EFFECTS OF THERMAL EFFLUENT 549 



cam pattern. One time-cycle controller and mixing valve controlled 

 the seawater flow to each group of 20 experimental tanks. The 

 mixing valve was supplied with warm effluent water through a 

 common line from two centrifugal pumps, each drawing from a 

 separate effluent sump unit. Each of these sumps, in turn, received 

 part of the drain flow from 10 experimental tanks, and the 

 remainder of the drain water was discharged to waste. Thus the 

 system recycled part of the warm water to conserve heat but had a 

 substantial and continuous supply of makeup water (10 to 30 

 liters/min) from both the effluent and ambient- temperature labora- 

 tory supply lines. The flow rate of mixed seawater to the individual 

 tanks averaged 8 liters/min (range, 6 to 10 liters/min). 



Continuous temperature records were made in the experimental 

 tanks and were compared with temperatures called for by the cam 

 patterns. Under normal operating conditions the accuracy of the 

 controller system in duplicating field-temperature records was within 

 ±0.5° C. An air stone operating continuously at a high flow rate in 

 each tank maintained dissolved oxygen concentrations at normal 

 levels and increased vertical and horizontal mixing. 



Twenty control tanks, identical with those used for the treat- 

 ment groups, were maintained at the generating-station laboratory 

 and were supplied with water at varying ambient temperatures from 

 the intake screenwell of the station. Twelve control tanks at the 

 Scripps Institution of Oceanography were supplied with water of 

 varying ambient ocean temperatures. Flow rates, oxygenation, and 

 other conditions were the same as those in the treatment tanks. 



Banks of fluorescent lights providing natural and uniform 

 illumination were controlled by standard clock timers, which were 

 reset weekly to follow sunrise and sunset times. Sunlight entered the 

 laboratories through translucent windows, providing supplementary 

 light and reasonably natural twilight conditions. 



Estimates of the relative amounts of thermal-effluent and 

 ambient-source water entering the treatment tanks, which were made 

 from reading in-line flow meters, indicated that the test animals were 

 held in varying effluent mixtures similar to those at the two field 

 locations. This suggests that the animals were exposed to water 

 conditions simulating the field locations fairly closely in temperature 

 and chemical composition. Significant biological effects observed in 

 the treatment tanks were assumed to be the combined results of 

 temperature and chemical composition of the water. 



Because of the location of the laboratory, the ambient-intake and 

 thermal-effluent water used in the experiments came from the 

 cooling-water system of different generating units than those simu- 



