Due to the ongoing climate and energy crisis, wastewater is becoming increasingly important as a source of renewable energy. In urban areas, heat recovery from the sewer is considered a promising approach, as the locations of supply and demand are close to each other. In this context, it is crucial that negative impacts on temperature-sensitive processes in the downstream wastewater treatment plant are strictly avoided. To support the necessary planning and authorization, this paper presents a model-based approach to assess the thermal energy level of the wastewater at any given location in the sewer, taking into account the influent temperature constraints of the wastewater treatment plant. The entire modelling is based on the open-source software SWMM 5, which was extended by a temperature model. The concept showed its practicability and informative value during a full-scale field application in the sewer systems of the Austrian city of Graz. All modelling is based on freely available software, which makes the approach easy transferable to other cities with comparable infrastructural boundary conditions.

  • Model-based assessment of wastewater thermal energy.

  • Use of open-source software SWMM 5 with temperature model.

  • Practical application and validation in Graz, Austria.

  • Extensive measurement campaign for temperature data.

  • Development of heat recovery maps for sewer networks.

In an era where the transition from fossil-based energy sources to renewable, eco-friendly alternatives is paramount, European nations are vigorously exploring new resources to support this shift and reduce their reliance on energy imports. This search is particularly critical in the context of heating, a sector the European Commission (2016) identifies as the largest in the European Union's energy landscape, a status it is poised to maintain. Recognizing the significant thermal energy present in wastewater, a fact well documented by researchers like Hao et al. (2019) and Neugebauer et al. (2022), the European Parliament and the Council (2018) have recently acknowledged wastewater as a viable renewable energy source.

The concept of wastewater heat recovery is not new in international research, with studies by Culha et al. (2015) addressing its technical aspects comprehensively. Hepbasli et al. (2014) extend this discussion to potential sites for application within wastewater systems. Two primary methodologies emerge in this realm: in-sewer heat extraction (encompassing both private and public sewers) and effluent-based extraction at wastewater treatment plants (WWTPs). From a combined wastewater and energy management standpoint, the latter, effluent-based approach is preferable, as it avoids the operational complexities of sewer systems and WWTPs with the simultaneous availability of large quantities of even treated wastewater. However, the in-sewer method often presents an advantage in terms of energy efficiency, primarily due to shorter distances for heat transport.

Despite the general consensus in literature, a distinction between these two approaches is often not explicitly made. Yet, it is reasonable to surmise that most existing wastewater heat recovery installations today are in-sewer based. Recently, however, more and more literature on heat recovery in the effluent of WWTPs has come to the fore. This can be explained by the fact that greater heat potentials can be tapped here due to the aforementioned points. With regard to implementation challenges the primary concern with in-sewer heat extraction lies in its potential impact on temperature-sensitive nitrogen removal processes in conventional WWTPs, an aspect highlighted by Wanner et al. (2005). Presently, three methodologies are employed to assess this impact: the allegation alternate method, a straightforward but less precise technique described by Kretschmer et al. (2016); a more detailed approach combining wastewater and energy considerations, introduced by Huber et al. (2020); and sophisticated wastewater temperature models, offering high accuracy at the cost of complexity. For situations requiring precise predictions of in-sewer heat recovery's impact on WWTP influent temperature (such as when the temperature nears WWTP design thresholds), the third approach is preferred.

Continuous research in this field has led to the development of initial models for predicting in-sewer wastewater temperature changes, with Dürrenmatt and Wanner's TEMPEST model (2014) and Abdel-Aal et al.'s model (2014) being notable examples. These foundational models have paved the way for ongoing advancements in both development and application. Efforts to streamline input data collection have been made, with Elías-Maxil et al. (2017) proposing a more streamlined model that omits in-sewer air heat transfer. Comparisons between mechanistic and conceptual models, such as those conducted by Saagi et al. (2021), are also notable for incorporating hydraulic flow calculations. Addressing data challenges, Golzar et al. (2020) employed artificial neural network techniques to predict wastewater temperatures. Concurrently, specific input parameters, such as in-sewer air heat transfer (explored by Abdel-Aal et al. (2021)) and heat transfer around buried pipes (investigated by Shafagh et al. 2022) are undergoing further examination. Household-level wastewater production modelling, a key component in planning in-sewer heat recovery, is being advanced by researchers like Sitzenfrei et al. (2017) and Wärff et al. (2020). Figueroa et al. (2021) have contributed a comprehensive thermal-hydrodynamic model using the Stormwater Management Model (SWMM), enhancing the understanding of thermal processes within sewers. This area of research is further enriched by Hadengue et al. (2021), who integrated thermo-hydraulic modelling across various scales, Pelda & Holler (2019), who developed methodologies for spatially mapping in-sewer heat recovery potentials and Abdel-Aal et al. (2018) who applied a heat transfer model in combination with a hydrodynamic sewer model to investigate different heat recovery scenarios across a sewer system considering wastewater minimum temperature requirements in the network and for the inflow of the WWTP.

The paper at hand adds to this body of knowledge by presenting a large-scale application of a thermo-hydraulic model, both spatially and temporally, within the extensive public sewer network of Graz, Austria. This network spans approximately 858 km, encompassing around 26,000 pipe sections and as many manholes. For calibration and validation purposes, data were collected from 2 soil temperature sites, 15 air temperature sites within sewers, and 17 wastewater temperature sites over periods ranging from 14 to 30 months. The research aimed not only to predict the impacts of specific in-sewer heat extraction sites on downstream WWTP inflow temperatures but also to create heat extraction maps for the entire sewer network. These maps, determining maximal heat extraction levels at any point in the system, are based on WWTP-specific wastewater temperature limits. This information is crucial for identifying optimal in-sewer heat recovery sites for targeted implementation. The methodology employed integrates thermal and hydrodynamic modelling using widely available tools, making it adaptable to other cities with similar infrastructure. The paper also shares valuable insights gleaned from this extensive measurement campaign.

Case study

The City of Graz, Austria, was selected as the case study site for this research. The sewer operator expressed interest in developing a map to determine whether thermal energy extraction from any point in the public sewer network, up to a defined maximum amount, could be approved. This initiative was particularly relevant given the bottleneck presented by the city's WWTP. Currently, the WWTP, located in the south of Graz, is operating at over 10% above its planned maximum treatment capacity due to population growth. Consequently, reducing the inflow temperature of the wastewater to the treatment plant by more than 0.1° K was deemed undesirable. However, this stringent approach still permits thermal energy extraction within the sewer network without significantly lowering the inflow temperature, as measured by the current temperature sensors at the WWTP.

In Graz's public sewer network, a comprehensive analysis was conducted. The total network spans 858 km, comprising 577 km of combined sewers and 226 km of sanitary sewers, accounting for 803 km under study. The sewer pipes exhibit a wide range of diameters, from 250 to 2,000 mm. The average distance between manholes in this network is approximately 39 m, with a mean sewer depth of 3.2 m. Graz's urban setting, with its population of 295,000 in 2022 across 127 km², provides a unique backdrop for this study. The wastewater treatment plant in Graz handles a peak dry weather inflow of 1.2 m³/s, or about 60,000 m³ daily. Many of the sewers, particularly those with smaller diameters, experience low flow rates, especially during night time. According to the Austrian guideline document (ÖWAV 2021) and in agreement with the responsible sewer operator, specific parameters were set to a minimum dry weather flow of 10 L/s and a minimum diameter of 400 mm, so that only sewers with a significant heat recovery potential were analyzed. Based on these criteria, the study focused on 900 sewers out of the total 26,000, representing a total length of 48 km, which met the minimum flow and diameter requirements for further analysis, as shown in Figure 1.
Figure 1

Selected public sewers with a diameter greater or equal 400 mm and a minimum dry weather flow of more than 10 L/s in the case study Graz (Background image: © Stadt Graz).

Figure 1

Selected public sewers with a diameter greater or equal 400 mm and a minimum dry weather flow of more than 10 L/s in the case study Graz (Background image: © Stadt Graz).

Close modal

Measurement approach

A long-term measurement campaign was carried out as the basis for the simulation study (one and a half year). The measurement approach follows two concepts: first, the efforts of the project aimed at the selected public sewers with a diameter greater or equal to 400 mm and a minimum flow of 10 L/s; second, the needs of the model, as the calibration should occur in areas where the simulations are expected to provide answers that are as close to reality as possible. Four measuring sections were selected (Figure 2). In the north of Graz, one section was situated in an industrial area with high-temperature gradients and included four measurement points (Industrial section). Another section was located in a residential area with low temperature gradients (Residential section). In both sections, the geological situation was comparable. One section was selected on the main sewer near the inner-city with three measurement points (Inner-city section, and the last section was on the main sewer in front of the wastewater treatment plant with four measurement points (WWTP section)).
Figure 2

Sections and nodes to measure wastewater temperatures, air temperature inside the sewers and soil temperature at different depth levels. A total of four measuring sections and additionally three individual measuring points for wastewater and sewer air temperature were installed. Soil temperatures were measured at locations at depths of 0.80, 1.15, and 1.50 m (Background image: © Stadt Graz).

Figure 2

Sections and nodes to measure wastewater temperatures, air temperature inside the sewers and soil temperature at different depth levels. A total of four measuring sections and additionally three individual measuring points for wastewater and sewer air temperature were installed. Soil temperatures were measured at locations at depths of 0.80, 1.15, and 1.50 m (Background image: © Stadt Graz).

Close modal

Air temperature and sewage temperature were measured at all selected sections. Additionally, three measuring points were selected: one in the east, one in the centre, and one in the west of Graz. Soil temperatures were measured in the west and in the centre of Graz at depths of 0.80, 1.15, and 1.50 m. Based on the measured values, an annual hydrograph for the mean sewer depth in Graz was non-linearly extrapolated. Sewage temperatures were measured at the eastern and western measuring points. To measure air and sewage temperatures, Hobo sensors and data loggers were used. The sensors were affixed to ropes inside flexible pipes to position them within the medium.

Modelling approach

Various temperature models available in the literature were analyzed. Among these, the TEMPEST model by Dürrenmatt & Wanner (2014), which was further developed by Figueroa et al. (2021) and Hadengue et al. (2021), as well as the model by Abdel-Aal et al. (2014), have been proven applicable for dynamic in-sewer temperature calculations. Although the TEMPEST model offers a more comprehensive physical description of in-sewer temperature processes, both models account for the most influential heat transfer processes between the sewage, sewer pipe wall, soil, and sewer air. Ultimately, the model by Abdel-Aal et al. (2014) was selected because, unlike the TEMPEST model, all its parameters are accessible or measurable in practical applications.

For the parallel coupling with a hydrodynamic sewer model using an explicit numerical scheme, Abdel-Aal et al.'s model was reformulated to meet the requirements. The implementation was done in the open-source code of the EPA SWMM 5.1 model. After a comprehensive sensitivity analysis, which identified the most influential processes on the resulting wastewater temperature in the inlet of the wastewater treatment plant as a result of thermal energy extraction in the upstream sewer network, the final simplified model with Equation (1) for thermal processes along conduits and Equation (4) for thermal mixing in junctions, was implemented. The thermal resistance sewage – sewer air and the thermal resistance sewage – soil is calculated according to Equations (2) and (3).
(1)
TS is the soil temperature, Ta is the sewer air temperature, ρ is the density, cp is the specific heat capacity, Q is the flow, Rws is the thermal resistance sewage – soil, Rwa is the thermal resistance sewage – sewer air, L is the conduit length, P is the thermal capacity in/out, T0 is the upstream temperature, T1 is the downstream temperature
(2)
hwa is the heat transfer coefficient wastewater – air, width refers to the sewer pipe width at water level, length refers to the length of the sewer pipe
(3)
tp is the wall thickness of conduit, ds is the soil depth to which heat is transferred, kp is the thermal conductivity pipe wall, ks is the thermal conductivity soil, wet.p refers to the Wetted perimeter of sewer pipe, length refers to the length of the sewer pipe
(4)
T0 is the junction temperature, Ti is the inflow temperature of conduit i, Qi is the inflow rate of conduit i.

This model maps the change in wastewater temperature due to the temperature exchange between wastewater and air, between wastewater and soil (first term in Equation (1)), and between different flows of wastewater (Equation (4)). It also accounts for temperature changes due to wastewater heat extraction for thermal energy use (second term of Equation (1)).

Sensitivity analysis

In this study, the Morris screening method was chosen for sensitivity analysis. The calculations were performed in Python using the Python package SALib (Herman et al. 2017). To integrate SWMM 5.1 with Python, we utilized the Python package swmm-api developed by Pichler (2022), enabling the coupling of various algorithms with SWMM 5.1. The Morris method, first proposed by Morris (1991) and later modified by Campolongo et al. (2007), is detailed in Saltelli et al. (2004). This method is designed for low computational cost and yields the average and standard deviation of local sensitivities across different parameter spaces. It categorizes parameter effects as either negligible, linear and additive, or non-linear, involving interactions with other factors. The method computes two sensitivity measures, and . A high value indicates a significant effect of the parameter on the model output, averaged over other parameters, signifying sensitivity. A high value suggests the parameter 's effect is greatly influenced by the input space point, indicative of non-linearity or interactions with other parameters. The results from the Morris screening are qualitative and do not quantify the parameter's contribution to output variance. We adopted the scaled version of the sensitivity measures as described by Neumann (2012). The used value ranges for the sensitivity analysis are summarized in Table 1.

Table 1

Value ranges of the model parameters used for the Morris method

ParameterDescriptionUnitDeltaLower limitUpper limit
densitySoil Soil density [kg/m³] – 1,500 2,100 
specHcSoil Specific heat capacity of soil [J/kgK] – 600 1,700 
k_Soil Thermal conductivity soil [W/mK] – 0.3 2.5 
k_Pipe Thermal conductivity of pipe [W/mK] – 0.8 2.3 
Thickness Pipe wall thickness [m] – 0.03 0.35 
DWFPattern Waste water temperature annual median [°C] – 15 19 
AirPattern Sewer air temperature monthly media [°C] ±3 – – 
SoilPattern Soil temperature monthly median [°C] ±1 – – 
Inflow Inflow [m³/s] – 0.3 1.3 
ParameterDescriptionUnitDeltaLower limitUpper limit
densitySoil Soil density [kg/m³] – 1,500 2,100 
specHcSoil Specific heat capacity of soil [J/kgK] – 600 1,700 
k_Soil Thermal conductivity soil [W/mK] – 0.3 2.5 
k_Pipe Thermal conductivity of pipe [W/mK] – 0.8 2.3 
Thickness Pipe wall thickness [m] – 0.03 0.35 
DWFPattern Waste water temperature annual median [°C] – 15 19 
AirPattern Sewer air temperature monthly media [°C] ±3 – – 
SoilPattern Soil temperature monthly median [°C] ±1 – – 
Inflow Inflow [m³/s] – 0.3 1.3 

Model calibration

After the sensitivity analysis, the temperature model underwent manual calibration, building upon a hydrodynamic sewer network model of Graz that had been previously calibrated and implemented in SWMM5.1. The calibration encountered challenges due to data loss during measurement campaigns, primarily caused by the depletion of batteries in some sensors – owing to the high temporal resolution of 5 min – and damage to others, necessitating replacement during monthly data readouts. Data were collected at all 16 measurement sites for sewer air and sewage temperature. However, not all stations provided continuous data throughout the year. Nonetheless, using the algorithm described by Pichler (2018), it was possible to extract daily patterns for sewer air and wastewater temperature, as well as monthly patterns for sewer air, wastewater, and soil temperature for all stations. Based on the different daily and monthly pattern in residential, industrial, and inner-city areas, these patterns could be allocated to the corresponding urban zones.

The decision was made to simulate one day for each month from May 2019 to May 2020 to refine the scaling of daily and monthly patterns. This approach facilitated the calibration of both diurnal and annual wastewater temperature cycles. Adjustments were applied to the start and end points of each measurement site (residential, industrial, and inner-city), except for the Graz-WWTP measuring section. The latter's influence on the network's last measurement point served as a reference for subsequent hazard analysis. The Nash–Sutcliffe Coefficient and the coefficient of determination R² were employed as indicators of calibration quality across all measurement tracks (Hauduc et al. 2015).

Derivation of in-sewer heat extraction maps

To derive wastewater heat extraction maps for the entire sewer network all in-sewer wastewater temperatures and flows were simulated by the fully calibrated thermo-hydrodynamic model. A second-order polynomial was used to approximate all possible combinations j of the wastewater temperature at any node and the corresponding inflow temperature at the WWTP , see Equation (5).
(5)
Tjx is the wastewater temperature at any node, TjWWTP is the inflow temperature at the WWTP.
For each node in the sewer network simulations were made with , and to get three value pairs To derive the values of Equation (1) has to be solved with
(6)
and
(7)
To get the useable span of temperature at any location of the system when a maximum permissible difference of temperature of the inflow to the treatment plant is given (to take into account of wastewater treatment specific temperature limits) Equation (8) was used.
(8)
Tperm TP refers to the maximum permissible difference of temperature of the inflow to the treatment plant.
To calculate the useable thermal energy on a given permissible difference of temperature of the inflow to the treatment plant at any node of the sewer system Equation (9) can be applied.
(9)
Px is the useable thermal energy, is the density water, Qx is the flow rate at node x, cp is the specific heat capacity fluid.
The complete sequence of the developed approach is shown in Figure 3.
Figure 3

Approach for calculating the required inverse functions for each channel section, with temperature change, e.g., by heat extraction, corresponding inflow temperature at the WWTP, SWMM5-T = adapted SWMM5.1 Engine with implemented temperature model.

Figure 3

Approach for calculating the required inverse functions for each channel section, with temperature change, e.g., by heat extraction, corresponding inflow temperature at the WWTP, SWMM5-T = adapted SWMM5.1 Engine with implemented temperature model.

Close modal

Temperature measurements

The primary objective of the measurement campaign was to provide suitable data to calibrate the temperature model and to derive the required daily and annual patterns for wastewater, sewer air and soil temperature. The main input variables required here are the temperature curves of the soil temperature at different depths, the sewer air temperature in different sewer sections and the wastewater temperature for the same sewer sections. To obtain as comprehensive as possible data sets for the calibration, the intention was to generate at least one complete annual cycle at all measuring points. This objective was achieved as far as possible.

Figure 4 demonstrates the almost complete annual course of the soil temperatures at the two continuously operated measuring stations together with the corresponding air temperatures. The different temperatures at different soil depths (here 0.80, 1.15, and 1.50 m) are recognizable.
Figure 4

Illustration of the ground temperatures measured at the two-ground temperature measuring points in Graz (see Figure 2) and the corresponding air temperatures.

Figure 4

Illustration of the ground temperatures measured at the two-ground temperature measuring points in Graz (see Figure 2) and the corresponding air temperatures.

Close modal
Figure 5 shows the seasonal pattern of the in-sewer air and wastewater temperature for a measuring point in the north of the city (residential section) and for a measuring point in the south of the city just before the central wastewater treatment plant (see Figure 2). The precipitation measured in the immediate neighbourhood of the measuring points is also shown for both measuring points.
Figure 5

Illustration of in-sewer air and wastewater temperature for the measure section residential (in the north) and in front of the WWTP together with precipitation measured in the immediate neighbourhood of the measure sections.

Figure 5

Illustration of in-sewer air and wastewater temperature for the measure section residential (in the north) and in front of the WWTP together with precipitation measured in the immediate neighbourhood of the measure sections.

Close modal

The comparison of the measuring station in the north of Graz with the measuring station situated about 8 km to the south and in the flow direction downstream just before the wastewater treatment plant shows the following phenomena:

  • (1) The amplitude of the daily pattern is greater in the north than in the south.

  • (2) The influence of precipitation is visible to the same extent at both measuring stations.

In addition, temperatures are approximately 2 K warmer throughout the entire year at the measuring section just before the wastewater treatment plant compared to the north of the city. This phenomenon is attributed to the discharge from an industrial area, which includes a large laundry, among other facilities.

Sensitivity and model calibration

The Morris method sensitivity analysis, which concentrated on the average wastewater temperature at the WWTP's inlet, yielded significant insights. Figure 6 presents simulation results for a typical winter day, identified as the most critical condition for WWTP processes. Similar patterns were observed in other time periods, with the ranking of model parameters by their influence on wastewater temperature at the WWTP influent remaining consistent, despite minor quantitative variations. Predominantly, the dry weather flow temperature (DWFPattern) was the most influential parameter, succeeded by the in-sewer air temperature (AirPattern), as demonstrated by their respective μ* values. Most other parameters, with μ* values ranging between 0.1 and 0.25, had a lesser impact on the uncertainty of the average inflow temperature at the WWTP. Soil density, with a negligible value, showed minimal effect on the average temperature. The elevated σ value for AirPattern indicates a tendency towards significant non-linearity or interactions with other parameters, a trait also observed in DWFPattern to a lesser extent. The histogram in Figure 7 illustrates the results from uncertainty propagation, displaying a broad distribution of the output variable between 11 and 17 °C.
Figure 6

Morris results on the left showing μ* and σ on the x- and y-axis respectively for a winter day. Histogram on the right showing result from uncertainty propagation.

Figure 6

Morris results on the left showing μ* and σ on the x- and y-axis respectively for a winter day. Histogram on the right showing result from uncertainty propagation.

Close modal
Figure 7

Calibrated model output versus measured wastewater temperature in front of the WWTP for May 2019–May 2020.

Figure 7

Calibrated model output versus measured wastewater temperature in front of the WWTP for May 2019–May 2020.

Close modal

Following the outcomes of the sensitivity analysis, the temperature model was manually calibrated. This process retained the distinct daily and monthly patterns for various city areas (residential, industrial, and inner-city) as derived from the measured data, without altering their characteristics. A unique monthly scaling factor was applied to each pattern in the model to fine-tune these patterns. The discharge rates remained unaltered, as they are part of an already calibrated hydraulic model. This calibration method resulted in convincing Nash–Sutcliffe efficiency coefficients, nearing 0.9 across all measuring stations (residential, industrial, inner-city, and in front of the WWTP). Given the very good Nash–Sutcliffe efficiency coefficients already achieved at all measuring sections and the low sensitivity of the other temperature model parameters, no further adjustments were made. Figure 7 shows the calibrated model output at the entrance of the WWTP, highlighting temperatures crucial for subsequent analyses.

Heat extraction maps

We varied the permitted wastewater temperature decrease (ΔTperm TP) from 0.01 to 0.5°K to generate maps illustrating the in-sewer heat recovery potentials, as depicted in Figure 8. Each map highlights the potential for thermal energy extraction from each sewer section, ensuring the defined maximum reduction in wastewater temperature is not exceeded. These results enable suitable locations for thermal energy recovery within the system to be determined, taking into account a permissible maximum temperature change in the inlet to the WWTP.
Figure 8

In-sewer heat recovery potentials considering wastewater treatment performance as a restriction of the difference in wastewater temperature with heat recovery and no heat recovery, ΔT(perm TP) = 0.01, 0.1 and 0.5 K (top to bottom). (Background image: © Stadt Graz).

Figure 8

In-sewer heat recovery potentials considering wastewater treatment performance as a restriction of the difference in wastewater temperature with heat recovery and no heat recovery, ΔT(perm TP) = 0.01, 0.1 and 0.5 K (top to bottom). (Background image: © Stadt Graz).

Close modal

In this study, we present a methodology focused on wastewater heat recovery within sewer networks. Recognizing the possible negative impact of reduced inflow temperatures on wastewater treatment processes (Wanner et al. 2005; Abdel-Aal et al. 2018), our method analyses wastewater heat recovery potentials at various locations in the sewer network considering wastewater treatment specific temperature limits. This approach accounts for wastewater heat extraction in the network and potential reheating processes due to contact with surrounding media en route to the WWTP. Simplified modelling approaches presented by Kretschmer et al. (2016) and Huber et al. (2020) were not utilized as they omit significant physical processes in complex networks and long travel times to WWTPs. We instead evaluated and adapted detailed, dynamic models from literature (Abdel-Aal et al. 2014; Dürrenmatt & Wanner 2014; Figueroa et al. 2021), implementing Abdel-Aal et al.'s (2014) formulation in a hydrodynamic sewer network model using SWMM 5.1.

An extensive measurement campaign, recording wastewater and in-sewer air temperatures at multiple locations over an entire year, supported the method's development. These data provided crucial insights into temperature dynamics within the sewer network.

Our sensitivity analysis and model calibration confirmed the selected approach's accuracy in replicating observed processes, achieving our study's objectives. It showed that wastewater temperatures in the sewer network, both before and after heat extraction, significantly affect WWTP influent temperature. Furthermore, dominant temperature exchange processes between wastewater and sewer air, and between wastewater and the sewer pipe/soil, must be considered, especially for longer flow paths to the WWTP, to capture essential reheating dynamics vital for maximizing in-sewer wastewater heat recovery potentials.

The adapted model, implemented in SWMM5.1, was effectively used to generate heat extraction maps, with each map detailing maximum permissible wastewater temperature reductions at WWTP inlets. Detailed analysis revealed that two factors impact potential heat recovery and optimum extraction location: existing temperature in conjunction with discharge, and remaining flow time to the WWTP. Key findings include:

  • 1. Higher permitted wastewater temperature reductions at WWTP inlets increase heat recovery (extraction) potential.

  • 2. Lower permitted wastewater temperature reductions enhance heat recovery (extraction) potential of peripheral, remote sewer lines.

  • 3. Even with higher permitted wastewater temperature reductions, maximum heat recovery (extraction) potential depends on both discharge volume and remaining flow time to the WWTP.

These general observations can be validated for specific extraction points within the Graz sewer network. A comparison of the sewer lines with the highest extraction potential across different permitted temperature reductions, as illustrated in Figure 8, reveals that the lines with the greatest potential for lower permitted temperature reductions upstream of the wastewater treatment plant are situated further north. It is also evident that, depending on the permitted temperature reduction, some peripheral sewer lines exhibit a lower potential. It is important to consider that, in addition to flow distance and potential recovery, the volume available for extraction plays a crucial role (see Equation (9)). For instance, the sewer section in the northwest of the city, where the industrial area measuring points were located, does not possess the highest extraction potential despite receiving higher temperature discharges. This discrepancy is attributed to the relatively low discharge volume in this section of the sewer.

Our methodology facilitates the generation of heat extraction maps, indicating potential extraction sites across the sewer network for specified maximum wastewater temperature reductions at WWTP inlets. This process, however, requires extensive simulations. If a heat recovery measure is implemented, simulations must be repeated to account for altered conditions, and maps must be updated. It is crucial for sewer network operators to be aware of the heat recovery capacity installed or planned.

Transferring this method to other sewer networks is technically straightforward, as the model was implemented within the freely available simulation platform SWMM5. Model parameters are designed to be generally applicable. However, our sensitivity analysis underscored the importance of temperatures within the sewer network, including their annual variation. Therefore, successful application of this method necessitates recording actual state temperatures through a measurement campaign.

In conclusion, this study successfully introduces a novel design methodology for thermal energy recovery within sewer networks, effectively addressing the challenges in wastewater management and renewable energy utilization. Through an intricate model-based approach, incorporating extensive field measurements and dynamic modelling, this research proposes a robust design tool to aid in the development of city-wide strategies for efficient in-sewer wastewater heat extraction, ensuring minimal impact on wastewater treatment processes. The key outcomes include:

  • The successful implementation and calibration of an open-source thermo-hydrodynamic sewer model.

  • A developed methodology enabling the calculation of maximum insewer wastewater heat recovery potentials and identification of optimal extraction locations.

These outcomes underscore the importance of considering sewer network temperatures and flow dynamics to optimize heat recovery. The model's adaptability and scalability, combined with its applicability in various urban infrastructures, make it a valuable asset for harnessing renewable energy from wastewater systems.

This paper summarizes core findings of three master theses elaborated under supervision of the first author. These works are quoted as Pichler (2018), Schlagbauer (2018), and Felder (2021) in the References section. Preliminary results have been presented at the IWA Watermatex 2023 conference. This conference contribution is quoted as Muschalla et al. (2023) in the References section. Parts of the research were funded by Holding Graz Wasserwirtschaft.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.

Abdel-Aal
M.
,
Smits
R.
,
Mohamed
M.
,
De Gussem
K.
,
Schellart
A.
&
Tait
S.
2014
Modelling the viability of heat recovery from combined sewers
.
Water Sci. Technol.
70
(
2
),
297
306
.
Campolongo
F.
,
Cariboni
J.
&
Saltelli
A.
2007
An effective screening design for sensitivity analysis of large models
.
Environ. Modell. Software
22
,
1509
1518
.
Culha
O.
,
Gunerhan
H.
,
Biyik
E.
,
Ekren
O.
&
Hepbasli
A.
2015
Heat exchanger applications in wastewater source heat pumps for buildings: A key review
.
Energy Build.
104
,
215
232
.
Directive (EU)
2018
2018/2001 of the European Parliament and of the Council of 11 December 2018 on the promotion of the use of energy from renewable sources (recast). Brussels, Belgium
.
Elías-Maxil
J. A.
,
Hofman
J.
,
Wols
B.
,
Clemens
F.
,
van der Hoek
J. P.
&
Rietveld
L.
2017
Development and performance of a parsimonious model to estimate temperature in sewer networks
.
Urban Water J.
14
(
8
),
829
838
.
European Commission
2016
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions – An EU Strategy on Heating and Cooling. Brussels, Belgium
.
Felder
T.
2021
Sensitivitätsanalyse eines Temperaturentnahmemodells im Kanalsystem der Stadt Graz (Sensitivity analysis of a temperature extraction model in the city's sewer system)
.
Master Thesis
,
Graz University of Technology
,
Graz, Austria
.
Figueroa
A.
,
Hadengue
B.
,
Leitao
J.
,
Rieckermann
J.
&
Blumensaat
F.
2021
A distributed heat transfer model for thermal-hydraulic analyses in sewer networks
.
Water Res.
204
,
117649
.
Hao
X.
,
Li
J.
,
van Loosdrecht
M. C. M.
,
Jiang
H.
&
Liu
R.
2019
Energy recovery from wastewater: Heat over organics
.
Water Res.
161
,
74
77
.
Hauduc
H.
,
Neumann
M. B.
,
Muschalla
D.
,
Gamerith
V.
,
Gillot
S.
&
Vanrolleghem
P. A.
2015
Efficiency criteria for environmental model quality assessment: A review and its application to wastewater treatment
.
Environ. Modell. Software
68
,
196
204
.
Herman
J.
&
Usher
W.
2017
SALib: An open-source Python library for sensitivity analysis
.
Journal of Open Source Software
2
(
9
),
doi:10.21105/joss.00097
.
Hepbasli
A.
,
Biyik
E.
,
Ekren
O.
,
Gunerhan
H.
&
Araz
M.
2014
A key review of wastewater source heat pump (WWSHP) systems
.
Energy Convers. Manage.
88
,
700
722
.
Muschalla
D.
,
Sprung
W.
,
Reinstaller
S.
&
Kretschmer
F.
2023
A model-based assessment of in-sewer heat recovery potentials
. In:
11th IWA Symposium on Modelling and Integrated Assessment
,
23–27 Sep 2023
,
Quebec City, QC, Canada
.
Neugebauer
G.
,
Lichtenwoehrer
P.
,
Huber
F.
,
Stoeglehner
G.
&
Kretschmer
F.
2022
Potentials and integrated suitability pre-assessment of wastewater treatment plants as local energy cells
.
Front. Environ. Sci.
9
,
785557
.
ÖWAV
2021
ÖWAV-Arbeitsbehelf 65: Energetische Nutzung des thermischen Potenzials von Abwasser. Austrian Standards plus GmbH, Wien, Austria
.
Pichler
M.
2018
Automated quantification of the pollutant load dynamics of the central storage tunnel in the inlet to the wastewater treatment plant Graz
.
Master Thesis
,
Graz University of Technology
,
Graz, Austria
.
Pichler
M.
2022
swmm-api: API for reading, manipulating, and running SWMM-Projects with python (0.3). Zenodo. https://doi.org/10.5281/zenodo.7054804
.
Saagi
R.
,
Arnell
M.
,
Reyes
D.
,
Wärff
C.
,
Ahlström
M.
&
Jeppsson
U.
2021
Modelling temperature dynamics in sewer systems – comparing mechanistic and conceptual modelling approaches
.
Water Sci. Technol.
84
(
9
),
2335
.
Saltelli
A.
,
Tarantola
S.
,
Campolongo
F.
&
Ratto
M.
2004
Sensitivity Analysis in Practice: A Guide to Assessing Scientific Models
.
John Wiley & Sons Ltd
,
Chichester, West Sussex England
.
Schlagbauer
P.
2018
Thermische Energiegewinnung aus Abwasser – Entwicklung eines Temperaturmodells für SWMM (Thermal energy recovery from wastewater - Development of a temperature model for SWMM)
.
Master Thesis
,
Graz University of Technology
,
Graz, Austria
.
Shafagh
I.
,
Shepley
P.
,
Shepherd
W.
,
Loveridge
F.
,
Schellart
A.
,
Tait
S.
&
Rees
S. J.
2022
Thermal energy transfer around buried pipe infrastructure
.
Geomech. Energy Environ.
29
,
100273
.
Sitzenfrei
R.
,
Hillebrand
S.
&
Rauch
W.
2017
Investigating the interactions of decentralized and centralized wastewater heat recovery systems
.
Water Sci. Technol.
75
(
5
),
1243
1250
.
Wärff
C.
,
Arnell
M.
,
Sehlén
R.
&
Jeppsson
U.
2020
Modelling heat recovery potential from household wastewater
.
Water Sci. Technol.
81
(
8
),
1597
1605
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).