Floating Flocs Team Summer 2009 Research Proposal

Introduction and Theory

There is a problem in some AguaClara water treatment plants with flocs floating to the surface of the water in sedimentation tanks. One hypothesis is that the water entering the sedimentation tank is supersaturated, and because total dissolved gas pressure is greater than the local solution pressure, gas comes out of solution and forms bubbles. The bubbles form on floc particles and bring them to the surface, causing the flocs to float instead of settling to the bottom of tank. Since flocs are swept out with the "clean" water and escape the sedimentation tanks, they eventually pollute the effluent water. As a result, the quality of the drinking water produced by the plant is degraded. One solution may be to send the water through a sand filter before it enters the sedimentation tanks to remove excess gas in the water, thereby preventing bubble formation on flocs.

In the treatment plants, water would be sent upward through a bed of sand suspended in the grit chamber. The sand filter is intended to provide surface area on which bubbles can form. The bubbles would accumulate, forming larger bubbles that could quickly float to the surface of the water in the grit chamber. The anticipation is that enough bubbles can form and leave the water in the grit chamber so that the water flowing into the sedimentation tanks would have a lower dissolved gas content and less potential to form bubbles.

Previous research has found that smaller grain sizes are more effective in gas removal than are larger grains in the same volume of sand, since small grains provide more surface area per unit volume on which bubbles can form. Other parameters, such as optimal bed depth and bed expansion, will be determined experimentally. Additionally, since dissolved gas content fluctuates in natural bodies of water, such as the sources of water for AguaClara plants, the dissolved air content of the influent water into the experimental system will be varied to explore its effects on gas removal. The purpose of this team is to test a variety of parameters (listed below) to design an effective sand filter that can be implemented in AguaClara plants.

Parameters

Size of sand grain

Research performed last semester focused on studying the effects of sand grain size on gas removal. We found that smaller sand grains provide more surface area per unit volume than larger grains, but grains that are too small are carried to the surface with bubbles and are washed away. Grain size also affects pore size of the sand bed, which affects the distance a gas molecule must diffuse to reach a surface. This also may affect aggregation of gas molecules to form larger bubbles. Further study of optimum grain size will be explored through MathCAD modeling, or through experimentation, if that is deemed necessary.

Bed depth (i.e., height of sand column before expansion occurs)

A range of depths will be explored to determine the relationship between bed depth and gas removal. Experiments have shown that increasing bed depth increases gas removal. Future experiments will explore the exact relationship to determine whether gas removal will increase with bed depth or whether gas removal levels off at some range of bed depth, etc.

Bed expansion (which depends on the flow rate of influent water)

The expansion of the sand bed affects its porosity, which affects the distance a gas molecule must diffuse to reach a surface, and also may affect aggregation of gas molecules to form larger bubbles. We would like to see the effects on gas removal using different expansions while holding other parameters constant.

Dissolved air concentration

We would like to determine whether effectiveness of gas removed depends on the concentration of dissolved gases in the influent water. We plan to adjust the amount of supersaturation by varying pressure in the aerator.

Experimental Design

Experimentation will be conducted using a new sand filter setup modified from the Spring 2009 setup. A description of the set-up and the Process Controller methods used to conduct experiments can be found here.

Experiment: Bed Depth

The Spring 2009 Floating Flocs team conducted some preliminary research into the effects of varying bed depth (i.e., the amount or height of sand in the sand column). An experiment showed that a greater bed depth increased gas removal. This result was expected, as a greater surface area of sand should allow more bubbles to form and accumulate. We would like to conduct further research into the effectiveness of various bed depths to model the relationship between bed depth and gas removal.

Using the setup detailed above, we plan to conduct this experiment by beginning with a small amount of sand in the column and running the apparatus on the "On" state (described in the description of the system here) in Process Controller to measure and record gas removal. We would repeat with increasing depths of sand until a relationship can be clearly defined. We plan to test bed depths within the range of 10 cm - 50 cm of Sand 40 (grain diameter 0.42 - 0.59 mm) with 50% expansion.

Research conducted in Spring 2009 suggested that gas removal increases with bed depth, but the sample was too small to make any definite conclusions about the relationship over a larger range. We predict that very small bed depths will be less effective at gas removal. As depth increases within some range that is yet to be determined, gas removal probably will increase. At very large bed depths, we believe, the amount of gas removal will begin to taper off; as the concentration of gases in solution decreases, fewer bubbles will form, and the rate of change of gas removal will decrease.

Experiment: Bed Expansion (and Flow Rate)

This experiment will attempt to examine the fluidized bed expansions for a filter media under varying flow rates, with other parameters held constant. We will adjust bed expansion by adjusting the flow rate through the sand filter. Expansion of a bed of sand affects the distance that the gas molecule has to undergo to reach a solid surface. Consequentially, this affects the bubble formation process. By conducting this experiment, the team hopes to quantify the relationship between the bed expansion at a certain flow rate and gas removal to find optimal conditionals of bed expansion in the sand filter. Presently, we are considering using Sand 40 with a grain diameter 0.42 - 0.59 mm. The range of expansions to be tested will be determined after the bed depth experiments are conducted. A bed depth that yields moderate gas removal will likely be chosen and the physical limitations of the sand column will govern the expansion range.

Our hypothesis is that if the bed expansion is too low, the gas bubbles might not be able to form properly or might be trapped inside the sand bed. On the other hand, we think that if the bed expansion is too high the gas removal rate may begin to decrease as a result of the increased distance that the gas molecules would have to travel to diffuse.

Experiment: Dissolved Gas Concentration

Because of seasonal changes in temperature, the dissolved air concentrations of influent water at AguaClara plants may not be consistent throughout the year. Because of this, we are interested in determining if there is a relationship between the dissolved gas concentration and the rate of gas removal via the sand filter. This experiment is designed specifically to measure the changes in the gas removal rate of the sand filter as a direct consequence of changes in the dissolved gas concentration of the influent water.

For this experiment, we will assume that the residence time in the new aerator will allow the dissolved gas concentration to equilibrate with the pressure maintained in the aerator. Since dissolved gas concentration is a function of pressure, the pressure in the aerator will be adjusted to achieve different dissolved gas concentrations in the influent water while keeping all other variables constant. It is likely that the range of pressures tested will be between 20 kPa - 100 kPa. The pressure is adjusted within the aerator by adjusting the "Aerator Min Air Pressure and Aerator Max Air pressure setpoints" in the process controller configuration file. Presently, we are considering using Sand 30 with a diameter of 0.59 - 0.84 mm with a bed depth of 30 cm and 50% bed expansion.

The initial experiment will be run with the aerator pressure set to 100 kPa. The experiment will be left to run for a day to observe and record the behavior of dissolved gas removal and to determine an appropriate runtime for subsequent experiments. (This will be determined by the time it takes for the bubble collector to refill a certain number of times)

The current relationship between the volume of gas removed over time considering sand grain size or bed depth is found to be fairly constant. When the dissolved air concentration of the water is adjusted, the gas removal rate probably will be the same initially, then will decrease or level off as the gas concentration in the water reaches saturation at atmospheric pressure. This hypothesis is based on the notion that water that is not supersaturated with respect to the local absolute pressure and the gas composition of potential bubbles will not form bubbles. We predict that, as the concentration of gases decreases, fewer bubbles will form and fewer will grow large enough to float away from the sand filter.

Additional Tasks

A mathematical model of the system suggests that 18 ml of gas should be able to be removed by the sand filter system. However, experiments performed last semester indicated a removal of 5.09 ml/L with Sand 40 (grain size 0.42 mm - 0.59 mm) and even lower removal with larger sand grains. While it is likely that the conditions of the sand filter may not be suitable to remove 18 ml/L, it is also likely that other components of the system are not functioning ideally. It was assumed that the dissolved gas concentration in the aerator had enough time to equilibrate with the 2 atm of absolute pressure in the apparatus; however, that may not be a completely valid assumption. Additionally, the bubble collector may not be collecting all the bubbles that leave the sand filter.

In order to better understand what is happening in the system in term of gas removal, we plan to measure the oxygen concentration in the water before and after the pressurized aerator using a dissolved oxygen probe. Additionally, the oxygen concentration will be measured before and after the sand filter and the bubble collector. The values collected will be compared to the theoretical model to better assess the functionality and efficiency of the system.