Coagulant dose as an operating parameter in a solid contact clarifier

Solid contact clarifiers are widely used in water treatment works worldwide due to their outstanding performance in solid removal and minimum area requirement. These solid contact clarifiers use the principle of facilitating flocculation, sedimentation, and clarification in a single unit. Mixing, internal solids recirculation, gentle flocculation, and gravity sedimentation are all combined into the single clarifier unit.

Different mechanisms are used to achieve particle removal in clarifiers. Clarifiers that include a mixing zone and a clarification zone is one type. Particle contact and floc formation are achieved in the mixing zone, and flocculated particles are settled in the clarification zone. The mixing zone is designed with a conical shape, where the upward flow of raw water will have a decreasing mixing intensity (G), enabling hydraulic flocculation. In some other clarifiers, flocculation is achieved using mechanical mixing devices that are placed in the mixing zone.

Pulsator^® is another type of clarifiers where the coagulated water in the inlet chamber is subjected to a periodic up and down motion (pulse) by the action of a vacuum pump. Water flows through a perforated pipe system out of the inlet chamber to the Pulsator^® tank. A motion opposite to the inlet chamber occurs in the Pulsator^® tank. A sludge blanket is formed in the tank, which enables the required flocculation and solids contact. Suspended particles contained in the raw water are agglomerated in the sludge blanket (Binnie et al. 2002).

Flocs created by agglomeration of smaller particles enables gravitational settling of particles. The type and dose of coagulants to be used in clarifiers depend on the raw water quality and the ambient conditions. The optimum coagulant dose is the lowest (least cost) dose which will produce a readily settleable floc to remove turbidity efficiently in a reasonably short time, remove excess color from the water, and have suitable filterability properties. Thus, finding the optimum coagulant dose will ensure best performing cost-effective clarifier operation.

Su et al. (2004) found that coagulant dose would affect the sludge blanket height in a clarifier. The response time and the optimal operating condition of the blanket that produces the lowest effluent turbidity corresponds to the optimal coagulant dose, determined by the Jar Test.

Hurst et al. (2010) studying the variables affecting the performance of a floc blanket using a laboratory-scale reactor showed that depending on the raw water turbidity, an optimum coagulant (alum) dose can be found. For the range of doses tested in the experiment, they have shown that under-dosing of alum could decrease the size and the amount of flocculated particles entering the floc blanket, whereas overdosing showed little effect on the floc blanket performance. The settling velocity of the coagulated flocs and the upward flow velocity would control the stability of the blanket. Stringent operational control is required to prevent sludge carryover (Gregory et al. 1996; Head et al. 1997).

Chen et al. (2003) reported that when the changes in solid concentration, zeta potential, floc size and capillary suction time were monitored, a relatively stable blanket to the shock load in raw water turbidity was observed when the turbidity was high (>100 NTU). When the raw water turbidity was low (<10 NTU), the blanket was rather unstable. With varying raw water quality, appropriate coagulant dose shall be used to get the desired effluent quality. They have reported complete washout when the poly aluminium chloride (PACl) dose was insufficient.

Previous studies have shown the importance of finding the appropriate coagulant dose. As explained above, the laboratory flocculation test, Jar Test, is the most commonly used method to find the optimum coagulant dose to be used, be it plain sedimentation, clarifiers with hydraulic mixing or clarifiers with mechanical mixing or Pulsators^®. However, the flocculation tests are not sufficient to transpose the results to a full-scale level where continuous changes to raw water quality occur due to seasonal as well as diurnal changes.

Degrémont (2007) has defined cohesion as a parameter which characterizes the sludge blanket. A sludge layer submitted to an upward water current expands and occupies an apparent volume roughly proportional to the upward velocity of the water entering the blanket, according to a ratio that characterizes the cohesion of the sludge blanket. The Sludge Cohesion Coefficient (SCC) is an indicator of blanket condition and serves as a check for optimum coagulant dose. The value of SCC is expected to be between 0.8 and 1.2 (2.9–4.3 m/hr) for quickly settled consistent sludge and less than 0.3 (1.1 m/hr) for flocculate that are fragile, light and rich in water (Degrémont 2007).

The objective of this study was to determine the applicability of the Jar Test results in solid contact clarifiers and give recommendations on determining the optimum coagulant dose. The Sludge Cohesion Test (SCT), which determines the cohesivity of the sludge blanket, and the Jar Tests were conducted for raw water with varying turbidity. Synthetic raw water prepared in the laboratory was used in the experiments. Based on the outcome, the applicability of the Jar Test results in determining the optimum coagulant dose is discussed, and recommendations are given.

Review of past data from several water treatment plants in the Kandy district, where the study was undertaken, showed that natural raw water turbidity varies within a range of 10–500 NTU throughout the year due to seasonal environmental changes and other activities in the watershed. Raw water turbidity as high as 1,000 NTU also has been reported on a few occasions, but these are not typical values for the rivers in the Kandy district. Therefore, it was decided to measure the SCC values using raw water with varying turbidity up to a maximum of 550 NTU.

Considering the large number of variable parameters affecting natural water, which will influence the outcome of the test results, it was decided to carry out the series of tests using synthetic raw water. Stock solution for the synthetic raw water sample was prepared by adding 10 g of Bentonite clay powder into 1 L of tap water. The suspension was well-stirred in order to ensure uniform mixing of Bentonite particles. The solution was left for 24 hours to allow for complete hydration of particles. The synthetic turbid stock solution was diluted immediately with tap water to get the desired raw water turbidity values.

The coagulant commonly used in the water treatment plants is PACl, which can be used in a wider pH range since it is already hydrolyzed. Therefore, it was decided to use PACl as the coagulant for the experiments. A PACl solution (1% stock) was prepared by dissolving 1 g of PACl in 100 g (ml) of distilled water. The required PACl dose for testing was obtained by pouring measured quantities of the above stock solution into 1 L beakers filled with raw water.

The six beakers in the Jar Test apparatus were filled with raw water of specific turbidity. The measured quantities of PACl from the stock solution was added to the beakers to obtain doses varying from 5 mg/L to 60 mg/L.

The beakers were subjected to a rapid stirring for 2 minutes at a speed of 150 rpm, and the speed was reduced to 30 rpm, and slow stirring was continued for 20 minutes. Stirring was stopped, and parts of the samples were poured into 250-mL measuring cylinders, and the balance was allowed to settle for 20 minutes in the beaker.

The turbidity of the supernatant of each beaker was measured after settling.

The sludge samples poured into the 250 mL measuring cylinders were allowed to settle for 10 minutes. Apparent sludge volume in the cylinder was made to be 50 mL after the 10-minute period by siphoning off the excess sludge.

A small funnel of which the stem was extended by a glass tube, the end of which was located about 10 mm above the bottom of the cylinder, was introduced into the cylinder, as shown in the experimental set up (Figure 1).

A volume of 100 mL of the supernatant from the Jar Test was collected and poured lightly into the cylinder through the funnel, making sure no air bubbles were drawn along. Water was introduced in a discontinuous manner by small quantities, and the excess liquid was allowed to run off by overflow from the top of the cylinder. The supernatant of the beakers after flocculation (Jar Test) was used for the SCT to ensure no variation in pH or temperature between the contents of the cylinder and the water introduced through the funnel. The expanded height of the sludge column and time taken were recorded for each pour of 100 mL (Degrémont 2007).

Seven series of tests were conducted for seven raw water turbidity values starting from 250 NTU to 550 NTU. The PACl dose was varied from 5 mg/L to 60 mg/L. The lowest PACl value of 5 mg/l was selected considering the minimum PACl dose used in the water treatment plants.

When the raw water turbidity was less than 200 NTU, it was not possible to secure a sufficient sludge volume to conduct the test within the given PACl range. For raw water with turbidity >450 NTU, the PACl dose range had to be extended up to 60 mg/L in order to capture the highest SCC.

A sludge layer submitted to upward water current expands and occupies an apparent volume roughly proportional to the velocity of water, according to a ratio that characterizes the cohesion of the sludge. The curve representing the variation in velocity (v) according to the volume of expanding sludge (V) is a straight line (Figure 2). The intercept of the linear regression line (K) is characteristics of cohesion of the sludge known as the sludge cohesion coefficient, SCC (Figure 2).

Depending on the characteristics of sludge, the SCC varies. Measuring SCC gives valuable information on how precipitates behave in a solid contact clarifier.

From the results of each test run, the linear regression line was generated between the velocity of flow (ν) and the apparent volume (V).

The value of the SCC of each test sample was determined using the intercept of the trend line (Figure 3).

Supernatant turbidity and SCC of each sludge sample were plotted against the coagulant dose. Figure 4 illustrates the variation of supernatant turbidity and SCC against the coagulant dose of sludge samples prepared with raw water turbidity of 500 NTU and PACl dosage varying from 15 mg/L to 70 mg/L.

A set of similar plots were developed for sludge samples prepared with raw water having varying turbidity in the range of 250 NTU to 550 NTU (Figures 5(a) and 5(b)).

Figures 5(c) and 5(d) gives the results obtained from the 20 verifying samples.

At low raw water turbidity, the amount of sludge formed was less, and the sludge was light and fragile. With higher raw water turbidity a sufficient amount of sludge was formed. However, it is noted that depending on the PACl dose the characteristics of the sludge varies. Sludge formed with PACl dose closer to optimum (PACl dose at lowest supernatant turbidity) was consistent, and the raising and lowering of sludge blanket was easy to distinguish. Sludge formed with other PACl doses was light and fragile compared with the sludge at the optimum PACl dose.

The correlation between PACl dose and the SCC value was tested using SPSS. When all the SCC values were considered irrespective of the raw water turbidity or the PACl dose, the SCC values recorded vary within the range of 0.59–3.31 m/hr. Mean SCC is 1.98 m/hr, and the standard deviation is 0.70 m/hr. All the SCC values were within the 95% confidence interval.

A quadratic curve fit was made for supernatant turbidity vs. PACl dose and average SCC vs. PACl dose at different raw water turbidity values. Table 1 gives the R² value and the level of significance for the quadratic relationships obtained.

A high R² value is recorded, indicating a strong correlation between the raw water turbidity and the two parameters considered.

The PACl dose at which lowest supernatant turbidity is recorded in the Jar test and PACl dose at which the highest SCC is recorded was plotted against the raw water turbidity. (Figure 6).

The quadratic curve fits established for optimum coagulant doses vs raw water turbidity values show a significant relationship with R² = 0.97 and p-value < 0.01. Figure 7 shows the relationship between the optimum PACl dose found by the Jar Test and the optimum PACl dose at which the highest SCC is recorded.

In practice, the optimum PACl dose obtained from the Jar Test is used for operating the sludge blanket clarifiers. However, these tests show that this may not be the case, particularly for the higher range of raw water turbidity, beyond 300 NTU. These results indicate that there is a strong correlation between the two optimum PACl values, and a linear relationship can be found between them with R² = 0.9 and p-value < 0.05.

Knowledge of the sludge cohesion co-efficient gives an understanding of the behaviour of a particular sludge blanket.

Low SCC values indicate a light and fragile sludge blanket whereas high SCC values indicate a quickly settled, consistent sludge blanket. At lower raw water turbidity ranges it is hard to establish a sludge blanket. Similar characteristics have been observed in the sludge samples prepared with higher turbidity and PACl doses, which are much different to the respective optimum dosage.

At a given raw water turbidity there was an optimum coagulant dose that produced the highest SCC and hence a consistent sludge blanket. With increasing raw water turbidity the optimum PACl dose at which the supernatant turbidity is at a minimum also increases. Similarly, the PACl dose at which the highest SCC is observed also increases. However, when the raw water turbidity was higher than approximately 300 NTU, PACl dose at the highest SCC was less compared with the standard Jar Test optimum.

A correlation is observed between the Jar Test and SCC optimum PACl doses, which appears to be a quadratic relationship according to the tests conducted so far. However, this needs to be studied further using a full-scale clarifier to confirm the laboratory results and establish a reliable relationship.

SCC is an appropriate indicator to understand the characteristics of a sludge blanket. This study shows that with increasing raw water turbidity the upward flow sludge blanket clarifier units can be operated with an optimum coagulant dose, which is less than the dose obtained from the laboratory flocculation test (Jar Test).

The relationship established can be used to determine the optimum dose to be used in an upward flow sludge blanket clarifier unit after conducting the Jar Test.

The authors wish to acknowledge the valuable assistance given by the Lab Assistant and all the technical personnel at the TSK Project in conducting the tests.