Abstract
The developed CFD-based model has been used to study the velocity profiles of OBS. The model is very useful tool in view of the fact that the OBS is having complicated hydraulic patterns that always cause a high internal recycle flow or the cycling of mixed liquor throughout the system. The developed CFD model has also been used to assess the current operational performance that is related to aerations shaft of the OBS. To date, limited information is available on the use of CFD simulation for operational performance study. The CFD analysis performed in the research may enhance the use of CFD in this field. The CFD application in wastewater industries has expanded and the confidence level of using it has also increased. This research has also lead to more studies related to this field. All pros and cons of this research can be used as the guideline for a better CFD application in the related research. When further understanding is achieved in CFD-based process modelling, anyone working in this related field may possibly accomplish a better knowledge or experience to put forward possible development in the CFD models themselves.
Introduction
Biological wastewater treatment is viewed as the most important and very complex unit processes in wastewater treatment [4, 1]. In order to deal with the complexity of the biological processes, it is crucial to understand the entire concept of the processes. Oxidation ditch (OD), which is an Extended Aeration (EA) system, consists of oval-shaped or a ring channel equipped with aeration devices. Screened wastewater flows into the channel and is combined with the return activated sludge [5]. Anoxic and aerobic condition is developed and maintained in zones up-stream and downstream of the rotor [6]. OD process is a highly reliable process and competent of treating shock loads without having an effect on the effluent quality [5]. The Orbal process is one of the modified oxidation ditches that applying a sequence of concentric channels inside a single biological reactor. OD is better than other biological treatment system because of its unique mixing performance [7].
The Orbal Biological System (OBS) is a modification of OD, which is equipped with mechanical aeration and mixing devices. OBS uses a series of concentric channels within the same structure [5], with the outer channel having half of the total volume. Screened wastewater enters the outer channel and flows from there to middle channel and finally to inner channel before the mixed liquor flows to the clarifier. Return activated sludge (RAS) from the secondary clarifier is also added to the outer channel. The channels are interconnected and the flow is directed in an inward direction.
3D model case setup
3D model case setup
Geometry layout of 3D model.
The flow in a biological reactor such as OBS is complex and usually cannot be accurately represented by simple techniques such as this current process model. In order to get a clearer picture of the real conditions inside the OBS, it is necessary to attempt to model the distribution of (Oxygen Utilization Rate) OUR throughout the channels. This objective can be achieved using more advanced modelling techniques that potentially gives a better description of the system. In this case, the CFD was chosen as the tool to extend the preliminary process model.
A three-dimensional CFD model was developed to represent the OBS. Unsteady state simulations or also known as transient model was carried out to take into consideration the time variations inside the ditch. Two-phase models (liquid-gas) was developed and an open channel model was defined to capture the actual flow process. The developed model has the ability to determine the properties of the flow such as velocity at any location inside the OBS. It also has the capacity to visualize the distribution of the Hydraulic Residence Times (HRT) throughout the channels. In CFD modelling, the target spatial region was meshed into interconnected elements and the flow field was calculated for each element by solving the governing equations [8]. Grid independent tests were performed in order to make sure that the selection of grid size is optimal. Three dissimilar sets of meshes with the number of cells of 889,303, 989,933 and 1,077,463 were selected to simulate the velocity magnitudes at the sampling points inside the OBS. As a result, the second set was chosen for all further computations, in view of the low difference of velocity magnitudes simulated with the best possible mesh (989933 cells) and the refined mesh (1077463 cells). For the purpose of calibrating and validating the models, sampling and water quality analyses were performed.
In general, the water entering the OBS consists of impurities in the form of suspended particulates, or other fluid impurities. This model did not incorporate these impurities. In order to consider these impurities, it comes under the domain of multiphase flow regimes, which requires the defining of more accurate description of their properties and their composition, which is a difficult scenario considering the extent of OBS and varying nature of impurities that enter the OBS channel.
Contours of velocity magnitude (m/s) at 4-meter depth.
Contours of velocity magnitude (m/s) at 3-meter depth.
Contours of velocity magnitude (m/s) at 2-meter depth.
Contours of velocity magnitude (m/s) at 1-meter depth.
Contours of velocity magnitude (m/s) at cross section near to shafts’ location.
Vectors coloured by velocity magnitude at 4-meter depth.
Vectors coloured by velocity magnitude at 4-meter depth (near to aeration shafts’ location).
Vectors coloured by velocity magnitude (m/s) at 4-meter depth (near to penstocks’ location).
In this study, volume of fluid (VOF) approach of the multiphase model was applied. Solver setting and boundary condition of the 3D model simulation are summarized in Table 1. The segregated solver of Fluent 15.1 was used with the default parameter settings applied. 3D geometrical model of the OBS is shown in Fig. 1.
Figure 2 shows the contours of velocity magnitude as observed on the top surface of the ditch. It is evident from this figure that, the maximum flow velocity is at the mixers, which helps to maintain the flow after the aeration discs. Figures 3–5 show the flow distribution at various depths inside the ditch. This is main advantage with 3D CFD modelling. The flow features can be obtained at any depth or location in the mixing tank. Further to this observation, Fig. 6 provides a cross sectional views of velocity profiles at different locations near the shafts. It can be seen that the velocity distribution varies from a maximum on the top surface to the minimum at the bottom of the tank. The first half of the tank shows high velocity gradient inferring better mixing due to the action of the shafts. The vector plot as can be seen in Fig. 7 depicted the right vectors of the flow. The flow’s vectors near to the aeration shafts (refer to Fig. 8) and vectors of the flow through the penstocks (pictured in Fig. 9) have also indicated that the model has successfully represented the OBS. They represent flow of water from the outer channel into the middle channel as well as from the middle channel to the inner channel.
In contrast to the 2D velocity, the maximum velocity obtained was much lesser in the 3D simulation. The scenario is due to the influence of the 3D movement of the fluid, where there are variations in flow velocity throughout the ditch in both vertical and horizontal directions. Furthermore, in 3D model, transverse directions were influenced by the spinning speeds of the aeration discs inside the ditch [8], where the rotational plane of the mixers tends to disturb the overall velocity flow of the ditch.
Velocity magnitudes given by 3D CFD model.
Comparison of the velocity magnitude given by 3D CFD model and onsite observation.
Average velocity magnitude at different depth.
Velocity magnitudes at each sampling point have been extracted and transferred into line graph as demonstrated in Fig. 10. Based on the simulation outputs, the average flow velocity throughout the ditch was 0.76 m/s. Maximum velocity value was 0.851, while the minimum value was 0.591. Velocity magnitudes all over the ditch were found to be heterogeneous with the highest velocity value revealed at the downstream of the outer channel. This may be related to the highest number of aeration discs inside the outer channel and also caused by the operation of aeration discs at the curve bend before the sampling point of the downstream of outer channel. Downstream of the inner channel has shown the lowest velocity value among all the sampling locations. This may be due to the inertia force and centrifugal force, where the fluid has been dragged towards the weirs and energy loss has occurred [10]. The variations of the flow velocity which caused high and low velocity zones inside the ditch have been pointed out by previous workers [3, 8, 9, 10]. The overall results have shown an acceptable match to the actual onsite measurements (as shown in Fig. 11), thereby validating our 3D CFD model. Thus, this advanced CFD model can now be used for more detailed study on hydraulic behaviour of OBS.
The simulation outputs have also established the overall velocity magnitude at various depths of the OBS as shown in Fig. 12. The average velocity at the top (4 m height) and bottom (1 m height) of the OBS were 0.825 m/s and 0.795 m/s, respectively. It can be noticed that, the values of velocity at various depths were slightly varied. The velocity at the top of the ditch was higher than the bottom part. However, the velocity values at 1 m, 2 m and 3 m were not much different. This may be attributable to the consistent acceleration produced by the aeration discs. This means that the current submergence of the aerations discs has given a good mixing pattern where the velocity values at 1 m, 2 m and 3 m are almost consistent.
Generally, the solids build up at locations where the velocities are relatively low [2]. The higher velocities at the bottom of the ditch may possibly help to keep solids in suspension. Since the mixers were producing a good mixing pattern where the velocities in vertical directions were not much varied, maybe another aspect that can be focussed on is the speed of the aeration discs. The speed of the aeration discs can be increased to prevent the sludge settling at the bottom of the ditch. As pointed out in the previous research, increasing the rotational speed of the aeration discs will decrease the variations of the flow speed [8]. Onsite measurement makes it almost impossible if not difficult to gather velocity distribution at different depth of wastewater level. Now, using the 3D CFD model, the velocity profiles at different depth of the system are easily measurable.
Footnotes
Acknowledgments
The authors gratefully acknowledge support from Universiti Sains Malaysia and financial support from the grant of Higher Institution Centre of Excellence (HICoE) (311/PREDAC/44039010). Many thanks to the director, all colleagues and staff of REDAC for their wonderful collaboration throughout this study. Their assistance would be greatly appreciated.