Home

TTCI report: user guide for variable charges model

1. 4 A EEE 20 Locomotives Other Figure 16 Effect on Passenger Vehicles of Operating at Balance Speed Figures 17 and 18 respectively show the percentage change in variable usage costs for empty and loaded freight vehicles 20 15 Seg eee oe eee eae eee 10 A ooo E 5 poem j 4 Y a f 0 E a Ex TF O yyy gt Pod cS A PAS o v E ig pu Pus An dd a 10 SN 15 eee eee eee 20 Locomotives Wagons Figure 17 Effect on Empty Freight Vehicles of Operating at Balance Speed Prepared by TTCI UK Ltd 27 March 2008 Page 27 m P Methodology to Calculate Variable Usage Charges for Control Period 4 am Qu M URL Meteoro Collate Variable Usage Charges for Control Perd 20 1590 eS SSS O SS ee gene 10 HHH eS RH RRR HERS RES hh a RES ROSES RSE eS RS RE SRR eee Re LP RR LL LL C ol io oka a a aes oO x ai lt 0 Ez iio ooo PS E TAR ht 2 Wiss ue ym at tar ES nto Autor Z t e gt P wi 716 75 Dae hg w A A 3 ew g CU E 36 o Y ej Y e 1 Ye bere 4 10 HRS RR HERR RE RS RE SRR SRR SRR HORE RSS manman Eee RS 15 A ES RS RES RO REG RO RS ROR HS RE RHO RE ES a aaa 20 Figure 18 Effect on Loaded Freight Vehicles of Operating at Balance Speed The effect of operating all freight trains at the balance
2. 0 5 4 4894 777777777 zou uz d oae o oa oe oe oo 0 9 9 9 oe id be en oo Season e ale eot toto o dl ow o vee 2 v E ae oe a a ee Ow E aU ccce c o i wo 10 i Se SS c DIEI PO Pe I Oe Pe Pe INE ee eee 15 4 SPSS Si SSess Sele SSP qeu tust Ss Sess sSe sees Sess ses esses 20 Locomotives Other Figure 19 Effect on Passenger Vehicles of No Lubrication Figures 20 and 21 respectively show the effect of increasing the coefficient of friction on variable usage charges for empty and loaded freight vehicles 20 C bos 15 E 4 g 10 pee eoe ADAM e z ads t a o e pe dou EX UIT aduer Ww s x w po o a 1 et M e E 0 e 2 ote LE B 9 c m a ae a aye t INS B dM 4 Porque epe dot eN ieee ee ace o 10 DI 15 EF E 20 Locomotives Wagons Figure 20 Effect on Empty Freight Vehicles of No Lubrication Prepared by TTCI UK Ltd 27 March 2008 Page 29 20 15 4 10 4 5 tomo l M ma inn cae es d 5 NET S Change in KGTM JA w e de e EL we FA e amp
3. 3 ama o4 e y wo ae 40 4 a s 60 4 tee ee eee ee 80 po a 100 Locomotives Wagons USM Exp 0 5 USM Exp 1 0 Figure 9 Effect of Unsprung Mass Exponent on Empty Freight Vehicle Variable Usage Charges 100 ee Ce 40 20 A lu c sm xi A gt o id 0 man AS an 2 RA sm 0 we ee tumama e Y ee ar we Q 9 y f ul Lu 20 Mem p m co m Cd GR 2 E IMS Tw v 40 Eo a 60 poe eee ee ee oe ee ee ee er eee eee 80 100 Wagons USM Exp 0 5 m Figure 10 Effect of Unsprung Mass Exponent on Loaded Freight Vehicle Variable Usage Charges Prepared by TTCI UK Ltd 27 March 2008 Page 16 m y UA P Methodology to Calculate Variable Usage Charges for Control Period 4 Figure 9 shows that due to their relatively high unsprung masses charges for freight locomotives generally increase when the unsprung mass exponent is increased The increases are similar to those for passenger locomotives Figure 9 and 10 show the variable usage charges for freight wagons generally reduce when the significance of the unsprung mass is increased This is because they typically have a lower than average unsprung mass For freight wagons with a higher than average unsprung mas
4. Ltd 27 March 2008 Page 56 u Pa Methodology to Calculate Variable Usage Charges for Control Period 4 a 8 Sur M Uns o IM Ltd 3 The T gamma values for the curving classes Coach 50 40 Coach 50 50 Coach 50 60 and perhaps other classes which are used in APPENDIX C are missing The net damage for these and other has been interpolated between adjacent vehicles with slightly different parameters 4 In chapter 2 different methods for the calculation of the vertical track damage are compared and a sensitivy analyses of several exponents have been performed Finally it is recommended in the report that the method regarding vertical track damage should be revised until CP5 From the supplier s point of view it would be preferred if the methods for vertical and tangential track damage remain unrevised or only with little adjustments for at least 10 years i e CP4 and CP5 That would the suppliers allow to have stabil design criteria for new vehicles Additionally the time of CP4 could be used to monitor the usability of the new tangential track damage method If it is necessary slight changes of the tangential track damage method could be introduced for CPS3 Noted If the vertical track damage method will be revised the new method should also consider that the vertical forces on the track are not only dependent on the unsprung mass but also on the inertia of the unsprung mass around the x axis because the vertic
5. 4 hy amp West Coast Trains Ltd letter 1 February 2008 Clearly on the WCML with Tilting Rolling Stock operating over curves at up to 300mm CD the average figure of 40mm is somewhat unrealistic particularly as such high CD s help our tilting vehicle s curve performance We would therefore suggest that in order to make the RSD term representative for WCTL in terms of the applicable T Gamma figures Traction Force exerted x Creepage that Network Rail re calculates in incremental steps a FULL range of CD s for each vehicle i e Class 390 and 221 This when applied to the CD s themselves as determined by the linespeed profile for each vehicle s route will thus ensure the RSD is correctly weighted for each tilt applicable vehicle Although route based charges are no longer proposed for CP4 a network average cant deficiency for tilting trains could be considered although any re calculation of T Gamma values is not necessarily too onerous to undertake our belief is that the VAMPIRE modelling tool used to calculate curve values etc does not currently encapsulate Class 390 and Class 221 values However there would be the possibility subject to cost that should Alstom be approached they may be able to undertake such T Gamma re calculations for us utilising their ModKat programme This would thereby ensure that any RSD value for WCTL would be robust and fair This would be the preferred approach Alstom calcu
6. 40 Passenger 40 50 Bogied coach or multiple unit Coach 50 50 Passenger 50 50 Bogied coach or multiple unit Coach 50 60 Passenger 60 50 Bogied coach or multiple unit Coach 64 30 Passenger 30 64 Bogied coach or multiple unit Coach 64 40 Passenger 40 64 Bogied coach or multiple unit Coach 64 50 Passenger 50 64 Bogied coach or multiple unit Coach 64 60 Passenger 60 64 Bogied coach or multiple unit Coach 80 30 Passenger 30 80 Bogied coach or multiple unit Coach 80 40 Passenger 40 80 Bogied coach or multiple unit Coach 80 50 Passenger 50 80 Bogied coach or multiple unit Coach 100 40 Passenger 40 100 Bogied coach or multiple unit Coach 128 30 Passenger 30 128 Bogied coach or multiple unit Coach 128 40 Passenger 40 128 Bogied coach or multiple unit Coach 128 50 Passenger 50 128 Bogied coach or multiple unit Tilting 50 50 Passenger 50 50 Tilting coach or multiple unit Loco2 50 Passenger Bo Bo locomotive Loco3 50 Passenger Co Co locomotive NACO empty Freight Empty wagon with swing motion bogies NACO loaded Freight Loaded wagon with swing motion bogies Pacer 10 Passenger 2 axle Shunter 3 axle Y25 empty Freight Empty wagon with Y25 bogies Y25 loaded Freight Loaded wagon with Y25 bogies Vehicle mass and primary yaw stiffness values are given only when necessary to distinguish between vehicle types Prepared by TTCI UK Ltd 27 March 2008 Page 22 e s_ A UA P Methodology to Calculate Variable Usage Charges for Control Period 4 A
7. A E a a a Eg etuer 100 Locomotives Wagons USM Exp 0 5 1 USM Exp 1 0 Figure 8 Effect of Unsprung Mass Exponent on Passenger Vehicle Variable Usage Charges Figure 8 shows that increasing the unsprung mass exponent increases the variable usage charges for passenger locomotives This is due to the relatively high unsprung mass of those locomotives With a linear dependency on unsprung mass which is not unreasonable the variable usage charges for passenger locomotives could increase by up to 70 The effect of the unsprung mass exponent on passenger multiple unit vehicles depends on whether the vehicle has powered or unpowered axles For powered axles the unsprung mass is relatively high and charges increase if the significance of unsprung mass is increased For unpowered axles the opposite is true Figures 9 and 10 show the effect of increasing the unsprung mass exponent on empty and loaded freight vehicles respectively Prepared by TTCI UK Ltd 27 March 2008 Page 15 Change in KGTM Change in KGTM m E UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 av SERERE L mmml 100 a le BH a 60 iim O ioe oe seecer eaters 40 4 20 f pao seca aL ELLE HEEL LLL Po z B o E sJ n a 5 p t Ue pases cn dent ei me a Canin Je e oem pit pe m mm aS um Km ma a amp mer 20 Pp
8. Cost 14 500 x CrackDamage 2 000 x WearDamage for 0 lt Ty lt 175 N B22a Substituting the values from Table B1 into Equation B20 gives Cost 12 500 x WearDamage for Ty gt 175N 22b where CrackDamage and WearDamage are defined in Equations B8 and B9 respectively in terms of Ty Prepared by TTCI UK Ltd 27 March 2008 Page 71 m E UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 References 1 Burstow M C October 2003 Whole Life Rail Model Application and Development for RSSB Development of an RCF Damage Parameter AEA Technology Rail Derby 2 Network Rail October 2005 Inspection and Maintenance of Permanent Way Network Rail Standard MR SP TRK 001 Issue 02 London UK Prepared by TTCI UK Ltd 27 March 2008 Page 72
9. Part 10 1980 Steel concrete and composite bridges Code of practice for fatigue British Standards Institution United Kingdom American Railway Engineering and Maintenance of Way Association Manual for Railway Engineering Chapter 15 Steel Structures Washington D C 2006 Burstow M C October 2003 Whole Life Rail Model Application and Development for RSSB Development of an RCF Damage Parameter AEA Technology Rail Derby Network Rail October 2005 Inspection and Maintenance of Permanent Way Network Rail Standard MR SP TRK 001 Issue 02 London UK Bevan A August 2005 Development of an Anti RCF Wheel Profile Rail Technology Unit Manchester Metropolitan University 110 10 Manchester UK Prepared by TTCI UK Ltd 27 March 2008 Page 39 Prepared by TTCI UK Ltd Methodology to Calculate Variable Usage Charges for Control Period 4 27 March 2008 Page 40 m E UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 Appendix A RESPONSES TO STAKEHOLDERS WRITTEN COMMENTS Stakeholder comments are shown in italics Responses are shown in regular text Virgin West Coast Trains 27 June 2007 Comment 1 1 We acknowledge that the current method for charging Train Operators for the damage that their vehicles cause to the Network is reasonable and that it gives some incentive to Operators to minimise such damage through the usage and ongoing establishment of track fri
10. Standards EMGTPA 0 47 UIC Code 714 0 28 berg and Andersson Up to 3 0 Railway Group Standards and the UIC Code have a similar approach to the factor K that accounts for axle load In both cases it depends on the mix of traffic but it typically varies linearly with axle load Thus the exponent used by these standards is typically 1 0 berg and Andersson raise the total vertical force static plus dynamic to a value of 3 0 for the settlement and component fatigue factors in their model Thus the effective exponent for the static axle load is somewhat less than 3 0 Table 4 summarises the exponents used for axle load In each case the formula for damage includes a gross tonnage factor that has the effect of increasing the axle load exponent by 1 0 Table 4 Summary of Axle Load Exponents Source Exponent B RI ORR Variable Usage Charging 0 49 1 49 Railway Group Standards EMGTPA 1 00 2 00 UIC Code 714 1 00 2 00 berg and Andersson Upto3 0 Up to 4 0 Again there is a wide variation and this can be expected to have a significant effect on the variable usage charges for high axle load vehicles The results of a sensitivity study on the axle load exponent are presented in subsection 2 4 2 Railway Group Standards and the UIC Code use the factor K to account for differences in unsprung mass between certain types of vehicles For freight vehicles designed to give low track forces typically
11. be less This would result in lower track access charges if a bottom up approach would be used However Figure 19 shows that the track access charges could be reduced for most of the passenger vehicles by up to 2 when there is no lubrication considered and the vehicles are causing more damage on the track This is due to the fact that by using the top down approach the total sum of the track maintenance costs is a fixed value which is just distributed to the different vehicles Therefore the benefit of lubrication does not have a positive effect regarding lower track access charges and there seems to be no incentive to Network Rail to invest in track lubricators for the damage reduction A bottom up approach would have this incentive The ORR sets efficiency targets to network Network Rail These incentivize Network Rail to reduce track maintenance and renewal costs by for example using lubricators Additionally we could not find any incentive to Network Rail to maintain an appropiate track quality within these methods This would also be important to have because it is generally known that a poor track quality may lead to further damages e g rcf can be caused by major track irregularities The costs for repairing the additional damages seems to be included into the top down approach Prepared by TTCI UK Ltd 27 March 2008 Page 59 m y UA P Methodology to Calculate Variable Usage Charges for Control Period 4 The ORR sets tra
12. by TTCI UK Ltd 27 March 2008 Page 12 Change in KGTM Change in KGTM ___ Pa Methodology to Calculate Variable Usage Charges for Control Period 4 U ma ai ai iu A 100 80 4 T si A E Shee eee CU AN 40 as zt a pl o amp A AAA ae 5 3 0 jon ooo a 20 D AEN ads ye 20 1 om e Sn a omen 1 nm x f 9 e e oy e e Vest site e AP s 40 PA Ange eT vr Fat A i a a n n P g 60 2p Xx emt a 2 8 s y oe ts e Angst js 80 i 9WMM a A e ore 2 dow o a ms 100 Locomotives Wagons Axle Load Exp 1 0 Axle Load Exp 2 0 Figure 6 Effect of Track Axle Load Exponent on Empty Freight Vehicle Variable Usage Charges 100 a gem m 80 rw o RT A tv F 1 4 EMI E Ti as a son SIUS y ma m hi a f prb HE A af 40 dima MU foresee A So e r a A u Q4 oa 4 o Qi e 20 O ind Ata eo e Me EPA a y Ed ae a e o Bp cost e QU o e 20 UNS VUES AA PME y RE ies by M iue 4 bd a 40 E A Poo e p m4 a e a a L1 L 60 7 Wo eee Hee See eee ee eee a i 80 SB 100 Wagons Axle Load Exp 1 0 Axle Load Exp 2 0 Figure 7 Effect of Track Axle load Exponent on Loaded Freig
13. estas RRS qa uiid iade oo i 10 Mrs v 2 i a o e ef e o gt e ce ea t eS e Y 3 hi 24 X ug t Pad a y Modes E 0 M e oo e i oam mun an o po Mi gt S S i A D 10 s n 15 ee 20 Locomotives Wagons Figure 22 Effect on Empty Freight Vehicles of Y25 P10 Wheel Profiles 20 15 10 5 0 5 c c c cc coccccoccccoccccocococccocccocc2 c 2c2 ee 15 20 Figure 23 Effect on Loaded Freight Vehicles of Y25 P10 Wheel Profiles Prepared by TTCI UK Ltd 27 March 2008 Page 31 lt A ur HUS TM Y IN Methodology to Calculate Variable Usage Charges for Control Period 4 ti amp Changing the wheel profile clearly makes the most difference to the vehicles whose wheels have been changed In this example changing the Y25 profiles from semi worn P5 to semi worn P10 has slightly increased the Y25 variable usage charge The percentage increase is greater for the empty wagons compared to the loaded wagons because the charges are less for the empty wagons The change is less than approximately 2 from the base case T
14. face of the high rail where it is assumed to be 0 2 A sensitivity study on this assumption is presented in this report 3 It could be argued that RCF is replaced by sidewear in sharp curves but this should be offset by track lubrication is it should not be the means by which the top corner of table 1 in TTCI UK NR Report No 07 003 is filled in Correct In sharp curves 7y is high enough to cause wear even though the curve is assumed to be lubricated 4 The proposed approach seems to create a perverse incentive to Network Rail to allow track to deteriorate since worse track attracts higher access charges as well as causing more fatigue damage to the vehicle There is no perverse incentive since the cost of maintaining track in deteriorated condition is more than the benefit from any increase in variable usage access charges The specification and control of track quality is handled outside the access charging regime 5 I am puzzled as to why the variable use percentages are so low because this implies that a one train a year railway would cost around 70 of the cost of the existing railway There are some obvious costs such as repair of structures and drainage which are train operation independent and possibly the effects of frost dry weather on vertical geometry but I struggle to think of other things Variable usage access charges are based on marginal cost 1 e the additional cost of operating a vehicle on a network already
15. other benefits such as the control of corrugations and the enhancement of short wavelength irregularities around welds that reduce vertical track forces Should this not reduce the cost element of rail grinding applicable to rsd 95 of the rail grinding cost is assumed to be due to rail RCF Rail rcf cracks are assumed to exist uniformly throughout curves although there are situations where this is true there are there not many situations where this isn t the case It is assumed that the entire curve is ground if RCF is uniform or in patches Rail grinding the cost of rail grinding is assumed to be carried out everywhere on the network The rationale for this proposal together with Ty values contained in Appendix D of Doc 2 make it clear that track with a curve radius greater than 1500m the range of track for which trackside lubrication should not be considered is most unlikely to need any grinding at all this amounts to around 80 of the network Should this not reduce the cost element of rail grinding applicable to rsd No 95 of rail grinding is assumed to be driven by RCF The methodology calculates the amount of rcf damage and wear depending on curves radius It is assumed that the costs and the processes for controlling rsd in S amp C can be averaged in with plain line Given that the cost of rsd in S amp C is around 40 of the total cost for rsd is this a valid assumption Prepared by TTCI UK Ltd 27 March 2008 Pa
16. quality could be as specified in the standards but cracks could still be caused by modes I and 2 Agreed For Control Period 4 vehicles will be charged based on their propensity to produce Mode 0 RCF Further improvements to the charging regime could be made for CP5 if deemed necessary said that route based charging would incentivise good behaviour design Prepared by TTCI UK Ltd 27 March 2008 Page 51 72222 Ula zu M GHz Agreed Network average charging with the new term for rail surface damage will incentivise improved curving performance Route based charging would provide a stronger incentive on routes with a high proportion of curves Methodology to Calculate Variable Usage Charges for Control Period 4 amp hy suggested this a single worn wheel profile could be challenged where wheels wear characteristic is different to that shown A sensitivity study on the effect of wheel profile is presented in this report Alternative wheel profiles will be considered if there is evidence they are representative asked that as there are not many wheel profiles why were the common profiles not utilised to make the assumptions Although there are many new profiles the differences between worn profiles are assumed to be insignificant A sensitivity study on the effect of wheel profile is presented in this report Railway Industry Association InfraTIG Minutes 20 June 2007 observed that axle spacing as
17. shift in vertical forces due to cant deficiency in curves The service data available at the time was used to calibrate the models The introduction of new methods of renewal maintenance and inspection and new materials has meant the historical service data is no longer relevant Rail failure statistics for example have changed in recent years A recalibration of the existing methodology is required In addition the coefficients and exponents in Equation 1 appear to have been derived for part of the former London North Eastern Zone of Railtrack There is an underlying assumption that these are valid for the network in general 2 2 Structures The existing methodology for recovering the variable usage costs from structures is the same as that for track except a different formula for damage is used For structures Equation 3 is used Equivalent Structures Damage C A S GTM 3 where C is 1 20 for two axle freight wagons and 1 for all other vehicles A is the axle load tonnes S is the operating speed miles hour Equations 1 and 3 show that the axle load and speed exponents are higher for structures compared to track damage It can also be seen that the unsprung mass term is not present in the damage formula for structures This is because the dynamic forces that are a function of unsprung mass are attenuated by bridge superstructure before they reach the bridge structure and cause damage 2 3 Alternative Methodologies
18. variable usage charges is to reduce the charges to the freight sector by approximately 11 6M year Table 19 shows that this represents a swing of 4 2 of costs from the freight to the passenger sector Prepared by TTCI UK Ltd 27 March 2008 Page 33 m zu P Methodology to Calculate Variable Usage Charges for Control Period 4 MA vr B US Z Ltd 4 ROUTE BASED CHARGING Currently variable usage charges are calculated on a network average basis The factor to convert an individual vehicle s damage to cost is derived by comparing the total damage for the network with the total variable cost for the network The new methodology for including a rail surface damage term has also been applied on a network average basis The advantage of this approach is that it is more cost reflective than at present and it incentivises operators to use trains that reduce rail surface damage The new methodology would be even better if it were route based instead of being based on the network average In a route based approach vehicles would be charged for the rail surface damage they caused on curves typical of those they normally encountered in service The network average approach assumes that all vehicles go around every curve on the network Table 20 shows the six route classes defined in Network Rail s ICM Three hundred strategic route sections SRS which vary from 1km to 546km in track length are defined in the ICM Table 20 Infra
19. well as axle load was a significant factor to include in any revised equivalent damage equation Agreed A full review of the vertical damage charging methodology is recommended Doubt was expressed that the relatively small tranche of the TAC to which changes would be applied may not be large enough to incentivise the substantial investment entailed in the development of new lighter trains The cost of a train with good curving performance is not necessarily higher than one with poor curving performance It was observed that mixing freight traffic and passenger services with various stopping patterns would invalidate the cant profile calculation on individual curves A sensitivity study on the effect of speed in curves is presented in this report Railway Industry Association email 30 May 2007 1 I had understood that RCF is worst for a range of middle sized curves but table 1 in TTCI UK NR Report No 07 003 seems to suggest that the factor is increasingly more severe as curves become tighter except when operating at cant excess Table 1 shows Ty which is not the same as RCF Ty is transformed by the proposed damage function into wear and or RCF Prepared by TTCI UK Ltd 27 March 2008 Page 52 _u_ Pa Methodology to Calculate Variable Usage Charges for Control Period 4 e AA Sur M Us o m Lid 2 How is coefficient of friction dealt with A value of 0 4 is assumed for all contact points except the gauge
20. 1 100 51 62 1 15 101 130 63 81 1 25 131 160 82 99 1 35 161 200 100 124 1 40 201 250 125 155 1 45 More than 250 More than 155 1 50 Figure 1 shows the factors from Tables 1 and 2 plotted on linear axes The RGS factor of 1 60 for the speed band 101 to 125 mph has been replaced with 1 70 The RGS value is a long way from the trend formed by the other values and is assumed an error Figure 1 also shows logarithmic trends fitted to the data points The exponents for these trends are 0 47 and 0 28 for the RGS and the UIC respectively The reason for the difference in the two standards is unknown Prepared by TTCI UK Ltd 27 March 2008 Page 5 1 6 e RES o UIC Damage Factor A N X e 0 50 100 150 200 Speed mph Figure 1 Speed Factors from Railway Group Standards and UIC Code berg and Andersson have recently described a model for determining the effect of traffic on track costs They reviewed 21 track deterioration models before producing their own model Thus the berg and Andersson model may be considered as the state of the art in track deterioration modelling It was developed for the track and operations in Sweden These are similar enough to those on Network Rail controlled infrastructure for the model to be applicable here The Oberg and Andersson model consists of three terms to account for 1 Track settlement which drives tamping costs
21. 10 4 uL 20 Figure 21 Effect on Loaded Freight Vehicles of No Lubrication Figures 19 to 21 show that with the exception of some locomotives variable usage charges are relatively insensitive to assumptions about the coefficient of friction between the wheel and the rail The net effect of assuming the higher values of coefficient of friction would be to transfer usage charges of 330k from passenger to freight Again in comparison with total variable costs of 273M this is insignificant 3 2 4 Wheel Profile In the base case the freight wagons with Y25 bogies are modelled with semi worn P5 wheel profiles These wagons sometimes run with P10 profiles The variable usage costs were recalculated for all vehicles with the Y25 bogies having wheels with semi worn P10 profiles The effect on passenger vehicles was found to be negligible change in p mile less than 0 1 Figures 22 and 23 respectively show the effect on empty and loaded freight vehicles of changing the wheel profiles on the Y25 bogies to semi worn P10s Prepared by TTCI UK Ltd 27 March 2008 Page 30 m Pa Methodology to Calculate Variable Usage Charges for Control Period 4 am Qu MUz Meteoro Calculate Variable Usage Charges for Control eod Change in KGTM __ Pa Methodology to Calculate Variable Usage Charges for Control Period 4 A AAA aS T TM T sam 20 A cessa er Gees
22. Alternative methodologies have been investigated to give continued assurance that the existing equivalent damage formulae are valid In particular alternative values for the axle load speed and unsprung mass terms in Equations 1 and 3 have been studied 2 3 1 Track In a former Railway Group Standard a definition was given for Equivalent Million Gross Tonnes per Annum EMGTPA The definition is based on one given by the International Union of Railways UIC EMGTPA is used today to specify design maintenance renewal and inspection of Network Rail s track Since EMGTPA is a method for combining the effects of speed axle load and vehicle type to determine the volume of work required it is also relevant to the cost of that work Prepared by TTCI UK Ltd 4 27 March 2008 4 Page 4 Methodology to Calculate Variable Usage Charges for Control Period 4 Equation 4 gives the formula for EMGTPA EMGTPA S K MGTPA 4 where S is a speed factor and K is a factor to account for different axle loads Tables 1 and 2 show the speed factors used by Railway Group Standards RGS and the UIC respectively Table 1 Speed Factors from Railway Group Standards Speed mph Factor Up to 40 1 00 41 50 1 10 51 60 1 20 61 75 1 35 76 100 1 55 101 125 1 60 126 140 1 80 Table 2 Speed Factors from UIC Speed kph Speed mph Factor Up to 60 Up to 37 1 00 61 80 38 50 1 05 8
23. Equation 3 which is used for allocating the variable cost of structures has axle load raised to the power 3 83 When the axle load component of gross tonne miles is included the exponent becomes 4 83 Fatigue damage in steel bridges is typically dependent on stress raised to the power 3 0 The exponent for concrete and masonry bridges is not well defined There may not even be a logarithmic relationship between damage and axle load for these types of structures If it is assumed that stress range in a steel bridge member is proportional to axle load then the axle load exponent in Equation 3 would be expected to be around 2 0 The value of 3 83 seems high A sensitivity study on the effect of this value is presented in subsection 2 4 4 In general stress range in a member of a steel bridge is not directly proportional to axle load It also depends on the spacing between axle loads Thus the equivalent damage formula for structures could be improved by including a term for axle spacing The approximation appears to have been made that except for the 2 axle freight wagon the variation in axle spacing is insignificant for vehicles operating on Network Rail controller infrastructure The current speed exponent in the equivalent damage formula for structures Equation 3 is 1 52 This is consistent with AREMA guidelines for speed limits on bridges Formulae for setting speed limits use exponents of 1 0 and 2 0 for concrete and steel bridges respect
24. Methodology to Calculate Variable Usage Charges for Control Period 4 UK NR Report No 08 002 By John Tunna Richard Joy Xinggao Shu and Ben Madrill TTCI UK Ltd 27 March 2008 Disclaimer This report was prepared for Network Rail by TICI UK Ltd a subsidiary of the Transportation Technology Center Inc TICI Pueblo Colorado USA It is based on investigations and tests conducted by TICI with the direct participation of Network Rail to criteria approved by them The contents of this report imply no endorsements whatsoever by TICI UK Ltd of products services or procedures nor are they intended to suggest the applicability of the test results under circumstances other than those described in this report The resuls and findings contained in this report are the sole property of Network Rail Intellectual property disclosed in this report is vested in TICI UK Ltd NR isa licensed user They may not be released by anyone to any party other than Network Rail without the written permission of Network Rail TICI UK Ltd is not a source of information with respect to these tests nor is it a source of copies of this report TICI UK Ltd makes no representations or warranties either expressed or implied with respect to this report or its contents TICI UK Ltd assumes no liability to anyone for special collateral exemplary indirect incidental consequential or any other kind of damages resulting from the use orapplication of this report o
25. al track irregularities of the left and the right rail are usually not parallel see Six Klaus Abh ngigkeit dynamischer O Kr fte von der Massenverteilung im Fahrwerk Elektrische Bahnen Vol 104 2006 No 11 pp 530 534 Thus it is suggested to consider an additional factor It is also suggested that the new method allows stakeholders to submit own calculations of the vertical track forces for a new vehicle this should be similar to the method for the tangential track damage It could also be useful to calculate these damage values on a route based approach too The route sections should be the same as for the tangential track damage method chapter 4 Otherwise the vertical forces on the track are usually measured during the admission tests of a new vehicle type in some countries These measured results could also be used for determing the vertical damage Such a method has already been introduced for calculating the track access charges of the Austrian Federal Railways for locomotives Additionally it should be mentioned that both methods vertical track damage and tangential track damage do not consider the track damage which is caused by the lateral track forces Y forces These forces might cause damage on the track regarding its lateral alignment when the vehicle is hunting or during curve Prepared by TTCI UK Ltd 27 March 2008 Page 57 m y UA P Methodology to Calculate Variable Usage Charges for Control Perio
26. ally properties of the whole tribological system and therefore dependent on the wheel and rail material and other parameters For example the values for the wear damage which can be found in the literature have a wide range How will be dealt with this issue when new materials are introduced e g head hardened rails A recalibration of the usage charges may be required if and when a significant proportion of the rail on the network is head hardened 14 Table B1 How have these values been assumed especially the last two rows The first four values have been taken from current practice and standards The last two values are from Ref B1 15 APPENDIX B This appendix is showing how the track maintenance costs are related to the track damage this seems to us like a bottom up aproach because to each amount of rcf or wear damage an amount of costs is matched see equation B16 Prepared by TTCI UK Ltd 27 March 2008 Page 60 gt aAWwLZ Lt Methodology to Calculate Variable Usage Charges for Control Period 4 N amp However as far as we understand the top down aproach the total track maintenance cost is related to the total damage total costs k total damage with total damage sum of the damages of the entire fleet Hence the factor k is calculated depending on the different amounts of the total damage of the entire fleet After k has been determined the costs for a vehicle can be calculated by the
27. ature Table 15 Rail Profile Names Curve Radius m High Rail LowRail Up to 750 0 750 H ban 0 750 L ban 751 to 1500 750 1500 H ban 750 1500 L ban 1501 to 3000 1500 3000 H ban 1500 3000 L ban More than 3000 Measured semi worn wheel profiles are used in the calculation of Ty Table 16 shows the profile names for the wheel profiles used for different types of vehicle Prepared by TTCI UK Ltd 27 March 2008 Page 24 m m al Methodology to Calculate Variable Usage Charges for Control Period 4 amp e EINE oy A e 2 Table 16 Wheel Profile Names Vehicle Type Profile Left Wheel Right Wheel Passenger Worn P8 F mod L whl F mod R whl Freight Worn P5 P5 20001009 0131 whl P5 20001009 0131 whl Freight Worn P10 P10 20040826 0061 whl P10 20040826 0061 whl All wheel and rail profiles are available in electronic format from Network Rail All combinations of wheel and rail profiles have been analysed to check for any unusual effects introduced from their interaction The check has been made by examining rolling radius plots for discontinuities The calculations of Ty are made on right hand curves Subsection 3 2 4 presents the results of a sensitivity study on the choice of wheel profiles 3 2 Sensitivity Studies 3 2 1 Rail Grinding and Renewal Costs The rail surface damage function derived in Appendix B converts damage to cost It does so by c
28. be MNm rad Mega Newton meters per radian 7 Pages 23 25 Several assumptions are described which have been used for the Teamma calculations It might be reasonable if a party wants to submit their own calculations for a particular vehicle to use these assumptions too if they apply for a specific vehicle This would allow to have comparable results between the different parties However the information on these pages give not the necessary full information for these calculations For example It is stated on page 24 that the friction coefficient might be reduced at the gauge face of the high rail However no information can be found where the border between these two friction coefficients is defined Is there a transition area between these two values Our simulations have shown that the Tgamma values at the wheel can change significantly only by changing the transition area between the friction coefficients on the tread and on the flange Additional questions regarding track irregularities or the calculation of the Teamma values maximum value or average for a given distance may arise Thus is it highly recommended that Network Rail should issue a document which gives detailed guidelines how the Tgamma values have to be created by any party With these guidelines any party should be able to perform these calculations without any remaining question to Network Rail of course the assumptions of these Prepared by TTCI UK Ltd 27 Marc
29. by reducing the unsprung mass the factor is reduced by 0 1 For axles that are powered and whose mass is therefore increased by the mass of the motor or gearbox an additional factor is included In the RGS this factor increases linearly with the power of the motor used to drive the axle The UIC Code does not distinguish between motors with different power Instead it applies a factor to all powered axles to account for the increase in unsprung mass The constant factor is in the middle of the range used by the RGS Prepared by TTCI UK Ltd 27 March 2008 Page 7 m E UA P Methodology to Calculate Variable Usage Charges for Control Period 4 berg and Andersson recommend using computerised vehicle dynamic models to determine the effect of unsprung mass on dynamic forces In the absence of such models the suggested high frequency dynamic forces depend on the square root of unsprung mass Since unsprung mass only affects one of the components of force in the berg and Andersson model the effective exponent will be somewhat less than 0 5 Table 5 summarises the unsprung mass exponents used by the various sources described above The results of a sensitivity study on the unsprung mass exponent are presented in subsection 2 4 3 Table 5 Summary of Unsprung Mass Exponents Source Exponent ORR Variable Usage Charging 0 19 Railway Group Standards EMGTPA 1 00 berg and Andersson Up to 0 5 2 3 2 Structures
30. ccess to the identified Network Rail documents i e transparency is required It is intended to make these documents available for the consultation process In section 6 paragraph 2 the document implies that the means of calculating rcf damage from freight trains might not yet be proven It is unclear why the document believes that there is a need for further investigation of rail surface damage from freight vehicles ATOC believes that appropriate techniques exist to estimate the track damage done by passenger vehicles and that these techniques are applicable to freight vehicles without the need for further research The further work on freight vehicles involves checking the appropriateness of the key assumptions that were made to produce the initial estimate of charges In section 6 Table 7 takes the total value for variable track access charges identified in Table 6 and reapportions it taking rail surface damage into account There is minimal increase of overall costs for the freight sector but a 50 increase for the passenger sector relative to 1999 2000 costs The costs being compared by ATOC are affected by changes in traffic and costs from 1999 2000 to 2005 06 and the introduction of the new rail surface damage term V T TAG Minutes 22 5 June 2007 Minute 2 1 asked if the constant J that converts damage to cost within the surface damage component was dependant on route Both network average and route specific
31. city As a responsible Supplier we optimised the bogie correctly but the train is unstable on certain tight to gauge sections of track The train operator considers that we did not take into account the known poor quality of NR track and NR say that track is within standards The problem is that the standards do not control conicity Should now pay for the increased cost of bogie maintenance on this well optimise solution The Ty values are calculated using typical mid life wheel and rail profiles They could be recalculated using specific profiles if agreed between the operator and Network Rail Track gauge specifications and control are outside the scope of variable track usage charges 18 requested by NR We complied with this requirement by offering a 30 reduction in a typical bogie yaw stiffness reflecting that this is a 120 kph application Fortunately in view of our Hull Train experience we asked NR for information on the worn rail condition and gauge Together with worn P8 wheel information we concluded that the actual conicity is between 0 6 to 0 75 With this conicity we are unable to offer any reduction in yaw stiffness If the new charges had already been in place who would have paid the additional TAC it should not be the Operator it should not be BT because we are prepared to optimise therefore I assume that NR would pay until the track has been fixed The rail surface damage component of variable usage charge wo
32. ck quality targets to Network Rail Track quality levels are also specified in Network Rails standards 10 Table 17 Page 33 Are there also tables for the different route and curvature classes available which are similar to Table 11 This would allow to optimize the design of a new bogie to a reasonable curve distribution Route and curvature classes are no longer being considered for CP4 11 Page 36 last paragraph The route based charges have the advantage that the vehicle design can be adjusted on the requiremnts of the track for a specific route it is supposed to be introduced However when the curve distributions are separated into 15 different classes it does not seem to be reasonable that the traffic mix is still assumed on the network base We would recommend to determine the values of the traffic mix for each route and curving class or at least for each route class Noted 12 Page 48 It is stated that the influence of the traction and brake forces are not included into this calculation method However we know from our experience that the traction and brake forces have a major influence either on the amount of the Tgamma values and also on the position of the damage on the wheel and the rail Thus this should be considered in some way This effect was considered It was decided that the complication of including traction and braking was not justified 13 Figure B1 The curves of the crack damage and the wear damage are actu
33. d 4 negotiation The Austrian method is considering this issue Thus if the method will be revised it is suggested to introduce a new method regarding the lateral track damage These comments will be considered in any future review of vertical damage usage charges 5 Serveral tables in chapter 2 e g Table 6 or Table 7 show the effect on the distribution of the variable usage charges as a result of the sensitivy analyses The total sum of these variable chages is 379 545 k year Does this value also include the variable usage charge for the tangential damage How much are the variable usage charges for the vertical damage and the tangential damage These tables concern sensitivity studies on the existing usage charge that only considers damage from vertical forces 6 Table 12 Page 22 The 4th column of this table shows the primary yaw stiffness Is this the primary yaw stiffness or the effective primary yaw stiffness which is also considering the effect of axle spacing within the bogie How is this primary yaw stiffness defined which is shown in this table It would be necessary to have there a calculation method in order to determine this value for new vehicles because this could be a criterion for new designs Additionally we are a little bit confused about the unit MNm mrad is that correct These values are the primary yaw stiffness of an axle on a bogie with a 2 6 m wheelbase The units stated are incorrect They should
34. d annual uplifts are agreed each year with the ORR Prepared by TTCI UK Ltd 27 March 2008 Page 44 m y UA P Methodology to Calculate Variable Usage Charges for Control Period 4 Comment 5 4 1 Freight and Passenger What VWC fail to understand is why there is such a major increase in VAC costs inc RSD as against a minimal increase for Freight In table 4 the change in traffic volume for Freight is about 40 lower than for Passenger but in table 7 the RSD charge is only 10 of that for Passenger The shift from passenger to freight is caused by a there being less increase in freight traffic compared to passenger and b freight traffic produces less RSD cost than passenger traffic Comment 5 4 2 Freight on Tighter Curves Taking into consideration that RSD should only apply to approx 20 of the network see 5 1 3 above and that freight vehicles tend to run round tighter curves combined in our particular case regarding tilting rolling stock how can such a high RSD charge be appropriate The numbers presented in Table 7 of the report are for network averages In this case all vehicles are assumed to go around all curves in the network Appendix D of the report shows how the charges might vary by route Comment 6 1 As we have emphasised above we are not in disagreement to the principle of RSD charging but it must be administered in a correct and impartial way To undertake this it must be foundered on a straigh
35. damage Agreed The proposed use of Ty tables determines the rail surface damage caused by different vehicle types Any vehicle type almost regardless of its parameters that operates on a predominantly straight route will not cause significant levels of rail surface damage MKIV coaches on ECML are a good example of this point A charging regime that is applied across the network would apply high charges to MkIV coaches because of rail surface damage that they would cause if they operated on the WCML Nineteen percent of the ECML between Kings Cross and Doncaster has a curve radius less than 2500m MkIV coaches are commonly understood to cause rail surface damage on these curves Prepared by TTCI UK Ltd 27 March 2008 Page 48 ae Sr B Us Lt Methodology to Calculate Variable Usage Charges for Control Period 4 hy amp Section 6 3 Technical Report Sections 5 and 6 In section 5 Network Rail data is used to nominate the variable track usage charge that would be applicable in 2005 6 Table 6 based on those set for 1999 2000 Table 2 and with updates to reflect traffic growth and changes in Network Rail costs The increases seem to be based entirely on references that are internal Network Rail documents that are not publicly available This section of the document is crucial as it establishes the size of the variable track access pie see Figure 1 but does so in a way that cannot be understood without a
36. e Primary Yaw Curving Class Operator Mass Stiffness Description tonne MNm mrad 2Axle empty Freight 2 axle wagon 2Axle loaded Freight 2 axle wagon 3Piece empty Freight Wagon with 3 piece bogies 3Piece loaded Freight Wagon with 3 piece bogies Artic2 80 Passenger Eurostar trailer coach Artic3 Passenger Eurostar powered trailer Class60 Freight Co Co locomotive Class66 Freight Co Co locomotive with steering linkages Coach 8 Passenger 45 8 Bogied coach or multiple unit Coach 12 30 Passenger 30 12 Bogied coach or multiple unit Coach 12 40 Passenger 40 12 Bogied coach or multiple unit Coach 12 50 Passenger 50 12 Bogied coach or multiple unit Coach 12 60 Passenger 60 12 Bogied coach or multiple unit Coach 15 40 Passenger 40 15 Bogied coach or multiple unit Coach 16 30 Passenger 30 16 Bogied coach or multiple unit Coach 16 40 Passenger 40 16 Bogied coach or multiple unit Coach 16 50 Passenger 50 16 Bogied coach or multiple unit Coach 23 30 Passenger 30 23 Bogied coach or multiple unit Coach 23 40 Passenger 40 23 Bogied coach or multiple unit Coach 23 50 Passenger 50 23 Bogied coach or multiple unit Coach 24 30 Passenger 30 24 Bogied coach or multiple unit Coach 24 40 Passenger 40 24 Bogied coach or multiple unit Coach 24 50 Passenger 50 24 Bogied coach or multiple unit Coach 48 40 Passenger 40 48 Bogied coach or multiple unit Coach 48 50 Passenger 50 48 Bogied coach or multiple unit Coach 48 60 Passenger 60 48 Bogied coach or multiple unit Coach 50
37. e S teeter 2 2 2 MUCOSA AAA ds 4 2 3 Alternative Methodologies cceccec ccce cec cec ecce 4 2 91 PACK ut cerei e cet ee Coe iL te e 4 2 9 2 OCIOSO ROT eae n A A opu aro t esa 8 2 4 Sensitivity to Damage Model Assumptions sssssssse 9 2 4 1 Speed EXPONEN iaa 9 2 4 2 Axle Load Exponent cccccccccccceeneeeeeeeeeeeeeeeeeeeeeeeeees 12 2 4 3 Unsprung Mass Exponent ceeeeeeeeeeeeeeeeeeeeeeeeeeeeeees 14 2 4 4 SIUS ote te o o p Rip ed eR DR pU pO on So pont lee 17 2 9 Sensitivity Summary sce ceeoee coco ose oce depen ee toler nee EH eene een hee ER ee ER eek e cn Cet 18 2 6 KecommieridallOris re ceteri Fen qtu con tan et med ea diia 19 3 Proposed New Methodology eenen 20 3 1 Rail Surface Damage Methodology seeeessssssssss 21 3 2 Sensitivity Studies 0000 25 3 2 1 Rail Grinding and Renewal Costs sssssssss 25 3 2 2 Cant Deficiency i oid ecoute ecd ea dude eod eu ede edes 26 LO EUbrIOSIO aii ert e e DR RERO eases 28 BZ AAA P EE 30 3 3 Freight Charging Categories eee 32 3 4 Traffic Sector RESUS 3 03 02 nda 33 4 Route Based Charging ii deeded eee 34 DY CONC So 38 FRETERENIGES DU 39 Appendix A Responses to Stakeholders Written Comments 41 Appendix B Rail Surface Damage Function cccseeeeeeeeeeeeeeeeeeeeee
38. e Variable Usage Charges for Control Period 4 5 CONCLUSIONS A high level review of the existing methodology for calculating variable usage charges provided no justification for immediate changes However it revealed several improvements that could be made e The source of the exponents in the equivalent damage formulae is undocumented e Reasonable alternative exponents give charges that can differ from the current ones by up to 100 e Axle spacing and lateral forces are not considered e Masonry and concrete bridges are not considered e The damage models are not calibrated to current service data A full revision of the existing methodology is recommended This should result in state of the art damage models calibrated to the latest service data The new methodology should be fully documented and traceable The revision should be completed in time for Control Period 5 A new methodology has been developed to recover costs of damage caused by tangential wheel rail forces Stakeholder comments and suggestions have been considered in the development Sensitivity studies on the key assumptions show that the new methodology is robust In contrast to the existing methodology reasonable alternative assumptions give charges that vary by less than 10 For all except a few special cases the variation is less than 5 A methodology has been developed for calculating variable usage charges at the national or route level It uses traffic vol
39. e apportionment of variable usage costs which is independent of the amount of the variable usage cost Prepared by TTCI UK Ltd 27 March 2008 Page 54 m y UA P Methodology to Calculate Variable Usage Charges for Control Period 4 15 The report discusses IPR trade secrets BT is fully prepared to provide T gamma tables such as Table 1 in Section 3 in TTCI UK NR Report No 07 003 We have already provided this information to NR for the Class 221 Of course we would not expect this information to be widely circulated or published However we don t not consider that measuring suspension stiffnesses and then publishing them respects IPR Some stakeholders have said suspension stiffness does represent IPR 16 The cost associated with the existing vertical track charge is fully in the hands of the train manufacture It only depends on the vehicle design speed axleload amp unsprung mass Each Supplier is free to optimise as he see fit Therefore it is a fair system although the incentives to fund innovation may be too small Noted I7 optimisation As you are well aware the compromise between stability and curving performance has always been most important parameter that is a required input conicity How can a vehicle Supplier optimise the train as required by the new RSD term if NR is not taking any responsibility for rail conicity Our problems on Hull Trains are well known Our Contract specified 0 4 operating coni
40. e assumptions made in the analysis Network Rail should demonstrate that these costs are de coupled from the other drivers for grinding and renewal work such as condition of ballast sleepers pads fastners etc and condition of welds It is not clear that use of ICM is able to deliver the appropriate cost data to be compatible with this analysis The ICM is used to calculate variable usage costs for track and structures A percentage of these costs is assumed to be due to tangential forces It is proposed to recover this percentage with the new methodology Other drivers for track and structure maintenance and renewal are handled by the existing methodology based on vertical forces Table Bl As with grinding depth Network Rail should provide data to support the assertion that a mean of 10mm side loss limit reflects the condition of rail that is renewed because of side loss Varying these assumptions would shift costs between grinding and renewal A sensitivity study on ratio between these costs is presented in the report The sensitivity to the assumed profiles is considered small in comparison with existing accepted sensitivities Table Bl ATOC assume that these crack and wear damage rates are simply indicative values These values were taken from the Whole Life Rail Model Ref B1 Prepared by TTCI UK Ltd 4 27 March 2008 Page 65 gt e71 Vr B Us Lt Methodology to Calculate Variable Usage Charges for Control Period
41. e costs have been estimated by considering each line item in the track maintenance and renewals budgets Table 11 shows the results of this analysis The costs shown in Table 11 are network totals after the Office of Rail Regulation cost reduction targets have been applied Table 11 Rail Surface Damage Costs CP4 Year 1 Total RCF Squats Wear Other Plain Line Renewals Percentage 2 5796 1 35 3 91 92 17 Cost k year 530 631 13 622 7 140 20 762 489 108 S amp C Renewals Percentage 5 15 0 00 7 69 87 16 Cost k year 181 415 9 348 0 13 943 158 124 Maintenance and Inspection Percentage 5 10 2 79 1 05 91 06 Cost k year 470 774 24 031 13 128 4 940 428 675 Totals Percentage 3 97 1 71 3 35 90 96 Cost k year 1 182 821 47 001 20 268 39 645 1 075 907 Thus 92 17 of plain line renewals costs 87 16 of S amp C renewals costs and 91 06 of track maintenance costs will be handled using the existing methodology i e using Equation 1 The rail surface damage costs will be handled with the new methodology described in subsection 3 1 Structures costs which amount to 371 320 k year will continue to be handled with the existing methodology i e using Equation 3 Signalling costs are 121 620 k year The variable part of these will continue to be added to the track costs for variable usage charging The annual cost of rail surface damage from Table 11 is 106 9M App
42. ed 2Axle_loaded 3 Wagon equipped with three piece 3Piece_empty bogie 3Piece_loaded Bogie wagon with enhanced three 4 piece bogie e g swing motion or NACO empty parabolic four wheel wagon NACO loaded Basic wagon with primary springs Y25 empty 3 e g Y25 suspension type Y25 loaded Wagon equipped with enhanced Y25 empty 6 primary springs i e low track force Y25 loaded bogies TF25 types or axle motion 7 Wagon equipped with enhanced No such wagons are currently in primary springs and steering operation Prepared by TTCI UK Ltd 27 March 2008 Page 32 Methodology to Calculate Variable Usage Charges for Control Period 4 3 4 Traffic Sector Results Table 18 shows the total variable usage charges for the passenger and freight sectors These charges are based on the 2006 07 traffic volumes and mix of vehicle types Table 19 shows the percentage contributions of each sector towards the total charge Table 18 Traffic Sector Charges k year Existing Vertical New Including Forces Only RSD Term Passenger 178 692 190 277 Freight 94 507 82 922 Total 273 199 273 199 Table 19 Traffic Sector Contributions Existing Vertical New Including Forces Only RSD Term Passenger 65 4 69 6 Freight 34 6 30 4 Table 18 shows that with the assumptions stated here the effect of introducing the new rail surface damage term into the
43. eees 67 m E UA P Methodology to Calculate Variable Usage Charges for Control Period 4 1 INTRODUCTION AND OBJECTIVES TTCI UK Ltd under contract to Network Rail has continued to develop the methodology for calculating variable usage charges for track access This report reviews the existing methodology before describing a new methodology The new methodology accounts for the variable usage costs arising from tangential forces between the wheel and the rail Sensitivity studies on both the existing and new methodologies are described Detail is given on applying the new methodology to routes rather than the network in total This report builds on earlier work and leads to a final set of proposed variable usage charges for Control Period 4 CP4 These charges are published separately Section 2 is a review of the existing methodology used to calculate variable usage charges The intention is to give an overview and to identify any changes that can be justified for CP4 A complete revision of the existing methodology is outside the scope of this report Section 3 describes the proposed new methodology which still uses the existing methodology for a proportion of track costs It introduces a new method for calculating variable usage charges to recover the costs of damage caused by tangential forces Tangential forces are responsible for rail wear rolling contact fatigue and squat defects Variable usage costs arising from these forces are a
44. endly vehicle types Similarly the charging regime puts an onus on Network Rail to maintain and renew track in the most cost effective manner through greater optimisation of possession opportunities Noted Comment 1 2 The report proposes that the current variable charge which reflects only vertical forces will also need to incorporate a lateral tangential element as well Although the Variable Access Usage Charge remains effectively the same it is proposed that the methodology which addresses track and structure damage from vertical forces also needs to incorporate the damage from lateral and longitudinal effect Variable usage charges would be more cost reflective if they incorporated damage from tangential forces Comment 1 3 VWC accepts that this revised methodology could be implemented if certain fundamental issues are resolved which did not result in Train Operators potentially bearing the cost of Network Rail s inefficiencies in maintaining the Network The existing track access charging regime is applied nationally to fleets across the entire network also see part 3 1 below which means that where major projects may have taken place ie the WCML particularly with the ongoing high levels of renewal supported through the gradual implementation of enhanced maintenance regimes we still continue to bear the costs of a route where wear amp tear costs are much lower than the national rate for any given fleet This argumen
45. equired in choosing where to locate curvature class boundaries The intention is to separate the 300 SRSs into 15 groups to keep variable usage manageable Inevitably some SRSs will be close to the boundary between two groups and so the charges that apply may differ if a different boundary was chosen A different curvature distribution has been used in calculating the variable usage charges for each combination of route class and curvature class In each case a curvature distribution representative of the SRSs in that group has been derived The variable track and structures costs for each combination of route class and curvature class have been calculated from the sum of costs for the SRSs in each group Traffic information is not currently available at the SRS level in sufficient detail for variable usage charging The total tonnages of passenger and freight are available for each SRS These are summed for each combination of route class and curvature class The mix of passenger and freight traffic is assumed to be the same as the network total The number of each type of vehicle is adjusted to give the appropriate passenger and freight tonnages for each combination of route class and curvature class The results of calculating variable usage charges for the particular curvature distribution and cost in each group are published separately Prepared by TTCI UK Ltd 27 March 2008 Page 37 10096 m E UA P Methodology to Calculat
46. er vehicles and a Ty column for cant excess used for freight vehicles This might reduce the current prospective finance shift between the two sectors A sensitivity test should be undertaken It is now assumed that passenger vehicles operate with cant deficiency and freight vehicles operate at balance speed or with cant excess A sensitivity study into the effect of this assumption is presented in this report asked how tilting trains would be assessed An alternative set of Ty numbers will be agreed for special cases such as tilting trains Stated that a clear understanding is required of how the charging regime is built up in order to give confidence in it and drive the appropriate behaviour This report and other publications are intended to provide that understanding questioned whether the issue of wheel profile had been considered Mid life worn wheel profiles are used in the calculations A sensitivity study on the effect of wheel profile is presented in this report asked if this a review of the vertical force component of the track access charge was being done Prepared by TTCI UK Ltd 27 March 2008 Page 50 gt e71 el Lt Methodology to Calculate Variable Usage Charges for Control Period 4 hy a An initial review of the existing methodology is presented in this report A detailed review is recommended before Control Period 5 said that manufacturers were being encouraged particu
47. erent usage charge Section 6 1 Technical Report Section 2 The document asserts that rail surface damage depends on the curving performance of a vehicle paragraph 2 of section 2 2 This is misleading because it fails to mention that rail surface damage is also highly dependent on the performance of Network Rail in managing the infrastructure and the standards and parameters to which this work is delivered If track is maintained by Network Rail to a reasonable geometric quality there will be no identifiable rail surface damage rcf or side wear on track that is straight or curved with a radius in excess of 2 500m Typically only around 20 of the network will consist of curves with a radius of less than 2 500m Furthermore the degree of rail side wear on curves that are tighter than 1 000m is highly dependent on both the installation and the performance of track lubricators a Network rail responsibility The established philosophy is to base usage charges on a network maintained according to standards If the standard calls for lubricators on curves less than a certain radius of curvature then charges should be based on the lubricators being in place and operational The use of T gamma tables and the distribution of track in bands of curvature deal with the issue of less RSD on straight and less curvy track Section 2 refers to a complex technical analysis in Appendix A that purports to link calculations of Teamma the curving per
48. ess but this is balanced by a reliable low maintenance solution Within Europe high yaw stiffness bogie such as TGV are fitted with on train wheel flange lubrication does NR consider that on train lubrication should be encouraged by a track charge incentive This question requires further study In principle if benefits could be identified from on train wheel flange lubrication these would be shared between Network Rail and the train operator Siemens email 2 January 2008 1 it is stated that parties would have to submit the accompanying modelling file if they want to change the T gamma tables by submitting own calculations Additionally it is stated that a similar process will apply for new trains Our question is how is it secured that these data will be handled confidential Which type of file is needed just a file which is describing the model e g a Word Dokument or a file which can be used by a simulation progam The intention is that parties can provide their own T gamma tables provided they can demonstrate the calculations have been performed correctly The supporting evidence would be kept secure and confidential by Network Rail 2 The Class 185 DMU cars are each equipped with one motor bogie and one trailer bogie How has this be considered for the assumption of the curving class and the determination of the unsprung mass The Class 185 has been considered as having both bogies powered Prepared by TTCI UK
49. etc 2 Component fatigue which drives renewal costs for rails sleepers etc 3 Rail wear and RCF which drive grinding and rail renewal costs Terms 1 and 2 include a dynamic vertical force parameter that can be assumed proportional to speed For both terms the dynamic vertical force parameter is raised to the power 3 0 For the settlement term this power comes from collaborative European railway research For the fatigue term this power is typical of the relationship between applied stress and fatigue life of welded metal components Since the wear and RCF term does not include a speed factor although some assumption about speed in curves must be made and the other terms include a static force parameter the overall effect of speed will be an exponent somewhat less than 3 0 Table 3 summarises the speed exponents used by the various sources described above Clearly there is a wide variation and this can be expected to have a significant effect on the variable usage charges for high and low speed vehicles The results of a sensitivity study on the speed exponent are presented in subsection 2 4 1 Prepared by TTCI UK Ltd 27 March 2008 Page 6 m P Methodology to Calculate Variable Usage Charges for Control Period 4 ZH e HUS ii ie a m E UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 Table 3 Summary of Speed Exponents Source Exponent ORR Variable Usage Charging 0 64 Railway Group
50. f maintaining and renewing track Included is the cost of grinding and rail renewal that will always be part of railways operations Comment 5 1 3 Effect of Curve Radius on Rail Surface Damage It is an established fact that Network Rail s maintenance standards require the attainment of a reasonable level of geometric quality This quality if sustained would result in no identifiable RSD damage either vertically or horizontally to the network on track that is straight or with a curve radius above 2500mm This results in only approx 20 of the network having some resultant vehicle damage wear amp teat reducing further to those sections of track with a radius of less than 1000mm which again in terms of installation and lubrication is Network Rail s responsibility The T gamma tables on which the methodology is based account for the reduction in forces at large curve radii The damage at each band of curve radius is factored by the length of track in the band Thus the effect of curve radius on RSD is handled correctly Comment 5 1 4 Handling Pendolino Trains Furthermore in conjunction with the VWC Class 390 Pendolino fleet there is no reference within the report that when such rolling stock operates in tilt mode at linespeed it mitigates against the effect of RCF whilst compensating for any lack of bogie steering capability This concept has been demonstrated during track force testing conducted with Pendolinos and has been pre
51. following equation costs k damage vehicle miles Thus we could not find the link between these approaches We would like to ask for further information regarding this issue A hybrid approach has been followed The total variable cost that should be recovered through the usage charge is known It is apportioned between the vehicles using the network according to engineering analysis of the damage they each cause Freightliner Ltd letter 13 February 2008 Wagons and Laden Weight We are concerned with regard to the Average Loaded Weights given to the intermodal wagons within the model It would appear that the weights given are broadly the maximum gross weight allowed on those wagons We understand that this maybe appropriate for other commodities e g coal but request recognition that intermodal wagons are rarely loaded to their full weight capacity Weight assumptions will be checked Speed of Class 86 and Class 90 locomotives We would suggest that on the freight side of the coin that locomotive speeds are limited to the speed assumption for the freight wagons they haul Furthermore we would suggest that there is no difference between the loaded or empty running speed of intermodal trains Speed assumptions will be checked Absence of track friendly TF25 bogie It is with regret that we noticed the absence of the TF25 bogie that is considered to be track friendly and as such rewarded for use We are curren
52. formance of a vehicle to track damage costs At the heart of this analysis is the belief that grinding is carried out in proportion to the amount of crack damage Whilst this forms part of Network Rails aspiration for their rail grinding programme it is believed that Network Rail s real performance in grinding rails is quite different to their aspirations Therefore a fundamental part of the technical analysis is not rooted in fact e Network Rail does not have enough grinding trains to cover all of the network some track will never be ground e Network Rail grinds both straight track where there should be no need to grind for rail surface damage and curved track where only curve radii below 2 500m will suffer rail surface damage and require grinding The charges are based on the long run grinding requirements and not what is currently achieved There appear to be a number of detail inconsistencies within Appendix A Further specific comments on Appendix A are welcome Prepared by TTCI UK Ltd 27 March 2008 Page 47 m A UA P Methodology to Calculate Variable Usage Charges for Control Period 4 The document has a key technical assumption that rail surface damage is caused primarily by tangential forces paragraph 2 of section 2 2 This is untrue Rail surface damage also includes the growth of defects known as squats that are not linked to tangential forces Around 40 of all rail replacements are due to squats Any co
53. g axle Further factors are applied for freight wagons to account for coal spillage and suspension types The coefficients and exponents in Equation 1 were derived from a parametric study using Mini MARPAS Results were calculated for many combinations of speed axle load and unsprung mass Power law relationships were then fitted to these results It should be noted that the axle load exponent in Equation 1 is effectively 1 49 since expanding the gross tonne mile term gives Equation 2 GTM A x number of axles x miles operated 2 As detailed in the 1999 ORR review the methodology for track costs makes several simplifying assumptions In summary it ignores the effects of e Tangential wheel rail forces generated in curving which require knowledge of wheel and rail profiles curvature and cant Inclusion of these effects is one of the principal changes proposed for Control Period 4 e Wheel flats and out of roundness Incentives are given to train operators outside the variable usage charge to minimise these types of wheel defects Prepared by TTCI UK Ltd 27 March 2008 Page 3 m E UA P Methodology to Calculate Variable Usage Charges for Control Period 4 e Axle spacing although a factor is included in the structures methodology see subsection 2 2 e The design of vehicle s secondary suspension except in a simplified way This may affect vehicles propensity to cause cyclic top track damage e Lateral forces and the
54. ge 62 m y UA P Methodology to Calculate Variable Usage Charges for Control Period 4 The simplifying assumption that has been made is that a vehicle with the propensity to cause RSD in curves will have a similar propensity to cause it in S amp C e Rail grinding is assumed to remove rail surface to match the depth of rcf cracks when cracks are at a critical size This does not seem to be a practical way of managing rail grinding The document makes no assessment of the possible benefits of earlier grinding when cracks are smaller and growing more slowly The use of the word critical in the report was an error The assumption is that grinding takes place when the cracks can be removed in a typical grinding cycle 0 5 mm depth of grind e The proposal calculates Ty and hence wear and rcf damage using surprising values for friction coefficients A sensitivity study on coefficient of friction is presented in the report The sensitivity to the assumed values is considered small in comparison with existing accepted sensitivities 4 4 1 NR should advise ATOC on the extent to which current knowledge of rcf calculations of rail rcf on TPE routes might affect the proposed method for calculating TACs If the effect is considered to be substantial further actions may be necessary Route based charging is no longer being proposed for CP4 4 4 2 ATOC would like NR to demonstrate that the cant deficiency assumption used in the proposal will st
55. ge and thus should not be associated with RSD costs Although Squats may be initiated by some other cause they are propagated by the same forces that produce rolling contact fatigue and wear Comments 5 2 1 and 5 2 2 Route based Charging The data that is identified in this section purports to representing just one type of rail vehicle over a number of differing curvatures and cants The various connotations of RSD effect caused by just one vehicle operating over a number of track type variables gives weight to applying a vehicle route specific charge as a national average figure per fleet vehicle type would be unrepresentative of the actual effect mused by such vehicles Therefore route based charging is necessary but to what depth and complexity does a charging regime have to be to capture most or all of the RSD It is agreed that introducing a RSD term strengthens the already existing argument for route based charging The RSD methodology needs to be at least as deep and complex as the current vertical damage methodology Comment 5 3 1 Cost Transparency It is apparent from the base data used in this section that most of the data where understood is derived from internal Network Rail sources which does not allow for transparency and the consequential ability to agree or refute the figures being used To this end how can VWC be sure the figures correctly represent accurate traffic growth and uplifted rates The traffic an
56. grinding cost and the annualised renewal cost Cost 7 GrindingCycles x Grinder Casi Renewail os B2 Year Year RailLife Since grinding is performed when the cracks in the railhead reach a critical size i e RCF Damage 1 GrindingCycles _ RCFDamage Axles B3 Year Axle Year Rail is commonly renewed when the material lost on the side of the rail reaches a specified limit Thus RailLife pier Osa SideLossLimit B4 SideLoss Year SideLoss Axle x Axles Year Now Cost _ Cost Axles B5 Year Axle Year Substituting Equation B3 B4 and B5 into Equation B2 gives Cost RCFDamage r SideLoss RenewalCost B6 Axle Axle Axle SideLossLimit where RCFDamage _ CrackDamage WearDamage Axle Axle Axle for 0 lt Ty 175N B7a RCFDamage _ 0 for Ty 2 175 N B7b Axle and Prepared by TTCI UK Ltd 27 March 2008 Page 68 _u_ UA Methodology to Calculate Variable Usage Charges for Control Period 4 za QM COD OARS hy ort SVEN B8a Axle D LAU OMRE CrackDamageRatelT y 15 for Ty 15 N B8b Axle and Perae ON B9a Axle WearDamage a WearDamageRate T y 65 for Ty gt 65 N B9b xle At Ty 175 N the crack and wear damage rates are equal Thus 160 WearDamageRate 110 x CrackDamageRate B10 Now the amount of material lost from the side of the rail for each axle is the sum of the depth ground and the depth worn for each axle SideLoss DepthGround n Dep
57. gure 3 100 8096 4 2 2 nn nn en nn enn nen en eee ene ene O A Change in KGTM 60 4 naa 80 4 100 Locomotives Wagons Speed Exp 0 28 Speed Exp 2 0 Figure 3 Effect of Track Speed Exponent on Empty Freight Vehicle Variable Usage Charges Prepared by TTCI UK Ltd 27 March 2008 Page 10 Z a E LIA Methodology to Calculate Variable Usage Charges for Control Period 4 TM AA 100 BO Yo ee ee i ee ie epee 60 ea D a ees o x el E gt c S 20 55 m a S aU RC mo aU 7 ap ee o a T m a m um Oo a ee aare a n ai 40 Lx Poem lee ppc a E vine E s E 50 SAA NN aa 100 Wagons Speed Exp 0 28 Speed Exp 2 0 Figure 4 Effect of Track Speed Exponent on Loaded Freight Vehicle Variable Usage Charges Figures 3 and 4 show that increasing the speed exponent in the equivalent track damage formula generally reduces freight vehicle variable usage charges This is because freight vehicles generally operate at lower speeds than passenger vehicles Thus increasing the significance of speed penalises the high speed vehicles to the benefit of the slower speed freight vehicles Freight vehicle variable usage charges are sensitive to the exponent of speed used in the equivalent track damage formula They can change by 20 to 50 of the current
58. h 2008 Page 58 gt e71 wl Lt Methodology to Calculate Variable Usage Charges for Control Period 4 hy amp guidelines may be changed if they do not apply in a specific case e g higher cant deficiency for tilt trains All the information required will be made available We would also suggest that Nework Rail will publish a parameter sheet and the Tgamma results of a benchmark vehicle Any party could then check their calculations first with this benchmark vehicle if these results are comparable with theose which are published by Network Rail Noted 8 Tables 13 and 14 We would like to please Network Rail to transmit us the profile files How are the profiles for the left and the right wheel different Which difference does this have on the curve negotiation right hand or lefi hand curves The left and right wheel profiles used for the freight vehicle modeling are the same The left and right wheel profiles used for the passenger vehicle modeling are very similar and there is no significant difference between the T gammas generated on the left and right hand curves 9 Remark to Chapter 3 2 3 Especially Figure 19 shows a disadvantage of the top down approach It is generally known that due to the reduction of the friction coefficient by lubrication the rcf and wear damage regarding to the Tgamma model is reduced Hence also the costs for the track maintenace for the repair of the rcf and wear damage will
59. here is a smaller effect on the other vehicles even though their wheel profiles did not change This is to compensate for the general increase in charges for the wagons with Y25 bogies Although it is proposed to base the usage charges on semi worn profiles it is recognised that some profiles such as the WRISA2 wheel profile can have a significant effect on rail surface damage When these profiles have proven successful the variable usage charges for vehicles using them could be suitably adjusted 3 3 Freight Charging Categories Freight vehicles come in many combinations of wagon body and bogie type Primary yaw stiffness and axle spacing are still key parameters in determining curving performance but the inclusion of friction damping and clearances in the suspension complicates the calculation of freight vehicle curving performance Eight categories are currently defined to overcome the complexities of freight vehicle charging for vertical forces Table 17 shows how these categories are proposed to be mapped onto curving classes for the new rail surface damage methodology Table 17 Freight Vehicle Curving Classes Category Description Curving Class Class60 for conventional 3 axle bogies Class66 for 3 axle bogies with steering linkages Freight locomotives 4 Four wheel 2 axle wagon with 2Axle_empty pedestal type suspension 2Axle_loaded 2 Four wheel 2 axle wagon with leaf 2Axle_empty springs friction damp
60. herefore believe that generic examples of route primary secondary etc based on the four curvature types mixed straight etc are a reasonable place in which to start considering a RSD charge This will maintain transparency through a simplistic but relative charging methodology akin to the Train Operators operational area This recommendation has been adopted Comment 3 1 One specific example pertinent to VWC and representative of those issues surrounding the technical complexity of introducing a RSD charging mechanism is in conjunction with tilting train operation This is explained in detail in section 5 1 4 below The proposed new methodology can handle specific cases such as tilting trains Prepared by TTCI UK Ltd 27 March 2008 Page 42 7 ares Lt Methodology to Calculate Variable Usage Charges for Control Period 4 hy amp Comment 5 1 1 VWC considers that certain technical assertions made by TTCI in this part of the report are incorrect This comment is assumed to be an introduction to the specific comments that follow Comment 5 1 2 Effect of Track Maintenance and Standards RSD is not only caused as a factor of railhead wear amp tear it is also reliant upon the level of track maintenance and the standards to which such work is undertaken This issue is more general than the introduction of a rail surface damage RSD term The methodology aims to recover the long run variable cost o
61. ht Vehicle Variable Usage Charges Prepared by TTCI UK Ltd 27 March 2008 Page 13 Methodology to Calculate Variable Usage Charges for Control Period 4 Figure 5 shows that increasing the value of the axle load exponent in the equivalent track damage formula increases the variable usage charge for passenger locomotives and generally reduces it for other passenger vehicles Figures 6 and 7 show the same is true for freight locomotives and loaded freight wagons Figure 6 shows the effect on empty freight wagons is to generally reduce variable usage charges Increasing the significance of axle load can be expected to increase variable usage charges for heavy axle load vehicles The alternative values of the axle load exponent used here are not unreasonable They result in 10 to 60 increases in charges for locomotives Charges for passenger coaches and multiple units can be up to 60 lower Freight wagon charges show the greatest sensitivity to the axle load exponent in the equivalent track damage formula Empty freight wagons because of their low axle loads generally benefit from increasing the value of the exponent The reduction in variable usage charge can be as much as 80 when the exponent is raised to 2 0 The variable usage charges for loaded freight wagons generally increase significantly when the axle load exponent is increased The increase can be as much as 100 The exceptions are certain combinations of wagon types and commodit
62. ife leaving fresh material at the running surface of the rail A principal purpose of grinding is to remove cracks when they are small and before they propagate to transverse rail breaks Prepared by TTCI UK Ltd 27 March 2008 Page 64 m A UA P Methodology to Calculate Variable Usage Charges for Control Period 4 n yq _ HS e Rail wear is a consequence of vehicle induced forces on curves that are lubricated according to Network Rail standards It is an accepted principal that variable usage charges are based on the network maintained according to the required standards Even lubricated rails wear Appendix B Given the earlier assumption that the exclusive purpose of grinding is to eliminate rcf damage it seems illogical for the analysis to assume that grinding and rail wear will occur together in one location There is a region in the whole life rail model T gamma between 65 and 175 N where wear is occurring but cracks are growing at a faster rate In this case it is necessary to supplement natural wear by grinding to remove RCF Equation B13 The appearance of this equation in the document is preceded by two qualifying statements This suggests that equation B13 can only be used in situations where the qualifying statements are true This derivation has been changed to avoid the need for the qualifying statements Table Bl It is important that these values renewal and grinding costs are related to th
63. ification charges are proposed Subsections 2 1 and 2 2 describe the methodologies used for recovering the variable usage costs from existing track and structure Subsection 2 3 describes methodologies used in other railways and subsection 2 4 presents a sensitivity study on the choice of methodology 2 1 Track The methodology for track that has been used since the start of Control Period 1 is based on the British Rail Research Mini MARPAS track deterioration models There are seven basic track damage submodels 1 Rail Maintenance 2 Rail Life 3 Sleeper Life 4 Track Geometry Maintenance 5 Ballast Life 6 Switch amp Crossing Ti Track Inspection Each model calculates damage from a specified traffic The units of damage may be a volume of work or the proportion of an asset s life consumed Damage is converted into cost using unit cost data for the various types of work e g re railing inspection Theoretically the Mini MARPAS models could be used to calculate usage charges directly This would be the so called bottom up approach Costs could be calculated for a base case The marginal cost of a small increase in the number of a particular vehicle could also be calculated This would give the marginal cost of operating one more of that particular vehicle Prepared by TTCI UK Ltd 27 March 2008 Page 2 Methodology to Calculate Variable Usage Charges for Control Period 4 Y IN ti amp The charges
64. ill permit the calculation of adequately cost reflective TACs A sensitivity study on cant deficiency is presented in the report The sensitivity to the assumed values is considered small in comparison with existing accepted sensitivities Exceptions to the general assumption on cant deficiency will be made for special cases such as tilting trains 4 4 3 ATOC would like NR to demonstrate that a broader and more probable set of friction coefficient values do not produce significant changes in TAC relative to those based on friction coefficients assumed in the document This request will be considered 4 4 4 ATOC would like NR to demonstrate that the rail and wheel profiles assumed will still permit the calculation of adequately cost reflective TACs A sensitivity study on freight vehicle wheel profiles is presented in the report The sensitivity to the assumed profiles is considered small in comparison with Prepared by TTCI UK Ltd 27 March 2008 Page 63 m A UA P Methodology to Calculate Variable Usage Charges for Control Period 4 existing accepted sensitivities Assuming typical mid life wheel and rail profiles is considered a reasonable approach to deriving cost reflective variable usage charges and is consistent with the philosophy of basing these charges on the as maintained condition of track and vehicles 4 5 ATOC would like NR to provide comparisons between the proposed route and traffic data and data from ACTRAF and e
65. ively Prepared by TTCI UK Ltd 27 March 2008 Page 8 m A UA P Methodology to Calculate Variable Usage Charges for Control Period 4 2 4 Sensitivity to Damage Model Assumptions 2 4 4 Speed Exponent The current formula for equivalent track damage Equation 1 raises speed to the power 0 64 Figure 2 shows the effect on passenger vehicles of changing this exponent to 0 28 and 2 0 Two values are plotted for each of the 117 different types of passenger locomotives coaches and multiple units One value is the percentage change in variable usage charge when the speed exponent is reduced from the base case value of 0 64 to 0 28 The second value is the percentage change when the exponent is increased to 2 0 Thus the scatter in results indicates the sensitivity of the charge to the speed exponent Higher sensitivity produces more scatter 100 E a 80 E 60 L d D LI E T m 2 40 5 MM i e om e 0 amp 20 y doo 2 E ee e Oa ee o Le sa e o hed A e S 0 mm we DG ei e oT A Cn oo oU mwa ot Garr eee ee wem c o O SK wa BS oe P 20 e PEE e n PTT o m En Es G 1 a m o 40 m P Paeesettesuate Ni ocectesenst eese i m mu TUS eer Sn JM o m 30 am O 100 Y Locomotives Wagons Speed Exp 0 28 E Speed Exp 2 0 Figure 2 Effect of Track Speed Exponent on Passenger Vehicle Variable Usage Charge
66. larly by the revised TAC to provide bogies with good curving performance but when stability problems caused by high equivalent conicity between the wheel and rail are encountered eg because of tight gauge track or worn rail profile manufacturers Operators are being left with a problem This was not a tenable position and someone must accept responsibility for controlling maximum conicity levels The variable usage charges are based on the assumption that the infrastructure is maintained according to Railway Group Standards V T TAG Minutes 23 17 July 2007 Minute 4 TG questioned why mode 2 defects would not be included A vehicle s propensity to cause rail surface damage is expected to be demonstrated by its steady state curving performance said that the propensity to cause RCF Mode 0 1 and 2 could be caused by different parts of the wheel Agreed The steady state approach has been chosen to give an appropriate level of complexity said that he understood that squats were initiated subsurface and that it was vertical loads that caused them Although there are different views on what initiates a squat it is generally agreed that tangential forces are responsible for propagating cracks from the initial defect questioned how S amp C costs will be dealt with The proportion of S amp C costs due to damage from tangential forces has been estimated and included in the variable usage charge said that track
67. lations would be subject to technical audit It is also widely acknowledged by the Industry that the WCML is the busiest Freight route in the country not necessarily in terms of the heaviest hauled services but in terms of the timetable quantity length and volume of services operated Combine this with the intensive use south of Rugby and in the NW with passenger services and you immediately see that wear amp tear is high Although appropriate maintenance and renewal provisions are will be in place there is nevertheless the issue that any VTA charge needs to be not only cost reflective in terms of average RSD but also relevant to the tonnage and damage covered over any particular route This gives more credence to our belief that individual route vehicle value application is necessary Route based charges are no longer proposed for CP4 Prepared by TTCI UK Ltd 27 March 2008 Page 66 E UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 Appendix B RAIL SURFACE DAMAGE FUNCTION Rail surface damage is calculated using a function developed by AEA Technology Rail Figure B1 shows this function separated into its two components of crack damage and wear damage 70 60 50 P 40 30 20 10 0 Damage 1 0 S axle 0 50 100 150 200 250 Ty N Crack Damage Wear Damage Figure B1 Separated Rail Surface Damage Functions The paramete
68. ll passenger vehicles except those that tilt are assumed to operate with a cant deficiency of 40mm Tilting passenger vehicles are handled as special cases AII freight vehicles except those whose maximum speed are 45mph are assumed to operate at balance speed Freight vehicles limited to 45mph are assumed to operate with 20mm cant excess The values of cant deficiency and excess were obtained from an analysis of cant curvature and line speed on Network Rail controlled infrastructure Figure 12 shows the distribution of cant deficiency for the network assuming vehicles can reach the local line speed The average of this distribution is approximately 40mm 1 200 000 _ 1 000 000 0 p S 800 000 x i E 600 000 5 s 5 400 000 5 d 200 000 x o de 2 d d d xP SP qe A Cant Deficiency mm Figure 12 Distribution of Cant Deficiency Assuming Local Line Speed Figure 13 shows the distribution of cant deficiency on the network assuming the maximum speed is limited to 60mph The average of this distribution is approximately zero 1 600 000 WO don RP Ett A 30800 A Sena 1 000 000 2 2 4 222 lot A BR 600 000 lMllbld amp 83psp 400 000 ecco AEN a Length of Track yards 200 000 HAM IA AA a Wwe SP 9o e Pd deg Cant Deficiency mm Figure 13 Distribution of Cant Deficienc
69. n for Network Rail controlled infrastructure that has been used to calculate rail surface damage costs The data is interpolated from the percentage points on the cumulative distribution of curvature taken from Network Rail s ICM The ICM gives cumulative percentages of track for curve radii 800 1500 and 2500m Table 13 Curve Distribution for Network Rail Controlled Infrastructure Radius m Track km 0 300 83 301 500 332 501 700 775 701 900 1872 901 1100 1187 1101 1300 1187 1301 1600 1187 1601 2000 1878 2001 2400 1712 2401 2800 1712 2801 3500 1712 3501 5000 4280 More than 5000 11088 Total 29094 Equation 4 is applied to the steady state Ty from all wheels on the high rail before summing to give the total damage for the vehicle For the purpose of Prepared by TTCI UK Ltd 27 March 2008 Page 21 m UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 calculating variable usage charges for Control Period 4 all vehicles operating on the network have been assumed to be represented by one of the curving classes shown in Table 14 Computer models have been used to determine the 7y produced by each curving class for the central curve radius in each band shown in Table 13 Table 14 Curving Classes Used to Determine Ty Vehicl
70. nge of different plan view damage mechanisms eg wear or rcf Agreed 12 Care is needed that the parameter used incentivises the use of vehicles that are actually beneficial from a system view and does not lead to perverse behaviour I am particularly concerned that use of a network average could lead to vehicles running on routes for which they are totally unsuitable Some form of charge tailored to route or route type seems necessary to avoid this The proposed rail surface damage variable usage charge will incentivise the use of vehicles with good curving performance Providing high speed stability can be achieved where necessary this is believed to be a benefit for the system 13 Some of the input data used seems somewhat arbitrary eg wheel rail friction levels cant deficiency balancing speed wheel profiles new for some vehicles worn for others and I would wish to see a sensitivity study to clarify whether the relative charges between different vehicles are significantly affected by these choices These sensitivity studies are included in this report 14 Section 2 figure 1 in TTCI UK NR Report No 07 003 appears to suggest that the total variable charge will remain the same but in future will be made up of two parts Vertical Tangential BT would like to see an increase in the variable part together with a reduction in fixed part to provide more incentive and funding for technical innovation This report discusses th
71. of review in that e vehicle type charges are national rather than route specific as recommended by V TSIC e an estimate of average or typical operating speeds is used for each vehicle type leading to under and over recovery Agreed Section 5 Requirements Network Rail propose to establish a new track access charging regime that extends the existing regime to include track damage costs and vehicle parameters that relate to rail surface damage Passenger train operators might support this change provided that a new regime e Is based on a broadly accepted technical rationale e Delivers true incentives to introduce vehicles that cause lower levels of track damage e Cost changes are embraced within existing franchise agreements The proposed methodology attempts to satisfy the first two requirements It is assumed that existing arrangments it has not been possible to determine what change in the proposed RSD VTA charge term would occur as a result of reducing the primary yaw stiffness of the MKIV fleet The use of T gamma tables is an attempt to simplify the data requirements while still reflecting the differences in performance between vehicles Reducing the Prepared by TTCI UK Ltd 27 March 2008 Page 46 gt e71 Sr B Us Lt Methodology to Calculate Variable Usage Charges for Control Period 4 hy a primary yaw stiffness of the Mk IV vehicle would result in a different T gamma table and hence a diff
72. onsidering the variable cost of grinding to remove cracks and the variable cost of rail renewal when too much material has been lost by grinding and wear In the process it accounts for the beneficial effects of grinding on rail life The assumed unit costs of grinding and rail renewal are 2 000 km and 250 000 km respectively The absolute values are not important for variable usage charging but charges may be sensitive to the ratio of these values The effect of doubling and halving the ratio of grinding to rail renewal costs has been studied Figure 15 shows the effect of halving and doubling the unit cost of grinding rail on passenger vehicle variable usage charges The effect is relatively small The absolute maximum percentage change is less than 396 from the base case Prepared by TTCI UK Ltd 27 March 2008 Page 25 m E UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 20 15 4 EEE i S ATE AE ANNE AEA EA A EE E e afe E ee SS a Ue iE 2 z o uus EDD gud cc EHI II amp o Va Rum nn ma a a 0 puro P e V ARA PT a TEN a LLL C EE c E o Dv Max DE A 15 A SSS SSeS SSeS SSS SSeS SSeS SSeS SSeS SSeS II 20 Locomotives Other Grinding Cost 1 000 km Grinding Cost 4 000 km Figure 15 Effect on Passenger Vehicles of Rail Grinding Unit Cost Similar results were found for freight vehicles Thus it can be concluded that the new
73. operating at current traffic levels 6 Curving performance is a function of more than yaw stiffness and a rough formula but beware formulae when thinking about curving is yaw Stiffness x wheelbase wheel radius x wheel radius Agreed The use of Ty tables avoids the need to use curving performance formulae 7 I don t understand the tables with straight missed curvy etc bars but some of the comments do bear out my fist comment about inclusion of really sharp curves The tables show the distributions of curvature for Straightish Mixed and Curvy track Any route on the network will fall into one of these groups Prepared by TTCI UK Ltd 27 March 2008 Page 53 eee al P Methodology to Calculate Variable Usage Charges for Control Period 4 mL vr B US TM Ltd 8 How are the traction characteristics of locomotives taken into account These are not considered because it is impractical to know when traction and braking is applied at particular curve radii 9 Tilt trains working at high cant deficiency reduce propensity to generate rail RCF How will the charge regime allow for this factor A Ty table will be used that reflects the particular curving performance of tilting trains 10 In principle I have no problem with the inclusion of a variable charge for plan view behavior Noted 1l The use of Teamma energy or some derivative of it seems sensible as it can encompass a ra
74. pproximately 1096 of the track variable usage costs and a similar magnitude to structures variable usage costs Section 4 describes an approach to route based variable usage charging as an alternative to the existing charging based on total network traffic and costs The new methodology can be applied the network totals but there are further advantages of applying it to specific routes Section 5 presents conclusions Comments and questions on the earlier work have been received in writing from stakeholders These together with responses are listed in Appendix A Prepared by TTCI UK Ltd 27 March 2008 Page 1 m m zu P Methodology to Calculate Variable Usage Charges for Control Period 4 MA vr BUS m Lid 2 REVIEW OF EXISTING METHODOLOGY In 1999 the Office of the Rail Regulator commissioned a review of the proposed access charging methodology for Control Period 1 That review noted that variable usage costs were separated into the following four headings 1 Track 2 Structures 3 Signalling 4 Electrification The model used for track and structures costs was the most advanced The variable part of the signalling costs was small and was lumped in with track costs for charging purposes The variable usage charge for electrification was combined with the charge for traction Signalling and electrification charges are not discussed any further in this report No changes to the handling of signalling and electr
75. r its contents EXECUTIVE SUMMARY TTCI UK Ltd under contract to Network Rail has continued to develop the methodology for calculating variable usage charges for track access This report reviews the existing methodology before describing a new methodology The new methodology accounts for the variable usage costs arising from tangential forces between the wheel and the rail Sensitivity studies on both the existing and new methodologies are described Detail is given on applying the new methodology to routes rather than the network in total This report builds on earlier work and leads to a final set of proposed variable usage charges for Control Period 4 CP4 These charges are published separately The sensitivity study shows the existing methodology based on vertical forces is very sensitive to the assumptions used to derive the charging formulae The new methodology in contrast is relatively insensitive to its assumptions A full revision of the existing methodology is recommended This should result in state of the art damage models calibrated to the latest service data The review and calibration should be fully documented and traceable The revision should be completed in time for Control Period 5 Table of Contents 1 Introduction and Objectives sssssssss HH 1 2 Review of Existing Methodology ccccccccccccccccccccccccnncnnnnnnnnnnnninnnnnnnininons 2 211 a tea ti te ei tame a eR M atone ARE eR EIN E T
76. r on the X axis of Figure Bl is a combination of lateral and longitudinal forces Tia and Tiong and creepages yi and Yiong as defined in Equation BI Ty ee ae Trong Long B1 Total rail surface damage is assumed to be a combination of crack damage and wear damage Figure B1 shows there are four regions to consider 1 For 0 lt Ty lt 15 N there are no maintenance or renewal costs associated with rail surface damage 2 For 15 lt Ty lt 65 N cracks are removed by grinding The amount of grinding is proportional to the amount of crack damage Eventually enough material is ground from the head of the rail to require the rail to be renewed 3 For 65 lt Ty 175 N less grinding is required to remove cracks since some are removed by wear Grinding and wear combine to remove Prepared by TTCI UK Ltd 27 March 2008 Page 67 m E UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 material from the head of the rail and the rail eventually requires renewal 4 For Ty gt 175 N grinding is no longer required since cracks are being removed completely by wear Eventually the wear reaches the point where the rail requires renewal The function for calculating the equivalent miles travelled by a vehicle in terms of rail surface damage is derived by considering the annual cost due to rail surface damage on a mile of track In general the annual cost for a mile of track will be the sum of the annual
77. rail surface damage methodology is relatively insensitive to the ratio between grinding and rail renewal unit costs 3 2 2 Cant Deficiency In the base case the passenger vehicles have a cant deficiency of 40mm and the freight vehicles with a maximum speed of 45mph have a cant excess of 20mm The sensitivity to speed in curves was checked by recalculating variable usage costs assuming all vehicles operate at the balance speed Figure 16 shows the percentage change in variable usage costs for passenger vehicles when they are assumed to operate at balance speed in curves For passenger vehicles the overall effect of reducing speed in curves is to increase the variable usage charge Except for the first data point which is the Class 08 shunter the change in variable usage charge is between 6 and 5 Prepared by TTCI UK Ltd 27 March 2008 Page 26 Se UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 am Qe Mus noter casino Variable Usage Charges for Control Ped d TM itii aa E 20 15 4 gt 10 4 2 5 Dh poLteeeHee Ecco ee E VL CR ss e Fi se p 0 oo dir s SA x ee e o e ee E 0 i A APR eee e 9 0 0 0 00 PR ye c F SSmo ev amp e B 5 7 o lt o 10
78. resulting from this bottom up approach depend on the unit costs used and the validity of the track deterioration models There is no guarantee that the sum of the individual vehicle usage charges would equal the total variable usage cost for the network An alternative is a top down approach In the top down approach formulae are used to calculate the equivalent damage caused by an individual vehicle The equivalent damage allows one vehicle to be compared directly with another A factor to convert damage to cost is derived by calibrating the total equivalent damage from all vehicles operating on the network with the total variable usage cost that is to be recovered Once calibrated the conversion factor is used to convert any individual vehicle s equivalent damage to its cost and hence variable usage charge The advantage of the top down approach is that it guarantees to recover the total variable usage cost It does so by apportioning this cost between the vehicles depending on their propensity to generate damage The formulae used to calculate damage are still based on the Mini MARPAS models Equation 1 shows the formula currently used to calculate the equivalent damage for track related costs Equivalent Track Damage C A Ss USM GTM 1 where C is 0 89 for loco hauled passenger stock and multiple units and 1 for all other vehicles A is the axle load tonnes S is the operating speed miles hour and USM is the unsprung mass k
79. roximately 73M of this total is assumed to be variable with traffic Network Rail s Infrastructure Cost Model ICM calculates the variable costs for the various types of asset Table 12 shows the results from the ICM in March 2008 Thus the variable costs of rail surface damage 73M year are approximately 30 of the total variable track costs Prepared by TTCI UK Ltd 27 March 2008 Page 20 LEER E Pa Methodology to Calculate Variable Usage Charges for Control Period 4 Table 12 Variable Usage Costs from ICM March 2008 Asset M year Track 243 4 Structures 23 0 Signalling 6 8 3 1 Rail Surface Damage Methodology Rail surface damage has been shown to be a function of the tangential forces 7 and creepages y between the wheel and the rail This function depends on the curving performance of a vehicle and cannot be expressed as a simple relationship with vehicle properties Thus the equivalent of Equations 1 and 3 for tangential forces is Equivalent Rail Surface Damage f Ty VM 4 where VM is the miles travelled by a vehicle The function f 7y is derived in Appendix B It only needs to be calculated once for each type of vehicle operating on the network The function combines the surface cracking and wear damage produced by a vehicle Equation 4 is applied to each of several bands of curvature and the results are weighted by the length of track in each band Table 13 shows the curve distributio
80. s Figure 2 shows that increasing the speed exponent in the equivalent track damage formula generally increases the variable usage charge for passenger locomotives and reduces it for other passenger vehicles The opposite happens when the exponent is reduced The variable usage charge tends to increase for high speed vehicles when the exponent is increased because they are being increasingly penalised for their high speeds Variable usage charges are clearly very sensitive to the value of the speed exponent For the highest and lowest speed vehicles the usage charge can change by Prepared by TTCI UK Ltd 27 March 2008 Page 9 ASA ET eee aa Methodology to Calculate Variable Usage Charges for Control Period 4 7H vr B Us m TM Lid almost 100 with alternative and reasonable values of the speed exponent For the majority of passenger vehicles the charges can vary between 30 and 60 Y Y Figures 3 and 4 show the effect on empty and loaded freight vehicles respectively of changing the speed exponent in the equivalent track damage formula Figure 3 shows a data point for each of the 293 different combinations of freight locomotive and commodity and a data point for each of the 1720 combinations of empty freight wagon and commodity Figure 4 shows a data point for each of the 1190 combinations of loaded freight wagon and commodity Note that freight locomotive results are shown only with the empty wagon results Fi
81. s charges increase when the unsprung mass exponent is increased Increasing the unsprung mass exponent to 1 0 giving a linear dependency on unsprung mass can change freight wagon variable usage charges by 20 to 30 Table 8 shows the net effect of changing the unsprung mass exponent on the variable charges to the passenger and freight sectors Clearly the choice of unsprung mass exponent has an effect on charges although it is not as significant as the effects of the speed and axle load exponents Table 8 Net Effect of Unsprung Mass Exponent on Variable Usage Charges k year Unsprung Mass Exponent 0 19 Base Case nd 10 Passenger 178 692 177 438 174 759 Freight 94 507 95 762 98 440 2 4 4 Structures The equivalent damage formula for structures Equation 3 currently raises axle load to the power 3 83 Variable usage charges have been recalculated with this exponent changed to 2 0 and 3 0 The effect on passenger vehicle charges was found to be relatively small between 10 and 10 The effect is less than that for the axle load exponent in the equivalent track damage formula because variable structures costs are only approximately 10 of variable track costs A similar result was found for freight locomotives and empty freight wagons When the axle load exponent in the structures equivalent damage formula was reduced the variable usage charges for empty freight wagons generally increased The increa
82. s used by other railway administrations It should result in a set of charges that are fully auditable Sensitivity studies should be performed to test that assumptions do not have a significant effect on the results The recommended review to the existing charging methodology would require e aliterature survey e analysis of maintenance and renewal costs e track and structures deterioration model development e parametric studies e development of new equivalent damage formulae e sensitivity studies e re calculation of variable usage charges and e documentation It is reasonable to expect this to be completed in time for Control Period 5 Prepared by TTCI UK Ltd 27 March 2008 Page 19 m E UA P Methodology to Calculate Variable Usage Charges for Control Period 4 3 PROPOSED NEW METHODOLOGY In recent years the industry has recognised the importance of rail grinding and ongoing rail renewal to deal with rail surface damage The cost of this work is significant and is currently included in the general track costs recovered through the variable usage charge for track To make usage charges more cost reflective and to incentivise train operators to use vehicles that reduce rail surface damage Network Rail proposes to introduce a term to account for rail surface damage in the methodology for calculating variable usage charges Rail surface damage costs arise due to rail wear rolling contact fatigue and squat defects Thes
83. se was less than 1096 for an exponent of 2 0 The greatest sensitivity to the structures axle load exponent was found to be with the variable usage charges for loaded freight wagons Figure 11 shows the change varied between 10 and 20 Prepared by TTCI UK Ltd 27 March 2008 Page 17 m zu P Methodology to Calculate Variable Usage Charges for Control Period 4 4H e HUS is ai 100 80 a 40 20 0 20 Change in KGTM T PS 60 80 100 Wagons Struct Exp 2 0 Struct Exp 3 0 Figure 11 Effect of Structures Axle load Exponent on Loaded Freight Vehicle Variable Usage Charges Table 9 shows the net effect of changing the structures axle load exponent on the variable charges to the passenger and freight sectors Clearly the choice of structures axle load exponent has an effect on charges although it is not as significant as the effects of the track speed and axle load exponents Table 9 Net Effect of Structures Axle load Exponent on Variable Usage Charges k year Structures Axle load Exponent 3 83 20 an Base Case Passenger 183 690 180 986 178 692 Freight 89 509 92 213 94 507 2 5 Sensitivi
84. speed in curves is generally to reduce variable usage charges The variation for locomotives and empty wagons is between 3 and 6 The variation for loaded wagons is between zero and 9 These results show that variable usage charges are not very sensitive to assumptions about operating speeds The largest absolute variation for an individual vehicle was found to be 8 The net effect of assuming balance speed compared to the base case would be to transfer 1 75M from freight to passenger charges This is relatively insignificant compared to the total variable usage cost of 273M 3 2 3 Lubrication In the base case the coefficient of friction is assumed to be 0 4 for all wheel rail contact points except the gauge face of the high rail where it is assumed to be 0 2 Variable usage charges were recalculated with the coefficient of friction changed to 0 45 at all contact points Figure 19 shows the effect of this change on passenger locomotives coaches and multiple units These results use the curve distribution for the network as do all the other results in Section 3 Prepared by TTCI UK Ltd 27 March 2008 Page 28 Se UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 a Qu ee OOD gt itii HA 20 15 a a a cc cc Dc c d bed ri y L rer ee ee T EL AELE LAE qa ea Ea rata EAC 1095 e v 5 1 a o S 0 ew
85. stablished curve histograms demonstrating that there are appropriate levels of agreement Route based charging is no longer being proposed for CP4 Appendix B The document seems to make the following assumptions in working logic all of which are implicit rather than explicit e Rail replacement is exclusively because of reductions in rail size that are a consequence of grinding and of rail wear That portion of the rail replacement costs due to tangential forces is due to this A portion of rail replacement costs is also handled by the existing variable usage charging methodology e Grinding and the associated cost are applied wherever rail rcf cracks achieve a 2mm length Varying this assumption would shift costs between grinding and renewal A sensitivity study on ratio between these costs is presented in the report The sensitivity to the assumed profiles is considered small in comparison with existing accepted sensitivities e Rail rcf cracks with a 2mm length exist in quantities in one location that it is cost effective to grind It is assumed that once RCF appears in one location on a curve it won t be long before it is sufficiently widespread to justify grinding the whole curve e Grinding delivers no other benefit than the elimination of rail rcf damage 95 of the benefit from grinding is assumed to be the elimination of RCF e The purpose of grinding is to eliminate rail surface material that has an exhausted fatigue l
86. structure Cost Model Route Classes Route Class London amp SE Primary Secondary Rural Freight Non NR Rather than have the complexity of a separate variable usage charge for each of the 300 SRSs three curvature classes have been defined The definitions vary by route class as Table 21 shows Thus there are 15 combinations of route class and curvature class Variable usage charges have been calculated for each of these combinations Table 21 Curvature Class Definitions 9o of route with curve radius lt 1500m Curvature Class Route Class Curvy Mixed Straightish London amp SE More than 6096 26 to 60 Less than 26 Primary More than 58 18 to 58 Less than 18 Secondary More than 60 22 to 60 Less than 23 Rural More than 63 23 to 63 Less than 22 Freight More than 60 29 to 59 Less than 29 Prepared by TTCI UK Ltd 27 March 2008 m Pa Methodology to Calculate Variable Usage Charges for Control Period 4 AA Aur Big neret Calculate Variable Usage Changes for Control ere TM gt Ltd The reason for varying the curvature class definitions by route class is to avoid having SRSs close to the boundaries between classes The boundaries have been chosen to lie in natural gaps in the curvature distributions for each route class see Figures 24 through 28 a bin size of 1 percent has been used Each SRS is considered in turn The total length of
87. sts associated with replacement of rails because of squats should be independent of rail surface damage See comments on squats above Section 6 2 Technical Report Sections 3 and 4 Section 3 identifies a table of data that describes the curving performance of one vehicle type on a range of curve radii and a range of cant deficiency values This demonstrates that with an assumed perfect track geometry curving performance can vary by a factor of 400 between tight curves and gentle curves Further variation by a factor of 6 is given by the extent of cant deficiency on any given curve Clearly the extent of rail surface damage that can be caused by any individual vehicle on a curve is highly dependent on the curve that it operates over the cant of the track in the curve and the operating speed of the vehicle on that curve This table Table 1 forms a persuasive argument that rail surface damage costs cannot be allocated on a network wide basis assuming that all vehicles operate over all curves with assumed levels of cant and of operating speeds The RSD damage term could be included with network average charges if desired by making the same type of assumptions that are currently made in the vertical damage methodology If track access charges are to be linked to the costs of rail surface damage and are to incentivise the use of the most appropriate train configurations the access charges must bear on vehicle types that do the most
88. t for route based charging is noted Prepared by TTCI UK Ltd 27 March 2008 Page 41 gt aAWwALZ Lt Methodology to Calculate Variable Usage Charges for Control Period 4 N amp Comment 1 4 We would therefore suggest that in order for us to adopt such a change that proposals are put forward for the charge be based upon simple route specific criteria this will enable greater alignment with the complex application of such a Rail Surface Damage charge RSD on curved track as well as specific to the speed and type of individual rolling stock Noted Comment 2 1 It is clearly obvious from the report that the complexity of such a charge would lead to issues surrounding its transparency in terms of methodology and cost application The lack of transparency could lead to uncertainty in the application of such a charge particularly regarding its ability to incentivise the Industry to optimise vehicle design and implementation The proposed methodology is a compromise between a detailed analysis and an over simplified approach to the issue of calculating rail surface damage from individual vehicles Transparency is provided by describing the method of calculation together with its inputs Comment 2 2 It must also be highlighted that by moving to a route based regime it will add complexity if utilised against Network Rail s Infrastructure Cost Model which segregates the network into 300 individual route segments We t
89. tforward route based regime rather than a national regime which could not take into consideration the differing vehicle types volumes of traffic and maintenance regimes appropriate to each Train Operator and the route s to over which they operate There are definite advantages in introducing RSD charging even on a national basis There are further advantages in introducing it on a route basis Comment 6 2 Furthermore the document takes no account of those principles as to how and where RSD is to be maintained grinded and or renewed There is an onus on Network Rail to uphold its infrastructure responsibilities the same way VWC is incentivised to minimise track damage through the use of track friendly and tilting rolling stock this is not clearly evident from the report The report states that variable usage charges are based on infrastructure and vehicles maintained to satisfy the industry standards and working instructions Prepared by TTCI UK Ltd 27 March 2008 Page 45 m y UA P Methodology to Calculate Variable Usage Charges for Control Period 4 Comment 6 2 VWC is not therefore in a position to support this proposal without improved transparency and regard to those issues highlighted above Responses to the issues highlighted by VWC are given above Further work is underway to improve transparency ATOC and GNER 29 June 2007 Section 2 Precedent even the relatively simplistic VTA model used today is worthy
90. thWorn B11 Axle Axle Axle where DepthGround _ DepthGround Grinding Cycles Axle GrindingCycle Axle B12 _ DepthGround RCFDamage GrindingCycle Axle For 15 lt Ty lt 175 N DepthWorn _ WearDamage B13 DepthGround RCFDamage Thus DepthWorn WearDamage e DepthGround Axle RCFDamage Axle B14 _ DepthGround WearDamage g GrindingCycle Axle Prepared by TTCI UK Ltd 27 March 2008 Page 69 Sa xy Methodology to Calculate Variable Usage Charges for Control Period 4 Bs m LIA za wr Us Substituting Equations B11 B12 and B14 into Equation B6 gives15 lt Ty lt 175 N RCFD SON x REDIERE x GrindingCost Axle Axle B15 DepthGround feret doc RenewalCost GrindingCycle Axle Axle SideLossLimit and finally for 15 lt Ty 175 N Equation B16 Cost RCFDamage x GrindingCost RenewalCost B16 DepthGround gt RCFDamage WearDamage x SideLossLimit GrindingCycle For Ty gt 175 N DepthWorn _ iu WearDamage B17 Axle Axle At Ty 175 N the depth worn per axle from Equations B14 and B17 should be the same Thus _ PTOA B18 GrindingCycle Substituting Equations B11 B17 and B18 into Equation B6 gives for Ty gt 175 Cost DepthGround WearDamage RenewalCost B19 Axle GrindingCycle Axle SideLossLimit and finally for Ty gt 175 N Equation B20 D R Cost GOA x WearDamage x Acs B20 GrindingCycle SideLossLimit Thus an equa
91. the SRS is added to the bin that corresponds to the percentage of track in that SRS that has a curve radius less than 1500m The y axis is then the total length of SRSs that fall in each percentage bin Length of Track km Straightish 160 140 120 Length of Track km o e of Track lt 1500 m Radius Figure 25 Distribution of Curvature in Primary Classes Prepared by TTCI UK Ltd 27 March 2008 Page 35 Methodology to Calculate Variable Usage Charges for Control Period 4 Straightish wy 32841 Jo uj6ue3 of Track lt 1500 m Radius Figure 26 Distribution of Curvature in Secondary Classes Curvy Mixed Straightish o o o o wo o N wy 32811 jo yyBua7 96001 2086 06 98 08 908 02 99 09 GS 0S SY 07 9oS 260 GC 07 GL 04 S 0 of Track lt 1500 m Radius Figure 27 Distribution of Curvature in Rural Classes Page 36 27 March 2008 Prepared by TTCI UK Ltd m E UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 l 1 I 4 I 1 l 1 l l Length of Track km X 3 5 2 5 D9 DS DS 9 O9 D9 QS O9 O9 Q9 B9 Q9 JS O9 2 9 SE SE O O O LO O LO O O O DOD O LO O LO O DOD O 1 O w A NN cO NM s oOo O O O N N o 0 O 0 of Track lt 1500 m Radius Figure 28 Distribution of Curvature in Freight Classes Figures 24 through 28 shows that considerable judgement is r
92. tions have been derived that gives the cost of rail surface damage in terms of known variables and the wear index Ty generated by a vehicle Table B1 lists assumed values for the variables in Equation B15 Prepared by TTCI UK Ltd 27 March 2008 Page 70 m E UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 Table B1 Constants and Assumed Values a 2 000 km Cost to grind a km of track may involve more than one pass oe 250 000 km Cost to renew the rail for a km of track one rail only Depth 0 5 mm grinding Depth of material removed from the gauge face of the rail per Ground cycle grinding operation may involve more than one pass a Meet 10 mm Maximum allowable gauge face wear at the gauge point of the rail Crack Damage 0 2 x 10 axle Amount of crack damage accumulated per axle pass Rate Wear Damage 0 291 x 10 axle Amount of wear damage accumulated per axle pass Rate Note that 1 The results are sensitive to the ratio of grinding cost to renewal cost not their absolute values 2 Side wear limit is derived from Reference 2 It is based on a nominal head width of 70mm and an average limit on head width of 60mm before corrective action is taken 3 Crack and wear damage rates are taken from Reference 1 Substituting the values from Table B1 into Equation B16 gives15 lt Ty lt 175 Cost 14 500 x RCFDamage 12 500 x WearDamage B21 Using Equation B7
93. tly seeking approval for release of data regarding the TF 25 we use and will share the data as and when we can Additionally the exclusion of LTF13 creates concern however we are unable currently to provide actual data for these A category for TF25 bogies will be added when this information has been received and processed General Completeness We suggest that all bogie types are included as this is an opportunity to get the spread of charges proportionately cost reflective of the pre efficiency target and so encourage positive track friendly investment over the next control period Prepared by TTCI UK Ltd 27 March 2008 Page 61 M m m UA rae Methodology to Calculate Variable Usage Charges for Control Period 4 hy amp This would require detailed knowledge and modelling of all freight bogies which is not considered practicable in time for CP4 Appendix to ATOC letter 30 January 2008 4 3 ATOC would value clarifications in the following areas Rail life is assumed to be a function of sidewear only are there not a number of other factors that control rail life including the effects of vertical track forces The methodology is only used to apportion variable costs due to lateral forces The existing methodology is used to handle costs including reduction in rail life due to vertical forces Rail grinding the cost of rail grinding is assumed to be done exclusively because of rail rcf are there not
94. ty Summary Table 10 summarises the results of the sensitivity study Table 10 Summary of Sensitivity Results Exponent Base Case Range Tested Range of Effect Track speed 0 64 0 28 to 2 0 100 to 90 Track axle load 0 49 0 49 to 2 0 85 to 90 Track unsprung mass 0 19 0 19 to 1 0 50 to 90 Structures axle load 3 83 2 0 to 3 83 15 to 5 Prepared by TTCI UK Ltd 27 March 2008 Page 18 m E UA P Methodology to Calculate Variable Usage Charges for Control Period 4 2 6 Recommendations The sensitivity study has shown how variable usage charges are strongly affected by the values of the exponents in the equivalent damage formula for track The speed exponent in the equivalent track damage formula lies in the middle of a range of reasonable alternative values Although the value should be reviewed in the future there are no strong arguments to change it at present The exponents used for axle load and unsprung mass in the equivalent track damage formula are low in comparison to other published sources TTCI recommends that they be changed after a full review of the charging methodology for vertical forces It has not been possible to trace fully the derivation of the existing equivalent damage formulae for track and structures A thorough review of the methodology for calculating variable usage charges due to damage from vertical forces is recommended The review should consider methodologies and model
95. uld be based on a typical mid life wheel and rail profiles not on the specific conicities on the route in question unless agreed between Network Rail and the train operator Prepared by TTCI UK Ltd 27 March 2008 Page 55 eee zu P Methodology to Calculate Variable Usage Charges for Control Period 4 vr B US 7M Ltd 19 This incentive for reduced yaw stiffness will certainly result in problems for the Train Operator and Supplier if nobody has a Contractual responsibility to control the operating conicity Each Tender Contract will need to define the Contractual track parameters for the actual route then the Supplier can optimise the bogie and the Operator will know who pays if there are problems in service Agreed A new standard on conicity would seem appropriate regardless of any changes to the variable usage charging regime 20 I agree with your comments on tilting trains and the benefits of operating at higher cant deficiency However the Train Operator has little control over the actual allowable speed on the curves since this depends on the track quality The Class 221 spends very little time operating at enhanced permissible speed so Virgin would be unable to realise the full reduction in the TAC that the train can provide The rail surface damage component of the variable access usage charge would be based on typical actual operating speeds and cant deficiencies 21 Some suspension types have a high primary yaw stiffn
96. ume cost and curvature data from Network Rail s ICM The 300 SRSs in the ICM have been grouped into 15 combinations of route class and curvature class Although variations in traffic volume have been considered the mix of passenger and freight vehicle types in each group is not currently known It has been assumed to be the same as the network average Prepared by TTCI UK Ltd 27 March 2008 Page 38 N 10 11 m E UA P Methodology to Calculate Variable Usage Charges for Control Period 4 REFERENCES Tunna J April 2007 A Revised Methodology for Variable Usage Charging UK NR Report No 07 003 TTCI UK Ltd Colorado USA November 1999 Railway Infrastructure Cost Causation BAH review for ORR http www rail reg gov uk server show ConWebDoc 5255 Broomhead S F G A Hunt and D R Thompson September 1999 Usage Cost Model v2 AEA Technology Rail report S AM PILO RP 5 Derby United Kingdom Railtrack October 1999 Categorisation of Track Railway Group Standard GC RT5023 Safety amp Standards Directorate Railtrack PLC London International Union of Railways January 1989 Classification of line for the purpose of track maintenance UIC Code 714 an Edition UIC Paris Oberg J and E Andersson June 2007 Determining Deterioration Cost for Railway Tracks Proceedings of the International Heavy Haul Conference IHHA Inc Virginia Beach USA BS 5400
97. value when alternative and reasonable exponents are used Table 6 shows the net effect of changing the speed exponent on the variable charges to the passenger and freight sectors The base case usage charges shown in Table 6 are the totals for the passenger and freight sectors over the whole network Table 6 Net Effect of Track Speed Exponent on Variable Usage Charges k year Track Speed Exponent 0 64 0 28 Base Case 20 Passenger 168 442 178 692 214 040 Freight 104 757 94 507 59 159 Table 6 shows that the track speed exponent has a significant impact on charges Choosing a value at the top or bottom of the range changes the balance of charges made by each traffic sector Prepared by TTCI UK Ltd 27 March 2008 Page 11 ee zu P Methodology to Calculate Variable Usage Charges for Control Period 4 M vr B US M IM Lti a 2 4 2 Axle Load Exponent The value of the axle load exponent in the existing equivalent track damage formula is 0 49 Figures 5 6 and 7 respectively show the effects of changing this value to 1 0 and 2 0 for passenger empty freight and loaded freight vehicles 100 80 60 40 a 20 pr 0 20 40 Change in Pence per Mile 60 80 100 Locomotives Wagons Axle Load Exp 1 0 m Axle Load Exp 2 0 Figure 5 Effect of Track Axle Load Exponent on Passenger Vehicle Variable Usage Charges Prepared
98. values of this constant are being calculated asked why category one freight locomotives had been split into those with and without steering linkages as they would differentiated by their Ty value Specific models of the Class 60 and 66 locomotives have been used because of their different curving performance Prepared by TTCI UK Ltd 27 March 2008 Page 49 m E UA Pa Methodology to Calculate Variable Usage Charges for Control Period 4 said he saw the advantage of the categories and the Ty tables could be used as a default Noted questioned why the Ty tables for new trains had to be published expressed concern that if a new train was built then a competitor could reverse engineer and so establish the train configured Ty tables are used to avoid publishing details of vehicles suspension designs Since there is more than one way of producing the same Ty it is not true that the design of a vehicle s suspension can be deduced from its Ty values asked whether Ty was evaluated with traction forces Traction and braking forces are not included To do so would require assumptions about occurrences of their application during curving asked why the zero cant deficiency column had been used for passenger trains which tend to operate at higher cant deficiency Suggested that using column zero was doing a disservice to passenger vehicles and perhaps a Ty column for cant deficiency should be used for passeng
99. viously accepted by Network Rail during various WCML technical forums A default T gamma table will be used for these vehicles based on their assumed primary yaw stiffness VWC are invited to submit an alternative table and provide details of the modelling and testing used to derive it It should be noted that track force testing does not measure directly the forces that cause RSD Prepared by TTCI UK Ltd 27 March 2008 Page 43 gt a 8 Sr B Us Lt Methodology to Calculate Variable Usage Charges for Control Period 4 i amp Comment 5 1 5 Grinding Effort It is also apparent that the Grinding programme with which Network Rail performs its function in managing damage to track could not consistently be applied to the entire network as there is insufficient yellow plant to affect this maintenance nationally However this could be managed more effectively by only grinding track which is straight and has a radii of less than 2500mrn in turn increasing the ability to cover more of the applicable network Network Rail s grinding program is focused on sections of track that require grinding Likewise the RSD usage charging methodology recovers cost from required levels of grinding Comment 5 1 6 Squats Furthermore it has become evident that RSD is not necessarily primarily caused by tangential forces Up to 40 of all track damage is caused through defects known as Squats which are not linked to rail surface dama
100. y that even in the loaded condition have low axle loads e g Domestic Automotive Table 7 shows the net effect of changing the axle load exponent on the variable charges to the passenger and freight sectors Clearly the choice of track axle load exponent has a significant impact on the balance of charges Table 7 Net Effect of Track Axle load Exponent on Variable Usage Charges k year Track Axle load Exponent 0 49 Base Case 1 0 2 0 Passenger 178 692 166 010 135 155 Freight 94 507 107 189 138 044 2 4 3 Unsprung Mass Exponent The unsprung mass exponent in the current equivalent track damage formula is 0 19 Variable usage charges have been recalculated with the exponent changed to 0 5 and 1 0 Figure 8 shows the effect on passenger vehicles Prepared by TTCI UK Ltd 27 March 2008 Page 14 _u _ Pa Methodology to Calculate Variable Usage Charges for Control Period 4 Ar AP cam umo urge Cargos for Contro fon d IM Ltd 100 80 7 ih A 5 E anun u ha 3 A E M Ge 5555555555455955555555554555 54 5 T e a 20 1e m Be ae es a spesse 99 Meme E e 1 i Ws E a 0 4 Lho aj E o ro TA Se eevee a oo th en ual c ve w yy E ID Mh hd e e ow u Wilh Lanocdenouis cse pee ou so a WEIL AN S ooo gt m aE LL K Em soon m 40 4 2 22 22 2 rA E atid 80 4 aE
101. y Assuming 60mph Maximum Speed Prepared by TTCI UK Ltd 27 March 2008 Page 23 _u _ Pa Methodology to Calculate Variable Usage Charges for Control Period 4 mS D gt Neon to Calcuta Verso Usage Chores or Control merda _ Tw Ltd Figure 14 shows the distribution of cant deficiency on the network assuming the maximum speed is limited to 45mph The average of this distribution is approximately 20mm 1 600 000 1 400 000 1 200 000 1 000 000 800 000 600 000 Length of Track yards 400 000 200 000 Cant Deficiency mm Figure 14 Distribution of Cant Deficiency Assuming 45mph Maximum Speed The results of a sensitivity test on the assumed value of cant deficiency are presented in subsection 3 2 2 Network Rail standards require all curves with radius less than or equal to 1500m to have the gauge face of the high rail lubricated Gauge face contact does not usually occur for curves with radius greater than 1500m at the cant deficiencies used in these simulations Thus in the calculation of 7y the coefficient of friction between the wheel and the rail is assumed to be 0 4 at all contact points except the gauge face of the high rail where it is assumed 0 2 The results of a sensitivity study into the effect of lubrication are presented in subsection 3 2 3 Measured semi worn rail profiles are used in the calculation of Ty Table 15 shows the profile names for each of the four different ranges of curv

TTCI report: user guide for variable charges model

Contents

Download Pdf Manuals

Related Search

Related Contents