….let's share our railways knowledge


By on April 6, 2021

Synopsis: The paper examines the dynamics of the CONRAJ train running on the Western Railway. Based on analysis of data obtained by foot plating, a train resistance curve for the CONRAJ train is constructed. The authors conclude that there is a considerable mismatch between the characteristics of the train and that of the locomotive used to haul it. To remedy the situation, a new design of locomotive within the tap-changer- DC traction motor framework but with a different gear ratio is suggested.


Containerized cargo is the norm in the international transport business. Goods packed in containers can be transported by several modes such as ships, trains and lorries to reach the destination fast and safe. In India, the Container Corporation of India (CONCOR) is the premier agency involved in this multi-modal logistics business. CONCOR undertakes the transportation of containerised traffic by rail or road. Container traffic on Indian Railways (IR) has been growing at an impressive pace in the past decade. At present, daily eight pairs of container trains run between Jawaharlal Nehru Port (JNPT) in Mumbai and major centers in the hinterland such as Tughlagabad and Ludhiana. These trains now account for nearly 40% of the total number of goods trains on the Western Railway (WR) route between Mumbai and Mathura. Container trains are gaining ground on other routes too all over IR.

WR has introduced a special train since 1998 aptly named ‘CONRAJ’: meaning CONtainer- RAJdhani. Expectations from the train are obvious- devoted exclusively to containerized cargo – running fast like the famous Rajdhani Express. The tariff structure puts a premium of 2.5% on certain improvements over the contracted transit time of 48 hours between Mumbai and Tughlakabad and an equal penalty otherwise. The train comprises specially designed wagons that can accommodate the containers and are fit to run at 100 Km/h. WR has a practice of allotting preferably a WAG-7 or a WAG-5 locomotive that is fit to run at 100 Km/h to this train. Nearly two-thirds of the trains are run by WAG -7.

Several officers dealing with operation on WR had observed that the train could not attain the speed 100Km/h on level track let alone on rising gradients in spite of the rather high hp per trailing ton. This is what attracted our attention to the problem. During extensive foot plating on the train, our observations confirmed that it could never attain the speed of 100 Km/h save on a few stretches of falling gradients. The maximum speeds attained with WAG-7 locomotive on level ranged between 80-85 Km/h. It should be noted that a high balancing speed and quick acceleration at start and after negotiating speed restrictions contribute towards increasing the average speed of this transit time-sensitive traffic.

Technical Characteristics of Container Train

CONRAJ train occupies nearly full loop length. It comprises 9 units of a 5 wagons consist that is about 69 meters in length making the total train length of 621 m. Two end wagons have CBC couplings and three middle one are connected with slack-less drawbars. Wagons have the following special features:

  1. Lightweight welded design of under-frame. Tare weight: BLCA 19.1t, BLCB 18.0t, Payload:61t, Length: BLCA 13.63m, BLCB 12.21m.
  2. Low platform container flat wagons to carry high-cube or Tallboy containers on routes where clearances would otherwise make this impossible.
  3. Though the  Casnub bogie used on  BOX-N  wagons is used wheel diameter is lower at 840 mm (new).
  4. Air brakes, single pipe feed, automatic load sensors.
  5. The maximum permitted speed: 100 Km/h.

Train Load: Containers are loaded on these wagons. Containers can be of several types dimensions also vary considerably. ISO standards permit considerable flexibility in the dimensions of containers. Maximum permissible loading in containers is restricted by the handling capacity of cranes at dockyards and is restricted to 61 t. Though the theoretical maximum load of the train works out to be 3600 t, actual loading depends on the commodity loaded and extent of filling. The train carries the load in both directions. During the period of foot-plating i.e. August –December 2002, we had found the trainload varied from 1640 t to 2480 t. Therefore we chose to make observations on trains at the higher end of the load spectrum i.e. 2015t, 2250 t and 2480 t. Further, we made calculations for the heaviest one i.e. 2480 t train.

To get an idea of a recent trends, the data of trainloads passing through Vasairoad station between March 1st and 7th was analysed. Table-1 gives the summary.



Statistic Value
No of trains 82
Min. Load (t) 815
Max. Load (t) 3251
Median Load (t) 2245
Mean Load (t) 1993
Std Dev Of load (t) 524
% of trains with less than 2480 t load 93
% of trains with 91 Units 85


Table-1 confirms the earlier trend. The variation in the load from train to train is too large making it difficult to make a decision about the motive power. Still, it is clear that a vast majority of the loads are in the middle bracket of 1900t –2400t. Our calculations for 2480t will be an excellent guide as 93% of the trains are lighter than this. This train is much lighter than the conventional goods train of 4750 t (58 BOX-N) but heavier than, though not by big margin, the longest passenger train 1430t (24 coach). The characteristics of this traffic make it a class by itself.

Train Resistance: A logical first step towards analyzing the problem was to understand the resistance offered by the train. We approached RDSO but could not obtain the required train resistance formulae. As a second-best measure, we decided to collect data by foot plating on the train. A lot of data was collected during foot plating to understand the resistance offered by the train as accurately as possible. Observations made on three different train-locomotive combinations are shown at annexure-1. During the run TM voltage, TM current, speed, and gradients were recorded. Care was taken to record data only after the speed had stabilized for a particular gradient. Table-2 indicates the train resistance for three different trains arrived at after analysing data in annexure-1.


Speed (Km/h) Train Resistance offered by the trains of different weight excluding locomotive (t)
2480 t 2250 t 2015 t
70 6.17 5.18 5.91
80 8.34 8.00 7.69
85 9.68 9.57 8.81
90 11.00 11.07 NA
95 12.55 12.50 NA
97.5 13.56 NA NA


It is clarified that all the data has been collected using the normal instruments fitted on the locomotives. No special instrumentation was done. In spite of rather unsophisticated instrumentation and procedure, the remarkable consistency of observations and their conformity with observations made by others is impressive. This effort does not replace a proper dynamometer trial to determine the train resistance yet it provides an insight into the behavior of the train so essential to analysing the issue at hand further.

A smooth train resistance-speed curve was developed based on the values calculated as above. For better appreciation, we compared it with that offered by a loaded BOX-N rake of the same weight. Fig-1 shows that the observed train resistance curve departs from that predicted by the formulae for BOX-N train of the same weight as the speed rises past 50 Km/h. In interpreting this curve, let it be recalled that the train resistance comprises two parts: One, mechanical resistance (bearing resistance + rolling resistance) that is proportional to trainload; and, two, aerodynamic resistance that is independent of train load but depends on the shape, size and spacing of vehicles. Hence the train resistance for BOX-N train shown in Fig-1, which has been calculated using RDSO formulae for loaded BOX-N wagons giving total resistance value per ton of trainload, maybe slightly lower than actual as load here is much less than the usual load of the BOX-N train.


Comparison of Train Resistance for CONRAJ and BOX-N wagon trains of 2480 t

In addition to this, the difference in train resistance of the two trains may be due to much higher aerodynamic resistance occasioned by the following factors:

  1. Large gaps between the units of five vehicles (Fig-2) allowing swirling of wind behind the last container of unit ahead and pushing of the wind column by the first container of the following unit.
  2. Uneven gaps in height and width between consecutive containers arising due to differences in sizes of containers.
  3. The uneven surface of containers increasing the friction along sides.


Indeed all the above factors could vary from train to train as the size of containers varies widely. Even the number of containers varies. Therefore even with properly designed trials, it will be possible to determine only a range within which train resistance may vary. Thankfully 90% of the trains carry 91 units indicating some standardization that has implications for train resistance.

Significantly this results in power demand from the locomotive rising steeply with speed. Table –3 gives the estimated power demand of the train. It is clear that a 5000 hp locomotive may be necessary if we want to operate in the higher end of the speed range.


2480 t Container Train on a level gradient

Speed (Km/h) Train Resistance excluding locomotive resistance (t) Power required to haul the train only (hp)
70 6.17 1600
80 8.34 2470
85 9.68 3050
90 11.00 3670
95 12.55 4420
97.5 13.56 4900

Suitability of WAG-7

The moot question, with which we began, can now be answered. First, it does have the required power rating. But a closer examination of the characteristics of the train and the locomotive is necessary. Tractive-effort speed characteristic curve for WAG-7 (gear ratio 16:65) included in its manual [1] at p-104 (No. DTE/GEL-1074) shows four stages of field weakening (77%, 61%, 50% and 40%). Actually, only three field shunts provided that reduce the field in stages up to 83%, 70% and 58% only. In Fig-3, the suitably modified curve is shown with the train resistance curve for 2480 t CONRAJ train superimposed on it. Stages of field weakening actually provided are also indicated.

TE vs Speed curve for WAG-7 GR 16:65

Fig- 3

It is obvious from the above curve that the starting and continuous tractive effort developed by the locomotive is far in excess of the requirement. Indeed the locomotive never draws a high starting current that is permitted. However, balancing speed on level is 87 Km/h and falls to 77 Km/h on 0.25% rising grade. No wonder, therefore, the repeated observations about speeds achieved discussed earlier. It is also worth noting that the locomotive develops 5000 hp between 44 – 63 Km/h – constant power zone- beyond which the power output falls. That explains the common observation by WR personnel that the locomotive is never able to draw its rated current even though full voltage is applied and field shunted. Thus high gear ratio of the locomotive makes produces much more tractive effort than needed for starting and accelerating the train but restrains it from delivering its rated power when the load demands. This observation should not be surprising, as WAG-7 was not designed for this application in the first place.

What is the solution?

The solution lies in modifying the characteristics of the locomotive to suit the requirements of duty. Let us resolve the apparent paradox of power output first. The power output of the locomotive is sufficient. But it should be available at the range of speed at which it is demanded by the load. From the understanding of train resistance developed earlier, such range is between 65-100 Km/h. The constant power zone can be made to coincide with the desired speed range primarily by changing the gear ratio and can be further fine-tuned by adjusting the number and values of shunting resistors. Fig-4 shows the impact of a change gear ratio. The gear ratio of 21:60 makes the full power available in the speed range of interest i.e. from 62-91 Km/h. In Fig-4, maximum field shunting to the extent of 58% field has been considered. But it can be increased further. Traction motor permits a 40% field. Even if we reduce the field up to 50%, it will be possible to maintain the constant power output up to 100 Km/h.


Comparison for Power of GR 16:65 and 21:60


True, that this will result in a lower tractive effort at starting. But even the reduced tractive effort corresponding to maximum permissible current for HS-15250 traction motor (1350 A) works out to 35 t. This is sufficient to meet the toughest of starting requirements. Starting tractive effort required can be estimated in the same way as that for other goods trains. Even to start the train at 0.8% rising grade, the total TE(s) required is estimated as under:

Starting Resistance (train)    2480*4 = 9920 Kg

Starting Resistance (loco)    123*6=738 Kg

Grade Resistance    1000/125*(2480+123) = 20824Kg

Total Resistance    31482 Kg or say 32 t

Normally a container train should be able to start on 0.67% rising gradient. Even the heaviest train of 3251 t, can be stated with a locomotive developing 35 t Tractive effort at the start.

The locomotive with the above gear ratio will be able to balance the train at a speed of 96 Km/h on level i.e. 8 Km/h higher than the present WAG-7. But what is more important is such locomotive maintains its rated power output down to 62 Km/h. Even on 0.25% rising gradient the balancing speed will still be 85 Km/h; on steeper rises of say 0.5% speed does no fall below 65 Km/h. On the whole, an increase in average speed during a run to the extent of 8-10 Km/h is easily attainable. Fig-5 shows the balancing speed that can be achieved with locomotives having the same HS-15250 traction motor but for different gear ratios. As can be seen that the earlier version of WAG-7 with a gear ratio of 18:64 gives a higher balancing speed. WAP-4 provides a much higher balancing speed of 101 Km/h but may not be able to provide comparable acceleration. Therefore the gear ratio of 2.857 or something close to it seems to be optimum.

Fig – 5!


Remember that the above analysis pertains to a heavy train. Since most trains are lighter, the achievement will be better. As for the few heavier ones, we have taken care of the starting requirement. Some drop-in balancing speed can be tolerated.

Is it worth the effort?

First the need for a higher average speed. Presently average speed is about 43 Km/h. An increase of 10 Km/h in average speed can cut the running time from Vasairoad to Tughlakabad to 28 hours. This coupled with the tackling of few more problems discussed later can bring down the wagon turn-round by 30-40%. This is a splendid achievement as the availability of wagons is the main constraint at the present. Some estimates suggest that at present CONCOR is able to pick up only about 20% of the total unloaded cargo at JNPT. This figure can easily rise up to 60% if CONCOR can provide wagons.

Second, the number of locomotives required? Already about 40 locomotives are being utilized for CONRAJ alone. Besides, there are other container trains such as those linking Kolkata, Chennai and Rajkot ports to various city centers.

As far as the traffic is concerned, even at the present level, this is an important segment distinct from conventional ‘goods traffic’. Containers are carrying about 15 million tons of traffic. Container traffic, therefore, has the potential to become a commercially significant segment by itself for IR and of course the largest segment for WR in a few years. In addition to the international traffic, domestic traffic is also picking up. Table-3 shows the growth in recent years.


Growth of container traffic

Serial No.     Year (From)     Year (To)     International     Domestic     Total     Growth
1 1996 1997 424741 278801 703542
2 1997 1998 491481 230238 721719 25.8%
3 1998 1999 576790 225156 801946 11.1%
4 1999 2000 664491 243330 907821 13.2%
5 2000 2001 755670 291304 1046974 15.3%


Past growth, as well as the current trends in globalisation of the Indian economy, augur well for this segment of traffic. At present availability of suitable wagons, not the availability of traffic is a constraining factor. Therefore, it can be expected that the traffic will treble in next five years. The growth can be even higher. Immediate growth may come on the western route given the location of ports. Development of more ports and linking of all ports by railway lines- a recent initiative- will open a new era for container traffic. In short, this is a sunrise sector of traffic.

Container traffic is here to stay and grow and grow. It has special requirements from locomotives that are not met by any of the present designs. It is time to think of a ‘Container’ locomotive as distinct from conventional ‘goods’ or ‘passenger’ classification. This should not necessarily mean that one is talking of a new to the world product. Indeed a locomotive can be conceived, designed and manufactured using the same assemblies that we are so familiar with.

Locomotive Design

We have already discussed the required tractive – effort –speed characteristics and proposed a method to achieve the same. Let us now examine some important additional issues.

  • It is possible to evolve a design within the conventional tap-changer, DC motor framework.
  • It is not necessary to have a heavy locomotive from an adhesion point of view. So possibilities of reduction in weight should be explored.
  • Since aerodynamic resistance is a key parameter, the possibility of reducing the same by suitable styling of the front of the locomotive needs to be explored.
  • Dynamic braking is required and will be more effective given the characteristics of the train. Dynamic brake presently provided on WAG-7 locomotive may serve well to enable the train to negotiate speed restrictions.
  • Containers are air brake wagons. So only an air brake will be required. Present stock as a single pipe system, however, a twin-pipe system is desirable All modern features like air dryers, unloading of compressors etc. need to be provided.
  • Unstated assumption in the discussion so far has been using of HS-15250 traction motor. This is justified from reliability considerations in addition. However, the option of TAO-659 with tapered roller bearings can be considered. This will have the advantage of reducing the un-sprung mass and overall weight. Of course, the characteristics will have to be worked out again in case such a decision is made.
  • While a high adhesion bogie is not necessary as such, bogie used under WAG-7 is still justified from considerations of reliability, maintainability and manufacturability.
  • The container locomotive will not be required to operate in multiple-formation. Therefore all the related circuitry will not be needed.

This list is not exhaustive. More ideas can be assimilated while designing the locomotive.

These features will not any increase the cost of the locomotive; not even call for any significant change in the manufacturing procedure. But the performance of the locomotive would certainly improve. In fact, such minor variations have been done several times in the past to suitably adjusting the locomotive characteristics for an application even if the number of locomotives required was small. In this case, the volume is likely to be much higher. We must answer one more question.

Whether any of the existing locomotives will do the job?

Well WAP-7 quickly flashes in mind for its characteristic is similar to the one proposed in this paper. Power output is more than required. To that extent, it will be underused. (It is not necessarily a disadvantage.) However, it is comparatively expensive in terms of the initial cost. But it will outperform conventional locomotive in dynamic braking performance including saving of energy and maintenance cost.

Other Issues

During the study, we found that traction change at the Mumbai end is a major factor increasing the transit time. One change is from 25 KV AC to 1500 V DC at Virar, the other from 1500 VDC to Diesel at Jasai, just 7 Km short of JNPT. Operation of the train in the small patch needs attention to improve the turn round of wagons.

Having dedicated locomotives for this important segment of traffic will also help in controlling maintenance problems. During our foot plating, we found several locomotives, in which field shunting/dynamic braking was not operative, hauling the trains.


This study leads to the following conclusions:

  1. The resistance offered by container train is significantly different from that offered by other goods trains particularly at higher speeds primarily due to the shape and sizes of containers. Indeed it varies from train to train.
  2. Characteristics of container train are in the middle of the continuum between conventional ‘goods’ or ‘passenger’ trains in terms of load as well as required speed. Therefore any of the present locomotives would not match perfectly. Characteristics of the WAG-7 locomotive are not ideally suited for the application.
  3. A suitable characteristic can be achieved by changing the gear ratio. In addition, several design features of the locomotive have been proposed.

It is recommended that the characteristics of the train may be verified through a proper trial. Since container trains are all set to multiply in the future, the design of a locomotive that meets the requirements perfectly needs to be evolved. Towards this end, a design has been suggested within the framework of conventional tap-changer technology.


The authors are thankful to Shri Sanjiv Bhutani, Sr DEE/TRO, Mumbai Central, and Shri. Sunil Mathur, Sr. DEE/TRS, Mumbai Central for suggesting this project and rendering necessary assistance. Constant encouragement by Shri MUM Vara Prasada Rao is gratefully acknowledged.



  • CLW, Chittaranjan, Maintenance manual of WAG-7 locomotive.
  • Characteristics of HS-15250 TM (Document no 10T827-668)
  • RDSO, Technical Circular no 27 dated 27.07.1998.
  • CLW, Chitttaranjan, Characteristics of WAP-7 locomotive.
  • Website www.concorindia.com


Analysis of observed data to compute train resistance

Train Load: 2015 t; Locomotive WAG-7 (123 t)

Gradient (+) Up    (-)


Current (A) Voltage (V) Input (hp) Output (hp) Speed (Km/h) Tractive Effort t) Loco Res.(t) Grade Res.(t) Train Res. (t)
1 2 3 4=(2)*(3)*6 5=(4) * 0.9 6 (7)= 8 (9)= (10) =
270*(5)/ (2015+123) (7)-(8)-
(6)*1000 /(1) (9)
200 625 750 3770 3393 60 15.27 0.43 10.69 4.15
260 600 750 3619 3257 65 13.53 0.47 8.22 4.83
420 550 750 3318 2986 70 11.52 0.52 5.09 5.91
900 500 750 3016 2714 75 9.77 0.57 2.38 6.83
2500 500 750 3016 2714 80 9.16 0.62 0.86 7.69
-2500 500 750 3016 2714 85 8.62 0.67 -0.86 8.81


Train Load: 2250 t; Locomotive WAG-7 (123 t)

Gradient (+) Up    (-)


Current (A) Voltage (V) Input (hp) Output (hp) Speed (Km/h) T.E (t) Loco Res.(t) Grade Res.(t) Train Res. (t)
200 725 750 4373.32 3936 65 16.35 0.47 11.87 4.01
300 650 750 3920.91 3529 70 13.61 0.52 7.91 5.18
500 600 750 3619.30 3257 75 11.73 0.57 4.75 6.42
1300 570 750 3438.34 3095 80 10.44 0.62 1.83 8.00
-2000 525 750 3166.89 2850 85 9.05 0.67 -1.19 9.57
-650 500 750 3016.09 2714 90 8.14 0.72 -3.65 11.07
-400 480 750 2895.44 2606 95 7.41 0.78 -5.93 12.56


Train Load: 2480 t ; Locomotive WAG-5A (118.8 t)

Gradient (+) Up    (-)


Current (A) Voltage (V) Input (hp) Output (hp) Speed (Km/h) T.E (t) Loco Res.(t) Grade Res.(t) Train Res. (t)
275 600 660 3184.99 2866 57.5 13.46 0.41 9.47 3.58
375 585 650 3058.31 2752 62.5 11.89 0.45 6.94 4.50
800 475 750 2865.28 2579 70 9.95 0.52 3.25 6.17
Level 425 750 2563.67 2307 77.5 8.04 0.59 -0.33 7.77
-1800 410 750 2473.19 2226 80 7.51 0.62 -1.45 8.34
-650 400 690 2219.84 1998 85 6.35 0.67 -4.00 9.68
-500 385 685 2121.11 1909 87.5 5.89 0.70 -5.21 10.40
-400 350 685 1928.28 1735 90 5.21 0.72 -6.51 10.99
-325 320 685 1763.00 1587 92.5 4.63 0.75 -8.01 11.89
-285 300 680 1640.75 1477 95 4.20 0.78 -9.13 12.55
-250 250 750 1508.04 1357 97.5 3.76 0.81 -10.41 13.36



WR Kota division made the following measurements. SEC TRIALS ON KOTA DIVISION IN JAN 2000

WAG-9 Loco was used in all the trains. All figures are in KWH/1000 GTKM

SECTION Container (1300-1400) t BXN loaded (3300-3400) t BTPN (3900-4000) t Box N (4700-4800) t
Kota-Nagda 23.2 7.2 7.5
Kota-Gangapur 18 9.1 9.0 7.5
Gangapur –Mathura 25 7.39
Kota-Asacuarl 4.2 5.3

The following expression gives the best fit to the train resistance data quoted in this report


  1. A more accurate determination requires a much larger volume of data than available this report
  2. The pronounced component of air resistance is reflected co-efficient of V2.
  3. Higher specific resistance explains higher SEC.

Post a Comment

Your email address will not be published. Required fields are marked *