NEW CANALIZATION OF THE NEDERRIJN AND LEK MAIN REPORT Design of a weir equipped with fibre reinforced polymer gates which is designed using a structured design methodology based on Systems Engineering
25 January 2013 : Henry Tuin
New canalization of the Nederrijn and Lek Main report
Colophon
Title: New canalization of the Nederrijn and Lek – Design of a weir with fibre reinforced polymer gates which is made using a structured design methodology based on Systems Engineering
Reference: Tuin H. G., 2013. New canalization of the Nederrijn and Lek – Design of a weir with fibre reinforced polymer gates which is designed using a structured design methodology based on Systems Engineering (Master Thesis), Delft: Technical University of Delft.
Key words: Hydraulic structures, weir design, dam regime design, Systems Engineering, canalization of rivers, fibre reinforced polymer hydraulic gates, Nederrijn, Lek, corridor approach, river engineering.
Author: Name: ing. H.G. Tuin Study number: 1354493 Address: Meulmansweg 25-C 3441 AT Woerden Mobile phone number: +31 (0) 641 177 158 E-mail address: [email protected] Study: Civil Engineering; Technical University of Delft Graduation field: Hydraulic Structures
Study: Technical University of Delft Faculty of Civil Engineering and Geosciences Section of Hydraulic Engineering Specialisation Hydraulic Structures CIE 5060-09 Master Thesis
Graduation committee: Prof. drs. ir. J.K. Vrijling TU Delft, Hydraulic Engineering, chairman Dr. ir. H.G. Voortman ARCADIS, Principal Consultant Water Division, daily supervisor Ir. A. van der Toorn TU Delft, Hydraulic Engineering, daily supervisor Dr. M.H. Kolstein TU Delft, Structural Engineering, supervisor for fibre reinforced polymers
: ARCADIS & TUDelft i
New canalization of the Nederrijn and Lek Main report
Preface & acknowledgements
This thesis is the result of the master Hydraulic Engineering specialization Hydraulic Structures of the faculty of Civil Engineering and Geosciences of the Delft University of Technology. The graduation topic is proposed by ARCADIS and is inspired upon the project RINK-SSC (Risico Inventarisatie Natte Kunstwerken- Sluis en Stuw Complexen). The research is performed under guidance of the Delft University of Technology and in cooperation with ARCADIS. The research is performed from April 2012 till January 2013.
The Nederrijn and Lek are canalised by three weirs which are placed near the villages of Driel, Amerongen, and Hagestein to which they are named. These weirs regulate the distribution of water over the rivers Nederrijn and the IJssel and regulate the water levels in the Nederrijn and the upstream section of the Lek for navigational purposes. A reliability and availability assessment performed in context of the RINK project stated that the weirs are in bad condition; the reliability and availability do not correspond to the current requirements. Therefore, the weirs need to be renovated or replaced by new weirs.
The subject of the graduation research is the development and design of an alternative for the present canalization of the Nederrijn and Lek using Systems Engineering. A design methodology based on Systems Engineering is developed and tested for the Nederrijn and Lek recanalization case study and a more detailed design of a fibre reinforced polymer weir gate is made which is able to regulate the water levels and the discharges. The resulting design and the tested design methodology are described in this main report. More detailed information and calculations are included in the appendices.
First of all, I would like to thank my graduation committee for their guidance. Special thanks to Hessel Voortman and Ad van der Toorn for their daily and fruitful guidance, their critics, and for challenging me by raising the bar a few notches. I would also like to specially thank Henk Kolstein for introducing me into the world of fibre reinforced polymers and providing me with literature and contacts for interviews. Furthermore, I would like to thank my fellow students of the Delft University of Technology and colleagues of ARCADIS for the fruitful discussions and joyful moments. At last, I would like to thank my parents, brothers, fiancée, and other family for supporting me during my entire study.
Henry Tuin
Amersfoort, January 2013
ii ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
Abstract
The Nederrijn and Lek are presently dammed by three weirs, build in 1960 till 1970, which are located near the village of Driel, Amerongen, and Hagestein after which they are named. Weir Driel regulates the distribution of water at the IJsselkop for dammed conditions (discharges lower than 2350 m3/s at Lobith). A minimum discharge of 285 m3/s has to be diverted into the IJssel by weir Driel for the fresh water supply of the northern part of the Netherlands (the section of the Netherlands which is located above the fictive line Amsterdam-Nijmegen) and for generating sufficient draught for commercial shipping. The weirs Amerongen and Hagestein regulate the upstream water levels for commercial shipping; minimum levels of +6.0 m NAP respectively +3.0 m NAP are maintained by these weirs. The water levels of the IJssel decrease to 1 metre or lower when weir Driel fails in dry summers which results in an obstructed waterway for commercial shipping and an insufficient fresh water supply to the northern part of the Netherlands.
Problem statement ARCADIS made an assessment for the reliability and availability of the weirs in the framework of the project RINK-SSC (Risico Inventarisatie Natte Kunstwerken Sluis Stuw Complexen). The assessment indicated that the weirs do not meet the reliability and availability norms for the present situation and that structural parts of the weirs have exceeded their technical life span. Therefore, the weirs need to be renovated or replaced within ten years from now. Furthermore, a secondary problem statement is present which concerns the application of Systems Engineering for large scale design projects. Presently, a major ‘transition’ of the substantiation of design choices made in the political domain and design choices which are made by engineering firms, which are based on the Systems Engineering methodology, is present. Design choices made before the ‘transition’ are not well traceable and results in ambiguities during the design process.
Approach This graduation research is focused on the replacement of the existing weirs. The weirs could be replaced at the same location, but a broader scope is chosen within the graduation research. The aim of the graduation research is providing an alternative design for the present canalization taking into account the changed environment and use of the waterway. This goal is reached by designing an alternative for canalization using a design methodology based on Systems Engineering (SE). Therefore, a methodology for the application of SE for large scale hydraulic projects is developed and tested for the design of the recanalization. The developed methodology is also applied at large scale design choices for interventions in the present ‘wet infrastructure’ which are, in fact, normally made in the political domain. A toolset is composed from the toolbox of SE methods for the development of the alternative design. The design is elaborated in five sequential design levels. The first design level starts at delta level, the second continues at Nederrijn-Lek level, the third proceeds at the weir location level, the fourth considers the weir build-up, and the structural design of the gates is made for design level five. The gates are designed using fibre reinforced polymers (FRP) which is a self-imposed secondary goal for the graduation research. A small requirement set is developed for each design level in order to obtain a well-structured and clear design method which is applicable for large scale hydraulic design projects.
Results of the recanalization design The result of the research is an alternative design of the present canalization. The Nederrijn and Lek are chosen for recanalization due to the lowest impact for the implementation of the recanalization with
: ARCADIS & TUDelft iii
New canalization of the Nederrijn and Lek Main report
respect to the other rivers. It is possible to canalize the Nederrijn and Lek using two weirs when the reach located in between the Amsterdam Rijnkanaal and the Lekkanaal is downgraded to a recreational river. The upstream weir is located near the village of Driel and has to maintain the present dam regime to reduce the implementation works and time. The downstream weir is located at a floodplain near the village of Culemborg and has to regulate a minimum upstream level of +5.0 m NAP. The weir is subdivided in three gaps of 41 metres wide each. Three FRP submerged segment gates have to control the discharge and the upstream water levels of which an impression is given in Figure 1. The gates are composed of sandwich panels and four shear webs which are located in between the panels.
Figure 1 Impression of the designed Culemborg weir
Results of the developed and tested SE methodology The application of separate design levels with a limited set of requirements each turned out to be applicable for large scale hydraulic design projects for the first four design levels. So, the SE methodology could also be applied for decision making on a large scale which are normally made in the political domain. A well-structured and clear design methodology is obtained, which performed well for these levels. The method had to be adjusted for the structural design level (level five) because the set of requirements exponentially expands which results in a non-clear design method. Furthermore, using a functional analysis for the structural design results in a non-meaningful definition of the functions of the elements.
Conclusions It is possible to re-canalize the Nederrijn and Lek using two weirs. The upstream weir is located near the village of Driel and the downstream weir near the village of Culemborg. Weir Culemborg is equipped with three fibre reinforced weirs which are able to retain a head of 5.5 metres which is present for the minimum flushing discharge and low water conditions (OLR). The developed design methodology is applicable for the conceptual design which also covers the design choices which are normally made in the political domain with some additions. So, it is possible to position the mentioned ‘transition’ before the policy making process. The methodology has to be adjusted for the structural design (level five) due to the exponential increase of requirements. A limited set of requirements per design level is made on which a design choice is made. The limited sets of requirements per design level results in a clear design process.
Recommendations The application of the material and conversion factors for the determination of the design strain according to the CUR96 are not consistent with the practice. Therefore it is recommended to perform a more detailed research on the design limits. It is recommended to apply the requirements developed in the conceptual design (design level one till four) for the structural design (level five) because applying the SE method for a structural design (design level five) results to an exponential increase of requirements.
iv ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
Contents
Colophon ...... i
Preface & acknowledgements ...... ii
Abstract ...... iii
List of figures and tables ...... iii
Glossary of terms and abbreviations ...... v
Project map ...... vi
1 Introduction ...... 1 1.1 Introduction to the Nederrijn and Lek canalization ...... 1 1.2 Introduction to Systems Engineering ...... 2 1.3 Problem definition ...... 2 1.3.1 Problem definition for the Nederrijn and Lek ...... 2 1.3.2 Problem definition for the application of Systems Engineering ...... 3 1.4 Goal ...... 3 1.5 Scope ...... 4 1.6 Reading guide ...... 4
2 Present situation ...... 6 2.1 History of the Nederrijn and Lek ...... 6 2.1.1 Normalisation of the Rhine branches ...... 6 2.1.2 Canalization of the Rhine branches ...... 7 2.1.3 Implemented canalization ...... 7 2.1.4 Dam regime ...... 8 2.2 Design of the present Nederrijn weirs ...... 8
3 Design methodology ...... 9 3.1 Design approach used for recanalization ...... 9 3.2 Application of the V-model ...... 10 3.3 Design steps of the design loop ...... 11
4 Design level 1 - Rhine delta ...... 13 4.1 Project area ...... 13 4.1.1 Transport of water ...... 13 4.1.2 Navigation ...... 14 4.2 Requirements analysis ...... 14 4.2.1 System boundaries ...... 14 4.2.2 Main requirements ...... 15 4.3 Functional analysis ...... 15 4.3.1 Functions ...... 15 4.3.2 Objects ...... 16 4.4 Design synthesis ...... 17
: ARCADIS & TUDelft v
New canalization of the Nederrijn and Lek Main report
4.5 Conclusion ...... 18
5 Design level 2 - Nederrijn-Lek configuration ...... 19 5.1 Project area ...... 19 5.2 Requirements analysis ...... 22 5.2.1 System boundaries ...... 22 5.2.2 Main requirements ...... 22 5.3 Functional analysis ...... 23 5.3.1 Functions ...... 23 5.3.2 Objects ...... 24 5.4 Design synthesis; variant study ...... 24 5.4.1 System Aspects ...... 25 5.4.2 Ranking ...... 26 5.4.3 Remaining variants ...... 26 5.4.4 Evaluation of variants (MCA) ...... 29 5.5 Conclusion ...... 29 5.5.1 Design choice ...... 30 5.6 Hydraulic model of the Nederrijn ...... 30
6 Design level 3 - weir location ...... 33 6.1 Project area ...... 33 6.2 Requirements analysis ...... 34 6.2.1 System boundaries ...... 34 6.2.2 Main requirements ...... 35 6.3 Functional analysis ...... 35 6.3.1 Functions ...... 35 6.3.2 Objects ...... 36 6.4 Design synthesis ...... 37 6.4.1 Approach channels ...... 37 6.4.2 Weir ...... 37 6.4.3 Filtering of potential locations ...... 37 6.4.4 Phasing of the construction ...... 38 6.5 Conclusion ...... 39
7 Design level 4 - weir design ...... 40 7.1 Project area ...... 40 7.2 Requirements analysis ...... 40 7.2.1 System boundaries ...... 40 7.2.2 Main requirements ...... 41 7.3 Functional analysis ...... 41 7.4 Design synthesis ...... 42 7.4.1 Design aspects ...... 42 7.4.2 Evaluation of variants ...... 44 7.5 Conclusion ...... 44
8 Design level 5 - gate design ...... 45 8.1 Loads ...... 45 8.2 Fibre reinforced polymers (FRP) gates/beams ...... 46 8.2.1 FRP stiffness properties...... 46
vi ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
8.2.2 Deflection of FRP beams ...... 46 8.2.3 Estimation of gate weight and gate costs ...... 47 8.2.3.1 Design stress ...... 47 8.2.3.2 Costs of a steel and FRP gate ...... 47 8.3 Gate design ...... 48 8.3.1 General gate design ...... 48 8.3.2 Dimensioning of the FRP gate ...... 49 8.3.2.1 Outer plates design ...... 50 8.3.2.2 Shear web design ...... 51 8.3.2.3 Shear elements of the InfraCore® panel ...... 51 8.3.2.4 Torsion ...... 52 8.3.2.5 Local deflections ...... 52 8.3.2.6 Combined stress capacity of the panels ...... 52 8.3.3 Design checks ...... 52 8.3.4 Connections for large span FRP gates ...... 52 8.3.5 Vibrations of a steel and FRP gate ...... 53 8.4 Technical drawings of weir Culemborg ...... 53
9 Evaluation of the design methodology ...... 54 9.1 Application of criteria ...... 54 9.2 Verification ...... 54 9.3 Preview to lower levels ...... 55 9.4 Definition of design levels and system boundaries ...... 56 9.5 Development and application of requirements ...... 56
10 Conclusions and recommendations ...... 57 10.1 Conclusions ...... 57 10.1.1 Design of the recanalization of the Nederrijn and Lek ...... 57 10.1.2 Developed design methodology ...... 58 10.2 Recommendations ...... 58 10.2.1 Design of the recanalization of the Nederrijn and Lek ...... 58 10.2.2 Design methodology ...... 59
References ...... 61
Appendices (provided on CD): A. Literature study K. Verification of the horizontal water B. Information and methods level assumption C. Area analyses L. Cost estimation of weirs D. Stakeholder analysis M. Weir location E. Requirements analyses N. Preliminary weir design F. Lists of requirements O. Hydraulic models of the weir G. Functional analyses P. Load definitions H. Design criteria Q. Waves and water pressures I. Redesign of waterways R. Gate weight and gate costs J. New configuration of the Nederrijn and S. Weir foundation Lek T. Global weir gate design U. Gate design
: ARCADIS & TUDelft vii
New canalization of the Nederrijn and Lek Main report
List of figures and tables
Figure 1 Impression of the designed Culemborg weir ...... iv Figure 2 Project maps of the five design levels. The red lines represent the systems boundaries (based on: Google Maps and ANWB topografische atlas)...... vi Figure 1-1 The schematised Nederrijn and Lek system (ARCADIS, 2010) ...... 2 Figure 2-1 Implementation works for the present canalization of the Nederrijn (Google Maps) ...... 7 Figure 2-2 Front and side view of weir Hagestein (Rijkswaterstaat, 1955) ...... 8 Figure 3-1 Applied V-model for the recanalization (based on: (Rijkswaterstaat & ProRail, 2009) & (Department of Defense, 2011)) ...... 10 Figure 3-2 ‘Normal’ design method (left); design method used in the graduation research (right) ...... 11 Figure 3-3 15 steps of requirements analysis (Department of Defense, 2011) ...... 12 Figure 4-1 Distribution of discharge for dry periods (left) (Noordhoff Atlasproducties, 2011) and for maximum discharge (right) (Brinke, 2004) ...... 13 Figure 4-2 Inland waterway cargo transport (Noordhoff Atlasproducties, 2011) ...... 14 Figure 4-3 Functional tree for the Rhine delta ...... 16 Figure 4-4 Functional flow block diagram (FFBD) for the Rhine delta ...... 16 Figure 4-5 Object tree of 'The Rhine branches' system ...... 16 Figure 5-1 Project area characteristics (MW-N=Merwede kanaal northern section, MW-S=Merwedekanaal Southern section, LK=Lekkanaal, AR-N=Amsterdam-Rijnkanaal Northern section, AR-S=Amsterdam Rijnkanaal Southern section) ...... 20 Figure 5-2 Harbour and river characteristics of the Nederrijn and Lek for the present fully dammed situation ...... 21 Figure 5-3 Functional tree for the Nederrijn-Lek configuration ...... 24 Figure 5-4 Object tree for 'The water supplier' system ...... 24 Figure 5-5 Basis drawing used for the brainstorm sessions for the Nederrijn-Lek configuration ...... 25 Figure 5-6 Remaining variants after the selection ...... 26 Figure 5-7 Variant “2w;Driel&Culemborg;com.ship&recr.” Harbour and river characteristics ...... 27 Figure 5-8 Variant “2w;Driel&Culemborg;com.ship&recr.” Pumping stations and inlets ...... 28 Figure 5-9 Variant “2w;Driel&Culemborg;com.ship&recr.” Lock characteristics ...... 28 Figure 5-10 Cost/performance ratio including weighting for the Nederrijn-Lek configuration ...... 29 Figure 5-11 Water levels upstream from Culemborg for the minimum dam regime ...... 32 Figure 5-12 Minimum and maximum dam regime for weir Culemborg ...... 32 Figure 6-1 Project area characteristics for the selection of the weir location (Google Maps) ...... 34 Figure 6-2 System boundaries (indicated by the red lines) for the selection of the weir location (Google Maps) ...... 35 Figure 6-3 Functional tree for the weir location ...... 36 Figure 6-4 Object tree of ‘The water retainer’ system ...... 36 Figure 6-5 Potential weir locations in between the Amsterdam Rijnkanaal connection and weir Hagestein (Google Maps) ...... 37 Figure 6-6 Layout of the new weir complex at the floodplains near Culemborg (Noordhoff Atlasproducties, 2011) ...... 39 Figure 6-7 Impression of the floodplains of the chosen location of the weir (photo of the author) ...... 39 Figure 7-1 System boundaries for the weir design ...... 41 Figure 7-2 Object tree for the weir structure ...... 42
: ARCADIS & TUDelft iii
New canalization of the Nederrijn and Lek Main report
Figure 7-3 Selected gates (From left to right: vertical lifting gates, submerged segment gate, flap gate, visor gate) ...... 43 Figure 7-4 Cross section of the submerged segment gate which is further elaborated in design level 5 ...... 44 Figure 8-1 Load definitions ...... 45 Figure 8-2 General gate cross section (left) and a cross section with stiffening ribs (right) ...... 49 Figure 8-3 Mechanical schemes of the gate ...... 49 Figure 8-4 Sandwich panels ...... 50 Figure 8-5 Gate cross section ...... 51 Figure 8-6 Shear elements of the InfraCore panel ...... 51 Figure 8-7 Failure modes of sandwich beams. (a) Face yielding/fracture, (b) core shear failure, (c) and d) face wrinkling, (e) general buckling, (f) shear crimping, (g) face dimpling and (h) local indentation (Zenkert, 1995) ...... 52 Figure 8-8 Bolted connection of a shear web and the outer plates ...... 53 Figure 9-1 Relation between requirements, design and criteria ...... 54 Figure 9-2 Preview to lower design levels ...... 55 Figure 9-3 Feedback/redesign loop which is present for a ‘dead end’ in the design process of the Eastern Scheldt barrier (de Gijt & van der Toorn, 2011)...... 56
Table 4-1 Filtered table of options for the selection of the Rhine branches ...... 17 Table 6-1 Assessment for the filtering of the locations ...... 38 Table 7-1 Filtered morphological map for weir Culemborg ...... 43 Table 7-2 Weir variants ...... 43 Table 8-1 Laminate stiffness properties (stichting CUR, 2003) & (Snijder, 2012) ...... 46 Table 8-2 Design stresses of the applied material...... 47 Table 8-3 Gate weight and costs ...... 48 Table 8-4 Deflections in metres for a ranging gate dimension ...... 50
iv ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
Glossary of terms and abbreviations
Abbreviations 2w Two weir variants 3w Three weir variants A Cross-sectional area [m2] AR Aspect requirement Com.ship Commercial shipping variant for design level 2 Com.ship&recr Combination variant for commercial shipping and recreation for design level 2 CUR Civieltechnisch centrum Uitvoering Research en Regelgeving CUR96 Recommendation 96 issued by the CUR relating on the application of FRP DoD Department of Defence E General tensile modulus [GPa]
E1 Tensile modulus in the main load bearing direction [GPa]
E2 Tensile modulus in the transverse direction [GPa] EB-beam Euler-Bernoulli beam ER External interface requirement FFBD Functional flow block diagram FR Functional requirement FRP Fibre reinforced polymer(s) G General shear modulus [GPa]
G12 Shear modulus of the laminate [GPa] I Moment of inertia [m4] IR Internal interface requirement l Span of the gate [m] MCA Multi Criteria Analysis NAP Dutch reference level (Dutch: Normaal Amsterdams Peil) P-K Pannerdensch Kanaal q Distributed line load at the middle of the gate [N/m] RA Reliability and availability Recr Recreation variant for design level 2 RINK-SSC Risico Inventarisatie Natte Kunstwerken – Sluis Stuw Complexen. RR Realisation requirement RvR ‘Room for the River’ project RWS Rijkswaterstaat SE Systems Engineering SLS Service ability limit state T-beam Timoshenko beam ULS Ultimate limit state v Deflection at the middle of the gate [m] WKC Water power station (Dutch: Water kracht central) WWII Second World War
Greek symbols ε Strain γ Angular rotation
: ARCADIS & TUDelft v
New canalization of the Nederrijn and Lek Main report
Project map
Design level 1
IJssel
Lek Nederrijn
Pannerdensch Kanaal Noord
Merwedes Waal
Design level 2
Design level 3 Design level 4
Design level 5
Figure 2 Project maps of the five design levels. The red lines represent the systems boundaries (based on: Google Maps and ANWB topografische atlas).
vi ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
1 Introduction
The graduation research focusses on two main topics which are the recanalization of the Nederrijn and Lek and the application of Systems Engineering (SE) for large scale hydraulic engineering projects. A design methodology based on SE for large scale hydraulic engineering projects is adjusted in order to design the recanalization of the Nederrijn and Lek. The developed design methodology is tested for the Nederrijn and Lek recanalization case study. Recommendations for the application of the methodology and the design of the recanalization are presented in this main report. First an introduction is given for the Nederrijn and Lek. The formation and the purposes of the present canalization are described in section 1.1. Secondly in section 1.2 a short introduction is given of SE which is applied during this graduation research. The problem definition is described in section 1.3, and the goal is described in section 1.4. The scope which is defined in the work plan prior to the graduation research is presented in section 1.5. This scope is defined in order to define the borders of the graduation research. The reading guide of the report is included in section 1.6.
1.1 INTRODUCTION TO THE NEDERRIJN AND LEK CANALIZATION
Throughout history, rivers are used as fresh water supply and as transport route. People tried to shape and influence the rivers to increase their navigational and fresh water benefits and their safety level. This also counts for the Rhine branches located in the Netherlands. From 1850 till 1916 normalisation works were implemented to increase the depth for navigational needs. Canalization was an option when the depth remained insufficient after normalisation. Three weirs were constructed in the Rhine delta to improve the depth of the IJssel and to steer the discharge distribution at the IJsselkop during dry periods. The weirs were constructed in the Nederrijn near the villages of Driel, Amerongen, and Hagestein to which they are named. The weir complexes are presented in Figure 1-1.1 The upstream weir generates a minimum water level of +8.3m NAP at the IJsselkop for realising a sufficient depth in the IJssel and diverting sufficient fresh water into the IJsselmeer. Weir complexes Amerongen and Hagestein regulate the water level of the Nederrijn and the upper section of the Lek. Navigation is hardly possible in the IJssel, Pannerdensch Kanaal, Nederrijn, and Lek without use of these weirs due to insufficient draught in periods of low discharge (Ministerie van Verkeer en Waterstaat, 1957). The available draught could reduce to 0.80 meters for the IJssel which was measured before the canalization of the Nederrijn (Til, 1961). Furthermore, insufficient water is diverted towards the IJsselmeer when weir Driel is out of service in periods of droughts. This result in depletion of the Netherlands located at the north of the fictive line of Amsterdam-Nijmegen.
1 Seven weirs are constructed in the Maas, but the Maas is not part of the Rhine Delta and is therefore beyond the scope of this graduation research. Some of these weirs and the working of a weir are described in appendix A.6.
: ARCADIS & TUDelft 1
New canalization of the Nederrijn and Lek Main report
Figure 1-1 The schematised Nederrijn and Lek system (ARCADIS, 2010)
1.2 INTRODUCTION TO SYSTEMS ENGINEERING
Systems Engineering (SE) offers an integrated and structured set of methodologies which can be used for successfully realising a (large scale) project. SE is based on the concepts of thinking in systems. The systems engineering method is applied for many projects like the Apollo program. A system is an integrated set of cooperating elements which have to realise a pre-defined goal. The core elements of SE are: Specifying a need in a structured manner. Designing an appropriate solution for this need in a structured manner. Realising the solution in a structured manner. Maintaining the realised solution on a structured manner. Verifying and validating the process and the designed solution in a structured manner Managing the system for the whole life-time in a structured manner. A large scale design project is decomposed in several design levels and a separate design loop is implemented per design level. Every design level imposes a set of requirements for the specific design level. The derivation of the requirements and the design choices that are made are traceable when the methodology is applied well. SE is a method that is not ready to use, but it is a large toolbox containing many developed tools. Every design project can be approached using a unique toolset from the SE toolbox (Rijkswaterstaat & ProRail, 2009) & (ARCADIS, 2008).
1.3 PROBLEM DEFINITION
The problem definition for this project is twofold. A problem definition is defined for the Nederrijn and Lek (1.3.1), and a problem definition is defined for the application of SE methodology (1.3.2).
1.3.1 PROBLEM DEFINITION FOR THE NEDERRIJN AND LEK
The Nederrijn weirs are getting older and with their aging their reliability and availability decreases over time. Rijkswaterstaat (RWS) awarded ARCADIS the project RINK-SSC in order to determine their reliability and availability of the weir complexes. The assessment indicated that the weirs do not meet the desired reliability and availability levels for the present situation. Furthermore, parts of the structures have exceeded their technical life span and do not fulfil to the design checks. So, the weirs do not meet the desired RA levels and design checks (ARCADIS, 2010). The problem can be summarised into the following problem definition:
The existing weirs located in the Nederrijn are in an insufficient condition and their reliability and availability decreases over time. (ARCADIS, 2010), which is undesirable.
2 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
1.3.2 PROBLEM DEFINITION FOR THE APPLICATION OF SYSTEMS ENGINEERING
The Dutch engineering sector applies an ‘integral design process.’ The ‘integral design process’ is composed of a design loop per conceptual, feasibility, preliminary, and detailed design level. The scale and the rate of details increases per design level. So, first a global design with minor details is made for the conceptual level; and at last a detailed design is made including many details. Each design loop is composed of an analysis, synthesis, simulation, evaluation, specification, and a review loop. The ‘integral design process’ is applied successfully for large projects like the Eastern Scheldt barrier (de Ridder, 2012). Nowadays, the Dutch engineering firms are obliged by the government to apply the Systems Engineering (SE) methodology. The SE methodology is already applied successfully in various projects of which the Apollo program is the best known. Initially, SE in the Netherlands was applied as a tool to prepare Design and Construct contracts. Later on, SE was implemented earlier in the design process coupled to an Environmental Impact Assessments and public decision-making. The design choices made during the tender phase were substantiated using SE, resulting in traceable requirements. However, the requirements and design choices resulting from the Environmental Impact Assessment were not traceable because SE was not applied for the Environmental Impact Assessment. The opportunity of the application of SE in an earlier stage of the project is a better transition between the Environmental Impact Assessment and the tender phase, because the requirements and design choices are traceable for the Environmental Impact Assessment and for the tender phase. A major ‘transition’ between the two phases could be eliminated when applying SE for both phases. Presently, the SE methodology is applied for the Environmental Impact Assessment and the tender phase, but SE is still not applied for policy making. A major ‘transition’ is now present between the substantiation of a design choice in the political domain and for the Environmental Impact Assessment executed by the engineering firms that follows. So, the major transition in the decision making process is not solved but relocated in between the political decision making and the Environmental Impact Assessment. Secondly, the description of SE in the Dutch SE manuals is very general; the methodology could be applied for the design of hydraulic structures but also for medical operations in a hospital. This general description results in a vague and ambiguous design methodology and the necessity to further specify the design process on a project-by-project basis.
ARCADIS requested the author to develop, apply, and test a design methodology based on the SE methodology. The SE methodology has to be: applicable to make strategic, large scale design choices for interventions in the present ‘wet infrastructure’ which, in fact, are made in the political domain. applicable to realise a continuous line of design choices from large scale design choices with few details to small scale design choices with many details. So, it has to be investigated whether the mentioned transition could be positioned before the policy making process or even be eliminated from the design process. This results in the following problem definition for the application of SE within this graduation research:
Currently, a major ‘transition’ of the substantiation of design choices made by politics and for the design process implemented by engineering firms is present. The ‘transition’ and the general written SE manuals results in ambiguities during the design process.
1.4 GOAL
The aim of the graduation research is twofold, namely: 1. Provide an alternative design for the present canalization of the Nederrijn and Lek. 2. Develop and apply a design methodology based on SE for large scale hydraulic design projects.
: ARCADIS & TUDelft 3
New canalization of the Nederrijn and Lek Main report
The alternative design has to regulate the flow over the branches and has to regulate the water management functions of the reach. The alternative is designed top down taking into account the present environmental influences and the present functions of the Rhine branches. The alternative design results into a changed (hydraulic) situation which could hamper municipalities, organisations, entrepreneurs, nature, etc. which are located along the reach, but also results in benefits like reduced delay for navigation. The available time for the implementation of a new alternative design is limited due to the bad state of the weirs. Therefore, a set constraint of the project is: alternative designs must be implementable within one decennium from now. A methodology for the application of SE within large scale projects for the redesign of waterways is developed for the design of the recanalization. The new canalization is elaborated in five sequential design levels characterised by geographical systems boundaries. The first design level starts at Delta level, the second continues at Nederrijn-Lek level, the third proceeds at the weir location level and at last the weir build-up is designed. Furthermore, the gates of the weir are designed from fibre reinforced polymers (FRP), which is a self-imposed secondary goal of the graduation research. To realise the goal, sub-goals are defined. The sequential sub-goals are defined to accomplish the goal, which are: Analyse the present water system. Develop a methodology for implementing SE in a multilevel hydraulic design project. Determine the river branch which could be canalised. Determine an alternative for the canalization of the Nederrijn and Lek. Determine the exact location of the weirs needed for the canalization. Determine a suitable configuration of the weir. Determine a suitable structural design of the hydraulic gates in FRP.
1.5 SCOPE
During graduation, the river branches of the Rhine delta which are located in the Netherlands are taken into account; the Rhine branches located upstream from Lobith are therefore not taken into account in this graduation research. An alternative for the present canalization is designed using a Systems Engineering (SE) methodology to order the design process. SE is not a set methodology but a toolbox with separate methods. A toolset from the toolbox is composed for developing a design methodology to structure the design process for the design of the canalization from the design of the Dutch Rhine delta till the design of a weir gate.
1.6 READING GUIDE
First background information for the graduation research is given in chapter 2 and chapter 3. The formation of the present canalization is described in chapter 2 in order to get a better understanding of the formation and functioning of the Nederrijn-Lek canalization. In chapter 3 the developed design methodology is described. This methodology is tested for the recanalization case study in the subsequent chapters covering the design process. The design process is divided into five design levels. The first design level concerns the design at delta level. A choice is made between the Rhine branches for recanalization. This design level is described in chapter 4. The second design level, which is described in chapter 5, concerns the design at reach level. The characteristics of the Nederrijn-Lek reach are determined and an overview is given of the best options for recanalization. At the end of this chapter a configuration is chosen for further elaboration. Chapter 6 describes the third design level. An overview of possible locations for a new complex and the exact location of the downstream weir are designed. The fourth design level, which is described in chapter 7, concerns the build-up of the weir. The total width of the weir, the amount of openings, hydraulic gate type, and accurate discharge control are designed in this chapter. Chapter 8 treats the design of the
4 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
hydraulic gate. The characteristics of FRP and the structural design of the gate are given. Chapter 9 describes the evaluation of the design methodology for the five design levels. The comments made in the intermezzos are further elaborated and a comparison is made between the Systems Engineering methodology of RWS, the applied methodology, and the design methodology of the Technical University of Delft. The report ends with the conclusion and the recommendations which are given in chapter 10. The conclusions and recommendations are given for the tested SE methodology and the resulting design of the weirs.
: ARCADIS & TUDelft 5
New canalization of the Nederrijn and Lek Main report
2 Present situation
Background information of the history of the Nederrijn and the present structures is given in this chapter. The history of the past 200 years of the Rhine delta and specially the Nederrijn and Lek canalization is described in section 2.1 to provide an overview of the made design choices and formation of the present canalization. The structural design of the weirs which are built in the 1950’s is described in section 2.2.
2.1 HISTORY OF THE NEDERRIJN AND LEK
The major improvement works of the main rivers started during the 19th century. The shortcomings of the 19th century river system and the normalisation works are described in 2.1.1. Secondly the canalization works are described in 2.1.2. The implemented canalization is described in 2.1.3 and at last the dam regime which is presently used is described in 2.1.4. More details about the history and the present weirs are given in appendix A. One major river improvement before the 19th century which has to be mentioned first is the excavation of the Pannerdensch Kanaal in 1707 which is indicated in Figure 2. This channel had to distribute the water over the Rhine and IJssel and the Bijlands Kanaal.
2.1.1 NORMALISATION OF THE RHINE BRANCHES
The Dutch river systems had major shortcomings around 1850. The summer levees were too high, islands were situated in the summer beds, large sandbars were located in the river, many trees were located at the floodplains, the Maas and the Waal joined each other at the castle of Loevestein (65 kilometres upstream from Hook of Holland) which caused failure of the levees along the Maas, and the river was relieved by spillways which caused large scale planned floods for high waters. These shortcomings resulted in major floods which took place in 1809, 1820, 1855 and 1861. Several committees were founded between 1800 and 1850 to protect the hinterland from flooding. Two inspectors of the Waterstaat, Ferrand and van der Kun, compared the recommendations of the committees and concluded that the rivers had to be able to transport the water and ice by themselves and not by relieving the river by spillways. They recommended to normalise the river in order to erode the islands and sandbars away and deepen the waterway. Furthermore Ferrand and van der Kun advised to excavate the Nieuwe Merwede and to disconnect the Maas from the Waal to solve the downstream blockage of water at Loevestein (Heezik, 2006). The normalisation of the Rhine was implemented between 1850 and 1896. Bends were cut off and groins were constructed. The normalisation works were finished in 1916. The result of the normalisation was a three metre deep river from the border of Germany to the city of Dordrecht and a normalised Nederrijn and Lek of two metres deep. However, the depth of the IJssel remained insufficient after normalisation. The only solution for improving the navigability of the IJssel appeared to be canalization (Woud, 2006).
6 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
2.1.2 CANALIZATION OF THE RHINE BRANCHES
The draught of the IJssel remained insufficient during summer after normalisation. Statistics from 1901 till 1950 showed a depth of less than 1.5 metres during 30 days during an average year. Therefore, ‘Studiedienst Bovenrivieren’ made a plan for the canalization of the IJssel. The goals of the canalization of the IJssel were: Improving the draught of the IJssel. Diverting more fresh water into the Lek for drinking water purposes. Diverting more fresh water into the Lek for counteracting the salt intrusion from sea. The disadvantages of the canalization of the IJssel were the blockage of fresh water from the IJsselmeer and an obstructed waterway to the North. The weirs would temporally be opened to supply the IJsselmeer in periods of drought. This plan was presented at the Dutch parliament, but was never approved due to the outbreak of the Second World War (WWII) (Rijkswaterstaat; Directie bovenrivieren, 1979). During WWII a plan was developed for the canalization of the Nederrijn which was an alternative solution for the IJssel canalization. The main goals of the canalization of the Nederrijn were (Blokland, 1957): Improving the navigability of the IJssel, Lek, and Pannerdensch Kanaal. Improving the fresh water supply of the northern part of the Netherlands by diverting fresh water into the IJssel. The main disadvantage of the canalization of the Nederrijn was the increased salinization of the western part of the Netherlands due to the decrease of fresh water resulting from the canalization. Ir. J. van Veen stated to close off some estuaries to counteract this disadvantage and to improve the safety of the South- Western delta. However, this was not possible due to political-social circumstances. But, after the floods of 1953, the political-social circumstances changed dramatically and the Deltaplan was approved, which resulted in several (partly) closed off estuaries. The Dutch parliament decided to implement the canalization of the Nederrijn, because the main disadvantage was solved. Furthermore, non-obstructed waterways are present to the North (IJssel) and to the West (Waal) in this alternative (Ministerie van V&W, 1957). So, the canalization of the Nederrijn in combination with the Deltaplan was the only option to increase the depth of the IJssel and increase the flow of fresh water towards the northern part of the Netherlands without too much hindrance for navigation (Til, 1961).
2.1.3 IMPLEMENTED CANALIZATION
Three weirs were built in the Nederrijn, which are presented in Figure 2-1. Weir Driel is the control instrument for the distribution of water and the other two weirs regulates water levels of the Nederrijn. Backwater effects caused by weir Driel at the Pannerdensche Kop had to be prevented. Therefore the river bed level of the Pannerdensch Kanaal and the upstream reach of the Nederrijn were lowered. Furthermore, bends had been cut off near the village of Doesburg and Rheden to increase the water demand of the IJssel (Ministerie van V&W, 1990).
Figure 2-1 Implementation works for the present canalization of the Nederrijn (Google Maps)
: ARCADIS & TUDelft 7
New canalization of the Nederrijn and Lek Main report
2.1.4 DAM REGIME
A minimum discharge of 285m3/s for the IJssel is the aim of the present dam regime which is realised for nine months a year. This situation is maintained for discharges higher than 1300m3/s and lower than 1500m3/s at Lobith. A discharge of 250m3/s for the IJssel is maintained for discharges at Lobith lower than 1300m3/s. A flushing discharge of 25m3/s is maintained for the Nederrijn to guarantee a sufficient water quality in the Nederrijn and Lek (Rijkswaterstaat, 2011). A higher discharge flows into the Nederrijn when the discharge at Lobith exceeds 1500m3/s and the discharge of the IJssel is maintained at 285m3/s. Weir Driel is fully opened for a discharge of 2350 m3/s at Lobith and weir Amerongen and Hagestein are fully opened at a discharge of 3000 m3/s (Brinke, 2004). For a fully open river, 21% of the discharge at Lobith flows into the Nederrijn. This corresponds to a maximum discharge of 3.376 m3/s after completion of the Room for the River project (RvR) (Ministerie van V&W, 2007).
2.2 DESIGN OF THE PRESENT NEDERRIJN WEIRS
The structural design of the Nederrijn weirs is described in this section. The three weirs do have similar designs, therefore only the design of weir Hagestein is given. A more detailed description of the structural design is given in Appendix A.5. The Nederrijn weirs are equipped with two visor gates with a span of 48 metres each. The vertical clearance underneath the gates is 9.1 metres which is available 5 metres from the pylon and abutments when they are fully lifted. A tunnel located at the upstream side of the complex provides a passage from bank to bank and to the mid pylon when the gates are lifted. The lifting towers are placed at the mid pylon and at the abutments. The hoisting equipment is accessible by stairs and is placed in the engine room at top of the pylons. The visor gates are used for the coarse discharge control and an extra cylinder valve is placed in the mid pylon for the accurate discharge control. Weir Amerongen and Hagestein are founded on a strong layer of sand. The soil conditions for weir Driel ware bad; a thick layer of clay had to be excavated and replaced by sand in order to found weir Driel directly on the subsoil (Rijkswaterstaat, 1955).
Figure 2-2 Front and side view of weir Hagestein (Rijkswaterstaat, 1955)
8 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
3 Design methodology
The design of the recanalization of the Nederrijn and Lek starts with a new design of the Dutch river branches and ends with the design of a weir gate. The quantity of information to be processed for the recanalization of the Nederrijn and Lek is large. Therefore, a design methodology has to be applied to structure the design process and to give an insight in the complexity of the object which has to be designed. The design methodology applied within this graduation research is called ‘Systems Engineering’ (SE) which is based on the principle of thinking in systems. The design process is divided in multiple design levels for which a decision is made per design level. The design approach which is used for the recanalization is described in section 3.1, the application of the developed model is described in section 3.2, and the design steps per design level are described in 3.3.
3.1 DESIGN APPROACH USED FOR RECANALIZATION
The methodology of SE provides a subdivision of the elementary design process into products, processes, interfaces, project boundaries, and functions. An iterative process form large to small-scale and vice versa is used in this methodology. The process could be described like a V for an ideal situation as presented in Figure 3-1. Not only one V is present for a design process, but several V’s for distinct specialisations are proceed and several V’s are aborted during the project because they turned up to be unfeasible. The parallel V’s are indicated by the blue arrows in Figure 3-1. The design levels are horizontally arranged and specified for the recanalization of the Nederrijn and Lek. The first row represents the design at large scale and the lower rows represent the detailed design. The purple dots within the green ellipses represent the design steps of the design loop which are executed within a design level. The three design steps are: the requirements analysis the functional analysis and allocation of the requirements the design synthesis. The ‘requirements loop’ connects the ‘requirements analysis’ with the ‘functional analysis and allocation’ and the ‘design loop’ connects the ‘functional analysis and allocation’ with the ‘design synthesis.’ Furthermore, a verification arrow is presented which connects the design synthesis to the requirements analysis. A more detailed description of SE is given in appendix B.1.
: ARCADIS & TUDelft 9
New canalization of the Nederrijn and Lek Main report
Figure 3-1 Applied V-model for the recanalization (based on: (Rijkswaterstaat & ProRail, 2009) & (Department of Defense, 2011))
3.2 APPLICATION OF THE V-MODEL
Only the first diagonal with the green ellipses of Figure 3-1 is executed because the realisation phase is not included in the graduation research. Therefore, only the development phase of the V-model is executed. Multiple conceptions for the application of the V-model and SE are present. Figure 3-2 defines two conceptions. The first conception is presented at the left of Figure 3-2. All the requirements for the project are defined during the start of the project. Specifications for the delta level and the weir design level are simultaneously defined. The five design levels are also executed simultaneously which results in ambiguities during the project. The second conception, which is developed by the author, defines the requirements applicable for a specific design levels as presented at the right of Figure 3-2. A small set of requirements is defined per level, which impacts the design solution at a specific design level. The design resulting from design level 1 is considered in design level 2. More detailed requirements are developed during the design loop of level 2 which are applicable for the Nederrijn-Lek configuration. The result of design level 2 is verified for design level 1, and a preview is made for design level 3 in order to prevent unexpected events during the design process. At a certain point (after design level 4) the dimensioning of the elements takes place. The allocated requirements to the elements specified for design level 4 are used as top requirements for design level 5. This transition point is located at the border of the concept design (design level 1 till 4) and the preliminary design of the gate (design level 5). No extra requirements are elaborated for design level 5 because the set of requirements expands exponentially at this design level and the SE method is not suitable anymore. Using the top requirements for the structural design of the gate is a better option for the elaboration of the gate in order to keep the design process clear. Furthermore, it is less meaningful to define the structural components in functional terms. Therefore, the area analysis, requirements analysis, and the functional analysis are not performed for design level 5 and a design is made using the top requirements of design level 4.
10 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
Figure 3-2 ‘Normal’ design method (left); design method used in the graduation research (right)
Intermezzo The designs of the Nederrijn weirs are equal for Driel, Amerongen, and Hagestein. They have all been equipped with an underflow gate which resulted in reduced production costs. An underflow gate is best suitable for discharge regulation and an overflow gate for water level regulation. Therefore, weir Driel could best be equipped with an underflow gate (it has to regulate the discharge distribution) and weir Amerongen and Hagestein with an overflow gate (these weirs have to regulate the upstream water levels). However, a decision for weir type at design level 4 was already made before the functions of the weirs were defined in level 1 or 2. This also counts for the foundation of weir Driel. A design founded directly on the subsoil was made for the complexes before the exact locations were determined. The soil conditions at Driel were bad, so extensive soil improvements were implemented to be able to found weir Driel instead of developing an alternative design on piles. Again a decision was made at design level 4 before the decisions for design level 2 and 3 were made.
3.3 DESIGN STEPS OF THE DESIGN LOOP
The requirements analysis is performed using the 15 steps proposed by the Department of Defence (DoD) which are presented in Figure 3-3. The requirements analysis starts with an input of objectives, description of the surroundings and the stakeholders’ needs. The area analysis and the stakeholder analysis, which are presented in appendix C and D, are used as input and analysed in the requirements analysis per design level. The result is a list of functional requirements (FR), performance requirements (PR), internal constraints requirements (IR), and external constraints requirements (ER) which are fully listed in appendix F per design level. In the main report only the most important requirements which are impacting the design solution the most are presented.
: ARCADIS & TUDelft 11
New canalization of the Nederrijn and Lek Main report
Figure 3-3 15 steps of requirements analysis (Department of Defense, 2011)
The functional analysis is the step between the overview of requirements and the design of the objects. Due to the corridor shape of the river branches a functional flow block diagram (FFBD) is used. The FFBD defines the task sequence and the relationship between functions. The start of the sequence is the upper boundary of the corridor and the end of the sequence is the lower boundary of the system. The steps between the upper and lower boundary represents the sub-functions of the system (ARCADIS, 2008). The system functions are identified and transformed in system objects which have to perform the functions. Subsequently, the requirements are allocated to the system objects (CROW, 2011). The full functional analysis per level is given in appendix G. Only the main functions and objects are briefly described in this main report. Design criteria are used for the qualitative or quantitate comparison of variants. A qualitative design choice is made without use of a Multi Criteria Analysis (MCA) which defines a revenues/costs ratio per variant. A quantitative design choice is made with use of a MCA. The design criteria per level are given in appendix H and the description of a MCA is given in appendix B.2.
12 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
4 Design level 1 - Rhine delta
This chapter contains the description of the design made for the Rhine delta. The design process starts with a description of the project area in section 4.1. The project area is presented in the drawing of design level 1 which is included in the project map (Figure 2). The area is analysed resulting in a set of requirements which are given in section 4.2. Thereafter, the functional analysis is described in section 4.3. In this section, the functions of the system are defined and objects are derived from the functions. The requirements are allocated to the derived objects. The resulting design is described in section 4.4 which is based on the allocated requirements to the objects. A conclusion is provided in section 4.5.
4.1 PROJECT AREA
The project area of the first design level covers the main Rhine river branches located in the Netherlands which are presented in Figure 2. A new canalization can be implemented in the Waal, IJssel, and Nederrijn-Lek. The headlines of the area analysis are presented in this section; the full area analysis for design level 1 is covered in appendix A.3.
4.1.1 TRANSPORT OF WATER
The transport of water is one of the most important aspects of the area analysis because it is the main function of the rivers. Two extremes for the transport of water are distinguished, namely extremes for low discharges and extremes for high discharges. The distribution of water for both cases is presented in Figure 4-1. A discharge at Lobith of 1.200m3/s or lower which represents low discharges is present for 10% a year (ARCADIS, 2010). The percentages given in the right figure of Figure 4-1 holds for high discharges and an undammed Nederrijn-Lek (discharges higher than 3000m3/s). The maximum discharge after completion of the RvR project is 16.000m3/s and 18.000m3/s for 2050. This increase of discharge is set for the climate change and the river improvements implemented in Germany (Hermeling, 2004).
Figure 4-1 Distribution of discharge for dry periods (left) (Noordhoff Atlasproducties, 2011) and for maximum discharge (right) (Brinke, 2004)
: ARCADIS & TUDelft 13
New canalization of the Nederrijn and Lek Main report
4.1.2 NAVIGATION
The use of the waterways by commercial shipping is the other aspect which is important for recanalization. The main transport routes are defined for the recanalization. The shipping corridors which are mostly used are: the Waal the waterway connection from Rotterdam to Antwerp the connection of Amsterdam to the Waal via the Amsterdam Rijnkanaal (AR-kanaal) the connection of Rotterdam with Amsterdam via the Lek, Lekkanaal, and the AR-kanaal.
The IJssel and the Nederrijn are of minor importance in comparison to the above mentioned waterways. The reach of the Lek between the Lekkanaal and the AR-kanaal is even less used as presented in Figure 4-2. So the main waterways are the Waal, the connection with Antwerp, the Lek till the Lekkanaal, and the AR-kanaal.
Figure 4-2 Inland waterway cargo transport (Noordhoff Atlasproducties, 2011)
4.2 REQUIREMENTS ANALYSIS
The 15 steps of requirements analysis applied for the Rhine delta are fully elaborated in appendix E.1. Only the system boundaries (section 4.2.1) and the main requirements (section 4.2.2) are given. The main requirements are impacting the design solution the most. Other requirements which are impacting the design solution less are listed in the full list of requirements which is included in appendix G.1.
4.2.1 SYSTEM BOUNDARIES
The system boundaries go beyond the geographical boundary of a system. The system boundaries define which part of the system is controllable by the engineer, and which part is not. Furthermore, the interactions between system elements are described. Design level 1 consists of one large system indicated by the blue lines of the first figure presented in Figure 2. The red lines of Figure 2 represent the system boundaries. The following components are part of the system: the Boven-Rijn the IJssel the Waal bifurcation the IJsselkop the Merwedes bifurcation the Pannerdensche Kop the Noord confluence of the Lek and the Noord the Pannerdensch Kanaal confluence of the Beneden Merwede the Nederrijn and the Noord. the Lek
14 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
The upstream boundary condition at Lobith contains the rate of discharge and navigation entering the Netherlands at Lobith and the downstream boundary condition at the Haringvliet, Nieuwe Waterweg and the IJsselmeer contains the rate of discharge and navigation leaving the Rhine branches. The boundaries located along the system (floodplains, waterway connections, levees) contain: The discharge entering the main river branches. Physical constrictions of the river profile.
4.2.2 MAIN REQUIREMENTS
A full list of requirements for this design level is given in appendix F.1. Only the main requirements resulting from the requirements analysis are presented in this section. The main requirements are impacting the design solution the most. The requirements which are impacting the design solution less are listed in appendix F. Furthermore, the requirements which are listed in appendix F are specified in more detail. For example, the minimum discharges for an ecologically healthy situation (FR 2) are presented in this appendix. The requirements which have the largest impact on the design are: FR 2) An ecologically healthy situation must be present. FR 3) Sufficient draught must be present in the waterways for navigation. IR 2) Navigation on the main transport route (the Waal) may not be hampered by objects. ER 1) The combined discharge capacity must be equal to the maximum discharge entering the Netherlands. ER 2) The discharge capacity of the downstream water system must be equal to the maximum discharge of the system. RR 2) The new canalization must be implemented within one decade.
4.3 FUNCTIONAL ANALYSIS
The system ‘The river branches’ representing the Rhine delta has to be defined in functional terms which results in a functional tree. The functional tree is the basis for the object tree to which the requirements are allocated. The objects of the object tree are derived from the functions using a functional flow block diagram (FFBD). The full functional analysis is not given in the main report but included in appendix G.1.
4.3.1 FUNCTIONS
The functions of the system ‘The river branches’ are decomposed into two sub-functions in the second breakdown as presented in Figure 4-3. The second break down contains two functions, namely: Transportation of water. Water has to be transported from the upstream boundary to the downstream boundary. A certain amount of water has to be transported and distributed for realising the safety norms, water quality norms, and water quantity norms. Providing navigation. Navigation has to use the system by moving through the system. The vessels need to have sufficient clearance, draught, and width for navigating over the Rhine branches.
: ARCADIS & TUDelft 15
New canalization of the Nederrijn and Lek Main report
Figure 4-3 Functional tree for the Rhine delta
A functional flow block diagram defines the task sequence and the relationships between functions (CROW, 2011) & (ARCADIS, 2008). The start of the sequence for this level is the upper boundary of the system which represents the water inflow of the system. The end of the sequence is the lower boundary of the system which represents the outflow of water from the system. Several steps are located in between the upper and lower boundary represented by three blocks in Figure 4-4. Every block represents a sub- function of the system. First, water enters the system from the upper boundary which is the entrance of the Rhine at Lobith. The water is transported to the water distribution objects (Pannerdensche Kop and IJsselkop) over distinct branches. Subsequently water is transported to the downstream boundary which is the IJsselmeer, the Haringvliet, and the Nieuwe Waterweg.
Figure 4-4 Functional flow block diagram (FFBD) for the Rhine delta
4.3.2 OBJECTS
The object tree is a top down approach arranging the objects from coarse to fine. The objects should be defined as “solution free as possible” in order to avoid a ‘tunnel vision.’ The system ‘The Rhine branches’ is split up in two sub-objects based on the FFBD. The system consists of a water transportation object, and an object which determines the discharge regulation over the river branches as presented in Figure 4-5. The set of requirements is allocated to the objects. The object tree with the allocated requirements is included in appendix G.1.
Figure 4-5 Object tree of 'The Rhine branches' system
16 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
4.4 DESIGN SYNTHESIS
A new canalization could be implemented in the Waal, IJssel, and the reach Nederrijn-Lek. Three options are set per reach, namely a decrease in discharge, an increase in discharge, and an equal discharge. These options are set for dry and wet periods as presented in Table 4-1 (the present situation is defined to be equal to the situation after the implementation of the project ‘Room for the River’ (RvR)). A qualitative assessment is made for filtering the most suitable options on basis of the derived requirements and functions. The filtered table is also included in Table 4-1. The full reasoning for the filtering is presented in appendix I, the headlines of the substantiation are presented in the sequential enumeration.
Table 4-1 Filtered table of options for the selection of the Rhine branches
Reach Dry periods Wet periods Lower discharge Lower discharge Waal Equal discharge Equal discharge Higher discharge Higher discharge Lower discharge Lower discharge IJssel Equal discharge Equal discharge Higher discharge Higher discharge Lower discharge Lower discharge Nederrijn-Lek Equal discharge Equal discharge Higher discharge Higher discharge
Equal discharges for the IJssel, Waal and the Nederrijn-Lek for dry and wet periods are possible because the present water distribution over the Rhine branches performs well. Waal: o Lower discharges for a free flowing river for dry periods are not desirable because the depth and draught decrease. The Waal is the main transport route, so a decreased discharge hampers the navigation significantly. Furthermore, implementing weirs hinders the navigation, so the canalization of the Waal is not desired. o Lower discharges for wet periods are not desired because the Pannerdensche kop has to be redesigned in order to divert more water into the IJssel which results in large scale river improvements. Therefore, a lower discharge for wet periods is not feasible. o Higher discharges for dry periods are not preferable because less water is diverted in the IJssel and Nederrijn which is detrimental for the fresh water supply to the IJsselmeer and draughts. So, a higher discharge for dry periods is not desired. o Higher discharges for wet periods are desirable. In this case, the Waal relieves the other branches and the IJsselmeer. Space is already reserved along the Waal for the implementation of the second phase of the Room for the River project. A higher discharge is possible till a maximum discharge of 18.000m3/s which is the limit of the second phase of the Room for the River project. IJssel: o Lower discharges for dry periods are not desirable because a minimum flushing discharge of 110 m3/s must be present for a canalized IJssel in order to supply the IJsselmeer during dry periods. This discharge is higher than the present minimum flushing discharge of the Nederrijn. So, more water is extracted from the other rivers which reduce the depth which is disadvantageous (Rijkswaterstaat; Directie bovenrivieren, 1979). o Lower discharges in the IJssel for wet periods are advantageous for relieving the IJsselmeer when the desired discharge capacity of the Afsluitdijk sluices and the storing capacity of the IJsselmeer are not met. The extra capacity has to be found in other river
: ARCADIS & TUDelft 17
New canalization of the Nederrijn and Lek Main report
branches which results in river improvement works and changes at the distribution points. These river improvements would take longer than 10 years, and are therefore considered not achievable within the current context. o Higher discharges for dry periods are advantageous for the IJssel, but disadvantageous for the Waal. A higher discharge results in lower depths which is disadvantageous for navigation. So a higher discharge for dry periods is not achievable. o Higher discharges for wet periods can be implemented because space is already reserved for river improvement works but is not desired for the IJsselmeer when the discharge capacity of the Afsluitdijk sluices and the storing capacity are not met. Nederrijn-Lek: o Lower discharges for dry periods are not desirable because the minimum flushing discharge which is present in the Nederrijn-Lek for realising a sufficient water quality for dry periods cannot be reduced further. o Lower discharges for wet periods can be implemented but includes a changed water distribution at the Pannerdensche Kop and the IJsselkop. Changing the lay-out of the two bifurcations would take longer than ten years. So, lower discharges for wet periods are not feasible and are therefore considered not achievable within the current context. o Higher discharges for the Nederrijn and Lek are possible for dry periods, however the depth of the IJssel decreases, which hampers the navigation and the water supply of the IJsselmeer. Therefore, this option is not feasible. o Higher discharges for wet periods are not desirable due to the narrow river profile of the Lek. Measures which could be implemented are not easily implementable and results in high costs. Increasing the capacity of the IJssel of the Waal is a better solution.
4.5 CONCLUSION
The redevelopment of the reach Nederrijn-Lek while preserving the present distribution of water at the IJsselkop is chosen because the impacts of this solution are lower with respect to the impacts of the redevelopment of the Waal and the IJssel or a changed dam-regime. The amount of additional river works in other river branches for the redevelopment of the reach Nederrijn-Lek are less compared to additional river works for the redevelopment of the Waal and IJssel. The building time and the building costs of the redevelopment of the reach Nederrijn-Lek are probably lower compared to the building time and costs of the Waal and IJssel. Therefore, the recanalization of the Nederrijn and Lek is the best option for the present canalization.
18 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
5 Design level 2 - Nederrijn-Lek configuration
This chapter describes the design of the configuration variants of which the project area which is presented in the drawing of design level 2 included in the project map (Figure 2). The number of weirs, the heads of the weirs, and their locations are varied to generated variants and the variants are compared with each other. The weir sections are modelled as water basins with a horizontal water level. This simplification is justified for the minimum flushing discharge which is substantiated in appendix K. The same cadence as for design level 1 is used. The design process starts with a description of the project area in section 5.1, subsequently the main results of the requirements analysis is given in section 5.2. Thereafter, the functional analysis is described in section 5.3. The main requirements and the allocation of the requirements are presented in this section. A (sub) conclusion is made in section 5.5 for the Nederrijn and Lek configuration. The functioning of the configuration is verified using a 1D static river model of which the result is described in section 5.6.
5.1 PROJECT AREA
The most important objects and characteristics which are present in the project area are described in this section. The full project area analysis is given in appendix C.2 and a short overview of the findings of this appendix is given below. The objects located along the Nederrijn and Lek are described in appendix B.3. The most important objects and characteristics in the project area which are impacting the design solution are: crossing bridges longitudinal profiles of the river and connecting channels levees inlets harbours pumping stations downstream water levels and discharge draught of the vessels upstream water levels and discharge. Figure 5-1 gives an overview of the Nederrijn and Lek with the crossing bridges (orange lines), the pumping stations and inlets (yellow arrows), and the navigational channels which are connected by sluices and barriers to the Nederrijn and Lek which are indicated by purple lines.
Intermezzo Defining the system boundaries is not an exact science but depends on the conception of the project team or designer. Two examples considering the definition of the system boundaries for this design level are given below: 1) The Nederrijn and Lek are spanning from the city of Arnhem till the city of Krimpen. Many subsystems are located along this 100 kilometres long reach. A system boundary has to be positioned between these subsystems. First, the lower system boundary was chosen near the village of Schoonhoven because sufficient draught for commercial shipping is available at this location. However, the system boundary splits the Lek in two parts at this location which is a kind of arbitrary; sufficient draught is also available at the village of Bergambacht, so the lower
: ARCADIS & TUDelft 19
New canalization of the Nederrijn and Lek Main report
boundary could also be chosen at this location. A more obvious lower boundary would be the downstream end of the Lek because the Lek ends at this location. Therefore, the lower boundary is chosen near the city of Krimpen and not near the city of Schoonhoven. 2) Water boards located along the Amsterdam Rijnkanaal drains their polder water into the Amsterdam Rijnkanaal in wet periods and extracts water from the Amsterdam Rijnkanaal in periods of drought. The extracted water from the Amsterdam Rijnkanaal originates from the Nederrijn and Lek, so the water boards located along the Amsterdam Rijnkanaal could be taken into consideration in this design level. However, the characteristics of the extraction and drainage could also be taken into consideration by the external interface requirements to reduce the complexity. Therefore, the locks of the Amsterdam Rijnkanaal are considered as system boundaries located along the Nederrijn and Lek including the characteristics of the systems located along the Amsterdam Rijnkanaal in the external interfaces.
Figure 5-1 Project area characteristics (MW-N=Merwede kanaal northern section, MW-S=Merwedekanaal Southern section, LK=Lekkanaal, AR-N=Amsterdam-Rijnkanaal Northern section, AR-S=Amsterdam Rijnkanaal Southern section)
The height and the clearance of the crossing bridges are set for the maximum design discharge and not for a dammed river. Therefore, the bridges do not impact the design of the new canalization. So, the bridges are located high enough for the implementation of a new canalization without adaptation works. The pumping stations and inlets which are indicated by the yellow arrows in Figure 5-1 are taken into consideration for the recanalization. These objects have to be redesigned for the implementation of the recanalization which is taken into account by key-ratios for the total project costs. The channels located along the rivers are connected by sluices and a barrier to the Nederrijn and Lek. The sluices do have a sill and clearance level generated by crossing bridges or lifting gates. The dammed water levels in combination with the vertical clearance and draught have to remain in between the sill level and the gate level. Furthermore, the southern section of the Amsterdam Rijnkanaal (called the Betuwepand) is in open connection with the Nederrijn for dammed situations. A barrier called Barrier Ravenswaaij closes when the water level of the Nederrijn exceeds +5.55m NAP and the open connection is lost. Vessel class Va and recreational boating class AM are using the Nederrijn and Lek. The draught of the Va class is 4.2 metres for the Nederrijn and Lek and the draught of the AM class is 1.8 metres. These draughts are indicated for the present situation which is presented in Figure 5-2. The lower orange line (line 1) represents the maximum bed level, the pink line (line 2) represents the minimum water levels for the recreational class, and the pale blue line (line 3) represents the minimum water level for the Va class.
20 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
Figure 5-2 Harbour and river characteristics of the Nederrijn and Lek for the present fully dammed situation
The draught is insufficient when the trapped line representing the dam regime (line 4) crosses the line 3 representing water levels for commercial shipping or line 2 representing the recreational boating. So, the present bed level of the Nederrijn and Lek is already insufficient in between 12km till 20km, 43km till 48km, and 68km till 82km which is verified by RWS (Alarm- en Berichten Centrum Arnhem, 2012). Currently, the bed levels are artificially lowered by dredging at these locations. The green line (line 5) and the purple line (line 6) represent the levels of the summer levees and floodplains. They have to be raised if the water levels regulated by the weirs exceed these levels. The water levels regulated by the weirs remain lower than the winter levees for all cases for the present situation, therefore the winter levees are not presented in Figure 5-2. Also harbours are indicated in Figure 5-2. The harbours are classified in: large commercial harbours, large recreational harbours, small commercial harbours, and small recreational harbours. These harbours and their location are taken into consideration for the implementation of the new dam regime by key-ratios. It is assumed that harbours have to be redesigned when the new water levels which are regulated by the weirs are lower than the present water level.
: ARCADIS & TUDelft 21
New canalization of the Nederrijn and Lek Main report
5.2 REQUIREMENTS ANALYSIS
The 15 steps of requirement analysis applied for the Nederrijn and Lek corridor are fully elaborated in appendix E.2. Only step 6 of the requirements analysis is presented in section 5.2.1 to keep the main report brief. The main requirements which are impacting the design solution the most are given in 5.2.2. The full list of requirements is given in appendix F.2.
5.2.1 SYSTEM BOUNDARIES
Design level 2 consists of one large system, from now on called ‘The water supplier’ which spans from the bifurcation IJsselkop till the confluence of the Lek and the Noord as shown in Figure 2 and Figure 5-1. The following components are part of system ‘the water supplier:’ bifurcation the IJsselkop reach IJsselkop-Wijk bij Duurstede reach Wijk bij Duurstede-confluence Lek & Noord levees (design heights) along the Nederrijn and Lek connections of channels along the Nederrijn and Lek pumping stations and inlets located between Arnhem till the confluence.
Boundaries are presented by the red lines in Figure 5-1. The upstream boundary conditions contain: discharges of the IJssel discharges of the Nederrijn navigation entering or leaving the system. And the downstream boundary contains: water levels caused by an combination of tide and runoff water transport capacity heights of levees navigation entering or leaving the system.
The boundaries located along system ‘The water supplier’ are: the winter levees plus an extra 100 meters behind the winter levees the locks of the connecting channels.
5.2.2 MAIN REQUIREMENTS
The main requirements which are impacting the design solution the most are presented in the following enumeration. These requirements originate from appendix F.2. FR 4) The system has to realise sufficient draught in the IJssel.2 FR 7) The same kind of navigation has to be able to use the system. FR 7.1) CEMT class VIa has to be able to navigate on the Lek. FR 7.2) CEMT class Va has to be able to navigate on the Nederrijn. FR 7.3) Recreational boating class AM has to use the Nederrijn and Lek. IR 5) Harbours along the Nederrijn and Lek must be accessible for the new situation. IR 6) Minimal lock dimensions must remain equal or larger with respect to the present situation. ER 3) The system has to facilitate a well-functioning water system of the connected waterways. ER 4) Cross system functions may not be hindered. ER 4.3) Navigation using the crossing waterways may not be hindered.
2 This is implemented by maintaining the same dam regime for the upstream weir.
22 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
ER 4.3.1) Navigation may not face an increase in delay or hindrance compared to the present situation. RR 3) A minimum of pumping stations and inlets should be redesigned for the implementation of the new dam regime.
5.3 FUNCTIONAL ANALYSIS
The Nederrijn and Lek corridor is described as the system ‘The water supplier’ for which the functions are defined in this section. The functional tree is the basis of the object tree for which the requirements are allocated. The objects of the object tree are derived from the functions using a functional flow block diagram (FFBD) which is included in appendix G.2. An example of a FFBD used for design level 1 is presented in Figure 4-4. The full functional analysis for design level 2 is presented in appendix G.2.
5.3.1 FUNCTIONS
System ‘the water supplier’ is subdivided into five functions and the functions are again subdivided in the third row of the functional tree which is presented in Figure 5-3. The five top functions are: Retaining water. Water has to be retained for providing two second level sub functions which are retaining water at the IJsselkop and retaining water at the Nederrijn and Lek. Water has to be retained for steering sufficient discharge into the IJssel and generating sufficient water levels in the Nederrijn and Lek. Transporting water. This function is subdivided in three lower level functions which are the discharge of upstream river water, the discharge of polder water, and the supply of water for the inlets. Facilitating navigation. Navigation has to use the system as well. A subdivision is made for moving through the system, and crossing the system. These functions have to be fulfilled by the new canalization Facilitating nature. Two lower level functions are added to this function, namely providing ecological linkage (in the waterway or at the floodplains) and to provide natural areas within the system. Facilitating users. This function considers the recreational possibilities and employment possibilities for people living in close contact with the system. These two sub functions have to be provided as well by the main system.
: ARCADIS & TUDelft 23
New canalization of the Nederrijn and Lek Main report
Figure 5-3 Functional tree for the Nederrijn-Lek configuration
5.3.2 OBJECTS
The system ‘The water supplier’ is split up into three objects which are based on the FFBD of appendix G.2. The system consists of water transportation object, an object which retains water, and an object which determines the discharge distribution over the IJssel and Nederrijn which are presented in Figure 5-4.
Figure 5-4 Object tree for 'The water supplier' system
The first subsystem fulfils the water transport function, navigational function, natural function, and user functions. This subsystem is the geographical connection between the other objects. The second subsystem fulfils the function ‘water retaining at the IJsselkop’ and distributes the water over the IJssel and the Nederrijn and Lek. The third sub object fulfils the function ‘water retaining at the Nederrijn and Lek’ and regulates the depths of the Nederrijn and Lek for the navigation and the nature connections. The allocated requirements are not included in the main report but are included in appendix G.2.
5.4 DESIGN SYNTHESIS; VARIANT STUDY
The amount of weirs, the locations and the heads of the weirs is varied for the design synthesis. The combination of weirs and river works have to fulfil the functions presented in section 5.3.1. Infinite configurations (any number of weirs located at any locations) are possible. Therefore aspects are developed which guides the design process during the brainstorm sessions. These aspects are described in
24 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
section 5.4.1. Many variants are developed during the brainstorm session based on the aspects. A ranking of the variants executed in 5.4.2 resulted in an overview of seven remaining variants which are developed in further detail. The remaining variants are described in 5.4.3 and are evaluated using a Multi Criteria Analysis which is described in 5.4.4. Furthermore a configuration variant is chosen for further elaboration for design level 3. The full design synthesis for design level 2 is described in appendix J. The headlines of the design synthesis are given in this section.
5.4.1 SYSTEM ASPECTS
A first brainstorm session (appendix J.1) indicated that one weir configurations and many weir configurations are not feasible. The river works for one weir solutions are too extensive and are not feasible to be implemented within one decade. A solution with many weirs is not feasible because this solution is not applicable within the set time constraint of ten years. Furthermore, the disadvantages for commercial shipping and recreational boating are larger compared for the existing situation. The system aspects which are developed for the second brain storm session (appendix J.2) for the Nederrijn and Lek are: ‘closable but open Rijnmond’, ‘commercial shipping’, ‘separation of recreational boating and commercial shipping’, and ‘ecology.’ Variants which make no sense like 3 weirs in between the Lekkanaal and the Amsterdam Rijnkanaal were no option and are eliminated. The schematisation of the project area used for the brain storm sessions is presented in Figure 5-5.
Figure 5-5 Basis drawing used for the brainstorm sessions for the Nederrijn-Lek configuration
A weir at the downstream section of the Lek is a demand for the ‘closable but open Rijnmond’ system aspect. This weir should be able to retain water for low discharges and to divert water into a new side channel for high discharges. Furthermore a minimal depth for commercial shipping should be present. The optimization of the navigational functions is the aim of the ‘commercial shipping’ system aspect. The minimum depth for commercial shipping should be present in the Nederrijn and Lek and the vessels have to pass as little locks as possible for the main shipping routes. Commercial shipping is fully removed from the Nederrijn and the upper section of the Lek for ‘separation of recreational boating and commercial shipping.’ The minimum draught of the Nederrijn and the upstream section of the Lek must remain sufficient for recreational boating. The advantages of this system aspect are the reduction of weir complexes and lock dimensions. The commercial shipping are using the Nederrijn and Lek mainly as secondary route towards Germany for saving fuel because no counter flow velocity is present at this reach and have the Waal as main transport route. So, damming off the Nederrijn and Lek does not impose major impacts on navigation at national level. The last system aspect is ‘ecology.’ As little weirs as possible are placed in the Nederrijn and Lek in order to improve the fish migration. Furthermore large ground water deviations must be avoided.
Intermezzo The handbook of SE states that variants have to be generated using a morphological map including the number of weirs and the weir locations as variables. Any number of weirs could be located at any location of the Nederrijn resulting in infinitive variants which have to be assessed. Therefore, system aspects are defined to guide the design
: ARCADIS & TUDelft 25
New canalization of the Nederrijn and Lek Main report
synthesis and to reduce the number of variants. Applying system aspects for developing variants is not prescribed in the SE theory. Therefore, the SE theory has to be extended with system aspects to guide the design synthesis.
5.4.2 RANKING
Developed variants based on the system aspects are filtered using the requirements listed in section 5.2.2. Variants which do not meet the requirements are deleted immediately. Furthermore requirements which perform worse at certain system aspects with respect to other variants are eliminated. This elimination is further described in appendix J.2. The closable open Rijnmond system aspect is not as good as assumed. Sufficient depth for commercial shipping without weirs is already present till the present weir of Hagestein; therefore no weirs are needed at this reach. So, every variant with a weir placed downstream from weir Hagestein is not an option. Also a filtering based on the upstream weir position is performed. This weir has to regulate the water level at the IJsselkop for the present dam regime. Furthermore, the upstream weir may not be located too far away from the IJsselkop to limit adaptation costs and not too close to IJsselkop to limit the amount of weir complexes. Therefore, the upstream weir should be located nearby Driel. The downstream weir has to be placed upstream from weir Hagestein and downstream from the Amsterdam Rijnkanaal crossing to maintain sufficient water levels for the crossing of the Amsterdam Rijnkanaal. Other configurations with a different location of the downstream weir do not meet the requirements and are eliminated. At last, an option composed of four or more weirs is more expensive with respect to an option composed of three or fewer weirs which is underpinned in appendix L. Variants composed of four or more weirs are therefore eliminated.
5.4.3 REMAINING VARIANTS
The remaining variants are drawn in Figure 5-6. The number of weirs per variant is presented before the semicolon, the locations of the weirs are presented in between the semicolons, and the system aspect is presented after the second semicolon of the subscripts of the distinct figures. The performance and the costs of the seven variants are determined for the determination of the most suitable variants. The performances are determined by scoring the variants for the design criteria presented in appendix H. The costs are determined on basis of the amount of work which has to be executed. The chosen configuration is presented as reference for the determination of the amount of work for the implementation of the new canalization.
Figure 5-6 Remaining variants after the selection
26 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
The chosen variant is a combination variant of the system aspect commercial shipping and recreation and is indicated by the dash-dotted rectangle in Figure 5-6. The upper section of the Lek (in between the Lekkanaal and Amsterdam Rijnkanaal) remains accessible for recreation. A minimal water depth of 1.8 meters is present at this reach. The reach in between weir 1 and 2 has a minimal depth of 4.2 meters, so commercial shipping is able to use the Nederrijn from the upstream side of the Amsterdam-Rijnkanaal to sail towards the IJsselkop and vice versa. During fully opened conditions, a water level of 4.2 meters is also present between the crossing of the Amsterdam-Rijnkanaal and the Lekkanaal, so vessels are able to sail through this reach. The impact on the surroundings of this variant is indicated in Figure 5-7, Figure 5-8, and Figure 5-9.
Figure 5-7 Variant “2w;Driel&Culemborg;com.ship&recr.” Harbour and river characteristics
The trapped water level representing the water levels regulated by the weirs undercuts the minimum water level for commercial shipping and overshoots the line representing the floodplains. Furthermore, the water levels for harbours changes. It is assumed that the lay-out of a harbour has to be modified for a water level decrease, and that harbours are able to cope with a water level increase because higher water levels are present for high discharges. According to Figure 5-7 the following has to be changed/adapted: The Nederrijn has to be deepened from km 15 till km 30 with 1,5 meters in order to create sufficient draught. The Lek has to be deepened from km 65 till km 85 with 1,5 meter in order to create sufficient draught for navigation sailing from the Lek towards the Lekkanaal. Levees have to be raised from km 50 till km 55 with 0,5 meter in order to keep the river in the summer bed.
: ARCADIS & TUDelft 27
New canalization of the Nederrijn and Lek Main report
1 large commercial harbour at km 25 has to be adapted for a water level decrease of 1 meter. 1 large recreational harbour at km 25 has to be adapted for a water level decrease of 1 meter. 3 small commercial harbours (km 18, km 31, km 32) have to be adapted for a water level decrease of 1 meter. 1 small recreational harbour (km 58) has to be adapted for a water level decrease of 3,5 meters.
Figure 5-8 Variant “2w;Driel&Culemborg;com.ship&recr.” Pumping stations and inlets
The inlets and pumping stations located along the Nederrijn and Lek are presented in Figure 5-8. The position of the inlets and pumping stations are given for the present dammed conditions. It is assumed that pumping stations and inlets have to be redesigned when the new dammed regime, which is indicated by the stepped line, does not corresponds to the presented pumping stations. According to Figure 5-8 the following has to be adapted: 1 pumping station has to be adapted for a water level decrease of 1 meter. 2 inlets have to be adapted for a water level decrease of 1 meter. 1 inlet has to be adapted for a water level increase of 2 meters.
Figure 5-9 Variant “2w;Driel&Culemborg;com.ship&recr.” Lock characteristics
The sill level and the maximum water level for the vertical clearance and the open passage for barrier Ravenswaaij are given in Figure 5-9. No adaptations works have to be implemented when the trapped line remains in between the purple triangles and the red crosses. The water levels at km 48 remains in between the upper and lower boundary according to Figure 5-9, so no adaptations have to be implemented for the
28 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
Amsterdam-Rijnkanaal. The trapped line undercuts the red crosses at km 68 and km 70; however this section remains undammed and corresponds to the present situation. The downstream water level is chosen equal to the conventional discharge (OLR). These locks are not influenced by the new canalization, and their improvement costs are not included in the comparison of variants.
5.4.4 EVALUATION OF VARIANTS (MCA)
The costs of the seven remaining variants presented in Figure 5-6 are determined in appendix J.5 using the methodology presented for the chosen configuration in section 5.4.3 and key ratios. Furthermore the rating of the performances per variant is determined. A scale from 1 to 5 for the design criteria presented in appendix J.5.1 is used and a weight reference is introduced. The score is obtained by dividing the (weighted) performances by the costs. The result for the seven remaining variants for the weighted evaluation is presented in Figure 5-10. An unweighted assessment is also included in appendix J.5.
Figure 5-10 Cost/performance ratio including weighting for the Nederrijn-Lek configuration
Three best variants are abstracted from Figure 5-10 using a method which is described in appendix B.2. The three best variants are: variant “2w;Driel&Culemborg;com.ship&recr” variant “3w;Driel&Amerongen&Hagestein;com.ship” variant “2w;Driel&Culemborg;recr.”
5.5 CONCLUSION
Variant “2w;Driel&Culemborg;recr” is considered as less preferable with respect to the other two variants, because the (social) impact of this variant is significantly larger with respect to the other two variants. The impact of this variant is significantly larger because the harbour of Wageningen has to be closed and the harbour of Arnhem is cut off from the main supply route from the west, which is unfavourable and could result in major social disturbances. Furthermore, the parallel route is fully blocked for commercial shipping, so a spare route for commercial shipping in case of hindrance of the Waal is not present anymore (Beaufort, 2012). Variant “3w;Driel&Amerongen&Hagestein;com.ship” has the best performance/costs ratio and the highest performance score. The impact on the surroundings is limited; no water level changes takes place for the dammed situation. Furthermore, a full passage for commercial shipping is present so vessels are not hindered for this variant. This variant has higher costs with respect to variant “2w;Driel&Culemborg;recr” and variant “2w;Driel&Culemborg;com.ship&recr” because one weir extra has to be build.
: ARCADIS & TUDelft 29
New canalization of the Nederrijn and Lek Main report
Variant “2w;Driel&Culemborg;com.ship&recr.” is considered as an alternative for the new canalization of the Nederrijn and Lek. A parallel route for the Waal is still present, and commercial shipping could use the Nederrijn in order to reach the harbours of Wageningen and Arnhem. The reach in between the connection of the Lekkanaal and Amsterdam-Rijnkanaal is closed off for commercial shipping but this has minor impact on commercial shipping because this reach is of minor importance considering the low amount of tonnages transported over this reach and the limited (harbour) connections (Noordhoff Atlasproducties, 2011).
5.5.1 DESIGN CHOICE
A variant composed of 3 weirs which are located near Driel, Amerongen, and Hagestein in combination with sufficient draught for commercial shipping is less interesting in the context of the graduation research because it is a rehearsal of the design of the existing structures. The same requirements could be used for the design and the same locations could be used for the implementation of the weirs. It is more interesting to elaborate the weir characteristics and design for a new location with different requirements in the context of this graduation research. Furthermore the thought of the separation of commercial shipping and recreational boating in order to reduce the maintenance costs is not that peculiar. This idea was already been investigated by Rijkswaterstaat (Havinga, 2012). Therefore, a variant composed of 2 weirs which are located near Driel and Culemborg with sufficient draught for recreation in between the Lekkanaal connection and the Amsterdam Rijnkanaal connection is chosen for further elaboration.
5.6 HYDRAULIC MODEL OF THE NEDERRIJN
A check is performed using a hydraulic model to investigate whether weir Culemborg is able to regulate the water levels of the Lek and Nederrijn. A suitable dam regime for weir Culemborg is obtained when the water levels of the maximum dam regime are higher than the minimum dam regime for every discharge. Furthermore, a sequential calculation is performed to determine whether a weir located near Culemborg is able to regulate the upstream water levels using an underflow or overflow gate. The results for an overflow or underflow gates are used as a preview for the sequential design levels. The hydraulic model is based on the following assumptions (Voortman, 2011): The Nederrijn and Lek are schematised as prismatic channels. The slope of the Nederrijn and Lek is constant till the transition to the horizontal part near Wageningen. The flow is stationary; water levels and discharges do not change over time. The downstream water level of weir Culemborg is equal to: o An average water level of +0.55m NAP for low discharges. o Equal to the equilibrium water level for higher discharges. The dam regime is calculated for three limitations which are present at the upstream reach of weir Culemborg. The three limitations are: A water level at barrier Ravenswaaij of +5.0m NAP which may not be exceeded for dammed situations. This limitation is called the ‘Ravenswaaij limitation.’ A minimum depth of 4.2 metres which must be present in the river for navigation. o The transition from an inclined river bed and a horizontal river bed near the city of Wageningen results in a minimum depth which is presented in Figure 5-11. This limitation is called the ‘Wageningen limitation.’ Equal water levels at weir Driel with respect to the present dam regime which must be realised to maintain the same dam regime for the new canalization. The model description, the calculations and the results of the hydraulic model are included in appendix O. The model indicates that weir Culemborg is able to regulate the water levels in between weir Culemborg
30 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
and Driel for the described limitations for an over- and underflow gate. The resulting dam regimes for weir Culemborg are presented in Figure 5-12. A suitable dam regime is obtained for every water level located in between the red line representing the maximum dam regime and the blue line representing the minimum dam regime. The green line represents the downstream water levels and the pink line represents a minimum water level for navigation. The weirs are opened for a discharge higher than 568m3/s for the minimum dam regime. The upstream and downstream water levels at weir Culemborg are equal for this discharge. The Nederrijn and Lek are dammed for 310 days per year on average for the new canalization
Intermezzo A validation and verification check results only in a ‘yes or no’ evaluation according to the SE handbook. However, the performed check using the hydraulic model results in a minimum and maximum dam regime of weir Culemborg and not in a yes or no. So, only a bandwidth is defined for which weir Culemborg is able to regulate the desired water levels and not exact upstream water levels of weir Culemborg. Defining a bandwidth is not in accordance with the SE methodology, so the SE methodology has to be adjusted for this aspect.
Secondly, calculations are made to investigate whether weir Culemborg is able to regulate the water levels using an underflow or overflow gate. These calculations are not strictly part of this design level but are a preview for the sequential design levels considering the weir dimensions and the gate design. This preview prevents surprises from happening in sequential design levels and investigates the feasibility of an underflow or overflow gate for weir Culemborg. A new configuration had to be chosen when the conclusion of this preview was: “weir Culemborg is not able to control the water levels with an underflow or overflow gate at this location for this reach,” The performed preview is not prescribed by the Systems Engineering handbook, so the Systems Engineering methodology has to be adjusted for this aspect.
: ARCADIS & TUDelft 31
New canalization of the Nederrijn and Lek Main report
Barrier Ravenswaaij
Wageningen limitation
Figure 5-11 Water levels upstream from Culemborg for the minimum dam regime
Suitable dam regimes
Figure 5-12 Minimum and maximum dam regime for weir Culemborg
32 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
6 Design level 3 - weir location
Chapter 5 concludes with a design choice for a two weir configuration which is elaborated in further detail in this chapter. The upstream weir has to be placed near weir Driel, or weir Driel can be renovated or renewed. 3 The downstream weir has to be placed near the village of Culemborg. This weir should be located downstream from the Amsterdam Rijnkanaal and upstream from weir Hagestein which is presented in the drawing of design level 3 presented in the project map (Figure 2). From now on, the design of the downstream weir is the emphasis of the graduation researched. Designing two weirs would take too much time within the set time limit for the graduation research. The exact location, the width and length of the approach channels, and the minimum and maximum width of the weir needs to be determined in one design level because these aspects influence each other. The exact location depends of the length and shape of the approach channels and the total width of the weir. The same cadence as for design level 2 is used. The design process starts with a description of the project area in section 6.1, subsequently the main results of the requirements analysis is given in section 6.2. Thereafter, the functional analysis is described in section 6.3. The resulting design is described in section 6.4. In the end, a conclusion is made in section 6.5.
6.1 PROJECT AREA
The most important objects and characteristics of the project area are described in this section for design level 3. The full project area analysis is given in appendix C.3. A short overview of the findings of this appendix is given below. The area of interest is presented in Figure 6-1. The distances are indicated with blue crossing lines at every 2 kilometres form the crossing of the Amsterdam-Rijnkanaal till weir Hagestein. The Lek starts at the village of Wijk bij Duurstede which is located at the crossing of the Amsterdam-Rijnkanaal. This starting point is set as reference (0 km). The Lek flows downstream through the area of interest and leaves the area at weir Hagestein. Several villages are located along this reach which are: Ravenswaaij, Beusichem, Culemborg, Everdingen, and Hagestein. The floodplains are mainly agricultural lands and natural areas which are no part of Natura2000. Also a sand pit is located upstream from Culemborg. The bottom level of this sandpit is -25m NAP (Reverda, 2012). The Southern floodplains near Beusichem are used for recreational purposes. A campsite, recreational harbour, a ferry connection, and several buildings are located here. A height restriction is present at the northern floodplains and at the river near the village of Culemborg imposed by a railway bridge and a ramp with many small spans and pylons. The river section is a dammed river for discharges lower than 568 m3/s and a free flowing river for higher discharges. The normalised width of the river varies around 140 metres within the project area.
3 The renovation or new weir design of weir Driel is beyond the scope of the graduation research.
: ARCADIS & TUDelft 33
New canalization of the Nederrijn and Lek Main report
Figure 6-1 Project area characteristics for the selection of the weir location (Google Maps)
6.2 REQUIREMENTS ANALYSIS
Only the system boundaries and the characteristics of the system boundaries are given for the requirements in the main report in order to keep the main report brief. The full requirements analysis with the 15 steps for requirements analysis is presented in appendix E.3. The most important requirements which are impacting the design the most are presented in this section. A full list of requirements applicable for design level 3 is described in appendix F.3.
6.2.1 SYSTEM BOUNDARIES
The system to be analysed for design level 3 spans from Wijk bij Duurstede till weir Hagestein as presented in Figure 2 and Figure 6-2. The following components are part of the system: the reach from Wijk bij Duurstede till weir Hagestein the levees and floodplains along this reach the harbour connections the connections of pumping stations and inlets.
An upstream boundary at Wijk bij Duurstede and a downstream boundary at weir Hagestein are presented. The upstream boundary conditions contain: Water levels caused by runoff for open river conditions. Navigation entering or leaving the system. Discharge entering the Lek. Minimum water levels at the constriction at Wageningen. Minimum water levels at weir Driel. Maximum water levels at Ravenswaaij. The downstream boundary conditions contain: Water levels caused by a combination of tide and runoff. Water transport capacity. Navigation entering or leaving the system.
The boundaries located along the system are: Winter levees aligned along the river. Objects like harbours, campsites etc. located along the river.
34 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
Figure 6-2 System boundaries (indicated by the red lines) for the selection of the weir location (Google Maps)
6.2.2 MAIN REQUIREMENTS
The main requirements which are impacting the design solution of the layout of the weir the most are presented in the following enumeration. The full list of requirements is included in appendix F.3. FR 10) Vessel class Va must be able to sail through the system for open conditions. FR 10.1) A width equal or larger with respect to the Lek must be present for the object. FR 11) Recreational class AM must be able to sail through the waterway. FR 11.1) A width equal or larger with respect to the Lek must be present. FR 12) Flow velocities must be limited for navigation. FR 12.1) Flow velocities must be lower than 2 m/s averaged over the cross section. IR 8) The structure should be situated in such a way that a straight line of sight with the navigational openings is present. IR 9) The structure should be situated in a position that minimizes cross currents. ER 6) The capacity of the system must be equal to the capacity of the Nederrijn. ER 6.1) The discharge capacity must be equal to the discharge capacity of the Nederrijn. ER 6.2) The sediment transportation capacity must be equal to the capacity of the Nederrijn.
6.3 FUNCTIONAL ANALYSIS
The new objects which have to be placed in the reach spanning from the Amsterdam Rijnkanaal till weir Hagestein is described as the system ‘The water retainer.’ The functional tree of ‘The water retainer’ is the basis for the object tree to which the requirements are allocated. The objects of the object tree are derived from the functions using a functional flow block diagram. The FFBD of design level 3 is included in appendix G.3.
6.3.1 FUNCTIONS
The ‘the water retainer’ is decomposed in four sub functions as presented in Figure 6-3. The four top level functions of the system are: Transportation of the water. The first sub function is the transportation of water from the upstream boundary towards the ‘water retention object’ and the second function is the transportation of water from the object towards the downstream boundary. Transportation of sediments. The first sub function is the transportation of sediments from the upstream boundary towards the ‘water retention object’ and the second sub-function is the transportation of sediments from the object towards the downstream boundary.
: ARCADIS & TUDelft 35
New canalization of the Nederrijn and Lek Main report
Retaining water. Water has to be retained for providing the second level sub-function which is the regulation of discharge and of the upstream water level. Water has to be retained at the upstream side of the system for preventing the river from running dry and realising sufficient draught for navigation. Facilitating navigation. Navigation has to navigate through the system. In the first sub-function, a vessel is using the system by sailing upstream or downstream. Enough capacity, clearance and width have to be available for the navigation using the system. Furthermore, good nautical conditions must be provided like sufficient width and sight.
Figure 6-3 Functional tree for the weir location
6.3.2 OBJECTS
The system ’The water retainer’ is split up in two sub-systems which is based on the FFBD. The system is composed of a transportation object and an object which retains and regulates the water level. The derived system objects are presented in Figure 6-4.
Figure 6-4 Object tree of ‘The water retainer’ system
The first sub-object (waterway connections) fulfils the water transportation function, sediment transportation fiction, and the navigational function within the system. The second sub-object fulfils the water retention function. This object retains water, regulates the upstream water level, and regulates the rate of discharge. Furthermore sufficient space needs to be reserved within the retention object for the passage of vessels
36 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
6.4 DESIGN SYNTHESIS
The design synthesis of this design level covers the choice of the location. The choice of weir location is influenced by the available space. Therefore, the dimensions of the approach channels and the weir are first described in 6.4.1 and 6.4.2 thereafter, variants are described in section 6.4.3
6.4.1 APPROACH CHANNELS
Enough space needs to be available for the approach channels and for the weir complex. A general approach channel length is not available; the design of the approach channel needs to be verified by simulations before implementation. The PIANC guidelines state that an approach channel needs to be aligned for a straight line of site with the navigational openings of the structure. The length of the approach channels of weir Driel, Amerongen, and Hagestein are in the order of 1000 metres which is also applied for the length of the approach channels at each side of weir Culemborg. Furthermore, the channels must have sufficient capacity and space for vessels and discharge. Therefore, the channels are chosen 140 metres wide which is equal to the width of the Lek near the village of Culemborg.
6.4.2 WEIR
The total width of the weir openings in must be smaller than the width of the river because the weir must be a constriction of a river (PIANC, 2006). Undesired sedimentation would take place at the weir when the total width of the openings of the weir is larger than the width of the river. Therefore, the width should be smaller than 140 metres. The minimum width of the weir openings in total is determined using a hydraulic model which is described in appendix O.2. The flow velocities and set up for a ranging width till a discharge of 1500m3/s are calculated. Discharges higher than 1500m3/s are exceeding the summer levees and more area is available for the flow. Therefore, it is assumed that the maximum rate of flow for the main summer river bed is reached at 1500m3/s. The total width of the opening of the weir has to be larger than 100 metres in order to limit the flow speed to 2m/s which results from the hydraulic model. So, the width of the weir needs to be larger than 100 metres and smaller than 140 metres.
6.4.3 FILTERING OF POTENTIAL LOCATIONS
The yellow line presented in Figure 6-5 indicates the weir location resulting from design level 2. A location nearby this yellow line needs to be found for the implementation of the downstream weir complex. Seven potential locations are also indicated in Figure 6-5. The filtering of the seven locations is given in Table 6-1. A more detailed description of the weir locations is given in appendix M.
Figure 6-5 Potential weir locations in between the Amsterdam Rijnkanaal connection and weir Hagestein (Google Maps)
: ARCADIS & TUDelft 37
New canalization of the Nederrijn and Lek Main report
Table 6-1 Assessment for the filtering of the locations
Location Description Assessment The upstream approach channel is located at the Amsterdam Rijnkanaal. Changing the crossing of the Amsterdam Rijnkanaal The weir complex is situated results in a changed nautical situation. Furthermore this 1 just downstream from the location is located 8km from the design location which results Amsterdam Rijnkanaal. in extra dredging works. Therefore, this location is not an option. A new river branch is The reach to be excavated is larger with respect to location 4 excavated at the floodplains and 6, so more river works have to be implemented which of Beusichem. The weir is 2 results in higher costs. Furthermore, the recreational area and located in between the the buildings have to be replaced. Therefore this location is not recreational area and the an option. winter levees. The weir complex is A straight line of sight is available for a new weir complex at constructed at the present the present river. Furthermore, a local constriction is available 3 river upstream from the at this location which is advantageous. Therefore this location design location. is a feasible location. The weir complex is A straight channel with a straight line of sight could be constructed at the floodplains excavated in the floodplains of location 4. Sufficient space is upstream from the design 4 available at the floodplains for construction works. The soil location. The floodplains are works of this reach are smaller than the soil works for location mainly used for agricultural 2. Therefore this location is a feasible location. purposes. The sand pit is too deep for constructing a new weir complex. The weir complex is The filling costs of the sandpit outweigh het excavation costs of 5 constructed at the sand pit of a new channel. Furthermore, the sand pit is located in a bed Culemborg. which is disadvantageous for the straight line of sight for navigation. Therefore, this location is not an option. The weir complex is A new reach runs underneath the railway bridge and ramp. constructed at the floodplains Navigation is not possible without adaptations of the railway 6 near the village of bridge due to the restricted height. Therefore, this location is Culemborg. not an option. The weir complex is The weir would be located in a bend for this location. This constructed at the river bend 7 bend and the railway bridge hamper the straight line of sight near the village of for shippers. Therefore, this location is not an option. Culemborg.
6.4.4 PHASING OF THE CONSTRUCTION
Not enough space is available at the floodplains of location 4 for realising the weir and the lock complex with separate approach channels Therefore, location 3 and 4 are both used for the new weir complex. Weir Culemborg has to be constructed ‘in the dry’ at the floodplain first. After realisation of the weir, the river is diverted towards the weir and the lock and hydro plant (if it is profitable) could be constructed in the old river bed. In this way the waterway is not (fully) blocked. The map of the weir location is presented in Figure 6-6. The weir channel presented in Figure 6-6 is drawn at the floodplains and the constriction for the lock complex is drawn in the present waterway. Weir Amerongen and Hagestein remain operative during construction of the weir and lock. Therefore a water level of +3,0m NAP is maintained at weir Culemborg during construction. The bed level at weir Culemborg is located at -2,2m NAP, so a depth of 5,2 metres is present during construction which is sufficient for navigation. Weir Amerongen and Hagestein are out of service when the weir and lock are finished and the new dam regime is implemented.
38 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
Figure 6-6 Layout of the new weir complex at the floodplains near Culemborg (Noordhoff Atlasproducties, 2011)
6.5 CONCLUSION
The ‘weir Culemborg’ is located at the northern floodplains as presented in Figure 6-6 and Figure 6-7. The approach channels are about 1 kilometre long and provide a straight line of site with the navigational openings of the weir. The width of the weir must be larger than 100 metres for limiting the flow speed and smaller than 140 for counteracting sedimentation at the sill. After the realisation of the weir complex, the old river branch can be dammed for the construction of the lock complex. The design of the lock complex is beyond the scope of the graduation research. The weir design is designed in further detail in the next paragraph.
Figure 6-7 Impression of the floodplains of the chosen location of the weir (photo of the author)
: ARCADIS & TUDelft 39
New canalization of the Nederrijn and Lek Main report
7 Design level 4 - weir design
The result of design level 4 is the design of the weir structure. The number of openings, gate type, and the need for an accurate discharge control are the key elements of this chapter. The same cadence as for the level 3 design is used. The design process starts with a description of the project area in section 7.1, subsequently the main results of the requirements analysis is given in section 7.2. Thereafter, the functional analysis is described in section 7.3. The resulting design is described in section 7.4. In the end, a conclusion is made in section 7.5.
7.1 PROJECT AREA
The description of the project area is given in this section. A more extensive project area analysis is given in appendix C.4 and has more clarifying figures. The main points of the area analysis are described below. The weir complex is planned at the northern floodplains as indicated in Figure 6-6. The floodplains are mainly used for agricultural purposes. A small natural area is indicated in Figure 6-6 but details of this natural area are not found. The northern floodplains are part of an ‘island’ which is circumscribed by the Lekkanaal, Amsterdam Rijnkanaal, and Lek. The project area is accessible by 6 bridges with a maximum width of 2 lanes. Furthermore, the floodplains are accessible by the Nederrijn and Lek. The Nederrijn and Lek remain dammed by the present weirs during construction, so the construction site is available by navigation. Furthermore, the sill is located at -2.2m NAP and a sand layer starts at -7.5m NAP, so the soil characteristics for the weir are rather good for foundation works.
7.2 REQUIREMENTS ANALYSIS
Only the system boundaries and the characteristics of the system boundaries are given for the requirements in the main report in order to keep the main report brief. The most important requirements which are impacting the design the most are presented in this section. A full list of requirements applicable for design level four is described in appendix F.4. Traditionally, weir gates were constructed using wood for small spans and steel for large spans. Presently, a new material could also be used for the design of a weir gate, which is fibre reinforced polymers (FRP). The application of this material for the weir gate is chosen as a self-imposed requirement and is included in the main requirements listed in the report.
7.2.1 SYSTEM BOUNDARIES
The system to be analysed in design level 4 spans from the approach channel at the upstream side of the weir till the approach channel at the downstream side of the weir as presented in Figure 7-1. The following components are part of the weir system: the upstream approach channel the FRP gates the weir foundation the weir superstructure
40 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
the weir inflow structure the downstream approach channel. the weir outflow structure
Boundaries are presented in Figure 7-1. An upstream boundary is present at the connection of the Lek to the upstream approach channel. A downstream boundary is present at the connection of the downstream approach channel with the Lek. The upstream boundary conditions contain: Water levels caused by runoff for open river conditions Navigation entering or leaving the system Discharge entering the approach channel from the Lek The aimed upstream water levels. The downstream boundary contains: Water levels caused by a combination of tide and runoff Water transport capacity Navigation entering or leaving the system. The boundaries located along the system are: the embankment of the channel maximum design water level.
Figure 7-1 System boundaries for the weir design
7.2.2 MAIN REQUIREMENTS
A full list of requirements which are abstracted from the full requirements analysis are listed in appendix F.4. Only main requirements which are impacting the design the most are presented in this section. The main requirements used for the design are: FR 14) Vessel class Va must be able to pass the weir for open configurations FR 14.1) A width of 29 metres must be present for one lane traffic per opening. FR 14.2) A width of 41 metres must be present for two lane traffic per opening. FR 14.3) A vertical clearance of 9.2 metres must be present. FR 14.4) A depth of 4.2 metres must be present. FR 15) Recreational class AM bust be able to pass the weir for open configurations. FR 15.1) A width of 25 metres must be present. FR 15.2) A height of 3.75 metres must be available. FR 15.3) A depth of 1.80 metres must be available. FR 16) A water level difference of 5.5 metres must be retained. FR 17) The structure must remain stable for all conditions. FR 18) The structure must be protected for bed degradation. FR 21) A minimum discharge of 25m3/s must be regulated during fully dammed conditions. FR 21.1) A watertight closure of the gates with the foundations has to be realised for a fully dammed operation. AR 11) The gates have to be designed using fibre reinforced polymers (FRP). IR 12) The gates have to be adjustable from +5.00m NAP to -2.2m NAP.
7.3 FUNCTIONAL ANALYSIS
The object to be designed for design level 4 is the ‘water retention object.’ This object can be subdivided into five sub-objects which are the weir complex, lock complex, fish passage, river banks and floodplains.
: ARCADIS & TUDelft 41
New canalization of the Nederrijn and Lek Main report
The focus for design level 4 within the graduation research is aimed at the weir and not at the other elements. Therefore the other elements are beyond the scope of the graduation research. The functions of the weir are: Regulation of upstream water levels o Retaining water Regulation of discharge o Providing a minimum discharge and higher discharges Providing a passage for navigation o Providing enough space for the navigational openings.
An object tree of the weir structure is given by PIANC and presented in Figure 7-2. The allocated requirements for the sub objects are not given in the main report but are included in appendix G.4.
Figure 7-2 Object tree for the weir structure
7.4 DESIGN SYNTHESIS
The design of the weir is performed using ‘design aspects’ which are representing the width of the openings in combination with the number of openings, the gate type, and the accurate discharge control. These design aspects are given in section 7.4.1. Variants are developed using these design aspects for which the evaluation is given in section 7.4.2.
7.4.1 DESIGN ASPECTS
Three categories of design aspects are available for the weir design, namely the width of the openings in combination with the number of openings, the gate type, and the accurate discharge control. The total width of the opening should be larger than 100 metres and smaller than 140 metres which results from design level 3. The total width is subdivided in several openings. The width of a single opening must be at least 29 metres for one lane commercial shipping and 41 metres for two lane commercial shipping. These widths are used as design aspects for the category ‘opening width and number of openings.’ Several combinations of opening widths and number of openings are presented in a morphological map included in appendix N.3. Various gate types are available for the control of flow. An overview of weir gates is given in appendix N.2. The gate types are subdivided in underflow and overflow weirs. The applicability of both gate types for weir Culemborg for a width ranging from 100 metres till 140 metres is investigated in appendix O using the dam regime for the upstream and downstream water levels. Both types of gates were not able to regulate the upstream water levels for the minimum dam regime. The theoretical gate position had to be lower than the bottom for an overflow weir and higher than the water level for a underflow weir for discharges just lower than the maximum discharge of 568m3/s. This was caused by a nearly zero head just before opening of the gates. A higher upstream water level than the minimum dam regime has to be maintained in order to generate a head over the weir just before the maximum dammed discharge. This head is still present when the gates are opened and decreases gradually during the lifting process. On
42 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
basis of appendix O.3 and O.4 can be concluded that an underflow weir and an overflow weir are able to regulate the upstream water levels for the adjusted minimum dam regime. The last ‘building block’ is the accurate discharge control. The sensitivity of discharge for an underflow and overflow gate is determined in appendix 0.5. The distance for which an overflow gate has to be lowered in order to realise a minimum discharge is much larger with respect to the distance for which an underflow gate has to be lifted. Therefore it is better to use an overflow gate for accurate discharge control than an underflow gate. Furthermore an underflow gate cannot be used for the minimum discharge control because the gap size is too small which results in damage caused by vibrations. Therefore, a weir complex equipped with underflow gates has to be equipped with an extra accurate discharge control which has to pass the minimum discharge and which has to fine tune the rate of discharge of the main gates. A filtering of the morphological map presented in appendix N.3 is performed, which results in the morphological map presented in Table 7-1.
Table 7-1 Filtered morphological map for weir Culemborg
Design aspect Options One large opening Number of openings and 4x 29m width 3x 41m Submerged segment gate Vertical lifting gate Gate type Flap gate (mechanically operated) Visor Gate Using overflow gates Accurate discharge control An separate accurate discharge control is needed when using underflow gates
The only suitable gate type for a ‘1 opening weir’ is the flap gate because this gate can be divided into several smaller gates operated with a piston from the foundation which bears the forces per running metre. The span of the other gates become large and become more vulnerable for vibration due to the flow with respect to smaller spans. Therefore, a gate with one large span for a weir is not preferred. The resulting variants are presented in Table 7-2. The building block ‘accurate discharge control’ is not included in Table 7-2 because the choice for accurate discharge control depends on the gate type. Sketches of the weir under consideration are presented in Figure 7-3. The performance and the costs of these weirs are determined in order to make a decision.
Figure 7-3 Selected gates (From left to right: vertical lifting gates, submerged segment gate, flap gate, visor gate)
Table 7-2 Weir variants
Submerged segment gate Vertical lifting gate Visor gate Flap gate 1 opening - - - x 3x 41m x x x x 4x 29m x x x x
: ARCADIS & TUDelft 43
New canalization of the Nederrijn and Lek Main report
7.4.2 EVALUATION OF VARIANTS
The evaluation of the nine variants is performed using a Multi Criteria Analysis (MCA). The costs of the variants are estimated using reference projects which are described in appendix N.4.3. The performance of the variants is based on a weighted and unweighted score determined using a MCA for the design criteria resulting from the manual of PIANC. The three best variants are: A flap gate with 3 openings of 41 metres. o This is the cheapest solution but has also the lowest score of the selected variants. A visor gate with 3 openings of 41 metres o The costs of this variant are between the flap gate and the submerged segment gate and the performance score is also in between the performance of the flap gate and submerged segment gate A submerged segment gate with 3 openings of 41 metres. o This variant has the highest performance, but also the highest costs.
7.5 CONCLUSION
The submerged segment gate is chosen for further elaboration despite the high (implementation) costs. It is more interesting for further elaboration within the graduation research with respect to a visor gate which is already installed and has a higher performance with respect to the flap gate.
A preview of the submerged segment gate which is elaborated in further detail in chapter 8 is given in Figure 7-4. The gate orientation is drawn for an open river and for a fully dammed river. The height of the gate is 8 meters which is 0.8 metres more than the head difference for guarantying a water tight closure.
Figure 7-4 Cross section of the submerged segment gate which is further elaborated in design level 5
44 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
8 Design level 5 - gate design
At last the submerged segment gate which is presented in Figure 7-4 is elaborated. A submerged segment gate consists of rotation disks and a beam which retains water. The focus of the design of the submerged segment gate is the elaboration of the beam which is located in between the rotation disks. The rotation disks are not elaborated within the graduation but only mentioned. The loads which the gate has to bear are described in section 8.1, a short introduction to fibre reinforced polymers (FRP) is given in section 8.2, and the design is elaborated in section 8.3. The top requirements for the elaboration of the gate originates from the requirements set of design level 4. No requirements are elaborated using the SE method because the set of requirements expands exponentially at this design level so; the SE method is not suitable anymore. Therefore, the design is based on the allocated gate requirements of design level 4.
8.1 LOADS
The main loads which are acting at the gate are the upstream and downstream hydrostatic pressures and the wave pressures which are presented for a general gate cross section in Figure 7-4. The load definitions, the calculation of the wave height, and the calculation of the forces are elaborated in appendix P and appendix Q. The separate distributed forces at the gate generated by the upstream water level, downstream water level and wave pressures are combined in one resultant force acting as a line load at the middle of the gate. The shift of the three separate resultant forces towards the middle of the gate generates a resultant moment which causes torsion. The torsional moment is neglected for the first calculations and the gate is designed for the maximum bending moment and maximum shear force generated by the line load. Later on, it is verified whether the gate is able to bear the torsional moment. The load is calculated for the service ability limit state (SLS) and the ultimate limit state (ULS) for a gate height of 8 metres. The resulting SLS load is 292kN/m and the ULS load is 376kN/m.
Figure 8-1 Load definitions
: ARCADIS & TUDelft 45
New canalization of the Nederrijn and Lek Main report
8.2 FIBRE REINFORCED POLYMERS (FRP) GATES/BEAMS
FRP elements are produced by combining fibres and resins. The fibres are necessary for the strength and stiffness and the resin is used for supporting the fibres, protecting the fibres from the outside environment, and to transfer the stresses from a fibre to another fibre. Codes and design manuals are only made for glass fibre reinforced composites in civil engineering. Therefore, glass fibre reinforced polymers are used for the design of the gate. A resin which cures at room temperature is used to avoid the application of a curing oven. An oven for the production of a weir gate needs to be large which results in high production costs. Therefore, a design choice is made for unsaturated polyester resin which cures at room temperature which has good strength characteristics for low costs. More information of the reinforcement, resins, and production techniques is given in appendix B.4.
8.2.1 FRP STIFFNESS PROPERTIES
The stiffness of a FRP element is determined by the characteristics of the reinforcement, the orientation of the fibres, and the volume fraction of fibres. The fibres can be spread evenly over the two main directions (0° and 90°) or could be divided unequally resulting in an orthotropic element. FRP lamellas are produced by injecting resin into a fibre mat or woven fabric. The stiffness properties of the FRP lamellas are determined by the fibre content, fibre alignment, and resin content. A laminate is produced by ‘stacking’ multiple layers of reinforcement and injecting the multiple layers with resin. The stiffness properties of the laminate are calculated with the ‘classical laminate theory.’ Applying the classical laminate theory is quite complicated and would take too much time within the graduation research. Therefore, laminates proposed by the CUR96 and FiberCore are used. An isotropic laminate is used for elements which bears the shear force because the shear force capacity of an isotropic laminate is higher than an anisotropic laminate. An anisotropic laminate is used for a tensile or compressive force. The direction with the highest fibre content is aligned with the direction with the largest tensile or compressive force. The stiffness properties of the laminates proposed by CUR96 and FiberCore are presented in Table 8-1. The maximum stress of a laminate is determined by the maximum strain ‘ε’ which is 1.2% for dry environments and 0.27% for wet or humid environments according to CUR96. The maximum angular rotation ‘γ’ for dry environments is 2.4% and for wet environments 0.54%. The representative stresses are calculated by applying hooks law for normal strain and angular rotation which are presented in appendix B.4.7.2.
Table 8-1 Laminate stiffness properties (stichting CUR, 2003) & (Snijder, 2012)
Stiffness Quasi isotropic laminate Anisotropic laminate Anisotropic laminate Unit property (CUR96) (CUR96) (FiberCore)
E1 GPa 18,6 25,8 31
E2 GPa 18,6 15,9 15,9
G12 GPa 7 5,6 5,6
8.2.2 DEFLECTION OF FRP BEAMS
The shear modulus of FRP elements is 11 times lower with respect to the shear modulus of steel. Therefore, shear deformations of a FRP structure are not negligible (stichting CUR, 2003). The main assumption for an Euler Bernoulli beam (EB-beam) is the Bernoulli-Navier hypothesis (plane cross-sections remain planar and normal to the beam) which is relaxed for the Timoshenko-beam (T-beam) which includes shear deformation (Simone, 2011). Myosotis equations are derived for calculating the deflections caused by a point or line load. Standard myosotis equations are available for EB-beams, but not for T-beams with significant shear deformation. Therefore, extended myosotis equations are derived for a simply supported beam, a pin supported-fixed beam, and a fixed-fixed beam of which the results are presented in appendix
46 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
T. The extended myosotis equation of a simply supported T-beam is presented in Equation 8-1 as reference. The first term of the right hand side represents the bending deformation and the second term represents the shear deformation. The contribution of the shear deformation ranges from 6% for a simply supported beam till 30% for a fixed-fixed beam for a span of 41 metres.
Equation 8-1 Extended myosotis equation for a Timoshenko beam
8.2.3 ESTIMATION OF GATE WEIGHT AND GATE COSTS
The gate cross section is modelled as two distinct plates for which the plates bears the moment which is generated by the resultant line force for the estimation of gate weight and costs. The dimensions are estimated for a steel and FRP gate with the laminate proposed by CUR 96. The resulting dimensions are used for the estimation of gate weight and costs which are used for the determination of the feasibility of a FRP gate. The calculations are fully described in appendix R.
8.2.3.1 DESIGN STRESS
The design stresses for laminates are calculated using material and conversion factors and the strain limits which are described in appendix R.1.3 for SLS and ULS. The results are presented in the first and second row of in Table 8-2. The design stress for SLS is very low and question marks have been raised for the application of this design stress. The application of this stress limit and the material and conversion factors is checked using literature and consults with experts. This study showed that no material and conversions factors are used for the design of the already build FRP lock ‘the Spieringsluis’ (stichting CUR, 2003) for SLS, and that the material and conversion factors are already included in the stress reduction from 1.2% to 0.27% according to FiberCore. A too conservative design would be obtained when a material and conversion factor is used for the SLS design of the CUR96. The resulting design stress without conversion and material factors for SLS is presented in the last row of Table 8-2.
Table 8-2 Design stresses of the applied material.
Tensile Design limit Representative Conversion and Design strain Design stress modulus state stress material factor [GPa] [%] [N/mm2] [-] [N/mm2] 25,8 SLS 0,27 70 2,9 24 25,8 ULS 1,20 310 2,9 117 25,8 SLS 0,27 70 - 70
A choice has to be made between the SLS design stress with factors and the SLS design stress without factors. A design choice is made for the SLS stress without factors because this is presently used for the design of hydraulic structures. Further investigation has to be performed for the limit stresses and strains and the application of factors in order to determine the right design limit for FRP structures.
8.2.3.2 COSTS OF A STEEL AND FRP GATE
The dimensions of a FRP gate are determined by the maximum deflection limit for SLS without use of conversion and material factor. No clear deflection limits are available for the design of hydraulic structures. Therefore, a maximum deflection of 1/150 of the span and 1/300 of the span are used for the estimation of weight and costs for the FRP design. The deflections of the steel gate remain below the stated deflection limits, so the steel gate is designed for the strength limit. The mass and the costs of a hydraulic gate for a span of 41 metres are presented in Table 8-3. From this table, it is concluded that FRP gates are
: ARCADIS & TUDelft 47
New canalization of the Nederrijn and Lek Main report
more expensive with respect to steel gates, and FRP gates are lighter for larger deformation limits. FRP gates become more costly and heavier for smaller deformation limits with respect to steel gates. The advantages of FRP gates are lower maintenance costs and a longer lifespan. It has to be investigated whether the FRP gates are cheaper over a life time. However, this is beyond the scope of the research. Furthermore, model tests have to be performed for the design of the FRP gate. The model tests results to €150.000 - €200.000 extra costs which make the FRP gate even more expensive with respect to a steel gate (Kolstein, 2013).
Intermezzo A requirement could dominate the resulting design. In this case, functional requirement FR 21) results in a small deflection limit and an expensive connection for guarantying a water tight closure between the gates and the sill. An alternative design could be made to comply with the criteria: ‘the structure may not be too expensive.’ So, a hydraulic gate could be designed for a larger deflection limit to reduce costs. This alternative has to be discussed with the client and a requirement could be changed or expired. The coupling of the criteria via the design to the requirements and vice versa is not included in the Systems Engineering methodology. Therefore, the Systems Engineering methodology has to be extended with this coupling between the criteria and requirements via the design.
Table 8-3 Gate weight and costs
Material Deflection limit Mass Key ratio Costs [-] [ton] [€/kg] [million €] FRP 1/150 100 12 1,2 FRP 1/300 150 12 1,8 Calculated for the strength limit, deflections are lower Steel 140 6,5 0,9 than 1/150 and 1/300
8.3 GATE DESIGN
The preliminary gate design is described in this section. First the general gate design is described in 8.3.1. Subsequently, a more detailed design is made in section 8.3.2. The design of the outer plates, the shear webs and the panels are described in this section. Design checks for the panels are also performed in 8.3.2. Furthermore, design checks are performed for local instabilities, deflections and strength in 8.3.3. The connections of the FRP gates are described in 8.3.4. The layout of the connections is based on the consults with experts and is not calculated due to the time constraint of the graduation research. At last the dynamic behaviour of a FRP gate is determined and compared with the dynamic behaviour of steel gates in section 8.3.5.
8.3.1 GENERAL GATE DESIGN
The general design of the weir gate covers the gate height, the gate cross section, and the mechanical scheme of the gate. The span of the gate is 41 metres which is already determined in section 7.4.2 and gate thickness needs to be in the range of 3 metres according to appendix T. The gate should be at least 7.2 metres in order to divert the water which is based on Figure 7-4. However an extra height of 0.8 metres is applied for the water tight closure which is necessary for the fully dammed operation. A maximum deflection limit of 100mm is used for the design of submerged segment gates in order to guarantee a water tight closure for fully dammed operation (Sloten, 2012). The gate cross section presented in Figure 7-4 is drawn solid which this is not an efficient design. A closed geometry composed of an upstream and downstream plate as presented in Figure 8-2 is preferable for the guidance of flow and is more efficient as a solid cross sections. Stiffening webs are added to this cross
48 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
section in order to improve the stiffness and to bear the shear forces. The cross section with stiffening webs is also presented in Figure 8-2.
Figure 8-2 General gate cross section (left) and a cross section with stiffening ribs (right)
Four options which are presented in Figure 8-3 are available for the mechanical scheme of the gate, namely: 1. a pin supported connection at both sides of the gate to the rotation disks a. Both disks do have a pin supported connection with the abutments. 2. a pin supported and fixed connections of the gate to the rotation disks a. Both disks do have a pin supported connection with the abutments. 3. a fixed connection at both sides of the gate to the rotation disks a. Both disks do have a pin supported connection with the abutments. 4. a fixed connection at both sides of the gate to the rotation disks. a. One disk is pin supported, the other disk sliding supported.
Figure 8-3 Mechanical schemes of the gate
The first scheme is not an option because this system is statically instable. The second and third schemes are statically determined and undetermined which is better than the statically underdetermined scheme. However, the gate-rotation disk connection has to transfer a moment which is less preferable for a FRP structure. No moments have to be transferred at the connections for the fourth mechanical scheme. Furthermore, scheme fourth is statically determined which has the preference above a statically undetermined structure. Therefore the fourth mechanical scheme is chosen for the design of the gate.
8.3.2 DIMENSIONING OF THE FRP GATE
The gate is designed for the maximum head only due to the approaching deadline of graduation. The force at the gate generates a maximum bending moment at mid span and a maximum shear force at the
: ARCADIS & TUDelft 49
New canalization of the Nederrijn and Lek Main report
bearings. These maximum forces are used for the design of the gate. The outer plates (8.3.2.1) are uncoupled from the shear webs (8.3.2.2) to simplify the calculation. The outer plates bear the moment and the shear webs bear the maximum shear force for this schematisation.
8.3.2.1 OUTER PLATES DESIGN
The outer plates bear the bending moment caused by the distributed load acting at the gate. The upstream plate is under compression and the downstream plate is under tension. The maximum deflection and maximum stresses are calculated for a gate width ranging from 2.5 metres till 3.5 metres and a plate thickness ranging from 0m till 0.2 metres of which the results are presented in appendix U. The plate thickness of the largest width of 3.5 metres needs to be 0.12metres which is impossible to produce; the thickness of FRP plates which are normally produced is 5 centimetres and could be stretched to 7 centimetres (appendix B.4). The resulting wall thicknesses are significant larger, so a massive FRP plate is not an option. Therefore a sandwich panel is used of which an example is given in Figure 8-4.
Figure 8-4 Sandwich panels
Two faces of 5 to 7 centimetres thick each are used as outer faces of the panel in order to obtain sufficient thickness for the bending moment capacity. The gate width of a gate composed of sandwich plates is varied from 3 metres till 3.75 metres and the thickness of the faces are varied from 0.05 metres till 0.07 metres. The resulting deflections are presented in Table 8-4. The red marked deflections are larger with respect to the deflection limit of 0.1 metres. The dark green marked deflections are just below the deflection limit and the light green marked deflections are well below the limit. A smaller gate width lowers the radius of the rotation disks and the dimensions of the recess of the foundations for the submerged segment gate. An example of the recess is given in Figure 7-4. Furthermore a thickness of 0.07 metres is less preferable because the fabrication process of this thickness is still under investigation and the thickness is based on the expected limits. Therefore the thickness of the FRP sandwich plate is chosen to be 0.06 metres in combination with a 3.5 metre wide gate.
Table 8-4 Deflections in metres for a ranging gate dimension
Width of the gate Thickness of FPR faces [m] [m] 0,05 0,06 0,07 3 0,157 0,133 0.120 3,25 0,128 0,109 0,094 3,5 0,107 0,090 0,078 3,75 0,091 0,077 0,066
A disadvantage of sandwich plates applied in hydraulic structures is the possibility of loss of bond between the faces and the core due to repetitive impact caused by ship collision or wave impact. The bond is lost which results in failure of the sandwich element. An improvement which solves this problem is the InfraCore® panel developed by FiberCore. Glassfibre mats are interwoven between the two faces which prevent the loss of bond. Therefore, the InfraCore® panel is applicable for hydraulic structures and is used for the design of the gate. More information for the InfraCore® panel is included in appendix B.4.4 and appendix U.1.2.1.
50 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
8.3.2.2 SHEAR WEB DESIGN
The distinct shear webs located in between the outer faces bear the maximum shear force. A design consisting of 4 shear webs and a design consisting of 5 shear webs are elaborated in appendix U.1 for which the design of 4 webs is chosen. A minimum face thickness of 0.0125 metres is needed for a cross section composed of 4 shear webs and 0.01 metres for 5 shear webs. However secondary forces are also present at the shear webs, therefore a thickness of 2 centimetres is chosen for a cross section composed of 4 shear webs to be conservative. The resulting cross section for the outer plates and the shear webs is presented in Figure 8-5.
Figure 8-5 Gate cross section
8.3.2.3 SHEAR ELEMENTS OF THE INFRACORE® PANEL
Shear elements are located in between the two outer faces of the InfraCore® panel which are presented in Figure 8-6 which bear the local shear forces. The spacing of a shear panel of an InfraCore® panel is 15 centimetres and the thickness of the shear element is in the order of a couple of millimetres. The maximum shear stress for SLS and ULS remains below the limit of 38N/mm2 for a spacing of 15 centimetres and a thickness of 5 millimetres which is used for further elaboration of the gate.
Figure 8-6 Shear elements of the InfraCore panel
: ARCADIS & TUDelft 51
New canalization of the Nederrijn and Lek Main report
8.3.2.4 TORSION
The schematisation of the separate resultant forces described in section 8.1 implied an extra torsional moment which was neglected first. The extra shear force generated by the torsional moment is 0.07N/mm2 for SLS and 0.09N/mm2 for ULS which is determined in appendix U.1.4. The shear stress is significantly lower with respect to the maximum shear stress of 30N/mm2 for SLS, so the negligence of the torsion in the calculations presented in section 8.3.2.1 and 8.3.2.2 is justified.
8.3.2.5 LOCAL DEFLECTIONS
The local deflections of the outer panels are checked for a panel spanning in between the shear webs. The span of the panel spanning in between the shear webs is 1.5 metres as indicated in Figure 8-6. The deflections remain lower than 1 millimetre which is calculated for a two side clamped panel and a four sided clamped plate in appendix U.1.5.3 and U.3.1.
8.3.2.6 COMBINED STRESS CAPACITY OF THE PANELS
The combined stress capacity for the InfraCore® panel is determined for the tensile stress and compressive stress for SLS and ULS. The combined stress capacities check which is executed in appendix U.1 states that the stresses remain within the applicable range. So the designed panel is able to bear the stresses for a maximum head.
8.3.3 DESIGN CHECKS
More detailed design checks for the presented cross section of Figure 8-5 are performed in order to verify the capacity and performance of the designed gate. Eight failure modes for sandwich panels are identified by Professor Zenkert of the University of Stockholm (Zenkert, 1995) which are presented in Figure 8-7. Not every failure mode presented in Figure 8-7 is applicable for the design of the InfraCore® panel because the foam core is replaced by FRP shear elements. Checks are performed in appendix U.2 and in appendix U.3. These checks indicate that mentioned failure modes are not present for the designed cross section.
Figure 8-7 Failure modes of sandwich beams. (a) Face yielding/fracture, (b) core shear failure, (c) and d) face wrinkling, (e) general buckling, (f) shear crimping, (g) face dimpling and (h) local indentation (Zenkert, 1995)
8.3.4 CONNECTIONS FOR LARGE SPAN FRP GATES
The outer panels and the shear webs of the FRP gates have to be connected to each other. The connections could be realised by glue or by bolts. The length of the strips which have to be glued together are large (till a length of 41 meters), so the first part of the glue is already hardened before the glue is applied at the
52 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
other side of the gate, therefore glued connections are not feasible for the FRP gate. The other option is a bolted connection. The bolt is the weakest part of the connection when the surrounding FRP body is 3 times as thick as the diameter of the bolt (Snijder, 2012) which is applied for the FRP gate of weir Culemborg. A close up of a connection executed using bolt is presented in Figure 8-8. The figure from which Figure 8-8 originates is included in the A2 technical drawings and appendix U.4.
Figure 8-8 Bolted connection of a shear web and the outer plates
The connection of the gate with the rotation disk has to be rigid according to the mechanical scheme. Therefore, bolts are applied at the downstream side of the gate and at the end faces of the gate in order to obtain a rigid connection. A figure with the side connections is included in appendix U.4. The given configuration of bolts and the dimensions of the connections are just an impression. Further investigation has to be executed in order to calculate the exact number of bolts.
8.3.5 VIBRATIONS OF A STEEL AND FRP GATE
The eigenfrequencies of the FRP submerged segment are determined in appendix U.5. Also the eigenfrequencies of a steel gate which is designed in appendix R.1.2 are determined in this appendix and compared with the designed FRP gate. A calculation is performed for a 1 degree of freedom system and for a simply supported bending beam which are located in a water body. The surrounding water body vibrates due to the vibrations of the gate. An added mass is taken into account for including the vibrations of the surrounding water body. The eigenfrequency and periods of the one degree of freedom system and a beam for the first vibration mode including the added water mass are presented below: The eigenfrequency and the period of the FRP gate are:
o
o The eigenfrequency and the period of the steel gate are:
o
o The results of the FRP gate and the steel gate are nearly equal. This is caused by the modulus of rigidities of the FRP and the steel gate and the quantity of the added mass: The EI of the FRP gate and the steel gate are nearly the same, so the spring stiffness of both gates is also nearly the same. The total effective masses of the FRP gate and the steel gate are nearly the same which is caused by the large added mass. The vibrations of a FRP gate matches the vibrations of a steel gate and corresponds for both materials with the vibration characteristics of existing hydraulic structures which is verified in appendix U.5.4. However further investigation has to be performed for the vibrations of a FRP gate to determine the impact of the large thickness and the flow characteristics of a submerged segment gate and the behaviour of a Timoshenko beam for vibrations.
8.4 TECHNICAL DRAWINGS OF WEIR CULEMBORG
Drawings of the weir are presented on the sequential A2 drawings.
: ARCADIS & TUDelft 53
New canalization of the Nederrijn and Lek Main report
9 Evaluation of the design methodology
Remarks about the design methodology are made in the intermezzos of the previous chapters. This chapter gives an overview of the remarks and are described in further detail. Furthermore, the SE methodology is compared with the ‘integral design process’ which is described in section 1.3.2.
9.1 APPLICATION OF CRITERIA
Variants are made for an object using the requirements which are developed for the specific design level. The developed variants are ranked using criteria to select the best variants. The criteria are used to assess the developed variants which are designed according to the pre-defined requirements according to the Systems Engineering methodology. So, the criteria are not coupled to the requirements via the design and vice versa for the Systems Engineering methodology. However, one requirement could dominate the resulting design. As example; a deflection limit could dominate the dimension of a hydraulic gate resulting in high costs. An alternative design can be made to comply with the criteria: ’the structure may not be too expensive.’ So, a hydraulic gate could be designed for a larger deflection limit with respect to the requirements to reduce costs. This alternative design has to be discussed with the client and a requirement could be changed or expired (Voortman, 2013). This process, which is indicated in Figure 9-1, is not included in the Systems Engineering methodology.
Figure 9-1 Relation between requirements, design and criteria
9.2 VERIFICATION
The validation and verification prescribed by the SE handbook results in a ‘yes or no’ evaluation which is indicated in Figure 9-2. However, only a ‘yes or no’ evaluation is not always sufficient and more information has to be included in the validation or verification assessment like a bandwidth for which the structure complies with the requirements. The bandwidth is explained by the following example:
54 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
The hydraulic model which is described in section 5.6 defines a bandwidth including a maximum and minimum water level at weir Culemborg as outer limits for which weir Culemborg is able to regulate the upstream water levels. The result of the assessments for the verification check ‘is weir Culemborg able to control the desired water levels’ is not in a ‘yes or no’ assessment but a ‘yes, if… (the water levels remain in between the maximum and minimum water levels).’ Defining a bandwidth is not in accordance with the SE methodology, so the SE methodology has to be adjusted for this aspect.
Figure 9-2 Preview to lower design levels
9.3 PREVIEW TO LOWER LEVELS
The SE methodology prescribes only a validation check for which it is verified whether the design at a detailed level complies with the design made in a previous/higher level. As example: it is verified whether the gate which is designed in design level 5 fits in between the pylons of which the dimensions and spacing are designed in design level 4. This is indicated by the validation arrow of Figure 9-2. The SE methodology does not prescribe a preview to lower levels to investigate the feasibility of the design for the sequential design levels to prevent surprises in the design from happening. Therefore, a preview to a lower level could be made before the design of the lower level is made. Two previews were made in this graduation research which are described in the following enumeration: 1. A preview was made from design level 2 to design level 4 to investigate the application of an overflow or underflow gate for weir Culemborg. It had to be checked whether it was possible to regulate the calculated dam regimes resulting from design level 2 using an overflow or underflow gate. A new configuration of the Nederrijn and Lek had to be made when the conclusion of this preview would be: ‘it is not possible to regulate the water levels with an underflow and overflow weir at this location.’ A ‘shameful’ step backward, which is indicated in Figure 9-3, would have to be made when the preview was not executed in design level 2 and the conclusion for the controllability of the overflow or underflow gate at design level 4 would be: ‘it is not possible to regulate the water levels with an underflow and overflow weir at this location.’ This ‘shameful’ step backward result in extra time and costs for the redesign of the configuration of the Nederrijn and the weir. Therefore, it is advised to implement preview investigations to lower levels in the design process to prevent ‘surprises’ from happening in lower design levels. 2. A second preview was made from design level 4 to design level 5 to investigate the application of a fibre reinforced polymer gate for weir Culemborg. The global dimensions and the costs of the weir gate were determined for a ranging span. In this way, it was investigated for which span the dimensions remained ‘reasonable’ and for which span the costs remained low. The output was used to check whether it was possible to design a fibre reinforced polymer gate for the navigational openings of the weir which would be 29 or 41 metres wide. A ‘shameful’ step backward would have to be made when the result of the preview would be: ‘it is not possible to
: ARCADIS & TUDelft 55
New canalization of the Nederrijn and Lek Main report
design a fibre reinforced gate for a span of 41 metres’ if a 41 metres wide opening was chosen in design level four.
Figure 9-3 Feedback/redesign loop which is present for a ‘dead end’ in the design process of the Eastern Scheldt barrier (de Gijt & van der Toorn, 2011).
9.4 DEFINITION OF DESIGN LEVELS AND SYSTEM BOUNDARIES
Design levels and system boundaries are defined during the project. First, the project was divided in several design levels. This division of the project in several design levels is present for the Systems engineering theory and the ‘integral design process.’ A system boundary is present for every design level. The system boundaries define which part of the system is controllable by the engineer and which part is not. Defining the system boundaries is not an exact science but depends on the conception of the project team or designer. It is an iterative process and the system boundaries could change during this iterative process for defining the system boundaries. Also the system boundaries are part of the Systems Engineering theory and the ‘integral design process.’
9.5 DEVELOPMENT AND APPLICATION OF REQUIREMENTS
The SE methodology requires a requirements analysis, functional analysis, and a design synthesis per design level. This methodology is applicable for the first four design levels covering the conceptual design and has to be leaved for design level five covering the structural design of the gate. Many objects like connections, shear webs, outer plates, etc. are present for this level. Each object has to be specified using the requirements analysis, functional analysis, and design synthesis which results in ambiguous definitions of the functions and requirements of a shear web or connection. Furthermore, no extra requirements are elaborated for design level five because the set of requirements expands exponentially at this level due to many objects like connections etc. resulting in an ambiguous list of requirements. Therefore, the top requirements for the design of the gate are extracted from design level four. These top requirements include the deformation requirements and the load requirements. This transition point between the conceptual design and the structural design is not included in the SE theory and has to be added to the Systems Engineering theory. Therefore, the gate has been designed using these top requirements in an iterative design process corresponding to the ‘integral design process’ for design level five.
56 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
10 Conclusions and recommendations
The problem definition was twofold and therefore the conclusions and recommendations are also twofold. A conclusion and recommendation is made for the recanalization case study and for the developed design methodology separately.
10.1 CONCLUSIONS
10.1.1 DESIGN OF THE RECANALIZATION OF THE NEDERRIJN AND LEK
The recanalization of the Nederrijn and Lek using two weir complexes equipped with fibre reinforced polymer gates is a feasible alternative for the present canalization. The FRP gates are able to bear the water level differences and the resulting moments and shear forces. Furthermore, a submerged segment gate is able to regulate the water levels in between weir Driel and weir Culemborg which is calculated using a 1D hydraulic model. The upstream weir of the new canalization variant is situated near the village or Driel in order to regulate the present dam regime. The present weir located at Driel could be renovated or a new weir could be constructed near the old weir which is not elaborated in this research. The downstream weir has to be situated nearby the village of Culemborg which is located 55 kilometres downstream from the IJsselkop to regulate the water levels of the reach Driel-Culemborg. The upstream floodplain in between the village of Culemborg and Beusichem is chosen after an assessment taking into account the geometry of the weir. The downstream weir, which is named after Culemborg, is composed of 3 openings of 41 metres wide each with a submerged segment gate spanning in between the pylons. An opening of 41 metres is needed for the passage of commercial shipping for a fully open situation. The total width of the openings of the weir is 123 metres which is necessary for counteracting the sedimentation at the sill and reducing the flow velocity for a fully open situation for navigation. The submerged segment gate is composed of fibre reinforced polymers (FRP) which is a self-imposed requirement. The geometry of the gates is determined in order to investigate whether or not a FRP gate is feasible for the application for weirs. The resulting gate is 8 metres high and 3.5 metres thick. The outer plates are composed of sandwiches with faces of 6 centimetres and a core of 15 centimetres thick which bears the moment. 4 Webs are located in between the outer plates which bear the shear forces. The webs and the outer plates are bolted together because the application of glue is less advantageous due to the large dimensions of the gate. The resulting dimensions are within the range of applicability of FRP structures.
: ARCADIS & TUDelft 57
New canalization of the Nederrijn and Lek Main report
10.1.2 DEVELOPED DESIGN METHODOLOGY
A methodology based on SE is developed for large scale hydraulic design projects and is tested for the Nederrijn and Lek recanalization case study. The application of separate design levels with a limited set of requirements each turned out to be applicable for large scale hydraulic design projects for the first four design levels covering the conceptual design. The methodology based on SE had to be extended with a preview and a coupling of the criteria via the design to the requirements for designing the recanalization. Furthermore, it is possible to use the Systems Engineering methodology for large scale decisions which are normally made in the political domain. So, it is possible to position the transition of the substantiation of a design choice before the policy making process. The advantage of the relocated ‘transition’ is a clearer and less ambiguous transition between the substantiation of a design choice in the political domain and for the Environmental Impact Assessment executed by the engineering firms that follows. No extra requirements are defined using the area-, requirements-, and functional analysis for design level five, which covers the structural design of the gate. This methodology was not used for this level because it is ‘not meaningful’ to specify the structural components in ‘functional terms.’ As example: it is not meaningful to define the functions of a shear web or bolt in a functional analysis. Furthermore, many detailed components (like bolted connections, shear webs, etc.) would have to be identified and described resulting in an ‘explosion’ of requirements. Therefore, the allocated top requirements for design level four are used as input for the structural design of design level five. The top requirements have been elaborated during the structural design resulting in the design of the gate of weir Culemborg.
10.2 RECOMMENDATIONS
10.2.1 DESIGN OF THE RECANALIZATION OF THE NEDERRIJN AND LEK
The configuration variant is chosen on basis of the performance/cost ratio. Only the implementation costs and not the maintenance costs are taken into consideration. Therefore, it is recommended to execute a life cycle cost analysis in order to determine the costs for the whole life time. Furthermore, the costs for the configuration variant are established on key-ratios. It is recommended to perform a more detailed research for more exact costs for the design choice for the configuration variants.
A design choice is made for constructing the weir at the floodplains and the lock in the river bed. It is recommended to investigate the possibilities for constructing the weir and the lock in the floodplain in one construction pit and damming the river after completion.
The cross section of the gate is calculated for a fully dammed condition for OLR conditions, which includes an upstream water level of +5.0m NAP and a downstream water level of -0.5m NAP. It is assumed that these water levels are normative. However, other loads are acting at the gate during transport, installation, maintenance, gate position for an open configuration, and all the positions of the gate in between a fully dammed and fully open configuration. It is recommended to check the designed cross section for these load cases in order to check the capacity of the FRP gate. Furthermore it is recommended to apply a 3D finite element method for the behaviour and transfer of stresses of the gate.
Not many guidelines are yet available for the design of FRP gates. In this research the recommendations of the CUR and the handbook of the Spieringsluis are used. The given recommendations of the CUR are conservative and not in accordance with the practice. The CUR recommends to use the reduced strain limit of 0.27% for wet environments in combination with a material and conversion factor. The handbook of the Spieringsluis and FiberCore uses only the 0.27% strain limit without material and conversion factor.
58 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
It is recommended to apply further investigation for FRP structures in humid environments in order to determine the right design limit and to review the CUR96
The eigenfrequencies of the FRP gate are similar to the eigenfrequencies of a comparable steel gate. This is caused by the large added water mass and the nearly equal EI. However further investigation is recommended for the vibration characteristics of a FRP gate due to the large thickness, the flow characteristics of submerged segment gate, and the low shear modulus: The impact of the flow of water underneath the gate has to be determined. A gap is present when the gate is not fully closed. Therefore the gate acts for a limited degree as an underflow weir. So, a high flow velocity underneath the gate is present when the gate is just opened. The impact of this flow for vibrations has to be determined in further investigation. The characteristics of the flow over the gate have to be determined. The impact of the overflow for vibrations has to be determined. Furthermore the hydraulics at the stilling basin has to be determined for a submerged segment gate weir. The FRP gate is modelled as an Euler Bernoulli beam without shear deformations for the calculation of the eigenfrequencies. It is recommended to perform more detailed calculations for a Timoshenko beam.
The lay-out of the connections of the webs and the outer walls and the connection with the rotation disks are based on an interview with a specialist. The connections are not dimensioned during the research. It is therefore recommended to elaborate the connections in more detail.
Only a preliminary design is given of the FRP element which spans in between the rotation disks. It is recommended to elaborate the weir complex in more detail.
10.2.2 DESIGN METHODOLOGY
The developed design method is tested for the Nederrijn and Lek recanalization case study only. The river branches do have a corridor shape with clear boundary conditions at the upstream and downstream boundary. It has to be tested whether this design methodology is applicable for projects without a corridor shape like the design of a key-wall. Furthermore it has to be verified whether this methodology is applicable for non-hydraulic civil engineering projects like the design of buildings or roads.
Secondly, it is recommended to apply the area analysis, requirements analysis, and functional analysis only for the concept design which covers the first four design levels of this graduation research. The area analysis, requirements analysis, and functional analysis have not to be applied for the structural design because this results in an explosion of requirements and non-meaningful descriptions of the structural elements in functional terms. The requirements defined for the concept version (design level four) have to be used as input for the structural design and expanded during the elaboration of design level five.
Thirdly, it is recommended to apply a ‘preview investigation’ to investigate the solution space in lower design levels to avoid surprises in the design process which is not included in the present SE methodology. Therefore, the SE methodology has to be extended with this ‘preview investigation.’
At last, it is recommended to relate the design criteria via the design to the requirements and vice versa which is not included in the SE methodology. Presently, the design criteria are only used for the assessment of the variants and ranking of the variants. However, one requirement could dominate the resulting design and could result in high costs. Therefore, an alternative design can be made for the
: ARCADIS & TUDelft 59
New canalization of the Nederrijn and Lek Main report
criteria: ‘the structure may not be too expensive.’ The alternative design has to be discussed with the client and a requirement could be changed or expired resulting in a cheaper design.
60 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
References
Alarm- en Berichten Centrum Arnhem, 2012. Maximale diepgang schepen in de Nederrijn en baggerachterstanden [Interview] (04 06 2012). ARCADIS, 2008. SE - Doen; practische handleiding voor toepassing in projecten, Amersfoort: ARCADIS. ARCADIS, 2010. RINK SSC - Integrale rapportage - complex Amerongen, Amersfoort: ARCADIS. ARCADIS, 2010. RINK SSC - Integrale Rapportage - complex Driel, Amersfoort: ARCADIS. ARCADIS, 2010. Statestiek Driel-Lobith, Amersfoort: ARCADIS. Beaufort, G., 2012. KIVI NIRIA kringavond; masterplan Beaufort [Interview] (12 06 2012). Blokland, P., 1957. Stuwcomplex te Hagestein. Cement, Issue Nr 9-10, pp. 367-374. Brinke, W. t., 2004. De beteugelde rivier, Diemen: Veen Magazines. CROW, 2011. Handboek specificeren, Ede: CROW. de Gijt, J. & van der Toorn, A., 2011. Structures in Hydraulic Engineering; Barriers, weirs, quay walls, and jetties, Delft: Technical University of Delft. de Ridder, H., 2012. Integraal ontwerpen, Delft: Technical University of Delft. Department of Defense, 2011. Systems engineering fundamentals, Fort Belvoir, Virginia: Defense Acquisition University Press. Havinga, H., 2012. Nieuwe kanalisatie van de Nederland; informatief gesprek [Interview] (21 06 2012). Heezik, A. v., 2006. Strijd om de rivieren. Den Haag: HNT Historische producties. Hermeling, V., 2004. Corrigeren van de afvoerverdeling op splitsingspunt 'de Pannerdensche Kop', Delft: Technical university of Delft. Kolstein, M., 2013. Groenlichtpeiling [Interview] (21 Januari 2013). Ministerie van V&W, 1957. Overwegingen tot kanalisatie van Beneden-Rijn en Lek, Den Haag: Ministerie van V&W. Ministerie van V&W, 1990. Stuwen is Varen, Ermelo: Ministerie van V&W. Ministerie van V&W, 2007. Ontwerpbelastingen voor het rivierengebied. [Online] Available at: www.rijkswaterstaat.nl Ministerie van Verkeer en Waterstaat, 1957. Overwegingen tot kanalisatie van Beneden-Rijn en Lek. 131 ed. 's Gravenhage: Ministerie van verkeer en waterstaat. Noordhoff Atlasproducties, 2011. De bosatlas van Nederland Waterland. 1e ed. Groningen: Noordhoff Uitgevers. PIANC, 2006. Design of movable weirs and storm surge barriers, Brussel: PIANC General Secretariat. Reverda, R., 2012. Depth of the sand pit near Culemborgq [Interview] (27 06 2012). Rijkswaterstaat & ProRail, 2009. Leidraad voor Systems Engineering binnen de GWW sector, Den Haag: Werkgroep Leidraad Systems Engineering. Rijkswaterstaat; Directie bovenrivieren, 1979. IJsselkanalisatie; een verkenning, Arnhem: Ministerie van Verkeer en Waterstaat. Rijkswaterstaat, 1955. Nota betreffende stuwcomplex Hagestein, Den haag: Ministerie van V&W. Rijkswaterstaat, 2011. Waterhuishouding en waterverdeling in Nederland, Den Haag: Ministerie van V&W. Simone, A., 2011. An introduction to the analysis of slender structures, Delft: Delft university of Technology. Sloten, J. v., 2012. Hogesterktebeton in de waterbouw; roteerbare waterkering van hogesterktebeton. Cement, pp. 28-30. Snijder, K., 2012. Interview met constructeur van FiberCore [Interview] (22 10 2012). stichting CUR, 2003. Aanbeveling 96; Vezelversterkte kunststoffen in civiele draagconstructies, Gouda: Stichting CUR.
: ARCADIS & TUDelft 61
New canalization of the Nederrijn and Lek Main report
Til, K., 1961. Algemeen overzicht van de Rijnkanlisatie. Cement, Issue PT 22-09-'61, pp. 697b-705b. Voortman, H., 2011. Verificatie capaciteit spuimiddel, Amersfoort: ARCADIS & DHV. Voortman, H., 2013. Evaluatie van de ontwerpmethodiek [Interview] (23 01 2013). Woud, A. v. d., 2006. Een nieuwe wereld. Amsterdam: Uitgeverij Bert Bakker. Zenkert, D., 1995. Sandwich Construction, West Midlands: Engineering materials advisory services LTD.
62 ARCADIS & TUDelft :
New canalization of the Nederrijn and Lek Main report
: ARCADIS & TUDelft 63