the results are consistent with an electrical connection between the canal and perched aquifer discharge zones (Pelton, 1995).

Figure 23 Misa a la masse map (omitted from on-line edition)

Description/Purpose

In 1994 Vector Engineering conducted a Wenner array study along the Fossil Gulch Canal with the goal of collecting data to aid assessment with the relation between the Fossil Gulch Canal and the underlying perched aquifers. Figure 24 illustrates the location of the Wenner array. The scope of the study includes a segment of the Fossil Gulch Canal between the inlet and Fossil Gulch Pond about 2.8 miles downstream. The Wenner array is very commonly used to identify flat-lying interfaces between lithologies. Resistivity data was collected directly below the canal to map subsurface anomalies. Pelton (1995) evaluated the Wenner study for accuracy and acceptable methodology.

Results

Vector Engineering 1994, states that from the start of the unlined canal and progressing 9,000 feet downstream resistivity decreases with depth indicating an increase in moisture content as illustrated in Figure 25. Resistivity increases with depth from 9,000 feet onward to the Fossil Gulch Pond indicating a decrease in moisture content suggesting this portion of the canal is not leaking as much as the previous 9,000 feet. Low resistivities are inferred by Vector to have high permeability even though saturated clay-rich units are known to have a low resistivity. Vector states that the low permeability of clay-rich units is believed to act as an aquitard inhibiting saturation of the clays and resulting in high resistivities for the "dryer" clays. Also, a large resistivity difference between shallow and deep measurements indicate lower permeability while if the geologic unit has high permeability the resistivity values would be similar indicating the same percentage of water saturation (Vector Engineering, 1994).

Figure 24 Wenner plan view map (omitted from on-line edition)

Figure 25 Wenner Profile Results (omitted from on-line edition)

Wenner array data at one location correlates with an anomaly from the Misa-a-la-masse data. Figure 23 illustrates a location with high electrical potential from the Misa-a-la-masse study and correlates with low resistivity from the Wenner study. Resistivity drops below 20 OHM-Meters between stations 2,500 and 3,000 from the Wenner array while electrical potential from the Misa-a-la-masse study was measured at 59 millivolts near the same location (Vector Engineering, 1994).

Conclusions

Vector Engineering claims that resistivity measurements collected along the canal are effective for representing zones with high water saturation and high water saturation implies a lithology, other than clay, transmitting water; the clays would inhibit downward migration of water thus high saturated conditions will not occur in the clays. These interpreted zones of high water saturation are assumed to be associated with canal leakage.

One flaw, according to Pelton (1995), in the Wenner study is the basic premise stated by Vector Engineering as:

"As the water saturation of a given porous medium increases, it has a larger and larger affect (sic) on the resistivity of the material. By measuring the resistivity at two different depths the degree of water saturation is evaluated at the two separate depths. In the same medium (e.g. sand) low resistivities indicate a higher water saturation and high resistivities indicate a lower water saturation (Vector Engineering, 1995)".

The intention of the Vector study was to map the estimated apparent resistivity at two different depths and then, based on the Archie water saturation equation, qualitatively infer the water saturation in the shallow subsurface. The Archie water saturation equation can be used to relate water saturation to measured resistivity in a porous medium if clay is not present (Schlumberger Educational Services, 1987). If clay is present then the Archie water saturation equation cannot be used. Lithology and water resistivity of the shallow subsurface need to be constant along the canal in order to use the Archie equation to interpret the data in terms of lateral changes in water saturation. It is common for low resistivities to be associated with low permeable clay rich lithologies and even slightly moist clay conducts well due to ion active clay minerals (Milsom, 1989). A change in apparent resistivity would accompany any lateral facies change or chemical change in a sedimentary unit (Pelton, 1995).

Description/Purpose

The goals of the Schlumberger study were to collect geophysical data to aid assessment with the relation between the Fossil Gulch Canal and the underlying perched aquifers. Figure 26 illustrates the location of each Schlumberger station marked with a triangle and defines the scope of the study which includes a segment of the Fossil Gulch Canal between the inlet and Fossil Gulch Pond about 2.8 miles downstream. Data was collected using an expanding electrode array to map subsurface resistivities and construct geoelectric cross sections and plan view maps. Pelton (1995) then evaluated this survey for accuracy and acceptable methodology.

Results

Considerable variation exists in the subsurface resistivity data throughout the study area likely due to the complex paleo-flood plain deposits and ground water flow geometry. There are low resistivities at one hundred feet below the Fossil Gulch Canal but the Shoestring Basalt flow is very resistive and variations within the flow are inversely proportional to porosity (i.e., low resistivities correlate to higher porosity) (Vector Engineering, 1994). Processing of the raw data was performed with a computer program named ABEM's "Super VES" to generate a multi-layered model for each Schlumberger station. The computer model suggests various geo-electric units between stations which are interpreted as sandy and clay-rich inter-fingering lithologic units with a relatively continuous basalt layer throughout most of the survey area (Vector Engineering, 1994).

The raw resistivity data was computer processed and used to produce cross section diagrams. When the computer program is anchored to well logs the cross section diagrams appear to be reasonable. Debatable results occur when the model is not anchored to well logs and the program has connected three separate resistivity responses at different elevations into one lithology. If the same data processing methods were used to create the plan view maps of resistivity as the vertical cross sections (which is likely the case) then there is a high degree of uncertainty in the Schlumberger maps.

Figures 26 and 27 illustrate the plan view Schlumberger study results after computer processing for 100 and 150-foot depths respectively below land surface. The triangle symbols mark the location of each Schlumberger station and the crosshatch pattern is interpreted as high water content by Vector Engineering. In both figures there is a general northwest/southeast trend to the high water content patterns along with apparent northeast/southwest trending axes from the low and medium water content patterns.

Conclusions

According to Pelton 1995, it is important to realize the unique methodology used to create layered resistivity models from the computer program ABEM"s "Super VES" which produced debatable results. Also, a typical Schlumberger curve (apparent resistivity versus AB/2) may be fit by more than one layered resistivity model and each model must be considered one of many possible models that could fit the data equally well (Pelton, 1995). The vertical electrical sounding diagrams from the Schlumberger array study are suspect due to the data post-processing by the computer program.

Figure 26 schlum at 100 ft (omitted from on-line edition)

Figure 27 schlum at 150 ft (omitted from on-line edition)

Table of Contents
Chapter 1 | 2 | 3 | 4 | 5
Appendix A | B | C | D

Questions or Comments?