Forth Replacement Crossing - Report 1 - Assessment of Transport Network

2 EXISTING CONDITION OF THE FORTH ROAD BRIDGE

2.1 BACKGROUND AND DESCRIPTION OF BRIDGE

When opened in 1964, the Forth Road Bridge was the longest span suspension bridge outside the USA, and the fourth longest in the world.

The bridge has a main span of 1006 metres and the side spans are each 408 metres long. The main span deck is an orthotropic steel deck comprising a 12.7 millimetres plate stiffened with longitudinal troughs. The surfacing comprises approximately 38 millimetres thick mastic, on a waterproofing layer. The side span decks comprise a 203 millimetre thick reinforced concrete slab with 38 millimetre thick mastic. Both the main and side span decks are supported on longitudinal steel beams that span between steel cross girders spaced at 9.144 metre centres. The cross girders are supported by two longitudinal stiffening trusses, which are supported via the hangers by the main cables. The hangers are comprised of wire ropes and are at 18.288 metre centres. The main cables are 600 millimetre nominal diameter and are each made up of 11618 galvanised wires each 4.98 millimetre diameter that transfer the loads from the bridge to the main and side towers and also down to the north and south anchorages. The anchorages are tapered rock tunnels filled with concrete and post tensioned. The cables at the main towers, side towers and in the anchorage chambers are seated directly on cast steel saddles. The saddles on the main tower are fixed directly onto each tower leg whilst those at the side towers and anchorages are fixed to steel rocker boxes that are free to rotate in the direction of the bridge axis.

The main towers, which are of steel box construction are 156 metres in height above water level and are a five cell structure in plan formed from three fabricated steel boxes that are joined by cover plates. The legs of each tower are connected by cross girders at the top, and just below deck level, and by diagonal stiffened box bracing above and below the deck. The approach viaducts to the bridge are steel twin box girders with a reinforced concrete deck slab. The eleven span south viaduct is 438 metres long and the six span north viaduct is 253 metres long.

Figure 2.1: The Forth Road Bridge

The design of the suspension bridge uses a stiffening truss rather than a box girder as subsequently used on the Severn and Humber bridges. The bridge carries two lanes in each direction, with wide segregated footway/cycletracks mounted on cantilevers outside the cables.

The bridge is managed by the Forth Estuary Transport Authority (FETA) and has a dedicated maintenance unit which carries out routine inspection and maintenance activities. Major tasks are usually undertaken by external contractors.

The bridge is listed by Historic Scotland as Grade A.

2.2 FACTORS WHICH INFLUENCE THE CONDITION OF THE BRIDGE

2.2.1 Introduction

It is important to understand the factors which have influenced the condition of the bridge as it stands and in the future. In the following section the various factors will be discussed and these include:

• weather and climate;
• increase in traffic since construction and in the future;
• deterioration, which is influenced by the number of vehicles, their gross and axle weights, and the detailing of the steelwork; and
• maintenance.

2.2.2 Weather and Climate

The Forth Road Bridge is at a relatively northerly latitude and is subjected to high winds blowing in from the west. In addition the presence of the North Sea to the east results in the Forth being a relatively cold body of water. This leads to foggy weather during the spring and summer months with high relative humidity. With the presence of salt water this helps to contribute to highly corrosive conditions and also reduces the length of the painting season. The conditions also make maintenance of the bridge a challenge to FETA’s maintenance team.

During winter, the carriageway is susceptible to icing with glycol being used as a deicing material on the bridge carriageways, as the use of road salts would be detrimental to the steel bridge structure.

2.2.3 Increase in Traffic Loading and Traffic Growth

Both the Forth Road Bridge and the Severn Bridge were designed to the loading specified in British Standard 153, which was derived from a train of up to five 22 ton and eight 10 ton vehicles followed by lighter vehicles. During the early life of the bridge, the amount of goods moved by road in the UK almost doubled and the number of goods vehicles with a gross weight over 28 tons rose from an insignificant number to 90,000.

This has lead to the frequent occurrence of convoys consisting of closely spaced HGVs which has resulted in loading effects greater than that originally anticipated.

A vehicle weigh-in-motion system is located just north of the bridge and this is used to generate up to date loading criteria that are used in bridge strength assessments.

The following vehicle weights have subsequently been introduced:

• in 1983, 38 tonne articulated vehicles five axle (with a drive axle maximum weight of 10.5 tonnes);
• in January 1999, the allowable weight limit was increased to 40 tonnes on five axles (with the axle weight limit increased to 11.5 tonnes);
• since 1994, six axle (max drive axle weight of 10.5 tonnes) 44 tonne vehicles have been allowed on a very limited number of UK roads; and
• in February 2001, six axle 44 tonne vehicles were allowed on all UK roads (with a max drive axle weight of 10.5 tonnes).

In addition to the growth in traffic loading set out above, the growth in traffic using the bridge has increased at a relatively steady rate of 2-3 per cent growth each year. At present the bridge carries over 24 million vehicles per annum or on average over 66,000 vehicles per day. Detailed information on traffic is given later in this report.

It should be noted that some overweight Heavy Goods Vehicles (HGVs) with a gross vehicle weight in excess of 44 tonnes use the bridge. FETA monitor vehicle weight and movements using a weigh in motion system located on the north side of the bridge.

Vehicles using the bridge affect it in different ways. In general, those elements directly supporting a vehicle will be affected most by individual axle weights, for example deck stiffeners. Cross girders carrying the deck panels will be most influenced by whole vehicle weights, whereas major elements such as the towers, main cables and stiffening truss are affected by a long queue of heavy vehicles.

Another factor which has influenced the condition of the bridge is the introduction of the high pressure single tyre on HGVs which replaces the twin wheels on trailer axles. The high pressure tyre results in increased pressure or stress on the surfacing which leads to a more rapid deterioration of the surfacing.

2.2.4 Deterioration

There are a number of deterioration mechanisms to which parts of the bridge will be subjected; these are particularly influenced by the weight and volume of traffic.

Deck Surfacing

In common with other long span bridges built in the 1960s and 70s, the bridge has a 38 millimetre thick surfacing of mastic asphalt. This was adopted as a lightweight solution, which is an important factor in the economics of a very long span bridge. The original surfacing lasted over 20 years, but successive replacements have proved not to last nearly as long. The surfacing is discussed in more detail in Section 2.4.4.

Expansion Joints and Bearings

The suspended spans of the bridge have major expansion joints at the main towers to accommodate the large range of movements. Other smaller joints are located at the side towers and along the approach spans. These, together with the bearings along the crossing, would be expected to be replaced several times during the life of the whole structure. Refer to Sections 2.4.6 and 2.4.7 for details.

Fatigue damage

Fatigue damage to the bridge is driven by three major factors:

• stress range at critical detail. The magnitude of the stress in the bridge elements changes as vehicles pass over the bridge. This stress range, or change in stress, produces fatigue damage in the bridge structure. During design, the steel elements and connections are detailed to maximise the fatigue life. The critical details which are affected by vehicles are located on the steel deck as noted in the bullet point below;
• number of stress cycles. Another important consideration is the number of stress cycles. The stress cycle refers to the change in stress described in the bullet point above. As the number of stress cycles increases, the rate of fatigue damage increases; and
• fatigue detail classification. A further consideration is the fatigue detail classification. At the design stage, all steel elements and connections are assigned a classification in accordance with British Standards. The potential for fatigue damage at an element varies and depends on the classification.

In the past fatigue has not been a major concern for the Forth Road Bridge, with the exception of the hangers. However, it may become more important in the future. The most critical fatigue details on the bridge are likely to be located on the steel deck on the main span. The deck plate and trough thickness would be regarded as thin when compared with current practice, and so to prolong its life it has to rely on composite action with the surfacing.

As discussed above, the bridge was originally designed to BS 153 and the deck panels were load tested as these elements were not specifically covered in the code. One of the most critical details is at the junction of the main troughs and the transverse members. This is an area where regular repairs are carried out. The top of the deck panels was visible when the surfacing was removed recently and there was no sign of cracking to the transverse butt welds between the deck panels, indicating that the expected fatigue life of the deck has not yet been exceeded. There was also no sign of corrosion. The expected fatigue life of the deck will be used up more rapidly than originally designed for as a result of the increase in loading and frequency of traffic. It is likely that the continued use by HGVs will lead to further and possible widespread fatigue cracking in the deck as the expected fatigue life of the deck becomes expended.

2.2.5 Maintenance

FETA’s workforce carry out most routine maintenance with the exception of specialist servicing of plant. Minor alterations to steelwork and weld repairs are carried out by FETA. Staff install and maintain most of the electrical installations, such as the toll system, CCTV cameras, weigh in motion, ice alerts, the anemometer and the variable message signs.

All maintenance painting of the structure is currently carried out by FETA staff.

During the summer months, the steelwork of the suspended span truss is washed down to remove surface salts and guano.

Since opening a number of improvements have been made to facilitate routine maintenance. These include underdeck walkways, temporary underdeck grid flooring, new maintenance gantries and rails and compressed air and water mains along the length of the bridge.

2.2.6 High Wind Management Strategy

FETA operate a high wind management strategy which is summarised below:

Gusts exceeding 35 mph.	Impose 40 mph speed limit on bridge
Gusts exceeding 45 mph.	Close bridge to double decked buses
Gusts exceeding 50 mph.	Close bridge to high-sided vehicles
Gusts exceeding 65 mph.	Close bridge to all vehicles except cars
Gusts exceeding 80 mph.	Close bridge to all traffic

FETA have provided statistics for traffic diversions as a result of high wind for the years 1999 - 2005. The data records the following information:

• days on which strong winds lead to traffic diversions of wind susceptible vehicles;
• for each day the time period during which the diversions are imposed;
• for four classes of vehicles, the number of vehicles diverted. The 4 classes are categorised as follows:
• Class 1: Motorbikes;
• Class 2: Cars/ Light vehicles not exceeding 3.5 tonnes and buses carrying not more than 16 passengers;
• Class 3: Buses carrying more than 16 passengers; and
• Class 4: HGVs exceeding 3.5 tonnes. The data is summarised in Table 2.1 below:

Table 2.1 Summary of high wind traffic diversions

Year	Number of days per year where strong wind leads to diversion of wind susceptible vehicles
1 April 2005 – 31 March 2006	10
1 April 2004 – 31 March 2005	31
1 April 2003 – 31 March 2004	1
1 April 2002 – 31 March 2003	8
1 April 2001 – 31 March 2002	19
1 April 2000 31 March 2001	6
January 1999 – December 1999	5

It can be seen from the above data that high wind leads to disruption to bridge users and is an extremely important factor when considering the performance of the bridge. The number of days affected varies considerably from year to year and the diversions normally occur during the winter months.

Anecdotal evidence suggests that the diversion of traffic from the Forth Road Bridge due to high winds, causes very significant congestion on and around the bridge, and at Kincardine. Traffic modelling to examine the effect of closing the Forth Road Bridge to HGVs on the surrounding road network will be undertaken later in the Study and reported in Report 3 on Option Generation and Sifting.

2.3 STRENGTHENING AND OTHER CONTRACTS COMPLETED

2.3.1 Introduction

Since the bridge opened several major capital projects have been carried out to replace, strengthen, or improve elements of the structure. These projects were necessary due to deterioration, changes in traffic loading, design code changes and a risk assessment of shipping impact; these are listed below:

• strengthening of Viaduct Box Girders;
• main Tower Wind Bracing strengthening;
• main Tower strengthening;
• hanger replacement; and
• construction of Pier Defences.

2.3.2 Strengthening of Viaduct Box Girders

The bridge was originally designed to sustain the loadings specified in BS 153. During the 1960s and early 70s the limitations of BS 153 with regard to the then current forms of steel construction became apparent and a temporary weight restriction was imposed. During 1976, an assessment of the viaducts to the Interim Merrison Rules was carried out and a strengthening scheme was put forward. The Interim Merrison Rules were introduced throughout the UK after further research had been carried out into the apparent weakness of steel box girders. They reflect the interim design standards of the time.

The works were completed in financial year 1981/82 allowing the permanent removal of the temporary weight restriction.

2.3.3 Strengthening of Main Tower Wind Bracing

As a result of an appraisal of the effects of an increase in the wind load derived from BS 5400 (which replaced BS 153), the main tower cross bracing members and the deck level lateral thrust members were upgraded. The works were completed in 1992.

2.3.4 Main Tower Strengthening

The requirement to strengthen the main towers was identified as a result of increases in live load and wind loading as well as differential temperature effects. In addition, the decision was taken to allow in the design of the strengthening for the possibility of a future fifth lane of traffic (light vehicles and cars only) running in the central reserve.

The upgrade of the towers involved erecting new substantial steel columns within the existing tower legs and transferring a proportion of the dead load to the new columns. The load was distributed into the new column by a jacking force of 4000 tonnes just below the saddles at the tower tops and a further 2000 tonnes applied just below deck level. Also included in the work was the strengthening of the saddles.

The works were substantially completed in June 1997.

2.3.5 Hanger Replacement

The hangers originally consisted of a pair of nominally vertical steel wire ropes looped over the suspension cable and were attached to the top chord of the steel deck truss by pairs of mutual sockets.

On 1 September 1995, one of the two ropes making up the hanger at mid span on the west cable was found to have frayed and traffic restrictions were put in place until the erection of temporary load bearing steelwork.

It was apparent from an examination of test specimens that the weakness in the ropes lay close to the sockets where a number of corroded and broken wires were found.

The design and detail of the hanger socket was changed from the original so that making up each new hanger was a pair of ropes, looped over the existing cable band, as shown in Figure 2.2. Each rope had a single socket at the ends and was fixed to a new deck detail with a single horizontal pin, as can be seen in Figure 2.3.

This was an improvement on the old detail as it allowed for any future replacement of one of the pair of ropes in the event of damage. In the original detail if one rope was damaged the pair had to be removed.

Figure 2.2: Hanger Looped over Cable

Figure 2.3 New Connection to Truss

Studies were carried out in the early 1990s to determine the effects of a ship collision. It was advised that whilst the mass concrete piers could withstand the impact, it was possible for certain vessels to impact on a highly stressed tower leg before striking the pier.

The works commenced in October 1996 and comprised steel sheet piling driven to rock or set into sand and gravel, forming large circular cofferdams on the navigational channel side of each tower. The cells were filled with rock and capped with a reinforced concrete slab.

2.4 CONDITION OF MAJOR STRUCTURAL COMPONENTS

2.4.1 Introduction

In the following section the condition of the existing bridge is outlined and an indication of the maintenance/strengthening works planned for the major structural components is provided. Reference has been made to FETA’s Capital Plan up to 2020/2021 published on FETA’s web site. In this section an indication of the how the bridge may be expected to perform in the future is provided.

A summary of the Capital Plan maintenance of principal items and the programmed schedule for these are shown below in Table 2.2.

Maintenance Activity	Time Period for Completion
resurfacing main/side span north	2007-15
resurfacing main span south	2012-20
resurfacing viaducts and north approaches	2007/2008 and 2014/2016
viaduct painting	2008-13
main tower painting	Present – 2010
main cable dehumidification (installation)	2007-09
parapet and barrier replacement	Present – 2010
main tower deck joints and cycle track joints	2008/2009
truss end linkages	2009/2010
bearing replacement viaducts and side spans	2009-12
suspended span painting	2010-21

2.4.2 Main Cables

Over the life of the bridge regular maintenance of the main cables has been undertaken, which has included periodic repainting. Regular external inspections have been carried out with no significant deterioration or egress of water noted.

During the last 15/20 years much investigative work has been carried out in the USA into the condition of main cables, with many bridges being found to have significantly deteriorated cables. Whilst there was some comfort in the fact that the Forth Road Bridge was somewhat younger than many of the American bridges investigated, FETA took a decision to follow the recommendations in the National Co-operative Highway Research Programme (NCHRP) Guidelines, and to carry out a first internal inspection of the main cables.

Cable inspection walks were undertaken which showed that the paintwork was overall in a good condition, although there were numerous locations with small defects such as those shown in Figure 2.4, and that some of these were exacerbated by defects either in the wrapping wire or crossing main cable wires below. The latter problem was noted at the area adjacent to the cable bands, but there was no evidence of water coming out of the cable at these locations.

Figure 2.4: Cracks in paint

It is noted that Faber Maunsell was lead consultant for the main cable investigation. Four low level inspection locations were recommended (two at mid main span, two diagonally opposite at ends of side spans) and four high level (two main span, two side span), as shown in Figure 2.5. Following the initial findings, a further two locations were added.

Figure 2.5: Cable Inspection Locations

Inspection Results

Following removal of the protective wrapping wire and red lead paste, lines of wedges were driven into the cable to reveal the condition of the wires as far as the centre of the cable. A number of broken wires were found in some locations, the maximum number being 31. Each wire was closely inspected and assigned a condition in accordance with the NCHRP Guidelines, ranging from Stage 1 (good) to Stage 4 (poor with significant areas of steel corroded).

In the worst location there were 55 per cent Stage 3 (some steel corrosion) and 37 per cent Stage 4. This finding has serious implications. If left unchecked, in time the Stage 3 wires will become Stage 4, and some Stage 4 wires may become cracked, ultimately leading to fracture. Examples of the findings of the cable inspection are shown in Figure 2.6.

Figure 2.6: Cable Inspection Findings

External Corrosion

Broken wires

Internal Corrosion

Strength Evaluation

A number of wires were removed from the cable for laboratory testing and over 700 tensile tests were carried out. This information was used to generate strength distributions of the wires in the various corrosion stages. The majority of wires tested achieved the minimum specified strength. Of those falling below the threshold, most were marginally under as a result of some loss of cross sectional area, and the remainder were as a result of cracks being present. The number of cracked wires is an important factor in the strength calculation.

The cable strength was evaluated using a NCHRP statistical model. It was recommended that the current strength loss was between 8 and 10 per cent.

In addition to the estimation of current strength, FETA was very interested in the projected future strength over a period of 15 years. Whilst a method for future strength estimation is provided in the NCHRP, it is somewhat tentative and is only intended to predict a further 10 per cent of the bridge’s life. In the case of the Forth Road Bridge this equals just four years.

The calculation depends on a factor which describes the rate at which a wire deteriorates from one stage to the next, determined from accelerated testing in a laboratory. This has been determined up to Stage 4 with confidence; however, the generation of cracked wires has not as yet been replicated. Therefore a range of assumptions have been made with the most pessimistic using the same rate as for progressing from Stage 1 to 4 and an optimistic one using twice this time interval.

It is appreciated that there is considerable uncertainty in future strength calculations; however, this information is very helpful in determining a long term cable strategy.

Cable Safety Factors

The cables were originally designed using a working stress approach with a factor of safety against tensile strength of 2.5. This corresponds to American practice of the day. It appears that the cables were originally slightly overdesigned and including the final weight on completion being less than expected gives an original safety factor of 2.64.

The current factor of safety has been evaluated using the cable strengths determined using the NCHRP model. These are shown in Figure 2.7 and include the calculation of cable strength in the as-new condition as well as the range of estimated loss of strength of between 8 and 10 per cent.

Figure 2.7: Current Factor of Safety

Some thought was given to the minimum acceptable factor of safety that is required to keep the bridge open to the public. There appears to be a consensus in the US that traffic should not run on a suspension bridge for any period with a factor of safety of less than 2.0, although it is known that a few bridges have been open below 2.0 for a short time whilst measures were taken to reduce loads or increase strength. However, it was agreed to set the minimum acceptable factor of safety as 2.0.

Having confirmed that the bridge cables were safe today, attention focussed on future predictions. Whilst it was appreciated that the future strength predictions could be considered as conjecture, it is nevertheless a vital piece of information for FETA to help plan future cable management strategy, with associated budgetary requirements.

The predicted changes in cable loading and strength and the corresponding factor of safety are given in Figures 2.8 and 2.9 below.

Figure 2.8: Predicted Cable Loads and Strengths

Figure 2.9: Predicted Future Factor of Safety

The more conservative case predicts that traffic restrictions would be required in 2014 (eight years time), and the other 2020. It is not feasible to reduce dead weight on the Forth Road Bridge. Indeed other studies have shown that the stiffening truss is currently theoretically understrength and will require some major strengthening soon. In the worst case it is envisaged that live load restrictions would be phased in, for example by firstly restricting HGVs in order to maintain a minimum factor of safety of 2.0 in the main cable. In this way it may be possible to continue to operate with restrictions on HGVs for five more years. The cable inspection and strength evaluation processes were audited by the Scottish Executive’s consultant, Flint & Neill Partnership (FNP). FNP broadly agreed with FETA’s methodology and findings.

Acoustic Monitoring

As this was a first internal inspection only a relatively small proportion of the cables were subject to detailed intrusive inspection. It could not of course be assumed that the worst panel had been found.

Further, the only data on the long term deterioration of the cables known with certainty are the two points comprising firstly when the cable was completed as-new, and secondly the time of the recent inspection.

The rate of deterioration of the cable is the rate at which wires pass from one stage to the next. This can only be fully measured through internal inspections undertaken at intervals. However, it is possible to record the final stage of a wire’s life when it breaks using acoustic monitoring techniques and this has been done with increasing levels of confidence on a number of other suspension bridges. A further benefit of this technique is that the entire length of the cables can be monitored.

Work started on site in April 2006 and was commissioned in August 2006. Figure

2.10 shows the equipment in place on the main cables. During the first three months of operation a total of three wires have been recorded as having broken.

Figure 2.10: Acoustic Monitoring System

Dehumidification

The future strength analysis predicts that if the cables continue to deteriorate at the present rate, the bridge will become unserviceable within a relatively short timescale. It was therefore important for FETA to consider methods for slowing down or halting the corrosion.

One possible method of protection is by oiling, which provides protection to individual wires by coating them, and also by filling the interstices between them. This has been used extensively in the US but it has a number of drawbacks including leakage of excess oil and it is an expensive operation as the entire length of cable needs to be opened to permit oil to be introduced (followed by compaction and re-wrapping), and the oil will, over time, gradually dry out making future re-oiling a challenging task.

A more recent development is cable dehumidification, which has been used on all the Honshu Shikoku suspension bridges in Japan. Systems have been fitted to new structures as well as to existing bridges. Similar systems have been retrofitted to three European bridges, although two of these have stranded cables². There is strong evidence that these systems can significantly reduce the amount of moisture held within the cable. If the relative humidity is reduced below a critical threshold, the rate of corrosion will diminish to a very low level or may even cease.

Further inspections of the cable will be made to verify the effectiveness of the dehumidification and to check the rate of corrosion. The NCHRP Guidelines recommend that, as broken wires have been found, the next internal inspection should be carried out in five years time, but this assumes that no intervention has been made. After completion of the installation of the dehumidification, the monitoring of the humidity of the exhaust air will provide some confidence that the cable is remaining dry. In addition, the acoustic monitoring will be providing additional information on wire breaks.

However, recognising the great importance of this issue, it would be prudent to carry out inspections. A strategy will be developed by FETA as time progresses depending on information being generated by the two systems. One possibility would be to target one or two particular panels which have been inspected already.

Figure 2.11: Cable Dehumidification System

Injection Sleeve on Main Cable on Minami Bisan Seto Bridge, Japan

Dehumidification Plant on Akashi Kaikyo Bridge, Japan

Faber Maunsell has carried out a feasibility study and design, with the intention of installing and commissioning the whole system by mid 2009.

Replacement or Augmentation of the Main Cables

Whilst FETA is optimistic that the proposed dehumidification will halt the corrosion of the cables, there is no absolute guarantee that this will be the case. Therefore, FETA had to give consideration to replacing or augmenting the main cables should it become necessary in the future. However, there are several significant engineering difficulties as follows:

• the main towers, including existing cable saddles, have already been strengthened and new load paths need to be identified at the tower tops;
• potential problems with the integrity of the existing anchorages, and the study would need to investigate methods of determining the condition of the post-tensioned strands;
• new anchorages will be required. It is noted that development on land to the north and south of the bridge make the provision of additional anchorages difficult; and
• the existing truss is overstressed and will also require to be strengthened.

A critical question is the extent of the carriageway restrictions that will be required to carry out replacement or augmentation of the main cables.

Given the above, FETA decided that a feasibility study should be carried out and consultant Fairhurst has recently been commissioned to undertake this work. The study has a 13 month programme and it is planned to provide input into the Forth Replacement Crossing Study during Spring 2007.

2.4.3 Painting

Painting of the bridge represents a major challenge to FETA in terms of providing suitable access and adequate containment. The ongoing painting also represents a major portion of the maintenance budget.

Up until 1980/81 maintenance painting on the bridge had been carried out using a number of different paint systems, and was limited to relatively small areas of blast cleaning and overcoating.

Following the introduction of a new permanent access system to the truss in 1980/81, major maintenance painting was undertaken on the suspended structure between 1980 and 1993. This major maintenance included the removal of the existing paint system using blast cleaning and the application of new paint systems.

From 1994 until 1999 areas of existing sound paint were washed down and overcoated using a chlorinated rubber system.

In 1999 it became apparent that the chlorinated rubber paints with a high volatile organic compound (VOC) were being phased out in response to the 1990 Environmental Protection Act. Also new legislation dictated that there should be a significant reduction in the amount of abrasive and paint entering the Forth.

Trials were carried out late in 1999 to determine whether or not the existing aged paint system could be over-coated. Unfortunately, the results of pull-off testing of this option were poor, with recurring cohesive failure of the underlying layers. Trials were also carried out on site to determine the most effective method of access and containment as well as trialling a two-pack epoxy paint system. In order to achieve adequate adhesion it was determined that it was necessary to remove all the existing aged paint system and apply the new protective system on the grit blasted steel or remaining metal spray. It was also found that a very high quality of containment was necessary to prevent paint escaping.

Estimates for painting the bridge were developed by FETA as well as by the Steel Protection Consultancy and by Balfour Beatty ltd who were painting the adjacent Forth (Rail) Bridge. The costs must be considered as approximate but all three are indicated that the total cost of painting the 202,000 plus square metres of the suspended span truss by the private sector including access, containment, blasting and the protective systems required, would be in the order of £65 million and could be achieved within a 12 to 15 year programme.

The Steel Protection Consultancy (SPC) were commissioned by FETA to review different several scenarios, including a "do nothing" approach, and spot blasting specific areas to compare with the full protection option. In their report SPC recommended the removal of the existing paint by blasting and the application of a new system. SPC further commented that if painting was delayed there would be an increased possibility that, when the future blasting and painting is carried out, the service life of the paint system would be reduced.

The complexity involved in painting of the bridge can be easily demonstrated by the "Dropped Object Canopy" which has been installed in order to protect bridge users during painting of the main towers. This is shown in Figure 2.12.

A difficulty for FETA in planning the painting work is the possibility of the future works that may be required. If the stiffening truss is to be strengthened, this is likely to involve a significant amount of welding work which will lead to paint damage. Also, if cable replacement or augmentation has to be implemented then more damage to the paint is likely to occur.

Figure 2.12: Main Tower Painting Dropped Object Canopy

In summary, the painting of the existing steelwork has become a very onerous task and needs to be carried out under a rolling programme in order to avoid storing up future problems and to maintain the condition of the bridge.

2.4.4 Surfacing

Another major onerous task faced by FETA is the maintenance of the bridge deck surfacing.

Due to the increased volume of traffic and increased weight of HGVs over the years, the surfacing on the bridge is failing sooner and having to be replaced at more frequent intervals. In addition, it is considered that surfacing life is also being shortened due to the detrimental effect of the introduction of the high pressure single tyre on HGVs, which replaces the twin wheels on trailer axles. The high pressure tyre results in increased pressure or stress on the surfacing which leads to more rapid deterioration of the surfacing.

It should be noted that some overweight HGVs with a gross vehicle weight in excess of 44 tonnes use the bridge. The increased axle loads will contribute to the increased rate of deterioration of the surfacing.

Another contributory factor leading to the deterioration of the surfacing is the detailing of the joints in the longitudinal deck stringer beams. (Refer to section 2.4.7 for more detail). FETA report that the excessive movement at these joints leads to increases in the stresses applied to the surfacing which tends to reduce the service life of the surfacing.

The road surfacing laid on the steel deck plates of the main span of the bridge is 38 mm thick mastic asphalt laid on top of a waterproofing membrane. The surfacing requires to be impervious, durable, skid resistant and of minimum weight. The mastic asphalt is also required to adhere well to the steel plates as this gives the composite action that reduces the magnitude of the stresses in the steel deck and increases the fatigue life of the deck welds.

The main span carriageway was first resurfaced in 1988 (northbound) and 1991 (southbound), having provided between 24 and 27 years service. At the same time the original waterproofing layer was also replaced with a proprietary sprayed membrane. The northbound was resurfaced using epoxy asphalt; however, this did not perform well and was soon replaced with mastic asphalt.

The northbound was resurfaced again in 1998 and is programmed to be repeated in 2007. It was found that during the 1998 operations that on removal of the old mastic asphalt damage occurred to the waterproofing membrane, which had to be repaired, considerably extending the time taken to undertake the works.

When the southbound was resurfaced in 2001, after 10 years service, efforts were made to simplify the site work by leaving a small thickness of the mastic adhering to the waterproofing. Research had been carried out in a laboratory which appeared to demonstrate the viability of the proposal. However, inspection of the surfacing during 2003 showed that the new layer of mastic had not properly bonded to the layer that had been left.

A further resurfacing contract was let in 2004, with the intention of removing all the mastic plus the waterproofing. High pressure water jetting was used to remove the material, but it was soon found that proper cleaning off took a very long time. To progress the work it was agreed to plane down to just above the waterproofing layer and then remove any remaining loose material with water jetting, followed by overlaying with new mastic asphalt. Whilst this did not achieve the optimum technical solution, it did reduce the disruption to traffic during the works.

A consequence of this experience was the realisation that the life of the surfacing was likely to be a maximum of eight years. To maintain the surfacing in the future, FETA envisage a phasing of work alternating northbound and southbound carriageways, each on an eight year cycle. The Capital Plan shows northbound work in 2007 and 2015, and southbound in 2012 and 2020.

In summary, it can be seen that the life of the deck surfacing is very limited under the current traffic loading. The replacement of the mastic asphalt and waterproofing is a very disruptive operation, and potentially the steel deck plate on the main span can be damaged during the procedure. However, if the surfacing is allowed to deteriorate there is a bigger risk that the deck plates could corrode and the fatigue life will become reduced as a result of the loss of composite action between the plate and the mastic.

FETA considers that the introduction of the HGVs’ single wheel tyres has increased the rate of deterioration of the surfacing. If the number of HGVs is reduced with the introduction of a replacement crossing, and assuming adequate traffic management systems are in place, it would be reasonable to predict that the life of the bridge surfacing would be increased.

Resurfacing causes major disruption in the vicinity of the bridge. Refer to section 2.4.11 for details of lane and carriageway closures arising from resurfacing.

2.4.5 Anchorages

The main cable anchorages of the Forth Road Bridge comprise post-tensioned strands in rock tunnels backfilled with concrete. The arrangements differ on the two sides of the Firth of Forth due to the different geological conditions. The north anchorages were built in accordance with the original design and comprise tapered rectangular tunnels driven through whinstone (basalt) rock. However, on the south side the rock was found to be much softer in places (shale and sandstone) and so the rectangular cross section was changed to a circular shape as it required less support during excavation.

Construction records which have only recently come to light suggest that, during backfilling of the working chambers of the south anchorage an explosion occurred. This took place after the strands had been post-tensioned and was believed to be due to a sudden and unexpected release of methane gas into the workings. To prevent further explosions, the chambers were allowed to naturally fill with ground water.

Figure 2.13: Layout of South Anchorage

Shortly afterwards, water and gas was observed to issue from the upper sockets. Initially the gas was thought to be methane. However, a similar phenomenon was noticed at the north anchorages where no methane had been found. The gas was tested and found to comprise mainly hydrogen. It was thought that the brackish ground water had reacted with the zinc coating on the strand wires, and that the gas travelled up along the interstices between the individual wires making up the strand. As the reaction with the zinc progressed it produced a product that tended to block the interstices and hence it was believed that the reaction would gradually slow and stop.

Advice was sought on the possible chemical processes causing the hydrogen and it was considered that initially the zinc coating was reacting with water. Whilst this is normally a slow reaction, there is a very large surface area of zinc, and also some acceleration was possible under the alkaline conditions present. Later samples of water taken from the upper sockets suggested that a more serious electro-chemical reaction was taking place.

Evidence from construction records suggest that remedial measures were planned involving sealing the lower ends of the strands. This would involve re-excavating the access shafts and removing the concrete from the working chambers. Records suggest that this was done at the north anchorage and that the top groups of strands were exposed and found to have no signs of corrosion. It is not clear whether the same was carried out at the south anchorages.

Tests were carried out in an attempt to establish the susceptibility of the strand to hydrogen induced embrittlement. The available records suggest that the strand was found to have a low susceptibility, but it was not fully reported.

It appears that no further action was taken at time of construction. It is recorded that the access shafts were backfilled with a mixture of shale and grout.

The condition of the post tensioning strands in the anchorages is an area of concern. There is no access to the base of the rock tunnels and the only inspection possible is to monitor for movement at the strand anchor bearing plate, which is now being carried out by FETA. A study to replace or augment the main cable has been commissioned by FETA and, as part of the contract, suitable methods of assessing the condition of the post tensioning inside the rock tunnels will be examined.

2.4.6 Bearings

The suspension bridge deck and approach viaduct decks are provided with bearings to permit movement arising from temperature, traffic and wind effects.

At the main towers, the deck is supported on truss linkages which allow longitudinal movement. FETA report that some of the bearings at the ends of the linkages may have seized and require replacement or refurbishment. This operation is likely to require short term carriageway closures overnight or at weekends and these may result in severe disruption to traffic. The lateral restraint bearings at the main towers were upgraded during the main tower wind bracing strengthening contract.

The bearings supporting the ends of the side spans are the original bearings, as are the viaduct bearings located on top of each pier and the abutments. On the shorter viaduct piers roller bearings are fitted (Figure 2.14) and it is known that some of the guide teeth have broken. Linear rocker bearings are fitted on the taller piers and movements are accommodated by flexure of the piers. These bearings have been inspected regularly but no major work has been carried out on them. A Principal Inspection is due to be carried out on all the bearings in 2007 and it is expected that they will all need to be replaced. Part of the complexity of the bearing replacement may be the need to strengthen the box girder to enable the bridge deck to be lifted up to permit bearing exchange. It is hoped that these bearings can be replaced without major disruption to traffic. However, this cannot be confirmed until feasibility studies have been carried out.

Figure 2.14: Roller Bearing on Approach Viaduct

2.4.7 Main Suspended Span Deck Joints and other Carriageway Joints

The principal expansion joints in the main carriageway are located either side of the main towers. These are special joints to cater for the very large movements and are now showing signs of wear and tear and are due to be replaced in 2008/2009. Experience of replacing similar joints on other major UK bridges suggests that long duration continuous carriageway closures will be required to carry out this work. For the main tower joint replacement work, FETA are studying options. However, other bridges such as Erskine Bridge with similar joints have required significant continuous carriageway closures over a period of weeks, and similar extended closures cannot be ruled out for the Forth Road Bridge.

The suspended roadway deck comprises 18.3 metre long steel panels which are supported by 4 longitudinal stringer beams per carriageway. The stringer beams are supported in turn by cross girders with movement joints in the stringer beams introduced to prevent the deck participating in the global behaviour of the bridge. The movement joints are offset from the cross girders by approximately 1.52 metres which result in the stringer beams behaving as relatively flexible propped cantilevers. As a result of the movement at these joints, high dynamic loads from heavy modern truck axles are produced causing wear and tear on the bearings and joints.

This detail has resulted in a significant maintenance problem for FETA. The shear pins within the half joints have failed and the horizontal beam surfaces have worn. FETA has tried to use bridging sections at the half joints to minimise the movement and limit the damage to the joints, beams and surfacing. FETA is examining longer term solutions to this problem but are aware that these may involve disruption to traffic.

Whilst the work is being carried out to replace the joints at the main towers it is believed that the opportunity will be taken to replace the viaduct carriageway comb joints. These are also the original joints and have been repaired a number of times but are getting beyond economical repair. Deck joints and comb joints are shown in Figures 2.15 and 2.16.

Figure 2.15: Main Tower Deck Joint

Figure 2.16: Viaduct Comb Joint

2.4.8 Parapets and barriers

The parapets and barriers on the crossing were designed before any national standards were introduced. There are several different forms of parapet and barrier in use on the bridge. On the suspended spans there are pedestrian parapets on both sides of the footway/cycle tracks, and the carriageways have a unique grillage and rail system that is designed to improve aerodynamic stability by allowing venting of air flow above and below the deck. The approach viaducts use the same outer pedestrian parapet, but have a post and rail parapet to separate the footways from the traffic and a standard open box safety fence in the median.

Following an assessment of the existing parapets to current standards, it has been established that there is a general shortfall in the strength and geometric requirements. The pedestrian parapets have been found not to be high enough for cyclists, and the spacing of infill bars is greater than permitted, so there is a risk of very small children passing through the gaps.

The suspended span outer carriageway barriers cannot be readily assessed to current standards, so it is planned to carry out off site dynamic impact tests to determine their containment. FETA point out that during the life of the bridge no cars have breached the barriers, but some HGVs have overturned onto the barriers, which has led to some local damage. The post and rail parapets at the edge of the viaduct carriageways have been found to be understrength and it is planned to replace them. This will cause disruption to bridge users. For the suspended span barriers lane closures will be required during the weekend or overnight. For the approach viaduct parapet replacement work, FETA report that it may be possible to carry out this work using narrow lanes to maintain two lane operation in each direction.

Parapets and barriers are shown in Figure 2.17.

Figure 2.17: Parapets and Barriers

Pedestrian

Suspended span grillage

Viaduct

2.4.9 Stiffening Truss

A strength assessment of the stiffening truss has been carried out. FETA report that the majority of the length of the top and bottom chord members is theoretically overstressed under a number of global loading conditions including dead load plus wind load at mid main span. Other components of the stiffening truss have lower amounts of theoretical overstress. An independent check on this work is currently being commissioned.

As stated in section 2.4.2 of this report, FETA has commissioned a feasibility study of possible methods of augmenting or replacing the main cables. The study will also look at the implications for the stiffening truss. If the cables were to be augmented or replaced, it is likely that some truss strengthening will still be required.

It is likely that the truss will require strengthening with additional plates. To carry out this work, it is likely that lane closures will be required to deliver materials to the required location. The specific method of working will be up to the Contractor, but it is likely that the Contractor would be required to deliver the materials during the night or weekend to minimise daytime lane closures.

2.4.10 Toll Plaza

The toll booths and canopies were replaced in 2006 together with a renewal of the toll registration equipment and office equipment. Following this work, FETA report that the capacity of each of the seven booths is 520 vehicles per hour and the combined traffic flow capacity of the toll plaza is in excess of the two northbound lanes they feed.

2.4.11 Effect of Lane and Carriageway Restrictions arising from Maintenance Work

FETA control all lane and carriageway closures and use their own staff to implement the traffic management. FETA tries to maximise the use of any carriageway possession, for example by carrying out some of their own tasks whilst the carriageway is closed for use by a contractor. Long summer daylight hours are used to the maximum. Maintenance work is carried out whenever necessary overnight to minimise disruption to bridge users.

Unscheduled carriageway closures include the incident in December 1997 when a hanger socket bolt failed due to an unforeseen metallurgical condition. A southbound slow lane closure was required for several days to allow for repairs and also reduced the live load on the bridge. Significant tailbacks were experienced in Fife which even affected traffic in Dunfermline town centre.

For carriageway resurfacing in 2004 FETA carried out a consultation with stakeholders. They were offered the option of a full carriageway closure for 16 weekends or for 33 continuous days. The feedback was strongly in favour of the weekend closures as this would not affect the commuter traffic. However, the tourist industry expressed concerns and so no weekend closures were scheduled for July or August. This rule is now applied for all future contracts.

Information has been sought on delays and queue lengths occurring as a consequence of the weekend closures however no information, other than anecdotal, exists.

For the main cable inspection, the majority of carriageway closures were overnight, which caused little disruption. However, a few tasks could not be undertaken at night for safety reasons and so a few weekend carriageway closures were required.

The dropped object canopy at the south main tower was mostly erected during nighttime carriageway closures. However, one operation required the complete closure of both carriageways for a short period between 00:30 and 03:30. This operation will also be repeated at the North Tower.

The acoustic monitoring system was installed with only a few weekend carriageway closures.

Resurfacing is planned for the northbound carriageway during 2007. It is planned to close the northbound carriageway for 16 weekends next summer, but excluding July and August.

The main cable dehumidification project will commence on site during 2007. The majority of carriageway closures required to move access equipment will be undertaken overnight to minimise the impact on traffic. In order to make maximum use of the northbound closures planned, work will be carried out on the adjacent cable. The dehumidification work is scheduled to continue in 2008 and 2009, and a further 16 weekends of closures per year has been made available to contractors.

For the main tower joint replacement work, FETA are studying options. However, other bridges such as Erskine Bridge with similar joints have required significant carriageway closures.

For the approach viaduct parapet replacement work, FETA reports that it may be possible to carry out this work using narrow lanes.

2.5 SUMMARY OF CONDITION OF THE FORTH ROAD BRIDGE

2.5.1 Main Cable

The condition of the main cable is well documented. Significant corrosion has been detected in the cable after the first internal inspection. The current estimated loss of strength is approximately 8-10 per cent. The original design factor of safety for the main cable was 2.64 and it is estimated that, using the most pessimistic deterioration model, the factor of safety will fall below the acceptable level of 2.0 in approximately 2013/2014. The rate of deterioration is being monitored by a recently installed acoustic monitoring system. In order to slow down or arrest the deterioration, a cable dehumidification system is currently being developed.

A lot of confidence has been placed in the dehumidification system. However, if it is not successful in arresting or reducing the rate of deterioration of the main cable, load restrictions will need to be put in place by first restricting HGVs in order to maintain a minimum factor of safety of 2.0 for the main cable. In this way it may be possible to continue to operate with restrictions on HGVs for 5 more years.

2.5.2 Painting

With the phasing out of high VOC paint systems, it has been determined that the best paint system is a two-pack epoxy system. To apply this paint and to achieve adequate adhesion it is necessary to completely remove the existing paint. The two-pack system requires improved containment compared with the original paint and, combined with increasingly more stringent health and safety requirements, the costs of repainting are high. Painting discrete areas which are badly affected has been studied but it has been concluded that this is not cost-effective.

The Steel Protection Consultancy (SPC) were commissioned by FETA to review several different scenarios including a "do nothing" approach and spot blasting specific areas to compare with the full protection option. In their report SPC recommended the removal of the existing paint by blasting and the application of a new system. SPC further commented that if painting was delayed there would be an increased possibility that, when the future blasting and painting is carried out, the service life of the paint system would be reduced.

In summary the painting of the existing steelwork has become a very onerous task and needs to be carried out under a rolling programme in order to avoid storing up future problems.

2.5.3 Surfacing

Due to the increased volume of traffic and increased weight of HGVs the surfacing on the bridge is failing sooner. In addition, it is considered that surfacing life is also being shortened due to the detrimental effect of the high pressure single tyre on HGVs. The high pressure tyre results in increased pressure or stress on the surfacing which leads to more rapid deterioration.

Another contributory factor leading to the deterioration of the surfacing is the detailing of the joints in the longitudinal deck stringer beams. FETA report that the excessive movement at these joints leads to increases in the stresses applied to the surfacing which tends to reduce the service life of the surfacing.

Resurfacing creates major disruption to the bridge users.

In order to avoid damaging the steel deck plate on the main span it has been found that when resurfacing, the existing mastic had to be planed as close as possible to the waterproofing before relaying an upper layer of mastic. FETA estimate that the maximum life of the surfacing is 8 years and envisage a phasing of work alternating northbound and southbound carriageways on an eight year cycle. If resurfacing is not carried out regularly there is a real risk that the composite action with the deck plate will be lost with a resulting reduction in the fatigue life of the deck plate.

2.5.4 Anchorages

During backfilling of the working chambers of the south anchorage after the strands had been post-tensioned an explosion occurred. Remedial works were carried out at the north anchorage but records for the south anchorage are unclear. There is a possibility that the strands will have undergone hydrogen embrittlement and the condition needs to be determined.

Due to the inaccessibility of the strands the only possible method of determining if there is any movement is at the strand anchor bearing plate. As part of the cable replacement study, suitable methods of verifying the condition are being studied. The implications if there is a loss of strength of the strands is serious due to the complexity of any possible remedial works.

2.5.5 Bearings

The lateral restraint bearings at the main towers have already been refurbished during the main tower wind bracing strengthening contract. It would be reasonable to expect these bearings to have up to 40 year’s life.

At the main towers, the deck is supported on truss linkages which allow longitudinal movement. FETA report that some of the bearings at the ends of the linkages may have seized and require replacement or refurbishment. This operation is likely to require short term carriageway closures and these may result in severe disruption to traffic.

A Principal Inspection is due to be carried out on all the bearings in 2007 and it is expected that they will all need to be replaced. Part of the complexity of the bearing replacement of the approach viaduct bearings may be the need to strengthen the box girder to enable the bridge deck to be lifted up to permit bearing exchange. It is hoped that these bearings can be replaced without major disruption to traffic. However, this cannot be confirmed until feasibility studies have been carried out.

2.5.6 Main Suspended Span Deck Joints and other Carriageway Joints

There is evidence of deterioration of the main expansion joints located either side of the main towers and it is expected that these will in 2008/09. Experience of replacing similar joints suggests that long duration carriageway closures will be required.

The movement joints in the longitudinal stringer beams are subject to wear and tear as a result of the high movements at these locations. FETA have installed bridging sections to minimise the movements. FETA are examining longer term solutions to this problem but it is likely that these solutions will lead to disruption to traffic.

The viaduct carriageway articulation and comb joints also need to be replaced as these are getting beyond economical repair.

2.5.7 Parapets and Barriers

Following an assessment of the existing parapets to current standards, it has been established that there is a general shortfall in the strength and geometric requirements. The pedestrian parapets have been found not to be high enough for cyclists, and the spacing of infill bars is greater than permitted, so there is a risk of very small children passing through the gaps.

It is planned to test the suspended span outer carriageway barriers off site. The parapets at the edge of the viaduct carriageways have been found to be understrength, and it is planned to replace them.

2.5.8 Stiffening Truss

The top and bottom chords of the stiffening truss are reported to be overstressed under certain load combinations. An independent check on this work is currently being commissioned.

2.5.9 Toll Plaza

The toll booths and canopies were replaced in 2006 together with a renewal of the toll registration equipment and office equipment. Following this work, FETA report that the capacity of each of the seven booths is 520 vehicles per hour and the combined capacity of the toll plaza is in excess of the bridge.

2.5.10 Effect of Lane and Carriageway Restrictions arising from Maintenance Work

FETA control all lane and carriageway closures and use their own staff to implement the traffic management. FETA tries to maximise the use of any carriageway possession, for example by carrying out some of their own tasks whilst the carriageway is closed for use by a contractor. Long summer daylight hours are used to the maximum. Maintenance work is carried out whenever necessary overnight to minimise disruption to bridge users.

2.6 CONCLUSION

Although the Forth Road Bridge has been maintained throughout its lifetime, it is showing signs of deterioration, mainly as a result of the growth and increase in weight of traffic together with the influence of the weather and climate.

Following the internal inspection of the main cables, severe corrosion was discovered and the cables are estimated to be working with a loss of strength of 8 to 10 per cent. The current cable factor of safety is between 2.2 and 2.3. Predictions indicate that, at the present rate of corrosion, the factor of safety could fall below the acceptable value of 2.0 in 2013/2014. It is expected that the installation of the dehumidification system in 2007-08 will arrest the deterioration. It is also recognised that, in the worst case where the dehumidification does not work, steps such as the restriction of HGVs would need to be phased in.

In several other aspects, regular disruptive maintenance to preserve the integrity and life of the bridge is required. The condition of the strands in the anchorages is unknown and some form of testing needs to be devised to provide confidence that the strands have not deteriorated through hydrogen embrittlement. A suitable form of testing is being studied but the method and feasibility has not yet been determined.

Strengthening is required to the stiffening truss. Although fatigue has not been an important issue to date it is likely that the fatigue sensitive details in the main deck will need major maintenance in the future.

Resurfacing and painting are extremely disruptive operations and both operations need to be carried out regularly to maintain the integrity and life of the structure.

In common with all bridges, the support bearings need to be replaced several times through its lifetime. The existing parapets do not comply with current requirements and it is likely that they will need to be replaced. The suspended span outer barriers will be tested offsite to determine their level of containment.

The bridge continues to carry an average of 66,000 vehicles per day. There are still many years of life in the bridge but with high ongoing maintenance liabilities, provided that the deterioration of the main cable can be arrested. With the studies on the main cables now under way, it is considered that the next critical issue to be studied and resolved is the condition of the anchorage strands. As noted in 2.4.2, FETA has commissioned a study into the replacement or augmentation of the main cable and the study will investigate methods of determining the condition of the post-tensioned strands.

Ongoing painting of the bridge is required whatever the live loading applied to the bridge. The volume and weight of the live load has a critical effect on the surfacing, deck, main cable, hangers, joints and bearings.

In summary, there will be an increased requirement for disruptive maintenance in the foreseeable future, due to increases in traffic levels, HGV proportions (and weight), and remedial measures that require to be undertaken on the bridge structure itself.

Contents   « Previous   Next »