Management of the Holyrood building project · 1. The subject of my report is the management of the...

92
Prepared for the Auditor General for Scotland Management of the Holyrood building project Insert sub head June 2004

Transcript of Management of the Holyrood building project · 1. The subject of my report is the management of the...

Page 1: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Prepared for the Auditor General for Scotland

Management of theHolyrood building project Insert sub head

June 2004

Page 2: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Auditor General for ScotlandThe Auditor General for Scotland is the Parliament’s watchdog forensuring propriety and value for money in the spending of public funds.

He is responsible for investigating whether public spending bodiesachieve the best possible value for money and adhere to the higheststandards of financial management.

He is independent and not subject to the control of any member of theScottish Executive or the Parliament.

The Auditor General is responsible for securing the audit of the ScottishExecutive and most other public sector bodies except local authoritiesand fire and police boards.

The following bodies fall within the remit of the Auditor General:

• departments of the Scottish Executive eg the Health Department• executive agencies eg the Prison Service, Historic Scotland• NHS boards and trusts• further education colleges• Scottish Water• NDPBs and others eg Scottish Enterprise.

AcknowledgementsWe gratefully acknowledge the Corporate Body’s permission to use theimages on pages 10, 11, 14, 15, 37, 42, 50 and 62. All other images arecopyright of Audit Scotland.

Audit Scotland is a statutory body set up in April 2000under the Public Finance and Accountability (Scotland) Act2000. It provides services to the Auditor General forScotland and the Accounts Commission. Together theyensure that the Scottish Executive and public sector bodiesin Scotland are held to account for the proper, efficient andeffective use of public funds.

Page 3: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Contents1

Key findingsPage 4

Part 1. IntroductionPage 10

Part 2. Project overviewPage 14

The new Parliament is a striking andvery complex building with highquality materials and some unusualdesign and construction featuresPage 14

The Parliament has been procuredusing construction managementPage 14

The £431 million current estimatedproject cost includes a large marginfor risk and contingencyPage 16

The project has suffered successiveslippage in its forecast completiondatePage 16

Construction is now well advancedPage 16

In summer 2000, compared to earlierdifficulties, there seemed a firm basisfor the Holyrood project to moveforwardPage 18

There was at best only qualifiedagreement of the necessary costplan in November 2000Page 20

In November 2000 estimated riskwas significantly more than theavailable budgetPage 21

By June 2001 the Corporate Bodyhad concluded the £195 milliontarget was not achievablePage 23

Contents

In June 2001 a fresh Parliamentarymotion relaxed the earlier budget andcompletion targetsPage 24

The procurement of works changedradically because of programmepressures in 2000 and 2001Page 24

As construction intensified in 2001programme monitoring andcoordination became more importantPage 27

Cost and time problems caused theCorporate Body to consider stoppingthe project to take stock inNovember 2002Page 28

From January 2003 the aim was tocomplete the project as quickly aspossible as the best way ofcontaining costsPage 30

The First Minister announced aninquiry into the increasing costs anddelay affecting the project in June2003Page 32

Revised arrangements for costreporting commenced in July 2003and a fee cap was agreed in August2003Page 32

Construction reached a peak inDecember 2003Page 32

The current programme was agreedin February 2004 for substantialcompletion in August 2004 at amaximum cost of £431 millionPage 35

Ninety per cent of final accounts forconstruction remain to be finallysettledPage 35

Part 3. The reasons for laterdeliveryPage 37

This analysis concentrates on whythe project did not achieve the 2000target for completion by the end of2002Page 37

The main cause of the slippage isdelays in design of a challengingproject delivered against a tighttimetable using an unusualprocurement routePage 39

Under construction management theclient retains construction riskPage 39

The complexity of the Holyroodproject has involved major challengesin programming the constructionPage 40

The client maintained a drive for theearliest achievable completion datePage 43

Design slippage was a major factor indelaying the overall programmePage 46

Part 4. The reasons for increasedcostsPage 50

There is a question whether the costplan underpinning the £195 milliontarget was adequatePage 50

In most of the trade contracts theestimated final cost greatly exceedsthe original cost plan allowancePage 51

Some of the increased costs are dueto the extended construction periodPage 51

Page 4: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

2

Some of the increased costs are dueto a very high level of designdevelopmentPage 53

The main reasons for cost increasessince 2000 are design developmentand delay in the construction processPage 56

Uncompetitive procurement hascontributed to increased costsPage 56

The focus on programme deadlinesdrove procurement and led to highercostsPage 56

Part 5. Project management andcontrolPage 62

Throughout the project there wastension between the objectives oftime, quality and costPage 62

There were problems in ensuringclear leadership and control on thepart of the client organisationPage 65

After June 2001 projectmanagement did not work within anoverall budget or approved costceiling for the whole projectPage 69

Cost reporting and financial controlshould have been better developedat all levels of the Holyrood projectPage 71

The approach to risk managementwas not fully consistent with goodpracticePage 72

Capping has controlled consultantsfees but only late in constructionPage 73

Project management did notimplement constructionmanagement fully in accordance withusual practicePage 74

Appendix 1. Trade Contract casestudiesPage 78

Appendix 2A. East framePage 80

Appendix 2B. MSP bay windowsPage 82

Appendix 2C. Specialist glazing(Stages 1&2)Page 84

Appendix 2D. Foyer frame andglazingPage 86

Appendix 2E. Assembly roofPage 88

Page 5: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

3

Exhibit 1: The Auditor General’s2000 report on the HolyroodParliament building project Page 12

Exhibit 2: Changes in the forecastcost and completion of the Holyroodproject Page 13

Exhibit 3: The new Parliamentbuilding Page 15

Exhibit 4: Key features of constructionmanagement procurement Page 17

Exhibit 5: Holyrood project costsreported to the Finance Committee2000 to 2004 Page 17

Exhibit 6: Holyrood project – analysisof current forecast costs Page 18

Exhibit 7: Holyrood – mainprogramme revisions 2000 to 2004Page 19

Exhibit 8: AGS 2000recommendations on risk and costcontrol Page 20

Exhibit 9: Uncertainty in theNovember 2000 cost plan for theHolyrood building Page 22

Exhibit 10: Estimated area of theHolyrood building Page 22

Exhibit 11: The Parliament’sHolyrood resolution 21 June 2001Page 25

Exhibit 12: Slippage in programmes4 and 5 in 2001 and 2002 Page 26

Exhibit 13: Audit findings on theFlour City contract Page 27

Exhibit 14: Tender approval for 20main Holyrood construction contracts Page 29

Exhibit 15: Slippage in programme 6in 2002 Page 31

Exhibit 16: Slippage in programme 7in 2003 Page 33

Exhibit 17: Holyrood inquiry – FirstMinister’s announcement June 2003Page 34

Exhibit 18: Slippage in completion ofthe main parts of the Holyroodbuilding Page 38

Exhibit 19: Key features of traditionalconstruction Page 39

Exhibit 20: Post tender changes incontract costs Page 41

Exhibit 21: Design development ofthe foyer roof Page 42

Exhibit 22: Problems ofinterdependency and complexity ofthe Holyrood project Page 43

Exhibit 23: Examples of programmequalifications Page 44

Exhibit 24: Initial contract delays foreight large works contracts Page 47

Exhibit 25: Design variations for fivecritical works contracts Page 47

Exhibit 26: Analysis of impact ofdesign variations on the mainassembly buildings contract Page 48

Exhibit 27: Increase in the forecastcost of the Holyrood project sinceSeptember 2000 Page 51

Exhibit 28: Initial and currentestimated costs for 58 Holyroodconstruction contracts Page 52

Exhibit 29: Holyrood – examples oftime delay resulting in extra costsPage 54

Exhibit 30: Holyrood – estimatedconstruction costs from 2000 to2004 Page 55

Exhibit 31: Design development ofthe foyer roof and glazing packagePage 55

Exhibit 32: Main reasons forconstruction cost increases sinceSeptember 2000 Page 57

Exhibit 33: Audit examination of 20works contracts Page 57

Exhibit 34: Competition for 20Holyrood works contracts Page 58

Exhibit 35: Tender recommendationto start on site for 11 Holyroodcontracts Page 60

Exhibit 36: Examples of contractswith significant design uncertainty atcontract stage Page 60

Exhibit 37: AGS recommendationson project management September2000 Page 63

Exhibit 38: Extracts from ScottishParliament building user brief Page 64

Exhibit 39: Achievement of prioritiesfor the Holyrood project Page 65

Exhibit 40: Traditional projectorganisation Page 66

Exhibit 41: Holyrood projectorganisation from June 2000 Page 67

Exhibit 42: Role of the HolyroodProgress Group Page 68

Exhibit 43: Design teammanagement Page 69

Exhibit 44: Project execution plansPage 70

Exhibit 45: Corporate Body capitalbudget and expenditure 1999 to2004 Page 70

Exhibit 46: Comparative unitconstruction costs for high qualitybuildings Page 71

Exhibit 47: Estimated constructioncosts for the Holyrood project 2000-04 Page 73

Exhibit 48: Special features ofworking with a signature architectPage 77

List of exhibits

Page 6: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

There were difficulties associated with the construction of a very complex, densely developed, unusual buildingagainst very tight deadlines.

4

Key findings

Page 7: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

5

1. The subject of my report is themanagement of the project toprovide the new Scottish Parliamentbuilding (the Holyrood project):

• Part 1 briefly describes myprevious reports on the Holyroodproject.

• Part 2 outlines what hashappened to the project since myreport of September 2000.

• Part 3 examines why the projectwas not completed by the earliertarget of December 2002 andwhy there was subsequentslippage in the programme.

• Part 4 concentrates on whyforecast project costs more thandoubled from £195 million1 inSeptember 2000 to £431 millionin February 2004.

• Part 5 assesses the managementand control of the Holyroodproject in the four years since my2000 report.

2. The Holyrood project was anextremely difficult and complexbuilding project. The estimated costincreased by some £220 million andthere has been 20 months slippageover the last four years. In examiningwhy this happened, it is important totake account of the quality of thebuilding, which seems likely tosatisfy the high standards specified in1998.

3. The difficulties of delivering theHolyrood building using the‘construction management’ methodof procurement lie at the heart of theproblems that arose. In my 2000report I commented on the strengthsand weaknesses of the constructionmanagement method of procuringbuilding projects. I drew attention tothe fact that with this method mostof the risks stay with the client ratherthan transferring to the contractors.

4. My report concerns themanagement and control processesapplied to the Holyrood project. Thisincludes the role of the HolyroodProgress Group but my report

focuses on the performance ofproject management for which thePrincipal Accountable Officer of theScottish Parliamentary CorporateBody is responsible. In undertakingthis examination I did not seek toform an opinion on whether anyindividual contractor has been atfault. Project management isresponsible for managing itsconsultants and contractors andassessing their performance.

• The client is the ScottishParliament Corporate Body onbehalf of the Parliament. TheHolyrood Progress Group hasassisted the Corporate Body in itsfunctions as client.

• Project management in this caseis the Chief Executive of theParliament (who is the ‘projectowner’) and the project teamunder the leadership of theproject director. Projectmanagement is responsible formanaging and delivering theproject with advice and guidancefrom the Progress Group.

1 Equivalent to £209 million including £14 million estimated landscaping costs in 2000.

Page 8: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

6

• The design team are thearchitects and the otherconsultants appointed by projectmanagement. The architects areEMBT/RMJM Limited, which is apartnership between architectsbased in Barcelona andEdinburgh. The other principaldesign team members are OveArup and Partners and RMJMScotland Limited.

• The cost consultant and quantitysurveyor is Davis Langdon andEverest.

• The construction manager isBovis Lend Lease (Scotland)Limited. They are responsible forcoordinating the design team andthe organisation of the site. Theyalso manage all construction workscontracts, but the client remainsthe employer for all contracts.

5. Lord Fraser was appointed in June2003 by the First Minister and thePresiding Officer to conduct aninquiry into the Holyrood project. Myreport will be available to Lord Fraseras he prepares his report which isdue for completion in September2004.

Slippage

6. The main cause of the 20 monthsdelay to the project since September2000 was the production of detaileddesign variations and the late supplyof information during theconstruction process.

7. There were difficulties associatedwith the construction of a verycomplex, densely developed, unusualbuilding against very tight deadlines.In some cases trade contractorswere responsible for some elementsof design. Both the architects andsome trade contractors did notdeliver on time some criticalelements of the design work.

8. Project management required avery demanding timetable forcompletion. The constructionmanager consistently sought toachieve the required early targetcompletion dates, although with thebenefit of hindsight the programmeset in December 2000 for completionby December 2002 was probablyunachievable. By September 2002 orApril 2003 the construction managerand project management should alsohave recognised its targets wereunlikely to be achieved.

9. Project management should havedone more to address the rootcauses of problems, which wereadversely affecting the cost andprogramme. The constructionmanager repeatedly preparedconstruction programmes, whichincluded assumptions andcommitments by the design teamand contractors that weresubsequently not achieved. Underconstruction management the clientultimately bears most constructionrisk. As slippage became evidentproject management did test eachnew revision of the programme andsought to ensure that the designteam, the construction manager andthe trade package contractors hadthe resources and commitment todeliver on time. But it was unable tofind the means to manage theserisks effectively.

Cost increases

10. The client did not significantlyalter its requirements after themiddle of 2000. The size and layoutof the building were not materiallyaltered. But the cost of realising thisdesign escalated enormously.

11. The main reasons forconstruction cost increases after2000 were design development anddelay in the construction process.The design development was entirely

related to realising the detail of thebuilding and aspects such as thequality of finish and the palette ofmaterials that were used, inaccordance with the client’srequirements. Construction costs(including irrecoverable VAT) accountfor 72% of the £431 million totalproject costs. Between 2000 and2004 construction cost estimates(including VAT) rose from £140million to £311 million, an increase ofmore than 220%.

12. Detailed development of theapproved design added £80 million(including VAT)2 to construction costs.Design development is the processwhen the design of a buildingevolves in parallel with the tenderingand appointment of contractors andsubsequent building work on site.Design development carries a risk ofcost increases. There should beadequate allowance for this riskwhen the design and its costs areapproved at ‘Stage D’ of thearchitectural design. The risks shouldthen be managed. In the Holyroodproject, however, designdevelopment has driven up thecosts. For this project, designdevelopment became a process ofcosting a developing design ratherthan developing the design within acost limit.

13. Because many of the workspackages were let when there wasuncertainly about the work involved,it was difficult to achieve goodcompetition and deliver value formoney. My examination included areview of 20 of the main constructioncontracts for the Holyrood project,representing 56% of the estimatedconstruction cost. Thirteen of these20 major contracts had three orfewer tenders. This was an unusuallylow number for contracts of this size.For 11 of the 20 major Holyroodcontracts examined there wassignificant uncertainty about the

2 Estimate by Audit Scotland based on the audit examination.

Page 9: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Key findings 7

detailed design at the tender stage.In normal contracts most of the costsare fixed at the point when thecontract is awarded, but this was notpossible in these contracts becausesignificant parts of the design and/orthe full extent of the works had notbeen determined at that point. Thisrestricted price competition and ledto much more negotiation with thecontractors than was desirable.

14. The decisions to award contractswith a large degree of uncertaintywere the result of the client’sprogramme requirements. Byawarding contracts for work withuncertain scope and design the clientwas in a weak position to resistsubsequent claims from contractorsfor extra time-related costs. And incases where the performance oftrade contractors might not havebeen satisfactory there was littleopportunity to attribute delays tothese individual contractors becauseof delays occurring elsewhere in theoverall programme.

15. The uncompetitive procurementof works packages has allowedcontractors’ claims for prolongation(ie, the extra costs of doing workover a longer period), disruption anddelay, which have added £86 million(including VAT)2 to the constructioncosts. It is not possible to say howmuch more it has cost to procurework priced mainly by negotiationwith single contractors rather thanwork obtained through competitivepricing of work. The costs fordisruption and delay are not forimprovements in the designspecified by the client or extrafeatures that added value to theproject. These extra costs are theconsequence of not fully preplanningthe construction works.

Project management and control

16. Although it is likely that a highquality building is being delivered, thetime and cost objectives have notbeen met. The same quality couldhave been achieved for less if thewhole design and constructionprocess had been better executed. Inany construction project it isnecessary to balance quality withtime and cost. In 1998 the clientrequired that the building should becompleted by summer 2001. Timewas a priority. Quality has beenequally important throughout theproject. The client also set cost limitsthat applied in the early stagesalthough it was unclear about howimportant cost was compared withtime and quality.

17. The organisation of the Holyroodproject should have had a singlepoint of leadership and control wheredecisions could be taken about howto balance time, cost and quality aspart of the client decision-makingprocess. Leadership and control ofthe project was not clearlyestablished. Normally leadership andcontrol should reside with the projectdirector (sometimes called theproject sponsor) who is a keymember of project management.The client gives the project directorthe responsibility for making theproject happen. In the Holyroodproject there was no single point ofleadership and control. Responsibilityand accountability for managing theindividual aspects of time, cost andquality was not clearly allocatedbetween the various different parties.

18. The parties leading the project didnot fully agree a cost plan. In April2000 the Parliament had set a £195million budget for completing theproject. In my 2000 report Irecommended that projectmanagement, the design team andthe construction manager should

agree a cost plan to provide a soundbasis for managing the remainingstages of the project within thisbudget. Although a draft plan wasprepared in Autumn 2000 most ofthe information in it was an indicatorof the costs rather than a reliableestimate of the costs.

19. Project management and theclient did not use normal budgetarycontrol procedures to allow abalanced consideration ofaffordability, quality and time in thedecision-making process. In June2001 the Parliament approved amotion that can be interpreted asremoving the previous overall budgetconstraint of £195 million. After thisdecision by the Parliament, projectmanagement did not establish analternative overall budget or approvedcost ceiling that would allow thecosts of the project to be properlymanaged. Because there was not abudget based on sound costestimates after June 2001, there wasa risk that project managementwould concentrate on achieving highquality and tight deadlines withouttaking full account of the costimplications.

20. There should have been bettercost reporting and financial control. Inmy 2000 report I recommended thatproject management should reviewand report project costs regularly(monthly) to the client on acomprehensive and systematic basis.The Accountable Officer advised theAudit Committee in October 2000that he had implemented thisrecommendation. But subsequentfinancial reporting of the project wasnot always comprehensive orsystematic. For example, regularreporting to the Parliament’s FinanceCommittee of the total estimatedcosts of the project (including, forexample, the landscaping costs) didnot start until July 2003.

Page 10: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

8

21. Risk management for theHolyrood project was not goodpractice. In my 2000 report Iconcluded that accounting for riskwas insufficient. I showed thatcontrary to good practice there wasno quantified allowance for the majorrisks facing the project. Irecommended that this should beestablished and the results used as abasis for an action plan to managethe risks. Project managementintroduced a process for quantifyingrisks from October 2000 and thenconducted a number of risk reviews.However, in the Holyrood project thegeneral approach was to accept costincreases and include them in theforecasts as the risks materialised.Because there was no agreedbudget limit after June 2001, there islittle evidence that forceful actionwas taken to prevent or reduce theincreases in cost.

22. In my opinion projectmanagement could have taken moreaction at an earlier stage to controlexpenditure on consultants. In my2000 report I suggested that beforethey appointed consultants, projectmanagement could have exploredmore carefully alternative feearrangements including financialincentives linked to delivering valuefor money. Percentage fees do notalign the objectives of the client withthe commercial objectives of theconsulting firms because the more aproject costs the more consultantsare paid. In 2000 the estimated feecost was £23 million3, approximately19% of the approved £119 millionconstruction cost includingcontingency at that time. The clientsecured fee capping for one of itsconsultants in 2000 and for the otherconsultants in 2003. Fee costs arenow forecast at £50 million4, 19% ofthe approved construction cost.

23. The Corporate Body limited itsexposure to increases in consultants’fees in 2003, but this was very late inthe programme, after the fees hadincreased significantly. Theagreement to the fee capping at thislate stage in the project did notprovide a timely incentive toconsultants to control costs andprogramme. Prior to fee capping inJuly and August 2003, there was noregular reporting of the significantexpenditure on the consultants’ feesfor the Holyrood project. Projectmanagement did not seek to convertits construction manager’s fee to afixed lump sum before July 2003although a clear opportunity topursue this was available from 2000.

24. Although project managementraised some significant questionsabout some aspects of some of itsconsultants’ work, it should havesystematically assessed theirperformance. An assessment wasneeded to safeguard public fundsand to ensure fee costs providedvalue for money. If projectmanagement was able to showthere had been significantunderperformance by any of itsconsultants there may be options forrecovering some of its additionalcosts.

25. Construction management is anunusual method of procuringconstruction projects in the publicsector and it has not been usedbefore for a major public buildingproject in Scotland. The experienceand expertise in constructionmanagement was not present in theearly stages of the Holyrood projectand therefore the risks andchallenges were not fully appreciatedby the client and projectmanagement.

Lessons from the Holyrood project

for public sector procurement

26. In the recent history of Scotlandthere has not been a public buildingproject as complex or as difficult todeliver as the Holyrood project. Themain lesson is, however, one thatcould be applied to all significantbuilding projects, namely that theform of contracting must always bechosen with care, with a soundappreciation of the risks and benefitsof each of the procurement options.

27. The different forms of contractingare intended to transfer risk to thosebest able to manage it. Underconstruction management, wheredesign is incomplete and uncertainwhen construction starts, the risksstay with the client. It is essentialtherefore that the client managesdesign development and has aproject team that gives a key role toprofessionals who are experienced inthis construction method.

28. In general, however, constructionmanagement is unsuited for mostbuilding projects in the public sector.There is sound advice available fromseveral sources in government andthe construction industry on differentforms of contracting.

29. In recent years the Office ofGovernment Commerce hasintroduced ‘gateway reviews’ formajor public sector procurementprojects. These reviews allow for aqualified team to scrutinise thebusiness need for a project at keystages in its lifecycle, before keycontracts are awarded, to provideassurance that it can progresssuccessfully to the next stage. Thisprocess is now applied acrossgovernment5.

3 Excluding site organisation costs of £5 million in 2000.4 Excluding site organisation costs of £18 million currently.5 The decision to proceed with Holyrood was made before the requirement to follow the gateway process was introduced.

Page 11: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Key findings 9

30. In all projects, care should betaken to put in place a paymentregime that provides incentives tocontractors to perform against cleartargets for quality, time and cost.

31. In complex public sector projects,the client should ensure that there isa single point of control andleadership for the project, withexplicit authority and responsibilitygiven to the person in charge.

32. In all major projects there shouldbe an agreed project budget and aset of key performance indicatorsthat should be used to measureperformance during the life of theproject. For example theperformance that is being deliveredin relation to cost, time and qualitycan be compared with theperformance that has been deliveredin similar projects elsewhere.

33. Whatever construction method ischosen, sufficient time should beavailable for the planning stage,before construction starts. Goodplanning will involve (a) getting theconstruction sequence right to avoiddelays and extra costs, (b) assessingand managing project risks and (c)using value management to assessthe contribution of each part of theconstruction process to removewaste and inefficiency. There mustalways be sufficient time forprocurement to allow the client’srequirements to be adequatelydefined so that it may obtain fixedand firm prices for the work in acompetition.

Page 12: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

I gave an undertaking in June 2003 that I would examine and report again on the key features of project management...

10

Part 1. Introduction

Page 13: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

11

1.1 This report concerns themanagement of the project toprovide the new Scottish Parliamentbuilding at Holyrood. This is my thirdreport on aspects of the Holyroodproject.

1.2 My report in September 2000(Exhibit 1 overleaf) explained why thethen estimated costs of the Holyroodproject had increased to just over£200 million and why the thenprojected completion date hadslipped to December 2002. Itquestioned aspects of projectmanagement including theprocurement strategy, the forms ofcontract adopted, cost control andrisk management.

1.3 In December 2002 I reported onvarious contract management issuesfollowing the termination of the FlourCity contract that had come to lightduring my annual audit of theCorporate Body.

1.4 I gave an undertaking in June2003 that I would examine andreport again on the key features ofproject management, the escalatingconstruction costs and the delays.My further report is intended tocomplement the report of the inquiryby Lord Fraser of Carmyllie, which isdue for publication later in 2004.

1.5 The Audit Scotland team wasassisted by Gardiner & Theobald, oneof the UK’s largest providers ofindependent professional advice andservices to the property andconstruction sectors.

1.6 Completion of the Holyroodbuilding project is a moving target(Exhibit 2 overleaf). Even at thisadvanced stage of constructionfurther cost increases and slippagecannot be ruled out. Because ofuncertainty about the final costs andcompletion of the new Parliamentbuilding it is likely that I will reportagain on the management of theproject and its consequences at afuture date.

Page 14: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

12

Exhibit 1The Auditor General’s 2000 report on the Holyrood Parliament building project

Source: The new Scottish Parliament building: An examination of the management of the Holyrood project, September 2000

The report described the delays and cost increases that occurred between the start of the project and thepublication of the report in September 2000. The target total cost had risen from £90 million at the outset to £195million, as agreed by the Parliament in April 2000. Within the £195 million target, estimated construction costs hadincreased from £50 million to £108 million and other costs rose from £40 million to £87 million. In addition, it wasnoted that £14 million would be required for landscaping and road works.

Almost half the construction cost increases was due to a 47% increase in the area of the building. The rest of theincrease was due to other factors such as the complexity of the approved design, improvements in the quality ofthe building, security needs, and the costs of Queensberry House. Consultants’ fees could reach £26 million andthe report suggested ways that fees might be better controlled at that time (September 2000).

The Auditor General said that most of the spending had yet to be incurred and there were risks and uncertaintiesthat needed careful management if the building was to be completed on time and within budget.

The Auditor General criticised the arrangements for project management’s cost reporting to the client. It was notedthat before June 1999, reports included only core construction costs and did not contain systematic and consistentreferences to the financial provisions which were necessary for contingency, fees, furniture and fittings, and VAT. Inaddition, for most of the project’s life, an important shortcoming had been that the reporting of construction costsdid not include a separate allowance for risk.

The report pointed out that there should have been an analysis of how to use incentive structures to promoteadded value in the design and construction processes. Financial bonuses for early completion and penalties fordelays were common practice in major construction projects where deadlines were tight.

The report said that the handover of the project from the Scottish Executive to the Corporate Body of theParliament in June 1999 should have been better managed. The Auditor General said it was unfortunate that anindependent review had not taken place, since at that point the design was not firmly fixed and the costconsultants were estimating construction costs significantly above the budget.

Reference was also made to project management – the officials responsible for managing and delivering theproject. While project management had strengths, the report questioned whether they had at all times the bestmix of skills needed for this demanding project.

The report said that slippage of the completion date (from July 2001 to December 2002) was due to a number offactors, including the architects’ difficulties in delivering the original brief to a very tight deadline and unforeseenchanges requested by the Corporate Body to meet the emerging needs of the new Parliament in its first year ofoperation.

On a positive note, the report said that arrangements for oversight and management of the project, including theestablishment of a Progress Group in June 2000, had been strengthened. However it said much remained to bedone.

Key recommendations included the need to:

• establish an action plan to manage risk

• agree a cost plan

• set milestones to help monitor progress until the project is completed.

Page 15: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 1. Introduction 13

Exhibit 2Changes in the forecast cost and completion of the Holyrood project

Originally

January 1998AGS’s first

report on the

Holyrood project

September 2000

First Minister

announces

inquiry

June 2003

Latest position

May 2004

Estimated project cost £90 million £209 million £375 million £431 million

Forecast completion July 2001 December 2002 November 2003 August 2004

Source: Audit Scotland

Page 16: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

one of the most challengingassignments tackled both inengineering design andconstruction terms.

• Themed irregular window andconstruction forms, including theindividual ‘thinking module’ baywindows for the MSP Block andthe large granite ‘trigger panels’on the east side buildings.

• The foyer roof occupying theformer Queensberry Housegarden, with leaf themed roof-lights.

• The distinctive curved ceilings ofthe committee rooms and thevaulted concrete ceilings of thepublic area below the debatingchamber.

• Materials and finishes, employingoak, granite and stainless steel,must be durable and of a qualityto reflect the building’s status andthe required lifespan of at least100 years for its fabric.

2.3 Satisfying the architects’ designconcept for these and other featureshas set many challenges for the

specialist contractors, projectmanagement and the design teamthemselves.

2.4 Exhibit 3 and the images at thestart of each part of this reportillustrate the main components of thenew Parliament building and theassociated landscaping.

The Parliament has been procured

using construction management

2.5 In July 1998 project managementchose construction management todeliver the new Parliament. It did sojudging that traditional routes ofdesign and then build could not meetthe July 2001 target completion dateset by the client at that time, theSecretary of State.

2.6 Construction management offersan opportunity to compress projectduration by commencingconstruction work before all detaileddesign is settled. To permit this tohappen there is no main contractorfor the fixed price delivery of asettled design. Instead the clientcontracts directly with individual tradecontractors appointed at differenttimes, and appoints a professional

2.1 This part of the reportsummarises the physical features ofthe new building and the positionnow achieved with regard to itscompletion. It analyses the mainevents in the progress of theHolyrood building project since 2000,when we last examined themanagement of the building project.

The new Parliament is a striking

and very complex building with

high quality materials and some

unusual design and construction

features

2.2 The different components of thebuilding range from the substantiallyreconstructed Queensberry House,dating from the 1650’s, to theupturned boat-like towers and curvedplan assembly buildings (on the eastside) and to the more rectangular butstill highly distinctive MSP officeblock (on the west side). Thebuildings exhibit distinctive designfeatures, including:

• The exposed ‘glulam’ (laminatedoak) roof beams in the debatingchamber, connected by 110uniquely specified stainless steelnodes. The chamber roof was

14

Part 2. Project overview

Page 17: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 2. Project overview 15

Exhibit 3The new Parliament building

Source: Audit Scotland

The site plan for the Holyrood parliament complex isshown below. The site is flanked on the west by theMSP block (lower left image). To the northwestQueensberry House (upper left image) will house thepresiding officer and parliament administration functions.It overlooks the foyer, which provides circulation spacefrom the MSP offices to the main complex.The east of the site (lower right image) comprises thecomplex of buildings where the business of theparliament will be conducted, in the committee roomsand the debating chamber.

MSP Building (looking to Salisbury Crags) Main building complex (looking from south)

Queensberry House and foyer roof

Tower 4

Tower 3

DebatingChamber

PressTower

CanongateTower

QueensberryHouse

PublicEntrance

58-60Canongate

Restaurant/Dining area

MSPBuilding

Tower 1

Tower 2

Foyer

Page 18: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

16

construction management firm tocoordinate the process. Becausethere is no main contractor there isno opportunity for the client totransfer responsibility for managingmost of the design and constructionrisk, including that from co-ordinationof design and trade contract work.The client retains responsibility forcontrol of the design process andavoids paying for risks that do notmaterialise (Exhibit 4). Whateverconstruction method is adopted theclient always pays for risk.

2.7 We analyse the complexinteraction between the main partiesresponsible for completing the newParliament building (Exhibit 41) in Part5. In summary the main parties are:

• The client organisation,comprising the ScottishParliamentary Corporate Body (asthe legal client); projectmanagement including theHolyrood project team; and theHolyrood Progress Group, whohave a role to work with theCorporate Body in the realisationof the project.

• The consultants for the project,comprising principally: thearchitects, who have leadresponsibility for design of thenew building; other members ofthe design team; the costconsultant; and the constructionmanager.

• The individual works packagecontractors. Under constructionmanagement (Exhibit 4) there isno single main contract coveringall the construction works.Instead there are some 60individual construction workspackage contracts, whichtogether comprise all thenecessary works. Although theconstruction manager overseesand coordinates each individual

trade contract, the client remainsthe employer for all contracts.

The £431 million current estimated

project cost includes a large

margin for risk and contingency

2.8 The current forecast of theestimated cost of the Holyroodproject is £431 million. As is wellknown, the project has sufferedsuccessive increases in forecastcosts since inception. Exhibit 5illustrates the increases in totalproject costs affecting the projectsince my report in 2000.

2.9 There is inherent uncertaintyabout the estimated outturn costbecause in most cases the costconsultant, the construction managerand project management has yet toreceive, negotiate and agree finalaccounts from the main tradecontractors. At May 2004 the projectmanagement had approved finalaccounts for 21 of some 60 mainconstruction contracts. The contractsums agreed so far total £26 million,10% of the forecast final constructioncost including the risk estimate.

2.10 In Part 4 we show thatadditional payments to contractorsfor extensions of time and out ofsequence working is a main reasonfor the increased costs of the projectsince 2000. The main risk to thecurrent forecast is that if work is notcompleted by August 2004 additionaltime related costs would be incurred.

2.11 However, while the precisefinancial commitment is thereforeuncertain, the current estimate ofcost (Exhibit 6 overleaf) includes amaximum £28 million as a reserveagainst construction risks (based onprofessional advice from the costconsultants). The constructionreserve is 10% of the currentconstruction cost commitmentincluding VAT. Project management

has included a further £10 millionprogramme contingency. This isavailable to cover any other variancein costs, while the main non-construction item (£50 million fees) isalready capped. Overall, the total riskreserve is £38 million, which is asignificant margin particularly at thisadvanced stage of construction.

The project has suffered

successive slippage in its forecast

completion date

2.12 In parallel with the costincreases since 2000 there havebeen five successive revisions of theprogramme for the completion of theHolyrood project. In total, forecastcompletion of construction hasslipped by 20 months fromDecember 2002 to August 2004.Exhibit 7 (overleaf) charts eachprogramme issue betweenprogramme 3 in July 2000 and thecurrent programme 7B issued inFebruary 2004.

Construction is now well

advanced

2.13 At this stage we cannot saywhether the whole building will besatisfactorily completed in summer2004. Currently (June 2004)construction is well advanced. Allparties are working towards thetarget to complete construction of allcritical building elements so thatusers – MSPs and the CorporateBody’s staff – may start to occupythe building later in the Summer2004. The programme for completionis tight and so some parts of thebuilding – including the debatingchamber and the main towers –construction will not be completedand areas declared safe foroccupation until July or August 20046.Project management has approvedthe retrofit of some building features,in either October or DecemberParliamentary recesses.

6 An occupation certificate or certificate of completion is needed before any new building can be occupied. It provides formal confirmation that an approved inspectoraccepts that the building work has been carried out in accordance with the warrant obtained prior to work starting and with the building standards regulations, in sofar as this can be ascertained from inspection.

Page 19: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

17Part 2. Project overview

• Used for completing fast track projects where there is a high degree of design uncertainty.

• There is no main contractor for the fixed price delivery of a settled design.

• Design and construction work overlap leading to an earlier start on site. Later changes in design may beaccommodated without necessarily incurring a high premium on construction costs, provided that there is nochange or disruption to work that has already commenced.

• The client still bears most of the design and construction risk, including coordination.

• The client enters separate agreements with the designer and the construction manager.

• The client appoints all construction contractors and the construction manager receives a fee for managing this work.

• The construction manager is responsible for the management and coordination of design and constructionworks, encouraging teamwork and ensuring ‘buildability’ is considered at an early stage.

• In June 1999 the Treasury recommended that this route should only be used where there is a very clear valuefor money case for doing so.

Exhibit 4Key features of construction management procurement

Exhibit 5Holyrood project costs reported to the Finance Committee 2000 to 2004

Source: Audit Scotland

0

50

100

150

200

250

300

350

400

450

Mill

ion

(£)

209

248 255274 281

309325

338

375

431

401369

Mar

ch 2

000

Apr

il 200

0M

ay 2

000

June

200

0Ju

ly 2

000

Aug

ust 2

000

Sept

embe

r 200

0O

ctob

er 2

000

Nov

embe

r 200

0D

ecem

ber 2

000

Janu

ary

2001

Febr

uary

200

1M

arch

200

1A

pril 2

001

May

200

1Ju

ne 2

001

July

200

1A

ugus

t 200

1Se

ptem

ber 2

001

Oct

ober

200

1N

ovem

ber 2

001

Dec

embe

r 200

1Ja

nuar

y 20

02Fe

brua

ry 2

002

Mar

ch 2

002

Apr

il 200

2M

ay 2

002

June

200

2Ju

ly 2

002

Aug

ust 2

002

Sept

embe

r 200

2O

ctob

er 2

002

Nov

embe

r 200

2D

ecem

ber 2

002

Janu

ary

2003

Febr

uary

200

3M

arch

200

3A

pril 2

003

May

200

3Ju

ne 2

003

July

200

3A

ugus

t 200

3Se

ptem

ber 2

003

Oct

ober

200

3N

ovem

ber 2

003

Dec

embe

r 200

3Ja

nuar

y 20

04Fe

brua

ry 2

004

Mar

ch 2

004

Apr

il 200

4M

ay 2

004

Note: equivalent to£195m excludinglandscaping costs

Total risk adjusted project costs,including landscaping costs

Source: Official Report

Page 20: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

18

In summer 2000, compared to

earlier difficulties, there seemed a

firm basis for the Holyrood project

to move forward

2.16 In September 2000 I reportedthe significant challenges that hadpreviously affected the project:

• Concerns in 1998 and 1999 aboutthe overall size and consequentlythe cost of the initial designs.

• Uncertainty about cost estimatesthroughout 1998, 1999 and early2000, with predicted costs whichconsistently exceeded theapproved budget.

• Major changes in therequirements for the area andlayout of parts of the building,with the client unable to freeze itsrequirements and consequentlythe design until June 2000.

2.17 At the time of my September2000 report on the management ofthe project, construction was at acomparatively early stage.Construction of the MSP buildingframe had commenced but the main

assembly building structure to theeast of the site was not ‘out of theground’.

2.18 In September 2000, however,while much remained to be done toensure completion, there seemed afirmer basis for the project to moveforward:

• In June 2000 the Corporate Bodyas client had approved the finalscheme design proposals fromthe design team (known as the‘Stage D’ report).

• The Corporate Body had selectedthe members of the newHolyrood Progress Group, with aremit to work with the progressof the project, and the Group hadstarted its business.

• The Corporate Body approved theStage D design on the basis ofconfirmation from all the mainparties that the building could becompleted by the end of 2002within a target construction costof £108 million (£119 millionincluding constructioncontingency). This figure was

2.14 While it is likely that thebuildings will be in a state capable ofbeing partially occupied the risk isthat by the current target forcompletion of August 2004significant areas may be incompletein other ways. Consequently theconstruction manager and the tradecontractors may have to spendseveral months or even longer in thebuilding continuing after handover tothe client to complete unfinishedworks and snagging.

2.15 The Corporate Body isimplementing a migration strategy forusers based on the assumption thatthere is enough time to completeconstruction to allow occupation anduse of the building in the latesummer 2004. The Chief Executiveestablished a dedicatedimplementation team of staff topromote effective planning anddelivery of the move to the newbuilding. There is close coordinationbetween the Holyrood project teamand the consultants with the newimplementation team.

Exhibit 6Holyrood project – analysis of current forecast costs

Source: Report to Finance Committee May 2004

£241m

£28m

£20m£10m

£18m £6m

£42m

£68m

Item Cost

Construction £241m

Fees and site organisation £68m

Irrecoverable VAT on construction £42m

Reserve (including VAT) £28m

Fit out £20m

Programe contingency £10m

Landscaping £18m

Site, demolition and archaeology £6m

Grand total £431 million

Page 21: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 2. Project overview 19

Exhibit 7Holyrood – main programme revisions 2000 to 2004

Note: Programme 3C did not include external landscaping

Source: Audit Scotland

MSP

bui

ldin

g

Que

ensb

erry

Hou

se

Ass

embl

y bu

ildin

gs in

clud

ing

deba

ting

cham

ber

Exte

rnal

land

scap

ing

MSP

bui

ldin

g

Que

ensb

erry

Hou

se

Ass

embl

y bu

ildin

gs in

clud

ing

deba

ting

cham

ber

Exte

rnal

land

scap

ing

MSP

bui

ldin

g

Que

ensb

erry

Hou

se

Ass

embl

y bu

ildin

gs in

clud

ing

deba

ting

cham

ber

Exte

rnal

land

scap

ing

MSP

bui

ldin

g

Que

ensb

erry

Hou

se

Ass

embl

y bu

ildin

gs in

clud

ing

deba

ting

cham

ber

Exte

rnal

land

scap

ing

MSP

bui

ldin

g

Que

ensb

erry

Hou

se

Ass

embl

y bu

ildin

gs in

clud

ing

deba

ting

cham

ber

Exte

rnal

land

scap

ing

MSP

bui

ldin

g

Que

ensb

erry

Hou

se

Ass

embl

y bu

ildin

gs in

clud

ing

deba

ting

cham

ber

Exte

rnal

land

scap

ing

Programme 3C

– July 2000

Programme 4B

– January 2001

Programme 5B

– January 2002

Programme 6B

– October 2002

Programme 7A

– August 2003

Programme 7B

– February 2004

Original start tofinish for

construction ofmain assembly

buildings

Current start tofinish for

construction ofmain assembly

buildings

September 1999October 1999

November 1999December 1999

January 2000February 2000

March 2000April 2000May 2000June 2000July 2000

August 2000September 2000

October 2000November 2000December 2000

January 2001February 2001

March 2001April 2001May 2001June 2001July 2001

August 2001September 2001

October 2001November 2001December 2001

January 2002February 2002

March 2002April 2002May 2002June 2002July 2002

August 2002September 2002

October 2002November 2002December 2002

January 2003February 2003

March 2003April 2003May 2003June 2003July 2003

August 2003September 2003

October 2003November 2003December 2003

January 2004February 2004

March 2004April 2004May 2004June 2004July 2004

August 2004September 2004

October 2004

Page 22: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

20

act as the single focal point for day-to-day management of theParliament’s interest in the project,with responsibility for securing thedelivery of the project to programme,within budget and to the specifiedquality.

2.21 Despite these promising signsthe project was to encountersignificant difficulties over thefollowing three years, as analysedbelow.

There was at best only qualified

agreement of the necessary cost

plan in November 2000

2.22 From inception in June 2000 theProgress Group generally met atleast every fortnight to fulfil its remiton behalf of the Corporate Body andthe Parliament. In July, August andSeptember 2000 the Groupdiscussed, amongst other matters,the overall forecast cost of theproject and ways of controlling this.In July 2000 the Group examined and

discounted as expensive andimpracticable the option ofabandoning the constructionmanagement contract and enteringsome form of guaranteed maximumprice contract of a means oftransferring risks to a contractor. TheGroup concluded the existingcontractual arrangements weresatisfactory.

2.23 In August and September 2000early tender results for some workspackages were available which weremore costly than had been forecast.The Group decided that the designteam, the construction manager andthe cost consultants must agree acost plan urgently. The cost plan wasneeded to translate the overallconstruction cost approved at StageD in June 2000 into more detailed,individual cost allowances, to alloweach component trade contractpackage to be controlled. Preparationof the cost plan had commenced inMarch 2000 but the parties had notreached the necessary agreement.

consistent with the Parliament’soverall budget of £195 million,which the Parliament hadestablished in its April 2000resolution7.

2.19 In my September 2000 report Imade several recommendations tohelp ensure that the ScottishParliament building would bedelivered on time and within budget.These included threerecommendations on themanagement of risk and the controlof overall project costs. In evidenceto the Audit Committee in October2000 the Corporate Body’sAccountable Officer indicatedacceptance of theserecommendations, which he statedwere being implemented (Exhibit 8).

2.20 In November 2000, shortly afterthe AGS 2000 report, Mr Alan Ezzitook up his appointment as theHolyrood project director, followingan open recruitment competition8.The purpose of this key position is to

Source: AGS report September 2000 paragraph 27; evidence to the Audit Committee October 2000

Exhibit 8AGS 2000 recommendations on risk and cost control

AGS recommendation, September 2000 Accountable Officer

response, October 2000

Project management should instruct the construction manager and the costconsultant to prepare as soon as practicable a risk analysis, which should identifyall remaining risks to the project and their potential impact on costs and deadlines.This analysis should take account of the main risks identified in Part 1 of thisreport and quantify the most likely outcomes as well as best and worst cases.The results should be the basis for an action plan to manage the remaining risks.

Project management should look again at the overall cost provision in the light ofthe risk analysis. They should ensure that, in accordance with good practice, thereis a proper, separate allowance for risk in the estimate.

Project management, the design team and the construction manager must agreea cost plan taking account of risks and uncertainty, to provide an effective basisfor managing the remaining stages of the project.

Accepted, currently in progress

Accepted, currently in progress

Done, as part of routinemanagement

7 Official Report 5 April 2000: That the Parliament notes the report of the Scottish Parliamentary Corporate Body on the Holyrood project; approves its terms, anddirects the Corporate Body to establish a progress group comprising representatives of the Parliament and relevant professionals to work with the Corporate Body to(a) finalise the design; (b) complete the project by the end of 2002 within a total budget of £195 million and (c) report regularly, or as from time to time may berequired, on progress including on expenditure to date and estimated completion costs to the SCPB and to members.

8 The appointment followed the departure of the previous project sponsor, Mrs Doig, in May 2000. Dr John Gibbons, the chief architectural advisor and a member ofthe Progress Group, had filled the project director post temporarily from May to November 2000.

Page 23: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 2. Project overview 21

2.24 Discussion of the necessarycost plan by all parties in Autumn2000 highlighted uncertainty and costdifficulties. In particular, for manypackages there was insufficientlydetailed design information to allowthe necessary costings to be donewith confidence (Exhibit 9 overleaf).

2.25 The cost consultant and theconstruction manager’s independentmeasurement of the gross area ofthe design also revealed an additional2,200m2 and 2,400m2 respectively(7% and 8%) compared to 30,600m2

which the architect had reported (andthe client had approved) for theStage D design (Exhibit 10 overleaf).

2.26 Uncertainty about detaileddesign information and thereforecosts is unavoidable at designapproval (Stage D) but the scale ofuncertainty in this case appears tohave been comparatively high. Todeal with the uncertainty theconstruction manager recommendedthat the cost plan should be adoptedonly on the basis that allowanceswithin it were strictly regarded astarget rates or budget ceilings, withinwhich the design team should berequired to work.

2.27 Despite the uncertainty initialcontract tender values subsequentlycame in close to – on average some5% above – the cost plan estimateafter allowing for inflation. Howeverthe eventual final contract valueswere substantially – on average 92%– above the cost plan estimate.

In November 2000 estimated risk

was significantly more than the

available budget

2.28 In October 2000, in accordancewith my recommendation (Exhibit 8),the project team instructed the costconsultant to organise a riskworkshop with the participation of allmain parties. Similar workshops hadpreviously taken place in 1998 and

1999, which identified risks but didnot quantify their potential cost. Theworkshop in October 2000 for thefirst time placed monetary valuesagainst each identified risk. The costconsultant reported these values inNovember 2000, producing a totalestimated risk cost of some £61 million (£51 million excludingVAT).

2.29 The purpose of the riskworkshop was to provide amanagement tool to monitor andhelp control project risks. The £61million was the combined impact ofthe identified risks, if they were notcontrolled or managed effectively. Itwas significantly higher than the £11 million contingency provisionassociated with the £108 millionconstruction cost within the £195million target approved by Parliament.

2.30 Although by definition risk isuncertain, some 70% of the riskidentified by the October 2000workshop was for items that werecategorised as ‘highly likely’ ie,assessed by the workshop membersas having at least a 95% probabilityof occurring. Each of these itemscarried with them varying levels oflikely impact on programme.Moreover the risk evaluation did notattempt to evaluate the monetaryvalue attached to the risks to thetime schedule. If it had it would haveadded further risk cost.

2.31 Despite these results projectmanagement and the HolyroodProgress Group seemed to be of theview that the project could becompleted within the availablebudget. In November 2000 theProgress Group reported to theCorporate Body that the riskassessment had been completedand that the design team had beeninstructed to work to manage thoserisks they could control within theavailable budget.

2.32 One obvious risk affecting thecost plan was the impact of inflation.Combining the £108 millionconstruction cost target with the £11 million contingency gave a totaltarget provision for construction of£119 million. At the time of mySeptember 2000 report projectmanagement considered that £119million was a cash limit within whichthe contract costs must bemaintained. However, my September2000 report noted the £108 millionderived from cost estimates basedon constant March 1998 prices. Ittherefore excluded inflation over thesubsequent period to completion inDecember 2002, ie, almost fouryears. My September 2000 reporthad estimated inflation could addconservatively 10% (£11 million) onto estimated costs. In November2000, as part of the risk workshopnoted above, the cost consultantestimated inflation could add £13 million.

2.33 Although £11 millioncontingency was available within thebudget to meet construction costinflation it also was required to meetall other construction risks, whichwere significant.

2.34 These cost difficulties were notdecisively resolved between theparties. In November 2000 projectmanagement reported to theProgress Group that all consultantshad agreed the £108 million packagecost plan subject to the availability of£11 million contingency. Despite thisagreement the £119 millionconstruction costs was not used asthe basis for calculating payments tothe consultants, except in one case,which indicates there was no strongagreement between the parties onconstruction costs:

• As an incentive to contain costsand deliver within budget all ofthe consultants’ fee remuneration

Page 24: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

22

Exhibit 9Uncertainty in the November 2000 cost plan for the Holyrood building

Source: Audit Scotland

Source: Audit Scotland

Exhibit 10Estimated area of the Holyrood building

Source Estimated gross internal area

Architect’s Stage D report June 2000 30,600 m2

Calculated by the cost consultant from Stage D drawings, August 2000 32,800 m2

Calculated by the construction manager from Stage D drawings, September 2000 33,000 m2

Uncertainty about building details and therefore costs is unavoidable at design approval stage for any building.However the scale of uncertainty affecting the Holyrood project appears to have been high.

An unfortunate fact is that the Enric Miralles, who had lead responsibility for the architectural design, had died inJuly 2000, shortly after the client approved his proposals.

In Autumn 2000 the design of the new Parliament building was comparatively undeveloped. Although Stage Ddesigns had been approved the necessary more detailed information in many areas was not available.

For example, for parts of the site the only drawings indicating construction proposals were at 1:250 and 1:200scale while procurement of the work generally required drawings at 1:20 or 1:10 scale to allow the necessaryconstruction details to be specified clearly.

The degree of detailed information available at this stage varies from project to project but it is not unusual to haveplans and drawings at substantially larger scales than 1:200.

The cost plan estimates agreed in November 2000 for Holyrood consequently reflected a high degree ofuncertainty. The cost plan included a large number of lump sum allowances, where there was insufficient detail toestablish either a quantity or a cost rate.

Where quantities could be established they were often broad brush, particularly with regard to buildings to the eastof the site where the design was least fully developed.

An example of this is Assembly roof building – roof structures and finishes. From the design this roof had ameasured area of 3,816m2, and the cost plan included some £3.816 million for the whole item ie, the estimatedroof area costed at £1000/m2. This item was assumed to include components such as the support structure,upstands, waterproofing, thermal circulation, cladding, flashings, walkways, eaves details and fascias. Becausethere was no specification of these individual elements there was no indication of the allocation of the £1000/m2

between them.

Page 25: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

23Part 2. Project overview

was linked to varying degrees to acalculation of the approvedconstruction cost rather than theeventual outturn cost9.

• In August 2000 projectmanagement did agree aprofessional fee cap with theirstructural engineer consultants,based on £108 millionconstruction cost. The fees to thestructural engineers (including thenegotiated costs of later,additional services) total £4million, 8% of the total feesexpenditure of £50 million.

• However project managementdid not agree a professional feecap based on construction costswith any other consultants until2003, when estimatedconstruction costs weresignificantly higher.

By June 2001 the Corporate Body

had concluded the £195 million

target was not achievable

2.35 In January 2001 projectmanagement advised the ProgressGroup on the tender results for majorparts of the new building includingthe assembly building concreteframe and the MSP building claddingand roof. Tenders were over the costplan budget. Taking account of thisand other cost pressures, the Grouprequired the design team and theproject director to complete a sitewide savings exercise. The objectivewas to reduce construction costswhilst maintaining overall quality andfunctionality.

2.36 In February and March 2001various savings options wereidentified and evaluated. Changeswith an estimated value (potentialsavings) of some £3 million wereinstructed but the client rejected

others valued at £2 million ascompromising the integrity and/or thequality of the design.

2.37 The process of seeking andevaluating construction cost savingsshowed substantial tensionsbetween the Progress Group, theproject director, the design team andthe cost consultant, particularly withregard to the budget for constructioncosts.

• In April 2001 the project directoradvised the Progress Group thatin his opinion the range of outturncosts for the project taking intoaccount inflation and risks wouldbe £211 million to £249 million ie,between 8% and 28% higherthan the £195 million project costtarget.

• Also in April 2001 the ProgressGroup informed the CorporateBody’s Accountable Officer of itsmounting concerns about theeffectiveness both of the designteam and of project managementin working together to deliver onbudget and programme andrespecting the client’srequirements.

2.38 In May 2001, with the advice ofthe Progress Group, the AccountableOfficer started regular meetings withthe principal members of hisconsultant advisors to help managethe project. The Progress Group wasconcerned about poor coordination,evidence of misunderstanding ofresponsibilities with regard to costcontrol and whether there wassufficient resource devoted to theproject. The Group also questionedthe effectiveness of the relationshipbetween the project director, thearchitect and the cost consultant. Inthe Group’s opinion there was aneed for the team to be welded back

together and for project managementto act as a catalyst for improvedcoordination and joint working.

2.39 The Accountable Officerconsidered that the ‘principals’meetings’ with his consultantsimproved the atmosphere. Howeverhe and the Progress Group wereconcerned about continued apparentweaknesses in coordination of thedesign work and in their view severeshortcomings in cost information,particularly with regard to designproposals received in May for thefoyer roof10. The Group questionedthe effectiveness of the projectdirector and in June 2001 advised theAccountable Officer that it had lostconfidence in him. The AccountableOfficer concluded the situation wasirretrievable and agreed with theproject director that he would leavethe project before the end of hiscontract.

2.40 The Accountable Officerimmediately appointed a new projectdirector in June 2001. There was nocompetition for this post, becausethe Accountable Officer judged that arecruitment exercise would result inthis critical post being vacant at acritical time, and a candidateconsidered suitable, Miss SarahDavidson, was available immediately.The new project director wasappointed, like her predecessor, asthe single focal point for day-to-daymanagement of the project. TheAccountable Officer changed theprevious requirement that the postholder should have substantialexperience in managing majorprojects. Instead roles andresponsibilities within projectmanagement were adjusted toprovide the project director withadditional professional and technicalsupport. The client adviser’s role wasextended to act as the main contact

9 The precise mechanism varied in accordance with individual contracts. The entire construction manager fee was a percentage of project cost, which was convertibleto a fixed lump sum on agreement of the cost plan. The fee for the design team members was a percentage of contract cost, with around 60% of the total feebased on the approved cost at intermediate stages before construction commenced.

10 The cost consultant’s view is that project management did not relay critical Progress Group opinion and it took it as read that project management understood thecost issues it had reported.

Page 26: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

24

between the project director and thedesign team on technical matters.

2.41 Regarding progress on site in2000 and early 2001, the design andconstruction method was that workwas to commence on the west sideof the site (the MSP building). It wasthen to proceed to QueensberryHouse and finally to the assemblybuildings (the debating chamber andcommittee towers) to the east of thesite. Construction had started in 1999and continued to progress during late2000 and 2001.

• In December 2000 work ondemolition and reconstruction ofQueensberry House commenced.

• In January 2001 the MSP buildingwas ‘topped out’, indicating theframe of the building had reachedits fullest height. In the samemonth winning tenderers for boththe MSP cladding package andthe MSP roof package wereselected.

• Also in January 2001 the majorcontract for the construction ofthe assembly building frame wasapproved, and work on thesubstructure for the eastern partof the site was well advanced.

• In February 2001 the ProgressGroup recommended approval ofthe Stage D design of thelandscape scheme associatedwith the new building11.

2.42 In March and May 2001 thewhole project team completed threemore risk workshops, as in October2000 providing estimated risk costs.The estimated additional design andconstruction risk including fees andforecast inflation costs was similareach time, in the range £46 to £49million including VAT.

2.43 In May 2001 the ProgressGroup reported their conclusions on

risk to the Corporate Body. Withoutspecifying any alternative singlefigure for the total cost of the projectthe Group demonstrated that riskcould significantly increase costscompared to the existing budget of£195 million. Even setting aside theoverall impact of inflation and riskallowances above the contingency, itstated the project was likely toexceed the budget by up to £24 million; the total excessdepended on assumptions, whichthe Group considered wereimprecise and imponderable.

2.44 The Group did not suggest theCorporate Body seek a new cappedfigure from the Parliament. It advisedinstead that Parliament be invited tonote the current position and therelevant variables. The CorporateBody was invited to seek authorityfrom Parliament that the projectshould proceed on the basis ofregular reports to the FinanceCommittee on budget, to inform theCommittee’s consideration of theannual budget process.

2.45 The Group’s main reason for notproposing a new capped total projectcost was that it would be misleadingto suggest a high degree of certaintyabout cost was possible. The Groupconsidered a new cap figure “wouldshow our hand both to potentialtenderers and to the design teamjust at the very time we areendeavouring to keep the latterbuckled down to working withinbudget.”

In June 2001 a fresh Parliamentary

motion relaxed the earlier budget

and completion targets

2.46 In June 2001, based on theCorporate Body’s report andrecommendation, Parliamentapproved a motion, which can beinterpreted as removing the previousoverall budget constraint of £195 million (Exhibit 11).

The procurement of works

changed radically because of

programme pressures in 2000

and 2001

2.47 Much of the Progress Group’sand the Corporate Body’s main focusduring 2000 and the first half of 2001was on design and cost issues.However important programmeissues also arose, including:

• Slippage in the production by thedesign team of necessary designinformation to allow procurementof individual works packages toproceed according to plan. Thiswas a particular concern affectingthe east side of the site, wherethe most complicated parts of thebuilding were to be constructed.

• Delays and re-sequencing of theplanned programme forQueensberry House arising fromunforeseen structural andconservation issues affecting itsdesign. Issues included whether itwould be necessary toreconstruct the Belvedere Tower,and questions about the height ofa critical building element, thewallhead.

2.48 In December 2000 theconstruction manager’s strategicprogramme review (based onprogramme series 3 issued in July2000) highlighted the likely impact ofdelays and slippage across theproject. Consequently theconstruction manager prepared andproject management approved arevised programme 4, to beconsistent with the previouslyapproved December 2002completion date.

2.49 In accordance with the client’sdesire successive versions ofprogramme 4 continued to targetcompletion of the main buildings(including the debating chamber) byDecember 2002. This was within a

11 At that time the necessary land was in the ownership of Historic Scotland. Subsequently the Corporate Body and Ministers agreed that the title for this land shouldtransfer to the Parliament. This brought responsibility for ownership of the land, approval of the design of the landscape works and control of costs within one body.

Page 27: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 2. Project overview

engineering in the debating chamberroof. Project management instructedthe construction manager toinvestigate the feasibility of a revisedprogramme, later issued asprogramme 5, which would mitigatethe delay to a more acceptable April2003 completion target for thedebating chamber. Programme 5Btargeting completion of constructionby April 2003 was duly approved andissued in January 2002 (Exhibit 12overleaf).

2.52 The pressure to maintainprogramme resulted in significantchanges to the procurement of workscontracts. The original procurementplan was conventional. It was totender and award each constructionworks package in a single stage,based on tender/design informationproduced by the design team.

2.53 In 2000 and early 2001 projectmanagement approved a range ofmitigation measures recommendedby the construction manager, in anattempt to maintain the plannedcompletion date. The mitigationmeasures changed radically howpackage contracts were to be

designed, procured and managed.The particular methods chosen variedat different times, but over the wholelife of the project have included:

• Using large ‘provisional sum’12

allowances to enable packages tobe tendered and awarded earlierthan would otherwise be possiblebecause of incomplete designinformation.

• Extending the responsibility oftrade contractors for design, tovarying degrees, thereby reducingthe overall workload for thedesign team. Some contractorswere required to finalise detaildesign only, within a conceptdesign from the design team. Inother cases contractors had alsoto finalise the development of theconcept design.

• Associated with trade contractordesign, using ‘two stage’ tenders for a few selectedpackages. At the first stagetenderers were invited to pricework where design informationwas sufficient and appointed onthat basis. At the second stage

compressed programme, which theconstruction manager emphasisedwas conditional upon all partiesachieving specified actions byspecified dates.

2.50 It was accepted that thelandscaping work, which was notcritical to occupation, could extend toApril 2003. After construction of themain buildings was complete therewould need to be more time forfitting them out. This work too wasproposed for the period betweenDecember 2002 and April 2003. Thiswould permit occupation of the newParliament around the time of theelections due in May 2003.

2.51 In August and September 2001the construction manager reported toproject management that theprogramme 4 target of completion byDecember 2002 could not beachieved. This took into accountcontinuing slippage in the supply ofsome design information, difficultiesarising from the underperformance ofone contractor responsible forcladding on the MSP block andparticular design issues on somepackages, for example the complex

25

Source: Official Report

Exhibit 11The Parliament’s Holyrood resolution 21 June 2001

That the Parliament notes the terms of the Scottish Parliamentary Corporate Body’s (SPCB) report of 13 June2001; notes that £60 million of the construction costs for the new Parliament building has been committed to dateand that a further £57 million (at 1998 prices) remains to be let; further notes that building industry inflation iscurrently estimated to be adding at least 16% to the costs of packages remaining to be let and that under theconstruction management contract there are additional and not fully quantifiable risks to which the project may beexposed between now and completion; directs the SPCB, through the Holyrood Progress Group, to work with thedesign and project teams to complete the project without compromising quality, while managing risks rigorously,and requires the SPCB, on a quarterly basis, to provide information to the Parliament’s Finance Committee on theprogress of the project in respect of inflation and materialisation of risk in order to inform the committee’sconsideration of the annual Budget Bill.

12 A provisional sum is a sum provided for work or for costs that cannot be entirely foreseen or detailed when the tender documents are issued to contractors.

Page 28: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

26

Exhibit 12Slippage in programmes 4 and 5 in 2001 and 2002

Source: Audit Scotland

Programme 4 – December 2000 to March 2001

Between December 2000 and March 2001 the construction manager prepared a revised programme 4 insuccessive versions 4A, 4B and 4C.

Programme 4 incorporated mitigation measures and all parties were required to commit to achieving specifiedactions by specified dates. On this basis the construction manager forecast that completion of the assemblybuilding could be achieved on target in December 2002.

However, while main construction was targeted to finish by this date, fit-out, landscaping and commissioningworks would need to extend to April 2003 before the building could be entirely handed over to the client.

Also construction completion by December 2002 was conditional upon key design information for three claddingand glazing packages being released on 12 February 2001, which was not achieved.

In August and September 2001 the construction manager reported to project management that programme 4,approved at the start of 2001, was no longer achievable. This took into account continuing slippage in the supply ofsome design information, difficulties arising from the underperformance of one contractor responsible for claddingon the MSP block and particular design issues on some packages, for example the complex engineering in thedebating chamber roof.

The construction manager advised that even if all the outstanding design issues were resolved without delaycompletion based on programme 4 would be November 2003. Project management instructed the constructionmanager to investigate the feasibility of a revised programme, which would mitigate the delay to a more acceptableApril 2003 completion target for the debating chamber.

Programme 5 – September 2001 to January 2002

Between September 2001 and January 2002 the construction manager issued programmes 5A and 5B. Theyproposed a revised procurement programme and other revisions to achieve the April 2003 completion target.

Page 29: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

27Part 2. Project overview

2.54 My audit of the CorporateBody’s accounts for 2001/02examined questions arising from theuse of letters of intent and interimcontracts to allow works to start atthe earliest possible date. I examinedthese matters because of thetermination of one contract (FlourCity) in October 2001 (Exhibit 13).Otherwise I consider theeffectiveness and propriety of thevarious mitigation measures notedabove in Part 5 of this report.

2.55 The construction manageradvocated other strategies topromote delivery of the project ontime and cost. In particular theconstruction manager requested adesign freeze at critical points, theearliest in July 2000 (for the criticaleast frame package only), to avoiddisruption and slippage fromrevisiting of designs once approvedby the client.

2.56 These various mitigationapproaches were introduced tosupport completion of theprogramme by December 2002, inaccordance with the constructionmanager’s revised and compressedprogramme 4 issued in January 2001.

2.57 In May and June 2001respectively the Progress Groupreported to the Corporate Body andthe Corporate Body reported to theParliament on the overall progress ofthe project. Neither offered anycommentary on the prospects forcompletion of the project against thetarget of December 2002 other thanit was on the advice of theconstruction manager. However inAugust and September 2001 theconstruction manager reported toproject management that theprogramme 4 target of completion byDecember 2002 could not beachieved.

As construction intensified in 2001

programme monitoring and

coordination became more

important

2.58 By December 2001, althoughdelays had affected almost everypackage, project management hadon the advice of the cost consultantand the construction managerapproved tenders for 33 workspackage contracts (out of the total ofsome 60 construction workscontracts for the project). Thisincluded tender recommendations

the successful tenderer from thefirst stage was invited to providea single tender for the remainingsecond stage work, once thenecessary design information hadbecome available. This allowed anearlier start on the initial worksthan otherwise would be possible.

• Re-scoping selected packages,meaning that work planned to betendered was awarded bynegotiation with contractorsalready appointed and working onsite, thereby saving timeotherwise required to completethe competition, appointment andcontractor mobilisation processes.

• Revising the order ofprocurement of packages, toallow a small number of largecritical packages to be progressedat the earliest possible dates. Re-sequencing and reducing the timeallowed for some critical activities.

• Increasing use of letters of intentand interim contracts to allowworks to start at the earliestpossible date, before a fullcontract and commercial termshave been agreed.

Source: The 2001/02 Audit Of The Scottish Parliamentary Corporate Body, Auditor General for Scotland, December 2002

Exhibit 13Audit findings on the Flour City contract

• Because of the termination of the Corporate Body's contract with Flour City Architectural Metals (UK) Ltd (FlourCity) in October 2001, Audit Scotland examined the contract as part of my audit of the Corporate Body'saccounts for 2001/02.

• In the Flour City contract I was concerned about the Corporate Body’s reliance on interim contracts and theabsence of a thorough financial assessment of Flour City at key stages. The audit questioned the effectivenessof the selection procedures in that case and highlighted that the full trade contract with Flour City was notsecured until August 2001, almost eight months after post tender negotiations started and only two monthsbefore the Corporate Body terminated its contract in October 2001.

• My audit showed that at that time similar risks – the use of interim contracts and delays in securingperformance bonds and parent company guarantees – affected many other packages.

• In September 2002 I informed the Accountable Officer of what I considered then were the key concernsemerging from this work. He accepted that some interim contracts were allowed to continue long after tradecontracts should have been finalised and that were significant delays in obtaining some performance bonds andparent company guarantees. Fortunately, except in the Flour City case, none of the risks implicit in this situationappeared to have crystallised. Following my audit the Accountable Officer took action to ensure that wherenecessary full contracts, bonds and guarantees were put in place and to prevent similar risks arising again.

Page 30: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

28

for 15 out of the 20 mainconstruction contracts for theHolyrood project (Exhibit 14). Thetotal value of work awarded byDecember 2001 was £98 million, 78%of the total estimated final constructioncost at that time excluding risk.

2.59 In October 2001 there weretensions between projectmanagement and the architecturaljoint venture about designmanagement and coordination. Thethree partners of the architecturaljoint venture had disputed theirrespective roles in relation to designmatters, and project managementconsidered the dispute wasinterfering with the progress of thedesign work. Project managementrequired the joint venture to vestresponsibility and authority for thecontrol of design delivery to a singlenamed director. After exchangesinvolving legal advisors on both sides,the parties agreed specific roles andresponsibilities within the jointventure for management of thedesign process, which satisfied theproject management’s requirementsfor clearer direction and control.

2.60 In October 2001, partly inresponse to the project slippage,project management and theProgress Group concluded additionalresources were required to allowproject management to increasemonitoring, manage change andprovide information and decisionswithin a compressed timetable.

2.61 Between October 2001 andFebruary 2002 project managementroles were redefined. In November2001 project managementcommissioned Turner TownsendProject Management to examine itsorganisation and makerecommendations. Taking account ofTurner Townsend’srecommendations three extra fulltime project managers joined the

client’s project management teambetween July 2001 and March 2002(although in January 2002 one projectmanager also left the team). Inaddition three other project managersand a forensic programmer weremade available to the team part time,as the need arose from December2001.

2.62 Between November 2001 andFebruary 2002 the forensicprogrammer investigated the basis ofthe construction manager’sprogramming recommendations, andsought improvements both inresources dedicated to this work andthe methods of analysis.

2.63 In addition to these increases inproject management resources tworoles that Turner Townsend hadidentified as desirable were filledfrom existing resources:

• A cost consultant to ensure thedelivery of a robust cost plan, andmanage the cost dimension ofthe project’s risk register – amember of project managementassumed this role.

• A design manager, to oversee theproduction of design information,and manage and coordinate thedesign change and approvalprocess – one of the architecturaljoint venture’s partners agreed toundertake the duties of designteam manager.

Cost and time problems caused

the Corporate Body to consider

stopping the project to take stock

in November 2002

2.64 During 2002, the ProgressGroup’s main focus was its concernsabout the continuing difficulties withslippage from cost increases,combined with its dissatisfaction atthe quality of cost and programmeinformation it was receiving.

2.65 In March 2002, completion wastargeted for April 2003 underprogramme 5 issued the previousautumn. However the ProgressGroup had real and deep concernsabout achievability within thistimescale. It was dissatisfied withthe construction manager mainly forproviding, in the Group’s view,optimistic programming advice. Theconstruction manager’s view is thatits programming always sought tomeet the client’s desire to achievethe earliest possible completion datebased on assurances andcommitments given to it and theclient by the design team and tradecontractors.

2.66 In June 2002 the Group notedwith alarm the level of additional riskin the latest cost review led by thecost consultants. Risk workshopsand cost reports in the subsequentmonths confirmed the forecast of asignificant increase in potentialproject costs, potentially another £34 million. In March the CorporateBody had reported to the FinanceCommittee a £6 million increase inpotential project costs to £265million13. The Group was dissatisfiedwith the cost consultants for, in theGroup’s opinion, not managing costseffectively14.

2.67 Also in June 2002, inconsultation with the Progress Groupand project management, theCorporate Body concluded the nextquarterly report to the FinanceCommittee due the same monthwould exclude the risk figures. Thiswas because the figures wereuncertain and the Corporate Bodywould provide updated figures in thenext report due in September. TheCorporate Body’s report to theFinance Committee in Septemberindicated a potential increase inmaximum costs to £295 million15.

13 Excluding £14 million landscaping costs.14 We discuss the roles and responsibilities in Part 5, including responsibility for managing costs. 15 Excluding £14 million landscaping costs.

Page 31: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

29Part 2. Project overview

The Exhibit shows the timing, scope and cost of the 20 highest cost trade contracts for the Holyrood building. Thesecontracts together are now estimated to cost £177 million, 78% of the £226 million total construction cost excludingrisk.

The date of project management’s acceptance of the tender recommendation is only an approximate measure of thetiming of the decision to proceed with each contract. In many cases project management issued letters of intent tothe contractor to allow work to start as soon as possible and full contracts were not agreed until months after tenderapproval, while commercial terms were finalised.

The estimated cost in each case is from the cost consultant’s reports to project management in 2004. In some casesestimates at the time of tender recommendation were different, because of subsequent increases or reductions in theexpected scope of the work in each case.

Exhibit 14Tender approval for 20 main Holyrood construction contracts

Source: Audit Scotland

Con

trac

t com

mitm

ent (

£ m

illlio

n) Cost plan allowanceplus estimated inflation

Trade contract value

Estimated final cost(excluding risk)

Description

Sept

embe

r 199

9 - S

ubst

ruct

ure

conc

rete

wes

t

Sept

embe

r 199

9 - T

ower

cra

nage

Febr

uary

200

0 - M

SP c

oncr

ete

fram

e

Febr

uary

200

0 - S

caffo

ldin

g

Nov

embe

r 200

0 - Q

ueen

sber

ry H

ouse

wor

ks

Janu

ary

2001

- Su

bstru

ctur

e ea

st

Janu

ary

2001

- A

ssem

bly

build

ing

conc

rete

fram

e

Apr

il 200

1 - M

SP c

arpe

ntry

& jo

iner

y

Sept

embe

r 200

1 - E

ast e

lect

rical

Oct

ober

200

1 - F

oyer

roof

and

gla

zing

(sta

ges

1&2)

Oct

ober

200

1 - A

ssem

bly

build

ing

roof

ing

Oct

ober

200

1 - S

tone

cla

ddin

g

Nov

embe

r 200

1 - M

echa

nica

l & p

lum

bing

eas

t

Dec

embe

r 200

1 - A

ssem

bly

win

dow

s

Dec

embe

r 200

1 - M

SP b

ay w

indo

ws

Janu

ary

2002

- To

ilet/f

itnes

s/ch

angi

ng fi

t-out

May

200

2 - S

peci

alis

t gla

zing

(sta

ges

1&2)

Aug

ust 2

002

- Zon

e 1

fit o

ut

Sept

embe

r 200

2 - Z

one

2 fit

out

Oct

ober

200

2 - Z

one

3 fit

out

0

5

10

15

20

25

30

35

40

September 1999 –first major workscontract approved

December 2001 –15 of 20 major workscontracts approved

Page 32: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

30

2.68 In October 2002 projectmanagement approved programme6B for issue, aiming for completionof the debating chamber by June2003. In earlier drafts theconstruction manager had suggestedNovember 2003 was more realisticbut after presentation to the ProgressGroup project, management hadrequested measures to better thisdate. Programme 6B as issued didforecast June 2003 completion, butremained heavily qualified (Exhibit 15).

2.69 In November 2002 the ProgressGroup had advice from theconstruction manager calling forfundamental changes to theproduction and management ofdesign information by the designteam. At the same time thearchitectural joint venture provided arebuttal of this. Project managementcontinued to question how wellresources were being targeted in allareas of the project.

2.70 In November 2002 projectmanagement and the ProgressGroup briefed the Corporate Body onprogress with the project. Furtherincreases in potential project costswere expected.

2.71 In December 2002, in the lightof the difficulties facing the projectthe Corporate Body discussed threepossible courses of action: stoppingthe project; instructing a further‘Spenceley’ type review; and puttinga cap on costs. The Corporate Bodywas concerned that no one knewwhat was happening, when theproject would be complete and howmuch it would cost.

From January 2003 the aim was to

complete the project as quickly as

possible as the best way of

containing costs

2.72 Previously, in October 2002 thecost consultant had advised the

Progress Group that the main risks tocosts was associated with timingissues. The Progress Group andproject management advised theCorporate Body that the costconsultant’s advice was that thebiggest single risk to cost was delayand the aim should therefore be todrive the project forward to theearliest possible completion.

2.73 In December 2002:

• The Corporate Body reported tothe Finance Committee that theestimated project costs includingpotential risks costs had increasedto £311 million16. This was 14%more than the maximum £274million potential cost reported tothe Committee in January 2002

• With regard to the programme,the latest completion target(programme 6B final) was June2003 (albeit in the same monththe Corporate Body had alreadyreported to the FinanceCommittee that some importantareas could not be completeduntil August 2003).

2.74 Just a month later, in January2003, the cost consultants advisedthat the most likely outturn was £324million17. Because of the problemswith the project the Progress Groupconsidered project managementshould seek to clarify the managementstructure of the design team, and togive the construction manager a leadrole in driving progress forward.

2.75 In subsequent months theconstruction manager continued tosuggest a range of measures toimprove project management anddesign team effectiveness, in theinterests of achieving the programmeand cost objectives. Actions included,for example:

• In January 2003 the design teamintroduced twice weekly

meetings to resolve designissues, with participation fromboth project management and theconstruction manager.

• Project management requestedthat the architectural joint ventureprovide a named deputy for itslead partner responsible foroversight of design, to expeditedecision-making when necessary.

• There was renewed emphasis onpriorities and identifying andresolving detailed design relatedand interface/interdependencyissues, as part of the continuingprogramme monitoring and reviewprocess. In April 2003 at theconstruction manager’ssuggestion, project managementwrote to all parties seeking toimpose a ‘design freeze’ on allpackages at the end of that month.

• Monitoring of individual tradecontractors performance, includinginitiatives to step in and assistmanagement where necessary.

2.76 In April 2003 constructionactivity on site was starting to peak.There were 1,100 people working onthe site, including trade contractorsand their suppliers, projectmanagement and representatives ofall the consultants.

2.77 Also in April 2003, however, theconstruction manager reported thatprogress was four months behindprogramme 6B final issued in January2003. A new and final programme 7was therefore required to beprepared and agreed. Programme 7was first prepared in May 2003 andissued in August 2003. The May2003 draft programme soughtcompletion by September 2003, butwhen the final version was issued inAugust 2003 the agreed target forcompletion of the debating chamberwas April 2004. This was ten monthslater than the June 2003 target for

16 & 17 Excluding £14 million landscaping costs.

Page 33: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 2. Project overview 31

Exhibit 15Slippage in programme 6 in 2002

Programme 6A interim May 2002

This interim programme was extremely qualified. The construction manager stated there were too manyimpediments outstanding to predict a definitive date to completion. Design development was still ongoing, scopewas changing and the construction manager and the trade contractors were not confident in supplying programmecommitment.

The construction manager agreed with project management that the interim programme would be issued whilstmore certainly was sought and commitment obtained from the design team.

Programme 6B, for Queensberry House and the MSP building only, August 2002

Although the construction manager anticipated that sufficient information would be available over the course of thefollowing two to three months to enable it to produce definitive programmes, in the event this was not achieved.This issue was for Queensberry House and the MSP building only and these programmes were supplementedwith a list of assumptions and programme risks. This programme referred to an April 2003 end date, highlyqualified because of the extent of information and issues yet to be resolved.

Full programme 6B issued for review September 2002

This showed the debating chamber completing in August 2003 with overall completion in November 2003. Thisprogramme was presented to the Progress Group in October 2002, after which project management instructed theconstruction manager to put in place measures to better the debating chamber access date.

Revised programme 6B issued October 2002

This identified the debating chamber completing In June 2003. Again the programmes were highly qualified andwere based upon the design team achieving critical dates. The programme also incorporated accelerationmeasures to a number of key trade packages.

The programmes for the MSP building and Queensberry House issued in August were in delay and had not beenissued to the trade contractors. The construction manager agreed to prepare a second edition of the programmes,which would take account of further design development and trade contactor delay.

Complete programme 6B final issued in December 2002 (to trade contractors in January 2003)

With regard to the MSP building and Queensberry House the programme included instructions to accelerate theworks, re-sequence the works and increase resources to achieve key dates, which would attract a significant costpremium. With regard to the more critical east side of the works at that time there were just six months to theforecast June 2003 completion of the debating chamber, and the construction manager reported a continuing trendof critical path slippage. In April 2003 the construction manager reported progress was four months behindprogramme 6B final.

Source: Audit Scotland

Page 34: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

32

completion in the previousprogramme 6 and seven monthslater than the first draft programme7A had proposed only four monthspreviously (Exhibit 16).

2.78 In May 2003 the costconsultants advised that because ofthe extension of time there would besignificant additional project costs, asyet unquantified.

The First Minister announced an

inquiry into the increasing costs

and delay affecting the project in

June 2003

2.79 In June 2003 the First Ministerannounced an independentinvestigation into the escalating costsand construction delays that areassociated with the new Parliamentbuilding. Following discussions withthe Presiding Officer, the FirstMinister announced that Lord Fraserof Carmyllie would lead theinvestigation (Exhibit 17 overleaf).There was a preliminary hearing inSeptember 2003 and the HolyroodInquiry commenced its mainproceedings in October 2003.

Revised arrangements for cost

reporting commenced in July 2003

and a fee cap was agreed in

August 2003

2.80 In June 2003 the costconsultant estimated the additionalcosts could be some £37 million,assuming a November completiondate. The Progress Group wasdissatisfied with this advice, whichwould result in total project costsincreasing to £359 million18. TheGroup was dissatisfied with how thecosts had been reported, in its viewthe cost consultant appeared simplyto have reacted to the newprogramme. The cost consultant’sview was that everything wasprogramme driven and every changehad a cost attached. The role of the

cost consultant was to advise oncosts but it did not give instructions.The cost estimates reflected hugechange since January 2003 in howthe site operated.

2.81 The new Corporate Body19

reported this increase in costs to theFinance Committee in June 2003.Reflecting its aim to be astransparent as possible, theCorporate Body commenced regularmonthly reporting to the FinanceCommittee from that month. FromJuly 2003 its reports on the costs ofthe project included landscapingcosts, previously excluded.

2.82 In July 2003 the Progress Groupalso instigated changes to the costconsultant’s reporting of cost andrisk. Hitherto the cost consultant hadreported fortnightly to each meetingof the Group with a detailed,package-by-package estimate of costexcluding risk. While the costconsultant had also preparedseparate estimates of risk costs, ithad done so less frequently – onaverage every six weeks betweenMarch 2001 and July 2003. From theend of July 2003 the cost consultantprovided an additional new reportfortnightly to the Progress Group,detailing risk at a package-by-packagelevel. The Group found the costconsultant’s frank commentsincluded in these reports helpful. Itagreed these reports should not beissued to the design team or theconstruction manager because, ifthey were the comments maybecome less open.

2.83 In July and August 2003 theCorporate Body reported to theFinance Committee on the progressof negotiations with the consultantsto cap their fees. In summary projectmanagement negotiated agreementswith each of the parties with the aimof ending the direct link between theconsultants’ fees and how much the

project would cost in total. At thetime of preparing this report all theconsultants had signed thenecessary contract variationdocuments to formalise this change,with the exception of thearchitectural joint venture.

2.84 We examine the value formoney of the fee arrangements inPart 5 of this report.

Construction reached a peak in

December 2003

2.85 On site, between August andDecember 2003 resolution of designissues, and site congestion andinterdependencies continued to bethe dominant considerations.

2.86 There were continuing concernswith the performance of individualtrade contracts, where work onindividual critical packages wasdelayed with knock-on affects forother contractors. For individualcontracts issues included, asillustrative examples:

• a lack of resources and delays inreceiving necessary materialsdelaying work on the MSPbuilding cladding

• a lack of resources delaying fit outworks in Queensberry House

• difficulties in fitting windowspreventing completion of stoneand roof works by other contracts

• internal fit out being delayed bywindow installation delays.

• difficulties in finalising necessarydesign details for fixing bracketingfor precast cladding panels

• a sub-contractor going intoadministration

18 Excluding estimated landscaping costs of £14 million. 19 From May 2003 the members of the Corporate Body were George Reid MSP (Chair), Robert Brown MSP, Duncan McNeil MSP, John Scott MSP and Andrew Welsh MSP.

Page 35: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 2. Project overview 33

Exhibit 16Slippage in programme 7 in 2003

Source: Audit Scotland

Programme 7A draft April 2003

The significant slippage identified in April meant a new programme series 7 was required. Programme 7A draftincorporated a 31-page schedule of issues to be resolved, package-by-package. The construction managerreported that completion of the debating chamber, now targeted for September 2003, was extremely ambitious.

In May 2003 project management wrote to the construction manager instructing them to release programme 7Ain draft to trade contractors and to immediately report progress against this new programme. The basis forreleasing the programme was that project management had got reassurances from the design team on theircommitment to achieving its requirements. Project management highlighted that changes to the design at thisstage would be unacceptable to the client “and the team must be focused on the delivery of the remaining designinformation and resolving the existing issues with trade package contractors”.

In July 2003, immediately following discussion at a Progress Group meeting the previous day, the constructionmanager advised project management it could finalise milestone dates for the completion of the MSP building,Queensberry House and back of house areas, but not for the remaining elements of the project due to the degreeof remaining key risks. The primary risk which the construction manager highlighted at that stage related to thespecialist glazing package and its knock-on impact on all other trades.

Programme 7A final August 2003

After a review of the outstanding information and approval by project management, the construction managerissued the revised programme. It showed a revised completion date of the end of August 2004 with completion ofthe debating chamber in April 2004.

Therefore in the four months between the issues of 7A draft and final the programme for the debating chamberhas slipped seven months, from September 2003 to April 2004. The reasons for this include a degree of continueddesign development, the construction complexity of the chamber and earlier programming optimism.

Programme 7B February 2004

Slippage on programme 7A became evident almost immediately. Consequently in February 2004 programme 7Bshowed a completion of the chamber in August 2004, which remains the target. In adopting this programme theconstruction manager continues to highlight the continued risks in issuing updated design information.

Page 36: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

34

Source: Official Report

Exhibit 17Holyrood inquiry – First Minister’s announcement June 2003

Extracts from the Official Report Thursday 19 June 2003:The First Minister (Mr Jack McConnell): I thank the Presiding Officer for giving me the opportunity to make astatement today on my plans for an investigation into the Holyrood building project. …

I consider that an independent investigation into the escalating costs and construction delays that are associatedwith the new Parliament building should be initiated because the Holyrood building project, more than any otherissue, overshadows the many real achievements of this young Parliament.

Following my discussions with the Presiding Officer, I announced last week that Lord Fraser of Carmyllie hasagreed to investigate the matter on our behalf. I am today making public the remit for the investigation, which Ihave agreed with the Presiding Officer, Lord Fraser and the Auditor General. …

The investigation will review the policy decisions that were taken in relation to the project prior to and since itstransfer to the Scottish Parliamentary Corporate Body on 1 June 1999. The investigation will build on the AuditorGeneral's existing findings in respect of procurement strategy and cost control and contractual and projectmanagement arrangements and extend consideration of those issues to cover the subsequent stages of theproject.

The investigation will produce a full account of the key decisions and factors that have determined the costs andvalue of the Parliament throughout the life of the project. It will also identify the lessons that are to be learned forthe procurement or construction of major public buildings in the future.

The investigation will report to the Parliament and to ministers as soon as reasonably practicable, taking account ofthe Auditor General's intention to examine the economy, efficiency and effectiveness with which resources havebeen used at all stages of the Scottish Parliament building project.

Lord Fraser's report will be submitted to the Executive and the Parliament. The Auditor General's previouslyplanned value-for-money audit will be presented to the Parliament as normal. The Parliament will determine itsown processes for dealing with the report and the audit document, including such committee consideration andthe calling of witnesses as is considered appropriate.

Page 37: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 2. Project overview 35

• additional work to roof lights andto the complex roof structure inthe debating chamber, preventingremoval of temporary accessscaffolding within the chamberthereby delaying other workrequiring access to the chamber.

2.87 In October 2003 as constructionapproached its peak, the ProgressGroup and project managementplaced the highest priority onachieving the programme. Thearchitectural joint venture raisedconcerns about quality sufferingbecause of this. As an example ofthis, in October 2003 the architectproposed to remove and redo thestainless steel roof on tower 3because the final appearance of partsof this roof did not meet what it hadenvisaged. The Progress Grouprejected this proposal because theroof had been produced in line withits specification and the changesrequested by the architect wouldcause unacceptable delay and extracosts.

2.88 In December 2003 constructionwork was at its peak. There were1,500 people on site every day.Substantial construction progress had been made;

• The MSP building wassubstantially complete, fullyfunctional and serviced.Queensberry House was alsosubstantially complete, althoughmatters such as fitting someexternal doors, completion ofplaster work and associatedjoinery and electrical work,completion of some flooring anddecoration was outstanding.

• The progress on these buildingsallowed resources to moveacross to complete work on theremainder of the site, includingthe debating chamber, thecommittee towers and theCanongate buildings. Here

external cladding, window androof installation remained inprogress.

2.89 In November and earlyDecember 2003 the Progress Groupnoted its serious concerns regardingthe performance of the design team.Project management was dissatisfiedwith the degree of cooperationachieved with the architects and theirspeed of decision making, which itconsidered was inhibiting the closedown of packages and overallcompletion of the project. Becauseof the strength of these concerns theChief Executive of the CorporateBody initiated meetings with theChairman of one of the owners ofthe architectural joint venture. Heobtained assurances in January 2004that there would be improvements inresourcing to address the concernsabout leadership of the design team.In February 2004 the Progress Groupwas advised that design teamperformance had improved.

The current programme was

agreed in February 2004 for

substantial completion in August

2004 at a maximum cost of

£431 million

2.90 By January 2004, althoughprogress towards overall completionwas substantial, slippage continued.The construction manager reportedthat there was delay of between twoand nine weeks on everyconstruction milestone withinprogramme 7A and it was preparingprogramme 7B. The constructionmanager advised that the overall aimto complete construction by June2004 remained feasible. But a newprogramme was required to allowwork to be resequenced taking intoaccount the latest information withregard to progress and resourcesavailable. Programme 7Bconcentrated on ten key packagecontracts that would critically affectcompletion.

2.91 In February 2004 programme7B (the current programme) wasapproved and issued. It aimed forcompletion of all areas excludinglandscaping – which is not critical tobuilding occupation – at the latest byAugust 2004. This includedcompletion of the debating chamberby end July 2004. The programmewas riskier than previously, becauseinevitably with a comparatively fixedend date there was less time tocomplete the same work. Itincreased the number of buildings tobe worked on concurrently right upto the August completion date.

2.92 In February 2004 advice fromthe cost consultant was that forecastcosts would continue to increase,particularly from the extension oftime arising from the movementfrom programme 7A to 7B. In itsreport to the Finance Committee inFebruary 2004 the Corporate Bodyreported increases of £15 million inestimated construction costsincluding VAT plus an increase in therisk estimate of £14 million (including£2 million for landscape risks). Theoverall potential maximum cost ofthe project therefore increased from£401 million to £431 million.

Ninety per cent of final accounts

for construction remain to be

finally settled

2.93 In April 2004 projectmanagement had approved orreceived recommendations for finalaccount settlement for 21 contractswith a combined cost of £26 million.For some 90% of the constructionwork, therefore, final accounts haveyet to be settled. The CorporateBody has also yet to settle finalaccounts with its consultants.

Page 38: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

36

2.94 Agreeing final accounts includesthe process of negotiating:

• claims from contractors againstthe client for extra costs, forexample as a result of delayoutside the contractors’ control oradditional work

• claims for set off (reduction) bythe client against contractorswhere the client has incurredextra costs because of thecontractors’ poor work or underperformance.

2.95 Project management aremonitoring action by the costconsultant and construction managerwhose responsibilities include theinitial examination of claims andrecommendations to the client aboutsettlement. The consultants haveaudited claims for delay costs as thework has progressed. If agreementon any account cannot be reached bynegotiation with a contractor there isthe option for either party to seekadjudication, arbitration or litigation.

2.96 In April 2004 projectmanagement concluded that thegeneral approach should be to advisethe Corporate Body in theSummer/Autumn of 2004 on issuesin relation to the handling of all claimsand seeking guidance on theframework to be adopted. Projectmanagement recognise there isscope for obtaining legal advice onsome matters relating to claims inadvance of providing advice to theCorporate Body. At the time ofpreparing this report projectmanagement had identified severalpossible claims cases where actionmay be required, but there was noagreed register of issues or actionsto be considered. Projectmanagement was considering anindependent review of all claimnegotiations, because of the need tobe certain that very significantincreases in package costs could bejustified in each case.

Page 39: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

This analysis concentrates on why

the project did not achieve the

2000 target for completion by the

end of 2002

3.1 In 1998 the completion date forthe new Parliament building was July2001. In my September 2000 report Iexplained the reasons for the delaysin the forecast completion untilDecember 2002. One major difficultyarose from achieving a settled StageD design. Another problem arosefrom difficulties encountered by thearchitects in complying with theoriginal demanding brief to a tighttimetable. There were alsounforeseen changes requested bythe client. These changes wereprincipally an adjustment to thelayout of the debating chamber inJuly 1999 and a significant increasein the total space required in the newbuilding in autumn 199920.

3.2 In 2000 the Parliament set atarget to complete the project by theend of 2002. Part 2 summarises the20 months slippage, which has since

affected the project. In the earlystages, the overall constructionprogramme 3C (July 2000) plannedfor completion by December 2002.Successive programme issues haveshown delays, culminating in thecurrent programme 7B, in which thetarget for completion is August 2004(Exhibit 7).

3.3 Two significant factors should beconsidered in relation to the slippageof the Holyrood project since 2000:

• The client did not altersignificantly the userrequirements for the buildingonce it approved the Stage Ddesign in June 2000. Themeasured construction costs ofaccepted change requests madeby the client since Stage D issome £0.6 million. This is just0.2% of the current constructioncost estimate.

• The slippage since September2000 is 20 months in a 42-monthperiod. It has affected all parts of

the construction of the building(Exhibit 18 overleaf). Taking intoaccount the very small amount ofclient variation in this time, theslippage is exceptional whencompared with other large-scaleconstruction projects21.

3.4 In examining the progress of theproject including the reasons forslippage, I did not seek to form anopinion on whether any individualparty or contractor has been at fault.It is the responsibility of the client tomanage its consultants and itscontractors and to assessperformance. The client should avoidduplicating the work commissionedfrom its consultants. TheAccountable Officer is accountablefor the significant public funds spenton these contracts.

37

Part 3. The reasons for later delivery

20 The new Scottish Parliament building: An examination of the management of the Holyrood project, AGS/2000/2, Paragraphs 2.7 to 2.921 See, for example, Review of Large Public Procurement in the UK by Mott MacDonald for HM Treasury in 2002. This study examined seven large non-standard public

buildings. It recommended that, for the purposes of risk analysis/project planning, the maximum risk allowance for slippage during the construction of non-standardbuildings should be 39 per cent compared to the estimate at outline business case stage. By comparison, the construction slippage now expected on the Holyroodproject will be some 72 per cent.

Page 40: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

38

Source: Audit Scotland

MSP

bui

ldin

g

Que

ensb

erry

Hou

se

Tow

ers

1 an

d 2

Tow

ers

3 an

d 4

Can

onga

te b

uild

ing

Deb

atin

g ch

ambe

r

Ligh

twel

l/pub

lic s

tair

Fina

l han

dove

r

Exhibit 18Slippage in completion of the main parts of the Holyrood building

26 monthsslippage

20 monthsslippage

19 monthsslippage

19 monthsslippage

20 monthsslippage

19 monthsslippage

19 monthsslippage

20 monthsslippage

Forecast duration

Actual Duration

September 1999October 1999

November 1999December 1999

January 2000February 2000

March 2000April 2000May 2000June 2000July 2000

August 2000September 2000

October 2000November 2000December 2000

January 2001February 2001

March 2001April 2001May 2001June 2001July 2001

August 2001September 2001

October 2001November 2001December 2001

January 2002February 2002

March 2002April 2002May 2002June 2002July 2002

August 2002September 2002

October 2002November 2002December 2002

January 2003February 2003

March 2003April 2003May 2003June 2003July 2003

August 2003September 2003

October 2003November 2003December 2003

January 2004February 2004

March 2004April 2004May 2004June 2004July 2004

August 2004September 2004

October 2004

Page 41: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 3. The reasons for later delivery

• The architects and some tradecontractors had problems indelivering some critical elementsof the work within the requiredtimescales.

Under construction management

the client retains construction risk

3.6 There are difficulties andcomplexities in any large constructionproject and all forms of procurementinvolve risk. Good practice is toselect the form of contract whichensures risks are allocated to thosebest able to manage them22. Underconstruction management the clientaccepts most of the risk associatedwith coordinating construction.

• All contracts are placed directlybetween the client and the tradecontractors – there is no maincontractor role.

• The client consequently retainsinterface risk, although managedthrough the construction manager.

• A construction manager isappointed as a consultant,

coordinator of the design team,manager of trade contractors andconstruction adviser to the client.

• At the same time the clientretains control of the designprocess and should avoid payingfor estimated risks that do notmaterialise.

• Under any form of contract theclient should pay for the risktransferred to other parties.

3.7 Under a ‘traditional’ buildingcontract (Exhibit 19) the design of thewhole building is taken to sufficientdetail23 for a competition for thewhole or greater part of theconstruction works before a singlemain contractor is appointed.Developing the design to this level ofdetail takes time. However the mainpotential advantage to the client iscertainty. Once a design has beenapproved there is scope for the workto be tendered on the basis thatmost programme and delivery riskremains with the contractor providedno changes are instructed.

The main cause of the slippage is

delays in design of a challenging

project delivered against a tight

timetable using an unusual

procurement route

3.5 There are five main reasons forthe slippage affecting the projectsince 2000:

• There are inherent risksassociated with the constructionmanagement procurementmethod but the client organisationdid not have experience in thisprocurement route.

• There were difficulties associatedwith the construction of a verycomplex, densely developed,unusual building against very tightdeadlines.

• The original timetable forcompletion was compressed withno room for slippage on thecritical path.

• At a later stage in the project, sometrade contractors were givenresponsibilities for design issues.

39

Exhibit 19Key features of traditional construction

Source: Audit Scotland

• Client has separate agreements with designer and main contractor. This separation may discourage innovationand the early identification of ‘buildability’ issues.

• Design and construction work are sequential, extending timescales.

• Construction is 80% tendered before work starts on a lump sum basis, providing greater cost certainty oncethe design is completed.

• There may be a lower degree of construction risk transfer from the client to contractors compared to designand construct, PFI or prime contracting methods. Suppliers prices will reflect the degree of risk transferachieved.

22 In my 2000 report I explained the Scottish Office chose construction management for the Holyrood project in 1998 after due professional consideration but thereshould have been a systematic assessment of risks in this route.

23 RIBA plan of work stage F.

Page 42: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

40

3.8 Construction management isquite different to a traditional buildingcontract. The main benefit is shortertimescales. Constructionmanagement compresses the overalldesign and construction programmeby allowing procurement andconstruction to proceed before thedesign of all packages is completed.It is intended for use when there is acompelling value for money case forcompleting a building project at theearliest possible date (eg, to benefitfrom forecast high rental incomes, orto achieve unmoveable opening dates).

3.9 Under construction managementcontractors commence works beforeall design is completed. There arerisks in this, for example inproceeding with the foundations andsubstructure of a building without acomplete understanding of thesuperstructure design. Because theclient contracts with the tradecontractors directly there is no maincontractor to accept and managerisks from programming andcoordination of contractors. Thedesign team must envisage thetotality of the design, accommodateany uncertainty in later, dependentpackages and progress long lead-time items (eg, lift packages) early inthe procurement process. Thereneeds to be experience, team spiritand cooperation to manage the risks.

3.10 The advantage of constructionmanagement is that the design ofpackages of work may be changedup to the point when the contract isawarded without necessarilyincurring additional costs for thosepackages. Once packages are let,their content should be fixed.Unforeseen change after a contractis let is likely to incur delay andadditional costs.

3.11 The audit examination does notprovide a detailed assessment of allthe changes made during theHolyrood project. That would be anenormous undertaking. The

Corporate Body has estimated therehave been some 10,000 proposedchange orders issued over thecourse of the project. Many of thesechanges could be small and somemay have arisen before the Stage Dapproval. However the verysignificant impact of change in eachcontract at awarded is evident from asimple comparison of the cost of thecontract at tender/contract awardstage and the current forecastoutturn (Exhibit 20).

The complexity of the Holyrood

project has involved major

challenges in programming the

construction

3.12 Before tenders for the mainbuilding work in the east of theHolyrood site could be obtained, thedesign team had to develop the June2000 Stage D design into detailedpackage designs to provide sufficientinformation for tendering. This was avery large task for the design teamand the construction manager.

• The Stage D design was not afully developed and coordinateddesign. The client accepted thisas an architectural Stage D,though some elements, forexample the foyer roof, were nomore than concept design (StageC). The client recognised thestructural and service designinformation was not equivalent toStage D. Despite the complexstructures and organically shapedbuildings, only comparativelysmall-scale drawings wereprovided at Stage D. Exhibit 21(overleaf) illustrates designdevelopment of the foyer roofsince the Stage D design, and theextent of the initial uncertainty.

• Programme 3C which envisagedcompletion by December 2002was issued on 22 July 2000.There were very short timescalesfor subsequent design details tobe provided. For example, the

target date for sending out thetenders for the assembly buildingsuperstructure package – thelargest and most costly package –was 11 August 2000. This wasjust seven weeks after the StageD design was approved without aStage D engineering design.

3.13 In addition to the challenges ofthe design programme particulardifficulties arose because:

• unforeseen structural andconservation matters affectedQueensberry House. There weredelays in securing agreement onthe reconstruction that wasrequired to take into account thefact that Queensberry House wasa listed building and securing therequired listed building consents

• because of the complexity ofmuch of the work, theinterdependency of manyelements and congestion on thesite, there were unforeseenproblems and delays. Exhibit 22(overleaf) illustrates a few examplesof these unforeseen difficulties.

3.14 The construction manager’sresponsibilities are detailed in itscontract with the Corporate Body. Itsresponsibilities include:

• liaison and consultation with allparties

• coordinating the services of thedesign team with the executionand completion of the project bythe trade contractors

• securing agreement to the costplan by the client and the costconsultant as soon as practicable

• ensuring that the project is executedand completed within the costplan and construction period

• programming at all levels ofconstruction

Page 43: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 3. The reasons for later delivery 41

Exhibit 20Post tender changes in contract costs

Source: Audit Scotland

Number of

contracts

Change in outturn costs compared to

original cost

Estimated

cost of these

contracts at

tender

approval

Current

estimated

final cost of

these

contracts

Average

increase (%)

5 Increased by between three and eighttimes £6.7 million £28.0 million 315%

9 Increased by between 95% and 172%ie, doubling or almost tripling £26.1 million £54.7 million 110%

13 Increased by 52% to 91% £41.1 million £69.3 million 69%

14 Increased by 21% to 49% £37.4 million £49.9 million 33%

11 Increased by 5% to 19% £12.9 million £14.0 million 9%

6 Little or no change £4.8 million £4.8 million -1%

58 All contracts £129.0 million £220.6 million 71%

(excluding landscaping and fit out)

Page 44: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

42

Source: Audit Scotland

Exhibit 21Design development of the foyer roof

These images illustrate how the design of one part of the Holyrood building – the foyer roof – has developedbetween June 2000 Stage D design approval and as it has now been built.

Between 2000 and 2004 the estimated cost of the foyer roof increased from £1.8 million to £7.3 million. Appendix2D provides more details of the progress of the contract for building the foyer frame and glazing.

The computer-generated image on the left, lookingwest to the MSP block, is how the architectpresented the concept for the foyer roof to the clientin June 2000.

The structure is distinguished by the extensivepresence of what appear to be curved pane rooflights,with minimal supporting frame, set into what appearsto be a cast structure consistent with the light greyconcrete used extensively on the site.

The roof is a single span construction with nosuggestion of intermediate supporting columns.

The image does not suggest blast performance was aprimary consideration at this stage.

The computer-generated image on the right waspresented to the client in 2001.

Rather than a single span roof the design nowincludes substantial supporting pillars and a heavytubular steel superstructure.

Extensive use of oak and stainless steel linings hasnow superseded the previous simple light grey finish.

Oak beams now support the spine of the roof lightsbut the frame supporting each pane remains minimal.

The photograph to the left shows the foyer roof underconstruction in early 2004.

Compared with the design in 2000 the roof is muchmore substantial in terms of its structural strength and mass.

The roof light panes are now supported in substantialframes, themselves reinforced by secondary struts oftubular steel.

Page 45: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 3. The reasons for later delivery

2, the construction manager was veryactive in developing a wide range ofmitigation measures to respond toprogramme difficulties as they arose.In 2002 project management hadraised questions on the advice of itsforensic programmer about theconstruction manager’s methodologyfor compiling and managing thenecessary construction programmesbut in general the audit found noadditional reason to question themethodology. One shortcoming wasthat the risk analysis for theprogramme was not equivalent to therisk analysis undertaken in relation tothe project costs. The risk workshopsin October 2000 had identified risksto time schedule associated withother project risks, which werecategorised as ‘highly likely’ (Part 2).The construction manager regularlyidentified and reported risk to theprogramme, but did not assess theeffects if delivery targets were not met.

The client maintained a drive for the

earliest achievable completion date

3.17 Between December 2000,when programme 4A was issued,and February 2004 the construction

manager undertook eight programmerevisions to address the problems ofdesign, procurement andconstruction slippages threateningthe successive completion dates.The client mostly rejected theconstruction manager’s initial draftprogramme revision each time andmaintained a drive for completion bythe earliest achievable date. Evenwhere the construction manager’sinitial programme was qualified theclient sought revisions to advancethe completion date. Programmerevisions repeatedly incorporatedassumptions about design andconstruction performance that thedesign team and contractors agreedwere achievable but weresubsequently not achieved. Theclient relied on the programmes andadvice recommended by itsconsultants, set completion targetsbased on these programmes andreported them to Parliament. Exhibit23 (overleaf) shows some examplesof the programme qualificationsbetween 2000 and 2003.

• management of trade contractors.

The construction manager must actin the client’s best interest at all times.

3.15 The six main programmes thatthe construction manager issuedwith client approval between July2000 and February 2004 (Exhibit 7)are only part of its very significantprogramming activity. Its detailedprogramming includes theestablishment, negotiation,monitoring and maintenance ofprogrammes for all 60 workscontracts, together with alllandscaping and fit out works. Theconstruction manager has providedcommentary and analysis of reasonsfor variation, as part of its normalprogramming activity. Allprogrammes and programmereissues have been subject to reviewand testing by project management.

3.16 Audit Scotland, with its expertconsultant, reviewed the constructionmanager’s design, procurement andconstruction programming. In theconsultant’s opinion, the scope andnature of this construction managementservice is normal. As shown in Part

43

Source: Audit Scotland

Exhibit 22Problems of interdependency and complexity of the Holyrood project

• In the case of the large and important specialist glazing package negotiations with a preferred tenderer forthe work appointed in July 2001 had extended without resolution into 2002. There were consequentlysignificant delays for this package and knock-on delays for other packages for which specialist glazing was onthe critical path. Eventually, in May 2002 project management chose to terminate the negotiation with thepreviously preferred tenderer, and appointed immediately another contractor by negotiation.

• By the end of October 2001 the construction manager was reporting that the MSP building cladding wasfifteen weeks in delay. A recovery strategy was put in place in November 2001, which sought to mitigate theimpact of the failed Flour City contract24. This involved re-procurement of the MSP cladding in six separatepackages, resequencing of the works and the adoption of temporary weather protection to allow internalfinishes to progress. These measures were intended to safeguard the then target completion date ofSeptember 2002 for the MSP building.

• Another difficulty arose from the complex design for the roof of the debating chamber. In July 2002 theProgress Group had advice that the beams forming part of the roof were ready to start installation. Installationwas assessed to be seven weeks in delay at that time. In September the Group received advice thatdifficulties with the installation of these beams could add 13 weeks further delay. Again this had an impact onother packages.

24 My December 2002 report on the 2001/02 audit of the Corporate Body describes the circumstances leading to the termination of Flour City’s contract in October2001 and associated contract management issues.

Page 46: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

44

Exhibit 23Examples of programme qualifications

December 2000

The programme 3C overview for the assembly buildings identified that the design team were fifteen weeks behindprogramme. Delays to the procurement of critical packages could amount to a four to six months delay

beyond the December 2002 completion target.

The construction manager sought to mitigate the extent of the delay by revising the design and procurementstrategies and rescheduling the construction sequence for the assembly buildings. This resulted in a sixteen-weekextension to this part of the programme but the target completion of December 2002 was maintained. Theconstruction manager’s report states that, “cladding and roofing packages continue to experience major problemssecuring adequate design information. This may have a delay/effect to critical client sign-off and contract lead-intimes”.

August 2001

The construction manager reported that programme 4C was no longer achievable and that even if all outstandingdesign issues were resolved then completion based on the then proposed procurement and construction

sequence would be November 2003. At that point project management requested that the construction managerinvestigate the feasibility of a revised strategy, which would allow access to the debating chamber ie, completion

by mid April 2003.

September 2001

The programme Series 5A ‘draft’ again sought to re-sequence the works. It put in place a revised procurementstrategy and mitigation measures to maintain the client’s completion aim of April 2003. It is to be questioned whether advancing completion of the debating chamber by some seven months was asensible move at this time The previous month the construction manager had reported that completion

could run out to November 2003. Significant outstanding design issues remained to be resolved.

May 2002

The construction manager issued programme 6A ‘interim’, which was extremely qualified. It stated, “there are too

many impediments outstanding to predict a definitive date to completion”. Design development was stillongoing, scope was changing and the construction manager and the trade contractors were not confident insupplying programme commitment. The construction manager agreed with project management that the interimprogramme would be issued whilst more certainly was sought and commitment obtained from the design team.The programme was highly qualified in terms of the extent of information and issues yet to be resolved but

the end date remained as April 2003.

A fifteen-page document accompanied programme 6A ‘interim’, highlighting assumptions and potentialimpediments to programme security. The impediments focussed upon design team delivery of information and theneed for a ‘marked improvement’ in design coordination.

At this time the construction manager anticipated that sufficient information would be available over the followingtwo to three months to enable it produce definitive programmes. In the event programme 6B was issued forQueensberry House and the MSP building only and these programmes were supplemented with a list ofassumptions and programme risks.

December 2002

Final version of programme 6B issued. Target for debating chamber completion – June 2003.

Page 47: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 3. The reasons for later delivery 45

Source: Audit Scotland

January 2003

The construction manager’s private & confidential report to project management suggested strategies to addresscurrent problems and improve project management and design team effectiveness, such as:

• resolution of a fee dispute with the architects and allay their fears of future retribution for delays

• attendance by project management at all meetings to arbitrate over disputes

• eliminate revisiting of designs previously agreed

• clarification of design team members roles and responsibilities. Reallocation of non-performing members

• increase design team openness on all issues with the rest of the team

• the future involvement of Barcelona at least limited

• the client needs to send a regular, strong and unequivocal message as to their goals and aspirations to theconstruction manager and design teams

• architect to act immediately and to the letter on any client decisions

• appoint additional project management resources.

The construction manager’s report shows that fundamental management and organisation issues were continuingto have a significant impact upon the project four years after work had commenced. It was prepared just sixmonths from the then forecast completion date of June 2003.

Page 48: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

46

3.18 The construction manager isresponsible for programmemanagement including exercising allthe proper skill care and diligence tobe expected for the work. It mustapply reasonable endeavours toensure that the project is executedand completed with the project costplan and within the constructionperiod, albeit it is not responsible forachieving programme. The architect’sresponsibility is to provide allproduction information (except whereagreed otherwise) and to apply itsbest endeavours to achieve theproject timetable. Against abackground of continued design andprocurement slippage projectmanagement and the ProgressGroup challenged both these partiesrigorously but no decisiveimprovement was achieved.Programmes were issued on thebasis that they were targets subjectto critical issues being achieved orresolved. When programmes did notachieve the targets set new targetswere sought and approved but thefundamental problems for non-performance were not overcome.

Design slippage was a major

factor in delaying the overall

programme

3.19 As described in Part 2 therewas a considerable uncertainty aboutthe programme throughout theproject. Targets for packages wereput in place but not achieved.Mitigation measures and new targetswere introduced in an effort tomaintain overall completion but theyalso failed in many cases.

3.20 Uncertainty and slippage arosepartly because:

• there was significant slippage inthe production of necessarydesign information both by thedesign team and by some tradecontractors that were responsiblefor significant design elements

• in some individual packagesdesign development wasextensive, disruptive toprogramme and threatenedhigher costs.

3.21 Audit Scotland’s consultantreviewed the progress of the twentylargest contracts. They trackedmovement between the date theclient approved contract action ineach case and the target for thisevent in programme 4 issued inJanuary 2001. Broadly speaking, thelead-time up to approval of contractdepends on the flow of designinformation. Delays up to this pointare largely a result of insufficientdesign information being supplied tomaintain the tender programme.

3.22 On average these twentycontracts were each delayed by 37weeks. Four of the contracts withdelays of between 19 weeks and 46weeks were the most critical on theproject (Exhibit 24). Many of thedifficulties caused on site over thelast year relate to the assemblybuilding’s cladding and the foyerglazing. If these packages had beenprocured on time, then more timecould have been devoted to tradecontractor design development,manufacture and the resolution ofinterfaces on site.

3.23 Not only were there delays inachieving contract approval, butmany packages were awarded withsignificant elements of workidentified as provisional sums orattracting significant variation aftercontracts were placed. Althoughcontracts were approved, asignificant degree of design activityhad yet to take place in some cases.

3.24 The most significant impact onthe programme during constructionhas been the timing and volume ofvariations. The most complex andtherefore programme-critical area ofthe project was the assembly

building. Within this, the five mostcomplex packages have been thosefor: the assembly frame; the foyerframe and glazing; specialist glazing;cladding and windows; and roofing.

3.25 As Exhibit 25 shows, these fivecontracts have required significantvariations, as measured by howmany ‘notifications of proposedchange’ instructions were raisedduring the work. Exhibit 25 alsoshows the period of variationextends well beyond the originalprogramme period for construction ineach of these cases.

3.26 A detailed month-by-monthanalysis of the most critical packageon the project – the assembly frame– is shown in Exhibit 26 (overleaf).This package is currently over 15months late and there weresubstantial variations taking placemonths after the original plannedcompletion of December 2001. Thisvolume of change must have causedproblems for managing the project,with consequences for theprocurement and delivery ofmaterials and the construction works.

3.27 The main causes for delaysaffecting the project are summarisedbelow:

• The primary cause of the 20months delay to the project sinceSeptember 2000 was the issue ofdetailed design variations and lateinformation during construction. Insome cases trade contractorswere responsible for someelements of design subject tofinal design team approval. Theprocess of design team approvalalso introduced significant delay.

• Delays resulted from late andincomplete release of designinformation for tenders andtender packages being returnedover the cost plan allowance. Inmany cases trade contractors

Page 49: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 3. The reasons for later delivery 47

Source: Audit Scotland

Exhibit 24Initial contract delays for eight large works contracts

Contract

reference

Name of

contract

Date client

authorised

contract

Slippage to

client

authorisation

Critical

slippage?

Trade contract

value at

tender stage

2600 MSP concreteframe

September1999

23 weeks No £4.2 million

2205 Substructureeast

September2000

19 weeks Yes £6.9 million

3320 Foyer roof andglazing

July 2001

31 weeks Yes £5.8 million

3350 Specialistglazing

July 2001

46 weeks Yes £7.2 million

3525 Assemblywindows

June 2001

24 weeks Yes £3.9 million

3528 MSP baywindows

April 2001

37 weeks No £2.7 million

6015 Mechanical &plumbing east

April 2001

28 weeks No £5.7 million

7015 East electrical April 2001

21 weeks No £4.2 million

Source: Audit Scotland

Exhibit 25Design variations for five critical works contracts

Contract

reference

Name of

contract

Original

programme period

(programme 4)

Number of

draft change

orders

Period of

variations

2605 Assembly buildingconcrete frame

November 2000 toDecember 2001

1,800 January 2001 toMarch 2003

3320 Foyer roof andglazing

November 2001 toApril 2002

117 August 2001 toApril 2004

3350 Specialist glazing September 2001 toMay 2002

322 July 2001 to date

3525 Assembly windows June 2001 toMarch 2002

288 July 2001 to date

3645 Assembly buildingroofing

October 2001 toMay 2002

485 July 2001 to date

Page 50: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

48

Note: The chart shows 1,516 variations between December 2000 and March 2003. Variations due to additional work not included in the original forecastprogramme, such as the boundary wall, are excluded.

Source: Audit Scotland

Forecast

Programme

Towers 1 & 2

Towers 3 & 4

Debatingchamber

Canongate

Actual

Programme

Towers 1 & 2

Towers 3 & 4

Debatingchamber

Canongate

Exhibit 26Analysis of impact of design variations on the main assembly buildings contract

Number of change orders in monthN

ovem

ber 2

000

Dec

embe

r 200

0

Janu

ary

2001

Febr

uary

200

1

Mar

ch 2

001

Apr

il 20

01

May

200

1

June

200

1

July

200

1

Aug

ust 2

001

Sept

embe

r 200

1

Oct

ober

200

1

Nov

embe

r 200

1

Dec

embe

r 200

1

Janu

ary

2002

Febr

uary

200

2

Mar

ch 2

002

Apr

il 20

02

May

200

2

June

200

2

July

200

2

Aug

ust 2

002

Sept

embe

r 200

2

Oct

ober

200

2

Nov

embe

r 200

2

Dec

embe

r 200

2

Janu

ary

2003

Febr

uary

200

3

Mar

ch 2

003

120

115

110

105

100

95

90

85

80

75

65

60

55

50

45

40

35

30

25

20

0

Page 51: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 3. The reasons for later delivery 49

were responsible for someelements of design subject tofinal design team approval. Theprocess of design team approvalin these cases also introducedsignificant delay.

• Because of the client’s desire forthe best achievable completiondates there was non-productiveand out of sequence working thatexacerbated the delays.

• The construction managerdemonstrated its commitment tothe project by consistentlyseeking to achieve early targetcompletion dates. With thebenefit of hindsight, the originalprogramme was unachievable. BySeptember 2002 or April 2003(when respectively programmes 6and 7 had been issued), theconstruction manager shouldhave recognised its targets wereunlikely to be achieved.

• It is not clear that projectmanagement did enough toaddress the root causes ofproblems, which were adverselyaffecting the cost andprogramme. The constructionmanager repeatedly preparedconstruction programmes, whichincluded assumptions andcommitments by the design teamand contractors that weresubsequently not achieved.Because all parties agreed thatthe basis for each programmewas achievable the repeatedprogramme slippage raisedfundamental questions about theperformance of all parties, whichno party appears to haveaddressed effectively. Underconstruction management theclient ultimately bears mostconstruction risk but it was unableto find the means to managethese risks effectively.

Page 52: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

totalling £311 million (72% of thetotal project costs)

• increased fees to advisers andsite organisation costs of £68 million (16% of the totalproject costs).

• smaller increases in the remainingfit out, landscaping, siteacquisition and programmecontingency costs, which nowtotal £44 million (10% of the totalproject costs).

There is a question whether the

cost plan underpinning the £195

million target was adequate

4.3 Under the RIBA plan of work fordelivering construction projects,Stage D detailed proposals shouldinclude complete development of theproject brief. The project brief isconcerned with agreeing concepts,performance and parameters such astime and costs. Good practice is thatthe project brief at Stage D willinclude a confirmed cost plan. TheStage D proposal for Holyroodincluded a cost estimate and was

approved on the basis that the costconsultant and the constructionmanager would prepare a packagebased cost plan consistent with theestimate.

4.4 In September 2000 Irecommended that projectmanagement, the design team andthe construction manager shouldagree a cost plan taking account ofrisk and uncertainty to provide asound basis for managing theremaining stages of the project(Exhibit 8).

4.5 In October and November 2000the construction manager provided acommentary on the cost plan. Ithighlighted there was insufficientdesign information to provide reliablecost estimates. It stated the designwas complete architecturally butincomplete structurally and significantrisks remained. It advised projectmanagement to obtain acommitment from the design teamthat it would complete the designwithin the target construction costwithin the draft cost plan.

4.1 In 2000 the Parliament set a£195 million budget for completingthe project. In my 2000 report Ishowed this was much higher thanthe estimate at the start of theproject in January 1998, mainlybecause of:

• a 47% increase in the size of thebuilding, largely at the client’srequest

• the additional complexity of theapproved design compared to theoriginal ideas for the building,resulting in a 48% increase in unitconstruction costs compared tothe initial estimate.

4.2 This part of my report examineswhy, since 2000, forecast projectcosts have more than doubled againto the most recent estimate of £431million. Exhibit 27 shows the mainparts of the total current £431 millionproject costs and how they haveincreased since 2000. The largestincreases arise from:

• increased construction costs andassociated irrecoverable VAT, now

50

Part 4. The reasons for increased costs

Page 53: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 4. The reasons for increased costs

In most of the trade contracts the

estimated final cost greatly

exceeds the original cost plan

allowance

4.7 Forty-one of 58 individual tradecontracts have an estimated finalcost 21% or more above the costplan allowance (Exhibit 20). Thesecontracts accounted for 91% of theestimated contract expenditure. Afew large value contracts account fora large part of the total variance inconstruction costs (Exhibit 28overleaf).

Some of the increased costs are

due to the extended construction

period

4.8 Time is money. In constructionprojects, once contractors have beenappointed, any time slippage canmean additional cost. Extra costsarise from prolongation, delay anddisruption. Prolongation costs are theextra cost of doing the same amountof work over a longer period. The

client will usually bear these costs ifprolongation occurs because ofthings outside a contractor’sresponsibility. For example, ifcontractor A suffers a delay inreceiving design information from athird party it may claim an extensionof time. If the client accepts theclaim it may reimburse contractor Afor labour and overhead costs for thelonger period.

4.9 Similarly, delay or disruption costsmay arise where the contractor’sagreed programme of work is upsetfor reasons outside its control and itcosts money to recover the position.For example, if work by contractor Ais delayed it may prevent necessaryaccess by contractor B to an areaboth contractors must work in,disrupting B’s work programme.Even if contractor B can make up losttime it may incur and claim additionalcosts in doing so, for example fromnecessary overtime payments orother acceleration costs.

4.6 The shortcomings of the costplan in November 2000 were that:

• Normally a Stage D budget orcost plan would be regarded as alimit that must not be exceeded.This seems to be how the clientperceived it for the Holyroodproject in 2000. However muchof the information in the cost plancould only be regarded as anindicative target rather than areliable prediction of cost.

• It was known in November 2000that the impact of inflation waslikely to be between £11 millionand £13 million. At this level itwould consume all the availablecontingency of £11 million. Therewas, consequently, no allowanceat all for risks, although all partiesaccepted risks were very likely tooccur. In November 2000 the riskworkshop quantified additionalcosts for risks at some £61 million, for which there wasno allowance in the budget.

51

Exhibit 27Increase in the forecast cost of the Holyrood project since September 2000

Note: £209 million is the £195 million target plus £14 million landscaping costs excluded from the target. The £108 million construction cost target was basedon March 1998 prices. The current reported £241 million construction costs is based on outturn prices ie, including inflation estimated at some £19 million Source: Audit Scotland

Estimate in AGS reportSeptember 2000:Total £209m(see note below)

£ m

illlio

n

Con

stru

ctio

n co

sts

Fees

and

site

orga

nisa

tion

cost

s

VAT

on c

onst

ruct

ion

Furn

iture

, fit

out e

tc

Land

scap

e an

dan

cilla

ry ro

ad c

osts

Con

stru

ctio

n re

serv

e/co

ntin

genc

y in

clud

ing

VAT

Site

acq

uisi

tion,

dem

oliti

on,

arch

aeol

ogy

0

50

100

150

200

250

Report to Finance CommitteeApril 2004: Total £431m

108

241

26

68

2842

17 20 14 1811

37

5 6

Page 54: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

52

Exhibit 28Initial and current estimated costs for 55 Holyrood construction contracts

Note: Values for the three smallest contracts have been excluded from the upper chart

Source: Audit Scotland

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

Estim

ated

con

trac

t cos

t (£

mill

lion)

Current estimateof final cost(excluding risk)

Initial contract value

26 contracts each costing less than £2 million

0

5

10

15

20

25

30

35

40

Estim

ated

con

trac

t cos

t (£

mill

lion)

29 contracts each costing more than £2 million

Note differentscale

Page 55: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 4. The reasons for increased costs 53

4.10 Exhibit 29 (overleaf) presentssome illustrations of extra costsarising from slippage and disruptionon the Holyrood project.

4.11 Exhibit 30 (overleaf) charts thegrowth in estimated costs for thewhole project from 2000 to 2004. Itshows that most of the increase inconstruction costs became apparentlater in the project, from late 2001onwards when most of the mainconstruction contracts had beencommitted25. In October 2001 a riskworkshop identified for the first timea significant cost associated with thetime related risks and this was thelargest single ingredient of additionalcost in subsequent risk workshops.Also from October 2001 the costconsultant’s fortnightly cost reportsstarted to show a steady increase inthe base construction costscompared to the cost plan.

Some of the increased costs are

due to a very high level of design

development

4.12 Design development is thenormal process when the design of abuilding evolves in parallel with thetendering and appointment ofcontractors and subsequent buildingwork on site:

• Once the detail design proposals(Stage D) are confirmed, thedesign team must develop finalproposals (Stage E) to allow allcomponents and elements of thework to be coordinated.

• The design team must alsoprovide production information(Stage F) with sufficient furtherdetail to allow tenders to beobtained. Production informationat tender stage should besufficient to allow tenderers toprice the work accurately.

• Once a contractor is appointed itmay require more information

from the design team. Theadditional information reflects thedifference between what isneeded to price a job and tophysically deliver the building.

4.13 Design development may covera range of the aesthetic, technical,performance and functional aspectsof a design. In the Holyrood project itinvolved architectural, structural andbuilding service issues. Specialistsand trade package contractors providedsome of the design input and therewas a need for coordination of thewhole. Under the constructionmanagement approach for the Holyroodproject the process was complicatedbecause more design developmenthappened at the same time as (ratherthan in advance of) construction.Design development included obtainingthird party approvals for the design asit developed, for example from theplanning authority, Historic Scotland,and the Fire Officer. An importantfeature for the Holyrood project wasthe need to develop the design tosatisfy security and blast considerations.

4.14 Design development carries arisk of cost increases. But in anyproject there should be adequateallowance made for this aspect in theStage D budget or cost plan, therisks should be managed and thereshould be a change control processto authorise necessary change.

4.15 For the Holyrood project thesystem for authorising changes wasreasonable but it provided a systemprimarily for monitoring not activelymanaging and restraining costs.Before project managementapproved any change to a tradecontract the construction managerwas required to report on anyprogramme impact and the costconsultant was required to report anycost impact. But the reporting of costand programme implications ofchanges that were imminent orinevitable did not offer choice.

4.16 In general terms, for effectivecontrol and management a costmanager/cost controller must anticipateand ‘flag up’ potential change at theearliest opportunity and the designteam or the contractor must providealternative solutions to give the clienta real choice. In the Holyrood project,however, increased costs have beenincurred partly because designdevelopment has driven the project.It seems that, for this project, designdevelopment became a process ofcosting a developing design ratherthan developing the design within a cost.

4.17 The different components of theHolyrood building feature many noveland complex features. For manypackages realising the design meantthat the cost of the works increasedsignificantly compared to the Stage Ddesign and the November 2000 costplan. It seems that there was not afull appreciation of the complexity ofthe design early enough in theproject. Only as the design evolveddid it become possible to estimateaccurately the cost of realising it. Thetime pressure was such that onceaspects of the design were realisedand the costs understood often therewas no alternative but to proceedbecause the potential savings frompausing and seeking a more economicdesign were judged to be outweighedby the potential costs of delay anddisruption from not proceeding.

4.18 In some cases complexarchitectural requirements involvedpreviously untested buildingsolutions. In these cases thedifficulties were made worse by tighttolerances and multiple complexinterfaces between packages. It wastime consuming and expensive todevelop solutions.

4.19 A striking example of the impactof this is the foyer roof and glazingcontract (Exhibit 31 overleaf).

25 As shown in Part 2 (Exhibit 14) 15 of the 20 largest contract tenders had been authorised by December 2001, and by this time 78% by estimated value of all workhad been awarded.

Page 56: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

54

Source: Audit Scotland

Exhibit 29Holyrood – examples of time delay resulting in extra costs

Assembly windows package

In September 2003 the construction manager recommended an extension of time of 43 weeks, bringing thecompletion date to 3 September 2003.

This is only an interim award and further extensions have since been approved.

The reasons for this extension include:

• increased complexity in terms of window position and geometry

• contract to design, manufacture and install roof windows was a large addition to the original contract award

• design complexity exacerbated when the original production slot was lost due to design information takinglonger to produce

• after missing the production slot the timber window manufacturers closed down for summer recess – 4 weeks

• manufacturing changes due to the addition of structural silicone – blast design requirements amended significantly

• installation complexity and re-sequencing of works – the complexity of the installation of the steel frames hasincreased due to the tolerance issue between adjoining structures resulting in additional setting out time,installation time for the steel frames and abortive time due to the adjustments that have been required toexisting concrete openings.

Draft change orders 1 to 130 are estimated to have added £6 million to the package – this is the greatest effect onthe contractor’s ability to progress the works.

The estimated contract value at tender acceptance was £3.9 million. In early 2004 the estimated final costexcluding risk is £13.0 million, an increase of £9.1 million (233%). Further sums are at issue. For commercialreasons these are not disclosed here pending agreement of the final package accounts.

Specialist glazing

This package covers 11 areas of the site, including the glazing for the debating chamber, the public stair and thenorth lightwell. The contractor has encountered a range of problems in delivering the package that have led to thegrant of an extension of time.

The original programme for the delivery of this package was nine months. If, as now planned it completes in July2004 the contractor will have engaged on it for some 34 months. There have been 322 change instructions issuedand from an initial tender value of £7.2 million, increases of £3.8 million (53%) have been approved.

The main delays affecting this contract have been:

• the development of design

• shortage of contractor resources deployed on site

• location of elements of the package caused significant problems for the contractor due to the presence of othertrade contractors working in the same space, thus denying unrestricted access or complete access at times

• there were also problems with access to cranage facilities in work areas. These access problems contributed toout of sequence working and the prolongation experienced by the contractor.

Page 57: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 4. The reasons for increased costs 55

Exhibit 30Holyrood – estimated construction costs from 2000 to 2004

Source: Audit Scotland

Source: Audit Scotland

Exhibit 31Design development of the foyer roof and glazing package

£100

mill

lion

0

1

2

3

4

Sept

embe

r 200

0

Dec

embe

r 200

0

Mar

ch 2

001

June

200

1

Sept

embe

r 200

1

Dec

embe

r 200

1

Mar

ch 2

002

June

200

2

Sept

embe

r 200

2

Dec

embe

r 200

2

Mar

ch 2

003

June

200

3

Sept

embe

r 200

3

Dec

embe

r 200

3

Mar

ch 2

004

Total costs

Total project costs reportedto Finance Committee

Construction costs (shaded)

Construction risk

Increased constructioncosts over cost plan

Construction inflation

Construction cost plan

Late in 1999 the foyer area was introduced as a comparatively late addition to the concept design of theParliament. It would provide necessary additional space for the client and remove the need for circulation betweenthe MSP block and the assembly buildings complex to go through Queensberry House as previously planned.

The Stage D cost plan in November 2000 included £1.5 million for this part of the building. There was very littledesign information. The cost plan allowed a lump sum of £0.7 million (with no quantities or rates) plus £0.8 millionfor the glazed roof costed at £500/m2.

The work was procured using the two-stage process. At the first tender stage in June 2001 a price of some £2 million was obtained and accepted. As the requirements were clarified and design developed between thecontractor and the architect, estimated costs increased to £4.2 million, £4.9 million and then in December 2001£6.2 million. This included £1.2 million for oak and steel finishes alone, almost the whole value of the original costplan allowance.

In December 2001 the Progress Group concluded there was little scope to save costs as any redesign wouldsignificantly delay the overall programme and increase costs elsewhere. It accepted the developing designreluctantly and allowed development to continue.

Exhibit 21 illustrates how the design of the area developed between 2000 and as it has now been constructed.

The current estimated cost of the contract is some £7.4 million excluding risk. The extra costs appear to be theresult of the inability of the contractor to price the contract fully at tender with the limited design informationavailable; the degree of design development on the package post tender; and extensions of time for delays due toother packages not completing on time and cranage access.

Page 58: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

56

4.20 We examine the additionalcosts that may be attributed todesign development and the otherunderlying causes in the followingsection.

The main reasons for cost

increases since 2000 are design

development and delay in the

construction process

4.21 Despite the large scale andcomplexity of the Holyrood project,there is no record that links particularfeatures of the design developmentprocess with specific estimates ofincreased costs. There was norequirement to assign changes incost to any category as theyoccurred, to allow the underlyingreasons to be summarised andunderstood. The change controlprocess ensured the financialcommitment for each contractstayed within the overall financiallimit for that contract at any timealthough the limits for almost everycontract could and did increase.

4.22 The audit did not examineindividually the reasons for the10,000 changes that projectmanagement has approved over thecourse of the project. Neverthelessconstruction costs increases can beshown to fall into four main areas(Exhibit 32):

• Inflation. The November 2000cost plan was prepared usingconstant March 1998 prices. TheAudit Scotland consultants madea mathematical assessment ofthe impact of the subsequentincrease in market prices. Theydid so by adjusting the March1998 cost prices for inflationusing published construction pricechange indices. Inflation hasadded £19 million to theconstruction cost estimated inNovember 2000.

• Client management of the brief.The Corporate Body has notintroduced significant additionalaccommodation requirementssince 2000. The cost of allchanges requested by the clientsince the Stage D design report inJune 2000 is £0.6 million.

• Prolongation, disruption and delay.Because contractors who claimextra costs for extensions of timemust demonstrate a case fordoing so information about thesecosts is available from contractrecords. The Audit Scotlandconsultants analysed the costconsultant’s records includinginterim account statements for allpackages, current trade contractorclaims and risk review informationto assess this cost. Their estimateof the costs of prolongation,disruption and delay is some £73 million. This includes most ofthe cost consultant’s current riskestimate of £24 million.

• Design development. Becausethe total increase in costs isknown the cost of designdevelopment is the balancingfigure, £68 million. Excluded fromthis heading is £4 million, which isproject management’s estimateof the extra costs incurred as aresult of the demise of Flour Cityin 200126.

Uncompetitive procurement has

contributed to increased costs

4.23 Audit Scotland and itsconsultants’ examined theprocurement and management of asample of 20 of the Holyrood tradecontracts (Exhibit 33). They looked at:the trade contractor selection andaward process; the status of designat tender stage; the initial estimatesfor each package; and thecommercial terms of each trade

contract. They also examined costreporting, cost management andforecasting, change controlprocedures and management of thepackages in the construction phase.

4.24 There was some competition in17 of the 20 contracts. A single stagecompetitive tender followedexpressions of interest at a pre-qualification stage in most cases.Although in most cases theconstruction manager soughtreasonable sized bid lists generallyfewer bidders tendered thanplanned. Of 17 packages tenderedcompetitively only five had tendersreturned by all of the firms on the bidlist. The 20 contracts each have anestimated outturn cost of between£1 million and £40 million. Forcontracts of this value it would benormal to have at least four tenders.Thirteen of the 20 packages did notmanage to achieve this level ofcompetition (Exhibit 34 overleaf).

4.25 In the consultants’ opinion, thechallenging scope of the work interms of design content or thecomplexity of the design may havediscouraged tenderers. For exampletwo of the six tenderers for theassembly building frame contractwithdrew citing lack of resources. It is possible that the two tenderersconsidered they did not have theresources or expertise to deal withthis very large and complex package.

The focus on programme

deadlines drove procurement and

led to higher costs

4.26 As noted above 13 of the 20packages had only three or fewertenders. But perhaps a much greaterchallenge to obtainingcompetitiveness and value for moneywas encountered through the lettingof packages with large elements ofuncertainty about scope.

26 My December 2002 report on the 2001/02 audit of the Corporate Body describes the circumstances leading to the termination of Flour City’s contract in October2001 and associated contract management issues.

Page 59: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 4. The reasons for increased costs 57

Source: Audit Scotland

Exhibit 33Audit examination of 20 works contracts

Note: These figures exclude VAT

Source: Audit Scotland

Exhibit 32Main reasons for construction cost increases since September 2000

Item Cost Source

Approved construction budget in 2000 £108 million November 2000 cost plan

Add:

Inflation £19 million Calculated from published construction inflation indices

Brief development £1 million From client’s change control records

Prolongation, disruption and delay £73 million From cost consultant’s contract records

Flour City demise £4 million From cost consultant’s contract records

Design development £68 million Balancing item

Current estimated construction cost

including risk

£273 million

The sample of 20 works contracts for the Holyrood project included:

• a range of different work types: structural, cladding, roofing, joinery, fit-out, services, hard landscaping etc

• all the contractors with the largest share of work

• a mix of contracts awarded at different times

• a few contracts where final accounts had been agreed

• many higher value contracts; the estimated final cost of the 20 contracts is £153 million or 56% of the totalconstruction expenditure excluding risk

• packages from different parts of the site including the MSP building, the assembly building and QueensberryHouse.

Appendix 1 details the 20 contracts examined.

Appendix 2 summarises the detailed findings from the examination of five of these contracts as an illustration ofthe scope, nature and progress of individual contracts.

Page 60: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

58

Source: Audit Scotland

Exhibit 34Competition for 20 Holyrood works contracts

Package Tenders

requested

Tenders

received

Procurement

route

Estimated

outturn cost

Hard landscaping 6 6 Single stage £5.6m

Mechanical and plumbing east 4 4 Single stage £9.1m

MSP building roofing 5 4 Single stage £1.0m

Stone flooring 5 4 Single stage £1.8m

Assembly building frame 6 4 Single stage £39.8m

Toilet and fitness area fit-out 6 4 Single stage £2.9m

Electrical east 3 3 Single stage £8.0m

Zone 2 fit-out 5 3 Single stage £7.8m

Assembly building windows 7 3 Single stage £13.0m

MSP building windows and cladding 3 2 Single stage £2.8m

MSP building carpentry and joinery 5 2 Single stage £4.0m

Queensberry House blast doors and windows 5 2 Single stage £1.2m

Zone 1 fit-out 3 2 Single stage £14.1m

Substructure west 3 3 Two stage £3.0m

Substructure east 3 3 Two stage £7.2m

Foyer roof and glazing 7 4 Two stage £7.4m

Assembly building roofing 6 3 Two stage £7.5m

Specialist glazing 1 1 Negotiated £11.0m

MSP building bay windows 1 1 Negotiated £3.4m

Assembly building rooflights 1 1 Negotiated £1.9m

Total - - - £152.5m

Page 61: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 4. The reasons for increased costs 59

4.30 To meet the requiredprogramme packages went out totender generally at the dates requiredin the tender event schedule, even ifthe amount of design informationwas less than would normally beexpected. The further advanced thedesign is prior to tendering, thegreater level of cost certainty thatcan be achieved. But as shown inPart 3 late and incomplete supply ofdesign information was a significantfactor in many Holyrood contracts.

4.31 Based on data in theconstruction manager’s tender eventschedule there was a long periodbetween the initial tender and thesubsequent start on site in 11contracts examined (Exhibit 35overleaf). The time between approvalof the tender recommendation andthe subsequent start on site stronglysuggests that the design was notsufficiently detailed at tender stageand that design input was requiredfrom the trade contractor. It wasalmost as if, once a package hadbeen awarded, it then had to wait forthe design to catch up.

4.32 Tendering work with anuncertain scope adds to overall costbecause it reduces the level ofcompetitiveness in achievingeventual price certainty. The cost risklies with the client and not the tradecontractor. The trade contractor is ina very strong position to set the finalcost of the provisional, undefineditems. Costs are subject to reviewand approval by the constructionmanager, the cost consultant andproject management. But the scopefor controlling costs for theseelements of the work is inevitablycompromised in the absence of acompetition. If there wasdisagreement with a supplier aboutcosts, the client could in theoryterminate the negotiation and seekanother supplier. But programmepressures meant this was not anattractive option.

4.33 The trade contractor can also,through pricing levels, ensure thatany risks on costs taken to win thecontract can be absorbed oreliminated in the subsequent designdevelopment. The trade contractordoes not need to go out to a range ofmaterial suppliers or sub-contractorsto achieve favourable prices within itscost allowances, as there is nocompetition.

4.34 The inevitable consequence ofletting packages before the designhad been fully developed wastherefore a loss of truecompetitiveness. Instead of securingas large a proportion of the overallcosts as possible on a fixed pricebasis, design uncertainty led to largeprovisional items, which led to amuch greater degree of negotiationthan was desirable. In theconsultant’s opinion seven of the 20trade contracts reviewed weretendered when design was not as faradvanced as they would normallyexpect. Whilst the rationale forawarding these packages in this waywas driven by programme, it had aninevitable impact on cost (Exhibit 36overleaf).

4.35 On 18 of the 20 packagesreviewed trade contractors had somedesign responsibility. On four ofthese packages the trade contractorswere required to finalise thedevelopment of the concept inaddition to the detailed design,subject to final design team approval.While specialist trade contractordesign input was necessary todevelop certain elements of thedesign this input ideally should havebeen separated from theconstruction works, which shouldhave been separately competitivelytendered once the design reachedthe required stage. This did nothappen at all in the Holyrood project.

4.27 The architects had acceptedresponsibility for completing allaspects of the design. But from 2000project management, on theconstruction manager’srecommendation, increasinglyadopted trade contractor design forsome key works packages. Underthis approach the architects remainedresponsible for most of the conceptdesign, but trade contractors tookgreater responsibility for developingthe detailed design informationnecessary for construction.

4.28 This provided the opportunity toenhance the ‘buildability’ of thedetailed design and speed upcompletion of the work. Tradecontractors brought their owndesigners in to supplement thearchitect’s resources. They couldprovide design solutions to oftencomplex requirements that theywere more confident they couldbuild. But this approach carried ahidden extra cost. The mixture of thecommercial approach adopted – suchas the use of two stage tenders andprovisional sum allowances – anduncertainty about the scope of thework had the result that for a greatpart of the work the final price wasset by negotiation with a singlecontractor not by a competition(although the contractor had usuallybeen selected by competition at anearlier stage).

4.29 To meet the programme theconstruction manager developed atender event schedule to drive alongthe process of trade contractorselection and award. The scheduleset out key milestones in the overallprocurement process of each tradepackage, from pre-qualification stageto agreement of bid lists, preparationof tender enquiry documents,obtaining tenders, tenderrecommendations, contract awardsand start on site.

Page 62: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

60

Exhibit 36Examples of contracts with significant design uncertainty at contract stage

Source: Audit Scotland

Source: Audit Scotland

Exhibit 35Tender recommendation to start on-site for 11 Holyrood contracts

Package Client approves

tender

recommendation

Start on site

date

Period from

approval to

start on-site

Assembly building frame 23 January 2001 12 March 2001 2 months

Foyer roof and glazing 1 March 2001 22 July 2002 41/2 months

Assembly building windows 2 December 2001 15 April 2002 41/2 months

MSP building bay windows 21 December 2001 27 March 2002 3 months

MSP building windows and cladding 21 March 2002 2 September 2002 61/2 months

MSP building roofing 30 January 2001 25 June 2002 5 months

Assembly building rooflights 6 August 2002 17 March 2002 7 months

Stone flooring 8 February 2002 24 June 2002 41/2 months

MSP building carpentry and joinery 4 April 2001 16 July 2001 3 months

Queensberry House blast doors and windows 23 September 2001 19 February 2002 5 months

Electrical east 25 September 2001 10 December 2001 21/2 months

On the assembly building frame, out of a contract sum of £17.9 million there were provisional sums of £3 millionand £0.4 million for the structural steel work and the glulam beams respectively. These were two major areas ofrisk, where the amounts allowed in the contract proved to be completely inadequate, due to the complexity of thedesign when it was finally developed, well after the trade contractor had been appointed. The current estimatedoutturn cost of this contract is £39.8 million, excluding risk.

On the site-wide package for fitting out toilets and fitness areas, the scope of work to the MSP building andQueensberry House was reasonably well defined, although the works to Queensberry House were remeasurable.However the drawings for the assembly buildings were at a scale of 1:250, which would not have been acceptablefor obtaining a fixed price. The provisional, remeasurable elements of this package totalled £0.9 million out of atotal contract sum of £1.5 million. The current estimated outturn cost of this contract is £2.9 million, excluding risk.

The site-wide contract for stone flooring was let on the same basis of design noted above. In this case theprovisional, remeasurable elements of this contract totalled £1.1 million out of a total contract sum of £1.4 million.The current estimated outturn cost of this contract is £2.5 million, excluding risk.

Page 63: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 4. The reasons for increased costs 61

4.36 After construction started theconstruction manager held regulardesign meetings, cost meetings andprogress meetings for each package.However, the effectiveness ofmanagement and control has almostcertainly been compromised by thestrict requirement to adhere toprogramme targets that could be saidto have been unachievable. Drivingtrade contractors to comply withunachievable programmes increasescosts, in terms of prolongation costsand also in disruption costs.

4.37 Eight of the 20 packagesreviewed were awarded on the basiseither that the outstanding designwould be completed within anagreed period after award or simplyas quickly as possible. In thesecases, for coordination, theconstruction manger ensured that aschedule was agreed that showedthe latest dates that the tradecontractor needed design informationfrom the design team. Achievementof the design programme slippedfrom the agreed dates and thisdesign slippage within individualpackages contributed to the overalldelay, as a delay on one package canaffect other packages, and so on.

4.38 Even where the performance oftrade contractors may not have beensatisfactory, there was in somecases little opportunity to attributedelays to them, because of largermovements in the overallprogramme. For example, on theassembly windows package thesatisfactory performance of the tradecontractor was questionedthroughout. The constructionmanager regularly wrote to the tradecontractor to record dissatisfactionwith progress of the works.

However, the construction managerhad to recommend a 42 weekextension of time for the tradecontractor when a new strategicprogramme was issued whichextended the time period forcompletion by 42 weeks.

Page 64: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

5.3 The estimated cost has increasedby some £220 million over the lastfour years. It has therefore beenimportant to examine whathappened, but answering thisquestion should take some accountof the quality of the building.Although the audit has not attemptedto evaluate the quality achieved,project management’s view is thatthe end result is likely to satisfy fullythe high quality standards in the userbrief of 1998 (Exhibit 38 overleaf).

Throughout the project there was

tension between the objectives of

time, quality and cost

5.4 In 1998 the client required thatthe building should be completed bysummer 2001. Time was a priority.Quality has been equally importantthroughout the project. The user briefdemonstrates that the objective wasto provide a high quality building ofwhich the Scottish people could beproud.

5.5 Construction management wasseen to be the only method of procuringthe Holyrood building that coulddeliver high quality within the deadlines.

5.6 The client also set a budget atthe outset. But there were conflictingmessages about how important costwas compared with time and qualityconsiderations.

• Construction management, thechosen procurement method, isnot very well suited to any projectwith a fixed cost constraint.

• The client considered that therewas a fixed budget for the projectfrom the start, but it did notcommunicate clearly what thebudget was27.

• In the early stages, full estimatedcosts were not reported to theParliament and this did nothappen in a systematic way until 2000.

• Parliament set a fixed budget in2000 but by 2001 the clientregarded it as no longerachievable. Subsequently, theclient reported successive costincreases to Parliament and didnot seek to set any new financialtarget or limit.

5.1 In this part I assess themanagement and control of theHolyrood project over the four yearssince my report of September 2000.

5.2 In that 2000 report I examinedthe strengths and weaknesses ofproject management and compliancewith good practice within the publicsector. I concluded that there shouldhave been greater recognition givento the importance of managing riskand that accounting for risk wasinsufficient. I said that the reportingof estimated costs was unsystematicand incomplete. I questionedwhether project management hadthe right mix of skills. I suggestedthat there should have beenincentives for the consultants toavoid cost increases and ensuredelivery on time. I alsorecommended improvements toproject management andgovernance. In evidence to the AuditCommittee in October 2000 theAccountable Officer indicatedacceptance of theserecommendations, which he statedwere being implemented (Exhibit 37).

62

Part 5. Project management and control

27 The client stated in 1998 that the budget was £50 million. This excluded costs such as professional fees, landscaping and irrecoverable VAT then estimated at £40 million.

Page 65: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 5. Project management and control 63

Source: AGS report September 2000 paragraph 27; evidence to the Audit Committee October 2000

Exhibit 37AGS recommendations on project management September 2000

AGS recommendation, September 2000 Accountable Officer response, October 2000

Now that the design of the building is firm, project managementshould pass more responsibility to Bovis as the constructionmanager and avoid the danger of duplicating the services thatBovis are commissioned to provide.

Done, as part of routine management

A single authoritative point of contact between the client andproject management must be confirmed. Similarly all instructionsto the construction manager and the design team on the client’sbehalf should come only from a single authoritative point withinproject management.

Project Director, formerly Project Sponsor

Project management have a key role to oversee and monitordelivery of the project and represent the client’s requirements anddecisions. Project management should identify and agree majormilestones or targets for the remaining project period for thepurposes of reporting and monitoring progress, both with theclient and with the design and construction teams.

Accepted, currently in progress

Project management should review and report project costsregularly (possibly monthly) to the client on a comprehensive andsystematic basis. Estimates should include all relevant costs ie,including construction (works package) costs, construction riskallowance, consultants fees, construction manager fees andcosts, furniture and fit out costs for the new building, any non-construction risk allowance that may be necessary, and VAT. Thereshould be a succinct commentary which draws attention tovariances since the last report and provides explanation whereverpossible. The team’s report should include the costs associatedwith the project that will be met by other public bodies.

Accepted and already implemented

In September 2000 the Clerk of the Parliament advised theCorporate Body of the results of his wider review of governancearrangements for the organisation as a whole. The Clerk and theCorporate Body should consider whether there is any need infuture for independent advice and reporting on the Holyroodproject.

Holyrood Progress Group already established.Outsourced internal audit is expected toprovide assistance on governance.

Page 66: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

64

The Scottish Parliament building presents the design team with a unique opportunity to make a significantcontribution to the design of a building which marks a very significant milestone in Scotland’s political history. (The aim should be to) … reflect the aspirations of the Government and the people of Scotland for this building.

The design should respect its historic surroundings, … at the same time be a building which reflects the prevailingmood at the end of the century and the millennium. It will be the first landmark building of the new millennium. Itshould have a resonance of quality, durability and civic importance of which the Scottish people can be proud.

The views to and from the building … … must play a role in integrating the building into the wider context …

There is an opportunity to … … produce an environment in which there is considerably less intrusion from throughtraffic.

The building must be accessible to all …

The design of the building must reflect the opposing requirements of openness and security …

The Scottish Office is committed to promoting good environmental practice in terms of building design, constructionand management. … … It is assumed the Parliament will share the Government’s approach in this area.

The Parliament is a living changing organisation … the building must be flexible and have the capacity to acceptchanges in organisation, space requirement and management.

A budget has been set for the building of the project to £50 million at March 1998 levels. This should enable thedesigner to reflect the requirements of the brief and also to provide quality in construction and design, and securevalue for money.

The design must take account of the latest advances in technology within the budget constraints … …… … People must be able to see and meet their representatives within the building … …

It is the aim of the Secretary of State to have the building completed and ready for occupation by the Parliament inthe year 2001. … …

This project presents an opportunity for the design team to produce a landmark building reflecting the aspirations ofScotland as a nation, with a building of quality and value.

Source: Scottish Parliament building user brief, October 1998

Exhibit 38Extracts from Scottish Parliament building user brief

Page 67: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 5. Project management and control

of skills for this demanding project.But it noted that the Corporate Bodyhad adopted the recommendationsof the Spenceley report and othermeasures intended to strengthen themanagement and oversight of theproject. Exhibit 41 (overleaf)summarises the project organisationas it has been since June 2000,when the Holyrood Progress Groupwas established.

5.10 The successful management ofa project normally requires that thereshould be single point of accountabilityand control where decisions can betaken about how to balance time,cost and quality as part of the clientdecision-making process. Normallythis control should reside with theproject sponsor/director.

5.11 Overall leadership for a projectwould normally reside with theproject director to whom the clientgives responsibility – within specifiedboundaries – for making the projecthappen.

5.12 The organisation of theHolyrood project did not provide the

necessary clear direction andleadership. The organisation chart inExhibit 41 (overleaf) suggests clearlines of responsibility, control,communication and accountability. Butthe real position was much less clear:

• The Corporate Body is the legalclient and in June 2000 formallydelegated responsibility forcompleting the Holyrood projectto the Accountable Officer.

• The Accountable Officer hasdelegated responsibility for day-to-day control and oversight of theproject to the project director withadvice and guidance from theProgress Group.

• The project director’s jobdescription states that the personacts as the single focal point forday-to-day management of theParliament’s interest in theproject, with responsibility forsecuring the delivery of the newbuilding complex to programme,within budget and to the specifiedquality. In practice the projectdirector has acted as the senior

5.7 How well the tensions of time,quality and cost are managed has asignificant impact for any project.Exhibit 39 illustrates the intendedpriorities with regard to time, qualityand cost of the Holyrood project andwhat actually happened.

There were problems in ensuring

clear leadership and control on the

part of the client organisation

5.8 Exhibit 40 (overleaf) illustrates atraditional model of good projectteam organisation, command andcommunication. This was the modeladvocated by HM Treasury in 1997.The model remains relevant, althoughin 2003 the Office of GovernmentCommerce recommended clientsshould follow procurement methodswhich provide an integrated supplyteam not separate agreements withindividual consultants, contractorsand specialist suppliers28.

5.9 The original project managementorganisation broadly reflected thistraditional model. My 2000 reportasked whether project managementhad always had the best possible mix

65

Source: Audit Scotland

Exhibit 39Achievement of priorities for the Holyrood project

Priority Intended Actual

Time First priority Failed – significant slippage

Cost Fixed budget Failed – significant increase

Quality High quality required Achieved

28 Achieving Excellence in Construction Procurement Guide 2 – Project Organisation (OGC 2003) states that "A traditional project structure is not integrated; it separatesout the responsibilities of each party. For this reason it should not be followed unless it demonstrates significantly better value for money than the recommendedprocurement routes".

Page 68: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

66

Source: Essential Requirements for Construction Procurement, HM Treasury, 1997

Exhibit 40Traditional project organisation

Client Adviser

(may be required for non-technical sponsor;

generally externalconsultant)

Project Board

User Panel

(including functional andoperational stakeholders

Investment

Decision Maker

Project Owner

Project Sponsor

Project Manager

(generally external consultant)

ContractorSuppliersConsultants

(designers, costconsultants, planning

supervisor)

Page 69: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 5. Project management and control 67

Source: Audit Scotland

Exhibit 41Holyrood project organisation from June 2000

Design Team

Architects

EMBT/RMJM

Service Engineer

RMJM Scotland Ltd

Works Package

Contractors

Structual Engineer

Ove Arup

Holyrood

Progress

Group

Construction Manager

Bovis Lend Lease(Scotland) Ltd

Client

Scottish ParliamentaryCorporate Body

Project Management

Clerk of Scottish Parliament

Project owner and PrincipalAccountable Officer

Holyrood Project Team

Headed by the ProjectDirector/Sponsor

Cost Consultant

Davis Langdon & Everest

Page 70: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

68

The Accountable Officer has notreferred any matter of disagreementto the Corporate Body.

• The construction managermanaged the constructionprogramme but did not havecontrol over costs or quality. Theconstruction manager is aconsultant and therefore cannotinstruct third parties such as thedesign team or the tradecontractors (except on behalf ofthe client), because it has nodirect contractual relationship withthem. The construction manageradvised project management andthe client on specific actions toprotect programme and cost butits advice was not always actedupon. For example, in October2000 it asked project managementto obtain a written confirmationfrom the architects that it wouldcarry out any redesign required toachieve the target rates andbudgets set for the tradepackages in the then draft costplan. Project management did notaction this request.

• Similarly the role of the costconsultant is to monitor costs and

to provide specialist advice but asa consultant it cannot instructother parties regarding costs.

• Design team management is acritical function in a large buildingproject (Exhibit 43). The architectis responsible for leading thedesign team but projectmanagement and the ProgressGroup have been dissatisfied withthe architect’s performance in thisrespect. At times leadership ofdesign has not been clear or therehas been conflict about the role.Enric Miralles died in July 2000yet it was 15 months, in October2001, before project managementinsisted on organisational changeswithin the architectural jointventure to clarify roles andresponsibilities.

5.13 The project execution plan is avital control procedure in any majorconstruction project (Exhibit 44overleaf). In my 2000 report I saidthere should have been such a plan.Subsequently the roles,responsibilities and accountabilities ofthe client, the project team includingthe consultants were set out in adraft project execution plan in

project administrator coordinatingthe interests of the various partiesand the high level communicationand reporting. There has beenless emphasis on leading andmaking decisions because theProgress Group must guide thedirector.

• The senior project manager/project manager supports theproject director. The currentsenior project manager instructsthe construction manager and thedesign team but is not a singleauthoritative point of command.He was appointed in 2001 buthas not had a formal jobdescription since taking up the role.

• The Holyrood Progress Group isan advisory and monitoring groupsupporting the Corporate Bodyand the Accountable Officer. Inpractice it quickly evolved into adecision-making body (Exhibit 42).While strictly the Progress Grouphad no executive power its termsof reference required theAccountable Officer to act on itsadvice or, if he disagreed, to referthe matter to the Corporate Body.

Exhibit 42Role of the Holyrood Progress Group

Formally the Holyrood Progress Group is an advisory and monitoring group supporting the Corporate Body and theAccountable Officer.

In practice it quickly evolved into a decision-making body. For example:

• in August 2000, following advice from the architect, the Group concluded that oak should be used in window frames

• in March 2001, the Group rejected cost saving proposals from the architects and recommended the use ofKemnay as opposed to Portuguese granite

• in January 2002, the Group concluded that granite rather than slate should be used for cladding materials.

The Progress Group demonstrated leadership on many important issues. For example:

• the Group instructed the cost consultant to seek opportunities for savings in the MSP block in November 2000

• in April 2001, the Group reported its ‘serious concerns’ regarding the performance of the design team to theAccountable Officer. A letter was sent in September 2001, clarifying the client’s expectations of the architects

• in June 2001, the Group informed the Accountable Officer of its concerns over the performance of the thenproject director.

Because the Progress Group is not formally a decision-making body it is not required to be accountable to any other body.

Page 71: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 5. Project management and control

2001, however, the Parliamentapproved a further motion, which canbe interpreted as removing theprevious overall budget constraint of£195 million. At the same time theParliament’s motion called forrigorous risk management (Exhibit 11).

5.16 Once Parliament had decided toremove the overall cash limit of £195million, project management did notestablish an alternative overall budgetor approved cost ceiling that wouldallow the costs of the project to beproperly managed. I am concernedthat project management did notestablish an alternative overall budgetor approved cost ceiling for the purposesof managing the costs of the project.Because project management set nobudget after the Parliament set asideits £195 million limit, it is not clearhow considerations of affordabilitywere taken into account.

5.17 I have noted above theimportance of – and inevitabletension between – the three prioritiesof time, cost and quality. Managedcorrectly, with an understanding ofthe relative weighting or importanceattached to each by the client, asuccessful outcome can be achieved.

By not setting a project budget after2001 there was a significant risk thatproject management would ‘losesight’ of the aim to contain costs andwould not have a budget limit as aperformance target. I am concernedthat project management and theclient did not use budgetary controlprocedures to allow proper financialconsiderations to be given dueweight in decision-making, togetherwith those regarding programmedeadlines and the quality, complexityand aesthetics of the building.

5.18 In the annual budget of theCorporate Body the provision forcapital expenditure has substantiallyexceeded actual expenditure everyyear (Exhibit 45 overleaf).

5.19 Exhibit 46 (overleaf) shows theunit construction costs of theHolyrood project compared to asample of other high quality buildingscompleted in recent years for amixture of public and private sectorclients. There has been a dramaticincrease in unit costs of the Holyroodproject since 2000, which is muchmore expensive than most recenthigh quality buildings.

October 2000. The draft plan was adetailed, comprehensive documentsetting out clearly policies, strategies,lines of communication, keyinterfaces and responsibilities. Butalthough there were significantchanges in project organisation sinceOctober 2000 the plan has neverbeen updated nor issued beyond adraft document, illustrating theunclear definition of roles.

After June 2001 project

management did not work within

an overall budget or approved

cost ceiling for the whole project

5.14 If there is an approved budgetthen project management mustrespond to cost projections thatexceed the budget by taking steps toreduce the cost or by securingadditional funds. The choice shouldbe based on a reassessment of valuefor money at the new projected costlevel.

5.15 In April 2000 the Parliamentapproved a budget for the project of£195 million. At the time projectmanagement considered this was acash limit within which the contractcosts must be contained. In June

69

Source: Audit Scotland

Exhibit 43Design team management

The Holyrood project design team comprised the architect (EMBT/RMJM), the structural engineer (Arup), theservices engineer (RMJM Scotland), the quantity surveyor (DLE) and various specialist consultancies. The architectwas designated as the design team leader. Its responsibilities include co-ordination of the work and activities ofdesign team members.

The design team needs to work together, in an integrated way, to produce the overall design solution of the building.This is an iterative process as, for example, the architecture may need to alter to suit the structural solution; similarlythe structure needs to accommodate the services engineering. Bringing all of these disciplines together to producethe ‘best’ solution for the client is an art. It requires an understanding of the various disciplines, an appreciation of thedesign processes and a reliance on effective relationships. Making all of the design happen and delivering the designinformation to the required programme is called design management. The leader of the design team needs to ‘own’the overall vision of the design and harmonise the various strands of design towards the realisation of that vision in atimely manner.

It is not unusual for design leadership to be placed with the architect. Indeed, traditionally the architect was alwaysthe design leader. It is only over the last 20 years, with the advent of fast-track forms of design, procurement andconstruction, that project managers and construction managers have entered the scene and that design managementhas taken on a higher level of criticality. Design management becomes particularly important when design periodsare ‘squeezed’ or shortened, as is frequently the case with fast-track approaches such as construction management.

It is important under construction management for a very close partnership to be developed between the architectand the construction manager, so as to ensure that the design programme and the procurement programme arecompatible and that the works are not delayed due to the progress of the design.

Page 72: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

70

Exhibit 45Corporate Body capital budget and expenditure 1999 to 2004

Source: Audit Scotland

Exhibit 44Project execution plans

The project execution plan should set out how the client intends the project to be managed, the roles andresponsibilities and any delegations of authority. The plan should be a controlled document, updated as necessarythroughout the life of the project. Good practice is that the project owner should sign off the plan before theappointment of any consultants or contractors. It should name individuals for each role and delegated authority,rather than merely state the post responsible. Accountability is far more effective where individuals responsible are named. The plan must be kept up-to-date, when individual post-holders change the plan should be updated and re-issued.

Note: Capital budget includes an element of capital spend on non-Holyrood projects

Source: Corporate Body annual financial statements

Ann

ual e

xpen

ditu

re (£

mill

ion)

0

20

40

60

80

100

120

140

160

180

Capital budget

Actual spend

Financial year

2003/041999/2000 2000/01 2001/02 2002/03

2417

63

26

84

40

136

100

178

133

Page 73: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 5. Project management and control

provided between June 2001 andJune 2003 reported all the othermain cost items. There was no otherregular reporting of total project costuntil July 2003 when the CorporateBody started routine monthlyreporting to the Finance Committeeon total project costs.

5.22 The cost consultant reportedestimated construction costs in detailevery fortnight to projectmanagement and the ProgressGroup. The scope and format ofthese reports varied over time, andthe Progress Group has stated that itfound the reporting formatunsatisfactory at different times.

5.23 The cost consultant considersproject management did not inform itof any critical opinion expressed bythe Progress Group at the time. Thecost consultant took it as read thatproject management understood thecost issues it had reported and wassatisfied with the clarity andcompetency of its cost advice.

5.24 The cost consultant wasresponsible for monitoringconstruction costs only and providing

risk analysis. It was not responsiblefor monitoring other costs such asfees, VAT and some of thelandscaping and fit-out costs.Although the cost consultant alsoprovided an overall estimate ofproject costs to project managementthis did not accurately measure feesand VAT, which the cost consultantwas not responsible for monitoring.In any event project management didnot report these overall estimates tothe Progress Group.

5.25 Although the cost consultant’sreports highlighted variances inconstruction costs when theyoccurred, there was little narrativeand no written analysis of reasons forchange. Until July 2003, when theCorporate Body introduced regularmonthly reporting on progress to theParliament’s Finance Committee,project management providedwritten reports on the costs of theproject to the Progress Group andthe Corporate Body on an exceptionbasis only, which was relatively rare.

5.26 There was no regular or detailedreporting of the significantexpenditure on the consultants’ fees

Cost reporting and financial

control should have been better

developed at all levels of the

Holyrood project

5.20 In my 2000 report Irecommended that projectmanagement should review andreport project costs regularly(monthly) to the client on acomprehensive and systematic basis.The Accountable Officer advised theAudit Committee in October 2000that he had implemented thisrecommendation. But subsequentfinancial reporting of the project hasnot always been comprehensive orsystematic.

5.21 Project management providedestimates of the overall project costto the Progress Group and theCorporate Body, in the form of draftsof the reports that the CorporateBody subsequently provided to theParliament’s Finance Committeefrom June 2001. Until July 2003 thereports were provided only quarterlyor less often (Exhibit 5). None of thereports before July 2003 providedinformation about landscaping costsand only four of the eight reports

71

Exhibit 46Comparative unit construction costs for high quality buildings

Note*: Where construction cost is not available the total project cost is reported. Unit costs have been updated to current outturn costs, using the BCIS index.Source: Audit Scotland

Estimated Gross floor area Unit cost

construction or total project*

cost

Holyrood – current position £273m 30,600m2 £8,922/m2

Holyrood – previously (Stage D £119m 30,600m2 £3,889/m2

June 2000)

Portcullis House £161m 22,811m2 £8,600/m2

New British Library Phase 1 £511m* 112,650m2 £6,600/m2

City of London financial centre – £180m* 67,350m2 £3,100/m2

private sector client

Large multinational HQ – £220m 87,680m2 £2,600/m2

private sector client

Greater London Authority building £43m* 18,000m2 £2,600/m2

London insurance HQ – £185m* 79,000m2 £2,300/m2

private sector client

Page 74: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

72

etc for the Holyrood project (currentestimated outturn cost £68 million).Project management did not resolveimportant questions about the basisof each consultant’s feeremuneration until August 2003comparatively late in the project. Iexamine this issue later in this part ofmy report.

5.27 Until July 2003, reporting by theAccountable Officer to the CorporateBody usually focussed on the valueof current construction contractcommitments and recent awards.Because no budget for the overallcost of the project had been setthere was no routine report orcommentary on overall project costsand trends or the constituentelements, relative to what would bein any budget.

5.28 In any project it is important that all the parties have a clearunderstanding of how budget, riskand contingency relate to each other:

• A budget is an approved sumallocated for a project (a pot ofmoney). Only the authority thatapproves a budget can vary itonce set. A budget is not thesame as a forecast, which willvary through the life of the projectas circumstances unfold.

• A risk is an event that may occurand may have an impact on theoutcome of the project.

• A contingency is an allowancewithin the budget for unidentifiedcosts, including risks. Similarly aforecast is likely to include acontingency.

5.29 At different times projectmanagement and the ProgressGroup acted to contain or reducecosts. However the normal financialdiscipline of named individuals being

accountable for controllingexpenditure within limits specified inan approved budget was not presenton this project. It seems that projectmanagement regarded the regularreports from the cost consultant onconstruction costs and on risk costsas setting a construction budget,when they were no more thanforecasts. An example of this is thatbefore payment project managementchecked each trade contractor’sinvoice to ensure it did not raise totalpayments above the total estimatedcost for the package reported by thecost consultant. The danger ofconfusing forecasts with budgets isthat forecasts will become self-fulfilling if effective action is not takento contain the cost.

5.30 Exhibit 47 illustrates how theestimated construction costsincluding risk reported to theProgress Group and the Parliamenthave steadily increased since 2000.Generally estimated constructionexpenditure and construction riskcosts reported to the Parliament hasinitially been lower but has then risento the same level as the costconsultant’s forecast.

The approach to risk management

was not fully consistent with good

practice

5.31 In my 2000 report I concludedthat accounting for risk wasinsufficient. I showed that contrary togood practice there was noquantified allowance for the majorrisks facing the project. Irecommended that this should beestablished and the results used as abasis for an action plan to managethe risks.

5.32 In Part 2 of this report I showthat project management introduceda process for quantifying risks fromOctober 2000 and has conducted riskreviews since then. However in

some important respects riskmanagement for the Holyroodproject has not been consistent withgood practice. Under the draft projectexecution plan prepared in October2000 it was project management’sresponsibility to arrange and convenerisk management meetings atrelevant intervals to review progresson all aspects of risk as they affectedthe project.

5.33 Twelve risk workshops did takeplace, between October 2000 andDecember 2002. In accordance withgood practice representatives fromproject management, the designteam and the cost consultantsparticipated. An experienced riskmanagement specialist employed bythe cost consultant led these events,in accordance with the costconsultant’s contractual responsibilityto undertake and lead riskassessments as part of its service.

5.34 A fundamental feature of riskmanagement is to allocate or assigneach risk to whichever party is bestplaced to manage it. Usually a namedindividual or party (an owner) isassigned to mitigate or manage outthe risk, with the aim to lead actionto eliminate, or at least reduce theimpact of, specific risks and thereforeminimise expenditure on‘contingency’ or ‘risk allowance’.

5.35 Although the risk workshops forthe Holyrood project did identifyowners for specified risks there wasno monitoring or feedback onsubsequent action. It is not clear thatthe workshops from 2000 toDecember 2002 succeeded in theaim to reduce or manage risks out.

5.36 From December 2002 riskmanagement operated differently.There were no further workshopswith the previous wide participation.There was thereafter no systematicbasis for any action by project

Page 75: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 5. Project management and control

5.37 The approach used at Holyroodappears to be an unusual way tomanage risks. It seems the approachhas been to tacitly accept increasesto the cost of the project rather thanforcing action to prevent the increases.

Capping has controlled

consultants fees but only late in

construction

5.38 In my 2000 report I suggestedthat before they appointedconsultants, project managementcould have explored more carefullyalternative fee arrangements with itsconsultants including financialincentives linked to delivering valuefor money.

5.39 Project management hadappointed its main consultantadvisers on broadly similar fee termsin 1998 and 1999. Each consultant’sfee remuneration was wholly ormainly a percentage value of theapproved construction cost of theproject. The tender proposal that thesuccessful firms had made duringthe process leading to theirappointment set the percentage fee

that applies to them. In 2000 theestimated fee cost was £23 million,approximately 19% of the estimatedconstruction cost includingcontingency of £119 million at thattime. The client secured (partial) feecapping in 2000 and (morecompletely) 2003 and fee costs arenow forecast at £50 million, 19% ofthe approved construction cost29.

5.40 Percentage fees do not align theobjectives of the client with thecommercial objectives of theconsulting firms because the more aproject costs the more eachconsultant is paid. Percentage feesare necessary when the scope andoverall cost of a project has a highdegree of uncertainty, but it isadvantageous to a client to fix thefees as soon as it can.

5.41 In August 2000 projectmanagement agreed a professionalfee cap with their structural engineerconsultants, based on £108 millionconstruction cost. The fees to thestructural engineers have sinceincreased because of additional workbut are less than 10% of the total

management to manage out risk,although the cost consultantcontinued to report its assessment ofthe cost of risk in accordance with itsterms of engagement. In its reportsthe cost consultant assigned a costto each risk and reported the totalcost, which it called a reserve foroutstanding construction items. Thisreserve was reported in parallel withthe cost consultant’s report on thecurrent construction commitmentand project management generallyaccepted the combined results of thetwo reports as the best estimate ofthe total eventual construction cost.Over time, as individual risks havematerialised, the cost consultant hasreported a reduction in the total riskreserve offset by a correspondingincrease in the cost of the currentconstruction commitment. Howeverfor most of the time the costconsultant has also identified andcosted new risks, which increasedthe risk reserve – it was ‘topped up’,as it were. The net result was thatbetween 2000 and early 2004 therewas a steady increase in the overallforecast construction cost.

73

Exhibit 47Estimated construction costs for the Holyrood project 2000-04

Source: Audit Scotland

29 In addition to its fee the construction manager is also reimbursed for its staff costs and for the costs of site organisation. Its fee and its staff costs were convertible toa lump sum on agreement of the cost plan. The fee costs stated here therefore include the construction manager’s fee and staff costs but not its site organisationcosts. The estimated site organisation costs have increased from £5 million in 2000 to £18 million currently.

Cost consultant’scost reports to theProgress Group

Corporate Bodyreport to FinanceCommittee

100

120

140

160

180

200

220

240

260

280

Estim

ated

con

stru

ctio

n an

d co

nstr

uctio

nris

k co

sts

(£ m

illio

n)

June

200

0

Sept

embe

r 200

0

Dec

embe

r 200

0

Mar

ch 2

001

June

200

1

Sept

embe

r 200

1

Dec

embe

r 200

1

Mar

ch 2

002

June

200

2

Sept

embe

r 200

2

Dec

embe

r 200

2

Mar

ch 2

003

June

200

3

Sept

embe

r 200

3

Dec

embe

r 200

3

Mar

ch 2

004

Page 76: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

74

fees expenditure of £50 million.However project management didnot agree a professional fee capbased on construction costs with anyother consultants until 2003, whenconstruction costs were significantlyhigher than in 2000.

5.42 Prior to fee capping, each of theconsultants’ fee remuneration waslinked partly to a calculation of theapproved construction cost ratherthan the eventual outturn cost. Insummary:

• The construction manager feewas a percentage of project costwhich, together with itsreimbursable staff costs, wasconvertible to a fixed lump sumon agreement of the cost plan.

• The fee for each design teammember was individually andseparately agreed, but in eachcase as a percentage of costs.Around 60% of the total fee wasbased on the approved cost atintermediate stages beforeconstruction commenced.

5.43 To calculate the percentage feepayable to each consultant at eachstage it was necessary to agree anamount that was the approvedproject cost prevailing during thatstage. As each stage was reachedproject management should havecertified or formally stated theapproved project cost. They did notdo so.

5.44 Between 2000 and 2003 projectmanagement pursued variousinitiatives to cap the architect’s fees:

• In March 2002 the architect wasclaiming fees based on £140million construction costs.Although the architect was thenwilling to consider converting to acapped fee no agreement wasachieved. As the estimated finalconstruction cost continued to

rise this prospect became lesslikely.

• Later during 2002 and 2003 therewas disagreement betweenproject management and thearchitects as to which levels ofconstruction costs should beapplied to the measurement of itsfees at each stage. No agreementwas reached.

• In July 2003 the Corporate Body’sinitiative seeking to cap fees withall its consultants overtook thisdispute. This resulted in anagreement with the architect inSeptember 2003 to a fixed lumpsum fee for its services. Howeverthe necessary contract variationdocuments to formalise thischange had not been agreed atthe time of preparing this report.

5.45 Fee caps have been agreed forthe services of the other consultants:

• A fee cap was agreed with theservice engineers (RMJMServices) in July 2003 andformalised in February 2004.

• A fee cap was agreed with thecost consultant in August 2003and formalised in January 2004.

• A cap for the constructionmanager’s site management staffcosts and for its constructionmanagement fee was agreed inAugust 2003 and formalised inJanuary 2004. These caps arequalified and the constructionmanager is entitled to seek orclaim additional fees or costs inspecified circumstances.

5.46 The Corporate Body has nowlimited its exposure to increases inconsultants’ fees. However it did sovery late in the programme and afterthe fees had increased significantly.The agreement of the fee caps atthis late stage in the project did not

provide an effective incentive toconsultants to control costs andprogramme. Even now, thereremains a risk of increases of fees tothe architects (as the cap agreementhas not been finalised) and of fees tothe construction manager (due to thequalified nature of the agreed capwith them).

5.47 I am concerned that projectmanagement did not seek to convertits construction manager’s fee to afixed lump sum before July 2003.This opportunity was part of thecontract with the constructionmanager and there seems no goodreason why project management didnot pursue this opportunity to containcosts as part of the development ofthe cost plan in autumn 2000.

5.48 I am also concerned thatalthough project management raisedsome significant questions aboutsome aspects of some of itsconsultants’ work, it has notsystematically assessed theirperformance. This is importantbecause if project management wasable to show significantunderperformance by any of itsconsultants it should also considerwhat options (if any) it may have forrecovering some of its additionalcosts. There is a risk that any inactionby project management so far couldlimit any recourse the CorporateBody may otherwise have had.

Project management did not

implement construction

management fully in accordance

with usual practice

5.49 Parts 3 and 4 of this reportdiscuss the reasons for the selectionof construction management forprocuring the Holyrood project.Construction management canprovide important benefits to a clientbut there are particular time and costrisks flowing from it, which must beaddressed. My 2000 report showed

Page 77: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 5. Project management and control 75

that the original procurementstrategy for the new Parliament wasincomplete because it did not assessthe risks implicit in the chosenprocurement route (designerappointment and subsequentconstruction management) or howbest to manage those risks.

5.50 Construction management is acomparatively specialisedprocurement approach. It is unusualwithin the public sector particularly inScotland. Audit Scotland’sconsultants compared key featuresof the organisation of the Holyroodproject with usual constructionpractice. This included comparisonswith seven constructionmanagement or similar projects fromacross the UK with outturn costs ofbetween £38 million and £300million.

5.51 Although the Holyrood project ismore costly than the selectedbenchmarks the comparisonsuggests important differencesconcerning roles and responsibilitieswithin the Holyrood project againstthe usual practice.

Project reporting

5.52 At Holyrood, the constructionmanager produces a regular progressreport as required by its contractualobligations, initially every fortnight,latterly monthly. Its main focus isprogress and programme. The costconsultant reports separately to theclient and the other consultants onproject costs. Since 2002 the costconsultant has estimated the cost ofproject risks for the client but thedetailed analysis is not shared withthe other consultants.

5.53 Normal practice is for theconstruction manager to produce aregular (normally monthly) report tothe client including progress againstprogramme, cost against cost planand quality against specification. Inother words the construction

manager provides an integrated viewof performance against the broadgoals of project cost, time andquality. There would not normally beseparate reports to the client fromother members of the project team,which may be divisive and make itdifficult for the client to reach anintegrated view of the project.

Cost management

5.54 Normally under constructionmanagement, the cost consultant isonly responsible for costmanagement and reporting untiltrade packages are let. Theconstruction manager takes overresponsibility for cost management,reporting and control progressively astrade packages are let (in the sameway that it assumes control ofprogramme, quality, changes,variations, etc). The cost consultantretains a monitoring role and watchingbrief on costs, on behalf of the client.

5.55 At Holyrood, the constructionmanager has not taken responsibilityfor cost management and reportingand control, apparently because itcould not agree the cost plan withoutqualification in November 2000. Thecost consultant has thereforeretained responsibility for cost advicethroughout the project butresponsibility for cost control is notclear. Nor has the constructionmanager assumed the usualintegrated responsibility for cost andprogramme management.

Scoping of trade contracts and

interfaces between trade

contractors

5.56 Contracts for Holyrood havebeen let with scope which hassubsequently changed significantly,as shown in Part 4. Normally underconstruction management there iscareful scoping of trade contracts. Itis important to ensure that togetherthe trade packages includeeverything required for theconstruction of the whole project.

Normally under constructionmanagement trade packages wouldbe let with no, or very few,provisional sums.

5.57 The consequence of awardingcontracts with a high proportion ofprovisional sums is that in suchcircumstance the client has poor costcertainty. While tenders may appearto be in budget, the prices are notsecure, as is evident in the Holyroodproject.

5.58 For construction managementto be successful the commonboundaries between the tradecontracts and how they will integratewith each other must be carefullydefined. This will include respectiveresponsibilities for achievingtolerances, for access, for safety andfor protection of completed works.Poor interface definition and/ormanagement will result in delay andadditional costs.

5.59 At Holyrood, the constructionmanager has placed a high degree ofeffort on managing the commonboundaries between contracts.Nevertheless, there are many casesof delays having been caused byproblems at the boundaries – forexample scaffolding having to bemoved multiple times to permit thefixing of windows and adjacentglazing panels.

The architect’s role

5.60 Under constructionmanagement it is essential to selectan architect who can envisage thewhole and the detail at the sametime. This is to enable the scopingand design of trade packages with along lead-time to be completed inadvance of adjacent and dependantpackages.

Page 78: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

76

5.61 The architectural joint venture iscomprised of two reputable firms ofhigh quality and was selected after acompetition in 1998. However theoriginal procurement strategy for theHolyrood project resulted in theselection of the architect before theclient had opted for a constructionmanagement contract. The selectioncould not therefore make anyallowance for the design team’sexperience and ability to workeffectively under constructionmanagement. There are specialfeatures of working with a signaturearchitect such as Enric Miralles,which can present significantchallenges under constructionmanagement (Exhibit 48).

Risk management

5.62 Effective risk analysis andmanagement is fundamental to allforms of project procurementincluding construction management.But construction management placesa higher risk with the client thanother forms of procurement becausethere is no main contractor role. Thissuggests that under constructionmanagement the client should placea greater premium on riskmanagement than under otherapproaches.

5.63 As noted above, however, theclient’s approach to risk managementon the Holyrood project was not fullyconsistent with good practice.

Fee arrangements

5.64 Often under constructionmanagement consultants areappointed on a percentage fee basis,but the fees are converted to lumpsums when the cost plan is agreedbetween the client and the projectteam.

5.65 As noted above, although feecapping was introduced for theconsultants, in most cases this wasin 2003 well beyond the point firstenvisaged.

Page 79: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Part 5. Project management and control 77

Source: Audit Scotland

Exhibit 48Special features of working with a signature architect

A signature architect is high-profile individual who is very clearly identified personally with a building and its design.The top (household) names in architecture can be regarded as signature architects.

Clients often choose to use a signature architect to create a special building, which is of symbolic importance andwhere a ‘statement’ is being made through the architecture. There are clearly benefits from doing so. But therecan be disadvantages too. These may include:

• more likelihood of the design being novel, unusual in style, more complex and perhaps with unusual materialsbeing selected

• may be more difficult to communicate the design to tenderers/contractors

• may be more difficult to estimate costs and price tenders

• may be more difficult to construct the building and more difficult to manage the construction

• design may take longer to produce

• architect more likely to push for higher quality levels, with less likelihood of compromise

• architect has a high reputation risk, which may influence his flexibility

• architect likely to develop a close relationship with the highest levels of the client organisation (Chairman orMinisterial level), thereby making it more difficult for the project sponsor and the project team to manage thedesign and the project.

Page 80: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

78

The following pages of the reportsummarise Audit Scotland's detailedfindings from its examination of 20individual trade package contracts.This Appendix 1 provides summarydata about the 20 contractsexamined.

Appendix 2 summarises detailedfindings from the examination of fiveof these contracts as an illustration ofthe scope, nature and progress ofindividual contracts.

Appendix 1. Trade contract case studies

Package Description Contractor Date of Basis of Cost Estimated

contract appointment Plan outturn

cost

1 Substructure Foundations etc O’Rourke Sep 99 Two stage £3.2m £3.0m west for MSP building

2 Substructure Foundations etc O’Rourke Jan 01 Two stage £6.5m £7.2meast for assembly

buildings

3 Assembly Including debating O’Rourke Jan 01 Single stage £16.9m £39.8mbuilding frame chamber and towers

4 MSP building – Coverite Jan 01 Single stage £0.7m £1.0mroofing

5 MSP building – Ultimate Mar 01 Single stage £2.3m £4.0mcarpentry and joinery

6 Electrical east Assembly building Forth Electrical Sep 01 Single stage £6.8m £8.0melectrical work

7 Queensberry – Drawn Metal Sep 01 Single stage £0.8m £1.2mHouse blast doors and windows

8 Foyer roof and – Mero Oct 01 Two stage £1.5m £7.4mglazing

9 Assembly Including Canongate Coverite Oct 01 Two stage £4.4m £7.5mbuilding roofing and towers

10 Mechanical and Plumbing etc. for Rotary Nov 01 Single stage £6.4m £9.1mplumbing east assembly buildings

11 Assembly Including press tower Drawn Metal Nov 01 Single stage £2.6m £13.0mbuilding and Canongate windows

12 MSP building – Baydale Dec 01 Negotiated £1.5m £3.4mbay windows

13 Toilet and fitness Fittings for MSP Mivan Jan 02 Single stage £1.4m £2.9marea fit-out building

14 Stone flooring Flooring for MSP Vetter Feb 02 Single stage £1.3m £1.8m building

Page 81: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Appendix 1. Trade Contract case studies 79

Package Description Contractor Date of Basis of Cost Estimated

contract appointment Plan outturn

cost

15 MSP building – Mero Mar 02 Single stage £1.0m £2.8m windows and cladding

16 Specialist Assembly buildings Mero May 02 Negotiated £1.9m £11.0mglazing specialist glazing

17 Hard Perimeter walls, O’Rourke Aug 02 Single stage £4.4m £5.6mlandscaping roads, etc

18 Assembly – Space Decks Aug 02 Negotiated £0.2m £1.9mbuilding rooflights

19 Zone 1 fit-out Fittings for chamber, Mivan Aug 02 Single stage £5.0m £14.1mtowers 1 & 2, presstower

20 Zone 2 fit-out Fittings for Canongate, Mivan Sep 02 Single stage £2.9m £7.8m foyer, towers 3 & 4

Total £71.7m £152.5m

Note: Cost Plan refers to the adjusted cost plan figure. The Estimated Outturn cost is the estimated final cost of the package including inflation but excluding risk elements. The combined risk element for the 20 packages is £40.3 million. Figures are rounded to the nearest £0.1 million.

Source: Audit Scotland

Page 82: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

80

Procurement

1. The package was procured via asingle stage tender issued to sixtenderers on 13 October 2000. Dueto the quantity and complexity ofpost-enquiry tender addenduminformation issued the tender returndate was extended by ten days to 20 November 2000.

2. At pre and mid-tender meetingsthe team noted that O’Rourkedemonstrated the most enthusiasmand the best understanding of theproject requirements. Two of theoriginal tenders withdrew from thetender process and four tenderswere returned on 20 November2000, with prices ranging from £17.7million to £19.8 million. This contrastedwith the DLE pre-tender estimate forthe package of £23.5 million and thecurrent cost plan allowance, plusinflation, of £16.1 million.

3. The initial review of the tendersassessed that O’Rourke (lowesttenderer) best satisfied the awardcriteria, while the highest tender wasuncompetitive and discounted.O’Rourke was given a post-tender

interview; with the two others held inreserve.

4. Further revisions to the tenderinformation on finish to the concretein the towers was given to O’Rourkethe post-tender period and theyadvised a significant cost impact,which the team had anticipated. Thisinformation was given to theremaining tenderers for comparisonand their poor response confirmedthe team’s earlier decision.

5. The construction manager’s tenderrecommendation was not issued toproject management until 19 January2001 as the post-tender period wasprolonged with attempts to mitigatethe overspend on the package.Target savings of £1.4 million wereidentified, the bulk of which weresubsequently achieved beforecontract award and O’Rourkereceived a letter of intent on 23 January 2001 and started on siteon 12 March 2001. Removal andclarification of bid qualifications andthe introduction of additionalprovisional sums increased thetender to £17,964,654.

6. The trade contract sum includedprovisional sums, totalling £4.68million. Additional elements of thestructures elements totalling £3.78million, were based on provisionalquantities re-measurable. Thistotalled £8.46 million, some 40% ofthe contract sum.

Design issues

7. It was acknowledged prior totendering that the ideal level of designinformation would not be available atthe programmed dates. In December2000 design was 14 weeks behindprogramme and the package was letwith much design to be completedbut work started on site at thescheduled date in the programme.

8. At package award, O’Rourkeissued an Information RequiredSchedule (IRS) indicating key datesby which they required designinformation to progress their owndesign elements and proceed withthe works. But the designprogramme slipped from the agreeddates and detailed design wascarried out whilst the works werebeing built.

Appendix 2A. East frame

Contractor O’Rourke (Scotland) Limited

Scope of works The package covers the structural frame to the assembly buildings. The structures comprise in-situ concrete columns, beams, vaults and walls, precast concrete stairs and wall panels, post-tensioned and reinforced concrete floor slabs, structural steel frames and trusses and laminated structural timber roof structures. The package includes some trade contractor design responsibility, relating to concrete reinforcement, concrete vaults, pre-cast concrete, structural steel roof structures and laminated structural timber roof structures.

Programme Planned ActualLetter of intent date 22 January 2001 23 January 2001Start on site 12 March 2001 12 March 2001Contract date – 14 May 2002Programme period March 2001 - May 2002 March 2001 - still on site

Costs

Adjusted cost plan + inflation £20.2 millionTrade contract value £23.5 million including £5.5 million for contingency and riskEstimated final costs £39.8 millionRisk still at large No – works substantially completed.

Page 83: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Appendix 2A. East frame 81

9. In July 2001 the constructionmanager was registering concernabout design information stilloutstanding on the package and thepotential critical adverse impact onthe Strategic Programme. Concernswere that final cost and programmecould not be established withoutformalising the extent and thereforevalue of the provisional sum items.

10. O’Rourke prepared a paper inOctober 2001 to facilitate somerecovery of the programme andsummarised information requiredissues. It records that 49% of theirRequests For Information (RFIs)related to conflicts betweenstructural and architectural drawingsand 32% related to areas whereinformation provided by the designteam was considered deficient. Atthat point O’Rourke were typicallyraising 50-60 RFI’s per week.

Programme

11. Works started on site on theprogrammed date of 12 March 2001and, according to the contractprogramme, should have completedby 20 May 2002. O’Rourke remainedon site in May 2004.

12. With the trade contractor startingon site on the agreed date, inadvance of key design elementsbeing finalised, detailed design wascarried out as works were built.Developing areas of complex designand the geometry required to buildthe structures, whilst progressing theworks, proved challenging.Construction, co-ordination andsolving buildability issues impactedon the design progress and had aknock-on affect on the overallprogramme.

Costs

13. The Estimated Final Account forthis package as at Cost Report No86, 16 April 2004 was £39,825,000,representing an increase of 122% onthe original trade contract sum. Costsincreased due to a number of factors,including mainly:

• design development

• disruption

• prolongation

• scope changes

• provisional items resolution at ahigher level than originally allowedfor. For example:

- glulam works final cost of£2,612,429 against an originalallowance of £800,000

- structural steelwork final cost of£5,448,105 against anallowance of £4,000,000

- boundary walls added into thepackage at a cost of £3,000,479

- other provisional sums generallyincreased by £929,201

- stainless steel nodes in theglulam roof increasing from 90nodes at some £2,000 per nodeto 111 bespoke nodes at anaverage of £11,000 per node.

14. Sundry other changes also addedto the cost. There were 1,740 DCO’s(Draft Change Orders) raised on thispackage by 31 October 2003 and atotal of 141 were awaited.

Page 84: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

82

Appendix 2B. MSP bay windows

Procurement

1. Procurement for this package wasinitially secured through acompetitive tender won by Flour CityArchitectural Metals (UK). Followingthe collapse of Flour City the packagewas secured through negotiation of afixed lump sum against a singletender for £2,515,739. The packagewas let following tender scrutiny foran agreed £2,655,399.

2. The tender was qualified onprogramme commitment and wastechnically non-compliant. Baydalepromised to endeavour to meet theconstruction manager’s programmebut would not be contractually boundto this. The construction managerrecommended the client’s acceptanceto finalise the order at the earliest, toenable the works to start, withprogramme improvements sought onan ongoing basis during the works.

3. In March 2003 blast doors weretransferred into the package from theDrawn Metal package; and, in April2003, Reid’s Close and the car parkroof were added. Further, later,additions were the assembly building

rain screen cladding, privacy screensand the panelling to Zen Garden. Atacceptance, costs associated withblast testing and additional insurancewere uncertain and excluded.

Design issues

4. A number of design issues aroseon the package relating to:

• increased costs from the additionand delivery time fromauthorisation for windowactuators

• membranes and rainscreens

• decisions on painting of blastdoors

• manufacture of blast doors instainless steel or mild steel

• potential delays in manufacture

• operation of blast doors ie,manual system – risk highlightedby Booth

• risks associated with 25 yeardesign specification.

Programme

5. On 16 September 2002 an interimapproval of extension of time wasmade based on a number of headsof claim:

• late contract signing meant amissed production slot at theFrench window factory

• late selection of stain colour

• late availability of stone cladding

• sundry site related delays, suchas scaffolding obstructing accessetc

• additional and enforced activity ofsurveying rainscreen fixing pointsindividually

• draft change orders.

6. The total claim of 18 weeks 2 dayswas awarded and completion wasextended to 16th May 2003.Although no further extension claimswere allowed Baydale remained onsite in December 2003 completingwork and resolving issues.

Contractor Baydale Architectural Systems

Scope of works The package comprises the bay windows and adjacent stainless steel panels to the MSP Offices at the Scottish Parliament Building, Holyrood, Edinburgh. This package also includes the supply and installation of galvanised mild steel framing to support the windows, flashings, privacy screens, louvers and window actuators. There are 114 west elevation bay windows and 14 north east windows.

Programme Planned ActualLetter of intent date 20 December 2001 (recommended) 27 March 2002Start on site 3 December 2001 (shown on programme) 27 March 2002Contract date –Programme period 3 December 2001 - 9 September 2002 27 March 2002 - present

Cost

Adjusted cost plan + inflation £1.8 millionTrade contract value £2.7 millionEstimated final costs £4.4 million

(including some £1.0 million for package additions)Risk still at large Yes

Page 85: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Appendix 2B. MSP bay windows 83

7. The blast doors are critical to theproject and the construction managerinitially indicated that it was notsatisfied with the level of detailcontained in the contractor’sprogramme; but it accepted that anelement of uncertainty surroundedthe design and acknowledged thatthe design team had contributed todelays in concluding the design ofthe blast doors.

Costs

8. The initial contract award valuewas £2,655,399 and it expanded toan estimated final value of£3,455,339, excluding the additionalblast door and Reids Close Works.The total variance equates to£943,600 or 37.5% of the originalcontract sum. Additional work addedto this package (for escape blastdoors, assembly rainscreen cladding,Reid’s Close coping units, privacyscreens and panelling to the ZenGarden) has added some £1 millionmore to these costs.

9. The average cost of windowscontracted for with Flour City was£5,633 each compared withBaydale’s £17,174 each.

Page 86: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

84

Appendix 2C. Specialist glazing(Stages 1&2)

Procurement

1. The nature of the glazing specifiedby the design team severely limitedthe range of contractors capable ofdelivering the package.

2. Early programmes had scheduledtendering enquiry issues for August1999, September 1999 and thenJune 2000. Pre-qualificationquestionnaires were not issued untilNovember 2000. The package tenderwas switched to a two-stage tenderroute because design information forthe package was not complete.Tender enquiries were issued to sixbidders on 14 March 2001 and threedeclined to bid. Returns wererequested for 18 April 2001 but thatdate was extended at the request oftenderers until 4 May 2001.

3. Tenders were received fromSchneider, Space Decks and FlourCity, with bids ranging from £2 million to £8 million, the lattersignificantly over the cost plan target.As the lowest bidder, Schneider didnot price for preliminaries, their bidwas not technically compliant.However, despite being unable toagree preliminaries figure withSchneider the construction managerrecommended progressing with

them; and the design team believedthat the Schneider’s outline proposalcould be developed to a workablesolution in the second stage. Theclient approved the Design Team’sTender Report on 25 July 2001 andengaged Schneider to progressdesign works to enable a lump sumcontract to be concluded. Despiteprolonged negotiations Schneider didnot produce sufficient detail ofprogramme and resources for thecontract to satisfy the client, whoformally withdrew their offer on 5 October 2001.

4. Project management took legaladvice on progressing the package,which cautioned replacement ofSchneider without following theprocurement regulations. To progressthe package, the client issued furthertender invitations on a negotiatedbasis to Schneider and Mero (UK),who were engaged for the foyer roofpackage. However, issues overSchneider design ownership haltedthis and the client continued tonegotiate with Schneider throughNovember to April 2002. With thedelay in negotiations, the client re-engaged with Mero in April 2002 as acontingency. On 1 May 2002, whenSchneider rejected the terms of a

final contract offer negotiations werestarted with Mero and a contractagreed on 22 May with an estimatedtarget value of £7.2 million.

Programme

5. The package contractorcontributed to the preparation of theconstruction manager programmethrough the provision of sterile30

programme data but could notmaintain progress to the variousprogrammes set by the constructionmanager. There are a number ofreasons for this, including:

• delays in the production of designinformation from the designteam; iteration of the contractor’sdetailed design to satisfy thearchitect’s design concept

• resolution of design issues tomeet blast performance on non-standard designs and thenecessary design testing

• sourcing glazing suppliers for thefeature wood laminated windowsin one area

• delays in resolving issues over the glass choice to lighttransmission and acoustic

Contractor Mero (UK)

Scope of works This package covers 11 areas in the parliament complex; including the extensive glazing to the front wall of the debating chamber, the north lightwell and the public stair.

Programme Planned ActualLetter of intent date – 22 May 2002Start on site 8 October 2001 1 August 2002Contract date – 21 May 2002Programme period September 2001 - May 2002 July 2001 - July 2004

Cost

Adjusted cost plan + inflation £2.4 millionTrade contract value £7.2 millionEstimated final costs £11.0 millionRisk still at large Yes

30 Programmes produced assuming no constraints on the timing of the work or access.

Page 87: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Appendix 2C. Specialist glazing (Stages 1&2) 85

performance of glazing in thedebating chamber

• general problems of out ofsequence working particularly inthe North Lightwell area

• continuing design developmentby the architect throughout thecourse of the programme to meetthe design concept

• resolution of multiple interfacearrangements with other packagecontractors

• deploying insufficient resourcesfor installation to meet ambitiousprogramme schedules.

6. While the design team preparedthe concept drawings and specifiedthe performance of the glazing thepackage contractor was responsiblefor developing the design through tothe detail; including the multipleinterfaces with other packagecontractors. The complexity of thedesign concept for some areas wasnot appreciated from the early 2Ddrawings eg, the glazing on the NorthLightwell, where the glazing meets indifferent planes. Much of the designinvolves bespoke interfaces that havenot been used before.

7. At the outset the contract was tobe completed by May 2002. Asnoted above the award of thecontract was delayed. The client hasgranted an extension of time on thepackage, which is now scheduled tocomplete in July 2004.

Cost increases

8. The cost plan for the package wasbased on an elemental allocationfrom the available budget using thelimited design information in theStage D proof cost plan at October2000, to which the architect hadagreed. Design development for theeast side was poorly developed andallowed no appreciation of thecomplexity of the solution thearchitect might subsequently adoptand the client accept.

9. The contract agreed with Mero isbased on actual costs plus a fixedpercentage for preliminaries andprofits. The use of cost plus contractremoved the need for competition insupply of materials for the package.Therefore the client is required to paythe price for the materials sourced bythe package contractor, who has nofinancial incentive to look for the bestprice. With more than 99 per cent ofthe materials manufactured inGermany and Austria the client hasborne exchange rate risk but noestimate of the extent of this ispresently available. The contractorcould not identify for Audit Scotlandany cost increases due to changes inblast requirements. Blastrequirements were incorporated intothe design as it developed.

10. At present the contract has run53 per cent over cost on the basis ofagreed commitments but furthersums are still at issue. Forcommercial reasons these are notdisclosed here pending agreement offinal package accounts.

Page 88: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

86

Appendix 2D. Foyer frame andglazing

Procurement

1. Due to the lack of designdevelopment the package wastendered in a two-stage process withthe trade package extended to coverthis. The tender issue date was 9 May 2001 with returns by 11 June2001. Under pressure from alltenderers for furtherer details ofdesign issues, the tender date wasextended by two weeks.

2. Four bids were received for thepackage and two were noncompliant by not pricing for specifiedelements. The two competing bidswere very close and the contractwas awarded to the lowest tenderfollowing post tender clarification ofcosts and pricing. The winningtenderer, Mero (UK), had experienceof designing to a target sum andshowed a willingness to move to alump sum if the client should agreethis. However a lump sumagreement was never achieved asdesign complications developed.

3. The client approved the award ofthe contract in October 2001 andMero proceeded on an interimcontract issued 28 November 2001.

A contract was finalised on 12 April2002.

Programme

4. The programme originally set forthis package allowed 16 months. Thepackage took 22 months to achievepractical completion in April 2004.Sterile31 programmes for the packagewere developed by Mero andfactored into the overall projectprogramme by the constructionmanager. The programmes were notmet due to a number of factors,including:

• insufficient design information

• design development exchangesbetween the design team and thetrade contractor

• blast testing

• access delays due to otherpackages

• availability of cranage

• resourcing by the contractor.

Design issues

5. The design at tender stage wasbased upon the architects designconcept images, which looked toextensive roof lights on a lightstructure, with no indication ofsecondary supports for glazing, asthe structural engineers had notengineered these at that stage. Thiswas developed in stage 2 of thetender. The original design did notaddress blast loading requirementsnor the added complications of blastdebris in the form of the extensivegranite cladding. These demanded astructure which was much moresubstantial than that which waspresented at Stage D. The structureeventually developed comprisedsubstantial steel sectional supports.

6. The tender documents envisageda narrow silicone glass joint that wasincapable of sustaining blast loadings.Subsequently, a joint was developedon the advice of Arup FaçadeEngineering that required the use ofseveral metal extrusions for jointingin addition to the secondary framingsupporting the blast loads.

Contractor Mero (UK)

Scope of works This package covers the foyer, which occupies the former ‘A listed’ garden of Queensberry House and provides the circulation space from the MSP building to thecomplex on the east side of the site. The package covers the frame supporting the roof and the distinctive leaf motif roof lights.

Programme Planned ActualLetter of intent date 30 July 2001 10 October 2001Start on site 4 March 2002 22 July 2002Contract date – 1 October 2001Programme period November 2001 - April 2002 August 2001 - April 2004

Cost

Adjusted cost plan + inflation £1.8 millionTrade contract value £5.8 millionEstimated final costs £7.4 million Risk still at large Yes

31 Sterile programmes assume no constraints on working and design information supplied as programmed.

Page 89: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Appendix 2D. Foyer frame and glazing 87

Cost issues

7. Due to the limited level of conceptdesign development the packagecost plan stage was based on anelemental allocation from theavailable budget. At that point thearchitects had committed to designwithin a price of £500/m2.

8. In April 2001 the emergingconcept design development wasnot adhering to the cost plan, andwas costed by DL&E at £3,300/m2.Because of increasing programmepressures to complete by November2003, HPG accepted the developingdesign and allowed development tocontinue.

9. Suggestions were made by thecontractor to standardise elements ofthe design to reduce manufactureand installation costs but thearchitect rejected these. The originaldesign concept envisaged the use oflead and copper, both ductile metalsand readily suited to shaping oncomplex forms designed into thisstructure. The use of non-ductilestainless steel in place of lead andcopper increased the complexity inmanufacture, fitting and costs toachieve the architecturalrequirement.

Page 90: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

88

Procurement

1. The package was procuredthrough a two-stage process, withcontractors selected through a pre-qualification process and approved bythe design team and client. InDecember 2000 the constructionmanager wrote to the design teamindicating that the package was notready to be issued for tender. InJanuary 2001 there were indicationsthat the package could produce asignificant overspend.

2. Six tenders were invited on 4 May2001 but only two were returned byclose on 18 June 2001. On 2 October 2001 the constructionmanager recommended acceptancethe Coverite first stage bid, with amaximum value of £114,342 fordesign works to be carried out duringan eight-week period.

3. Second stage negotiations were tocommence on 9 October 2001 andcomplete on 4 December 2001 butwere extended to March 2002. Theintention was to culminate in anagreed lump sum price for the tradeworks not exceeding £6.5 million.

4. Negotiations on the full contractactually took until 26 July 2002,when the construction managerrecommended accepting a contractwith Coverite for £6.96 million. Thisexceeded the target sum andincluded a provisional sum allowanceof £2,200,000 some 32% of thecontract value.

Design issues

5. Issues on detail design arose onthe package in June 2001 andfundamental design issues, such aspurlin spacing remained outstandingin November 2001.

6. Design workshops wereundertaken throughout the contract.By August 2002, the constructionmanager was recording frustration atthe time being taken by the architectto conclude the detailing of the greenroofs.

7. In June 2003 the Architectexpressed great concern to projectmanagement regarding approval byproject management to asubstandard detail regarding thevapour barrier. This important issuebetween architect and project

management remained prominentinto late 2003. An independentcondensation risk assessment wasconducted. The architect expressedextreme concerns to projectmanagement regarding lack ofprogress in resolving the matter.

8. There are a significant number ofdesign issues recorded throughoutthe progress meetings relating todesign information and coordinationaspects. By 14 October 2003 thedesign team had issued 389instructions on the package.

Programme

9. The planned programme period forthe package was from January 2002to Mid September 2002. On 14 September 2002 an extension oftime of some 59 weeks was grantedtaking target completion to 17 November 2003.

10. The extension of time wasgranted on the basis that Coveritewere given late access and were notprovided with free/uninterruptedworking, hence they were requiredto undertake the work on the basisof a multiple visits. However, further

Appendix 2E. Assembly roof

Contractor Coverite

Scope of works The package comprises the design, manufacture, supply and installation of stainless steel roofing to the assembly buildings. The scope of work includes roof purlins, structural steel decking, insulation and waterproofing, stainless steel roofing and gutters with integral roof edge protection system. The roof package also includes for the design, manufacture, supply and installation of rooflights to each roof.

Programme Planned ActualLetter of intent date – Tender recommended - 26 July 2002

Stage 2 to proceed with design - 2 October 2001 Start on site – 28 January 2002 Contract date – 26 July 2002Programme period January 2002 - September 2002 January 2002 - present

Cost

Adjusted cost plan + inflation £5.3 millionTrade contract value £7.0 millionEstimated final costs £7.5 millionRisk still at large Yes

Page 91: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Appendix 2E. Assembly roof 89

work continued on site and by midJanuary 2004 design issuesremained outstanding andcompletion was scheduled then formid-June 2004.

Costs

11. The initial tender was for £6.2million included provisional and primecost sums of £2.2 million. However,following design development, afixed price contract was accepted atsome £7.0 million, exceeding the£6.5 million target value originally setby the client.

12. In early 2004 the total contractcommitment stood at £7.5 million, anincrease of some 38% over theagreed cost plan plus inflation.Further sums are at issue. Forcommercial reasons these are notdisclosed here pending agreement ofthe final package accounts.

Page 92: Management of the Holyrood building project · 1. The subject of my report is the management of the project to provide the new Scottish Parliament building (the Holyrood project):

Audit Scotland110 George StreetEdinburgh EH2 4LH

Telephone0131 477 1234Fax0131 477 4567

www.audit-scotland.gov.uk ISBN 1 904651 42 9 AGS/2004/7

Management of theHolyrood building project Insert sub head