short to preclude the slightest variation in water resistivity,, a simu- 

 lation of water level change provided experimental data representative 

 of the water level-versus-power relationship. These data are presented 

 graphically in Figure 17. 



Relative values of stream bed and stream water resistance were 

 measured at the four electrical weir sites as described in an earlier 

 section. Results indicated clearly thats (a) in these locations the 

 ratio of stream bed to water resistance was l.lj : 1 or greater, and 

 (b) this ratio increased with the depth to which the resistance probe 

 was forced into the stream bed. These findings represent highly desir- 

 able conditions for the operation of an electrical weir. Where bottom 

 resistivity is high, excessive power loss will not occur in the stream 

 bed because its higher relative resistance will inhibit the flow of 

 current through it. Therefore, it is desirable to locate all electrical 

 weirs in an area where the ratio of stream bed to water resistance is 

 greater than 1, as measured by the method described earlier or by any com- 

 parable method. 



(7) Power sources 



The following three systems of electrical weir power supply, 

 all of which proved effective and satisfactory, were tested during the 

 course of the 1952 experiments s 



System 1 ( Commercial line power only ); This method of opera- 

 tion, as tested at Hibbards' Creek, did not permit the escapement of lam- 

 preys. No known power failure occurred, and at no time did the line 

 voltage fall to a value sufficiently low to render the barrier ineffective. 



System 2 ( Commercial line as normal , generator as standby power 

 source ); This system, as used at the Kewaunee River weir~was successful 

 in maintaining continuous power. Since no line power failures occurred, 

 the operation of the standby generator was checked frequently by manual 

 interruption of the line power. The time interval preceding restoration 

 of power to the weir by the standby generator was estimated to be consider- 

 ably less than 1 second. This was demonstrated by the fact that the dis- 

 continuity in power was only dis cernible with difficulty in the slight 

 deflection of a voltmeter connected across the weir terminals. The time 

 interval required to transfer the electrical weir load back to the normal 

 source of power upon its restoration was not sufficiently great to detect 

 visually in an incandescent light bulb connected across the weir terminals. 

 In neither case were lampreys capable of escaping through the weir in the 

 small interval of time during which the power was broken. The standby gen- 

 erator, which was operated a total time of approximately 170 hours, func- 

 tioned satisfactorily without breakdown or need for repairs of any kind. 

 This generator had ample capacity for the operation of floodlights and warn- 

 ing lights during periods of maximum power consumption of the electrical weir. 



hX 



