Heat Test

Unknown macro: {toggle-cloak}

Abstract

In order to confirm the accuracy of the data, the effects of heat energy that is continually added to the settle sample by the turbidimeter light are analyzed for significance. If the temperature gradient is great enough, it causes density currents that resuspend the flocs. It is found that little difference is observed between runs, that heat currents are unlikely a major source of error with the given conditions, and experimental processes do not need to be revised to account for such changes.

Unknown macro: {toggle-cloak}

Introduction

The Spring 2007 Research and Development Team suggested that convection currents created by light heating the water in the sample tube may be affecting the turbidity data in experiments. The turbidimeter measures the turbidity from the scattering of light through the sample of water as a beam of light is passed through the tube. The turbidimeter and its light are currently left on throughout the entire test and could be heating the sample, which may cause convection currents that hinder the sedimentation of the flocs and re-suspend them in solution. This would only affect the effluent turbidimeter data since the solution through the influent turbidimeter is continuously flowing during relevant testing. The purpose of this test is to either rule out convection currents as a concern or to revise future methods to protect against them.

Unknown macro: {toggle-cloak}

Methods

The experimental setup was run using Process Controller software. A Process Controller method was modified so that we can compare the settled turbidity values for the case with the light left continuously on with the case of the light only turned on briefly at the end of the Settle state. It automatically looped through four states: "1- Clean Flocculator", "2- Clean Effluent Turbidimeter", "3- New Flow and Sample", and "4- Settle State." In this first sequence the turbidimeter was left on throughout states 3 and 4 to measure the settled turbidity throughout the first sequence. Then, the method was extended to replicate states one through three as states five through eight. A ninth state was added called "9- Measure Settled Turbidity." State nine was created to allow the turbidimeter to be turned off for the duration defined by set point "Off Time" in state eight and turned on for the "On Time" of two minutes. After the tests are run the data measured during the "On Time" in "9- Measure Settled Turbidity" and the last two minutes of "4- Settle State" were compared.

The raw water flow rate was 2 mL/s. The clay and alum flow rates are determined by the afore-mentioned flow rate, their respective tubing sizes, the stock turbidity or concentration, and a goal influent turbidity which was set at 100 NTU. Clay stock turbidity was set at 2800 NTU and alum stock concentration was 5 mg/L. A 12-inch glass tube was used in this experiment, and PVC pipe was fitted to cover the exposed part of the effluent turbidimeter sample tube to minimize the affects of ambient light in the lab affecting the turbidity readings.

Unknown macro: {toggle-cloak}

Results

The data for the turbidity recorded during the settle state was analyzed. First the day fraction was converted to an elapsed time for each run. The statelog was used to determine the time at which the settle state started and ended. The first data of each settle state was referenced as the start time for the corresponding run. The following day fractions were modified to multiply the difference from the start time by 1440 minutes per day. The elapsed time for each run was 18.16 minutes.

In analyzing the data the first sequence of the cycle where the turbidimeter is on throughout is denoted as run "a" and the second sequence as run "b." Only the last two minutes of the "a" runs were compared to the data from the "b" runs. If convection currents are present we expected to see a significant difference between the "a" and "b" runs in terms of absolute turbidity values and the range of data.
In the first run the data from the "b" run shows a larger spread of data by about 0.1 NTU. It can be seen that the data from both runs is mostly between 1.2 and 1.4 NTU. In the second and third runs the data from both sequences are intertwined. They show similar patterns are have about the same ranges. These results strongly suggest the absence of convection currents.In the last run of the test the "a" run shows a greater spread of data by about 0.3 NTU. This is the opposite of the first run. By inspection you can see that the data from run "a" centers around 1.4 NTU while the data from run "b" centers around 1.2 NTU. This is the same range as in the first run.

Unknown macro: {toggle-cloak}

Conclusion

Where there was a difference between the absolute turbidities between sequences in the same run where the in one the turbidimeter is on throughout and in the other where it is only on at the end the difference was small. There existed no pattern in the differences in range between the runs. It can be concluded that convection currents are not affecting the data recorded in experiments and in the future it will not be necessary to turn off the turbidimeter. Differences in data between and within runs can be attributed to the difference in the timing, size, and orientation of the flocs in solution falling past the turbidity sensor where the size of the flocs created in each run should fall in the same range.

Back to Tube Floc Research