The Water Nexus research programme is on the lookout for alternative water sources to achieve water self-sufficiency regions. The motto of the programme is ‘salt where possible, fresh where necessary’. But which water source do you opt for and how do you get the water to users?
Zeeuws-Vlaanderen is not self-sufficient when it comes to water. The groundwater is mostly brackish and the North Sea and Westerschelde are salt water bodies, so fresh water is now pumped to the area from the Biesbosch over a distance of 120 km to meet water requirements. Zeeuws-Vlaanderen is reliant on this external water supply. In times of drought, this water source (i.e. the reservoirs in the Biesbosch) may find themselves under pressure, which can lead to water shortages for residents, industry and/or agriculture. There are several alternative and renewable water sources available locally, including fresh, brackish and saline groundwater, rainwater and treated waste water. Using these water sources could reduce dependency on the Biesbosch and make the water supply future-proof.
To use these alternative water sources, a switch to a decentralised water supply is an option – this would require large investments in new infrastructure, a complex design and complex implementation. Alternative sources must be selected, as well as the most cost-efficient way to move water to where it needs to be used.
The aim of the Water Nexus research programme is to investigate how agro-industrial regions can become self-sufficient in water, with salt water and waste water seen as part of the solution, rather than a problem. A new tool, WaterROUTE, is helping to design new ways of supplying water and thus support decision-making. The starting point is to develop decentralised water systems that are in balance with local ecosystems. WaterROUTE visualises water availability and regional water demand with hydrological models, combined with methods of GIS and mathematical programming.
Sources in focus
The WaterROUTE tool has been tested on a case study in Zeeuws-Vlaanderen. The first step is to identify the alternative water sources with the maximum renewable water availability for each source (see Figure 1). There are 25 locations with availability of fresh groundwater and several towns and villages with rainwater and waste water.
The availability of fresh groundwater is mapped with a 3D groundwater model linked to a salt transport model (Van Baaren et al., 2016, Willet et al., 2020). Potential fresh groundwater wells are mapped with an upper limit for salinity of 1500 mg chloride per litre. Extraction from the wells must not affect local ecosystems through salinisation or a severe drop in groundwater. We chose a maximum drop of 50 mm as the input parameter for the model (Oude Essink and Pauw, 2018). Adjustment is possible if stricter requirements prove necessary.
A total of 6.1 million m3 per year of renewable fresh to brackish groundwater is available.
In cities and villages, 4.6 million m3 of rainwater and 4.2 million m3 of waste water could, theoretically, be collected. This is based on precipitation data from the dry year 2018, rainwater harvesting from all roofs with an efficiency of 85 per cent and decentralised waste water treatment with a volume based on daily water consumption per inhabitant (118 litres).
Figure 1: Alternative water sources in Zeeuws-Vlaanderen and the most cost-efficient pipelines between them
WaterROUTE then maps all possible connections between water users and water sources based on the lowest relative cost of constructing pipelines (Figure 1). Each land use (cities, natural areas and agricultural land) has different construction costs – the cost of building pipelines along existing infrastructure such as roads is, for example, lower than through cities or natural areas. The relative costs were determined in consultation with experts in the field of water distribution networks. The relative costs for different types of land use are then converted to a cost grid with GIS. The potential connections with the lowest cost are generated using a ‘least cost path algorithm’.
From water demand to a decentralised supply network
As a final step, WaterROUTE takes water demand (quantity and quality) to calculate the pipeline network that will meet that demand, based on cost minimisation. The result is an overview of the most suitable water sources.
In the optimisation, the model determines which pipelines should be used and how much water should flow through each pipe, with a piece-wise linear cost function minimised for each pipe. The cost function is based on the cost of building new pipelines and pumping costs.
For the Zeeland-Vlaanderen case study, we assumed an industrial water user in the Terneuzen region. In this article, we show the results for an annual water demand of 4 million m3. We have run the model for three scenarios: (1) Groundwater only, (2) Groundwater and rainwater and (3) Groundwater, rainwater and decentralised treated waste water. The current situation, with supply from the Biesbosch, is not included.
More sources, smaller network
Figure 2 shows the different networks required to move the requisite amount of water to the water user. With more alternative water sources, the requisite water network would become significantly smaller:
- Scenario 1 – 100% groundwater: 63.7 kilometres of pipeline;
- Scenario 2 – 54% groundwater and 46% rainwater: 37.1 kilometres of pipeline;
- Scenario 3 – 34% groundwater, 30% rainwater and 36% treated waste water: 23.2 kilometres of pipeline.
Compared to scenario 1, costs in scenario 2 would fall by a quarter and in scenario 3 by half. In our example, the operational costs for the pipelines are based on the energy costs for pumping (€0.2/kWh) and the investment costs depend on the required pipe diameter and the total length of the pipelines. In general, a minimum flow velocity of 0.4 m/s and a maximum flow velocity of 1.5 m/s have been assumed. (Mesman and Meerkerk, 2009). For WaterROUTE, the upper limit of 1.5 m/s is significant because after that point, pumping costs increase dramatically. The key figures that we have used for the construction of new pipelines are €50/m per 100 mm diameter – so a 300 mm pipe would cost €150/m (Willet et al., 2021).
When extracting groundwater, the operational costs concern the energy to pump the water from under the ground (€0.2/kWh) and the investment costs for drilling the wells. For each location, we have assumed a cluster of small-scale wells, where the costs depend on the depth of each well (we calculate €50/m).
In order to have the collected rainwater available all year round, we have assumed storage in open basins that are 3 metres deep. The investment costs for this comprise the costs for the earthworks and the construction of the waterproofing membrane (initial investment costs of €15,000 per basin, with costs increasing by €2.40 per m3 of realised capacity). In addition to the earthworks, we have also calculated the purchase costs for the (agricultural) land (€75,000/ha).
For waste water, we have assumed that it is properly treated for reuse at the water treatment plant and that no additional costs are included.
In Zeeuws-Vlaanderen, fresh groundwater is scarce while brackish groundwater is abundant. Consequently, we have also looked into the use of brackish groundwater. In this scenario, the WaterROUTE model calculates the most suitable sources based on the maximum permissible salt concentration (customised water). We ran the model for moderately brackish water with maximum concentrations at the end user of 375, 400, and 425 mg chloride per litre (Willet et al, 2021). The cost of the network has been shown to decrease significantly as saltier water is mixed with fresher water closer to the water user.
The model can also be used to clarify which is cheaper: saltier water with low transport costs but high desalination costs, or fresh water with high transport costs but low desalination costs. In areas where the quality of water sources is changing due to salinisation, this functionality can be used to design water systems to be robust in the long term.
Figure 2. The cheapest WaterROUTE network and the most suitable alternative sources for an industrial water user in Terneuzen (4 million m3). Scenario 1 (top) uses only groundwater, scenario 2 (bottom left) uses groundwater and rainwater and scenario 3 (bottom right) uses groundwater, rainwater and decentralised treated waste water. The current situation is not included.
Decentralised water systems with alternative water sources can offer a solution to meet the water demand of all water users in a region, and make the water supply climate-proof. The WaterROUTE tool enables research into decentralised systems and can thus facilitate design and decision-making by comparing different options (including the current one). The model can be applied to any regional water system. However, it is essential to know what alternative sources are available, what sustainability criteria apply and what costs need to be taken into account. The model can include the costs relevant to the user and can be expanded to include, inter alia, the costs of waste water treatment or underground storage.
Locally available alternative water sources can contribute to a climate-resilient water supply. This requires a switch to a decentralised water supply, but this need not be something that is more expensive. Both the design and decision-making are, however, complex and it is difficult to properly evaluate different scenarios. With the new WaterROUTE tool, we have compared the availability of local groundwater, rainwater and waste water for Zeeuws-Vlaanderen with the water consumption of a large industrial water user. By using more alternative water sources, the required pipeline network becomes smaller, with lower costs overall.
Mesman, G.A.M., Meerkerk, M.A., 2009. Evaluatie ontwerprichtlijnen voor distributienetten / Evaluation of draft guidelines for distribution networks KWR, 48 pp. http://api.kwrwater.nl//uploads/2017/10/KWR-09.073-Evaluatie-ontwerprichtlijnen-voor-distributienetten-vertakte-netten.pdf.
Oude Essink, G.H.P., Pauw, P.S., 2018. Evaluation and in-depth study of groundwater extraction rules in the province of Zeeland / Evaluation and in-depth study of groundwater extraction rules in the province of Zeeland, 170 pp. https://publicwiki.deltares.nl/display/ZOETZOUT/Modelstudies (accessed 16/01/2020).
Van Baaren, et al. 2016. Verzoeting en verzilting van het grondwater in de Provincie Zeeland, Regionaal 3D model voor zoet-zout grondwater, Deltares report 1220185 / Salinisation and groundwater in the Province of Zeeland, Regional 3D model for fresh-saline groundwater, Deltares report 1220185, 86 pp. https://publicwiki.deltares.nl/download/attachments/55640066/1220185-000-BGS-0003-r-Verzoeting%20en%20verzilting%20freatisch%20grondwater%20in%20de%20Provincie%20Zeeland-def.pdf?version=1&modificationDate=1490345125944&api=v2 (accessed 27/05/2021).
Willet, J. et al. 2020. Water supply network model for sustainable industrial resource use a case study of Zeeuws-Vlaanderen in the Netherlands. Water Resources and Industry 24, 100131.
Willet, J. et al., 2021. WaterROUTE: a model for cost optimization of industrial water supply networks when using water resources with varying salinity. Water research, 117390.