...
The Spring 2010 ANC Control team inherited the original apparatus used by the Fall 2009 team. The original apparatus was two 1" pipes connected at a 45 degree angle, as depicted in Figure 1.
Figure 1
...
The vertical column addressed the issue of feeding the lime but in order to improve the reactor's performance and address some hypotheses, the team built an additional apparatus with an upper slanted segment of 2" diameter instead of 1". By doubling the diameter of the tubing, the design reduces the upflow velocity in that segment by a factor of four. This allows for a lower capture velocity, which addresses the issue of lime being lost with the effluent and allows higher flow rates. Figures 2 and 3 depict the two apparatuses constructed by the Spring 2010 ANC Control Team.
Figure 2
The redesigned apparatus took the original apparatus and added a vertical 1" column by which lime is to be fed.
Figure 3
A second apparatus was built with an almost identical design, only 2" columns for the upper half of the apparatus were implemented.
...
In designing the new apparatuses, the team made the vertical column taller than the diagonal column to ensure that the former, which is open to the atmosphere, would not overflow when water was pumped through. However, the team discovered a problem when it began running experiments. At the end of the effluent tube, there is a candy-cane shaped piece shown in Figure 4 that the water flows through which backs up the flow enough that the pH probe can be immersed. This creates a free surface open to the atmosphere at the elevation of the sink. Basically, the only other free surface in the system exposed to atmospheric pressure is the water in the vertical feeding tube, which should be at a much higher elevation. This creates a siphon which drains the reactor to the elevation of the pH probe, as shown in Figure 5.
Figure 4
Figure 5
The team attempted to correct this problem by sealing the top of the vertical column with a cap. It was hoped that this would create a vacuum above the water in the vertical column and hold that surface at a constant elevation, depicted in Figure 6. However, the cap would not seal well enough to solve the problem, and the solution meant that the reactor would be drained each time the cap was removed for lime feeding.
Figure 6
Instead, the team altered the design by opening the top of the diagonal column to atmospheric pressure, shown in Figure 7.
Figure 7