Article date 02 August 2012
Revised 26 October 2016
This article hypothesises a new railway line from Watford to Welwyn and Hertford via St Albans Abbey. The specific purpose is to discuss possibilities that would avoid the need to provide a bypass of St Albans Abbey station, since such a bypass would destroy a scenic area including views of the River Ver and would have a disruptive effect on the village of Sopwell. To avoid this disruption, we have investigated scenarios in which all trains are routed via St Albans Abbey. This article reports on our work on this matter.
Our work consists of a number of scenarios. These scenarios differ in a number of respects, including the extent to which the existing line from Watford to St Albans is upgraded. The existing line from Watford to St Albans Abbey, known as the Abbey line, is currently single track i.e. permitting only a train in the direction Watford to St Albans Abbey or in the direction St Albans Abbey to Watford, but not two trains in opposite directions at the same time. The single track line thus limits the frequency of services, currently provided by a single train running from Watford to St Albans Abbey, then returning to Watford. From one Watford departure to the next is a time interval of no less than 45 minutes. Stations on the Abbey line are : Watford Junction, Watford North, Garston, Bricket Wood, How Wood, Park Street, St Albans Abbey. All our scenarios commence from the starting point of upgrading the Abbey line to dual track from Watford Junction to How Wood, however they differ in respect to the anticipated upgrade of the line from How Wood to Park Street and from Park Street to St Albans.
Our work consisting of a number of scenarios, these differ, as we have outlined, in the extent to which the existing line from Watford to St Albans is upgraded, specifically, from How Wood to St Albans. They also differ in the extent to which St Albans Abbey station is expanded to provide additional platforms.
The current service from Watford Junction to St Albans Abbey stops at intermediate stations and takes 16 minutes. Some years ago a request was made for an official journey time from Watford Junction to St Albans Abbey non-stop. The official answer was 11 minutes. There are 5 intermediate stations on the Abbey line and the journey time for a stopping service is 16 minutes. We note that the non-stop journey time is equivalent to saving one minute for each intermediate station, thus reducing the journey time from 16 minutes to 11 minutes. This is the means by which we have approximated progress along the railway line of a non-stopping service i.e. not stopping at intermediate stations. This applies to all scenarios. In fact we allow 12 minutes for all non-stop services using the existing line, also with time allowed where necessary for trains to pass at Park Street.
There is a review at the end of this article.
The simulations referred to in this article are animations and are available here (opens in new window) : List of animations
This simulation of train services for St Albans assumes that the existing Abbey line from Watford Junction to St Albans Abbey is upgraded from the existing single track to dual track from Watford Junction until just short of How Wood. Thus the intermediate stations would also be re-designed to include two tracks and two platforms. These intermediate stations are: Watford North, Garston and Bricket Wood. At How Wood, only one platform is assumed and single track working from How Wood to Park Street is assumed to be retained in order to reduce construction costs.
The existing single track line to St Albans Abbey serving a single platform would be upgraded to provide three tracks serving three platforms. The existing platform is one of two that used to be in operation. Reinstatement of the former platform would provide us with a required second platform. We assume that the required third platform would be on the southern side of this pair, although a third platform on the northern side would provide us with more or less equivalent functionality.
The Abbey line would become dual track, served by two platforms at Watford Junction. It seems reasonable to suggest that the existing passenger foot tunnel serving the majority of existing platforms would be extended to serve the two new platforms. For this reason, we have suggested the two new platforms as forming an island pair, with a single access from these platforms (probably both steps and lift) to the passenger foot tunnel.
The existing station at Park Street is a single platform and a single track. Opposite the platform is an embankment leading down to the River Ver. In order to contain costs, we have assumed that this is unchanged. On the St Albans side of Park Street, there used to be a siding (as old photographs show). We are assuming that the line would become dual track at this point i.e. immediately beside the existing station, with a second platform provided alongside. Thus there would be two platforms at Park Street however they would both be accessed from the same side of the track. It is possible that customer access to the new platform might be via the existing station access unchanged by walking along the existing platform to reach the new platform.
Although we have not undertaken an investigation of the locality, it is perhaps possible that a new station access could be provided resulting in more direct access to the two platforms, probably by turning to the left for the new platform and to the right for the existing platform. We have not investigated this. There is also a small car park as part of the existing station access, noteworthy since there is generally a lack of car parking provision currently at stations along this line.
There is a pedestrian crossing of the Abbey line between Park Street and St Albans Abbey stations in the vicinity of Cottonmill Road.
Our simulation for this scenario shows the line from Park Street to the connecting line to the Midland Main Line (note 15 in our article Options for railway lines in the St Albans area ) as being straight. This is not accurate since there would be a curve in the track. This is corrected in a later simulation. This scenario includes a red-coloured service from St Albans City to Watford.
The "Saints link" is incorporated into this simulation, used to provide a secondary route between Park Street and St Albans Abbey.
How Wood station would remain much as it is, more or less unchanged, with a single track through the station. Immediately on the Watford side, the line would become dual track. There is a pedestrian crossing of the line beside How Wood station.
At Bricket Wood, there would be dual tracks through the station and two platforms, as there used to be. The infrastructure for this simulation includes a facility for trains to overtake, although not used.
At Garston, there would be dual tracks and two platforms. We have not investigated how this would be achieved.
At Watford North, there would be dual tracks and two platforms. We have not investigated how this would be achieved, however there is a vehicle level crossing of the existing single track line beside the station. With a likely increase in the frequency of services and the line upgraded to dual track, the level crossing would potentially become unworkable, requiring provision of a road and foot passenger bridge.
St Albans Abbey is a terminus station. Trains must change direction. In normal operation, this is achieved by the driver walking to the other end of the train. Time has to be allocated for this, whether the train is in service or not. In the simulation, the green-coloured trains to Welwyn use this normal arrangement at St Albans Abbey and six minutes are allocated for this turn round.
The significance of the different colour-coded services is documented in the timetable specific to this scenario. Both the blue-coloured Euston shuttle trains and the orange-coloured trains to Welwyn use a different arrangement, which we refer to as a driver allocation or driver sharing system. These trains change direction in three minutes. Upon arrival at St Albans Abbey, a driver is waiting on the platform and boards the train in order to continue the service, the driver who brought the train in to the station alighting. The driver who alights then walks to the other end of the appropriate platform in order to take over as the driver of a subsequently-arriving train. To minimise non-productive time, drivers are shared between the blue-coloured Euston shuttle service and the orange-coloured service to Welwyn. Under franchising arrangements, train services are provided by Train Operating Companies (TOCs). The driver allocation system implies that the blue-coloured Euston shuttle and orange-coloured service to Welwyn would be provided by the same Train Operating Company.
A limitation to services as illustrated by this simulation results from the single track section between How Wood and Park Street. In this simulation, we have assumed such a single track section in order to contain costs, the existing line between How Wood and Park Street being elevated, crossing Watling Street, with housing in the vicinity of the line and also the River Ver in proximity opposite Park Street station. In other words, this section would be relatively costly to upgrade to dual track.
Two platforms at Park Street enable services to be synchronised at Park Street, this arrangement being in order to maximise occupancy (usage) of the single track section between How Wood and Park Street.
The major drawback in this scenario is that we are not attempting to fully exploit the line to Welwyn. Only a limited number of services are running in each half-hour period whereas more services could potentially run. This leads us to our subsequent scenarios.
Our second scenario aims to more fully exploit the new railway line to Welwyn. For this purpose more platforms have been added at St Albans Abbey and in two banks. The upper bank, envisaged as located where the existing platform is to be found, contains four platforms. The lower bank contains two platforms. The original intention was that the upper bank would be used for services from Watford to Welwyn and the lower bank used for services from Welwyn to Watford. There would be a footbridge between the two banks of platforms.
When we considered driver allocation, we noted that the driver of a train from Welwyn to Watford next driving a train from Watford to Welwyn would simply cross by means of the footbridge and go to the appropriate platform. However, in the opposite direction, the driver of a train from Watford to Welwyn next driving a train from Welwyn to Watford would need to walk the length of the upper bank platform, cross by means of the footbridge then walk the length of the lower bank platform. We refer to these as a short walk (SW) and long walk (LW) respectively.
We did not find it easy to derive train timetables for this scenario. Attempting to allow a longer time for the Long Walk than for the Short Walk served only to make the challenge even greater. Rightly or wrongly, this resulted in an abandonment of the original intention of using the upper bank for services from Watford to Welwyn and the lower bank for services from Welwyn to Watford. Instead, services have been platformed pragmatically in an attempt to increase the likelihood that long walks are avoided : a driver arriving at an upper bank platform is intended to next drive a train from an upper bank platform, a driver arriving at a lower bank platform is intended to next drive a train from a lower bank platform. Drivers go to the appropriate platform, then walk the length of one platform.
A second change from our first scenario is that we have added a service from St Pancras to St Albans Abbey continuing to Welwyn. This service is coloured pink.
Our first scenario included a red-coloured service from St Albans City to Watford. In our second scenario, where we have attempted to capitalise more fully on the new railway line to Welwyn, it so happens that there are no free paths for a service from St Albans City to Watford, therefore such a (red-coloured) service is not included. Such a service could be added, but would need to be at the expense of at least one of those included and illustrated.
A reasonably-sized car park at St Albans Abbey might be incorporated into the neighbouring commercial estate, with a suitable pedestrian link to the station.
A station St Albans East has been added. We have not defined a precise location for this station, but it is intended to be approximately three minutes journey time eastwards from St Albans London Road. It is perhaps possible that this station could be located in proximity to Morrisons, with the superstore car park by arrangement also serving as a station car park, becoming a dual-storey car park if necessary. This station might potentially be useful for commuters to Euston commencing their journey by car, it being noteworthy that there is a shortage of car parking provision currently at stations on the Abbey line. It is likely that a service to St Pancras would also call at this station. This station would not necessarily be located in proximity to Morrisons.
Our first scenario consisted of a single simulation (animation). Our second scenario differs in that there are three. These three simulations are distinguished by the intended destination of the Euston shuttle. The green-coloured service to Welwyn has characteristics that differ between these simulations.
In the first simulation for this scenario, the Euston shuttle terminates at London Road. The shuttle changes direction in five minutes at St Albans Abbey, where the green-coloured service to Welwyn changes direction in six minutes and from Welwyn in seven minutes. Both of these are achieved by the driver walking to the other end of the train.
In the second simulation for this scenario, the Euston shuttle terminates at St Albans East. The shuttle changes direction in three minutes at St Albans Abbey, where the green-coloured service to Welwyn changes direction in six minutes and from Welwyn in seven minutes. When a shuttle from Watford arrives at St Albans Abbey, the driver remains in place. A second driver boards at the other end of the train then drives the train to London Road and St Albans East. Both drivers remain in place at St Albans East. The first driver then drives the return service to London Road and St Albans Abbey then alights. The second driver drives the train onwards. The driver who has alighted walks to the other end of a platform ready to board the next-arriving shuttle where the procedure recommences. There is one more driver than shuttle trains. By this means, the shuttle changes direction rapidly both at St Albans Abbey and at St Albans East.
In the third simulation for this scenario, the Euston shuttle continues to Welwyn. The shuttle changes direction in three minutes at St Albans Abbey, where the green-coloured service to Welwyn changes direction in three minutes and from Welwyn in two minutes. Drivers are allocated between the shuttle and the green-coloured service, the mechanism being as was used in our first scenario where it was used for the shuttle and the orange-coloured service.
An inspection of all three simulations for this scenario shows numerous timing dependencies between the services. Defining how recovery would occur from late running or other unwanted events, e.g. services arriving in the wrong order, would be difficult. In practice, therefore, not all services shown would run. To enable recovery from events such as late running, only a selection of the services shown would in fact run. The same applies to the third scenario, which follows.
Our third scenario is based on our second scenario. The only difference is that the line from How wood to Park Street is dual track, whereas it is single track in our second scenario. The layout at St Albans Abbey is identical.
We did not find it easy to derive train timetables for our second scenario. We started on the task of deriving train timetables for this, our third, scenario. We found that this was going to be an appreciable task, so we switched to a different approach : starting with the train times for our second scenario, we attempted to add additional services. In our second scenario, train frequencies were limited by the single track section between How Wood and Park Street. This restriction no longer applying in this scenario, we determined what additional services might run, that is, without changing the timetables from our second scenario. The result was one additional service. This additional service could only run when the Euston shuttle runs to St Albans East or continues to Welwyn. In the case of the Euston shuttle terminating at London Road, it blocked the path of the additional service.
We have, therefore, produced two simulations for this scenario. The first of these is for the shuttle running to St Albans East, and with an additional train service. The second is for the shuttle continuing to Welwyn, and with an additional service. In all other respects, these two simulations are identical to the corresponding simulations of our second scenario. The timetable for the additional service (as also platforming at St Albans Abbey) is identical in the case when the shuttle runs to St Albans East and in the case when the shuttle continues to Welwyn. The additional service is off-green-coloured. (Any inspiration concerning how to better describe the colour which we have described as off-green could be of interest.)
Our second scenario featured a design for St Albans Abbey incorporating two banks of platforms. The original intention was that the upper bank would be used for services from Watford to Welwyn and the lower bank used for services from Welwyn to Watford. However there was an issue with driver short walks and long walks.
Our approach in this scenario is to keep two banks of platforms and to place one bank of platforms above the other. The station is a dual-level station. This eliminates the issue of short walks and long walks.
At ground level, there are five platforms including for services from Welwyn to Watford. Below ground, there are four platforms which are for services from Watford to Welwyn. In this scenario, we are making a greater effort to make fuller use of the new line to Welwyn (strictly, to provide a fuller set of paths) therefore the line from How Wood to Park Street is dual track. Moreover, we are also assuming that the Abbey line from Watford is not only dual track to How Wood and to Park Street but also all the way to St Albans Abbey. At the point that Abbots Avenue crosses the line it is likely that one track would be above the other and continue in this manner from here to St Albans Abbey. We assume that the bridge that carries the A414 over the Abbey line is of sufficient span to permit dual tracks beneath it.
A dual-level station St Albans Abbey would necessitate excavation activity. Being in proximity to Roman Verulamium, time would need to be allocated for archaeological studies.
As promised earlier, we have corrected the line from Park Street to the connecting line to the Midland Main Line to show a curve in the track i.e. more accurately reflecting on-the-ground reality.
A connecting chord onto the HS2 to HS1 link line is shown. The HS2 to HS1 link line, itself, is not shown, only the connecting chord onto the HS2 to HS1 link line is shown. We have elected to connect it into the Abbey line between How Wood and Bricket Wood although we think the connection could be located on the south i.e. Watford side of Bricket Wood. Having made good use of the new line to Welwyn and having also included a service from St Pancras to Welwyn via St Albans Abbey (coloured pink), we attempted to include a service (red-coloured) from St Albans City. We found that the only paths available necessitated tight timings : from St Albans City, the service has to fit in with a pink-coloured service from St Pancras in order to make use of a free path at Park Street. In the other direction, to St Albans City, the only path available necessitates passing the shuttle to Euston on the right at Park Street and the timing of this service, running in front of the shuttle from Euston, means that this train could not have arrived at Park Street from Watford : to arrive just in front of the shuttle at Park Street from Watford would imply overtaking the all-stations-stop shuttle somewhere between Watford and Bricket Wood. Instead, we have elected to consider this red-coloured service as connecting onto the Abbey line via the connecting chord from the HS2 to HS1 link line. Logically therefore the service in the other direction uses the connecting chord from the Abbey line onto the HS2 to HS1 link line.
We note that the timings of the only available paths for a service via St Albans City are tight. In practice, these paths might not be used at all and so services via St Albans City risk being at the expense of other services (via Park Street) to Welwyn i.e. there is a trade-off between the two. This would seem to be a natural implication of the infrastructure. It may be that there would be advantages if a connecting line from the Midland Main Line connected to the Abbey line north of How Wood in place of north of Park Street (i.e. note 5 in place of note 15 in our article Options for railway lines in the St Albans area ) however we are aware that a Midland Main Line connection to the Abbey line north of How Wood might be obstructed by the proposed Strategic Rail Freight Interchange at Radlett.
It seems to make little difference in terms of path availability whether the connecting chord from the HS2 to HS1 link line connects onto the Abbey line to the north of Bricket Wood or to the south of Bricket Wood, the only difference being whether Bricket Wood might be a station stop.
We have clarified which junctions are not flat. The Midland Main Line crosses the line to Welwyn by means of a bridge (already in existence and what a fine bridge it is). The line from Park Street and How Wood onto the connecting chord to the HS2 to HS1 link line is shown as crossing over (although it could be under) the line from Bricket Wood to How Wood : indeed we certainly think this should not be a flat junction.
This scenario incorporates a 4 trains per hour service coloured in yellow, demonstrating that we have plenty of paths available on the line to Welwyn. In other words, we have a design for St Albans Abbey that enables good use to be made of the line to Welwyn. We found it much easier to derive a timetable for this scenario than we did for the second scenario.
The shuttle is illustrated as running to Welwyn. Because there are fewer challenges in deriving a timetable for this scenario than in our second scenario, little needs to change if the shuttle runs to St Albans East or to London Road instead of to Welwyn. Hence there is only one simulation for this scenario.
Our article New rail services suggests an assignment of train colours to train services. Potentially only the green and blue services would be routed via Watford Junction, the orange, yellow and off-green services running instead via the connecting chord from the HS2 to HS1 link onto the Abbey line. This makes no difference to the animation in that the purpose of the animation is to show timings at St Albans Abbey and which are not affected by whether services run via Watford Junction or via the HS2 to HS1 link line. Earlier we demonstrated that the red-coloured service was by choice a service via the HS2 to HS1 link line in order to avoid having to overtake the Euston shuttle. Other services via the HS2 to HS1 link line instead of via Watford Junction do not affect the timings of this scenario.
We have kept unchanged the times of the Euston shuttle, passing at Park Street and so necessitating two platforms at Park Street. However this was designed to maximise usage of a single track section between Park Street and How Wood and which does not apply in this scenario. Potentially therefore there is a possibility of re-timing the Euston shuttle and so also all other services such that only a single platform is required at Park Street. Potentially, for example, a shuttle to Euston might depart from St Albans Abbey at the same time as a shuttle from Euston arrives at St Albans Abbey. This might form the basis of a further scenario. Currently it is left as a possibility for the future.
All scenarios should be regarded as no more than introductions, designed to outline various possibilities.
We note that Park Street station complicates matters, is little-used and is in proximity to How Wood station.
The terrain from Watford Junction to How Wood station is reasonably flat. Upgrading much of the line from the current single track (i.e. permitting only a train in the direction Watford to St Albans Abbey or in the direction St Albans Abbey to Watford, but not two trains in opposite directions at the same time) to dual track (i.e. permitting trains in each direction at the same time) would seem to be feasible. In contrast, from the St Albans side of How Wood station until the approach to St Albans Abbey station the terrain is far from flat : the line is on an embankment, then in cutting (in the vicinity of Abbots Avenue, St Albans) prior to returning to flat terrain for the approach to St Albans Abbey. To limit costs, it is perhaps better to assume that no upgrade to dual track would occur from the St Albans side of How Wood station until the approach to St Albans Abbey station. This has the effect of constraining how we might envisage an enhanced railway infrastructure.
We have assumed that the bridge that carries the A414 over the Abbey line is of sufficient span to permit dual tracks beneath it. However, as stated in the previous paragraph, we might envisage scenarios in which no upgrade to dual track would occur from the St Albans side of How Wood station until the approach to St Albans Abbey station. If Park Street station remained open, there would only be one platform.
Note 20 of our article Options for railway lines in the St Albans area documents an alternative possible connection on to the Midland Main Line that is not currently incorporated into any scenarios in this article.
All scenarios assume conventional rail. However, it is possible that a line from Watford and St Albans to Welwyn and Hertford might be constructed solely for use by light rail.