Decentral recovery of water and nutrients from wastewater: A case study in Flanders

Leestijd: 8 minuten

Decentral recovery of water and nutrients from wastewater: A case study in Flanders



Abstract: Greywater upgrading can be a valuable source of reusable water when other water sources, such as rainwater, are unavailable or scarce. Moreover, source separation at the decentral level allows for local recovery and reuse of nutrients such as phosphorus and nitrogen. De Kruitfabriek (the Gunpowder Factory) is an event location in a former industrial building in Vilvoorde, Belgium, where a research and demonstration project is running for decentral treatment of wastewater and recovery of resources. Wastewater is separated into greywater, urine and blackwater. The greywater is treated by a constructed wetland and nanofiltration to produce reusable water. The urine is treated by a struvite precipitation reactor to recover phosphorus and nitrogen as fertilizer. Unique about this project is that the recovered resources are applied and reused locally, to support a short-loop circular economy.

Keywords: Decentralized greywater reuse; Nanofiltration; Resource recovery; Source separation; Struvite



In Flanders, as in many urbanized and industrialized regions, the water cycle is under pressure, resulting in periodical droughts and floods. There is a need for technological measures that can be taken on a decentral level. Buffering of rainwater for low-grade use (e.g., washing machines, toilet flushing) is already common practice. In densely urbanized regions, however, rainwater availability is limited. Greywater upgrading can be a valuable source of reusable water in cases where other water sources are unavailable or scarce.

Municipal wastewater can be source-separated into different fractions: greywater (from showers, washing machines, dishwashers, sinks), and blackwater (from toilets), which can be further separated into brown water (faeces) and yellow water (urine) and yellow water (urine). Source separation of wastewater allows a more efficient recovery of resources, because greywater comprises 70% of the water volume and nearly all of the recoverable heat are present in the greywater. Macronutrients mostly end up in urine (80% of N, 53% of P, 70% of K), and faeces contain 47.5% chemical energy1,2.

Phosphorus is a crucial nutrient for the fertilization of crops, but natural reserves are limited, and the ores geographically restricted. When wastewater is separated at the source, the mineral struvite (NHMgPO·6HO) can be precipitated from the urine fraction before subsequent treatment. Struvite precipitation from urine has been investigated and demonstrated in practice3,4, but needs to overcome a number of challenges, both technologically and in terms of marketability.

Source separation at the decentral level allows for local recovery and reuse of nutrients such as phosphorus and nitrogen. It is not always possible, however, to implement a successful decentral resource recovery project. Much depends on the specific situation, and careful consideration is needed of the constraints in space, economic balance, technology readiness, product quality and willingness of the end user to cooperate. Here, De Kruitfabriek is presented in detail as a specific case study on decentral recovery of water and nutrients in an urban setting.


Case study: De Kruitfabriek

De Kruitfabriek (the Gunpowder Factory) is an event and business location in a former industrial building in Vilvoorde, Belgium. The site functions as a living lab where innovative concepts for urban housing projects can be tested. Four partners (city developer Matexi, sewage treatment utility Aquafin, technology provider NuReSys, and the city of Vilvoorde) are currently running a research and demonstration project for decentral treatment of wastewater and recovery of resources. Wastewater is separated at the source into greywater, male urine from the urinals, and blackwater. Greywater is treated in a constructed wetland and a nanofiltration unit, to remove bacteria, viruses, organic pollution, and divalent ions. Urine is treated in a struvite precipitation reactor (Figure 1). This project is unique in the sense that the recovered water and nutrients are re-applied locally, to create an effective short-loop circular economy.

Figure 1. Schematic representation of the water and nutrient cycle at De Kruitfabriek.

This living lab focuses on closing the water and nutrient cycle on a local scale, with special attention to dissemination and education: the +40.000 annual visitors experience from up close how their wastes are transformed into resources, which they are invited to reuse themselves. Figure 2 shows the nanofiltration pilot in the dedicated filtration room, during construction.

Figure 2: Nanofiltration pilot (right) and other equipment in the filtration room at De Kruitfabriek.

Operational experiences in De Kruitfabriek

The water and nutrient installations at De Kruitfabriek were brought into operation in the first half of 2019. The constructed wetland and nanofilter installations for greywater recycling have been fully operational since September 2019. In the past six months (September 2019 – February 2020), an average of 430 L/d of greywater was produced and treated. Of this treated greywater, an average of 113 L/d (26%) was effectively reused for flushing the toilets, cleaning, irrigating garden patches, and in the washbasin of the bicycle workshop. The unused treated greywater (74%) consisted of constructed wetland effluent overflow and nanofilter permeate overflow. The relatively high fraction of overflow to the sewer was due to overproduction of treated greywater (i.e., a higher supply than demand). Nonetheless, the nanofilter was operated discontinuously. This meant that, on certain occasions, the constructed wetland effluent buffer was overflowing to the sewer, while the nanofilter permeate buffer was empty, and the toilets were flushed with tap water. This illustrates the intrinsic difficulty in matching supply and demand, and the need to design for variable flows.

Table 1 summarizes the characteristics of the untreated greywater, the constructed wetland effluent, and the nanofilter permeate for different parameters, as well as the removal efficiencies for these parameters between the treatment steps. Overall, the combination of the constructed wetland and nanofilter achieved a good removal of BOD and TSS, with a decrease in BOD from 1240 to 60 mg/L (95%) and in TSS from 830 to 5 mg/L (99%). Whereas this quality may be sufficient for low-end uses such as toilet flushing, it should be noted that the treated greywater did not reach the effluent quality standards for centralized WWTPs, as specified in the Urban Waste Water Treatment Directive5, depending on the agglomeration size. This may present a challenge in case greywater is reused in urban areas where centralized sewerage is available and strict effluent discharge standards apply. Local treatment and reuse of greywater in, for example, garden irrigation, may be considered as discharge of effluent. Therefore, it is recommended to make a strict distinction between reuse applications that result in discharge into sewers (e.g., toilet flushing), versus reuse applications that result in discharge to the environment (e.g., garden irrigation). Furthermore, waste streams should be handled appropriately. For example, membrane concentrates, backwash water and excess raw greywater should be discharged into combined sewers or sanitary sewers in case of separated systems, while overflow of excess treated water should be avoided by pausing the treatment process during periods of low demand.

Table 1: The constructed wetland removes a major fraction of pollutants, but additional treatment by the nanofilter is needed to achieve better effluent quality. N.a.: no data available.

The permeability of the nanofilter membranes was followed up over time, as an indication of membrane fouling. Figure 3 depicts the membrane permeability of the membrane over time, expressed as L/m²/h/bar at 20°C, with indication of the type of feed water (tap water, greywater, or constructed wetland effluent) and chemical cleaning occasions. The permeability started at a level of 5 L/m²/h/bar with tap water, and quickly dropped to around 3 L/m²/h/bar when the membranes were fed with greywater. It was not possible to restore the permeability to the original level, neither by changing the feed (to tap water or constructed wetland effluent), nor by chemical cleaning (basic cleaning with 200 ppm NaOCl + NaOH at pH 11, or acidic cleaning with 2% citric acid + HCl at pH 2). The sawtooth pattern in permeability was due to a hydraulic cleaning after each run, indicating that hydraulic cleaning is sufficient to keep the permeability roughly stable over time. When treating constructed wetland effluent, the permeability remained between 3 and 4 L/m²/h/bar throughout the remainder of the project.

Figure 3. Permeability of the nanofilter membrane throughout the cumulative run time of the installation.

The production of struvite was monitored by analyzing the inflow and outflow of the struvite reactor. The orthophosphate content of the urine is shown in Table 2, along with the calculated formation of struvite. An average urine inflow of 41 L/d was obtained (i.e., urine + urinal flushing water), from which an average of 28 g/d of struvite was produced. Over 268 days of operation, this amounted to a cumulative struvite production of 7.6 kg. It should be noted that this is a stoichiometric calculation of struvite production, based on the orthophosphate difference between the reactor inflow and outflow. In practice, a lower amount of struvite was harvested (amount not yet reported). The difference between struvite formation and struvite harvesting can be attributed to the washout of small crystals through the effluent, and possibly buildup of precipitate along the reactor walls and piping.

Table 2: Phosphorus content of urine, and struvite crystal formation.

Overall, this project indicates that it is feasible to treat and reuse greywater at a decentral level with a small constructed wetland + nanofilter installation, and recover nutrients from source-separated urine in the form of struvite. Further work will focus on assessing the quality and consistency of the nanofilter permeate, and obtaining the required legal permits for marketing the produced struvite.


1.    Jönsson, H., Baky, A., Jeppsson, U., Hellström, D. & Kärrman, E. 2005 Composition of urine, faeces, greywater and biowaste for utilisation in the URWARE model. Urban Water Report 2005:6, Chalmers University of Technology, Gothenburg.

2.     STOWA 2011 Source-separation in the urban water infrastructure. Report 2012-W14. Stichting Toegepast Onderzoek Waterbeheer (STOWA), Utrecht.

3.     Derese, S. 2018 Towards innovative technologies for nutrient recovery from human urine. PhD thesis, Ghent University, Ghent.

4.     I-QUA 2019.

5.     Council Directive 91/271/EEC of 21 May 1991 concerning urban waste-water treatment

Project by:

F. Meerburg,a, B. Saerensa, D. Cortoisb, W. Moermanc, L. van de Putted, P. Fraipontd, K. Lombaerte, J. Severynsa, M. Weemaes*,a


aAquafin NV, Dijkstraat 8, 2630 Aartselaar, Belgium.

bDe Kruitfabriek, Steenkaai 44D, 1800 Vilvoorde, Belgium.

cNuReSys BVBA, Hoekstraat 3, 8540 Deerlijk, Belgium.

dStad Vilvoorde, Grote Markt, 1800 Vilvoorde, Belgium.

eMatexi NV, Franklin Rooseveltlaan 180, 8790 Waregem, Belgium.

Social Share Buttons and Icons powered by Ultimatelysocial