...
The vertical flow hydraulic flocculator has been used to test tapered and uniform baffle configurations during previous semesters, but the two configurations have never been compared with similar raw water conditions. For this reason the Fall 2008 Pilot Plant Flocculation team focused on determining which flocculation configuration is optimal. The tapered flocculation configuration model is based on the G-theta Model, while uniform configuration is based on the energy dissipation model. It has recently been determined that the Gtheta model used to design the tapered flocculator is inaccurate. Gtheta describes mixing in the viscous subrange. Mixing in a turbulent flow hydraulic flocculator is better modeled by energy dissipation rate in the flocculator. The energy dissipation model suggests that the difference in performance between tapered and uniform flocculators might not be that large. The uniform versus tapered flocculator tests were performed to confirm this theory. In our final results, the tapered baffle configuration produced marginally better turbidity than the uniform baffle configuration. However, due to low temperature conditions, the incoming turbidity was too low in order to produce conclusive results.
Introduction and Objectives
Methods
Results
Several tests have been run with the uniform baffle configuration. The initial results were inaccurate due to a clog in the inlet valve of the raw water sampling line. The following graph is the turbidity profile obtained on 10/20/08.
Recently development in Aguaclara technology have shown that the Gtheta model may be inappropriate to describe flocculation. A main goal of the pilot plant flocculation team was to determine the optimal flocculation configuration: Tapered or Uniform Baffle Spacing. Using the Gtheta model a tapered configuration would expose flocs to high shear initially and as flocs grew they would be exposed to less shear. Using an energy dissipation model, the the tapered configuration still would be optimal, but the difference in performance is predicted to be small. However, it should be noted that the energy dissipation model analysis was based on the assumption of constant floc density. Changes in that assumption to a more realistic model could significantly alter the model predictions.
The Pilot Plant flocculation team approached this problem by first determining the optimal spacing for the tapered flocculation configuration. It was then determined that the two configurations would only be comparable if the uniform spacing was equal to the last section of the tapered spacing. The tapered and uniform spacings were compared consecutively. The tests were then run in the same day to ensure similar environmental conditions and comparable data.
An important aspect of the vertical flow hydraulic flocculator is the alum dosing One of the main issues of this semester was determining appropriate alum dosing for cold water. This problem was approached by inquiring the staff at the Water Filtration Plant about their Polyaluminum Chloride dosing. Their dosage was then converted so that the Pilot Plant can insure proper dosage with a simple equation based on the plant's dosage.
Sampling Method
Overview
This sampling method in the flocculator at the Pilot Plant has been used for several semester's research. Alum is used as a coagulant and the dose can be set using process controller. Raw water running through flocculation tank can be sampled at any desired location. Sampling lines run to turbidimeters from three mobile tube settlers. The tube settlers mimic sedimentation. Additionally a sampling line runs from the raw water inlet to a turbidimeter. Turbidity data is stored in a excel spreadsheet using process controller.
Alum Dosing
Because the pilot plant takes water directly from Fall Creek, environmental conditions change all the time and affect the incoming turbidity to the plant as well as the chemical composition of the particles causing turbidity. It is incredibly important to determine the best alum dose for each test to ensure the formation of good flocs and collect appropriate data. The alum dose can also be related to the plant's PAC dosage using the relationship 1 part per million is equivalent to 0.182 mg/L.
Tube Settlers
Tube Settlers are used to mimic sedimentation before the turbidity of water in the flocculator is measured. They can be moved to different locations along the flocculator in order to examine effluent turbidity at various stages of flocculation.
Data Collection and Analysis
Data is collected using process controller. It is stored in excel spreadsheets and then analyzed.
Results
The team had difficulty finding the correct alum dose at low temperatures and incoming turbidity this semester. The team attempted to find an appropriate relationship between Alum and PAC dosing. This relationship would have helped the team take advantage of the plant operators alum dosing expertise by converting current PAC doses to alum doses when running a test. Unfortunately the molecular formula of the PAC product used at the WFP is a trade secret, so a molar ratio of aluminum in PAC and alum was not determined.
The best dosages for given tests were much lower than initially expected. At average incoming turbidity of 2.3 NTU, it was determined that the optimal alum dose that day would be around 4 mg/L (Figure 1). As can be seen in the figure below this alum dose gave the best effluent quality. The differences in NTU between 4 mg/L and 9 mg/L are about 0.3 NTU.
Wiki Markup |
---|
{float:left}
h6. Figure 1:
!Alum.JPG|width=500!
{float} |
The configuration of the uniform and tapered flocculator was changed to agree with the energy dissipation model of flocculation. Previously the end of the tapered flocculator the baffle spacing was 15 cm, this resulted in a energy dissipation lower than the recommended value of 0.4 mW/kg. In order to achieve the recommended value of energy dissipation the spacing in the flocculator was changed to 13.2 cm. This value was calculated using the equation:
Include Page | ||||
---|---|---|---|---|
|
Where epsilon is the energy dissipation (7.5 mW/kg), Q is the flow rate (see below), b is the baffle spacing (see below), w is the width of the flocculator (30.5 cm), pi cell is the ratio of length to width of the zone of energy dissipation in the flocculator (2), and Cp is a loss coefficient for a 180 degree turn (3).
The spacing of the uniform flocculator was altered to match the last section of the tapered flocculator so that floc break up would not occur due to shear in the uniform flocculator. The resulting energy dissipation values for the tapered flocculator are based on a flow rate of 100 L/min:
Section | Spacing(cm) | Number of Baffles | Energy Dissipation (mW/kg) |
---|---|---|---|
1 | 3.4 | 2 | 98.7 |
2 | 3.9 | 4 | 50.3 |
3 | 5.8 | 5 | 10.9 |
4 | 10.1 | 11 | 1.2 |
5 | 13.2 | 11 | 0.4 |
A turbidity profile for the uniform and tapered flocculator was determined using this alum dose. The results are inconclusive- neither configuration produced an effluent with a lower turbidity than the raw water entering the flocculator. Furthermore in the graph shown below the flow rate was decreased to 51.5 L/min producing an energy gradient of only 0.055 mW/kg throughout the uniform flocculator (Figure 2). The residence time of both flocculators at this flow rate was 15 min.
Wiki Markup |
---|
{float:left|border=2px solid black}
h6. Figure 2:
!Turbidityprof.jpg|width=500!
{float} |
Conclusions
The turbidity profiles along the flocculator show little difference between the uniform and tapered configuration. Neither configuration shows a better effluent turbidity than the raw water turbidity. The tapered flocculator performed slightly better than the uniform flocculator, this is to be expected as more reasonable energy gradients were occurring in the tapered flocculator. Although the energy gradients are extremely low, clearly flocculation is occurring as seen by the initial increase in turbidity. This increase shows that flocs too small to settle out but large enough to increase the turbidity of the flocculator were forming. In addition, these flocs began to increase in size and number and proceeded to settle out.The accuracy of this result is questionable. The tank had to be cleaned before the measurement could be taken. Cleaning may have increased turbidity in the last section of the flocculator. We observed an incoming turbidity of 1 NTU, which is comparable to the raw water that enters the CUWFP. Residual sludge iin the end of the tank could be contributing to the final turbidity creating some error. However a decrease in turbidity can been seen as water travels through the flocculator indicating more flocs are settling later in the flocculator.
The team will develop a more comprehensive profile for the uniform baffle configuration, sampling turbidity at more locations in the flocculator. A comprehensive turbidity profile must also be created for the tapered baffle configuration.Fall 2008Fall 2008