Breaking News
Quick Links
Map Based Road Projects Finder - Feedback
We are currently piloting a new map based tool to help you find road projects. Please let us know what you think
The Project Control Framework
This framework sets out how we, together with the Department for Transport, manage and deliver major road improvement projects.
A417 Cowley to Brockworth Bypass Improvement
7 ENGINEERING COMPARISON OF OPTIONS
7.1 This chapter considers the engineering issues associated with each of the three options being considered. Because of the importance of geotechnical and groundwater issues for this scheme, this section is divided into two sections: Geotechnical Engineering and Highway Engineering. However, all assessments carried out to-date should be treated as preliminary and require more extensive investigation and analysis to refine the points made.
Geotechnical Engineering
Options 1 and 2 - Crickley Hill
7.2 Above the escarpment route options would be constructed over the limestones of the Lower Inferior Oolites, which are likely to provide relatively simple construction conditions, although some minor solution features may be present. It should also be noted that the Geotechnical Feedback Report for Birdlip Bypass (Gloucester County Council, 1989) indicates that toward the edge of the escarpment the limestone becomes cambered and a series of deep fissures (sub-parallel to the escarpment) were encountered during construction of the present route south of the Air Balloon Roundabout.
7.3 Where options cross the unstable escarpment slopes, it is considered that a combination of cutting, particularly within side-sloping ground, and embankment would be required to provide suitable vertical alignments. Special stabilisation and drainage measures would also be required for highway construction. Although not prohibiting construction within the escarpment slopes, the presence of this significant thickness of unstable ground on the slopes below the escarpment would provide considerable engineering and construction difficulties.
7.4 The slopes below the escarpment consist of a complex mass of landslipped material, masked by a mantle of soliflucted Head deposits. In general, the fines content of the colluvial deposits increases westwards towards the Vale of Gloucester. Three main types of colluvial deposit have been identified as forming the slopes and comprise the following:
- Angular limestone gravels, containing some cobbles and boulders of oolitic limestone together with a varying sand content (dominating the upper slopes);
- Clayey silt and fine sand; and
- Silty clays (dominating the lower slopes).
7.5 Where the ground comprises silts and sands, slope angles of up to about 20o (in most groundwater conditions) may be achieved. However, where silty clays are present, i.e. below, say, Cold Slad, a slope of approximately 10o (typical of much of the middle and lower slopes) is considered to be only marginally stable, assuming the presence of relic shear zones. Therefore, it is considered that even minor changes to the groundwater regime or disturbance could reactivate failure, which could affect a significant area up slope.
7.6 Stability analysis carried out as part of this study suggests that, within the colluvium, unsupported cut slopes should not exceed approximately 2m. Where cut slopes are required in excess of 2m, a retaining measure would be needed.
7.7 In general, groundwater conditions are not well understood, but Hutchinson (1991) has identified some areas of artesian and near-artesian groundwater. In addition, during the construction of the present alignment in the 1960's an area of the slope below Crickley Hill is recorded to have failed due to a significant groundwater issue. This section was stabilised with the use of deep counterfort drainage.
7.8 The extent of colluvial (landslipped) materials within the study area can be seen from the constraint mapping to generally cover the area from the top of the escarpment to the floor of the Vale of Gloucester. These slopes are likely to comprise soliflucted and landslipped material, comprising a mixture of the overlying and underlying strata, with a number of randomly located springs, indicating the complex nature of the hydrogeological regime. They comprise unstable ground on which special stabilisation and drainage measures would be required for highway construction. However, various options are available for the proposed highway improvements below the escarpment. Each option would affect the marginally stable valley slopes to a varying degree. At this stage and in light of the uncertainties regarding depth of existing potential failure surfaces, groundwater pressures and soil properties, the amount of stabilisation works required would be dependant upon the route alignment adopted.
0 to 400 metres from the Balloon roundabout
7.9 Improvement options over the initial 400 metres from Air Balloon Roundabout, are likely to involve on-line widening on one or both sides of the existing carriageway within the existing highway boundary. However, Option 2 includes some off-line construction in the vicinity of the existing roundabout together. Existing cut slopes in this area would appear to be stable and, therefore, slopes at the existing gradient (approximately 30º) should be assumed at this stage, where these can be accommodated. Where options require slopes to be steeper than approximately 30º retaining measures, probably in the form of walls, would be required.
7.10 No significant engineering constraints are anticipated over the initial section and it is proposed that widening at existing slope angles to both the north and south could be achievable. Similarly, where new or deeper cuttings are proposed, these could be achieved at the angles of the existing slopes.
7.11 Where required, widening of the shallow cut slopes on both the northern and southern sides of the existing carriageway could be achieved by cutting at similar slope angles and maximising the use of the relatively wide verges, particularly to the north of the existing carriageway. Where there is a need to create new cuttings similar slope angles to those already adopted within the limestone should be employed. Further investigation would be required where steepening of the existing slopes is required or where new or deeper cuttings are proposed. To achieve widening within existing land-take, low retaining structures may be required to support cut slopes and minimise land take.
7.12 The northern slope, below Crickley Hill, is observed to be within the Oolitic Limestone Group. Improvement options are constrained by these slopes and would require construction of an anchored retaining wall (probably by top-down construction) together with deep drainage measure to stabilise the slope.
7.13 With the slopes above the northern boundary of the A417 being designated as SSSIs, together with them being nationally important rock exposures, widening options to the south side of the existing route are considered the only ones viable. 400 metres from the Air Balloon Roundabout to Grove Farm Access and Cold Slad
7.14 Below approximately 400m from the roundabout the ground drops away steeply on the southern side of the existing carriageway. The carriageway is, currently, partly supported by an embankment at about 18º supported by a minor retaining measure. To the north, the cut slopes comprise 30º to 40º separated by wide benches within the Oolitc Limestone Group.
7.15 The road improvement options over this section are again constrained to the north by the slopes below the Crickley Hill escarpment. Widening options of these slopes would require cutting back of a substantial part of this rock slope at a similar angle. This is not considered possible due to the adjacent SSSI, National Nature Reserve and National Trust land and, therefore, construction of an anchored wall (by top-down construction) and deep drainage would be required to stabilise the slope. However, founding such a wall on potentially unstable ground below the limestone, i.e. from approximately 450m west of the Air Balloon Roundabout, may require substantial piles to support the wall and to stabilise the slope. Stabilisation methods, such as piled dowels and drainage techniques, may also be required in combination with the retaining measures in order to provide a stable solution.
7.16 Off-line widening to the south could be achieved over this section of the route by the use of earthworks and retaining walls between carriageways. The use of a split-level carriageway in Option 2 may be considered, thus, reducing the volume of imported fill and, hence, the potential destabilising effect of filling on these slopes. However, careful consideration is required to the founding of any retaining measures used between the carriageways within the potentially unstable ground, particularly where the carriageways start to converge toward Air Balloon Roundabout (it is recognised that the retained height would decrease over this transition). This option would potentially involve the demolition of the former Cotswold Restaurant, although has the advantage of having little impact on the existing road during construction as well as not impacting upon the SSSI to the north. As a result of the recorded artesian groundwater pressures in this area, it is likely that a provision of a drainage blanket would be required beneath any earthworks to ensure dissipation is achieved.
7.17 Provision of a piled carriageway may be considered should land-take issues constrain the route of this section.
7.18 Further culverting of the stream would be required in Option 2. Agreement from the Environment Agency would be required to culvert or divert the stream below the embankment.
7.19 The Crickley Hill escarpment, again, constrains the improvement options over this section. Widening to the northern side of the existing carriageway may also cause considerable disruption to the traffic during the construction phase and may require single lane traffic flow during some phases of the works.
7.20 Widening to the south could be achieved over this section of the route by the use of earthworks and retaining walls between carriageways. Embankment widening of up to 6m high together with the use of a split-level carriageway may be considered. This potentially involves the demolition of the former Cotswold Restaurant. This option has the advantage of having little impact on the existing road during construction as well as not impacting upon the SSSI.
Cold Slad / Grove Farm Access Road to Dog Lane
7.21 The ground is likely to comprise a mixture of the colluvium types described in previous sections, but it is likely to be dominated by clays and silts on these lower slopes. The ground and groundwater conditions are likely to be highly variable over this section of the route and, as a result of previous ground movements, a detailed investigation would be required to provide sufficient data to design earthworks and stabilisation / retention measures.
7.22 The improvement options over this section are again constrained to the north by the slopes below the Crickley Hill escarpment. However, as a result of the wide verge, minor widening to the north may be achieved without disturbing the existing slopes.
7.23 Widening to the south could be achieved over this section of the route by the use of earthworks and possibly retaining walls between carriageways. The use of a split-level carriageway may be considered, thus reducing the volume of imported fill and, hence, reducing potential destabilising effects of filling on these slopes. Further culverting or realignment of the stream would be required should a southern widening option be taken forward. As a result of the recorded artesian groundwater pressures in this area, it is likely that a provision of a drainage blanket (together with other slope drainage measures) would be required beneath any earthworks to ensure dissipation is achieved.
7.24 Widening works to the south would encounter deep soft organic deposits within the valley floor. Construction of embankments over these materials are likely to lead to considerable, potentially uneven, settlement and may interfere with the existing groundwater regime as a result of the stream re-alignment/culverting. These settlements may affect the existing carriageway, particularly where the toes of existing fill slopes are close to the valley centre. Likely differential settlement between the existing and new carriageways may be significant and careful consideration of these would be required in design of the widening options.
7.25 The anticipated engineering problems associated with widening to the south, while still significant, are considered to be less than those associated with widening to the north of the existing alignment.
7.26 The road improvement options over this section are again constrained to the north by the slopes below the Crickley Hill escarpment. However, minor widening to the north, within the existing wide verge, may be achieved without disturbing the existing slopes.
7.27 Widening to the south could be achieved over this section of the route by the use of earthworks and retaining walls between carriageways. The use of a split-level carriageway may be considered.
7.28 Significant widening works to the northern side of the existing carriageway may also cause considerable disruption to the traffic during the construction phase and may require single lane traffic flow during some phases of the works.
7.29 The anticipated engineering problems associated with widening to the south, while still significant are considered to be less than those associated with widening to the north of the existing alignment and, therefore, widening to the south is considered more appropriate.
Dog Lane to the Brockworth Bypass.
7.30 The ground is likely to comprise a mixture of the colluvium types described previously, although clays and silts will dominate. The groundwater conditions are likely to remain highly variable over this section of the route as a result of the highly variable ground conditions and a detailed investigation would be required to provide sufficient data for geotechnical design.
7.31 Widening to the north of the existing carriageway is constrained by the marginally stable shallow angle slopes below Crickley Hill. However, minor widening may be achieved by a series of retaining measures, particularly below the section of Dog Lane that runs close to the existing alignment (1200 to 1420 metres from the Air Balloon Roundabout).
7.32 Widening could be achieved to the south of the existing carriageway by the construction of further earthworks. Again a split-level carriageway may be considered. Comments made previously regarding the provision of adequate drainage measures and differential settlement of the carriageway would apply here.
7.33 Agreement from the Environment Agency would be required to culvert or divert the stream below the embankment.
Option 3 - Tunnel
7.34 The tunnel option has been assumed to consist of two road tunnels, connected at approximately 100 m intervals by cross passages (DMRB:BD 78/99). Both tunnels would have twin lanes. At approximately 2.8 km, the tunnel is long by UK standards, but is modest compared with many existing road tunnels outside of the UK.
7.35 The tunnels are considered to be within the practical limit for longitudinal ventilation (less than 5 km) and a ventilation shaft/shafts is not considered, at this stage, to be necessary.
Secondly, the Cotswolds Escarpment at this location is quite steep and is recorded as being unstable. Therefore, ventilation shafts would be difficult to accommodate on the lower parts of the tunnel. Lastly, ventilation shafts would cause additional visual impact within the AONB.
7.36 The cross passages should not be used to ventilate the road tunnels in the event of a fire in the road tunnels. Each tunnel would require its own separate ventilation system. Some ventilation control equipment may be located in the cross passages.
7.37 It is assumed that the tunnels would have an excavated diameter of 14 m, similar to the Round Hill Tunnel at Folkestone. Some allowance could also be made for additional verge/walkway width, if required, on the slow lane side, as with the Southwick Hill Tunnel at Brighton. The excavation height would be about 10m. The cross passages are assumed to have an excavated diameter of about 4m and a length of less than 25m.
7.38 It is assumed that the tunnel bores have been assessed on the basis that no side ducts would be required to supply ventilation air and as such are giving bore diameters as follows:
- Two lane tunnel - excavated diameter 11.5m
- Two lane tunnel with walkways - excavated diameter 13.8m
7.39 The road tunnels would be designed to the requirements of BD78 (DMRB:BD78/99). However, the recent series of fires within road and rail tunnels in Europe has resulted in new recommendations from the UN. There is also new guidance as to the requirements for tunnel ventilation and smoke control due from the EEC. This guidance is yet to be published and, when released, could seriously affect the way that tunnel ventilation systems are designed. While it would be inconsistent with the requirements of BD78, consideration should be given to a local variation in the distance between adjacent cross passages.
Summary of Geology within the Tunnel Corridor
7.40 The published geological mapping suggests that the south-east portal would be excavated in landslip material and Fullers Earth. The tunnels would probably pass through the Fullers Earth at shallow depth and then encounter in sequence the Upper, Middle and Lower Inferior Oolite, Cotteswold Sands, Upper Lias, Marlstone Rock, Middle Lias and possibly the Lower Lias. The lower, 850 to 900 m and the north-west portal itself would be driven through a major landslip. Some landslip material and surface deposits (drift, scree and head) may also be encountered at both portals. Some weak organic material may also be present at levels within and below the lower sections of the slips. The limestones are generally more permeable than the underlying Lias and the contact is assumed to form the spring line.
7.41 Towards ground surface, the Oolitic limestones is recorded as becoming thin/very thinly bedded and having numerous vertical fissures. The upper 0.5 to 1.5 m is fractured. Bands of silty/sandy clay and rare, thin marl also occur. The fissures are not usually continuous from bed to bed. Some fissures were found in surface excavations to have an open width of 20 to 300 mm and an estimated depth of over 17 m. The open width tends to decrease with depth. The excavation of surface trenches with a backactor resulted in considerable overbreak. Cutting slopes 1:1 to 1:2 and rock embankment slopes of 1:1.5 are reported. The Upper and Lower Inferior Oolites are limestones. The Middle Inferior Oolite is a limestone with clay and sand.
7.42 As described in Section 3, the slopes below the Cotswold Escarpment and those located near the proposed south-eastern portal comprise landslipped material. Two forms of landslip may be encountered by the tunnel:
- Translation and rotational slides of limestone blocks due to slips occurring in the underlying Lias. These tend to form terraces between beds of limestone in the upper part of the slope; and
- Near surface translation slides. These tend to be derived from the Fullers Earth, Oolite, Upper and Middle Lias and to consist of limestone gravel, cobbles and boulders (some over 15 m in length) within unstratified clay, silt and fine sand. Polished shear planes are frequent in the cohesive materials. 'Experience indicates that all these colluvial deposits, particularly where more clayey, would be interlaced in many directions by pre-existing slip surfaces at residual shear strength and that their existing margin of stability is generally low. Thus, quite modest disturbances, in the form of cuts and fills for example, can be expected to lead to a reactivation of slip movements' (Hutchinson, 1991). Slips have occurred in this area following periods of heavy rainfall and where shallow cuttings have been made. 'In a more stable part of the area, a cut of 3 m or more in depth may not produce a significant failure, whereas in a less stable part, a cut of as little as a metre depth could cause a landslide' (Hutchinson, 1991).
Construction Issues
7.43 The portals are some 3 km apart and the roads between them are not considered suitable for continual movement of construction plant. Other than for an initial section of clays, the upper 1/3 of the tunnel would be excavated in hard limestone rock and the lower 2/3 section of the tunnel would be excavated in clays and silts and, probably, landslipped material. The methods of excavation and support, and the plant to be used for these two tunnel sections are likely to be different. Each portal/tunnel section should therefore be considered as being in effect two separate sites.
Brockworth Portal
7.44 The tunnels are likely to encounter landslip material (consisting of gravel to boulder sized pieces of limestone within unstratified clays, silts and fine sands) at the Western portal and probably in the first 870 m of the tunnel drives. There are several private properties and public roads built on the landslip material. Several streams and springs are also present. It should be assumed that the landslips are in a marginal state of equilibrium and the portal should be designed such that the least possible ground deformation occurs during construction.
7.45 It is likely that the two portals would be constructed in a similar way to the UK portals of the Channel Tunnel, which were excavated through a glacial landslip on the side of Castle Hill (Harris, Hart, Varley and Warren 1995) and the Serre la Voute Road Tunnel, Turin. The slips at Castle Hill were assumed in the design to be only marginally stabile. Following a major site investigation; which included detailed orientated core logging, instrument installation (surface survey points, inclinometers and piezometers) and an exploration tunnel, substantial toe weighting was placed over the toe of the slips to increase their apparent stability. The portals to the three tunnels at Castle Hill were constructed within a box with end walls and side-walls, formed of bored piles. Tension piles were also installed through the basal slip plane below the subsequent floor of the box to prevent heave due to the excavation of the ground within the box. The roof slab was cast on top of the piles at ground surface. The ground within the box was then excavated in benches from the top down; the excavated material being deposited above and around the box to limit the net change in ground load on the slip planes caused by the excavation. The tunnel eyes were made through the piles in the end walls, while the tunnels were driven towards the portal from the opposite side of the hill. Drainage galleries were subsequently installed to further increase the long term stability of the slipped mass. The lining of the running tunnels was built in 5 m long, reinforced sections through the slips. These allow minor movements along the slip to continue in the long term, without damage to the lining.
7.46 There is little information on the location and current stability of the landslipped materials below Crickley Hill. The slipped mass at Crickley Hill appears to contain more soft, non-cohesive material, flowing water and groundwater standing (although this is very variably) within 1 m of ground surface. The material would appear to be more sensitive to shallow excavation than those encountered at Castle Hill and it would, probably, be necessary to lower the water table by pump wells prior to portal excavation. Moving the portal further to the north-west could put the portal in the Lower Lias, outside of the slipped materials, but would also increase the length of the tunnel within the water-bearing silts of the Middle Lias.
7.47 Although a site investigation would be required, slip planes within clay materials are very difficult to locate and to evaluate and, therefore, it should be assumed that the slopes have only a marginal stability and require additional long term instrumentation and stabilisation measures. The topography, scenic location and diversity of the slips would probably not allow wide scale toe weighting. However, deep drainage may be a viable solution. In addition, it may be possible to install gravity drainage in the Lias Clays, without significantly affecting the overlying springs seen within the landslipped material. Where the slope drainage does effect the surface springs, it may be possible to use the drainage to re-supply water to the streams and to any properties currently abstracting from well or spring sources.
Cowley Portal
7.48 The geological mapping suggests that the southern portal and first 100 m of the tunnels would be excavated in a shallow landslip, probably formed of weathered Fullers Earth. The slip may only effect the portal and first few metres of the tunnel, but its apparent direction of movement (approximately 45º to the proposed axis of the tunnel) complicates the design and it would be preferable either to move the portal or to remove the slip material in an open cut.
7.49 It is probable that the portal would need to be excavated by first constructing a bored pile, or other structural head wall, to secure the ground at the tunnel face and then excavating the portal in open cut with protected (probably shotcreted and drained) side walls. Excavation with normal surface excavation plant would probably be adequate until the limestone is reached. A single cutting pick and a de-mountable impact hammer should also be provided to assist in the excavation of any rock bands and the underlying limestone until a full face of secure limestone is established.
7.50 The low intersection angle between the Oolitc Limestone and the over-lying Fullers Earth would probably result in some blasting as the limestone rises in the lower face. Continuous lattice arch, shotcrete and rock dowel support are likely to be required and some spilling; particularly close to ground surface, is likely. The Fullers Earth is probably an aquaclude and the dip of the underlying, permeable, limestone may result in initial sub-artesian water inflows from the limestone. Forward and downward probing would be necessary as the contact is approached in order to assess likely water ingresses.
Pilot Tunnel
7.51 Most of the tunnel alignment would encounter the Upper Lias clay and the Inferior Oolite limestone, but some Middle Lias, a weakly cemented silt, may also be encountered towards the lower end of the tunnel; particularly if the portal is located out of the landslip, within the Lower Lias. The ground pressure (from approximately 100 m of overburden and 50 m of groundwater) is likely to be very much greater than the mass strength of the ground and squeezing conditions can be expected in the Middle Lias. Prior drainage through a pilot tunnel would increase face stability during the excavation of the larger road tunnels; although drainage could also result in ground shearing and this would require careful consideration during the detailed design. It is considered at this stage that a pilot tunnel should be excavated in advance of the excavation of the portal and the breakout of the road tunnels from the portal. While this tunnel could be excavated with a roadheader and supported with lattice arches and shotcrete in the Lias, the consequences of a face collapse in this critical area close to the main basal slip plane are considered to be too high for this method of excavation and, therefore, it is recommended that the tunnel is excavated with a road header mounted in a shield and supported with a one-pass, bolted, segmental concrete tunnel lining. Ports in the lining should be provided to permit drainage holes to be drilled towards the road tunnels and cross passages. Filters may be required to prevent the silt from washing into the pilot tunnel.
7.52 The north-west portals would take some time to establish and it is recommended that the pilot tunnel should be driven from the base of an adjacent shaft while the portal box and stabilisation work are being undertaken. The shaft should be lined with concrete segments and the ground/lining annulus grouted to restrict ground movement. The stability of the shaft floor during excavation (possible boiling) needs to be considered in the detailed design. The shaft would be replaced by a portal as part of the permanent drainage of the slope.
Oolitic Limestone
7.53 Excavating down grade would result in the flow of any groundwater infiltration, rain or construction water towards the face, from where it would need to be pumped out to permit face excavation. This is not considered to be a critical factor; at least for the initial section within the Oolitic Limestone. Excavation up grade would result in gravity drainage away from the face towards the Brockworth Portal and is, therefore, preferred for the softer Lias deposits.
7.54 Groundwater is anticipated in the oolitic limestone and is likely to flow laterally along the sub-horizontal discontinuities; particularly above low permeability beds such as marl bands, and vertically along the open fissures. Clay beds and clay filled fissures (possibly including faults) are likely to act as aquacludes, resulting in perched water. The Upper Lias can be considered to be an aquaclude, such that the groundwater within the overlying limestone is likely to impose an additional pore water pressure within the Lias and landslides. Excavation of the pilot tunnel is likely to reduce the pore water pressure within the Lias prior to the excavation of the larger road tunnels, although as a result of relative difference in permeability, drainage of the Lias may only result in a marginal benefit.
7.55 The bedding is sub-horizontal (+/-50 dip towards the south-east, Hutchinson, 1991). Frank Graham's 1989 Crickley Hill Tunnel Study, includes a geological section; which shows the strata to have an apparent dip of about 20 towards the south in the plane of the tunnel section, which agrees with the published geological mapping and data presented in Gloucestershire County Council's 1989 Geotechnical Feedback Report. The bedding would, therefore, dip towards the face where the tunnel is excavated down slope from the south-east towards the north-west. This has a positive benefit; in that a forward probe hole would be more likely to intersect any beds of water bearing or soft strata before they appear in the tunnel face. Such strata would first be encountered in the lower part of the face, rather than in the roof and would be more likely to drain before intersecting the roof. There is a risk in excavating the face in the up slope direction that any water bearing strata or beds acting as aquacludes to water bearing strata above, could be undetected by the probe holes and approach the roof at the face above the section of tunnel previously excavated. A face collapse could, therefore, run back for some distance (given the low strata/tunnel intersection angle) from the face.
7.56 The Brockworth Portal and landslip additional stabilisation works are likely to take longer to construct than are the Cowley Portal and tunnel excavation within the Fullers Earth. The Oolite limestones are also likely to be easier to excavate and support than the Lias deposits. Excavation of the two road tunnels at the upper end is, therefore, likely to be well advanced before excavation of the road tunnels at the lower end begins.
7.57 The limestone is probably too strong for excavation with roadheaders. Impact hammers could be used, if the bedding and joint planes are close enough for the hammer to breakout the intervening blocks. The most likely method of excavation would be by 'drill and fire'. The sub- horizontal bedding in the limestone is likely to result in a flat, stepped longitudinal roof profile. The sub-vertical jointing is likely to control the lateral profile; blocks falling from the shoulders and walls. A stepped, square profile is likely to result. Tensioned rock bolt support with local mesh is anticipated. Lattice arches and shotcrete are not generally suitable for these conditions but would be required to some extent to support broken, faulted and clay/marl ground. Arches are designed to deform under load and this can result in long term problems (for example, Lewes Road Tunnel in chalk rock). Some face grouting may be required to seal any fissures giving large water inflows, but this is unlikely to be a major problem given the proximity of the tunnel to the escarpment edge. A D-shaped tunnel profile with a flat invert is recommended.
7.58 As the Crickley Hill Tunnels would be permanently lined with a concrete lining, it is reasonable to reduce the support to that suitable for the temporary condition. Any instability would be dominated by block falls and therefore the rock bolts and arches need to be maintained, whereas the amount of shotcrete can be reduced. Under normal conditions it is anticipated that the upper heading can be advanced in 3 m rounds, the bench in 3 m+ rounds to suit the contractor's operational requirements. Most of the support in the heading would be provided by 3.5 m long (4.5 m long for the three-carriageway tunnel), tensioned rock bolts at 2 m centres. Some, 100 mm thick, mesh-reinforced shotcrete would be required where the ground is locally fractured. Elsewhere, open mesh, supported by rock bolts, would be sufficient. Some roof support; consistent with safety would be required close to the face. The rock bolts should be retensioned and the remaining support installed within 20 m of the face. Little, if any support would be required in the bench walls.
7.59 Slightly longer rock bolts of 4 to 5 m at 1.5 m centres, 100 mm thick, mesh-reinforced shotcrete and shorter advance lengths of 1.5 m may be required in the heading in the areas of more fractured ground. Some colliery arches would also be required. Most of the support would need to be installed at the face and should be completed within 10 m of the face. The arch legs would need to be extended after bench excavation. Advance lengths of 3.0 m+ would still be possible in the bench.
The Upper and Lower Lias
7.60 As with the Fullers earth, the Upper and Lower Lias clays could be excavated with either a roadheader or with surface excavators (Volvo EC series or similar). A demountable impact hammer and single cutting tooth should be provided to excavate any rock bands. Some explosives may be required to excavate hard bands of rock intersecting the invert and roof. The upper heading would probably need to be excavated in two sections with a temporary invert arch. Bands of sand would require foreprobing and possibly glassfibre face dowel support. A curved invert arch would be required. The detailed design should consider the excavation sequence and support requirements in detail.
7.61 Instrumentation in the pilot tunnel should be used to verify the excavation and support design for the larger road tunnels. Instrumentation in the road tunnels should be used to verify the design of the lining.
7.62 Taking the example of Castle Hill, it is initially considered likely that the tunnels would require 200 mm of fibre or mesh-reinforced shotcrete, lattice arches at 1.5 m centres, 4 m long rock dowels at 1.5 m centres and 4 m long spiles across the roof at 0.5 m centres. The face should be protected between each advance with shotcrete. The heading would require a temporary invert arch. The invert arch would require to be closed to prevent continued tunnel deformation.
The Middle Lias and Cotteswold Sand
7.63 The weakly cemented, water-bearing silts of the Middle Lias and Cotteswold Sand are likely to present the most challenging tunnel excavation problems of all of the ground types on the project. Partial face excavation, continuous face support and overlapping roof support are probably required. The intersection lengths are probably too short to make a Perforex solution viable, unless this system is used to excavate all of the none-hard rock sections.
7.64 These are both weakly cemented, probably water-bearing, sandy silts. Some strong, massive, medium to thickly bedded limestone bands may also be present. The in situ condition of these silts needs to be determined by a site investigation. Even a small amount of cohesion would have a positive effect on face stability and hence in the speed of construction and amount of support needed.
7.65 Assuming that the unweathered material at tunnel depth is encountered as a weakly cemented silt/sand, it is probable that the heading would need to be advanced as a partial face with a central buttress, under the cover of an umbrella of grouted or jet grouted pipe spiles. The heading would require an arch of at least 200 mm thick, mesh-reinforced shotcrete, supported by 'elephants feet' or jet grouted foundations. A temporary invert arch may be necessary. The face would need to be supported with shotcrete between each excavation cycle. Lattice arches and rock dowels would be required at 1 m centres.
7.66 The bench would need to be excavated in tandem with the heading to allow early closure of the invert.
Cross Passages
7.67 Cross passages would be excavated within each of the rock types. Wherever possible, the distance between adjacent cross passages should be reviewed where this requires their excavation within the Middle Lias or Cotteswold Sand. Where it is unavoidable, a method of excavation and support similar to that for the road tunnels is required.
7.68 Those cross passages eyes within limestone rock are likely to be excavated concurrent to the adjacent road tunnel bench. The subsequent rounds could be drilled while the bench (at some point remote from the cross passage being extended) is being mucked. The method of support would be similar to that for the road tunnels. Arches should be located such that the adjacent cross passage can be excavated without removing the arch.
7.69 Cross passages within the Upper and Lower Lias, and the Fullers Earth would probably be excavated by hand tools, unless the contractor has a small roadheader available.
Tunnel Ventilation Options
7.70 The general parameters for ventilation design would be taken from Department of Transport Existing Standards pending the receipt of revised requirements from the EEC, which are understood to be issued imminently.
7.71 There are three options for the ventilation of the tunnel, these are transverse ventilation, semi-transverse ventilation and longitudinal ventilation. The systems differ as follows:
Transverse Ventilation
7.72 Transverse ventilation comprises a supply and extract system with graded inlet and outlet jets over the tunnel length which supplies and extracts air over the whole length of the tunnel. This system requires both supply and extract ducts over the length of the tunnel bores. As a result of the length of the tunnel and the required ventilation air volume, vertical ventilation ducts would be required at points along the line of the tunnel. The spacing of the vertical supply and extract shafts is governed by their possible locations and a requirement to limit the dimensions of the bore. Therefore, if shafts are provided at the lower level and at the upper level the necessary duct capacity can be quartered providing that a uniform spacing is feasible. This would mean that with the proposed tunnel alignment one of the shafts would need to penetrate the landslip area at the lower end of the bores. The location of the other shaft at the upper end is not so critical as the ground is more solid.
7.73 The shafts would contain input and extract fans separate to each tunnel bore. Generally the airflow would be side, inlet and crown outlet so that fresh air would be introduced at low level and exhausted at high level. This system tends to ensure that combustion products in the case of a fire are induced upwards and are exhausted to leave the lower occupied areas of the tunnel clear of smoke and products of combustion.
Semi-Transverse Ventilation
7.74 In this option the air generally enters through the tunnel portals and is uniformly exhausted along the length of the tunnel. Similarly to fully transverse systems ventilation shafts would be required, but these would be extracted only with air being extracted at high level. Generally this system would maintain clear air at low level in the case of fire, though in this case the airflow would be from the portals to the point of extract. This system may require reversible fans to control smoke in the case of fire.
Longitudinal Ventilation
7.75 In this option air is moved unidirectionally, generally in the same direction as the traffic flow, by jet fans mounted at high level in the bores.
7.76 This system requires no shafts but the airflow is generally in the same direction as the traffic flow. In case of fire, this would tend to move smoke and products of combustion away from the source in the case of a vehicle fire. On the assumption that traffic would stop and that the traffic in front of the fire would exit the tunnel, this should provide reasonable smoke clearance. Sufficient redundancy is required to be provided to cover failure and maintenance plus the possibility of a fire affecting the fan operation. Generally such fans are capable of operating in temperatures of up to 300°C and can be reversible to permit the tunnel to be operated in either direction under normal operation. However, bearing in mind the slope of the tunnel, a fire in the downgrade bore could permit smoke to travel against the direction of traffic flow.
7.77 As a result of the geological instability of overlying rock and the environmental sensitivity of the area, options requiring ventilation shafts would not be taken forward.
Alternative Option
7.78 There is the possibility for a compound arrangement of semi-transverse ventilation with jet fan assistance. In this option the jet fans may be used to create opposing airflow in the case of fire and hence confine the smoke to a prescribed area of the tunnel where it may be extracted by the exhaust system. The disadvantage of this option is that ventilation shafts are still required, but potentially more effective smoke control may be maintained particularly in the downgrade tunnel.
Fire Fighting
7.79 Consideration would need to be given for the provision of fire hose reels and fire hydrants over the length of the tunnel bore. The availability of a water supply bearing in mind the level differential between lower and upper portals and the extent of the local distribution may possibly require the installation of a reservoir and pump near the southern portal.
7.80 The local water reticulation at the upper area (Cowley Portal) may be adequate.
7.81 Hydrant and hose reel spacing in each tunnel bore would need to ensure that there is overlap in prospective hose lengths so that more than one hydrant or hose reel may be applied to a fire and to allow for the fact that facilities may be in accessible due to the fire. The provision of hand held equipment need also be considered.
Drainage
7.82 Internal roadway drainage would need to be considered within the overall design and section of the tunnel. The provision of gullies at the tunnel entrance and exit would reduce the quantity of water carried into the tunnel leaving only driven rain through the portals and such as may be carried in be vehicles. However, drainage would be required in the tunnel bores to allow for the disposal of fire fighting water, foam etc. As the tunnel is used by vehicles, fuel and oil entrapment would be required prior to discharge.
Engineering Risk Evaluation
7.83 An engineering risk evaluation has been undertaken and is presented in Appendix C.
Highway Engineering
7.84 A detailed description of the three option taken forward for assessment is provided in Chapter 6. However, for ease of reference, brief descriptions of the Options are included below.
- Option 1 - Dual carriageway along the route with a traffic signal controlled layout at Air Balloon Roundabout and an additional lane added down Crickley Hill
- Option 2 - Dual carriageway from Cowley Roundabout to the bottom of Birdlip Hill, a grade separated layout replacing the existing Air Balloon Roundabout and an additional two lanes down Crickley Hill.
- Option 3 - A twin bore tunnel between Nettleton and the bottom of Crickley Hill. The existing A417 would remain but would probably be de-trunked.
7.85 A comparison of the construction issues and general engineering issues associated with the different options is included in Table 7-1 and Table 7-2.
7.86 The options have only been developed to the stage where outline design elements of the proposed routes have been identified. The detailed design of the scheme has not been undertaken and therefore the comparison of engineering issues is based on the preliminary information currently available.
7.87 As part of the examination of the potential options, consultation has taken place with the local Police and Fire Services to gain their views. Further details are provided in Chapter 11. However, the key issues raised by each are summarised below.
7.88 Gloucester Constabulary have formally expressed a preference for a conventional 'open air' highway layout. This is because of the large financial and operational burden to the service of a tunnel. They also expressed an opinion that the length of the tunnel would see many motorists choosing to use the alternative 'open air' routes through fear of the unknown.
7.89 Gloucestershire Fire Service have also formally expressed support for a surface option. They have expressed concerns regarding public safety and the risk to fire-fighters as well as the potential cost of providing dedicated services at the tunnel.
7.90 As part of the assessment of the Options, a preliminary safety review was carried out on Option 1, the at grade option with traffic signals, because of the safety issues associated with traffic signals on steep gradients. The review identified a number of safety issues that are particular to this Option.
7.91 The downhill gradient, of generally 5% or 6% on Birdlip Hill would encourage high approach speeds to the signals, particularly as they would need to be visible from some distance. This would give rise to overshoot and rear shunt type accidents. The traffic bound for the A436 would need to move out to the offside lane to go 'straight ahead', which is likely to give rise to further rear shunt type accidents.
7.92 The link road between Birdlip Hill and Crickley Hill would have a downhill gradient of about 9% leading to the left-hand bend to join the existing Crickley Hill. The gradient, although in character, would be well below desirable standards. The bend would also be well below desirable standards, particularly so after the steep downhill gradient and could lead to loss of vehicle control and HGV's overturning.
7.93 The traffic signals would require a speed limit, preferably even lower than the current 60 mph, however compliance is likely to poor and enforcement unlikely. In itself a lower speed limit is unlikely to reduce the problems to any great extent, as even 30mph on the link road could be excessive for some vehicles.
7.94 The combination of geometric features present with Option 1 is likely to result in a hazardous road layout. This cannot be overcome by careful detailed design.
