How Network Rail Plans Train Timetables - EMF 2026

[applause] Hello. Can everyone hear me? Yes. Yes, you can hear me. Excellent. That's always the first thing to check. Good. Who here likes trains? >> Excellent. You've come to the right talk then. So this is how I've got it up on the screens up there. How network rail plans trains or why does the 204 train from Cardiff to Chester go early. So my name is Charles. I am a operational planning specialist at Network Rail. That is my official job title. What that means in practice is I am one of the people who puts train timets together and for Network Rail. So, as you can see, I've I love problem solving and puzzles, which is great because train timetable validation is one great big jigsaw puzzle. And it's perfect job for me because I have loved trains my entire life. So, enough about me. You're not here to talk about me here about me. What is it that I do? So, I work for Network Rail. Do we have someone here from network rail? >> Yeah, we do. Excellent. I I've met one or two other people, rail industry people at EMF already this weekend. So, good to know I'm not alone here. So, for people who don't know so much about the rail industry, there's lots of different companies involved in it. So, Network Rail run all of the own all of the track, the infrastructure. What we mean by that is the track, signal boxes, bridges, tunnels, vioaducts, etc. And it's their job to maintain all of that. Then you have the train operating companies who run the actual trains. So start off with the main train operating companies, which are the public sector, publicly funded ones, the talks. This is all changing in the near to medium future as GBR becomes a thing. But at the moment, the structure of the industry is that you still have lots of separate organizations all running the different trains. Some of them are publicly owned, some of them are still privately owned. It's a mixture. And they are all funded by the different governments. Most of them by the Department for Transport, but in other parts of the country, they are devolved to Wales, Scotland, and also London and Mury Rail as well. And those that the train operating companies are the brands that you will see if you get on board a train. They will be brands that are familiar to you. So GWR, Cross Country, companies like that. The other group of operators we have on the network are the open access operators. These are private companies that are completely privately funded. They don't get whereas the department for transport and the devolved governments give very specific instructions to all the talks about what they want them to run. The open access operators look for gaps in the market where they think they can see a gap that the talks aren't filling and then if they can get space on the network they try to run trains to fill those gaps to make a profit. So for them we have Grand Central, whole trains, Lumo some of you probably heard of and Heathro Express is technically an open access operator as well. Then final group of operators on the network are the freight companies. So these are completely private essentially exactly the same model as the open access operators. They are completely privately run completely privately funded and they compete against each other to get contracts with different um customers. So they take things like aggregates, steel, intermodal containers, lots of different types of freight that run and they don't get any private funding at all. Uh sorry, any they're completely privately funded. Don't get any public funding at all. So, we've got all these different companies that all want to run trains on the same set of tracks owned by Network Rail. So, what they do is they do a thing called bidding for paths on the network and then Network Rail's job is to take all of the bids from all of the different companies and put them together. So, what is a bid? Well, first of all, it doesn't involve money, at least not at this stage. What it involves is them submitting their time t their preferred timetable to network rail. That network rail to see if they can get what they want from network rail. So there are two timetable changes a year, one in May, one in December. This is when what we call the base timetable changes, also known as the working timetable. This is the timetable before it has had any kind of short-term amendments which are done for things like engineering work, special events and so on. So twice a year about 9 months before the timetable change, all of the operators submit a complete bid to network rail which is a complete set of all of the timets that they want to run for all of the schedules that they want to have paths for. Network rail then has about 12 weeks in which to validate the bids. Now, what that means is taking all of the schedules that all of the different operators want to run, putting them all together, seeing where they work, seeing where they don't work, where they don't work, changing them, and then once they've done that, we offer them back to the toxin fox. At that point, there is a little bit further back and forth where they can say if they do or don't like changes that we've made to them. Um, and we keep talking to them throughout the whole process. So it's not so the the mantra is no surprises on offer day. So we try to keep them in the loop discuss with them what they're do what we're doing so that it's not a complete surprise what we've been doing to their trains when it comes to offer day because if you try to offer them something that doesn't work for them that's no good. So what are we actually doing? So I said that all of the train operators submit their complete timets to network rail. So if you think of a timet, you probably think of something like this. So imagine you were you were going to EMF and you wanted to come here by train. Some of you may have actually done this. You'd probably look up something like train line. Other journey planning apps are available. Um and you'd get a screen that looks a little bit like that. That's a journey planning screen. departures. You can pick your outward and return. Fine. Click on the train. You can see all the places it stops at. Fine. Nice and straightforward. This is what it looks like to network rail. It's got a complete list in gray on the right of all the places that the train is passing as well as the places that it's stopping. So what we do is we take not just the places where the trains stop but every point that it passes on its journey so that we can see how it's interacting with other trains at that point in time and then we plot it on a graph. So what we do is we put time on the bottom, the x- axis to scale. And then we put all the locations that we want to plan our train at on the y- axis. And then we do something like let's say we want to have a train that's going from Houston to Birmingham, stopping at all of these stations. Put something on the graph that looks a bit like that. And that is a train planning graph. Nice and simple, nice and straightforward. Got a train on there. When this graph is sloped, it's moving. Steepness of the line shows you how fast the train is moving. Speed is distance over time. When the line is flat, that's when it's stationary. Stop moving. Hopefully, that's nice and straightforward. And you can put a second train on the same graph. And you can see that if you have a second train that leaves Houston a bit later and arrives at Birmingham first, goes straight through the first train roughly halfway between Milton Kees and Rugby. So that's only going to work if you've got two sets of tracks and they can go past each other. If you've got two pairs, if you got two trains on the same track, that's not going to work. They're going to clash with it. crashes into each other. Just about batteries stopped working. Excuse me a second. Just about there. I'm going to take two seconds to just see if I can fix this, but Yeah, it's working. >> Excellent. Thank you. So, here's a timetable graph as as uh as shown on the train planning software TPS that Network Rail uses. And this is pretty much exactly the same as the first one. This is the West Coast mainline out of London, Houston, heading off towards Birmingham. I've only put as far as about Watford Junction on here because any more than that, you wouldn't fit it all on. And let's just take a look at what we can see here. We've got different lines. Each line is a train. Different colors represent different sets of tracks. So, West Coast Mainline, you've got four tracks, two in each direction. So, the green lines and the red lines are on different pairs of tracks. On this particular graph, the green lines are the fast lines and the red lines are the slow lines with stopping trains. The colors aren't particularly meaningful. They're fairly arbitrary, but and then where you've got trains that appear and disappear on the graph, that's where you've got trains turning onto that route at junctions or turning off that route at junctions. So, what we do is we look at these graphs and check if they work. So, what are we checking for? The first thing we're checking for is headways. Headways are the time between two consecutive trains. That is the minimum time that is the minimum time after one train departs before a second train departs when the second train is going to get green signals. So the way the signaling system generally works is that the track is split up into what we call signal sections. You can only ever have one train in one signal section at a time. Red signal in front of the first train means that the first train is protected by it. Yellow signal behind that warns the driver of the second train. If they get to that signal that the next signal after that is going to be at red, they need to start slowing down. Signal behind that is green. Sometimes you get more more than one yellow signal, but that's the kind of the most simple model. Most of the signaling does work basically like that. So with the headways, what we want to do is allow enough time for the first train to get far enough ahead that when the second train departs, it's going to get green signals all the way through its journey. And then we have a document called the train planning rules, timetable planning rules. This document lists all of the headways and all of the other rules that we use to check if two trains are valid. So the headways between two trains are different on different sections of track depending on the signaling. So that's a little screenshot from the line between Dutert and Oxford. And if you can read it, it says between Duta and Wolver 3 minutes headway or 4 minutes if you're following a stopping passenger train. Beyond Wolver as far as Hayford, it's a six-minute headway. That's because the signals there are further apart. There's less capacity on that section of line. If that were a more intensively used section of line, there might be a case for upgrading the signaling to reduce the headways to allow more trains per hour through that section of line. But we fund the signaling to the degree that we actually want to fit trains through it. So when you're looking at the graph and checking if the two trains are far enough apart to have sufficient headway, you also need to check the timetable planning rules to see what the headway is in that particular location. And that is how you look at the headway on a graph. Two trains, gap between them is the headway. Junction margins. Probably the second most common thing after headway that we're checking for. If you imagine driving a car and you want to turn right, you need to wait for a gap in the oncoming traffic before you can turn right. That's a junction margin. And it's exactly the same with trains. That is Dartford Junction in Kent pictured there. Templink train going over it. I bet someone knows what class of train it is. Um when we're planning the timetable, we need to say if a train is going to be going over Dartford West Junction at 11:07, then if you need 3 minutes after that, you can't plan another one to go past until 11:10. Say I can't remember what the junction margin actually is at Dart West Junction, which slightly spoils that. 3 minutes is typical. For some junctions it's longer, four or five minutes. Some it's more. And that is what a junction margin would look like on a graph roughly. One train turning off, another train passing in the opposite direction, gap between them at the point where one is turning off. That's your junction margin. So, let's put all this together and look at a real world example. I'm hoping I haven't lost you so far. So, we've got another graph. I'm getting whoops for the graphs. I'm liking graphs. Next time I do this, I'll put in more graphs. Woo! Yeah, whoop for the graph. [laughter] So this is the line between Heraford and Shrewsbury and it's mostly trains for transport for Wales trains that run up here but we also have various freight trains. Now this is what the graph looked like when all of the Tox and Fox bid the December 2024 timetable. So screenshot from a couple of years ago and as is usually the case with most timetable changes, it was mostly a rollover from the previous timetable which is why it mostly works. Transport for Wales didn't want to make many changes. Most of their trains are in the same place. Most of the freight was in the same place. Most timetable changes, unless you're doing a big recast, in which case there's usually pre-work involved, it's usually a case of just validating what has been changed from the previous timetable because people like continuity. They like to keep things the same where there's no need to change them. However, they always do make some tweaks. And on this one, can anyone see where there is a problem on that graph? Bit of a challenge. Maybe a bit small to see. >> It's probably an unfair question. So, >> I think that is a junction. I think that is fine. Sometimes the graphs, you can't actually tell what the trains are doing unless you click on them and look at the individual schedules because of where the junctions are. The graphs aren't always brilliant at visualizing it. is the green one. >> If anyone was noticing that bit, it's where we've got two trains going one through the other. So, we've got two trains on the graph, both leaving at about the same sort of time, getting closer and closer together until the lines cross over each other. That just does not work. So, one of these was a passenger train, the other one was a freight train. And the freight train was biz to have a long dwell outside Shrewsbury station to wait for in the next bit of its path. I think this one was going to Pennyford Cement Works if I remember correctly. So what do you do? Well, you could tell one of the operators that they're not going to get their train. Sorry, you can't run that freight train. Or sorry, you can't run that passenger train. That is an option. Sometimes there is no solution. and then you have to reject one of the trains. But if at all possible, you try to avoid doing that. So here's what I did. I put the freight train into a loop. At Wuerton, we have a a loop, which is a little bit of track that this that the freight train could recess into, and it could go in there, sit there for however long it was, about an hour and a half, and then carry on its way to Shrewsbury. And the long dwell that it had at Shrewsbury is removed and it carries on and it picks up its path. Great. Mostly great. You can't see it on there because so I don't think it's fair to ask you, but there is a there is still a problem with this. So when I was looking for solutions, I was thinking, okay, I can put the freight train into the loop. That's fine. But then I need to get it out again. If we go back to the previous one, I'm thinking, how long do I need to keep it in that loop until I get it out again? Well, I need a big enough gap on the graph for the freight train to come out slowly. And it's a slow big, heavy, slow freight train. Freight trains are slower than passenger trains generally, particularly when they're starting from a standing start. We're going to need a big wide gap to fit it into. There's a reasonably wide gap in the graph, but it's not quite wide enough. Unfortunately, that freight train, if it came out after the passenger train in front of it just passed it, it would get to Shrewsbury not quite in time before the passenger train behind it caught up with it. So, the only thing to do there was to rettime the passenger train that was in front of it. So, I looked at that and I thought, okay, this is the 14, this is the 422 from Cardiff to Chester. What can we do with it? Can it have a shorter dwell somewhere? No, it can't. It's got minimum dwells at all the stations throughout its journey. If I want it to go ear, can the freight train go later? No, it can't because it bumps into the next train behind it and it doesn't pick up its path after that. Can this passenger train go earlier? Yes, it can. But it's going to have to go earlier all the way from Cardiff. So, what Transport for Wales had bid for was a nice hourly service going just after 20 minutes past the hour. But to fix that, I had to change it so that the train that goes after 4:00 goes at 17 minutes past the hour instead. So you've got your 3:24, what was the 422, the 523, the 622? Sorry. Yeah, the 622. So if you're someone who lives in Cardiff, likes going to Shrewsbury, you know, great, there's always a train just after 20 the hour. Turn up at 20 4. Oh, wait. I've just missed it. Why does that one go earlier? Well, it's because of that freight train. So, I sent that back to Transport for Wales and they said, "hm, we don't like that. Can you look at something else?" And a couple of us looked at it and we said, "No, some sometimes it'd be like that. There's no that is the least worst solution overall." They still get their passenger train. If you can read the tiny writing, it takes a little bit longer than the other trains in their timetable to get there. It's about 5 minutes longer, but the freight train can still run. It gets a path. The passenger train can still run. It gets a path roughly the same time as the other trains. It's the least worst solution overall. So that was what we went with. Offered it to Transport for Wales and they said, "Yeah, okay. That's fine. We'll accept it." And if you look at the if you try to look at that today, it looks like that today. Thank you for coming to my talk. [applause and cheering]