You are:

+ Share | Print page

M4 in Wales. Castleton to Coryton (J29 to J32)

This 7.9 mile (12.7km) section is covered more extensively as its preparation took place as the apparent public wish for motorways was changing, and objections were rapidly increasing in number and content. It also covers the more general problems encountered, which are applicable to many of the other motorway lengths, and can be taken as representative of the attitudes and problems from both administrative and engineering standpoints.

At the end of World War II Glamorgan County Council had located a route for a new trunk road north of Cardiff, safeguarded by the then County Surveyor before the war. It was later rescheduled as a motorway but was regarded as a low priority section. As an interim measure traffic was to be routed via the A48 Eastern Avenue Bypass, completed in 1971, and a proposed Llantrisant Radial Road.

It was thought that the motorway would not be needed before the late 1980s but the Llantrisant Radial Road involved much demolition. Consequently the Secretary of State for Wales announced in 1971 that the M4 would be built, and the Llantrisant Radial shelved.

Consultant Howard Humphreys and Partners ceased work on the Radial, and commenced study of the M4 between St Mellons in the East and Coryton in the West. The protected line had for a long time been regarded as a limit to the urban development of Cardiff, whereas a more northerly route allowed expansion of the city. The City Planning Officer pressed hard for the change. Three possible lines then emerged north of Lisvane, and because of the importance of the issue Shankland Cox and Associates were appointed as Planning Consultants. Two other studies were commissioned. One in 1972 when UWIST (University of Wales Institute of Science and Technology) was asked to investigate traffic noise levels, and another in July 1972 when Brian Clouston and Partners were engaged to provide advice on Landscaping.

As a consequence of the detailed studies the most northerly route in the area of Lisvane was decided upon, adding about 4km to the overall length. The route was confirmed in 1975 following a lengthy Line Order Public Inquiry held in late 1974.

A Wimpey Fairclough(AMEC)consortium was awarded the construction contract and started work on the 21st October 1977. The road was opened to traffic on the 10th July 1980 after a total period of 9 years from commencement of initial studies to completion. This was a remarkably short time period, and resulted in a visit to Wales by a special committee set up to try to establish a system to speed up the overall programme for highway schemes, throughout the UK.

Public Consultation

The public was consulted at an early stage when the two routes selected for final analysis were presented at a public exhibition. This was before more formal ways of consulting the public were established. The exhibition was very well attended and the public encouraged to give their views and advice. A lively and useful interest was generated throughout.

The detailed treatment of the preferred route attracted considerable comment on: landscape impact; noise and intrusion (in particular the effects on the tourist attraction of Parc Cefn Onn); the danger to Wenallt reservoir immediately adjacent to a proposed deep cutting and near properties; and severance effects generally (in particular on Grade 1 land).

In general the exhibition was regarded as a successful exercise and alerted the Welsh Office and its consultants to points of objection likely to be encountered at a Public Inquiry.

The Public Inquiry

The public Inquiry was held into the main line scheme, side road orders and a variation order for the Tredegar park to St. Mellons motorway spur.

The CPO included extra land to make up a deficit of fill material on the adjacent Coryton to Miskin scheme to the west, whilst reducing surplus on Castleton to Coryton itself. It also facilitated the building of an interchange on the adjacent Tredegar Park to St. Mellons spur to the east. A Variation Order was needed to authorise the construction of that interchange which joined the A48M link to the M4.

All the Orders were published by the Secretary of State for Wales, on the 16th of April 1974 and attracted 176 objections. The Inquiry lasted from October until a few days before Christmas, the longest period experienced by the Welsh Office up to that time (and later to be exceeded only by the A55 Line Order Inquiry in North Wales). It indicated clearly the growing public awareness and strength of objection to major projects.

It was the most contentious of the M4 lengths in South Wales. The objectors were largely residents defending their properties rather than the more general objection of later years concerned with need.

Some objectors did support the alternative Eastern Avenue - Llantrisant Radial through route however, which also became an issue on the Coryton Miskin section. The arguments were less sophisticated than at later Inquiries, and a simple balancing of properties threatened on the competing routes.


During investigations leading up to the adoption of the preferred route, some objectors to the scheme described fogs and mists so dense and frequent as to provide serious dangers for fast traffic. In response regular observations were made starting on the 1st of January 1973. No evidence was found to support objectors claims. On the contrary the more northern route was the least susceptible.


All the general standards of the time were adopted, and statistics later showed that the dual 2 lane Castleton to Coryton section experiences about half the national average accident rate for motorways. This is very good as dual 2 motorways have been shown to be more hazardous than dual three.


Engineers were quite accustomed to considering environmental issues at this time, and had involved architects and planners. The engagement of environmental and ecological specialists is a more recent activity.

The concerns at that time were location, alignment, landform, landscaping and noise and visual intrusion. Experts were engaged as described earlier, and all helped in the presentation of the Public Exhibition of June 1974, and at the Line Order Public Inquiry.

Of the measures finally decided upon, one resulted from public comment requesting the construction of bunds at the top of embankments at Lisvane to reduce noise and visual intrusion.

The Landscape Advisory Committee (LAC) also had a major input as they did on all such schemes. As part of their advice the LAC commented on the sensitive issue of Parc Cefn Onn, and its access severed by the motorway line. The M4 is in shallow embankment as it crosses the extreme southern boundary of the park, then crossing over the Rhymney Valley railway line onto another short length of embankment. The Committee were concerned that the motorway embankment should not appear as a prominent feature when viewed from the south and east sides of the Park, and their recommendation to reduce the level was accepted.

The Committee also gave advice on special fencing and their opinions were accepted. This included proposals for sympathetic planting around the Wenallt Reservoir to soften and fade the new construction into the landscape.

On the east side of the Rhymney Valley split level carriageways were adopted, differing in level by a maximum of 1.6 metres. This not only reduced earthworks but also helped to fit the M4 into its surrounding landscape, whilst giving westbound motorists better views than otherwise would have been the case. In many instances of shallow cutting and fill the engineering slopes were flattened to remove the artificial appearance, and to return the land at the back of the motorway to agricultural use (a process known as daylighting).

Finally it was decided to grass side slopes under bridges over the motorway, to provide visual continuity. French drains with no outlet were used to keep the grass watered, and a robust shade and salt tolerant seed was used. The outcome has not been as successful as the designers wished, but an opportunity for remedial work will occur when widening takes place.

A major cutting at Lisvane about 5 Km east of Coryton was 14 metres deep and the 'notch' like appearance created was made worse by the 1 in 3 northern slope required. The 'notch' was closed effectively by the construction of an arch bridge carrying Graig Road over the cutting. The road could have been diverted with some cost saving, but the landscape advantage tipped the balance.

Extensive consideration was given to appearance and the amelioration of environmental hazards throughout the scheme length. Damage was kept to a practical minimum. Only twelve dwellings were demolished, a modest number considering the scale of the works, whilst other homes were protected from noise by double glazing, and from noise and visual intrusion by acoustic earth barriers.

Geology and Earthworks

The Geology of South Wales is varied and surprises can await the engineer unless he carefully records all the information he can, including local opinion. The geology of the underlying rocks is well recorded. Bedrock along the Castleton to Coryton Section consists of the Lower Old Red Sandstones of the Lower Devonian series as the route skirts along the mouths of the valleys. These are subdivided into three main stratigraphic units:

  • Rhaglan Marl Group, outcropping in the Rhymney Valley;
  • St Maugham's Group, outcropping over the remainder of the route;
  • Brownstone Group, with outcrops confined to the west of the route.

The first and second groups comprise marls, sandstones, and limestones, the sandstone characterised by the addition of coarse interbedded conglomerates, the most notable being the Llanishen Conglomerate. The Brownstone Group consists mainly of red brown sandstone and subordinate marl. All are gently folded and show generally north east to south west axial trends. The dominant fold is the Usk anticline intersecting the route on the eastern part of the Rhymney Valley

The rocks are overlain by unconsolidated glacial deposits, ranging from clays to gravels and termed Moraine. The largest deposit of alluvium occurs in the Rhymney Valley where terrace gravels overlie morainic drift.

The geological information was supplemented by soil investigation along the line of the road. The preliminary site investigation by Glamorgan County Council Materials Division was carried out between September 1971 and March 1972. 51 boreholes were sunk close to the probable location of structures.

The results of this preliminary investigation were used to prepare the specification for the main investigation which was carried out by Nuttall Geotechnical Services in August and September 1972. 75 boreholes, 13 trial pits and four trenches were sunk. Block samples were also dug from trial pits by Soil Mechanics Ltd, and tested in a large shear box to determine shear strengths and discontinuities. Piezometers were also installed in 25 of the 75 boreholes to measure water levels in the various horizons.

A supplementary survey of the Wenallt Reservoir area was carried out in June 1973 in collaboration with Edgar Morton and Partners. Eight boreholes, and twelve piezometers at selected levels in the boreholes, monitored ground water conditions.

Assessment of the main investigation resulted in further work in the Lisvane area, in order to confirm the dip of the marl bedrock.

Such recordings of geology, and a similar degree of site investigation, was carried out on all lengths of the M4 to determine the likely conditions cuttings were excavated and foundations dug. It also determined the suitability of materials as fill, and the extent of unsuitable material to be disposed of.


With the information provided by the geological and soils studies, earthworks were generally straightforward and in many places allowed steeper slopes than the standard 1 in 2. Some slopes and required slackening to provide stability however, so that a variety of side slopes resulted. The total volume of excavation was approximately 1.8 million cubic metres, and of this about 1 million was placed in embankments and noise bunds. The remainder was removed with the Contractor and farmers agreeing to its disposal off site.

The main event of interest was a slip at Wenallt during construction, where the M4 is in a 16 metre deep cutting passing between housing to the south and a 15 million gallon reservoir adjacent to the northern boundary . The reservoir was roofed with a reinforced concrete floor and mass walls. It had a chequered history, with foundation problems before and after construction. Uplift pressures under the floor, due to groundwater, had been remedied by the installation of a secondary floor and under- drainage in 1965.

Local residents had been anxious at the prospect of the cutting, and voiced their concern at the Public Inquiry. The information gathered by the Consultant, including information from papers published in the 1920s, had been extensive however. The north face of the motorway cutting was designed with a 1 in 2 slope in the reshaped reservoir spoil and the moraine and bedrock, but with a berm at the original ground level. The top of the cutting was 65 metres away from the reservoir wall, and in order to draw down the water table as cutting took place, thirty relief wells were installed, 14 at the berm and sixteen at the toe.

Despite careful monitoring, routine inspection on the 26th October 1978 when the cutting was almost completed, revealed a crack in the slope adjacent to the reservoir. It was opening at 2 mm per hour. Excavated material was placed at the toe of the slip whilst investigations took place, and the top of the cutting was trimmed back to ensure stability.

Investigation showed that the slip was due to an area of folding in the marl bedrock, between reservoir and carriageway. It was concluded that there was no risk to the reservoir. The slipped material was therefore removed, and replaced after the installation of a drainage layer. Slope drainage and added relief wells were installed in both faces of the cutting.

Geometric Standards

Full standard 120kph geometric design and visibility was adopted throughout, except at overbridges where a reduction in sight distance was permitted to achieve an economic layout. In all cases sight distances exceeded the minimum quoted in Tech.Memo.H/71.

In some cases verges were widened from 4.9 to 7.6 metres and the central reserve increased to 8.5 metres however, to obtain the full 300 metre visibility required.

The main straight gradient is at 3.5%, on the west side of the Rhymney Valley over a distance of 600 metres. The maximum gradient on a curved section was 3% over 200 metres however. Elsewhere gradients are less than 3%, except for one instance at the split level section where 250 metres of the westbound carriageway is on a down grade of 3.3 %.

Hard Shoulders were built to full carriageway construction, and the pavement designed to Road Note 29.

Bridge design

Generally, the side road overbridges were standardised in the form of a two span voided in-tensioned concrete. The Rymney River underbridge is of pretensioned U beam construction with a composite deck. Elsewhere a variety of structural forms was developed to accommodate particular site requirements where the standard forms were inappropriate.

M4. Craig Road arch bridge

In most cases it was possible to found on conventional spread footings at moderate depth, and this method was chosen first. Piled foundations were necessary at the Thornhill, Heol Hir and Rhymney Railway overbridges, however. In certain instances spread footing loads were transferred to load bearing strata using plum concrete.


The Welsh Office decided that some allowance should be made to facilitate widening to dual three lane in the future. The provisions asked for were that:

  • all overbridges should be increased in span to accommodate dual 3lane carriageways;
  • the widths of underbridge decks should be increased by 1.7 metres;
  • all hard shoulders should be constructed to full pavement depth.

Widening would be symmetrical and on the outside of the carriageways, except for two kilometres at the western end. Over the main length the 8.5 metre central reserve made it possible to add a third lane to the north side of each carriageway. The resultant reduction in the central reservation would shorten sight distance but would enable widening to be achieved, without the loss of a hard shoulder adjacent to the retaining walls on the south side of the motorway. These walls also function as abutments at the Rhiwbina Hill and Wenallt Road overbridges.


Two intermediate interchanges were considered but abandoned. The first was with the existing A469 but was removed as the Buchanan plan for Cardiff was abandoned. The second was to be the terminal of the proposed Pentwyn Link which formed part of the (Cardiff) Peripheral Distributor Road, the PDR. This was constructed later by South Glamorgan County Council however, in their dual carriageway Link Road contract, joining the M4 to the A48 Cardiff Eastern Bypass. This was financed by Private Developers as part of a very large scale housing development, with all preparation, design, and construction supervised by Engineers from the County Council (Ref B6-2).

Flood Study

The run off calculated for each section of the system was used to determine the size of the culverts required under the motorway, and to assess the effect on water courses and existing drainage systems. Calculations showed that the provision of culverts would have no significant effect on watercourse flows.

However, Cardiff City Council had observed flooding caused by run-off from other highway schemes in the Cardiff area, and cited examples where proposed culverts below the motorway were larger than those existing downstream. They were concerned that the M4 Castleton to Coryton scheme might also cause problems. The Welsh Office therefore instructed Howard Humphreys and Partners to carry out a flood study to identify and remedy any deficiencies. The study covered all the watercourses crossed by the Castleton to Coryton motorway section, and especially Nant (Welsh for brook) Cwmnofydd, Nant Glandulas and Nant Pontprennau, about which the council had expressed particular concern. The study showed that flooding arising from the motorway was highly improbable, and that generally the motorway run off would be rapid, and take place before contribution from the catchment areas also discharging through the motorway culverts.

The Contract

In September 1977 a contract for the construction of the Castleton to Coryton section of the M4 was let to the Wimpey-Fairclough consortium, in the sum of £10,519,000. This was substantially below the Engineer's November 1975 estimate of £15.000,000, based on motorway rates applicable at that time.

Although a point of considerable concern during the consideration of tenders, the difference probably reflected market changes which had taken place in the interim, and underlines the discussion on uncertainty in making assessments.

Wimpey Fairclough was also the Contractor for the adjacent Coryton to Miskin section, completed in December 1977, so that they were able to move seamlessly from one contract to the other. The extensive site offices were also well located for both contracts. Establishment costs were therefore minimised, and with the local knowledge already gained were probably major factors enabling the Contractor to undercut the next lowest tenderer by £1.5 million.

This supports the claim by contractors, covered in the earlier volumes, that there were advantages to be gained in serial contracting. This was not proven in the case of the Coryton to Miskin and Miskin to Pencoed tenders discussed later however. On these lengths contractors were invited to price the sections separately or as a combined contract. The lowest price was for two separate contracts.

There was also an advantage for the Client. There was a tacit agreement between the Welsh Office, South Glamorgan County Council and Howard Humphreys and Partners, that as far as practicable as many of the County's Resident Engineer staff from the Coryton to Miskin section as possible, should be employed on the Castleton to Coryton Section. Thus, whilst the Resident Engineer(RE) was a member of the Howard Humphreys permanent staff, the Deputy RE and two Section Engineers were recruited from South Glamorgan. Overall nearly half the Resident Engineer's staff were drawn from the County's site.

The works were started by the Prime Minister, Mr. James Callaghan, on Friday 21st October 1977, immediately after he had opened the adjacent Tredegar Park to St. Mellons section to traffic. In stripping the first sod it was planned that the Prime Minister, instructed by a Wimpey operator, would drive the bulldozer only a short distance. However his enthusiasm was such that he continued for well over a hundred metres, and almost disappeared over the crest of a slight hill in the field with security men in hot pursuit.

Problems with Precast Concrete Bridge Beams

Two of the major bridges, over the Rhymney Valley Railway near Parc Cefn Onn, and over the River Rhymney, were standard pretensioned U beams with slightly modified tops. These beams were 1.5 metres deep with lengths varying between 17.5 and 30 metres. They were manufactured off site at a casting yard in Bridgend, some 12 miles away.

The first indication that something was wrong came after a number of beams were in place. It was then observed that on two of them the tops were too close to the arise at one or more points. Initially it was not thought to be serious as the beams were well finished and there were no other apparent defects.

At the beginning of March 1979 a survey was carried out using a cover meter to determine the extent of remedial works, and the area of inadequate cover was greater than expected. More beams were inspected and it became clear that there was a very serious problem. While the minimum cover specified was 30 millimetres, the survey revealed that 50% of the beams had areas of less than 20 mms, and in a few places 5 to 10 mm.

Considerable discussion took place to look for acceptable remedies, with a shortage of time because of a requirement to meet a railway possession for launching the beams. In mid March the contractor proposed systems of repair that involved an epoxy wet coat for small cover deficiencies, and an epoxy fibre reinforced coat for greater deficiencies. He was required to carry out trials to demonstrate that the adhesion of the repairs was satisfactory, and to subject one of the beams to a flexural test. However, he also elected to place the remaining beams over the railway in early April, in advance of the evaluation.

It was soon apparent that the epoxy fibre repairs were not adhering satisfactorily to the concrete. The repairs were patchy, the colour match poor, and no information was forthcoming on long term durability. The Engineer finally rejected those beams not complying with the specification, at the end of April.

The Contractor continued to put forward alternative methods of repair where cover just failed to comply. A Vandex system for beams with a slight cover deficiency was agreed, with the proviso that the Contractor built a means of access into the structure, to facilitate inspection within the U beams. Such a provision was difficult to make, and the Contractor wisely, took the decision to replace all the beams.

The main cause of the cover deficiency was a combination of movement of the inner core of the mould during casting, and movement of the reinforcement during vibration of the concrete. With hindsight it is easy to ask why these deficiencies were not envisioned or detected earlier. However, in 1979 the practice was generally to test cover prior to concreting, and the use of a cover meter was restricted to examining cases where there were reasons to suspect inadequate cover. There was apparently no good reason to expect that the cover would not be satisfactory in this instance.

Subsequently the Department of Transport specification was revised, and clause 1714 now requires that all concrete work must undergo a cover meter survey. Such an instance indicates why the Works specification and the thickness of tender documents increased inevitably. Contractors have much to say on this issue, in Volume 2 of this series.


The Minister of State for Wales, Mr. Michael Roberts, opened the Castleton to Coryton section to traffic on Thursday the 10th July 1980.