SPONTANEOUS POTENTIAL
Between any two points on the surface of the earth, there generally exists a difference in electrical potential, commonly measured in volts (or, millivolts). There are two natural 'grounds' (natural sources or sinks that can absorb or supply more charge than we humans can generate). One is the ionosphere, where ionized gasses conduct electricity, and the other is the earth (hence, the origin of the term "ground"), where ions in soil moisture and ground water conduct electricity. There exists a tremendous potential difference between these grounds, a difference maintained by the atmosphere (an insulator). Because the air is not a perfect insulator, a slight electrical current slowly flows through the air, resulting in a near-surface gradient of several hundred volts/meter. This potential is difficult to measure because of the very low charge density; however, atmospheric disturbances can disrupt SP investigations if we are not careful when we plant our electrodes.
SP is due to a number of factors. A natural north-south flow of electrical current, driven by the dynamics of the ionosphere, generates what we call TELLURIC electrical currents. Since the earth is a good (but not a perfect) conductor, Ohm's Law implies a pattern of potential differences due to this current.
Electrochemical potentials can be generated by the battery-like effect of dissimilar fluids in contact with a uniform soil, or, from dissimilar rocks in contact with ground water. These result in anomalies - interesting variations in an otherwise dull, uninteresting earth. It is such variations, or anomalies, that we seek. Example: sulfide ore bodies oxidize when exposed to oxygen (e.g., when the water table drops and exposes previously saturated rocks to air). The resulting chemical reaction (sulfide to sulfate) makes the ore body act much like a large battery. Mapping SP has outlined some ore deposits (see Parasnis for examples). Example: acidic fluids leaking from a SUPERFUND site are chemically reactive compared to normal ground water, allowing a battery-like situation to develop (see Stierman, 1985, attached). Example: electrochemical reactions between drilling mud and fluids in pores of rocks penetrated by the drilling bit set up electrical currents detected by SP logging of boreholes.
A third mechanism responsible for natural voltage differences is streaming potential. Flowing water strips electrons from rock fragments and a potential difference occurs. In areas of significant topographic relief, there is often an SP pattern highly correlated with elevation (water generally flows downhill, even underground). Water flowing toward a cone of depression can also cause an SP anomaly. SP has been used to map geothermal fluid circulation in The Geysers field north of San Francisco, California. SP has been used to map fractures carrying water under dams.
METHOD
EQUIPMENT: 1 pair non-polarizing electrodes (porous pots); wire and high- impedance VOM or potentiometer. Note that all geophysical surveys require the use of basic measuring equipment such as surveying tapes to measure distance, a compass to measure bearings, base maps on which to plot the locations where measurements are taken, tools to dig holes, etc. Do NOT list all routine equipment in your report - just those tools specific to the method employed.
Non-polarizing electrodes: if two metal stakes are driven into the earth, they react with chemicals in the soil and the geoelectric field to set up a countercharge. We therefore require electrodes that do not polarize - porous pots containing saturated copper sulfate solution are the standard non-polarizing electrodes. These cost about $80/pair from Soiltest, or you can make your own from ceramic tips used for tensiometers (about $15 each), 15 cm of PVC tubing, a rubber stopper and length of copper tube. Copper sulfate can be purchased at many hardware stores or from industrial chemical supply distributors. From Soiltest, 1 lb. of copper sulfate may cost $10; in bulk, a 50-lb bag may cost about $50. Note: copper sulfate is the main chemical used to prevent algae from growing in waterbeds.
WARNING: copper is very toxic to fish. Take care not to contaminate areas where we are conducting our surveys. We are supposed to be the "good guys" - the people solving groundwater contamination problems, not the parties responsible for pollution.
A saturated solution is used so that the electrolyte in all porous pots is identical. A saturated solution is maintained by mixing up the solution well in advance and always adding more crystals than can be dissolved by the water in our container. If there are no crystals on the bottom of the electrolyte jar, the solution is probably not saturated and needs additional chemical. For very precise work (SP as an earthquake precursor?), a silver iodide solution with silver wires in place of copper tubes is used. These are very stable; however, they are expensive and tricky to assemble (silver solder has a high melting point and releases toxic fumes). Porous pots must be filled and allow to equilibrate several hours prior to being used. Fill the pots, place them in a container (so that leaking electrolyte does not mess up the lab) and connect the copper stems with wires.
Wires in the field: rubber or vinyl; there must be no leaks to ground or exposure to the atmospheric voltage gradient. Dirt or moisture on the cable reel can leak signal.
Voltmeter: digital VOM capable of measuring DC millivolts; must he very high input impedance.
Potentiometer: the best device for measuring SP. A VOM modifies the system it measures (although this is not significant if a quality meter is used). The potentiometer balances an external (unknown) voltage against an internal (reference) voltage by adjusting a calibrated voltage divider. When balance is achieved, no current flows through the indicator (dial) and the voltage (in millivolts) is read from a calibrated dial. Because no current is drawn from the system, the system measured is not modified by the measurement. On our potentiometer, there is a multiplier switch (x 0.1, x1, x10, x100) and a 3-digit dial; the right-hand digit wheel has tic marks between the numbers so that a total of 4 significant figures can be obtained. There is also a polarity switch; increasing the dial reading rotates the needle counterclockwise. If the dial reads "0) and the needle points to the left of the null position, you (1) flip the polarity switch; (2) re-null the needle while pressing the "NULL" button; (3) release the null button and make your measurement. NOTE IN YOUR FIELD BOOK: polarity switch, multiplier switch, dial reading - at least until you have done this so often that you do not make reading errors.
CONDUCTING AN SP SURVEY
SP measurements require good ground contact; dry or loose soils make SP readings erratic. Plant the pots with care, tamping soil firmly around them.
Review your field protocol; make a sketch map of the location, etc.
One electrode is planted to serve as a reference electrode. Potential differences will be measured with respect to this spot. Plant the second electrode next to the reference electrode and measure the SP; what you are measuring is a bias due to minor chemical differences between the porous pots (should be only a couple of mv). If the difference is greater than a few millivolts (I like 2 mv but tolerate 5 or 6), try a different pot. The MOVING ELECTRODE SHOULD BE CONNECTED TO THE POSITIVE (RED) TERMINAL OF YOUR POTENTIOMETER OR VOM. This is my protocol. In any event, this is the sort of detail that must be noted in the field book and mentioned in your report. The roving electrode is then moved from place to place. At each location, the pot is planted and allowed to sit for a minute or two. SP is measured and the location noted. It is usually best to grid survey points or to plan a profile across a suspected structure.
INTERPRETATION: One data point does not constitute an "anomaly"; very local, shallow conditions can cause spikes (positive or negative). You should plot data as they are collected so that spikes can be investigated by changing sample density. SP anomalies will provide targets for further electrical investigations, for measurements that are more diagnostic but which are slower or more difficult to make. Few people are successful in conducting SP surveys. One "secret" involves contact resistance - that is, how well is the electrode "connected" to the earth? After all, atmospheric electrical gradients are not properly grounded. When using a VOM, you can check ground 'resistance' by using the OHMS dial - a resistance below 3000 ohms is very good, a resistance over 20,000 ohms is very bad. The potentiometer becomes unstable if contact resistance is poor. If the needle moves when you touch the case, one of the electrodes is not well grounded. You are not likely to lead many SP surveys, but SP will have to be dealt with in any electrical study you do make and it may become more useful than commonly thought.
EQUIPMENT CHECK LIST: SPONTANEOUS POTENTIAL SURVEY
Potentiometer or VOM
Porous pots (at least 2), filled with saturated copper sulfate solution at least 24 hours prior to beginning survey; 1 bottle extra solution.
Insulated alligator clips (for attaching to porous pots) and banana plugs (for input to VOM or potentiometer).
Surveyor's tape, tool box, rock hammer or small shovel, field notebook and ballpoint pen, graph paper, map if needed.
DATA DISPLAY AND FILTERING
Plot SP as a function of location along your profile.
Numeric filters are sometimes used to smooth data. Other filters are used to highlight differences in the data.
Suppose your data along a profile can be represented by a series of numbers representing SP readings along some regularly spaced line.
SP = {12, 15, 46, 19, 24, 30, 35, 10, 5, 3, 4, 5, -1, -3, -4, 4, 12, 13, 16, 22, 25, 31, 24, 20, 15}
An operation called CONVOLUTION can be used to modify this series. Suppose you wish to smooth the series. This can be done by replacing each value with a certain fraction of that value added to a fraction of the neighboring values. For example, the filter {.3, .4, .3} adds 0.3 of the values on either side to 0.4 the center value. Applied to the series above, the new smoother series is
SP = {23.4, 28.6, 28.6, 24.3, 29.7, 26, 16, 5.9, 4.3, 2.9, 0.2, -2.7, and so on)
A smoothing filter should have a sum total equal to 1; when applied to a horizontal straight line, the values should remain the same (what is smoother than a straight line?). Note that the end points of the profile are lost. Numeric filters can yield odd results near the endpoints of a series - such end effects should not be interpreted as they are artifacts of a filtering method. The 3-point running average using coefficients {0.3, 0.4, 0.3} spreads a spike among its neighbors.
A different filter can be used to enhance differences in a series. The filter {-.5, 1, -.5} subtracts 0.5 the value on either side of a value from the center value. When applied to a straight horizontal line, this filter yields "0" for all values of the filter output. Note that a differencing filter should have this characteristic of summing up to 0.
When you use a filter to smooth data or to enhance differences, you must specify the filter in your report. It is not sufficient to state that "a numerical filter was used to smooth the data" because the reader has absolutely no idea as to the characteristics of your filter.
IN CASE OF SPONTANEOUS POTENTIAL MEASUREMENTS, DO THE FOLLOWING:
If not, just keep these notes for future reference.
TO DO: Display your readings on a graph. On a second graph, display your SP readings smoothed by a 3-point filter such as {.2, .6, .2} or {.4, .3, .4} or perhaps {.33, .34, .33}. Are the major trends in the SP data preserved while any spikes are attenuated? You are free to select any 3-point smoothing filter you wish but you must specify in your report the filter coefficients. I often read reports in which filtered data are displayed without any clue as to the nature of the filter. This is not acceptable practice.
CONVOLUTION: Let series A = {a1, a2, a3, - - - aN} and series B = {b1, b2, b3}. The convolution of A with B yields series C = {c1, c2, c3, - } with
c1 = a1*b1
c2 = a1*b2 + a2*b1
c3 = a1*b3 + a2*b2 + a3*b1
c4 = a2*b3 + a3*b2 + a4*b1
c5 = a3*b3 + a4*b2 + a5*b1
and so on. Note that series C has (N + 3 -1) elements, the last element being c(N + 2) = aN*b3. Recall the convolution spreadsheet used for synthetic seismograms in 4610/5610.. The teaching assistant is also familiar with a software product called PC-MATLAB that can perform the convolution operation, and I have listings of BASIC and FORTRAN routines that perform this operation.