Background Approximately 20% of the surface water in the EU is severely threatened by pollution from nitrates and phosphates caused mainly by agricultural activities. These cause redundant algae growth, harming fish and sometimes humans. One solution developed to clean polluted surface water is the vertical flow reed bed. The vertical flow reed bed consists of layers of sand, iron fillings and gravel - phosphate in the water binds to and is held by the iron particles. It is planted with reed and pipes and drainage systems are added. This had been shown to work effectively in certain conditions, but experiences with the filter had been limited to systems cleaning relatively polluted water and with a lifetime of only 5 to 6 years. Leidsche Rijn is a new urban area for over 90,000 inhabitants near Utrecht in the Netherlands, with a special sustainable surface water system involving about 82km of waterways. It is a closed system retaining and storing water in the area’s ground, ditches, canals and lakes that has no connections with the surrounding surface water, except for an outlet and inlet. However, since the closed system is fed by ground water and rainwater, urban pollution continues to get into the water system. Objectives This LIFE project, known as Verbal, aimed to demonstrate a successful system for improving the quality of relatively clean surface water for a complete urban area that could last at least 30 years. It sought to test a 1 ha filter to define the optimal set up of a future 6 ha vertical reed bed filter in the residential area Leidsche Rijn. It also sought to design a general applicable model for the design, maintenance and control of vertical flow reed beds in other settings. The project hoped to show the effectiveness of a vertical flow reed bed filter at removing phosphate from surface water. It also expected to remove some nitrates, bacteria, heavy metals and suspended solids. Results The Verbal project demonstrated that the innovative vertical flow reed bed filter is effective at cleaning low polluted surface water in a closed system in a large-scale urban area. Being a biological filtering system, it makes minimal use of chemicals and energy and can blend naturally into the landscape. The project constructed a vertical flow reed bed filter in Leidsche Rijn with one large filter bed of 2 220 m2 and 12 filter beds of 170 m2, with a depth of 1m. Every filter had a unique combination of variables such as the amount of iron and limestone, the filling frequency or the presence of reed to enable the effect of every single variable to be determined. To monitor the quality of the different water flows, a number of continuous measurements were taken, such as pH, transparency and oxygen. Additionally, regular water samples were taken to measure the levels of nitrate and phosphate. To simulate a period of thirty years (the period of time needed to remove phosphates that have leached into the soil) in two years, the concentration of phosphate was artificially increased in two filter beds up to 80mg/l. During the two years of measuring, the removal of dissolved and non-dissolved phosphate was found to be very high. The concentration of phosphate in the effluent of all filters was on average clearly below 0.05 mg/l. The project demonstrated that the filter is a good alternative for cleaning surface water in urban areas and even suggests that the system could treat a higher surface load compared to the original design parameters. The collected data provided sufficient information to construct an optimal full-scale filter and fit it in to the water system at Leidsche Rijn, where it will continue to be used by the regional government in order to provide clean surface water its residents. It is anticipated that the filter will initially serve to clean surface water in the canals within the residential area and will ultimately ensure sufficiently clean water in a lake area that will be used for swimming, canoeing, and other watersports. As a major deliverable to promote transferability, the project produced a general applicable model for the vertical reed filter that may be applied to other areas in Europe. This allows the optimal dimensions and required equipment, management and maintenance of a filter in any given setting to be determined, together with an estimate of the investment and annual management costs. In financial terms, the beneficiary expects the verticle flow method to be particularly cost effective (compared to alternative methods such as horizontal reed bed filters and chemical disposal) in areas where land is expensive. Disclaimer : This « results » section should be considered as a draft until the Commission has completed its evaluation . This project has been selected as one of the 17 "Best" LIFE Environment projects in 2008-2009.