1. Station cost
AM’s immediately drew attention to the huge disparity between the CSAG cost estimate for a 2 platform station of £350 k and the Network Rail ballpark figure of £5 m. The main reason for this huge disparity is the different scale of the stations considered. I would like to base the discussion on the cost of the new station at Llanharan, which is currently under construction for a cost of £4.3 million. I obtained my information from the Welsh Assembly Government website at http://new.wales.gov.uk /topics/transport/PublicTransport/Rail/LlanharanStation/?lang=en.
Llanharan station is to consist of 2 platforms 4 carriages long, linked by a footbridge with long ramps suitable for wheelchair users, and a 54 bay car park. On the other hand, CSAG is proposing a station with platforms only 10 metre long and no footbridge. If it is assumed that the footbridge with the long access ramps costs £1 m, then the rest of the station costs £3.3 m. It can be argued that most of the remaining costs vary pro-rata as the platform length, because the fencing and lighting along the platform will vary in this way, and items such as shelters and the car park will coarsely sized in proportion to platform length. Thus, assuming that the length of the platforms at Llanharan is 95 metres, the cost of the 10 m platforms at Carno may be roughly estimated at £3.3/9.5 = £0.35 million, the same as our original figure.
Conclusion: The difference between the £350 k and £5 m cost estimates lies almost entirely in the different station specifications considered – that is the platform length and the presence or absence of a DDA compliant footbridge. CSAG therefore submit that the discussion can be restricted to the question of what platform length is actually needed, which fundamentally reduces to a safety issue.
2. Safety of a short platform station
Mt Ian Baxter explained that there was a safety risk associated with short platforms, because the conductor-guard might open all the doors by mistake, and someone might try to alight from one of the doors away from the platform. I thought it would be useful to estimate how many accidents might arise from this scenario.
As members of the committee are probably aware, it is only within the last couple of years that slam-door trains have ceased to run on the national network. Prior to that, passengers on such trains were free to open the door when the train was moving, let alone stationary, and several accidents resulted every year.
I have obtained summary statistics of railway passenger accidents for the period 1996/7 – 2000/1 from Appendix 1 “Key safety facts” in the “Annual Safety Performance Report 2000/1” which is available at the Rail Standards and Safety Website, http://www.rssb.co.uk/. This includes the number of fatalities and major injuries sustained as a result of accidents occurring when alighting from trains or falling out of carriages, which average 1 and 25.6 respectively each year over the five year period.
If we pessimistically assume that all these accidents occurred when passengers were trying to leave slam-door trains at stations when their carriage was not at the platform, then we can estimate the number of accidents that might occur at the short platform at Carno. It is also necessary to assume what proportion of trains had slam-doors as opposed to sliding doors during this period, and I am conservatively taking this figure as 25%.
Taking the total passenger kilometres travelled per annum as 35.7 billion, the annual number of passenger kilometres travelled in slam door trains comes to 8.9 billion.
Carno station is predicted to generate 400,000 extra passenger kilometres per year, so the expected number of accidents per annum due trying to alight away from Carno station platform would be 26.6(400,000/8.9 x109) = 0.0012 if it were served by slam-door trains. In fact the station would be served by sliding door trains, so if is pessimistically assumed that the conductor opens all the doors by mistake once in 50 stops on average, the number of accidents per annum reduces to 0.000024, or once every 42,000 years. The calculation is set out in the attached spreadsheet entitled “Carno short platform accidents 1”
This predicted accident rate of 0.000024 due to the short platform should be compared with the predicted reduction in road accidents if Carno station were to re-open of 0.176 per year – see the attached spreadsheet “Environmental benefits 1 Rev A”, which has not been submitted before. Note that the difference between 0.000024 and 0.176 is a ratio of about 7000.
Based on the difference in cost between a 4 carriage long and a 10 metre long platform of about £1.3 million pounds, the cost of preventing a death or serious injury over a 100 year period by building a 4 carriage long platform comes to £1.3/(100 x 0.000024) = £542 million and the cost of preventing a death over the 100 year period is £542 x 26.6 = £14,408 million! Clearly the £1.3 m would be better spent in improving road safety, and it would be criminal to waste it on making an already very safe railway even safer.
Conclusion: The anticipated frequency of alighting accidents due to the short platform is once every 42,000 years on pessimistic assumptions about conductor error. It is concluded that the road safety benefits of having a station vastly outweigh the tiny reduction in rail safety resulting from having a short platform rather than a full length one, which is likely to be prohibitively expensive.
3. Benefit to the Cambrian line timetable of siting the loop at Carno
The admission by Mr Mike Bagshaw of Arriva Trains Wales that Carno would be the optimum position for the loop was of tremendous significance. This, of course, was our group’s contention and the availability of a 5 minute time saving on the round trip time from the loop to Birmingham and back if the loop is moved from Talerddig led us to advocate it because it would significantly improve line punctuality.
The set of slides on the “Cambrian Frequency Enhancements – GRIP 4 ‘Single Option Development’ Study” presented by Mr Ian Baxter to the Shrewsbury-Aberystwyth Liaison Committee at Machynlleth on July 27th includes one entitled “Performance of an Hourly Service”. This models performance between Aberystwyth and Shrewsbury only and shows an improvement from 86.81% PPM for a 2 hourly service on today’s infrastructure, which rises to 92.37% for an hourly service on the enhanced infrastructure with the loop remaining at Talerddig. However, this is not a satisfactory basis to compare different passing loop patterns, because the existing service runs to Birmingham and is subject to delays between Wolverhampton and Birmingham due to congestion, resulting in a PPM of only about 60%. Thus the benefit of having the halfway loop closer to Birmingham has not been measured in the performance modelling at all!
The excuse given for not modelling performance over the whole route to Birmingham was that the December 2008 timetable was not known. Yet the existing timetable was known, and could have been used instead to get a useful comparison between Talerddig and Carno.
Could Network Rail be asked why they did not use the existing Wolverhampton-Birmingham timetable to model performance on the complete Aberystwyth-Birmingham route in order to evaluate how much the existing very poor performance would improve with the loop moved to Carno?
CSAG contend that the Welsh Assembly Government is missing a once-in-a-generation opportunity to make a step-change improvement in the currently unacceptable punctuality on the line. Could Mr Tim James be asked why the potential extent of this improvement has not even been evaluated?
The Ministerial decision on passing loop pattern
The Minister has stated in replies to critics of his decision in favour of Talerddig that “There were a number of value for money and operational considerations involved in the choice between Talerddig and Carno as the optimum site for this infrastructure”, but gave no details. Could the Minister be asked to set out what these considerations were? As regards cost, Mr Ian Baxter indicated that the net additional capital cost of moving the loop to Carno was £0.6 m, but failed to answer our question about the additional maintenance costs associated with the 2 km longer dynamic loop at Welshpool.
If the loop remains at Talerddig, Network Rail will be faced with the additional maintenance cost of the 2 km of new track required for the longer dynamic loop. We do not have access to any track maintenance costs for relatively lightly trafficked lines, but a rough estimate can be made from published figures. Based on the UK total track length of 31105 km (Network Rail, 2006 Annual Return, Table 47) and the Network Rail 2004-5 annual maintenance expenditure of £1271 million (Office of Rail Regulation Annual Assessment of Network Rail, Table 15), the annual maintenance expenditure per track kilometre comes to about £40,000 per km. Because of the low traffic intensity, it is likely that maintenance expenditure on the Cambrian line is significantly less than the average, so it is suggested that the maintenance figure per track mile on the Cambrian line is taken as £20,000 per km, giving an annual maintenance expenditure on the extra 2 km of track at the Welshpool dynamic loop of £40,000. Discounting 60 years worth of additional maintenance expenditure back to the present (at 3.5% over the first 30 years and 3% over the next 30 years) results in an additional contribution of £40 k x 34.34 = £1.4 million to the project whole-life costs. This far exceeds the £0.6 m capital cost saving of retaining the Talerddig loop as opposed to relocating it at Carno.
Conclusion: The decision to retain the loop at Talerddig as opposed to relocating it to Carno is wrong on two major counts:
- It fails to realise the major punctuality benefits associated with saving 5 minutes on the round trip time from the loop to Birmingham and back
- It is more expensive in whole-life cost terms
4. Option of combining a single platform with the passing loop
After we had submitted our written evidence, we realised that it would be possible to combine a single platform with a passing loop at Carno and still provide a 2 hourly interval service to Carno after the introduction of the hourly service serving the other stations. When a train stopped at the station platform on one track of the passing loop, the passing train would run through on the other track without a platform.
This arrangement would obviously cost much less than the two platform station envisaged originally, particularly if the latter were deemed to require a footbridge. Based on the £346 k cost estimate for two 10 m platforms and car park, the cost of one 10 m platform and car park would be £346 x 0.6 = £208 k, giving an increased Financial Benefit Cost Ratio of £0.887/£0.208 = 4.26. Even a 2 carriage long platform (corresponding to the length of most trains after the introduction of the hourly service) would cost only about £0.825 m based on the Llanharan costings, giving a Financial Benefit Cost Ratio of £0.887/£0.825 = 1.08.
During the questions session, AM’s appeared to accept the £5 million ball-park cost estimate given by Network Rail.
Conclusions:
- A single platform would be sufficient to provide Carno with a 2 hourly train service in conjunction with a passing loop.
- The station would cost only about £200 k if the platform were 10 m long, giving a Financial Benefit Cost Ratio of 4.3.
- The Llanharan costing indicates a £1.65 m cost for a 4-carriage long platform (which would be fully compliant) and a £0.825 m cost for a 2-carriage long platform which would be fully compliant most of the time.
5. Feasibility of inserting Carno station into the hourly train service timetable in the absence of a loop
Now that the Minister has announced that Talerddig passing loop is to be enhanced instead of constructing a new loop at Carno, it is vital to establish whether a station stop can be inserted in the timetable at Carno once the hourly service has been introduced, unless this decision is quickly reversed.
Network Rail stated at the Cambrian Line Liaison Committee meeting on July 31st that the selection of Talerddig as the loop location instead of Carno would neither make it easier to re-open the station at Carno nor make it more difficult and the Minister has stated in replies to critics of his decision in favour of Talerddig that “Nothing in this makes the potential future development of a station at Carno less practicable or possible.” We do not agree.
In view of this, I submitted the following questions to Mr Ian Baxter of Network Rail the day before the meeting: “Network Rail stated at the Cambrian Line Liaison Committee meeting on July 31st that the selection of Talerddig as the loop location instead of Carno would neither make it easier to re-open the station at Carno nor make it more difficult. Could Network Rail explain how an additional station stop can be fitted in between passing loops optimally spaced for an hourly service in each direction? Presumably if a target running time of 28 minutes (30 minutes less 2 minute safety margin) is required between loops, then an extra stop costing 3 minutes just cannot be accommodated.” Mr Ian Baxter failed to answer this question.
Neither did Mr Tim James from the Welsh Assembly Government Rail Team submit evidence as to how the Carno stop could be inserted.
We would be grateful if the committee could obtain answers to these vital questions.
We would also ask you to note that our local Assembly Member, Mr Mick Bates, has tried repeatedly since March to secure a meeting between our group and the Minister for Economy and Transport to discuss the loop siting before any decision was made, but was unsuccessful.
6. Lack of Information
One of the great difficulties we are labouring under is lack of information from the authorities. We have tried for months to obtain a copy of the GRIP 3 report, which reported on work completed in December. We am sure that the information in the GRIP 3 “Option selection” report and the GRIP 4 “Single option development” report (which presumably should now be available) would be helpful to us and to your committee.
The lack of information available to us is made explicit in the list of questions we submitted to Mr Ian Baxter on Monday and to Mr Tim James on Tuesday during my presentation. Of the eight question for Mr Baxter (see attachment), answers were only given to the first two during the meeting, and Mr James neglected to respond to the three questions addressed to him, which were as follows:
a) Stakeholder Consultation: Why were SARPA’s representations that passing loops should be located in centres of passenger demand ignored? No response was received to SARPA’s letter of October 24th (copy attached).
- b) GRIP 3 Next Steps: Why was the Carno loop option not pursued in GRIP 4 as proposed by Network Rail? Although Network Rail originally gave an undertaking to provide a written answer to this question, they recently stated that CSAG should seek the answer from WAG, who were the client for the feasibility study.
- c) GRIP 3 Report: Why has this report not been made public in the interests of transparency, despite many requests over many months?
We believe that the provision of answers to the pertinent questions we have raised would assist your work, and we trust the committee will be able to obtain them.
With reference to the cost breakdown I requested in the last of my questions to Network Rail, we can report that this breakdown is provided in full for the Talerddig option on one of the slides on the “Cambrian Frequency Enhancements – GRIP 4 ‘Single Option Development’ Study” referred to above. The costs are as follows:
Dyfi Junction Loop + 0.6 m line and platform lift £2.72 m
Talerddig Enhanced Loop including retained Engineer’s Siding £1.14 m
Welshpool 4 km Dynamic Loop to Fron LC £4.99 m
Weig Lane AOCL conversion – MCB CCTV + track condition works £0.63 m
Sub-Total #1 £9.48 m
Network Rail Costs £0.65 m
ERTMS Incremental Signalling Costs £0.40 m
TOC compensation costs £0.25 m
Sub-total #2 £10.78 m
Generic 20% risk at GRIP 4 (post QRA) £2.15 m
TOTAL £12.93 m
In his written evidence, Mr Ian Baxter stated that the cost of the enhancement scheme with the loop at Carno was £0.6 m more (cf £1.0 m more at GRIP 3), so the GRIP 4 breakdown of costs for the latter scheme should be available. Please can you assist with obtaining the information?
7. Conclusions
1. Station costs: The difference between the £350 k and £5 m cost estimates lies almost entirely in the different station specifications considered – that is the platform length and the presence or absence of a DDA compliant footbridge. CSAG therefore submit that the discussion can be restricted to the question of what platform length is actually needed, which fundamentally reduces to a safety issue.
2. Safety of a short platform station: The anticipated frequency of alighting accidents due to the short platform proposed at Carno is once every 42,000 years on pessimistic assumptions about train conductor error. It is concluded that the road safety benefits of having a station vastly outweigh the tiny reduction in rail safety resulting from having a short platform rather than a full length one, which is likely to be prohibitively expensive.
3. Benefit to the Cambrian line timetable of siting the loop at Carno: The decision to retain the loop at Talerddig as opposed to relocating it to Carno is wrong on two major counts:
- It fails to realise the punctuality benefits associated with saving 5 minutes on the round trip time from the loop to Birmingham and back
- It is more expensive in whole-life cost terms
4. Option of combining a single platform with the passing loop:
- A single platform would be sufficient to provide Carno with a 2 hourly train service in conjunction with a passing loop once the hourly service is introduced on the line.
- The station would cost only about £200 k if the platform were 10 m long, giving a Financial Benefit Cost Ratio of 4.3.
- The Llanharan costing indicates a £1.65 m cost for a 4-carriage long platform (which would be fully compliant) and a £0.825 m cost for a 2-carriage long platform which would be fully compliant most of the time.
5. Feasibility of inserting Carno station into the hourly train service timetable in the absence of a loop: The Minister has stated that retention of the loop at Talerddig would not prejudice the opening of Carno station, but neither Network Rail nor the Welsh Assembly Government Head of Rail advanced any evidence to support this.
8. Further questions
In addition to the 8 questions we submitted to Network Rail in advance of Tuesday’s meeting and the 3 posed to Mr Tim James on the day, many of which have been highlighted above, the following additional questions have been raised in this document:
1. Could Network Rail be asked why they did not use the existing Wolverhampton-Birmingham timetable to model performance on the complete Aberystwyth-Birmingham route in order to evaluate how much the existing very poor performance would improve with the loop moved to Carno?
2. CSAG contend that the Welsh Assembly Government is missing a once-in-a-generation opportunity to make a step-change improvement in the currently unacceptable punctuality on the line. Could Mr Tim James be asked why the potential extent of this improvement has not even been evaluated?
3. Could Mr Tim James from the Welsh Assembly Government Rail Team submit evidence as to how the Carno stop could be inserted into the hourly service with the loop remaining at Talerddig.
Tony Burton 7.9.07