För ett regn med 10-års återkomsttid och 10 minuters varaktighet kommer att få en nederbördsökning på 4 % enligt den modell som används i P104 jämfört med modellen i P90 som användes vid projekteringen. Om intensitets och varaktighetskurvor ritas upp med hjälp av regndata från Uppsala, kommer det att medföra en nederbördsökning på 24 % jämfört med P90. Ökningar mellan P90 och P104 kommer inte att göra att kapaciteten hos magasinen överskrids eftersom ökningen är så liten. Dock så tyder studierna på att vattnet inte alls tar de vägar som har antagits vid projekteringen och som normalt antas vid dagvattenprojekteringar av detta slag. Det troliga är att större delen av den nederbörd som faller aldrig kommer att nå fram till dikena och infiltrera ner i dessa. Vid små nederbördsmängder kommer större delen av vattnet att infiltrera i stödremsan nära asfaltskanten. Om det kommer större mängder nederbörd kan infiltrationskapaciteten hos stödremsan överskridas, då kommer den yta varvid vattnet infiltrerar att sprida sig till övre delen av dikesslänten. Nederbörden som infiltrerar genom stödremsan kommer att perkolera ned genom marken med en i huvudsak vertikal gradient, men en del av vattnet kommer att transporteras horisontellt in i vägkroppen. Den nederbörd som tar sig in i vägkroppen kommer där att leds till terrassdräneringarna via fiberduken i botten av vägterrassen. Det finns en risk för att vattnet som kommer in i vägterrassen orsakar skador på vägen. Även om nederbörden inte går via de stenfyllda dikena så kommer inte reningseffekten att gå förlorad utan reningen kommer att ske i vägkroppen samt ytskikten och troligtvis bli minst lika hög som den hade blivit om nederbörden tagit den väg som det projekterades för. Det skulle kunna vara så att dikena inte kommer att ha annan effekt än sänkor för grundvattnet och säkra avåkningsytor. / The purpose of this work is to describe the hydraulics in rock-filled stormwater ditches and surface runoff from roadways, and to make a statement about the degree of purification based on other research. There has been no research in either impact or cost savings / higher expenses by using this type of system compared to a conventional stormwater systems for motorways. If this type of solution is cost effective it would mean savings for future construction in which this solution is chosen. A comparison was made of how well the calculations based on P90, which were used during the design stage with the upcoming P104. P90 and P104 is published by the interest association Swedish water and can be viewed as guidelines for water and wastewater industry how to make the design of stormwater systems. For a rainfall of 10-years return period and a 10-minutes duration, the precipitation will increase 4%, according to the model used in the P104 compared to P90. If the intensity and duration curves are plotted using rainfall data from Uppsala, it will increase the precipitation 24% compared to P90. Increases between P90 and P104 will not exceed the capacity of the reservoirs since the increase is so small. If these models are inaccurate and precipitation intensity will be equal to the intensity and duration curves, the capacity of the reservoirs will be exceeded. However, as the studies indicate that the water does not take the way that have been adopted in designing and normally assumed by stormwater projections of this kind. It is likely that most of the precipitation that falls will never reach the ditches and infiltrate into these. For small amounts of precipitation, the majority of the water will infiltrate in the support strip near the edge of the asphalt. If there are large amounts of rainfall the recharge capacity of the support strip will be exceeded, then the surface water will infiltrate in the upper trench slope. Precipitation that infiltrates through the support strip will be percolating down through the soil with an essentially vertical gradient, but some of the water will be transported horizontally into the road structure. The precipitation that enters the road structure will be led to the collector pipes via fiber cloth in the bottom of the embankment. There is a risk that the water that enters the embankment causes damage to the road. Various injuries that can occur are: reduced friction between the grains in the unbound filling materials, redistribution of loads, pumping, depressions, increased cracking of the carriageway and damage that may occur when freezing during the cold season. Even if the water does not go through the stone-filled stormwater ditches, the cleaning effect will not be lost, purification will take place in the road structure and probably be at least as high as it had been if the precipitation took the way that it was designed for. It could be so that the ditches will have no other effect than sinks for ground water and safe running off surfaces. According to the contractor the cost savings were about 15% compared to a conventional stormwater systems. This figure is probably on the low side when Ramboll came in late in the projection and had to redo parts of the design of the system during construction time.