Towards Circular Water Neighborhoods: Simulation-Based Decision Support for Integrated Decentralized Urban Water Systems
, Diederik van Duuren 2,3, Henk-Jan van Alphen 1, Jos Frijns 1, Dionysios Nikolopoulos 4 and Christos Makropoulos 1,4Round 1
Reviewer 1 Report
The reviewed article describes current and important issues related to the well-thought-out management of water distribution systems. This study presents advanced simulation-based framework for the quantitative performance assessment of decentralised systems at a neighbourhood scale, where different technologies can be linked together to provide beneficial effects across multiple urban water cycle domains. A very interesting and valuable aspect of the described simulation is the ability to assess how climate change can affect the water distribution system. The work undoubtedly qualifies for printing because it brings a lot of valuable and interesting information. The article is also very well developed in terms of graphics.
However, the text of the article notices minor editing errors and some issues that require additional explanation.
1. Small editing errors should be removed:
- unnecessary bullets in the lines: 193 and 194,
- unnecessary number 8 in the line 290,
- incorrect numbering in the lines: 384, 391, 395 – should probably be 1, 2, 3,
- in Table 4 in the row regarding RWH buffers, the Storage value should probably amount
to 300 m3, because it is the sum of 60 + 190 + 50,
- the text in line 488 is probably sub-chapter 3.5,
2. Table 4 lacks information about the depth of the rainwater tank (Water square). It is possible to mark the rainwater tank in the drawing above table 2.
3. Does the SUPERLOCAL system predict water intake for watering green areas? In simulation, water for this purpose should also be provided.
4. In a way, it is in figure 6, but it would be worth to assess the value or percentage of the results for SUPERLOCAL during the wet, normal and dry year. This applies to the annual amount of precipitation.
5. In the final part of the work, there is no discussion, i.e. references of results obtained to other such studies.
6. Did the simulation show some weaknesses in the circular water neighborhood of SUPERLOCAL?
In general, I rate the article positively and recommend it for publication in Water
Author Response
The reviewed article describes current and important issues related to the well-thought-out management of water distribution systems. This study presents advanced simulation-based framework for the quantitative performance assessment of decentralised systems at a neighbourhood scale, where different technologies can be linked together to provide beneficial effects across multiple urban water cycle domains. A very interesting and valuable aspect of the described simulation is the ability to assess how climate change can affect the water distribution system. The work undoubtedly qualifies for printing because it brings a lot of valuable and interesting information. The article is also very well developed in terms of graphics.
However, the text of the article notices minor editing errors and some issues that require additional explanation.
Authors’ response: We would like to thank the reviewer for his/her positive feedback and his/her constructive comments. The minor issues mentioned are addressed in the points seen below.
1. Small editing errors should be removed:
- unnecessary bullets in the lines: 193 and 194,
- unnecessary number 8 in the line 290,
- incorrect numbering in the lines: 384, 391, 395 – should probably be 1, 2, 3,
- in Table 4 in the row regarding RWH buffers, the Storage value should probably amount
to 300 m3, because it is the sum of 60 + 190 + 50,
- the text in line 488 is probably sub-chapter 3.5,
2. Table 4 lacks information about the depth of the rainwater tank (Water square). It is possible to mark the rainwater tank in the drawing above table 2.
Authors’ response: The aforementioned editing errors have been fixed by changing the relevant text styles. Table 4 has been updated with the right value in the RWH buffers, while the depth of the rainwater tank has been added to it as well.
3. Does the SUPERLOCAL system predict water intake for watering green areas? In simulation, water for this purpose should also be provided.
Authors’ response: Garden water uses are included in SUPERLOCAL as part of the demands, which can be also seen in Figure 3 (i.e. with the bottom garden hose icon in the lower middle of the graph, where the household demands are described). This water is part of the total household demands and covered by either harvested RW in the first design option or by the mixed RWH-GWR water buffer in the second design option. These garden demands are modelled with a distinct water component in UWOT (named ‘OU’, an abbreviation for outside uses), which can be seen in the upper and middle left of Figure 4.
4. In a way, it is in figure 6, but it would be worth to assess the value or percentage of the results for SUPERLOCAL during the wet, normal and dry year. This applies to the annual amount of precipitation.
Authors’ response: The reviewer raises an interesting point on how results such as runoff reduction vary over dry and wet years. This can be implicitly seen in Figure 6 but was not analysed in the text in order to conserve space and focus on higher-resolution (i.e. event-scale) runoff reduction instead. Still, this variability is interesting to mention; to address this point, the analysis part of lines 471-474 has been reworked to also discuss the influence of annual dryness/wetness on the results.
5. In the final part of the work, there is no discussion, i.e. references of results obtained to other such studies.
Authors response: Wherever possible, a comparison of SUPERLOCAL modeling results with other reference studies or datasets is done while the results are initially presented in Chapter 3. For instance, modelled water demands are compared to an external data report (reference number [45]) in lines 419-422. A model-based comparison of annual runoff is given in lines 429-446. Likewise, the reliability levels that are presented are contrasted against other studies (reference number [50]) in lines 482-483. However, the authors agree that these comparisons need to be summarized in the final part of the work as well. To address this point, a new section has been added in lines 653-659, where the results seen in this study are discussed against a number of other studies on RWH and GWR systems. To provide more insight against other studies, a number of new references have been added as well [58-61].
6. Did the simulation show some weaknesses in the circular water neighborhood of SUPERLOCAL?
Authors response: Each one of the design options in SUPERLOCAL has certain weaknesses which are evident through the simulation model results. The results have shown a trade-off between system autonomy and rainwater retention/runoff reduction that is evident when the two solutions (SCEN_A and SCEN_B) are compared (e.g. through Figure 6): SCEN_A offers a system that is weaker in terms of WW reduction and autonomy from central services, while SCEN_B offers a system that is less efficient in runoff reduction and water storage. This is also discussed in lines 661-671 of the final chapter.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The paper is an interesting exercise on an issue that, day after day, assumes growing importance in many areas of the planet.
The main comments refer to the basic criteria or conditions assumed in the application of the model. For example, in line 186 a reference is made to nutrient recovery, which is considered correct, since the future sustainability of the planet will be based on the water-energy-food triangle. In this context, phosphorus is the main resource to be considered and there are already several publications on the possibility of urine recovery at the building level. However, it is noted that this aspect is no longer referred to in the text or integrated in the model.
As far as efficient products are concerned, the taps were not considered. Why?
The choice of vacuum toilets is also very controversial for the residential sector and should be duly justified. Generally, its application is only interesting for toilet-intensive locations. In the residential sector, maintenance costs, energy costs and, sometimes, noise problems, make it rarely an option. Furthermore, when using solutions such as rainwater harvesting or greywater reuse, the water balance in the building does not justify this option because the available volumes of recovered water are perfectly compatible with the use of efficient toilets (4 or 5 litres, for example).
Author Response
The paper is an interesting exercise on an issue that, day after day, assumes growing importance in many areas of the planet.
The main comments refer to the basic criteria or conditions assumed in the application of the model. For example, in line 186 a reference is made to nutrient recovery, which is considered correct, since the future sustainability of the planet will be based on the water-energy-food triangle. In this context, phosphorus is the main resource to be considered and there are already several publications on the possibility of urine recovery at the building level. However, it is noted that this aspect is no longer referred to in the text or integrated in the model.
Authors’ response: Resource recovery (phosphorus, nitrate etc.) is a key characteristic of the SUPERLOCAL project. However, at the moment not enough data was available from the SUPERLOCAL pilot to provide a solid basis for accurately modelling these material flows. Moreover, as discussed in lines 607-615, UWOT currently limits simulation to water flows and not nutrient transport. The combination of low pilot data availability on nutrient recovery and modeling limitations meant that this particular aspect, although important, was omitted from the analysis. The authors agree that this parameter is important for sustainability and aim to include it in future model iterations, both in the simulation and in the corresponding KPIs (see also the Discussion section, lines 607-630). This inclusion will be further facilitated by the operational data that will progressively become available from the SUPERLOCAL pilot at later phases. The authors also acknowledge that the omission of nutrient recovery from the analysis needs to be explained better; lines 376-379 in Section 3.1. have been added, explaining model limitations. Moreover, lines 612-613 have been added to reinforce discussion on model improvements with regards to nutrient recovery.
The choice of vacuum toilets is also very controversial for the residential sector and should be duly justified. Generally, its application is only interesting for toilet-intensive locations. In the residential sector, maintenance costs, energy costs and, sometimes, noise problems, make it rarely an option. Furthermore, when using solutions such as rainwater harvesting or greywater reuse, the water balance in the building does not justify this option because the available volumes of recovered water are perfectly compatible with the use of efficient toilets (4 or 5 litres, for example).
Authors’ response: The reviewer at this point remarks on the technological mixture of SUPERLOCAL that is selected by its stakeholders, focusing on the choice of vacuum toilets over other solutions. The exact mixture of technologies, which includes vacuum toilets, was decided and finalised by the four project partners, as mentioned in Table 1. This decision is arguably beyond the scope of this study, which has the aim to describe the simulation and KPI framework of the occurring combined RWH-GWR system. These technological choices stem from multiple reasons that have to do with the project aims of SUPERLOCAL, as well as local conditions and market dynamics.
The vacuum toilets have been mainly chosen for resource recovery, in order to implement the ‘new sanitation’ concept. When using only 1 litre per flush the density (biomass) of the black water (BW) waste stream can be converted in different products using a digesting installation (mesophilic UASB) that is sketched in Figure 3. As a (positive) side-effect the water demand for the toilets is very low; nonetheless, the selection criteria for these toilets was not the low amount of water used. To make this vacuum system more economic, viable food grinders are connected as well. This is not allowed in a regular sewage system in the Netherlands (as conventional sewage is not designed for slurry with high organic loads), however with a vacuum system it becomes feasible as an independent BW stream that can be treated locally. Indeed noise problems might occur, however, several specific adjustments have been made in the houses and recently two new silent vacuum toilets (new technologies) entered the market and they look promising; these will be also tested in SUPERLOCAL.
Concerning the remarks on water balance, the stakeholders believe that using toilets that use a significant larger amount of water (4 or 5 litres instead of 1) has a negative impact on the local availability of water. Part of the project contains apartment buildings, meaning a limited roof area can be used for harvesting rainwater with multiple families living below this roof. Hence, for the location and household distribution of SUPERLOCAL, more drastic water saving devices (i.e. the choice of vacuum toilets) can be justified due to the site specific ratio of surface available for rainwater harvesting per person. But, as mentioned above, the main reason is resource recovery, and diluting the slurry with the minimal possible amount of water is a prerequisite for efficient resource recovery in the BW stream.
As this paper focuses on the simulation framework and the corresponding KPIs and not the pilot itself, the authors believe that a substantially more extensive justification than the one already provided in the paper - and in the aforementioned response - on the choice of technologies and vacuum toilets falls out of scope of this study. Minor additions to clarify the choice of vacuum toilets have been made; lines 191-192 have been added in Section 2.2, in order to link the choice of vacuum toilets with the goals on nutrient recovery.
The authors would like to conclude this response by noting that, beyond explaining why vacuum toilets were selected, it is important to consider that the proposed quantitative UWOT-KPI simulation framework is able to model multiple different technologies, and non-vacuum toilets can be readily modelled as well. In fact, the model architecture allows for any toilet with a specific demand (in L/use) to be readily included in the scenarios. As such, it is not dependent on that particular technological choice.
Author Response File: Author Response.docx
