Impact of Variable Technological and Quality Factors on the Efficiency of Filtration Processes Using Dynasand Filters and Lamella Separator

The selected research object is the Water Treatment Plant (WTP) in Stary Sącz, which treats water from the Dunajec River and from 16 infiltration wells fed with water from the river through a system of three infiltration basins. The quality of water in the Dunajec River meets the quality requirements of WTP equipment of the A2/A3 category in compliance with the Regulation of the Minister of Maritime Economy and Inland Navigation of August 29, 2019. [19]. These waters require high-efficiency physical and chemical purification or biological methods, in particular, oxidation, coagulation, flocculation, decantation, filtration, adsorption on activated carbon, and disinfection by ozonation or final chlorination [19].

Raw water from the surface water intake from the river is pumped to the volumetric coagulation vertical settling tanks located on the premises of WTP in Stary Sącz. Then a coagulant is added to the raw water before the settling tanks in an automatic mode, depending on the turbidity of the raw water and on flows. At that point, the pre-treated water is directed to the filtration system consisting of 10 open self-washing DynaSand DS5000 filters with a capacity of 50 m³·h⁻¹ each. In order to reduce water losses caused by a large volume of rinse water (approx. 40 m³·h⁻¹) generated after rinsing the filter bed on DynaSand filters, the rinse water is transferred to the separator Johanson Lamella LS 60−05+F where it is treated and then again returned to the technological system, i.e. to the pipeline before sand filters. The treated water after the filtration process in DynaSand filters is pumped to a siphon venting column (VC) with a diameter of 1.000 mm and a height of 3.500 mm before the filters. Afterwards, it flows by gravity to the buffer tank and to the water-ozone contact tank, where the minimum time of ozone-water contact of approx. 20 minutes is ensured. Ozone is produced on-site from the air. After the ozonation process, the water is transferred by pumps to eight pressure filters filled with activated carbon. After filtration, the water goes through a physical disinfection system based on a UV lamp and it is collected in clean water tanks. Chlorine (I) is dosed into the water as the final disinfectant. The water treated in this way is passed to the water supply system by a set of high-pressure pumps. The maximum capacity of the technological system is approx. Q = 690 m³·h⁻¹.

Water from the infiltration intakes in the amount of max. 190 m³·h⁻¹, depending on its physical and chemical quality, may be subjected to the following treatment stages (discussed earlier): filtration on DynaSand filters, ozonation, sorption on activated carbon and disinfection with UV rays and chlorine, or only ozonation and sorption on filters together with disinfection with the UV lamp and chlorine gas or, in the case when its quality is good, it is subjected only to physical disinfection with UV rays and chemical disinfection with chlorine gas. [20] The WTP technological diagram is shown in Figure 1.

Figure 1.

The collection of water samples for testing was carried out on a full technical scale, in a regular manner (although with different frequencies during the year), maintaining retention times on DynaSand filters and on the Lamella separator. The samples for testing were collected in the period from 2018 to 2022 at the following points: (1)

before the DynaSand filters,

In total, 2597 water samples were collected and tested. Water quality was assessed on the basis of tests performed by an accredited laboratory of Waterworks L.L.C. (accreditation no. AB 980) and by a technological laboratory located at the Water Treatment Plant in Stary Sącz. The tests were performed in compliance with the applicable Regulation of the Minister of Health of 7 December 2017 on the quality of water intended for human consumption [21]. In order to properly assess the effectiveness of DynaSand filters, sixteen selected physico-chemical indicators were analyzed (temperature, pH, specific electrolytic conductivity, turbidity, dry residue, color, absorbance at 254 nm wavelength, oxidisability with KMnO₄, total iron, manganese, chlorides, sulphates, phosphates, ammonia, nitrites, nitrates) and three bacteriological indicators (coliform bacteria, Enterococcus Faecalis, Clostridium perfringens). Turbidity was measured by the nephelometric method (PN-EN ISO 7027-1:2016-09), pH was tested by the potentiometric method (PN-EN ISO 10523:2012), electrical conductivity at 25°C by the conductometric method (PN-EN 27888:1999). The color of the water was tested according to PN-EN ISO 7887:2012; Ap1:2015-06, absorbance UVA_254nm according to PN-84, C-04572, the oxidisability with was tested by the titration method. Chlorides, sulphates, phosphates and ammonia were tested by ion chromatography (PN-EN ISO 10304-1:2009/AC:2012). Iron and manganese were tested using atomic absorption spectrometry with electrothermal atomization (graphite cuvette) PN-EN ISO 15586:2005 and PN-EN ISO 17294-2:2016-11.

Microbiological tests were conducted in accredited water and wastewater testing laboratory using the plate method (deep culture) and membrane filtration method, based on standard PN-EN ISO 7899–2:2004 for Enterococcus faecalis and standard PN-EN ISO 9308–1:2014-12 + A1:2017-04 for Coliforms bacteria and Escherichia coli, and in accordance with Directive 98/83/EC [22] and currently Directive (EU) 2020/2184 [23] for Clostridium perfringens.

In order to assess the treatment effectiveness of rinse water in the Lamella separator, pH, specific electrolytic conductivity, turbidity, color and UVA_254nm were analyzed. In addition to the impact of variable quality parameters of raw water on the filtration and separation processes, the technological factors were also assessed, i.e. the impact of variable water flows (retention and reaction times) on the filtration process on filters, and thus on the quality of filtrate, as well as the impact of rinse water quality on the quality of water subjected to filtration and the quality of filtrate itself. In the qualitative and quantitative analyses of water samples before sand filters, a collective sample was taken into account, which consisted of inflowing water after volumetric coagulation settling tanks and water pumped from infiltration wells, as well as rinse water returned to the system from the Lamella separator.

The results of tests and statistical analyses of physico-chemical parameters aimed at assessing water purification effectiveness on DynaSand filters and treatment efficiency of rinse water in the Lamella separator are presented in Table 1.

Physico-chemical parameters of water before the filtration process change dynamically along with the change in the parameters of the water collected from the Dunajec River.

The temperature of surface waters is characterized by high amplitude because it depends on variable air temperature in the cycle of four seasons. The lowest temperatures before filters were recorded in winter months (min. 0.70, average 11.80), while the highest in July and August of a given year (max. 22.90) (Table 1). The filtration processes on DynaSand filters were more effective at higher temperatures (summer, autumn), which is also confirmed by the results of tests conducted at WTP. This effect is connected with chemical processes taking place in the filter bed, e.g. contact coagulation as well as biological processes. As can be seen from Table 1, the maximum pH value of water before the filters over the last 5 years was 8.70 in 2021, and the lowest value, i.e. 7.20, was recorded in the same year. We can observe fluctuations in pH in one-year periods, which is caused by seasonality (drought period after rainfall, development of biological life in water collected from the river). In effect of the processes taking place on the filters, including coagulation in the DynaSand filter bed, a decrease in the pH of water after a given process was observed (min. 7.10, max. 8.30). The specific electrolytic conductivity was the result of the mixture of surface and infiltration waters (min. 270.00, average – 391.00, max. 578.00), similarly to the dry residue (Table 1).

One of the main pollutants in surface waters is turbidity, which is usually removed by filtration [21]. High values of turbidity, color, UVA_254nm in the analyzed WTP as well as oxidisability could be observed in spring during snowmelt and in summer during torrential rainfall and flooding. In the examined period of five years, the lowest water turbidity before filters (in winter) was 0.57 NTU, and the average one was 5.94 NTU (Table 1). The maximum value before the filters of 133 NTU was recorded on July 19, 2018, after heavy rainfall, when the turbidity of the water in the Dunajec River was 2.550 NTU. After the filtration process, the turbidity reached a value of 0.81 NTU (Table 1). For DynaSand filters, the limit value of turbidity recommended by the manufacturer is approximately 50 NTU. Above this value, the filtration bed may be punctured by carrying out the excess suspension, which is not effectively removed from the filtrate. The results of tests on a technical scale carried out at WTP in Stary Sącz confirm this thesis (Table 1).

It is worth emphasizing that in the examined period of 5 years, the average turbidity values after sand filters were 0.20 NTU. For the 869 tested water samples, the standard for water intended for consumption, which is 1 NTU [21] was not exceeded (Table 1, Figure 2). Turbidity was reduced on DynaSand filters from 63.3% to a maximum of 99.6%, with an average of 93.2% (Table 2).

Figure 2.

Similar relationships were found for substances that impart color to the water, for dissolved organic compounds designated as UVA_254nm and for oxidisability with KMnO₄. Such compounds were reported in high concentrations in water after rainfall. The highest tested value of color recorded on July 19, 2018 before sand filters was 182 mg Pt·dm⁻³, UVA_254nm – 1.450, and oxidisability – 4.10 mg O₂·dm⁻³. The filters reduced the color index to 8.99 mg Pt·dm⁻³, against the standard of 15 mg Pt·dm⁻³ (Figure 3) [21], UVA_254nm to 0.257 and oxidisability to 2.30 mg O₂· dm⁻³ (Table 1). The lowest tested color value after sand filters was 1.8 mg Pt· dm⁻³, the average was about 4.2 mg Pt·dm⁻³, and the highest was 8.99 mg Pt·dm⁻³ (Table 1), with an average reduction of 69.7% and a maximum of 98.3% (Table 2).

Figure 3.

DynaSand dynamic bed filters effectively removed total iron and manganese from water. The average total iron content in water before the filters was 0.1370 mg·dm⁻³ and max. 2.4800 mg·dm⁻³. After the filtration process, the average values for total Fe were 0.0112 and max. 0.1300 (standard 0.200 mg·dm⁻³) [21], and for Mn – on average 0.0004 and max. 0.0017, which yielded a reduction in the filtration process of an average of 86.2% to the max. 100% for total Fe, and an average of 96.4% to a maximum of 100% for Mn. The tests of the level of chlorides and sulfates in water before and after filtration indicate a slight reduction of chlorides (on average 4.8%, max. 25.0%) and sulfates (min. 10.3%, average 19.65%, max. 33.6%). The filters reduce the concentration of phosphates in a satisfactory percentage (on average 49.6% and max. 100%). Ammonia is reduced with an average efficiency of 76.9% and max. 100%, nitrites – on average 90.2%, max. 100%. Nitrates are removed from water to a lesser extent, on average 9.5%, max. 44.1% (Table 1, Table 2).

High concentrations of bacteriological indicators such as coliform and Enterococcus faecalis in the Dunajec River waters attest to water contamination with household sewage from the yet unsewered catchment areas, located above the water intake. Clostridium perfringens is found in the soil, and it freely enters surface waters. The processes taking place in the filter bed allow for the effective reduction of bacteriological indicators. Coliform bacteria were removed with an average efficiency of 93.3% and max. 100%, Enterococcus faecalis were removed on average in 91.9%, maximum in 100%, while Clostridium perfringens – on average at 97.5% and max. in 100% (Table 3).

The high removal efficiency of bacteria and algae in the dynamic bed of sand filters in Egypt has been confirmed by [9].

Water from infiltration intakes is characterized by high physical and chemical stability. The quality of infiltration water is similar to that of groundwater. In terms of color, turbidity, pH, electrolytic conductivity, nitrates and organic compounds, the water does not raise any objections. Due to the irrigation of infiltration wells with water from the Dunajec River, heavy metals and single coliform bacteria, Enterococcus faecalis and Clostridium perfringens may be found periodically in the water. Therefore, the infiltration water is directed in the technological system to DynaSand filters and then to the UV lamp.

In filters with a dynamic sand bed, as well as in pressure sand filters or in those with an activated carbon bed, physical, chemical and biological processes take place. Filters effectively purify water by operating in an uninterrupted mode, with constant flows and constant filtration speed. The rise in the flow of raw water onto the filters with low values of physico-chemical indices (mainly turbidity, color, UVA_254nm) and proper dosing of the coagulant (depending on the flow and turbidity of raw water) does not cause any disturbing changes in the quality of filtrate at the WTP in Stary Sącz (Figure 4). Only a long-term rise in the critical parameters (turbidity, suspension, color, UVA_254nm) of raw water subjected to filtration at high water flows (from 350 m³·h⁻¹ to 466 m³·h⁻¹) resulted in an increase in water quality indicators after the filtration process to the value of 0.81 NTU at WTP in Stary Sącz (Table 1, Table 2). The above relationships are illustrated in Figure 3, using the example of turbidity.

Figure 4.

The analyses of the operation efficiency of DynaSand filters, depending on water retention times on the filters and on filtration speed (directly dependent on flows) confirmed the earlier results of the analyses. At low values of physico-chemical parameters of raw water inflowing onto the sand bed, the reduction of retention times (reactions) on the bed, and hence the rise in filtration speed did not cause any significant differences in the quality of the filtrate (0.20 NTU – 0.35 NTU) (Figure 5). Only the rise in turbidity, suspension and organic compounds, with the concurrent shorter reaction time of water with the bed, reduced the operation efficiency of the filters through faster colmatation of the bed and raised resistance as well as increased turbidity of the filtrate. Similar results were obtained for UVA₂₅₄, oxidisability, total Fe and Mn.

Figure 5.

DynaSand self-washing filters continuously generate an average of 40.3 m³· h⁻¹ and a maximum of 46.3 m³·h⁻¹ of rinse water after washing the filter sand (Table 2). In order to reduce water losses in the technological system, a compact Lamella separator was used, which is designed to purify rinse water, which is then returned to the system before DynaSand filters. The coagulant is dosed by adjusting the dose, depending on the turbidity of water before Lamella and on the current flow (the lowest – 147 ml·h⁻¹, the average – 319 ml·h⁻¹ and the highest – 760 ml·h⁻¹ (Table 2)) obtained from the flowmeter readings (online). The turbidities of rinse water before the Lamella separator ranged from 9.23 NTU, an average of 36.38 NTU, to a maximum of 278 NTU (2021). After the separator, the obtained turbidities were from 2.70 NTU (reduction of 35.50%), on average – 10.45 NTU (reduction of 67.16%) to a maximum of 55.60 NTU (reduction of 97.79%) (Table 1, Table 2). The color of rinse water was on average 72.2 mg Pt·dm⁻¹ and max. 288.0 mg Pt·dm⁻¹ (2021). The UVA_254nm tested before the separator ranged from 0.151 to 1.037, an average of 0.306. After Lamella, the values from 0.093 to 0.392 were obtained, with an average of 0.158. The percentage of color removal averaged 63.3% with a maximum of 90.3%, while that of the UVA_254nm index was 46.2% and 79.9%, respectively (Table 1, Table 2).

In the multi-year period, the test results of turbidity, color and UVA_254nm demonstrated no negative impact of the reversed rinse water (in the amount of approx. 40.0 m³·h⁻¹) on the efficiency of filtration processes and filtrate quality. In Figure 6, Figure 7 and Figure 8, no direct relationship was observed between water parameters after Lamella and before DynaSand filters. The filtration process was mainly determined by the quality of raw water from the Dunajec river pumped into the system as well as by its quantity. Post-coagulation suspension flowing onto dynamic bed filters may also contribute to the rise in the efficiency of the filtration process.

Figure 6.

Figure 7.

Figure 8.

Yet, it was found that the parameters of raw water have a direct impact on the degree of contamination of the filter sand, and thus on the quality of rinse water generated by DynaSand filters and on the quality of rinse water after the Lamella separator. The efficiency of the Lamella separator is the result of the flow, turbidity of rinse water and coagulant dosed so far in an adjustable manner. The results of the analyses show that optimizing coagulant dosing through the application of a turbidity meter measuring (online) the turbidity of rinse water, and by adjusting the pump dosing coagulant onto Lamella in terms of flow and turbidity in the automatic system, can increase the effectiveness of the rinse water treatment. The reaction of the control system to the rise in the turbidity of rinse water will be immediate, which ultimately will intensify the cleaning process of rinse water.