and permitted following the distribution of the hottest portion of 

 the effluent stratum for several consecutive hours (Figs. 17 and 18). 

 Figures 15 and 17 are typical of the isotherm distribution observed 

 during March, 1971. At that time the effluent stratum was suspended 

 in mid-water and moved regularly with the tide. Figures 16 and 18 

 show the typical effluent stratum in May, 1971 extending from fifteen 

 feet to the bottom of the canal. Tidal influence was small during 

 the periods of observation. 



COPPER AND NICKEL 



Copper and nickel discharged from the desalination plant increased 

 to a maximum in January, 1971 and subsequently decreased until 

 August, 1971. The copper and nickel came from corrosion in two 

 areas of the plant: 1) from the first set of monel tube-bundles 

 receiving the well water directly and 2) from copper-nickel (monel) 

 separatory screens in the deaerator. Although monel normally is 

 highly resistant to saltwater corrosion, the combination of heat, 

 high water velocity, low pH, and excessive H2S caused rapid removal 

 of copper and then nickel from the metal alloy. 



There are five tube-bundles in the desalination plant. Most of the 

 corrosion, however, occurred in the tube-bundle which received the 

 raw saltwater supply from the pumps (see Figure 2, page 18). This 

 water contained 4 to 13 ppm H2S (as normal for the large saltwater 

 aquifer which underlies the Florida Keys). After treatment in the 

 atmospheric decarbonator (aeration and acidification) the brine 

 passed through the air ejector (to remove excess dissolved gas) and 

 spilled down into the brine chambers through large separatory screens. 

 The water, at this stage, had its lowest pH and the highest concen- 

 tration of hydrogen sulfide. Most of the internal corrosion occurred 

 at this point. From June, 1970 to June, 19 71 the discharge averaged 

 1,766 ppb copper. Copper in the well water averaged only 56.32 ppb. 

 Analysis showed the copper in the effluent to be 78.4% ionic, 3.4% 

 particulate, and 18.2% organically complexed (Table II). 



Copper discharge was greatest during periods when plant operation was 

 not stable and particularly after the plant resumed operation follow- 

 ing maintenance periods. The ionic analysis of the effluent shol^m in 

 Table III was taken on February 3rd and 4th, 19 71 following resumption 

 of plant operation after maintenance. Samples were taken for twenty- 

 four hours after start-up on February 3rd, 19 71. Copper, nickel, and 

 iron were all present in high concentrations following the onset of 

 desalination operations. At that time, the salinity was lower than 

 normal (39.14 0/00 as indicated by sodium levels) and the effluent 

 did not sink. By midnight, six hours later, the effluent reached a 

 salinity of 47.59 0/00 and peaked at 54.21 0/00 at 0300 on February 

 4th, 1971. Thereafter, the effluent continued discharging at its 

 average value of 52 0/00. 



58 



