Experimental procedure

The pressure filtration experiments were performed for all types of slurries in three different modes: (1) filtration of the slurry through the filter media only, (2) through the filter media, which was previously coated with filter aids, and (3) the slurry in which filter aids were mixed before the tests. Vacuum filtration was only done for the Al/Fe slurry.

50

9.3.1 Electrocoagulation experiments and sampling

65 L of pure water and 5 L of the prepared solution were filled into the reactor and mixed with speed of 300 rpm for 10 minutes. After that, the pH value of the initial solution (70 L) was measured with a pH-Electrode SenTix 22. Hydrochloric acid at a concentration of 1 M was added to the solution to adjust the pH value to 3 (the optimal value determined previously) while mixing.

Before the experiment was started, an initial sample of the solution was taken from the reactor for further analysis. After that the electrodes were placed into the reactor covering it with the reactor head, power supply cables were connected to corresponding electrodes, stirrer speed was adjusted to 150 rpm, cooling system was switched on and the current was applied to the electrodes and simultaneously the stopwatch was switched on. The moment when the current was applied to the system was the starting moment of the EC-test. The total time of all EC-tests was 5 hours. In order to avoid the effect of electrodes passivation, the polarity of electrodes was changed every once hour during the experiments with Fe/Fe and Al/Al electrode combinations. The time when there was no current applied to the system was not considered as a time of reaction. The polarity was not changed when Al/Fe electrode combination was used.

Along with the initial (“zero”) sample of the solution (taken before the start of the EC-test) samples of the forming suspension were taken after 30 min, 1, 2, 3 and 5 hours of the treatment. Zero sample was taken once in an amount of 20 mL whereas the volume of other samples was 40 mL.

The samples were analysed in-situ to determine pH, conductivity and reduction-oxidation potential. The actual voltage and temperature in the reactor suspension were also monitored at the time of the sampling. Further studies included determination of the particle size distribution, the density of the samples, the solid content by mass fraction, density of the solid fraction, scanning electron microscopy and X-ray diffraction analyses.

9.3.2 Pressure filtration

By testing three different electrode combinations, three types of slurries were obtained. They consisted of iron oxide species (Fe/Fe slurry obtained in the EC with Fe /Fe combination), of aluminum oxides (Al/Al slurry, Al /Al combination) and of both iron and aluminum (Al/Fe slurry, Al cathode/Fe anode). Since the concentration of solids in the suspensions right after the EC is not high enough to perform the cake filtration experiments, the slurries were collected after 48 hours when most of the solids had sedimented. This means that the initial suspension obtained from the EC-tests was divided into supernatant solution and concentrated slurry at the bottom of the reactor.

51

Clarified liquid was drained and the concentrated slurry was discharged through the bottom valve of the reactor and was collected for the filtration experiments. In those cases where iron electrodes were used (Fe/Fe and Al/Fe arrangements), the floated solids were also presented and formed a top layer that varied in width from approximately 1.0 – 1.5 cm for Fe/Fe suspension to 3.5 – 5.0 cm for Al/Fe slurry. In both cases the floated layer was removed manually after the electrodes were taken out of the reactor in order to avoid the dissolution of the fine solids from this layer into the clarified liquid.

The filtration experiments were carried out with the constant pressure filtration mode for all slurry samples. The volume of samples was varied: 50, 100, 200 and 300 mL of slurry was taken for the filtration under the pressure difference of 2, 4 and 6 bars for each of the volume. Filtration tests performed at pressure of 6 bars were conducted twice to verify the repeatability of the results.

Before the filtration tests were started, temperature and pH of the slurry were measured with a laboratory thermometer and the pH-Electrode SenTix 22 respectively. The slurry was mixed by a metallic agitator powered by an electric motor at 10 rpm during the whole duration of the tests.

Filter medium disks were first dried and kept in a desiccator and weighted. Then the filters were submerged into pure water for two hours to remove air from the material and weighed in a saturated state. After adjusting the pressure difference to the selected value, assembling the filter and starting the LabVIEW program, the slurry sample of particular volume was quickly poured into the filter chamber by using a funnel. The filtration test began when the gas was allowed to flow into the chamber making the liquid to penetrate through the filter medium disk solids to form a filter cake on the surface of the disc.

The gas feeding was stopped when there was no more slurry to be filtered in the chamber. Data on the mass of the filtrate and cake thickness measured by a special stick and a ruler were acquired from each experiment. Both the cake and filter media were removed from the filter and weighted twice: in a wet state after being discharged and in the dry state after being kept in an oven at 105

oC for 24 hours. The concentration of solid matter in the slurry (solid content by mass fraction) was determined according to Eq. 4.11 after withstanding of the slurry sample with a defined mass in the oven at 105 ^oC for 24 hours and measuring the mass of the dried solids. The density of the solids was determined by a standard method with a pycnometer. All the data collected were used in the calculations of such parameters as filtration concentration, moisture ratio and moisture content of the cakes, filter medium resistance and cake resistance, porosity and compressibility index.

52 9.3.3 Filtration with filter aids

Filtration tests with addition of filter aids were performed for Al/Fe slurry only. This type of the slurry was chosen because of the results on the removal efficiency obtained in the EC experiments which were the highest in the case of using Al/Fe electrodes. Filtration tests were conducted under constant pressure for both pre-coat and body-feed tests. When the pre-coat filtration experiments were carried out, firstly a filter aid slurry was prepared was prepared and filtered. Diatomite, milled cardboard, RHA X+100 and RHA Grade 4 were used separately as a material to form the pre-coat layer. After being fed on the filter media disk, FA forms a layer of solids which has its own resistance. For the purposes of minimization of the resistance and material saving, the thickness of this layer needs to be as thin as possible to provide higher permeability but thick enough to retain the particles, thus operating similarly to depth filters and preventing the filter medium blinding. Based on the data of the solid content of the Al/Fe slurry which was 2,05 % by weight and the observed thickness of the cakes it was decided to prepare the FA-slurry with the same concentration of solids, 2.05%. The volume of the filtered Al/Fe slurry was constant, 200 mL, for each experiment; volume of the FA-slurry in the pre-coat tests was 100 mL.

Pre-coat filtration. All samples of the FA-slurry were made with a similar procedure. FA in amount of 2.05 grams were mixed manually with 100 grams of pure water until uniformity of the slurry was observed. Then the suspension was additionally mixed with a magnetic stirrer for 2 minutes at the speed of 450 rpm. Temperature and pH of the FA-slurry were measured before each test. The filter was assembled in the same way as it was made for the filtration of Al/Fe slurry. The only difference was the value of the pressure difference that was adjusted to 1 bar in order to obtain less dense layer of FA on the filter disk.

In the first set of experiments only the FA-slurry (sample volume of 100 mL) was filtered at 1 bar.

For the obtained FA-cakes thickness, moisture content (MC) and moisture ratio (MR) were determined and the resistance of the FA cakes was calculated by using data recorded by the LabVIEW system. In the second set of experiments FA-slurry was deposited onto the filter medium the same way as it was in the first set. Then the filtration operation for Al/Fe slurry was repeated at 6 bars according to the procedure described above. As previously, the data used in further calculations was recorded by the LabVIEW system.

Body feed filtration. Only three of the four mentioned material were used in the body-feed mode, namely, diatomite, RHA X+100 and RHA Grade 4. Milled cardboard was not possible to use

53

because this slurry stacked in the inlet tube when being poured into the filter chamber. The amount of particular FA added as the body feed was 50, 100 and 200% by the weight of solid content of the slurry. The slurry with FA was firstly mixed manually until the visual appearance of uniform structure was achieved and then with the magnetic stirrer for 2 minutes at the speed of 450 rpm.

The collected sample of 200 mL was poured into the filter chamber and filtration process was done as described above. The pressure difference was kept at 6 bars for each amount (%) of each FA and the most favourable FA concentration was determined. Additionally, the filtration tests were done at the pressure difference of 2 and 4 bar for this concentration for each FA.

9.3.4 Vacuum filtration

Before pouring the slurry into the filter cell, the vacuum chamber was closed and the required pressure difference was adjusted. The slurry was then poured into the cell, and the filtration started when the slurry valve was opened. The slurry deposited on the filter medium was exposed to vacuum created in the vacuum chamber. The filtrate was collected into the cup placed on the scale inside the chamber. The data was recorded by LabVIEW system.

Vacuum filtration tests were made only for Al/Fe slurry and the volume of each sample was 200 mL. Pressure differences of 0.1 and 0.5 bar were applied in the tests. In the first test, the vacuum pressure difference was established to the minimum value of 0.1 bar. Al/Fe slurry was poured onto the clean filter medium and the slurry valve was quickly opened. To investigate the effect of the operating pressure on the process the second and third tests were made pressure differences of 0.5 bar and the valve was opened more smoothly. In the second case the slurry was also poured onto the clean medium whereas in the third experiment the medium with a cake deposited at the second test was used to determine the possibility to improve the filtrate quality.