Nutrient Removal Project: Chemical Phosphorus Removal

Nutrient Removal Project: Chemical Phosphorus Removal

CEE 453

May 11, 2004

Jill Crispell

Stephanie Wedekind

Sarah Rosenbaum

Abstract

Our nutrient removal research involved the addition of metallic salts to determine their effectiveness in removal of phosphorus through the process of precipitation. Our objective was to reduce the influent phosphorus concentration of 4 mg/L to an effluent concentration of 50 g/L. We hypothesized that the removal of phosphorus would increase with increasing metallic salts concentrations. An initial test was done to compare the ability of alum and ferric chloride to remove phosphorus as well as the effect of reaction products on phosphorus removal. We found that alum removes phosphorus more effectively than ferric chloride and that the reaction products do react with phosphorus. In our main experiment, we tested the amount of phosphorus that was removed using four different alum concentrations. We found that as the concentration of alum increased, phosphorus removal also increased. With the concentrations that we used, we were unable to reach our goal of 50 g/L phosphorus in the effluent. However, we still obtained high percent removals. More tests should be done to determine the effect of higher alum concentrations than were used in this experiment, and the effect of using a mixture of metallic salts.

Introduction

We performed a research project in our plant involving chemical phosphorus removal techniques by experimenting with the operation of the plant. The plant was set up to model a clarifier with treated wastewater as the influent, and will test the effectiveness of phosphorus removal by precipitation with metallic salts. We obtained periodic measurements of the phosphorus concentration over time and optimized phosphorus removal by redesigning our plant. We shared information and reagents with the group of Alex Mikszewski, Brett Bovee, and Peter Burns when testing phosphorus levels using the spectrophotometer. We obtained information for our experiment from Lagoon Systems in Maine: An Informational Resource for Operators of Lagoons ( and the lab titled “Phosphorus Measurements” by Monroe Weber-Shirk.

Objectives

Our main objective was to reduce the concentration of phosphorus in the effluent of the wastewater treatment plant by precipitating the phosphorus with varying concentrations of metallic salts. A project criterion was that the total phosphates should not exceed a concentration of 50 g/L (as phosphorus) in a stream entering a lake or reservoir (Weber-Shirk, 2003). This is based on EPA suggestion for phosphorus level discharge. Our hypothesis was that as the concentration of a metallic salt added increases, the concentration of phosphorus in the effluent should decrease from 4 mg/L to an asymptote less than 50 g/L.

Materials and Methods

Setup:

Figure 1: Lab setup.

Our plant is designed to be a tertiary treatment for a wastewater treatment plant. The clarifier is a 6 liter tank (at 4-Liter capacity). A peristaltic pump with a flow rate of 450 mL/min is attached to the tank to pump in tap water, phosphorus solution, and metallic salts. The phosphorus solution is pumped in from the refrigerator because this was the setup from our original plant. The flow rate was determined by measuring the amount of water pumped into a beaker in one minute. The tank is situated on a stirrer, which allows mixing to take place when necessary. An elbow joint is used in the tank to remove as much water as possible during effluent cycles. The plant is automatically controlled by the Process Controller software developed by Dr. Monroe Weber-Shirk of Cornell University. The Process Controller software uses rules, states, and set points developed by our group.

Reagents:

Phosphorus solution: 200 mg/L KH2PO4

Alum solution: 400 mg/L Al2(SO4)3 o 14H2O

Ferric Chloride: 200 mg/L FeCl2

Procedures:

1. An initial experiment was done with each metallic salt to determine if the products of the reactions would continue to react with phosphorus, further removing it.

• In each cycle of the process, 125 mL of the phosphorus solution was added to the tank with 4 L of tap water, to make the phosphorus concentration in the tank 4 mg/L.
• In only the first cycle of the process, 125 mL of the alum solution or 136 mL of FeCl2 was added to bring the concentration in the tank to 12.5 mg/L alum or 6.8 mg/L FeCl2. In order to pump the correct amount of metallic salts into the clarifier, a code was written in LABVIEW which took inputs of pump flow rate and volume of metallic salts to be added and gave an output of the amount of time that the peristaltic pump needed to pump the metallic salts into the clarifier to obtain the desired concentration.
• After the water and reagents were added to the tank, the mix was stirred in the reactor for 2 hours.
• After mixing, the tank settled for 1.5 hours to allow the phosphorus and metallic salt products to precipitate.
• Next, the effluent drained, leaving a small amount of water, 1.2 L, and the precipitates in the bottom of the tank. The effluent sample was caught in a 125 mL bottle in the drain.
• After the first cycle in which metallic salts were added, 4 more cycles were run, without adding metallic salt, only adding water and phosphorus.
• After taking these samples, we analyzed them using the spectrophotometer as described in Phosphorus Measurement. The results of this experiment were used to determine the parameters for the main experiment.

2. In the main experiment, different concentrations of alum were used to determine which concentration removes phosphorus most efficiently. The setup was the same as in the first experiment except that the tap water was coming out of our own separate jug of tap water instead of the large tank that the rest of the class was using. This was because we discovered after performing the first experiment that the class’s water contained ferric chloride. The same states were used as in the first experiment, except a rinse state was added at the beginning of each cycle to clean out the remaining phosphorus and salts. In the rinse state, the tank filled to 4 L with tap water, then was drained. .

• 125 mL of the phosphorus solution was added to the tank with 3.86 L of tap water, to make the phosphorus concentration in the tank 4 mg/L.
• The concentrations of alum tested were 100 mg/L, 125 mg/L, 150 mg/L and 250 mg/L. Three cycles were done using each of these concentrations and an effluent sample was taken from each cycle.
• The samples from this test were analyzed using the spectrophotometer to determine the amount of phosphorus remaining in the effluent.

Results and Discussion

Typical values of phosphorus in wastewater range from 0.5 mg/L to15 mg/L with an average of 4.1 mg/L in a study in Michigan or 3.3 mg/L in a study in Minnesota (Lopez 2004). We decided that a phosphorus concentration of 4 mg/L would be reasonable based on these values.

We determined the amount of metallic salts necessary to add to the plant for complete reaction with the phosphorus using chemical equations (Lopez 2004). Utilizing stoichiometry, we found from Equation 1 that 12.5 mg/L of alum was needed to react with the 4 mg/L of phosphorus. From Equation 2, we found that 6.7 mg/L was needed to react with the 4 mg/L of phosphorus.

Al2(SO4)3 o 14H2O + 2PO43- 2AlPO4 + 3SO42- +14 H2O 1

FeCl3 + PO43- FePO4 + 3Cl-2

From our first experiment, we measured the concentration of the phosphorus using the spectrophotometer. The initial samples were treated with the metallic salts – either alum or ferric chloride – to determine which salt was more effective in phosphorus removal in our plant. After the first run of the respective alum and ferric chloride tests, no more metallic salt was added to the initial experiments. The samples with the respective metallic salts added had a lower concentration than the samples without metallic salts.

Since the phosphorus concentration in the plant was 4 mg/L, we expected the untreated samples to have concentrations near this concentration. However, this was not the case for either the alum or the ferric chloride. The alum made a significant reduction in the concentration of phosphorus, but samples without additional metallic salts varied (Figure 2). The ferric chloride did not have a large effect on the phosphorus concentration (Figure 3). Ferric chloride is most effective in removing phosphorus when the pH is in the range of 4.5 to 5.0 (Lopez 2004). Alum is most effective when the pH is in the range of 5.5 to 6.5 (Lopez 2004). The pH of our plant was higher than either effective metallic salt range at an approximate pH between 8.3 and 8.4. This is closer to the optimum pH range for alum than for ferric chloride. This could in part explain why alum was more successful than ferric chloride in our initial experiment.

Figure 2. Concentration of Phosphorus in the Plant Effluent vs. Alum Sample.

Figure 3. Concentration of Phosphorus in the Plant Effluent vs. Ferric Chloride Sample.

However, after these tests were taken, it was discovered that the influent water contained ferric chloride. This contributed to the varied results for the alum samples because the ferric chloride was reacting along with and then without the alum to remove phosphorus in the first test. To alleviate this problem, we pumped tap water into the plant from a large jug separate from the water supply for the other labs. We narrowed our research to only use alum, as ferric chloride did not seem to effectively remove phosphorus in our preliminary testing.

The products of the alum and phosphorus reaction continued to react with the influent phosphorus (Figure 2). Thus the concentration of alum is not constant or consistent with the concentration added to the plant. Similarly, the products of ferric chloride reaction continued to react with the influent phosphorus (Figure 3).

In order to confirm all of the products from the previous cycle were removed from the tank, we changed the cycles of the plant to include a rinse and a rinse effluent cycle that rinsed and cleaned the tank. The effluent was also removed from the bottom of the tank using an elbow joint to empty as much water as possible before the next cycle.

In the main experiment, we compared the effects of differing concentrations of alum for precipitating the phosphorus (Table 1).

Table 1. Phosphorus Concentration of Sample Effluents

sample / phosphorus conc. mg/L
1: 100 mL alum / 1.35
2: 100 mL alum / 1.32
3: 100 mL alum / 1.15
average / 1.27
standard deviation / 0.09
percent removal / 68.17
1: 125 mL alum / 1.16
2: 125 mL alum / 1.1
3: 125 mL alum / 1.05
average / 1.10
standard deviation / 0.04
percent removal / 72.42
1: 150 mL alum / 1.16
2: 150 mL alum / 0.836
3: 150 mL alum / 1.1
average / 1.03
standard deviation / 0.14
percent removal / 74.20
1: 250 mL alum / 0.542
2: 250 mL alum / 0.596
3: 250 mL alum / 0.954
average / 0.7
standard deviation / 0.22
percent removal / 82.57

We found that as the alum concentration in the plant increased, the phosphorus concentration decreased and percent removal increased based on samples with 100 mL, 125 mL, 150 mL and 250 mL of alum respectively. These results are graphed in Figure 4. Although we did not meet our goal of 50 g/L, our high percent removal indicates that the alum is effective. In our last test with 250 mL alum, sample 3 appears to be an outlier. This may be due to the fact that this sample was warm and recently taken, while samples 1 and 2 had been refrigerated and sitting for a number of hours.

Figure 4. Average Concentrations of Phosphorus in Effluent Relating to Concentration of Alum Added

Generally, the trend appears to be linear and decreasing. More testing is necessary to determine whether or not the data may reach a minimum and become asymptotic.

When setting up the clarifier, we kept the setup the same as it had originally been when we were using it with activated waste to remove BOD. We discovered that the water we were pumping into the plant had FeCl2 in it, which reacts with phosphorus. We changed the setup so that the water being pumped into the plant was pumped from a large jug of tap water instead of from the tank. With the previous setup, the activated sludge was left at the bottom of the tank, and we tested to see if this would work with the precipitates. However, this did not work because the products from the first cycle continued to react with the phosphorus, thus making the actual concentration of metallic salts in the clarifier very difficult to measure. To alleviate this problem, we installed an elbow joint in the effluent hole in the tank, which drained almost all of the water during the effluent cycles. We then designed a different rinse and rinse effluent cycle to clean and drain the clarifier. This efficiently removed all of the products from the previous reaction, so we were able to accurately control the concentration of the metallic salts.

If we had more time to work on this experiment, we would have liked to do more experiments using higher concentrations of alum. If higher concentrations were tested, the results would probably be more asymptotic, and our goal phosphorus concentration of 50 g/L may have been attained. Also, we would like to do an experiment that tests the phosphorus removal using a mixture of alum and ferric chloride. Using the two metallic salts in combination might be a good solution in a wastewater treatment facility because their abilities to remove phosphorus vary with pH. Because the pH of wastewater is not always the same, a mixture of the two salts might achieve higher phosphorus removal in a wider range of pH values.

Bibliography

Lopez, Ernesto and Charles Pycha. Lagoon Systems in Maine: An Informational

Resource for Operators of Lagoons. Technical Support Section Water Compliance Branch U.S. EPA 5WCT-15-J. 14April. 2004. (

Weber-Shirk, Monroe. Phosphorus Measurements. 2003