The relationship between energy density (W-hr per pound of 

 magnesium), temperature, and spacing is shown in Figure 6. Energy 

 density is shown to be a strong function of temperture and a relatively 

 weak function of plate spacing. Based on theory it was expected that 

 energy density would increase with temperature. The experimental 

 results, however, show a marked decrease at the higher temperatures. 

 This effect at the high reaction rates is caused by unused magnesium 

 sloughing from the plates. The sloughed magnesium can be seen as small, 

 dark-colored particles circulating in the electrolyte. Figure 6 shows 

 that an energy-efficient reaction rate occurs between 100°F and 150°F 

 (38° and 66°C). 



The effects of plate spacing and temperature on power density are 

 shown in Figure 7. As expected, the reaction rate proceeds more rapidly 

 at higher temperatures and closer spacings. 



The electrolyte condition is described in terms of pH, salinity, 

 and density. Theory predicts that the reaction rate is reduced by high 

 pH, low salinity, and increased density (resulting from reaction products 

 mixing with the seawater electrolyte). To some extent the electrolyte 

 density and pH can be controlled. But, in general, the heater must be 

 designed to accommodate the natural variability of the seawater enviro- 

 ment. 



The pH, salinity, and conductivity of the electrolyte were measured 

 to determine quantitatively how these parameters affected cell perfor- 

 mance. For these parametric tests, the pH changed relatively little, 

 because the comparatively large volume of seawater diluted the reaction 

 product concentration. In addition, since seawater is a very good 

 buffer, large quantities of Mg(0H)2 would be required to make significant 

 pH changes. Generally speaking, the reaction rate was affected as 

 expected. Table 2 summarizes the effects of these and other parameters 

 on cell performance, but no quantitative trends were identified. 



Reaction Rate Decay Characteristics. To verify the results of the 

 parametric tests, a series of large-scale tests were run in the acrylic 

 vessel. In these tests fresh electrolyte was added at a controlled rate 

 (125 ml/min), and an equal volume of slurry [Mg(0H)2 and water] was 

 removed to maintain constant electrolyte pH and density. The results 

 of a typical test are shown in Figure 8. The broken line shows the esti- 

 mated decay in cell power as predicted from parametric tests and based 

 on anode consumption and increasing electrode gap. These results ver- 

 ified the fact that reaction is strongly dependent on electrode gap. 



The power output during the first hour was significantly higher 

 than predicted. The increase is attributed to several factors: clean 

 anodes and low electolyte pH and density. Initially, the anodes are 

 clean and free from reaction products, but during the first hour, magne- 

 sium hydroxide accumulates on the anode surface. These deposits inhibit 

 the reaction process at the anode and subsequently reduce the overall 

 reaction rate. Also, fresh seawater, which is a very good buffer, has a 



