2018 International Conference on Modeling, Simulation and Optimization (MSO 2018) ISBN: 978-1-60595-542-1

 

Simulation of Oxidation Ditch with STOAT

Oleyiblo Oloche James1,*, Er-yan Zhao1, Tao LI1 and Gan WANG2

1Anhui Guozhen Environmental Protection Technology Joint Stock Co., Ltd, Hefei 230088, China 2Anhui PLAKeco Ecological Sci. & Tech. Co., Ltd, Hefei, China

*Corresponding author

Keywords: Activated sludge, Dynamic simulation, Oxidation ditch, STOAT.

Abstract. This paper examines the use of the modified version of ASM model referred to as IAWQ#2B implemented in STOAT simulation package to model the performance of oxidation ditch process. STOAT is a free simulation package for comprehensive modeling and simulation of long-term treatment plants. The dynamic model was calibrated and validated using real operating data from the wastewater treatment plant. The model was able to produce good match in terms of COD, TN, MLSS concentrations, and effluent TSS without adjustment to any default model parameters apart from the wastewater treatment plant operating parameters, and the influent wastewater characteristics which were determined during the data collection campaign. Additionally, the use of chemical polishing for total phosphorus removal (TP) was reasonably predicted by the model. The model was able to identify ‘on-off aeration operation’ as the reason for ammonia spike at the WWTP.

Introduction

The main benefit of using a computer model to simulate wastewater treatment plant (WWTP) is its ability to evaluate different operational conditions in order to develop an optimum operational strategy. Models are useful tools for troubleshooting and optimization of existing WWTPs [1,2,3,4,5,6]. Variants of ASM models on the platform of different simulators have been used for the purpose of optimizing the biological nutrient removal (BNR) process in municipal WWTPs in several research work [7,8,]. However, information of such application of ASM variants of models on the platform of STOAT simulation package in such research publications is rare. In 2010, Water Research Centre (WRc) has made its dynamic wastewater treatment modelling program, STOAT, available as freeware in order to encourage the uptake and understanding of wastewater treatment process modelling. STOAT is a simulation package for comprehensive modeling and simulation of long-term treatment plants. It is based on the generally accepted activated sludge models ASM1, 2, 2d and 3, as well as STOATs’ own models. It can simulate both simple and complex processes, for municipal, as well as industrial WWTPs. In addition, STOAT uses Takacs model to evaluate settling and separation processes. This study aimed at assessing STOAT capability to simulate the performance of full-scale WWTPs. An oxidation ditch process was chosen for the assessment.

Oxidation ditches are variants of the activated sludge system. They are single sludge wastewater treatment systems, which is capable of achieving carbon oxidation, nitrification, denitrification and phosphorus removal in single biomass slurry because of the presence of aerobic, anoxic and anaerobic zones [9]. Oxidation ditches has a number of advantages over the activated sludge systems because of their unique features, and mode of operations. Aerators are located in series along the ditch channel, providing alternating aerobic and anoxic zones along the ditch. Thus, simultaneous removal of organic and nitrogenous matter takes place continually in oxidation ditches. It requires only 10 to 15 minutes for wastewater to recirculate around the ditch. Therefore, biomass undergoes rapid transformation between aerobic and anoxic conditions, and consequently, encourages the growth of various types of microorganism in the oxidation ditches. The recirculation flowrate in cubic meter per hour (m3/h) is about 4-6 times the total volume in cubic meter (m3).

115

Oxidation ditches have good mixing and good buffer against shock loads because of the high internal recirculation rate, coupled with the high turbulence induced near the aerators. Oxidation ditches has low food to micro-organism ratio because they are operated at an extended aeration mode (high solids residence time), and they produce well stabilized sludge that has little odor problems [10]. In addition, the high rate of nitrate recirculation in oxidation ditches usually leads to a significant reduction in the amount of oxygen needed for oxidation, as nitrate, instead of oxygen, is used as a terminal electron acceptor in the denitrification processes, thus, making the operation of oxidation ditches cheaper than the operation of other conventional WWTP's. Generally, oxidation ditches produces very low effluent nitrogen concentration because of the efficient simultaneous nitrification and denitrification that take place within the system [11].

Materials and Methods

The WWTP is located in Cai Tian Pu, Anhui Province People’s Republic of China. The plant is operated and maintained by Anhui Guozhen Environmental Protection Technology Joint Stock Co., Ltd. (GZEP) in corroboration with the company’s environmental engineering design and research institute (GZED). The system comprises of two-train oxidation ditches (OD), each of which has pre-anaerobic selector. The total volume of each OD is 18675 m3, whereas the volume of each pre-anaerobic selector tank is 3679.5 m3. Each OD operates in alternating anoxic aerobic modes, and a secondary clarifier is connected to each OD train. Each of the clarifier has a total surface area of 1808.64 m2. The return sludge is pumped to the inlet of the reactor and mixed with the influent. Each OD is equipped with three aerators. Prior to this study, the WWTP was having issues with ammonium spike, though not violating the effluent criteria. This case study is a report on one of the various trainings on the application of STOAT for WWTP simulation. The simulation was performed in accordance with the “Good Modeling Practice” which was developed by Rieger et al. [12].

In order to familiarize with the WWTP layout, to identify the locations of important sampling and monitoring locations, to obtain input from plant operators regarding equipment hydraulics and process limitations in the plant based on their operating experience, and to evaluate the plant performance based on the available historical data, two of the trainings were conducted at the WWTP site.

Seventeen days sampling campaign was carried out in March for the purpose of proper wastewater characterization and model calibration, whereas seven days sampling campaign was conducted in June to validate the model. Influent flow rate in cubic meter (m3) to the treatment plant was measured online. STOAT accepts flow in cubic meter per hour (m3/h). Samples of influent composition were taken every 2-hour for 24-hour during the sampling campaign. The parameters of interest are: total COD (non-filtered COD), and filtered through 0.45-μm and 1.2-μm diameter (soluble influent COD, filtered and flocculated), BOD5, volatile fatty acids (VFA), ammonia nitrogen (NH3-N), nitrate nitrogen (NO3-N), total Kjeldahl nitrogen (TKN), total phosphorus (TP), ortho-phosphate ( 3

4PO ), total suspended solids (TSS), volatile suspended solids (VSS) and non-volatile suspended solids (NVSS) concentrations. The parameters were determined in accordance with the standard methods [13]. Also, in order to describe the biological conversion as a function of reactor length, dissolved oxygen (DO) was measured over the length of oxidation ditches at different locations. The average sampling results are presented in Table 1.

Apart from volatile fatty acids (VFA), STOAT differentiates four influent COD’s, these are: soluble biodegradable COD (SS), soluble non-biodegradable COD (SI), particulate biodegradable COD (XS), and particulate non-biodegradable COD (XI).

116

Table 1. Average influent wastewater characteristics during sampling campaign in March.

Parameter Max Min Average SD SRT, d 31 24 27 - Flow, m3/d 54016 20772 36648.4 11858 CODtot, mg/L 343 102 248 78.5 SS 121 36 87.5 24.7 SI 30.5 9 21 8.2 XS 148 44 107 39 XI 37 11 25.5 10.2 BOD5, mg/L 209 86.6 161.3 22.4 TKN, mgN/L 54 14.8 37.6 9.8 NH3, mgN/L 43 11 29.3 7.2 NO3, mgN/L 2.8 0.11 0.83 0.49 TP, mgP/L 5.38 1.56 3.55 0.67 PO4-P, mgP/L 3.98 0.88 2.41 0.24 TSS, mg/L 223 111 156 51 VSS, mg/L 167 72 117 42 NVSS, mg/L 55 24 39 14 PH 7.72 7.52 7.64 0.04 Temperature 13 10 11.2 -

In STOAT, oxidation ditches are specified as a number of stages in series, and the number of stages depends on the particular ditch being modelled. It is assumed that the effluent for the aeration basin always leaves in the last stage, and all the stages are equal in volumes. In this study, the ditch is divided into 10 stages, with the effluent taken from the last stage. The actual dissolved oxygen (DO) concentrations for each of the compartments, as measured at the WWTP during the measurement campaign were assigned, and controlled in STOAT by modifying the KLa values for each of the compartments. The minimum and maximum DO values for the aerobic compartments were 0.5 and 1.4mg/l, respectively. The STOAT configuration of the WWTP is shown in Figure 1.

Figure 1. STOAT configuration of the Cai tian pu WWTP.

Results and Discussion

Model Calibration and Validation

Model calibration and validation are two crucial steps involved in process simulation [14]. In wastewater treatment plant modelling, the calibration step is essential, and it’s generally performed using steady state mode. The model does not consider the effect of time in steady state simulation mode, because, it is assumed that the plant has reached the steady state operating conditions. The aim of steady state calibration is to obtain average effluent concentrations in the same order of magnitude with measured data, and to have the correct solids balance to enable good initial conditions for the dynamic simulation [15]. However, in STOAT, there is no steady state simulation package.

Therefore, in order to calibrate the model, sewage flow rate and the influent compositions which were determined during the data campaign were used to generate influent profile. The model was

117

run repeatedly for fifteen days until a dynamic equilibrium was reached. This is indicated by obtaining similar effluent results for subsequent cycles [16]. Dynamic equilibrium run results are shown on Table 2. The differences in effluent concentrations between the subsequent runs (runs 2 to 6) of all the parameters were less than 0.1mg/L except for COD which was approximately 0.36mg/L. Apart from the effluent concentrations, the MLSS in the bioreactor also reached dynamic equilibrium (Table 2).

Table 2. Results of dynamic equilibrium.

Parameters TSS COD NH3 NO3 TP TN MLSS

R1 10.73 23.25 1.88 8.47 0.21 10.6 4995 R2 10.43 25.85 0.43 8.74 0.29 9.62 4962 R3 10.19 26.37 0.33 8.66 0.33 9.55 4933 R4 10.15 26.80 0.30 8.64 0.32 9.60 4932 R5 10.12 27.10 0.29 8.63 0.34 9.63 4910 R6 10.11 27.28 0.28 8.63 0.33 9.56 4924 Avg diff 0.08 0.36 0.04 0.03 0.01 0.04 16.5

Note: Avg diff = average difference

Having reached the dynamic equilibrium, the next step was to compare the model and the calculated results. Discrepancies in the results would require the modeler to determine which default parameters should be modified to achieve a better fit. The model should be re-run with the adjusted parameter(s), if there is any, until a better match is reached. Three important points were considered when determine the goodness of fit, these are: the average values of the measured and simulated data, the peaks and trough of the measured and the simulated data, and the timing of the peaks and trough. However, a comparison of the results (Table 3) shows no substantial differences between the measured and simulated values that could require modification of any model default parameters.

Table 3. Average measured and simulated effluent concentrations.

Parameter COD NH3 NO3 TP TN MLSS Measured 27.64 1.14 - 0.29 9.2 5168 Simulated 26.80 0.30 8.64 0.32 9.60 4932

The only notable difference was between measured and simulated ammonia values (Table 3), the model over predicted ammonia, and this was possibly caused by the on-off aeration operation system. Hence, further simulation was performed with the implementation of the on-off aeration operation system, and the results are presented in Table 4. Though there was great improvement, nevertheless, the model slightly over predicted ammonia. However, it should be noted that whereas constant values of dissolved oxygen (DO) concentrations were maintained in the model for each of the oxidation ditch compartments during the entire simulation time, the same could not be said to have been maintained in the real plant, and this could be the likely reason for the differences.

Table 4. Average measured and simulated effluent concentrations for on-off aeration operation.

Parameter COD NH3 NO3 TP TN MLSS Measured 27.64 1.14 - 0.29 9.2 5168 Simulated 26.53 0.98 7.83 0.33 8.9 4928

The second step in the calibration process is the dynamic simulation. In STOAT, this is accomplished by comparing the peaks and troughs of the measured and the simulated data, and the timing of the peaks and troughs. Hence, the simulation results of the on-off aeration operation are plotted alongside with the measured data (Figure 2).

118

Figure 2. Dynamic simulation graph showing measured and simulated peaks, troughs and timing.

The model was validated, using the data collected in June (Table 6), and the validation results are presented in Table 5 and Figure 3, respectively.

Table 5. Average measured and simulated effluent concentrations for model validation.

Parameter COD NH3 NO3 TP TN MLSS Measured 27.1 1.38 - 0.32 6.2 4311 Simulated 26.79 0.59 4.78 0.32 5.54 4226

Figure 3. Model validation graph showing measured and simulated peaks, troughs and timing.

Table 6. Average influent wastewater characteristics during sampling campaign in June.

Parameter Max Min Average SD SRT, d 29 22 25 - Flow, m3/d 65071 54283 61619 3280 CODtot, mg/L 238 164 206 25.8 SS 78 53 67 8.5 SI 28 19 25 3 XS 89 63 79 10 XI 39 27 34 4.3 BOD5, mg/L 130 94.2 110 15.5 TKN, mgN/L 38 18.5 30.5 6.4 NH3, mgN/L 24.7 11.2 21.5 4.4 NO3, mgN/L 2.1 0.1 0.9 0.65 TP, mgP/L 4.53 2.23 3.7 0.93 PO4-P, mgP/L 3.26 1.2 2.48 0.34 TSS, mg/L 182 102 153 24.9 VSS, mg/L 135 78 112 17.9 NVSS, mg/L 56 32 48 7.7 PH 7.68 7.49 7.58 0.06 Temperature 27 22 26 -

An analysis of both the calibration and validation results (Table 3, Table 4 Table 5, Figure 2, Figure 3, respectively), shows that the deviation between the measured and simulated output variables by STOAT are very negligible, this is an indication that the model calibration was performed correctly [17]. Therefore, STOAT ability to simulate long-term treatment plant performance was further tested by applying the calibrated model to a two-month routinely measured data. The simulation results, comparing the measured and simulated values is shown in Figure 4.

119

Figure 4. Results of using two-month routinely measured data.

It is obvious from Figure 4, the model reasonably predicted the behavior of the wastewater treatment plant with the routinely measured data, indicating that the WWTP was well represented in the model, and the calibration was properly performed.

Conclusions

In this work, the calibration and validation of the STOAT free dynamic simulation package to a full-scale Oxidation ditch wastewater treatment process was performed successfully. The model was able to produce good match between measured and simulated variables of interest. The success of the model calibration is attributed to proper influent wastewater characterization which was performed during the sampling campaign, as well as, the good representation of the treatment process in the STOAT simulation package. The model proved to be capable of mimicking processes taking place at the wastewater treatment plant. The model was helpful in identifying ‘on-off aeration operation’ as the reason for ammonia spike at the WWTP. This work should encourage the use of free simulation packages in pursuit of reducing water pollution problems and the cost of treatment plant upgrade. By doing so, we can easily share knowledge on the existing free simulation packages and thereby combating water and environmental pollutions at a cheaper cost.

Acknowledgement

The authors are grateful to Anhui Guozhen Environmental Protection Technology Joint Stock Co., Ltd., for supporting this work. Included in the acknowledgment is the staff of Cai tian pu WWTP for not only participating in the training exercises, but also provided assistance in data collection and analyses. All the staff who attended the training is acknowledged for their various inputs. Special thanks to Engineer Li Tao, the departmental head of Guozhen Environmental Engineering Design and Research Institute, for organizing and coordinating the entire training program. WRc plc is also acknowledged for making STOAT as a freeware and for providing a platform for STOAT users.

References

[1] Guisasola, M. Pijuan, J.A. Baeza, J. Carrera, J. Lafuente. Improving the start-up of an EBPR system using OUR to control the aerobic phase length: a simulation study. Water Sci. Technol., 53(2006) 253-262

[2] K.J. Kennedy, A.M. McHarg. Optimization of municipal wastewater for biological nutrient removal using ASM2d. J. Environ Eng. Sci., 6(2007): 31-43.

[3] A. Griborio, J. Rohrbacher, R. Taylor, P. Pitt, R. Litmer. Evaluation of Wet Weather Strategies and Clarifier Optimization Using State-of-the-Art Tools. In: Proceeding of the 81th Annual Water Environment Federation Technical Exposition and Conference, Chicago, IL. p. 4 (2008): 957-4 968.

[4] H. Andres, M. Newbigging, J. Bicudo. Using a Wastewater Process Model to Evaluate the Capacity of the Galt WWTP Under Increased Load from an Industrial Source. Proceedings of the 37th Annual WEAO Technical Symposium, Blue Mountain, Ontario, 2008.

120

[5] X.F. Zhou, Y.L. Zhang, M.D. Hu, W. Shi. Optimization design for nutrients removal process with activated sludge 2D model. CIESC J., 60(2009b): 3 122-3 129. (in Chinese with English abstract)

[6] O.J. Oleyiblo, J. Cao, Q. Feng, G. Wang, Z. Xue, F. Fang. Evaluation and improvement of wastewater treatment plant performance using BioWin. Chin. J. Oceanol. Limnol. 33(2014): 468-476

[7] M.S. Moussa, A.R. Rojas, C.M. Hooijmans, H.J. Gijzen, M.C.M. van Loosdrecht. Model- based evaluation of nitrogen removal in a tannery wastewater treatment plant. Water Sci. Technol. 50(2004) 251–260.

[8] D. Brdjanovic, M. Mithaiwala, M.S. Moussa, G. Amy, M.C.M. van Loosdrecht. Use of modelling for optimization and upgrade of a tropical wastewater treatment plant in a developing country. Water Sci. Technol. 56 (2007) 21–31.

[9] WEF. Biological and chemical systems for nutrient removal, A special publication, Municipal subcommittee of the technical practice committee, Water Environment Federation, 1998 (WEF), USA.

[10] A.T. Van der Geest, W.C. Witvoet. Nitrification and denitrification in a carrousel system, Progress in Water Technology, 8(1977): 654-660.

[11] D. Orhon, N. Artan. Modeling of activated sludge systems, Technomic Publishing Company, Inc. Lancaster, Pennsylvania, USA. 1994.

[12] L. Rieger, G. Langergrabe, T. Ohtsuki, A. Shaw, I. Takacs, S. Winkler. Guidelines for Using Activated Sludge Models. IWA Scientific and Technical Report No. 22, 2012. IWA Publishing, London.

[13] APHA, AWWA & WEF 2005 Standard Methods for the Examination of Water and

Wastewater, 21st Edn. American Public Health Association, Washington, DC, USA.

[14] J. J.W. Hulsbeek, J. Kruit, J.P. Roeleveld, M.C.M. van Loosdrecht. A practical protocol for dynamic modelling of activated sludge systems. Water Sci. Technol., 45(2002): 127-136.

[15] J. Cao, O.J. Oleyiblo, Z. Xue, Y.M. Otache, Q. Feng. Achieving low effluent NO3-N and TN concentrations in low influent chemical oxygen demand (COD) to total Kjeldahl nitrogen (TKN) ratio without using external carbon source. Chin. J. Ocean. Limnol. (2015) 33: 468.

[16] A.J. Stokes, C.F. Forster, J.R. West, M. De Bel. Modelling of activated sludge process at Chorley WWTP using STOAT. In ISEB4: Proceedings of the 4th International Symposium on Environmental Biotechnology, 75-78, Noordwijkerhot, The Netherlands, 2000.

[17] E. Liwarska-Bizukojc, R. Biernacki, D. Gendaszewska, S. Ledakowicz. Improving the operation of the full scale wastewater treatment plant with use of a complex activated sludge model. Environ. Protect. Eng. 39(2013) 183–195.

121