Chapter 8: Utilities

Utilities

8.1 In addition to providing means of travel from place to place, roads are also used to a considerable extent to provide routes for the various utility services required for homes and businesses. These include water, sewerage, electricity and gas and also the various communication media such as telephone lines and fibre optic cables. The carriageway also requires to accommodate drainage of the road itself, and electricity supplies for street lighting and road signs. We are all familiar with the roadworks required to carry out works on these services buried under the road and the inconvenience that results. That inconvenience would be all the greater if it were necessary to access services under a tram line. The works to reinstate track and track bed would be more difficult to achieve and could result in greater disruption. As this would clearly be unacceptable, it was necessary to remove almost all the utilities under or close to the new Edinburgh tram line before it was laid.

8.2 Even viewed in isolation, this element of the works itself ran over budget and took longer than expected, and it is relevant to examine the reason for this. The effects went further than the works to the utilities, however, as the delay caused knock-on effects with the infrastructure works, which led to further delay and expense there.

8.3 The number of utilities buried in the roads – and the facts that the period during which they had been installed extended over more than 100 years and there was no overall plan of what had been installed – meant that the task of moving the utilities was significant. The first stage was to identify which utilities lay under the road, and in precisely which locations, in order to identify what had to be moved. When the construction of the tram line was first planned it was known that utilities records were unlikely to be complete and accurate and that it was likely that unexpected utilities and underground obstructions would be found when the ground was excavated [Mr Barclay PHT00000031, pages 24–26; Mr Malkin ibid, page 138; Mr Casserly PHT00000022, page 63; Mr Reynolds TRI00000124_C, page 0059, paragraph 188; Mr Chandler PHT00000020, page 128]. This had been the experience in other recent tram projects in the UK and was particularly likely to be the case in a historic city like Edinburgh, where numerous layers of utilities had been laid, diverted and abandoned in the ground over many years. Water and gas pipes and electricity cables started to be placed in the roadways well over 100 years ago. The means of keeping records of this were, by today’s standards, primitive. Although there is now a statutory requirement to log accurately the position of utilities, fully computerised records and databases are a recent addition [Mr Barclay PHT00000031, pages 27–28]. In addition, the condition of apparatus would sometimes not be known until it was uncovered. The problem, however, was not just one of identifying which utilities must be moved to provide a clear path for the trams; it was also necessary to identify where the replacement pipes, cables and sewers could be placed. This was not straightforward, as the areas in question were likely already to be congested with existing utilities and there were constraints as to the distance that had to be maintained between different apparatus and the required depth of cover.

8.4 The statutory regime that regulates utilities placed in roads is contained in the New Roads and Street Works Act 1991 (the “1991 Act”). Section 143(1) of the 1991 Act states that where apparatus in a road will be affected by works, such as the Tram project, the utility undertaker and tram authority shall take such steps as are reasonably required—

“(a) to identify any measures needing to be taken in relation to the apparatus in consequence of, or in order to facilitate, the execution of the authority’s works,

(b) to settle a specification of the necessary measures and determine by whom they are to be taken, and

(c) to co–ordinate the taking of those measures and the execution of the authority’s works, so as to secure the efficient implementation of the necessary work and the avoidance of unnecessary delay.”

8.5 Regulations made under the 1991 Act state that if there is a dispute between the utility undertaker and the transport authority in relation to the above matters, it may be settled by the Scottish Road Works Commissioner (the “Commissioner”) [Road Works (Settlement of Disputes and Appeals against Directions) (Scotland) Regulations 2008 (SSI 2008/89), reg 2(2)(e)].

8.6 The 1991 Act also makes provision for a code of practice to be issued. When the Act was passed, this would have been issued by the Secretary of State for Scotland but, since devolution, it has been the responsibility of the Scottish Ministers. Such a code of practice, giving practical guidance on the matters mentioned in section 143(1) of the Act and the steps to be taken by the authority and the utility undertaker, was issued in June 1992 [Code of Practice on measures where apparatus is affected by major works (diversionary works), approved by the Secretaries of State for Transport, Wales and Scotland, June 1992, TRI00000302] (the “Code”).

8.7 The Code contains no provisions that relate specifically to tram works. Appendix C to the Code sets out recommended procedures for consultation, planning and the execution of any utility works that may be required as a result of the authority’s works [ibid, pages 0057–0065]. It envisages that execution of any necessary work to the utility will be carried out by the utility undertaker. Prior to execution of work, it has a procedure for exchange of information between the parties. It requires that the authority submit details of the proposed scheme to the undertaker and that the undertaker responds with preliminary details of the effects on its apparatus and budget estimates for the works. The estimate should normally be provided within 20 days. When the design of the scheme has been finalised, it should be submitted to the undertaker and within 25 working days the undertaker should provide a description of the measures that are required, a detailed specification of the works required, a detailed estimate and programmes and timescales for the work. The effect of the Multi-Utilities Diversion Framework Agreement (“MUDFA”) contract supplanted the detail of these stages, and I refer to them simply to indicate the pre-existing mechanisms that envisaged an exchange of information between the two parties. Even if the Appendix C procedures had been used, there was a need for exchange of information to identify which works were required and the likely costs. In the Tram project, prior to the Mar Hall mediation, neither side fulfilled the expectations of the Code regarding the provision of information.

8.8 For completeness I note that, in evidence to the Inquiry, Mr Rumney noted that an attempt was made to update the Code in 2003, but that it was held up pending a court decision in the case of British Telecommunications plc v Gwynedd Council, which concerned whether the costs of preparing a detailed estimate for the proposed diversion works were part of the costs that the Act required to be shared [TRI00000283, pages 0005–0006, paragraph 32]. Although that decision was given by the Court of Appeal in 2004 to the effect that these were part of the “allowable costs” [[2004] EWCA Civ 942] a new Code has not been issued. Mr Rumney considered that, as a result, the Code is now markedly out of date and does not reflect changes in legislation since it was adopted. During the preparation of this report the Solicitor to the Inquiry obtained information from the Commissioner about the current position in relation to the Code. I am extremely grateful to the Commissioner for his assistance, which confirmed that the Code is still in force but that there has been a series of Advice Notes relating to diversionary works, the most recent being Advice Note No. 2010/1 by the Highways Authorities and Utilities Committee [HAUC (UK)] [GOV00000030]. As disclosed in correspondence between the Solicitor to the Inquiry and officials in Transport Scotland [TRI00000315], the Code is the subject of a joint review by the Department for Transport (“DfT”) and Transport Scotland. See Appendix 5.

The MUDFA concept

8.9 As was noted in Chapter 5 (Procurement Strategy), the report by the National Audit Office entitled ‘Improving public transport in England through light rail’ [CEC01708649] noted that the diversion of utilities tended to be an expensive part of tram or light rail schemes, and it recommended that the DfT fund projects only where the promoters had adequate proposals for managing the risks associated with such diversions. (See paragraph 5.5 above, where I refer to Mr Kendall’s statement to the Inquiry on this subject.) Accordingly, Transport Initiatives Edinburgh Limited’s (“tie’s”) procurement strategy for the Edinburgh project sought to address the risks associated with the diversion of utilities by completing design and utilities diversion works in advance of the infrastructure works rather than having them undertaken by the infrastructure contractor. The intention was that the infrastructure contractors should, in general, be handed a utility-free zone in which they could undertake their construction. This entailed that excavations would be carried out to enable the utility works to be done, the road would then be reinstated and, later, when the infrastructure works were done there would be further excavation, works and reinstatement. The result was that there would be disruption to road users and businesses on the road twice: once when the utility works were carried out and again when the tram works were carried out. The view was taken that despite this it was a preferable way to have the works carried out in terms of delivering the Tram project. It was also considered that the overall period of disruption would be shorter, as the roads would not remain in a disrupted state in the period between concluding the utilities works and commencing the infrastructure works.

8.10 There were some exceptions to the principle that the utility work would be carried out in advance of infrastructure works where it was always intended that the works would be carried out by the infrastructure contractor (“Infraco”). Other than those instances, the prior completion of the MUDFA works was one of the pricing assumptions contained in the Infraco contract. However, as noted in Chapter 12 (Contract Close), it was apparent that there was slippage in the MUDFA programme before the Infraco contract had been signed. The timetable slipped further as the works were carried out on the utilities.

8.11 In paragraph 8.6 above, I have noted that the assumption in the Code promulgated under the Act is that the alterations to utilities will be carried out by the utility companies themselves. The intention for the Tram project, however, was that, a single contractor appointed by tie would carry out all the utilities works except where they related to the diversion of apparatus requiring specialist skills such as high-pressure gas mains, high-voltage electricity supplies and certain telecommunications cables [Final Business Case, version 2, CEC01395434, Part 5, page 0109, paragraph 7.72]. This was a novel approach. Although other projects had sought to move utilities prior to infrastructure works getting under way, in those instances each utility company remained responsible for diverting its own apparatus [eg in the Croydon and Manchester tram schemes, as discussed in the advice note produced by UK Tram – Phase 2, Guideline 3: ‘The Causes and Control of Cost Creep and Cost Escalation’ (June 2010), TRI00000303].

8.12 In pursuit of this approach, on 4 October 2006, MUDFA was concluded between tie and Alfred McAlpine Infrastructure Services Limited (“AMIS”) [CAR00000300, Parts 1–2]. Following the acquisition in February 2008 by Carillion plc of Alfred McAlpine plc, AMIS later changed its name to Carillion Infrastructure Services Limited and then to Carillion Utility Services Limited. There was no change in the party to the contracts. In this Report, I have generally sought to use the name of the company as it was at the time of the event or events that I am describing, but they may be regarded as being interchangeable.

8.13 The full scope of the services to be supplied by AMIS was set out in Schedule 1 to MUDFA [ibid, Parts 1–2, page 0146 onwards]. The contract provided for the MUDFA works to be undertaken in two phases: a pre-construction services phase followed by a construction services phase. The pre-construction phase involved AMIS in preparing for the construction phase of the MUDFA works, including planning the works, planning the traffic management arrangements, preparing procedures, setting up site offices and materials compounds, resource planning and establishing relationships with the various parties [Mr Malkin TRI00000056_C, pages 0009–0010]. The pre-construction deliverables are set out in Schedule 1 (scope of services) to MUDFA [CAR00000300, Part 1, pages 0162–0164, clause 2.52].

8.14 During the pre-construction phase AMIS was to receive the Issued for Construction (“IFC”) utility design drawings and plan the utility diversion works. In response to a tender query dated 8 May 2006 by AMIS, tie indicated that it was anticipated that detailed utility design would be complete between 25 April and 21 July 2006 and, after comments on the detailed design by tie, the utility companies and the MUDFA contractor, IFC utility design would be available between 13 September and 21 December 2006 [letter dated 23 August 2007 from AMIS to tie CEC01702113]. In the event, as will be discussed below, there was significant delay in producing the IFC utility design.

8.15 In the course of the evidence, reference was made to sections of the route where utility diversion works were planned or undertaken. To assist the reader’s comprehension of that evidence and of references in this report to work planned or undertaken in a particular area identified by number it should be understood that the route of line 1a was divided into various sections, which are illustrated in Table 8.1. References in this report to sections by number refer to the numbers allocated to each section in that table. Sections 3 and 4 are omitted from Table 8.1 because they relate respectively to the section from Haymarket to Granton Square (line 1b) or the later extension from Granton Square to complete the northern loop of the network.

8.16 MUDFA was a re-measurement contract rather than a fixed-price contract. This meant that at the outset there was no statement of what the final cost should be. Instead, the contract recorded the rates that would be paid for the work to be carried out and the sum due would be determined once it was known which works had actually been undertaken. This was done because at the outset it was not possible to predict with any accuracy what work would be required. After the adverse experience in Nottingham, if a contractor was to undertake to carry out all the work for a fixed price, in order to avoid potentially huge losses, they would require to submit a price that took account of the anticipated worst-case scenario. Such a precautionary approach was thought likely to inflate the price unnecessarily but, even with such an approach, contractors might decide that they were not willing to bear the risk that things turned out to be even worse than expected. This could make it difficult to find contractors willing to carry out the works, resulting in a difficulty in attracting sufficient interest to result in competitive bids when the contract was put out to tender.

8.17 A report to the City of Edinburgh Council (“CEC”) on 21 September 2006 noted that although it was not possible to finalise the scope of the utilities diversion work until the MUDFA contractor had been appointed and the pre–construction stage had been completed, it was estimated that the total value of MUDFA would be in the order of £50 million [CEC02083472, page 0004, paragraph 3.13]. Mr Barclay said that, at the time, the MUDFA works for this project constituted the largest multi-utility project ever undertaken in Europe. He considered that people failed to appreciate the magnitude of the MUDFA project. The process was extremely complex and he was of the view that insufficient time had been allocated at the outset before work started. With hindsight, he considered that design should have started a year earlier and that in future tram projects utilities should be diverted “well in advance of the tram infrastructure team commencing their work” [TRI00000024_C, pages 0032–0034, paragraphs 97 and 103].

8.18 Pre-construction services were expected to start in late July/August 2006, with construction programmed from January 2007 to April 2008. As was noted in paragraph 8.14 above, it was planned to issue preliminary design for all sections by the end of 2006. To complete the utility designs it was necessary to have the infrastructure design, as it would dictate which apparatus had to be moved. It was reported that, by mid-autumn, the new road layout, overhead line equipment (“OLE”) pole positions, the track bed alignment and cross-sections along the full length would be known. Feedback from some of the utility companies, however, was that the programme was very tight and would be difficult to achieve [CEC01872859, page 0002, paragraph 4.4].

8.19 In addition to MUDFA, between August 2005 and June 2008, tie (and, in some cases, CEC) entered into a number of agreements with utility companies in relation to which information and approval would be sought and provided and how the MUDFA works would be carried out. These agreements included those with Scottish Water (in August 2005) [CEC01015132]; Transco and Scotland Gas Networks (“SGN”) (in October 2005) [CEC01275865]; British Telecom (“BT”) (in February 2006) [CEC01580825]; and Scottish Power (in June 2008) [CEC01370893].

8.20 MUDFA contained embargo periods during which utility works could not be undertaken in certain parts of the city. In particular, no works were to be undertaken in the city centre (between Haymarket and London Road) between early August and early September (for the Edinburgh Festival) or between early December and early January (for the Christmas period) [CAR00000300, Part 1, page 0205, Schedule 2, clause 8.5]. In addition, it was agreed with Forth Ports that MUDFA works would not be undertaken between October and January, so as to avoid the Christmas season. There was evidence that if, for whatever reason, the window of opportunity for carrying out works in these areas was missed, there could be a significant delay. In addition, because of a high demand for gas in winter, certain works could not be done to the gas network between November and March each year, and allowance had to be made for delays associated with emergency work to utilities unrelated to the Tram project [Mr Bell TRI00000109_C, pages 0014–0015].

8.21 The programme contained in Schedule 8 to MUDFA [CAR00005833] envisaged that the pre-construction services would be undertaken between 31 July and December 2006 and the MUDFA construction works would thereafter take approximately 14 months between March 2007 and May 2008. At the time when MUDFA was entered into it was anticipated that the Infraco works would commence at the beginning of 2009, which, if the diversion works had been completed by May 2008, would have allowed for a gap of approximately seven months between the programmed completion of the MUDFA works and the planned commencement of the Infraco works. The date for starting pre-construction work was, however, clearly impossible, as MUDFA was not concluded until October 2006, so, from the very outset, it must have been clear that the programme was slipping. In 2007 a revised MUDFA programme was agreed, which provided for the MUDFA construction works starting in July 2007 and finishing by the end of 2008, ie a period of approximately 18 months, with almost no buffer between then and the start of the Infraco works.

8.22 Even this new period was not realistic for the completion of the required works. The length of the on-street section in the Edinburgh Tram project (ie from Haymarket to Newhaven) was approximately eight kilometres. Mr Rumney gave evidence that a total duration of 18 months for the MUDFA works did not sit with the duration of diversion works in comparable projects. In Manchester, where there was a 2,000-metre on-street section, and in Birmingham, where there was an 800-metre on-street section, the works took about 2 years to complete [TRI00000283, page 0019, paragraphs 95–97]. In view of erosion of the buffer period between completion of the MUDFA works and the commencement of the Infraco works, this meant that from the outset it should have been obvious to tie’s senior officials that problems were inevitable and that the MUDFA works would impinge on the Infraco works.

8.23 Moreover, I accept Mr Barclay’s evidence that inadequate time was allowed to accommodate delays associated with inadequate records of utilities and unexpected obstructions or cavities. Even without the benefit of hindsight, tie ought to have anticipated such delays, bearing in mind the known difficulties at that time relating to the records of existing and redundant utilities and the uncertainties of what lay below the ground in the historic city of Edinburgh. tie should also have realised that the time allocated for the diversion was inadequate when account was taken of the time taken for much shorter lengths of on-street diversions in Manchester and Birmingham. In these circumstances tie ought to have instructed Parsons Brinckerhoff (“PB”) to commence design a year earlier to ensure that the diversion of utilities was completed before the infrastructure works commenced. Alternatively, once it was apparent that there was slippage in the programme of diverting utilities, tie ought to have delayed the procurement process for the Infraco contract and should not have signed that contract until the diversion of utilities had been completed. tie’s failure to take either course of action and its insistence upon concluding the Infraco contract before the utilities had been diverted was a serious failure in its management of the project.

Identification of utilities

8.24 I have noted in paragraph 8.3 above that the records of the location of utilities were known to be incomplete and inaccurate but those records were nonetheless what had to form the starting point for identifying the work that had to be done. By August 2002, available plans had been obtained from utility companies, and the information that they contained was collated to produce a set of composite utility plans [CEC01692684, pages 0009–0011, paragraph 2.1]. The poor quality of the records obtained was subject to frequent comment [see, eg, Mr Bell TRI00000109_C, page 0011; Report to MUDFA Sub–Committee on 12 March 2007, CEC01694523].

8.25 As it was known that the records were not complete, consideration was given to means by which a better picture of the location and volume of utilities might be obtained. Ground-penetrating radar (“GPR”) is a technology that makes it possible to survey apparatus buried under the roads without disturbing the surface. It was known that it is not 100 per cent accurate, as the multiple layers of apparatus that may exist beneath the road surface, particularly in a city centre, make it difficult to analyse the data and predict accurately which utilities exist and where they are [Mr Chandler TRI00000027_C, pages 0084–0085, paragraph 332]. Inaccuracies are more likely with smaller elements of apparatus [Mr Malkin PHT00000031, page 139; Mr Chandler PHT00000020, page 133; Mr Reynolds PHT00000019, pages 157–158] and may depend also on the material that the utility apparatus is made of, and the type and moisture content of soil containing the apparatus [Mr Barclay TRI00000024_C, page 0039, paragraph 119].

8.26 The known limitations of records of utilities were such that at the outset there was an intention to conduct a survey of the entire on-street section of the network, but this intention was abandoned because of “programme and budget constraints” [Technical Note by Halcrow, dated 4 April 2006 (the “Technical Note”), PBH00004905, page 0006, paragraph 4.1]. The programme concern was that the survey would take nine months and that this could not be accommodated within the intended timescales for the MUDFA works. As to the cost, the full survey was estimated to cost in excess of £2 million, whereas the reduced scope survey that replaced it cost just £220,000. The change in cost gives an indication of just how reduced the scope of the alternative was. It considered only areas of severe congestion such as junctions, major services that would generate greater cost and areas where there was very restricted land space. Although these would be areas of sensitivity where it could be difficult to accommodate the changes, they were not necessarily areas where there was most likely to be unrecorded apparatus, and no account was taken of what would happen if unrecorded apparatus were to be discovered in other areas.

8.27 Where surveys were carried out, the surveyors investigated the utilities using three different methods. There was an initial study of the drawings, followed, in the case of cables, by induction of a current into the cables that could be tracked at the surface, and finally the repeated use of GPR at each location to identify apparatus at different depths [ibid, pages 0007–0008, paragraph 4.6]. The outcome of these surveys was recorded as follows:

• “Water and gas apparatus is relatively accurate
• Others utility apparatus showing high levels of positioning inaccuracy between plans and survey
• All utilities – significant volume of additional apparatus identified
• High number of chambers, tunnels and other underground structures identified” [ibid, page 0008, paragraph 4.9].

8.28 In general, these findings confirmed that the records were not accurate. The Technical Note made the following further conclusions.

• Because utility design to date had proceeded on the basis of the composite plan produced on the basis of the contents of the records kept by utility companies, it carried a significant risk of abortive design work, including negotiations with those companies.
• The presence of significant apparatus not identified on plans raised severe risk on the MUDFA scope of works and programme [ibid, page 0009].

8.29 It also expressed the view that the decision to restrict the extent of the surveys was an acceptance of the risks of discovery of unknown apparatus. I do not think that there had been any express, or even conscious, acceptance of this risk but as the works progressed it was certainly the case that the prediction that there would be unknown apparatus was borne out.

8.30 I consider that it was a clear error to reduce the scope of the survey and that this should have been obvious to everyone at the time – particularly in view of the warning in the Technical Note and also in view of the experience with other UK tram projects. It was appreciated that the records were not reliable, which meant that it should have been obvious that unexpected apparatus, underground chambers and other obstructions would be encountered as well as apparatus in positions other than those expected. It would have been obvious also that if this occurred during the infrastructure works it would give rise to delay and additional cost. This should have been apparent to tie before Schedule Part 4 (“SP4”) to the Infraco contract was concluded in early 2008 and that proceeding on the basis of a restricted survey increased the risk of encountering unexpected conditions underground, resulting in delay and in additional costs. There does not appear to have been any attempt to compare these costs and these delays with the consequences of having the more extensive survey carried out. This was a material failing.

8.31 Following the results of the limited survey exercise, consideration was given to the need for trial pits/slit trenches to verify the apparatus in the ground and for surveys being carried out to identify voids under the roads. It had been originally envisaged that in addition to surveys System Design Services (“SDS”) would instruct trial pits to confirm the results of the surveys and to inform the design. In March 2006, SDS produced a survey programme that included provision for utilities trial pits to be undertaken between August and October 2006 [TIE00201989, page 0008, line 289]. The minutes of a meeting on 1 November 2006, however, noted that although SDS was to forward a schedule of trial pits that it required in order to complete design, tie and SDS would discuss and agree how the trial pits would be procured, eg through the engagement of the utilities contractor under MUDFA or by another route [TIE00677359, page 0004, item 5.4]. A set of drawings for trial pits prepared by SDS and sent by tie to AMIS in late November 2006 showed 77 possible trial pits between Haymarket and Newhaven [TIE00207331; TIE00207334; TIE00207333]. A separate summary document appears to suggest that it was proposed to proceed with only 23 of these trial pits [TIE00207336]. In March 2007, SDS said that the advice from its designers was that trial pits were not required and it would not, therefore, be carrying out any of the trial pits that were originally planned [CEC01663114, page 0003, item 2.16]. Consistent with that position, a surveys programme produced by SDS in March 2007 contained no provision for utilities trial pits [CEC01686232]. Within tie there was a view that this was a tactic to avoid doing work at the time and instead do it later when there might be additional payment [email from Mr Dent to Mr Barclay, 28 March 2007, CEC01638353]. The System Design Services contract (“SDS contract”) did not, however, expressly require SDS to instruct or undertake trial pits for utilities, and it took the view that it was not required to instruct or pay for trial pits to be undertaken.

8.32 In late 2007 and early 2008, tie eventually instructed AMIS to carry out the works on trial pits but the delay in settling this matter meant that they were carried out later than had been planned. This meant that although trial pits were carried out, they were not, in general, carried out until after utility IFC design drawings had been issued, and shortly before the MUDFA diversion works were to be carried out. That meant that the results of the trial hole investigations (including, in particular, the discovery of unknown utilities and underground obstructions) were not known until shortly before the diversion works took place, which caused delay while the design was revised, and agreed with the utility companies, to take account of unknown utilities etc.

8.33 Mr Barclay gave evidence that more surveys and trial holes were undertaken on the Tram project than on other projects he had worked on [PHT00000031, page 41]. However, there was evidence from a number of other witnesses that more trial pits should have been dug [Mr Chandler PHT00000020, pages 129–135 and 137; Mr Casserly PHT00000022, pages 79–81; Mr Malkin TRI00000056_C, page 0035; PHT00000031, pages 141–142].

8.34 Clearly, at this stage I have the benefit of knowledge of the problems that arose from unidentified utilities. Nonetheless, having regard to the known problems with records of utilities, the importance within the procurement strategy of completing utility works before infrastructure works started and the short timescale that would be afforded to the utility works, I consider that it ought to have been apparent that if the trial pits were not dug timeously, the result would be delay and expense. It was an error not to proceed with the pits as originally planned, and having more would have provided SDS, tie and AMIS with more accurate and complete information. As was noted in paragraph 8.30 above, the responsibility for this omission rests with tie.

8.35 In early 2007 a decision was taken to undertake a trial site for the MUDFA construction phase at Casino Square, Leith. Mr Barclay gave evidence that that location was chosen as it was a non-critical area with relatively few utilities, low stakeholder engagement and low profile and would allow a start to be made there while SDS focused on producing utility design for the more critical areas, ie the main arterial roads [TRI00000024_C, page 0024, paragraph 72]. Mr Malkin explained that there was some ambiguity about the meaning of “trial holes”. Trial holes were dug to support MUDFA works but were also dug to make sure that the apparatus in the area had been positively identified and that the (diversion) works to be done were fully understood [PHT00000031, pages 132–134]. Work on the trial site was started in early April 2007 and trial holes were dug as part of the MUDFA works [CEC01638569, page 0009]. Shortly prior to commencement of the excavations Adien undertook a GPR survey of the area. Although four items of utilities apparatus were detected by the survey, 17 items of apparatus were discovered when the trial site was excavated. The unknown apparatus, which was not shown on the approved for construction (“AFC”) drawings, included six pipes, one cable, two ducts with cables in them and two drains. Moreover, the 90 mm gas main was not at the location shown on the survey [tie emails dated 5 April 2007 (CEC01639398) with spreadsheet listing the anticipated and discovered, apparatus (CEC01639399)]. By letter dated 5 April 2007, tie wrote to SDS expressing concern in relation to the unidentified apparatus and the location accuracy of known utility assets. In response to these concerns expressed by tie, by email dated 10 April 2007, Adien reported that its site procedures had not been fully followed at the trial site by one of its team of surveyors. Accordingly, Adien decided to undertake, at its own expense, site walkover surveys in the other sites that that team had surveyed, in order to provide confidence that the trial site error was a single error and had not been repeated at other sites [TIE00184030]. See also Adien’s feedback report [PBH00010316]. The outcome of the trial site meant that not only was it established that the records of the utility companies were inaccurate, but that it would not be possible to rely on GPR surveys to uncover the true position.

8.36 In my view, the experience at Casino Square should have alerted tie in May 2007 to the consequences for the programme of MUDFA of the failure to dig slit trenches or trial pits at the pre-construction phase and of the unreliability of the information about the nature and extent of the work that would be required to divert utilities along the route of the tram. tie failed to appreciate that it was probable that there would be significant delay in completing the diversion of utilities, which, in turn, would result in delay in the award of the Infraco contract if the procurement strategy were to be maintained. Alternatively, if tie did appreciate the implications for the procurement strategy of these findings, it failed to take appropriate action, such as re-programming MUDFA and delaying the award of the Infraco contract. In either event, it indicates tie’s failings as the company responsible for procuring and delivering the project on behalf of CEC. At that time there was concern that the newly elected Scottish National Party (“SNP”) Government wished to cancel the project, and it may be that tie was reluctant to take any action that would delay the award of the Infraco contract due to fear that such action might result in the cancellation of the project. However, any such concern by tie was irrelevant to its obligation to procure and deliver the project in accordance with the procurement strategy, and in that respect to ensure that councillors at CEC were made aware of the likely delays in the completion of the project.

Design of the utility works

8.37 Clause 2.4 of MUDFA stated:

“Notwithstanding that the SDS Provider shall be responsible for the design and specification of the MUDFA works … the MUDFA Contractor shall be responsible for its input into the design and specification of the MUDFA Works or any part thereof …” [CAR00000300, Part 1, page 0035, paragraph 2.4.]

8.38 The responsibility of SDS for the design and specification of the MUDFA works was not mirrored by the obligations of PB in terms of its contract with tie. The schedule of services to be provided by PB in terms of the SDS contract contained a short paragraph on “Utilities”, which provided that SDS was to “provide assistance to tie with the management of an advanced utilities diversion programme”. This was stated to include:

• “assessing the need for and acquiring relevant data relating to the presence and location of all buried and above ground utility services;
• agreeing the need for and extent of diversions;
• undertaking critical design and developing a strategy for all utilities diversions to minimise diversion requirements and out-turn costs;
• … preparing C4 cost schedules;
• preparation of documentation (including the contract terms) associated with the proposal to appoint a single service agreement with a specialist contractor to carry out advanced utility diversions;
• activities required to support the utilities diversion process including, but not limited to, traffic management plans/traffic regulation orders, site meetings and all necessary re-designs;
• management of unidentified diversions and design re-work/modifications on an as required basis;
• on-site attendance on an as-required basis; and
• attendance at all meetings on an as-required basis.” [CEC00839054, pages 0091–0092, schedule 1 (scope of services), paragraph 3.2.1.]

8.39 Where any necessary works to the utilities were to be undertaken by Infraco, rather than the MUDFA contractor, SDS also had specific responsibility “for the determination and design” of all utility diversions [ibid, page 0092, paragraph 3.2.2]. Thus the responsibility for the design of all utility diversions rested with PB, as the SDS provider, whether the MUDFA contractor or Infraco was responsible for undertaking the necessary work.

8.40 Schedule 12 to the SDS contract specified the functional requirements that PB had to meet in respect of civil engineering works. The final paragraph of the Schedule titled “Utilities” was in the following terms:

“Wherever possible, the Edinburgh Tram Network shall be designed such that there is no requirement to divert existing public utilities. Where this is not achievable, then diversionary works shall as far as possible be undertaken as part of an ‘advanced works’ programme with the relevant public utility organisation.” [ibid, page 0319, paragraph 2.6.]

This design principle was intended to limit the diversion of public utilities to those locations where diversionary works could not be avoided.

Initial delay in designs

8.41 Mr Malkin said that the principal difficulty for AMIS during the pre-construction phase was the delay in receiving utility design from SDS and that design, once available, did not contain the necessary details to allow AMIS to properly plan and purchase materials and fittings to support the construction phase. He said that the non-availability of detailed utility drawings virtually eradicated the benefits of carrying out the pre-construction services phase and resulted in the commencement of the main MUDFA works being delayed, compounded the overall programme and led to out-of-sequence working based on drawing availability, which consequently required additional resources and service support, and increased cost [TRI00000056_C, pages 0007–0010].

8.42 There was awareness within tie of the design problems. A progress report by Mr Harper for a meeting of the Design, Procurement and Delivery Sub-Committee on 8 November 2006 noted that there was concern about the impact that the timing of the delivery of utility diversion design would have on the implementation of MUDFA works. It was noted that AMIS had written to tie indicating that the quality of design was far below what it would have expected at that stage and indicating that that might have an impact on its ability to deliver its first programme. Concern was also expressed about the risk of an increase in the project estimate as a result of Scottish Power’s request for five additional feasibility studies in specified areas [CEC01803371, pages 0014–0015].

Difficulties in obtaining input and approval from utility companies

8.43 In order to complete the design, it was necessary that the utility whose apparatus was being moved agreed to what was proposed. If the utility in question did not agree to the change, the design would have to be altered and the process to obtain approval would have to start again. If the design thus amended was objectionable to another utility, a further round of design and approval would take place, and so on until agreement was reached. It is immediately apparent that this was labour intensive and a recipe for delay. Even if this was carried out just once for each alteration to a utility, this would be very time-consuming. In practice it was more difficult as, at the same time that the utilities design was being developed, changes were being made to the infrastructure design that required consequential changes to the utilities diversions. When this occurred, even if the process for utility design had been brought to a conclusion, it would have to start again.

8.44 The input from the utility companies was clearly critical to the process. It was necessary that they consider the designs as they were developed and provide a note of their views. In practice, the teams within the utility companies with the task of responding to these requests were not resourced to be in a position to respond within the required timescale. The design programme had been prepared on the basis that they would reply within 20 days but, by February 2009, SDS was allowing 40 days. The position was not helped by the iterative process for obtaining approval described above or by the necessity of seeking approval afresh when the scope of work required to the utilities was changed to accommodate changes to the infrastructure design. In that situation, utility companies that had been through a number of iterations would be told that it was necessary to start again.

8.45 In considering this situation, it might be thought that the burden of responding to these requests was simply a price that the utility companies should have to bear for being able to place their business assets in the roads. I consider that there is force to this argument, and I return to it below. However, even taking this approach and accepting that an iterative process is to some extent inevitable, it would be reasonable for the infrastructure design to be clear at the outset such that all parties were aware of what had to be moved.

Consequences of late designs

8.46 Mr Malkin said that, in early 2007, it was apparent that the SDS utility design drawings were going to be significantly late and, with AMIS having already mobilised its project management team, consideration was given to a proposal whereby it would be tasked to carry out project works. Ms Craggs was concerned that this proposal might not be the best option as AMIS had been procured on the basis of its reputation for moving utilities and it might not be the most suitable contractor to undertake advanced works packages such as at Ingliston Park and Ride and excavation works for the depot at Gogar. She was also concerned that awarding these advanced works packages to AMIS might breach procurement legislation [TRI00000029, pages 0103–0104, paragraph 288]. Nevertheless tie agreed to the proposal and AMIS undertook the bulk excavation works at the depot and the diversion of a 600 mm water main using resources and expertise that would otherwise have been sitting underutilised [Mr Malkin TRI00000056_C, page 0019; Mr Bell TRI00000109_C, pages 0020 and 0145].

8.47 At a meeting between tie and AMIS in March 2007 it was agreed that, as a consequence of late designs and associated data, a phased transition would take place rather than the distinct completion of the pre-construction services phase before the construction services phase, which would result in the pre-construction phase being completed in parallel with the construction phase, as detailed utility design became available [CEC01630357, page 0002, paragraph 3].

8.48 By letter emailed on 20 June 2007, AMIS advised tie that delay and disruption caused by late provision of designs and information had resulted in a reduction in its turnover of £6.1 million and loss of profit and overhead recovery to date of £530,000. It considered that the information that had been provided meant that revision 05 of the MUDFA programme was “untenable” and it was concerned that the loss would continue to accumulate unless it was provided with sufficient designs and information to seek to recover the schedule [CEC01636546; CEC01636547]. It repeated its concerns in a letter of 28 June 2007 [CEC01691617].

8.49 In a strongly worded letter to tie dated 28 February 2007, SDS expressed concerns at being asked again to programme the production of utilities design to re-align the deliverables for the MUDFA programme [CEC01800436]. Its concern was in part that this would require further input from the utility companies, with the problems noted above. It was also concerned that tie appeared to be developing an early programme of MUDFA works for the road sections that was out of sequence with the development of finalised roads infrastructure design and OLE design on which it should have been based. Its concern was that tie had no regard for the consequences of this. It was also concerned that delays by CEC in deciding the preferred road designs following the charrette process meant that tie’s programme would require MUDFA works to be undertaken before the apparatus re-design was completed. It enumerated five consequences that it said would inevitably follow from the approach taken by tie. These were:

“(i) Delay to Statutory Utility Companies’ approval of SDS’ Detailed Design submissions.

(ii) A very negative response from the Statutory Utility Companies when they then discover that they had been requested to formally approve or they have already approved SDS’ apparatus diversion Detailed Designs submissions only for them to change later and require re-approval. This may well impact on their willingness to approve our subsequent Detailed Designs submissions or our AFC submissions.

(iii) The potential for costly re-work by MUDFA in the event that apparatus that they had already diverted using SDS’ AFC drawings referred to above, proves to be too shallow or in the incorrect location for the final Detailed Design alignment and road layout once issued.

(iv) The potential for MUDFA to claim standing time from tie whilst SDS’ updated designs are reapproved by the Statutory Utility Companies.

(v) A programme mismatch between tie’s programme for execution of MUDFA works and SDS’ ability to deliver updated IFC drawings to MUDFA.” [ibid, page 0002.]

8.50 Both the nature of the works that were to be carried out and contractual structures put in place created a number of interfaces between the various parties: tie, SDS, AMIS, the utility companies and CEC. There were various interactions, flows of information and approvals. This imposed requirements for co-ordination, communication and management. The responsibility for this rested on tie, and it is apparent that it did not adequately manage the process with the result that there was considerable delay and consequent expense.

Timing of MUDFA works

8.51 A report to the Utilities Sub-Committee on 4 April 2007 noted that AMIS had issued a draft MUDFA programme revision 04 for comment, which showed the main MUDFA works starting on 2 July 2007 (being three months later than the previous programme, which was noted to be driven by design and work order requirements). The revised programme showed the main MUDFA works for phase 1a (Edinburgh Airport to Newhaven) being completed by early January 2009, which was six months later than shown on the previous programme (revision 03) [CEC01638569, pages 0009–0010]. By mid-May, revision 05 had been agreed, which showed an end date for the phase 1a works of November 2008 [CEC01566088, page 0002, paragraph 2.1].

8.52 At this time, in light of concerns about the initiation of the utility diversion programme, tie decided to implement AMIS’s proposal of a risk and trade-off (RATs) programme [letter dated 26 April 2007, CEC01691204]. Mr Barclay explained that in areas where there were few utilities and only simple utility diversions were required (eg at the trial site at Casino Square), AMIS excavated trial holes and then produced initial, or feasibility, design for diverting the utilities which was passed to SDS for checking and approval. While that was an attempt to fast-track design in areas in which only simple diversions were required, it quickly became clear that it would not be feasible to proceed by that method as the main works were carried out, which involved more complex diversions [Mr Barclay PHT00000031, pages 66–67; TRI00000024_C, page 0025, paragraph 74].

8.53 In addition to all these difficulties arising out of design, the MUDFA works became subject to an external delay. As was noted in Chapter 3 (Involvement of the Scottish Ministers), an election to the Scottish Parliament took place on 3 May 2007, following which an SNP administration was formed. Mr Swinney asked the Auditor General for Scotland to undertake a review of the project and the Edinburgh Airport Rail Link (“EARL”) project. Pending completion of that review Transport Scotland ordered that, with a few key exceptions, utilities diversion work be halted [Audit Scotland Report, CEC00785541, page 0020, paragraph 67]. Any contribution that this might have made to the delay in completion of utilities works would, however, have been minimal. The delay in the issue of IFC utility design meant that it was unlikely that AMIS would have been able to progress meaningfully the MUDFA works at that time [Mr Gallagher TRI00000037_C, pages 0121–0122, paragraph 362; Mr Barclay PHT00000031, pages 67–68; Mr Casserly TRI00000111_C, pages 0010 and 0021].

8.54 For completeness, it may be noted that, following the publication of the Audit Scotland Report, Mr Swinney publicly expressed concern about the utilities diversion works for the Tram project. The following day, on 21 June 2007, Mr Gallagher wrote to Mr Swinney, seeking to provide reassurance in relation to the utilities works [CEC01677601]. The letter stated that although Mr Gallagher agreed with Mr Swinney that utilities were a major concern, the MUDFA programme had the highest focus within tie and that Mr Gallagher was confident that tie could manage the MUDFA programme within the budget of £60 million. The letter made no mention of the difficulties and delays with the MUDFA works noted above, and in my view it was misleading. Mr Gallagher was – or should have been – aware of the problems.

MUDFA works in the period to contract close

8.55 The Inquiry heard evidence that the easier sections of the works were undertaken first, with the more difficult sections to be undertaken later in the programme. Mr Casserly, for example, gave evidence that it was anticipated that there were elements of the utility diversions, particularly around Ocean Terminal, which would be easier to do as they were on privately owned land, the owner (Forth Ports) had good records of the utilities in the area and there were expected to be fewer utilities in those areas. These works were, therefore, carried out first, commencing at Ocean Drive, Leith (section 1A) in July 2007, while IFC utility design was awaited from SDS for the other areas [PHT00000022, page 62. See also Mr Barclay PHT00000031, pages 98–99].

8.56 In view of both the state of the records of apparatus and the limitations on the surveys carried out, it was not surprising that, once the works were started, utilities were uncovered that had not been expected. When this happened, MUDFA required AMIS to take steps to identify the owner of the apparatus and, as instructed by tie, to apply for and obtain all approvals as might be required to allow works to be carried out in respect of the apparatus in such a way as to minimise disruption to the programme [CAR00000300, page 0042, clause 3.12]. Similarly, if adverse physical conditions or artificial obstructions were encountered, tie required to give written instructions to AMIS on how the adverse physical conditions or artificial obstructions were to be dealt with [ibid, pages 0055–0056, clause 10]. Clearly, the effect of finding unrecorded apparatus was that the scope of works increased. Even before the works had started, in a letter dated 7 February 2007, AMIS noted that although a total of 24,662 linear metres of diversions had been anticipated at the tender stage, the anticipated total was now 38,967 linear metres – an increase of 58 per cent [CEC01831536, page 0003]. Moreover, Mr Casserly gave evidence that although it was always known that there would be unexpected utilities, the problem turned out to be worse than was expected, and that that had become apparent by the end of 2007 [PHT00000022, page 64].

8.57 In this period the major concern remained the late submission of SDS detailed utility designs [CEC01831536]. Even one month after the works started, AMIS’s monthly progress report for August 2007 noted its main concern as being the lack of IFC utility design drawings (with approximately 285 drawings being behind schedule for work ordering) and other materials essential to progress the works [CEC01683946, pages 0005, 0007–0008, 0010 and 0013]. The same concern was voiced in the Progress Reports for September 2007 [CEC01684924, pages 0004, 0007–0009, 0012 and 0017], October 2007 [CEC01527921] and November 2007 [CEC01523817, pages 0004 and 0011]. There are no progress reports after that date as tie advised that it would take over that role [Mr Malkin TRI00000056_C, page 0016].

8.58 The works in this period also exposed the difficulties in working with the utility companies. The smooth operation of the MUDFA works was dependent on co-operation from the utility companies and the response from them was described by Mr Reynolds, of PB, in an email of 28 September 2007 to Mr Gallagher, as “patchy” [CEC01714281]. Agreements between tie, CEC and the utility companies required the utility companies to provide the successful bidder for the SDS contract (PB) with detailed information before the award of MUDFA to enable PB to complete the utilities diversion designs. He said that although information had been provided within the required timeframe by a number of utility companies, in several cases it had not been provided to the expected level of detail. In particular, BT Openreach had repeatedly failed to meet the required sectional completion dates, to the extent that several packages were still outstanding long after MUDFA was awarded, and this had resulted in serious delay to all subsequent milestones, including final delivery of the IFC utility drawings. He said that once composite drawings had been prepared by SDS they were circulated for review and approval to each of the utility companies and, although the programme had been based on a four-week period for such approval, in practice, four weeks had proved to be too short a period for Scottish Water, with the result that final IFC milestones had slipped further. He said that problems with approval of SDS designs had also arisen due to the delay to the conclusion of an agreement between tie and SGN that meant it had not been possible for SDS to secure final approval. Responsibility for managing the interaction with the utility companies had been retained by tie, and Mr Reynolds’ weekly report dated 28 September 2007 noted that tie had been reminded of this and had therefore been applying increasing pressure to both Scottish Water and BT [PBH00029122, page 0001, paragraph 3.2].

8.59 Further difficulties arose as a result of changes to the design of the Infraco works. An example of such works was that utilities had been moved based on the original thickness for the track slab, but the thickness of that slab had subsequently increased after the appointment of the Infraco contractor. Rather than readjusting the utilities, a decision was taken that the Infraco contractor would deal with those utilities that might require adjustment [Mr Casserly PHT00000022, pages 65–66; TRI00000111_C, page 0004].

8.60 Mr Chandler gave evidence that, quite often, the more complex utility diversions were left for resolution at a later date. That was because the advice and agreement of utility companies for what was to be done with these utilities either had not been asked for or had not been provided. SDS considered that these were matters for tie to resolve. In addition, the utility design had not been finalised in a number of locations because the location and design for OLE had still to be finalised such that it was not clear where to put the diverted utilities so as to avoid conflict [Mr Chandler TRI00000027_C, pages 0081, 0116 and 0159, paragraphs 318, 476 and 648–649 respectively]. The result of these decisions was that when the MUDFA contractor completed its works there were utilities still in the ground that would require to be moved before the Infraco contractor could undertake its work. Both SDS and tie were aware of that position.

8.61 Revision 06 of the MUDFA programme was adopted in November 2007 [CEC01520590]. The letter sending this to tie noted that, of 1,250 IFC utility drawings, 157 had been provided for completed and current worksites and 17 had been provided for review purposes. The remaining 1,076 had been assessed on the basis of first-generation designs dated August 2006 and prior to their issue by tie to AMIS in October 2006 had not been developed technically. An email from Mr Hickman in tie dated 12 December 2007 identified a number of areas in which there was a conflict between the completion dates in MUDFA revision 06 and the commencement of the Infraco works [CEC01452538]. It was suggested that in the off-street sections the conflict could be mitigated by transferring the remaining utilities scope to the Infraco contractor. The effect of leaving works to be done with the infrastructure works merely moved the problems into the future. By way of illustration, the opening statement from Mr Walker at the Mar Hall mediation noted that in a 700-metre section of works at Shandwick Place, 302 utilities had not been diverted despite the MUDFA contractor having undertaken work in that section. The remaining works had been left for the Infraco contractor to attend to as they were considered too difficult [Mr Walker TRI00000072_C, page 0083, paragraph 151].

8.62 By May 2008, when the parties were in a position to sign the Infraco contract, there were already delays in the MUDFA works. There were a number of areas where MUDFA was the early start constraint for Infraco. Mr Malkin of AMIS gave evidence that, around March 2008, he was aware that MUDFA programme revision 06 would require to be revised as a result of the progress achieved, the extent of ongoing changes and the late provision of design and design-related information. He was of the view that a more realistic end date for the MUDFA works was around March 2009. He explained that the level of confidence in the MUDFA programme was decreasing as a result of understanding the complexities of the city-centre works and the dependence on supporting organisations to provide the prerequisite details to support the works [TRI00000056_C, pages 0041 and 0046–0048]. There was widespread knowledge within tie also that the objective of having MUDFA works complete prior to Infraco works commencing would not be achieved. Mr Barclay said that by the first quarter of 2008 he was aware that the MUDFA works were not going to finish within the timescales envisaged. Mr Bell said that he was aware of difficulties and that there would be overlap with the Infraco works [PHT00000024, pages 178–180; TRI00000109_C, pages 0056, 0062–0063 and 0067]. Mr Gallagher said that he was aware that the MUDFA works would not be completed in their entirety before the Infraco works commenced but he considered that they would not impact on those works [PHT00000037, pages 122–124]. As will be noted in Chapter 12 (Contract Close), however, the difficulties and delays with the MUDFA works and the work that had been left to be dealt with during the Infraco works were not apparent in the reports prepared for financial close.

8.63 One of the pricing assumptions for the Infraco contract was that the utilities diversions were to be completed in accordance with the requirements of the Infraco programme, save for utilities diversions to be carried out by the Infraco contractor pursuant to the expenditure of the provisional sums noted in Appendix B [Pricing Assumption 24, SP4, USB00000032, page 0008]. Pricing Assumption 32 – that programming assumptions in Schedule part 15 of the Infraco contract would remain true in all respects – was also material, as the programming assumptions included an assumption that diversion of utilities would have been completed for each of the sectors by specified dates and there would be no slippage in the MUDFA programme [USB00000081, pages 0002 and 0005, paragraphs 3.1 and 9.1.3]. If these pricing assumptions were not true, that would give rise to a Notified Departure. In addition, clause 22 of the Infraco contract provided that unidentified utilities uncovered by the Infraco contractor would amount to a tie change or compensation event [CEC00036952, Part 1, pages 0088–0090].

MUDFA works in the period after contract close

8.64 As the work continued after contract close, so did the problems. According to Carillion, these problems included traffic management constraints, incomplete design, existing utilities that were unforeseen, congested and/or at shallower depths than permitted, unforeseen/unknown underground obstructions and technical queries that remained unanswered and insufficient work available on site to deploy the entire workforce. This will be discussed in paragraph 8.67 below [CEC01140099; CEC01162082]. Many of the same factors were identified within tie as being causes of the delay. A paper produced by Mr McEwan [email dated 28 July 2009 (CEC00762213) with attached paper (CEC00762214)] referred to the following factors:

• traffic management;
• the re–alignment of the tram at Haymarket;
• unexpected buried structures;
• the discovery of former graveyards in Constitution Street and Elm Row;
• the design approval process;
• the extent and number of unexpected utilities;
• additional works required by utility companies;
• increased scope of the works;
• the condition and extent of existing basement structures extending out from buildings below the road; and
• imposed embargos.

8.65 Mr Bell of tie also made reference to the unknown and unexpected utilities and obstructions, traffic management issues, embargos and delays with design but also considered there was some inconsistent performance from the contractors [PHT00000024, page 191]. Work was further delayed when, in the second half of 2008, an Infraco/SDS design safety audit resulted in the whole Haymarket area being re-designed, with the need to re-divert previously diverted utilities, and additional works along large parts of the route.

8.66 At the Tram Project Board (“TPB”) meeting in July 2008, Ms Clark reported that the team was still working to get MUDFA finished by the end of 2008 [CEC01053601, page 0006, paragraph 2.5]. Mr Gallagher blamed the delay on poor logistics and management rather than design problems and said that the TPB should not be unduly worried about progress. It seems to me that this flew in the face of the evidence that had existed for some time. On any view, what followed indicated that the time slippage was not controlled and is another indication of Mr Gallagher’s poor management that appears to have been driven by unrealistic optimism, as was apparent in his proposed solution to design delay being to be “right first time”. As was noted in Chapter 6 (Design (to May 2008)), that philosophy failed to appreciate the iterative nature of design and the interdependence of the design of different sections of the route. By October 2008, Carillion was forecasting that works would not be complete until November 2009 [CEC01140099]. That the delays could not be made good was apparent when, after agreement of the MUDFA revision 7.9 programme, by letter dated 24 March 2009, tie formally granted an extension of time for substantial completion of the works to 1 April 2009 and the longstop date for the assessment of liquidated damages in terms of clause 45 of MUDFA was revised to 3 August 2009 [CAR00000560]. By letter dated 30 April 2009, tie sent MUDFA programme revision 08 to BSC. This anticipated completion of the MUDFA works by 17 December 2009 [CEC00322635, page 0002]. The papers for the TPB on 10 February 2010 included a report to Transport Scotland that noted that utilities (including telecoms) were programmed to be completed by September 2010 [CEC00474418, page 0049]. The report to Transport Scotland included in the TPB papers for the following month noted that the forecast date for the completion of utilities (including telecoms) had slipped to December 2010 [TIE00894384, page 0048]. Despite the consistent trend of these reports, a report by the Director of City Development to CEC’s Tram Sub-Committee on 22 March 2010 giving an update on the MUDFA works said that they would be complete by September 2010 [CEC01891483].

8.67 In October 2008, Carillion noted that the unexpected utilities and the delay in design meant that it was unable to deploy its resources in an efficient manner. It claimed that there was not sufficient work for its staff. It also claimed that the result of this was that it had demobilised 35 operatives and was likely to demobilise a further 30. Moreover, the combined effect of the Christmas embargo on undertaking work at certain locations, the threat of extending the embargo to Leith Walk in December 2008 and the lack of IFC designs to enable work to proceed on sites unaffected by the embargo meant that Carillion would have little option other than to demobilise the majority of its 250 resources due to lack of available work fronts between early November 2008 and late January 2009 [CEC01140099]. Carillion noted its concern that, having demobilised these resources, there was no guarantee that they would be available in future.

8.68 As early as January 2009 there was a view within tie that attempts should be made to restrict the scope of Carillion’s work. In an email dated 28 January 2009 [CEC01145982; CEC01145983] Mr McEwan recommended that tie look to cut back to an “irreducible minimum” the amount of work that tie could expect Carillion to carry out, with works being transferred from Carillion to the Infraco contractor.

8.69 It was apparent that Carillion was also unhappy with the contract. During the works, on a number of occasions arguments were advanced as to why it would be appropriate to change the basis on which it was paid under the contract. In particular, it sought to move away from the rates and prices agreed in MUDFA to payment on a cost plus basis (ie where it would be paid the actual costs that it incurred plus a management fee). Mr Malkin gave evidence that the rates and prices tendered by AMIS had assumed that AMIS would be provided with detailed drawings and work orders to enable it to plan, execute and value the works efficiently and that AMIS considered that the inability of tie and SDS to provide AMIS with all of the work ordering information to which it was entitled under MUDFA resulted in the entire contract premise being unworkable [TRI00000056_C, pages 0030 and 0039]. This was, however, resisted by tie [Mr Barclay PHT00000031, pages 69–72]. Nevertheless, tie did accept that delay and disruption between October 2007 and September 2008 had arisen as a result of the number and extent of unidentified services encountered, design delivery, resolution of technical queries and traffic management arrangements and that there was a contractual entitlement as a result of these factors. Consequently, on 19 and 23 March 2009, tie and Carillion entered into a settlement agreement [CAR00000243] in terms of which tie agreed to pay Carillion £1.2 million (comprising £1,050,000 in respect of delay and disruption between October 2007 and 30 September 2008 and £150,000 for measurement items). This was immediately before the extension of time referred to in paragraph 8.66 above. Within tie a view had formed that Carillion was losing money on MUDFA [Mr Bell TRI00000109_C, pages 0115 and 0117]. Mr Barclay explained his view that AMIS had based its tender on a certain rate for its workforce, but found that to carry out the work at the time that it was required it had to use sub-contractors who charged rates higher than those set out in MUDFA [Mr Casserly TRI00000111_C, page 0050; Mr Malkin TRI00000056_C, page 0014].

8.70 A presentation to the TPB on 6 May 2009 noted that, for performance, quality and cost reasons, it was proposed to close the Carillion contract, with the preferred option being to remove the entire utility diversion works at section 1A between Newhaven and Tower Place bridge, and all other outstanding works as at 31 July 2009, from the Carillion scope, and to award these to suitable contractors following competitive tender [CEC01026346; CEC00633071, page 0026]. It was envisaged that the management of the outstanding MUDFA works would take place in conjunction with the infrastructure works, “ensuring optimal synchronisation of the overall programme aims”. It was considered that the transfer of works to other contractors would result in a reduced delivery and construction period and a cost saving.

8.71 In September 2009, Carillion made a claim for losses suffered as a result of delay and disruption from September 2008. With more than 1,600 technical queries and 3,700 change items raised to date, Carillion anticipated a final account value in excess of £70 million. The document stated that tie had repeatedly failed to follow the work order procedure set out under MUDFA whereby, prior to starting the works, AMIS ought to have received a fully compliant design (other than unforeseen obstructions) for which there would be an agreed price and programme finalised by a works order confirmation notice [CEC00790177, pages 0002–0003].

8.72 By the time of the TPB meeting on 21 October 2009, commercial discussions with Carillion were under way regarding its exit from the project. On 4 December 2009, tie and Carillion entered into an agreement bringing MUDFA to an early end [CAR00000145; see also the unsigned agreement, CAR00000429, page 0008 onwards, for more legible versions of the schedules. Although neither of these versions of the agreement bears a date, the later agreement referred to in paragraph 8.74 below indicates that it was signed on 4 December 2009.]. Under the agreement Carillion was relieved of any responsibility for carrying out the remaining works under MUDFA and, as at 4 December 2009, was treated as having completed all of the works required of it (with the exception of the works detailed in Appendix 1 of the agreement). The defects correction period under MUDFA was extended to 3 December 2011 [CAR00000145, page 0005, clause 13]. Farrans and Clancy Dowcra were appointed to undertake the remaining utilities diversion works.

8.73 The minutes of the TPB meeting on 21 October 2010 noted that the utilities works in the Airport to Haymarket section were complete. The utilities works in the Haymarket to Newhaven section were complete save for some telecoms cabling and transfers, testing/commissioning/abandonments of transferred services and the Baltic Street diversions (1,500 metres) [CEC00014175, pages 0009–0010].

8.74 On 10 November 2010, tie and Carillion entered into a further minute of agreement settling outstanding claims between them [TIE00094413]. Under this second agreement tie agreed to pay Carillion a further, and final, amount of £5,824,000 plus any applicable VAT(which amount was in respect of both utilities work carried out by Carillion and Carillion’s claim for delay and disruption from September 2008 onwards). This amount reflected agreement of Carillion’s final account in the amount of £62,500,000. An email dated 19 November 2010 from tie attached a spreadsheet giving a breakdown of that amount [TIE00094459; TIE00094460]. An amount of £56,676,756 had previously been paid, resulting in an outstanding balance of £5,823,243 (£62,500,000 – £56,676,756). tie reserved its right to claim against Carillion for future defects (ie excluding defects that were notified by tie to Carillion, or known to tie, prior to the date of the agreement) [TIE00094413, page 0002, clause 5.1].

After Mar Hall

8.75 As will be considered in Chapter 19 (Mediation and Settlement), the agreement ultimately reached at the conclusion of the Mar Hall mediation between tie and Bilfinger Berger, Siemens and CAF was that Bilfinger Berger Siemens (“BBS”) would complete the tram line from the Airport to St Andrew Square/York Place. Significantly, even at this stage it was a stipulation in the Heads of Terms following the mediation that BSC would take none of the risk of all the utility diversion works required to complete the on-street works between Haymarket and St Andrew Square/York Place [CEC02084685, page 0003, paragraph 5.4]. Although this was a continuation of the position that had existed up to that date, much of the remainder of the approach to the utilities was to change. For the remainder of the works there were to be further investigations, a new governance structure, a new way of working, a new contractor and a new contract.

8.76 Probably the key change that was made following Mar Hall was to the relationship between the utilities works and the infrastructure works. Whereas a core concept within MUDFA had been that the works to utilities would be completed in advance of the infrastructure works and the roads reinstated, so that the Infraco contractor would have a “clear field” in which to work, the approach after Mar Hall was that a “bow wave” approach would be adopted. This entailed that the utilities works would be done just ahead of the infrastructure works. The utilities contractor went in just before BSC, excavated down to formation level and resolved the utility conflicts before BSC then came on site to undertake the infrastructure works. This removed the need to reinstate the road surface and then to re-excavate for the main works. BSC agreed to change its approach to allow multi-working on the various sites, ie with the utility and Infraco contractors working with and around each other [Mr C Smith TRI00000143_C, pages 0072–0073].

8.77 After the mediation, McNicholas Construction Services Limited (“McNicholas”) was appointed to carry out the remaining utility works. It was appointed under the New Engineering Contract standard form engineering contract rather than a bespoke contract as had been used before. It was to be paid on a cost-reimbursable basis rather than in accordance with an agreed schedule of rates [CEC01889514, page 0003]. This was the basis on which Carillion had sought to be paid. Although this form of contract had been selected before Turner & Townsend was involved, Mr Weatherley said that the advantage of it was that it was not necessary to agree the cost of each item of additional works as it was required so that the contactor was able to respond quickly, thereby minimising the impact of utilities works on the Infraco contract [TRI00000103, pages 0031–0032, paragraph 51].

8.78 As noted in Chapter 20 (Post-Mar Hall), there was a general change to the way in which the project was managed, with the removal of tie, the introduction of Turner & Townsend as project managers, and the commencement of new governance arrangements. In relation specifically to the utilities works, a weekly utilities meeting was established to discuss and agree utilities issues. In addition, a utilities project manager was appointed, who acted like a clerk of works. He had a roving role to move up and down the site; problems discovered on the ground were immediately brought to his attention and there was a short line of communication from the utilities project manager to Mr C Smith to enable issues to be resolved or recorded quickly [Mr C Smith PHT00000053, pages 138–146].

8.79 Mr Weatherley explained that a representative from each of the utility companies was co-located with the on-street works team, including experienced project managers, commercial managers and McNicholas, to identify what the best solution would be to the utility diversions where conflicts arose [PHT00000046, page 86; TRI00000103, pages 0028 and 0030, paragraphs 44 and 48]. He gave evidence that when Turner & Townsend was mobilised there remained a significant quantity of unresolved legacy design issues with Scottish Water and unresolved utility clashes with the proposed tram works [ibid, page 0024, paragraph 35].

8.80 A report to CEC on 25 August 2011 noted that further investigations in relation to utilities had been instructed on key sections of the on-street works between Haymarket and York Place. These were intended, in particular, to identify conflicts arising as a result of the finalised design, including the locations for OLE poles [TRS00011725, pages 0002–0003]. These further investigations had identified approximately 550 potential utilities conflicts (although it was not believed that all of those lay on the critical path of the tram). Mr Weatherley said that Turner & Townsend undertook a desk-top exercise, and dug slit trenches, to try to understand the likely scale of the risk from utilities [TRI00000103, page 0028 and 0030, paragraphs 44 and 48]. The first location that was excavated was at Haymarket, and that uncovered a number of unknown utilities. This influenced Turner & Townsend’s decision to undertake quite extensive surveys elsewhere in the on-street section. The full extent of utilities-related issues and conflicts only became clear, however, as the on-street excavations took place. Where conflicts were identified it was not always the utility works that were changed. Some of the infrastructure works were re-designed to avoid clashes with utilities.

8.81 As was noted in paragraph 8.80 above, as investigations continued, a more complete picture emerged of the utilities work that remained to be done. The minutes of the joint project forum on 18 October 2011 noted that, as a result of the digging of test trenches, an increase in the number of potential conflicts from 550 to 895 had been identified [CEC01890987, page 0007, paragraph 6.2]. By November, the minutes noted that the number of clashes was 1,127 [CEC01890994, page 0006, paragraph 4.1]. The numbers of conflicts that are quoted from time to time are not easy to reconcile. In the progress report from Turner & Townsend in November 2011 it was noted that there were currently 1,128 utility conflicts, of which 811 were “live” [CEC02085662, page 0013]. The Turner & Townsend progress report for 9 November to 8 December 2011 noted that the total number of known utility conflicts remained at approximately 820, but that a total of 1,645 utility issues had been identified [CEC01891191, page 0003]. These issues included unverified potential clashes and legacy utility issues, largely associated with Scottish Water.

8.82 The scope of work that remained to be done on utilities after Mar Hall cannot be reconciled with various statements in project reports, at meetings of the TPB and otherwise that the MUDFA works were largely complete. Mr Foerder suggested that assuming that the reports were not deliberately inaccurate, it suggests that tie did not at any stage have a grasp of the full scope, cost, timescale or impact of the utility diversions required [TRI00000118, pages 0043–0044, paragraph 9.7]. I agree with this assessment.

8.83 In relation to the utilities work that was required in September 2011, in July 2014 Turner & Townsend produced a draft preliminary report that identified a fundamental difference of opinion between tie and BSC over the definition of a “utility” and what “utility free” meant [WED00000103, page 0056]. The view taken by tie was that something should be considered a utility only if it was live and had been provided by a utility company (ie Scottish Power, SGN, Scottish Water or BT) and certain communications companies. This excluded existing street lighting, power and data supplies to bus stops etc. and legacy ducting. In contrast, BSC had defined utilities as including both “live” and “dead” services, as provided by the utility companies and communications companies, and including street lighting, power and data supplies to bus stops, along with all other “live” power and communication feeds [ibid, pages 0056–0057, paragraph 4.9.1]. Which of these two views was taken would have a material impact on the scope of works to be carried out, and I find it remarkable that the difference of opinion emerged only at this very late stage.

8.84 Despite these problems, in the period after Mar Hall the necessary work was carried out more smoothly, with less confrontation with the contractor and within the required timescale. Mr Foerder said that the new approach led to better planning of resources and fewer abortive works [TRI00000118, page 0116, paragraph 21.3.4]. Although this was very different from what had happened during MUDFA, not everything changed. As will be discussed in paragraphs 20.92–20.93 below, the estimate of the cost to complete utilities works was £2.91 million, of which £1.1 million was for legacy works in Leith Walk [WED00000092, page 0001], but Mr C Smith considered that the information provided by tie in that regard was unreliable. In the post-Mar Hall budget £5 million was allocated to utilities risks [TRI00000153, page 0020]. The final account for the utility works completed after Mar Hall exceeded £20 million [ibid, page 0065].

8.85 With the benefit of hindsight, the “bow wave” approach appears to have much to recommend it over the initial approach, but I think it would be inappropriate simply on that basis to reach a judgement that it should have been the approach used from the outset. As was noted in paragraph 8.76 above, the “bow wave” approach required a large degree of co–operation between utility contractors, infrastructure contractors, designers, utility companies and the employer – and that did not exist prior to Mar Hall. In my view, therefore, the critical changes made were those that brought about a way of working that was flexible, responsive, co–operative and agile. It is clearly a process in which a number of different parties have valid interests, all of which must be accommodated. The original way of seeking a consensus was cumbersome and unwieldy and where there is any significant extent of works to be carried out in a city centre, it will pose a substantial obstacle. This was made much worse by the absence of a clear statement as to where design responsibility lay and the delays in design that affected the project generally.

8.86 Clarifying where the design responsibilities lie is a matter of proper drafting of the contracts. I have considered the issues that led to problems with design generally, and what might have been done to address them, in Chapter 6 (Design (to May 2008)) and Chapter 7 (Design Following Novation and Contract Close). In relation to providing more co-operation from the utility companies, I consider that the code of practice under the 1991 Act should be updated to include such an obligation where major transport projects are undertaken. The utility companies may object that this imposes a burden and cost to them in relation to transport projects that are not specifically in their interest. Although that may be true, I consider that the imposition of a more onerous duty to co-operate and to participate in meetings to have matters resolved can be seen as part of the price that is to be paid for the otherwise free use of the roads to accommodate their business assets. Whether the costs are placed on the utilities or on the promoter of transport schemes the reality is that they will ultimately be borne by the public in the form of increased utility charges or council tax on the one hand, or increased costs of public transport projects on the other. I readily accept that an argument can be made that these costs should initially be said to arise from the transport project such that they should be allocated to the promoters. Ultimately, however, the primary function of roads is to permit passage and my view is that this should be given priority in situations where it is necessary to balance the interests of transport projects and the interests of utility suppliers.

8.87 The issue of whether it was the MUDFA works or the Infraco design issues that had the greatest impact on the Infraco works does not admit of an answer. Clearly, both had an impact. Not all the delays in the MUDFA works were attributable to AMIS/Carillion, however. This is most clearly demonstrated by the fact that they were granted extensions of time under the contract to complete their works [CAR00000022; CAR00000398; CAR00000163]. In addition, in a letter dated 9 April 2008 in recognition of the issues that had adversely affected AMIS’s contract up to 30 September 2007, tie agreed to pay AMIS an additional £991,142.95 [CAR00000074].