The WATS model

The WATS model (Wastewater Aerobic/anaerobic Transformations in Sewers) is a deterministic in-sewer process model for simulation of organic matter and sulfur transformations. The model consists of a number of non-linear differential equations describing the transformation of a number of different wastewater compounds. A software version of the WATS model has been developed by the SPN-Group and allows the dynamic simulation of a branched sewer network, including the possibility for stochastic simulation.

What does the WATS model simulate?

The WATS model is a deterministic model dealing with transformations taking place in gravity sewers as well as pressure mains.

	Biological transformations of wastewater COD. The organic matter (COD) undergoes changes during transport. How and to what extend these changes take place depends on a number of conditions related to the layout of the sewer and the composition of the incoming wastewater. Transformations under aerobic, anoxic and anaerobic conditions are covered.
	Biological and chemical transformations of sulfur compounds. When conditions become septic – e.g. when wastewater is pumped or conveyed in gravity sewers with low reaeration – H2S is formed. When the H2S is subjected to oxygen, it will be oxidized biologically as well as chemically. Oxidation takes place in bulk water and biofilm, but also on moist surfaces above the water surface. When oxidized on moist surfaces of the sewer, sulfuric acid is the end product, corroding concrete and metal.
	Gas transfer at the water/gas interface. In gravity sewers, the turbulence of the flowing water causes a continuous supply of oxygen from the sewer atmosphere into the wastewater. If wastewater becomes septic or not is determined by the balance between the oxygen supply and the oxygen consumptions of the water phase, biofilm and sediments. H2S becomes a problem when released to the sewer atmosphere, where it can cause corrosion of moist surfaces.
	Simulation of transport and transformation in both the water phase and the gas phase. Transformations in the water phase takes place in bulk water, biofilm and sewer sediments, while transformations in the gas phase only take place on the moist sewer surfaces. Bulk water and gas is transported with different transport velocities within the sewer. Furthermore, gas is lost by ventilation into the urban environment.
	Odor problems. H2S ventilated out of the sewer system can cause significant odor problems. However, other odorous compounds formed in the absence of oxygen do also contribute to the problem. Examples hereof are mercaptanes, indole and VFAs.
	Treatment plant concerns. Odorous compounds cause problems at treatment plants. Furthermore, H2S causes corrosion of concrete and metal but can also adversely affect activated sludge microorganisms. The biological transformations of COD can be beneficial or unfavorable for the treatment plant, depending partly on the type of treatment plant and partly on the type of transformations taking place in the sewer networks.
	but WATS does not address sediment deposition/erosion and not receiving water impacts.

WATS focuses on dry weather problems and not wet weather problems. It deals with microbial and chemical transformation processes.

How does the WATS model simulate sewer processes?

WATS simulates microbial and chemical transformation processes of organic matter, oxygen, oxidized nitrogen compounds, and sulfurous compounds (Figure 1 ).

Figure 1. Microbial and chemical processes in a gravity sewer.

Crucial for simulating in-sewer processes is the ability to predict if conditions in the water phase are aerobic (oxygen is present), anoxic (no oxygen but nitrate is present) or anaerobic (no oxygen and no nitrate is present) (Figure 2 ). To do so, the processes involved in the consumptions of oxygen and nitrate must be modeled.

Figure 2. Oxidation and reduction of wastewater components in sewers

Aerobic conditions in the water phase When oxygen is present, heterotrophic microorganisms cause the main part of the oxygen consumption. They oxidize organic matter (COD) to yield energy for growth and maintenance. At the same time, the organisms use the COD as constituents in the formation of new biomass. Most of the COD present in wastewater is not immediately suited as substrate for the biomass. The molecules are typically too large to pass the cell walls and must be broken down into smaller, more readily degradable compounds. This process is called hydrolysis, and hereby the main bulk of the wastewater COD can be made available for the biomass. There is, however, always a small fraction of COD that is inert and cannot be made available for the biomass. This fraction is included in the slow hydrolysable substrate fraction (Figure 3 ).

Figure 3. Aerobic transformation of organic matter

Hydrogen sulfide formed under anaerobic conditions becomes oxidized under aerobic conditions. The oxidation process is chemical as well as biological. Both process rates are significant under sewer conditions. The oxidation products are a mixture of oxidized sulfur forms, containing but sulfate and sulfur.

Anoxic conditions

A significant fraction of the heterotrophic organisms being active under aerobic conditions, cause nitrate reduction under anoxic conditions. Parallel hereto, the organisms use COD for growth and maintenance purposes. When the biomass uses nitrate instead of oxygen, nitrate is reduced via some intermediates to molecular nitrogen (Figure 4 ). Under sewer conditions, the intermediate nitrite tends to accumulate in the water phase. Consequently, when simulating nitrate uptake, nitrite formation and uptake must also be modeled (Figure 5 ).

Figure 4. Reduction pathway of nitrate

Figure 5. Simplified reduction pathway of nitrate

Anaerobic conditions

In the absence of both oxygen and nitrate, the previously mentioned heterotrophic organisms are no longer active. Instead other organisms take over. Some of these organisms ferment organic matter. Others reduce sulfate and oxidize COD. Another group of microorganisms produce methane from the COD. The first process – fermentation – gives rise to end products that are both desirable and undesirable. The end products have typically an unpleasant odor, however, they are excellent substrate for aerobic and anoxic treatment plant organisms. The second group – the sulfate reducing organisms – produce sulfide. Sulfides being related to a number of problems like odors, health hazards, toxicity towards microorganisms and corrosion.

Rooting of gas and water

The rooting of water is simulated as stationary flow without dispersion. The reason for not using the full Saint-Venant’s equations is that the WATS model simulates dry weather conditions. Here flow variations occur slowly and backwater is seldom an issue. Conveyance of gas are also modeled as stationary flow.

Which input data does WATS need?

All WATS model components and parameters can be measured. It has been a key concern during the development of the model also to develop methods that allow the determination of the model parameters. In order to achieve a good simulation, as detailed a knowledge on the local conditions must be obtained. This includes a thorough characterization of the wastewater and other characteristics of the system to be modeled.

For the WATS model as well as any other model, it must be kept in mind that The Model Output will Never be Better than the Model Input. If only limited data can be made available, experience and knowledge on wastewater composition depending on catchment characteristics must be applied as a substitute.

It is seldom – or more correctly: it is never – possible to determine all inputs and all process parameters for a given sewer. This is partly due to the amount of information needed being rather large, partly due to a large natural variability of wastewater quality and process parameters, and partly due to the fact that we often want to simulate future scenarios. This problem can be dealt with by taking a stochastic approach. When the statistical distribution of component and parameter values can be obtained, stochastic simulation of the in-sewer transformations can be made. The result of such simulation is the statistical distribution of e.g. component concentrations or corrosion rates. E.g. statements can be made like: with 95% probability the concrete pipe walls of this sewer will corrode less than 1 mm/year.

The WATS model is a deterministic model. It has a level of complexity similar to the hydrodynamic models used for simulation of stormwater runoff. For the pipe hydraulics, WATS needs the same data as any hydrodynamic model. In addition hereto, WATS must receive input data on the quality of the wastewater and – for corrosion and odor modeling – on the ventilation of the sewer atmosphere (Table 1 ).

Table 1. WATS input data

* The biodegradability of the COD can be measured by respiration activity measures. Oxygen uptake rate (OUR) measurements do – together with a conceptual understanding of the processes taking place causing the oxygen uptake – yield the needed information. But also VFA measurements are beneficial. In lack of better data, the ratio of COD to BOD can be used as a rough estimation of the wastewater quality.

Setting up WATS

The model can – from a numerical point of view – handle any branched system. It is, however, typically not necessary to simulate the upstream end of the system. Catchments should be organized into manageable “lumps”. The size of these lumped catchments depends on local conditions, but is typically larger than 1000 Person Equivalents per lump.

Calibration of WATS

A number of model parameters need to be calibrated to the actual location and quality of wastewater. Some of the parameters are obtained when characterizing the quality of the wastewater by means of oxygen uptake rate (OUR) measurements. Other parameters are chosen based on experience and system operational information, e.g. if there has been observed permanent sediment beds or H₂S and odor related problems. Other parameters should be calibrated by comparing to actual measurements of upstream and downstream conditions. Where this is not feasible – e.g. because the sewer line only exists in the design phase – a large database of experienced parameter values is available.

A software version of WATS

A software version of the WATS model has been developed by the SPN-Group and allows the dynamic simulation of a branched sewer network, including the possibility for stochastic simulation. The SPN-Group will assist you to select model scenarious, seting up the model and calibrating it. The SPN-Group can also be of assistance in finding and analyzing problems and solutions related to in-sewer transformatinons.