168 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 39 



the absorber; thus a maximum temperature differential is provided. 

 This surge quickly dies away as unheated water rushing through the 

 absorber fills the vertical riser while hot water from the tank is drawn 

 into the vertical drop; thus a temporary minimum temperature dif- 

 ference is created. Another smaller surge occurs later for similar 

 reasons, and then, except for a minor surge at 11: 15, the circulation 

 builds up steadily to a maximum at noon. It gradually falls off 

 thereafter as the tank water warms up and the solar intensity de- 

 creases. The flow stops rather suddenly when the absorber becomes 

 shaded. Because of nocturnal cooling of the vertical riser which 

 enters the tank 4 feet above the pipe to the bottom of the absorber, 

 slight reverse circulation occurs, reaching a maximum about sunrise. 



The temperatures of the water leaving and entering the absorber 

 are shown by curves B and G^ respectively. These refle<;t the flow 

 surges previously described. The temperature difference between B 

 and C does not indicate properly the driving force for the circulation, 

 because of the importance of the water density in the vertical riser 

 shown in plate 1 from points B to C and in the vertical part of the 

 cold pipe from tank bottom to pipe- absorber bottom at point A of 

 plate 1. The relation observed in figure 5 between the outlet tempera- 

 ture B and tank-center temperature D shows the mutual dependence 

 of the two. This 120-gallon storage should have had an absorber 

 nearly three times as large as the experimental 40.3 square feet, so that 

 the daily rise in tank temperature would be much steeper. Then B 

 and C would also rise much more sharply during the day because the 

 water will not flow from the absorber into the tank unless the absorber 

 is hotter than the tank. Practically, the tank temperature D shown in 

 figure 5 is what might be expected in a full-sized installation when 

 about 100 gallons of hot water are gradually drawn during the day. 



The rate of heat input to the storage tank is influenced by the) 

 temperature of the tank water, because a high inlet temperature 

 means greater losses from the absorber box. This change, however, 

 is not so important in limiting the maximum storage-tank tempera- 

 ture during nonuse of hot water as is the continual heat loss 24 

 hours a day from the hot tank and absorber to the colder air and 

 surroundings. 



The temperatures inside the absorber box at the top and at the 

 bottom are included in figure 5 to indicate the heat transfer by air 

 convection from the black bottom to the pipes. The minimum box 

 temperature was 4° F. below minimum air temperature and the 

 glass was frosted at sunrise although the pipe did not quite reach 

 freezing. This absorber was exposed during the entire winter of 

 1935-36 at Davis and did not burst although a minimum air tem- 

 temperature of 25° F. was recorded. This, however, cannot be con- 

 sidered safe practice. 



