the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
City-scale heating and cooling with Aquifer Thermal Energy Storage (ATES)
Haegyeong Lee
Philipp Blum
Kathrin Menberg
Abstract. Sustainable and climate-friendly space heating and cooling is of great importance for the energy transition. Compared to conventional energy sources, Aquifer Thermal Energy Storage (ATES) systems can significantly reduce greenhouse gas emissions from space heating and cooling. Hence, the objective of this study is to quantify the technical potential of shallow low-temperature ATES systems in terms of reclaimable energy in the city of Freiburg im Breisgau, Germany. Based on 3D heat transport modeling, heating and cooling power densities are determined for various hydrogeological subsurface characteristics and ATES configurations. High groundwater flow velocities of up to 13 m d-1 cause high storage energy loss limiting power densities to a maximum of 3.2 W m-2. Nevertheless, comparison of these power densities with the existing thermal energy demands shows that ATES systems can achieve substantial heating and cooling supply rates. This is especially true for the cooling demand, for which a full supply by ATES is determined for 92 % of all residential buildings in the study area. For ATES heating alone, potential greenhouse gas emission savings of up to about 70,000 tCO2eq a-1 are calculated, which equals about 40 % of the current greenhouse gas emissions caused by space and water heating in the study areas’ residential building stock. The modeling approach proposed in this study can also be applied in other regions with similar hydrogeological conditions to obtain estimations of local ATES supply rates and support city-scale energy planning.
- Preprint
(2350 KB) - Metadata XML
-
Supplement
(559 KB) - BibTeX
- EndNote
Ruben Stemmle et al.
Status: closed
-
RC1: 'Comment on hess-2023-62', Jannis Epting, 30 Apr 2023
Dear colleagues,
The paper "City-scale heating and cooling with Aquifer Thermal Energy Storage (ATES)" presented by Stemmle et al. represents an interesting case study that builds on previously developed methods. A new aspect is the simultaneous consideration of heating and cooling.
One major drawback is the missing information on the city-scale model (setup, parameterization, boundary conditions, calibration / validation, etc.). It is also not apparent whether this information has been published elsewhere.
The use of an “ambient temperature of 12 °C” should be justified. The use of an "ambient temperature of 12 °C" should be justified. In Freiburg, groundwater temperatures vary greatly, and the starting point will be different depending on the aquifer region.
The latter two aspects also lead me to decide that major revisions are needed.
More sepecific comments include:
p. 1, l. 13: provide information on the aquifer, unconsolidated gravel aquifer, …
p. 1, l. 13: specify «limiting power densities» aquifer
p. 2, l. 53: GWHP was already introduced, check whole text
p. 3, l. 65: “2D numerical box models” include: F. Bottcher, A. Casasso, G. Gotzl, K. Zosseder, TAP - thermal aquifer Potential: a quantitative method to assess the spatial potential for the thermal use of groundwater, Renew. Energy 142 (2019) 85e95.
p. 3, l. 76: GHG was already introduced, check whole text
p. 3, l. 94: specify “lower hydraulic conductivity”
Fig. 1a: missing scale bar (maybe zoom in, include Voges); m asl
Fig. 1b: flow direction of river Dreisam (maybe zoom in, show 10m equidistance); m asl
Fig. 1c: requires x-scaling, legend for lithologies, turn m asl 180°
Section 2.2.: Required? No new equations are developed. Reference could be sufficient.
Table 1: 4.4 E-3; 6 E-5 (reference personal communication?); References for standard values appropriate?
Fig. 2 could be merged with Fig. 1
Table 2: Value representation? Simplify & avoid repetitions.
p. 7, l. 170: Why estimated?
p. 7, l. 174.176: What is the difference between uniform and constant thickness?
p. 8, l. 193: Specify “substantial”
Fig. 4: turn labeling right axis asl 180°
p. 8, l. 211: When is thermal equilibrium reached? This measure maybe is more appropriate compared to the lifetime.
p. 10, l. 232-233: Discuss “ambient temperature of 12 °C”
p. 10, l. 249: Table 3?
p. 12, l.301: It would make sense to also calculate the SCOP as seasonal operation is investigated.
p. 13, eq. 14: see Epting et al. 2018
p. 17, l. 421-428: Compare to Epting et al. 2018 & 2020Good luck and all the best,
Jannis Epting
Citation: https://doi.org/10.5194/hess-2023-62-RC1 -
AC1: 'Reply on RC1', Ruben Stemmle, 22 Jun 2023
Dear Dr Epting,
thank you very much for your comments. We are convinced that your feedback can contribute to improve the quality of our manuscript. We therefore will also acknowledge your review in our Acknowledgments. Below you can find the answers to each of your comments.
"Dear colleagues,
The paper "City-scale heating and cooling with Aquifer Thermal Energy Storage (ATES)" presented by Stemmle et al. represents an interesting case study that builds on previously developed methods. A new aspect is the simultaneous consideration of heating and cooling.
One major drawback is the missing information on the city-scale model (setup, parameterization, boundary conditions, calibration / validation, etc.). It is also not apparent whether this information has been published elsewhere. "
Answer: We agree. You are referring to crucial aspects regarding the city-scale model. The required information can be found in the Supplement to the manuscript, where we state details on the model parameterization, boundary conditions, calibration etc. The Supplement can be accessed through the preprint webpage. We also attached the supplement pdf file to this reply.
"The use of an “ambient temperature of 12 °C” should be justified. The use of an "ambient temperature of 12 °C" should be justified. In Freiburg, groundwater temperatures vary greatly, and the starting point will be different depending on the aquifer region.
The latter two aspects also lead me to decide that major revisions are needed."
Answer: We agree. The choice of assuming a uniform ambient groundwater temperature of 12 °C is discussed in more detail in the revised manuscript in Chapter 3.5 “Limitations of the box model approach”.
The mean value of 12°C was chosen after consulting the Environmental Protection Authority of Freiburg, which is in charge of regulating groundwater withdrawals. Naturally, the temperature of the ambient groundwater varies during the course of a year and between different parts of the city. Unfortunately, these variations are largely unknown for the city area, as detailed spatial and temporal monitoring data is not available. However, the assumption of a uniform temperature across all box models is deemed appropriate regarding the simplified modelling approach in this study. As injection and reinjection temperatures of the ATES systems, as well as the criterion for the delineation of the thermal plumes refer to relative temperature differences, the impact of the chosen absolute ambient temperature will be minor with respect to the influence of e.g. groundwater flow velocity.
However, modelling of site-specific storage systems, e.g. for planning purposes, would require more detailed information including the ambient and local groundwater temperatures. This aspect is now discussed in the revised manuscript.
"More sepecific comments include:
p. 1, l. 13: provide information on the aquifer, unconsolidated gravel aquifer, …"Answer: We agree and included this information on the aquifer in the corresponding passage of the abstract in our revised manuscript version.
"p. 1, l. 13: specify «limiting power densities» aquifer"Answer: Done. The loss of stored thermal energy due to groundwater flow is an important factor that can reduce the extractable thermal energy and thus the power density of ATES systems. In this case, the maximum achievable power density considers the energy loss is 3.2 W m‑2. We specify this more clearly in the revised manuscript.
"p. 2, l. 53: GWHP was already introduced, check whole text"
Answer: Done. We checked the whole text and now introduce the abbreviation only once in the new manuscript version.
"p. 3, l. 65: “2D numerical box models” include: F. Bottcher, A. Casasso, G. Gotzl, K. Zosseder, TAP - thermal aquifer Potential: a quantitative method to assess the spatial potential for the thermal use of groundwater, Renew. Energy 142 (2019) 85e95."
Answer: Done. The cited paper is now referenced at the corresponding text passage.
"p. 3, l. 76: GHG was already introduced, check whole text"Answer: Done. We checked the whole text and now introduce the abbreviation only once.
"p. 3, l. 94: specify “lower hydraulic conductivity”"
Answer: Done. We now specify this part by adding a reference to Table 1, where hydraulic conductivities are given for the model area.
"Fig. 1a: missing scale bar (maybe zoom in, include Voges); m asl"Answer: Done. We adjusted the figure accordingly.
"Fig. 1b: flow direction of river Dreisam (maybe zoom in, show 10m equidistance); m asl"Answer: Done. We adjusted the figure accordingly.
"Fig. 1c: requires x-scaling, legend for lithologies, turn m asl 180°"Answer: Done. We adjusted the figure accordingly.
"Section 2.2.: Required? No new equations are developed. Reference could be sufficient. "Answer: It is true that no new equations are developed in section 2.2. Nevertheless, we still prefer to state the relevant equations in the text, which is in accordance with other similar studies.
"Table 1: 4.4 E-3; 6 E-5 (reference personal communication?); References for standard values appropriate?"
Answer: Done. We adjusted the numbers accordingly. In our experience, it is reasonable to reference personal communication with authorities such as the Baden-Württemberg State Office for Geology, Raw Materials and Mining (LGRB) or the Environmental Protection Authority of Freiburg. These authorities have profound knowledge of the subsurface and groundwater conditions in the study area.
"Fig. 2 could be merged with Fig. 1"Answer: While it is in principle possible to merge these two figures, we prefer to keep them separate. Fig. 2 shows the delineated hydrogeological regions, which is an important step of the implemented methodology. In contrast, Fig. 1 provides relevant geographical and geological information on the study area. Separating these information also helps to keep a clear structure in the manuscript.
"Table 2: Value representation? Simplify & avoid repetitions."Answer: Done. We simplified Table 2 by removing unnecessary repeated values that are already stated in Table 1.
"p. 7, l. 170: Why estimated?"Answer: The stated groundwater flow velocity of 29.1 m d-1 for region 4 is not measured nor modelled using the city-scale model. It is calculated from the hydraulic conductivity, hydraulic gradient and effective porosity. We adjusted the text to make this clearer.
"p. 7, l. 174.176: What is the difference between uniform and constant thickness?"Answer: There is no difference in this context. We therefore only use the word “uniform” for both cases.
"p. 8, l. 193: Specify “substantial”"
Answer: We chose to not state more specific values here, since the precise extent of the thermal losses is part of the results and thus elaborated on later in the manuscript. Chapter 3.1 “Thermal recovery” provides exact values in more detail.
"Fig. 4: turn labeling right axis asl 180°"Answer: Done. We adjusted the figure accordingly.
"p. 8, l. 211: When is thermal equilibrium reached? This measure maybe is more appropriate compared to the lifetime."
Answer: The time needed to reach thermal equilibrium in the ATES box models strongly depends on the hydrogeological region, the aquifer formation and the ATES well configuration. For this reason, we use a uniform modelling period of 30 years across all box models as well as for the city-scale model. This time period is based on the typical expected ATES systems lifetime. According references are stated in the manuscript.
"p. 10, l. 232-233: Discuss “ambient temperature of 12 °C”"
Answer: Done. In the revised manuscript, we discuss now our assumption of an ambient groundwater temperature of 12 °C in Chapter 3.5 “Limitations of the box model approach”. Please also refer to the more detailed answer on this aspect above.
"p. 10, l. 249: Table 3?"Answer: Corrected. This table should be indeed Table 4.
"p. 12, l.301: It would make sense to also calculate the SCOP as seasonal operation is investigated."
Answer: While the ATES operation is simulated using several seasonal extraction and injection phases over the course of a year, the evaluation of the extracted energy is done for whole seasons in order to determine the thermal recovery. In heating mode, the considered energy used to calculate the power density includes an additional term resulting from heat pump operation with an assumed COP of 3.5. This value of 3.5 is also the average COP during the heating season, i.e. the SCOP.
"p. 13, eq. 14: see Epting et al. 2018"
Answer: We added two references for eq. 14, including also Epting et al. (2018).
"p. 17, l. 421-428: Compare to Epting et al. 2018 & 2020"Answer: We now discuss the power density values and the underlying methodology from Epting et al. (2020) in this text passage. However, we decided not to include the study by Epting et al. (2018) in this part of the discussion, because this specific study deals with a more theoretical thermal potential of shallow geothermal resources. Also, the values in Epting et al. (2018) are not referring to a specific technology, which hinders meaningful comparison to our results.
"Good luck and all the best,
Jannis Epting"
Thank you for your time and feedback on our work!
-
AC1: 'Reply on RC1', Ruben Stemmle, 22 Jun 2023
-
RC2: 'Comment on hess-2023-62', Anonymous Referee #2, 14 May 2023
This manuscript describes a case study in which the shallow geothermal system Aquifer Thermal Energy Storage (ATES) is analyzed for the city of Freiburg in Germany. The work is largely presented as case study, and the question is why a reader in Florence, Oklahoma City, or Nanjing should be interested in reading it. If this manuscript addresses readers who are really interested in the city of Freiburg, the German journal “Grundwasser” would be the right place to publish it, because the latter journal made for a German audience. A paper in HESS, by contrast, should present either new insights in hydrological system behavior, or new methods of general interest that can be applied at many places. I honestly don’t see that the authors have advanced our general understanding of heat transport in the subsurface, but I also have difficulties to recognize methodological breakthroughs. Using simplified box-shaped models with a uniform ambient hydraulic gradient, is not really new, and the concept of the power density has been introduced before. So what is really new and of interest to a broad readership?
I have to admit that my specialty is not in the design of shallow geothermal systems, but I understand the underlying concepts and governing equations quite well. Having said that, I am not convinced about the concept of the power density. The natural geothermal heat flux in the region is about 0.1 W/m2. If you plaster the domain with many geothermal wells, all of them extracting heat, and you insulate the top (as done by the authors), the long-term power density will equal the natural geothermal heat flux of 0.1 W/m2. The trick of ATES is of course that heat is injected into the ground in the summer and extracted in the winter, which makes even the scenario of a plastered domain with an insulated top less dependent on the natural geothermal heat flux. What the authors do, is to consider a rectangular box with a single or two geothermal well pairs or triplets, run the system for 30 years, and then outline the 0.5 K contourline of the temperature anomaly. This design is not representative if you consider a second/third/fourth row of well pairs/triplets that will interfere. This is a simple matter of superposition, which applies to linear transport. A realistic small-scale scenario representative for an entire city would use periodic boundary conditions, in which the heat flux leaving the domain at the downstream end enters at the upstream end. If you run the system until you reach dynamic steady state, you see how the geothermal systems influence each other in the middle of a periodic domain. This would make much more sense then outlining an arbitrarily chosen contourline of a lonesome geothermal system. A systematic analysis of heat fluxes (how much is recycled, what is the difference in the incoming versus the outgoing heat flux with ambient flow if you don’t assume periodicity, what is the heat flux at the non-insulated top boundary) would make more sense to me then the power density, which only gives the illusion of a practical metric.
I could imagine that the work could be converted into a methodological study, outlining how small-scale models can be used to predesign large-scale optimization of geothermal systems. But the manuscript is not written in such a way, and a real large-scale optimal design is not attempted at all.
Notwithstanding these fundamental concerns, I have a series of minor comments:
- Line 34: “thermal energy” rather than “environmental”
- Line 55: “revealed” (past tense)
- Line 89: Shouldn’t it be “alluvial fan” rather than “cone”?
- Lines 92-94: Outside of Baden-Württemberg, nobody knows the Neuenburg and Bresigau formations. Give at least their age (their both from the Pleistocene)
- Equation 3: You use a scalar dispersive thermal conductivity, but later on you use a longitudinal and a transverse dispersivity. This implies applying the Scheidegger parameterization, and it implies anisotropic dispersive heat transfer. The corresponding equations are missing.
- Equation 5, line 131: if p is supposed to be the pore pressure, you need to add gravity to Darcy’s law: u = -K*grad(p/(ρ*g)+z). Alternatively, you can define p as the total potential (pore pressure plus ρ*g*z).
- Line 132: I assume that you neglect the temperature dependence of hydraulic conductivity, which stems from the temperature dependence of viscosity. We are talking about a factor of two for 10 K temperature difference.
- Equation 6: I assume that you use the specific storativity to relate the given time derivative to the rate of change in pore pressure/potential.
- Table 1: It’s “transverse dispersivity” rather than “transversal”, and the coefficient is about two orders of magnitude too large. Stochastic theory reveals that the longitudinal dispersion coefficient (if you believe in Fickian dispersion) increases by orders of magnitude upon upscaling, whereas the transverse dispersion coefficient is hardly affected at all.
- Figure 3 and its discussion: The heat-related boundary conditions are incomplete. What is the bottom boundary? How do you treat temperature at the up- and downstream faces?
- Section 2.6: I have expressed my concerns regarding the power density above. In fact, the results of the city-scale model confirm the importance of the interference across boxes. I am not convinced.
- Lines 348-349: is the observation that ambient flow influences the thermal recovery in ATES systems new at all?
- Figure 7: This plot shows a linear dependence of TR on ln(u). It should be possible to approximate the corresponding coefficients by analytical considerations, but the authors not even do the linear regression.
- Lines 431-432: For the given simple setup with uniform ambient flow I am not convinced that numerical models truly outperform analytical expressions.
- Lines 451-454: I expressed doubts about the thermal insulation on top already further up. In fact, a large fraction of the heat in shallow geothermal systems comes from the surface. It’s still reasonable to prefer these systems over air heat pumps because you change the timing.
- Line 468 ff.: Here the discussion becomes by far too local. I value the cities of Ludwigsburg (nice baroque castle), Lausanne (Olympic museum on Lake Geneva), and Geneva (impressive fountain in the lake), but the authors simply list a couple of case studies. You need to concentrate on general principles and methods instead!
- Section 3.4: I am not convinced about the savings of greenhouse-gas emissions either. This all depends on the specific energy mix, for the existing heating systems, and for electricity production. On the long run, all heating has to be done with renewable energy. In Freiburg, deep geothermal energy would be an option. If you exclude that, you get stuck to heat pumps. The rght comparison in my opinion is to based on the electric-energy demand for heating/cooling with shallow geothermal systems versus air heat pumps. The (hopefully renewably produced) electricity saved is the long-term argument for ATES.
- Section 3.5: I am not convinced that the city-scale model is really computationally too demanding. In this model you can address the interferences of individual systems, and you don’t rely on parallel ambient flow. The goal should be to optimize the big system. If you can show how to do that in general (with the help of box models for predefine or not), you have a HESS-worthy publication.
Citation: https://doi.org/10.5194/hess-2023-62-RC2 -
AC2: 'Reply on RC2', Ruben Stemmle, 23 Jun 2023
Dear colleague,
thank you very much for your thorough feedback on our manuscript. Below, we respond to each of your comments.
"This manuscript describes a case study in which the shallow geothermal system Aquifer Thermal Energy Storage (ATES) is analyzed for the city of Freiburg in Germany. The work is largely presented as case study, and the question is why a reader in Florence, Oklahoma City, or Nanjing should be interested in reading it. If this manuscript addresses readers who are really interested in the city of Freiburg, thöe German journal “Grundwasser” would be the right place to publish it, because the latter journal made for a German audience. A paper in HESS, by contrast, should present either new insights in hydrological system behavior, or new methods of general interest that can be applied at many places. I honestly don’t see that the authors have advanced our general understanding of heat transport in the subsurface, but I also have difficulties to recognize methodological breakthroughs. Using simplified box-shaped models with a uniform ambient hydraulic gradient, is not really new, and the concept of the power density has been introduced before. So what is really new and of interest to a broad readership?"
Answer: While the concept of power density has been introduced before (as stated in our manuscript), our study expands the concept of power densities to determine the technical potential of ATES on a city-scale level for the first time. To that extent, the study’s aim is not to improve the understanding of subsurface heat transport, however to provide a novel methodology, which joins existing approaches into a new and novel framework, and showcasing its applicability on an exemplary study area. While the study area is naturally geographically constrained, the developed methodology can easily be adopted for other cities and regions. Thus, we are convinced that our study will be of interest for many researchers and practitioners alike, as new strategies for urban energy and subsurface planning are urgently needed to increase the share of renewable energy in urban areas.
"I have to admit that my specialty is not in the design of shallow geothermal systems, but I understand the underlying concepts and governing equations quite well. Having said that, I am not convinced about the concept of the power density. The natural geothermal heat flux in the region is about 0.1 W/m2. If you plaster the domain with many geothermal wells, all of them extracting heat, and you insulate the top (as done by the authors), the long-term power density will equal the natural geothermal heat flux of 0.1 W/m2. The trick of ATES is of course that heat is injected into the ground in the summer and extracted in the winter, which makes even the scenario of a plastered domain with an insulated top less dependent on the natural geothermal heat flux."
Answer: We believe there is a misunderstanding regarding the operational principle of ATES, the concept of power density and the energy fluxes in the shallow subsurface.
First, we would like to emphasize that the geological subsurface and the groundwater are the storage medium of an ATES system, rather than the primary source for extracting thermal energy. In contrast to conventional shallow geothermal systems, ATES offers the possibility to store and re-use for example excess waste heat, which makes the amount of storable energy largely independent of the natural heat fluxes (see also response to comment #15). Indeed, we simulate ATES systems with a varying, but seasonally balanced extraction and injection of thermal energy.
Secondly, while the geothermal heat flux is of the same physical unit as the power density (W m‑2), the power density as utilized in our manuscript refers to the energy usable with a specific technology. Since ATES systems use energy from other sources than the subsurface, they can have much higher power densities than the natural heat flux. Assuming a balanced ATES operation, the obtained power density values are also true for long-term system operation, and allow quantitative comparison with other energy sources with respect to space requirements (e.g. Kammen & Sunter, 2016).
"What the authors do, is to consider a rectangular box with a single or two geothermal well pairs or triplets, run the system for 30 years, and then outline the 0.5 K contourline of the temperature anomaly. This design is not representative if you consider a second/third/fourth row of well pairs/triplets that will interfere. This is a simple matter of superposition, which applies to linear transport. A realistic small-scale scenario representative for an entire city would use periodic boundary conditions, in which the heat flux leaving the domain at the downstream end enters at the upstream end. If you run the system until you reach dynamic steady state, you see how the geothermal systems influence each other in the middle of a periodic domain. This would make much more sense then outlining an arbitrarily chosen contourline of a lonesome geothermal system. A systematic analysis of heat fluxes (how much is recycled, what is the difference in the incoming versus the outgoing heat flux with ambient flow if you don’t assume periodicity, what is the heat flux at the non-insulated top boundary) would make more sense to me then the power density, which only gives the illusion of a practical metric."
Answer: First, we would like to clarify that we do not simulate well triplets, but in fact ATES with one, two or three well doublets depending on the system configuration. Thus, interference of these well doublets is an integral part of the operational scheme and modelled numerically. This is now clearly stated in the revised manuscript.
The reviewer raises a valid point in saying that neighboring ATES with multiple well doublets might also interfere. However, German regulations (amongst others) require planning and operation of geothermal systems in a way that avoids thermal interference between individual systems above a certain threshold temperature (e.g. 0.5 K). Thus, in accordance with the aim of the study to provide a planning methodology for ATES systems, we do not allow for such thermal interference in our approach, which is also true for many other international studies in this field (e.g. Gizzi et al., 2020; Piga et al., 2017; Visser et al., 2015). The limitations of our approach in this respect are discussed in the manuscript, including a note that mutual thermal interference might actually have an overall positive effect on system efficiencies on a large scale.
In our opinion, a detailed analysis of the heat fluxes of the box models would not provide added value to the assessment of the ATES potential, for the reasons already mentioned in our previous response.
"I could imagine that the work could be converted into a methodological study, outlining how small-scale models can be used to predesign large-scale optimization of geothermal systems. But the manuscript is not written in such a way, and a real large-scale optimal design is not attempted at all."
Answer: Presenting a novel methodology for city-scale assessment of the technical ATES potential comprises indeed a major part of our study, and our manuscript is structured accordingly. To highlight this issue, we reformulated our objective of this study.
Large-scale optimization is beyond the scope of our study for several reasons. First, to make optimization on this scale meaningful, thermal interference between individual systems would have to be considered, which is currently not realistic (see response above). Also, prioritizing optimal large-scale thermal use over individual system performance would require city-scale energy management above ground, i.e. city-wide energy distribution systems for heat and cold, which are often not feasible in reality. However, our approach can serve as a first step for city-wide planning of ATES, and thus also for optimizing the thermal use of the subsurface in the future. This option is now discussed in the revised manuscript.
Furthermore, a single run of the city-scale model takes about 50 hours (as stated in the manuscript), which hinders the applicability of comprehensive optimization schemes, which often require hundreds of iterative simulation runs. While it might be possible on an HPC cluster, it would significantly impede the applicability of our approach for municipal energy planning.
"Notwithstanding these fundamental concerns, I have a series of minor comments:
1. Line 34: “thermal energy” rather than “environmental”"
Answer: We agree that this expression fits better within the context of the sentence and use the term “thermal energy” in the revised manuscript.
"2. Line 55: “revealed” (past tense)"
Answer: Done. We adjusted the word to past tense in the revised manuscript.
"3. Line 89: Shouldn’t it be “alluvial fan” rather than “cone”?"
Answer: Correct. We now use the more common term ‘alluvial fan’.
"4. Lines 92-94: Outside of Baden-Württemberg, nobody knows the Neuenburg and Bresigau formations. Give at least their age (their both from the Pleistocene)"
Answer: Done. We now state the Pleistocene age of both formations in the new manuscript version besides describing their hydrogeological characteristics relevant to our study.
"5. Equation 3: You use a scalar dispersive thermal conductivity, but later on you use a longitudinal and a transverse dispersivity. This implies applying the Scheidegger parameterization, and it implies anisotropic dispersive heat transfer. The corresponding equations are missing."
Answer: True, we use a longitudinal and a transverse dispersivity and the corresponding equations were missing. Hence, we now state the equations for calculating the disperse thermal conductivity from the longitudinal and the transverse dispersivities in the revised manuscript.
"6. Equation 5, line 131: if p is supposed to be the pore pressure, you need to add gravity to Darcy’s law: u = -K*grad(p/(ρ*g)+z). Alternatively, you can define p as the total potential (pore pressure plus ρ*g*z)."
Answer: Since we do not set topographic elevations in our simplified box models, we do not account for gravitational effects in the box models. Preliminary simulations during development of the box model approach showed that including gravitation in the models had only a very minor and therefore neglectable effect on the thermal plume propagation, which is the relevant aspect for determining power densities.
"7. Line 132: I assume that you neglect the temperature dependence of hydraulic conductivity, which stems from the temperature dependence of viscosity. We are talking about a factor of two for 10 K temperature difference."
Answer: The temperature dependence of viscosity and therefore hydraulic conductivity is indeed neglected in our study. This is a common approach for numerical simulations of heat transport processes in shallow groundwater for low-temperature geothermal systems (e.g. Hecht-Méndez et al., 2010; Beernink et al., 2022; Bidarmaghz et al., 2020; Previati et al., 2022).
The factor between hydraulic conductivities at the maximum (18 °C) and the minimum (6 °C) temperatures in our study is actually around 1.4 (rather than 2). Also, these “extreme” temperatures occur over a very limited spatial area, which makes the effect of varying hydraulic conductivities on the entire plume length negligible.
"8. Equation 6: I assume that you use the specific storativity to relate the given time derivative to the rate of change in pore pressure/potential."
Answer: Our models simulate the thermo-hydraulic behavior of ATES systems in an unconfined aquifer. In this type of aquifer, the specific storage resulting from compressibility of groundwater and aquifer matrix is orders of magnitude smaller than the storage effects from the matrix porosity. We therefore only consider the porosity as a storage parameter in the sense of the specific yield.
"9. Table 1: It’s “transverse dispersivity” rather than “transversal”, and the coefficient is about two orders of magnitude too large. Stochastic theory reveals that the longitudinal dispersion coefficient (if you believe in Fickian dispersion) increases by orders of magnitude upon upscaling, whereas the transverse dispersion coefficient is hardly affected at all."
Answer: Done. We changed the wording mistake and use the correct term “transverse dispersivity” now in Table 1.
Previous studies by Zech et al. (2015, 2018) suggest that the commonly used scale-dependency of the longitudinal dispersivity is not valid. The dependence of longitudinal and transverse dispersivity on the length scale of the transport process is however a field of active and current research (e.g. Di Dato et al., 2022; Park & Lee, 2021; Pophillat et al., 2020; Younes et al., 2020). Furthermore, only few reported values for thermal dispersivities values obtained by field experiments exist (Stauffer et al. 2014). Hence, for the current study with the focus on the development of a novel methodology for city-scale assessment of the technical ATES potential, we assume commonly used thermal dispersivities values reported in the literature (e.g. Stauffer et al., 2014).
"10. Figure 3 and its discussion: The heat-related boundary conditions are incomplete. What is the bottom boundary? How do you treat temperature at the up- and downstream faces?"
Answer: We agree and added the missing details on the heat-related boundary conditions (BC) and include an adjusted figure in the revised manuscript. Like for the top boundary, we implemented a no heat flux BC at the model bottom. This is also true for the upstream and downstream model faces, which is line with the box model approach of preventing any thermal interferences between individual ATES systems.
"11. Section 2.6: I have expressed my concerns regarding the power density above. In fact, the results of the city-scale model confirm the importance of the interference across boxes. I am not convinced."
Answer: We already addressed the reviewer’s concerns regarding the power density above. The conservative way of delineating the spatial extent of the thermal impact of individual ATES systems prevents any significant thermal interferences between ATES systems. This is also the common approach for planning, regulatory approval and operation of such systems. Accordingly, interference is not accounted for in the box models.
"12. Lines 348-349: is the observation that ambient flow influences the thermal recovery in ATES systems new at all?"
Answer: The influence of the ambient groundwater flow on the thermal recovery, i.e. storage efficiency of ATES systems has indeed been studied previously as stated in Section 2.4.2. We here discuss these effects, because they impact the choice of optimal ATES design, which is a new observation, and because the thermal recovery directly influences the power density, whose estimation is the key aspect of this study.
"13. Figure 7: This plot shows a linear dependence of TR on ln(u). It should be possible to approximate the corresponding coefficients by analytical considerations, but the authors not even do the linear regression."
Answer: While the idea of inferring a dependence of the thermal recovery on the groundwater flow velocity is interesting, this is beyond the scope of this study.
In our proposed methodological workflow, the thermal recovery is a byproduct from the numerical modeling results, which is worth discussing as it represents a main efficiency criterion for ATES. Yet, inferring a robust statistical relationship between the two factors would require a very comprehensive modelling exercise, without adding significant value to the presented approach and objective of this study.
"14. Lines 431-432: For the given simple setup with uniform ambient flow I am not convinced that numerical models truly outperform analytical expressions."
Answer: We are not aware of any analytical solution that offers equivalent possibilities as our numerical setup. Using numerical models for simulating ATES systems with numerous extraction and injection wells operated under groundwater flow conditions and according to seasonally varying pumping schemes with reversing pumping directions and passive storage phases is considered state of the art (e.g. De Paoli, 2023, Duijff et al., 2021, Regnier et al., 2022).
"15.Lines 451-454: I expressed doubts about the thermal insulation on top already further up. In fact, a large fraction of the heat in shallow geothermal systems comes from the surface. It’s still reasonable to prefer these systems over air heat pumps because you change the timing."
Answer: As explained in a previous response, the extracted (or recovered) heat in an ATES does not stem from the ambient (sub)surface, but mainly from waste heat from e.g. space cooling or industrial cooling processes and excess heat from CHP plants. ATES systems are therefore not dependent on the heat input (or recovery) from the surface, however rather benefit from the insulating effect of the subsurface.
"16. Line 468 ff.: Here the discussion becomes by far too local. I value the cities of Ludwigsburg (nice baroque castle), Lausanne (Olympic museum on Lake Geneva), and Geneva (impressive fountain in the lake), but the authors simply list a couple of case studies. You need to concentrate on general principles and methods instead!"
Answer: We agree and adjusted the corresponding part in Section 3.3, where we compare our supply rates with previous publications. The revised manuscript presents a reduced text passage focusing on relevant similarities and differences among the studies. The discussion of specific geographical references was removed.
"17. Section 3.4: I am not convinced about the savings of greenhouse-gas emissions either. This all depends on the specific energy mix, for the existing heating systems, and for electricity production. On the long run, all heating has to be done with renewable energy. In Freiburg, deep geothermal energy would be an option. If you exclude that, you get stuck to heat pumps. The rght comparison in my opinion is to based on the electric-energy demand for heating/cooling with shallow geothermal systems versus air heat pumps. The (hopefully renewably produced) electricity saved is the long-term argument for ATES."
Answer: We agree that the savings of GHG emissions are governed by the specific heat and electricity mix. Hence, we use the latest available data on the city’s energy mix for space heating from 2020 (see section 3.4). Emissions from ATES systems are adopted from Stemmle et al. (2021), and account for the German electricity mix as of 2014.
We also agree that different options for renewable heat (and cold) supply have to be considered in future in order to fulfill different needs, such as deep (hydrothermal) geothermal systems for high-temperature heat, which are indeed an option in the city of Freiburg. Given the diversity of the current (and presumably future) heat mix a comparison solely against air-source heat pumps (ASHP), which currently make up less than 1% of Freiburg’s heat mix, does not appear meaningful to us. However, we added a short discussion comparing the energy demands of ASHP and ATES systems to the revised manuscript.
"18. Section 3.5: I am not convinced that the city-scale model is really computationally too demanding. In this model you can address the interferences of individual systems, and you don’t rely on parallel ambient flow. The goal should be to optimize the big system. If you can show how to do that in general (with the help of box models for predefine or not), you have a HESS-worthy publication."
Answer: Please see our response on the major comment regarding large-scale optimization above.
References:
Beernink, S., Bloemendal, M., Kleinlugtenbelt, R. and Hartog, N. (2022)Maximizing the use of aquifer thermal energy storage systems in urban areas: effects on individual system primary energy use and overall GHG emissions. Applied Energy, 311, 118587.
Bidarmaghz, A., Choudhary, R., Soga, K., Terrington, R.L., Kessler, H. and Thorpe, S. (2020) Large-scale urban underground hydro-thermal modelling – A case study of the Royal Borough of Kensington and Chelsea, London. Science of The Total Environment, 700, 134955.
De Paoli, C., Duren, T., Petitcler, E., Agniel. M. and Dassargues, A. (2023) Modelling Interactions between Three Aquifer Thermal Energy Storage (ATES) Systems in Brussels (Belgium). Applied Sciences, 13(5), 2934.
Di Dato, M., D’Angelo, C., Casasso, A. and Zarlenga, A. (2022) The impact of porous medium heterogeneity on thermal feedback of open-loop shallow geothermal systems. Journal of Hydrology, 604, 127205.
Duijff, R., Bloemendal, M. and Bakker, M. (2021) Interaction Effects Between Aquifer Thermal Energy Storage Systems. Ground water, 61(2), 173-182.
Gizzi, M., Taddia, G., Cerino Abdin, E. and Lo Russo, S. (2020) Thermally Affected Zone (TAZ) Assessment in Open-Loop Low-Enthalpy Groundwater Heat Pump Systems (GWHPs): Potential of Analytical Solutions. Geofluids, 2020, 1-13.
Hecht-Méndez, J., Molina-Giraldo, N., Blum, P. and Bayer, P. (2010) Evaluating MT3DMS for heat transport simulation of closed shallow geothermal systems. Ground Water, 48(5), 741–756.
Kammen, D.M. and Sunter, D.A. (2016) City-integrated renewable energy for urban sustainability. Science, 352, 922-928.
Mueller, M.H., Huggenberger, P. and Epting, J. (2018) Combining monitoring and modelling tools as a basis for city-scale concepts for a sustainable thermal management of urban groundwater resources. Science of the Total Environment, 627, 1121-1136.
Park, B.-H. and Lee, K.-K. (2021) Evaluating anisotropy ratio of thermal dispersivity affecting geometry of plumes generated by aquifer thermal use. Journal of Hydrology, 602, 126740.
Piga, B., Casasso, A., Pace, F., Godio, A. and Sethi, R. (2017) Thermal Impact Assessment of Groundwater Heat Pumps (GWHPs): Rigorous vs. Simplified Models. Energies, 10 (9), 1385.
Pophillat, W., Bayer, P., Teyssier, E., Blum, P. and Attard, G. (2020) Impact of groundwater heat pump systems on subsurface temperature under variable advection, conduction and dispersion. Geothermics, 83, 101721.
Previati, A., Epting, J. and Crosta, G.B. (2022) The subsurface urban heat island in Milan (Italy) - A modeling approach covering present and future thermal effects on groundwater regimes. Science of The Total Environment, 810, 152119.
Regnier, G., Salinas, P., Jacquemyn, C. and Jackson, M.D. (2022) Numerical simulation of aquifer thermal energy storage using surface-based geological modelling and dynamic mesh optimisation. Hydrogeology Journal, 30, 1179-1198.
Rivera, J.A., Blum, P. and Bayer, P. (2017) Increased ground temperatures in urban areas: Estimation of the technical geothermal potential. Renewable Energy, 103, 388-400.
Stauffer, F., Bayer, P., Blum, P., Molina-Giraldo, N. and Kinzelbach W. (2013) Thermal Use of Shallow Groundwater. 287 pages, CRC Press.
Visser, P.W., Kooi, H. and Stuyfzand, P.J. (2015) The thermal impact of aquifer thermal energy storage (ATES) systems: a case study in the Netherlands, combining monitoring and modelling. Hydrogeology Journal, 13, 507-532.
Younes, A., Fahs, M., Ataie-Asthiani, B. and Simmons, C.-T. (2020) Effect of distance-dependent dispersivity on density-driven flow in porous media. Journal of Hydrology, 589, 125204.
Zech, A., Attinger, S., Cvetkovic, V. et al. (2015) Is unique scaling of aquifer macrodispersivity supported by field data? Water Resources Research, 51(9), 7662-7679.
Zech, A., Attinger, S., Alberto, B. et al. (2019) A Critical Analysis of Transverse Dispersivity Field Data. Ground water, 54(4), 632-639.
Zhu, K., Blum, P., Ferguson, G., Balke, K.D. and Bayer, P. (2010) The geothermal potential of urban heat islands. Environmental Research Letters, 5, 044002.
Citation: https://doi.org/10.5194/hess-2023-62-AC2
Status: closed
-
RC1: 'Comment on hess-2023-62', Jannis Epting, 30 Apr 2023
Dear colleagues,
The paper "City-scale heating and cooling with Aquifer Thermal Energy Storage (ATES)" presented by Stemmle et al. represents an interesting case study that builds on previously developed methods. A new aspect is the simultaneous consideration of heating and cooling.
One major drawback is the missing information on the city-scale model (setup, parameterization, boundary conditions, calibration / validation, etc.). It is also not apparent whether this information has been published elsewhere.
The use of an “ambient temperature of 12 °C” should be justified. The use of an "ambient temperature of 12 °C" should be justified. In Freiburg, groundwater temperatures vary greatly, and the starting point will be different depending on the aquifer region.
The latter two aspects also lead me to decide that major revisions are needed.
More sepecific comments include:
p. 1, l. 13: provide information on the aquifer, unconsolidated gravel aquifer, …
p. 1, l. 13: specify «limiting power densities» aquifer
p. 2, l. 53: GWHP was already introduced, check whole text
p. 3, l. 65: “2D numerical box models” include: F. Bottcher, A. Casasso, G. Gotzl, K. Zosseder, TAP - thermal aquifer Potential: a quantitative method to assess the spatial potential for the thermal use of groundwater, Renew. Energy 142 (2019) 85e95.
p. 3, l. 76: GHG was already introduced, check whole text
p. 3, l. 94: specify “lower hydraulic conductivity”
Fig. 1a: missing scale bar (maybe zoom in, include Voges); m asl
Fig. 1b: flow direction of river Dreisam (maybe zoom in, show 10m equidistance); m asl
Fig. 1c: requires x-scaling, legend for lithologies, turn m asl 180°
Section 2.2.: Required? No new equations are developed. Reference could be sufficient.
Table 1: 4.4 E-3; 6 E-5 (reference personal communication?); References for standard values appropriate?
Fig. 2 could be merged with Fig. 1
Table 2: Value representation? Simplify & avoid repetitions.
p. 7, l. 170: Why estimated?
p. 7, l. 174.176: What is the difference between uniform and constant thickness?
p. 8, l. 193: Specify “substantial”
Fig. 4: turn labeling right axis asl 180°
p. 8, l. 211: When is thermal equilibrium reached? This measure maybe is more appropriate compared to the lifetime.
p. 10, l. 232-233: Discuss “ambient temperature of 12 °C”
p. 10, l. 249: Table 3?
p. 12, l.301: It would make sense to also calculate the SCOP as seasonal operation is investigated.
p. 13, eq. 14: see Epting et al. 2018
p. 17, l. 421-428: Compare to Epting et al. 2018 & 2020Good luck and all the best,
Jannis Epting
Citation: https://doi.org/10.5194/hess-2023-62-RC1 -
AC1: 'Reply on RC1', Ruben Stemmle, 22 Jun 2023
Dear Dr Epting,
thank you very much for your comments. We are convinced that your feedback can contribute to improve the quality of our manuscript. We therefore will also acknowledge your review in our Acknowledgments. Below you can find the answers to each of your comments.
"Dear colleagues,
The paper "City-scale heating and cooling with Aquifer Thermal Energy Storage (ATES)" presented by Stemmle et al. represents an interesting case study that builds on previously developed methods. A new aspect is the simultaneous consideration of heating and cooling.
One major drawback is the missing information on the city-scale model (setup, parameterization, boundary conditions, calibration / validation, etc.). It is also not apparent whether this information has been published elsewhere. "
Answer: We agree. You are referring to crucial aspects regarding the city-scale model. The required information can be found in the Supplement to the manuscript, where we state details on the model parameterization, boundary conditions, calibration etc. The Supplement can be accessed through the preprint webpage. We also attached the supplement pdf file to this reply.
"The use of an “ambient temperature of 12 °C” should be justified. The use of an "ambient temperature of 12 °C" should be justified. In Freiburg, groundwater temperatures vary greatly, and the starting point will be different depending on the aquifer region.
The latter two aspects also lead me to decide that major revisions are needed."
Answer: We agree. The choice of assuming a uniform ambient groundwater temperature of 12 °C is discussed in more detail in the revised manuscript in Chapter 3.5 “Limitations of the box model approach”.
The mean value of 12°C was chosen after consulting the Environmental Protection Authority of Freiburg, which is in charge of regulating groundwater withdrawals. Naturally, the temperature of the ambient groundwater varies during the course of a year and between different parts of the city. Unfortunately, these variations are largely unknown for the city area, as detailed spatial and temporal monitoring data is not available. However, the assumption of a uniform temperature across all box models is deemed appropriate regarding the simplified modelling approach in this study. As injection and reinjection temperatures of the ATES systems, as well as the criterion for the delineation of the thermal plumes refer to relative temperature differences, the impact of the chosen absolute ambient temperature will be minor with respect to the influence of e.g. groundwater flow velocity.
However, modelling of site-specific storage systems, e.g. for planning purposes, would require more detailed information including the ambient and local groundwater temperatures. This aspect is now discussed in the revised manuscript.
"More sepecific comments include:
p. 1, l. 13: provide information on the aquifer, unconsolidated gravel aquifer, …"Answer: We agree and included this information on the aquifer in the corresponding passage of the abstract in our revised manuscript version.
"p. 1, l. 13: specify «limiting power densities» aquifer"Answer: Done. The loss of stored thermal energy due to groundwater flow is an important factor that can reduce the extractable thermal energy and thus the power density of ATES systems. In this case, the maximum achievable power density considers the energy loss is 3.2 W m‑2. We specify this more clearly in the revised manuscript.
"p. 2, l. 53: GWHP was already introduced, check whole text"
Answer: Done. We checked the whole text and now introduce the abbreviation only once in the new manuscript version.
"p. 3, l. 65: “2D numerical box models” include: F. Bottcher, A. Casasso, G. Gotzl, K. Zosseder, TAP - thermal aquifer Potential: a quantitative method to assess the spatial potential for the thermal use of groundwater, Renew. Energy 142 (2019) 85e95."
Answer: Done. The cited paper is now referenced at the corresponding text passage.
"p. 3, l. 76: GHG was already introduced, check whole text"Answer: Done. We checked the whole text and now introduce the abbreviation only once.
"p. 3, l. 94: specify “lower hydraulic conductivity”"
Answer: Done. We now specify this part by adding a reference to Table 1, where hydraulic conductivities are given for the model area.
"Fig. 1a: missing scale bar (maybe zoom in, include Voges); m asl"Answer: Done. We adjusted the figure accordingly.
"Fig. 1b: flow direction of river Dreisam (maybe zoom in, show 10m equidistance); m asl"Answer: Done. We adjusted the figure accordingly.
"Fig. 1c: requires x-scaling, legend for lithologies, turn m asl 180°"Answer: Done. We adjusted the figure accordingly.
"Section 2.2.: Required? No new equations are developed. Reference could be sufficient. "Answer: It is true that no new equations are developed in section 2.2. Nevertheless, we still prefer to state the relevant equations in the text, which is in accordance with other similar studies.
"Table 1: 4.4 E-3; 6 E-5 (reference personal communication?); References for standard values appropriate?"
Answer: Done. We adjusted the numbers accordingly. In our experience, it is reasonable to reference personal communication with authorities such as the Baden-Württemberg State Office for Geology, Raw Materials and Mining (LGRB) or the Environmental Protection Authority of Freiburg. These authorities have profound knowledge of the subsurface and groundwater conditions in the study area.
"Fig. 2 could be merged with Fig. 1"Answer: While it is in principle possible to merge these two figures, we prefer to keep them separate. Fig. 2 shows the delineated hydrogeological regions, which is an important step of the implemented methodology. In contrast, Fig. 1 provides relevant geographical and geological information on the study area. Separating these information also helps to keep a clear structure in the manuscript.
"Table 2: Value representation? Simplify & avoid repetitions."Answer: Done. We simplified Table 2 by removing unnecessary repeated values that are already stated in Table 1.
"p. 7, l. 170: Why estimated?"Answer: The stated groundwater flow velocity of 29.1 m d-1 for region 4 is not measured nor modelled using the city-scale model. It is calculated from the hydraulic conductivity, hydraulic gradient and effective porosity. We adjusted the text to make this clearer.
"p. 7, l. 174.176: What is the difference between uniform and constant thickness?"Answer: There is no difference in this context. We therefore only use the word “uniform” for both cases.
"p. 8, l. 193: Specify “substantial”"
Answer: We chose to not state more specific values here, since the precise extent of the thermal losses is part of the results and thus elaborated on later in the manuscript. Chapter 3.1 “Thermal recovery” provides exact values in more detail.
"Fig. 4: turn labeling right axis asl 180°"Answer: Done. We adjusted the figure accordingly.
"p. 8, l. 211: When is thermal equilibrium reached? This measure maybe is more appropriate compared to the lifetime."
Answer: The time needed to reach thermal equilibrium in the ATES box models strongly depends on the hydrogeological region, the aquifer formation and the ATES well configuration. For this reason, we use a uniform modelling period of 30 years across all box models as well as for the city-scale model. This time period is based on the typical expected ATES systems lifetime. According references are stated in the manuscript.
"p. 10, l. 232-233: Discuss “ambient temperature of 12 °C”"
Answer: Done. In the revised manuscript, we discuss now our assumption of an ambient groundwater temperature of 12 °C in Chapter 3.5 “Limitations of the box model approach”. Please also refer to the more detailed answer on this aspect above.
"p. 10, l. 249: Table 3?"Answer: Corrected. This table should be indeed Table 4.
"p. 12, l.301: It would make sense to also calculate the SCOP as seasonal operation is investigated."
Answer: While the ATES operation is simulated using several seasonal extraction and injection phases over the course of a year, the evaluation of the extracted energy is done for whole seasons in order to determine the thermal recovery. In heating mode, the considered energy used to calculate the power density includes an additional term resulting from heat pump operation with an assumed COP of 3.5. This value of 3.5 is also the average COP during the heating season, i.e. the SCOP.
"p. 13, eq. 14: see Epting et al. 2018"
Answer: We added two references for eq. 14, including also Epting et al. (2018).
"p. 17, l. 421-428: Compare to Epting et al. 2018 & 2020"Answer: We now discuss the power density values and the underlying methodology from Epting et al. (2020) in this text passage. However, we decided not to include the study by Epting et al. (2018) in this part of the discussion, because this specific study deals with a more theoretical thermal potential of shallow geothermal resources. Also, the values in Epting et al. (2018) are not referring to a specific technology, which hinders meaningful comparison to our results.
"Good luck and all the best,
Jannis Epting"
Thank you for your time and feedback on our work!
-
AC1: 'Reply on RC1', Ruben Stemmle, 22 Jun 2023
-
RC2: 'Comment on hess-2023-62', Anonymous Referee #2, 14 May 2023
This manuscript describes a case study in which the shallow geothermal system Aquifer Thermal Energy Storage (ATES) is analyzed for the city of Freiburg in Germany. The work is largely presented as case study, and the question is why a reader in Florence, Oklahoma City, or Nanjing should be interested in reading it. If this manuscript addresses readers who are really interested in the city of Freiburg, the German journal “Grundwasser” would be the right place to publish it, because the latter journal made for a German audience. A paper in HESS, by contrast, should present either new insights in hydrological system behavior, or new methods of general interest that can be applied at many places. I honestly don’t see that the authors have advanced our general understanding of heat transport in the subsurface, but I also have difficulties to recognize methodological breakthroughs. Using simplified box-shaped models with a uniform ambient hydraulic gradient, is not really new, and the concept of the power density has been introduced before. So what is really new and of interest to a broad readership?
I have to admit that my specialty is not in the design of shallow geothermal systems, but I understand the underlying concepts and governing equations quite well. Having said that, I am not convinced about the concept of the power density. The natural geothermal heat flux in the region is about 0.1 W/m2. If you plaster the domain with many geothermal wells, all of them extracting heat, and you insulate the top (as done by the authors), the long-term power density will equal the natural geothermal heat flux of 0.1 W/m2. The trick of ATES is of course that heat is injected into the ground in the summer and extracted in the winter, which makes even the scenario of a plastered domain with an insulated top less dependent on the natural geothermal heat flux. What the authors do, is to consider a rectangular box with a single or two geothermal well pairs or triplets, run the system for 30 years, and then outline the 0.5 K contourline of the temperature anomaly. This design is not representative if you consider a second/third/fourth row of well pairs/triplets that will interfere. This is a simple matter of superposition, which applies to linear transport. A realistic small-scale scenario representative for an entire city would use periodic boundary conditions, in which the heat flux leaving the domain at the downstream end enters at the upstream end. If you run the system until you reach dynamic steady state, you see how the geothermal systems influence each other in the middle of a periodic domain. This would make much more sense then outlining an arbitrarily chosen contourline of a lonesome geothermal system. A systematic analysis of heat fluxes (how much is recycled, what is the difference in the incoming versus the outgoing heat flux with ambient flow if you don’t assume periodicity, what is the heat flux at the non-insulated top boundary) would make more sense to me then the power density, which only gives the illusion of a practical metric.
I could imagine that the work could be converted into a methodological study, outlining how small-scale models can be used to predesign large-scale optimization of geothermal systems. But the manuscript is not written in such a way, and a real large-scale optimal design is not attempted at all.
Notwithstanding these fundamental concerns, I have a series of minor comments:
- Line 34: “thermal energy” rather than “environmental”
- Line 55: “revealed” (past tense)
- Line 89: Shouldn’t it be “alluvial fan” rather than “cone”?
- Lines 92-94: Outside of Baden-Württemberg, nobody knows the Neuenburg and Bresigau formations. Give at least their age (their both from the Pleistocene)
- Equation 3: You use a scalar dispersive thermal conductivity, but later on you use a longitudinal and a transverse dispersivity. This implies applying the Scheidegger parameterization, and it implies anisotropic dispersive heat transfer. The corresponding equations are missing.
- Equation 5, line 131: if p is supposed to be the pore pressure, you need to add gravity to Darcy’s law: u = -K*grad(p/(ρ*g)+z). Alternatively, you can define p as the total potential (pore pressure plus ρ*g*z).
- Line 132: I assume that you neglect the temperature dependence of hydraulic conductivity, which stems from the temperature dependence of viscosity. We are talking about a factor of two for 10 K temperature difference.
- Equation 6: I assume that you use the specific storativity to relate the given time derivative to the rate of change in pore pressure/potential.
- Table 1: It’s “transverse dispersivity” rather than “transversal”, and the coefficient is about two orders of magnitude too large. Stochastic theory reveals that the longitudinal dispersion coefficient (if you believe in Fickian dispersion) increases by orders of magnitude upon upscaling, whereas the transverse dispersion coefficient is hardly affected at all.
- Figure 3 and its discussion: The heat-related boundary conditions are incomplete. What is the bottom boundary? How do you treat temperature at the up- and downstream faces?
- Section 2.6: I have expressed my concerns regarding the power density above. In fact, the results of the city-scale model confirm the importance of the interference across boxes. I am not convinced.
- Lines 348-349: is the observation that ambient flow influences the thermal recovery in ATES systems new at all?
- Figure 7: This plot shows a linear dependence of TR on ln(u). It should be possible to approximate the corresponding coefficients by analytical considerations, but the authors not even do the linear regression.
- Lines 431-432: For the given simple setup with uniform ambient flow I am not convinced that numerical models truly outperform analytical expressions.
- Lines 451-454: I expressed doubts about the thermal insulation on top already further up. In fact, a large fraction of the heat in shallow geothermal systems comes from the surface. It’s still reasonable to prefer these systems over air heat pumps because you change the timing.
- Line 468 ff.: Here the discussion becomes by far too local. I value the cities of Ludwigsburg (nice baroque castle), Lausanne (Olympic museum on Lake Geneva), and Geneva (impressive fountain in the lake), but the authors simply list a couple of case studies. You need to concentrate on general principles and methods instead!
- Section 3.4: I am not convinced about the savings of greenhouse-gas emissions either. This all depends on the specific energy mix, for the existing heating systems, and for electricity production. On the long run, all heating has to be done with renewable energy. In Freiburg, deep geothermal energy would be an option. If you exclude that, you get stuck to heat pumps. The rght comparison in my opinion is to based on the electric-energy demand for heating/cooling with shallow geothermal systems versus air heat pumps. The (hopefully renewably produced) electricity saved is the long-term argument for ATES.
- Section 3.5: I am not convinced that the city-scale model is really computationally too demanding. In this model you can address the interferences of individual systems, and you don’t rely on parallel ambient flow. The goal should be to optimize the big system. If you can show how to do that in general (with the help of box models for predefine or not), you have a HESS-worthy publication.
Citation: https://doi.org/10.5194/hess-2023-62-RC2 -
AC2: 'Reply on RC2', Ruben Stemmle, 23 Jun 2023
Dear colleague,
thank you very much for your thorough feedback on our manuscript. Below, we respond to each of your comments.
"This manuscript describes a case study in which the shallow geothermal system Aquifer Thermal Energy Storage (ATES) is analyzed for the city of Freiburg in Germany. The work is largely presented as case study, and the question is why a reader in Florence, Oklahoma City, or Nanjing should be interested in reading it. If this manuscript addresses readers who are really interested in the city of Freiburg, thöe German journal “Grundwasser” would be the right place to publish it, because the latter journal made for a German audience. A paper in HESS, by contrast, should present either new insights in hydrological system behavior, or new methods of general interest that can be applied at many places. I honestly don’t see that the authors have advanced our general understanding of heat transport in the subsurface, but I also have difficulties to recognize methodological breakthroughs. Using simplified box-shaped models with a uniform ambient hydraulic gradient, is not really new, and the concept of the power density has been introduced before. So what is really new and of interest to a broad readership?"
Answer: While the concept of power density has been introduced before (as stated in our manuscript), our study expands the concept of power densities to determine the technical potential of ATES on a city-scale level for the first time. To that extent, the study’s aim is not to improve the understanding of subsurface heat transport, however to provide a novel methodology, which joins existing approaches into a new and novel framework, and showcasing its applicability on an exemplary study area. While the study area is naturally geographically constrained, the developed methodology can easily be adopted for other cities and regions. Thus, we are convinced that our study will be of interest for many researchers and practitioners alike, as new strategies for urban energy and subsurface planning are urgently needed to increase the share of renewable energy in urban areas.
"I have to admit that my specialty is not in the design of shallow geothermal systems, but I understand the underlying concepts and governing equations quite well. Having said that, I am not convinced about the concept of the power density. The natural geothermal heat flux in the region is about 0.1 W/m2. If you plaster the domain with many geothermal wells, all of them extracting heat, and you insulate the top (as done by the authors), the long-term power density will equal the natural geothermal heat flux of 0.1 W/m2. The trick of ATES is of course that heat is injected into the ground in the summer and extracted in the winter, which makes even the scenario of a plastered domain with an insulated top less dependent on the natural geothermal heat flux."
Answer: We believe there is a misunderstanding regarding the operational principle of ATES, the concept of power density and the energy fluxes in the shallow subsurface.
First, we would like to emphasize that the geological subsurface and the groundwater are the storage medium of an ATES system, rather than the primary source for extracting thermal energy. In contrast to conventional shallow geothermal systems, ATES offers the possibility to store and re-use for example excess waste heat, which makes the amount of storable energy largely independent of the natural heat fluxes (see also response to comment #15). Indeed, we simulate ATES systems with a varying, but seasonally balanced extraction and injection of thermal energy.
Secondly, while the geothermal heat flux is of the same physical unit as the power density (W m‑2), the power density as utilized in our manuscript refers to the energy usable with a specific technology. Since ATES systems use energy from other sources than the subsurface, they can have much higher power densities than the natural heat flux. Assuming a balanced ATES operation, the obtained power density values are also true for long-term system operation, and allow quantitative comparison with other energy sources with respect to space requirements (e.g. Kammen & Sunter, 2016).
"What the authors do, is to consider a rectangular box with a single or two geothermal well pairs or triplets, run the system for 30 years, and then outline the 0.5 K contourline of the temperature anomaly. This design is not representative if you consider a second/third/fourth row of well pairs/triplets that will interfere. This is a simple matter of superposition, which applies to linear transport. A realistic small-scale scenario representative for an entire city would use periodic boundary conditions, in which the heat flux leaving the domain at the downstream end enters at the upstream end. If you run the system until you reach dynamic steady state, you see how the geothermal systems influence each other in the middle of a periodic domain. This would make much more sense then outlining an arbitrarily chosen contourline of a lonesome geothermal system. A systematic analysis of heat fluxes (how much is recycled, what is the difference in the incoming versus the outgoing heat flux with ambient flow if you don’t assume periodicity, what is the heat flux at the non-insulated top boundary) would make more sense to me then the power density, which only gives the illusion of a practical metric."
Answer: First, we would like to clarify that we do not simulate well triplets, but in fact ATES with one, two or three well doublets depending on the system configuration. Thus, interference of these well doublets is an integral part of the operational scheme and modelled numerically. This is now clearly stated in the revised manuscript.
The reviewer raises a valid point in saying that neighboring ATES with multiple well doublets might also interfere. However, German regulations (amongst others) require planning and operation of geothermal systems in a way that avoids thermal interference between individual systems above a certain threshold temperature (e.g. 0.5 K). Thus, in accordance with the aim of the study to provide a planning methodology for ATES systems, we do not allow for such thermal interference in our approach, which is also true for many other international studies in this field (e.g. Gizzi et al., 2020; Piga et al., 2017; Visser et al., 2015). The limitations of our approach in this respect are discussed in the manuscript, including a note that mutual thermal interference might actually have an overall positive effect on system efficiencies on a large scale.
In our opinion, a detailed analysis of the heat fluxes of the box models would not provide added value to the assessment of the ATES potential, for the reasons already mentioned in our previous response.
"I could imagine that the work could be converted into a methodological study, outlining how small-scale models can be used to predesign large-scale optimization of geothermal systems. But the manuscript is not written in such a way, and a real large-scale optimal design is not attempted at all."
Answer: Presenting a novel methodology for city-scale assessment of the technical ATES potential comprises indeed a major part of our study, and our manuscript is structured accordingly. To highlight this issue, we reformulated our objective of this study.
Large-scale optimization is beyond the scope of our study for several reasons. First, to make optimization on this scale meaningful, thermal interference between individual systems would have to be considered, which is currently not realistic (see response above). Also, prioritizing optimal large-scale thermal use over individual system performance would require city-scale energy management above ground, i.e. city-wide energy distribution systems for heat and cold, which are often not feasible in reality. However, our approach can serve as a first step for city-wide planning of ATES, and thus also for optimizing the thermal use of the subsurface in the future. This option is now discussed in the revised manuscript.
Furthermore, a single run of the city-scale model takes about 50 hours (as stated in the manuscript), which hinders the applicability of comprehensive optimization schemes, which often require hundreds of iterative simulation runs. While it might be possible on an HPC cluster, it would significantly impede the applicability of our approach for municipal energy planning.
"Notwithstanding these fundamental concerns, I have a series of minor comments:
1. Line 34: “thermal energy” rather than “environmental”"
Answer: We agree that this expression fits better within the context of the sentence and use the term “thermal energy” in the revised manuscript.
"2. Line 55: “revealed” (past tense)"
Answer: Done. We adjusted the word to past tense in the revised manuscript.
"3. Line 89: Shouldn’t it be “alluvial fan” rather than “cone”?"
Answer: Correct. We now use the more common term ‘alluvial fan’.
"4. Lines 92-94: Outside of Baden-Württemberg, nobody knows the Neuenburg and Bresigau formations. Give at least their age (their both from the Pleistocene)"
Answer: Done. We now state the Pleistocene age of both formations in the new manuscript version besides describing their hydrogeological characteristics relevant to our study.
"5. Equation 3: You use a scalar dispersive thermal conductivity, but later on you use a longitudinal and a transverse dispersivity. This implies applying the Scheidegger parameterization, and it implies anisotropic dispersive heat transfer. The corresponding equations are missing."
Answer: True, we use a longitudinal and a transverse dispersivity and the corresponding equations were missing. Hence, we now state the equations for calculating the disperse thermal conductivity from the longitudinal and the transverse dispersivities in the revised manuscript.
"6. Equation 5, line 131: if p is supposed to be the pore pressure, you need to add gravity to Darcy’s law: u = -K*grad(p/(ρ*g)+z). Alternatively, you can define p as the total potential (pore pressure plus ρ*g*z)."
Answer: Since we do not set topographic elevations in our simplified box models, we do not account for gravitational effects in the box models. Preliminary simulations during development of the box model approach showed that including gravitation in the models had only a very minor and therefore neglectable effect on the thermal plume propagation, which is the relevant aspect for determining power densities.
"7. Line 132: I assume that you neglect the temperature dependence of hydraulic conductivity, which stems from the temperature dependence of viscosity. We are talking about a factor of two for 10 K temperature difference."
Answer: The temperature dependence of viscosity and therefore hydraulic conductivity is indeed neglected in our study. This is a common approach for numerical simulations of heat transport processes in shallow groundwater for low-temperature geothermal systems (e.g. Hecht-Méndez et al., 2010; Beernink et al., 2022; Bidarmaghz et al., 2020; Previati et al., 2022).
The factor between hydraulic conductivities at the maximum (18 °C) and the minimum (6 °C) temperatures in our study is actually around 1.4 (rather than 2). Also, these “extreme” temperatures occur over a very limited spatial area, which makes the effect of varying hydraulic conductivities on the entire plume length negligible.
"8. Equation 6: I assume that you use the specific storativity to relate the given time derivative to the rate of change in pore pressure/potential."
Answer: Our models simulate the thermo-hydraulic behavior of ATES systems in an unconfined aquifer. In this type of aquifer, the specific storage resulting from compressibility of groundwater and aquifer matrix is orders of magnitude smaller than the storage effects from the matrix porosity. We therefore only consider the porosity as a storage parameter in the sense of the specific yield.
"9. Table 1: It’s “transverse dispersivity” rather than “transversal”, and the coefficient is about two orders of magnitude too large. Stochastic theory reveals that the longitudinal dispersion coefficient (if you believe in Fickian dispersion) increases by orders of magnitude upon upscaling, whereas the transverse dispersion coefficient is hardly affected at all."
Answer: Done. We changed the wording mistake and use the correct term “transverse dispersivity” now in Table 1.
Previous studies by Zech et al. (2015, 2018) suggest that the commonly used scale-dependency of the longitudinal dispersivity is not valid. The dependence of longitudinal and transverse dispersivity on the length scale of the transport process is however a field of active and current research (e.g. Di Dato et al., 2022; Park & Lee, 2021; Pophillat et al., 2020; Younes et al., 2020). Furthermore, only few reported values for thermal dispersivities values obtained by field experiments exist (Stauffer et al. 2014). Hence, for the current study with the focus on the development of a novel methodology for city-scale assessment of the technical ATES potential, we assume commonly used thermal dispersivities values reported in the literature (e.g. Stauffer et al., 2014).
"10. Figure 3 and its discussion: The heat-related boundary conditions are incomplete. What is the bottom boundary? How do you treat temperature at the up- and downstream faces?"
Answer: We agree and added the missing details on the heat-related boundary conditions (BC) and include an adjusted figure in the revised manuscript. Like for the top boundary, we implemented a no heat flux BC at the model bottom. This is also true for the upstream and downstream model faces, which is line with the box model approach of preventing any thermal interferences between individual ATES systems.
"11. Section 2.6: I have expressed my concerns regarding the power density above. In fact, the results of the city-scale model confirm the importance of the interference across boxes. I am not convinced."
Answer: We already addressed the reviewer’s concerns regarding the power density above. The conservative way of delineating the spatial extent of the thermal impact of individual ATES systems prevents any significant thermal interferences between ATES systems. This is also the common approach for planning, regulatory approval and operation of such systems. Accordingly, interference is not accounted for in the box models.
"12. Lines 348-349: is the observation that ambient flow influences the thermal recovery in ATES systems new at all?"
Answer: The influence of the ambient groundwater flow on the thermal recovery, i.e. storage efficiency of ATES systems has indeed been studied previously as stated in Section 2.4.2. We here discuss these effects, because they impact the choice of optimal ATES design, which is a new observation, and because the thermal recovery directly influences the power density, whose estimation is the key aspect of this study.
"13. Figure 7: This plot shows a linear dependence of TR on ln(u). It should be possible to approximate the corresponding coefficients by analytical considerations, but the authors not even do the linear regression."
Answer: While the idea of inferring a dependence of the thermal recovery on the groundwater flow velocity is interesting, this is beyond the scope of this study.
In our proposed methodological workflow, the thermal recovery is a byproduct from the numerical modeling results, which is worth discussing as it represents a main efficiency criterion for ATES. Yet, inferring a robust statistical relationship between the two factors would require a very comprehensive modelling exercise, without adding significant value to the presented approach and objective of this study.
"14. Lines 431-432: For the given simple setup with uniform ambient flow I am not convinced that numerical models truly outperform analytical expressions."
Answer: We are not aware of any analytical solution that offers equivalent possibilities as our numerical setup. Using numerical models for simulating ATES systems with numerous extraction and injection wells operated under groundwater flow conditions and according to seasonally varying pumping schemes with reversing pumping directions and passive storage phases is considered state of the art (e.g. De Paoli, 2023, Duijff et al., 2021, Regnier et al., 2022).
"15.Lines 451-454: I expressed doubts about the thermal insulation on top already further up. In fact, a large fraction of the heat in shallow geothermal systems comes from the surface. It’s still reasonable to prefer these systems over air heat pumps because you change the timing."
Answer: As explained in a previous response, the extracted (or recovered) heat in an ATES does not stem from the ambient (sub)surface, but mainly from waste heat from e.g. space cooling or industrial cooling processes and excess heat from CHP plants. ATES systems are therefore not dependent on the heat input (or recovery) from the surface, however rather benefit from the insulating effect of the subsurface.
"16. Line 468 ff.: Here the discussion becomes by far too local. I value the cities of Ludwigsburg (nice baroque castle), Lausanne (Olympic museum on Lake Geneva), and Geneva (impressive fountain in the lake), but the authors simply list a couple of case studies. You need to concentrate on general principles and methods instead!"
Answer: We agree and adjusted the corresponding part in Section 3.3, where we compare our supply rates with previous publications. The revised manuscript presents a reduced text passage focusing on relevant similarities and differences among the studies. The discussion of specific geographical references was removed.
"17. Section 3.4: I am not convinced about the savings of greenhouse-gas emissions either. This all depends on the specific energy mix, for the existing heating systems, and for electricity production. On the long run, all heating has to be done with renewable energy. In Freiburg, deep geothermal energy would be an option. If you exclude that, you get stuck to heat pumps. The rght comparison in my opinion is to based on the electric-energy demand for heating/cooling with shallow geothermal systems versus air heat pumps. The (hopefully renewably produced) electricity saved is the long-term argument for ATES."
Answer: We agree that the savings of GHG emissions are governed by the specific heat and electricity mix. Hence, we use the latest available data on the city’s energy mix for space heating from 2020 (see section 3.4). Emissions from ATES systems are adopted from Stemmle et al. (2021), and account for the German electricity mix as of 2014.
We also agree that different options for renewable heat (and cold) supply have to be considered in future in order to fulfill different needs, such as deep (hydrothermal) geothermal systems for high-temperature heat, which are indeed an option in the city of Freiburg. Given the diversity of the current (and presumably future) heat mix a comparison solely against air-source heat pumps (ASHP), which currently make up less than 1% of Freiburg’s heat mix, does not appear meaningful to us. However, we added a short discussion comparing the energy demands of ASHP and ATES systems to the revised manuscript.
"18. Section 3.5: I am not convinced that the city-scale model is really computationally too demanding. In this model you can address the interferences of individual systems, and you don’t rely on parallel ambient flow. The goal should be to optimize the big system. If you can show how to do that in general (with the help of box models for predefine or not), you have a HESS-worthy publication."
Answer: Please see our response on the major comment regarding large-scale optimization above.
References:
Beernink, S., Bloemendal, M., Kleinlugtenbelt, R. and Hartog, N. (2022)Maximizing the use of aquifer thermal energy storage systems in urban areas: effects on individual system primary energy use and overall GHG emissions. Applied Energy, 311, 118587.
Bidarmaghz, A., Choudhary, R., Soga, K., Terrington, R.L., Kessler, H. and Thorpe, S. (2020) Large-scale urban underground hydro-thermal modelling – A case study of the Royal Borough of Kensington and Chelsea, London. Science of The Total Environment, 700, 134955.
De Paoli, C., Duren, T., Petitcler, E., Agniel. M. and Dassargues, A. (2023) Modelling Interactions between Three Aquifer Thermal Energy Storage (ATES) Systems in Brussels (Belgium). Applied Sciences, 13(5), 2934.
Di Dato, M., D’Angelo, C., Casasso, A. and Zarlenga, A. (2022) The impact of porous medium heterogeneity on thermal feedback of open-loop shallow geothermal systems. Journal of Hydrology, 604, 127205.
Duijff, R., Bloemendal, M. and Bakker, M. (2021) Interaction Effects Between Aquifer Thermal Energy Storage Systems. Ground water, 61(2), 173-182.
Gizzi, M., Taddia, G., Cerino Abdin, E. and Lo Russo, S. (2020) Thermally Affected Zone (TAZ) Assessment in Open-Loop Low-Enthalpy Groundwater Heat Pump Systems (GWHPs): Potential of Analytical Solutions. Geofluids, 2020, 1-13.
Hecht-Méndez, J., Molina-Giraldo, N., Blum, P. and Bayer, P. (2010) Evaluating MT3DMS for heat transport simulation of closed shallow geothermal systems. Ground Water, 48(5), 741–756.
Kammen, D.M. and Sunter, D.A. (2016) City-integrated renewable energy for urban sustainability. Science, 352, 922-928.
Mueller, M.H., Huggenberger, P. and Epting, J. (2018) Combining monitoring and modelling tools as a basis for city-scale concepts for a sustainable thermal management of urban groundwater resources. Science of the Total Environment, 627, 1121-1136.
Park, B.-H. and Lee, K.-K. (2021) Evaluating anisotropy ratio of thermal dispersivity affecting geometry of plumes generated by aquifer thermal use. Journal of Hydrology, 602, 126740.
Piga, B., Casasso, A., Pace, F., Godio, A. and Sethi, R. (2017) Thermal Impact Assessment of Groundwater Heat Pumps (GWHPs): Rigorous vs. Simplified Models. Energies, 10 (9), 1385.
Pophillat, W., Bayer, P., Teyssier, E., Blum, P. and Attard, G. (2020) Impact of groundwater heat pump systems on subsurface temperature under variable advection, conduction and dispersion. Geothermics, 83, 101721.
Previati, A., Epting, J. and Crosta, G.B. (2022) The subsurface urban heat island in Milan (Italy) - A modeling approach covering present and future thermal effects on groundwater regimes. Science of The Total Environment, 810, 152119.
Regnier, G., Salinas, P., Jacquemyn, C. and Jackson, M.D. (2022) Numerical simulation of aquifer thermal energy storage using surface-based geological modelling and dynamic mesh optimisation. Hydrogeology Journal, 30, 1179-1198.
Rivera, J.A., Blum, P. and Bayer, P. (2017) Increased ground temperatures in urban areas: Estimation of the technical geothermal potential. Renewable Energy, 103, 388-400.
Stauffer, F., Bayer, P., Blum, P., Molina-Giraldo, N. and Kinzelbach W. (2013) Thermal Use of Shallow Groundwater. 287 pages, CRC Press.
Visser, P.W., Kooi, H. and Stuyfzand, P.J. (2015) The thermal impact of aquifer thermal energy storage (ATES) systems: a case study in the Netherlands, combining monitoring and modelling. Hydrogeology Journal, 13, 507-532.
Younes, A., Fahs, M., Ataie-Asthiani, B. and Simmons, C.-T. (2020) Effect of distance-dependent dispersivity on density-driven flow in porous media. Journal of Hydrology, 589, 125204.
Zech, A., Attinger, S., Cvetkovic, V. et al. (2015) Is unique scaling of aquifer macrodispersivity supported by field data? Water Resources Research, 51(9), 7662-7679.
Zech, A., Attinger, S., Alberto, B. et al. (2019) A Critical Analysis of Transverse Dispersivity Field Data. Ground water, 54(4), 632-639.
Zhu, K., Blum, P., Ferguson, G., Balke, K.D. and Bayer, P. (2010) The geothermal potential of urban heat islands. Environmental Research Letters, 5, 044002.
Citation: https://doi.org/10.5194/hess-2023-62-AC2
Ruben Stemmle et al.
Ruben Stemmle et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 508 | 144 | 23 | 675 | 55 | 9 | 7 |
- HTML: 508
- PDF: 144
- XML: 23
- Total: 675
- Supplement: 55
- BibTeX: 9
- EndNote: 7
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1