power to the stream bed were minimized through the provision of adequate 

 clearance between distal ends of electrodes and stream bed to allow for 

 any sagging of the electrode suspension caused by expansion of cables or 

 settling of tower bases. Damage to, or permanent displacement of, elec- 

 trodes by floating debris and mechanical damage to wiring was practical- 

 ly non-existent. 



Hinge mountings as used in the multiple-row electrode systems 

 and the short cable and pipe-cap hangers used in the single row and 

 ground line systems proved to be inexpensive and mechanically sound (see 

 "Detail" in Figs. 6 and 11). It was found to be absolutely essential 

 that only hinge mountings be used with any multiple-row, suspended elec- 

 trode system. This type of mounting restricts the movement of the elec- 

 trode to a single plane of deflection (downstream) and prevents contact 

 between electrodes of adjacent rows with resultant short-circuiting of 

 the device. The simpler cable and pipe-cap mounting is entirely satis- 

 factory for use with the single row, suspended electrode and ground line 

 system since random movement and contact among the single row of sus- 

 pended elements produces no damaging short-circuits. 



(6) Effects of water level , water resistivity, and 

 "bottom resistivity on power drain 



A direct, though non-linear, relationship exists between power 

 consumption of an electrical weir and the water level of the stream in 

 which it is installed, assuming water resistivity and applied voltage to 

 be constant. Conversely, at a given water level, power consumption of 

 such a device is directly proportional to the square of the applied volt- 

 age and inversely proportional to the water resistivity. Changes in the 

 power demand of an electrical weir or variation in load at other points 

 on a commercial power line will, in extreme cases, cause fluctuations of 

 ± 6 volts on a nominal 117 volt AC line. Water resistivity, in turn, is 

 influenced by such factors as dissolved and suspended solids content, pH, 

 and water temperature. Frequent measurements of water level and power 

 consumption made at each of the weir sites did not reveal the exact nature 

 of the water level-versus-power relationship. This fact may be understood 

 when it is realized that neither the direct effects of voltage variation 

 nor the sometimes diverse effects of the variables governing water resist- 

 ivity, could be controlled (e.g., water temperature might rise coincident 

 with a rise in dissolved solids content, these factors having opposite in- 

 fluence on water resistivity). Measurements of water temperature, dis- 

 solved solids content, and pH were made at each experimental site at every 

 possible opportunity in an endeavor to correlate these variables with power 

 consumption. Determination of the true correlation was precluded in this 

 case by the interaction of these variables and lack of a sufficiently large 

 number of measurements. However, in a test conducted at the Carp Creek 

 weir site at a constant voltage level over a period of time sufficiently 



Uo 



