h1. Chemical Dose Controller
h2. Abstract
At the Cuatro Comunidades plant, the [chemical dose controller (CDC)|Linear Chemical Dose Controller] is used to set the plant's alum dose. 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 CDC at the plant was monitored and improvements were made to the original model.
{float:right|border=2px solid black|width=600px}
!CDC.jpg|width=600px!
h5. Figure 1a. CDC schematic 1b. Original Cuatro Comunidades CDC
{float}
\\
h2. Introduction and Objectives
The original CDC used in the Cuatro Communidades plant was the first generation of the technology used by AguaClara, (Figure 1). The CDC is made up of three components--- a flow control module (FCM), lever arm and a sutro weir. The alum flow rate, and therefore dose, is set by the available head between the FCM inlet valve and outlet hose. The available head is changed by moving the outlet hose along the lever arm. Once the hose is positioned on 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 which is determined by the sutro weir. 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 in selecting 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. After observing the original CDC, an improved model was constructed. The theoretical and actual flow rates for both CDCs are compared below.
h2. Results and Discussion
h3. Original CDC
{float:right|border=2px solid black|width=400px}
!OriginalCDCGraph.gif|width=400px|height=350px!
h5. Figure 2
{float}
The alum flow rate of the original CDC was measured every time the operator changed the dose. The measurement was taken by filling a graduated cylinder for thirty seconds. As can be seen in figure 2, the actual dose delivered by the CDC was highly variable close to the lever arm's fulcrum was consistently under dosing further along the lever arm. The CDC was consistently under dosing away farther away from the fulcrum. 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. The range of doses provided by this CDC was appropriate for most incoming turbidities (<200 NTU) to the plant. However, the dose was not adequate for the high turbidity raw water coming to the plant after a rainstorm. Additionally the variability in the dose made it impossible to associate a particular distance along the lever arm with a specific alum dose.
{float:left|border=2px solid black|width=300px}
!la_attachment.jpg|width=300px!
h5. Figure 3: Attachment of FCM hose to PVC lever arm
{float}
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. 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 3). The notches on the slit meant the operator could only choose incremental doses. 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. 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 4).
{float:right|border=2px solid black|width=350px}
!1lawater.jpg|width=350px!
h5. Figure 4: The end of the lever arm is almost submerged
{float}
Additionally, the sutro weir often clogged with leaves if it was not constantly monitored and cleaned. Leaves and grit blocked the open surface area of the sutro weir and increased the height of water in the entrance tank when there was not change in the plant flow rate. 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.
h3. Aluminum CDC
{float:left|border=2px solid black|width=400px}
!aluminumleverarm.jpg|width=400px|height=350px!
h5. Figure 5: Aluminum CDC
{float}
A new lever arm was constructed and monitored. The lever arm was made of a rust resistant bar of aluminum (Figure 5). The FCM outlet hose was slid along the lever arm a fixed in place by a screw. This sliding system gives the operator more flexibility when choosing an alum dose. 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 6).
{float:border=2px solid black|width=600px}
!measurementdevice.jpg|width=600px|height=350px!
h5. Figure 6: Flow rate measurement device
{float}
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 7.
{float:border=2px solid black|width=600px}
!aluminumgragh.jpggif|width=600px|height=350px!
h5. Figure 7
{float}
The data in figure 7 was taken after the FCM hosing had been cleaned and was taken over the span of several hours rather than several weeks as was the case in figure 2. Each actual dose point represents the average of three trials. The results show that the dose is increasing linearly with plant flow rate as expected. However, the CDC was still incapable of reaching the higher end of the design flow rates. The slope of the actual doses measured with the aluminum CDC is closer to the theoretical line than the slope of the actual doses given by the PVC model. The slope could be increased even further by shortening the FCM outlet hose. The length of the hose controls the head loss in the system. Decreasing this head loss would increase the maximum flow rate. Another way to increase the maximum dose delivered by the CDC would be to lengthen the lever arm.
From figure 7, it is impossible to say whether the variability in the chemical dose at each location on the lever arm improved. The system used to measure the flow rate was more precise. The data was only taken at one plant flow rate and the system had just been cleaned. The data taken with the original PVC lever arm was taken at multiple plant flow rates. However, the original data showed a low correlation coefficient, 0.29, between percent error in the dose and the plant flow rate. Where percent error was calculated by:
An the correlation coefficient was found in using the Excel CORREL function which uses the equation:
{float:right|border=2px solid black|width=600px}
!hosefitting.jpg|width=600px|height=350px!
h5. Figure 8: Connection of hose to lever arm
{float}
Given this low correlation coefficient it is unlikely that the plant flow rate has a significant effect on the error in dosing.
A final adjustment was made after this test to reduce variability in the dose--- the hose was directly attached to the lever arm and the short aluminum bar was removed (figure 8). 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. This will help maintain a constant dose at each lever arm 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.
h2. Conclusions
The connection between the hose and the lever arm should be rigid.
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).
h2. Conclusions
The connection between the hose and the lever arm should be rigid.
The hose needs to be changed to reach correct maximum dose (or longer lever arm)
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. | 