technical report volume 2: river modelling report … · technical report volume 2: river modelling...

Comhairle Chontae na Mhí

MEATH COUNTY COUNCIL

Comhairle Chontae Fhine Gall

FINGAL COUNTY COUNCIL

RIVER TOLKA

FLOODING STUDY

TECHNICAL REPORT VOLUME 2:

RIVER MODELLING REPORT

October 2003

River Tolka Flooding Study Technical Report No. 2 River Modelling Report

074515001Rp0032ADE.doc i Rev F01

TABLE OF CONTENTS

1 INTRODUCTION.............................................................................................................................. 1 1.1 APPOINTMENT OF MCOS ............................................................................................................ 1 1.2 PROJECT APPRECIATION ............................................................................................................. 1

1.2.1 Background/Flooding History............................................................................................. 1 1.2.2 Study Context .................................................................................................................... 3

1.3 OBJECTIVES OF THE REPORT....................................................................................................... 3 2 CATCHMENT DESCRIPTION......................................................................................................... 5

2.1 OVERVIEW................................................................................................................................... 5 2.2 RIVER NETWORK......................................................................................................................... 5 2.3 INTERACTION WITH OTHER STUDIES.............................................................................................. 5

3 DATA COLLECTION....................................................................................................................... 6 3.1 REPORTS.................................................................................................................................... 6 3.2 RAINFALL DATA........................................................................................................................... 6 3.3 FLOODS RECORDS/LEVELS.......................................................................................................... 6 3.4 FLOW INFORMATION .................................................................................................................... 7 3.5 PROPOSED RIVER GAUGES........................................................................................................... 8 3.6 TOPOGRAPHICAL DATA / ORTHO-PHOTOGRAPHY .......................................................................... 9

3.6.1 Orthophotography.............................................................................................................. 9 3.6.2 Topographical Survey........................................................................................................ 9

3.7 PIPE NETWORKS ......................................................................................................................... 9 3.7.1 Drainage Networks .......................................................................................................... 10

3.8 FLOOD PROTECTION WORKS / RIVER ALTERATIONS ..................................................................... 11 3.8.1 Historical Works............................................................................................................... 11 3.8.2 Recent / Proposed Works................................................................................................ 11 3.8.3 Study Interim Works ........................................................................................................ 11

3.9 TIDAL INFORMATION .................................................................................................................. 11 4 LAND USE AND PLANNING ........................................................................................................ 12

4.1 EXISTING LAND USE .................................................................................................................. 12 4.2 CONSTRAINTS MAPPING ............................................................................................................ 12

4.2.1 Sites & Monuments; Special Areas of Conservation; Special Protection Areas and Natural Heritage Areas:.................................................................................................................. 12 4.2.2 Catchment Landuse / Corine Data: ................................................................................. 12 4.2.3 GSI Bedrock / Groundwater and Aquifers: ...................................................................... 12

4.3 FUTURE DEVELOPMENT WITHIN THE CATCHMENT ........................................................................ 12 5 HISTORICAL EVENT ANALYSIS ................................................................................................. 14

5.1 ANALYSIS OF FLOOD EVENT OF 14TH – 15TH NOVEMBER 2002...................................................... 14 5.1.1 Rainfall Records............................................................................................................... 14 5.1.2 Affected areas.................................................................................................................. 14 5.1.3 Impact on River Tolka Flood Study ................................................................................. 18

5.2 ANALYSIS OF FLOOD EVENT OF 1954 ......................................................................................... 19 5.2.1 Rainfall Records............................................................................................................... 19 5.2.2 Affected areas.................................................................................................................. 19

5.3 ANALYSIS OF FLOOD EVENT OF NOVEMBER 2000 ....................................................................... 19 5.3.1 Rainfall Records............................................................................................................... 19 5.3.2 Affected areas.................................................................................................................. 19

5.4 FLOOD FREQUENCY ANALYSIS .................................................................................................... 19 5.4.1 Introduction ...................................................................................................................... 19 5.4.2 Hydrological data series .................................................................................................. 20 5.4.3 Fitted Distribution based on Peak Over Threshold Series .............................................. 20


074515001Rp0032ADE.doc ii Rev F01

5.4.4 Annual Maxima Series..................................................................................................... 21 5.4.5 Expansion of Annual Maxima Series ............................................................................... 21 5.4.6 Catchment Characteristics Analysis ................................................................................ 21 5.4.7 Summary.......................................................................................................................... 23

5.5 SEA LEVEL / FLOOD INTERACTION.............................................................................................. 24 5.5.1 Tidal Frequency Analysis................................................................................................. 24 5.5.2 Sea levels ........................................................................................................................ 26

6 LONG TERM CLIMATE CHANGE IMPACTS............................................................................... 28 6.1 INTRODUCTION .......................................................................................................................... 28 6.2 GDSDS PREDICTIONS FOR CLIMATE CHANGE................................................................................ 29 6.3 APPLICATION OF CLIMATE CHANGE TO THE RIVER TOLKA .............................................................. 30 6.4 JOINT PROBABILITY OF TIDAL AND FLUVIAL FLOODING................................................................... 31 6.5 SUMMARY.................................................................................................................................. 31

7 MODEL CONSTRUCTION ............................................................................................................ 32 7.1 OVERVIEW ................................................................................................................................ 32 7.2 MODEL DEVELOPMENT .............................................................................................................. 32

7.2.1 Modelling Software .......................................................................................................... 32 7.2.2 Phased Model Development............................................................................................ 32 7.2.3 User Defined Flags .......................................................................................................... 33

7.3 HYDROLOGICAL MODEL ............................................................................................................. 33 7.3.1 Rainfall/Runoff Model ...................................................................................................... 33 7.3.2 Subcatchment Detail........................................................................................................ 35

7.4 HYDRAULIC MODEL ................................................................................................................... 35 7.4.1 Cross section Representation.......................................................................................... 36 7.4.2 Hydraulic Structure Representation................................................................................. 36 7.4.3 Flood Plain Representation ............................................................................................. 40

8 MODEL CALIBRATION / VERIFICATION.................................................................................... 42 8.1 HYDROLOGIC CALIBRATION........................................................................................................ 42

8.1.1 November 2002 ............................................................................................................... 42 8.1.2 November 2000 ............................................................................................................... 43 8.1.3 August 1986..................................................................................................................... 44

8.2 HYDRAULIC CALIBRATION ........................................................................................................... 44 8.2.1 November 2002 ............................................................................................................... 44 8.2.2 November 2000 ............................................................................................................... 46

9 DESIGN EVENT ANALYSIS ......................................................................................................... 47 9.1 DESIGN MODEL......................................................................................................................... 47

9.1.1 Design Rainfall................................................................................................................. 47 9.1.2 Critical Duration ............................................................................................................... 47

9.2 EXISTING................................................................................................................................... 47 9.2.1 River Flooding.................................................................................................................. 47 9.2.2 Tidal Flooding .................................................................................................................. 48 9.2.3 Tidal Gate Analysis .......................................................................................................... 50

9.3 FUTURE DESIGN ........................................................................................................................ 50 9.3.1 Urban Development ......................................................................................................... 50 9.3.2 Urban Development & Climate Change .......................................................................... 50

9.4 SENSITIVITY ANALYSIS............................................................................................................... 51 9.4.1 Sensitivity Analysis of Roughness Coefficients ............................................................... 52 9.4.2 Sensitivity Analysis of Percentage Development ............................................................ 52

10 CATCHMENT MANAGEMENT.................................................................................................. 54 10.1 FUTURE CATCHMENT MANAGEMENT FRAMEWORK .................................................................... 54 10.2 FLOODPLAIN MANAGEMENT FOR THE RIVER TOLKA ................................................................... 55

10.2.1 Floodplain Mapping ......................................................................................................... 55 10.2.2 Flood Awareness and Emergency Planning.................................................................... 55 10.2.3 Flood Forecasting and Warning....................................................................................... 56 10.2.4 Local Flood Protection ..................................................................................................... 56


074515001Rp0032ADE.doc iii Rev F01

10.2.5 Maintenance of Watercourses ......................................................................................... 57 10.2.6 Sustainable Drainage Systems........................................................................................ 58 10.2.7 Flood Risk Concepts........................................................................................................ 58 10.2.8 Planning and Building Control ......................................................................................... 60 10.2.9 Summary.......................................................................................................................... 60

11 CATCHMENT MANAGEMENT ALLEVIATION OPTIONS ....................................................... 62 11.1 OVERVIEW OF OPTIONS CONSIDERED...................................................................................... 62 11.2 RIVER IMPROVEMENT WORKS................................................................................................. 63 11.3 FLOOD DEFENCE WORKS....................................................................................................... 64 11.4 RIVER DIVERSIONS ................................................................................................................ 65 11.5 ATTENUATION........................................................................................................................ 65 11.6 SUMMARY OF FLOOD MANAGEMENT OPTIONS CONSIDERED....................................................... 68

12 PROPSED FLOOD ALLEVIATION SCHEME........................................................................... 70 12.1 OVERVIEW ............................................................................................................................. 70 12.2 MEATH COUNTY COUNCIL AREA – PROPOSED WORKS............................................................... 71

12.2.1 Upstream of Bennetstown ............................................................................................... 71 12.2.2 Works at Bennetstown..................................................................................................... 71 12.2.3 Bracetown ........................................................................................................................ 72 12.2.4 Gunnock House (River tolka)........................................................................................... 72 12.2.5 Dunboyne Town Castle Stream Reach; .......................................................................... 73 12.2.6 Loughsallagh to Clonee ................................................................................................... 74 12.2.7 Clonee Village Area ......................................................................................................... 75

12.3 FINGAL COUNTY COUNCIL AREA – PROPOSED WORKS............................................................ 76 12.3.1 Huntstown to Parlickstown............................................................................................... 76 12.3.2 Littlepace.......................................................................................................................... 78 12.3.3 Mulhuddart – Blanchardstown (Map Sheet 5, Final Report) ........................................... 78 12.3.4 Blanchardstown to Finglas Section ................................................................................. 79

12.4 DUBLIN CITY COUNCIL AREA – PROPOSED WORKS .................................................................... 80 12.4.1 Finglas Road to Glasnevin Bridge River Section............................................................. 80 12.4.2 Glasnevin Bridge to Dean Swift bridge........................................................................... 80 12.4.3 Dean Swift Bridge to Drumcondra Road Bridge .............................................................. 81 12.4.4 Drumcondra Road Bridge to Luke Kelly Bridge............................................................... 82

12.5 GENERAL RECOMMENDATIONS................................................................................................ 85 13 MODEL HANDOVER ................................................................................................................. 86

14 CONCLUSION............................................................................................................................ 87


074515001Rp0032ADE.doc iv Rev F01

LIST OF APPENDICES

APPENDIX A Catchment Overview Plan APPENDIX B Data Collection Appendix B1 Relevant Reports Appendix B2 Rainfall Data Appendix B3 Available Flood Records Appendix B4 Data Collection Figures APPENDIX C Model Construction Appendix C1 Sub-catchment Details Appendix C2 Flow Control Structures Appendix C3 Spatial Isohyetals APPENDIX D Calibration and Verification Appendix D1 Flow Calibration Appendix D2 Level Calibration Appendix D3 Design Event Analysis APPENDIX E Flood Management Assessment

LIST OF FIGURES Figure 1.1 Extent of Flooding on 6/11/00 looking downstream from Loughsallagh Bridge Figure 1.2 Comparative photograph, looking downstream from Loughsallagh Bridge Figure 3.1 1986 Flood – Botanic Avenue Chart 5.1a Fitted Distributions – EV1 Chart 5.1b Fitted Distributions – EV2 Chart 5.2 Comparison of growth curves Figure 5.1 February 2002 Tide Levels Figure 5.2 Tide / Flooding Coincidence Figure 5.3 Tide Levels for Dublin Bay (Malin Head) Figure 7.1 Interpolated Section in InfoWorks along the Pinkeen Stream Figure 7.2 Schematic of an Arch Bridge at Damastown Figure 7.3 Schematic of a USBPR Bridge at Annesley Road Figure 7.4 Triple Barrel Culvert at Blanchardstown Road North Figure 7.5 Schematic of an Orifice on the Pinkeen Stream Figure 7.6 Weir at Glasnevin Road Figure 7.7 Dean Swift Bridge modelled as a Bernoulli Loss Figure 7.8 Temporary works along the Tolka during the DART bridge construction,

modelled by a general headloss Figure 7.9 Floodplain Storage in Clonee Figure 9.1 Design Tidal Run Comparison on Phase 1 Model Figure 9.2 Design Tidal Run on Proposed Scheme


074515001Rp0032ADE.doc v Rev F01

Figure 10.1 Source, Pathway, Receptor Model with Local Flood Protection Measures

LIST OF TABLES Table 1.1 Available Historical Flood Data Table 4.1 Catchment Development Table 5.1 Adjustment Factor Comparisons Table 5.2 Summary of Catchment Flow Analysis Table 5.3 Results of Analysis Table 5.4 Sea Level Rise for Dublin Bay Table 7.1 Catchment Characteristics input in the model Table 7.2 Sub-catchment Detail of River Tolka Table 7.3 Classification of Structures used in InfoWorks RS Table 8.1 Rainfall Data input for Urban Catchments Table 8.2 Calibration Comparison for 14th & 15th November 2002 Table 8.3 Calibration Comparison for 6th November 2000 Table 9.1 Hydrograph Scaling Factors Table 10.2 Risk Concepts Table 11.1 Summary analysis of principle flood relief options Table 11.2 Impact of Attenuation Options upon Peak Flood Levels and Flows for November

2002 flood


074515001Rp0032ADE.doc 1 Rev F01

1 INTRODUCTION

1.1 APPOINTMENT OF RPS-MCOS In 2001 Dublin City Council, acting on behalf of the Local Authorities in the Greater Dublin area, commissioned a major study of foul and storm drainage needs in the area having regard to current needs and for development horizons to 2011 and 2031. The work was undertaken by a Consultancy Consortium known as the Dublin Drainage Consultancy. This work will culminate in 2003 in the publication of the Greater Dublin Strategic Drainage Study (GDSDS).

As an extension of this study, the River Tolka Flooding Study was commissioned by Dublin City Council, in association with Fingal County Council, Meath County Council and the Office of Public Works in 2002. The study arose from concerns regarding increased flooding risk to properties along the River Tolka arising from a significant flood in November 2000, when many properties were inundated particularly in parts of Meath and in the Dublin City Council area. The River Tolka Study has been carried out by M.C. O’Sullivan & Co. Ltd. (RPS-MCOS), on behalf of the Dublin Drainage Consultancy.

1.2 PROJECT APPRECIATION

1.2.1 Background/Flooding History The Tolka has a long flat, mainly rural catchment with minimal reaction to short duration heavy rainfall storms. However, the Tolka does have a significant history of flooding after long steady rainfalls, with lands adjacent to the river inundated a number of times during the last century. The extent of the flooding during the 6/11/00 flood, estimated as a one in twenty year event, is shown in Figure 1.1. The long flooding history is reflected in the Tolka’s Gaelic name ‘An Tulca’ which means The Flood.

Figure 1.1 - Extent of Flooding on 6/11/00 looking downstream from Loughsallagh Bridge


074515001Rp0032ADE.doc 2 Rev F01

Figure 1.2 - Comparative photograph, looking downstream from Loughsallagh Bridge

The November, 2000 flood revised public awareness of the acute flood risk from the River Tolka to properties along the river, including relatively recent developments. As a result, this study was commissioned to analyse the problem and examine solutions. From the available data Table 1.1 lists recorded floods occurring since 1880. A full hydrological assessment and results of historical flood analysis are included in Chapter 5

Table 1.1: Available Historical Flood Data

Ranking Date Estimated Flow at the outlet of

the Tolka

Ranking Date Estimated Flow at the outlet of the

Tolka 1 14/15th November 2002 97m3/s 10 20th September 19461 48 m³/s

2 8th December 19541 85 m³/s 11 23rd November 18981 45 m³/s

3 6th November 20002 76 m³/s 12 12th November 19151 42 m³/s

4 28th October 1880 71 m³/s 13 3rd April 1909 37 m³/s

5 19652 59 m³/s 14 5th February 1946 Minor Flood

6 26th August 19863 57 m³/s 15 3rd January 1948 Minor Flood

7 12th November 19011 57 m³/s 16 19th December 1932 Minor Flood

8 1st September 19311 54 m³/s 17 17th November 1916 Minor Flood

9 19682 49 m³/s

This table shows that the major flooding has generally been confined to the period between late August and early December. Due to the flatness of the catchment and its retentive vegetation, spring and summer storms have seldom produced flows of any major significance. Recent flood history (2000 flood in particular) indicated that the catchment could be classified in terms of flood risk generally in accordance with the following:

1 From 1955 Dublin Corporation Report on the 1954 Flood estimated using debris line at weir upstream of Finglas Bridge. 2 Recorded at Botanic Gardens station. The rating curve has been developed for flows up to 87m3/s, flows in excess of this should be treated with caution. 3 Recorded at Drumcondra Station.


074515001Rp0032ADE.doc 3 Rev F01

�� Flooding had occurred in the town of Dunboyne on the Castle Stream branch of the River

Tolka which required that this branch be modelled in detail. Door to door surveys indicated that a number of properties not flooded during the November 2000 event were close to being flooded and could be considered at risk to flooding during an event of larger magnitude. In addition to this, following the event, a number of recent developments had been constructed in low lying areas, which could be at risk.

�� The sensitivity to flooding in the Clonee area needed to be examined together with lands

zoned for development in Clonee and Dunboyne areas. Areas upstream of Clonee Bridge have a history of flooding; areas downstream of the bridge were flooded from a secondary flow path due to exceedance of the river capacity in this area.

�� Through Mulhuddart and Blanchardstown, much of the flood plain of the River Tolka has

been maintained free from development (with the exception of Gleeson’s pub at Mulhuddart Bridge) Therefore flooding associated with river levels did not appear to be a significant risk to existing developed areas.

�� In the DCC area, a significant developed area of low lying lands below the Botanic

Gardens had a propensity for flooding. Careful model construction and analysis was required through Glasnevin and Drumcondra including consideration of numerous bridges, weir structures and other potential controls in the channel in this area to assess flood risk in what is a complex system. Between Drumcondra Bridge and East Point Business Park, the Tolka needed to be considered in terms of inter-tidal conditions with historical references to flooding in the Clonliffe / Ballybough area.

1.2.2 Study Context The recently experienced strong economic growth in the Irish economy, the continuing development of the national road network, availability of public transport services and zoning of land for industrial and commercial development to provide sustainable employment for new populations, has been fuelling the continuing expansion of the Dublin area and is resulting in major development pressure in the greater Dublin area and surrounding counties. Currently land use within the upper catchment is a combination of lightly forested and agricultural land. In the middle catchment residential and commercial development competes with agricultural use but an increasing amount of land is being developed for residential and commercial use. In the lower catchment, the land use is almost entirely residential, public parks and commercial. As the lower and middle catchments in particular, and certain concentrated areas within the upper catchment, are witnessing significant increases in land use change, the risk of flood damage to property and infrastructure is likely to have increased. Prior to this study, the lower river reaches had been the subject of a number of investigations the most significant of which were those carried out in 1955 and 1986. In both cases the study reports identified and recommended specific remedial measures to be taken that would alleviate or reduce the incidence of flooding. While some of the 1955 proposals were implemented, post 1955 recommendations were not implemented as the downstream effects could not be quantified. In November 2000, the river discharge again reached flood levels and a significant amount of damage was sustained by developments in the upper and middle catchments and by certain areas in the lower catchment that have a propensity for flood damage. There had never been a fully inclusive analysis of the river’s behaviour under a full range of potential floods and certainly there has been no investigation of the catchment’s response to severe rainfall as it is currently developed and as planned to be developed. In recognition of these pressures, Dublin City Council, in association with Fingal County Council, Meath County Council and the Office of Public Works commissioned this comprehensive study.

1.3 OBJECTIVES OF THE REPORT The objective of the Report is to describe the comprehensive flood analysis of the River Tolka, Castle Stream and Twin Pinkeen Streams and how it developed, from a modelling perspective. The analysis included the development of a robust hydraulic and hydrological mathematical model representing the characteristics of the River Tolka catchment. The model has the facility to map flood risk for design


074515001Rp0032ADE.doc 4 Rev F01

rainfall events. Using model output and taking account of climate change and of current and future land developments. From this analysis a flood relief scheme has been developed to protect existing property. Recognising the downstream effects of improvement works, the impacts of local flood alleviation measures have been assessed in the context of the entire river system and measures recommended for mitigating these impacts. Additionally catchment wide schemes such as attenuation and diversions have been assessed. The scheme is considered to offer a sensible level of protection for all areas currently developed. This has involved assessment for the November 15th, 2002 flood and the traditional once in 100 year flood (1% risk of occurrence in any year), to cater for development to 2031 and a reasonable degree of climate change, with a reasonable margin of freeboard. A detailed technical description of the model construction process is outlined. The accuracy of the flow and level calibration and sensitivity to model parameters is assessed and the results are portrayed in the report.


074515001Rp0032ADE.doc 5 Rev F01

2 CATCHMENT DESCRIPTION

2.1 OVERVIEW The Tolka River is the second largest river to enter Dublin City in terms of its length and catchment area, after the Liffey. As depicted in Figure A1.1, Appendix A, the river rises west of the city in County Meath near the Culmullin Cross Roads on the R125 roadway and is fed by a network of small tributaries as it flows through Batterstown, Rathbeggan, Quarryland, Piercetown, Blackbull, Dunboyne, Clonee, Mulhuddart, Blanchardstown, Finglas Bridge, Glasnevin, Drumcondra, North Strand and East Wall to enter the sea east of the DART depot at Fairview Park. It has a catchment area of 14,150 hectares, and drops 140m in 33.3 kilometres. This makes it a relatively flat river, with modest gradient from source to sea. The upper-catchment can be described as predominantly rural. The river in this area is little more than a stream with small meanders and low banks with a relatively flat bed gradient of about 0.4%. It is roughly 2.5 m wide near Batterstown and 5 m wide just upstream of Clonee Bridge. Despite this, the upper catchment is subject to occasional flooding with the flood plain extending up to 400 meters wide in places.

2.2 RIVER NETWORK The profile of the river changes noticeably as it drops from an open, rural catchment upstream of Clonee into the developing urban environments of Mulhuddart, Blanchardstown and Ashtown. Through the formalised Tolka Valley Park, Botanic Gardens and Griffith Park, it becomes somewhat wider and straighter, with generally higher and more defined grass banks. In its latter reaches through Glasnevin, Drumcondra and Marino, the river becomes increasingly canalised, which is characteristic of many inner city rivers where development extends almost to bank top. In this section, the riverbank varies from natural riverbank to an ad-hoc arrangement of walls of varying height and robustness. Below Drumcondra the river is also subject to tidal influence and the channel is wider with more formal riverside walls in the lower section.

2.3 INTERACTION WITH OTHER STUDIES In parallel with the river study, the GDSDS study includes modelling of the stormwater drainage networks connecting to the River Tolka. These studies will include consideration of the interaction between the local drainage networks and river flood conditions in order that secondary flooding risk through backpounding of the pipe networks is dealt with. The ultimate integrated flood relief scheme for the River Tolka must comprise of the recommended river engineering works, pipe network upgrading works as well as the overall catchment management measures proposed.


074515001Rp0032ADE.doc 6 Rev F01

3 DATA COLLECTION A significant amount of data has been procured and investigated; this data is detailed in the following sections.

3.1 REPORTS A number of relevant reports have been procured in relation to the study. These reports are detailed in appendix B -1, along with a brief synopsis.

3.2 RAINFALL DATA Rainfall data shown in Figure B – 2.1, Appendix B.2 is available from a number of meteorological stations located in and around the catchment area. Most historic stations record only daily rainfall with the exception of Dublin Airport and Casement Aerodrome. In addition to this, DCC have detailed station data dating back 5 years for four locations, and an additional 4 stations have been installed in the last 2 years. Rainfall data relevant to the study is included in Appendix B - 2

3.3 FLOODS RECORDS/LEVELS A number of sources for flood levels and records have been investigated. These include:

�� 14th/15th November 2002 Recorded levels during the flood, debris levels, surveyed levels provided by DCC, MCC and FCC, OPW and FCC video footage and photographs.

�� 31st November 2002 Recorded flow levels – In channel flood �� 6th November 2000 M3 study recorded levels during the flood, photographs and MCC, photo’s

and Levels �� 6th November 2000 Dunboyne aerial photography and levels �� 6th November 2000 DCC video and photo’s �� 1986 Slides – Drumcondra and Distillery Weir (c.f Figure 3.1) �� Anecdotal evidence from reports and visits to affected residents �� Dublin Corporation – Reported levels from 1955 flooding report. The 1954 album of photos is

missing Anecdotal evidence and flood level data is critically important in building a profile of a flood event and can greatly assist in understanding of flood mechanisms. A degree of caution is needed with regard to accuracy of data and interpretation due to:

�� Changes in flood levels/flows during a flood �� Influence of local factors

�� Wave effects due to vehicle influence

�� Interactions with the piped sewerage systems and blockages of screens, pipe inlets, etc, due

to debris

Large flood events have a major impact on those affected and very good information can be derived from local people including information on historical frequency, severity, timing, recorded level information and local changes to the drainage and river system over the years.


074515001Rp0032ADE.doc 7 Rev F01

This information can be particularly useful in assessing variations in levels that can be achieved in secondary flow paths and as a result of obstructions such as local blockages and boundary walls between properties. Flood debris levels also should be treated with caution as level variations may be recorded due to pile up of debris providing overestimation of flood levels and floating debris which settles at a lower level providing an underestimation. However an experienced hydraulic engineer can usually make a good estimate when viewing flood debris and taking account of the location, type of debris and local conditions. It is therefore particularly important to collect level information directly following a flood when the amount and quality of information is best. The available flood records are included in Appendix B3.

Figure 3.1- 1986 Flood - Botanic Avenue

.

3.4 FLOW INFORMATION A number of gauges have been installed on the river to monitor flows and levels. These include:

�� Botanical Gardens Station 09037 - EPA provided records from Sept 1999. A near continuous record is available. Prior to the 15th of November, 2002 flows recorded in excess of 30.5m3/s were based on an extrapolated rating curve. However on 15th November calibration was carried out by EPA staff and the rating curve is now developed for flows up to 87 m3/s. This should be noted when reviewing historical reports.

�� Drumcondra Bridge, Station 09019 - Records were received. The records are very intermittent and on review it was shown that all records with levels above mean flows have been removed. The report received, Summary of Hydrometric records: July 1977. An Foras Forbartha, indicates that due to silting and vegetation growth, the continuous records were not satisfactory for determining mean and minimum flows, although rare flood flows could be established. Records of use received include an estimate of the flood hydrograph for the 1954 flood at this station, and the flood hydrograph for the 1986 flood.

�� The K C O’Donnell report on the 1986 flooding has a rough sheet appended indicating annual peak flows on the Tolka River at Drumcondra. The peak flows are included for the years 1954-1969, 1976-1986, though records from 1970 to 1975 are not available.


074515001Rp0032ADE.doc 8 Rev F01

�� Batterstown Staff gauge 09018 - EPA provided records from March 1985 to Sept 1990. These are spot checks at low flows and are not useful to this study.

�� Griffith Park – some flow calibration was done using staff gauge, although this was at very low flows and is not useful for this study

�� Cardiff’s Bridge has some old records. Again these are at low flows only and are not useful for this study

�� MCOS water quality assessment recorded low flows only. (River Tolka draft Water Quality Management Plan)

In summary there is very little high quality flow information available to enable calibration to flow hydrographs within the catchment. There is however some good quality, high flow hydrograph records available in the lower catchment and these were used in calibration of the entire model. There are also a number of level records throughout the catchment. It is often found in modelling rivers that level calibration is of most importance. This is due to factors such as aerial floodplain increase and flood stage velocity increase, whereby large increases in flow can translate into moderate increases in level. Therefore, level information is the issue of most importance as flooding / no flooding are a function of level.

3.5 PROPOSED RIVER GAUGES A report on the suitability of providing two gauges located on the river in addition to the existing gauge at the Botanic Gardens and the existing tide station at Dublin Port was issued to Dublin City Council. It is proposed that these gauges would be linked to the Council’s telemetry system. In addition to this it is proposed that the flow monitoring stations be utilised for long term monitoring in conjunction with environmental parameters. The river was investigated taking into account the following general site criteria.

�� Site Security The river gauges should be located on private or local authority land with no public access or areas where access is restricted due to geographical location. If the gauging station is located on private land consideration has to be given to the negotiation of rights of access for installation and commissioning of the gauging stations and for regular maintenance.

�� Positioning of River Gauges The River Gauges should be positioned in areas where river debris will not dislodge the equipment or unduly affect water levels. If existing low flow weirs are used the infrastructure is already in place to ensure good river bank and river bed conditions as well as ensuring that the river gauges aren’t placed above the dry weather flow. The gauges should not be placed in locations that are subject to downstream influences causing the gauges to drown out during flooding events. �� Upstream / Downstream Riverbank condition The upstream and downstream condition of the river bank should be smooth to reduce the possibility of drag and to assist the free movement of water along the river profile. �� River Bed Conditions The river bed should be smooth with a gradual incline to assist in obtaining a smooth flow and should not be prone to scouring or depositing of excessive river material.

Due to the relatively low gradient of the river and the extensive flood plains, only 2 general areas meeting the above requirements were identified as suitable for providing additional flow information relative to the study and include: 1) Cardiff’s Bridge, Ashtown (Dublin City Council, Fingal County Council boundaries). This location

is the current catchment boundary between the semi-rural and urban developed areas. The recommended site in this location is Finglas (Factory) weir, 100m u/s of the road bridge on the N2 (Finglas Road). This site has a weir with a head difference of approx. 2.5m between upstream and downstream. While this site may seem a comparatively long distance downstream from Cardiff’s Bridge, the increase in catchment area is small, it is upstream of the confluence of the


074515001Rp0032ADE.doc 9 Rev F01

Finglas River and will not be influenced by it, as the Finglas river joins the Tolka River downstream of the weir.

2) N3 Clonee (Fingal County Council, Meath County Council boundaries). This area is the first

suitable site offering smooth flow conditions below the confluence of the upper catchment tributaries. The recommended site is downstream of the Clonee by-pass alongside the Kepak premises. A Non Standard Control structure would be required at this location if monitoring of low flows is also required. This location is well upstream of the next culvert, and is less likely to be effected by backwater than other locations in this area.

These proposed sites could be utilised in conjunction with the existing monitoring sites located at: a) Dublin Port, which provides tidal records for the hydraulic study and upon investigation by the project team was found to be satisfactory for the project needs, (this station was connected to Dublin City Council’s telemetry system in 2003) and; b) Botanic Gardens, which is currently maintained by the EPA and was also found to be satisfactory for the project needs (following satisfactory extension of the rating curve to 87m3/s).

3.6 TOPOGRAPHICAL DATA / ORTHO-PHOTOGRAPHY

3.6.1 Orthophotography Latest Ordnance Survey V1000 and V2500 mapping have been assessed and details of unmapped recent developments have been included, where available, from the relevant local Authority Building Control sections. The N3 Tolka study topographical Digital Terain Model (DTM) and Ortho-photography was utilised for the Meath Area along with the N3 Tolka study - cross-sections and Bridge sections of Tolka and Castle Stream (Dunboyne). A LIDAR DTM and Digital Surface Model (DSM) was produced for the remaining length of the Tolka and the tributaries modelled. Ortho-photography is available for almost the entire study area.

3.6.2 Topographical Survey A topographical survey for the remaining length of the river and the Scribblestown and Twin Pinkeen streams was carried out by BKS and subcontracted by them to Longdin and Browning. This survey included river cross–sections, bridge details, outfall pipe locations and house levels for areas initially identified as at risk to flooding. Photographs of cross sections and flow control structures were also provided. Cross sections were taken approximately every 50m, from left bank to right bank position looking downstream, and extending to the floodplain where necessary. Drawings and data files were provided. The data file was in ASCII Format so a direct import into InfoWorks RS was possible. An example of a cross section taken in ASCII format is shown in Appendix B – 4, Table B.4.1. The writing in italics is for information purposes only. Flow control structures were surveyed immediately upstream and downstream, extending to the floodplain, defining its profile and showing additional points where any significant change in level occurs.

3.7 PIPE NETWORKS Storm water pipe network requirements are generally being dealt with in the GDSDS storm model catchments, in addition to this: Dublin City Council provided updated digital Network details. These included a detail sheet of Distillery Weir. Hard copy storm water maps were provided by Fingal County Council dated 1990. Network details are not available for Dunboyne and Clonee. Outfalls were initially identified from the draft River Tolka Water Quality Management Plan report (MCOS) and river surveys have included invert levels and diameters for these and other outfalls identified. Wad River Diversion Plans (Ballymun Development) circa 1968 were received from Dublin City Council. The parallel drainage network studies in the GDSDS will enable the interaction of the network and the river system to be examined. Management of this interaction is a critical element of a secure River Tolka flood defence system in the future.


074515001Rp0032ADE.doc 10 Rev F01

3.7.1 Drainage Networks There are 6 main drainage networks entering the Tolka River. These are catchments 30-35, see Figure C1.2, Appendix C.1. Catchment 31 (Green) is the smallest region within the Tolka NDDS catchment at 235 ha bounded by the Finglas River to the north and east, the Scribblestown Stream to the west and the Tolka River to the south. It drains across the steep sided Tolka Valley Park via the Finglaswood Stream, the Scribblestown Stream and numerous smaller piped systems in the lower reaches. The largest ouflow from this catchment is the Finglaswood Stream and outfalls via a box culvert in Tolka Valley Park. 4 smaller catchments outfall in this area via 300-600mm Ø pipes. Catchment 32 (Red) is 373 ha, lies to the south of the Tolka River. It is bounded approximately by Botanic Rd to the east, the Royal Canal / Carnlough Rd / Navan Rd to the south and Ashtown Rd to the west. The Cemetery Drain that services Glasnevin (Prospect) Cemetery in Drumcondra is the only named watercourse in this region. The outfalls consist of a 600m Ø pipe outfalling at Glasnevin Bridge, a 450mm Ø pipe outfalling between Botanic Road and Mobhi Road, a 750mm Ø pipe outfalling at Finglas Road, several 450-600mm Ø pipes between Ratoath Road and Ashtown, and a 450mm Ø pipe at Tolka Valley Park. Catchment 33 (Orange) is 241 ha and lies between the Wad River Diversion to the east and the Finglas River to the west. This drains to the Tolka from a maximum high point of about 68 m MHD at Finglas East predominantly via the culverted Claremont Stream. It enters the Tolka through a 1370 x 1670 mm box culvert, 19m downstream of Glasnevin Bridge. Three other subcatchments of Catchment 33 enter the Tolka via 300-450mm Ø pipes. This region comprises of well-established urban areas mixed with a good proportion of institutional lands and sports fields in the middle to lower reaches. It is crossed by the regional and main roads of Finglas Rd Old, Griffith Avenue, Glasnevin Av, Ballygall Rd East / Beneavin Dr / Willow Park Cr / Sycamore Rd and Glasanon Rd. Catchment 34 (Blue) is 536 ha and the eastern most region bounded by the Wad River Diversion to the west, the Wad River catchment proper to the north and east, the Tolka River to the south and the river estuary to the south-east. This region drains from a maximum high point of about 48 m MHD near Dublin City University at the Collins Avenue watershed in the north to the Tolka River and estuary via several watercourses that have now all been culverted and often diverted from their original course. The most significant watercourses are the Marino Stream, 1050mm Ø pipe discharging at Fairview Park, Grace Park Stream, a 1350mm Ø pipe discharging south of Richmond Rd, Drumcondra Rd Upper / Swords Rd drain, 375mm Ø pipe discharging at Drumcondra Bridge and the Hamsptead Stream: 600mm Ø pipe, discharging downstream of Mobhi Rd. Catchment 35(Finglas River) is 1080 ha runs from north of the M50 to the River Tolka at the Finglas Road. The section of the catchment north of the M50 is rural, whereas the majority of the catchment south of the M50 is urbanised. The Finglas area is served by a partly separate sewerage system, the surface water system is composed of several branches draining to the Finglas River. There are 7 CSO’s within the foul/combined system, which discharge to the surface water sewers or directly to the Finglas River. Catchment 23 and 23A (Blanchardstown/Mulhuddart) is 843 ha in size and incorporates the drainage network from Blanchardstown and Mulhuddart. The majority of the catchment is urbanised. The main outfalls from this network are two 900mm Ø pipes in Tolka Valley Park, one upstream and one downstream of the Northern Cross Route, a 1050mm Ø and 450mm Ø pipe, discharging at Mill Road and James Connolly Memorial Hospital respectively, a 1000mm Ø pipe discharging behind Main Street, Blanchardstown, a 525mm Ø pipe discharging upstream of Snugborough Rd culverts, a 1050mm Ø pipe discharging upstream of Blanchardstown Road North and a 1200mm Ø pipe incorporating Mulhuddart, discharging upstream of Mulhuddart bridge. There are smaller outfalls from this catchment including 4 outfalls ranging from 225mm to 450mm Ø pipes. In addition to these formal culverted and piped drainage networks, rural and semi rural catchments have been identified and included in the model. These catchments are identified in Figure C1.1, Appendix C1.


074515001Rp0032ADE.doc 11 Rev F01

3.8 FLOOD PROTECTION WORKS / RIVER ALTERATIONS A number of river alterations have occurred since the 1954 major flooding event. These alterations have been included in the various models where necessary to ensure accurate representation of past existing and future events.

3.8.1 Historical Works �� The Glasnevin Bridge was rebuilt and completed in April 1968. �� River works were undertaken from the Estuary to Ballybough and completed 1966. �� Sections of East Wall have been reclaimed for the business park. �� The residents of Tolka cottages were relocated and the cottages were removed. The area

was subsequently built up and a park provided. �� A section of the river that ran where Botanic Avenue now exists was realigned through Griffith

park. �� Various minor bed cleaning works have been undertaken. �� Various culverts have been installed with the construction of the N3 motorway. �� Channel works were undertaken at the property surrounding Flat House Bridge. �� A number of developments have occurred in and around the wider flood plain. �� The Wad river diversion was undertaken circa 1968 as part of the Ballymun developments.

3.8.2 Recent / Proposed Works �� A section of the river in Dunboyne was excavated in 2001, following the November 2000 flood

event. New cross sections in this area were surveyed for insertion in the model for analysis of future flood risk.

�� Irish Rail has upgraded the bridge over the Tolka at East Wall Road in 2002 altering the hydraulic details in this area.

�� Works are underway on a proposed new bridge at the Dublin Port Tunnel site. �� A new pedestrian bridge was constructed downstream of Mulhuddart bridge during the study

period (no information was available for this bridge and it is not included in the model). �� A small bridge is proposed at Clonee for access to a new development. �� A bridge to James Connolly Memorial Hospital is proposed as part of the M50 upgrade to the

N3 interchange. �� Diversions and Culverts are proposed as part of the N3 Clonee to Dunshaughlin Road

Scheme.

3.8.3 Study Interim Works A number of interim works, maintenance cleaning and damage repair works have been undertaken or are proposed and these are dealt with separately within the report.

3.9 TIDAL INFORMATION Alexandra Quay tidal gauge records have been received from Dublin Port Authority for the period 1923 to date. Additional records were requested for the period 1844 – 1922 but Dublin Port Authority advised that these records were unreliable. The 1923 to 2002 records have been digitised and converted from L.A.T datum to Malin Head Datum. A sea level of some 2.95m (MH) was recorded on 1st Feb 2002 at 14.30 hours, the highest reading on record. This unprecedented event resulted in East Wall Road/North Strand (a section of the study area) being flooded as a result of the tidal section of the Royal Canal escaping from its channel. No flooding was reported in the tidal reaches of the River Tolka. Thus, in 2002, two separate independent flood events took place of unprecedented magnitude in the Tolka catchment, a major fluvial flood event and a major tidal event. A comprehensive analysis of all available tidal data outside the scope of this study is being undertaken for the Dublin Coastal Flood Risk Assessment Study (DCFRAS).


074515001Rp0032ADE.doc 12 Rev F01

4 LAND USE AND PLANNING

4.1 EXISTING LAND USE A comprehensive population and land use study has been undertaken as part of the Greater Dublin Strategic Drainage Study. One element of this land use study involved the assessment of the current development situation (2002) in the Greater Dublin Area through the construction of a database which identifies sites that have been developed since the production of the most recent 1:1000 OS mapping. The database also categorises the type of development at each site into residential, industrial, commercial, mixed use or science and technology. For the construction of the river model, this database, used in conjunction with the County Development Plans and the OS mapping provides a comprehensive knowledge of land use in the catchment. This information is illustrated in Figure B4.1, Appendix B4. Future Land Use scenarios (2011, 2031) are discussed in Section 4.3 of the report. The Corrine Land Use mapping in Figure B - 4.2, Appendix B4 also gives an overview of land use in the area.

4.2 CONSTRAINTS MAPPING Various data sets outlined below have been displayed in GIS map format to be used later in the study for identification of possible constraints to solution options and include;

4.2.1 Sites & Monuments; Special Areas of Conservation; Special Protection Areas and Natural Heritage Areas:

Figure B – 4.3, Appendix B4 displays the most up to date information available from the Dúchas website on the various protected and sensitive areas in the catchment. This was downloaded from the website in ArcInfo format and then converted in house to MapInfo format.

4.2.2 Catchment Landuse / Corine Data: This data shown in Figure B – 4.2, Appendix B4 is developed from an interpretation of Satellite Images from 1989-90 and gives a broad picture of land use during this period. Imagery Software has been used to define the land use depending on the pixels colours. Areas less than 0.25Ha are interpreted only. This data is relatively old and broad, however is included for information purposes. Land use patterns developed in the GDSDS have been used for model input data.

4.2.3 GSI Bedrock / Groundwater and Aquifers: Details of the geology, groundwater and aquifers in the catchment from the Geological Survey Ireland are shown in Appendix B4 - Figure B4.4 and Figure B4.5.

4.3 FUTURE DEVELOPMENT WITHIN THE CATCHMENT The GDSDS called for population and land use analysis for the three design scenarios of 2002 (existing), 2011 and 2031. A current development database was created for the 2002 scenario. The 2031 planning horizon is well beyond the periods considered by Local Authority Development Plans (2004/2005); The Strategic Planning Guidelines (SPG) (2011); the Dublin Transport Office’s Platform for change (2016). It was therefore required to carry out sensitivity analysis on the population projections which were supplied in the Strategic Planning Guidelines and the National Spatial Strategy to estimate the 2011 and 2031 populations for each county. The county development plans, industrial land use surveys, county housing strategies and biannual housing surveys were then used to identify land requirements to accommodate these increases in population. Indicative land area requirements


074515001Rp0032ADE.doc 13 Rev F01

can be assessed. These same design scenarios are used in the Tolka Flood Study. The resulting development scenarios are illustrated in Appendix B4 - Figure B4.6. It can be seen that the majority of the future development planned for the River Tolka catchment is proposed in Fingal County with significant development of Greenfield lands in this section of the catchment. Other areas of future development include infill and re-development to higher densities (esp. along future transport routes), in the Dublin City Council areas of the catchment as well as Rail and Road projects. Also envisaged is further development of Clonee and Dunboyne in County Meath along with the proposed N3 National Primary Route and associated interceptor roads. With the exception of the developments in Meath outlined above and possible future development in Batterstown, information to date indicates that the upper catchment will remain relatively unchanged and rural in nature. However, further sensitivity analysis was undertaken during the assessment of hydraulic model forecasts for the future scenarios. This was undertaken to ensure the affects of varying land use on future river levels could be accommodated in the scheme proposed. For further comprehensive details of Land use and planning the reader is referred to the GDSDS Final Report on Population and Land Use4 and The National Spatial Strategy5. Table 4.1 gives a brief summary of percentage of catchment developed and estimated to be developed for the various key dates utilised in this study.

Table 4.1 Catchment Development

Year Catchment Area (km2)

Area Developed (km2)

Percentage Developed (%)

1800’s 141.1 0 0

1900 141.1 2.0 1.4

1955 141.1 11.5 8.2

2002 141.1 27.7 19.6

2011 141.1 41.9 29.6

2031 141.1 49.0 34.6 These figures have been used in the development of the hydrological assessment of the catchment, and are refined during the modelling study to include detailed urban percentages for the sub-catchments. (c.f. Appendix C – Figure C1.1 for a detailed map indicating sub catchments and model parameters) The rate of development between 2002 and 2011 is significantly greater than that between 2011 and 2031. These estimates are partly due to a greater percentage of the catchment in Fingal County being recognised as fully developed by 2011 and therefore further development will be restricted (Blanchardstwown/Mulhuddart). In addition the SPG includes policies which limit development outside of the Dublin Metropolitan area to Primary and Secondary development centres. Since Meath is outside of the Metropolitan area and there are no development centres within the Tolka Catchment, future development here “will be limited to meeting local needs.” Notwithstanding this objective, sensitivity analysis was carried out using the modelling for alternative development scenarios.

4 Greater Dublin Strategic Drainage – Population and Land Use, Final Report January 2003, Dublin Drainage Consultancy, GDSDS/NE02057/094v1. 5 The National Spatial Strategy, Final Report, October 2001, Jonathan Blackwell and Associates in association with Roger Tym & Partners.


074515001Rp0032ADE.doc 14 Rev F01

5 HISTORICAL EVENT ANALYSIS

5.1 ANALYSIS OF FLOOD EVENT OF 14TH – 15TH NOVEMBER 2002

5.1.1 Rainfall Records The flood of 15th November, 2002, followed 2 days of very heavy rainfall. A previous rainfall event on 8th-10th November had resulted in a very wet catchment which, combined with Winter vegetation conditions, meant that a very high level of run-off took place on the 15th November, with little, if any infiltration/soakage into the ground. Rainfall data for Dublin Airport and Casement weather stations showed that rainfall commenced in the afternoon/evening of 13th November and reached peaks of 8.7mm/hr at Dublin Airport in the afternoon of 14th November and 8.3mm/hr at Casement Aerodrome at 1.00am on 15th November. The total rainfall depth measured at Dublin Airport over 53 hours was 87mm. 72 mm of rainfall was recorded over a period of 35 hours at Casement. Met Eireann estimate that a rainfall of this duration has a return period of approximately 50 years. However, subsequent investigation of Dublin Airport radar information combined with sub-catchment calibration indicates that some areas within the catchment sustained even heavier rainfall, with over 100mm of rain indicated as falling in the Dunboyne area during this period. Met Eireann estimate that rainfall of this quantity based on Dunshaughlin data to be in excess of 100 year return period. In summary, the November 2002 flood flows in the River Tolka were the result of a major rainfall event in circumstances which characteristically give rise to serious flooding in the River Tolka, namely, a rainfall duration of 1.5 – 2 days of appropriate intensity and with preceding wet catchment conditions. The modelling studies showed that recorded rainfalls (Dublin Airport and Casement) had to have been significantly exceeded to produce the flood flows at Dunboyne and this was confirmed by reference to radar data.

5.1.2 Affected areas The extents of flooding on the 15th – 16th November 2002 is illustrated on 26 aerial maps provided in Appendix B of the Final Report, beginning at Batterstown, ending at the ocean outfall at East Wall and are based upon site observations, surveyed debris levels and modelling results. The February 2002 tidal flooding is also depicted on these maps while the 1954 flood is illustrated on 4 additional maps and are provided for reference in Appendix C of the Final Report. It should be noted that large areas of secondary flooding occurred outside of the analysed and represented flood plain in low lying areas. Where this affected property and information is available these are generally noted within the report.

The Historical Flood Maps must only be used in conjunction with the Notes and Disclaimers provided in the Final Report and must not be used in isolation to avoid incorrect interpretation of the data provided.

Dublin City Council Administrative Area The flood mechanism in this area was almost identical to that experienced in 1954. However, in 1954 the low-lying areas of Ballybough, North Strand, East Wall and Fairview were more severely impacted due in part to the lower level of flood defence walls in these areas at that time (the raising of these walls was subsequently undertaken from the Estuary to Ballybough and completed in 1966). Flood levels in 1954 are also thought to have been influenced by the partial collapse of the CIE Bridge, which anecdotal evidence suggests became blocked with debris. Additionally the rebuilding of Glasnevin Bridge completed in 1968 has reduced the risk of the overflow path to floodwaters in this area and no overflows occurred via this route in November 2002. The following is an overview of the flood experience in 2002 between Finglas Road and Fairview:


074515001Rp0032ADE.doc 15 Rev F01

�� Finglas Factory Weir Area: Flooding of basement car park occurred at Tolka Vale due to the outlet of the storm water pumping system being inundated preventing discharge against the high flood level (refer to Frame 24, Appendix B, Final Report).

�� Finglas Bridge to Glasnevin Bridge: No property damage occurred in this relatively steep valley section of the river (refer to Frame 24 & 25, Appendix B, Final Report). However erosion of river banks took place due to high water velocities, and additionally significant alluvial material was deposited downstream of Finglas bridge during the flood recession. Some damage was sustained in the Botanical gardens with a number of mature trees being undermined, erosion to banks and walls and general damage to planted areas.

�� Glasnevin Bridge to Deans Swift Bridge: This short reach of the Tolka extends from Glasnevin Road Bridge located on Glasnevin Hill Road to Deans Swift Bridge located on Mobhi Road (refer to Frame 25, Appendix B, Final Report). No substantial property flooding occurred immediately upstream of Glasnevin Bridge. However immediately downstream of the bridge, the river overtopped its right bank. The resulting overbank spill caused localised flooding at the school and properties immediately adjacent. Flood water from this channel breach continued down Botanic Avenue towards Drumcondra Road Lower where it added to the flooding within that area.

�� Deans Swift Bridge to Griffith Park: From the Deans Swift Bridge to Griffith Park, flood flows were conveyed within the banks of the River Tolka (refer to Frame 25, Appendix B, Final Report). The residential properties located adjacent to the river over this reach are protected by flood walls and are generally above flood level. Any partial floodwaters in this area arise from overflows upstream described above. The river kept within its walls and floodplain of Griffith Park within the upper portion of this reach and prevented overbank flows from the Tolka reaching Botanic Avenue.The River began to break its banks and flow onto Botanic Avenue just downstream of the residential properties (85 – 155 Botanic Avenue) within the southern side of Griffith Park, near the Intersection of Botanic Avenue and Mannix Road. Downstream of this location, the area was completely inundated by overbank flows. From anecdotal evidence, including debris left at the entrance to the park, flood waters spilled through the entrance and over the low (300 – 400 mm high) roadside concrete wall. Flood waters from Griffith Park flowing into Botanic Avenue inundated the roadway and residential properties within this area.

The Woodville Road Footbridge at the downstream extent of this reach formed a constriction to main channel discharge capacity by trapping debris in its railings. The river narrows in this area with residential property boundaries located either side further confining the flood flows. At the footbridge, overbank flood flows can discharge to both North and South of the river. Flows spilling north of the footbridge exited to Millmount Road causing inundation of properties on this road. As the area filled up, flood water continued further downstream crossing the Drumcondra Road Lower and Millmount intersection into Clonturk Park and continuing down Richmond Road. Floodwaters overflowing South of Woodville Road Footbridge added to flooding in the Lower Botanic Avenue Area.

�� Woodville Road Footbridge to Drumcondra Road Bridge: Extensive flooding occurred within the reach upstream of Drumcondra Road Bridge inundating residential properties in the vicinity of Botanic Road and Millmount Road (refer to Frame 26, Appendix B, Final Report). In addition to the floodwaters described above, the River overtopped its banks within the old Tolka Park Cottages area inundating these grounds and Botanic Avenue. Flow entered onto Botanic Avenue by overtopping the outer perimeter wall and also flowed freely through the entrance to the gardens. As flood waters receded, flows from Botanic Avenue flowed back to the River Tolka via the lower area at Drumcondra Road.

While wall levels on the south bank were high enough to prevent inundation of flood waters within this reach, the inundation of properties behind the wall occurred from the unprotected area leading to the Woodville Road Footbridge and from Botanic Road due to flood waters escaping from Griffith Park. Wall levels immediately downstream of the footbridge on the north bank were below flood level and thus direct inundation occurred to properties behind this wall from the river. Furthermore, surcharging of local drainage pipe systems backpounded manholes and road gullies, this added to flooding of low-lying areas.


074515001Rp0032ADE.doc 16 Rev F01

�� Drumcondra Road Bridge to Richmond Road Industrial area: Extensive flooding also occurred within this reach from the Drumcondra Road to the Industrial Area located just downstream of Tolka Park (refer to Frame 26, Appendix B, Final Report). Flood waters also extended back to the low-lying residential areas surrounding Clonturk Park. The River Tolka is bounded by residential property immediately downstream of Drumcondra Road. On the south bank, both walls and natural land levels form limits to flooding in this region. On the north bank floodwaters reached the top of the embankment level behind Nos. 4-52 Richmond Road. However, no significant overtopping occurred. Flooding of properties over this section occurred from the overtopping of the walls at 60 to 68 Richmond Road, including the entrance to Tolka Park, and was further exacerbated by flows bypassing Drumcondra Bridge which had overflowed upstream from the Tolka at Griffith Park and bypassed the river downstream via Millmount Road.

�� Richmond road industrial area to Distillery Road Bridge: Distillery Road Bridge is located midway between the Luke Kelly Road Bridge (Ballybough Road) and Drumcondra Bridge (Drumcondra Road Lower) (refer to Frame 26, Appendix B, Final Report). Sections of the North Bank of this reach are below flood level with intermittent low walls allowing waters to enter the industrial area. On the south bank floodwaters are able to flow onto the river floodplain within the Holy Cross College sports grounds. Distillery Road Bridge was also overtopped allowing floodwaters to escape into areas on both sides of the bridge. Levels in this area were increased due to debris becoming entrapped in the bridge structure. Extensive flooding occurred on both sides of the river in the lower portion of this reach with the adjacent industrial and residential areas being completely inundated. Flooding occurred within the new apartment complex from floodwater entering across its western boundary and overtopping the flood wall upstream of Distillery weir. In addition to the out of channel flood flows described above, the industrial and residential areas located on either side of Distillery Road Bridge were flooded as a result of the unprotected area leading to the Distillery Road Bridge and via gaps and low points on the industrial side river wall.

�� Tolka Road: No significant flooding occurred within Tolka Road (refer to Frame 27, Appendix B, Final Report) due to sand bagging on the road preventing flood water entering from Distillery Road and the river remaining within its banks. However flood flows from Distillery road continued downstream to the low-lying areas of Ballybough, where basement flooding occurred. Significant scouring of the riverbank occurred resulting in undermining and failure of the river wall and loss of a backyard shed. The scouring occurred due to increased velocities caused by in-stream blockages from trees growing in the river.

�� Fairview Strand: Low lying areas in Fairview were inundated (refer to Frame 27, Appendix B, Final Report). This is likely to be due to inundation of storm water outlets servicing this area (back-pounding of the drainage system). �� Distillery Road Bridge to Outlet: No Flooding problem occurred in this area from the Tolka River. However Luke Kelly Bridge became surcharged due to debris build up partially blocking its waterway. This resulted in vibrations to the Bridge which caused some traffic concerns. No restrictions in flow were evident from recent works to the river in this area.

Overall, there was extensive property flooding because the flows exceeded the “in-bank” capacity of the Tolka through the DCC area, with some local factors (channel obstructions) contributing and with areas of secondary flooding due to backpounding of the piped drainage system in low-lying areas. Fingal County Council Administrative Area For much of its length in the Fingal County Council area, the Tolka is contained within an undeveloped linear park with development set well back from the river. This contrasts with the older developed areas in Dublin City Council where developments extend to riverbank. As a result, the extent of development at risk in the floodplain in the Fingal County area is relatively limited. Much of the floodplain surveyed after the 2002 flood provides safe storage of floodwaters, away from developed areas. This provides attenuation of river flows through the reach and is affective in reducing flood flows to the downstream reaches within Dublin City Council administrative area. Nevertheless, some river flooding of properties did occur in the Fingal section as summarised below:


074515001Rp0032ADE.doc 17 Rev F01

�� Littlepace: Littlepace is located approximately 600m east of Clonee along the Navan R156 Road and was flooded in conjunction with the flooding of Clonee (refer to Frame 13 & 15, Appendix B, Final Report). Substantial residential property flooding occurred within Littlepace along with flooding of the N3 road in this area. Flooding here was largely caused by Tolka overbank flows occurring upstream at Clonee along with backpounding from a tributary stream.

�� Damastown: This reach of the River Tolka is centred between the N3 and Damastown Culverts (refer to Frame 13 & 15, Appendix B, Final Report). No habitable property was inundated although two commercial properties were considered at risk (i.e. marginal floor level clearance over flood levels).

�� Damastown North: This reach centred on the Pinkeen tributary exhibited no direct property flooding. Commercial properties were at risk in this area during the flood (refer to Frame 15, Appendix B, Final Report).

�� Westpoint Business Park – Parslickstown: This reach of the River Tolka is centred between the Damastown Culverts and Westpoint Business Park (refer to Frame 15, Appendix B, Final Report). A number of the business park properties and a section of the N3 road were inundated within this area.

�� Mulhuddart: This reach extends from Westpoint Business Park culverts to Mulhuddart Bridge Road located on Church Road (refer to Frame 17, Appendix B, Final Report). Commercial properties and residential properties currently under construction were inundated near Church Road.

�� N3 – Blanchardstown: This reach extends from the Blanchardstown Road crossing to Snugborough Road Bridge (refer to Frame 18 & 19, Appendix B). Significant flooding occurred on the N3 National Primary Road, effectively closing the road.

�� Herbert Road – Blanchardstown: This reach extends from Snugborough Road Bridge to the Mill Road Bridge (refer to Frame 19, Appendix B, Final Report). Minor flooding occurred to several residential properties located on the north side of Herbert Road.

The flood mapping of the November 2002 flood, based on the anecdotal evidence collected, shows the extent of the floodplain in the Fingal area and the properties at risk. In summary, the floodwaters largely inundated the undeveloped floodplain over this section with local flooding of properties at a number of sites and extensive flooding of the N3 National Primary Road at Blanchardstown. Meath County Council Administrative Area In County Meath, the areas affected include Clonee on the borders with Fingal and the wide floodplain upstream on the River Tolka and its major Castle stream tributary which flows through Dunboyne. The impact of the 2002 flood in the Meath Administrative area is summarised below: �� Rural Catchment: Significant flooding occurred in the flat low lying areas of the rural catchment,

affecting agriculture, local roads and land flooding. Some local property flooding occurred in the Bennetstown area as a result of local drainage systems (refer to Frame 13 & 15, Appendix B, Final Report).

�� Bracetown (River Tolka): Relating to the reach from the intersection of the N3 and Navan Road

(R157) to downstream of Navan Bridge crossing of the Tolka (refer to Frame 6 & 7, Appendix B, Final Report). Several rural properties and access roads were flooded in this area.

�� Gunnocks House (River Tolka): Relating to a reach of approximately 1.2km of the River Tolka

from approximately 500m downstream of the R157/Navan Bridge to 300m upstream of Loughsallagh Bridge (refer to Frame 8, Appendix B, Final Report). Minor flooding of the N3 occurred, with significant inundation of the rural floodplain in this area. Secondary flooding of Gunnocks House occurred due to local drainage systems.

�� Dunboyne (Castle Stream): Relating to the reach of the Castle Stream through Dunboyne

between a point downstream of the disused railway bridge to upstream of the Newtown Bridge (refer to Frame 9 & 10, Appendix B, Final Report). Significant land flooding occurred within this reach affecting low-lying areas of Dunboyne, flooding a large number of residential properties and impacting on the village itself.


074515001Rp0032ADE.doc 18 Rev F01

�� Loughsallagh (River Tolka): Covering a 1.1km reach from Loughsallagh Bridge to Clonee

Bridge. Flooding of local properties occurred adjacent the R156 and in the floodplain in this low-lying flat area (refer to Frame 10 & 11, Appendix B, Final Report).

�� Clonee (River Tolka): extending from a point upstream of Clonee Bridge to the overpass bridge

to Damastown industrial area, downstream of Kepak (refer to Frame 11 & 13, Appendix B, Final Report). Extensive Property flooding occurred within Clonee, from overflowing of the Tolka upstream of the N3 culvert.

The floodplain mapping for the 2002 flood in the Meath County Council area is based on substantial flood level records and anecdotal evidence of areas inundated. It shows significant inundation of residential areas in both Dunboyne and Clonee. The flood mapping demonstrates that this event, corresponding with record (i.e. greater than 100 year 48 hour) levels of rainfall in the catchment area, resulted in concentrated flood storage at the confluence of the Tolka and Castle tributary, with downstream flows crossing the N3 limited by the capacity of the N3 culvert. The flooding mechanisms were inundation of the properties in the floodplain at Dunboyne, including marginal flooding of the commercial village area directly from the Castle stream, with the combined Tolka flows inundating Clonee village and flowing onwards to Littlepace.

5.1.3 Impact on River Tolka Flood Study The impact of this flood event on the River Tolka Study was to provide an extreme event baseline condition against which the mathematical model could be reliably validated. The study team directed considerable effort and resources to obtaining accurate information on flooding mechanisms and flood water levels. The model was then set up, tested and calibrated in an iterative process to replicate the rainfall/run-off and flooding response in the river system to match the recorded flood conditions. This has a number of benefits including:

�� It facilitates a more complete understanding of the flooding in each area including the performance of the river structures along the route;

�� It facilitates consideration of flood levels and conditions for a range of forecast flood events of different severity; different combinations of rainfall, tide levels and antecedent catchment conditions;

�� It facilitates consideration of the impact on flood levels of development in the catchment generally or in the floodplain of the river including changes in storage;

�� It facilitates consideration of improvements which might be carried out by way of channel enlargement, enlargement of culverts and bridges, bypass or storage attenuation options or other measures to reduce flooding risk. The model can simulate such improvements and critically, allows assessment of their impact on flood levels and flood conditions upstream and downstream.

The November 2002 flood event in the River Tolka was unprecedented in terms of its scale having regard to all known flooding records in the catchment. It produced flood flows and corresponding flood levels in the Dublin City Council area which were approximately 10% greater in scale than the highest previously recorded flood conditions associated with the 1954 flood. At the same time, the scale and extent of flooding upstream in the catchment, particularly in the Dunboyne/Clonee area at the confluence of the Tolka and Castle Stream has been demonstrated to correspond to rainfall and run-off conditions well in excess of previous design standards. As a result, the previous design norms for river bridges and culverts in this area require to be re-evaluated in the light of the November 2002 experience. This experience will also influence the management of future development in the catchment having regard to the spatial extent and storage volume of floodwaters in the area.


074515001Rp0032ADE.doc 19 Rev F01

5.2 ANALYSIS OF FLOOD EVENT OF 1954

5.2.1 Rainfall Records The second largest flood ever recorded on the Tolka occurred on 8th December 1954. The week preceding was not remarkable as regards rainfall at Dublin, but there was a relatively high rainfall at Dunshaughlin for this period. 69.3mm of rainfall fell in 17 hours on the upper catchment (Dunshaughlin) with 57.8mm of rain falling on the lower part of the catchment (Dublin Airport). The flow estimated at the Finglas Weir for the 1954 event is 85m3/s.

5.2.2 Affected areas The 1954 storm caused severe flooding with properties in excess of 1,000 damaged. The worst affected areas were those in the Drumcondra and Ballybough areas. Floodwaters broke out just upstream of Deanswift Bridge causing a flow path down Botanic Avenue and inundating houses in this area with floodwaters. Grifffith Park is also shown to be a weak section of the river, with flood waters entering Botanic Avenue, Millmount Avenue and Richmond Road through the park. The collapse of the DART Road Bridge added greatly to the flooding in the East Wall and Fairview areas. The improvement scheme set out in Dublin Corporation 1954 Flooding Report involved the improvement of the river from Glasnevin Bridge to the estuary at East Wall Road, as the flooding that occurred upstream of Glasnevin Bridge traversed open land at that time and therefore did not impair serious damage to properties. Works in this report were not specifically carried out; however available records indicate works as described in section 3.8 have been undertaken in the intervening years. Additionally improvements to levels of protection have been made to properties which have been redeveloped in the period since the flood (boundary walls raised). For a detailed analysis of the 1954 flooding see Dublin Corporation 1954 Flooding Report.

5.3 ANALYSIS OF FLOOD EVENT OF NOVEMBER 2000

5.3.1 Rainfall Records The flood of 6th November 2000 is the 3rd largest flood ever recorded on the Tolka. A flow of 76m3/s was recorded at the Botanic Gardens station. Hourly rainfall is available for Dublin Airport and Casemont Aerodrome for this period and total rainfall for the entire event is available for Leixlip and Dunshaughlin. 78.3mm of rainfall was recorded at Dublin Airport for a period of 40 hours beginning at 8am on 5th November. 95.3mm of rainfall was measured at Casemont Aerodrome, Baldonnel for the same period. 90.8mm and 76.3mm of rainfall was measured at Cellbridge and Dunshaughlin respectively.

5.3.2 Affected areas The worst affected areas from the November 2000 flood were parts of the upper catchment, in particular Clonee and Dunboyne. Many new houses in Beechdale Estate, Dunboyne and Main Street, Clonee flooded in November 2000 and again in November 2002. 62 properties in total were flooded from this event.

5.4 FLOOD FREQUENCY ANALYSIS

5.4.1 Introduction Flood estimates are required for the design and economic appraisal of flood mitigation measures and engineering protection works for the River Tolka. In advance of modelling, it is therefore important to attain an understanding of likely floods which will be exceeded at any given site on a given frequency (for example on average once in 100 years or 1% risk on average in any year). Two main approaches have been used in this study; 1) the statistical analysis of historical flood series and 2) investigation of rainfall data sets and resulting runoff related to catchment characteristics. By using both methods an understanding of the historical reaction of the catchment to rainfall events can be developed to ensure that specific parameters within the model and their influence are understood.


074515001Rp0032ADE.doc 20 Rev F01

In general these parameters are catchment area, basin slope, soils, catchment wetness, climate, rainfall intensity and duration and time scale of flood events. Flood hydrographs incorporating flood volume and time to concentration are also required in order to assess storage affects, constraints, the coincidence of tides, and the duration and depths of flooding at given areas. Existing event data analysis therefore forms the basis for the modelling strategy, whereby more comprehensive catchment rainfall runoff models are utilised and calibration and validation undertaken.

5.4.2 Hydrological data series The data available is limited to the lower catchment area in Dublin City Council. This data can generally be described in three distinct subsets:

1. Data from 1880 to 1955 consisting of eleven (11) estimated peak flows for flooding events from analysis undertaken in 1955. This data set represents only events which provided a large flow in the Tolka and therefore could be considered as a ‘Peak over threshold’ series. The accuracy of this data could be considered low, however the analysis of the 1954 event was comprehensive and this flow estimate of 85m3/s could be considered to have fair accuracy.

2. Thirty one (31) annual peak flow records from Drumcondra Bridge, station 09019 for the period 1955 to 1986. The accuracy of these records could be considered fair.

3. Three (3) annual peak flow records from Botanic gardens Station 09037 for 2000, 2001 and 2002. (Data revaluated with updated rating curve following November 2002 event). The accuracy of the data could be considered excellent.

Therefore intermittent records of varying reliability exist, containing 41 peak annual flows spanning a period of 122 years. This data provides a fair representation of the scale and frequency of floods in the Tolka dating back to 1880. It is unlikely that floods of greater severity have occurred during this period without note.

5.4.3 Fitted Distribution based on Peak Over Threshold Series It was determined that a threshold flow of approximately 35 m3/s has historically resulted in floods that have been of severity enough to warrant record in newspapers. This gives a series from 1880 to 2002 equal to 122 years. Using the assumption that the above flows are the maximum in this period, and then the maximum values in the 122 year series are ordered including the top 14 post 1880 peaks This series enables the development of a fitted distribution and determination of the a and u parameters from the records where x=u+ay, which then allows for fitting of the extreme value type 1 (Gumbel) distribution.

Chart 5.1 (a&b) Fitted Distributions

Fitted EV2 Distribution (k=-0.05)

y = 0.9816e0.0561x

1

10

100

1000

0 20 40 60 80 100 120

Flow

Ret

urn

Per

iod

Fitted EV1 Distribution

y = 14.61x + 6.7159

0

20

40

60

80

100

0.00 1.00 2.00 3.00 4.00 5.00 6.00

Y

Flo

w (

m3 /

s)


074515001Rp0032ADE.doc 21 Rev F01

From the results of the EV1 distribution the EV2 distribution can be determined whereby x = u+aW and W = (1–e-ky)/k. Here k is the regional shape parameter = -0.05 for Ireland. (The estimate for k based on the short record was found to be -0.047). The results of this method provide a fitted curve to historical recorded floods that are assumed to exceed all intermediate unrecorded ones and hence could be considered overly conservative. Generally, this data, based on a curve fitting for only 14 peak records could be considered skewed.

5.4.4 Annual Maxima Series An annual maximum series was developed for the full available record of 34 years. The Extreme Value Type 2 analysis of this series provided a good fit with catchment characteristics estimate of mean annual flood, and what could be considered a conservative estimate for higher return period events.

5.4.5 Expansion of Annual Maxima Series A number of methods are available for the extension of short records. The flood studies report generally recommends extension of very short records based on rainfall correlation with adjacent gauged catchments. Unfortunately work undertaken by J H Clark in 1955 indicates that a very poor correlation between rainfall and flooding exists in the Tolka catchment, it is noted that high rainfall (100mm) during summer months does not cause flooding however as little as 50mm during winter months can cause flooding. Additionally surrounding catchments generally have a more rapid response. Therefore it is not considered that an adequate model could be developed to improve the reliability of existing data. For similar reasons the method of extension of the record based on a regional estimate was not considered appropriate. The above series was expanded using a number of random samples to fill the missing annual maximum records, on the assumption that unrecorded years would have a reasonably even distribution of flood peaks for a period of 120 years ranging in severity up to the threshold developed above. This was undertaken to perform a final check on the skew resulting from sampling choice and distribution errors.

5.4.6 Catchment Characteristics Analysis A number of possible hydrological model strategies are available for determining model input hydrographs. The most widely accepted in Ireland are those based on FSR techniques which uses Catchment characteristics (CC) methods. To provide initial comparative data CC analysis was undertaken including sensitivity to various parameters and design flows. For this catchment, CC method appears to under predict flows when compared to historical data. This method uses Irish regional multipliers, for the determination of various return period floods, based on the derived mean annual flood. It is interesting to note that the catchment specific growth curve for the Tolka catchment is similar to the UK growth curve 6/7. There are now 9 regional curves for the UK (FSSR 146) and it is instructive that they are all steeper than Ireland and the eastern ones are steeper than the western regions. On comparison of the Tolka curve against the three eastern UK curves (values are provided in Table 5.1), it could be inferred that Dublin reflects the east – west differences seen in UK. This is an important result as from 1975 following FSR many bridges in the catchment are likely to have been generally designed in accordance with this parameter.

Table 5.1 Adjustment factor comparisons

Adjustment Factor Return Period 5 10 25 50 100 200

Ireland 1.20 1.37 1.60 1.77 1.96 2.14UK (standard) 1.22 1.48 1.88 2.22 2.61 3.06UK Curve 6/7 1.3 1.6 2.13 2.75 3.13 3.75UK Curve 5 1.32 1.65 2.21 3 3.5 4.2Tolka Design 1.4 1.8 2.3 2.8 3.3 3.6

6 Natural Environment Research Council, UK, - Flood Studies Supplementary Reports (1977-1988) No. 14:


074515001Rp0032ADE.doc 22 Rev F01

When choosing a design curve for the analysis of flooding in the River Tolka the following points require consideration;

�� A degree of conservatism is required to ensure that adequate levels of protection are provided by any proposed scheme

�� Additional conservatism is inbuilt to the analysis by the use of the design rainfall event which uses a normal distribution and therefore generally provides conservative water levels particularly upstream of flow constraints. This is due to high volumes centred on the peak of the event. (Twin peaked events may however provide higher levels in certain specific circumstances)

�� The design uses conservative rainfall parameters including a critical duration event and a 95% rainfall aerial reduction factor for the entire catchment.

�� If the design curve is too conservative it may not represent actual conditions and the financial benefits for any scheme may be affected. This could result in a lower level of protection becoming financially viable.

�� The positive benefits of implementation of sustainable drainage systems are currently difficult to model and therefore these benefits are not regarded when assessing future development, this provides an additional degree of conservatism, however if the design curve is too conservative, future implementation of SuDS may provide a lesser degree of attenuation due to the increased outflow allowed to mimic existing conditions.

�� The design curve must also take account of the errors associated with such an analysis and the limited reliability of the data.

The design curve must therefore be based on a multi criteria assessment of the data using a degree of engineering judgement.

Table 5.2 Summary of Catchment Flow Analysis Analysis Annotation Return Period Qm Q5 Q10 Q25 Q50 Q100 Q200 Catchment Characteristics -standard parameters associated with Tolka catchment CC 19 23 26 30 34 37 41 Catchment Characteristics Assuming Soil Class 5 CCs5 38 45 51 60 66 74 80 Catchment Characteristics Assuming Soil moisture defecit = 0 CC (SMD 0) 23 27 31 36 40 45 49 Catchment Characteristics, including urbanisation adjustments CC Urban 28 34 38 42 44 47 47 Catchment Characteristics, including urbanisation adjustments and assuming soil class 5 CC Urban (s5) 55 66 74 83 87 92 93 Catchment Characteristics, including urbanisation adjustments and assuming Soil moisture deficit = 0

CC Urban (SMD 0) 33 40 45 50 53 56 57

Gumbel Analysis without Nov 2002 EV1 (2001) 26 38 48 60 69 79 88 Gumbel Analysis with November 2002 included EV1 (2003) 14 32 47 66 80 93 107 Extreme Value Analysis type 2 with November 2002 included EV2 (2003) 14 33 50 71 87 104 122 Gumbel Analysis including only series above 35m3/s

EV1 (2003 POT series) 29 45 57 74 86 97 109

Extreme Value Analysis type 2 including only series above 35m3/s

EV2 (2003 POT series) 29 46 60 78 92 107 122

Gumbel Analysis expanding to 122 years

EV1 (2003 expanded series) 24 36 45 57 66 75 84

Extreme Value Analysis type 2 expanding series to 122 years

EV2 (2003 expanded series) 24 36 47 61 71 82 94

Error band Min 14 28 39 53 63 73 81 Error band Max 29 46 63 78 92 107 122 DESIGN DESIGN 28 38 50 63 77 91 101


074515001Rp0032ADE.doc 23 Rev F01

5.4.7 Summary The results of the analyses described above are tabulated and included in Table 5.2. This provides a summary of the analysis of historical catchment flows. While no one method can be considered definitive, the EV1 (Gumbel) and EV2 distributions are considered a good statistical method for arriving at flood estimates based on a sample of real world data. Three such analyses have been investigated to assess the skew resulting from sampling choice and distribution errors. Additionally, Catchment Characteristics method is widely used in Ireland for flooding studies and a number of parameters were checked to establish their sensitivity in the River Tolka. The results indicate that while the Catchment Characteristics method provides good agreement with predicted mean annual floods, the average Irish regional multipliers appear too low for higher return period events for the Tolka specifically. The Design curve chosen takes into account a number of factors including the errors associated with such an analysis and the limited reliability of the data. Of particular note is that a flow of 85m3/s being the largest flow within over 120 years record prior to the November 2002 flood would traditionally have been considered to have a return period in excess of 100 years. This flow is now considered to have a return period of approximately 90 years. The November 2002 flow of 97m3/s is now considered to have a return period in excess of 100 years and this is supported by analysis of rainfall runoff relationships in the model.

Chart 5.2 Comparison of growth curves

Return period comparison

0

20

40

60

80

100

120

140

Qm Q5 Q10 Q25 Q50 Q100 Q200

Return Period

Flow

(Q m

3/s)

CC CCs5 CC (SMD 0) CC UrbanCC Urban (s5) CC Urban (SMD 0) CC Urban (UK) EV1 (2003)EV2 (2003) EV1 (2003 POT series) EV2 (2003 POT series) EV1 (2003 expanded series)EV2 (2003 expanded series) Min Max DESIGN

Qm = Mean annual flood

The most appropriate method of modelling catchment flows, for use as inputs to the InfoWorks RS model is described further in Chapter 9 Therefore, analysis of historical flow data indicates that the catchment provides a significantly higher response to flooding than that which would generally be calculated by catchment characteristics methodologies, widely used in the design of culverts and bridge structures and indeed in assessing flood levels.


074515001Rp0032ADE.doc 24 Rev F01

5.5 SEA LEVEL / FLOOD INTERACTION

The tidal waterway is that subject to diurnal variation in water levels. The Tolka catchment discharges directly to the sea at East Wall / Clontarf. Analysis of information provided by Dublin Port Authority is provided in this section.

5.5.1 Tidal Frequency Analysis Statistical Analysis A statistical analysis of tidal recurrence was undertaken for annual maximum sea levels at Dublin Port for the period 1923 to Feb 2002, with the results listed below in Table 5.3:

Table 5.3 Results of Analysis

Level m OD (Malin Head) % Annual probability of occurrence Return Period

EV1 EV2 42.9% 2.33 2.30 2.31 20.0% 5 2.40 2.41 10.0% 10 2.48 2.50 4.0% 25 2.58 2.61 2.0% 50 2.66 2.70 1.0% 100 2.73 2.79 0.5% 200 2.81 2.89 0.2% 500 2.91 3.01 0.1% 1000 2.98 3.11

Water Levels on 1st February 2002 Figure 5.1 below sets out the predicted tide levels from Tide Tables and actual water levels between 31 January and 2 February 2002. The graph reveals that recorded water levels were consistently above those predicted, due to storm surge and wave effects. This illustrates the influence of weather factors on actual tide levels, i.e. extra tide level rise due to low barometric pressure, onshore winds and wave heights.

Figure 5.1 February 2002 Tide Levels

-3.00

-2.00

-1.00

0.00

1.00

2.00

3.00

4.00

30/01/200212:00

31/01/200200:00

31/01/200212:00

01/02/200200:00

01/02/200212:00

02/02/200200:00

02/02/200212:00

03/02/200200:00

03/02/200212:00

Time & Date

Wat

er L

evel

(m M

H)

RecordedPredictedFlood Event


074515001Rp0032ADE.doc 25 Rev F01

The above analysis indicates that the Highest Recorded Sea Level of 2.95m OD (Malin Head) has the following theoretical return periods:

EV1 758 years (average)

EV2 330 years (average)

Again this analysis is theoretical and accounts for incidence of storm surge in the recorded data. Results for return periods greater than 100 years should be treated with caution. This sea level has been assessed in the model; additionally an increase of 400mm (3.35m OD) to account for possible affects of climate change has been assessed. Climate change effects could reduce return period of extreme tide levels considerably. Tidal Coincidence Assessment of hourly sea level records surrounding peak flooding events was undertaken to establish coincidence. In general there is not enough data to establish a joint probability for these events. It has been established that due to the long narrow and flat nature of the Tolka river basin, flood hydrographs, where available for the Tolka River, can range up to 48 hours duration with flat peaks of up to 12 hours duration. Therefore, unlike some more flashy rivers, the likelihood of tidal coincidence is increased (1-2 high tides will coincide with a flood typically). The worst scenario for which records exist is 1954. In that case an estimated 90 year return period flood partially coincided with a 1.5 year return period tide. In that instance the flood was rising on the receding tide, and flooding was also influenced by the collapse of the CIE Bridge. Figure 5.1 indicates that due to the wide channel in the lower tidal reaches the increase in level associated with flow is marginal and the tidal level is the major governing factor relating to flood stage (level) in this area.

Figure 5.2 Tides / Flooding Coincidence

Typical Flow -Tide Coincidence AssessmentFebruary 2002 Tide with 2 Year return period flood

-2

-1

0

1

2

3

4

5

6

7

8

0.00 20.00 40.00 60.00 80.00

Time (hrs)

Stag

e (m

AD)

0

10

20

30

40

50

Flow

(m3/s

)

Stage at Luke Kelly BridgeStage at Tidal BoundaryFlow at Luke Kelly Bridge


074515001Rp0032ADE.doc 26 Rev F01

5.5.2 Sea levels Sea levels are recorded by the Dublin Port Authority and have been provided in metres to Malin Head Datum. The usual tidal range is provided by Figure 5.3 below:

Figure 5.3 Tide Levels for Dublin Bay (Malin Head)7

Tidal Reach

For a river or watercourse discharging to the sea, the influence of tides decreases as a tide moves up the watercourse. This is due to friction losses in the watercourse channel and results in a gradual decrease in the tidal range when compared to the ocean tidal range. For the Tolka River the maximum tidal range is confined to the reach below Distillery Weir. Sea Level Rise A separate study has been undertaken regarding climate change impact on future tides, and results are utilised in this study. Linear interpolations of the current results are included in Table 5.4 below; however the 400mm rise in sea level is applied for assessment purposes.

Table 5.4 Sea Level Rise for Dublin Bay

2011 2031 2051 2070-2100

Sea Level Rise 33mm 66mm 153mm 300mm

Land Movement - 3mm - 7mm - 15mm - 30mm

Total 36mm 73mm 168mm 330mm

Design Tail-water Levels Due to the lack of detail concerning the joint probability of tide and design flood flows, a simplified approach to inter-tidal flood conditions based on recommendations in the GDSDS Climate Change Document was undertaken to ensure that practical design scenarios are assessed, these include the following combinations of floods and tides:

1. 50 year river flood flows taken with 1.5 year tides (2.20m OD) (Approximate 1954 event)

2. 2 year river flood flow with 100 year tides (2.79m OD)

3. 0.5 year river flood flow with 200 year tides (2.89m OD)

4. Base flow with highest recorded tide plus 400mm (3.35m OD) (climate change scenario)

7 HAT Highest Astronomical Tide ML Mean Level MHWS Mean High Water Springs MLWN Mean Low Water Neap MHWN Mean High Water Neap MLWS Mean Low Water Springs

Note: Malin Head datum is 2.714 m above Poolbeg datum and 2.514 above L.A.T.

HAT MHWS

MHWN

ML

MLWN MLWS

2.70m

1.70m

0.80m

-0.20m

-1.34m

-2.25m


074515001Rp0032ADE.doc 27 Rev F01

It is likely that the joint probability of these combinations all exceed a 200 year return period criterion.

All additional analysis was undertaken with the following tail-water scenarios:

5. Calibration flows with recorded tide for event

6. Design flows with 1.24m OD tide (average tide)

The tide data was applied as a dynamic level hydrograph using a tide curve and applied to the model at a time to ensure the worst case is established. This was found to be the tide rising on the rising flood. Scenario 1 is an approximation of the 1954 event, and scenario 4 is an approximation of the February 2002 event with the added impact of possible future sea level rise.


074515001Rp0032ADE.doc 28 Rev F01

LONG TERM CLIMATE CHANGE IMPACTS

6.1 INTRODUCTION

The Dublin City Flood Risk Assessment study is reviewing the impact of climate change on tidal flood risk going forward. This section of the current report provides an over-view of potential climate change impacts on river flooding generally, based on research reviewed in the GDSDS study. The scheme as designed provides for alleviation of flooding to existing developments for extreme predicted flood events having regard to: �� Over 100 years of historical flood history in the Tolka. �� Allowance for existing and future development as predicted to 2031. �� A repeat of the November 2002 flood which was the largest flood in recorded history in the river

and which may be considered to incorporate all of the current factors contributing to flood risk (existing development, changes in the floodplain or river channel and current climatic conditions).

�� Tidal effects interacting with river flood flows. �� A modest level of “freeboard”, to give a safety margin (typically 300mm and occasionally greater).

This margin is necessary having regard to local variations in flood mechanisms, uncertainties in modelling and in flood prediction and to prevent overtopping from wind and wave action.

At the same time, the further possibility of increased flood risk in the future associated with climate change, specifically global warming resulting from increasing concentration of greenhouse gases in the environment has to be considered. This has been the subject of detailed study over recent years and the Inter-Governmental Panel on Climate Change (IPCC) recently concluded that atmospheric warming observed over the last 50 years (estimated 0.60C rise in temperature) is predominately due to human activities. Significant investigation has been carried out in the U.K. (under the U.K. Climate Impacts Programme) to provide climate predictions for the U.K. up to 2080. This has led to the publication in April 2002 of “Climate Change Scenarios for the United Kingdom – the UKCIP02 Report” with major inputs from the Tyndall Centre for Climate Research and the Hadley Centre of the U.K. Met Office. The report recognises the uncertainties in environmental predictions for climate change depending on future emission levels, the response of populations, economies, energy technologies and societies generally. A range of projections have been postulated and four of these alternative scenarios have been taken forward for detailed consideration representing low, medium-low, medium-high and high emission scenarios. The research reports stress that no probabilities can be assigned to any of these forward emission scenarios, the lowest one being based on international co-operation to reduce the emissions below current levels as proposed under the Kyoto protocol. All of them are regarded as plausible descriptions of socio-economic development that could affect future emissions of greenhouse gases. It is emphasised that global climate change and particularly reversal of current trends will be a very slow process. Carbon dioxide has an effective lifetime in the atmosphere of almost 100 years, so that any reduction in global emissions would take several decades for a response in terms of changing concentration in the atmosphere. For the four scenarios discussed, predicted increase in global temperature to 2080 would be in the range 2.00C to 3.90C. The Hadley Centre Global Climate Model has been developed to simulate changes in the climate due to each of the carbon emission scenarios discussed. Based on the global model, the Hadley Centre Regional Climate Model has been developed to predict climate change for the U.K., with an admitted high degree of uncertainty. Nevertheless, the model contains predictions for changes in tidal


074515001Rp0032ADE.doc 29 Rev F01

conditions and rainfall levels which could be applied to the Dublin area as a basis for estimating possible medium-long term impacts in the River Tolka catchment.

6.2 GDSDS PREDICTIONS FOR CLIMATE CHANGE

In light of this scenario, the GDSDS Study required the production of a Guidance Document to assist in identifying possible climate change implications for urban drainage in the Dublin area in the future. A draft document has been circulated for consultation which reviews the work done to date and in particular the latest predictions from the Hadley Regional Climate Model and its application to Ireland. The prediction which is considered applicable to Ireland is for an increase in annual average temperature by approximately 30 for the year 2100. This temperature effect will influence sea level rise due to thermal expansion and snow melt as well as increased energy available for storm events, giving an expectation of more intense rainfalls. It will also impact on evaporation, leading to lower base flows in rivers and increased river water temperatures. For the medium-high scenario, the U.K. Climate Impacts Programme 2002 (UKCIP02) suggest the following rainfall changes: �� Winter rainfall depths to increase by 20% with decrease in summer by 35 – 45% overall. �� The two-year daily rainfall event showing similar trends with a 20% increase in winter and 10%

reduction in summer. �� Daily data from the HADCM3 model for the Dublin region suggests that a two-year return period

event has an increase in rainfall of about 10% rising to nearly 25% for the 100 year event. If this were borne out, it would have a significant effect in reducing the return period of a given storm/flood event.

�� There is no hourly rainfall information available which could be used for drainage engineering

purposes. At present, the only reference is to daily rainfall information and considerable work is required if short duration rainfall predictions are to be made available.

Changes in sea level for engineering application must have regard to rate of sea level rise, rate of land fall/rise, surge height and wave height. The predicted rise in sea level in the UKCIP02 model, taking into account surge, is in the order of 300mm-400mm, with an additional 30mm due to the relative land movement. Therefore, the recommendations for climate impact on tides are: �� The predicted change in tide level will be an overall increase of 400mm by 2100. �� The predicted once in 200 year extreme tide level in Dublin Bay is 2.89m OD. However, in an

extreme tidal flood event in February 2002, a record level was recorded in Dublin of 2.95m OD. �� On that basis, a design extreme tide level for 100 year design life structures is suggested as

3.35m OD. (unless superseded by analysis undertaken in the DCFRAS) In the very long term, it is considered that sea levels will continue to rise and over several centuries, rises of 1m or more are anticipated. The GDSDS report recognises that full integration of these potential climate change impacts into design criteria for river and stormwater drainage and tidal defence structures would have major economic implications. To ignore these implications would be likely to result in a slow reduction in the level of service provided by drainage systems as conditions change. The following factors have to be considered: �� The degree of certainty with regard to the level of hydrological change is acknowledged by all

experts as being quite low. This is particularly true for extreme events that are critical for drainage design application.


074515001Rp0032ADE.doc 30 Rev F01

�� The implications of applying 10-25% uplift on design rainfalls would dramatically increase solution costs for dealing with existing flooding and drainage infrastructure generally. This is estimated as being in the order of 15% for new pipework but would be substantially more for upgrading of existing systems.

�� The uncertainty of the change in rainfall, together with the knowledge that these changes are

predicted to take place over the coming century implies that solutions to climate change possibilities need to be addressed in a flexible manner. The target should be that existing networks should at least meet target criteria based on current rainfall, with provision for future rainfall uncertainties related to cost-benefit considerations.

6.3 APPLICATION OF CLIMATE CHANGE TO THE RIVER TOLKA

In the River Tolka Study, it is possible to use the model to test the various climate change scenarios discussed in this chapter. Based on this analysis, the following conclusions are arrived at: �� Extreme tides; the greatest degree of certainty regarding climate change is likely to relate to

expected rise in extreme tides. On that basis, an extreme tide level of 3.35m OD is recommended. This level, corresponding with 2100 prediction, is marginal as regards existing flood defence wall levels in the lower tidal area of the Tolka and some flooding would be likely to occur. In the short term, therefore, the existing tidal defences are considered satisfactory but over time, continued monitoring of tide levels would be required with the likelihood that some raising of existing floodwalls would be necessary in the tidal region, if current predictions of climate change tidal increases are borne out during ongoing monitoring.

�� Extreme rainfall; the GDSDS climate change review advises that a precautionary position of

10% should be used given the current state of knowledge of increased rainfall depths. It is noted there is potential for 22% increase in rainfall depth for a 100 year return period extreme event. This is considered a conservative approach and its application would have major implications for the design of the Tolka scheme. The prediction relates to climate change out-turn at 2080 which is substantially beyond the design horizon for the scheme.

If a medium – low emission scenario was used, the rainfall depth increase for the Dublin area would be predicted to rise by 10-15% by 2080. There is already a significant natural variability in extreme rainfall patterns. Historical review of weather records identifies certain periods differing significantly from the norm, for example the extreme dry summers of 1975, 1976. This is recognised in section 7.6 of the UK CIP02 report, where it is stated that the increase in winter rainfall as predicted by the model for the year 2020 under the medium – high admission scenario is of the same order as would be expected from natural climate variability. This natural climate variability must be assumed to be encompassed by the historical records for the Tolka, including the extreme storm event of November, 2002. The report does not offer any comment on possible change in frequency of high return rainfall events which could lead to increased frequency of critical storms. In light of this situation, it would not be possible to justify major additional expenditure on a benefit – cost basis for possible climate change impacts over and above the capacity of the proposed scheme. The River Tolka Flood model was used to test the impact of the various climate change rainfall increases in excess of the 1 to 100 year return period design event. For a full 22% increase in rainfall depth, the impact would be catastrophic with serious exceedence of all of the main bridges and hydraulic structures along the channel. However, it was established that a 10% increase in rainfall depth could generally be accommodated within the freeboard which has been allowed for in the design and up to 15% extra flow could be accommodated in parts of the scheme. Therefore, if it is considered that freeboard is provided to deal with a range of uncertainties to include climate change, it can be seen that the level of freeboard available can accommodate up to 50% of the predicted climate change impact to 2080. Given the design horizon of the scheme of 2031, (and 2053 for economic analysis) this appears to be a reasonable situation.


074515001Rp0032ADE.doc 31 Rev F01

The provision of a scheme for the ultimate climate change prediction plus safety margin, would result in a wholesale upgrading of the capacity of the river and its associated structures. Instead, the proposed scheme which has been demonstrated to be cost effective in light of historical flood statistics, makes reasonable provision for medium- term climate change impacts. It has been demonstrated in the modelling analysis that the provision of additional storage attenuation at key sites along the river could reduce flood flows, with individual schemes (Tolka Valley, M50) potentially offering up to 5% reduction in peak flood flow through the city. Retro-fitting of such flood attenuation scenarios, therefore, could be considered in the future to offset climate change effects not provided for at this stage. Therefore attenuation measures (and associated spatial requirements) should be allowed for in development plans, in addition to the retro-fitting of such measures to existing development.

6.4 JOINT PROBABILITY OF TIDAL AND FLUVIAL FLOODING

In the inter-tidal area it is necessary to consider the possibilities of joint occurrence of critical tidal and river flooding events and a pragmatic approach to design for such combinations of events is proposed in the GDSDS report along the following lines. (Note: A more conservative approach as outlined in Section 5.2 has been assessed in this study.)

�� River and Coastal flooding; one year river flow with one hundred year tide plus one year tide with 25 year return period river flow

�� River and tidal flooding –inter-tidal area; two year river flow with one hundred year tide

and one year tide with fifty year river return flow

�� Storm Drainage; Assessment in conjunction with river/tide; two year river design with one hundred year tide and one year tide with ten year design drainage for the tidal interface and one year drainage with 50 year river flow together with one year river flow with 20 year drainage design storm for interaction of drainage system and river levels.

These requirements are recommended for the design of pipe drainage systems in urban areas. Where existing systems are analysed, a pragmatic approach is required where upgrading of the network would be required to meet the standards. This requires local risk assessment which would have regard to the risk and scale of property inundated which would justify expenditure on a cost benefit basis. These interactions of river and tidal conditions are accepted for design purposes for historical design storms. The additional climatic impact for river flooding has been dealt with as discussed above. For pipe networks, consideration is given to up to 10% increase in rainfall intensity where new construction is envisaged. This is considered to be a reasonable and pragmatic response to the current understanding of climate change impacts on flood risk in the context of the River Tolka study.

6.5 SUMMARY

To summarise, this section has reviewed general guidance for drainage Engineers as regards climate change impacts and considers the practical application of the draft guidance produced by the Dublin Drainage Consultancy as part of the GDSDS study. The study advises that a precautionary position of 10% should be used given the current state of knowledge of increased rainfall depths. Guidance from the EPA also suggests that provision of the order of 10% is reasonable. Ultimately, design decisions must have regard to the cost / benefit implications of the works recommended. In the case of the River Tolka, it is considered that the proposed scheme makes adequate provision for climate change effects, having regard to the current state of knowledge of these effects, the marginal capacity within the scheme and the design horizon being catered for (2031 development scenario and 50 year design life for economic assessment purposes). Provisions can be maintained for future enhancements in capacity or defence standard where appropriate.


074515001Rp0032ADE.doc 32 Rev F01

7 MODEL CONSTRUCTION

7.1 OVERVIEW The River Tolka model covers the main Tolka River from the Tolka Estuary to Batterstown (integrating the Clonee to Dunshaughlin Road Scheme model, assessing the River Tolka in Co. Meath) plus the potential urban reaches of the Castle stream in Dunboyne, and the Twin Pinkeen streams at the Fingal / Meath county boundary west of Mulhuddart. Other culverted rivers are included in the contributing urban catchments. Piped/culverted streams will be incorporated in GDSDS models and form contributing areas to the Tolka model. The Finglas River is modelled separately in the GDSDS, with its extensive culverted reaches. The catchment is incorporated as contributing area to the Tolka model however data was not available from the GDSDS to allow both models to be reconciled. An important factor in a robust model is the quality of the input data. Key data includes:

1) Up to date river cross section, bridge and other flow control structure records;

2) Information on floodplains, flood storage, and secondary flow path routes;

3) Calibration and verification data, key rainfall events and observed river levels.

Due to the surveys taken at the commencement of the study and in particular the flood that occurred during construction of the model, the quality of the input data for the model is very high.

7.2 MODEL DEVELOPMENT

7.2.1 Modelling Software Infoworks RS is an integrated hydrological and hydraulic modelling package developed by Wallingford Software Ltd. Infoworks RS incorporates the well established ISIS hydraulic simulation engine with the functionality of a GIS database management system. Infoworks RS main advantage is in the integration of much of the model build and run processes into the one package allowing for a simpler documentation of the whole modelling process. However, as found throughout this project, Infoworks RS is a relatively new software package and contains some facilities that require further refinement. Ultimately, it has been possible to utilise the model to simulate the hydrological and hydraulic processes of the river. Manual intervention has been required to develop appropriate output of results in terms of risk mapping, assessment and presentation of data.

7.2.2 Phased Model Development One of the deliverables of the River Tolka Flood Study is a robust InfoWorks RS model of the River Tolka and its key tributaries. The model build process was carried out in a series of stages. Import of Cross Sections The cross sections from the topographical survey were imported into InfoWorks. The surveys had been taken in ASCII file format, surveying sections from left bank to right bank looking downstream and hence a direct import with no manipulation of data was feasible. The HEC-RAS model was converted into InfoWorks by Wallingford Software Ltd. As cross sections in HEC-RAS were not correctly georeferenced, the river sections were imported with incorrect orientation. This was amended within InfoWorks by re-aligning the cross sections correctly.


074515001Rp0032ADE.doc 33 Rev F01

Tolka Tributaries The Castle Stream, Twin Pinkeens and upper and lower Tolka reaches were initially modelled separately, using appropriate stage time boundary conditions at the outflow points. This allowed several smaller simpler models to achieve a stable level of operation in advance of integration of these sub models to a singular model network. Flow Control Structure Input Structures were input and tested on a gradual basis. The structures from the HEC-RAS model were also input manually as a direct import was not possible.

7.2.3 User Defined Flags A series of flags have been used in the model to denote where the data originated, and ensure all the data sources are recorded. InfoWorks RS provides a number of default flags. Additional flags specific to the project were then added to coincide with the model construction. Table C.1.1, Appendix C shows the flags used with a brief description of them.

7.3 HYDROLOGICAL MODEL A number of possible hydrological strategies were investigated to determine the best approach to utilise in this study. An integrated model is considered the best approach. The Rainfall-Runoff models currently available in IWRS are as follows: 1. FEH Rainfall-Runoff Method

2. FSSR16 Method

3. US SCS Method

4. InfoWorks PDM

The modelling approach used is FSSR 168. Other methods were not used for the following reasons:

�� FEH is only applicable to the UK;

�� SCS is even more uncertain with the empirical selection of coefficients;

�� InfoWorks PDM requires a continuous time series of flow data and uses a black box approach to fit the catchment characteristics and therefore cannot be applied here;

�� The PR method used in GDSDS is not available in IWRS additionally the option of manually integrating detailed InfoWorks CS models was not feasible as the GDSDS results were not available. This would be an impractical and time consuming task however as InfoWorks models are not yet sufficiently integrated to allow auto communication between models.

Therefore, in the absence of recorded flows at any of the major sub-catchments, the FSSR16 method was used.

7.3.1 Rainfall/Runoff Model FSSR16 is the rainfall/runoff relationship used in the River Tolka Flood Study. The rainfall runoff model is based on the Unit Hydrograph Theory. The unit hydrograph was derived from rainfall and runoff records to define catchment characteristics can be used in the model. The FSSR16 boundary generates flow hydrographs for design events or will simulate runoff during historic events using recorded rainfall and other input data derived from calibration. Effective rainfall is transformed into a runoff hydrograph by convolution of the unit hydrograph.

8 Natural Environment Research Council, UK – Flood Studies Supplementary Reports (1977-1988) No. 16: The FSR rainfall runoff model parameter estimation equations updated (Dec 1985).


074515001Rp0032ADE.doc 34 Rev F01

The catchment characteristics required as input data for this method along with a brief description are as shown in Table 7.1. A detailed table of input parameters to the model is shown in Table C.1.2, Appendix C1. Table 7.1 Catchment Characteristics input in the model Catchment Characteristics Description Area (m2) Contributing Catchment Area Urban Fraction Fraction of the catchment with Urban Development (range: 0.0-1.0) SAAR (mm) Standard Average Annual Rainfall taken from FSR Maps. S1085 This is the average channel slope between 10 and 85% of it the

length of the main stream measured up from the outlet MSL (m) Main Stream Length, this should be measured with dividers of 0.1km

step length from the outlet to the source M5 2 Day (mm) 5 year return period, 2 day duration rainfall, taken from the FSR M5 25 Day (mm) 5 year return period, 25 day duration rainfall, taken from the FSR Jenkinson Ratio (r) Ratio of 60 minute M5 rainfall to 2 day M5 rainfall (M5-60min/M5-2D) CWI (mm) Catchment Wetness Index (computed or observed) A comprehensive explanation of the rainfall runoff relationship showing equations used can be found in Section C1 of Appendix C.


074515001Rp0032ADE.doc 35 Rev F01

7.3.2 Subcatchment Detail The overall Tolka Catchment is divided into individual sub-catchments as shown in Figure C1.1, Appendix C. These sub-catchments were derived using contour maps and examining tributaries and ditches and where they enter the main Tolka River. This information was then digitised. There are 35 subcatchments in total, 28 rural and 7 urban. The drawing displays the sub-catchment area and the percentage development up until 2031. Table 7.2 shows the catchments and corresponding area with percentage development input into the model. Table 7.2 Sub-catchment Detail of River Tolka

% Development Catchment No

Catchment Area (ha) 2002 2011 2031

1 364.65 2.5 3.7 3.7 2 460.18 0 1.1 3.2 3 180.34 2.6 5 5 4 381.8 0 1.1 4.9 5 253.44 3.1 6.5 6.5 6 1423.26 0 0 4.1 7 685.03 0 0 0 8 1091.71 0 0 0 9 742.86 0 0 0 10 76.91 6.9 20.3 20.3 11 85.62 12.5 39.5 39.5 12 34.77 7.1 7.1 7.1 13 121.14 1.3 47.2 47.2 14 166.46 0.5 0.5 0.5 15 1345.95 0.2 0.2 0.3 16 34.65 0 0 0 17 316.85 13.5 21.4 21.4 18 98.01 28.9 73.1 73.1 19 48.47 58.8 59.5 59.5 20 93.47 42.3 46.9 46.9 21 180.33 0.4 6.3 6.3 22 339.81 0 4.6 77 23 339.81 86.5 92.9 93 24 215.07 37.4 68.6 68.6 25 203.24 0 73.9 76.1 26 338.78 0 65.7 94.1 27 609.88 46.7 95.2 97.8 28 384.76 7.7 37.6 92.1 29 147.4 5.6 5.6 6.5 30 294.64 20.6 20.6 36.3 31 237.99 81 81 85.5 32 380.37 56.5 56.5 56.5 33 308.67 98.3 98.3 98.3 34 536.49 66 66 66 35 1022.3 36.3 39.4 42.7

7.4 HYDRAULIC MODEL The Hydraulic Model represents the conveyance of flow through the river. The InfoWorks RS model contains a detailed representation of river channel cross sections and flow control structures to adequately model the full length of the River Tolka. Secondary flow paths and storage areas have been examined to correctly represent the river processes as outlined below.


074515001Rp0032ADE.doc 36 Rev F01

The data collected as a result of the November 2002 flood was invaluable in ensuring that secondary flow paths were identified and calibrated correctly. Thus, significantly adding to the confidence in the model in this area.

7.4.1 Cross section Representation Cross sections were imported directly into the InfoWorks RS model using the data received from the topographical survey. The length between cross sections also imported directly. Interpolated sections were often required in the model when there was sparse cross sectional data, or cross section properties varied radically between sections, causing instabilities in the model. The InfoWorks RS feature inserts an interpolated section with only x,y co-ordinates of where it is positioned and not of every point along the cross section associated with it. Hence the interpolation was often carried out in HEC-RAS and then transferred to InfoWorks RS. Figure 7.1 shows the InfoWorks RS interpolated section between 2 river sections. Figure 7.1 Interpolated Section in InfoWorks along the Pinkeen Stream

River cross sections were extended to the river flood plain extent following identification of requirements, using the Digital Terrain Model (DTM).

7.4.2 Hydraulic Structure Representation Each flow control structure was input manually into InfoWorks RS. Appendix C.2 describes in detail each hydraulic structure in the model. Bridges and Culverts in InfoWorks RS There are 2 types of bridge structures in InfoWorks RS, the arch bridge and the USBPR (United States Bureau of Public Roads) bridge. The arch bridge takes surcharging into consideration and hence is used to model the majority of bridges on the River Tolka. It calculates afflux across the arches from the empirical methodology derived at HR Wallingford and is described in Afflux at Arch Bridges (1998). If the bridge surcharges and either overflows or has a flow path around it, a spill structure is required as an alternative path for the flood waters with details being derived from DTM and survey data, and checked with site visits to ensure correct representation. Input parameters for the arch bridge include; appropriate upstream and downstream cross sections, skew angle, cross section detail of bridge, and arch spring and soffit levels. A schematic of an arch bridge with a spill unit in parallel is shown in Figure 7.2.


074515001Rp0032ADE.doc 37 Rev F01

Figure 7.2 Schematic of an Arch Bridge at Damastown

A limitation of the USBPR Bridge is that it does not take surcharging into consideration. Therefore analysis was required to ensure that this method could be used at specific sites during the model construction and calibration phases. It computes the afflux at bridges using the methodology developed by the US Bureau of Public Roads. An example of where this bridge is used is the DART Road Bridge (See Appendix C2). Input parameters for the USBPR Bridge are; type of abutment (normal, 90° wing or vertical, 30° wing), alignment, cross section detail of bridge and soffit levels and detailed pier data. A schematic of the USBPR Bridge is shown in Figure 7.3. Both the arch and USBPR bridge have no length associated with them. This length of the bridge is then compensated by including it in the upstream and downstream links. Figure 7.3 Schematic of a USBPR Bridge at Annesley Road

To model a culvert in InfoWorks RS a series of network objects are used. These include; culvert inlet, rectangular/circular conduit and culvert outlet. The culvert inlet and outlet model the culverts entrance and exit losses respectively. The friction losses along the reach of the culvert are modelled by the conduit nodes. A schematic of how a 3 barrel culvert is modelled, is shown in Figure 7.4.


074515001Rp0032ADE.doc 38 Rev F01

Figure 7.4 Triple Barrel Culvert at Blanchardstown Road North

Input parameters for culvert inlets (i.e. inlet loss coefficient etc) vary depending on the culvert material and inlet type and can be found in the Culvert Design Manual (1997). Physical parameters i.e. invert level, diameter, for culverts are input in the circular/rectangular conduits. Outlet loss coefficient is also input into the culvert outlet.

Many of the culverts on the Twin Pinkeen streams have been modeled as orifices. This is due to their shortness in length and failure to converge if modeled as a culvert. The orifice uses the orifice equation if surcharged otherwise the weir equation is used.

The input parameters include; invert level, soffit level, upstream and downstream sill levels, and bore area.

Figure 7.5 Schematic of an Orifice on the Pinkeen Stream


074515001Rp0032ADE.doc 39 Rev F01

Weirs in InfoWorks RS There are 17 weirs along the River Tolka. Many of these are merely a slight step in bed level, in particular through Griffith Park and Botanic Gardens and become drowned during flood events and therefore do not impact on flood levels. However, they have been included in most cases to increase model stability. There is a feature in InfoWorks RS to model broad crested weirs, but because of all the irregular shaped weirs along the Tolka, many instabilities are derived from their use, therefore they have been modelled using spill units. The cross section detail of the crest of the weir is input in the model. In many cases it was necessary to lower the cross section directly upstream of the weir, so a noticeable drop in bed level occurred for the weir structure calculations to operate correctly. Figure 7.6 shows a schematic of a weir being modelled in InfoWorks. Figure 7.6 Weir at Glasnevin Road

Bernoulli Loss / General Headloss The Bernoulli loss is used to model head losses, such as those caused by changes in cross section across open channel constrictions or expansions. It may also be used to model severe bends in a river channel. In the Tolka model the Bernoulli loss is used mainly to model bridges. In some cases arch bridges were initially used but changed to Bernoulli loss during calibration if the model was predicting afflux in excess of actual calibration records. The Bernoulli equation is used to model the Bernoulli loss. Figure 7.7 Dean Swift Bridge modelled as a Bernoulli Loss


074515001Rp0032ADE.doc 40 Rev F01

A General Headloss is used to model any obstruction or bend in the river causing an energy loss. It calculates the difference in velocity head and allows the user to enter a coefficient to increase or decrease the magnitude of the energy loss to match actual calibration records. Figure 7.8 shows a schematic of a general headloss. Figure 7.8 Temporary works along the Tolka during the DART bridge construction, modelled by a general headloss

Table 7.3 Classification of Structures used in InfoWorks RS Structure Number present in model Arch Bridge 31 USBPR Bridge 2 Culvert 26 Orifice 10 Weirs 17 General Headloss 2 Bernoulli Loss 6

7.4.3 Flood Plain Representation Due to the flooding in November 2002, the knowledge of flood plain storage and additional flow paths increased greatly throughout the River Tolka Flood Study. This also contributes to the accuracy of the model. Floodplains may act as either:

�� storage where the flood waters may not significantly contribute to the main river flow conveyance or pond at low points

�� a significant flow path that contributes to the river flow conveyance In the first scenario the storage is generally modelled as a floodplain storage area. A storage cell with a series of levels with corresponding areas is entered. The floodwaters spill into the storage cell via a spill unit at an appropriate out of bank level. A schematic of a floodplain storage area in Clonee is shown in Figure 7.9.


074515001Rp0032ADE.doc 41 Rev F01

Figure 7.9 Floodplain Storage in Clonee

The spill just upstream of the N3 culverts in Clonee controls the flow into the flood storage. The flood storage is modelled as three storage cells, due to the dip in the road on Main Street, Clonee and the low lying area of Littlepace. The storage cells are connected via spill units and the crest level of the spill units control the flow. The flood storage area at Littlepace is connected to the river at Huntstown via an orifice. This shows where the storm drainage and flood waters of Littlepace re-enter the River Tolka. In the case of significant flow paths contributing to the conveyance, river sections were extended to the extent of the floodplain using the LIDAR DTM. In some cases an alternative flow route was examined but could not be stabilised so a series of flood storage cells with graduated spill levels were used.


074515001Rp0032ADE.doc 42 Rev F01

8 MODEL CALIBRATION / VERIFICATION The model was calibrated using two flood events; 14th / 15th November 2002 and 5th / 6th November 2000 and subsequently verified by the event on 26th August 1986. The event of 22nd October 2002 was initially used as a verification event but is considered to be of insufficient size for an extreme flow hydraulic model. Model calibration runs were undertaken to provide a best fit between the 2 events in terms of hydrograph peak and volume in comparison with the recorded hydrographs at Botanic Gardens. Model calibration was particularly focused on the November 2002 flooding event due to the scale of the event and due to the quantity, quality and reliability of the event data including rainfall flow and level information and the fact that the channel characteristics have not changed significantly since this event. Verification of the model was addressed in terms of historical rainfall, flows, hydrographs and flood levels from previous reports, photographs, Hydrometric station information and anecdotal evidence.

8.1 HYDROLOGIC CALIBRATION A very good fit was achieved without the need to adjust the timing or shape of the derived unit hydrographs. It is noted that an area of flow exists in all of the predicted hydrographs at the start of the event. Infoworks RS has the facility to incorporate a variable percentage runoff throughout the event, based on the FEH decreasing proportional loss model. This allows for an increasing Catchment Wetness Index (CWI) throughout the storm, and hence a greater runoff as the soil becomes more saturated. Unfortunately, however IWRS does not as yet have the facility to calibrate the increasing CWI during an event. Therefore CWI has been set at a slightly higher than average level to ensure that peak flows are met. It is likely that a higher variance of CWI during the event, (i.e. lower initial value and higher end value) would move this initial flow towards the centre of the hydrograph. The hydrologic and hydraulic calibration was to some extent an iterative procedure. The hydrologic calibration was achieved to a reasonable level of satisfaction and the hydraulic calibration was then assessed. The hydraulic calibration required adjustments which in turn altered the hydrological calibration.

8.1.1 November 2002 The flow calibration based on an extreme value event such as the November 2002 is considered excellent for flood study purposes. It should however be noted that the model is calibrated to winter storms with wet antecedent conditions, to the slight detriment of summer storm analysis, as this is considered to be a more conservative approach. This also results in a conservative analysis of flood storage effects when using design rainfall events. Results are shown in Graph 1, Appendix D.1 The event was preceded by 2 days of very heavy rainfall. A previous rainfall event on 8th -10th November resulted in a very wet catchment, which when combined with winter vegetation conditions, resulted in near impervious catchment conditions. Rainfall commenced in the afternoon/evening of 13th November and reached peaks of 8.7mm/hr at Dublin Airport in the afternoon of 14th November. The highest total rainfall for the event was 93.45mm recorded at a temporary GDSDS gauge in the South East of the catchment, followed by 85.8mm at the Dublin Airport Gauge. A further reading at Casemont Aerodrome, Baldonnel of 72.1mm indicated that the greatest rainfall for the event was received at the eastern extent of the catchment. Rainfall figures of 57mm, 52.5mm, 48.7mm and 55.4mm were measured at Ratoath, Dunshaughlin, Warrenstown and Celbridge respectively to the west of the catchment. Graph 4, Appendix D1 shows a comparison of rainfall recorded at Dublin Airport to flow measured at the Botanic Gauge.


074515001Rp0032ADE.doc 43 Rev F01

Total measured rainfall amounts were interpolated for every subcatchment and then compared to the total volume of flows measured at the Botanical Garden Gauge. This indicated that 88% of the rainfall contributed to flows above the baseflow. Six hourly radar precipitation accumulations of the event are shown in Appendix B2, Figure B2.2 - Fig B2.5. Met Eireann have emphasised that the radar plots only have an accuracy of 40% to 70%, and should not be used for determining rainfall amounts in the catchment. Radar images are however a useful tool in observing spatial distribution trends across the catchment. The radar plots indicate a comparative region of high rainfall intensity in the upper western catchments, where raingauge data was unavailable. The higher rainfall experienced around Dunboyne was further confirmed when the flows required to obtain a match with measured water levels on Castle Stream were found to be higher than the flows derived from measured rainfall in Dublin and Dunshaughlin. It is important to note that the rain gauge, which was decommissioned in Dunboyne would have been invaluable for calibration purposes as there is no rainfall data available within the upper catchment. Raingauge 1 located in Ashtown as part of the GDSDS was used as the temporal pattern for the rural part of the catchment. This raingauge recorded the highest depth of rainfall of all the available raingauges. It is also the closest available gauge to the upper parts of the catchment. The profile had a sharper peak than any of the other gauges, which resembled the recorded hydrograph at the Botanic gauge. The total depth of rainfall for each individual catchment was scaled up or down, depending on the spatial distribution. The isohyetal distribution derived for this event can be found in Appendix C3, Figure C3.3. Table 8.1 describes the rainfall used for the urban subcatchments. The data was input directly from the GDSDS Raingauges. Table 8.1 Rainfall Data Input for Urban Catchments Catchment No Raingauge No Location 29 3 Cappagh Orthapaedic Hospital 30 1 St Vincents, Ashtown 31 2 Patrician College, Finglas 32 5 Kartoncraft, Cabra 33 6 Clarehall Community Unit, Glasnevin 34 12 St. Vincent de Paul, Marino 35 4 Finglas Bring Centre The Antecedent Precipitation Index, which uses preceeding rainfall to the event was used to calculate an initial Catchment Wetness Index. This was used as an initial estimate but both the CWI and SPR were adjusted in parallel to achieve an accurate calibration plot. A CWI of 185 and a Standard Percentage Run Off of 70 was used for the November 2002 event.

8.1.2 November 2000 The flood of 6th November 2000 is the 3rd largest flood ever recorded on the Tolka. A flow of 76m3/s was recorded at the Botanic Gardens station. Hourly rainfall is available for Dublin Airport and Casemont Aerodrome for this period and total rainfall for the entire event is available for Leixlip and Dunshaughlin. The hydrologic calibration graph for this storm gave quite a good fit initially and only involved slight refinement of the catchment parameters to achieve the final calibration plot, shown in Graph 2, Appendix D1. Graph 5, Appendix D1 shows a comparison of rainfall recorded at Dublin Airport to flow measured at the gauge in Botanic Gardens. The temporal pattern based on rainfall recorded at Dublin Airport was used for the November 2000 event as the GDSDS raingauges were not in place at that time. The rainfall was then scaled for each catchment according to the spatial distribution as shown in Figure C3.2, Appendix C3. A Standard Percentage Runoff of 70 and a Catchment Wetness Index of 165 was used. The measured and modelled peak flows coincide and similar characteristics appear on this graph as in


074515001Rp0032ADE.doc 44 Rev F01

November 2002. The initial modelled flow is higher than the measured flow as previously due to the limitations in adjusting the CWI throughout the event.

8.1.3 August 1986 The model was verified using the August 1986 event, Hurricane Charlie. Hourly rainfall is available for Dublin Airport and Casemont Aerodrome, Baldonnel with daily rainfall available for Leixlip and Dunsanny. Graph 6, Appendix D1 shows a comparison of rainfall recorded at Dublin Airport to flow measured at the gauge in Botanic Gardens. Recorded rainfall at Dublin Airport was used as the temporal pattern, due to the lack of temporal data available. Isohyets were plotted to scale the distribution, as shown in Figure C3.1, Appendix C3. A Catchment Wetness Index of 50 and a Standard Percentage Run off of 70 was used for this event. The calibration plot is shown in Graph 3, Appendix D1, there appears to be a higher volume of modelled flow compared to measured flow. This inconsistency may be due to the fact that the model was calibrated to a much higher peak flow, for winter conditions and also the extent of calibration data available was not as broad. It is possible that the increased volume modelled is due to the spatial distribution of rainfall being much more scattered than represented in the model, or that the summer conditions resulted in greater retention and infiltration. However, as outlined previously the model is primarily required to represent winter flooding conditions.

8.2 HYDRAULIC CALIBRATION The hydraulic phase of calibration involved ensuring that the structural model was accurately representing recorded calibration levels. This involved a review of all floodplain storage, secondary flow paths, loss coefficients, and spill structures within the model and sharp bends along the river. Final adjustments included review of roughness values and adjustments to afflux parameters. There are a number of weirs located along the River Tolka, which are very good calibration points. These include the weir at the Botanic Gardens gauge, Finglas Factory Weir and River Road Weir, Blanchardstown. Hydraulic calibration plots of observed levels against modelled levels for November 2002 and November 2000 are shown in Graph 1-4, Appendix D2.

8.2.1 November 2002 The majority of levels taken for the November 2002 storm were at flow control structures, mainly upstream and downstream of bridges or at weirs. There is a wide extent of data available for this event, including levels, flood extents and knowledge of flow paths. The initial hydraulic calibration before any adjustments were made was quite accurate (approximately 600mm). Then the areas of higher inaccuracies were examined carefully, with respect to the method of flooding and the accuracy of the calibration data itself. Adjustments were made accordingly e.g.

�� Spill structures were altered (widened, shortened or lowered) to represent more accurately what occurred during the storm;

�� Energy losses associated with bends were modelled as a general headloss and the headloss coefficient adjusted until the observed levels were achieved;

�� Flow control structures were modelled in a different manner to produce better calibration results, i.e. a number of arch bridges were modelled as Bernoulli losses and gave a more accurate calibration.

Further calibration then involved adjusting mannings roughness along certain reaches of the river and if necessary altering the calibration coefficients of individual structures. Details of these changes or similar calibration adjustments can be found in Appendix C2 – Flow Control Structures. Table 8.2 shows a comparison of the calibration levels to the modelled levels recorded for the November 2002 event.


074515001Rp0032ADE.doc 45 Rev F01

Table 8.2 Calibration Comparison for 14th & 15th November 2002 Cross Section

Location Observed Level (mAD)

Modelled Level (mAD)

Difference (m)

T-003 East Point Business Park. 1.52 1.741 0.221 T-008 John Mc Cormack Bridge 1.66 1.824 0.164 T-015 Footbridge u/s Dart Bridge 2.38 2.277 -0.103 T-027 Luke Kelly Bridge 2.6 2.753 0.153 T-033 Distillery Rd Bridge 4.19 4.164 -0.026 T-036 Upstream of Distillery Weir 4.8 4.738 -0.062 T-039 Downstream of Tolka Park 5 5.131 0.131 T-044 Drumcondra Rd Bridge 6.39 6.194 -0.196 T-049D Millmount Villas 7.3 7.241 -0.059 T-052U Drumcondra Footbridge 7.8 7.774 -0.026 T-060 Weir in Griffith Park 8 7.927 -0.073 T-068 Dean Swift Bridge 10.33 10.437 0.107 T-B1U Glasnevin Rd Bridge 11 10.96 -0.04 B15a Gauge in Botanic Gardens 13.74 13.745 0.005 T-109 Finglas Road Bridge 17.75 17.781 0.031 T-113AD Weir beside Tolka Vale 19.63 19.603 -0.027 T-119BU Finglas Wood Park Bridge 21.5 21.563 0.063 T-120B Ratoath Rd Bridge 24.44 24.433 -0.007 T-124U Cardiff’s Bridge 26.2 26.162 -0.038 T-136D Scribblestown Rd Bridge 30.9 30.669 -0.231 T-137U Scribblestown Rd Bridge 32 31.848 -0.152 T-157 Northern Cross Route 37.7 37.666 -0.034 T-181U Blanchardstown Bypass 46.3 46.424 0.124

T-182a Herbert Rd, Blanchardstown 46.5 46.703 0.203

Int183c Herbert Rd, Blanchardstown 46.8 46.743 -0.057

T-189 Upstream of Blanchardstown Bypass 48.1 48.081 -0.019

T-191D Corduff Bridge 48.2 48.276 0.076 T-192U Corduff Bridge 49 49.228 0.228 T-195 Snugborough Rd 51.3 51.348 0.048 T-222bU Mulhuddart Bridge 54.5 54.677 0.177 T-231 Parslickstown Rd 56.5 56.504 0.004 T-233aD Parslickstown 57 57.165 0.165 T-243 Huntstown Culverts 59.7 59.599 -0.101 XS108A Damastown Culverts 60.4 60.343 -0.057 XS126A N3 Culverts, Clonee 62.3 62.332 0.032 XS156AD D/S of confluence, Clonee 63.7 63.765 0.065

XS205AU Castle Stream, near Loughsallagh 63.9 63.965 0.065

XS214A Castle Stream 64.6 64.486 -0.114 XS218A Castle Stream 65 64.755 -0.245 XS222A Castle Stream 65.5 65.334 -0.166 XS229AU Bridge d/s of Beechdale 66.4 66.155 -0.245

XS236A Larchfield Estate, Dunboyne 67.3 67.302 0.002

XS243A Larchfield Estate, Dunboyne 67.5 67.432 -0.068

XS250AD Rooske Rd Bridge 67.8 67.809 0.009 XS265A Maynooth Road Bridge 68.5 68.422 -0.078 It can be seen from the table that the largest difference is an underestimation of 245mm. The Mean Absolute Error of the data set is 95mm. This gives the correct indication of how accurate the calibration is.


074515001Rp0032ADE.doc 46 Rev F01

8.2.2 November 2000 The calibration levels available for the November 2000 storm are generally for the upper Tolka region. They proved quite beneficial as there were no calibration levels available in this region for the November 2002 event due to difficulty in obtaining access. The calibration for this event mainly involved adjusting mannings roughness coefficient for certain lengths of the river and altering the calibration coefficient of individual structures. Details of adjustments during calibration are shown in Appendix C2 – Flow Control Structures. Table 8.3 shows a comparison of calibration levels to modelled levels used to calibrate the November 2000 event. Table 8.3 Calibration Comparison for 6th November 2000 Cross Section

Location Observed (mAD)

Modelled (mAD)

Difference (m)

T-615 Macetown Rd Bridge (Pinkeen) 59.77 59.571 -0.199 T-624 Powerstown Rd Bridge (Pinkeen) 62.32 61.771 -0.549* T-137U Scribblestown Rd Bridge 31.86 31.367 -0.493* T-243 Culverts at Huntstown 59.7 59.282 -0.418* XS125A N3 Culverts, Clonee 61.39 61.665 0.275 XS126A N3 Culverts, Clonee 61.782 61.824 0.042 T-260D Clonee Bridge 62.42 62.675 0.255 XS1-48-A Downstream of Loughsallagh 63.4 63.375 -0.025 XS1-63-AU Loughsallagh Bridge 64.3 64.278 -0.022 XS1-139-AU Bennetstown Bridge 70 69.666 -0.334 XS1-161-AU Old Railway Bridge 70.43 70.494 0.064 XS1A-2-AU Flat House Bridge 72.9 72.813 -0.087 XS1A-38-A Black Bull Bridge 74.1 74.5 0.4 The November 2000 calibration is not as accurate as the November 2002 calibration. The Mean Absolute Error of the data set is 243mm. However 3 calibration levels marked * are based on anecdotal levels supplied and may be overstated.


074515001Rp0032ADE.doc 47 Rev F01

9 DESIGN EVENT ANALYSIS

9.1 DESIGN MODEL

9.1.1 Design Rainfall InfoWorks RS computes rainfall for a given design storm using the FSSR16 rainfall boundary and input parameters shown below. The computed 75% Winter Profile was used as this produces the worst case scenario and severe floods on the Tolka generally occur between October and March. The following parameters were required to generate the design rainfall: �� M5 2D Rainfall 2 day rainfall with a return period of 5 years �� M5 25D Rainfall 25 day rainfall with a return period of 5 years �� Jenkinson Ratio (r) Ratio of 60 minute M5 rainfall to 2 day M5 rainfall �� Country England England was chosen as the rainfall generator since the historical event analysis graph for the Tolka approximates the English growth curve more so than the Irish growth curve (see Chart 5.2, Chapter 5). The hydrograph scaling factor was then used to fit the specific growth derived as outlined in table 9.1. This was required as there are only 2 growth curves available in InfoWorks RS.

9.1.2 Critical Duration A series of runs were carried out to find the critical duration of the catchment. A graph was produced showing all critical durations tested using the 100 year 38 hour run as the base case scenario. The graph is shown in Appendix D3. The 26 hour storm gave flood levels quite similar to the 38 hour storm in the Dublin City area but increased flood levels of up to 150mm in the upper catchment. The 20 hour run gave increased flood levels in the upper catchment of up to 200mm but a reduction in flood levels in the Dublin City area. For this reason the 26 hour storm was chosen as the critical storm. The 12 hour, 32 hour and 44 hour storm were also tested. An areal reduction factor (ARF) was applied to the catchment so that point rainfall could be scaled for the overall catchment. This was read from Fig 5.1, Volume II of the Flood Studies Report using 26 hours as the duration (D) and the overall catcment area (A) as opposed to the sub-catchment area. An areal reduction factor of 0.95 was applied to the Tolka catchment for the 100 year design storm. Design runs were also carried out for the 10 year, 25 year, 50 year and 200 year return periods, with a critical duration of 26 hours and an Areal Reduction Factor of 0.95. A hydrograph scaling factor was applied to these design runs so the maximum flow for the design storm coincided with the flow in the historical event analysis graph in Chart 5.2 Chapter 5. Table 9.1 shows the factors which were applied. Table 9.1 Hydrograph Scaling Factors Design Storm Hydrograph Scaling Factor 200 Yr 1 100 Yr 1 50 Yr 0.9 25 Yr 0.82 10 Yr 0.8

9.2 EXISTING

9.2.1 River Flooding The existing scenario (November 2002 storm event) is compared to the 100 year event. This comparison is shown in Graphs 1 – 3, Appendix D3. These graphs represent the difference in levels


074515001Rp0032ADE.doc 48 Rev F01

modelled between the existing scenario design event and the November 2002 flood event. The following conclusions were made: Dublin City Council Administrative Area �� M50 to Finglas Factory Weir: There is no significant difference between the 100 year event and

the November 2002 event in this length of the river in the Dublin City Council area. The difference in flow in this area is minimal hence there is very little change in level.

�� Finglas Factory Weir to Distillery Weir: Flood levels in this region range between 0 and 200mm below the base case scenario. The November 2002 event reaches a higher peak flow than the 100 year design event. This is because urban catchments have a quicker response to higher intensities due to the extent of paved areas. The rainfall profile for the November 2002 event has a higher intensity than the 100 year design event and hence gives a higher outflow hydrograph.

�� John McCormack Bridge to Estuary: The 100 year flood levels in this area are between 200 and 300mm lower than the November 2002 flood event. This is due to the tidal boundary which was used. An average tide was used for the 100 year event; this is lower than the actual tide in November 2002. See section 9.2.2

Fingal County Council Administrative Area �� Keypac to Parslickstown: The 100 year flood levels in this area are between 100 and 200mm

above the November 2002 event. A greater flow is coming down the river during the 100 year event, leading to the overall increase in levels.

�� Mulhuddart: The 100 year flood levels at Mulhuddart Bridge are approximately 200mm above the November 2002 event. There is a flow difference of approximately 8 cumecs in this region, which causes an increase in flood levels in this area due to storage effects described below.

�� Blanchardstown Road North: The 100 year flood levels upstream of Blanchardstown Road North are approximately 500mm above the November 2002 event. The flow here, almost 5 cumecs higher than the November 2002 event. The conservative normal distribution of the design flow hydrograph increases the amount of flow surrounding the peak of the event when compared to the November 2002 event. This results in additional storage at the triple barrel culvert and increased levels.

�� Snugborough Road, Blanchardstown: The 100 year flood levels between Snugborough Road and Blanchardstown Road North are about 300mm higher than the November 2002 event. The flow coming down the river at Snugborough is approximately 3 cumecs higher for the 100 year event. More storage occurs at the triple barrel culvert as outlined above.

�� Herbert Road, Blanchardstown: The 100 year flood levels at Herbert Road averages approximately 100mm above the November 2002 event. This is due to a greater peak flow in this area.

Meath County Council Administrative Area �� Upper Tolka Catchment to N3 Culverts, Clonee: The 100 year flood levels in the upper Tolka

catchment range between 200 and 400mm above the November 2002 event. There is a larger flow in the upper catchment for the 100 year event, causing the rise in levels.

�� Castle Stream: There is no significant difference between the 100 year event and the November 2002 event on the Castle Stream. Flood levels for the 100 year event vary within 50mm of the base case scenario.

9.2.2 Tidal Flooding Dublin City Council Administrative Area The tidal reach of the Tolka extends to the Distillery Weir. A series of tidal/fluvial run combinations were carried out on the phase 1 model and their results shown in Figure 9.1. It can be seen from the graph that the tidal boundary dominates for most of the tidal reach, where current flood walls are generally adequate except for approximately 200m downstream of Distillery Weir, where inter - tidal conditions exist. Fluvial flooding dominates in this area; however the tidal boundary also has an influencing effect. With the exception of the inter - tidal area immediately downstream of Distillery Weir, the main risk occurring from tidal flooding is indirect tidal flooding resulting from storm sewers backing up (back


074515001Rp0032ADE.doc 49 Rev F01

pounding) or from surface water flooding (ponding) resulting from storm sewers having reduced discharge for local storm water drainage against flood levels. Figure 9.1 Design Tidal Run Comparison on Phase 1 Model

US of John McCormack Bridge

US of Dart Bridge

US of Annesley Bridge

US of Luke Kelly Bridge

US of Distillery Weir

Tolka Cottages

0

1

2

3

4

5

6

7

8

9

10

0 500 1000 1500 2000 2500 3000 3500 4000Chainage (m)

Wat

er L

evel

(m)

Q2yr+H2.79m

Nov 2002

Qbase+H3.35m

Qbase+H2.95m

Tidal Extent

Q50yr+H2.20m

Figure 9.2 shows the same series of tidal combinations and their results on the proposed scheme. The results are very similar and differ only within the inter – tidal area following the lowering of distillery weir by 1m to a level of 1.90m and the channel widening to 20m in this area. This considerably lowers levels and is described further in section 12.4.4 Figure 9.2 Design Tidal Run Comparison on Proposed Scheme

US of John McCormack Bridge

US of Dart Bridge

US of Annesley Bridge

US of Luke Kelly Bridge

US of Distillery Weir

Tolka Cottages

0

1

2

3

4

5

6

7

8

9

10

0 500 1000 1500 2000 2500 3000 3500 4000

Chainage (m)

Wat

er L

evel

(m)

Q2yr + H2.79mNov-02Qbase + H3.25m Qbase + H2.95m Tidal ExtentQ50yr + H2.20m


074515001Rp0032ADE.doc 50 Rev F01

9.2.3 Tidal Gate Analysis Analysis of the affects of providing tidal gates at the outlet of the Tolka was undertaken, to determine if these would provide benefit in terms of risk reduction. The brief also requires the assessment of implications of general development aspects including a possible marine lake at Clontarf. Analysis was achieved by providing a non-return outlet node on the tidal boundary. It was found that under climate change scenario (Sea level of 3.35m OD) only a 130mm reduction in levels could be achieved. It is concluded that the tidal area does not have enough volume available to store the river flows even under a low flow scenario of 8m3/s.

9.3 FUTURE DESIGN

9.3.1 Urban Development The impact of future development was assessed within the Tolka catchment for both the 2011 and 2031 scenarios. A breakdown of sub-catchment detail is shown in Chapter 7, Table 7.2. An analysis of the development to 2031 with the inclusion of the N3 Road Scheme was compared to the existing scenario for the 100 year event. The following effects in flood level were noted: Dublin City Council Administrative Area: The overall effect of development in the Dublin City Council area is minimal. This is due to two reasons 1) because there is very little predicted development in sub-catchments 29-35 in the Dublin City Council area up to 2031. The urban catchment is almost at its threshold with respect to future development. And 2) because the additional flows attributable to development in the upper catchments are somewhat moderated due to the retention of upstream constraints to flow, and the attenuation benefits these constraints offer. Fingal County Council Administrative Area �� Mulhuddart: There is an increase in flood levels of approximately 200mm at Mulhuddart. As there

is only a slight increase in development in Catchment 23 of 4.4% the rise in flood levels is caused by the increase in flow, which arises from the cumulative increase in urban development from areas upstream in the Tolka catchment.

Meath County Council Administrative Area �� Upper Tolka Catchment to Old Railway Bridge, Bennetstown: There is an increase in levels of

between 0 and 50mm in this area. This is due to the increased development in the upper catchments.

�� Bennetstown to Clonee: The flood levels in this reach up to 100mm above the existing scenario.

There is a significant increase in urban development in Catchment 10, Catchment 17 and Catchment 21 increasing the percentage run off, hence increasing flood levels.

�� Castle Stream: Future development appears to have very little effect on flood levels along the

Castle Stream. There is a large increase in percentage development in Catchment 11 and Catchment 18; however these sub-catchments are small in area hence an increase in run off leads to only a slight increase in flood levels (approximately 30mm).

9.3.2 Urban Development & Climate Change As outlined in the Chapter 6 the impact of climate change suggests an increase in rainfall of 22% for the 100 year event. This is considered to be a very conservative approach. If a medium - low emission scenario was used the increase in rainfall would be about 10-15%. Both of these scenarios are analysed in the model, also taking development to 2031 into consideration. The influence of climate change on tidal boundaries (increase of 400mm) is also accounted for. Climate Change – Scenario 1 (10%) The climate change scenario with an increase in rainfall of 10% with development to 2031 on the proposed scheme 100 year event was assessed against the proposed scheme with development to


074515001Rp0032ADE.doc 51 Rev F01

2031 - 100 year event and the difference in flood levels were assessed. The results are shown in Graphs 4 – 6, Appendix D3. The following conclusions were made. Dublin City Council Administrative Area Flood levels increased between 0 and 400mm in this region. This is due to the 10% increase in rainfall and the increase in tidal boundary data of 400mm. Fingal County Council Administrative Area Flood levels due to climate change increase from between 100 and 600mm in the Fingal area. Flood levels at Blanchardstown Road North and Snugborough Road are worst affected by climate change. This is due to the cumulative increase in flow as a result of climate change throughout the catchment and the increase in storage in this area. Meath County Council Administrative Area The effect of climate change in the Meath County Council area results in an increase of between 0 and 200mm in flood levels. This is due to the 10% increase in rainfall for the extent of the catchment. Climate Change – Scenario 2 (22%) The climate change scenario with an increase in rainfall of 22% with development to 2031 on the proposed scheme 100 year event was assessed against the climate change scenario with an increase in rainfall of 10% with development to 2031 on the proposed scheme 100 year event. The results are shown in Graphs 7 – 10, Appendix D3. The equivalent tidal boundary was used for this case as for Scenario 1. The following shows what impact this change had on flood levels in the catchment. Dublin City Council Administrative Area The conservative increase in rainfall of 22% leads to a further increase in flood levels of between 150 to 400mm. There is a marginal increase in flood levels in the tidal reach, as a result of the cumulative increase in rainfall. Fingal County Council Administrative Area There is a further increase in flood levels of between 150 and 1600mm. The areas worst affected by climate change are again Snugborough Road (900mm), Blanchardstown Road North (1600mm), and Mulhuddart (1400mm). Meath County Council Administrative Area The further increase in flood levels in the Meath County Council area is generally between 0 and 400mm. However as can be seen from the graph there is an increase in flood levels of 600mm upstream of Clonee Bridge. This bridge does not have sufficient conveyance to accommodate a rainfall increase of 22%.

9.4 SENSITIVITY ANALYSIS The sensitivity analysis of a river model is a key element of model calibration when extrapolating to extreme design conditions. It is based on extrapolation of reliable data to obtain a range within which the extreme values lie. However, in the case of the River Tolka, there is highly accurate data available for the most extreme event ever recorded, the event of 14th and 15th November 2002. The sensitivity analysis of the model was to a certain extent undertaken through the following procedures: �� Calibration and Verification; �� Design Event Analysis; �� Development of Solution Options; �� Climate Change Assessment. Further sensitivity analysis involving global changes in Mannings number and percentage development was undertaken and the results analysed.


074515001Rp0032ADE.doc 52 Rev F01

9.4.1 Sensitivity Analysis of Roughness Coefficients The sensitivity analysis was carried out on the proposed scheme network for the 100 year event. No development has been included. Mannings roughness coefficient was increased globally (channel, floodplain and flow control structures) by 35% to assess the sensitivity of the model. Graphs 10 – 12, Appendix D3 show the comparison of the proposed scheme prior to and after mannings number adjustments. The following effects were noted: Dublin City Council Administrative Area M50 to Scribblestown Road Bridge The increase in mannings roughness coefficient in this area results in an increase of 150 to 250mm along the river. This shows the model is quite sensitive to changes in mannings number in this area. Downstream of Scribblestown Road Bridge to Cardiffs Bridge The effect of increasing mannings number in this area is minimal. Flood levels increase and decrease by between 30 and 50mm in this area. The channel is quite smooth along this length of river so the 35% increase is marginal. Upstream of Finglaswood Bridge to gauge at Botanic Gardens The increase in mannings number for this length of the river results in an overall increase of between 150 and 250mm in flood levels. The model is quite sensitive to changes in mannings number in this area. Botanic Gardens to Griffith Park Increasing mannings number in this area had the effect of changing flood levels by ± 200mm. This shows the model is sensitive to changes in mannings number along this reach. Drumcondra Bridge to Estuary Flood levels from Drumcondra Bridge to the Tolka Estuary increased by an average of 100mm as a result of increasing mannings number. In some areas flood levels increased by up to 300mm and other areas levels decreased by 150mm. Fingal County Council Administrative Area Damastown to Blanchardstown Road North Flood levels increased by approximately 200mm along this reach as a result of increasing mannings number. Blanchardstown Road North to M50 Increasing mannings number by 35% has the average effect of increasing flood levels by between 100 and 300mm. In some areas the flood levels increase by up to 470mm and other areas the flood levels decrease by up to 100mm. The model is sensitive to changes in mannings number in this area. Meath County Council Administrative Area Upper Tolka to Clonee Flood levels increase along this reach by between 80-380mm due to changes in mannings number. There is an average increase of approximately 180mm in the Upper Tolka region. This shows the model is quite sensitive to changes in mannings number in this area. Castle Stream Increasing mannings number along the Castle Stream results in an average increase in flood levels of approximately 200mm. In some areas however the flood levels increase by up to 400mm. Hence, the Castle Stream is sensitive to changes in mannings number.

9.4.2 Sensitivity Analysis of Percentage Development The sensitivity analysis was carried out on the proposed scheme network for the 100 year event. Development to 2031 has been included and percentage development has been increased by 10% of its 2031 value. A comparison of the increased development to the 2031 development is shown in Graphs 13 – 15, Appendix D3. The following effects were noted: Dublin City Council Administrative Area


074515001Rp0032ADE.doc 53 Rev F01

Flood levels increase between 0 and 35mm as a result of increased development in the Dublin City Council area. The increase is minimal as there is a very high runoff already, prior to the increase in development. This shows the model is not particularly sensitive to increased development in this area. Fingal County Council Administrative Area Flood levels increase in the Fingal area by between 0 and 80mm. The increase is particularly evident at the Snugborough Road and Blanchardstown Road North culverts. This is as a result of the cumulative increase in flows due to the increase in development. Other areas in Fingal show little change in flood levels as the runoff prior to the increase in development is already very high. The model is not sensitive to changes in development in this area with such a high percentage runoff. Meath County Council Administrative Area Increasing 2031 development by 10% has very little effect in the Meath area. Flood levels increase by between 0 and 15mm. This is again due to the very high percentage runoff, leading to little change in the overall runoff with increased development.


074515001Rp0032ADE.doc 54 Rev F01

10 CATCHMENT MANAGEMENT

10.1 FUTURE CATCHMENT MANAGEMENT FRAMEWORK The examination of historical flooding in the Tolka Catchment demonstrates that ultimately, flooding has to be seen as a natural phenomenon associated with the variability of climate conditions, with inevitable large floods from time to time. The management of the catchment and in particular, the management of land-use development as it impacts on the floodplain and on the river channel can have an influence on the causes of flooding and its consequences. Natural river channels tend to have capacity for moderate flows, typically equating to an average annual flood, so that extreme floods have to be accommodated by overbank flow and storage in what is defined as the floodplain of the river. Historically, socio-economic development was closely related to the waterways which provided basic transport, water supply and wastewater disposal services in support of human activity. For that reason, most Irish towns are located on rivers, with the older developments often in or close to the floodplain reflecting the significance of the river for the original town development. At the time (18th and 19th centuries) the impact of fluvial flooding of properties was a minor nuisance in the overall context of the time. As the country continued to develop, significant problems began to arise associated with the relationship between urban development and rivers. These included pollution of rivers as effluent loads increased, loss of river environment due to engineering of river systems (culverting, canalisation and other modifications of the natural channel), as the need to prevent flooding of property became more important. Historically, Local Authorities have been responsible for public health and water quality issues in accordance with the legislative and policy framework set down by the Department of the Environment, Heritage and Local Government (DOEHLG), with the Office of Public Works (OPW) having responsibility for arterial drainage and flooding control at river catchment level. The adoption of the Water Framework Directive (WFD) provides for integrated catchment management embracing all aspects of river systems with the objective of establishing “good” water quality status by 2015. The scope of the Directive covers all aspects of water resources in a catchment including quantity and quality. In Ireland, the Directive will be implemented through the establishment and operation of a total of seven River Basin Districts (RBD’s), currently being established. The Eastern RBD (ERBD) managed by Dublin City Council as lead authority embraces the Boyne, Liffey and Wicklow River catchments and will include the Tolka Catchment. It is envisaged that the implementation of the ERBD will include the development of catchment management and sub-basin plans geared to delivering the objectives of the WFD. This integrated approach to catchment management will cover issues such as water quality and quantity, water resources management, floodplain management, fisheries and general habitat protection. It will also include the management of activities liable to impact on water quality and quantity in the basin including effluent discharges, influence of agriculture, forestry and other land-use activities. It is recommended that the River Tolka should be considered a priority catchment for the development of an integrated catchment management plan having regard to availability of a substantial baseline data set arising from this and other studies. In the context of flooding, a catchment management strategy developed in this way would involve all stakeholders in the development of a comprehensive understanding of the river system, in the setting out of objectives and in the implementation of action plans for more effective catchment management. As part of this exercise, a greater understanding and public awareness of flood issues would be developed including the risks of flooding, the factors which give rise to it and application of flood warning systems in the context of rainfall, river flow and water level monitoring


074515001Rp0032ADE.doc 55 Rev F01

10.2 FLOODPLAIN MANAGEMENT FOR THE RIVER TOLKA Effective floodplain management requires a range of measure to identify and seek to manage flood risk in the catchment. The application of these measures to Irish catchment is under review by OPW and Adamson (2003) provides an overview of the application of these measures in Ireland.

10.2.1 Floodplain Mapping Floodplain mapping is a fundamental component of floodplain management, requiring the production of flood risk maps to visually define flood risk areas as a basis for communicating flood risk to the public and in particular to assist in the management of future development. Floodplain maps for the area have been produced in conjunction with the Final Report, derived from the River Tolka model. These maps provide an indicative outline of the areas at risk for floods of defined return periods. They can be further used to define flood depths, flood storage volumes, etc., to assist in evaluating the impact of development in the catchment. The interpretation of flood risk maps requires realistic understanding of what they represent, the fact that they correspond with potential flooding on an extreme return period basis, that inevitably they are approximations based on direct correlation of flood levels and ground levels and may not fully reflect the flow paths and obstructions which may influence flooding on a localised level. Nevertheless, these maps provide a realistic basis for communication of flood risk to all stakeholders in a catchment, leading to a better understanding of measures which might be taken to minimise that risk, particularly in so far as this might affect existing properties and planned new developments. Therefore, floodplain maps would be an important element of an integrated overall catchment management framework for the Tolka.

10.2.2 Flood Awareness and Emergency Planning Awareness of flooding is an important component of floodplain management. Obviously the greater the awareness, the better peoples ability to prepare for and respond to flood events. Flood awareness in relation to the River Tolka is particularly high at present. However, over time this can lapse as the memory of the November, 2002 flood recedes. The ‘Tolka/Castle Stream’ channels in normal flow conditions give no indication of the scale of flooding which they can support in an extreme event. Floodplains have traditionally been seen as desirable places to live, where the risk of flooding was implicitly accepted. Where severe or frequent flood events occurred, flood defences were constructed or property floor levels were raised to give people some protection, but this has led to a perception that only structural measures are effective. There is however little appreciation of the risks associated with structural works, including potential backing up of floodwaters, scouring of foundations and loss of access. The causes of flooding and activities that can lead to increases in flood risk (e.g. increased runoff from impervious surfaces, development in the floodplain, etc) are generally not well understood. Furthermore, awareness of flooding potential falls in the years between major flood events, often leading to inappropriate development, e.g. river-bank buildings, extensions, etc. Ensuring continued flood awareness is a long-term but necessary task in floodplain management. Many countries have established awareness programs that comprise road signage, public education, emergency response planning, training, and support elements. The objective of emergency response planning should be to prepare for flood events and reduce the risk to life and property. Particular attention should be given to the following items:

�� Existing emergency plans: upgrade plans to take account of different types of flood events;

�� Communications: install redundancy or backup for all emergency response organisation communications;

�� Traffic routes: identify and maintain evacuation routes and routes for emergency response organisation vehicles;

�� Key utilities and institutions: protect pumping stations, electricity supplies, telecommunications, hospitals, etc from flood impacts;


074515001Rp0032ADE.doc 56 Rev F01

�� Flood Pathways: Identify flood pathways and include measures to section of flood risk areas, and provide flood return routes in the event of defence failure.

�� Testing of emergency plans: conduct regular testing of emergency plans, including “what if” scenarios.

Dublin City Council have recently undertaken a review of the Major Emergency Plan (MEP).

10.2.3 Flood Forecasting and Warning At present the flood forecasting system in Ireland relies on Met Éireann to issue storm warnings, and eyewitnesses to report where flooding is occurring. Emergency response is then initiated by the Local Authorities and emergency response organisations according to procedures already in place. A more effective flood forecasting system will rely on linking weather radars, rainfall stations, river/tidal gauging stations and eyewitness accounts to a central co-ordinating office, which can then issue warnings as appropriate. Total reliance on these systems would not be appropriate for the River Tolka, as a relatively small catchment with short time to peak of the flood. Flood forecasting is very catchment specific and requires a good knowledge of local conditions. Real-time flood modelling is an emerging technology to support forecasters but requires detailed planning, good management and judgement to ensure it works effectively. The flood warning system might comprise of media alerts, automated telephone/fax calls, public broadcast systems and door-to-door contact that can be utilised at different stages of the event. Dublin City Council has a well developed telemetry system to monitor the water supply and drainage services on a regional basis on behalf of the local authorities in the metropolitan area. The Dublin City Coastal Flood Risk Assessment Study (DCFRAS) is examining the requirements of a flood warning system. This system will integrate the recommendations of this report for a river flood module, integrating rainfall, flow and water level/tidal data, as a basis for flood forecasting. This telemetry based system will collect real-time data which can prompt alert and emergency response as appropriate.

10.2.4 Local Flood Protection There are now numerous methods available to provide local flood protection for property. It has been shown that modest investments in temporary or local protection can bring significant reductions in flood damage. Traditionally, flood protection was afforded by walls or embankments that acted as a first line of defence. Where defences failed or were non-existent, then emergency response organisations depended on the use of sandbags as a flood barrier. This is now recognised as a very reactive approach to flood defence. Local flood protection aims to address the residual risks that dedicated flood defences do not cover due to cost. Local protection measures should obviously be just one element of an integrated floodplain management plan and not be relied upon solely. There are three different type of local flood protection as given by Bramley & Bowker (2001) and shown in Figure 10.1: 1. Temporary and Demountable Barriers: enhance existing defences or protect otherwise

undefended areas by preventing floodwaters reaching the property. Available barriers include rigid panels, freestanding, weighted structures and filled containers with liners;

2. Moveable Barriers: designed to seal potential flood routes into a property. These include flood

boards in doorways, flood skirts, airbrick covers and non-return valves on pipes 3. Construction Techniques and Materials: several UK guides are available detailing for home

owners how to improve the flood resistance of property and to repair/restore property after flooding

4. Property Defences: Householders and businesses can fit removable barriers and maintain sand-

bags and other devices to further increase their security against floods in vulnerable areas.


074515001Rp0032ADE.doc 57 Rev F01

Figure 10.1: Source, Pathway, Receptor Model with Local Flood Protection Measures The application of local flood protection measures is not only site specific but very dependant on the flood warning system to be effective. There may also be some instances where demountable barriers can prove to be more cost-effective than permanent structural works, especially for small developments. As suggested by Crichton (2003) for the UK, the Irish government should look to investigating the viability of local protection measures and to providing some form of financial assistance to companies and property owners who wish to adopt them. Planning and building controls should take account of the use of local protection measures, particularly in the context of re-development of brownfield sites in the Dublin City Council area where achieving desirable floor levels might be problematic.

10.2.5 Maintenance of Watercourses Legal and environmental constraints severely limit the scope of major maintenance of rivers. The OPW, as part of its National review of flood management are examining these issues with a view to establishing appropriate policies for general application. Flooding can be worsened by blockages in a watercourse, particularly where items like fallen trees, shopping trolleys, large household items, etc become trapped at culverts or bridges, restricting flood flows. In particular anecdotal evidence suggests that flooding at Newtown Bridge during the November 2002 event may have been exacerbated by debris build up. Additionally large trees above Finglas Road Bridge were required to be removed during the flood to prevent a possible blockage. Regular watercourse maintenance is therefore an essential activity towards reducing local flood risk. Infill dumping in floodplains may also result in a reduction in floodplain volume and therefore should be discouraged. Presently, the responsibility for watercourse maintenance rests with the riparian landowner, whose property boundary will normally extend to the watercourse centreline. Irish legislation is not explicit however in setting out the rights and responsibilities of riparian landowners and there has been little education to date regarding this. In many cases, blockages can be caused by people dumping items into the watercourse, without regard to the consequences. In the past, many Local Authorities had dedicated river crews that regularly remove blockages on the larger rivers to reduce the risk of flooding or erosion, while maintenance on the smaller watercourses is usually carried out on an as-needs basis. However environmental constraints have made undertaking this work difficult. Although the maintenance of watercourses has been undertaken to some degree, there are several actions required to improve maintenance regimes:


074515001Rp0032ADE.doc 58 Rev F01

1. Legislative Review: the rights and responsibilities of both riparian landowners and the drainage authorities must be clarified;

2. Public education: people must be informed of the potential risks created by blockages in a

watercourses and to prevent future dumping; 3. Limit culverting: culverts block easily and are difficult to access, particularly when fitted with

trash screens. All culverts must be accessible for and subject to planned maintenance; 4. Maintain and improve watercourse access: an undeveloped riparian corridor should be left

adjacent to all watercourses to allow maintenance access and for environmental protection; 5. Tighter consent procedure for watercourse structures: many bridge and culvert structures

have been built that are undersized. Hence they can block more easily than properly designed and consented structures;

6. Balanced Approach: maintenance objectives must be balanced against environmental objectives

within an overall floodplain or catchment management framework; 7. Funding: Adequate maintenance funding is required on an ongoing basis to ensure that

objectives can be met. Strategies to deal with these required actions are a subject of the OPW National Flood Policy Review Group.

10.2.6 Sustainable Drainage Systems Urbanisation can have a profound effect on catchment hydrology. The conversion of relatively permeable, greenfield areas to impermeable surfaces (roads, roofs, etc) has meant that rainfall now runs off the land much more quickly and in greater volume. Piped drainage systems are intended to convey this runoff as quickly as possible to receiving waters, but in many catchments this has led to an increase in downstream flood risk, particularly for lower return period events. Sustainable Drainage Systems (SuDS) are effective technologies which aim to reduce flood risk, improve water quality and enhance biodiversity and amenity. They rely on imitating natural processes to slow down and store excess storm water runoff from urban developments. However, their impacts in preventing large scale flood effects can be over emphasised. SuDS should not be considered as a method of river flood protection for extreme events. However they can provide a reduction in the frequency and scale of smaller events and offer many environmental benefits. They are particularly useful in reducing local flood risk in the context of small to medium sized sub-catchments and in conserving low flows as well as in reducing pollutant loads. SuDS have been successfully used throughout the world for many years, particularly the US and Australia, and are progressively being introduced in the UK. Local Councils in Ireland are familiar with the principles of SuDS and some have developed guidelines, but they are still relatively new to their application. Land-take, maintenance responsibility, liability issues and cost are now shaping up to be the most significant issues affecting the implementation of SuDS in Ireland. Most critical is the issue of responsibility for their maintenance, both from a practical and legal point of view and resolution of this issue is critical to their future successful application. As part of the GDSDS, regional policies are being prepared that will facilitate a uniform and consistent approach to the provision of stormwater drainage infrastructure across the Greater Dublin area. These policies will recommend that SuDS be implemented on all new developments wherever practical and outline how to select and design such systems for different objectives. The policies also promote the importance of the catchment management framework, floodplain management principles and environmental impact assessment. It is recommended that these policies be implemented immediately upon their adoption by the local authorities.

10.2.7 Flood Risk Concepts One of the most difficult concepts to convey to the general public and indeed professionals is the risk of flood events occurring. This has been expressed in terms of return period in the foregoing sections.


074515001Rp0032ADE.doc 59 Rev F01

By determining the risk of flood events in terms of return periods, we introduce the concept of periodicity, the inverse of which is the probability in any one year. However, the same probability applies for any year, including the year directly following a large event. In other words, a once in 100 year event has an occurrence risk of 1% in any year but does not preclude occurrence of a similar event in two consecutive years. (Occurrence in 2002 does not suggest a similar event in 100 years in the future).Table 5.5 below indicates various methods for communicating risk in different situations, with particular regard to flooding.

Table 10.2 Risk Concepts

Mean

Annual Flood

Five year flood

Ten year flood

Twenty five year flood

Fifty year flood

One Hundred

year flood

Two Hundred

year flood

Five Hundred

year flood

One Thousand year flood

Return period 2.33 5 10 25 50 100 200 500 1000 % Annual probability

of occurrence 42.9% 20.0% 10.0% 4.0% 2.0% 1.0% 0.5% 0.2% 0.1%

Chance of occurrence in any

year 1 in 2 1/3 1 in 5 1 in 10 1 in 25 1 in 50 1 in 100 1 in 200 1 in 500 1 in 1000

Chance of occurrence in 70

year lifetime period* 94% 76% 51% 30% 13% 7%

Chance of occurrence twice in

70 Year lifetime period**

35% 18% 6% 1% 0.3%

*Valid for T greater than 10 years. **Valid for T greater than 25 years. River flood alleviation therefore must be considered in the wider context of risk. The risk of flooding from rivers, local storm water networks or from tidal events cannot be eliminated. However measures can be implemented to mitigate affects and reduce their frequency. Risk reduction must however be balanced with social, environmental and economic realities. For example there is a 7% chance that a flood in excess of a 1000 year return period will occur in the next 70 years. Designing for such events could effectively require relocation of significant urban populations and businesses. This would be the case in many low lying coastal and fluvial areas of Ireland. Therefore the community as a whole, including those organisations representing their well-being, and the insurance industry, must accept a realistic level of risk and take practical steps to mitigate any future impacts where feasible. In summary, the social, environmental and economic effects of endeavouring to eliminate all risk of flooding are so severe as to make this option impractical. The objective in this flood study of the River Tolka, therefore, is to establish a flood risk profile for the river, with particular reference to existing developments within the potential floodplain, and to design a scheme which will give the best possible level of protection which is considered to be the once in 100 year flood risk, i.e., a scale of flood for which the risk of exceedence would be no more than 1% in any year. The River Tolka flood alleviation scheme has been designed to account for the 100 year design flood using conservative rainfall runoff parameters, hydrograph shape and critical duration rainfall. Additional allowance is included for development to 2031, taking a conservative view in not including the positive benefits of the implementation of SuDS policies. Additionally a margin of freeboard is included, at, or in excess of 300mm on all defences. The November 2002 flood which is in excess of the 100 year design flood in terms of flow has also been assessed and the margin of freeboard provided. In addition to the structural protection measures other recommendations are made to provide an additional reduction in risk, through, planning, flood warning and integration of flood relief mechanisms within the Major Emergency Plans.


074515001Rp0032ADE.doc 60 Rev F01

10.2.8 Planning and Building Control Planning for flood risk must be given high priority at local, regional and national government levels. The preparation of a National Floodplain Management Policy by the OPW, in consultation particularly with the DoEHLG, should set the direction for future planning with regards to flood risk. Planning aspects fall under the remit of DoEHLG, although further guidance from both authorities will still be required going forward. Planning documents such as the National Spatial Strategy and the Local Development Plans should explicitly state how flood risk will be addressed through planning. Local Authorities must now zone areas identified as at risk appropriately commensurate with that risk. The flood maps produced could be expected to inform planning decisions in this way. Again, planning and development should be integrated with the catchment management strategy for the river. Developers and the public at large must be made aware of flood risk prior to new development. Riparian corridors should where practical be maintained free from any development to prevent loss of natural attenuation, to allow for effective river maintenance and to facilitate conservation of the river catchment. It is recommended that a minimum of 15-20m either side of the watercourse is retained, this will be required to be assessed on a case by case basis. The riparian corridors should be integrated into the development plans to indicate where these impact on minor watercourses, allowing developers to integrate the area into their development proposals at an early stage. Areas outside the riparian corridor but within the floodplain should only be developed if adequate compensation for loss of floodplain storage is made and if property floor levels are maintained 500mm above Q100 flood level. Many “brownfield” redevelopment sites are in “at risk” areas. Such redevelopment may be socially and economically desirable and where permitted should take full cognisance of the flood risk and incorporate appropriate flood protection measures to prevent inundation into or through the new development.

10.2.9 Summary In the context of an overall catchment management framework, therefore, a number of important recommendations are made which will assist in managing flood risk in the future, particularly for new developments as follows: �� Riparian corridors should where practical be maintained free from development along all

significant rivers and streams. It is suggested that a minimum width of approximately 30m of riparian corridors is desirable along significant rivers and tributaries and this should be increased where required by the floodplain extent. Arrangements for maintenance access should be provided along the river. It is recognised that provision will be required for essential crossings and developments of national importance such as transport corridors and the provision of these should be subject to flood impact assessment as part of the EIS and general development planning;

�� There should be strict compliance with the requirements of the Arterial Drainage Act regarding

Section 50 approvals for river structures such as bridges and culverts; �� In principle, the floodplain to the 1:100 year flood (1% risk per annum) contour should be

maintained free of development as far as practical. At a minimum, it is suggested that development should not encroach into the floodplain corresponding to the 1:50 year flood (2% risk per annum). This 1:50 year flood contour has been shown on the flood risk mapping to facilitate this condition and adherence to this principle would protect flood storage substantially in the floodplain. In any event, the guideline level for new developments in the flood risk area should be set at a minimum of 500mm above the 1:100 year predicted flood level;

�� Planning authorities should develop guidelines for permitted development in the different risk

category areas and should accommodate these principles in the Development Plans; �� It is suggested that there should be a coordinated planning approach to assessment and approval

of planning applications between the three authorities responsible for the Tolka Catchment;


074515001Rp0032ADE.doc 61 Rev F01

�� All significant developments impacting on the flood risk areas should be required to include a

Flood Impact Assessment, to identify potential loss of floodplain storage and how this will be offset in order to minimise impact on the river flood regime;

�� In other Countries consideration has been given to managed retreat from floodplains in certain

high risk areas as part of redevelopment to provide for lowering of water levels and property risk reduction in the future. This is not a practical option in the River Tolka floodplain. At a minimum, all redevelopment of brownfield sites within the floodplain should include provisions for maintaining effective protection against flood inundation from the river into or through the proposed development and should incorporate all other necessary measures to provide for effective drainage of the development.

These measures are considered desirable in the context of an overall management approach to the River Tolka Catchment. Nevertheless, it is recognised that further significant engineering measures are unavoidable at this stage in order to protect existing developments which are at risk and which cannot be relocated or otherwise defended in the context of general policy measures.


074515001Rp0032ADE.doc 62 Rev F01

11 CATCHMENT MANAGEMENT ALLEVIATION OPTIONS This chapter provides a quantitative analysis of the significant flood alleviation options considered and their resulting impact upstream and downstream. For a more qualitative analysis see Chapter 8 of the Final Report.

11.1 OVERVIEW OF OPTIONS CONSIDERED A key objective of the River Tolka Flooding Study is to develop cost effective proposals to reduce flood risk to existing development. The catchment management strategy outlined in the preceding section will provide some benefits in terms of better maintenance of the river channel, improved flood warning and response to flood events. However, the scale of flooding experienced in November, 2002 and in previous major floods can only be alleviated by implementation of significant engineering works in the areas affected. The principal options available for consideration are: �� Channel Improvement: widening and deepening of the river channel to increase its capacity to

accommodate flood flows. This option would require dredging of the channel to lower invert levels, optimise the available channel gradient and widen the channel cross section in order that water levels would be lowered sufficiently over the length of the river to avoid repeat flooding of properties currently at risk. The technical, environmental and economic implications of this option as the principal response to the brief are very significant;

�� Flood Defence Works: involving the construction of floodwalls and embankments to confine flood

waters to the river channel as the principle method of protecting properties at risk. These measures can be combined with minor or local improvements to the channel including deepening and widening where necessary to provide an adequate factor of safety against repeat flooding;

�� Bypass Overflow: consideration was given to the construction of a bypass to divert peak flows

from the River Tolka to the River Liffey. Such an option would appear to be technically feasible to divert flows upstream of Clonee to join the Liffey downstream of Strawberry Beds. Serious consideration of such a scheme would require a detailed study of the technical, environmental and economic implications, with particular focus on the degree to which such a scheme could alleviate the problem. Other local bypass channels were also considered.

�� Flood Attenuation Storage; involving the development of increased flood storage capacity in the

upstream catchment to retain flood waters in order that flows downstream to the Tolka River channel would be reduced. Such a scheme would require sufficient storage volume to effectively regulate forward flows for a wide range of rainfall events and would require appropriate flow controls to manage the flow regime within the capacity of the downstream system. As with other examples, a scheme involving flood storage could include combinations of channel improvements and flood defence works as part of an overall integrated scheme.

�� Property Relocation; where the costs of flood protection are considered excessive, the option of

relocating development out of the floodplain can be considered. Because of the extent of property damage associated with the November 2002 in the River Tolka, it is evident that this option is not socially or economically viable.

A multi criteria matrix was developed to represent the relative merits of principle options available as they relate to the River Tolka and is presented in Table 11.1. This matrix allows for a general summary understanding of the key environmental, cost and technical criteria used in assessing various principle scheme options in advance of detailed technical analysis.


074515001Rp0032ADE.doc 63 Rev F01

Table 11.1 Summary analysis of principle flood relief options

Risk Reduction Overview Option Details

Local Downstream

Environmental Impacts

Community Impacts Cost Sustain-

ability

Works located

in area at risk

Total

Major River Scheme In Dublin Area

3 3 -3 -3 -3 -1 0 -4 Channel Improvement

Local Channel Improvements 2 -1 -2 0 -1 1 3 2

Flood Defense Works

Walls Embankments 1 -1 -1 0 -1 1 3 2

Major scheme 3 3 -3 -2 -3 3 -3 -2 Bypass Overflow Local Diversions 3 -1 -1 -1 -1 2 -1 0

Major Scheme -1* 2 -2 -2 -3 3 -2 -4 Flood Storage Local Storage 0** 1 -2 -1 -2 3 -2 -3 Property

Relocation 1 0 0 -3 -3 2 3 0

* Increase in local risk directly upstream (-3) weighed against reduction in local risk directly downstream (2) = Total (-1) ** Increase in local risk directly upstream (-2) weighed against reduction in local risk directly downstream (2) = Total (0)

The results of the optioneering undertaken are outlined in Appendix E, Graphs (a) – (p), they should be read in conjunction with Table E1 – E6. These graphs are provided to represent the development of the scheme at various stages. They are therefore provided in a staged format as represented in Table E2 Model solution schedule, with the staged optioneering summarised in Table E3 Interim design incorporation. The charts therefore should be considered in a flip chart format with the changes resulting from different model summary runs compared to the previous model summary run when assessing impacts.

11.2 RIVER IMPROVEMENT WORKS River improvement works were modelled in the following areas and the following effects noticed (the reduction in levels is compared to the 100 year event): �� Luke Kelly Bridge to Distillery Weir: This area required minor channel improvements to relieve

local obstructions to flow as opposed to channel regrading. A maximum of 300mm was removed from the channel bed. This has the effect of reducing flood levels in this area by between 200 and 300mm.

�� Distillery Weir to Glasnevin Road Bridge: Again only minor channel improvements were carried

out. Works did not involve the adjustment of invert levels; however the channel profile was changed to give a constant invert throughout the cross section. Flood levels were reduced by up to 300mm in places as a result of this.

�� Finglas Road Bridge: Localised channel regrading, which involved the removal of flood debris was

modelled here. The channel profile was smoothened and mannings coefficient adjusted to reflect this. As a result head losses were reduced through the bridge. Flood levels were reduced locally by up to 500mm.

�� Mulhuddart: It was initially recommended to regrade from Mulhuddart Bridge to Blanchardstown

Road North. After further examination of model results it was found that these works added to the increase in levels at Blanchardstown Road North and Snugborough Road and did not reduce flood levels by a considerable amount in the Mulhuddart area. The option has since been changed to only include the regrading and widening works associated with the Mulhuddart Bridge improvements in this area.

�� Huntstown to Parslickstown: This area is quite overgrown and hence requires some minor

deepening and regrading. This has the effect of reducing levels of between 0 to 400mm in the area.


074515001Rp0032ADE.doc 64 Rev F01

�� Bennetstown to Damastown: Channel cleaning was modelled in this area, which involved adjusting

mannings for this length of the river. These works had only a slight effect on water levels. However, channel regrading; which involved modifying the channel profile was modelled between Clonee Bridge and Damastown. Flood levels were lowered by up to 400mm as a result of this regrading.

�� Castle Stream Confluence: Cleaning was undertaken from the confluence to Clonee Bridge. No

channel adjustments were made, it was sufficient to alter mannings coefficient to adequately represent the cleaning. This effected flood levels downstream of the confluence by up to 100mm.

�� Castle Stream – Dunboyne: As the Castle Stream is significantly undersized for the design flood

event, flood alleviation options were limited. Widening and deepening the channel was the most feasible approach. The bed levels were deepened by between 1.3 and 2.0m for the primary length of the stream and the associated bridges were underpinned. The model allowed this option to be tested and assess the effect the works had on downstream flood levels. The flood levels on the Castle Stream are reduced by up to 2.0m.

The works outlined above are noted to have adverse effects on the downstream areas, raising water levels by between 300 and 500mm at the culverts at Blanchardstown Road North and Snugborough Road. This impact on levels did not reflect downstream in the Dublin City Council area however, due to the storage provided by these culverts.

11.3 FLOOD DEFENCE WORKS The following flood defence works were implemented in the model (again the change in levels compare to the November 2002 event): �� Woodville Road: This bridge caused a large constriction on the river due to blinding of the

gaurdrails, which was evident during the flood in November 2002 (400mm afflux) and contributed to the flooding on Botanic Avenue and Millmount Avenue. The bridge was removed in the model, which removed the headloss associated with it. A new bridge with a higher conveyance capacity was designed for construction slightly upstream. The bridge is modelled as an arch bridge and has no afflux across it for the 100 year event. The overall effect in the Griffith Park Area is an increase in flood levels of about 200mm. This is due to the loss of flood plain storage on Millmount Avenue and the cumulative increase in flood levels from upstream areas.

�� Griffith Park: Gently sloping embankments in Griffith Park were constructed to contain flood waters

and protect houses in Botanic Avenue. This was modelled by shortening the extended cross sections and increasing the height of the spill units into the flood storage areas. The embankments were raised to 300mm above the 100 year flood level.

�� Downstream of Glasnevin Bridge: An embankment and retaining wall is modelled here to prevent

flood waters escaping to Botanic Avenue. The spill units joining the storage cells are adjusted to reflect this.

�� Mulhuddart: Mulhuddart Bridge has been upgraded in the model and a by pass culvert inserted. A

rectangular conduit was placed in parallel with the arch bridge. As a result the afflux has been reduced from 900 to 250mm across the bridge.

�� Loughsallagh Bridge: Due to the provision of embankments upstream of Loughsallagh Bridge flood

levels are increased in this area as a result of the reduction in spill area across the R157. This would result in increased risk to properties upstream. Therefore various bridge upgrade options were considered in the model. The result of this analysis is that Loughsallagh Bridge has been upgraded in the model to a single span bridge 10m wide and 1.56m higher than the original bridge. It is modelled as an arch bridge. The afflux is reduced from 600mm to 190mm following input of embankments. Upstream embankments are raised and the road crossing Loughsallagh Bridge is raised also.


074515001Rp0032ADE.doc 65 Rev F01

�� Clonee Village: The Clonee Stream was considered for diversion by means of a culvert to discharge downstream of the N3 culverts. Modelling this involved upgrading the culvert to an 1800mm diameter culvert and changing the layout of the outfall. Walls and embankments have also been constructed in Clonee to contain flood waters. This is modelled by increasing the spill levels spilling to flood storage.

11.4 RIVER DIVERSIONS The two principal river diversions considered were: �� A local bypass scenario of the River Tolka at Clonee; �� Large scale diversion from the Tolka to the River Liffey. The Clonee Bypass option on the River Tolka would involve the construction of an overflow bypass channel commencing upstream of the existing N3 crossing of the Tolka (Gunnocks House reach) and rejoining the river downstream of the Kepak factory. The bypass could be located to the east of the N3 dual carriageway, generally along the route of a small tributary stream. A trapezoidal channel, ranging from 8m to 4m wide, top to bottom, was modelled using the natural slope of surrounding lands. This option was considered so as to reduce flows through Clonee and the N3 culverts. This option reduced levels in the Clonee area by up to 180mm, providing to some extent a local solution. However the diversion showed adverse impacts downstream by increasing water levels in the Fingal area of approximately 100mm, resulting from a reduction in natural attenuation upstream of the N3 and having no effect in flood levels in the Dublin City Council area. The proposed large scale diversion of the River Tolka to the River Liffey would involve a diversion channel/culvert commencing at Loughsallagh and discharging to the River Liffey by way of a canal. The diversion was modelled using an abstraction node with a maximum of 45 m3/s reduction in peak flows. The results showed a significant decrease in flood levels in areas as far downstream as Snugborough Road, Blanchardstown. However the decrease in flood levels in critical areas in the Dublin City Council area was generally 200mm which would not significantly reduce the level of works required in Dublin City. Effective use of flood diversion would require automated real time control in order that the benefits of the bypass could be fully realised, depending on timing and intensity of rainfall. The loss of attenuation storage due to reduced levels significantly offsets the benefit of the bypass, particularly in the critical DCC area. Additionally as flooding was experienced at Strawberry Beds on the Liffey during November 2002, detailed assessment of flood risk in the Liffey would be required before this scheme could be recommended. In summary, although technically feasible, the bypass diversion options are not recommended based on technical performance assessment. Apart from technical considerations, these options would also have significant economic and environmental implications which could only be justified if they delivered proven benefits in terms of flood alleviation. A small scale drainage bypass is however recommended for the Clonee Stream.

11.5 ATTENUATION Storage attenuation schemes are often received by the public as an effective means of providing flood protection. In practice, during extreme weather events, natural attenuation areas can be full in advance of the peak run-off condition, so that they are rendered ineffective in attenuating the peak flow. Nevertheless, the provision of adequate storage by the construction of dams and spillways and associated engineering works could provide a measure of protection by reducing downstream flows.


074515001Rp0032ADE.doc 66 Rev F01

During the November, 2002 flooding event, an estimated 12,000,000m³ of water fell on the Tolka Catchment in the critical period. Although the land topography of the Tolka is generally not well suited to providing the required attenuation volumes, a number of possible attenuation areas were identified and investigated in detail in the model. The options were modelled by inserting a constriction (culvert/orifice) in the model and increasing its effect until the required water levels were achieved. Six potential attenuation schemes have been considered to reduce flood risk for existing conditions and the potential increased flood risk from climate change and future development. The attenuation schemes have been assessed on an individual basis to enable the effectiveness of each scheme to be thoroughly investigated. Additionally a combination of attenuation schemes was considered. A summary table of the results under peak flood levels and flows are provided in Table 11.1. Table 11.2 Impact of Attenuation Options upon Peak Flood Levels and Flows for November 2002 flood

Reduction in Peak Flood Level Compared with Existing Scenario

(mm) Attenuation Location Volume (m3)

Clonee Mulhuddart Snugborough Road

Finglas Bridge Drumcondra

M50 150,000 NA NA NA 50 80

Scribblestown 180,000 NA NA NA 40 45

Tolka Valley Park 197,000 NA NA NA 51 100

Upper Tolka 1 43,000 25 35 15 0 0

Upper Tolka 2 23,500 85 90 35 0 0

Combined Attenuation upon proposed Scheme 593,500 110 127 50 66 150

Note: Effectiveness of attenuation is not directly cumulative and reported results are based on their critical times of concentration that generally change between 26hrs to 38hrs due to attenuation effects. M50 Option The option of attenuation storage immediately upstream of the M50 crossing of the Tolka was considered (refer to Figure E1.3, Appendix E). This could be achieved by providing a flow control structure immediately upstream of the M50 motorway crossing, thereby regulating the flow through the M50 culvert and downstream to the Tolka. Looking at the levels in the area, it was estimated that a maximum peak flood level of 44.5m OD would be achieved. Any increase above this level would pose a flood risk to the interchange and upstream roads and development. At this level, the available flood storage would be approximately 155,000 m³. Modelling studies indicate that this storage could achieve reduction in peak flood flows for the November 2002 event of approximately 4% at the Botanic Gardens gauge, giving a reduction at Tolka Cottages, for example, of approximately 80mm in water level. This is a marginal benefit which does not significantly reduce flood risk generally. Scribblestown Option At Scribblestown, a storage scheme could be developed based on the relatively undeveloped floodplain adjacent to River Road (refer to Figure E1.2, Appendix E) by restricting the flows downstream of Scribblestown Bridge including the construction of appropriate embankments and spill structures, raising of road levels, etc. There is a significant degree of natural attenuation at present in this area due to limitations of Scribblestown Bridge and the disused Ashtown Bridge. Increased storage of approximately 180,000m³ could be achieved by raising maximum flood levels by an estimated 3.4m to a peak flood level of 35.2m OD. Ancillary works would include realignment of River Road to protect some adjacent properties which could be put at risk otherwise. Modelling analysis suggests that benefit in terms of reduced peak flows at Botanic Gardens would be approximately 4.7% for the November 2002 event, again providing only a marginal reduction in flood levels. It is suggested that this option does not offer adequate benefit to justify its implementation and may be unacceptable having regard to the potential increased risk of flooding to local properties. Tolka Valley Park Option


074515001Rp0032ADE.doc 67 Rev F01

Further consideration of additional storage attenuation focussed on Tolka Valley Park immediately upstream of Finglas Road Bridge (refer to Figure E1.1, Appendix E). The concept of the scheme involves the construction of embankments and spillways to increase flood levels for the November 2002 event by approximately 3.4m to a maximum level of 24.0m OD in the Tolka Valley Park. At this level, additional storage of approximately 197,000m³ could be provided. The impact on peak flood flows and levels is approximately 5% for the November 2002 event while further downstream in the critical Drumcondra area, the benefit would be less than 100mm reduced water depth for this event. It is suggested that this option does not offer adequate benefit to justify its implementation and may be unacceptable having regard to the potential increased risk of flooding to local properties. Upper Tolka 1 Option Within the upper Tolka reach of Quarryland, near Rathbeggan House, a storage scheme could be developed based on the relatively undeveloped floodplain adjacent to the Abandoned Railway by restricting the flows downstream (refer to Figure E1.7, Appendix E). The attenuation would require the construction of appropriate embankments and spill structures etc., with appropriate planning measures in place should the railway be brought back into use. Increased storage of approximately 43,000m³ could be achieved by raising maximum flood levels by an estimated 1.5-2.0m. Ancillary works would include an embankment to protect some adjacent properties south of the attenuation scheme which could be put at risk otherwise. Modelling analysis suggests that benefit in terms of reduced peak flows at Clonee would be approximately 2.6% for the once in 100yr flood with the proposed scheme fully constructed, again providing only a marginal reduction in flood levels. It is suggested that this attenuation option does not offer adequate benefit to justify its implementation. Upper Tolka 2 Option An attenuation scheme has also been considered at Piercetown, near the Black Bull Cross Roads (refer to Figure E1.6, Appendix E). A combined offline and online storage scheme could be developed based on the relatively undeveloped floodplain adjacent to Flat House Bridge on the River Tolka and the tributary immediately upstream of this location rising from the west. The attenuation area would be formed by restricting the flows downstream of Flat House Bridge and on the adjacent tributary. Works for this scheme would include the construction of appropriate embankments and spill structures, and raising road levels on Flat House Bridge. Increased storage of approximately 23,500m³ could be achieved by raising maximum flood levels by an estimated 1.5 to 2.0m. Ancillary works would include an embankment to protect some adjacent properties north and east of the attenuation scheme which could be put at risk otherwise. Modelling analysis suggests that benefit in terms of reduced peak flows at Clonee would be approximately 3.3% for the once in 100yr flood with the proposed scheme fully constructed, again providing only a marginal reduction in flood levels. It is suggested that this scheme does not offer adequate benefit to justify its implementation and may also be unacceptable having regard to the potential increased risk of flooding to local properties. Combined Attenuation The attenuation options considered above (i.e. M50, Scribblestown, Tolka Valley Park, Upper Tolka 1 and Upper Tolka 2) in isolation have been combined to assess the overall potential for flood alleviation. The total storage capacity of the combined system is approximately 600,000m³, that is approximately 4.5% of the total flood volume. Modelling analysis suggests that the benefit in terms of reduced peak flows at Clonee and Drumcondra would be approximately 4% and 11% respectively for the once in 100yr flood with the proposed scheme fully constructed, again providing only a marginal reduction in flood levels of 110mm and 150mm. As discussed above, it is suggested that this combined scheme does not offer adequate benefit to justify its implementation and may also be unacceptable having regard to the potential increased risk of flooding to local properties. The results of modelling indicate that even large-scale attenuation schemes offer limited reduction in flood risk throughout the catchment. Their impact in reducing flood levels is generally confined to local downstream reaches. The natural attenuation in these downstream reaches is then lost due to reduction in levels, which in turn leads to a reduction in the overall impact of the attenuation scheme.


074515001Rp0032ADE.doc 68 Rev F01

More critically, the exact real-time control of storage attenuation and forward flows and its location vis a vis vulnerable areas would be critical to the efficiency of attenuation in reducing flows, depending on the exact sequence and duration of rainfall. Additionally consideration of spatial and temporal variation in rainfall would be required if reliance was to be placed on such a scheme. There is a real risk that flood storage might already be full at a critical stage during a storm (e.g. for twin rainfall peaks) Overall, therefore, while some benefits could be achieved from increased flood storage in the catchment, particularly where flood storage can be located immediately upstream of vulnerable areas, the scale of works involved are very substantial, they have the potential to increase flood risk and public safety risk locally, while the benefits to be derived are relatively minor and cannot, of themselves, significantly mitigate the scale of flooding in the high risk areas. These options could however have a useful role in the future in augmenting the safety margin to take account of development or climate change impacts.

11.6 SUMMARY OF FLOOD MANAGEMENT OPTIONS CONSIDERED

Before considering recommended works in detail, therefore, it is possible to narrow the range of practical options based on the foregoing: �� Major dredging/channel improvement; does not offer a sustainable cost effective solution,

though it has the potential to provide a secure scheme with maximum reduction in flood risk to development. Because of economic, environmental and social impacts of such a scheme, only minor local channel improvements are considered, where these are necessary to remove critical flow controls on the river, in conjunction with flood defence works. An example is in the Dunboyne Village area, where the major constraints to flow from the existing channel and river structures have to be removed as part of an effective scheme to reduce flood risk in this area.

�� Flood walls/embankments; offer cost effective and practical benefit to protect low lying areas

from inundation, provided they are combined with the necessary modifications to the existing urban drainage system to prevent secondary flooding in these areas.

�� Channel diversions/bypass options; are not recommended on technical grounds primarily. �� Flood storage/attenuation options; have been tested and offer relatively modest benefit. In

general, it is proposed to retain as far as practicable, existing flood storage and constraints to flow to exploit attenuation in the existing floodplain. However, the development of additional attenuation storage volumes is unlikely to be effective in reducing flood risk to acceptable levels. The areas discussed above can remain as long-term options if required by climate change increases in rainfall.

Arising from this assessment, therefore, the following strategic solution options have been developed and examined in detail in this study:

1. Rural areas upstream of Clonee/Dunboyne; minimum alteration to river regime with restrictions on future development to conserve floodplains as far as practical and local protection options for isolated properties at risk.

2. Bennetstown/Bracetown area; Provision of minor earthen embankments and local re-alignment of existing roads to protect isolated properties at risk and to maintain transport routes during a flood event. Local channel maintenance is recommended to prevent obstruction in order to reduce flood risk locally. Planning policy should restrict future development in identified floodplains to conserve existing storage.

3. Dunboyne area; In this area, significant numbers of properties are at high risk of repeat flooding within the natural floodplain of the Castle Stream. This floodplain is largely dictated by the limited


074515001Rp0032ADE.doc 69 Rev F01

capacity for the existing channel and its associated structures. Therefore, improvement proposals in this area require widening and deepening of the existing channel and of the associated culverts.

4. Clonee area; The construction of appropriate embankments and walls are required to contain flood volumes in this area and to remove secondary flow paths causing inundation of property. Upgrading of the local drainage system is required to prevent secondary flooding in Clonee and in the Littlepace area. In general, retention of flood storage upstream of Clonee including the constraint of the existing N3 culvert appears desirable.

5. Mulhuddart/Blanchardstown area; Upgrading of existing flood defence barriers including extension of embankments and walls is required locally, together with modifications to local drainage systems to protect isolated properties at risk. Consideration is given in the detailed studies to local channel improvements and upgrading of existing bridge/culvert crossings, to reduce flood levels locally.

6. Glasnevin/Drumcondra areas; In this area, a combination of measures is proposed comprising of flood defence walls and embankments to contain flood levels and to cut off secondary flow paths which caused major inundation of properties in previous floods. Minor channel improvements are proposed to relieve local obstructions to flow, reconstruction of a number of bridges, modifications to Distillery weir and other works to provide an integrated comprehensive flood risk management scheme in this area. The scheme must be sufficiently robust to allow for additional future development upstream. In addition, the studies of the pipe drainage system in the GDSDS study will require to identify necessary alterations to the local pipe drainage system (flap valves, diversions and pumping stations) in order to prevent secondary flooding of low lying areas by back-pounding through the pipe network and to accommodate local flows from these low lying areas, including culverted streams and local run-off. A separate technical report will identify all of these measures in conjunction with completion of the GDSDS modelling studies of the local drainage systems.

These general proposals were developed and examined using the River Tolka Flooding model and the detailed proposals for flood relief on the River Tolka are discussed in the following chapter.


074515001Rp0032ADE.doc 70 Rev F01

12 PROPOSED FLOOD ALLEVIATION SCHEME

12.1 OVERVIEW Various combinations of works were identified in each area at risk from flooding and tested in the model for effectiveness in alleviating flood risk. The model was also used to assess the overall integrated impact on flood levels throughout the river for the combination of measures. The rationale for the recommended works in each area followed on from previous analysis where practical measures were identified on the basis of overview consideration of their practical implementation and effectiveness, acceptable environmental impact and likely cost effectiveness. Sub Model and Full Model testing of solution options were undertaken to determine their effects and also to determine sensitivity of the model. These options are amalgamated in a summary of model run assessments. Model results are compared to November 2002 flood levels developed for the existing case scenario to indicate general effects within the study reaches of the river network as a whole. The summary Model Run Assessment graphs are included in Appendix E to indicate the general development of the scheme. Detailed results of the individual model runs are available electronically within the model output files. The scheme proposals were assessed in terms of the design 100-year flood event (1% risk of exceedence per year), incorporating full development to the 2031 design horizon year. In addition, a suitable margin of freeboard has been applied to flood defence works, generally 300mm for hard defences (walls). Additional freeboard to 500mm is provided for soft defences (embankments) where these could be subject to alteration over the design life (settlement, etc). Where sensitivity analysis suggests a greater margin may be required this has been included (upstream of flow constraints, where rises in water levels are subject to attenuation affects from the constraint). The effects of widespread implementation of sustainable urban drainage practises for new development is not considered within the model and any benefits which such systems deliver for extreme flood events would be a further benefit. The model conditions assume similar catchment conditions to those which were obtained in November 2002, i.e., a high level of catchment wetness and run-off effectiveness with limited infiltration. In deriving the ultimate recommendations, a substantial matrix of model runs was undertaken comprising of various solutions and combinations of solutions. Overall, it is estimated that approximately 700 model and sub-model runs were undertaken during the model construction, verification and optioneering phase of the study. All options were considered in terms of variations (up or down) from recorded levels in the November 2002 event which has been taken as a baseline of extreme flood levels for the Tolka system, as this provides an extreme flood situation which is relatively well understood. Further sensitivity analyses were carried out for climate change and other factors and these are reported on later. This chapter reports on the recommendations arrived at following this analysis on an area by area basis. It is emphasised that these local measures form part of an overall integrated package of measures which was tested on the model as a whole and the implementation of which is required to be systematically proceeded with in order to meet the objectives of the scheme. Because of the scale of flooding experienced, the Local Authorities combined with OPW in a fast-tracked approach to implement immediate measures as identified in an interim report submitted shortly after the flooding. These measures are now well in hand and form an integral part of the overall scheme. Other measures in that report have been integrated to future recommendations, with some amendments based on the technical analysis and design decisions in the study. All of the works required are identified in this chapter, including works which are already completed, in progress or at planning stage.


074515001Rp0032ADE.doc 71 Rev F01

The proposed works recommended for implementation in this report are outlined in plan and section in the Drawings Volume 2, Final Report.

12.2 MEATH COUNTY COUNCIL AREA – PROPOSED WORKS

The flood risk areas of the River Tolka in Meath County Council are identified, together with proposed protection works, in Drawing Sheets 6 to 8, Final Report inclusive. Flood impacts and works associated with Castle Stream are represented on Drawing Sheet 7, Final Report. A considerable number of properties have suffered flooding within this area, with most located in Dunboyne due to flooding from Castle Stream and Clonee, due to flooding from the River Tolka. The proposed works in the Meath County area are considered working downstream along the river from Bennetstown to the Fingal County border:

12.2.1 Upstream of Bennetstown

This area is relatively undeveloped and design flows can be accommodated in the natural channel and floodplain. Flooding of properties was reported near Batterstown; however this has been confirmed to be a result of restrictions in a local side stream and is being dealt with separately by MCC. One property at Black Bull Bridge is identified as in the flood plain of the design flood, however floor levels are confirmed to be above flood level and this property has no history of flooding. The Proposed N3 National primary route dual carriageway interacts with the river in this section and this has been included in the model, the N3 upgrade includes proposals to attenuate all stormwater discharging from it and has been subject to a flood impact assessment. Large scale attenuation schemes where also considered in this area at Piercetown and Moyleggan as described in chapter 11.

12.2.2 Works at Bennetstown

Properties are considered at risk along the R157 Navan Road immediately south of the N3 junction, with three properties flooded in the 2002 flood. The N3 National Primary Road was also inundated locally. Works required to protect these properties include:

Item 1 Approximately 220m of road realignment of the Navan (R157)County Road parallel to the River sheet 8, Final Report (cross-section references 1-57A to 1-66A).

Item 2 Approximately 200m of earth embankment to a height of approximately 1m adjacent to the left bank of the river either side of the disused railway embankment (Drawing sheet 8, Final Report, cross-section reference 1-56A to 1-64A).

The works will require traffic management to deal with diversion of traffic from the existing road, reconstruction of fencing and services and provision of flapped drainage outlets from the protected area to the river. Channel cleaning to further reduce flood levels will assist in drainage from the relatively small local catchment area and reduce flooding on the N3. The removal of the I beam structures associated with the derelict railway bridge are recommended to reduce the risk of obstruction to the local road culvert. During the November 2000 and 2002 storm events it was noted that flood flows by-passed much of the developed area by means of overflowing down the disused railway. There are proposals to reopen this railway therefore it was important that any solutions developed accounted for the increased flows resulting from the removal of this bypass.


074515001Rp0032ADE.doc 72 Rev F01

12.2.3 Bracetown

Properties in this area are at risk from the River Tolka floodplain in the area between the R157 County Road and the disused railway embankment. To protect these properties, the following measures are proposed:

Item 3 Raise the vertical profile of the Navan (R157) County Road over a length of approximately 220m (Drawing sheet 8, Final Report adjacent river cross-sections 1-139-A to 1-152-A)

Item 4 Construction of an embankment approximately 180m in length to an average height of 1m connecting the realigned R157 County Road to the existing disused railway embankment. The flood embankment is to be constructed along the left bank of the River Tolka (Sheet 8, Final Report cross-section reference 1-134-A-to 1-139-A). This embankment and the realigned road will protect the properties north of the proposed embankment between the County Road and the disused railway line. It is then linked into the raised road and continues for 300m up the right bank of the river.

Item 5 As part of the road realignment, it is proposed to construct a bypass channel to the

Old Navan Road Bridge (R157 crossing of the River Tolka) to accommodate the full channel flow, offsetting the additional backing up as a result of works above.

�� It will be necessary to make provision for appropriate traffic management to facilitate

realignment of the road, renewal of fencing and road services and provision of appropriate drainage services to prevent impounding of storm water within the protected area. Cleaning of the channel bed is recommended to maximise the capacity of the R157 and Railway bridges and assist in drainage discharge locally.

The option of raising the road for 400m in this area was considered however as this obstructs the existing spill across the road significant backing up occurs, resulting in increased flood risk upstream. This was considered initially to provide additional attenuation storage in this area; however the downstream impacts were negligible, and did not warrant the increased risk to local properties. The additional consideration of this bridge being recognised by the local community, as the ‘sli’ (way) to Tara, provides some difficulties in undertaking works to the bridge. These works would be required to raise the road. Therefore the options outlined above allow for the bridge to overtop during rare flood events, maintaining the desired levels upstream. As an additional route to Dunboyne is proposed in the N3 dual carriageway scheme this option is considered adequate as alternative traffic routes will be available.

12.2.4 Gunnock House (River tolka) This 1.2km reach extends from approximately 500m downstream of the R157 Bridge to 300m upstream of the Loughsallagh Bridge. No habitable property within this reach is affected by the River Tolka Floods. However, some flooding has previously occurred over a small area of the N3 and flooding was reported at Gunnocks house however as this property is well above river levels this has been confirmed to be a result of restrictions in a local side stream and is being dealt with separately by MCC. The river is able to flow unimpeded over its natural floodplain without inundating property from flood flows. No significant flow constraints are identified within this reach of the river. Analysis of this reach indicates the natural floodplain extent and it is concluded that any future development should be either located outside this area, or alternatively above the design flood level with any associated loss of floodplain storage offset by matching storage.


074515001Rp0032ADE.doc 73 Rev F01

No flood alleviation measures are required in this reach. This portion of the N3 is currently planned for up-grade to a dual carriageway with proposed levels well in excess of flood levels.

12.2.5 Dunboyne Town Castle Stream Reach;

The works proposed in this reach relate to the section of the Castle Stream from upstream of the Maynooth Road through the town of Dunboyne to a point downstream of the disused railway embankment (Drawing Sheet 7, Final Report cross-section reference 2-71A to 2-15A). Within this section, extensive property flooding occurred and many other properties would be considered at risk in areas of Main Street, Castleview Estate, Woodview Heights, Hamilton Hall, Beechdale, River Court and Larchfield. The November 2002 flood and the previous smaller flood in November 2000 demonstrated that the existing channel and bridges on the Castle Stream cannot accommodate the run-off from its catchment of approximately 1,900 hectares. Moreover, the extent of development within the floodplain makes any solution based on flood defence works impractical. The proposed solution in this area involves lowering of stream bed levels, including underpinning of the existing bridges in order to lower flood levels sufficiently to prevent direct inundation of properties and to permit the local drainage system to work effectively. These works will release additional floodwaters to the floodplain upstream of Clonee which have to be dealt with in the context of protecting Clonee and the R156 area. The following works are proposed at Dunboyne:

Item 6 Widening and deepening of the Castle Stream channel between the Maynooth Road Bridge and a Local Access Bridge approximately 130m upstream of the Castle Stream confluence with the River Tolka. The total length of realigned river channel is approximately 1,860m over this section. Bed levels require to be lowered by between 1.3m and 2m over the principal section between the Maynooth Road Bridge and the Dunboyne Railway Bridge (Sheet 7, Final Report, cross-section reference 2-66A –2-5A). The channel works will include relocation of services, reinstatement and rehabilitation of the area following completion of the excavation works.

Item 7 Local regrading of the channel for a distance of approximately 230m upstream of the

Maynooth Road Bridge is proposed including cleaning of the culvert bed under this existing bridge (Sheet 7, cross-section reference 2-71A to 2-65A). These works will limit the floodplain and risk of flooding on the Maynooth Road.

Item 8 It will be necessary to underpin the Old Castle Bridge by approximately 1.5m. This

will require reconstruction of the abutments to the bridge and associated revetment to the river channel, and alternative option of local relocation and refurbishment could be considered.

Item 9 Deepening and underpinning or reconstruction of the Rooske Road Bridge (B61) will

be required (cross-section reference 2-50a) where it is required to lower the existing riverbed level by approximately 1.8m. The work will include as a minimum new abutments and revetment works connecting the bridge to the new river channel. The alternative option of complete replacement of the bridge could be considered.

Item 10 Underpinning of the existing Railway Embankment Bridge (B60) will be required to

lower the existing riverbed level by approximately 1.2m. This bridge is immediately downstream of the major housing estates affected by the flooding and lowering the channel here is an integral part of the scheme. As a minimum, new abutment works will be required combined with appropriate revetment to connect the bridge to the new river channel.

Item 11 Upstream of Dunboyne and the Castle Stream, outside of the study area proper, a

small bridge on the Summerhill Road (Newtown Bridge) has been noted as subject to blockages causing flooding to properties located directly upstream in the rural area.


074515001Rp0032ADE.doc 74 Rev F01

There are proposals to replace this bridge as part of constructing the Dunboyne slip road on the proposed N3 Clonee to Dunshaughlin Dual Carriageway scheme. As such, this work does not require to be included in the proposed River Tolka scheme.

�� Given the scale of works in this area, it will be necessary to accommodate services

crossings and to modify existing drainage outlets, where appropriate, to suit the re-aligned works.

Consideration should be given during detailed design to low flow rehabilitation measures such as provision of ripple pools, additional meanders and off line ponds or wetland areas, to improve the natural habitat and water quality. Measures should be included to ensure that maintenance access to the riparian environment is available whilst ideally this environment can be improved during construction of the necessary flood alleviation measures. However these improvements are not required for flood risk reduction and could be considered for integration to future development proposals in this area taking account of the recommended riparian corridors which to some extent provide for a linear park. This also provides for ongoing environmental protection and rehabilitation opportunities whilst offering potential educational and amenity uses (walk/cycle ways and nature areas) that can often be lost during urbanisation.

12.2.6 Loughsallagh to Clonee

This section relates to conditions at Loughsallagh, in the vicinity of the confluence of the main River Tolka River and the Castle Stream, downstream to the village of Clonee as far as the Clonee Bridge. In the Loughsallagh area, a very extensive floodplain is indicated for the design flood with significant numbers of properties at risk in the vicinity of the Dunboyne-Dublin Road (R156). There is also a risk of inundation of this road in the vicinity of Loughsallagh Bridge (B54) and downstream towards the roundabout immediately west of Clonee. Therefore, extensive embankment works are proposed to limit the floodplain in the vicinity of the R156 Road to protect both properties and road access to Dunboyne. The reduction in flood storage in this area results in some increase in flood levels with the result that the proposed embankments are designed to protect against flood levels marginally higher than those recorded in November 2002. The works required in the Loughsallagh area comprise:

Item 12 The construction of approximately 2,370 linear meters of embankments both north and south of the Dublin Road (R156) and extending along each side of the road over the area at risk (reference Sheets 7 and 8)

Item 13 Reconstruction of the Loughsallagh Bridge (B54) taking full design flow including

associated roadworks to maintain effective protection levels, services renewals and other ancillaries

�� Provision of local drainage to low-lying properties protected by flood embankments

between the R156 and the embankment immediately north of the confluence of the River Tolka and Castle Stream. Appropriate consideration will also have to be given to local drainage from other properties in the vicinity of these works.

Item 14 The analysis indicated a potential secondary flow path from the Loughsallagh

floodplain to the Clonee Stream which could increase flood flows to the south of Clonee village. This requires some local regrading using deposition of spoil from the dredging works on the Castle Stream and reinstatement of the lands to ensure effective cut-off of flood water access between the Castle Stream catchment and the Clonee stream catchment ( refer to Sheet 7, Final Report, Soil Deposition Area). These can be accommodated by a general raising and re-grading of the lands in order to minimise adverse impact on their general use as agricultural lands.

The raising of the road and provision of embankments upstream of Loughsallagh Bridge, serves to create an obstruction to the existing spill across the road and significant backing up occurs, resulting in increased flood risk upstream. This provides some additional attenuation storage in this area;


074515001Rp0032ADE.doc 75 Rev F01

however to ensure that flood risk is not significantly increased upstream and to facilitate upgrading of the road, Loughsallagh Bridge must be upgraded. A flow control structure was considered upstream of Clonee Bridge to ensure that relief measure in Dunboyne did not adversely impact on downstream areas at risk. It was found during modelling that the Clonee Bridge becomes surcharged, increasing levels in this area and is therefore providing this facility with no further works required.

12.2.7 Clonee Village Area

Clonee Village has suffered flooding primarily from inundation from the north and west where the River Tolka water levels exceeded bank level between Clonee Bridge (B53) on the R156 and the N3 Tolka culverts. This allowed waters to inundate the village causing major flooding of low-lying properties. Further secondary flooding has been associated with the Clonee Stream due to obstructions within the culvert and outlet backpounding. The works required in Clonee are:

Item 15 A flood defence wall over a length of approximately 300m is required on the right bank

of the River Tolka extending upstream from the N3 culvert. This wall will be constructed in private properties and will include reinstatement and re-routing of any drainage outlets/connections to the river (Sheet 7, Final Report, cross-section reference T-259 – T-256)

Item 16 Over the same section, it is proposed to widen and re-grade the channel of the River

Tolka, involving only minor deepening but also including removal of a wooden footbridge at approximately cross-section reference T-258 and its replacement with a local access bridge at appropriate levels to suit the flood defence works.

Item 17 Raising of lands to the northern boundary of the floodplain, immediately north of

Clonee between the R156 and the N3 to protect a number of existing properties at risk in this area and to integrate with an existing planning consent for this area.

Item 18 Provision of an embankment approximately 300m in length to the western side of the

R149 Road to prevent inundation of Clonee from the west.

Item 19 the provision of approximately 520m of new culvert to accommodate the Clonee Stream along the eastern boundary of the village, crossing the Navan Road (R156) and incorporating approximately 80m of tunnelled section across the N3 Dual Carriageway, out falling to the River Tolka downstream of the River Tolka N3 slip road culvert. The Clonee Stream Culvert is indicated on Drawing Sheet 6 (dotted blue) and the slip road culvert is at cross-section reference T-250 on the drawing. Construction of this culvert will include altering existing water-main services conflicting with the existing culvert and all other necessary accommodation works.

This option for Clonee Stream is the only option available as the current outfall is provided to the N3 road drainage (1200mm diameter) which in turn discharges upstream of the culvert at the N3 slip road. This system is therefore subject to backing up (flood levels 61.7m). In addition to this the current entry culvert upstream of Clonee is 750mm diameter which is not sufficient to convey peak flows. This results in backing up behind the village, allowing floodwaters to spill into local property. All options including pumping would therefore require upgrade of the existing system and a tunnelled discharge under the N3. Therefore a gravity solution of providing a bypass culvert integrated with the tunnelling for a proposed water-main in this area is considered the most reliable and sustainable solution. The foregoing represents the works required in the County Meath administrative area to reduce flood risk in Dunboyne and Clonee as well as individual properties in the vicinity of the River Tolka to the north and east of Dunboyne. It was noted that the entry culvert at Clonee crossing the Dual Carriageway and the downstream culvert crossing the N3 side-road were surcharged during the flood. However, in order to minimise the downstream impact, it is not proposed to upgrade these culverts, additionally assessment in the model


074515001Rp0032ADE.doc 76 Rev F01

indicates that the channel between the culverts would also require upgrading to achieve any benefit. Some works are proposed to streamline the entry to the upstream entry culvert in order to minimise head losses due to entry conditions, and cleaning of the intermediate channel is recommended to improve channel performance. The removal of the significant flooded area in Clonee and Littlepace along with channel improvement works results in increased flows downstream to the Fingal County Council Area. The existing risk along with these increased flows and accommodation of increased flows from future development in the catchment is therefore addressed.

12.3 FINGAL COUNTY COUNCIL AREA – PROPOSED WORKS

The flood risk areas of the River Tolka in Fingal County Council are identified, together with proposed protection works, in Drawing Sheets 2 to 6, Final Report inclusive, flood impacts and works associated with the two Pinkeen tributaries, are represented on Drawing sheets 10 and 11, Final Report. Though the main Tolka and its tributaries suffered serious flooding, this was largely accommodated within undeveloped floodplains with only a small number of properties inundated and other properties marginally at risk. The increased flows from improvements in Co. Meath are not mitigated immediately. There is no choice but to lower water levels in Co. Meath and there is an unavoidable increase in levels downstream in Co. Fingal as a result. Therefore consideration is given to these impacts in providing flood protection to properties in the Fingal area. The proposed works in the Fingal County area are considered working downstream along the river from the Meath County boundary to the Dublin City boundary. The works on these areas are briefly outlined as follows:

12.3.1 Huntstown to Parlickstown

This section of the River Tolka is shown on Drawing Sheet 6, Final Report and covers the River Tolka from the N3 Dual Caraigeway crossing to the downstream of Mulhuddart Bridge (B40) (cross-section reference T-252 to T-229). It includes consideration of the two Pinkeen Stream tributaries which join the Tolka at cross-section references T-233A and T-222 approximately. Kepak was not inundated in November 2002 due a floodwall being built after the November 2000 flood and sandbagging across the entrance road prior to the peak flood level. The RMF Ireland Ltd building also escaped being inundated due to floor levels being approximately 200mm higher than the flood peak. However both properties are at risk and improvements to protect Clonee translate into approximately 50mm increase in levels in this area. Therefore without modifications to the existing channel, water levels over this section would be marginally increased due to implementation of flood defence works upstream. In order to ensure flood risk in this area will not be increased compared with the previous flood, a marginal re-grading and widening of the river channel is proposed over a length of approximately 1800m. The extent of channel deepening is approximately 300mm on average to give a relatively uniform gradient and consistent cross-section profile to the channel (cross-section reference T-249 to T-232A). Overall, the works in this area are relatively minor and can be summarised as follows:

Item 20 Minor deepening and re-grading of the River Tolka over a minimum length of approximately 1,570m from the N3 side-road culvert to approximately 250m upstream of Parlickstown Road Bridge. It is proposed that the channel bed be cleaned for the remaining length to the road bridge culvert giving an overall length of approximately 1,800m for channel improvement (cross-section reference T-232 to T-249). In conjunction with the works, the obstruction caused by an existing sewer and associated manhole should be removed.


074515001Rp0032ADE.doc 77 Rev F01

These works will address the fact that the natural channel in this area has become significantly overgrown and deposition of materials over the years has increased bed levels. Reduction of flood levels is required to reduce the risk to RMF Ireland and Kepak, to improve efficiency of existing culverts and drainage from Westpoint Business Park and the N3. The regrading provides a 400mm reduction in levels at Kepak and a 500mm reduction downstream of Huntstown culverts tapering to no benefit at Parslickstown culverts. Refer to long section details provided in Drawing DG2206. Some cleaning work has already been undertaken by the Office of Public Works (OPW) but will require ongoing maintenance. Benching the cross section will provide some additional improvement, however it is likely that this will become significantly overgrown during summer months, and therefore would require more frequent maintenance than the deepening works recommended. Therefore, OPW works undertaken to date should be verified to ensure that design bed levels have been achieved particularly downstream of Huntstown Road Bridge as this area directly impacts on levels upstream of this culvert, and hence risk to Kepak. Consideration should be given during detailed design to low flow rehabilitation measures such as provision of ripple pools, additional meanders and off line ponds or wetland areas, to improve the natural habitat and water quality. Measures should be included to ensure that maintenance access to the riparian environment is available whilst ideally this environment can be improved during construction of the necessary flood alleviation measures. However these improvements are not required for flood risk reduction and could be considered for integration to future development proposals in this area.

Item 21 The Pinkeen tributary outfalls to the Tolka downstream of a culvert crossing of

Damastown North Road which is 2.2m in diameter. Backing up from this culvert creates a flood risk at the Rottapharm Factory Complex. It is therefore recommended that a new culvert comprising twin 2.5m diameter culverts be provided to ensure that there is no obstruction of the Pinkeen Stream outlet (Sheet 10, Final Report, cross-section reference T-701).

Item 22 A short section of embankment of maximum length 150m is required to

protect the N3 from possible inundation from the River Tolka (opposite cross-section reference T-233 – T-235). Fitting of flap valves or other measures to prevent backpounding of the N3 drainage system should be allowed for.

This is essentially required to provide the required 300mm freeboard, as cleaning of the existing drainage outfall and the proposed regrading above will reduce the local flood risk. It is possible that minor flooding will occur in this location with less than 1% probability PA. Therefore this would not justify a pumping station or raising of the N3, the embankment will ensure that this flooding is minimised to local runoff volumes for larger events.

Item 23 Approximately 360m of embankment is required to protect the Westpoint Business Park from inundation from the River Tolka (Sheet 6, Final Report, cross-section reference T-233 – T-235). These works are required given that some flooding has previously occurred within the Westpoint Business Park

�� The re-grading of the River Channel is this area will assist in giving free discharge

conditions to the Pinkeen(2) River tributary, given that the River Tolka bed levels are currently above invert levels of the Damastown Culverts on the tributary.

�� The works in this area will have to include consideration of the local piped drainage

system and necessary alterations to prevent secondary flooding from ponding.

�� The channel works will include relocation of services, reinstatement and rehabilitation of the area following completion of the excavation works.


074515001Rp0032ADE.doc 78 Rev F01

12.3.2 Littlepace The Littlepace housing estate to the south of the River Tolka was inundated with flood waters flowing from Clonee during the November, 2002 flood. Properties were flooded at Littlepace Crescent, Littlepace View and Littlepace Drive. At the adjacent roundabout, flood water depths reached up to 1.5m. It is considered that the flood alleviation works proposed for Clonee affecting the River Tolka and the Clonee Stream will provide significant protection against a repeat flood event at Littlepace. The local stormwater drainage system at Littlepace should be assessed for capacity and condition to ensure that it can operate efficiently in the event of a future storm event.

12.3.3 Mulhuddart – Blanchardstown (Map Sheet 5, Final Report)

This section of the river extends from upstream of Mulhuddart Bridge (B40) to downstream of the Mill Road Bridge (B30) adjacent to James Connolly Memorial Hospital. Within this section, the principal properties that flooded were Gleesons Pub upstream of Mulhuddart Bridge and an adjacent property and commercial properties located on Blakestown Road, inundation of the N3 motorway at Blanchardstown and some local property flooding south of the motorway upstream of the Blanchardstown Bypass Bridge. The modelling analysis demonstrates substantial differences in flood level across Blanchardstown North Bridge and Snugborough Road Bridge. Both of these bridges had the effect of holding back flood waters and increasing floodplain storage upstream. Apart from inundation of the N3 motorway, the consequences of this flooding were relatively minor for properties. Moreover, this storage is contributing significantly to attenuation of floodwaters downstream. Because the existing bridges are substantially surcharged, any increase in flood flows would have a significant impact in this area. Sensitivity analysis showed that an extreme climate change scenario involving 25% increase in flows would increase levels between 0.5m and 2.2m in the area. The Mulhuddart Bridge and pipe crossings in the vicinity of the bridge have some effect also in increasing flood levels upstream. While the headloss due to these obstructions is significantly less than that for the downstream bridges, it is important in terms of flood risk to the pub and adjacent property upstream of the bridge. Moreover, the works recommended for implementation upstream at Clonee are considered likely to increase flood levels at Mulhuddart Bridge by approximately 400mm in a worst case scenario. For that reason, it is considered necessary to implement works to reduce flood levels at Mulhuddart by enlarging the capacity of the bridge, upstream culvert and the downstream footbridge. Local flood defence works and measures to deal with the 335ha local catchment discharging upstream of the bridge will also be required. It can be seen from the interaction between flood levels here and works at Clonee that these works must be considered an integral part of the overall scheme to minimise flood risk in the area. It is noted that the Mulhuddart Bridge is a listed structure therefore the initial option of complete removal has been reconsidered, with regard to the environmental appraisal report. It proves extremely difficult to protect this low-lying area with a large number of constraints as outlined above. One option considered as an alternative to structural relief measures was the relocation of the Pub and associated ancillaries. This would allow the existing listed bridge structure to remain, and a reduction in flood levels could be achieved by increasing the width in which flood waters could overtop this bridge. The works required in the Mulhuddart area are therefore as follows:

Item 24 Upgrade Mulhuddart Bridge (B40) and provide a bypass culvert to the North to accommodate full channel cross section. Include associated services diversions and revetment to the river channel. Include underpinning/modification to footbridge currently under construction.

Item 25 Realign and marginally deepen the river channel to provide uniform longitudinal and

cross section profiles over a length of approximately 780m (cross section reference T-


074515001Rp0032ADE.doc 79 Rev F01

221 to T-212) and integrate widening to provide for bypass culvert. Include any modifications to existing structures or services affected by the works, including removal and re-routing of the upstream pipe bridge. Typical extent of invert deepening is approximately 0.3m with a focus primarily on uniform gradient and cross section

Item 26 Upgrade floods defence works to existing properties upstream of Mulhuddart Bridge,

which involves raising approximately 90m of existing wall and providing an additional 90m length of new flood wall/embankment (cross section ref. T-223 to T-221)

Item 27 Option 1: Re-route upper catchment drainage system (335ha) to downstream of

bridge. On the remaining local drainage system include a non-return valve to prevent back pounding and provide emergency pumping facilities for local drainage.

Item 27 Option 2: Seal off local area of upper catchment drainage system (335ha) and allow to

discharge under head. Re-route local catchment drainage and provide emergency pumping facilities.

At Blanchardstown, the works required to protect the N3 and local properties at risk include:

Item 28 Provide a continuous earthen embankment of approximately 390m in length parallel to the N3 (cross section Ref. T-195 to T-199). This embankment will extend upstream from Snugborough Road adjacent to the northern dual carriageway.

Item 29 Provide 2m high embankment or flood wall near property boundaries at

Herbert Road, Blanchardstown, over length of approximately 210m (cross section reference T-182a to T-183).

Item 30 Remove or renew the disused walk bridges in the Tolka Valley Park which are

in poor condition.

Item 31 Provide new drainage outlets from the N3 dual carriageway to discharge downstream of the Snugborough Road Bridge, including non-return flap valve arrangements or alternative drainage arrangements to protect the dual carriageway from ponding.

Item 32 Carry out detailed structural assessment of existing Corduff, Snugborough

and Blanchardstown Bridges having regard to significant water pressures and scour forces associated with repeat flooding similar to that which occurred in November 2002 to ensure that the bridges and embankments will satisfactorily cope with similar conditions in the future. Additionally both the Blanchardstown North Road and Snugborrough Road embankments should be investigated in terms of UK reservoir legislation. Overflow spill measures may be required to reduce the risk of failure in the event of a culvert blockage or larger flood.

12.3.4 Blanchardstown to Finglas Section

The remaining section of the River Tolka in the Fingal County area crosses the M50 and flows through relatively undeveloped lands (adjacent to Elm Green Golf Course and the National Food Centre) and continues through public parks crossing the Ashtown Bridge, Scribblestown Road Bridge, Cardiff’s Bridge, the Ratoath Road Bridge and Finglas Wood Bridge to the Finglas Road (N2). No works are considered essential over this section for flood risk alleviation. The area serves to provide natural flood attenuation under an extreme event, thereby moderating downstream flows to the Dublin City area. Of note both the Ashtown and Scribblestown Road Bridges are subject to significant afflux providing additional attenuation. The river road is subject to flooding, however it is understood that an alternative route is currently under construction as part of the Pelletstown Development.


074515001Rp0032ADE.doc 80 Rev F01

12.4 DUBLIN CITY COUNCIL AREA – PROPOSED WORKS

Due to the built up nature of the River Tolka floodplain within the Dublin City Council area, it is inevitable that significant flood alleviation and protection works are required over the section between Finglas Bridge (N2) and Luke Kelly Bridge. Downstream of Luke Kelly Bridge, river water levels are significantly influenced by tides, tide levels were not critical during the November 2002 flood. The model has been used to simulate a number of combinations of floods and significant high tides. Results from this analysis showed that current flood walls in the tidal reach of the river are generally adequate. The works in the Dublin City Council area are indicated on drawing Sheets 1 and 2, Final Report, which illustrated the study reach of the river between the Finglas Road Bridge (Cross Section Ref. T-109) to below the East Point Business Park Bridge (Cross Section Ref. T-001). The works in the Dublin City Council area are briefly outlined in the following sections.

12.4.1 Finglas Road to Glasnevin Bridge River Section

This section of the river from the Tolka Finglas Road Bridge (B17) to the Glasnevin Bridge (B12) extends from the crossing of the N2 Finglas Road dual carriageway and the river section adjacent to Glasnevin Cemetery and through the Botanical Gardens. No property flooding was experienced in this section, the principal consequences being uprooted trees and river bank erosion, and damage to planted areas in the Botanic Gardens. However some backing up of flood waters was experienced at the Finglas Bridge itself. Residents at Glasnevin Woods and in the Tolka Vale apartments were concerned at possible obstruction of the bridge due to trees and debris during the flood and additional concerns relating to an ESB pole in the floodplain. To reduce risk of flooding in this area, a number of measures are recommended as follows:

Item 33 A short section of wall (40m long) at Finglas factory weir adjacent to Tolka Vale apartments should be raised to give the required freeboard over the design flood levels.

Item 34 The entry and exit conditions from the Finglas Road Bridge should be improved by

streamlining the river channel, with removal of flood debris and local widening downstream of the bridge on the left bank. The extent of this work is local to the bridge for approximately 10-15m upstream and 20-30m downstream and should include general re-grading of the channel bed through this section and through the bridge. The object is to produce a more streamlined flow to and from the bridge to minimise local head losses.

In the Botanic Gardens area, general maintenance is required with removal of undermined trees and accumulated debris which has resulted from scouring of the river bed and damage to retaining walls due to the flood conditions, local protection measures or changes to the planted species may be required in the flood plain area to mitigate the affects of future floods on sensitive plant species. Tolka Vale apartment’s lower level car park is subject to flooding, and should be addressed by improvements to their pumping system. The recommended installation of a Gauging station at Finglas Factory Weir requires smoothing of the Weir crest. Therefore consideration was given to lowering the Weir to reduce levels in this immediate area. However taking account of archeological considerations and as only one property is at risk the option of increased wall height to provide additional protection directly above the Weir is recommended above.

12.4.2 Glasnevin Bridge to Dean Swift bridge

A short section of this reach between Glasnevin Bridge (B12) and Deans Swift Bridge (B11) experienced over topping of the river banks causing localised flooding at the school and properties


074515001Rp0032ADE.doc 81 Rev F01

immediately adjacent. Flood waters from this overspill continued down Botanic Avenue contributing to flooding in this area and to Drumcondra Road Lower. The works proposed in this area comprise:

Item 35 Approximately 100m of landscaped earthen embankment to a height of approximately 1.5m to the rear of houses 1-27 Botanic Avenue, and associated low wall. These works are under construction.

Item 36 Provision of approximately 50m of gravity flood wall approximately 1m high to the rear

of the school, also in progress.

Underpinning of Deans Swift Bridge was considered in the model to further reduce flood levels. Underpinning the bridge by approximately 600mm could be achieved. This would result in a 500mm reduction in levels directly upstream reducing to no improvement at the weir directly above Glasnevin Bridge in Botanic Gardens. The recommendation for underpinning will be subject to analysis of the local drainage system to determine if it offers any advantages in reducing drainage requirements.

Item 37 Inspection of Dean Swift Bridge indicates a degree of structural cracking and spalling of concrete. A detailed structural assessment of the bridge is recommended, removal of all trees and other vegetation growing on and adjacent to the structure and the implementation of appropriate remedial works.

�� Underpinning of the Dean Swift Bridge to reduce flood levels or adjustment of the local

drainage system to divert drainage outlets will be required to ensure that low lying areas will not be ponded in the future. The precise requirements will be determined in conjunction with the GDSDS Final Report.

12.4.3 Dean Swift Bridge to Drumcondra Road Bridge

Working downstream from Dean Swift Bridge, the housing developments (St Malachy’s Rd area) immediately on the right bank escaped flooding due to the existence of block work and masonry walls along the property boundaries with the river , additionally ground levels at 9.0m OD are generally above flood level. These walls comprise ad-hoc construction, the integrity of which for flood protection purposes is in doubt. Nevertheless, these walls appear to have withstood the flood levels during the November, 2002 flood. It is noted that partial floodwaters in this area arise from overflows upstream. Significant overspill from the River Tolka occurred to Botanic Avenue from the south side of Griffith Park near the intersection of Botanic Avenue and Mannix Road. Downstream of this location, a large area was inundated by flood waters from the Tolka. The Woodville Road footbridge was noted as a significant constriction to the main channel discharge by trapping debris in its railings. Flood waters north of the bridge also exited to Millmount Road on the left bank of the Tolka causing significant inundation of properties in this area. As these areas filled up, flood waters continued downstream crossing the Drumcondra Road Lower and Millmount intersection into Clonturk Park and continuing down Richmond Road. Flood waters overflowing south of Woodville Road footbridge also added to flooding in the lower Botanic Avenue area. Similarly, extensive flooding occurred in the reach immediately upstream of Drumcondra Road Bridge with the river over topping its banks within the old Tolka Park Cottages area. Accordingly, this area has been a priority for urgent remedial works. The section overlaps Sheets 1 and 2, Final Report, of the maps, with cross section references T-066 to T-044. The following works are proposed in this section:

Item 38 Replacement or upgrading of approximately 255m of riverside wall downstream from Dean Swift Bridge on the right bank. This could comprise rebuilding of existing walls or the construction of a riverside barrier inside the existing walls. Complications arise


074515001Rp0032ADE.doc 82 Rev F01

from the fact that existing walls in some cases support roofs to extensions, garden sheds, etc. These works also include the raising of the ‘Old folks apartments’ wall to 9.1m OD to prevent floodwaters entering the grounds.

Item 39 It is proposed that the left bank of the river be widened by approximately 5m over a

length of approximately 220m with replacement of the low riverside walls and associated adjustment of drainage services, paths and general reinstatement of the area. This work will assist in ensuring that water levels are not increased by the general reduction in storage upstream and would facilitate construction of a new flood wall on the right bank within the river channel.

Item 40 Approximately 250m of embankment within Griffith Park from Drumcondra Footbridge

along the southern boundary of the park tied into the footbridge (Cross Section Ref. T-060 to T-053). In addition, a short embankment of approximately 40m is being provided on the left bank upstream of the footbridge.

Item 41 Replacement footbridge for the Woodville footbridge in Griffith Park together with

realignment of paths and other services. This involves a pedestrian bridge of approximately 20m span some 2.5m wide. The new footbridge is integrated with the embankments previously described. These works also involve the replacement of the railings on the upstream footbridge for aesthetic purposes, however this will also offset local increases in levels as a result of the embankment constriction.

Item 42 Immediately downstream of Woodville Footbridge, new walls are required to

approximately 1.0m over garden level. A 25m length of wall is required to the right bank to raise existing wall levels behind No. 5 Woodville Road. A 60m overall length of wall is required in private properties to the left bank, behind Millmount Villas. These works are also under construction in the current preliminary phase of work.

Item 43 Upstream of Drumcondra Bridge, approximately 125m of embankment is required

together with approximately 65m of raised boundary wall with new railing to seal off the section of the River Tolka from Botanic Avenue.

Item 44 A new flood wall approximately 60m long and 2m high is required upstream from

Drumcondra Road Bridge to the back of the properties on Millmount Terrace to prevent floodwaters entering in this area.

Item 45 General provision is made for modification to the pipe drainage system to the lower

Botanic Avenue area to prevent ponding of this area behind the flood defence works. This will require modifications to the drainage system locally.

Consideration was given to underpinning the Drumcondra Road Bridge to further reduce flood levels in this area; however it was noted that flood levels are only reduced in the local area. Flood levels upstream are dictated by the channel width and widening would not be possible due to the proximity of development, additionally deepening of the channel would require underpinning or replacement of river walls, therefore these options were not considered further.

12.4.4 Drumcondra Road Bridge to Luke Kelly Bridge

Extensive flooding was experienced in the Richmond Road area and in the Tolka Road / Clonliffe Road / Poplar Row area as well as in Marino. This flooding resulted from flood waters emanating from upstream of Drumcondra Road as already described together with overflows directly to Richmond Road and in the vicinity of the Distillery Road Bridge. The following works are required over this river section, considering the river section working downstream from Drumcondra Bridge:

Item 46 Raising/reconstruction of an embankment behind Nos. 4 – 52 Richmond Road on the left bank of the river for a length of approximately 120m. This is a relatively low embankment to approximately 0.5m over existing levels to improve freeboard and fill localised depressions.


074515001Rp0032ADE.doc 83 Rev F01

Due to the large change in direction of flow at Tolka Park, revetment reinforcement is required to prevent erosion of the river outer bank and to help direct flows around this “gooseneck” river alignment. A new wall and embankment at the bend to Tolka Park Stand (i.e. behind 60-68 Richmond Road) is currently being constructed as part of the interim flood alleviation works.

Item 47 New wall/embankment to the left bank behind No. 6 – 68 Richmond Road linking to

the wall of Tolka Park stand. This requires a length of approximately 55m to a height of 2.5m.

The existing retaining walls adjacent to Tolka Park Football Ground consist of a low flood wall which appears to form a retaining wall for the secondary wall as part of the grandstand. The retaining wall is low grade and has suffered significant flood damage. It is recommended that this wall be upgraded or replaced with a structurally sound and continuous flood defence wall.

Item 48 New/upgraded river wall approximately 160m in length to improve protection to Tolka Park.

Additionally an embankment should also be constructed at the rear of the properties located at 112-128 Richmond Road. This would ensure property protection is increased and that flood waters cannot extend back towards Richmond Road and Grace Park Road. Richmond Road may require additional pumping systems to deal with localised drainage, particularly for the basements of these properties which are subject to regular local flooding. However these works will be less critical in terms of there protection of Richmond Road and Grace Park Road if Distillery Weir is lowered as detailed later.

Item 49 New embankment south of Tolka Park on the left bank for approximately 50m behind Nos. 110-126 Richmond Road.

Item 50 Provision is also made for a section of new wall behind Edgewood, Richmond Road

which would be less critical if the lowering of the Distillery Weir is proceeded with as recommended. This involves approximately 35m length of 1m high wall.

Item 51 The provision of a new wall behind the Industrial Estate (Edgewood to Distillery Weir)

involving approximately 270m of up to 2m high river wall to the left bank, this is required to replace existing low grade and non existent boundary walls, and to integrate with the widening works described below.

Item 52 It is proposed to lower the effective crest level of Distillery Weir by approximately 1m

from 2.87m OD to 1.9m OD and to refurbish the weir at this revised level. This work will include the widening of the weir to approximately 20m and is necessary to lower flood water levels upstream of the weir and to reduce the difficulties of dealing with local drainage in the area.

This work will result in a lowering of flood water levels of 900mm at the weir crest. The impacts of these works will extend back to Drumcondra Road Bridge where upstream flood levels will be reduced by approximately 50mm. A 300mm reduction in flood levels will be achieved opposite Tolka Park. Additional lowering is not considered as the Tidal divide would be impacted at levels below 1.9m. (Noting that MHWS tide level is 1.7 at the outlet.) This level is higher than the level of the existing downstream slot (1.6m OD), and it is possible that this represents the original weir level which has been heavily modified. Initial considerations were given to providing an adjustable weir to maintain the existing low flow regime upstream of Distillery Weir. However it is noted that tides would rarely extend upstream of the weir; and in these rare cases the extent would be limited to a maximum of 400 metres upstream. Environmental assessment of the proposed works indicates that bank side vegetation would be improved due to the lowering. Additionally, Technical Report Number 5; Environmental Assessment; Appendix A, pg 10 recommends that the weir be removed due to its current adverse influence on fish passage; however this is offset by archaeological implications. It is therefore recommended that adequate provision for fish passage be provided as an integral part of the works outlined above. Depending on Archaeological and drainage requirements as yet undefined this


074515001Rp0032ADE.doc 84 Rev F01

appears to be a reasonable recommendation; however some refinement may be required following further investigation in the detailed design phase to accommodate archaeological requirements.

Item 53 On the right bank, the construction of approximately 130m of boundary wall to protect

the new apartments (Clonliffe Square) has been proceeded with, to an effective height of 1.5m. These works would result in increased levels at distillery weir due to the removal of the secondary flow path. However the lowering of the weir will offset these increases.

Item 54 A further 15m of 1m high wall is similarly required to complete the protection at

Clonliffe Square (Cross Section Ref. T-036). This has also been proceeded with. It is noted that proposals to redevelop the Distillery Industrial Estate on the North bank are being considered. Lowering of water levels is required in this area to offset increases resulting from improvements upstream and to assist discharge of local stormwater drainage. Dredging was considered, however its impact is minimal under Tidal scenarios (1m dredging results in 100mm reduction in levels, and would require underpinning or reconstruction of both left and right bank riverside walls). It is therefore recommended that this narrow section of the river is widened on the north side by approximately 5 meters to 20m including the construction of a structurally sound and continuous flood wall extending downstream to the river bend (opposite112-114 Tolka Road). These works will provide a 400mm reduction in design flood levels in addition to the 100mm reduction for channel cleaning works, and assist in the protection of Richmond Road/Marino. Other options have been considered for a lesser widening in the lower half of this section to approximately 17m due to constraints identified with existing buildings. These result in a lesser reduction in levels of 250mm downstream of Distillery Road Bridge.

Item 55 Some widening of the left bank downstream of the weir has been allowed for to

improve the safety factor in this area (approximately 5m) with removal and replacement of existing intermittent and low grade walls along this reach.

The constraints provided by Distillery Weir, Distillery Road Bridge and additionally the improvements to the wall at Distillery apartments exacerbate flood levels through this narrow area. Of note the Distillery Road bridge is subject to trapping debris resulting in a localised increase in flood risk.

Item 56 It is recommended that the Distillery Road Industrial Estate Bridge be replaced at a higher level (minimum soffit level of 4.0m OD). This will require some horizontal realignment and re-grading to facilitate access to the bridge and retain access to existing properties (minor widening from 18.5m to 20m is included). A 200mm reduction in levels upstream of the bridge is provided due to removal of bridge blinding which is causing increased afflux.

Item 57 Downstream of Distillery Road Bridge, behind No. 20 Distillery Road, it is

recommended that approximately 20m of 1.5m high flood wall be provided.

Item 58 Trees and debris have been recently removed during reinstatement works to an

existing wall behind Nos. 112 – 114 Tolka Road. This wall was undermined due to scour resulting from the localised obstructions. Maintenance cleaning works have also included removal of significant amounts of debris and construction waste from under Luke Kelly Bridge resulting in improvements to the flow regime and levels in this area.

Item 59 General reinstatement of the River Channel and Riverside areas would be required as

part of these overall works. Item 60 Detailed consideration to the pipe drainage systems in the lower Clonliffe Road,

Poplar Row, Marino and Ballybough Road areas is required as part of the GDSDS drainage study to ensure that these low lying areas will not be ponded and can be effectively drained during a flood event. These works will be integrated to the River


074515001Rp0032ADE.doc 85 Rev F01

Tolka Flood Alleviation Scheme and provision is made in the project budget for implementing the additional drainage works.

Item 61 In the tidal area, drainage flap valves are recommended on all stormwater outlets,

with particular reference to East Wall Road, where properties are below the critical tidal flood range and require effective protection by the existing flood walls.

Tidal assessment indicates that with the exception of the items outlined above, No additional protection works are recommended downstream of Luke Kelly Bridge in the tidal reach of the river. In this section, no flooding took place directly from the river in the November, 2002 flood, although some low-lying areas were ponded by secondary flooding via the drainage system. This will be addressed as part of the GDSDS study. Additionally no flooding took place from the Tolka River in the February 2002 Tidal Surge event which is the largest sea level ever recorded in the Dublin area. Apart from these works, a strategy for general maintenance and operational management of the River Tolka channel is required to ensure that its hydraulic efficiency is maintained in future years. This will require appropriate structures and budgets to be agreed between the relevant Local Authority. Recent alterations to the river including the East Point Business Park Bridge, Upgrade of the CIE Bridge and proposed Dublin Port Tunnel works have been assessed and provide sufficient capacity to convey flows. Summary results of these assessments are provided in figures 9.1 and 9.2. This analysis indicates that extreme flood levels in this section of the river are dominated by tidal influence for levels above 2.5m OD. Therefore current channel widths are adequate subject to ongoing provision of appropriate maintenance cleaning works to ensure the existing flow regime is maintained. Tidal analysis indicates that when including for sea level rise for climate change, levels of protection are however marginalised in certain areas. Direct improvements to the freeboard levels could be provided by alterations to the Fairview Park Footbridge directly above the CIE Bridge and minor wall raising upstream of Luke Kelly Bridge and directly downstream of Distillery Road Bridge. However as current freeboard requirements are available ongoing monitoring is recommended to determine if climate change impacts are realised before these works are recommended as part of the current flood relief scheme. It is therefore recommended in the interim, that redevelopment of riparian sites and repairs to walls as outlined in the Preliminary Structural Survey Report provide a minimum top of wall level of 3.7m OD for all areas downstream of Distillery Road Bridge.

12.5 GENERAL RECOMMENDATIONS

General maintenance to walls and bridges are outlined in the Preliminary Structural Survey Report together with recommended works. Additionally flood damage repairs and maintenance recommendations are included in the Preliminary Geotechnical Report and include details of these works. Determination of the responsibility for the various works identified will be required including repairs to bridges and walls and the responsibility of the riparian landowners. The existing flood risk maps generally identify areas subject to remaining risk from failure of defences, or overtopping of walls and embankments in the event of a flood in excess of the established design. Consideration should be given to provision of failure mechanisms to section off affected areas of development as part of the Major Emergency Plan. Failure analysis should also include allowance for flood return paths should upstream defences be breached. This could be provided by manual gates or by the installation of a floating defence in walled or embanked sections. Additionally mitigation measures could be integrated into the detailed design of drainage upgrades to deal with such scenarios where appropriate. The implementation of the works recommended in this chapter will involve significant planning and design. As part of that work, site investigations will be required to establish soil conditions. Services investigations will be necessary to locate existing services which may need to be altered to facilitate the works. In addition, as already outlined, significant amendments to the local drainage works will be required to integrate these works satisfactorily to the proposed scheme.


074515001Rp0032ADE.doc 86 Rev F01

13 MODEL HANDOVER In order to carry out a model run in InfoWorks RS both the model network and appropriate event is required. The event contains the boundary conditions such as the rainfall, design event and tidal boundaries. The calibrated model consists of the existing case scenario before any interim solutions were undertaken. This model run is presented with the following events:

�� 14th & 15th November 2002 �� 6th & 7th November 2000 �� 200 Year Design Storm �� 100 Year Design Storm �� 50 Year Design Storm �� 25 Year Design Storm �� 10 Year Design Storm

The proposed flood alleviation scheme, as outlined in Chapter 12 is included in model handover and carried out with the following events:

�� 14th & 15th November 2002 �� 200 Year Design Storm �� 100 Year Design Storm �� 50 Year Design Storm �� 25 Year Design Storm �� 10 Year Design Storm

The proposed flood alleviation scheme with development to 2031 is carried out with the following events:

�� 200 Year Design Storm �� 100 Year Design Storm �� 100 Year Design Storm with Climate Change of 22% �� 100 Year Design Storm with Climate Change of 10% �� 50 Year Design Storm �� 25 Year Design Storm �� 10 Year Design Storm

The following tidal model runs were carried out on the lower half of the Tolka model network and have been included in the model handover:

�� Q2yr + H2.79m �� November 2002 �� Qbase + H2.95m �� Q50yr + H2.20m

(Note: Initial Conditions have been included also – these were taken from a successful model run and used when attempts at completing successive runs were ineffective.)


074515001Rp0032ADE.doc 87 Rev F01

14 CONCLUSION To summarise, this report has outlined the requirements for the River Tolka flood alleviation scheme, based on protecting existing developed areas in the study region of the river for at least the one in one hundred year flood event. The scheme has inbuilt capacity to deal with projected urban development to 2031 and within the safety margin (freeboard) of the scheme it can accommodate a significant degree of rainfall increase associated with possible climate change impacts (up to 10%). The implementation of the scheme involves completion of the existing preliminary works now in hand, to be followed by completion of the remainder of the river engineering works designated as Phase 1 and 2. In parallel, any requirements to modify the existing pipe drainage system in low lying areas will be identified in conjunction with the GDSDS study and these works will form an integral part of the overall scheme, being necessary to protect low-lying areas from secondary flooding and for which provisional costs have been included in the estimates. The recommendations include the provision of an integrated management and monitoring system for the River Tolka involving a co-ordinated approach to the management of the river and of future development, implementation of routine monitoring and maintenance and including automated monitoring of water flows, water levels and rainfall depths as an extension of the existing Dublin City Council telemetry system. It is recommended that the River Tolka flood alleviation scheme as outlined in this report be implemented in full. The River Tolka flooding model will be made available to Dublin City Council as lead authority for the project and should be maintained and updated as a key management tool in the overall management of the catchment.

technical report volume 2: river modelling report … · technical report volume 2: river modelling...

Documents