Paving Design Guide

Paving design for container terminals Design Guide

Royal Haskoning December 2010 Revision A 38999

  

   

Rightwell House Bretton Peterborough PE3 8DW 8DW United Kingdom +44 (0)1733 334455 +44 1733 262243

Telephone Fax

[email protected] E-mail www.royalhaskoning.com Internet

Document title

Paving design for container terminals Design Guide

Document short title Status Date Project name Project number Client Reference

A

20/12/10

Initial

8/11/10

Revision

Date

Container terminal paving Revision A December 2010 RH Technical committee 38999 Royal Haskoning 38999/R/301087/PBor

Sections 2.8, 5.7 & 6.4 revised

Description

PHL Beamish

RD Allen

PHL Beamish

RD Allen

RD Allen

PHL Beamish

Drafted by

Checked by

SUMMARY This guide considers the design of concrete block paving paving for container terminals. It is not intended to be a comprehensive guide to paving for container terminals, and the scope has been deliberately limited to the most common design adopted by RH. Nevertheless, a section reviewing the alternative types of paving is included and if it is considered useful, the scope could be widened in later revisions. Guidance is given on how to assess the loads and number of repetitions, and how to carry out the design. The method is based on using the “BPA manual”, though an approach using the Hipave software is also discussed.

Container terminal paving Revision A

-i-

38999/R/301087/PBor 38999/R/301087/PBor December 2010

CONTENTS Page 1

INTRODUCTION 1.1 Reason for the guide 1.2 Scope 1.3 Authors 1.4 Royal Haskoning experts

2

TYPES OF PAVING FOR CONTAINER TERMINALS 2.1 Introduction 2.2 Paving terminology 2.3 Concrete block paving 2.4 Asphalt 2.5 Grout filled asphalt 2.6 Conventional rigid concrete paving 2.7 Gravel beds 2.8 Runway beams and pad foundations 2.9 Rolled concrete or RCC (Rolled Compacted Concrete) 2.10 Large precast concrete panels

3 3 4 4 6 7 7 8 9 10 10

3

DESIGN METHODS 3.1 Reference documents 3.2 Software

12 12 13

4

DESIGN 4.1 4.2 4.3 4.4 4.5

General Container stacks RTG loading Occasional use of reach stacker Design life

14 14 14 14 16 16

Rubber tyred gantries Loads Number of repetitions, stationary RTG lifting a container Number of repetitions, RTG travelling over one spot Example calculation 1 Example calculation 2 Container stacking Tractor trailers and trucks General Loads, tractor trailers Loads, road going vehicles Dynamic factors Loads to be used for design Number of passes

18 18 18 21 22 22 25 25 26 26 27 29 30 31 31

5

LOADINGS 5.1 5.1.1 5.1.2 5.1.3 5.1.4 5.1.5 5.2 5.3 5.3.1 5.3.2 5.3.3 5.3.4 5.3.5 5.3.6


- iii -

1 1 1 2 2


5.3.7 5.4 5.5 5.6 5.7 6

Channelisation Reach stackers for handling full containers Small fork lift truck Hatch covers Proximity factors

OBSERVATIONS 6.1 Basecourse material 6.1.1 Introduction 6.1.2 CBM 6.1.3 Cement stabilised material 6.1.4 Wet lean concrete 6.1.5 Un-bound bases 6.2 Bedding sand 6.2.1 Grading 6.2.2 Drainage 6.3 Previous designs 6.4 16 wheel versus 8 wheel RTG’s 6.5 Tyre Pressures 6.6 Standard details 6.7 Specification 6.8 Environmental aspects


33 33 34 35 36 38 38 38 38 40 40 41 41 41 41 42 43 43 44 44 44


1

INTRODUCTION

1.1

Reason for the guide The paving for a container terminal represents a substantial cost, and a small difference in paving thickness can have a significant cost implication. In reality the design is not an exact science, but there can be pressure from clients to reduce the paving thickness by say 5% to reduce costs. Where this is the case, a detailed analysis comparing designs undertaken at different times or by different engineers can show up anomalies. Paving designs for container terminals paved with concrete block pavers over CBM (refer to section 2.3 below) are usually carried out using the “BPA manual”. However the answer depends on which edition one uses, how one assesses the loadings and how one interprets the manual. In addition the “BPA manual” method is a simplification simplification and does not always stand up to detailed rigorous examination. Based on Royal Haskoning’s extensive experience it is known that the BPA manual is conservative with respect to paving under container stacks. stacks. RH has therefore adopted an alternative method of design, justified on the basis of this experience. This document therefore has been prepared to provide a guide on the design process, so as to increase the repeatability of the designs. This guide should be treated confidentially and not issued outside RH maritime engineers.

1.2

Scope This guide covers the design of concrete block paving on cement stabilised bases. This is the most common form of paving designed by RH and is that recommended by the British Ports Manual (note however that the latter is sponsored by the concrete block manufacturers). Section 2 of this report however comprises comprises a review of the paving options available. The pavement’s function considered in this guide is to dissipate high concentrated surface loads into the formation without local failure or undue settlement in the form of surface rutting. The pavement will not inhibit consolidation or compaction related settlements resulting from sources such as liquefaction, container stacking or consolidation of reclamation fill on which the pavement is constructed; these should all be assessed separately. Pavement also has other functions that should be considered (e.g. skid resistance, the comfort of drivers) and these should be assessed separately. The guide is intended for both experienced and less experienced engineers. It is assumed that the reader has some knowledge of paving design, the BPA manual and container terminals, but is not an “expert” in this field. All designs should be checked Container terminal paving Revision A

38999/R/301087/PBor 38999/R/301087/PBor -1-

December 2010

and reviewed in accordance with Royal Haskoning’s QMS procedures; in particular it is assumed that any designs will be reviewed by an experienced engineer.

1.3

Authors This guide has been written by: Peter H.L. Beamish Roger D. Allen. Its preparation has been overseen by a group set up by the technical committee, comprising: Jonathan Tyler Chris Holt Tony Neal Alastair Reid Vendy Santruckova Peter Wright The above also contributed to the text, along with Philip Smith and Jan van Beemen.

1.4

Royal Haskoning experts The following can be consulted on paving issues: Jonathan Tyler Peterborough Chris Holt Peterborough


Container terminal paving -2-

Revision A

2

TYPES OF PAVING FOR CONTAINER TERMINALS

2.1

Introduction This section considers the alternative types of paving that can be used for container terminals, particularly for the container stackyard and immediately adjacent areas for container handling equipment. The type of paving will depend on the method of handling the containers in the yard. The main methods are: 1 2 3 4

Rubber tyred gantries (RTG’s), probably the most common method, and certainly the most common outside western Europe. Rail mounted gantries (RMG’s). This is particularly suited to automation, where the cranes are often called Automatic Stacking Cranes (ASC’s). Straddle carriers (SC’s). Common in Europe. Reach stacker or fork lift trucks (RS/FLT). This is generally only appropriate for lower throughput terminals because of the lower container stacking density, say no more than 50 to 100,000 TEU per annum. It also requires the strongest pavement.

The options for paving include: 1 2 3 4 5 6 7

Concrete block paving on cement bound material (CBM) or similar Asphalt on cement bound material (CBM) or similar Grout filled asphalt Conventional rigid concrete paving Gravel beds Runway beams and pad foundations Rolled concrete

The following has been used in the past and is included here in case one comes across it in an existing port. However in any assessment of the type type of new paving, it is suggested that it is ignored. 8

Large precast concrete panels (“Stelcon slabs”)

These are discussed below. It should be noted that many operators have strong (and differing) opinions on the type of paving to be used, based on their own experience. Therefore before selecting the type of paving to be used, one should discuss the subject with one’s client to establish their views so these can be taken into account. UNCTAD has published a monograph (No 5), titled “Container Terminal Pavement Management”, dated 1987. The authors are the same same as the earlier editions of the BPA Manual, see Section 3.1. This includes an assessment of alternative pavement types, which strongly recommends gravel beds as the best solution for container stacking. One of the authors is known for his strong advocacy of gravel beds, however a review of


38999/R/301087/PBor 38999/R/301087/PBor -3-

December 2010

existing container terminals shows that he is in a minority. It is not recommended that one refers to this UNCTAD monograph.

2.2

Paving terminology The pavement is typically considered to comprise three components: surfacing (or wearing course), structural base layer (or base course), and pavement foundation (or sub base and capping). These overlie the ground, referred referred to as sub grade. The discussion on the different types of surfacing covers the surfacing and base layer. A sub base of graded crushed rock is usually provided in all cases, unless the sub grade is particularly good quality.

2.3

Concrete block paving Concrete block paving typically comprises rectangular concrete pavers, 200mm by 100mm by 80mm thick, laid in a herringbone pattern on 30mm thick bedding sand on a base course of cement bound material or soil soil cement. These rectangular blocks are often referred to as “interlocking” although this term more properly refers to nonrectangular blocks that have definite interlocking shapes. In The Netherlands and Belgium and possibly elsewhere in Europe, concrete block paving in ports typically comprises rectangular concrete pavers, 200mm by 100mm by 120mm thick, on a 30mm thick thick bedding of fine crushed gravel (5mm). The crushed gravel bedding is for drainage of water penetrating through the groove between the blocks. It is understood that the thicker blocks have been found to cope better with with the port loading, however generally elsewhere the 80mm block as used in UK practice has been adopted. The thickness of the base course is varied varied to suit the design loadings. The base course is usually laid by a paving machine.



Revision A

Figure 2.1

Laying concrete blocks by machine

The concrete blocks provide good resistance to point loads, i.e. from the corner castings, but the paving is still semi-flexible. It can therefore cope with with limited differential settlements. The paving is usually cheaper than rigid concrete paving, paving, though not as cheap as gravel beds. The disadvantages of concrete blocks are: •

• •

It is not that good at coping with many vehicles turning or braking braking at the same point. (It is understood that this is less of a problem with the 120mm thick blocks blocks used in The Netherlands, because of the larger vertical surface); Good quality workmanship is required, both in block manufacture and in laying; Good quality blocks are required.

Further to the workmanship, one international operator reported informally that concrete block paving only worked satisfactorily if laid properly and in his experience there were few places where it could be laid properly. This is not RH’s experience, though it is noted that one exception quoted by this operator was designed and supervised by RH. Good block pavers are a high-tech product. The concrete mix, vibration, compression and curing must be fully controlled to ensure good quality. In some places achieving achieving the quality can be a problem. In Australia, problems of ants burrowing in the sand layer have been reported. The authors are not aware of this being an issue elsewhere. Stevedores from one operator complained that blocks did not give a smooth enough running surface for straddle carriers and the blocks were replaced by asphalt.


38999/R/301087/PBor 38999/R/301087/PBor -5-

December 2010

Our designs usually comprise a continuous paving, i.e. without runway beams. In general it is found that the difference between the design for the container stacks (based on 5 high) and that for the RTG runway areas is small, and therefore a standard design everywhere is justified in view of the simplified construction and operational flexibility.

2.4

Asphalt Asphalt paving has been successfully successfully used in container terminals. terminals. In container terminals it usually comprises an asphalt surfacing overlying a cement bound or soil cement basecourse. (In highway construction, however, the basecourse can also be bituminous material).

Figure 2.2

Asphalt paving, including damage from corner casting

The Asphalt/CBM option is typically cheaper than concrete block paving (CBP) to construct plus it is also reasonably reasonably easy to maintain and repair. This form of construction is relatively flexible and therefore to a limited extent it accommodates differential settlements in the underlying soils by “following” this settlement. settlement. However, the disadvantage of this option is that the Asphalt surfacing is not as durable as other surfacing materials and is likely to require replacing well before other surfacings such as CBP require repair. Its main disadvantage is that the corner castings can cause indentations in the surfacing. There are two two alternative approaches adopted by ports: • •

Accept the indentations within the container yard; Use a high high specification specification asphalt mix to provide better resistance to indenting. This approach has been successfully adopted in continental Europe, but it increases the cost.

Asphalt tends to be more susceptible to rutting. Where it is used in container terminals, concrete paving can be used in critical areas such as gate houses, where the risk of rutting is high,



Revision A

2.5

Grout filled asphalt Grout filled asphalt is an asphalt with relatively large voids into which a proprietary cement based slurry is poured. It is commonly referred referred to as Densiphalt or Hardicrete, which are trade names for the material. It provides a hard wearing surface surface that still has some flexibility. It still requires a base layer and is thus thus only an alternative to CBP or asphalt in most cases. The degree of penetration of the slurry into the voids is an uncertain factor, depending much on temperature, humidity and other factors. Quality control during application is difficult as penetration can not be observed from the surface. Quality control can only be done through destructive testing. Its use is beyond the scope of this guide and specialist advice should be sought from the suppliers. It is relatively expensive.

2.6

Conventional rigid concrete paving This comprises pavement quality concrete cast in bays with suitable joints between the bays. The concrete overlies a suitable sub base. The construction is similar to to standard concrete road construction, albeit the concrete is usually thicker. According to the design, the concrete can be reinforced with steel reinforcement or fibres.

Figure 2.3

Rigid concrete paving with reach stacker

This provides a hard wearing surface that accommodates the point loads from the containers well. This option typically has excellent durability and long term surface profile characteristics. It is also impermeable and resistant to oil spills. However this is a rigid form of construction, i.e. it does not cope well with settlements. As most container terminals are built on reclaimed land, ongoing settlement is often to


38999/R/301087/PBor 38999/R/301087/PBor -7-

December 2010

be expected. This can be mitigated to some extent by providing more joints in the paving, however this increases the cost. Rigid concrete is more difficult to break up and re-instate if there is a need to install underground services later. Rigid concrete paving is generally more expensive than concrete block paving, significantly so if a close spacing of joints is adopted because of settlement concerns. This type of paving is found in many terminals, particularly in established ports. However it is not usually used for new terminals on reclaimed land because of the settlement issues and costs. In such terminals, typically rigid concrete paving is used only for specific critical areas as follows: •

•

2.7

Where oil and fuel fuel spills spills can be expected, e.g. in workshop and and washdown washdown areas; Where concrete concrete block paving is generally adopted, rigid concrete paving can be used in areas that are particularly subject to lateral forces, e.g. from braking vehicles. Therefore in RH’s usual design, it is used at the gate house, where vehicles will all be braking at the same point.

Gravel beds Gravel beds comprise a thickness of gravel, where the container rests on the gravel, i.e. the corner castings sink into the gravel and the weight is supported on the edge beams of the container rather than the castings. They can be used with RTG’s or RMG’s. The gravel beds are used directly under the container stacks and RTG’s RTG’s run on concrete runway beams. Suitable paving has to be provided for the tractor trailers and other vehicles.

Figure 2.4

Laying gravel bed in India



Revision A

The advantage of gravel beds are: •

•

•

The paving construction is cheap compared compared to other types; (However if conventional paving is cheap, savings for the terminal area as a whole can be relatively small. This is due to the extra costs for piecemeal working, and extra works such as containment kerbs that are needed.) Usually the gravel gravel beds can act as soakaways, with significant savings in drainage costs. This bears a risk of ground contamination from leaking containers and this risk must be considered. Maintenance is usually simple, i.e. a case of raking raking the gravel gravel occasionally occasionally to to maintain level and topping up if necessary.

The main disadvantages are: •

•

•

The gravel can stick stick to the underside of the containers or in the corner castings. Gravel of a single size, usually 50mm, is used to prevent the latter, however the gravel can be damaged, splitting into smaller pieces. The gravel tends to spread throughout the terminal roads. (At one port, a full time road sweeper is used to control this.) While the maintenance is simple, it is required regularly.

Gravel beds should not be used for cold climates where where there is significant snow. The snow is compacted into the gravel (it cannot be completely cleared), and then freezes to the container along with the gravel.

2.8

Runway beams and pad foundations This system can be used for RTG and RMG operation. The containers are supported on concrete pad foundations and (as for gravel beds) runway beams and paving are provided for the RTG’s and terminal vehicles. Often beams are used rather than individual pads. The area between the concrete pad foundations has a light paving, so as to prevent problems with dust. This can be asphalt or gravel. The latter can be used as soakaways, resulting in significant significant savings on the drainage. drainage. This bears a risk risk of ground contamination from leaking containers and the risk must be considered. Where gravel is used, this system is sometimes referred to as gravel beds. However as the gravel is not supporting the containers, this is not strictly correct. Where settlements are expected, it is usual to layout the yard allocating slots for 40ft containers or 20ft containers but not to mix the two. two. This is because there have been instances where 40ft boxes were placed on plinths suitable for 20ft containers, the loaded plinths settled relative to the intermediate plinths and the boxes were then supported mid span rather than on the corner castings. With boxes then stacked on top, the bottom containers in effect broke their backs and failed.


38999/R/301087/PBor 38999/R/301087/PBor -9-

December 2010

This system has the following advantages: •

•

It can be cheaper than conventional paving, though not as economic as gravel gravel beds; It does not have the problem of gravel sticking sticking to the containers.

The disadvantages are: •

•

2.9

The layout is completely completely inflexible, i.e. it can only be used for the specified purpose and the container stacks cannot be moved. (Note however that in any container yard there is often only very limited scope for changing the layout, due to the constraints of the ridges and valleys of the drainage system and the location of lighting towers and other services.) Construction is fiddly and requires a lot of accurate setting out. This means that there is more room for errors, and the savings may not be as great as one might hope for.

Rolled concrete or RCC (Rolled Compacted Concrete) Rolled concrete has recently been trialled in some some places. It comprises a dry concrete laid and rolled in situ, i.e. similar similar to the CBM basecourse. However the differences are that the rolled concrete is used for both the base course and surfacing, and a stronger concrete is used. The technique has often been used used in dam construction and forestry haul toads in North America. The main concerns with this system are: • •

whether the surface is sufficiently hard wearing. It is very difficult to regulate the surface accurately. The resultant surface tends to be uneven, resulting in a poor running surface, puddles and uneven support to containers.

Until further evidence of good durability is available, rolled concrete is not recommended.

2.10

Large precast concrete panels This paving system comprises large precast concrete panels, often called Stelcon panels, that are laid on bedding sand. Usually the edges of the slabs are protected protected by steel angles and the slabs are reinforced.


Container terminal paving - 10 -

Revision A

Figure 2.5

“Stelcon” paving to left (after many years of service), rigid concrete paving to right

It was used because it provided the resistance to point loads of rigid concrete paving, but was more flexible and did not require complex joints. Disadvantages are: • • • •

Very difficult to lay such that a good bearing is achieved; Settlement and movements result in steps between the panels; The angles significantly increase the cost; In practice they have tended not to perform well.

This pavement type does not cope well with the high wheel loads on container terminals. If exposed to high wheel loads, the sand under the edges of the slabs compresses more than under the centre of the slab, as there is no shear transfer between the slabs. Because it is not equally supported any more, the slab starts tipping and finally cracks. With further traffic, the steel angle can break and cause severe damage to the tyres of vehicles. This type of pavement was used mainly in the 1960’s and early 70’s (later in the CIS), but has fallen out of favour due to the poor performance and high cost.


38999/R/301087/PBor 38999/R/301087/PBor - 11 -

December 2010

3

DESIGN METHODS

3.1

Reference documents The main reference used is “The Structural Design of Heavy Duty Pavements for Ports and other Industries”. There are several editions as follows: • • • •

1st edition published by British Ports Association, 1984 2nd edition published by British Ports Federation, 1986 3rd edition published by Interpave and British Ports Association, 1996 4th edition published by Interpave, 2007

The 3rd edition is often referred to as the BPA manual, however BPA are not associated with the 4th edition. One of the authors, John Knapton, has been involved in all all editions and we understand the design curves are based on his work. It should be noted that the 4th edition published in 2007 had several major typographical errors. Interpave has subsequently corrected most of these. The corrected version has the date December 2008 rather than December 2007 at the top left of the cover, but otherwise there is no admission that the text has been corrected. The user should make sure the latest edition is being used, by downloading it free from the Interpave website (www.paving.org.uk). There remains one significant “typo” in the 4th edition. In the graph for wheel loads, the ordinates (i.e. the numbers on the y axis) do not line up correctly with the grid. It is assumed that the grid is correct, i.e. at the top of the graph the numbers are displaced by one grid, giving unsafe answers if the numbers are used. The 3rd and 4th editions are superficially similar but give different answers1. A comparison of the design curves is given in in Appendix A: From these the following can be concluded: •

•

The 4th edition gives slightly thinner pavements for wheel loads. (NB. The wheel load numbers on the y axis are not in exactly the right position, so care has to be taken in reading the graph); For container stacking the 4th edition gives slightly thicker pavements.

It does not help that the design guides are not completely logical, but they are the best we have in UK practice. For ease, the reference is referred to as the BPA manual in this guide, even though this is not strictly correct for the 4 th edition. 1

rd

th

Note on difference between 3 and 4 editions. Whilst there is only a slight difference in the design curves for wheel loads between the editions, the curves have rather different derivations. They are both calibrated against BS 7533 (against the 1992 and 2001 editions respectively, but these appear to be the same). In the calibration the applied wheel loads are different. The software (finite element) is is different th rd th and the 4 edition element model is much larger than the 3 edition. Finally the the 4 edition introduces a factor of safety of 1.5 for stresses in the CBM. 38999/R/301087/PBor 38999/R/301087/PBor December 2010


Revision A

3.2

Software Royal Haskoning has purchased a computer programme, Hipave, that calculates paving thicknesses. Hipave is not a complete process – it is still necessary to make assumptions on such items as dynamic factors and container stacking weights, and it is often the practice to refer to the BPA manual for these items. The input for the wheel loads is not straightforward and it is necessary to input each axle as a separate load case, then combine them appropriately. The software requires a Dongle for the single user licence. This is held by James Morley in Maritime Peterborough AG. Because it in part relies on the BPA manual and method, it is not a completely independent method. If used correctly, one main technical benefit of Hipave appears to be that it calculates the effect of adjacent wheel loads more rigorously. In Hipave the critical design case is usually the horizontal tension in the bottom of the CBM base, and not the vertical bearing stress on top of the sub grade. However, both should be checked for completeness. Where Hipave is used correctly, the answer should be similar to that obtained from the BPA manual. If significantly different results are obtained, then it is likely that there is an error and both calculations should be carefully checked. RH NL uses different software, namely CARE. CARE. This also requires a dongle, see Vendy Santruckova. CARE appears to be based on similar principles to Hipave, but we have not carried out a detailed comparison.


38999/R/301087/PBor 38999/R/301087/PBor - 13 -

December 2010

4

DESIGN

4.1

General The design approach is based in part on the fact that previous designs have worked satisfactorily. For moving loads (RTG’s, etc) we have traditionally used the BPA manual, i.e. 3rd edition. However it is proposed by this guide that we should use the 4th edition in future. This will give slightly thinner pavements. The last two BPA editions approach the design as two parts: the structure and the foundation. The foundation design depends on the CBR2 of the sub-grade and for CBR < 5% involves adding a certain thickness of capping material to the sub-grade as prescribed in BPA’s Table 20. The structural design of the base layer then only depends on the applied loads.

4.2

Container stacks For container stacks, the BPA manual is considered conservative based on long experience of the designs we have used. The design approach adopted by RH is based on the premise that some local cracking under the container stack loads is acceptable. This is because the the loads are always applied in the same place and minor indentations are considered acceptable. This approach therefore does not check tensile cracking in the base of the CBM. However there are concerns about the design method adopted in the past, and this is currently being reviewed. Based on previous projects if the paving is designed for the RTG it should be satisfactory for the container stack. Where discrete concrete pads are used to support container stacks, these are designed as foundations. Due allowance has to be made for the the variation in location of the corner castings (i.e. where the load is applied), and on whether 20ft or 40 ft containers are being stacked.

4.3

RTG loading The BPA manual method is used for RTG loading. The BPA manual refers to a critical container weight of 22 tonnes for 40ft containers and 20 tonnes for 20ft containers. This can be used where there are millions millions of passes (e.g. tractor trailers along the quay), but where the number of passes is lower as for RTG’s, the maximum weight always gives the thickest paving. (Note that the critical container 2

There is sometimes misconception about CBR, which is intended to be a laboratory test to evaluate the mechanical mechanical strength of sub grades grades and sub bases. See Appendix B.



Revision A

weight is based on a container distribution profile from the 1970’s, and may no longer be correct, see Section 5.2.) For RTG’s, the following conditions are taken: 1

2

RTG lifting maximum weight container (30.5 tonnes if single lift). This is taken with no dynamic factors, as the RTG wheel loads should include an allowance for these. The “number of passes” is taken taken as the number of lifts at the cross row. (See below for remark remark on wind) RTG travelling with no container, but with operational wind and braking.

For the RTG lifting a maximum weight container, a critical decision is whether to include wind load with a maximum weight container. If wind load is included, even for only a small proportion of lifts, this results in a thicker pavement. RTG loading is discussed further in Section 5.1. Figure 4.1

RTG’s in container yard


38999/R/301087/PBor 38999/R/301087/PBor - 15 -

December 2010

4.4

Occasional use of reach stacker Some clients want the option of using a reach stacker to handle loaded boxes occasionally. This could be if an RTG breaks down, or if quay cranes place boxes on the quay, and a reach stacker is then used to place the the container on a trailer. Under normal operations however a reach stacker should not be handling full containers in an RTG terminal. Reach stackers carrying full boxes impose very high wheel loads on the pavement. Once dynamic factors are included, the required pavement is usually significantly thicker than required for the RTG operation. It is therefore proposed that in such circumstances the paving is designed for the RTG’s and container stacks, and it is calculated what reach reach stacker load can be taken. In a typical design, this might be a reach stacker carrying a 30 tonne container, but no increased wheel loads for braking or cornering or other dynamic factors. In such a case, reach stackers should be operated very carefully when handling heavy boxes. The client can then be presented with the options to make a decision. RH has not designed an RTG terminal that can fully cater for reach stacker loads, though for one project in Pakistan the thickness was increased to allow for a reduced reach stacker loading.

4.5

Design life Standard design life is a nominal 20 years. Generally the paving design is governed by the heaviest container loads (30.5 tonnes), because the number of repetitions repetitions is relatively low. Most containers are much lighter (an average container weight of 14 tonnes is typical). This means that in practice much of the paving should last considerably longer, with only those areas used for heavier containers being loaded as assumed by the design. Sometimes clients will be interested in the effect of whether the design lifetime should be 15, 20 or 30 years and we need to be able to make a recommendation. However this is difficult to answer consistently because the design loading is not specified precisely in the BPA manual and there is plenty of scope for the designer’s assumptions to affect the conclusions. The calculated design life can change significantly for example by changing the assumptions on the proportion of repetitions at a particular point subject to braking increments or wind loads, or the container dwell time in the stack. It may also be too theoretical an approach, considering that usage is likely to change compared to the assumptions over so many years. In any case for a 20 year design life, the number of repetitions is often below the minimum curve given in the BPA manual (250,000 repetitions)3. This means that with 3

th

rd

It is not clear why in the 4 Edition the lowest number of repetitions is 250,000. The 3 Edition effectively included a 1 repetition curve by plotting the container stacking curve on the same graph. It is noted that Knapton, in the book “In situ situ concrete industrial 38999/R/301087/PBor 38999/R/301087/PBor December 2010


Revision A

RTG loadings and the BPA method it is not possible to advise the Client what reduction in thickness is possible for a reduced design life. In conclusion, it is recommended that the 20 year design life is adopted, noting that: • •

•

•

20 years is the design standard; Much of the paving should last significantly longer because it is all designed designed for the maximum weight of container plus in some cases wind loads. As containers are generally much lighter, most areas will not be subject to this load; There are are many assumptions in the the design loadings, which mean mean that the design life should be treated as an indication, not an accurate figure; In any case, the number of repetitions may be below the threshold value in the BPA manual.

Only if a much longer design life is specified, at least 30 years, should we then review the effect of changing the design life on the paving design.

hardstandings” gives fatigue factors to allow for the number of repetitions varying between 1 and 25 million. Hipave does take into account fewer repetitions. Container terminal paving Revision A

38999/R/301087/PBor 38999/R/301087/PBor - 17 -

December 2010

5

LOADINGS

5.1

Rubber tyred gantries

5.1.1

Loads Experienced operators can usually provide information on the wheel loads from their RTG’s. Alternatively the RTG manufacturer may provide them. One issue to be considered is whether to design the paving for the RTG’s identified at the time of the design, or whether to allow for future proofing. This should be reviewed with the client, taking into account the following: •

•

The design life of the pavement is typically 20 years, which is similar to that of the RTG’s; At the time time of writing, the only foreseeable development that that would increase wheel loads significantly is twin lift RTG, i.e. an RTG that lifts 2 no full 20 ft containers at the same time.

Wheel loads are required for the following cases: A

a stationary RTG lifting a maximum weight container,

B

a moving RTG not carrying a container. It is assumed that an RTG will not carry a container along a stack in normal operation. (In practice, RTG’s may carry a container when doing housekeeping moves, but this should be rare.) For this, dynamic factors should be applied to the wheel load in accordance with Table 17 of 4th edition. (Note that the 3rd edition does not have dynamic factors for RTG’s). It is usual practice to apply a factor for both braking and uneven surface, i.e. a total factor of 20%. For braking, half the wheel loads increase by 10% and half decrease. For uneven surface at any one point all the loads could increase. The resultant is that half the wheel load passes increase by 20% and half do not change. Note that uneven surface is is not defined. For an RTG, the surface should be even when first constructed. However to allow for differential settlement, it is standard practice to include for uneven surface.

The treatment of wind loads has tended to be inconsistent. It is recommended that the operating wind is taken into account for the moving RTG, but only for say 5% of repetitions. (Calculations show that the exact percentage selected does not make much difference to the answer.) For the RTG lifting a container, the chances of a maximum weight container being lifted at the same time as the maximum operating wind occurring from the right direction (i.e. along the stack) needs to be considered. A view then needs to be taken as to whether to design for this case. The following should be taken taken into account:



Revision A

•

•

•

For twin twin lift RTG’s, the proportion of twin twin lifts lifts will be small. small. The number of lifts lifts of 50 or 61 tonnes is likely to be very small. For these lifts, it is suggested that the wind load is ignored. Single lift RTG’s are often designed for a 40 tonne lift, to allow for overweight boxes. Again there should be very very few overweight boxes, and for the 40 tonne lift again it is suggested that wind load is ignored. The likelihood of the operating wind occurring. The operating wind is either 20m/s or 25 m/s, and bears no relationship to how windy the terminal location is. There will be a wide variation between terminals of the likelihood of this wind occurring.

Where RTG’s can be subject to extreme winds, e.g. cyclones, tie downs will be required. In these cases significantly higher wheel loads can be experienced, but these will be in defined places, i.e. where tie downs are located. In these cases, “foundations” should be provided to cater for both these high wheel loads and the uplift forces. Examples of RTG wheel loads are given in Table 5.1. Figure 5.1

RTG dimensions as used in Table 5.1


38999/R/301087/PBor 38999/R/301087/PBor - 19 -

December 2010

Table 5.1

Examples of RTG Wheel loads

Ref + date of design

Width

Height Boxes

No wheels

Oman, ZPMC 2009

Dimensions

1 +7

6

8

7.9

2.7*

-

50 t

53.3

16

8.4

2.1*

1.08*

50 t

Morocco 2005

1+7

5

8

7.9

2.7

-

61 t

UK 2009

1+ 7

5

Egypt 2009

1+7

A

B

C

Lifted load

Wheel loads, tonnes Lifting With wind

Notes * Lifted load

Dimensions

5

No wind

With wind

36.6

45.0*

40.6*

23.9*

26.3

18.6

22.5*

19.9*

48.5

41

44.8*

34.3

No wind 32.3*

These loads are large. 25m/s wind speed

12.2*

16.1*

Estimated dims are conservative

26.8

30.5*

166t dead load. ZPMC. 25m/s wind speed Fantuzzi Reggiane

16

7.2

2.5

1

40t

18.18

16.27

13.54

11.69

16

7.5

3

1.5

40t

19.03

17.57

14.08

13.11

ZPMC

8

7.9

2.5

32

Not designed for wind

50t

45

Figure estimated by RH. All loads loads are are under under spreader. For comparison, comparison, the maximum weight of any any container container should be 30.5 30.5 tonnes, tonnes, whether whether it is 40ft or 20 ft. Previously the maximum weight of a 20ft container used to be only 25 tonnes. RTG’s designed for a single lift often have a capacity of 40 tonnes under the spreader to cater for f or over weight boxes. For twin lift RTGs, wheel loads are often stated for a lifted lif ted load of 50 tonnes, but they can sometimes lift 61 tonnes at reduced performance. See Figure 5.1


5.1.2

Comments

Travelling, no box


Revision A

Number of repetitions, stationary RTG lifting a container Two wheel loads are identified in Section 5.1.1, one with wind and one without wind. The case with wind represents a rare case, i.e. a maximum operating wind speed blowing along the stack combined with a maximum maximum weight lift. If it is adopted, it should only be used for a nominal number of lifts. The case with no wind is more more common. In calculating the number of repetitions, it is proposed that it is assumed that virtually all lifts are maximum weight, no wind. This is based on the premise that maximum weight containers could all be stored in one block together for the life of the terminal. It is recognised that this is not then consistent with using reduction factors for the container stacking. The number of lifts at any one point (20ft or 40ft position) can be calculated as follows: Nlift = 2 * y * B * F * HS * (365 / DT) * U Where Nlift = number of lifts y = design life in years B = number of containers in one cross row, e.g. 35 if 7 wide and 5 high F = factor of average fullness of stack. 0.6 would be a typical value.

5.1.2

Number of repetitions, stationary RTG lifting a container Two wheel loads are identified in Section 5.1.1, one with wind and one without wind. The case with wind represents a rare case, i.e. a maximum operating wind speed blowing along the stack combined with a maximum maximum weight lift. If it is adopted, it should only be used for a nominal number of lifts. The case with no wind is more more common. In calculating the number of repetitions, it is proposed that it is assumed that virtually all lifts are maximum weight, no wind. This is based on the premise that maximum weight containers could all be stored in one block together for the life of the terminal. It is recognised that this is not then consistent with using reduction factors for the container stacking. The number of lifts at any one point (20ft or 40ft position) can be calculated as follows: Nlift = 2 * y * B * F * HS * (365 / DT) * U Where Nlift = number of lifts y = design life in years B = number of containers in one cross row, e.g. 35 if 7 wide and 5 high F = factor of average fullness of stack. 0.6 would be a typical value. HS = factor to take into account extra house moves and shuffles. For every move putting a box in and out of the stack (productive move), there are additional moves to dig out boxes and to rearrange the stack (unproductive move). A typical figure for this would be 0.5 unproductive moves for every productive move. However the maximum wheel load requires the box to be at one side of the RTG, which would not normally happen with a shuffle. A figure of 1.3 is suggested. DT = average dwell time of containers in days. This varies depending on the terminal. 4 days is good, 7 days is typical and and 10 days is poor. The Client’s view should be sought and it should be borne in mind that dwell times may be high initially but improving them leads to increased loading on the paving. U = factor to allow for one part of the stack or yard being used more than others. This might be in the order of 1.2. (Note: this factor does not have a significant effect on the design, and not all designers think it is necessary to include it.). Depending on the layout of the RTG wheels, it is possible that one point is loaded by the RTG in two positions, i.e. when when working adjacent 20 ft or 40ft cross rows. This is relevant if the distance between RTG wheels corresponds to 20 ft box spacing (approximately 6.5m) in which case the number of repetitions could be doubled. This should be considered for the RTG’s being proposed by the operator and any new RTG that may be introduced later.


38999/R/301087/PBor 38999/R/301087/PBor - 21 -

December 2010

5.1.3

Number of repetitions, RTG travelling over one spot The number of times an RTG traverses over one point will depend on how well the yard is organised and the nature of the cargo, and is difficult to quantify with accuracy. However the number of traverses, Ntrav, can be related to the number of lifts each RTG undertakes, on the basis that each RTG is allocated a certain area of the yard. While in practice there will be some overlap between the areas each RTG services, this approximation is considered reasonable. The calculation is as follows: Ntrav = NRTG * y * T * C Where Ntrav = number of times the RTG traverses a particular point. For both 8 and 16 wheel RTGs, the number of repetitions is 4 times this. NRTG = number of productive lifts made by an RTG per year. year. It is assumed that unproductive lifts do not generally involve any traversing. This is true true of all shuffles, the majority of unproductive moves. y = required design life T = proportion of lifts where where the RTG has to move to another cross row. For a badly organised terminal T = 1.0, but this would not generally be allowed to happen and it is suggested that 0.5 is used. (This is based on the assumption that most transfers to road going trucks involve a traverse, but most transfers to the quayside do not. In a transhipment terminal, terminal, a lower figure of say 0.3 could be used.) C = factor to allow for the fact that when RTG moves, it will not always cross the same spot in the paving. For random movement C is theoretically 0.5 and to allow for the length of the RTG it is suggested 0.6 is used. Typically an RTG might undertake a total of about 55,000 moves per year, of which around 35,000 are productive moves. (Note: This figure is provided by one major operator. While it is considered reasonable, there is a wide variation variation in what ports achieve, and some ports achieve much higher values, e.g. 50,000 productive moves or higher. If possible, the operator should advise a figure.) If 35,000 is taken, the number of times the RTG passes a particular point per year is then: 35,000 * 0.5 * 0.6 = 10,500 per year. It is standard practice to apply the dynamic factor to all these movements, but this is conservative. This is discussed further in the example example calculation 1 below.

5.1.4

Example calculation 1 In this example, the design repetitions have been assessed for the 8 wheel RTG for Oman for a 20 year life. Wind load has been included for the lifted load. Note that this this is a relatively heavy wheel wheel load. Table 5.2 gives the results of the calculations. For the number of loaded lifts, a dwell time of 6 days has been taken. U has been taken as 1.0. 38999/R/301087/PBor 38999/R/301087/PBor December 2010


Revision A

N = 2* 20 *35 * 0.6 *1.3 * (365/6) *1.0 = 66,430 repetitions. For the RTG traversing, 10,000 passes of the RTG have been taken, for 20 years with 4 wheels. In Table 5.2, the design number of passes is calculated, ignoring any dynamic factor. The number of passes is converted to the maximum wheel load according to the ratio of the wheel load to the power of 3.75, as defined in the BPA manual. It is assumed that the wheel spacing and sub grade is such that the proximity factor is 1.0. Table 5.2

Number of passes of maximum wheel load, ignoring dynamic factors

Load case

Max wheel

Frequency

load, tonnes RTG lifting

Total With wind No wind

RTG traversing

No wind

Factor to

Passes of

repetitions

convert passes to max wheel load

max wheel load

1

66,430

53.3

0.05

3,322

1

3,322

45

0.95

63,109

0.53

33,451

1

800,000

40.6

0.025

20,000

0.36

7,207

23.9

0.025

20,000

0.05

988

32.3

0.95

760,000

0.15

116,171

Total With wind

No of

Total

161,138

The resulting number of passes is below the minimum given in the BPA charts, which is 250,000 passes. If the dynamic factors (braking and uneven surface) are applied to the RTG traversing load case, no wind only, then the results in Table 5.3 5.3 are obtained. (Note. As explained in Section 5.1.1, only half the wheel loads are increased as the braking load decreases half the loads, balancing the increase for uneven surface.)


38999/R/301087/PBor 38999/R/301087/PBor - 23 -

December 2010

Table 5.3

Number of passes of maximum wheel load, including dynamic factors

Load case

Max wheel load,

Frequency

No of repetitions

tonnes RTG lifting

Total With wind No wind

RTG traversing

No wind

Passes of max wheel

to max wheel load

load

1

66,430

53.3

0.05

3,322

1

3,322

45

0.95

63,109

0.53

33,451

1

800,000

40.6

0.025

20,000

0.36

7,207

23.9

0.025

20,000

0.05

988

32.3

0.475

380,000

0.15

58,085

32.3 + 20%

0.475

380,000

0.30

115,079

Total With wind

Factor to convert passes

Total

218,132

This shows the number of passes increases by 35%, but is still below the 250,000 threshold given in the BPA manual. It is considered that this approach is conservative but reasonable. Experience shows this example is typical. typical. From this example, the following conclusions conclusions are drawn: •

•

The number number of of repetitions repetitions is dominated by the RTG traversing, for which the figure is least well known. It is also affected significantly by the assumptions on the dynamic factors. It is apparent that any detailed calculations on changing the design life life are in practice rather nebulous, as one can change the number of passes and therefore design life by small changes in the design assumptions. In practice, the design life will be better than the 20 years used for most of the paving because of the conservative assumptions.

The effect on the design of a range of repetitions has been checked using the BPA manual, see Table 5.4. It is assumed that the CBR and wheel wheel spacing are such that the proximity factor is 1.0. Table 5.4 Case 1

Design paving thickness

Design wheel load

Number of passes

Design C8/10 thickness

53.3 t

250,000

500mm

Minimum curve for passes in BPA manual

400,000

510mm

An arbitrary, higher number of passes.

800,000

470mm

As example in Table 5.3, but

2 3

45 t

Comment

converted to wheel load without wind

From this, the following is noted:



Revision A

• •

5.1.5

The number of passes is not that critical to the final design. Case 3 illustrates the point that for RTG design etc, the maximum wheel load is the critical case. In reading the BPA manual, one could conclude that one should design for the RTG lifting a “critical container load” of 22 tonnes for a 40ft container (see Section 8.5), however this is not correct for RTG operations where the number of passes is relatively low.

Example calculation 2 In this example, the same RTG is considered as for Example 1, but the wind load is not taken with the maximum load. Table 5.5 summarises the results. Table 5.5

Number of passes of maximum wheel load, including dynamic factors

Load case

Max wheel

Frequency

load, tonnes RTG lifting

Total

1

With wind

0

No wind RTG traversing

45

No wind

Factor to

Passes of

repetitions

convert passes to max wheel load

max wheel load

1.00

66,430

66,430

1.0

66,430

1

800,000

40.6

0.025

20,000

0.68

13,597

23.9

0.025

20,000

0.09

1,864

32.3

0.475

380,000

0.57

217,108

32.3 + 20%

0.475

380,000

0.29

109,584

Total With wind

No of

Total

408,583

This gives a design paving thickness of 450mm, i.e. ignoring the wind allows a 50mm reduction in the paving thickness. As discussed above in Section 5.1.1, it is suggested that in this case, for a twin lift, the paving can be designed without wind as the lift lift is unusual. Ideally however this should be agreed with the client. Whatever is decided, the maximum design wheel wheel load will be the main determining factor in the thickness of the sub base, even if this case is taken for only 1 repetition.. repetition.. (Note: When using the 4th Edition of the BPA, the minimum number of repetitions has to be taken as 250,000.)

5.2

Container stacking The BPA manual gives container stacking loads based on the maximum permitted weight of containers, see Table 18 on page 40 in the manual. Reductions in container loads are given according to the height of the container stack, i.e. this assumes that the higher the stack the less likely that all containers will be heavy. However in practice similar containers will often be stacked together for operational reasons, and therefore this joint probability approach is not conservative.


38999/R/301087/PBor 38999/R/301087/PBor - 25 -

December 2010

It is also noted that the BPA manual gives a reduction depending on the height of the stack but no reduction for block stacking. Therefore for example if 4 containers are stacked on top of each other, a 30% reduction factor is allowed. However if these same four containers are then laid out on the ground next to each other, i.e. a block stack, then no reduction is allowed. This does not appear logical. There are therefore concerns about the reduction factors given in the BPA manual. However it is industry practice to use the BPA reduction factors, and we are not aware that this has caused any problems. The BPA manual also gives figures on the distribution of container weights. weights. These are based on data from the 1970’s when the container industry was relatively immature. While these figures give a guide, they should be treated with a good deal of caution, as the types of containerised cargo have evolved over time and will vary between terminals. In particular the the following is noted: •

•

•

•

The figures show show a maximum maximum weight of 25 tonnes for a 20 foot container. While this was correct at the time, subsequently the maximum allowed weight has been increased to 30.5 tonnes, the same as for 40ft containers. Far more cargo is now containerised than was was in the 1970’s. The increased trade in consumer goods has resulted in more lighter boxes, while this is counteracted by the fact that more bulk cargoes are being containerised, which are heavier. As the volume volume of trade has increased, the proportion of 40 ft boxes has increased. With increased trade, there is less cargo for which a 20ft box provides sufficient volume. Therefore it is more likely that 20 ft boxes are being used for heavy cargoes, where a 40 ft box can be only partially filled because of the weight restrictions. Hipave includes the the distribution distribution from the BPA BPA manual manual in its design data.

5.3

Tractor trailers and trucks

5.3.1

General Where one design of paving is used throughput, tractor trailer loading is not critical. However this information is included in case certain areas are to be designed for tractor trailers only. One particular scenario is that where containers are supported on gravel beds or pad foundations, and RMG’s are used or RTG’s are supported on runways. runways. In this option, the paving could be designed for tractor trailers only, IF the client does not require the paving to be designed for the occasional use of reach stackers. This option is outside the scope of this report, but information is given for reference. If the paving is only to be designed for road going trucks, then standard highway design or industrial paving methods should be adopted. For block paving, BS7533 Part 1 can be used. However if the paving has to cater for tractor tractor trailers or other vehicles that have higher wheel loads, then the BPA manual method should be used.



Revision A

5.3.2

Figure 5.2

Tractor trailers on quay, carrying 2 number 20 ft containers

Figure 5.3

Tractor trailers on quay

Loads, tractor trailers A typical tractor trailer is shown shown in Figure 5.2. In terms of the loading, such a trailer trailer can carry one 40ft container, maximum load of 30.5 tonnes, or two 20ft containers, maximum loading of 61 tonnes. The latter is much more more significant in terms of the paving design. However some trailers are designed to only carry 40 tonnes. Therefore where paving is to be designed for trailers one should check what type of trailers the paving is to be designed for. It is noted that with the increasing use of twin lift (i.e. quay cranes that can


38999/R/301087/PBor 38999/R/301087/PBor - 27 -

December 2010

lift two 20ft containers at the same time), most new trailers have a capacity of 61 or even 65 tonnes. Where the capacity is sufficient, it should be assumed that all 20ft containers will be carried in pairs by the tractor trailers, even if the RTG’s are not twin lift. In reviewing loading information for this report, it is apparent that a range of information has been used. Figure 5.4 shows what is considered a reasonable layout for a tractor trailer capable of carrying a 45 ft container. The Client should always be consulted for his views on trailer types and loads.

Figure 5.4

Tractor trailer layout

Table 5.6 gives the calculated axle loads for various weights of containers, based on the tractor trailer illustrated in Figure 5.4. Table 5.6

Calculated axle loads for tractor trailer

Container load tonnes

Axle load, tonnes Tractor front

Tractor rear

Trailer

0

5

5

3.8

Empty

30.5

5

15.3

13.9

Max 40 ft container

40

5

18.5

17.0

2 x 20 ft containers at critical load as per BPA

61

5

25.5

24


Comment

2 X 20 ft containers at max load


Revision A

The trailer can be parked by itself, i.e. without the tractor unit. In this case the trailer will will sit on the dolly wheels at the front (these (these can be seen in Figure 5.4). This will impose higher loads over a smaller contact area, and could be a critical load case. While this is standard operating practice in a Ro Ro terminal, it is not usual in a container terminal except in rail operations, and therefore only needs to be considered in exceptional circumstances. (It can be relevant relevant in USA style “grounded” operations.) On container terminals with rail facilities trailers with a strong fixed stand, and no dolly wheels, are often used. The stand is lifted from the ground by the tractor’s “fifth wheel”. The trailer with an import container is simply dropped under reach of the rail terminal crane and an export container on another trailer is then taken into the yard. Furthermore “out of gauge” OOG containers (i.e. containers that are larger than the standard container sizes) can be stored on such trailers. Although not intended for such use, the trailer’s stand is strong enough for landing the trailer before coming to a full standstill. The stand then skids over the paving for half a metre or so. This can happen quite frequently.

Figure 5.5

5.3.3

Trailer with fixed stand

Loads, road going vehicles Road going vehicles are rarely critical, unless an area of paving is designed only for road going vehicles. The type of road road going vehicle varies between countries, and therefore if paving is to be designed solely for such vehicles, research into the loadings should be undertaken. In such cases standard standard highway design methods should should be used. However it may be necessary to include the road going vehicles in the design of the container yard for the sake of completeness. Figure 5.6 shows a typical wheel layout for a road going truck in the UK, based on conventional tyres. tyres. (Super single tyres are becoming the norm in UK and western Europe, where each pair of tyres is replaced by one larger tyre at higher pressure. In practice the effect on the paving is not that great, due to the effect of the proximity factors.) Such a truck can carry a maximum load of Container terminal paving Revision A

38999/R/301087/PBor 38999/R/301087/PBor - 29 -

December 2010

30.5 tonnes. It can be assumed that a road going truck will only carry one full box, i.e. it is rare for a truck to carry two full 20 ft containers because of the public road weight restriction.

Figure 5.6

Road going truck

Loads are given in Table 5.7. Table 5.7

Road going trucks

Load

Axle load (tonnes)

Total

1

2

3

4

5

6

2

2

2

2

2

6

16

3 t empty container

2.5

2.5

2.5

2.75

2.75

6

19

30.5 t

7.2

7.2

7.2

9.5

9.5

6

46.5

Unladen

As noted elsewhere (see Section 5.7), the BPA proximity factor approach is not suitable for the 3x2 wheel grouping. It is suggested the pairs pairs of wheels are “merged” together and the grouping dealt with as a line of three loads.

5.3.4

Dynamic factors The BPA manual gives dynamic factors for tractor trailers and trucks as follows (see Table 17 of BPA manual) Braking Cornering Acceleration Uneven surface

+/- 10% +30% +10% +20%

Where effects are combined, these factors are added.



Revision A

It is usual practice to include the uneven surface factor, though no guidance is given as to the definition of an uneven surface. In a container terminal there will be “unevenness” “unevenness” due to the ridges and valleys required for drainage and any post construction settlements. It is therefore recommended recommended that the following values values are used: • •

Where vehicles may be cornering Elsewhere +30%

cornering + uneven surface =+50% braking/ accelerating + uneven surface =

For both braking and accelerating, the effect is +/-, i.e. the total load does not change. Braking will increase the wheel loads of the tractor unit and reduce the wheel loads of the trailer, and accelerating will be vice versa. An uneven surface can result in the wheel load always increasing at a particular point, so an increase should be applied to all wheel loads. loads. Cornering will result in all the wheel loads on the outside of the corner increasing, so again an increase should be applied to all wheel loads. 5.3.5

Loads to be used for design It is proposed that two cases are checked: 1 2

Vehicles carrying carrying the critical container load, as given in the BPA manual, for the full number of calculated passes; Vehicles carrying carrying the maximum allowed container container load, load, for the minimum number of passes, i.e. 250,000 passes.

Where the paving is to be designed for tractor trailers carrying two full 20 ft boxes, the number of tractor trailers carrying a single 40ft and two 20fts can be calculated as follows: No of trailers carrying 2 x 20ft No of trailers carrying 40 ft

= =

0.5* No of boxes *(2-TR) No of boxes * (TR -1)

Where TR =

TEU ratio, i.e. the TEU TEU throughput divided by the throughput in containers. This depends on the mix of 40 ft and 20 ft containers, and varies depending on the trade. 1.6 is a typical typical figure for deep sea trade but this should be established together with the terminal operator.

It should be assumed that all containers in any RTG block are full containers, not empty. Empty containers will usually be block stacked elsewhere using reach stackers or empty container handlers (similar to fork lift trucks, but with spreaders rather than forks). Where empties are stored in the RTG yard, they will be put together in one area. 5.3.6

Number of passes Tractor trailers run in the following areas:


38999/R/301087/PBor 38999/R/301087/PBor - 31 -

December 2010

A B

C

Within the stacks. The number of passes is not that high, but the route is is constrained In the main aisles. aisles. The number of movements movements is much higher than in the stacks, but the aisles are usually generously sized so the vehicle tracks are less channelled. In the quay area, both behind the quay and under the quay cranes.

For Area A, the number of passes can be reasonably accurately predicted based on the dwell time. In most cases this calculation is not required as the same paving will be used for the stacks and the RTG runways. Where it is required, the number of passes can be estimated by the following formula: N = 2 * y * B * F * (365 / DT) * L /TR Where N = number of passes of loaded tractor trailer or truck, based on one container per vehicle y = design life in years B = number of containers in one stack, e.g. 35 if 7 wide and 5 high F = factor of average fullness of stack. 0.6 would be typical DT = average dwell time time of containers in days. This varies depending on the terminal (and Customs procedures). 4 days is good, 7 days is typical and 10 days is poor. L = length of stack in TEU TR = TEU ratio, i.e. the TEU throughput divided by the throughput in containers. This depends on the mix of 40 ft and 20 ft containers, and varies depending on the trade. 1.6 is a typical figure. In almost all terminals, an overtaking lane is provided. The number of movements will be split between the lane under the RTG and the overtaking overtaking lane. Conservatively it is proposed to design both lanes for the full number of repetitions. The split between tractor trailers and road going trucks depends on the proportion of transhipment. Where all containers are Import export, i.e. no transhipment, then 50% of the passes will be the tractor trailer. Where it is100% transhipment, then 100% of the passes will be tractor trailers. For Area B, if the layout is efficient the number of passes will be the sum of the number of movements in each stack adjoining the roadway divided by 2. This assumes no transhipment. With transhipment, the tractor trailers are always going from the stack to the quay, so the number of passes at any one point is the sum of the number of movements in each stack to landward of that point. For Area C, an assessment based on the planned throughput of the terminal is required. The following should be noted:



Revision A

•

•

•

5.3.7

The roadway behind the quay cranes cranes will be used for moving up and down the terminal. One cannot assume that the containers for any one berth will necessarily be behind that berth. The area under the quay cranes will be divided into lanes. Each quay crane will be allocated a lane when working. Standard practice would be for the tractor trailers to use the allocated lane to access the quay. However an alternative method of working is to allocate lanes for travelling along the quay, and the tractor trailer then changes to the loading lane just before the quay crane. Lane usage under the quay cranes will not be uniform. The landward lanes will be less favoured for loading because the quay crane has to trolley further.

Channelisation The paths of the tractor trailers are constrained to a different degree in different areas of the terminal. For example when travelling under the RTG’s, RTG’s, the lanes are quite narrow and the vehicles will tend to follow very very similar paths. This effect is called channelization. According to the 2nd edition of the BPA manual, the design graphs assume that there is a high degree of channelization, and it proposes that the number of passes can be reduced for wider roads, where the vehicles would not follow exactly the same path. This reduction factor however does not appear in the 3rd edition, though the subject is discussed in general terms. In the 4th edition, the subject is again discussed in general terms (Section 8.8) but with a recommendation that “in some extreme cases” where wheels are restricted to very narrow lanes, “the number of repetitions be enhanced by a factor of five in design”. This is to avoid problems of rutting, rutting, but the logic behind this recommendation is not apparent, and it is suggested that it is ignored. For example, as far as we are aware, RH has not applied this enhancement to pavements designed for RTG’s. Rutting is usually a problem with the the surface layer, not the base layer. It is recommended that where there is a high degree of channelization, further consideration should be given to the running surface. This is one of the reasons why RH prefers to use rigid concrete paving at gatehouses. In the design of the main aisles of the container stackyard, it is considered that in general a reduction in the number of passes is appropriate. This is because the main aisles are invariably wide (between 19 and 30m is typical), and tractor trailers will follow a wide range of paths. It is proposed that the number of movements is reduced by 40%, unless the route is constrained.

5.4

Reach stackers for handling full containers Ideally the client should provide information on the reach stacker he envisages using. For fully loaded containers, a reach stacker will have a lifting capacity of at least 40 tonnes and the most likely capacity is 45 tonnes. However the size and loadings vary depending on what the capacity is in the second and third rows, i.e. a 45 tonne capacity machine is not a sufficient description. Container terminal paving Revision A

38999/R/301087/PBor 38999/R/301087/PBor - 33 -

December 2010

Typical axle loads for a range of Kalmar machines are given in Table 5.8. Table 5.8 Kalmar ref

Axle loads for Reach stackers Wheelbase M

Lifted load Tonnes

Axle load, tonnes

DRD420-60DS

6.0

0

33.1

31.0

42

94.7

11.4

DRD450-65S6

6.5

0

34.3

33.3

45

98.7

13.9

DRD450-80SX

8.0

0

51

48.7

45

126.1

21.5

Front

Comment

Back Smallest RS for full boxes Mid-range RS for full boxes Heaviest machine, capable of rd lifting 35 tonnes in 3 row. Unlikely to be used in RTG terminal.

Dynamic factors for braking, cornering, acceleration and uneven surface are given in the BPA manual. Note that acceleration is never critical. As discussed in Section 4.3, it is proposed that the designer establishes what combination of dynamic factors the pavement design can cater for, rather than design the pavement for a worst combination. It is suggested that the worst options to be considered would be uneven surface plus cornering. While in theory braking could also occur at the same time, this is considered too severe a case. As reach stackers would only be envisaged to be used in unusual circumstances, the minimum number of passes in the BPA manual can be taken.

5.5

Small fork lift truck Small fork lifts will be used on the quay to handle pallets of twistlocks, see Figure 5.7. These can be ignored in the paving calculations.



Revision A

Figure 5.7

5.6

Small fork lift truck handling pallets of twistlocks

Hatch covers Hatch covers will be stored under the quay crane, usually in the back reach, see Figure 5.8. (Hatch covers form the deck of the container vessel. Containers are stored stored both in the hold under the hatch covers, or on the deck on top of the the hatch covers. The hatch covers are removed to access the containers in the hold.)

Figure 5.8

Hatch covers stored on quay

Hatch covers have a range of sizes and designs. Weights are usually around 30 tonnes but can be up to 39 tonnes. (They will not be significantly heavier because they they are designed to be lifted by the container crane) At any one point, all the hatch covers across the beam of the vessel are likely likely to be stacked. This gives a maximum of 4 based on current designs, e.g. Emma Maersk, with with a maximum load of 4 * 39 tonnes. If the beam of vessels is restricted to panamax, then a load equivalent to 3 hatch covers could be considered.


38999/R/301087/PBor 38999/R/301087/PBor - 35 -

December 2010

The hatch covers will sit on on the paving on 4 legs. The paving can be designed as if for a container stack of equivalent load but: •

•

5.7

Allowing some some cracking cracking in the bottom of the CBM is not safe because the location of the loading points is not fixed. (However this approach has been used in the past without any reported problems). It is therefore recommended recommended that the container storage load graph is used.

Proximity factors The BPA manual includes a method of establishing the effect of adjacent wheels using proximity factors. These proximity factors depend on the CBR of the ground, see Table 19 in the manual. The method has been unchanged in the BPA Manual since the second edition (1986), is very approximate and should be used with caution. The method is widely used, however the following is noted: •

•

•

•

The formula used to calculate the the proximity proximity factors is is based on Boussinesq equations for an elastic half space. This must be considered a rough approximation as if Boussinesq were valid, it could be used for the paving design and it would not be necessary to carry out more advanced numerical calculations. . The approximation appears to assume a basecourse thickness of 300mm. This may have been a normal basecourse thickness in 1986 but in practice basecourses are nowadays generally thicker, which would give higher values of proximity factors. Effective thicknesses higher than those those estimated by the formula are indicated in the design charts for the second edition. The effective depth is estimated based on an E value for the base of 35,000 N/mm 2 which was the value given in the second edition. The fourth edition uses 40,000 N/mm2 (Table 15). This formula is not not therefore consistent with the rest 4 of the manual . (Poissons ratio for the sub-grade has also been re-assessed from 0.25 in the the third edition to 0.40 in the fourth edition. edition. Poissons ratio does not actually affect the values of the proximity factors, but will in general affect interactions between wheels in proximity.) The method adds together the tension tension stresses obtained in the bottom of the basecourse. Even within the limitations limitations outlined above, the method only works for loads in a straight line. Where there are 3 or more wheels not in a straight line, the stresses should be combined as tensors. For groups of loads not in a straight line the method method is incorrect but should be conservative. (e.g. for 16 wheel RTGs.)

4

These values of E are high compared to values measured in the field. The manual justifies the high values by stating that these are dynamic values, similar to those used in highway design. There is an argument that in container terminals the heavy loads are moving relatively slowly, and therefore the dynamic values are too high. high. However a reduced E value would result in lower proximity factors. 38999/R/301087/PBor 38999/R/301087/PBor December 2010


Revision A

The manual also takes each wheel as a single single pass. However research in US on aircraft landing gear shows that where the wheels are close together, they should be grouped as a single pass. It is recommended that for hand calculations for loads in a straight line, the BPA manual method may be used despite the above comments. Consideration should be given to changing the “starting thickness” of 300mm. If a more accurate analysis of the effect of adjacent wheels is required, then Hipave should be used. Hipave calculates the tensile stresses in the bottom of the basecourse rather than use proximity factors to estimate them. It also gives advice and can take account of the grouping effect when calculating the number of passes. It is emphasised that the BPA approach is approximate and must not be used as a method for calculating the dependency of paving thickness on the CBR of the subgrade. The method in the third and fourth editions for dealing with the CBR value is simply to increase the capping thickness for CBR less than 5%. For CBR 5% and above, the paving layers do not change. The basecourse layer thickness is independent of CBR in this design method.


38999/R/301087/PBor 38999/R/301087/PBor - 37 -

December 2010

6

OBSERVATIONS

6.1

Basecourse material

6.1.1

Introduction The BPA manual allows a wide wide range of materials to be used for the basecourse. The graphs are based on CBM3. Table 13 in the BPA manual gives Material Material Equivalence Factors for converting the calculated thickness for CBM3 to other materials. RH designs have to date always used cementitious cementitious material. The main options are: “CBM” Cement stabilised material Wet lean concrete

• • •

The decision on the type of material and strength to use is affected by the following considerations: • • •

The maximum thickness of material that can be laid whilst maintaining quality; The availability of local materials; Concerns about thermal and shrinkage cracking.

These are discussed in the following sub sections.

6.1.2

CBM CBM (cement bound material) comprises dry concrete that is mixed in a batching plant and laid by paving machine and compacted by rolling.



Revision A

Figure 6.1

Laying CBM by paving machine

Figure 6.2

Compacting CBM by roller

An important concern is how thick a layer can be laid and compacted in one pass, the compaction being the limitation. Modern plant manufacturers claim claim that their largest machines can place and compact up to 600mm in in a single pass. However, RH has never come across this in practice and it is suggested that our designs typically assume


38999/R/301087/PBor 38999/R/301087/PBor - 39 -

December 2010

a maximum of 300mm in a single pass is possible. As a typical thickness is between 400mm and 500mm, this means that the material has to be laid in two passes. Note that while traditionally this material is called CBM, in modern codes it is now called cement bound granular material (CBGM) The main choice usually is between CBM3 and CBM4, with cube strengths at 7 days of 10 N/mm2 and 15 N/mm2 respectively. From an academic point of view there there is nothing much between the two, but typically CBM4 will be cheaper as it only requires a small amount of additional cement but the overall thickness can be reduced by 20%. However, the performance of this thinner option is much more sensitive to fluctuations in strength of the CBM, the sub base, sub grade, tolerances on base thickness etc and is more likely to exhibit failure compared to a thicker layer of CBM3. CBM3. RH preference is therefore for CBM3. 6.1.3

Cement stabilised material This is similar to to CBM, except that the fill is used for the aggregate. It can be used where good quality fill is available, either gravel or sometimes coarse coarse sand. Because the aggregate grading is not carefully controlled, higher cementitious contents are required and the material is more variable. In some cases the the material has been mixed in situ. Cement is spread over the fill. It is then rotovated, water added if necessary and rolled. While it is possible to rotovate up to a depth of 550mm, there is concern about whether this technique is suitable for depths of basecourse greater than 300mm. The issues include: •

•

whether the cement can be be evenly evenly mixed mixed to provide a homogenous material, or whether the lower layers are weaker; how to adequately compact the material to its full depth. As discussed in Section 6.1.2, there is concern about compacting layers over 300mm thick.

This method is not therefore recommended for container terminals, but it may be appropriate where loads are less and therefore the paving is thinner. thinner. (However it is understood that this method has been used successfully in The Netherlands.)

6.1.4

Wet lean concrete This comprises weak concrete, usually laid by a paving machine and compacted by vibrators.



Revision A

Figure 6.3

Laying wet lean concrete by paving machine

The advantage of this option is that the thickness is only limited by the capacity of the machine, and it should be easier to obtain better quality. It is usually more expensive than the CBM, but the contractor opted to use it at one site where it was a technique he was familiar with and he had the equipment available. 6.1.5

Un-bound bases Where very good quality stone is available it is theoretically possible to use an unbound base under block paving. It is very difficult to control the quality of the stone and its placing. A number of failures failures have occurred with such bases, believed to be because of material and/or construction imperfections. Unbound bases must not be used.

6.2

Bedding sand

6.2.1

Grading It is important that that the bedding sand complies with the specified grading. grading. On one project in Ghana (not RH), the contractor used sand with the incorrect grading. The sand became waterlogged and the paving failed almost immediately. immediately. The concrete blocks had to be taken up and the bedding sand replaced.

6.2.2

Drainage Most rainwater runs off over the CBP surface to gullies and drainage channels, but some water can seep down through the grooves between the blocks. This water penetrates the bedding material and slowly runs off over the top surface of the impermeable sand cement stabilization to the valleys and accumulates there. As a result the 30mm bedding layer plus the grooves between the block pavers can become Container terminal paving Revision A

38999/R/301087/PBor 38999/R/301087/PBor - 41 -

December 2010

waterlogged and evaporation is very slow. With frequent traffic over the valley pavement, water flows up and down locally, eroding the bedding material till the pavers rest directly on the base course, causing mechanical damage and the failure of the pavement. To prevent this, holes (approx. 50 mm diameter) can be provided at say 1.5 to 2m spacing along the valley lines. Two possible details are: •

•

50mm uPVC pipes pipes mortared mortared into the CBM, extending through the sub sub base, with geotextile, see Figure 6.4 50mm holes drilled drilled through the base course filled with fine crushed crushed gravel (5mm), providing adequate drainage of the bedding layer into the sub layer.

Figure 6.4

Possible drainage detail for bedding sand

Such a detail is not necessary in dry climates. In wet climates such as UK, much paving has been laid without such drainage and performed satisfactorily. However there have been isolated problems and it is suggested that such a detail is now included.

6.3

Previous designs Table 6.1 gives some previous designs. These all used 80mm concrete blocks on 30mm sand on CBM on sub base.



Revision A

Table 6.1

Some previous designs

Project Morocco

Sub grade 10% CBR

Capping No

Sub base 150mm

CBM 460mm CBM3 design

Comments Twin lift RTG’s

360mm wet mix Middle East

7% CBR

No

200mm CBR 30%

450mm CBM4

RTG 45t, 6 high stacking

India

5% CBR

No

225mm crushed rock

420mm CBM4

40t RTG

N Europe

Varies

No

Varies

375mm CBM3

Experience has shown that such a thin pavement is satisfactory. However the fill is good quality, being gravelly sand.

Asia

6.4

?

No

?

450mm CBM4

16 wheel versus 8 wheel RTG’s 16 wheel RTGs are available and they have 4 wheels in each corner in a 2 x 2 arrangement. Some operators prefer them to to 8 wheel RTGs. It used to be such that only one manufacturer made them but now most manufacturers offer them as an option. Among the advantages that operators can perceive are: •

• •

The wheels can be smaller, using standard (truck) tyres, thus saving maintenance costs The machines are are more more stable and therefore the operating rate can can increase increase They do not need turning plates

Disadvantages can be: • •

More expensive Slightly wider thus taking up more yard space

A further advantage is that the paving needed for RTGs can probably be thinner. It is not possible to accurately design paving for 16 wheel RTGs using the BPA method because of the proximity factor issues mentioned previously (see Section 5.7). Unfortunately it appears that Hipave cannot analyse the situation with the four wheels either.

6.5

Tyre Pressures Section 8.2 of the 4th edition indicates that the damaging effect is proportional to the tyre pressure to the power 1.25. The reference pressure there is 0.8 N/mm2 but Container terminal paving Revision A

38999/R/301087/PBor 38999/R/301087/PBor - 43 -

December 2010

elsewhere in the manual, pressure is standardised to 1.0 N/mm2 and this value is used in the finite element model. The damaging effect equation is not really followed through and there is reference to solid rubber tyres, found on some trailers, developing 1.7 N/mm 2. (Section 8.6) This higher contact pressure is said to be dispersed satisfactorily through the pavement and standard design methods can be used. There are references elsewhere to heavily loaded radial tyres transmitting much of the load through the edges of the tyre due to local effects of tyre walls - rather different to the idealised uniformly distributed circular load used in analyses. Tyre pressure will also vary vary from the nominal value depending on the wheel load, ambient temperature, solar radiation and precision of maintenance inflation. Given the inconsistency of the manual, it is proposed to follow the general remarks, i.e. not make adjustments for tyre pressure.

6.6

Standard details Details of transitions between different paving designs and edge details will be required. It is suggested that one of the following projects is referred to for “standard details”. Project number

6.7

Project

Drawing nos

Comment

9S0226

Port Said

/TERM-R/1013 to 1015

Overseas project

9S5537

QICT

/LND/1220 to 1223

Has runway beams

9T0020

FSR

/02/1010 to 1015

UK, fully detailed

Specification The standard specifications, which are those used for Felixstowe, include specifications for paving. It is intended to update them to refer to EuroNorm execution and material codes. An important point to be addressed by the specification is that of pre-cracking. If the cracking is uncontrolled, large cracks can occur, which the bedding sand then fills up. This results in local settlement in the paving. Measures have to be taken to ensure that this does not happen, for example by pre-cracking the CBM to ensure lots of small cracks rather than a few large cracks.

6.8

Environmental aspects As there are more and more brownfield terminal developments, reuse of paving materials plays an increasing role. In a terminal refurbishment project, RH re-used 12 cm block pavers that had already been used for 30 years on a carrier terminal. In that period they had been relayed twice and for this project they were relayed for a third time, for use on a reach stacker terminal. Of course there was some loss due to breakage and chipping, but the majority were re-used.



Revision A

In the Netherlands an automated mobile installation is in operation, for cleaning, sorting and palletizing used block pavers, ready for mechanized paving, just like the new product. Another form of paving material re-use is crushing rejected pavers and other rejected concrete products like kerbs and sewage pipes, and re-using the material in the base layer. Similar is the re-use of the base course material for terminal refurbishment. After crushing, the material is used for a new base course, and the same strength can be achieved with less cement. In both cases the cost of new material plus waste disposal and the cost of re-use needs to be compared. In countries with high cost of waste disposal, re-use is not only environmentally friendly but may also be the most economical. For large projects mobilising an on-site crushing plant may be cost effective. For smaller projects crushing at an existing crushing plant is generally more economical.

=o=o=o=


38999/R/301087/PBor 38999/R/301087/PBor - 45 -

December 2010

APPENDIX A: Comparison of design charts in 3rd and 4th Edition Comparison of design charts of 3rd Edition (black) and 4th Edition (red).



Revision A

For block stacked containers, loads are (applying the BPA reduction factors): 4 high 5 high 6 high

853 kN 914 kN 1,097 kN


38999/R/301087/PBor 38999/R/301087/PBor - 47 -

December 2010

APPENDIX B: CBR The California bearing ratio (CBR) is a penetration test developed by the California Department of Transportation for evaluating of the mechanical strength of road sub grades and sub bases. The test is performed by measuring the pressure required to penetrate a soil sample restrained in a standard mould (152 mm diameter) with a plunger of standard dimensions (49.7 mm diameter). The CBR value is then quoted as a percentage of the pressure required to achieve an equal penetration on a standard crushed rock material, the harder the surface, surface, the higher the CBR rating. For example, soft clay may have a CBR of 1 or 2 whilst moist sand may have a CBR of 10. High quality crushed rock can have a CBR value of over 80. The standard material for this test is crushed and compacted California limestone which has a CBR value of 100. The CBR test is generally used in the laboratory to assess the likely performance of fill materials compacted to a certain reference density. The test can be carried carried out on undisturbed samples of soil but as the samples have to be 152 mm in diameter; they are effectively limited to large block samples (tube samples are generally 100 mm in diameter).

The CBR test can be carried out insitu but the value of the insitu test is questionable because it has to be carried out at the ground surface exposed at the time of testing. Further, as the test area is limited to a 50 mm diameter circle, the test results may not be representative of the ground as a whole. Finally, unlike the laboratory test, in the insitu test, the soil is not confined by a rigid mould. For these reasons, the CBR value for insitu materials is generally estimated on the basis of correlations with common insitu tests such as the dynamic probe, the SPT or the CPT. It should be appreciated that designs which are based on CBR values implicitly assume that the design value represents the worst case within the ground affected by the pavement; the design CBR should not just represent the characteristic of the surface of any particular layer of material.



Revision A

Paving Design Guide

Recommend Documents