CHAPTER 12 - DISCHARGE MEASUREMENTS USING TRACERS
6. Tracer-Velocity-Area Methods
Either salt or dye may be conveniently used in tracer-velocity-area discharge measurements with equal potential accuracy. The only difference is that different detection equipment is needed. Dyes have an added advantage in that they can be detected visually, allowing simpler measurements of less accuracy that may be sufficient for irrigation needs. However, when using any simpler method, the error checks and mixing problems of section 5 in chapter 12 should be considered.
(a) Salt-Velocity-Area Measurements
The salt-velocity-area method takes advantage of the fact that salt in solution increases the electrical conductivity of water. This method has been successfully used in open channels and pressure conduits of constant cross section.
Because of its high potential accuracy, the salt-velocity-area method is one of several methods accepted for turbine testing in American Society of Mechanical Engineers (1992). The equipment described in Thomas and Dexter (1955), consisting of injection system and the sensing electrodes (figures 12-1 and 12-2), are rather complex. Also, a turbulator is sometimes used to ensure adequate mixing of the injected salt tracer solution and the flow by the time they reach the first electrode station. Full details regarding the equipment required for techniques found satisfactory under field conditions are contained in Thomas and Dexter (1955).
Commonly, sodium chloride (NaCl) is the selected salt used in the tracer injection solution. Finely ground salt should be purchased for ease in mixing the solution. Enough salt must be added to significantly increase the electrical conductivity of the water so that concentrations can be measured accurately. The required amount of salt can be estimated by analyzing the water for existing background quantity of salt in the measurement flow, estimating the amount of flow to be measured, and using chemical handbook data from conductivity-salinity tables. Trial runs may be needed to determine the optimum amounts, which may vary with discharge depending on the range to be measured.
For a measurement, a quantity of salt tracer solution is forced into the stream under pressure to provide better initial distribution and assure thorough mixing before arrival at the detection stations. The pop valve injector used by Thomas and Dexter (1955) (figure 12-2) will provide the faster and better mixing required to produce code accuracy.
To determine velocity for equation 12-1, a pair of electrodes is installed in the cross section at each end of a measured length of channel well downstream from the injection system. The distance between the pairs of electrodes should be sufficient to ensure accurate measurement of the time of travel between them. The electrodes are electrically energized and connected to a central instrument that records the electrical conductivity at each pair of electrodes with respect to time.
A sample of a strip chart recording showing conductivity change that occurs as a salt cloud passes the electrodes is shown on figure 12-3. The recording shows a conductivity rise that indicates the passing of the salt solution cloud past each electrode station. In addition to their peaks, the cloud plots have a leading and a trailing edge of low conductivity approaching the baseline conductivity of the flowing water (figure 12-3). The time of cloud travel between electrodes is measured on the chart time scale between the centroids of the two plotted conductivity cloud areas above the background conductivity level. Digital recordings are more convenient than analog recordings for computer determination of area under the time-conductivity plots to determine the center of mass of salt clouds.
This method requires special equipment and experienced personnel and is relatively expensive. Care in selecting convenient reaches will help reduce time and expense in measuring length and determining an accurate average cross-sectional area.
(b) Color-Velocity Measurements
To achieve maximum accuracy using dye tracer solutions, the procedures similar to those described for the salt-velocity-area method must be followed. However, fluorometer detection or a set of visual color comparison standards must be used instead of equipment used for salt solution injections. Carefully following American Society of Mechanical Engineers Performance Test Codes (1992) will result in very accurate discharge measurements. Fluorometry combined with well designed multiple port injection and sampling port arrangements at two stations downstream from an injection station produces high accuracy.
The simpler but less accurate method using visual observation for tracer cloud detection in pipelines consists of determining the velocity of a dye tracer between two stations in the pipe. This velocity, used as the mean velocity of flow, is multiplied by the cross-sectional area of the pipe to give the discharge as shown in equation 12-1.
The simplified procedure ordinarily used in making the velocity measurements is described below. If possible, a small slug of concentrated dye solution is quickly injected or poured into the pipeline entrance where the pressure is relatively low. In pipes, a high-pressure system through fittings may be required to inject dye. Time observations are made at the instant the dye is injected and at its first and last appearance at the downstream station, usually at the pipe outlet. The mean velocity is computed from the mean time required for the dye to travel the known length of reach. Com-parisons with other measurement methods show this simplified color velocity method is accurate enough for irrigation measure-ments when properly done in relatively long pipes.
Simplification similar to the pipe flow case is possible in open channel flow. However, the color-velocity-area method in open channels has more limitations and drawbacks. The air entrained by surface velocities and spray above the surface may hinder detection of the position of the center of mass of the colored water in high-velocity flows (Hall, 1943). Also, slow flows are more likely to cause mixing problems.