Many of the Bureau’s dams are earth embankments built over 50 years ago. Some of these structures are located in areas of high seismic risk. Although long recognized that earth embankments need periodic inspection and monitoring, it has only been recently recognized that geophysical surveys can supplement the results of standard inspection and monitoring techniques.

Bureau of Reclamation personnel modified a seismic refraction technique, and have pioneered the use of that technique (the common offset seismic method) to evaluate the condition of earth embankment dams. This technique is currently being used as part of the Safety Evaluation of Existing Dams (SEED) Program in the western United States. Approximately 25 embankment evaluations utilizing this refraction technique have been made to date.

The common offset seismic method is an adaptation of the seismic refraction method. A standard seismic refraction line is laid out using 24 geophones at 10-foot intervals. Energy is input into the ground at a point located some distance from the first geophone, resulting in a single-ended refraction line. A plot of time versus distance (in seconds) will yield the apparent compressional wave velocities of each layer detected by the refraction line. Figure 1 presents a schematic diagram of a single-ended refraction line and the resulting time-distance plot. After a single-ended refraction line is completed, the shotpoint (point of energy input) and geophones are advanced across the dam crest or toe at 50-foot intervals, where another single-ended refraction line is completed. This process is continued for the total length of the dam crest or toe. This sequence results in a compilation of seismic energy arrival times at a number of different geophones from a number of different shotpoints. However, the distance from a shotpoint on the first seismic line to, for example, geophone number 9 of that seismic line will be the same distance to geophone number 9 of any seismic line from the shotpoint for that line. Since this represents a common geophone to shotpoint offset (distance), the method of data compilation is termed the common offset method. Figure 2 presents a schematic diagram of this method. The resulting data for a particular geophone (1-24) for all single-ended refraction lines is displayed as a common offset sort, as shown in Figure 3.

For uniform embankment conditions, these arrivals would form a straight line of equal time. Since the field and processing parameters are held constant for each profile, any anomalies observed in the seismic data must be the result of changes in the embankment or foundation materials.

Velocities determined from the single-ended refraction time-distance graphs are compared with variations observed on the common offset sort. For direct arrivals, variations should be very similar since all observed anomalies in the common offset sort must be due to velocity variations. Variations observed on more distant channels (generally refracted arrivals) can be due to changes in velocity, changes in thickness or encountering a different refractor.

The compressional wave velocity is affected by many conditions. However, in sediments (or compacted soil) the primary factors affecting the compressional wave velocity are density, porosity and saturation. In lithified materials (rock) , factors such as cementation, fracturing, alteration and stress generally have a greater affect on the velocity.

EXAMPLE 2

This rolled earthfill dam is roughly 115 feet high and 900 feet long. Figures 5a and 5b show the common off set sorts for a low- and a high-numbered channel respectively. Figure 5a indicates very uniform conditions in the near surface materials. Figure 5b indicates an abrupt change in arrival times on the right side of the figure. This would indicate a much faster velocity or a sudden decrease in depth. The time-distance analysis for this data indicates that all the arrivals are direct arrivals with the exception of the far end of the profile. It appears, according to construction drawings, that the anomaly observed on the common off set data is due to an abrupt thinning of the embankment materials over the abutment.

(1) Background potentials, which are primarily a result of, (a) two electrolytes of different concentration being in contact with one another, (b) electrolytes flowing through a capillary system or porous media, (c) an electrolyte in contact with a solid, and (d) electromagnetically induced telluric currents.

(2) Mineralization potentials, which are primarily the results of chemical concentration cells formed when conductive mineral deposits, such as graphite or sulfide, are intersected by the water table.

Mineralization potentials are almost always negative, and may have values up to several hundred millivolts. Background potentials can be either positive or negative, and usually have values of only a few tens of millivolts.

The background potentials developed by electrolytes flowing through a capillary system, or porous media, called electro-filtration or streaming potentials, are used for the study of seepage. As water flows through a capillary system, it collects and transports positive ions from the surrounding materials. The positive ions accumulate at the exit point of the capillary, leaving a net positive charge. The untransported negative ions accumulate at the entry point of the capillary, thus leaving a net negative charge. If the streaming potentials developed by this process are of sufficient magnitude to measure, the entry point and the exit point of zones of concentrated seepage may be determined due to the negative and positive (respectively) self potential anomalies.

Figure 8 indicates the type of self potential anomaly, in millivolts, that can be produced by an artesian water system, while Figure 9 presents the type of self potential anomaly that can be produced by ground water draining or flowing into a fractured bedrock system. Self potential surveys have been successfully used to map reservoir leakage and water seepage paths through embankments and foundation materials.

Self potentials are measured by the use of nonpolarizable electrodes embedded in soil or water. These types of electrodes are used to prevent electrode potentials from developing and masking the self potentials. The measured value of the self potential is recorded on a sensitive, high impedance multimeter (or d.c. voltmeter). The field procedure consists of placing one electrode (base electrode) in a quiet area, such as the reservoir. Another electrode is moved at selected intervals along each of the profiles, and the potential between the base electrode and the moving electrode is measured, recorded and plotted at each location. Where changes in the measured self potential of over 10 millivolts are observed along a profile, detailed measurements are made at small intervals to better define the limits of these anomalous zones. Figure 11 presents the results of a self potential profile made over a known leak in the downstream toe of a dam.

The results of a self potential survey are usually presented as a set of profiles, plotting self potential values against the distance of the measuring electrode from the base electrode. The interpretation of this data is mainly qualitative. The magnitude and sign (±) of self potentials will be affected by seepage flow (streaming potentials) and both lithological and structural factors.

Potentials resulting from seepage flow can generally be classified as follows:

1. Positive anomalies can represent subsurface water flow, and areas of water discharge.

2. Negative anomalies can represent areas of water infiltration.

Potentials resulting from lithological and structural factors can generally be classified as follows:

1. Positive anomalies can represent areas of a higher content of clayey material.

2. Negative anomalies can represent zones of accumulations of coarse detritus material.

3. A sudden change in the magnitude and/or sign of the anomaly can represent a geologic contact.

4. Negative anomalies in a homogeneous rock formation can represent areas of fracturing.

It is evident that to separate streaming potentials from lithologic and structural potentials, care must be exercised in the field to note all soil and geological changes along each profile. In many cases, this information may be sufficient to eliminate certain factors from causing a self potential anomaly. For example, to determine if a negative anomaly is the result of water infiltration or fractured rock may be very difficult; however the geological position of the anomaly, its size and shape, observations of local outcrops, and correlation of the anomaly with known subsurface geological conditions may all be factors that can be used to eliminate water infiltration as the cause of the anomaly.

The self potential surveys offer an inexpensive method of seepage detection, and, when correlated with seismic surveys, self potential surveys can provide added information on porosity, saturation and/or permeability changes in embankment materials.