Activated Sludge Modeling: Part II

In my previous blog, I talked about the capabilities of FLOW-3D’s new activated sludge model (ASM) and standard mathematical systems that describe and numerically solve the biochemical reactions inside the aeration tanks. In this blog, I will show FLOW-3D’s ASM solver results for the decay and growth of various influent species concentrations as they exit a wastewater treatment plant (WWTP) in Zele, Belgium. Accurately tracking the species and hydrodynamics calculations can equip wastewater treatment professionals to take quantitatively backed design and operations decisions.

The Zele WWTP

The Zele WWTP was constructed in Belgium in 1983 for 50,000 inhabitants. Typically, the influent at this WWTP consists of 40% household wastewater and 60% industrial wastewater. After the primary treatment process, the influent flows to the biological activated sludge treatment site, where it is mixed with the recycled activated sludge.
Schematic of a wastewater treatment plant
Schematic [2] of a WWTP from Zele, Belgium. The green box delineates the secondary treatment process.

The activated sludge tank or the aeration tank consists of one plug-flow aeration tank that is divided into 6 lanes of about 400 m3 each. The effluent from the aeration tank goes to two secondary clarifiers (SC1 and SC2) each of volume 2050 m3. The final effluent is discharged into a nearby stream. From underneath the secondary clarifiers, a portion of the activated sludge is recycled back into the aeration tank to increase the efficiency of secondary treatment.

We chose this WWTP for our case study because of the availability of the detailed information about the geometry of the secondary treatment components and the influent concentrations of various species. The information is detailed but not complete and this incomplete information will have significant consequences on the effluent concentrations, which I will discuss later.

Geometry, meshing and physics

The geometry creation and meshing were straightforward. FLOW-3D has a suite of primitive geometry shapes that we used to completely define the complete WWTP. These shapes are easy to generate and are error-free unlike some geometries generated using external CAD software. Similarly, the use of structured grids saved time dealing with typical errors associated with an unstructured grid generation.

The physics inside the aeration tank are complex and require solving a complete system of mass and momentum conservation equations (Navier-Stokes equations), species transport, reaction kinetics, oxygen dissolution and continuous density evaluation. FLOW-3D accounts for all these physics in a fully-coupled way for the most accurate calculations.

Zele WWTP setup in FLOW-3D
The Zele WWTP setup in FLOW-3D. The arrows indicate the flow direction and the influent enters the domain at the beginning of the green arrow.
Out of the three standard mathematical models: ASM-1, ASM-2 and ASM-3, researchers use ASM-1 mathematical model in this WWTP because it is simple yet covers the many important biochemical processes. The ASM-1 model considers the evolution of 13 species that are typically found in wastewater or are generated during the treatment process [Table 1].
SpeciesSpecies IDInitial influent concentration at Zele (mg/l)
Soluble inert organic matterSI7.5
Readily biodegradable substrateSS400.0
Particulate inert organic matterXI40.0
Slowly biodegradable substrateXS40.0
Active heterotrophic biomassXB,H120.0
Active autotrophic biomassXB,A5.0
Particulate products arising from biomass decayXP0.0
OxygenSO0.0
Nitrate and nitrite nitrogenSNO0.0
Ammonium nitrogenSNH15.0
Soluble biodegradable organic nitrogenSND8.2
Particulate biodegradable organic nitrogenXND11.3
AlkalinitySALKNot included

Table 1. List of species in the standard ASM-1 mathematical system and the corresponding measured initial influent concentrations at the Zele WWTP. Some of these initial concentrations are deduced and are likely to have large uncertainties associated. S and X represent soluble and particulate matter, respectively.

Each of these species depends on one or more biochemical processes, except for the inert species (SI and XI) which do not react. The influent and effluent concentrations of the inert species may vary due to settling such as in the case of XI. SALK was not measured at the WWTP and consequently ignored in this case study.

Effluent quantities of interest

The main effluent quantities of interest to wastewater engineers are total chemical oxygen demand (CODtot), concentration of ammonium nitrogen (SNH), concentration of nitrites and nitrates nitrogen (SNO) and total Kjeldahl nitrogen (TKN), where

  • CODtot = SI + SS + XI + XS

  • TKN ~ XND + SND + SNH

These quantities indicate the overall quality of the treated water.

Effluent quantitiesMeasured influent concentration (mg/l)FLOW-3D effluent concentration (mg/l)
CODtot600264.04
SNH1530.34
SNO01.86
TKN3537.28

The concentration of total COD, SNH and TKN should decrease as the wastewater flows through the aeration tank and exits the WWTP. This behavior is correctly predicted for the total COD [Table 2], but not quite for SNH and TKN. The concentration of SNO is expected to increase, which is correctly predicted by the ASM solver. The concentration of all the effluent species is shown in the animation below.

The animation shows the simulation results for the evolution of all the species in the Zele WWTP.

Sensitivity of results to the WWTP data

I attribute the incorrect evolution of some of the species in the effluents to the assumptions in modeling and the missing WWTP data. The uncertainty in the measured species concentrations in the influent; missing information about the initial concentrations; and, the missing data about the settling properties for the particulate matter are likely to have affected the concentrations of species in the effluent.

Similarly, incomplete geometry specifications can negatively affect the accuracy of hydrodynamics calculations inside the WWTP. Additionally, there was only partial information about the oxygen sparging into the aeration tank. Oxygen is a crucial component that has major impact on the decay and growth of other species.

It may not be always possible to measure all the data in a WWTP. In such cases a calibrated numerical model can be effectively used as a virtual laboratory to test different WWTP designs. This case study shows that it possible to track the concentrations of species in the secondary treatment part of a WWTP, especially in the aeration tank. And this is can be done while considering the hydrodynamics effects. In the presence of complete WWTP data and problem specification, engineers and designers can take better informed decisions about the WWTP plant operations and design optimizations.

We are open to collaboration with researchers and professionals in the wastewater treatment industry to further develop and calibrate our activated sludge model. Email me at adwaith@flow3d.com to discuss your WWTP projects and research.