Research of filtration and energy parameters of membrane bioreactors

. Membrane bioreactors nowadays are intensively applied in wastewater treatment worldwide. The paper refers to a research, which goal was to evaluate the optimal parameters of membrane bioreactor performance. A focus was given to a part of the research that considered operation of side-stream membrane bioreactor under pressure of 0.2 MPa and variable amount of backwashes. There were three types of feed water investigated with mixed liquor suspended solids concentrations of 1.6, 2.5 and 7 g/L, respectively. The article provides results of permeate production and their processing to obtain specific energy consumption for each combination of operation parameters.


Introduction
Application of membrane bioreactors for wastewater treatment, which became well known in 1980s, is now a matter of a significant interest from viewpoint of research and practical implementation [1,2]. The more detailed the research are, the wider and more solid data will be obtained and the more accurate will be the operation. Real performance, of course, cannot be fully predicted during the bench-scale, pilot-scale or even full-scale research, however, all of these can significantly contribute to sustainable operation [3,4]. The experience in operation of MBR in Russian Federation is not so solid, and these blanks and gaps should be filled [5]. It considers almost every aspect of MBR -contamination removal from wastewater (with typical for Russia values of pollutants), energy efficiency of the process, fouling mitigation, operation conditions and many others [6,7] If membrane bioreactors are taken into account submersible MBR are the most common solution in majority of cases. On the other hand, if to apply MBRs for reconstruction of small wastewater treatment plants (WWTP) the facilities' dimensions always matter and dimensions of auxiliary tanks in particular [8,9]. In this viewpoint side-stream membrane bioreactors can be focused on within a research as a probable option if to place them nearby in portable containers (and without auxiliary capital construction) [10].

Materials and methods
The research was performed following a goal to investigate operation of side-stream MBR in case of several possible operation parameters -mixed liquor suspended solids (MLSS), operation pressure, and amount of backwashes. The objectives of the research are: − to study suspended particles detention (of active sludge) using membrane methods at different values of operating pressure − to increase performance of membrane bioreactors, as well as to determine the optimal range of working pressure under which it is possible to prevent clogging of membranes and breakthrough of the contaminants in the filtrate.
The bench scheme is shown in figure 1. The feed wastewater (mixed liquor after biological treatment) is supplied from the feed water tank 1 to the membrane module 3 using the pump 2. The working pressure is regulated by means of a pressure tank 4, pressure switch 5, a bypass valve 10, and a pressure gauge 12. Permeate after the membrane module was collected in the samplers 7. Ordinary capillary ultrafiltration membranes with standard characteristics were used for the bench-scale research with pore size of 0.1 micron and total surface area of 1m 2 . Fig. 1. Scheme of the bench-scale model: 1 -feed water tank; 2 -pump; 3 -membrane module; 4pressure tank; 5 -pressure switch; 6 -valve; 7 -sampler tank; 8 -washing tank; 9 -washing water sampler; 10 -bypass valve; 11 -pressure tank for washing; 12 -pressure gauge Backwash is a mandatory option for the functioning of membrane modules in MBR. Deionized water after the reverse osmosis plant was used for backwashing and supplied from washing tank 8. Switching between feed and wash tanks was performed using valves 6. The working pump fed the wash water into the filtrate path; afterwards it was collected through the source water path to the wash water samplers 9. Switching of pipelines was carried out by means of valves 6. Pressure was controlled with pressure switch 5, a bypass valve 10 and pressure tank of washing water 11.
The whole plot of the research was divided into several sub-stages to compare obtained results in the following. The criteria of the division was operation pressure, while the MLSS value remained the same through the parts of the research. There were three concentrations of MLSS investigated within the research similar to those in conventional activated sludge systems. This if the MBR will be applied for reconstruction of existing WWTP parameters. The third value, MLSS=7 g/L MBR. This paper has a focus on a part of the research with operation pressure as our previous study showed that higher operation pressure stimulates breakthrough of suspended solids [11].

Results
The first obtained results were the volumes of permeate membrane filtration ( Figure 2). It is predictable, that higher efficiency ( volume) was achieved at low MLSS value (1.6 g/L). provided a slight (10-15%) decrease of permeate production, however, both may be considered as comparable. At the meantime, times when the MLSS value was risen up to 7 g/L That may signal that under operation pressure of 0.2 MPa the sharply depended on the growth of MLSS. plummets due to soar of fouling resistance cause by intensive growth of permeate production is witnessed at the initial period of filtration almost linear growth can be seen within first 3 minutes The next part is filtration between 3 and 8 minutes, when the permeate production grows approximately 70% under all three MLSS value. Duration of 8 minutes is a sort of checkpoint as far as the next 5 minutes period showed only 10% of permeate production growth. That means that after 8 minutes of filtration the cake layer on the membrane surface is dense enough to stop the filtration at any MLSS value. be described through the dependence between higher the MLSS value the more intensive is flux drop volume in certain time can be determined SS investigated within the research: MLSS=1.6 and 2.5 g/L are ional activated sludge systems. This approach seems reasonable, if the MBR will be applied for reconstruction of existing WWTP with current operation 7 g/L lies in a range of optimal values for submerged This paper has a focus on a part of the research with operation pressure of 0.2 MPa, as our previous study showed that higher operation pressure stimulates breakthrough of The first obtained results were the volumes of permeate related to variable duration of predictable, that higher efficiency (greater permeate volume) was achieved at low MLSS value (1.6 g/L). Slightly higher MLSS (of 2.5 g/L) 15%) decrease of permeate production, however, both sets of values may be considered as comparable. At the meantime, the permeate production dropped 3.2 LSS value was risen up to 7 g/L comparing to MLSS=1.6 g/L.
Growth of permeate volume through time n pressure of 0.2 MPa the permeate production is the growth of MLSS. In other words, the permeate production due to soar of fouling resistance cause by intensive pore clogging. The most production is witnessed at the initial period of filtrationalmost linear growth can be seen within first 3 minutes of filtration (for all three MLSS). The next part is filtration between 3 and 8 minutes, when the permeate production grows by MLSS value. Duration of 8 minutes is a sort of checkpoint as far as the next 5 minutes period showed only 10% of permeate production That means that after 8 minutes of filtration the cake layer on the membrane ltration at any MLSS value. The same process can also be described through the dependence between flow and time of filtration ( figure 3). The the more intensive is flux drop due to rapid pore fouling. Permeate can be determined if to draw straight line from abscissa axis and parallel to ordinate axis. Area of obtained figure corresponds to volume of filtered wastewater during selected time.  It is obvious that the shorter filtration cycle is other hand, the more often backwashes are applied, the larger is the volume of water required to implement efficient backwash and the lower is "useful" product water amount defined as a difference between total permeate parallel to ordinate axis. Area of obtained figure corresponds to volume of filtered filtration cycle (pressure= 0.2 MPa) our further investigation of operation conditions, we focused at three possible filtration cycle duration of 2, 4 and 6 minutes or 30, 15 and 10 different operation conditions the shorter filtration cycle is the larger is permeating volume. On the other hand, the more often backwashes are applied, the larger is the volume of water required to implement efficient backwash and the lower is "useful" product water amount permeate amount and water used for backwashing.
Backwash efficiency depends on filtration addition, backwash efficiency can be evaluated as percentage of fouling removed from membrane surface. Definition of corresponds to minimal operational costs was among the goals of our Experimental data processing for experiments are presented on figures 5 and 6. Figure 5 shows dependencies of specific membrane frequency of backwashes. Total permeate increase of suspended solids (and fouling, respectively)   Figure 6 demonstrates dependencies of backwashing frequencies) to suspended solids concentration in the feed water. tends to increase in water consumption due to increase of frequency of backwashes and washing time. Figures 7-9 gives examples of evaluation of "useful" during operation of the module for various backwash frequencies performed at Figure 7 refers to frequency of 30 backwashes, figur to 10 backwashes. There are four graphs at each of these figures flow of single cycle of filtration (Q 1f ); a flow used for backwash (Q that is a sum of Q 1f and Q w ; a useful flow (Q   Having an optimum operation mode determined the next step of the research was to define the entire energy consumed and the specific energy consumption per flow. Within this we assumed that according to Q-H function of centrifugal pumps e required to pump up 1 m 3 /h at 1 m of height is 0. energy consumption calculated for the useful  It is important to note, that specific energy consumption at low MLSS va less similar for different length of filtration cycle increase of specific energy consumption, however, the longer cycle of filtration less amount of backwashing and less consumed water correspondingly) specific energy consumption with approx. 2 times difference if we compare 30 and 10 backwash mode respectively.

Conclusions
1. Permeate effluent has the most intensity within the first 3 minutes of operation, no matter what is the value of MLSS.
2. After 8 minutes of filtration, permeate fouling on membrane's surface (at high MLSS values 3. Shorter filtration cycles provides higher overall amount of filtered water, however it also means more frequent backwashes that increase wa 4. Specific energy consumption has a strong relation to MLSS values and operation mode. Less frequent backwashing helps to reduce low MLSS values, and up to 50% at high MLSS values.
5. Further research is urgently needed to of pollutants in feed wastewater.
This work was financially supported by Ministry of Science and Higher Education of the Russian Federation (#MK-519.2019.8).
It is important to note, that specific energy consumption at low MLSS value is more or filtration cycle. Growth of MLSS provided a respective , however, the longer cycle of filtration (that means less amount of backwashing and less consumed water correspondingly) resulted in lower with approx. 2 times difference if we compare 30 and 10 has the most intensity within the first 3 minutes of operation, no permeate production plummets due to accumulated at high MLSS values in paeticular). Shorter filtration cycles provides higher overall amount of filtered water, however it also means more frequent backwashes that increase water needs for that purpose.
Specific energy consumption has a strong relation to MLSS values and operation helps to reduce specific energy consumption by 10% at low MLSS values, and up to 50% at high MLSS values.
research is urgently needed to fit obtained results to various concentrations This work was financially supported by Ministry of Science and Higher Education of the Russian