This paper presents the key elements that emerged from a 3-year study on 4 campsites. Each of them has its own Vertical-Flow Constructed Wetlands (VFCW) for on-site wastewater treatment. Each year, the pollution discharge peaked in the 5 to 6-week summer period but was zero for a 6-month period. The natural-concentrate wastewater is collected into a short sewer network. Hydraulic load and campground occupancy rate were tightly correlated. The effluent has a characteristically high nitrogen concentration due to the lower greywater-producing activity. These quantitative and qualitative elements were used to define the camper equivalent' as a daily pollutant load of about 100 L water, 90 gCOD, 35 gBOD5, 40 gSS, 11 gKN and 1.4 gPt. During the 5-6 weeks of peak visitor traffic, the full first stage receives an organic load of 200 gCOD.m-2.d-1 and 25 gKN.m-2.d-1, which is 2 to 3.5-fold higher than the loads conventionally applied in VFCW systems in France. Despite these heavy loads, COD removal remains constant (at 91%) over the season. During the summer peak, nitrification drops sharply but KN removal rate holds steady near 60%. This study demonstrates that VFCWs can robustly handle massive load increases. The paper concludes that total system size can be downscaled to 0.75 m²/camper. This design allows to reach minimum yields of 75% KN, provided that the system is built and run to state-of-the-art rules. | Cet article présente les éléments clés issus d'une étude de 3 ans sur 4 campings qui traitent leurs eaux usées de manière individuelle par des Filtres Plantés de Roseaux à écoulement vertical (FPRV) aux dimensionnements différents. La charge hydraulique et le taux d'occupation du camping sont étroitement corrélés. L'émission est nulle 6 mois par an et la pointe estivale dure 5 à 6 semaines. Les eaux usées, collectées par un réseau court sont naturellement concentrées. L'effluent présente notamment une très forte concentration en azote du fait d'une activité ménagère réduite. Ces éléments qualitatifs et quantitatifs permettent de définir l' « équivalent campeur » par une charge de pollution journalière d'environ 100 L d'eau, 90 g de DCO, 35 g de DBO5, 40 g de MES, 11 g de NK et 1,4 g de Pt. Au cours des 5-6 semaines de fréquentation maximale, la totalité du 1er étage reçoit une charge organique d'environ 200 gDCO.m-².j-1 et 25 gNK.m-².j-1, soit 2 à 3,5 fois plus que les charges classiques appliquées en FPRV en France. En dépit de ces fortes charges, l'élimination de la DCO reste constante (91 %) au cours de la saison. Lors du pic de fréquentation estival, la nitrification décroît fortement mais le rendement global se maintient à environ 60 %. La robustesse des FPRV face à une augmentation massive des charges appliquées est démontrée. L'article conclut par un dimensionnement réduit de 0.75 m2/campeur au total. Ce dimensionnement permet d'atteindre un rendement minimum de 75 % du NK, correspondant à la nitrification, sous réserve que les règles de l'art de construction et d'exploitation soient strictement respectées.

Subsurface horizontal flow constructed wetlands (HFCW) and vertical flow constructed wetlands (VFCW) have been compared regarding their efficiencies on microbiological indicators removal. Seven types of HFCWs and two types of VFCWs have been monitored for several years. Fecal coliforms, E. coli, somatic coliphages and F-specific bacteriophages were analyzed in the influent and effluents of each system. All constructed wetlands monitored were planted with Phragmites australis but have different design characteristics and operation variables. All data were statistically treated (SPSS v17) in order to analyse the effect of design and operation variables on each microbiological indicator. VFCWs were more efficient than HFCWs systems when considering removal rates per surface for all the microbiological indicators (cfu or pfu removed/m2.d). Nevertheless, considering mean removal efficiencies (in log units), results did not show statistically significant differences. The hydraulic retention time was the key parameter regarding filter disinfection capacity in both types of filters. Bacterial indicators were removed at a higher rate than viral indicators.

Phosphorus (P) removal in constructed wetlands (CWs) has received particular attention during the last decades by using specific materials which promote P adsorption/precipitation phenomena. Recent studies have shown an interest in using apatite materials to promote P precipitation onto their surface. About ten years of experiments at different scales (laboratory, pilot, full scale) have passed. The process is now open in France. This paper presents the first design recommendations according to the apatite mineral quality and treatment outlet requirements. Kinetics and their evolution with the material saturation degree are presented based on the k-C* model. The process can achieve 1 mgP.l-1 for long term while using surface P removal filter area of 0.5 m².p.e.-1. With a low land footprint and high P removal efficiency, the interest of the process also result in the orthophosphates accumulation on a phosphate mineral which makes easier to reuse phosphorus from wastewater once saturated.

We develop an original solid respirometric method to measure the biodegradation rates of COD and ammonium at different depths of a Vertical Flow Constructed Wetland (VFCW). The method is developed for a matrix operated under unsaturated conditions. The oxygen demand is measured versus time in a gas passing through a reactor containing a mixture of an organic matrix mixed with a bulking agent (wood). For the measurement of the nitrification rate, a solution of ammonia and hydrogenocarbonates is spread on the sample. We ran 15 experiments on different matrices with variable organic matter contents (sludge, VFCW gravels and sand). Within this paper we present the overall methodology, the factors that allow improving the accuracy (number of duplicates, blank, sample volume, bulking agent volume, gas flow, etc.) and the nitrification rates that were obtained.

Oxygen renewal, as a prominent phenomenon for aerobic bacteria activity, deeply impacts VFCW treatment efficiency. We introduce MIGRA, a new model able to simulate oxygen transfer in VFCW. It is based on a realistic two-phase flow module, that allows us to describe various loading sequences, and a transport module. The multi-component transport module is able to deal with convection/diffusion phenomena, inter-phase (air-water) mass exchange, and first-order kinetics. It is validated with analytical solutions and data from an experimental VFCW (column). We show the benefits of a two-phase flow modeling using a comparison with HYDRUS.

French vertical flow constructed wetlands treating raw wastewater have been developed successfully over the last 30 years. During the last years, SINT and Epur Nature developed the approach further by introducing a deep single-stage vertical flow bed that comprises both stages into one in the aim to reduce the footprint. This process, patented by Epur Nature as Bi-filtre®, consists in two vertical stages piled-up. While classical systems in France are designed with 2m².PE-1 (1.2m² and 0.8m² on the first and second stage respectively), the Bi-filtre® has a larger total filtration area with 2.5 m²/PE (1.5m² and 1 m² on the first upper and second lower stage respectively) but a foot print reduced to 1.5 m².PE-1. To favour aeration of the system, an intermediate natural aeration system is introduced at the interface between the first and second stage. The paper will present a study on a fullscale plant (Saint Michel l'Observatoire, 175 PE, France) running at full load in which different materials (sand, pozzolana of different particle size) were tested at the second stage. The aim was to identify the best materials to use to guarantee performance and avoid clogging on this unplanted and buried stage. The study consisted in measuring hydraulic conditions (inlet/outlet flows, infiltration rates, tracer experiments) correlated with 24h composite samples assessments. The results shows that such systems, if well designed, can guarantee an outlet quality of 35 mg.L-1 in SS, 125 mg.L-1 in COD, 25 mg.L-1 in BOD and 20 mg/L in KN with a total foot print of 1.5 m².PE-1. The paper will present the design recommendations and the limits of the systems taking into account a larger experience gathered on different existing plants (e.g. Chorges Bifilter®, 2 150 P.E, France).

This paper relates to the 1st studies conducted in France on Vertical Flow Constructed Wetlands [VFCWs]. This process was originally designed by Käte Seidel according to her previous research done in Krefeld (Germany) the so-called Max Planck Institute Process [MPIP]. Based on measurements campaigns done successively in Saint Bohaire and Pont-Rémy, it was clearly established that the vertical flow 1st stage filters were significantly more efficient than the horizontal ones. This acknowledgement of success was positively used to design a treatment plant in Gensac la Pallue, still in operation after 23 years, with the special feature that the VFCW are fed with raw sewage. This attribute was then spread to the design of French VFCW systems. The main reasons which explain why representatives of small communities are so enthusiastic about feeding with raw sewage are presented. The state of the art of French VFCWs is detailed and scientific arguments which nowadays lead to new fields of application and a better understanding of small scale processes which occurs in these filters are also suggested.

[Departement_IRSTEA]Eaux [TR1_IRSTEA]ARCEAU | The purpose of this work is to assess the effectiveness of two grass covers (buffer zone and grass-covered inter-row), to reduce pesticide leaching, and subsequently to preserve groundwater quality. Lower amounts of pesticides leached through grass-cover soil columns (2.7e24.3% of the initial amount) than the bare soil columns (8.0e55.1%), in correspondence with their sorption coefficients. Diuron was recovered in higher amounts in leachates (8.9e32.2%) than tebuconazole (2.7e12.9%), in agreement with their sorption coefficients. However, despite having a sorption coefficient similar to that of diuron, more procymidone was recovered in the leachates (10.2e55.1%), probably due to its facilitated transport by dissolved organic matter. Thus even in this very permeable soil, higher organic matter contents associated with grass-cover reduce the amount of pesticide leaching and limit the risk of groundwater contamination by the pesticides. The results of diuron and tebuconazole transfer through undisturbed buffer zone soil columns are in agreement with field observations on the buffer zone.

L'article présente l'application de l'Analyse du Cycle de Vie (ACV) à un système d'assainissement complet, incluant le réseau d'égout. Il présente d'abord les hypothèses retenues pour l'ACV concernant les deux systèmes de traitement des eaux usées ayant tous les deux la même charge nominale quotidienne de la DBO 5 et associés au même réseau d'égouts. Ces systèmes correspondent respectivement à (i) un "filtre planté de roseaux à lit vertical (VFCW)" pour lequelle un inventaire détaillé des flux émis et des ressources consommées a été élaboré et (ii) une STEP à « boues activées » boues » issue de la base de données d'ACV Ecoinvent LCI. Les résultats d'ACV soulignent l'importance de l'eutrophisation qui peut facilement s'expliquer par la suppression incomplète de l'azote total et du phosphore total dans un VFCW. De façon plus surprenante, l'impact du réseau semble considérable. Enfin, l'article analyse l'applicabilité et les limites de l'ACV pour les systèmes de traitement des eaux usées. / The article presents the application of Life Cycle Assessment (LCA) to a complete sanitation system including the sewer network. It first describes the LCA hypothesis which concerns two types of waste water-treatment plant with the same daily nominal load in BOD5 and associated to the same sewer network derived from the Life Cycle Inventory (LCI) database Ecoinvent. The two wastewater treatment systems compared are (i) a Vertical Flow Constructed Wetlands (VFCW) for which a detailed inventory was elaborated and (ii) an activated sludge stemming from the LCI database Ecoinvent. LCA scores of VFCW highlight the importance of eutrophication which can be easily explained by the incomplete removal of total N and total P in a VFCW. In a more surprising way, the impact of the network seems considerable. Finally, the article analyses the applicability and limitations of LCA for wastewater treatment with regard to water quality and the needed improvements of water status in LCA.