Chemical Dose Controller

At the Cuatro Comunidades plant, the chemical dose controller (CDC) adds alum to raw water in the entrance tank. The appropriate alum dose varies based on two factors, the incoming turbidity and the plant flow rate. The CDC is designed to maintain a constant dose regardless of changes in the plant flow rate.
The original CDC used in the Four Communities plant was the first generation of the technology used by AguaClara, (Figure ###). The CDC is made up of three components— a flow control module (FCM), lever arm and a sutro weir (FIGURE ###). The alum flow rate is set by the difference in height between the FCM inlet valve and outlet hose. This flow rate is constant regardless of the available head in the alum stock tank. The sutro weir is used to create a linear relationship between plant flow rate and the height of water in entrance tank. The lever arm unites the first two components of the CDC. The outlet hose of the FCM is connected to one end of the lever arm while a float balances the opposite end of the fulcrum. The float height varies with the water level in the entrance tank. When the float rises, the opposite end of the lever arm with the FCM outlet hose falls increasing available head in the FCM and therefore the alum dose. The coordination of plant flow rate and alum dose created by the CDC decreases the number of variables the operator must deal with to choose an appropriate alum dose.
Over the course of the summer, the alum flow rate was measured and the CDC system was monitored to detect any failure modes. The alum flow rate was every time the operator changed the dose by filling a graduated cylinder for thirty seconds. The results are shown in figure ###.

The dose controller showed great variability close to the fulcrum and was consistently under dosing further along the lever arm. Thirty centimeters away from the fulcrum the dose is supposed to be its highest 60 mg/L. However, the CDC was only dosing an average of 35 mg/L. This dose should not be adequate for turbidities above around three hundred NTU incoming turbidity. Additionally the variability in the dose made it impossible to associate a particular distance along the lever arm with a specific alum dose.
Some of this variability is inevitably due to the fact that the measurements were taken on different days and the hose may have been clogged when the measurements were taken. However, several defects in the lever arm design were also noted, and a new model was made. The original design consisted of a PVC lever arm with notches cut into it for each chemical dose (side view, figure ###). The dosing hose was moved up and down the lever arm by a loop of fishing line and a knot that could be attached to each slit, (figure ###). The notches on the slit meant the operator could only choose incremental doses rather than flexibly choosing a desired dose. The alum dose at each notch should have been constant. However, the fishing line connection could be adjusted within each notch so the dose was not consistent with location. Another observed issue was that the hose was slightly buoyant. This became an issue when the water level in the entrance tank was high. At low doses, the hose had enough room the hang freely above the water level but at high doses the hose had the tendency to float decreasing the distance between the hose outlet and the lever arm. Furthermore the lever arm was not located at a sufficient height in the entrance tank. At high flow rates the end of the lever arm was submerged, (figure ###). Additionally, the sutro weir often clogged with leaves if it was not constantly monitored. Leaves and grit blocking the surface area of the sutro available for water to flow through increased the height of water in the entrance tank. Because the alum dose was based on the height of water in the entrance tank and not directly on the plant flow rate, the alum dose increased as well even though the plant flow rate did not changed.

A new lever arm was constructed and monitored. The lever arm was made of a rust resistant bar of aluminum (Figure ###). The FCM outlet hose was slid along the lever arm a fixed in place by a screw. This system increased the flexibility of available alum doses. The hose hung from a short aluminum bar. The difference between theoretical and actual chemical flow rates was measured for this lever arm as well. A more accurate method was implemented to obtain the flow rate data. A clear PVC pipe was inserted between the chemical stock tank and the flow control module (Figure ###). To measure flow rate, the first valve from the stock tank to the PVC pipe was opened and the PVC pipe was filled. The line bypassing the PVC pip was then shut off and the time for the cylinder to empty to obtain the actual flow rate was measured. The results of measured flow rate before and after improvements were made to the chemical doser can be seen in Figure ###

This data was taken after the hosing had been cleaned and was taken over the span of several hours rather than several days. Each actual dose point represents the average of three trials. The results show that the dose is increasing linearly with plant flow rate. However, the CDC is still not capable of dosing at high turbidities. From this data it is impossible to say whether the variability in the chemical dose improved. The system of measuring the flow rate was more precise. The data was only taken at one plant flow rate and the system had just been cleaned. NEED LONGER HOSE...
A final adjustment was made after this test— the hose was directly attached to the lever arm and the short aluminum bar was removed (Figure ###). The short aluminum bar could be attached to the lever arm at an angle changing the distance between the hose and the FCM inlet. The change was made to fix the distance between the FCM inlet and outlet hose at each position. Additionally, the change increased the distance between the hose and the water level in the entrance tank so the hose will not be submerged at high plant flow rates.
Additional Improvements
Change in stock concentration over time was not accounted for in measurements of expected versus actual alum dose. The operator mixes the stock multiple times daily but some alum settles out of the solution in between mixing (Figure ### comparison of alum in tubes). The issue with leaves and grit can be improved by better monitoring, an new configuration of larger holes, a prescreening system before the sutro weir.