Model-Based Construction of Wastewater Treatment Plant Influent Data for Simulation Studies

Model-Based Construction of Wastewater Treatment Plant Influent Data for Simulation Studies

3.1. Verification of the Modelling Approach for Dry Weather Inflow Pattern

This modelling approach was used to model a selection of measured example diurnal cycles. For the analysis, the data from Langergraber et al. 2008 [5] (17 examples from Austria, 2 from Gerseveral) were supplemented by further daily profiles collected by members of the HSG Simulation working group ( In particular, larger plants from Gerseveral were included. In the presented results (see Figure 11 and Appendix C), 21 data sets from plants of wider range of sizes were analysed.

A second-order Fourier analysis was calculated to adjust the water flow rates. The analysed data sets only contain 2 h values for the water quantity; with this temporally not very high-resolution database, no significant improvement in mapping can be achieved with a higher-order Fourier series. The analysis calculates the parameters a0, a1, a2, b1 and b2.

The parameters Q i n f and t p l u g were optimized to adjust the C O D to the measured values. This means that the time series of the C O D concentration is completely determined by the mixture of infiltration water ( C O D ≈ 0) and wastewater ( C O D constant). A time shift to the time series of the water quantity and a further deformation of the pattern results from the transport through a storage volume.

The parameters T N , m a x , β and f U , m i n were optimised to adjust the T K N to the measured values. The time series of the T K N concentration is characterised by a proportion of C O D , representing organic bound nitrogen, in the grey water (constant concentration) and a very high concentration (approximately 400 gN/m3) in the urine. The time function of the urine flow rate is characterised by a constant proportion ( f U , m i n   Q U , M ) and a nitrogen peak in the morning ( 1 f U , m i n )   Q U , M . This nitrogen peak is determined by the two shape parameters T N , m a x (time) and β (the shape parameter of the Gumbel function and the width of the peak).

The parameter f P , C O D , G was optimised to adjust the P concentration to the measured values. Similar to the course of the T K N concentration, the course of the P concentration is characterised by a C O D proportional fraction ( f P , C O D , G )) in the grey water (constant concentration) and a very high concentration in the urine. By adjusting the parameter f P , C O D , G , the pattern can be varied, with larger values of the factor in the direction of the temporal course of the C O D concentration, with smaller values in the direction of the pattern of the T K N concentration. Unfortunately, it was only possible to adjust the P concentration for a small number of measured daily patterns due to the lack of P measurement values in several data sets.

This analysis impressively demonstrates that the dry weather inflow of wastewater treatment plants can be plausibly described with regard to C O D , T K N and P concentrations using the simple model approach presented. The results of the analysis of all 21 daily cycles considered can be found in the appendix. Based on the analysis carried out, the systematic dependencies of the adjusted shape parameters on the plant size can also be analysed. These correlations can help to estimate typical daily patterns even for locations for which no measurements are available.

The shape parameters T m i n ,   f Q , m i n , T m a x   and   f Q , m a x  

(for definition, see Figure 2) were determined from the time courses of the wastewater ( q w t = q u t + q g t ). The determined shape parameters are shown in Figure 12 over the plant size in population equivalents (PE calculated from the C O D load with the assumption of a PE-related daily C O D load of 120 g C O D /PE for raw wastewater and 78 g C O D /PE for pre-treated wastewater (data points with red label)).

In principle, it becomes obvious from Figure 12 that the time of the night-time minimum and the morning peak are shifted backwards with the size of the system. The dynamics of the daily cycle also decrease with the size of the system, with the minimum and the maximum approaching the mean value. Despite this significant correlation, a large variance in the values can be observed. This clearly speaks in favour of the necessity of using the real measurement data of the inflow for the planning and simulation studies of wastewater treatment plants. Only in an emergency should standard assumptions be used and can be calculated with the shown linear correlations. Figure 12, bottom right, shows the height of the wastewater maximum in relation to the average wastewater volume. The examples are roughly in the range specified in DWA A198, 2003, Figure 2 [21] but also outside this range. The influence of the number of connected person equivalents (PE) to a WWTP can be summarized by two effects. For larger number of PEs, a more equalized wastewater will be the produced.For larger numbers of connected PEs, the maximum flow rate will arrive later in the day at the WWTP. These arise partly from the size of the required sewer system (more distributed sources, longer travel time). In addition, cultural differences between urban and rural areas are responsible for this behaviour. Similar results have been reported already in Langergraber et al. 2008 [5].
The C O D concentration time series are adjusted using the parameters Q i n f and t p l u g . There is no plausible dependence on the plant size for the infiltration water fraction. As can be seen in Figure 13 (left), the values for the proportion of extraneous water ( Q i n f / Q m in relation to the total water volume) of the sample daily cycles fluctuate in the range of 10–70%. The temporal shift of the concentration curve in relation to the water retention volume ( t p l u g ) varies greatly. Larger values (0.1–0.2 d) are to be expected for larger plants. Daily variations in the pre-treatment process (data points labelled in red) typically show larger delays.
Figure 14 shows the calculated C O D concentrations of the examples. The expected values of approximately 1000 gCOD/m3 for raw wastewater and approximately 600 gCOD/m3 for pre-treated water (data points labelled in red) are confirmed.
The T K N concentration time series is adjusted to the measured values using the parameters T N , m a x , β and f U , m i n . Figure 15 shows the estimated parameters over the system size. Similar to the time of the maximum inflow volume, the time of the urine peak is clearly dependent on the system size. Compared to the maximum water volume, the urine peak is between 0–0.2 d ahead (Figure 15, bottom right). The proportion of the constant urine volume varies in the range of 0.3–0.65 with outliers. There is no visible dependence on the size of the plant. Pretreated wastewater tends to produce higher values for the constant fraction of urine flow rate ( f U , m i n ), which can be explained by further equalization in the pre-treatment stage. For the width of the urine peak ( β ), a dependence on the size of the plant is also visible. For larger plants, the urine flow is obviously more even, and the morning peak is less pronounced.
To adjust the phosphorus concentration pattern to the measured values, the parameter f P , C O D , G was optimized. As only some of the available daily data contain phosphorus concentration values, a value of f P , C O D , G = 0.008 gP/gCOD was assumed for the remaining examples. This value corresponds approximately to the mean value of the analyses for which P data was available. A correlation with the plant size is not expected and also not visible in Figure 16.

3.2. Experiences with the Method to Generate Long-Term Dynamic Simulation Influent Data

The method described here modifies the HSG method from Langergraber et al. 2008 [5] in order to improve the shortcomings described in the motivation section. This extends the scope of application of this method. For several typical municipal plants, both methods will yield comparable results. The inflow to municipal wastewater treatment plants can be described very well in most cases using both methods. The original method was used in a large number of simulation studies. The good-to-very-good agreement achieved between the simulated effluent values of the plants and the measured values is an indirect confirmation of the quality of the method.
In a simulation study carried out by the authors, the method was evaluated even further. Here, at the wastewater treatment plant, the continuous quality measurements in the influent during continuous operation (NH4-N, online PO4-P analyzer, daily 24 h composite samples with C O D , T K N and PO4-P) were available. In this case, the influent data synthesized with the method can be compared with measured values. Figure 17 compares the ammonium loads calculated from the measured data with the data synthesized using the method.
A very good reproduction of the inflow situation including the rain events can be created here. Differences result primarily from variations in the inflow loads that cannot be recorded with the presented method. A quantification of the quality of a generated dynamic influent pattern as presented in Figure 17 or a quantification of the quality of achieved fit of the validated model to the available effluent measurements remains a difficult problem. There was an intensive discussion in the HSG group about the valid methods and the error definitions for this task (Ahnert et al. 2009 [23]). Simple residuals are not sufficiently significant. A visual evaluation was used here as a pragmatic solution.

The method described is not applicable in the following cases. A necessary precondition is that the dry weather inflow is mainly created from normal human activities in urbanizations (municipal wastewater). If the wastewater is completely produced by industry or a large fraction of the wastewater arises from industrial sources, the assumptions regarding cyclic production patterns of greywater and urine will not hold. The method cannot be applied. A failure of the method was also observed in one case in which municipal wastewater was pumped to the wastewater treatment plant in a widely branched pressurized pipe system with very long residence times (>4 h) (wastewater treatment plant in Berlin). In general, the method might fail in situations where large fractions of the wastewater will be managed and stored in the sewer system. This will lead to a non-predictable change in the resulting inflow patterns.

One major limitation of the method results from the assumed measurement situation (only routine data). In this situation, the influent concentration (24 h composite sample) are not measured every day but only on some days (e.g., 50 times a year). As a consequence, a constant load for every day must be assumed. Stochastic variations and typical weekly patterns cannot be reproduced. A limited quality of the influent data arises. Better influent data quality can be achieved only with an additional (non-routine) measurement effort.

But, if applicable, the method drastically reduces the requirements for influent data down to a continuous influent flow measurement and a few 24 h composite samples.

Disasters Expo USA, is proud to be supported by Inergency for their next upcoming edition on March 6th & 7th 2024!

The leading event mitigating the world’s most costly disasters is returning to the Miami Beach

Convention Center and we want you to join us at the industry’s central platform for emergency management professionals.
Disasters Expo USA is proud to provide a central platform for the industry to connect and
engage with the industry’s leading professionals to better prepare, protect, prevent, respond
and recover from the disasters of today.
Hosting a dedicated platform for the convergence of disaster risk reduction, the keynote line up for Disasters Expo USA 2024 will provide an insight into successful case studies and
programs to accurately prepare for disasters. Featuring sessions from the likes of The Federal Emergency Management Agency,
NASA, The National Aeronautics and Space Administration, NOAA, The National Oceanic and Atmospheric Administration, TSA and several more this event is certainly providing you with the knowledge
required to prepare, respond and recover to disasters.
With over 50 hours worth of unmissable content, exciting new features such as their Disaster
Resilience Roundtable, Emergency Response Live, an Immersive Hurricane Simulation and
much more over just two days, you are guaranteed to gain an all-encompassing insight into
the industry to tackle the challenges of disasters.
By uniting global disaster risk management experts, well experienced emergency
responders and the leading innovators from the world, the event is the hub of the solutions
that provide attendees with tools that they can use to protect the communities and mitigate
the damage from disasters.
Tickets for the event are $119, but we have been given the promo code: HUGI100 that will
enable you to attend the event for FREE!

So don’t miss out and register today:

And in case you missed it, here is our ultimate road trip playlist is the perfect mix of podcasts, and hidden gems that will keep you energized for the entire journey


This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More