Infiltration Trenches in Scania

Water and Environmental Engineering Department of Chemical Engineering Infiltration Trenches in Scania A Study of the Hydraulics and the Pollutant Re...
Author: Augustus Carr
2 downloads 0 Views 4MB Size
Water and Environmental Engineering Department of Chemical Engineering

Infiltration Trenches in Scania A Study of the Hydraulics and the Pollutant Removal Effect

Master’s Thesis by

Sofia Vestergren April 2010

Vattenförsörjnings- och Avloppsteknik Institutionen för Kemiteknik Lunds Universitet

Water and Environmental Engineering Department of Chemical Engineering Lund University, Sweden

Infiltration Trenches in Scania A Study of the Hydraulics and the Pollutant Removal Effect Master Thesis number: 2010-06 by

Sofia Vestergren Water and Environmental Engineering Department of Chemical Engineering April 2010

Supervisor: Viveka Lidström, LTH/Ramböll Sweden AB Co-Supervisor: Lena Sjögren, Ramböll Sweden AB Examiner: Professor Jes la Cour Jansen

Picture on front page:

1

Postal address: P.O Box 124 SE-221 00 Lund Sweden

1. Infiltration Trench, Malmö, Sweden (Vestergren, 2010)

Visiting address: Getingevägen 60

Telephone: +46 46-222 82 85 +46 46-222 00 00 Telefax: +46 46-222 45 26 Web address: www.vateknik.lth.se

ii

Foreword and Acknowledgements The idea of studying the hydraulics and the pollutant removal effect of infiltration trenches in Scania was contrived by Lena Sjögren at Ramböll Sweden in Malmö, whom further developed the idea together with Viveka Lidström at the technical faculty of Lund University (LTH). A special thank is directed to both of them for supervising me on my work that has resulted in this report: Infiltration trenches in Scania – a study of the hydraulics and the pollutant removal effect. The project was running from early October 2009 to late March 2010. I would also like to thank my examiner Jes la Cour Jansen for his views. A special thank also to Elisabet Hammarlund, Ramböll Sweden AB in Malmö, for her supervision on geology and geotechnical matters. I would also like to show my gratitude to Åsa Peetz and Peter Wenzel at NSVA and Pär Hagstrand, Stefan Milotti, Jeanette Lindvall and Henrik Thorén at VA Syd for making the experiments possible by donating time and resources. Further on, I would like to thank Anna Fälth, Ramböll Sweden AB in Gothenburg, for letting me take part of her works. I would also like thank the following persons for, in one way or another, contributing to my work: Jennifer Molloy, EPA – terminology Mikael Persson, manager Biltema – granting infiltration trenches Marika Aurell, Buffin Real Estate Sweden AB – blueprints/geotechnical survey Ann-Marie Frisell, Helsingborg Stad – traffic statistics Gertrud Persson and Ylva Persson, LTH laboratories – laboratory work Paul Widenberg, DHI Växjö – weather data Sydsten AB (Dalby) – macadam samples Tommy Olsson, manager/reseach engineer, plant ecology, Lund University – laboratory work Sofia Dahl, NSVA – site facts I would also like to thank the persons participating in the questionnaire study: Stefan Lindgren, Hässleholms Vatten AB Bjarne Segersteén, Östra göinge municipality Mona Eliasson, Bjuv municipality Lars Svensson, Osby municipality Jan-Åke Persson, Trelleborg municipality Jesper Andersson, Sjöbo municipality Mats Ljung, Svalöv municipality Anna Palm, Hörby municipality Olle Johannesson, Landskrona municipality Anders Johnsson, Bromölla municipality Magnus Brom/Anders Horstmark, Eslöv municipality Per Nilsson, Lomma municipality Sofia Westergren Lund, April 2010

iii

iv

Summary Infiltration trenches are a common solution in Scania for handling storm water. The main purposes are to delay and temporary house stormwater and to even out the run-off peaks. The solution has however not been evaluated at any greater extent in Sweden and there are no common standards for how to construct the trenches. Several municipalities and consulting companies have requested an evaluation of the hydraulic function and the pollutant removing effect. The main objective of this report is to describe how an infiltration trench works by studying the hydraulic transport in the infiltration trench and the pollutant removing effect of the infiltration trench, with focus on heavy metals (cadmium, chrome, copper, lead, mercury, nickel and zinc). The project consists of a literature study, a questionnaire study and experiments. The aim of the questionnaire study was to review the experience of, and thoughts about, infiltration trenches among the Scanian municipalities. The experiments included flushing/flow measuring and sample-taking/metal analyses. During the former experiment, a tanker truck was emptied directly into the trench or on the contributing area while the outflow from the trench was measured digitally or with bucket/stopwatch in order to get a picture of the hydraulics. During the latter experiment, samples of filling aggregate was collected and washed with water or acid where after the liquids were analyzed in regard to heavy metal content in order to get an idea about the pollutant removal effect. There are different opinions about whether infiltration trenches should be constructed in impermeable soils or not due to the lack of infiltration, but studies have shown that the infiltration rate is higher than expected under these conditions. On the other hand, most Scanian municipalities consider the time-lagging effect being the main advantage of infiltration trenches. The main competing solution is still conventional pipes. Flushing/flow measuring experiments showed that water runs right through a new trench. A delaying effect can be reached with a throttling outlet. The hydraulic conditions changes over time as the trenches get clogged. Older facilities show a time-lagging effect, but also different types of errors related to clogging. Previous studies show that 50% of the infiltration trenches have a reduced efficiency after 5 years of running. Physical clogging can be avoided at some extent by pre-treatment. The catchment area should preferably be fully developed and vegetated before constructing the infiltration trench in order to prevent premature clogging. The pollutant removal effect of the infiltration trench is relatively unknown, but is assumed to have a high removal rate for particle-bound pollutants and a moderate removal rate for soluble pollutants. Both laboratory experiments and full-scale trenches show that a large part of the pollutions get stuck in the upper 1-3 decimeters. The highest pollutant removal effect seems to be reached when combining the infiltration trench with a swale. In the trenches included in this study, lead is the most common heavy metal (when washing with acid) in contrast to the common distribution in stormwater, where zinc is dominant. Most Scanian municipalities do not have a maintenance plan. Putting together a maintenance plan valid for Swedish conditions would be a useful measure and could also bring a development of the structure of the infiltration trench by enlightening errors. v

vi

Sammanfattning Makadamdiken är en vanlig dagvattenlösning i Skåne. Huvudsyftet med denna lösning är att fördröja och tillfälligt magasinera dagvatten samt att minska flödestopparna. Makadamdiken har inte utvärderats i någon högre utsträckning i Sverige och det finns ingen generell standard. Flera kommuner och konsultbolag har efterfrågat en hydraulisk och föroreningsmässig studie. Målet med denna rapport är att beskriva hur ett makadamdike fungerar genom att studera hydrauliken och föroreningseffekten – med fokus på tungmetaller (bly, kadmium, koppar, krom, kvicksilver, nickel och zink). Projektet består av en litteraturstudie, en enkätundersökning och experiment. Målet med enkätundersökning var att få en bild av skånska kommuners tankar och erfarenheter av makadamdiken. Experimenten omfattade såväl spolning/flödesmätning som provtagning/metallanalys. Under det tidigare experimentet tömdes en tankbil med vatten direkt eller indirekt i diket medan utflödet mättes digitalt eller med hink/stoppur. Detta för att få en bild av hydrauliken. Under det senare experimentet togs prover av makadam som sedan tvättades med syra eller vatten. Dessa vätskor analyserades sedan med avseende på tungmetaller för att få en uppfattning om makadamdikens renande effekt på dagvatten. Det råder delade meningar om huruvida makadamdiken bör byggas i impermeabla jordar eller ej eftersom dessa inte tillåter infiltration till omkringliggande jordar. Studier har dock visat att infiltrationen från sådana diken är store än väntat. Å andra sidan är det den fördröjande effekten som är den huvudsakliga fördelen med dessa diken enligt de skånska kommunerna. Konventionella ledningar ser de som den främsta konkurrerande lösningen. Experimenten i form av spolning/flödesmätning visade att vattnet rann rakt igenom de nyligen konstruerade diket och att man måste ha ett strypt utflöde för att få en fördröjande effekt. De hydrauliska förutsättningarna ändras dock med tiden till följd av olika grad av igensättning. Studier har visat att 50% av alla makadamdiken har förlorat kapacitet efter 5 års drift. Igensättningen kan till viss del förhindras genom avskiljning av sediment. För att inte riskera att dikes sätts igen i förtid bör ett exploateringsområde bebyggas och planteras innan makadamdiket konstrueras – så är dock normalt inte fallet. Makadamdikets renande effekt, med avseende på tungmetaller, är relativt okänd, men det antas ha en hög reningseffekt för partikelbundna föroreningar och en medelhög reningseffekt för lösta föroreningar. Både laboratorieexperiment och fullskalestudier har visat att en stor del av föroreningarna fastnar i de översta 1-3 decimetrarna. Störst grad av rening tycks man få om man kombinerar makadamdiket med ett gräsbeklätt svackdike. Dikena i studien visade högst halter av tungmetallen bly (vid syratvätt), vilket strider mot den vanliga fördelningen i dagvatten där zink är vanligast. Bland de kommuner som svarade på enkätundersökningen återfanns bara ett fåtal som hade någon form av strukturerat underhåll av makadamdiken. Genom att konstruera, och implementera, en underhållsplan för makadamdiken i svenska jordar skulle konstruktionen kunna förbättras genom att problemen belyses.

vii

viii

Contents 1

2

3

Introduction................................................................................................................................................................1 1.1

Background .......................................................................................................................................................1

1.2

Objectives ...........................................................................................................................................................1

1.3

Methodology .....................................................................................................................................................1

1.4

Limitations and Assumptions of This Thesis Work ..........................................................................2

1.5

Definitions .........................................................................................................................................................2

Infiltration Trenches ...............................................................................................................................................3 2.1

General Description .......................................................................................................................................3

2.2

Applicability ......................................................................................................................................................3

2.3

Advantages ........................................................................................................................................................4

2.4

Operation, Efficiency and Performance .................................................................................................4

2.5

Maintenance......................................................................................................................................................5

2.6

Costs .....................................................................................................................................................................7

2.7

Design ..................................................................................................................................................................8

2.7.1

Dimensioning of an Infiltration Trench .......................................................................................8

2.7.2

Other Construction Considerations ............................................................................................ 11

2.7.3

Design Summary................................................................................................................................. 16

Storm Water Quality ............................................................................................................................................ 17 3.1

Pollutants: Heavy Metals .......................................................................................................................... 17

3.2

Level of Sorption Related to Grain Fraction ..................................................................................... 19

3.2.1

Geotechnology ..................................................................................................................................... 19

3.2.2

Sorption.................................................................................................................................................. 20

4

Questionnaire Study ............................................................................................................................................. 23

5

Case Studies ............................................................................................................................................................. 25 5.1

Study Areas..................................................................................................................................................... 25

5.1.1

Svågertorp, Malmö............................................................................................................................. 25

5.1.2

Mariastaden, Helsingborg............................................................................................................... 26

5.2

Experiment Description ............................................................................................................................ 27

5.2.1

Metal Content Analysis .................................................................................................................... 27

5.2.2

Flushing/Flow Measuring .............................................................................................................. 29

5.3

Case 1: Biltema I ........................................................................................................................................... 30

5.3.1

Trench Description ............................................................................................................................ 31

5.3.2

Special Conditions.............................................................................................................................. 32

5.3.3

Metal Content Analysis .................................................................................................................... 32 ix

5.3.4 5.4

Case 2: Biltema II ......................................................................................................................................... 33

5.4.1

Trench Description ............................................................................................................................ 34

5.4.2

Metal Content Analysis .................................................................................................................... 35

5.4.3

Flushing/Flow Measuring .............................................................................................................. 36

5.5

Case 3: Kungshultsvägen Road............................................................................................................... 37

5.5.1

Trench Description ............................................................................................................................ 39

5.5.2

Metal Content Analysis .................................................................................................................... 41

5.5.3

Flushing/Flow Measuring .............................................................................................................. 45

5.6

6

Flushing/Flow Measuring .............................................................................................................. 33

Case 4: Pålsjö 3:1.......................................................................................................................................... 46

5.6.1

Geology ................................................................................................................................................... 47

5.6.2

Trench Description ............................................................................................................................ 47

5.6.3

Metal Content Analysis .................................................................................................................... 48

5.6.4

Flushing/Flow Measuring .............................................................................................................. 50

Resuming Results and Discussion .................................................................................................................. 53 6.1

Pollutant Removal Effect .......................................................................................................................... 53

6.2

Hydraulics ....................................................................................................................................................... 62

6.3

Comparison with Similar Stormwater Facilities............................................................................. 62

6.3.1

Performance Assessment of a Swale and a Perforated Stormwater Infiltration System, Toronto, Ontario ................................................................................................................ 62

6.3.2

International Competition in Purification of Stormwater................................................. 62

6.3.3

Local Stormwater Treatmet – Experience from Some Systems in Operation........... 63

6.4

Comparison with Competing Solution ................................................................................................ 65

6.4.1

Infiltration Trenches vs. Conventional Pipes.......................................................................... 65

6.4.2

Infiltration Trenches vs. Vegetated Filter Strips ................................................................... 65

6.4.3

Swales ..................................................................................................................................................... 66

6.5

Advantages of Infiltration Trenches .................................................................................................... 67

6.6

Disadvantages of Infiltration Trenches .............................................................................................. 67

6.7

Possible Improvements of the Design ................................................................................................. 67

6.8

Maintenance................................................................................................................................................... 68

6.9

Method Evaluation ...................................................................................................................................... 68

7

Conclusions .............................................................................................................................................................. 69

8

Suggestions for Further Work.......................................................................................................................... 71

9

References ................................................................................................................................................................ 73

Appendix A......................................................................................................................................................................... 77

x

Appendix B......................................................................................................................................................................... 79 Appendix C ......................................................................................................................................................................... 85 Appendix D ........................................................................................................................................................................ 87 Appendix E: Article ......................................................................................................................................................... 89

xi

xii

1 Introduction 1.1 Background An increasing amount of paved surfaces in our society leads to a decreasing possibility for the stormwater to naturally infiltrate. It also causes a rapid run-off with high peaks. The traditional way to handle stormwater is to lead it through underground pipe systems. Exploitation has however led to a number of problems, such as overload of the pipe system, basement flooding, need to overflow combined systems, large quantities to handle at the treatment plants and the necessity to re-dimension the already existing pipe networks. In the 70’s a new way of thinking was introduced. It included an environmental ambition to treat the stormwater and recharge it to the groundwater close to the source. Control measures to decrease stormwater quantities and peak-flows as well as to improve stormwater quality was taken. Sustainability is a keyword in this context and it refers to the planning and realization of the facilities and not to the service life of the facility itself. Infiltration trenches, wetlands, green roofs, swales, porous pavers and ponds are all part of what is called BMP – Best Management Practice (British: SUDS (Sustainable Urban Drainage Systems); Swedish: öppen dagvattenhantering). An infiltration trench is one possible solution for handling stormwater. The main purpose is to delay and temporary house storm water and to even out the run-off peaks. The solution has however not been evaluated at any greater extent and there are no common standards for how to construct the trenches. Several municipalities and consulting companies have requested an evaluation of the hydraulic function and the pollutant removing effect of these facilities.

1.2 Objectives The main objective of this report is to investigate and evaluate the function of infiltration trenches in Scania. Focus is on: 

The hydraulic transport in the infiltration trench before, during and after a rain event



The pollutant removing effect of the infiltration trench

1.3 Methodology The study consists of three complementing parts: 1. A literature study 2. Experiments 3. A questionnaire study The literature study includes a discussion about stormwater and its heavy metal content as well as the chemical prerequisites and processes behind sorption. It also deals with the hydraulics of the trenches. Former studies of infiltration trenches are presented and comparisons with alternative solutions are made.

1

The experiments include flushing/flow measuring in order to determine the hydraulic capacity of the trench as well as sample-taking/metal analysis of the filling aggregate in order to evaluate the pollutant removal effect of the trench. The experience of infiltration trenches among a number of Scanian municipalities was evaluated through a questionnaire study.

1.4 Limitations and Assumptions of This Thesis Work There are no discrete definitions for a lot of the terminology for this kind of stormwater solutions. One type blends into another with respect to what different types of practices are called. Infiltration trenches share the same structure as the Swedish makadamdike. Depending on the soil characteristics, the function of the facility varies. According to VAV P46, trenches in soils with a permeability of less than 2∙10-5 m∙s-1 are considered having no infiltration capacity and working purely as delaying volumes (VAV, 1983). This study focus on the south of Sweden, where the soils often contain high concentrations of clay and are close to impermeable and the infiltration trenches do not allow the water to infiltrate to the surrounding soils. The main purpose of these trenches is a delaying effect. Due to the lack of geotechnical survey results for some of the areas of interest, assumptions based on geological maps have been used for determining the hydraulic conductivity of the soils. This assumption is justified by the well known characteristics – impermeability - of the Scanian clays. The experiments took place during October/November 2009 and March 2010. To get a more just result, the experiment should have taken place all year long in order to consider weather-related varying conditions. A ‘first-flush’ not followed by a continuous rain would cause greater pollution levels in the infiltration trenches than a ‘first-flush’ followed by a continuous rain, since the latter would dilute. Further on, it is difficult to find dry conditions in Scania during the winter half-year. The changing weather conditions – autumn turning winter turning spring – may also have caused a different load to the trenches regarding salts, tire wear particles etc. The field work was delayed by a long period of temperatures below zero with heavy snowfalls. The snow melting may have caused either an accumulation of pollutants in the trenches or a flushout. The measuring devise have been a limiting factor due to calibration problems and the fact that the trenches have not been constructed in a way that easily allows monitoring. Most of the reference literature is from the U.S. and may therefore not give a correct understanding for the Swedish conditions (climate, geology etc.). It would have been preferable with more Swedish sources, but the number of Swedish publications within the area is limited.

1.5 Definitions BMP: Best Management Practice; control measures taken to decrease stormwater quantities and peak-flows as well as to improve stormwater quality EPA: United States Environmental Protection Agency Macadam: crushed rock in fractions 2-63 mm; commonly occurring fractions: 4-8 mm, 8-12 mm, 8-16 mm, 12-16 mm, 16-27 mm, 16-32 mm and 32-63 mm 2

2 Infiltration Trenches 2.1 General Description An infiltration trench is an underground volume filled with gravel, i.e. commonly macadam, Figure 1. The effective volume is made up by the pore volume – approximately 30% of the total volume if the filling aggregate is macadam. The emptying of the trench happens through infiltration to the surrounding soils or via a drainage pipe (Stahre, 2004). The trench has commonly a light bottom slope and the drainage pipe can run through the whole length of it or just collect water at the very end. Several Swedish stormwater investigations show that the time-lagging effect is held as the main objective for constructing infiltration trenches. However, according to the Tennessee BMP Manual (2004), infiltration trenches should mainly be used for water quality improvements and are not recommended for quantity control since they do not reduce runoff peaks very well.

Figure 1. Infiltration Trench Sketch (Vestergren, 2009).

2.2 Applicability An infiltration trench is a stormwater solution suitable when the catchment area is narrow and when there is no possibility to lead the water to a grassed area. It is however not a suitable solution if the groundwater level exceeds the bottom of the trench according to Stahre (2004), whilst U.S. governmental sources recommend a groundwater clearance of at least 0.6 m. According to VAV P46, infiltration trenches should not be considered unless the groundwater table is situated at least 0.5 m below ground (VAV, 1983). According to Minnesota Urban Small Sites BMP Manual (2001), infiltration trenches should be constructed not closer than 45 m from drinking water wells in order to avoid contamination; and 3 m down-gradient and 30 m up-gradient from building foundations in order to avoid seepage problems. This kind of requirements, however, is set by local authorities.

3

IDEQ (2005) do not recommend constructing infiltration trenches in low permeability soils due to slow infiltration. A Danish study of infiltration trenches in low-permeability urban soil from 1999, however, shows that the infiltration rate was higher than expected from prevailing conditions (Warnaas et al., 1999). Soils with more than 40% clay content are subjects to frost heave, which make them excluded in areas were frosts occur (IDEQ, 2005). According to EPA (1999), infiltration trenches have mainly been used in the warmer and less arid parts of U.S., but they are also considered being applicable in colder climate. In case of colder climate, a part of the trench should be constructed below the frost line and the trench surface has to be kept free from ice and compacted snow. This kind of facilities should not be constructed at sites where chemicals and hazardous substances are being kept (EPA, 1999). Infiltration trenches are not recommended in areas with heavy traffic loads, such as driveways and parking lots, because the soils become compacted and accessing the trenches for maintenance and monitoring is hard (IDEQ, 2005).

2.3 Advantages Infiltration trenches are, in contrast to many other BMP’s, suitable for retrofitting due to being easy to fit into the margin (IDEQ, 2005). In a permeable soil, the infiltration trench will allow groundwater recharge, which contributes to a decreasing risk of soil settlement and a decreasing load to the sewage net. These effects will however default in a low permeability soil such as clay and the infiltration trench will only work as a time-lagging storage. In that case, the similarity to a pipe trench is substantial (Larm, 1994). IDEQ (2005) consider infiltration trenches to be advantageous in comparison with other BMP’s, such as basins, when it comes to clogging.

2.4 Operation, Efficiency and Performance Clogging is defined as reduced porosity and permeability due to chemical, physical and biological processes (Bouwer, 2002). Due to field studies of stormwater injection wells in aquifers made by Pavelic et al. (1998) and Rinck-Pfeiffer et al. (2000), it can, according to Lindsey et al. (1992), be presumed that physical clogging is the dominating type of clogging in stormwater infiltration systems. Studies of infiltration trenches in Maryland, U.S., in the mid-80’s showed that 33% of the 2-year old infiltration trenches did not function due to clogging (Lindsey et al., 1992). Studies made in the same place in the early 90’s showed that 53% of the facilities did not work as they were designed, 36 % were partly or totally clogged and 22% showed slow filtration (EPA, 1999). The overall experience is that in 50% of the cases, the infiltration trench has a reduced efficiency after 5 years of running (Larm, 1994, EPA, 1999). The estimated lifetime is approximately 10 years (Larm, 1994). The pollutant removal effect of the infiltration trench is relatively unknown, but it is assumed that the facility has got a high removal rate for particle-bound pollutants and a moderate removal rate for soluble pollutants (Larm, 1994). Hatt et al. (2007) found, in a laboratory infiltration trench set-up, that at least 50% of the entering load of heavy metals was trapped in the top 0.30 m of the gravel. Further on, their experiments showed that a 0.5 m deep infiltration trench would remove an adequate amount of sediment and heavy metals and that the pollutant removal rate seems to be independent to the hydraulic loading and the level of clogging of the facility.

4

Since sediment and hydrocarbons may clog the trench, runoff containing these pollutions should preferably be pre-treated by being let over a vegetated filter strips, swales or other kind of sediment traps (EPA, 1999). According to IDEQ (2005), a filter strip should be at least 6 m wide. EPA (1999) sees great similarities between infiltration trenches and rapid infiltration facilities for wastewater when it comes to structure and therefore suggests that the pollutant removal effect should be similar. As can be seen in Table 1, the latter type has proven to have a very high removal rate when it comes to sediment, metals, bacteria and organics. According to EPA (1999), the removal rates for soluble metals should however be expected to be lower than 90% for infiltration trenches. A minimum drainage time of 6 hours is preferable in the infiltration trench in order to get a satisfying pollutant removal. CASQA (2003) claims the removal rate for sediments, nutrients, trash, metals, bacteria, oil and grease as well as organics to be high. As organic matter, clay and loam enters the trench and causes clogging it also increase the removal rate of soluble metals due to adsorption. Table 1. Typical Pollutant Removal Efficiency for Wastewater Infiltration Systems (EPA, 1999).

Pollutant

Typical Removal Rate (%)

Sediment

90

Total Phosphorous

60

Total Nitrogen

60

Metals

90

Bacteria

90

Organics

90

BOD

70-80

Fälth (2008) studied a series of passive filter systems for leachate treatment at a refuse dump, among them a macadam filter, and determined the pollutant removal effect by taking water samples before and after each filtration step. The heavy metals included in the investigation was arsenic (As), cadmium (Cd), chrome (Cr), copper (Cu), mercury (Hg), nickel (Ni), lead (Pb) and zinc (Zn). The reduction of zinc (Zn) turned out to be 6-80 % in the macadam section – the higher figure was however a one-time occurrence that may be caused by a measuring fault. A possible, but not significant, reduction of copper (Cu) and chrome (Cr) was also seen. Among the remaining heavy metals the results were fluctuating, why no conclusions could be drawn.

2.5 Maintenance Already in 1981, Stahre stated that infiltration trenches need continuous maintenance and that this maintenance often is neglected. According to EPA (1999), infiltration trenches should be annually inspected. This inspection should include assuring that the trench drains within the design time by monitoring the observation well. A small section of the top layer should also be removed in order to find out 5

whether the facility is clogged or not. The trench should also be inspected after major rain events and debris should be removed. If the infiltration trench is surrounded by a vegetated filter strip, it should be continuously maintained with respect to mowing and removing of unwanted vegetation, such as trees. Every 5 to 15 years, the infiltration trench needs to be reestablished. EPA has put together a routine for maintenance activities that includes maintenance tasks and frequency of tasks, Table 2. Examples of maintenance tasks are to inspect filter fabric for sediment deposits by removing a small section of the top layer, to mow and trim vegetation around the trench to maintain a neat and orderly appearance and to remove any trash, grass clippings and other debris from the trench perimeter. EPA has also put together a formulary to fill in when inspecting the trenches, Table 3. It includes possible defects, conditions when maintenance is needed and expected results when maintenance is performed. Examples of defects are standing water, trash and debris accumulation and sediments. IDEQ (2005) are stricter in their recommendations regarding monitoring the observation well and the sediment build-up. They claim that an infiltration trench should not only be monitored after rain events greater than 254 mm/24 h during the first year, but also monthly during the wettest months (October 15-April 15 in Idaho) and quarterly during the rest of the year. Table 2. Copy of EPA’s Table of Routine Maintenance Activities for Infiltration Trenches. Routine Maintenance Activities for Infiltration Trenches No.

Maintenance Task

Frequency of Task

1

Remove obstructions, debris and trash from infiltration trench and dispose.

Monthly, or as needed after storm events

2

Inspect trench to ensure that it drains between storms, and within 5 days after rainfall. Check observation well 2-3 days after storm to confirm drainage.

Monthly during wet season, or as needed after storm events

3

Inspect filter fabric for sediment deposits by removing a small section of the top layer.

Annually

4

Monitor observation well to confirm that trench has drained during dry season.

Annually, during dry season

5

Mow and trim vegetation around the trench to maintain a neat and orderly appearance.

As needed

6

Remove any trash, grass clippings and other debris from the trench perimeter and dispose of properly.

As needed

7

Check for erosion at inflow or overflow structures.

As needed

8

Confirm that cap of observation well is sealed.

At every inspection

9

Inspect infiltration trench using the attached inspection checklist.

Monthly, or after large storm events, and after removal of accumulated debris or material

6

Table 3. Copy of EPA’s Inspection and Maintenance List for Infiltration Trenches. Defect

Conditions When Maintenance Is Needed

Maintenance Needed? (Y/N)

Comments (Describe maintenance completed and if needed maintenance was not conducted, note when it will be done)

Results Expected When Maintenance Is Performed

1. Standing Water

When water stands in the infiltration trench between storms and does not drain within 5 days after rainfall.

There should be no areas of standing water once inflow has ceased. Any of the following may apply: sediment or trash blockages removed, improved grade from head to foot of infiltration trench, removed clogging at check dams, or added underdrains.

2. Trash and Debris Accumulation

Trash and debris accumulated in the infiltration trench.

Trash and debris removed from infiltration trench and disposed of properly.

3. Sediment

Evidence of sedimentation in trench. Less than 50% storage volume remaining in sediment traps, forebays or pretreatment swales.

Material removed and disposed of properly so that there is no clogging or blockage.

4. Inlet/Outlet

Inlet/outlet areas clogged with sediment or debris, and/or eroded.

Material removed and disposed of properly so that there is no clogging or blockage in the inlet and outlet areas.

5. Overflow Spillway

Clogged with sediment or debris, and/or eroded.

Material removed and disposed of properly so that there is no clogging or blockage, and trench is restored to design condition.

6. Filter Fabric

Annual inspection, by removing a small section of the top layer, shows sediment accumulation that may lead to trench failure.

Replace filter fabric, as needed, to restore infiltration trench to design condition.

7. Observation Well

Routine monitoring of observation well indicates that trench is not draining within specified time or observation well cap is missing.

Restore trench to design conditions. Observation well cap is sealed.

8. Miscellaneous

Any condition not covered above that needs attention in order for the infiltration trench to function as designed.

Meet the design specifications.

2.6 Costs Clearing, excavation, placement of the geo textile and filling material, installation of the monitoring well, and establishment of a vegetated buffer strip are parts of the capital cost as 7

well as the costs for planning, geotechnical evaluation, engineering and permitting (EPA, 1999). The estimated cost for constructing an infiltration trench in Sweden is 400-500 SEK∙m-1 (questionnaire study, ch. 4). EPA (1999) suggests that the annual maintenance cost is approximately 5-10% of the capital cost and the estimated cost for re-establishing the facility is 15-20% of the capital cost.

2.7 Design 2.7.1 Dimensioning of an Infiltration Trench One way to dimension an infiltration trench is to calculate an effective cross-sectional area of the backfill corresponding to the cross-sectional area of a drainage pipe dimensioned to handle the entire design rain (AMA Anläggning 07, 2008). Dimensioning the trench to be able to house the entire volume of the design rain is also an option (Persson, 2010). A Danish source from 1999 suggests 1 m3 of bulk volume per 30 m2 catchment area, i.e. connected impervious area (Warnaas et al., 1999), whereas VAV (1983) suggests that the infiltration trench should be designed in accordance with a detention facility. Some or all of the following parameters are to be considered when dimensioning an infiltration trench (VAV, 1983): 

Inflow



Discharge



Infiltration/ no infiltration

2.7.1.1 Inflow VV P46 recommends infiltration trenches to be designed for the same design rains as pipe systems (VAV, 1983). In Sweden, the design rains for stormwater pipes are commonly determined in accordance with Table 4. Table 4. Return Periods for Stormwater Pipe Dimensioning (Svenskt Vatten, 2004).

Type of Area

Filled Pipe Dimensioning Stormwater Pipe

Outside city development, gravitational surface drain possible

1 yr

City development, gravitational surface drain possible

2 yr

Outside city development, gravitational surface drain not possible

5 yr

City development, gravitational surface drain not possible

10 yr

8

A design rain is determined from the local conditions of the site. The climatologic analysis made by Dahlström in 1979 is commonly used in Sweden as a template when determining the design volumes. By analyzing the regional, climatologically patterns of intensive convective precipitation in 1931-1960, he developed an intensity/duration formula presenting the rain intensity based on 3 min-96 h duration and 1 month-10yr recurrence period. Based on regional conditions, Sweden was divided into different zones; each assigned a certain Z-value used for calculations (Appendix A). Dahlström (1979) provides i and t for a certain location and corresponding Z-value. Sjögren & Mårtensson (1982) suggests that the block rain given as design rain by using the method above should be increased with 25% in order to cover the precipitation falling right before and right after the block rain. The inflow volume Vin (m3) of an infiltration trench can be calculated in accordance with the following formula (VAV, 1983):

Vin  AT  1.25  i  10 3  t  3600 Where AT (ha) is the effective catchment area, I is the block rain intensity (L∙s-1∙ha-1) and t is the duration of the rain (h). In the U.S., infiltration trenches are held to be suitable for catchment areas less than 4 hectares, but if the contributing area is larger than 2 hectares another BMP should at least be considered (EPA, 1999). In some cases, depending on hydro-geological conditions, the infiltration trenches should be designed not for rain quantities but for snow melting quantities (VAV, 1983). 2.7.1.2 Discharge The drainage pipe diameter controls the flow out of the infiltration trench. Legal requirements – specific for each municipality or even each estate – may regulate the discharge. In Helsingborg the standards are different for different estates (Peetz, 2009). Malmö municipality has a general requirement (Malmö Stad, 2003): 

A real estate should be able to handle a 10-year rain with the duration of 1 hour



The maximum flow to the municipality pipe network may not exceed 20 L∙s-1∙ha-1

2.7.1.3 Infiltration/No Infiltration Infiltration can be assumed to take place in case the permeability of the soil is greater than 2∙10-5 m∙s-1. Since this work is dealing with impermeable Scanian soils, calculations including infiltration are left out. 2.7.1.4 Procedure Dimensioning an infiltration trench in an impermeable soil should be executed as follows (VAV, 1983): 1. A curve describing the relationship between required specific trench volume D (m3∙ha-1) and specific discharge E (L∙s-1∙ha-1) is constructed, Figure 2. The required specific trench 9

volume D is given by the maximum distance between specific inflow volume Vin and the specific discharge E. The required specific volume D is given by:

D  max[ 3.6  1.25  i  t  E  t  3.6]

Figure 2. The required specific trench volume D is given by the maximum distance between specific inflow volume Vin and the specific discharge E (Vestergren, 2010).

2. The effective catchment area AT (ha) is determined 3. The desired mean discharge qmean (L∙s-1) of the infiltration trench is set to a suitable value – commonly the assumption that the mean discharge corresponds to the discharge of a half-full trench can be made. The specific discharge E is calculated trough the mean discharge qmean and the catchment area AT:

E

q mean AT

4. The desired specific trench volume D is received from the curve describing the relationship between required specific trench volume D and specific discharge E 5. The required trench volume is calculated from the required specific trench volume D and the catchment area AT:

Vreq  DAT 6. The dimensions of the facility is has to meet the following relationship:

1.5bhL  10

Vreq n

where n is the effective porosity and b is the base and h is the height, Figure 3.

Figure 3. Infiltration Trench Dimensions (Vestergren, 2010).

2.7.2 Other Construction Considerations AMA Anläggning 07 (2008) suggests that the effective cross-sectional area Aeff (m2) of the backfill should be given by the corresponding cross-sectional area of a drainage pipe dimensioned to handle the entire design rain, Figure 4. That means that Aeff is given by:

where A (m2) is the cross-sectional area and n is the porosity of the filling aggregate. According to VAV P46, the geometrics of the trench are of little importance, since the discharge of the trench happen trough a drainage pipe. In order to prevent a too rapid discharge the drainage pipe can however be dimensioned to throttle the flow (VAV, 1983). Infiltration trenches are commonly 0.9 to 3.6 m deep (EPA, 1999). The side slope of the pit walls are recommended by the Swedish Road Authority to be 4:1, Figure 4 (AMA Anläggning 07, 2008). There are several different ways to determine the volume of the trench. In the U.S., the infiltration trench is commonly designed to handle the first flush, i.e. the first 1.3 cm of runoff, from the catchment area as the examples below show (EPA, 1999): 

In Maryland, the infiltration trenches are designed to handle the first 1.3 cm of runoff from the contributing impervious area



In Washington, the government suggests that the trench should be designed for the first 1.3 cm of runoff from the contributing impervious area or being able to handle a 6.4 cm rain

According to AMA Anläggning 07 (2008), the free horizontal measure dhorizontal between the drainage pipe in the bottom of the trench and the pit wall should be at least 0.35 m. The vertical 11

distance dvertical between water surface corresponding to design rain conditions and the road superstructure is suppose to be 0.30 m in order to protect the durability of the road construction, Figure 4 (VVTK, 2008). The slope of the catchment area should be less than 5% and be such that the run-off is evenly distributed and enters the trench as a sheet flow, Figure 4 (EPA, 1999) – Minnesota Urban Small Sites BMP Manual (2001) recommends slotted curbs to act as level spreaders.

Figure 4. Infiltration Trench Design Parameters: The Cross-sectional Area A, the Maximum Slope of the Catchment, the Free Horizontal Measure dhorizontal and the Vertical Distance dvertical (Vestergren, 2010).

The expected drainage time is varying. In Washington D.C., the infiltration trenches are designed to drain within 72 h (EPA, 1999) whereas IDEQ (2005) recommend a drawdown time of 24 h. In U.K., the standard is 50% empty 24 h after the rain event has ended and in Australia, the recommended drainage time varies from 12-60 h (Browne, Delict et al., 2008). In order to capture the water in the trench it should preferably be placed in a depression (EPA, 1999). 2.7.2.1 Filling Aggregate The recommendations for the diameter of the filling aggregate vary within the U.S., Table 5. In Sweden, the diameter of the filling aggregate in infiltration trenches is, in comparison, smaller. Grain fraction 16-32 mm is commonly used. Göbel et al. (2008) compared different grain sizes and performed experiments with regard to heavy metals. No differences were however found between sand (ø=0-2 mm) and gravel (ø=0-32 mm) of the same geological element, i.e. quartz.

12

Table 5. Size of Filling Aggregate – recommendations.

Source

Size of Aggregate (mm diameter)

EPA, 1999

25-76

CASQA, 2003

38-64

Montgomery County, 2004

38-76

Sweden (commonly)

16-32

2.7.2.1 Observation Well To enable monitoring of the infiltration trench an observation well should be part of the construction. EPA (1999) recommends a 10.2 to 15.2 cm diameter PVC pipe, anchored vertically to a foot plate at the bottom of the trench. According to Montgomery County (2004), it is suitable to have one observation well per approximately 45 m2 trench whilst IDEQ (2005) recommends one observation well every 15 m. 2.7.2.2 Protective Measures during Construction Infiltration trenches, as other stormwater solutions, are commonly constructed during an early phase of the developing of a new area. Heavy traffic and construction work can however cause great damage to the facility in terms of smearing and over-compaction of the filling aggregate and the surrounding soils, Figure 5 (Stahre, 1981). The catchment area should preferably be fully developed and vegetated before constructing the infiltration trench in order to prevent premature clogging (EPA, 1999). Minnesota Urban Small Sites BMP Manual (2001) recommends temporary diversion berms or silt fences as protective measures during the construction. Larsson (2010) mentions two recent projects in Scania where the infiltration trenches have been covered with geotextile and upon that, a load-carrying gravel layer or a temporary binder have been placed during the construction phase. This solution, however, requires a careful removal and replacement of the impermanent solution. According to Horstmark (2010), prospection and careful planning are measures used to protect the infiltration trenches during construction. Waiting to construct the sensitive and exposed parts of the infiltration trenches is also a possible option, but it requires providing lots of information to the concerned parts in order to make an understanding for how the stormwater handling is supposed to work.

13

Figure 5. During the Construction Phase, Infiltration Trenches Run a Great Risk of Errors such as (a) Snow Bank Deposition Causing Clogging, (b) Skittering Traffic Causing Compaction and Smearing or (c) Dislocation of Wells and (d) Cement and Other Construction Materials Causing Clogging (Vestergren, 2010).

2.7.2.3 Different Types of Top-layers In this work, the main focus is on infiltration trenches exclusionary filled with gravel. Different kinds of top-layers do however exist - pea gravel, grassed swales and pervious pavers are commonly put on the top in order to improve the operation in different ways, Figure 6. In order to delay the clogging of the trench, the top-layer of macadam can be substituted with pea gravel. The pea gravel will prevent stowing by trapping sediment and debris more efficiently in the upper layer of the facility (EPA, 1999, Montgomery County, 2004). An infiltration trench can also be placed under a swale in order to increase the available storage volume – commonly called Grass Swale-Perforated Pipe System. According to ASCE (2001), a grass covered infiltration trench should be covered with 0.3 m soil. According to landscape architects at Ramböll Sweden, however, the Scanian municipalities Eslöv, Helsingborg and Lund commonly use a 10 cm soil layer. Geotextile between the soil and the underlying macadam will 14

keep the latter from clogging (ASCE, 2001). Research of long-term operation facilities in urban areas has shown that the few upper cms of soil captures about 90 % of most of the pollution parameters (Sieker, 1998). According to Göbel et al. (2008) replacement of the top-soil needs to be done every 10-20 years due to accumulation of heavy metals. In case of high permeability soil with low adsorbing capacity 10 years is the prevailing lifetime of the top-soil, whilst replacement should be done every 15-20 years in case of low permeability soil with high adsorption capacity. By covering the infiltration trench with pervious pavers (also known as open-cell unit pavers or geoblocks) a load-carrying construction is formed, suitable for parking lots and other trafficked areas. The holes are preferable filled with gravel or macadam, but grass is also an option.

Figure 6. Infiltration Trenches with Different Kinds of Top-Layers: (a) Macadam, (b) Pervious Pavers, (c) Pervious Pavers, (d) Grass (Vestergren, 2009).

15

2.7.3 Design Summary A summary of the design constraints of ASCE (2001), EPA (1996) and Maryland (1999) gives an overview of the American standards for infiltration trenches, Table 6. Table 6. Summary of U.S. Constraints for Infiltration Trenches.

Constraints

ASCE, 2001

EPA, 1996

Maryland, 1999

Contributing Area

-

13 mm/h are recommended

Water 0.6-1.2 m clearance table/Bedrock

> 0.9 m clearance

0.6-1.2 m clearance

Proximity to building foundations

Minimum distance of 3 m down-gradient from buildings and foundations

(local requirements)

Minimum distance of 3 m down-gradient from buildings and foundations

Max depth

-

-

1.8-3.0 m depth depending on soil type

Maintenance

-

Periodic low-level

Moderate to high

16

3 Storm Water Quality The quality of stormwater varies due to the possible sources of pollutants. For example, the contaminants in stormwater from highly trafficked roads differ from the contaminants in water from a housing area – the former is however not necessarily more polluted.

3.1 Pollutants: Heavy Metals In this work, focus lay on seven specific heavy metals, namely: cadmium, chrome, copper, lead, mercury, nickel and zinc. These are chosen partly due to their occurrence in traffic environments and partly due to allowing comparison with former studies. In Table 7, the main sources of these heavy metals in stormwater and their biological effects are presented. All of them, except chrome, appear as divalent cations in solutions. Table 7. Sources of Heavy Metals in Stormwater (Dagvattenstrategi för Malmö, 2008).

Metal

Main Sources of Pollution

Biological Effects

Cadmium

Vehicles, tainted in zinc

Poisonous for humans and animals

Chrome

Vehicles, buildings

Negative effect on humans, animals and vegetation

Copper

Buildings (mainly roofs), vehicles

Poisonous for aquatic animals and vegetation

Lead

Chimney cowls, vehicles

Highly toxic for humans and animals; bio-magnifying

Mercury

-

Highly toxic for humans and animals

Nickel

Traffic, coatings

Negative effect on humans, animals and vegetation

Zinc

Buildings, vehicles, infrastructure

-

The metals can enter the infiltration trench in different ways: 

As solutes



As particle-bound

The classification of the heavy metal content in stormwater is, according to Stockholm Vatten (2001), done as presented in Table 8. The classification of low, medium and high heavy metal content in stormwater varies between species – when it comes to cadmium 1.5 μg∙L-1 is considered a high concentration, while the corresponding number for zinc is 300 μg∙L-1.

17

Table 8. Classification of Heavy Metal Contents in Stormwater (Stockholm Vatten, 2001).

Metal

Low levels

Medium levels

High levels

(μg∙L-1)

(μg∙L-1)

(μg∙L-1)

Cadmium

1.5

Chrome

75

Copper

45

Lead

15

Mercury

0.2

Nickel

225

Zinc

300

Based on a field study, Larm (2000) presents the average content of heavy metals in stormwater from areas with different traffic loads, Table 9. The different categories of roads are divided based on the daily average of cars passing over a year. Table 9. Average, Minimum and Maximum Content of Heavy Metals in Stormwater from Areas of Different Traffic Influences (Larm, 2000). Roads: Daily average over the year.

Unit

City centre

Parking lots

Roads

min

average

max

min

average

max

5000

10000 15000

Cadmium μg∙L-1

0.5

1

2

0.2

0.45

1

0.2

0.3

0.3

Chrome

μg∙L-1

4

5

20

3

15

20

1.0

1.8

2.6

Copper

μg∙L-1

20

30

60

25

40

70

31

51

59

Lead

μg∙L-1

10

40

230

10

30

50

14

17

21

Mercury

μg∙L-1

0.1

0.1

0.4

0.1

0.1

50

0.1

0.1

0.1

Nickel

μg∙L-1

5

10

20

1

4

7

1.15

1.8

2.5

Zinc

μg∙L-1

60

140

400

60

140

300

62

89

116

It is assumed that stormwater in residential areas with one-family households contain high concentrations of pesticides/herbicides compared with stormwater from other parts of the city. Even the streets in a residential area can sometimes pollute the stormwater more than a heavy trafficked highway with respect to copper, zinc and cadmium due to roof materials. Stormwater

18

from parking lots mainly contains pollutants from exhaust gas and wear of brakes, tires and paving (Dagvattenstrategi för Malmö, 2008). The stormwater policy of Helsingborg City assigns different kinds of requirements and attitudes towards stormwater treatment depending on recipient, Table 10. Table 10. Stormwater Treatment Depending on Recipients According to the Stormwater Policy of Helsingborg City (Dagvattenpolicy för Helsingborg stad, 2007).

Level of Pollution

Sensitive Recipient

Less Sensitive Recpient

Little

No treatment

No treatment

Treatment considered

No treatment

Treatment required

Treatment considered

(housing area) Medium (roads with less than 2000 vehicles per day/parking lots/industrial areas) Large (roads with more than 2000 vehicles per day)

3.2 Level of Sorption Related to Grain Fraction 3.2.1 Geotechnology The Swedish natural soils were formed circa 10,000 years ago by the ice sheet. Different processes have caused different kinds of soils and these are often classified by grain fractions, Table 11. Different compositions of fractions give different kinds of physical, technical and chemical characteristics. One characteristic that is dependent on the grain size is the permeability. Clay has a permeability of less than 10-9 m∙s-1. The number is increasing with increasing grain fraction. Fine gravel - such as macadam - has a permeability of 10-1-10-3 m∙s-1, Table 11.

19

Table 11. Grain Fraction and Permeability of Different Kinds of Soils (SGF, 1981, Stahre, 1981).

Fraction

Particle Size

Permeability

Permeability

mm

m/s

m/yr

Gravel

60-2

10-3-10-1

30,000-3,000,000

Sand

2-0.06

10-5-10-2

30-300,000

Silt

0.06-0.002

10-9-10-5

0.03-300

Clay

10-9