Improving the Efficiency of the Aeration Process in the Wastewater Treatment Plant in Masyaf City in Terms of Treatment Speed and Saving Electricity Consumption

Improving the Efficiency of the Aeration Process in the Wastewater Treatment Plant in Masyaf City in Terms of Treatment Speed and Saving Electricity Consumption

Eng. Shadi Ali

Phd sudent, Industrial Automation Engineering Technical Engineering College

Tartous University Tartous, Syria

Prof. Their Ahmad Ibrahim

Industrial Automation, Technical Engineering College Tartous University

Tartous, Syria

Prof.Assi. Mohsen Khatib

Industrial Automation, Technical Engineering College Tartous University

Tartous, Syria

Abstract Aeration process is the cornerstone of organic matter removal in wastewater treatment plants which work with activated sludge model, It is also one of the most energy intensive stages of the plant, so improving its efficiency has a significant impact on treatment quality and reducing operating costs.

The Masyaf wastewater treatment plant operates using an activated sludge model and uses a cascade dissolved oxygen control system to control the aeration process. However, it suffers from prolonged shutdowns due to its high electricity consumption. In this research, we propose a more efficient control system that is able to track biological load, which leads to reducing the amount of air which needed for treatment, in addition to reducing the treatment time and saving energy by 85%.

We implemented a prototype of the aeration process that includes the aeration tank and the secondary sedimentation tank. Then we built a control circuit capable of monitoring the values of dissolved oxygen and ammonium nitrogen, then sending these values to the computer to be processed in the Simulink environment using the proposed control system and issuing the appropriate control command to the aeration elements.

Keywords Wastewater Treatment Plants (ASM Model) electrical energy , aeration control.

1. INTRODUCTION

   The Aeration process in wastewater treatment plants (which operate using the activated sludge method) is considered one of the most important treatment stages, as it removes organic matter, which constitutes the majority of the biological load, but it consumes 50 to 90% of the plant's electrical operating costs [1], At this stage, air is pumped into the water coming from the primary sedimentation tanks through porous molds installed at the bottom of the tank and connected to pipes in which the air is pressurized. It should be noted that excessive aeration causes the formation of pin flakes that emerge with the treated wastewater flowing in, raising pollution indicators. It is also an expensive process in terms of energy consumption [2].

   In 2019, Drewnowski et al. conducted a study on wastewater treatment plants in Poland, examining control strategies and their impact on energy consumption. The researchers concluded

   that conventional control systems that rely on a constant oxygen value in the aeration basin are ineffective, and that the use of an aeration control system based on ammonium nitrogen can save 10-20% of energy consumption and improve the denitrification process. The researchers recommended the use of predictive control because it can ensure process quality and reduce operating costs [1].

   In 2013, researcher Raed Jafar and others conducted a modeling and simulation of the wastewater treatment plant in the Khirbet al-Maza area in Tartous (Syria) using the STOAT program. The researchers noticed that a high pollution load reached the plant in addition to excessive flows. The research concluded that increasing the ventilation stages does not affect the quality of treatment, but rather leads to excessive ventilation and high energy consumption. The researchers recommended designing control systems locally to suit the Syrian reality[2].

   In 2015, researcher Haitham Jnad and others monitored the distribution of dissolved oxygen concentrations in the Marj Mariban and Al-Ruwaimiyah treatment plants in Lattakia(Syria), which use the extended aeration system. The research showed a significant increase in the dissolved oxygen values in the studied aeration basins over the reference value (2 mg/l), which means increased energy consumption and the occurrence of operational problems. [3]

   In 2020, Medinilla et al. used an ammonia-based aeration control (ABAC) system at the No. 1 water recycling plant in Ontario, California. Energy consumption of the air pumps was monitored over a seven-month period, then compared to the system's performance before the ABAC system was implemented. A 9% reduction in energy consumption and an increase in nitrification process efficiency were observed. The researchers confirmed that the increased energy consumption was associated with increased airflow. [4]

   In 2019, Miyahara et al. conducted their study at the Sonomachi Water Treatment Plant in Tokyo. They noted that the ventilation process consumes 23% of the electrical energy consumed in the plant, and that this consumption can be reduced by reducing the amount of air flow. Therefore, a new technique was developed

   based on measuring the drainage load and controlling the air flow rate, and they were able to reduce the amount of air flow required by 10%. [5]

   In 2019, Farhana et al. conducted an energy audit to analyze the energy consumption of wastewater treatment in a plant using the activated sludge method in Malaysia. The study showed that the aeration process consumes the largest amount of energy in the plant. It was suggested to use a high-speed turbo compressor instead of air blowers and to install appropriate valves to control air flow and prevent leakage. As a result, energy consumption decreased by 42%. [6]

   In 2020, Ballhysa et al. reviewed the latest control methods in wastewater treatment plants. They found that control using PI leads to large amplitude fluctuations around the desired value. Predictive control can help reduce the error between the response and the desired value by a large percentage, thus improving the control quality. The research concluded that

   d [()] = [ + ] (1)

   dt

   Qin: inlet flow of aeration tank Qr: return flow to tank

   Qo: outlet flow of aeration tank, = H(t)

   R

   H(t) : Liquid level inside the tank A: Tank space

   R: radius of the tank, = . =

   We substitute the values of R and Qo and then perform the Laplace transform of Equation (1) to obtain Equation (2), which expresses the liquid level inside the basin in terms of the inflow and return flow.

   k

   controlling the oxygen concentration and nitrogen concentration in the aeration basin will lead to improved process quality and

   () =

   [s+1] [() + ()] (2)

   optimal operation of air blowers, which leads to lower aeration costs and thus lower operating costs. [7]

   In 2022, Y.Chen et al. used multivariable optimal control to reduce energy loss in the aeration process in a wastewater treatment plant which using an activated sludge system. The oxygen concentration DO and the sludge discharge quantity Qw were selected as control variables, while the substance concentration S and microbial concentration X in the aeration tank were selected as state variables, with restrictions on energy consumption. As a result, the proposed controller was able to reduce energy consumption by 20% while maintaining wastewater quality [8].

   In 2024, Zijian Wang et al. used nine methods to select the main operating parameters affecting energy consumption in wastewater treatment plants according to daily operation records, and built an intelligent operation management system based on genetic lgorithm by mapping the relationships between energy consumption and the main operating parameters (inlet water flow rate – total nitrogen in outlet water – ammonia loading rate). The results showed that energy consumption can be reduced by an average of 22% [9].

   In this research, we modeled the aeration process in one of the wastewater treatment plants used in Syria (Masyaf plant), which uses a cascade dissolved oxygen control system to control the aeration process. We studied the use of a cascade ammonium control system, and then worked on integrating the two previous control methods using multivariable predictive control.

   We also implemented a prototype of the aeration process (including the aeration tank and the secondary sedimentation tank with a capacity of 20 liters) and then linked this model to the predictive control system that we built in the Simulink environment (in Matlab) via an Arduino chip, which in turn reads the sensors and sends them to the serial port in the computer, to be processed and then send control commands to the surface aeration motor and the air blower.

   The liquid level control system was modeled in the Simulink

   environment after substituting the values of the constants, then adding a PID controller to control the level, as in Figure (1): [10]

   Fig. 1. Simulink model of liquid level control system

   B. The aeration system modeling

   The aeration system is very complex, non-linear, hybrid and time-varying [12]. The aeration system consists of: (air blowers

   – air tube – diffuser distribution tubes – diffuser system). The blowers pump air through the distribution tubes to a network of diffusers distributed on the branch at the bottom of the bioreactor, which diffuses the air to the wastewater inside the aeration tank [11].

   To calculate the average air flow rate at the blower output over a specific period (24 hours), we apply the following relationship:

   = 1 () (3)

   0

   Q: Air flow rate at the blower output

   T: The time period which the average will be calculated (24 hours).

   The transfer function of the aeration system (for a single blower) is shown by the relationship (4) [11]:

2. DESIGN OF THE ACTIVATED SLUDGE MODEL

   2.3

   () = = =

   (4)

   A. Modeling the liquid level control in the aeration tank.

   +1

   1059+1

   The rate of change in the amount of liquid inside the aeration tank is equal to the inflow into the tank minus the outflow according to Equation (1): [10]

   We apply the ASM1 model parameter values found in Table (1) which follow the technical specifications of the blower used.

   .

   TABLE I. Aeration system parameter values according to the

   ASM1 MODEL.

   Parameter
   Value	50	115	1432	0.74
   Unity			3	/3

   C. Modeling biological processes in the aeration tank

   In order to build the mathematical model, the basic system parameters must be determined, which are: inlet flow rate Qin, return flow Qr, BOD concentration, biomass concentration X, and dissolved oxygen DO. Note the relationships (8,5,6,7) [13].

   2 mg/l), and the dissolved oxygen sensor is used to monitor system changes within the aeration tank [13].

   The aeration control system (Aeration System) was designed in the Simulink environment and linked with the studied station model (Aeration Tank) which was designed in order to study the system response and monitor the changes in the dissolved oxygen and the amount of air flowing inside the aeration tank as shown in Figure (2).

   ( )

   dS(t) 1

   =

   () 1 [() + ()]()

   dt

   () () (5)

   dX(t) = 1 [()

   () + () ()] 1 [() +

   Fig. 2. Simulink model for a cascade dissolved oxygen control system

   dt

   ()]() + ()() (6)

   The model was run and the system response was monitored for

   ( )

   dDO(t) 1

   =

   () 1 [() +

   a constant dissolved oxygen value of 2 mg/l. Figure (3) shows

   dt

   the changes in the dissolved oxygen concentration for the

   ()]DO(t) + [ ()]

   sequential dissolved oxygen control system. We note that the

   () () (7)

   system fluctuates around the reference value of 2 mg/l and then stabilizes after about 15 hours at an ammonium nitrogen value

   d() = 1 [() + ()]X(t) 1 [

   () +

   of 3.4 mg/l. The air flow rate also ranges around 90 m3/min,

   dt

   which means a slowness in the treatment process and a waste of

   ()](t) (8)

   We note that there is a set of constants that differ from one station to another related to the size of the aeration tank, the amount of biomass entering the station, in addition to the highest permissible dissolved oxygen concentration.

   TABLE II. The values of the model constants were obtained from the technical specifications of the Masyaf plant.

   Parameter	Y
   Value	0.65	2mg/l	0.5mg/l	0.18
   Parameter
   Value	100mg/l	10mg/l	0.5mg/l	300mg/l

3. STUDY OF CONTROL SYSTEMS:

   The growth rates of nitrifying bacteria (Nitrosomonas and Nitrobacter) depend on the concentration of dissolved oxygen. It was found that when (DO < 2 mg/l) the nitrification process becomes less effective, and if (DO 3 mg/l) the properties of the settled sludge are not good [10].

   In this study, we investigated the use of three aeration control methods and applied them to an aeration model designed in the Simulink environment. We then compared the simulation results in terms of processing time and the amount of energy consumed.

   1. Cascade Oxygen Control System

      The cascade oxygen control system includes two control loops, the first of which controls the air blowers to deliver the appropriate amount of air to the aeration tank, while the second control loop sets the reference dissolved oxygen value to (DO =

      electrical energy consumed during the treatment process.

      Fig. 3. Response of a cascade dissolved oxygen control system

   2. Cascade AmmoniomControl System

      The biological load is constantly changing and unstable, and here the importance of controlling ammonia appears to obtain better and less expensive treatment results. For this system, an ammonia control loop is added to the previous system, as the reference value of dissolved oxygen in this method changes according to need. Figure (4) shows a model of the ammonia control system.

      Fig. 4. Simulink model for a cascade ammonia control system

      We note from Figure (5) that the concentration of dissolved oxygen increases gradually to settle at a value of 0.9 mg/l after about 9 hours, and the air flow rate ranges around 56 m3/min, which means a decrease in the value of the electrical energy consumed. We also note that the concentration of nitrogen settles after about 9 hours at a value of 3 mg/l, which means a reduction in the treatment time.

      Fig. 7. Response of the proposed predictive control system

4. CALCULATING THE ENERGY CONSUMPTION OF THE STUDIED CONTROL SYSTEMS

   To calculate the electricl power required for an air blower, the following equation can be used[12]

   Fig. 5. Response of a cascade ammonia control system

   1. Multivariable predictive control system

   = ×

   ×1000

   × 60 × 24 (9)

   The predictive controller reads the dissolved oxygen sensor and nitrogen sensor, predicts the system response and then gives commands to control the air blower according to the reference value of dissolved oxygen which is around (mg/l (2) and the reference value of nitrogen (1.5 mg/l) which in turn helps in reducing the amount of organic matter produced in the treated water.

   The predictive control model (MPC) in the Simulink environment was used to control the aeration process according to the biological load (ammonium nitrogen) and the dissolved oxygen value. The bulk model (designed aeration system and

   : Electrical power consumed per day (kw)

   Q: Air flow rate (m3/min).

   P: Pressure (Bar), which is 0.8 for the blower used.

   : Blower efficiency, ranges between (65% and 85%).

   We added 1000 to convert to kilowatts and 60 to calculate kilowatt-hours, then multiply by the number of operating hours to calculate daily consumption. This results in the electrical power consumed in the ventilation control systems studied previously, in sequence:

   biological processes inside the aeration basin) was included in the controller and appropriate restrictions were set for the ammonium nitrogen value (0-50 mg\l) and the dissolved oxygen

   ()

   = 90×0.8

   0.75×1000

   56×0.8

   × 60 × 15 = 86.4/

   value (0.5-2 mg\l). The number of samples that would be predicted was set (10 samples) and then the response was

   (4) = 0.75×1000 × 60 × 9 = 32.25/

   followed up. Figure (6) shows a model of the proposed predictive control system.

   ()

   = 33×0.8

   0.75×1000

   × 60 × 6 = 12.67/

   Fig. 6. Simulink model for the proposed predictive control system.

   Figure (7) shows the changes in the concentration of dissolved oxygen and the concentration of ammonium nitrogen for the proposed predictive control system. We notice that the concentration of dissolved oxygen increases gradually after the air blower is turned on, so that the system stabilizes at the value of 1.2 mg/l after about 6 hours, and the air flow rate ranges around 33 m3/min, which means a decrease in the value of the electrical energy consumed. We also notice that the concentration of ammonium nitrogen stabilizes after about 3 hours at the value of 2.3 mg/l, which means speed in processing the biological load.

   We compared aeration control methods and determined the best in terms of processing speed and energy efficiency, as shown in the following table.

   TABLE III. Comparison of aeration control methods

   control system	treatment speed	Energy consumption
   Cascade oxygen control	15 hours	86.4/
   Cascade ammonia control	9 hours	32.2/
   predictive control	3 hours	12.6/

   We note from the previous table, Table (3), that the predictive control of the ventilation process contributed to reducing the treatment time to three hours compared to other methods. This means increasing the amount of wastewater that can be treated, and reducing energy consumption by more than 85% compared to the method of sequential oxygen control (used in the station currently being studied).

5. DESIGN OF APROTOTYBE FOR AERATION PROCESS

We implemented a prototybe consisting of an aeration tank and a secondary sedimentation tank with the aim of testing the predictive control method. The model was built of glass and the tanks were connected by plastic pipes for the activated sludge to flow through them as shown in Figure (8). A 12V DC motor was used for surface aeration, and a small 5V air blower.

Fig. 8. Aprototybe for aeration process.

1. Design of a control panel

   The Arduino Uno was chosen to connect the MATLAB- designed model to the previous laboratory model via the serial port. The Arduino reads the dissolved oxygen and nitrogen sensors and sends them to the computer. The designed model then predicts the biological load and sends control commands to the Arduino, which contains relays that drive the air blowers and the sludge return pump. The sensors were replaced with variable resistors to overcome the slow change in response to the biological load. The following figure, Figure (9), shows the implemented control circuit, which includes the Arduino Uno, variable resistors (instead of sensors), a control relay, and a power supply circuit.

   Fig. 9. A control panel

   In order to connect the model designed in MATLAB to the external control circuit, the serial port "UART" on the computer and the "Instrument control toolbox" library in the Simulink environment were used.

   We built a Simulink model to receive data on the serial window, as shown in Figure (10). This model enables us to read the data received from the control panel and separate the sensor signal for both nitrogen and dissolved oxygen in preparation for linking them with the predictive controller. After the data is read and

   processed in the predictive controller, the control commands will be sent to the control circuit to turn the air blowers on or off.

   Fig. 10. A Simulink model for sending and receiving data on the laptob port

2. Model Test:

In order to monitor the behavior of the implemented control system and study the response, the ammonium nitrogen signal was recorded for 100 minutes using a variable resistor. The following figure, Figure (11), illustrates the Recorded signal of change in ammonium nitrogen concentration

Fig. 11. Recorded signal of change in ammonium nitrogen concentration.

We note that the concentration of ammonium nitrogen increases as a result of the variable drainage flow until it reaches a value of 100 mg/l, then it decreases again as a result of pumping air into the water and stabilizes after about 10 minutes at a value of

10 mg/l. In this signal, we assumed that the sludge flow continues so that the concentration of ammonium nitrogen increases again and then stabilizes at a value of 20 mg/l after about 60 minutes.

In order to observe the system response and due to the unavailability of the required sensors, the ammonium nitrogen signal was recorded on the output of the system designed for the station, as shown in Figure (12).

Fig. 12. Ammonium nitrogen concentration in the predictive control system.

By observing Figure (12), we find a great match with the applied signal, which means that the more accurate the model is, the more accurate and flexible we get in monitoring changes. The system also stabilizes after 60 minutes at an oxygen value of 0.5 mg/l and a nitrogen value of 20 mg/l, which are within the permissible limits for each.

DISCUSSION OF RESULTS

We built a model of the aeration process in an activated sludge wastewater treatment plant in the Simulink environment, and linked the model to three control systems to monitor the response of each system and the amount of electrical energy consumed.

It was found that the use of sequential ammonia control reduced the treatment time to nine hours and contributed to a good energy saving value. As for the predictive control, it reduced the treatment time by about 80% compared to the sequential dissolved oxygen control method used in the station (100 ×

()()), In addition, it enables us to predict the

()

amount of biological load and thus reduce the amount of air required for the treatment process, thus saving 85% in energ consumption (100 × ()()).

()

A control panel for the aeration process was implemented and linked to the designed model via the computer serial port. As a result, we were able to read the sensors and send control commands in real time.

CONCLUSIONS AND RECOMMENDATIONS

Based on the work we have done in this research, we have concluded that the use of predictive controllers improves the ventilation control process in wastewater treatment plants in terms of response speed and energy savings compared to other control methods used:

• The use of the proposed predictive controller leads to a reduction in processing time (3 hours).

• The ammonium nitrogen value stabilizes at 2.3 mg/l after approximately 3 hours, thus reducing the residence time inside the aeration tank, which contributes to increasing the amount of treated wastewater.

• The proposed predictive controller contributed to reducing electrical energy consumption (by 85%).

• The more accurate the designed model, the more effective the predictive controller we get.

For future work, we recommend integrating the proposed predictive controller with an AI technique such as fuzzy logic to increase the efficiency of the controller.

REFERENCES

1. J.Drewnowski, A.R.Skwarek, S.Duda and G. agód, Aeration Process in Bioreactors as the Main Energy Consumer in a Wastewater Treatment Plant. Review of Solutions and Methods of Process Optimization, 2019. doi:10.3390/pr7050311.

2. R.Jafar, H.Sabboh and M.Abd Alkader, Modelling and Simulation of Wastewater Treatment Plant in Small Agglomeration in The Coastal Zone, 2013. 35(9)Tishreen University Journal.

3. H.Jnad, A.Dabaliz and H.Mohammad, An evaluation of the distribution of DO concentrations in aeration tanks at Lattakia WWTPs, 2015. 37(2)Tishreen University Journal.

4. Victoria R. and all., Impact of Ammonia-Based Aeration Control (ABAC) on Energy Consumptio, 2020. doi:10.3390/app10155227.

5. M.Miyahara, M.Nakata, K.Toyooka and M.Takakura, Aeration Control by Real-Time Nitrification Control Utilizing Activated Sludge Model, 2019. Meiden Review series No.176.

6. N.F.Aziz, N.A.Ramli and M.F.Abd Hamid, Energy efficiency of wastewater treatment plant through aeration system, 2019. doi:10.5004/dwt.2019.24134.

7. N.Ballhysa, S.Kim and S.Byeon, Wastewater Treatment Plant Control Strategie, 2020. http://dx.doi.org/10.7236/IJASC.2020.9.4.16.

8. Y.Chen, H.Zhang , Y.Yin, F.Zeng and Z.Cui, Smart energy savings for aeration control in wastewater treatment, 2022. Energy Reports 8. 1711 1721.

9. Z.Wang. and all., XGB SEGA coupled energy saving method for wastewater treatment plants, 2024. Applied Water Science, Vol (14).

10. R.Katebi, M.A.Jonson and J.Wilkie, Control and instrumentation for wastewater treatment plants, 1999. Advances in Industrial Control, Springer, London, UK, ISBN: 1-852-33054-6.

11. M.Daod, F.Motawej and R.Abdullah., Improvement of Aeration System in Wastewater Treatment Plants using an Ammounium Cascade Control System, 2021. Al-Baath University Journal, Vol(43).

12. Y. Amerlinck, W. De Keyser, G. Urchegui and I. Nopens., Arealistic dynamic blower energy consumption model for wastewater applications,2016. Water Science & Technology, 74.7.

13. Amand, L., Ammonium Feedback Control in Wastewater Treatment Plants, 2014. 104. 256 pp. Uppsala: Acta Universitatis Upsaliensis.

ISBN 978-91-554-8900-7.