The four reasons why major projects fail in the UK: Reason Number 2: The speed of decision making

In a series of articles I am looking into why the UK has struggled to successfully deliver major projects. As we embark on a major programme of megaprojects this question has never been more relevant. HS2, Crossrail 2, Small Modular Nuclear Reactors, superfast broadband, submarine nuclear deterrence, highways and bridges. Investment of hundreds of billions of pounds. All coming off the back of a series of highly publicised failures and public concern at rising costs and serious delays. Clearly, this next series of megaprojects needs to be delivered differently otherwise we should not be surprised if the results are the same. MIGSO-PCUBED is the world’s largest project management consultancy and working in collaboration with academia and our industry partners we have undertaken a major research programme to understand the root causes of this failure to deliver, and in this series of articles I present our conclusions. Second in the series is:

The speed of decision making.

Background

The Built Environment Committee of the House of Lords recently conducted an inquiry into the government’s decision-making processes for major infrastructure projects.[1] It found that: “The process has become slower in recent years. An infrastructure planning application entered in 2012 would typically receive a decision in 17 months; by 2020 the average wait had risen to 22 months. We heard that it can take as long to gain planning permission as it does to build a project.” It also found that once approved the project could be further delayed by issues requiring government intervention to revisit decisions during the delivery of the project. The recent decision to cancel the HS2 northern link is an example of this.

Let’s compare two projects in different governance regimes: Hinkley Point C (HPC) nuclear power plant in Somerset and the similar Barakah 1 plant in the United Arab Emirates (UAE).

I make that 37 years planning and 10 years build (fingers crossed) for HPC versus 4 years planning and 9 years build for Barakah. And bear in mind that the HPC build duration is an estimate.

The UAE is, of course a “new entrant” with no prior experience of operating or constructing nuclear power plants. If we look a little further east at the Korean experience you can see from the graph below[2] that repeat build programmes bring learning benefits and it’s likely that the UAE’s performance will continue to improve. They plan a second series of 4 reactors and we should reasonably expect the build time to reduce further towards the Korean best practice of around 5 years.

?

If we ignore the first two attempts to build HPC (although we really shouldn’t) it still took 11 years from publishing the policy statement to approving construction. Just think about the threat this poses to a project once it has started. The project leader has to manage a project in a decision making regime that is easily capable of taking many years to make a major decision. So, if the project needs to make a major change in direction and needs to go back through the various governance layers to gain approval, the impact on the project timeline and cost will be huge, possibly requiring demobilisation of the workforce and even project cancellation. Again the HS2 northern leg is an example of this along with Wylfa and Moorside nuclear plants and many, many others.

Root Causes

Let’s look at what’s happening in a little more detail by examining two examples in the UK nuclear industry. The first example had a remarkably similar experience to HPC. In the early 1980s BNFL and the Nuclear Installations Inspectorate agreed a plan to process waste from Sellafield’s highest waste repository, the Magnox Swarf Storage Silos (MSSS). So started the SDP project, or the Sellafield Drypack Plant. This was initially progressed by a BNFL subsidiary, BNFL Engineering Limited (BEL). Construction started in the early 1990s but was aborted in 2002 due to technical problems. A second attempt was launched using a more basic technical solution and now called the Silos Direct-encapsulation Plant (still SDP!). Development work on this solution was cancelled in 2008 and the project was competitively tendered instead. Nuvia won this competition. Design and construction work restarted, but problems continued to dog the project and in 2012 the Nuvia contract was cancelled. A new competition was launched and a new consortium, Areva-Mace-Atkins (ama) took over in 2014. However, Sellafield decided just 18 months later that an alternative technical solution was preferable and cancelled the project completely in 2015. Over this time costs escalated from ￡200m to over ￡2bn and the start date slipped from 2003 to, well, never (the last forecast was 2022). Interestingly, each of these iterations of the project increased the number of stakeholders and slowed down the speed of decision making. Initially a vertically integrated BNFL team was delivering the project. This then became a client-contractor model and eventually became a client-consortium model. In this final arrangement a key decision required by the project leader to mitigate a key risk that significantly affected time, cost or scope would need to be approved by three corporate boards, two in the UK and one in France, before being passed to the operating company, Sellafield Limited. Then to the Government agency set up to manage UK nuclear decommissioning, the Nuclear Decommissioning Authority. Then to the relevant government department (Business, Energy and Industrial Strategy) which was advised by the Shareholder Executive and finally approved (or not) by HM Treasury. I make this 8 CEOs or Secretaries of State. Even worse, if the problem originates in the supply chain it is possible for 25 organisations to be involved in the decision.

The second example I want to look at highlights a different problem with dispersed decision making in non-vertically integrated supply chains. Towards the end of the last century a new submarine refitting facility was being constructed at Devonport dockyard in Plymouth. This was suffering from escalating costs and delays: several hundred million pounds over budget and delayed to the point that it was threatening the availability of submarines to meet the military need. A key contractor, with unique analytical capability, was approached to provide a revised technical analysis which if successful would reduce the need for major redesign and further delay of the facility. However, the contractor concluded the risk was too high to take this contract. Why? The value of the work to them was around ￡10,000, but if it subsequently proved to be in error, they could be liable for costing the MOD several hundred million pounds and potentially even losing the ability to put submarines to sea. Had this contractor been part of the MOD or Devonport then the work would have been done and would probably have shown that the redesign was not needed, saving substantial time and cost.

You will note that all the examples quoted in this article are nuclear examples. This is because there is precious little published data on governance processes. Companies protect this data and generally the important lessons are associated with project failures which they have no interest in publicising. Therefore all the examples quoted in this article are projects that I have worked on at executive level, either in a delivery or governance capacity.

The research

In order to develop insight into this issue I led a research project with the Nuclear Institute (NI) and the Association for Project Management (APM) via the Project Management Special Interest Group (Nuclear PM SIG) and we adopted a research approach based on a number of stages:

• Internal review and debate in the SIG leading to the development of a hypothesis. Most of the major UK nuclear organisations were involved in this debate.
• Publication of a blog via the APM website to engage with project management professionals and create a social dialogue via LinkedIn. We also spoke to senior representatives from the investment community to get their thoughts.
• Delivery of a webinar via the APM and NI to engage with a mixed group of project and nuclear professionals, including an open Q&A session.
• Survey of senior project and nuclear professionals to obtain the views of an informed professional group of insiders.
• Publication of an insights paper to report the findings of the study and potential next steps[3].

Bringing together this combined thinking of project managers from most of the nuclear organisations in the UK, the investment community and subject matter experts from consultancy and academia we developed a hypothesis which we intended to test: Complex nuclear projects need timely decision making to be successful but slow governance inherent in western nuclear projects makes this challenging. This is a major inhibitor to predictability and market-competitive time-cost performance.

I believe this hypothesis and the research we conducted is equally applicable to any non-nuclear major project in the UK where commercial, political and regulatory interests are involved. Which is probably all of them. ?

The survey of nuclear and project professionals found that no-one felt these projects were predictable and over 70% of respondents thought they would be late and over budget. Just think about this for a moment. These are the people currently leading these projects!

We analysed the stakeholder maps in the UK and the UAE and found that the UK has a far more complex set of stakeholders than the UAE. We found that the UAE, as you would expect is very lean, being new. If you want to make a major decision in the UK, then there are a lot of stakeholders that need to be consulted. All of them are very capable lobbyists and we’ve didn’t even consider the NGOs, the media and other influencers.

The UK nuclear industry has become progressively more fragmented over the 70 years of its existence. At one stage in its history, the peak of productivity in nuclear project delivery, the UK’s stakeholder map was not unlike the UAE’s, with a single developer/operator, one research body, one regulator and large engineering giants with vertically integrated supply chains. Over the years as we have privatised, globalised and decentralised we have created a large set of relatively small, but influential players. This makes decision making complex, with more stakeholders wishing to be involved, leading to decision paralysis in some cases. We compared the devolved and decentralised UK stakeholder map to the UAE’s chain of command and found that a single decision in the UK could involve over 25 organisations and require over 100 steps. In the UAE this is 4 or 5 organisations and less than 10 steps.

Another point we need to consider is the oft quoted statement that projects are temporary endeavours. This simply isn’t true for large megaprojects[4]. The “temporary” HPC project will have lasted around 50 years by the time it is completed. And in truth it is a programme of projects that will continue indefinitely. Meanwhile the so called permanent organisation that owns HPC has changed several times: CEGB to Nuclear Electric to British Energy to EDF to an EDF-Centrica partnership to an EDF-CGN partnership. It’s likely that CGN’s share will be sold before completion of the project due to UK concerns over China’s involvement in its nuclear programme. So there will have been seven “permanent” organisations but the “temporary” organisation has stayed constant. This ownership “churn” has a negative impact on decision making. When owners change, they often want to reassess major elements of the project, creating artificial decisions that don’t need to exist from a project perspective. These decisions can be significant and result in major delays if not managed well.

A more human example of the impact of this temporary-permanent question is a person I know who spent his whole career working on the SDP project discussed earlier. Over that period of time he worked for 6 different employers. Just before he retired, the project was cancelled. Try telling him it was a temporary endeavour! You need dedication to stick with these projects.

Guidance for Project Leaders

The purpose of our research and the objective of these articles is to provide advice and guidance to project leaders. Firstly to provide awareness of the importance of these 4 reasons for failure and secondly to offer solutions that can be applied by the project team. Much of the previous research on this topic focuses on macroeconomic, political and investor level causes of failure: ensuring the project starts right. However, when the project leadership team is appointed, there is little they can do about this. The project they inherit is the pretty much the one they’re going to have to deliver. We advocate a stoical approach for the project manager. Don’t worry about what you can’t change and focus on what you can.

With this in mind, we believe the fundamental root cause when the slow speed of decision making results in project failure, is a mismatch between planning and reality. Our guidance for project managers is to ensure the decision making cycle is fully understood and measured for the project in question. We have yet to find a major project team that has measured its decision-making process, either theoretically or using real life experience. This seems remarkable to me. Their schedules contain decision-making activities on the critical path. For reasons that defeat me these always seem to be 3 months or 6 months duration and are not based on analysis or measurement.

The project team should conduct a lean assessment of their decision-making processes and calculate the statistical average cycle time at 80% confidence levels. These calculated durations should be shown explicitly as an activity on the schedule for every key decision (including at each major technical milestone), especially where stakeholders outside the immediate project team need to be engaged. Stakeholders in the parent companies of the project consortium should be considered to be external to the project in this context.

Further the project should calculate the “critical decision making time” for their project. This is the average time, above which, the project becomes unstable and unpredictable, i.e. risk mitigations cannot be implemented due to the time taken to agree changes in approach. Finally, the project should measure the actual decision making times and ensure they remain safely below the critical limit. This concept is show in the model below.

If the measured average decision making time is as shown by the green bar above then the project can make decisions faster than the decision making window of opportunity. If it is shown by the red bar, then the project is doomed to be continually impacted by damaging risks. Rather like a boxer on the ropes, constantly taking punch after punch until the unequal struggle is stopped. I suspect many ongoing projects will find their schedules are unstable if they undertake this assessment. However, that’s not a reason for not doing it. In fact it’s a reason for doing it and sharing their analysis with the stakeholders responsible for the complex decision-making process.

Our second recommendation is with regard to the temporary-permanent dilemma discussed earlier. It is important that the project leaders fix the project’s governance processes as much as possible, to limit the impact of changes in ownership that will almost certainly occur during the life of the project. This can be done by setting the project up as a special purpose vehicle with its governance processes set in its articles of association or written into its safety case or similar long term and hard to change documents. Furthermore, parent company reps should only have non-exec roles on the board. Clearly, they can appoint the executive officers, but they shouldn’t directly control day-to-day project decisions. Most projects we have looked at give very little consideration to the permanent-temporary dilemma and even fewer put measures in place to protect the governance processes from unnecessary change.

Clearly there are other actions Government and investors can take beyond the scope of this research and we recommend that in agreeing new funding models Government and other senior stakeholders should seek to simplify the stakeholder map and ensure lean governance models can be implemented which can be forecast with confidence by the project manager. These governance arrangements should not merely be imported from the parent organisations. They should recognise the longevity of these project organisations and not treat them as transient teams in the way internal projects may be treated.

[1] https://committees.parliament.uk/publications/33849/documents/185245/default/

[2] https://www.oecd-nea.org/jcms/pl_14904/nuclear-new-build-insights-into-financing-and-project-management?details=true

[3] https://www.nuclearinst.com/write/MediaUploads/2765_-_Nuclear_PM_SIG_Report_No.5.pdf

[4] https://www.sciencedirect.com/science/article/abs/pii/S0263786317306385