Вы находитесь на странице: 1из 65

THROUGHPUT ENHANCEMENT

Report of the Multidisciplinary Team 1

Ministry of Railways
NEW DELHI

MAY 2005
((FFoorr iinntteerrnnaall cciirrccuullaattiioonn oonnllyy))
Throughput Enhancement Team 1 May 2005

THROUGHPUT ENHANCEMENT

Report Of The Multidisciplinary Team


Set Up Vide Railway Board Order
ERB-I/2005/23/12 Dated 11/4/05

TEAM MEMBERS

ED/LRDSS (Convener)
ED/CE((P)
ED/SIGNAL
DIR/CS-1/LRDSS
DIR/PLG/LRDSS
DIR/FIN /LRDSS
DIR/EE(RS)

MAY 2005
MINISTRY OF RAILWAYS
(RAILWAY BOARD)
RAIL BHAWAN
(For internal circulation only)

1
Throughput Enhancement Team 1 May 2005

TABLE OF CONTENTS

CHAPTER PAGE

CHAPTER 1 3
Outlines constructs for conceptualizing
throughput and line capacity, freight-
passenger interactions, speed differentials,
etc and inter-relationships therein.
Annexure 1.1 17
CHAPTER II 19
Reviews results from earlier work
undertaken in LRDSS using simulation
modeling.
Annexure 2.1 & 2.2 29
CHAPTER III 31
In the context of ‘global’ & ‘local’ factors,
‘hard’ & ‘soft’ options, presents a system -
dynamics model of interactions; analyses
speed differentials in case of passenger &
freight train operation based on actual data
for select routes; reviews HP/TL ratios for
freight, slack in passenger train scheduling,
etc. Suggests methodology for systemic
review, benchmarking and planning for
throughput enhancement.
Annexure 3.1/1 to 3.1/3 57
CHAPTER IV 60
Looks at some ‘software’ strategies for
improving throughput through reduced
speed differentials and other supporting
strategies outlined in the Terms of
Reference.

2
Throughput Enhancement Team 1 May 2005

CHAPTER I

BACKGROUND

This multi disciplinary team (no. 1) was set up by the Ministry of Railways vide order no
ERB-I/2005/23/12 dated 11/4/05 to analyze impact on throughput of various operational
initiatives (Annexure I.1).

The terms of reference assigned to the team are as follows:


To analyze the impact on throughput due to each of the following:-
1) Reduction in speed differential between passenger and good trains
2) Augmenting horse power to trailing load ratios
3) Tightening scheduling of passenger trains including stoppages at stations
4) Introduction of high speed turnouts, especially on high density corridors

INTRODUCTION
As per the Oxford dictionary throughput is the amount of material or items passing
through a system or process. In the context of rail transport, this refers to the passengers
and freight transported across a section in a period of time. In effect this has two
components, viz. the number of passengers and/or tons per train and the number of trains
per day passing over a section. Logically, therefore, there are three ways of increasing
throughput:

a. Increase the number of trains,


b. Increase traffic carried (i.e. tons and/or no. of passengers) per train so that each train
carries more,
c. A combination of a. & b. above.

3
Throughput Enhancement Team 1 May 2005

Given the terms of reference, the current exercise confines itself to ‘a.’ above, i.e.
increasing the number of trains per day on a section. and not on increasing tons /
passengers per train. In other words, throughput is synonymous with line capacity and
these terms can, therefore, be used interchangeably.

OPERATIONAL INITIATIVES

The four initiatives mentioned in the terms of reference are:


a. Reduction of speed differentials
b. Higher hauling power to trailing load ratios
c. High speed turnouts
d. Reducing slack on passenger trains

These alternatives seek to increase capacity through improved sectional fluidity rather
than through additional running lines, shorter block sections or improved signaling. For
analytical purposes, it is important to make this distinction since the choice variables
change with the strategy adopted. When we pursue a physical capacity enhancing
strategy, the choice variables relate to physical levels of infrastructure / assets and the
outcome could be higher speeds. On the other hand, if the objective function seeks to
improve fluidity through speed enhancement, the choice variables are speed parameters
though infrastructure improvements could at times be a precondition for achieving this.

In this report, it would be appropriate to first examine whether the set of strategies being
considered can be taken together as a consistent set. In other words, is there any conflict
between the four strategies that would entail forsaking one for the other or are they
mutually compatible? Logically, the next question is whether they are all complementary,
if they are compatible. Further, do we need all the strategies to achieve the desired goal or
would a subset suffice? A related question would be, are there other complementary
elements that are possibly not in this set but are the focus of another study by a different
committee. It is felt that it would be best if an attempt were made to address these

4
Throughput Enhancement Team 1 May 2005

questions at the outset and thereby identify a consistent of set of strategies that could
combine to form a holistic policy option for Indian Railways.

SPEED DIFFERENTIALS

Speed differential affects mobility as it causes trains have to queue up behind each other
in their pursuit to utilize the same resource1. A slower train ahead implies either slowing
down of other trains or a precedence event to allow the faster train to go ahead. In case of
the former, the effect is to reduce the average speed of all following trains and in the
latter, mobility is hampered on account of stoppages for the slower train to give
precedence. Clearly there are two ways of reducing speed differentials; one would be to
generally down scale the speeds of the faster trains and the other would be enhance the
speeds of slower trains. The outcome would be reduced speed differentials and generally
increased throughput, and the exact magnitude of this would depend on a number of
factors. In the present analysis, the focus is only on the latter method of reducing speed
differentials, i.e. enhancing the speed of slower trains, for the simple reason that any
strategy of reducing speeds to reduce differentials would be a retrograde step2.

The first relationship postulated is that throughput (T) is a function of speed differential
(Sd) and reduction in speed differential leads to an increase in throughput.

T = f (Sd )

In order to understand this relationship better, a definition of speed differential is


required. Typically, train-mix on a section or network is such that speeds can range

1
The problem becomes more complex when we consider crossings as well. To keep the analysis simple
we consider only unidirectional tracks (typically one way traffic on a double line section). This could be
justified since most of the sections that are critical are also likely to be on trunk routes with double line
facilities.
2
It may be possible that in some cases reduction of speeds of the faster trains actually increases
throughput more than increasing the speeds of slower trains. However, the committee feels that in such
cases those specific factors that lead to this result should be identified and addressed rather than
adopting a strategy of reducing speeds. Increasing speeds to reduce speed differentials would and ought
to be a dominant strategy compared to decreasing speeds.

5
Throughput Enhancement Team 1 May 2005

between smin and smax where s denotes speed. A simple measure could be to use the range,
i.e. the difference between smin and smax. However the relationship between the range and
throughput may not necessarily be monotonic, i.e. reduced range may not always lead to
increased throughput. It is possible that a smaller range combined with an unfavorable
train mix leads to a lower throughput. The train mix is important and therefore the entire
distribution of speeds affects throughput. What measure of differential would capture the
distribution of speeds? Can we use a single parameter that would have a monotonic
relationship with throughput? One measure could be to take the variance of speeds rather
than the range and assume that the relationship would be inversely monotonic3, i.e. as
variance decreases, throughput increases.

Sd ≡ var(s )
We can combine the first two relationships to a single relationship as

T = f (var(s ))
T = f (σ s2 )
where σ s2 denotes the variance.

One of the questions before this Team was how to estimate this function – the impact of
reduction in speed differentials on throughput. What will be the impact of reducing the
variance of speeds on the throughput of the network? There could be two ways of
estimating this relationship – one by observing reduced speed differentials over time and
using regression methods to look at the relationship between throughput and variance in
speeds, while controlling for other factors. The other method is to simulate running of
trains at different speeds over the network and measure the impact of progressively
reducing the speed differential. Unfortunately, the first method requires data that is not
readily available and the second requires software that is also not currently available. In
another section, this Team has leaned on previous analysis in LRDSS using the second
method to present some estimates of this relationship.

3
The assumption of monotonicity may not be too restrictive for the ranges of variance values that are
possibly being considered.

6
Throughput Enhancement Team 1 May 2005

The Team also deliberated on the issue of what speed parameters should be used for
doing the analysis. Passenger carrying trains have three kinds of speeds, a maximum
speed, a booked speed and a commercial speed. The question was which one of these
speeds should be taken to work out the variance in speeds within passenger trains and
across passenger and freight trains. Since freight trains do not run to schedules, actual
speeds of freight trains are taken to observe variance in speeds. For passenger trains, it
was decided to take commercial speed rather than booked speed while defining speed
differential. The rationale for this was that the comparison should be across actual speeds
and commercial speed of passenger carrying trains captures the train’s occupation of the
section and therefore capacity. Commercial speed is the outcome of booked / actual
speeds and stoppages and this is what affects the capacity of the section.

HIGH SPEED TURNOUTS

High-speed turnouts are a technological input at stations that allows trains to move
through stations (loop lines or lines other than main line) at higher speeds. The main
benefit of this input arises from reduced time losses as trains decelerate / accelerate to
traverse loop lines at stations and faster clearance of the rear block section. The severity
of the problem thus depends on the frequency of such events occurring where trains have
to stop and / or precede another by taking the loop line. The greater the frequency, more
would be reduction in losses due to deceleration and acceleration. In other words, if
reducing speed differentials can reduce the number of such precedence events, then high-
speed turnouts may not be required additionally. Besides the speed over turnouts, the
volume of traffic on turnouts (due to stoppages on other than mail lines or precedences or
movements through large passenger / freight terminals) is a critical factor affecting the
throughput. If the high-speed turnouts are set for maximum speed of st, then the
relationship can be shown as:

T = g ( s t ; nt )

7
Throughput Enhancement Team 1 May 2005

Throughput depends on the maximum speed on the turnouts (st) given the number of
turnout events (nt). For a given level of turnout events, throughput will increase as the
speed on the turnout increases, a direct positive relationship. As a corollary, reduced
speed differentials and provision of high-speed turnouts may not necessarily be part of a
single policy option; more of one could reduce the need for the other, since reduced speed
differentials potentially implies lower level of turnout events, nt.

HORSE POWER TO TRAILING LOAD RATIOS

How much power is right for pulling a load? Benchmarks could be obtained by posing
the question as to what is the horsepower required to enable the train to move at a certain
specified speed through the network. From the viewpoint of speed enhancing strategy the
issue is what is the horsepower to trailing load ratio required to maintain a certain pre-
specified level of speed, say s, across the network? This would have to be specified for
different types of track geographies. The issue is linked to the strategy of reduced speed
differences since s would depend on that. Thus these two strategies are complementary
and it would only be appropriate that these form part of a single policy option. Since this
is a ratio between hauling power and trailing load, differences in trailing loads can be
compensated by differences in horsepower to keep the ratio constant. However, track-
geography differences in gradient and curvature imply that there cannot be a standard
ratio for all trains and all segments of the IR network.

TIGHTENING SCHEDULES OF PASSENGER TRAINS

The strategy of reducing slack in the timetable is somewhat related to the issue of
reduced speed differentials. It could actually be a component, a subset of a speed
differential reducing strategy. The effect of reducing slack would be to lower the variance
in speeds since we consider only increasing speeds to reduce differentials. Given the fact
that we take the commercial speed of passenger carrying trains, the reduction in slack or
stoppages can be subsumed within the reduction in speed differential strategy.

8
Throughput Enhancement Team 1 May 2005

In conclusion, it is seen that three of the four strategies mentioned in the terms of
reference are complementary to each other; the fourth - high-speed turnouts - is
complementary under circumstances such as where train stoppages on other than main
lines and / or overtakes and / or through movements on other than main lines across
passenger / freight terminals occur frequently.

LINE CAPACITY AND THROUGHPUT

In the backdrop of the foregoing discussion, the four alternatives and the
interrelationships between them, it would be relevant to look at the concept of line
capacity in greater detail. Throughput and line capacity are interchangeable terminology
in the context of the present terms of reference. Thus, T denoting throughput can be
substituted with C denoting capacity and the relationships continue to hold true since for
a given tonnage level of trains, throughput and capacity are collinear variables.

MOTIVATION

In order to motivate discussion on the stated terms of reference, a simple example has
been constructed. The example seeks to highlight the inter relationship between speed
differential and line capacity and show that if trains with different speeds are run in a
mixed manner it would result in lower line capacity. And, further, as to how stoppages –
whether scheduled as in case of passenger trains or unscheduled as in case of a slower
train for giving precedence to a faster train - leads to further erosion of capacity.

Consider a section where all trains have the same speeds (i.e. variance of speeds is equal
to zero) and there are no stoppages. To simplify the model further, it is also assumed that
all the block sections of this section are equally spaced and similar in terms of terrain. In
such a scenario, the capacity of the section would depend only on the headway. In the
two-dimensional space of time and distance, sectional capacity is the number of parallel

9
Throughput Enhancement Team 1 May 2005

lines that start at the point of origin of the section and reach section finish point in the
span of 24 hours.

Suppose we assume that this headway is h, and then the number of train starts in a day
would be 1440/h. The capacity of the section is therefore simply the total time divided by
the headway time. Headway would depend on the speed of the trains, the length of
block sections and the time required for train operations (time taken to close the block
plus train start time).

1440
C =
h
The figures below are used to explain this in simple terms. Fig. 1.1 shows a case where
all trains run at the same speed of 120 km/hr. and the block sections are 10 kms length.
The parallel lines running across the box are the trains and clearly the smaller the
headway the greater the number of trains that can be accommodated in a specified time
period.

20 km

DISTANCE
h
kmp
120

Y-Axis
AT

10 km
Y-Axis

N
RAI
TT
FA S

0
10 20 TIME 40 50 60
30

Fig. 1.1: No Speed Differential (all trains running at 120 km/hr.)

10
Throughput Enhancement Team 1 May 2005

The assumed speed of 120 km/hr. leads to headway of 5 minutes given the length of the
block section. The capacity in this extreme case would be 288 trains in one direction for a
double line section.

The second figure (Fig. 1.2) shows another case with no speed differentials but slower
trains running at 60 km/hr. In this case headway increased to 10 minutes and capacity is
halved to 144.

20 km

DISTANCE
h
k mp
60

Y-Axis
AT

10 km
AIN
Y-Axis

TR
WO
SL

0
10 TIME 30
60

Fig. 1.2: No Speed Differential (all trains running at 60 km/hr.)

What if both these trains are run on the same section? Suppose, to begin with, these trains
are run in separate time corridors, with one half of the day being utilized for high-speed
trains and the other half for slower trains. In such a case the capacity would be around
216, the halfway point between 144 and 288. This is a case of speed differentials but no
interactions between the faster and the slower trains. The time corridors do not have to be

11
Throughput Enhancement Team 1 May 2005

equal and for any other combination the capacity would lie between 144 and 288
depending on the ratio of slow to fast trains. The reduction in capacity compared to the
first case is due to the fact that a third of the trains run at a lower speed. But the speed
differentials per se do not cause the reduction in capacity since the two trains do not
interact in their running.

Next, consider the case where the slower train and the faster train interact. In this
example, it is assumed that a slow train is invariably followed by a fast train (or vice
versa). In other words, the slow and fast alternate each other, thus maximizing the
number of interactions between the two speed types. We also assume that the faster train
is given precedence over the slower train. In this case, capacity would fall below 216 to
192 even though the ratio of fast and slow trains (the mix) is about the same; the
precedence events lead to loss of capacity (see Fig. 1.3).

20 km
ING
CED

DISTANCE
PRE

Y-Axis
kmph

10 km
20
ph

AT 1
km
60

N
TRAI
AT
IN

FAST
A
TR
OW
SL

0
40
10 20 50 60
TIME 30

Fig. 1.3: Speed Differential with two train types alternating and precedence for faster train

12
Throughput Enhancement Team 1 May 2005

Finally, consider a case where along with precedence to the faster train we also have a
stoppage for the faster train. The fast and slow train continues to alternate as in the earlier
case and the resultant pattern is shown in Fig. 1.4. Notice how stoppages lead to further
curtailment of capacity.

20 km

DISTANCE

Y-Axis
10 km
IN AT ph

D
120
km

PREC PING AN
60

G
EDIN
AT

A
STOP
AIN

T R
TR

FAST
kmph
OW
SL

0
40
10 20 50 60
30
TIME

Fig 1.4: Speed Differential with two train types, stoppage and precedence for faster train

These examples provide a simple way of understanding the nature of the problem as we
move away from a world with no differentials and no stoppages. When the objective
function is maximization of the number of lines running across the distance – time box,
events like precedence and stoppages act as constraints. The simple example provided
here also underlines the importance of simulation modeling for its ability to replicate real
world flows and study the effect of proposed changes by building different scenarios.

13
Throughput Enhancement Team 1 May 2005

Needless to say, real life situations are far more complex and dynamic. The following
illustrations provide a sample of control charts.

Fig 1.5: BARODA SURAT - Day 3 (0000 to 0800) Base Case Scenario

The above illustration is a simulated output of trains running on Baroda - Surat section.
Note the steady, streamlined flow of trains in the Surat - Baroda direction (top of
illustration to bottom of illustration). These are Mail / Express trains running with
uniform speeds. On the other hand, note that the all - stopping passenger train departing
from Vadodara at about 0315. This train, with a relatively low commercial speed, eats
into the sectional capacity so that “window” till the following train (shown in blue,
departure from BRC at about 0500) is lost.

14
Throughput Enhancement Team 1 May 2005

The following illustration presents a far more complicated section than the one presented
earlier.

Fig 1.6: GHAZIABAD- RURA (Day 2: 0000hrs to 0900 hrs): Base Case Simulation

Note:
a) The larger number of trains running on the section
b) Speed differentials: differences in the slope of lines as they cross each other
c) Red lines: depicting the all stopping slow passenger trains
d) Large number of overtakes to allow higher priority trains to take precedence

These two illustrations are used to depict the immense degree of complexity associated
with optimizing throughput on a real time basis and the need to control factors
responsible for speed differentials as well as lower average speeds.

15
Throughput Enhancement Team 1 May 2005

REPORT FORMAT

Available results of simulations relevant to the terms of the reference are recapitulated in
Chapter II. An in-depth analysis of average speeds and speed differentials based on recent
observed data pertaining to a cross section of rail links is presented in Chapter III.
Finally, some policy implications of the findings and proposals are highlighted in Chapter
IV.

16
Throughput Enhancement Team 1 May 2005

!" # datedJ $

#% #& '( ) * #! * ( & * #! * ( * '+ " * , + - #+ + - % & # . "


) !! #% / . ! # #& - #0 !#% * '( *1 & * % * !( 2 #1 0 * - % " ' . / " 0 . ) "
, * '# . &
0 '* # % * ! #% # #* #, & 3
. ! # #& - #0 !#% * ', *1
" *1 #!! - % & #& ) " ) !! #% / 3

44 5 - . #, #' - ' 6 * #! * ( * '+ 7 % , '# '


5 - . #, #' - ' 7 #, #! % / / 8 9 * #! * ( * '+ 1 : '
5 - . #, #' - ' #/ % * ! 9 * #! * ( * '+ 1 : '
$ #' - ' 7 9 * #! * ( * '+ 1 : ';
#' - ' 8 !/ 9 * #! * ( * '+ 1 : '
< #' - ' #% - 6!% - 9 * #! * ( * '+ 1 : '
= #' - ' ! - '#- * ! % / / 9 * #! * * '+ 3 4

" *1 #!! * % * !( 2 #1 0 * - % "' ./"0. +. * - " ) " ) !! #% / 3

# + . - # % #% & 0 + + #)) ' % #* !& > : % 0*&& %/ ' *%+ / +&


'* #% & 6
## . / 1 % #% / " '& 0 ' '* #!#% / ! * + '* # 6
### #/ " ' & - " + . !#% / ) 0 * & & % / ' '* #% & 9 #% - !. + #% / & 0 0 * / & *
9> & * # %&6*%+
#, % ' + . - # % ) " #/ " & 0 + . '% . & 9 & 0 - #* !!( % " #/ " + % & # (
' . &

Multi Disciplinarv Team

" *1 #!! - % & #& ) " ) !! #% / 3


5 - . #, #' - ' 7 * - " #% / 9 * #! * ( * '+ 5 - . #, 7 %, % '
#' - ' '* ))#- '* % & 0 ' * # % 9 * #! * (
* '+ 1 : '
5 - . #, #' - ' - " * % #- * ! % / / '* - # '!9
* #! * ( * '+ > 9 1 : '
$ 5 - . #, #' - ' ! - '#- * ! % / / 9
* #! * ( * '+ 9 1 : '

" *1 #!! * % * !( 2 #1 0 * - % "' ./"0. +. * - " ) " ) !! #% / 3


17
# + . - # % #% % . 1 : ' ) '* - # % - " * % / 0 #% & #% - !. + #% / '* - # % - " * % / 0 #% & ) ' )' #/ "
'* ))#- ? 0 #% / #% , # " ). ! - & 9 - & ) 0 '* # % * % + 1 * #% % * % - ) ! - & 6
+ . - # % #% % . 1 : ' ) - 1 #, & 6
##
Throughput Enhancement Team 1 May 2005

t'
-2
(iii) . Reduction in number of crews; and
(iv) Reduced running time of trains.

Multi Disciplinarv Team No.3.

Traffic R:esearch Directorate ofRDSO, Lucknow will analyze impact on throughput by


"Running Freight Trains to Schedule!Time Table. The Traffic Research Directorate of
RDSO will take assistance of other Technical Directorates including Loco Directorates
(both Electrical and Mechanical of ROSO).

2. The reports of all three Multi Disciplinary Teams must be submitted to the Board
by 15-05-2005.

3. The Headquarters of the Team Nos. I & 2 wilt be at New Delhi and Headquarter of
theTeam No.3 will be at Lucknow.

4. The Convenor(s) and Members of the Teams wilt be eligible to draw T AIDA as per the

existillg rules. ~
~. M
Under Secretary(E
Copy to: Railway B

I. The General Managers: All Indian Railways and Production Units.


2. TIle General Manager(Construction)fNortheast Frontier Railway, Guwahati. .
3. TIle Director General/RDSO, Lucknow. He may kindly ensure constitution of Multi Disciplinary
Team No.3 and submission of report by 15-05-2005 to Board.
4. TIle Director General, Railway Staff College Vadodara.
5. OSDfMR, PSfMR, PSfMSR(N), PSfMSR(V), EDPGfMSR(N), JDPGfMR.
6. Sr. PPSsfPSSsfPSs to CRB, FC, MS, ME, MM, MT, ML, Secy. DGfRPF, DGfRHS, All
AMsf Advisers, ED(Chg), EDTIM, EDME(fr), ED(LRDSS), EDCE(P), ED(Sig.), ' JS, JS(G), JS(E&P),
Dir(LRDSS)fCS-I, Dir(PIg.)/LRDSS, Dir(Finance)/LRDSS, DEE(RS), Dir.(Estt.),
JDIP, DS(P), DS(G), DPIO,US(A)! US(Protocol), US(Parl.), Railway Board.
7. The Chief Commissioner of Railway Safety, 16 Ashok Marg, Lucknow.
8. The F A&CAO, Northern Railway.
9. The Pay & Accounts Officer, Railway Board.
10. The General Secretary, AIRF & NFIR.
II. The General Secretary, IRCA, New Delhi.
12. TIle Secretary General, FROA & IRPOF, All India RPF Association. ,
13. Railway Board Secretariat Services Group'A'Officers Association, R.No. 324, Rail Bhawan,
New Delhi. ...
14. Cash-I, G, G(Pass), G(Acc.), SEE(Power), Chg., ERBdI, III, AC-II,..£;:a.rJ.., (),\'
r1\,{J ~~ion,
Library, Telecom and Stationery Branches, Railway Board.
15. The Convenor(s) and Members of the Committee. TIleConvenor(s) of the Committee is
18
requested to kindly make available a copy of Report of the Committee, when finalized, to ERB- I
for record.

SinhaflU)/ORDER/27-29 t'
Throughput Enhancement Team 1 May 2005

CHAPTER II

AVAILABLE RESULTS, DISCUSSIONS

Dynamic simulation exercises undertaken since the inception of LRDSS provide a wealth
of insight into issues listed in the present terms of reference. The results of these
analyses and their relevance to the current context are discussed in this section. It is
pertinent to note at the outset that the result of a combination of operational
initiatives highlighted in the terms of reference can and is different from the sum of
individual effects in many cases.

Elaborate and comprehensive simulation analysis was carried out for Sonnagar (SEB) –
Mughal Sarai (MGS) section (a freight dominant section at that time) using RAILS
(Railway Analysis & Interactive Line Simulator) software. It needs to be borne in mind
that the results obtained are generally specific to the nature of the section, mix of trains
and operational practices then prevailing on this section. On a different section with
different traffic mix and other operational parameters, the results – especially their
magnitude - are likely to be different.

IMPACTS ANALYSIS – A REVIEW OF SIMULATION RESULTS

From amongst a plethora of Scenarios simulated during Phase II of the LRDSS project,
this Team has culled out those that are relevant to the present terms of reference. These
scenarios aim at isolating the impact of various improvements, taken singly or in
conjunction with others, and are useful for determining the impact of various
improvements on throughput. A summary of the findings is attached as Annexure II.1.
These findings form the basis for the following discussion.

19
Throughput Enhancement Team 1 May 2005

MAXIMUM SPEEDS OF FREIGHT TRAINS

A number of Scenarios were simulated to analyze the effect of speed on transit times /
capacity in case of SEB-MGS section.

Pass Halts = 1 min


BOXN Speeds

BTPN Speeds

All Frt Speed

Pass Speed
Scenario

HP/TL

Gain
Scenario Description

1a All freight 80, loaded freight double headed 80 80 80 Y Y 19%


1c All loaded freight 2WAG7 BOXN/BTPN=90 90 90 Y Y 21%
1d All loaded freight 2WAG7 BOXN/BTPN=100 100 100 Y Y 33%
1e All trains 110 110 110 110 110 8%

Table 2.1: Impact of Change in Speeds – Freight & Passenger (commercial)

In the first set of Scenarios (Sl nos 1a to 1e of Annexure II.1, extracted in Table 2.1
above), the maximum speed of freight trains was initially kept at 80 km. /hr. with
passenger train halts reduced to 1 minute. This was found to give a gain of 19% in
capacity with double-headed WAG5 as well as with double-headed WAG7 locos. Then,
freight train speed was further raised to 90 km. /hr. and the gain over base case was found
to be 21%. Finally, speed was raised to 100 km/hr and gain in this case was found to be
33%. As such, from this set of simulations, it was found that 100 km. /hr was the point
where gains in terms of transit times and line capacity were optimized. Improving HP/TL
Load ratios coupled with higher maximum speeds for freight trains results in improving
the slope of the freight train lines and bringing these slopes closer to the commercial
speeds of passenger trains. As such, freight train speeds closer to the average commercial
speed of passenger trains would minimize the instances of freight trains stopping to give
precedence to Mail/Exp trains. This issue is dealt in greater detail in Chapter III of this
report.

20
Throughput Enhancement Team 1 May 2005

Scenario 1e brings out an interesting result: raising the speed of all freight and passenger
trains to 110 Km. /hr. without increasing HP/TL ratio commensurately or modifying the
passenger stoppages / schedules gives a gain of only 8%. This clearly indicates that
removal of speed differentials in terms of maximum running speeds alone without
implementing other steps like tightening up of passenger schedules (thereby improving
their commercial speed) and increasing HP/TL ratios is not likely to yield a substantial
gain.

Pass Schedule Tightened


Pass Halts = 1 min
BOXN Speeds

BTPN Speeds
Scenario

Gain
Scenario Description

5 Hi Sp freight double headed 100 100 13%


6 Hi Sp freight double headed pass schedule tightened 100 100 Y 15%

7 Hi Sp freight double head, pass halts reduced 100 100 Y 18%


8 Hi Sp freight double head, halts + schedule 100 100 Y Y 21%

Table 2.2 Impact of Tightening of Passenger Schedules

In another set of simulations (Scenarios 5 to 8 of Annexure II.1, extracted in Table 2.2


above), where BOXN/BTPN trains were double headed with WAG5/WAG7 locos, the
capacity gain with a speed of 100 km. /hr. was 13% with no change in passenger halts /
schedules and 21% with 1 minute passenger halts and tightened passenger schedules. As
such, it can be seen that the gain of 33% in the case quoted in the previous para (Scenario
1d) is the combined effect of the speed as well as the higher HP/TL ratio (all and not just
BOXN/BTPN freight trains were double headed in Scenario 1d).

The following table (Table 2.3) illustrates the impact of raising the maximum speeds of
freight trains to 100 km. /hr and its impact on line capacity.

21
Throughput Enhancement Team 1 May 2005

Hi-Spd wagons have same priority as M/E


% BOXN wagons run at 100 Kmph

BOXN Failures
Scenario

Gain
Scenario Description

2a 25% Hi speed BOXN 25 -2%


2b 50% Hi speed BOXN 50 0%
2c 75% Hi speed BOXN 75 0.60%
2d 100% Hi speed BOXN 100 3%
3a Hi Speed BOXN with 0 failures 100 0 9%
3b Hi Speed BOXN with changed priority 100 Y 9%
3c Hi Speed BOXN with 0 failures & changed priority 100 0 Y 10%

Table 2.3 Impact of phased introduction of High - Speed freight trains

In a different sets of simulation runs (Scenarios 2a to 2d and 3a to 3c of Annexure II.1,


extracted in Table 2.3 above) relating to varying fractions of just BOXN trains running at
100 km. /hr. (with base case HP/TL ratios) and all other freight trains running at existing
speeds, it transpired that there is, in fact, a capacity loss of about 2% if only 25% of
BOXN trains run at 100 km. /hr. Only when 75% of BOXN trains were made to run at
100 km. /hr. was there a reversal in negative gains to capacity. With 100 % of only
BOXN trains running at 100 km. /hr. there was a gain of just 3%. Further, when the
priority of hi-speed freight trains was equated with Mail Exp trains (Scenario 3b), then
the gain, over base case, went up to 9%. In yet another Scenario (Scenario 3a), increase
in BOXN speeds without change in train priority was found to give a 9% capacity
increase if failures of BOXN trains were altogether eliminated. Combination of these
two steps (0 failures for BOXN and change in priority) resulted in a total gain of 10%.
These results lead to the conclusion that increase in freight speeds would have to be
across the board for significant impact on capacity and would have to be implemented
relatively quickly to ensure that speed differentials within the freight block come down,

22
Throughput Enhancement Team 1 May 2005

instead of going up. Restricting higher speed potential to just a cross section of freight
trains will add to speed differentials and, thereby, depress fluidity. Preferably, increase in
speed of all freight trains should be combined with increase in HP/TL ratios and/or
change in train running priority to be of any significant benefit.

HORSE POWER TO TRAILING LOAD RATIO

Analysis of effect of HP/TL ratio was done through simulation studies on the SEB-MGS
section (Table 2.4). These studies indicated that the correlation between HP/TL ratio
and transit time (for free running time) is of the order of 0.98. However, this correlation
was found to be valid within a certain range only i.e. up to a value of 1.5. Increase in
this ratio beyond 1.5, especially on a flat terrain, was not found to show a commensurate
decrease in free running times.

Sl No Train Type HP/TL Transit Time (Min)


1 1 WDG2, 57 Ld BOXN 0.661 171
2 2 WDG2, 57 Ld BOXN 0.985 150
3 1 WAG7, 57 Ld BOXN 1.048 147
4 2 WAG7, 57 Ld BOXN 1.576 132
5 1 WDG2, 57 Ety BOXN 2.235 133
6 1 WAG7, 57 Ety BOXN 3.524 130

Table 2.4: Effect of HP/TL ratio on running time4

In Scenario 9a (Annexure II.1, extracted in Table 2.5 below), the gain from Hi Speed
BOXN/BTPN wagons with M/Exp/Pass halts reduced and schedule tightened was found
to be about 9%. Further in Scenario 9b, it was found that double heading all
BOXN/BTPN trains lead to a total gain of about 29%. As such, an incremental gain of
almost 20% was obtained from the increase in HP/TL ratios. However, comparison of
Scenario 2d (gain of 3% with hi-speed BOXN) with Scenario 5 (gain of about 13% from

4
Bharat Salhotra’s paper on “Comparative Analysis of Diesel vs. Electric Traction” (LRDSS Ph II).

23
Throughput Enhancement Team 1 May 2005

double headed hi speed BOXN/BTPN) shows an incremental gain of about 10% from
higher HP/TL ratios. This leads to the conclusion that benefits of higher HP/TL ratios
can be leveraged by using a combination of steps like reduction of stoppages and
tightening of M/Exp/Pass schedules.

Pass Schedule Tightened

M/E schedule Tightened


M/E/Raj Halts = 1 min
Pass Halts = 1 min
BOXN Speeds

BTPN Speeds
Scenario

HP/TL

Gain
Scenario Description

1d All loaded freight 2WAG7 BOXN/BTPN=100 100 100 Y Y 33%


2d 100% Hi sp BOXN 100 3%
5 Hi Sp freight double head 100 100 Y 13%
7 Hi Sp freight double head 100 100 Y Y 18%
9a Hi Spd options 100 100 Y Y Y Y 9%
9b Hi Sp freight double head 100 100 Y Y Y Y Y 29%

Table 2.5: Impact of High HP/TL ratios on Capacity

Further, on comparing Scenarios 7 (100 km. /hr. BOXN/BTPN double headed with
2WAG5 & and passenger stops reduced to 1 minute, gain = 18%) and Scenario 1d (100
km. /hr. BOXN/BTPN, all freight double headed, passenger stops 1 minute, gain = 33%),
it is seen that an incremental gain of almost 15% is obtained from increasing HP/TL
ratios all across the board for freight trains as compared to increasing HP/TL ratios only
for hi-speed BOXN/BTPN trains.

HIGH SPEED TURNOUTS

The impact of high-speed turnouts on a standalone basis was simulated for Gurpa –
Gujhandi section. It was found that raising the interlocking speed alone from 15 kmph to
50 kmph led to a marginal gain in transit time of about 5.9% on an average (Annexure
II.2). However, when the effect of high-speed turnouts was analyzed in conjunction with
increase in speeds of all freight trains as well as passenger trains to 100 km/hr as well as
zero failures and automatic signaling on MGS – SEB section (Scenario 10 a of Annexure

24
Throughput Enhancement Team 1 May 2005

II.1), it was found to result in 39% gain in capacity. In a different set of scenarios
evaluated for Baroda – Surat section, the switch from 15 kmph turnouts by 50 kmph
turnouts ceteris paribus was assessed to result in an increase in capacity from 83 trains to
92 trains per day5. It must be added however, that a lower number of precedence events
that, in turn, are a result of reduced speed differentials and lesser priority groups would
reduce the benefits of high - speed turnouts. This point is elaborated upon in Chapter III.

TIGHTENING SCHEDULE OF PASSENGER TRAINS

The effect of tightening of schedule of passenger trains was analyzed in two parts. First,
the effect of reduction of passenger halts to one minute each and second the effect of
taking out the slack in mail/exp/pass train schedules. In case of the second alternative,
the model was allowed to use the passenger (M/Exp/Pass) train arrival times as obtained
from the train speeds and priority based estimates.

Pass Schedule Tightened

M/E schedule Tightened


M/E/Raj Halts = 1 min
Pass Halts = 1 min
BOXN Speeds

BTPN Speeds
Scenario

HP/TL

Gain
Scenario Description

4a Pass halts Y 2%
4b Pass halts + schedule Y Y 7%
5 Hi Sp freight double head 100 100 Y 13%
6 Hi Sp freight double head 100 100 Y Y 15%
7 Hi Sp freight double head 100 100 Y Y 18%
8 Hi Sp freight double head 100 100 Y Y Y 21%
9a Hi Spd options 100 100 Y Y Y Y 9%
9b Hi Sp freight double head 100 100 Y Y Y Y Y 29%

Table 2.6: Impact of Tightening of Passenger Schedules

The reduction of passenger halts to 1 minute for the SEB-MGS section was found to lead
to a gain of 2% in capacity (Scenario 4a of Annexure II.1, extracted in Table 2.6 above).

5
Bharat Salhotra’s paper “Simulation Modeling of Baroda – Surat section: Evaluation of Investment
Alternatives” (Rail Transport Journal, Oct-Dec 2000)

25
Throughput Enhancement Team 1 May 2005

When passenger train schedules were modified to allow trains to arrive at the stations as
per train speed and priority (ie. arrival times did not have any additional recovery times)
and halt duration reduced to 1 min, a total capacity gain of almost 7% was obtained
(Scenario 4b of Annexure II.1, extracted in Table 2.6 above).

However, when these Scenarios were analyzed in conjunction with speeding up of freight
trains (BOXN/BTPN at 100 km. /hr. with higher HP/TL ratios i.e. double headed WAG7),
the incremental gain due to reduced passenger stoppage timings was found to be about
5% (Scenario 7 vs Scenario 5 of Annexure II.1, extracted in Table 2.6 above). In
Scenario 6 (Annexure II.1, extracted in Table 2.6 above), reduction of slack with high
speeds (BOXN/BTPN at 100 km. /hr. with higher HP/TL ratios) was found to give an
incremental gain of about 2%. Combination of reduced stoppages & reduction of slack
was found to give an incremental gain of about 8% (Scenario 8 vs Scenario 5 of
Annexure II.1, extracted in Table 2.6 above). There was a further gain of about 8% when
slack was also removed from M/Exp train schedules and stoppages reduced (Scenario 9b
vs Scenario 8 of Annexure II.1, extracted in Table 2.6 above).

As such it can be seen that length of passenger train stoppages and slack in the M/Exp
train schedules have an important bearing on sectional capacity.

FAILURES

Asset failure is not included in the present terms of reference. This factor is critical to
enhancing line capacity and throughput. The leverage that it offers in introducing
reliability of performance and consistency of operation is immense and hence, a section
has been included in this chapter. Impact of failures was simulated for a number of
sections and since the results obtained were rather significant, they are referred to here en
passe. Scenario 10a (Annexure II.1, extracted in Table 2.7 below) shows that increase of
all freight and passenger speeds to 100 km. /hr. with zero failures, 50 km. /hr. turnouts
and Automatic signaling leads to a gain of 39%. Finally, Scenario 10b shows that hi-
speed freight with tightening of M/Exp/Pass schedules with high HP/TL ratio and zero

26
Throughput Enhancement Team 1 May 2005

failures leads to a capacity gain of 73% even in the absence of hi-speed turnouts and
automatic signaling. In this context, Scenarios 2d and 3a are also relevant.

Pass Schedule Tightened

M/E schedule Tightened


M/E/Raj Halts = 1 min
Pass Halts = 1 min

50 Km. /hr. T/out


BOXN Failures
BOXN Speeds

BTPN Speeds

All Frt Speed

All Failures
Pass Speed

Signaling
Scenario

HP/TL

Gain
Scenario Description

2d 100% Hi sp BOXN 100 3%


3a Hi Sp BOXN with 0 failures 100 0 9%
10a Ideal 1 w/o halt/schedule mod 100 100 100 100 0 Y Auto 39%
10b Ideal Case 100 100 100 Y Y Y Y Y 0 73%

Table 2.7: Impact of Failures on Line Capacity

Simulation studies based on typical traffic patterns over a one - week period on different
sections showed equipment failures results in substantial loss of line capacity. Thus, in
one week, on Sonnagar - Mugal Sarai Section, a total of 105 train hours was lost on
account of failures of various types. This loss had a multiplier effect on the performance
of the section leading a cascading loss to the tune of 395 train hours. In other words, it
was found that the cascade factor was of the order of 3.5. Viewed differently, it implied
that any improvement in asset reliability leading to a reduction in failures by a factor of
one provides a very high leverage (of the order of 3- 4 times the reduction in direct
failures) in terms of line capacity or throughput.

The following percentage losses (in 1996-’97) of train hours, directly on account of
failures, were obtained from seven days control charts of the sections given below:

a. Sonnagar – Mughal Sarai : 11%


b. Vadodara – Surat : 18%
c. Tundla - Bayana : 20%

Failure reduction therefore, provides immense leverage for enhancing throughput.

27
Throughput Enhancement Team 1 May 2005

SUM UP

Analysis of a combination of factors shows that in case of various capacity enhancement


alternatives considered, the whole is not necessarily equal to the sum of its parts. For
example, the reduction in passenger halt durations to 1 minute was found to lead to a gain
of 2 % and opening up of the passenger schedules was found to lead to another 5% gain.
However, when use of high - speed wagons was considered alongwith, the reduction in
halts led to an incremental gain of about 5%. In this case, tightening of the passenger
train schedules only gave a further gain of 2 – 3%. In case of the HP/TL ratios also, it
was found that increase of HP/TL ratios with higher freight speeds gave an incremental
gain of 10%. On the other hand, when passenger train schedules were tightened up and
stoppages reduced with higher freight speeds, then increase in HP/TL ratios gave an
incremental gain of almost 20%.

Best, if not utopian, results were obtained when a multi-pronged strategy, including
introduction of high speed wagons, reduction of passenger train stoppages and tightening
of schedules, increasing HP/TL ratios to 1.5 and eliminating equipment failures was
simulated (Scenario10b).

28
Annexure II.1

Hi-Spd wagons have same


Pass Schedule Tightened

M/E schedule Tightened


M/E/Raj Halts = 1 min
Pass Halts = 1 min

50 Km. /hr. T/out


BOXN Failures

priority as M/E
BOXN Speeds

BTPN Speeds

All Frt Speed

All Failures
Pass Speed

M/E speed

Signalling
Remarks
Scenario

HP/TL
Gain
1a All frt 80, ld frt dbl headed 19% 80 80 80 Y Y

1c All ld frt 2WAG7 BOXN/BTPN=90 21% 90 90 Y Y

1d All ld frt 2WAG7 BOXN/BTPN=100 33% 100 100 Y Y

1e all trains 110 8% 110 110 110 110

2a 25% Hi sp BOXN -2% 100

2b 50% Hi sp BOXN 0% 100

2c 75% Hi sp BOXN 0.6% 100

2d 100% Hi sp BOXN 3% 100

3a Hi Sp BOXN with 0 failures 9% 100 0

3b Hi sp BOXN with changed priority 9% 100 Y

3c Hi sp BOXN with 0 failures & changed priority 10% 100 0 Y

4a Pass halts 2% Y

4b Pass halts + schedule 7% Y Y

5 Hi Sp frt dbl head 13% 100 100 Y

6 Hi Sp frt dbl head 15% 100 100 Y Y

7 Hi Sp frt dbl head 18% 100 100 Y Y

8 Hi Sp frt dbl head 21% 100 100 Y Y Y

9a Hi Spd options 9% 100 100 Y Y Y Y

9b Hi Sp frt dbl head 29% 100 100 Y Y Y Y Y

10a Ideal 1 w/o halt/schedule mod 39% 100 100 100 100 0 Y Auto

10b Ideal Case 73% 100 100 100 Y Y Y Y Y 0

29
Scenario Sheet ANNEXURE: II.2

Base Case Scenario Inputs Revised Scenario Inputs

Link Gomoh-Gujhandi Link Gomoh-Gujhandi

Length 104.65 Length 104.65


Gauge : Broad Gauge : Broad
Tracks Double Tracks Double
Terrain R1 ( 1 in 150 to 1 in 300) Terrain R1 ( 1 in 150 to 1 in 300)
Signalling MACLS with End cabins Signalling MACLS with End cabins
Traction Electric Traction Electric
Interlocking Speed = 15 Kmph Interlocking Speed = 50 Kmph

Traffic Characteristics Per Day


Trains/Week 566 81
Passenger 235 34
Freight 331 47

TOP
Description Scenario Base Case
Speed
Simulation Simulation
Train Type Results Results % Reduction in Travel Time TPC

1 1 Wag7, 58 Loaded BOXN 60 2.60 2.76 5.80 1.91


2 2 Wag5, 58 loaded BOXN 55 2.29 2.43 5.76 1.93
5 1 Wag7, 48 Loaded BTPN 60 2.78 3.12 10.90 1.88
6 1 Wag7, 58 Empty BOXN 65 2.31 2.48 6.85 1.70
7 1 Wag7, 35 Loaded BRN 55 2.69 2.85 5.61 1.98
9 1 Wag7, 40 Loaded BCN 55 2.54 2.83 10.25 1.98
11 1Wag7, 40 Empty BCN 60 2.41 2.59 6.95 1.83
1 Wag7, 40 Loaded
12 BOX/BCX 50 2.72 2.97 8.42 2.16
1 Wag7, 40 Empty
14 BOX/BCX 60 2.36 2.6 9.23 1.83
17 Light Engine 60 2.36 2.46 4.07 1.75
21 Passenger Trains 60 3.16 3.27 3.36 1.82
22 Rajdhani Express 110 1.06 1.1 3.64 0.98
23 Mail Express Trains 110 1.6 1.62 1.23 1.08
AVERAGE DELAY 66.15 2.24 2.38 5.88 1.76

30
CHAPTER III

INTRODUCTION

This chapter looks further at various factors impacting the operational initiatives outlined
in the terms of reference and, also, the interrelationships amongst them and line capacity.
Subsequently, in this chapter, the concept of speeds and speed differentials as also a
methodology for their assessment, review, benchmarking, etc is developed in the context
of actual observed data pertaining to a sample of sections.

In so far as the ground covered in Chapter II is concerned, it is important to expose the


dangers associated with generalization of results based on a limited analysis of a select
few sections. Also pertinent in this regard is to distinguish between two types of factors
that impact throughput.

a) Global factors, which are generally prevalent across the entire IR network. These
include train types, freight train consists, types of locomotives, their maintenance
standards and actual ratings and actual tractive effort vs. speed curves. Operating
rules that transcend across sections, zones & divisions include horsepower to trailing
loads, speed at turnouts, maximum speeds of freight and passenger trains etc.

b) Local factors, which are specific to a section. These include the physical
characteristics, grades, spacing between stations, length of critical block sections,
caution orders, operating train mix in terms of the number and types of passenger &
freight trains running over the section, corridor blocks as well as speed restrictions of
various types.

In the context of the present exercise, capacity of a rail network is the maximum number
of trains that can be run between pairs of points with an acceptable level of speed. This
definition of line capacity is based on the methodology6 adopted in Phase II LRDSS for

6
“Line Capacity Definition & Planning for Railroad Operations” – Bharat Salhotra ( Rail Transport
Journal, April June 1999)

31
assessment of potentially critical sections based on demand projections upto 2010-2011.
Since throughput and line capacity are being used interchangeably (see Chapter 1),
factors impacting capacity also impact throughput of the network.

Throughput is related to fluidity: the more fluid the movement over the rail network,
higher is the ability to run trains. A key objective of throughput enhancement must
necessarily imply a reduction of impedance that hinders or curtails the flow of traffic.
Fluidity of a network is impacted upon both by global & local factors, which have been
described above.

Because of the localized and tactical nature of train operations, it is important to resist the
temptation of offering global strategic solutions to improve fluidity, based on micro
analysis of a few sections, unless it is clearly established that these results are
representative enough and equally reflective of across the board operational performance.
Conclusions based on individual experiences or simulations in a select part of the
network, may often, when executed, not produce any substantive benefits in another part
of the network. A strategy that is most effective in enhancing throughput on a passenger
dominant link like Baroda - Surat, for instance, may not work with the same efficacy as
Sonnagar – Mugal Sarai or Ghaziabad - Rura, on which the traffic are more balanced in
nature. Segregation of factors into global & local is, therefore, fundamental to
understanding throughput.

FACTORS IMPACTING THROUGHTPUT

The focus in this section is on first analyzing throughput with a global perspective in
mind and then looking at the actual results as obtained on specific sections of Indian
Railways.

Throughput is impacted by two major features:

a) ‘Plant’ characteristics.
b) Headway

32
The following illustration provides a conceptual overview of factors impacting
throughput.

Throughput

Plant
Headway Characteristics

Gap b/w Slope of Priority of Signals (SA- Length


Lines Lines Freight 13) of Sections #
(SA-5 to Tracks
8)

Failures
Speed (SA-12, (FA- 1 to 4 Signals Signals
SA36,SA- 37) SA-61)
Speed at Station at Section
Differential Turnouts
(TA-1 to TA-4)
Commercial Speed
Max Speed Of Passenger
Of Freight (SA-1 to
5, 23,25,26,35)

Pass. Halts (SC-1,SC-2,


HP/TL Ratio SA-23,SA-25, SA-49)
(SA-43, SA-46,SA-49) Pass. Schedule (SC-2, SA-38,SA-39-SA-40,SA-49,SA-50)

Fig 3.1

Plant characteristics are the “hardware” of the railway network. These include type of
gauge, signaling system at sections and stations, number of tracks and quality of turnouts.
On the other hand, given a certain plant design, or given a certain “hardware” profile, the
single factor most responsible for throughput enhancement is headway. Headway is the
“software”, the key variable vital for maximizing throughput based on changes in factors
that impact slope of train lines (as depicted on control charts) as well as the gap between
them. Since this study is primarily limited to exploring impact of various operational
initiatives, modifications to plant characteristics and the concomitant impact on
throughput has not been analyzed.

Headway is the minimum time window that must exist between trains on a section, and
determines how many trains can be dispatched from a station. In the Absolute Block

33
System that is predominantly prevalent on IR, headway is impacted by the time it takes
the preceding train to cover the length of the block (which is a function of the booked
speed of the train, the plant characteristics of the section including the length of the
block). Ceteris paribus, the shorter the block, the less the headway and less is the
potential time window between successive trains. On the other hand, the lesser the time
taken to clear the block by the preceding train, or in terms of control charts, the steeper
the slope of the speed line, the less the headway. Speed differential – the gap between the
commercial speed of passenger trains and achieved speed of freight trains – therefore, has
a direct bearing on throughput. While passenger - train schedules and halts determine the
commercial speed of passenger trains, the speed of freight trains is dependent upon the
Horsepower (HP) to Trailing Load (TL) ratios as well as maximum permissible speeds
for the rolling stock. In either case, the slope of string lines (or train lines) is a key
determinant of throughput.

The second determinant of headway and hence throughput of a section is the variation in
the slopes of string lines. Given a section, and keeping the control chart view in mind as
the basis for further analysis, the variation in the gap between the lines is determined by
the variation in the slopes, i.e. by speed differentials. The lower the variation in speed
differentials, the more homogeneous is the flow of traffic and greater will be the fluidity
of the section. Variations in the gap between the control lines, therefore, is the second
most important factor having a direct impact on the headway achieved on a section as
well as on its throughput. Further, given a specific section, the steeper the control lines,
the less the headway and better the throughput. Speed of trains, therefore, has a direct
bearing on throughput.

These key factors, i.e. speed and speed differentials, depend on the following factors:

a) Horsepower to Trailing Load ratios of freight trains


b) Maximum Speeds of Freight Trains
c) Commercial Speed of Passenger Trains

These variables, depicted in red in the figure above, are fundamental to determination of
sectional fluidity and have, therefore, been elaborated on further in the following section.

34
HORSEPOWER TO TRAILING LOAD RATIOS

a) Horse power to Trailing Load Ratios of freight trains: The performance of freight
trains is impacted by the horsepower (HP) to trailing load (TL) ratio. The higher
the HP/TL ratios, the speedier are acceleration / retardation characteristics of
trains and, consequently, steeper are the slope of control lines for freight traffic on
control charts. In an earlier LRDSS study, HP/TL ratios in the range of 1.3 to 1.5
were determined as perhaps the most optimum for flat terrain under conditions
prevalent then. The study, which spanned detailed simulation over seven sections
of IR, concluded the following:

i) HP/TL ratios for ALL freight trains must be raised to ensure that
freight trains are able to achieve and sustain speeds of 100 km/hr
(close to booked speeds of passenger trains) over long stretches of
track and also achieve these speeds quickly.

ii) Increasing HP/TL ratios without introducing other significant


improvements will reap only marginal benefits in terms of
improving sectional fluidity and hence throughput.

iii) In a scenario of higher horsepower to trailing load ratios and


maximum speeds of freight train matching the booked speeds of
passenger trains, the introduction of modern technologies such as
high - speed turnouts and automatic signaling can significantly
improve the line capacity of a section.

b) On IR, HP/TL Ratios vary from 0.81 (one WAG5 hauling a loaded 58 BOXN
train) to 4.51 (one WAG9 Loco hauling empty 58 BOXN train). In reality,
however, since there is, in general, unidirectional flow of loaded traffic on a
section, the variation in case of freight trains varies from 0.81 to 2.11 in the
loaded direction and 2.92 to 4.51 in the empty direction.

35
DESCRIPTION HP/TL RATIO
LOADED TRAINS
58 Loaded BOXN headed with 1 WAG5 Loco 0.81
58 Loaded BOXN headed with 2 WDM2 Locos 1.02
58 Loaded BOXN headed with 1 WAG9 Loco 1.26
58 Loaded BOXN headed with 2 WAG5 Loco 1.63
58 Loaded BOXN headed with 2 WAG7 Loco 2.11
EMPTY TRAINS
58 Empty BOXN headed with 1 WAG5 Loco 2.92
58 Empty BOXN headed with 1 WAG7 Loco 3.79
58 Empty BOXN headed with 1 WAG9 Loco 4.51

Table 3.1: HP/ TL Ratios of Select Freight Train Configurations

This difference, in essence implies that a WAG9 headed freight train can achieve
speeds of 80 km/hr is approximately half the time and half the distance that it
takes for a WAG5 - headed loaded 58 BOXN train to achieve that speed. (It may
be noted that the gross weight of a loaded BOXN wagon was taken as tare +cc+2
tons. The ratios will have to be reworked for higher payloads - cc+8 – that are in
the process of being permitted on select routes for select commodities.)

c) The table below gives the HP/TL ratios of a select passenger train configurations.

PASSENGER TRAINS DESCRIPTION HP/TL RATIO

Commuter Train with 12 coaches headed by 1 WAM4 7.81


Commuter Train with 16 coaches headed by 1 WAM4 5.71
Shatabdi with 14 coaches headed by 1 WAM4 4.65
Shatabdi with 14 coaches headed by 1 WAP5 7.96
19 Coach Mail Express Train headed with 1 WAM4 3.97
19 Coach Rajdhani Express headed with 1 WAP5 6.17

Table 3.2: HP/ TL Ratios of Select Passenger Train Configurations

In the case of passenger trains, horsepower to trailing load ratios vary from a high
7.96 for a 14 coach Shatabdi Train to 3.97 for a Sampark Kranti / Mail Express
Train headed by a WAM4 Locomotive. Slow passenger trains (typically, trains
with three digit numbers, e.g. train no 302 or 742 on Sonnagar – Mugal Sarai
section) have horsepower to trailing load ratios and, consequently, a running

36
potential that is very similar to that of a Shatabdi Train. Given the high values for
HP/TL ratios of Passenger Trains in comparison to the Freight Trains, it is also
clear that in case of passenger trains, schedules and halts are the critical factors in
improving fluidity & throughput of a section.

d) Within the freight block, relative HP/TL ratios are perhaps as important as the
absolute values in terms of their impact on sectional performance. Having wide
variations in HP/TL ratios among freight trains – e.g. having some trains double
headed with WAG7 or WAG9 and, at the same time, some trains single headed
with WAG5 – can, in fact, lead to a reduction in throughput. Given that line
capacity is a “non inventoriable” asset, any loss on account of this relative
difference is a loss in perpetuity to the organization, the customers as well as the
nation. An incremental approach in improving HP/TL ratios in a phased manner
over an extended period of time may only result in further magnifying the relative
HP/TL differentials. Widening the dispersion of HP/TL ratios, which would be
inevitable in case of a slow phased improvement effort is therefore, likely to be
counterproductive in the interim period and may have long term repercussions as
well. This conclusion, based on simulation modeling, underscores a key
recommendation: the need to improve HP/TL ratios over a fairly short span of
time, rather than in an incremental, gradual manner.

e) With HP/TL ratios in the range of 4:1 to 5:1 in case of passenger trains, it is
generally possible to achieve maximum/booked speeds, starting from a stop, in
1.5 to 2 minutes. Given the current regimen of maximum speeds and passenger
train schedules as well as various allowances, HP/TL ratios in case of passenger
carrying trains are not considered as bottlenecks for enhancing line capacity /
throughput of a section.

MAXIMUM SPEEDS OF FREIGHT TRAINS

Along with reducing the dispersion of HP/TL ratios as well as improving this value in
the range of 1.3 to 1.5 (flat terrain), it is essential to improve the capability of wagons to
allow all freight trains to run at 100 km/hr. At present, this is not the case. Different

37
stocks (BOX, BOXN, BOBS, BOBR, BTPN, BCN etc.) are certified for a certain
maximum speed that is equal to or less than the booked speed of the slowest passenger
trains running on the section in many cases. In such cases, the potential benefits
achievable through higher acceleration & deceleration that typically accompanies higher
HP/TL ratios will not be fully realizable. Similarly, raising the maximum speeds of a
subset of wagons / trains (e.g. BOXN trains only) without taking steps to improve others
is likely to be counterproductive (Ref: Chapter II of this Report).

COMMERCIAL SPEEDS OF PASSENGER TRAINS

The Commercial Speed of a passenger train is the overall speed achieved by it over a
defined section. Commercial Speed depends not only on the slope of the train lines
between two stations on the Control Charts, but also the number and duration of
“horizontal holds” over a section.

Figure (Fig: 3.2) illustrates the concept of commercial speed, booked speed (of passenger
trains) and maximum speed (of freight trains) as well as speed differentials, through a
simple control chart drawn for a two section, three trains scenario.

Speed Differential
Commercial Speed of Pass. Trains & actual speeds of Freight Trains
e1 e2 e3

n
ai i n
Tr ra
ht tT
d
ee

g
ei i gh
Sp

Fr ath e
Fr ath
ial

P P
c
er
m
DISTANCE

m
Co

Halt
d
pee
S
edk
Boo

F1 P1 F2 s3
t1 t2 t3 t4
TIME

Fig 3.2

38
Freight Train F1 leaves the first station at time t1 and follows the path shown in black
based on its HP/TL ratio, maximum permissible speed on section and maximum speed
based on its ability to haul the load. It reaches its destination at e1. Passenger Train P1
follows trains F1. It runs at booked speed, halts as per the timetable and then proceeds at
the booked speed to the end of the section, reaching the end of the sections at time e2.

In the Absolute Block Section System, Train P1 can start from the 1st station only after the
F1 has cleared the block section, i.e. at t2. If the length of the “horizontal hold” of the
passenger train, i.e. the duration t3 to e1 is the Scheduled stoppage of train P1 at the
second station, it starts on time i.e. at e1 from this station. Train F2 can follow Train P1, as
soon as the latter has cleared the first block section i.e. at t3. However, in view of its
inability to match the commercial speed of the Passenger Train, it starts at time t4 and
arrives at the end of the 2nd block section at time e3 without having to ‘detain’ at the 2nd
station for ‘TOS’ of P1.

Given the passenger train schedule, the passenger train halts (both number and duration)
as well as the freight train speeds, throughput can be enhanced through minimizing the
gap between the commercial speed of the passenger train and the speed of the freight
train as actually achieved on the section. This can be done in three ways:

a) Increase the maximum speeds of all freight trains so that they are closer to the
commercial speeds of passenger trains. This requires wagons capable of running
at higher speeds as well as withstanding higher acceleration and deceleration
rates, higher horsepower to trailing load ratios so that effective acceleration and
deceleration times can be reduced and track configurations consistent with higher
maximum speeds of freight trains.

Keeping the halt duration of P1 at the second station and its commercial speed
unchanged, the picture that emerges is as depicted in Fig 3.3 on the following
page:

39
Speed Differential eliminated through improved Speed of Freight Trains
e`1< e1 e2 e`3< e3

ain

in
th t Tr

ra
ed

th t T
Pa igh

pe

Pa igh
lS
e

e
Fr

Fr
c ia
er
DISTANCE

m
m
Co
Halt

d e
Spe
ked
Boo
F1 P1 F2
t1 t2< t2 t3
TIME

Fig 3.3

Since e`1<e1, it in a sense alternatively allows the halt duration of P1 at the second
(intermediate) station to be reduced and, thereby, improve its commercial speed.

b) Increasing the speed of freight trains AND reducing the HALT length (duration
time of halt) of passenger trains can effectively reduce speed differentials even
further. Passenger & freight trains can then follow more closely and the gap
between the black freight lines (Freight Train Speed) and DOTTED passenger
line (Commercial Speed of Passenger Trains) reduces. This is illustrated in Fig.
3.4 on the following page:

40
Commercial Speed Improved through reduced Passenger Train Halt,
Freight Speeds re-aligned with improved Commercial Speed

n
rai

n
rai
Pa ght T

Pa ght T
ee
Sp
i
th
Fre

i
th
Fre
ial
erc
mm
Co
DISTANCE

Halt

d
pee
dS
oke
Bo
F1 P1 F2

TIME

Fig 3.4

c) Increase speeds of Freight trains AND increase slope of line of Passenger Train
(i.e. improved time tabling with lesser running slack) so that speed differentials
reduce. This results in maximum increase in fluidity of the section and is
illustrated below (Fig. 3.5):

Commercial Speed Improved through higher booked speed


Freight Speeds re-aligned with improved Commercial Speed
ain
Tr
ain

Pa ight
d
Tr

ee

th
e
Pa ight

Sp

Fr
th
e

ial
Fr

erc
mm
DISTANCE

Co

Halt
eed
ed Sp
Book

F1 P1 F2

TIME

Fig 3.5

41
d) Logically a fourth alternative could well be to reduce commercial speeds of
passenger trains to align with the speeds achieved by freight trains. However, this
alternative is not evaluated since it is considered a retrograde step; at least in a
global perspective, actual speeds achieved over IR are quite conservative.

Note
In the conceptual constructs sketched above, it is implicit that block operation time is
zero. This is purely by way of simplification. If block operation time is, say, k then Train
P1 in Fig 3.2 above will start from the first station at time t2+k instead of t2 and so on and
so forth. However, the conceptual underpinnings will not undergo any change. Strategies
for reducing / optimising k are continuously being evolved and implemented on IR. In the
context of track circuiting / provision of axle counters, etc the replacement of the
conventional lock and block instruments by non-cooperative ones, introduction of
automatic signaling (and possibly the moving block in future) especially on high density
routes, etc are some of the strategies already existing and being furthered on IR. In a
sense, reduced speed differentials will reinforce the case for these strategies to be
implemented more vigorously. Though the issue is of critical significance on a standalone
basis, it is strictly not germane to the scope of this exercise and is, therefore, not being
elaborated upon further here.

Each of the three strategies {i.e. a) to c) above}, taken together or independently, has
different implications in terms of economics, time to execute, delivery periods as well as
practicality. Further, these strategies need to be integrated and made part of a larger
strategic plan so that there is better alignment between overall business strategy and
operational strategies. It is, once again, critical to mention that the exact impact of each of
these strategies will vary from section to section since, as already mentioned earlier, train
operation is essentially driven by factors existent on the ground.

SYSTEM DYNAMICS

The sketch on the following page encapsulates the conceptual analysis for understanding
the inter-relations between various operational factors that contribute to throughput.

42
Variation in
booked speeds Variation in
Slack
Priority Groups + Impact of High + +
Speed Turnouts Variation in Slope
+ of train line
+
Precedences
+ + Variance in
Commercial Speeds
-
Throughtput Variation in
Variance in Duration of Halts
-
HP/TL Ratios
+ Variation in
Variance in Static time
achieved speeds +
+ Speed
differential
# of Passenger
Variance in Maximum Train Halts
Speed of Freight Trains

Fig 3.6

While there could be and, in fact, are other factors (such as failures, inter - station
distances, type of signaling system etc.) that impact throughput, the aforesaid system
dynamics analysis has been scoped out keeping the terms of reference before this
Committee in view.

Variations in Maximum Speeds of Freight Trains coupled with HP/TL Ratio variances
within the freight group lead to variance in achieved speeds, which in turn contribute, to
higher speed differentials across the freight group of trains. Similarly, variations in
booked speeds of passenger trains running on a section, along with variation in slack
provided on each section for different trains contribute positively to variance in
commercial speeds. Higher the variation in booked speeds, larger will be the variance in
commercial speeds. Further, higher the variation in static time (time when the passenger
train is not running), greater will be the variance in commercial speeds. Variance in
commercial speeds of passenger trains as well as variance in achieved speeds of freight
trains jointly contributes to higher speeds differentials, more overtakes, more hold backs
etc. On a section where the variances described above are minimal, the only factor that

43
causes trains to be held back for allowing other trains to overtake it are priority groups.
More priority groups mean more precedences, more starts, stops and lesser throughput.

In other words, the performance of a section depends, centrally, on average speeds and
standard deviations along with the distribution of speeds. In order to enable a fuller
appreciation of issues related to speed differentials, horsepower to trailing load ratios of
freight trains as also variation in commercial speeds of passenger trains, a sample study
of four sections was undertaken based on FOIS data spanning a one-week period. The
results of this study are presented in the following section.

SAMPLE SECTIONS ANALYSED

Sample sections were selected to analyze speed differentials as well as variations in


running times in various categories of traffic based on the following considerations:

a) The section, preferably, should have been studied earlier in LRDSS Phase II / III
since a detailed methodology was adopted then to select sections that were
representative of IR. Also, detailed simulation results would be available for analysis.

b) From amongst the aforesaid sections, a sample of three section types was short-listed
for the present exercise: one passenger dominant section (Baroda - Surat: 131 km),
one freight dominant section (Champa – Bilaspur: 53km), one mixed traffic section
(Sonnagar- Mugal Sarai: 122 km) and one section on which detailed simulation for
assessing impact of automatic signaling had been done (Ghaziabad - Rura: 370 km,
also passenger dominant).

For each of these sections, one week’s data (23/3/05 to 29/3/05) on freight movements -
departure times at start of section, the arrival times at end of section, the type of load, as
also the type & number of locomotives - in UP & DN directions separately was obtained
from FOIS. Passenger train schedules were extracted from divisional working timetables.
The data was then tabulated on an excel sheet and analyzed for mean and standard
deviation in terms of times taken separately in UP & DN directions and separately as well
as jointly for Freight and Passenger Trains.

44
RESULTS

The following table presents the average speeds and standard deviation in speeds (both in
km/hr.) on select sections on Indian Railways. Freight results are based on actual data
collected for the sample week (23/3/05 to 29/3/05) while in case of passenger trains is
based on scheduled rather than actual running times obtained from divisional working
timetables.

Traffic
Pass. Freight Mix
Trains Trains Ratio: Mean Mean St. Co-eff. Co-eff. Pass/
(24- (24- Pass. / Speed Speed St. Dev. Dev. of Var. of Var. Frt.
FROM TO Hrs) Hours) Frt (Pass.) (Frt.) (Pass.) (Frt.) (Pass.) (Frt.) Spd
FREIGHT DOMINANT:
BSP CHP 15 23 0.7 55 30 9 15 16 50 1.8
CHP BSP 17 30 0.6 39 18 10 12 26 67 2.2
BALANCED TRAFFIC
MGS SEB 22 23 0.96 72 35 18 11 25 31 2.1
SEB MGS 23 17 1.4 56 24 14 10 25 42 2.3
PASSENGER DOMINANT:
ST BRC 66 25 2.6 60 30 11 9 18 30 2.0
BRC ST 63 21 3 57 20 12 12 21 60 2.9

GZB RRH 49 21 2.4 75 30 14 9 19 30 2.9


RRH GZB 49 19 2.1 68 34 13 9 19 26 2.2

Table 3.3: DISTRIBUTION OF AVERAGE SPEEDS & STANDARD DEVIATION ACROSS


SELECT SECTIONS

Graphs for these sections illustrating the dispersion in travel times for different types of
traffic are placed at ANNEXURE 3.1.

The following conclusions can be derived on the basis of the aforesaid data:

a) On a freight dominant section, higher HP / TL ratios in the empty direction


(Bilaspur - Champa) result in higher average speeds of freight trains. Because of
the dominant operating mix in favor of freight traffic, passenger trains moving
in the empty freight direction have been timetabled to move at higher speeds as
compared to the timetabled speeds in the reverse or loaded freight trains
direction.

45
Scatter of Train Speeds -7 days (CHP-BSP) Scatter of Train Speeds -7 days (BSP-CHP)

70 120

60 Passenger Passenger

Commercial Speed ( Km./Hr.)

Commercial Speed (Km./Hr.)


100
50
x= 39 =10 80
40
60 x= 55 =9
30
Freight
x= 18 =12 40
Freight
20
x= 30 =15
10 20

0 0
1 51 101 151 201 1 51 101 151 201 251
Trains Trains

Fig 3.7

b) On freight dominant sections, HP/TL ratio of freight trains results in higher


freight speeds which in turn, appear to implicitly drive the time tabling strategy
for passenger trains. On Passenger Dominant sections, assuming that there is
adequate HP/TL ratio (1.3 and above) for freight trains, timetabling passenger
train must be the dominant strategy to improve sectional throughput. Improving
HP/TL ratios on freight dominant sections will have a direct bearing on
improving sectional fluidity as also allow stringent time tabling of passenger
trains.

c) On a Passenger Dominant section such as Baroda - Surat, passenger trains are


the key driver of freight trains performance and hence, commercial speeds of
passenger trains have a direct impact on speeds achieved by freight trains. Thus,
on Baroda - Surat, higher passenger train speeds result in higher average freight
speeds being achieved on the section. In a sense, higher passenger train speeds
appear to PULL up the graph of freight train speeds on a passenger dominant
section.

46
Scatter of Train Speeds (BRC-ST) Scatter of Train Speeds (ST-BRC)

90 90
80 Passenger 80 Passenger
Commercial Speeds (Km./Hr)

Commercial Speed (km./hr.)


70 70
x= 57 =12 x= 60 =11
60 60
50 50
40
Freight 40 x= 30 =9
30 x= 20 =12 30
20 20 Freight
10 10
0 0
0 20 40 60 80 0 20 40 60 80
Trains Trains

Fig 3.8

d) Standard deviation in commercial speed of Passenger Trains across these four


sections varies from 9 km/hr to 18 km. per hour, (Ref: Table 3.3) suggesting the
need to timetable passenger trains so that the spread of commercial speeds
(booked speeds + halts) is reduced. This would enable a more uniform flow of
passenger trains.

e) Standard deviation in case of Freight trains varies from 9 km. per hour to 15
km./hr (Ref: Table 3.3) – the highest variation being observed in freight
dominant sections and the lowest in passenger dominant section.

f) Co-efficient of variation (ratio of standard deviation to mean * 100) is used to


compare two distributions with different means. In the samples under study, it is
seen that passenger trains have a co-efficient of variation = 26 (BSP-CHP)
whereas in the case of freight trains, this figure jumps up to 66 in case of freight
dominant Champa - Bilaspur section. These results are likely to be witnessed
across other sections of IR as well. Containing this value at a low absolute value
and also within a narrow band should be possible in case of passenger trains
since they are, essentially, uniformly equipped to travel at same speeds. The
only reason for variability must be duration and number of halts which vary
from passenger train type to passenger train type. Part of this variation is
unavoidable; however, in terms of booked and maximum speeds at least, there
can be higher uniformity across train types.

47
g) Co-efficient of variation in case of freight trains may be attributed to different
HP/TL load ratios leading to variance in running times and effective speeds on
section, variance in maximum speeds for different wagons, as well as running of
passenger trains on a section. In order to isolate the impact of HP/TL ratios,
analysis of freight train timings on select sections revealed that having some
freight trains double/triple headed and others single headed has no impact on
their running timings. Analysis of sample data for BSP-CHP and GZB-RRH,
both in UP & DN directions reveals that double/triple heading, did not result in
lower sectional running times. Calculation of f-Statistic (used to test hypothesis)
revealed that it can NOT be stated with 98% confidence that double / triple
heading only SOME freight trains improves their speeds. This is an important
conclusion substantiated with simulation results obtained in Phase II (see
Chapter II).

h) It is interesting to note that despite a substantial variation in the commercial


speeds between the UP & DN directions on SEB-MGS, the coefficient of
variation for passenger trains in both directions is practically the same and ratio
of average passenger to average freight speeds is also uniform. It would appear
that providing higher HP/TL ratios coupled with reduced slack and improved
time tabling on the loaded direction, may be the best strategy to enhance
throughput on a section such as SEB-MGS that has balanced traffic.

Scatter of Train Speeds - 7 Days (SEB-MGS) Scatter of Trains - 7 days (MGS-SEB)

120
100
Commercial Speeds (Km./Hr.)
Commercial Speed (Km/Hr.)

Passenger 100 Passenger


80 x= 72 =18
80
60
60
x= 56 =14
40
x= 35 =11 40
x= 24 =10
20 20
Freight Freight
0 0
0 50 100 150 0 50 100 150 200 250
Trains Trains

Fig 3.9

48
i) In order to develop a better understanding of passenger timetabling, the
performance of passenger trains in UP & DN directions for SEB– MGS (balanced
traffic) as well as the GZB-RRH (passenger dominant) sections was analyzed.
The objective of this analysis was to estimate slack times that exist for each
passenger train, given its booked speed, the halt times and arrival & departure
times.

In the case of SEB-MGS, it was observed that time taken by passenger trains
varied between 88 minutes and 227 minutes. The lowest booked speed was 90
km/hr while the highest booked speed (Rajdhani Express) was 120 km/hr. Similar
variations in durations are observed on Passenger Dominant, BRC-ST section
(Lowest time = 96 min, Highest time = 258) as well as Freight Dominant, CHP-
BSP section. Additionally, there is a wide variation in the slack provided for
different passenger trains both in the UP & DN directions. In the case of MGS-
SEB, slack was calculated by deducting the total stop time, acceleration and
deceleration time (2 min for each acceleration & deceleration) from the time
tabled running time. It is observed that with this methodology, the range of slack
varies from a high 23% (Train 2308) to 2% (Trains 3010/ 2322) in the MGS-SEB
direction. In the reverse direction, slacks ranged between 8% and 43% with an
average slack of 29%.

Understandably, higher slack in SEB-MGS direction is due to the fact that Mugal
Sarai is the terminating node of a zonal railway and hence provisions on account
of EA, TR, etc. have been provided at the end of the section (i.e. short of Mughal
Sarai). Be that as it may, it is felt that there is scope for rationalizing the allocation
of slack based on state of the assets as well as their maintenance norms /
requirements.

In the case of GZB-RRH, only those trains that traverse the entire length have
been considered for slack analysis. A total of 10 pairs of passenger trains that
traverse only a part of this link have not been considered.

49
SLACK as percentage of running time
SECTION
MINIMUM MAXIMUM AVERAGE
GZB-RRH 16 43 29
RRH-GZB 12 36 21
SEB-MGS 8 43 29
MGS-SEB 2 23 9
Table 3.4
The time taken by Mail/Exp trains was found to vary between 216 to 563 minutes
in DN (freight empty flow) direction and 235 to 631 minutes in UP (freight
loaded flow) direction. Average speeds were 68 Km and 75 Km. /hr. respectively
for UP & DN directions; however, the standard deviation and coefficient of
variation in both directions are almost same. However, there is a wide variation in
the slack available in different trains, with higher slack being available in the UP
(loaded freight) direction than in the DN (empty freight) direction. It is also seen
from the working timetable that Engineering Recovery in UP & DN for each train
varies by over 25 % (36 min. in UP direction & 28 min. in DN direction).

j) The following table (Table 3.5) illustrates the variation between the slowest
moving and fastest moving passenger trains. As can be seen, there is a wide
(commercial) speed differential that exists within the passenger trains even in
terms of timetable; in reality, there could be even higher variations on day to day
basis. Even on a short 53 kilometer link like Bilaspur - Champa, the ratio of times
of the slowest and fastest moving passenger trains is as high as 2.16.

50
DISTANCE Mean Speed Min Time Max. Time Max Time/
FROM TO (Km.) (Km/Hr.) (min.) (min.) Min Time

Bilaspur Champa 53 55 45 78 1.73


Champa Bilaspur 53 39 58 125 2.16

Mugal
Sarai Sonnagar 121 72 71 236 3.32
Mugal
Sonnagar Sarai 121 56 88 227 2.58

Surat Baroda 131 60 96 258 2.69


Baroda Surat 131 57 92 270 2.93

GZB RRH 370 75 216 563 2.61


RRH GZB 370 68 235 631 2.69

Table 3.5: DISTRIBUTION OF SPEEDS & TIME DURATIONS FOR PASSENGER TRAINS

It is apparent that there are speed differentials between freight and passenger trains
because of reasons embedded in differences in booked speeds, in maximum speeds as
well as in slack across and within the passenger train group. An analysis of slack existing
for various trains highlights the need to develop tighter timetables and higher booked
speeds for passenger trains. For example, a slow moving 8 to 10 coach passenger train
that is hauled by a WAM4 can logically operate like a Shatabdi, in terms of its
acceleration, deceleration as well as booked speeds. Historical variation between booked
speeds & maximum speed could also be reviewed for reducing slack in passenger train
operations.

The first step in the direction of improving throughput must be to identify trains pulling
down average commercial speeds and raising standard deviations within the passenger
trains group. Not unlike the impact the critical block section on a link has on line
capacity, the train(s) with the lowest commercial speed(s) impacts adversely on the speed
of freight trains following it. A consistent process of review of such trains, aimed at
increasing their booked speeds, reduction / rationalization / elimination in the number and
duration of halts as well as reduction in any slack that is built into their schedules will
ensure that the speed differentials within the passenger trains group is optimized.

51
Variations in commercial speeds of passenger trains can (and need to) be reduced /
rationalized but cannot be altogether eliminated. Local conditions, at times, influence the
arrival / departure time of trains at designated stations for a host of reasons – providing
connections to other trains, capacity issues especially at terminals / junction points
restricting their ability to receive / dispatch trains in certain time corridors, meeting the
needs of certain groups of patrons such as office goers, vendors etc. Such variations,
therefore, can best be analysed and rationalized at the field level.

Throughput can well be enhanced by making changes in the hardware, i.e. Plant
Characteristics. However, in most cases, that is an expensive option. Modifying the
“software”, i.e. a time table of trains that is continuously evolving to reflect higher
average speeds with lower standard deviation over time, is a lower cost, low gestation lag
alternative that must be explored and exhausted before or along with investments
decisions aimed at changing the plant characteristics.

SUMMING UP

A review of impact analysis diagram, system dynamics chart, graphs and histograms
drawn for sections under study as also detailed analysis of slack in passenger - train
running on Sonnagar – Mughal Sarai and Ghaziabad - Rura sections leads us to the
following conclusions:

a) There is a wide speed differential between freight trains speeds and the commercial
speed of passenger trains.

b) Within each of the groups (Passenger & Freight), the variation in times/speeds
between the slowest & faster moving trains lies between 3.5 (passenger) to 15
(Freight).

c) Since every passenger train is ‘powered’ adequately and available rolling stock and
track technology is ‘good enough’, the enhancement of booked / maximum
permissible speeds of slow moving passenger trains to match those of fastest Mail
Express Train on the section must be undertaken to reduce variation in the passenger
trains group.

52
d) Variations in number of halts as also their duration adversely affect throughput on the
section. Feasibility and extent of possible rationalization needs to be studied in a
systemic, continuous and iterative process of review at the field level, both from the
operational as well as commercial angles. Thereafter, the conflicting interests of
different groups of rail users would need to be reconciled for facilitating
implementation.

e) Replacement of conventional all-stopping passenger trains by MEMU / DMU


services has been undertaken on some sections. In every case, this replacement
should, ideally, be accompanied by rationalization in their schedules, including the
number and duration of halts; if this rationalization is not undertaken, the envisaged
benefits are not likely to be realized. The existing cap on the maximum permissible
speed of MEMU / DMU trains would also need to reviewed and relaxed.

f) Review of seven days results on various sections reveals that double heading / triple
heading of a few freight trains does not directly lead to improved train running. In
fact, no direct correlation is observed between run times and double / triple heading
of freight trains. This leads to the conclusion that within the freight trains set, every
effort must be made to ensure that the HP/TL load ratios do not exhibit a wide
variation.

g) Along with systemic steps to reduce the variation in commercial speeds, HP/TL ratios
for flat terrain needs to be simultaneously improved to around 1.3 to 1.5 so that their
overall performance can match the commercial speeds of passenger trains. For graded
sections, this ratio should be higher.

A METHODOLOGY FOR SYSTEMIC REVIEW & PLANNING FOR


THROUGHPUT ENHANCEMENT

The output, whether of a nation, organization, division, route or section, is evaluated in


reference to performance norms and benchmarks. Benchmarks serve as standards for
improving current performance levels and provide useful feedback to executives to
improve their performance. They firmly establish a process of review and analysis of on a

53
consistent ongoing basis with the objective of “getting more out of less”. Performance
norms help to identify areas of shortfall in processes as well as end results and are useful
to determine training requirements for staff and officers. Benchmarks are not standards
and limits etched in stone; instead, as the organization improves, and as technology and
external environment undergoes changes, these benchmarks must continuously be
reviewed, inspected and, if required, mid course corrections applied to reflect higher
levels of expectations and achievements. As stated by Jack Welch, former CEO of
General Electric, “organizations get what they inspect, not what they expect”; unless ,
therefore, benchmarks are institutionalized as part of the process improvement exercise,
and regularly reviewed, they will not serve the purpose meaningfully.

The throughput, fluidity and overall performance of a section depend centrally on


average speeds and standard deviations and these two statistics should now form part of
the reporting system on Indian Railways. Not only are these effective indicators of
fluidity of a section, they also provide very useful feedback on performance. This
reporting statistic can serve as an effective benchmark for operations, since in an essence,
it determines the quality of sectional running, both cross-sectionally in the short term as
well as over a long time series. In summary, if reporting systems / statistics are modified
to include the spread of speeds, as well as mean and standard deviation of speeds
achieved by passenger and freight trains separately as well as by the (combined) total
traffic flow, a more complete reflection of the section’s performance would become
available.

The set of histograms shown in Fig. 3.10 illustrates the kind of reports that can and
should now be generated and continuously reviewed for each section of IR. In the
illustrated report, the overall spread in transit time on Baroda – Surat, separately for
passenger, freight trains as well as combined traffic is depicted. In addition, the mean &
standard deviation, in terms of minutes is also indicated in the report.

54
Freight Trains on ST-BRC (22/3/05 to 31/3/05)
Freight Trains on BRC-ST (22/3/05 to 31/3/05)
60 50

Frequency 50 40
40 x= 316 30 x= 287

Frequency
30 =327 20
=91
20
10
10
0 0

0
12

18

24

30

36

42

48

54

60

66

72

78
0

0
12

18

24

30

36

42

48

54

60

66

72

78
M in Min

Passenger Trains: BRC-ST Passenger Trains: ST-BRC


25
25
20
20
Frequency

x=148

Frequency
15 15 x= 137
=44 =35
10 10

5 5

0 0

100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
More
0

e
10

12

14

16

18

20

22

24

26

or
Min Min.
M

All Trains: BRC-ST (22/3/05 to 31/3/05) All Trains: ST-BRC (22/3/05 to 31/3/05)
35 50
30
x=191 40
Frequency

25
=173 x= 178
Frequency

30
20 =96
20
15
10 10

5 0
0
0

0
10

16

22

28

34

40

46

52

58
100 130 160 190 220 250 280 310 340 370 400 430 460 490 520 550 580 More Min.
Min.

Fig: 3.10
The “Performance Histogram”, as depicted on page, forms the core of a consistent
methodology aimed at throughput enhancement through ‘software’ improvements. The
fundamental objective and / or utility of such reports will vary across the organizational
level that is analyzing it. At the level of Sr. DOM, it could well serve as a tool to even
train speeds and review the performance of section / deputy controllers. Alternatively, at
the level of Zonal Railways and the Board, it will serve as a planning tool to review
timetabling & scheduling of passenger trains, developing strategies for streamlining
freight speeds with improved commercial speeds of passenger trains, etc. This statistical
analysis could also be integrated with the section/route – wise investment planning
process and utilized for generating ‘soft’ options to address throughput and line capacity
issues.

55
CONCLUSION

With the implementation of FOIS on Indian Railways, deriving statistics on speeds and
speed differentials is now achievable at least for freight traffic. In fact, it could be
feasible for FOIS to build in the Performance Histogram as part of the Management
Information System Reports that is prepared for each Division/Zone/ Board. In regard to
passenger trains, the Coaching Operating Information System (COIS) when implemented
will facilitate the management to systematically obtain the actual mean & standard
deviation in commercial speeds; however, in the interim, time tabled data can be utilized
for working out planned / scheduled mean & standard deviation of passenger carrying
trains on each section.

Performance indices (in terms of average speeds and standard deviation) for the freight
trains performance can easily be generated at the field level from FOIS and can serve as a
very useful feedback for operational performance as well as evaluation of fluidity on each
section. This direct feedback on performance must be reviewed periodically at various
levels not with the objective of policing, or controlling, but with the objective of
improving and facilitating a throughput enhancement process primarily through soft
options. To begin with, benchmarks for each section could be presented in terms of these
statistics and be used in time tabling as well as planning processes. High cost, long
gestation – lag investments may be a long term solution to easing pressures especially on
congested corridors along on the Golden Quadrilateral. However, in the immediate, short
and medium term, streamlining of operations through the optimisation - of - speed -
differentials route may perhaps enable IR to achieve Mission 700 million tons within a
defined, limited time frame.

56
ANNEXURE # 3.1/1

x= 86 x= 59
=21 =10

x= 103
=104 x= 206
=238

57
ANNEXURE # 3.1/2

ANNEXURE # 3.1/2

58
ANNEXURE # 3.1/3

RRH-GZB Freight GZB-RRH Freight


40 60
35
50 Mean = 830
30 Mean = 690 40 SD = 562

Frequency
Frequency

25 SD = 230
20 30
15 20
10
10
5
0 0

e
5
5
4
4
3
33
62
92
21
51
80
or
38
51
64
77
90
e
5

5
4
4

3
33

62
92

21
51
80

10
11
12
14
15
16
or
38

51
64
77

90

M
Minutes
10

11
12

14
15
16

Minutes
M

RRH-GZB Passenger 70 GZB-RRH Passenger


90
80 60
Mean = 311
70 Mean = 340 50 SD = 65
SD = 70
Frequency

60
Frequency

40
50
40 30
30 20
20
10
10
0 0
235
260
285
309
334
359
384
408
433
458
483
507
532
557
582
606
e
5

Minutes Minutes
or
23

28

33

38

43

48

53

58

180 RRH-GZB All Trains GZB-RRH all trains


100
160 90
Mean = 456
140 80 Mean = 497
SD = 219
120 70 SD = 421
Frequency

Frequency

100 60
50
80
40
60
30
40 20
20 10
0 0
0

0
00

00

00
0

0
00

00

00

20

30

40

50

60

70

80

90
20

30

40

50

60

70

80

90

Minutes Bin
10

11

12
10

11

12

59
CHAPTER IV

RECAP

In the context of the terms of reference for this study, it is seen that speed differential is
the most critical factor in so far as impact on line capacity and throughput is concerned.
Elimination / significant reductions in speed differential will negate the need for
arranging a number of precedence events and, thereby, obviate the requirement of
implanting a plethora of high-speed turnouts7 especially across high-density routes.
Moreover, the augmenting of HP/TL ratios in case of freight trains and tightening
passenger schedules are strategies, which impact favorably on line capacity and
throughput because they, essentially, reduce speed differential.

STRATEGIES

PASSENGER TRAINS SEGMENT

Within the category of passenger trains being operated on a route, there exists presently a
large differential as measured in terms of mean speed, their standard deviation (variance)
and coefficient of variation (standard deviation divided by mean multiplied by 100).
These differentials are arising on account of variations in both the maximum permissible
as well as booked speeds and / or the number and duration of stoppages across trains and
/ or the recovery times (Engineering, Traffic, etc) provided. In case of trains providing
stoppages on any section, its commercial speed is the critical determinant of capacity
especially in the given regime of inter se priority wherein all passenger - carrying trains
enjoy precedence over freight trains. Improvement in commercial speeds of the slower
amongst passenger - carrying trains will not only reduce speed differential within the
passenger trains group but will also lay the foundation for running freight trains at higher
booked speeds and, thereby, attaining higher average speeds.

7
It is a moot point whether the first take-off point and if need be others on the most frequently used route
after the main line, especially at large stations on high-density routes, should be equipped with high –speed
turnouts. The associated movement would however, have to be appropriately and unambiguously signaled.

60
For achieving reduction in speed differentials within the category of passenger - carrying
trains, it is the running of the all – stopping passenger trains that first and foremost
deserve attention. These trains cannot be wished away at least in the short run. How,
then, is an improvement in their commercial speed to be achieved?

a. Based on findings relating to sections specifically examined in Chapter III, it


is clear that the all-stopping passenger trains are booked at maximum
permissible / booked speeds that are significantly lower than those of the
fastest passenger - carrying train on the section. Ceteris paribus, if this
variation is removed, i.e. if the maximum permissible / booked speed of the
all-stopping passenger trains is raised to the level of the Rajdhani or Shatabdi
or whatever is the fastest train of the section, its as well as the average
commercial speed of the group as a whole will go up. In other words, it is the
all-stopping train that deserves most to be booked at the highest feasible /
permissible speed on the section. A back - of - the - envelope exercise shows
that if the slowest passenger train on MGS-SEB section (Train no.302) is, in
fact, booked at 120 km. /hr., its bare running time over the section will come
down by 20 minutes. In addition to this, use of appropriate rolling stock can
reduce the acceleration + deceleration times by about three minutes per
stoppage. This would give a further gain of 51 minutes (17 stoppages).
These steps would result in improvement in its commercial speed from 31
km. /hr. to 46 km. /hr. even with existing stoppages and provisions for
engineering and traffic recovery times. The execution of this strategy requires
availability / earmarking of coaching stock with the requisite speed potential
for the present day category of slow runners. Costs and benefits are easily
quantifiable on case-to-case basis and consequently the associated proposals
can readily be made part of the annual works program exercise.

b. Simultaneously, a review of the number and duration of stoppages in the


context of ground realities and based on, if required, local consultations with
user groups can also result in reduction in the number of horizontal holds on
the control chart and, thereby, result in higher average commercial speeds and

61
lower spreads in the category of passenger – carrying trains. Such exercises
are best conceived and executed at the field level.

c. It is the slow, all-stopping passenger train that also deserves technological


support for achieving superior acceleration / deceleration rates. If time loss on
account of each stoppage is reduced, the commercial speed over the link will
also improve. Providing appropriate rolling stock – both coaches as well
motive power - for these trains is, therefore, essential. If these trains are
served predominantly on loop (instead of mainline) platform lines at stations,
high-speed turnouts may also be considered for reducing total time ‘loss’ at
each stop / start.

d. In case the required realignment in the commercial speed of slow-runners


with the faster / fastest passenger trains on the section continues to pose as a
constraint to the process of throughput enhancement in the long run, it may
even be prudent to evolve and implement schemes for altogether doing away
with the all-stopping passenger train. This would entail protracted and, at
times, even tough negotiations with user groups, the local administration, road
(even other modes like inland waterways based on ground conditions)
operators, etc. At the rate at which the road sector is expanding / expected to
expand, railways, in any case, may not remain the preferred mode for the
relatively short hauls in the passenger traffic segment (other than suburban)
for long.

62
FREIGHT TRAINS SEGMENT

The realignment of freight speed potential with improved (higher mean and lower
standard deviation) commercial speed of passenger trains, as brought out earlier in this
report, has a beneficial impact on fluidity and / or line capacity / throughput. Is there an
optimum speed potential that can be targeted in case of freight services? Since the
commercial speed of passenger carrying trains vary both within a given section and also
across sections, there cannot possibly be one unique optimum. To begin with, however, it
may be prudent to target a freight speed potential that is higher than the lowest
commercial speed in the category of passenger - carrying trains. If this is done, the
running of freight trains in time corridors that are either ahead of or significantly after the
slowest passenger train will achieve higher average speed and greater congruence with
the average speed of the faster Mail / Express trains on the section. And, if the
commercial speed of the slowest passenger train is a later stage improved on account of
strategies listed in the preceding paragraphs, the number of non-interactive parallel lines
on the control chart will increase thereby improving the given infrastructure’s ability to
run more trains more efficiently.

In the conceptual constructs that form the basis of this analysis, not all factors specific to
particular sections have been captured. Some of the factors are:

a) The nature and extent of permanent speed restrictions on the section and their
impact on the feasibility / required transit time of various trains to achieve
maximum permissible / booked speed.
b) The nature and extent of utilization of engineering and other allowances.
c) The nature and extent of asset failures and other unusual incidences on different
routes.

The impact of these factors may be more critical on some routes and less on others. It is
for this reason that section-specific strategies aimed at removing speed differentials
would have to be formulated (and executed) at the field level itself. As pointed in the
preceding chapter, the estimation of the mean speed and its range and standard deviation

63
for individual trains, categories of trains and all trains together should serve as a starting
point and the basis for this analysis. Variations in speed (both mean and standard
deviation) across different trains on a given section coupled with a comparison of speed
curves across sections pertaining to the same Division or across other Divisions / Zones
would bring into focus the underlying reasons and, thus, pinpoint the strategy for
reducing speed differentials. The movement of these indices over time would indicate
whether the direction and speed of progress is right or not.

Divisions / Zonal Railways would be well advised to immediately formulate proposals


aimed at easing speed differentials, on high - density routes to begin with. The proposals
could then be reviewed and their implementation monitored at various levels, including
the Board.

64

Вам также может понравиться