Growing with the world's population is the demand for increased development of water resources.

The world population is growing geometrically, projected to reach 8 - 11 billion people by 2050. Even in North America, where population growth rates are considerably lower than most of the world, the 1999 population estimate of 303 million is expected to grow to 374 million by 2050. Growing with population is the demand for increased development of water resources. Groundwater resources figure prominently throughout North America for private and public drinking water supply, irrigation, manufacturing, and mining. The United States uses approximately 76.4 billion gallons of groundwater per day for these purposes. The U.S. Geological Survey (USGS) estimates that in 1995 38 percent of the water supplied by municipal water departments in the US came from groundwater sources. Groundwater supplies served more than 97 percent of the rural population with its drinking water.

In Canada the numbers are somewhat lower, where in 1981 approximately 11 percent of municipal water was supplied by groundwater and approximately 82 percent of rural water supplies came from groundwater. This is probably due to abundance of surface water sources in Canada. Groundwater usage is expected to increase during this century as need for water increases and available sites for surface reservoirs decrease.

Ancient Methods of Finding Groundwater

One of the oldest and most controversial methods for locating groundwater is dowsing. Dowsing, also known as water-witching, divining, rhabdomancy, or doodlebugging, is locating water supplies by use of dowsing rod or pendulum. These dowsing rods can be a forked stick (Y-rod), a rod bent at a 90-degree angle (L-rod), or a straight stick (bobber). The dowser is said to be led to water supplies by movement or action of the dowsing rod. The origin of the practice is not clear, but earliest known indication of its usage occurs on a 4500 to 5000-year old grave inscription in Brittany. The first known literature reference is found in a 1540 publication on mining, De re Metallica by Georgius Agricola.

Opinions on validity of the method range from avid support to complete skepticism. Many scientific studies on the practice have yielded no concrete evidence of its validity, but many continue to use dowsers as their preferred method of locating drilling sites.

Traditional Scientific Methods

One popular method of locating drilling sites for many decades is fracture trace analysis. Particularly in bedrock aquifers, groundwater tends to be concentrated in localized fracture zones. Fracture zones are areas of higher fracture density within bedrock of lower fracture density (Figure 1). Fracture zones are generally less resistant to erosion than unfractured rock, and are more likely to form topographic low areas or stream beds.

Fracture zones are expressed on ground surface over long areas, called fracture traces or lineaments. Because these features are large in scale, they are difficult to detect on the ground. Hence, the most common method of investigating fracture traces is through aerial photographs. On aerial photographs, fracture traces are usually expressed as tonal variations in soils, alignment of vegetative patterns, straight stream segments or valleys, aligned surface depressions, gaps in ridges, or other features displaying a linear orientation. Figure 2 illustrates an example of how some fracture traces appear on aerial photographs.

Statistical studies of wells located in carbonate terrain have shown wells in fracture traces tend to have higher yield. Figure 3 illustrates the productivity of fracture-trace wells is considerably higher than wells not located on fracture traces.

While using fracture trace analysis is an improvement over random drilling and can decrease uncertainty of obtaining high well yield, it is far from a conclusive method. At ATS we routinely become involved in water supply projects where unsuccessful wells had been drilled on features interpreted to be fracture traces. Not all linear topographic features are fracture traces, and not all fracture zones are expressed on the ground surface as linear topographic features.

Improving on Traditional Methods

Geophysical methods have been used for subsurface exploration for many years. Some of the most popular methods in recent years include seismic refraction, very low frequency radar (VLF), and electrical resistivity. Recent advances in electrical resistivity imaging make this method the most inexpensive and effective geophysical technique available.

By starting with an understanding of local geology, we conduct a fracture trace analysis to look for likely fracture zones. The final step is to apply electrical resistivity imaging to confirm or refute the existence of fracture zones.

The primary factors affecting resistivity of earth materials are porosity, water saturation, clay content, and ionic strength of pore water. In general, minerals making up soils and rock do not readily conduct electric current and most current flow takes place through the material's pore water. Therefore, resistivity decreases with increasing porosity and water saturation. Clay minerals tend to be conductive because of availability of free ions in sheet silicate structure of clay particles. Therefore, electrical resistivity of a material decreases with increasing clay content. Similarly, dissolved ions in groundwater make water more conductive to electric current. Thus, electrical resistivity decreases with increasing ionic strength.

Resistivity imaging (also called resistivity surveys, resistivity profiling, and resistivity tomography) is conducted by inducing an electric current into the ground between two electrodes, and measuring the potential at other electrodes. Numerous configurations of electrode placement are commonly employed, each with unique data characteristics. Figure 4 illustrates a schematic diagram of dipole-dipole configuration. A current is applied to two adjacent electrodes positioned a predetermined distance apart (distance a). Voltage across two other electrodes is measured simultaneously with the applied current.

The two sets of electrodes are always spaced a distance apart and the distance between the current and voltage electrodes is always a multiple of a (n* a). Depth of sampling increases with distance between the current and potential electrodes. Therefore, regardless of the configuration employed, the resistivity dataset always has a triangular shape with the deepest sampling occurring in the middle of the section (Figure 5).