UK's HS2 project and implications for large infrastructure projects

Time and cost overruns are a feature of infrastructure projects in particular and large projects in general. This post is about couple of recent reports from UK which points to some takeaways to address the problem. 

The High Speed 2 (HS2) is arguably the most high profile infrastructure project in UK, whose formal construction approval was given in April 2020. The project which has been years in the works, is a high speed rail connecting London with the north of England and the Midlands, and its construction has been split into three phases. A recent report by the UK's Institute for Government examined the evolution of HS2 and consolidated the takeaways. 

Like other such large projects, HS2's estimated cost has tripled even before construction has started. 

Equally important, its benefits-cost ratio (BCR) assessment has declined from 2.5 in 2012 to 1.5 by 2020.

The report makes several suggestions to address time and cost overruns - improving the analysis underpinning initial decisions (better quality DPRs, range of BCRs, alternative options), political commitments are not made before feasibility and cost-effectiveness are demonstrated, greater scrutiny of cost estimates, transparency in evolution of project cost, independent evidence checks, improving project delivery by understanding project better, flexible approach to planning, and systematic and rigorous oversight. 

Another report, by the UK's National Audit Office (NAO) offers some suggestions on more effective management of mega projects like those in transport, energy, and defence. It has this to offer as learnings on cost and schedule estimates,

Estimates made at the concept stage will necessarily be based on high-level information, which becomes more detailed as the scope of the programme is developed. The relative lack of information at a programme’s early stages means that any estimate will be highly uncertain and contain many areas of potential risk. For example, assumptions will need to be made about elements such as ground conditions, which cannot be known until detailed surveying takes place, and knowing how something will be built might require a certain level of detailed design to understand. Early costs will also likely lack input from suppliers... In addition, the use of early estimates as delivery targets can incentivise delivery bodies and suppliers to attempt to meet unrealistic expectations, which then drive behaviours that are detrimental to the successful delivery of the programme. These behaviours include overambitious attempts to find savings and meet risky schedule targets...

An estimate produced from early high-level information is unlikely to be suitable for setting a programme budget and schedule. Any early estimate of programme cost and schedule should be seen as provisional, should clearly recognise limitations and uncertainties, and be used only in an indicative fashion to guide long-term planning until a detailed design supported by industry pricing is available. Previous practice in government has been to publish cost and schedule estimates as single point figures, and it is now moving to the use of ranges to better reflect risks and uncertainties. While we welcome this development, decision-makers must understand what these ranges represent and how much confidence they can put in them. Bodies should work to understand what risks and scenarios might cause the programme to exceed these ranges, and revisit this analysis as the programme progresses...

Senior decision-makers should be alert to behaviours that suggest a schedule is becoming increasingly challenging. Persistent schedule replanning, the removal of scope and /or benefits, previously unplanned staging of a programme and excessive focus on individual project risks are signs that they need to examine closely the feasibility of the schedule. Organisations should consider early in a programme’s lifecycle what needs to be in place to meet their schedule targets, and periodically assess the likelihood of these requirements being in place. The assessment should include historical performance and productivity and should feed into an assessment of whether the benefits of meeting a deadline under pressure outweigh the risks.

The report consolidates several actionable recommendations which can easily be reduced to a checklist and integrated with the project planning, approval, and execution processes. In fact, the checklist could be made part of the Cabinet Note or the relevant approval Note, and officials should be held accountable for deviations from its adherence without satisfactory enough explanation. Further, any deviation from the original checklist should be brought to the notice of the project approving authority for revised approval.  

There is nothing innovative or new with the recommendations. But the report consolidates all of them into one place. The challenge is to integrate them into the project processes in a manner that maximises the likelihood of compliance and makes deviation personally very costly for those responsible. Having done that, it's a matter of executing the processes and hoping that an associated culture takes hold in the organisation. There are no substitutes for such systemic solutions.