Francis Paonessa reflects on his time at Network Rail
When Dr Francis Paonessa joined Network Rail in the summer of 2015 from Bombardier Transportation, where he had been managing director of the train manufacturing and servicing business in the UK, he did so to head up Infrastructure Projects (IP) – the national organisation charged with delivering major enhancements on the railway network.
Now, with new CEO Andrew Haines pushing forward with a policy of devolution and decentralisation, the work of Infrastructure Projects will be undertaken by the five regions and 13 routes that are being set up. As a result, Francis left Network Rail at Easter.
Before he left, he took the time to meet with Rail Engineer at the company’s offices in Derby to look back at his five years in railway infrastructure, Infrastructure Projects’ achievements in that time, and his personal experiences.
On the face of it, Francis’ experience at Bombardier, and before that at naval shipbuilder Vosper Thornycroft, had got him used to major manufacturing on fixed sites. The railway is not manufacturing, and it takes place on very long, thin sites, in the open, all over the country. On the face of it, a very different experience.
“Actually, I was expecting a lot of similarities,” Francis demurred. “The theme that ran through ship building, at Bombardier and here is that of large-project management and of engineering-led, low-volume manufacture.
“Ship building was the ultimate low volume manufacturer for the complexity of the design that you did. It was a massive design effort that resulted in the construction of one, two or three ships.
“Bombardier was very similar, with complex train integration and system integration, but for me it was a dream of then having maybe 100 of them to build. That was big volume and there was a lot of opportunity for continuous improvement that was much more difficult to do in ship building.
“Coming to Network Rail, I was expecting a large project management system, an integration-type business, but again, ultimately, very low volume manufacture. It’s an interesting mix of big projects like Thameslink or Birmingham New Street or some of the other stations where they are true one-offs, together with a lot of repetitive work where bringing my manufacturing experience in really helped.
“Track is very repetitive, and, actually, so is electrification more than anything – a logistics exercise rather than a complex design. There is nothing in electrification that is difficult in itself – it’s the amount of it that is difficult, and the integration with the railway and getting access to it, and working with the local community when you’re raising bridges, shutting bridges and demolishing bridges. There’s a lot of similarity, but also some things which are quite different.
“What I did find surprising was that so much of the enhancements were still in the development phase. That was not what I was expecting. It was fine up until reclassification, when the accounting process that we had put in place to cope with the fact that so much was in development was no longer fit for purpose. It switched life from being ‘Is it value for money? Is it efficient spend?’ to ‘Is it within budget?’ which sounds similar but is a fundamentally different approach.
“The railway had to switch from a whole approach, and a whole CP5, of ‘Is it efficient spend?’ to ‘There is a finite budget and you’ve got to design within it’. We were not prepared for that significant change. In fact, when I arrived – and that was only a couple of months before we reclassified – the conversation was more about borrowing money than it was about recognising what a fundamental shift that would make to the whole approach to the business.”
Reclassification caught many people out. When Network Rail became an arms-length central government body, it was thought that this was purely an accounting exercise and everything would stay much the same. It wasn’t, and it didn’t.
Previously, Network Rail had to plan several years ahead when it applied for government funding, which was provided on a five-year cycle. Some projects, particularly those which would be carried out over a long period of time, or which would be started late in a control period, were therefore costed at very early stages of the project, often before the design was complete or work such as ground surveys had been undertaken.
This was not especially Network Rail’s fault. Money needed for a project scheduled to take place in the third year of a control period had to be applied for up to two years before the start of that period, or five years before construction would take place. So much of the cost calculation was based on a very incomplete design.
As a result, the costings were often wrong, sometimes wildly so, and almost always underestimated. But that didn’t matter too much as Network Rail could borrow the shortfall by issuing its own bonds and other measures.
With reclassification, that all stopped. No longer could Network Rail borrow extra capital, it was to be on a fixed budget which would come direct from government.
Which left Network Rail in a pickle. While many projects were running well, and financially under control, several of the ‘early estimate’ projects were substantially underfunded, but still had to be completed. They were planned and costed under one regime, but were now to be financed under a new and different regime, and that would be impossible.
Network Rail chairman Sir Peter Hendy was asked to conduct a thorough review of the enhancement programme in England and Wales to see what could be delivered in an affordable and timely way within the funding period to 2019.
Hendy’s proposal was to complete all of the work that had been agreed, but over a longer timescale. So, the money allocated for the period up to 2019 (in CP5) would then fund a reduced amount of work, with the remaining projects funded from a new budget for CP6 (2019-2024).
As part of this exercise, Francis and his team were asked to reduce costs as much as possible.
“Firstly, we didn’t re-estimate the projects,” Francis explained. “There wasn’t the time to do that as part of the Hendy review. That’s a big misconception that people have, they think that Hendy was a whole reset of the process. It wasn’t. We used the estimates that we had at the time. What we presented to Peter, and then Peter ultimately recommended to the Department for Transport, was the prioritisation of the projects and the likelihood of them being delivered on time with the schedule that went with them.
“When you added it all up, we had £193 million worth of extra stuff to deliver than we had money for. So, we had to value engineer £193 million worth out.”
This figure was not about slippage, it was a genuine need to reduce the cost of current projects by £193 million. In addition, Sir Peter Hendy had assumed that about 10 weeks’ work, or £500 million, would slip into CP6 from CP5. But the £193 million is on top of that.
“It was interesting,” Francis recalled. “We classed value engineering as whether we could deliver the same passenger outputs or benefits for less. That drove a whole lot of innovation.
“In the end, we delivered about £240 million of value engineering, so we came out on the right side.
“Because we didn’t re-estimate Hendy, we worked with the DfT and University College London who did a whole piece of work looking at estimating of early rail projects to cope for optimism bias. If you apply those optimism bias percentages, which we’ve now included in later estimates to the Hendy portfolio, it would have added about another £1 billion to the cost. We had this massively significant headwind buried within the budget that wasn’t re-estimated.
“When we look at our value engineering, we also then delivered, on top of the £240 million, another £500 million of value engineering which was offset by headwinds.”
The team also looked into Cause Association Modelling (CAM). This looks at the links between risks, and how one risk can affect another and so skew the risk profile. The calculation was that, across the portfolio of work, an extra £300 million was needed to cover those risks.
“We spent the £300 million – we saved £500 million,” Francis claimed. “This offset headwinds in the portfolio and then we saved another £240 million through value engineering, which kept us within the Hendy budget. Having now just gone through the end of CP5, we’ve delivered the £15.3 billion of Hendy within a fraction of one per cent of the budget we set out over four years ago which, personally, I think is hugely impressive, given where we started.”
Back to renewals
In CP3 and CP4, work was concentrated mostly on renewals. CP5 saw a large upswell in the amount of enhancements taking place. Now, for CP6, the emphasis seems to be back on renewals.
“There’s still a lot of enhancement spend in CP6, let’s be clear,” Francis commented. “There’s about £10.5 billion that is inherent within the funding that we’ve got to deliver within CP6, which is still an enormous amount of money to go and spend on enhancements.
“But I don’t think we’ll see another control period like CP5. £18.7 billion invested in enhancements and £14.4 billion worth of renewals within Network Rail over the last five years. I think Nichols worked out we delivered 22 per cent of all UK infrastructure. That’s a lot for one organisation to do, particularly when we have the constraints of access to the railway.
“You asked at the beginning about the big difference between working in ship building and working in Bombardier, it’s ten times easier to work in a fixed site where you don’t have to break the work up into 40,000 chunks. We take about 40,000 possessions a year to deliver our work, that’s 40,000 opportunities to get it wrong and to either overrun or put the travelling public at risk.
“It’s very challenging to do and it’s totally unlike any other business I’ve been in. With a brownfield site, and we’ve been fortunate to work on some of those, Stafford was a really good example – where the vast majority of the work was off railway and the only railway interfaces were the turn off, then it’s a very different way of working and a very different way of thinking because you’ve got 24/7 to do it in. You haven’t got to wait for the last train to go by, take the possession, get onto the railway, transit to where you want to work and then get back off again.
“I think our average possession time is 5 hours and 35 minutes, it’s not a lot of time, given the demands and the expectation of handing back on time and handing back safely, but we are now managing to fit 38 per cent more spend into each possession minute than we were at the start of the control period.”
Reliability and contingencies
One of the other improvements that have come about over the last five years is in the reliability of plant, and the reduction in overruns and additional costs they can cause.
“The track team has put an enormous amount of effort in with the equipment providers to drive down our RRV failures,” Francis commented. “If an RRV fails in the dig, you’re really stuck because it’s not an easy piece of equipment to fix. It’s sat in the way and trying to get a 30-tonne bit of kit out of a hole isn’t that easy when you’ve got engineering trains on one side and you’re trying to work. And if you’ve only allowed for four hours of contingency time, you’re hard up against it with time.
“Typically, at the bank holidays, we lose about half a percent of the work because we don’t start it. It may be weather driven, usually, or there’s been some operational incident or a late running train. Whatever it might be, we’ll be saying we’ve passed the cut-off point in our contingency planning and we won’t start.
“Then we lose about another half a percent of work in trying to guarantee on-time hand back. That’s been pretty consistent now for the last three years or so. The reason it’s not turning into overruns is that the teams have been far better at planning their work in a very structured way so that we can get the core bit of work done.
“Welding and stressing might be a really good example for a track job. If we’re planning to hand it back at 50mph, we might not do the welding and stressing at the end. We might do it in midweek nights.
“We can get the track in the ground, we can get it plated and we can get it tamped and we can usually hand back at 50mph. Typically, within a week, we will have handed that back at the line speed. It’s that structuring of the way in which we do the work that means that we can cope with these failures. Otherwise, the only real way of doing it is to put more contingency time at the end and that way you just get less work done.
“Of course, there are some jobs where you just can’t do that. Bridge renewals, we’ve got two bridge renewals coming up at Easter so we will be taking off the track, taking out the bridge, rebuilding the bridge, putting the track back in the four-day period. You can’t do half the bridge. Things like bridges are particularly challenging because once you’ve started you’ve passed the point of no return.
“For those, we don’t have a bigger contingency, we just have a lot more planning. The basic problem with a bridge is we’ve got four days over a bank holiday, a weekend if they’re smaller ones. We’ve got the time we’ve got and we just have to make sure that we sequence the work in the way that allows us to get that done.
“Where we might build some more time in is if, say, it’s being delivered by road crane, which can suffer from wind issues, but we’re delivering more now by self-propelled modular transporters, so we don’t often have the wind issues. We don’t have to put that contingency in, but we might still have welding and stressing at the end. We might have hand rails, guard rails and all sorts of bits that we might be able to curtail.
“Frequently, we might have a four hours float and four hours of extra contingency and that, statistically over the last four years, has given us a very secure hand-back profile. The problem is, there is absolutely no guarantee that the time that we’ve got will cope with everything and that’s why we still have a few small overruns during the year.
“It isn’t luck we’ve ended up where we are. There has been a whole host of learning those lessons, process improvement, sharing of best practice right across the organisation that the teams have embraced and done a fantastic job in embedding and getting us to the point now where we have a far, far more dependable delivery.
“Also, our visibility going forward of where issues are, is far better. We have now actually got time to fix problems before they materialise into a possession overrun. Is that perfect? No, it’s not and it’s fairly difficult delivering the 11,500 projects that we have and £5.5 billion worth of stuff that we’ve done just this year, to sit and really shout that from the roof tops because we are not a perfect organisation.”
The real heroes
“Celebrating success, and particularly celebrating success with the projects that are the big flagship ones, is one of the big leadership challenges within the organisation, because it’s the teams that deliver renewals that are real heroes,” said Francis.
“They are in frequently very small teams. They might be delivering multiple small projects themselves. They’re doing a great job with the supply chain in delivering those, but they’re not London Bridge. Renewals are the vast majority of the possessions that we take and the teams deliver them, time and time again faultlessly, and people don’t notice. We renewed 300km of track last year and no one noticed. The signalling teams renew and replace hundreds of SEUs (signalling equivalent units) and people don’t really notice. The new signalling scheme, they don’t notice it having been re-locked to a different location or we’ve closed signal boxes or we’ve modernised – all that stuff just happens.
“Passengers are not interested, and rightly they shouldn’t be, but making sure the narrative within the business equally recognises that the teams that we’ve got working incredibly hard, delivering those far less exciting visual renewals as the one or two flagship projects that we might have ongoing, is an important balance.
“Of course, the external narrative rightly tends to focus on the one or two projects that don’t go as well as we’d hope. Because of the negative focus we get when it goes wrong, keeping those teams motivated who are in the spotlight working, day after day, on really difficult engineering projects, is challenging. Huge credit to the teams who get up every morning and come in and put their best work in on a project that’s difficult and, because of where it is, it’s frankly, no matter what they do, never going to be seen as a success.”
Failure and success
When Francis joined Network Rail in 2014, it was three months after reclassification had been announced but two months before it was implemented.
Six months later, at Christmas 2014, there were the well-publicised overruns at King’s Cross and Paddington that so-badly damaged Network Rail’s reputation. It was a real baptism of fire.
“Joining an organisation and three months in to find that you are forecasting to be several billion pounds overspent already three months into a new control period, is not good,” he remembered. “To then badly let customers down at Christmas was really not good. You suddenly find that you’ve got an organisation, in the broadest sense, of people who are trying their absolute hardest to do a good job, but aren’t delivering on that. At the same time, you’ve got a phenomenal workload to deliver. To transform an organisation whilst having a phenomenal amount to deliver is really difficult.
“Probably my only regret is that we’ve now got to a point where I’m extremely proud of the organisation, the capability that sits within it, the detail of the planning which underpins CP6, and it would have been fantastic to run that machine at the plans that we’ve got for CP6 for two years and absolutely smash it out of the park. I sadly won’t have the opportunity to do that, but the new regional managing directors will with the team that we’ve built.”
The devolution of IP, being promoted by new CEO Andrew Haines, isn’t, in Francis’ opinion, going to be as much of a change as people think.
“One thing to clarify, IP isn’t a centralised structure,” he explained. “Because we’re a national function, there’s quite a lingering perception that we’re a team that sits in the middle somewhere and does stuff across the country. We’re far from that. There are seven delivery teams – four regions, northern programmes, track and signalling, supported by five functions. 96 per cent of the people in IP sit in those delivery legs. About 186 people sit in the functions. We are probably the most devolved part of Network Rail already and the teams sit geographically dispersed.
“We’re sitting here today (in Derby) and part of my signalling design team are sitting downstairs and my East Midlands regional enhancements team are 800 yards down the road, which shows how widely 4,800 people in IP are geographically distributed.
“When we move into a devolved structure with the routes, a large percentage of people in IP won’t be moving anywhere. They will still be sat at the same desk in the same location next to the same people doing the same job but working within a new regional structure. What we’re doing is changing the accountability, and particularly the responsibility for the delivery activity, from it being a national function to a devolved regional structure. Accountability for enhancements and renewals will ultimately rest with the new regional managing directors.
Francis departed Network Rail, not before Easter, but immediately afterwards, having visited work sites and encouraged staff on his last day. He says he has no future plans at present, but no doubt he will pop up somewhere before too long.
Rail Engineer wishes him well in the meantime.