Yuku free message boards

For probably the first time ever, an attempt has been made to work out the cost incurred by Railways in issuing tickets through its own PRS network and the behavior of queue in the system. This costing has been arrived at on the basis of actual data as far as these were available, extrapolated data from similar situations and averaged data from Annual Statistical Statements 2006-07. For the purposes of study, Church Gate PRS has been taken up and the salary data of the employees available in bill unit 502 (PRS Church Gate) having 80 employees has been considered. One major component of cost which has not been considered here is the cost of land.

Costing has been done for direct cost of manpower, indirect cost of facilities to manpower, direct cost of consumables, tickets, electricity bill, AMC of hardware, AMC of AC plant, cost of setting up of PRS, depreciation cost of assets, opportunity loss cost of lease rentals and the cost of loss of national productivity due to wasted time in queue.

Additionally, comparison has also been done with the data punching speed of Data Entry Operators of EDP Centre and ECRCs of PRS. This has been done with a view to rationalizing the yardstick for ECRCs.

Lastly, quantitative technique of Queuing Theory has been used to find out the waiting time of a person in the queue for Multiple Queue-Multiple Servers Model. And this waiting time has been used for monetizing the loss of national productivity. Additionaly, waiting time of a customer in SIngle Queue-Multi Server Model has been calculated using the same data and it is shown that the waiting time of a customer in single queue, multiple server is much less than multi queue-multi server. This is likely to impact future PRS plan and designing the queue structure adopted there as we are more or less using multiple queue,multiple server model till now.

Based on the cost arrived and comparison done with speed of Data Entry Operators, a list of actionable recommendations have been proposed. The gist of the cost as well as summary is given hereunder

1. Average cost of issuing a ticket is about Rs. 22.

2. As marginal cost is lesser than the average cost, average cost of issuing a ticket should be brought down by increasing marginal utilization. This can be done by keeping PRS centres open 24 hours a day. The booking during night time can be outsourced on per ticket basis so as to give 100% utilisation of PRS assets which is now only 50% utilized being closed from 8 AM to 8 AM.

3. Net booking of tickets needs to be encouraged by passing marginal benefit to the customers by keeping tickets booked through IRCTC slightly less, say Rs. 20, than cost of booking through PRS. This will encourage more traffic to IRCTC and more net booking. Recently IATA has also banned paper tickets for air journey and encouraged only e-tickets.

4. Tickets booked through IRCTC should carry advertisement to compensate for loss of revenue on paper tickets due to advertisement.

5. Additional PRS Centre using railway resources should be opened only in village/rural areas having less internet density so as to fulfill Railways' social obligations.

6. Tie ups with Department of Post should be explored to book tickets through post offices. Progressively, this should also be extended to Banks.

7. Mobile/telephone number of customers should be fed into the system at the time of booking ticket which can be later utilised in case of emergency or for commercial purposes.

8. Present yardstick of booking tickets of about 120 forms per shift should be enhanced to 180 forms.

9. As waiting time of a customer in single queue, multiple server is much less, we should allow that kind of queue formation in PRS so that less customer time is wasted.

The cost of manpower is calculated on actual basis considering the total payment in billunit 502 which is for PRS(CCG). Hence we are treating billunit 502 as ONE individual for the purposes of calculation with necessary modifications whenever needed. Costs are taken on actual basis, as far as possible, on extrapolated basis from similar situations and from averaged basis from statistical statements. The various components of cost are calculated below on actual basis.

1. Salary: Total earnings for employees of bill unit 502 for a period of one year has been calculated here and presented below in a tabular format

An employee can avail of 30 days LAP, 20 days LHAP (monetarily equal to 10 days LAP) and 10 CLs (8 CLs for office staff and 10 for line staff) in a year meaning thereby that an employee can get paid leave of 50 days in a year. So in effect, his above salary of one year i.e. 365 days is basically salary payable for 365-50=315 days. Hence the yearly salary payable becomes 15205279*365/315= Rs. 17618815.35 and per day salary becomes Rs. 48270.73 for billunit 502. Moreover, to give effect to the impact of all leaves due, all the calculations which have been done on the basis of 'year' as a unit would be multiplied by the factor (365/315=1.1587) to incorporate the effect of leave. Thus in the proceeding paragraphs 'leave factor' would refer to this and its value would be 1.1587.

2. Leave Encashment: An employee is eligible for full pensionary benefits on completing 28 years of service. This will include 300 days of Leave encashment. Which means that one year of service ears him 300/28= 10.71 days of full salary. As we have already considered the effect of compounding due to monetization of leave, we need not cover this benefit or else it would lead to duplication.

3. Gratuity: It is 15 days salary for every completed year in service. A person is eligible for full benefit with 28 years' of service due to 5 years' grace. Hence 28 years' of service earns (33/2) months' salary. Thus one year's service earns 33/(2*28) months or (33*30)/56 days' salary= 17.68 days' salary. With the above rate of 48270.73 per day, the gratuity liability for every year of service for billunit 502 becomes Rs. 48270.73 *17.68= Rs. 853426.51. Rationalising it with 'leave factor' it comes to Rs. 853426.51*1.1587= Rs. 988865.29.

However taking a conservative estimate of costing of gratuity as per half months pay for every year of service will bring us to Rs. 48270.73*15= Rs. 724060.95 as the amount payable to employees of billunit 502 for one year of service. Rationalising it with 'leave factor of 1.1587 will make it Rs724060.95*1.1587= Rs. 838969.42.

To back-reference this calculation with the actual gratuity payable, we calculate the per day gratuity as per the maximum eligibility per employee which is Rs. 3.5 Lakhs presently. Hence for 80 employees, the maximum total gratuity liability would be Rs. 350000*80= Rs. 2,80,00,000 for 16 and half months i.e. 495 days. Hence maximum yearly accretion of liability due to gratuity would be Rs. (28000000*15)/495= Rs. 848484.85 and rationalized with 'leave factor' will make it Rs. 983139.4 which is almost equal to the calculation for gratuity taken on the basis of 28 years' service with 5 years' grace period and closer to (but more than) the calculation on the basis of rationalized yearly gratuity accretion on the basis of full gratuity liability on completion of 33 years' of service.

4. Pension: This will also be calculated on the basis of 28 years' of service and 33 years of service respectively. For 28 years, yearly pension liability is 1/28 times (Basic +DP) so that in the instant case the yearly accretion of pension liability at the current rate would be (Rs. 5596398+ Rs. 2808366)/28 = Rs. 300170.14 on which Dearness Relief of 47% is payable. That makes the yearly pension liability accretion of billunit 502 Rs. 300170.14*1.47= Rs. 441250.11. Now rationalized with 'leave factor', the figure of yearly pension liability accretion becomes Rs. 441250.11*1.1587= Rs. 511276.5.

Doing the same calculation with 33 years' service eligibility for full pension brings the calculation to (Rs. 5596398+ Rs. 2808366)/33= Rs. 254689.82 which with 47% Dearness Relief makes it Rs. 374394.03. Rationalising it with 'leave factor', the yearly accretion of pension liability becomes Rs. 374394.03*1.1587= Rs. 433810.36.

5. Pass & PTO: As pass and PTO can be taken for any permissible route which can be long or short, safely and stingily we can consider its monetization to be equal to 5% of gross annual salary. In the instant case of employees of billunit 502, it becomes Rs. 15205279*0.05=Rs. 760263.95.

6. Medical & Staff Welfare Etc.: The medical establishment, clubs etc and the concomitant expenditure is incurred for the welfare of staff and officers which can be safely distributed over the employees to arrive at the average cost per employee. This has been done by taking data for expenditure under Demand number 11 from Indian Railway Annual Statistical Statement 2006-07. Average has been taken for entire Western Railway as well as entire Indian Railway and to err on the side of caution, lower of the two has been taken as average yearly cost per employee. Expenditure on Staff Amenities in 2006-07 in Western Railway was Rs. 1295892000 in Broad Gauge, Rs. 71148000 in Metre Gauge and Rs. 14109000 in narrow gauge i.e. Rs. 1381149000 across all gauges whereas for the entire Indian Railways it was Rs. 16,65,17,97,000 in 2006-07 across all gauges. (Page 22/23 of Annual Statistical Statement 2006-07).The manpower statistics as per the Statistical Statement is 1,06,301 for WR and 13,17,306 for Indian Railway in 2006-07 (Page 427). Thus the average per employee yearly cost for Staff welfare activity comes to Rs. 1381149000/106301= Rs. 12992.81 for Western Railway and Rs. 16651797000/1317306= Rs. 12640.80. Hence we take the lower of the two i.e. Rs. 12640.80 as the yearly average staff welfare cost. Thus for billunit 502 having 80 employees, the yearly medical, welfare etc. cost becomes Rs. 12640.80*80= Rs. 1011263.71.

7. Cost of Hiring PRS Channel This cost is Rs. 64000 per annum for Church Gate PRS.

8. Cost of computer hardware etc: We can safely keep as Rs. 500000 in absence of original file. (Roughly this was the cost of computer hardware at Dadar PRS.) With the life of these assets as five years, we have the annual cost of Rs. 100000 here. Dividend liability of 7% on original cost is Rs. 35000. So annual cost is Rs. 135000.

9. Cost of AMC of computers, printers etc.: For 32 terminals, 27 printers and one dotmatrix printer, the cost of AMC per year as per existing contract is Rs. 33965 for PRS/CCG.

10. Cost of Ticket & forms In last one year 1576970 tickets have been booked from this CCG PRS and 259023 tickets have been cancelled which means that a total of 1835993 tickets have been utilized. The cost of one ticket roll is 490 per 1000. So this cost for the year comes to Rs. 899637.57.

11. Imprest expenses This expenditure for the year is Rs. 4804.

12. Electricity Bill: The average daily consumption of electricity in that PRS centre having 60 Ton AC is about 660 units. The yearly consumption at the rate of Rs. 8 per unit comes to Rs. 660*365*8= Rs. 1927200.

13. AC Installation: The installation of 2 AC systems of 30 T each was done in 1987 at the cost of Rs. 1900000 each i.e. Rs. 3800000. This system is now being replaced after 20 years and the file has been initiated. Hence the yearly depreciation cost becomes Rs. 3800000/20= Rs. 190000. Dividend cost at 7% on original cost is Rs. 266000. Thus total cost here is Rs. 456000.

14. AC AMC cost: The present AC system is under AMC with the party whose annual cost is Rs. 155940 less 40% = Rs. 93564 and cost of 4 supervisors given by the party for Rs. 259296 less 40%= Rs. 155577.6. Thus annual maintenance cost of AC of PRS/CCG is Rs. 93564+Rs. 155577.6=Rs. 249141.6.

15. Diesel Generator: One diesel generator was procured as standby at Rs. 1041000. Life of power generation machinery as per Finance Code,Volume 1 is 15+ years. Taking it to be 15 years, the annual depreciation cost becomes Rs. 1041000/15= Rs. 69400. Dividend liability at 7% is Rs. 72870. Thus total yearly cost is Rs. 142270.

16. AC AMC Supervisor: One railway staff of grade-I helper category is on duty specifically for that. The average annual wage per group C employee of Western Railway including his pension, gratuity as per Annual Statistical Statement of 2006-07 is Rs. 171959. (Page 407 of report).Adding cost of staff welfare as per para 6 above, we come to Rs. 171959+Rs. 12640.80= Rs.184599.80. Using 'leave factor' of 1.1587 makes it 213895.79.

17. Cost of Setting Up of PRS: As the original file of CCG PRS could not be located, we are extrapolating that cost from the cost of Dadar PRS on the basis of per unit cost incurred in Dadar. Expenditure on Dadar PRS is Rs. 35719967 for an area of 2868.12 sqm. Thus the cost per sqm becomes Rs. 35719967/2868.12= Rs. 12454.14. Using this figure, the cost of setting up of CCG PRS of 604.64 sqm becomes Rs. 12454.14*604.64= Rs. 7530271.21. The yearly cost of this money on the basis of 7% dividend liability is Rs. 527118.98. Moreover the life of civil engineering asset as per Finance code is 65 years so that yearly cost of depreciation becomes Rs. 7530271.21/65= Rs. 115850.33. Hence total yearly cost for this asset is Rs. 527118.98+Rs. 115850.33= Rs. 642969.31.

18. Opportunity Loss cost: As per the land licensing policy of the Railways, we charge 10% of the cost of the land per annum as land license fee. If the same space of PRS centre was licensed to some party we could have got 10% of the cost of the land as license fee. This is one major income which we are foregoing and will come as opportunity loss cost. The market cost of commercial property in this area is from Rs. 15000-22000 per sqft as per the website mumbaipropertyexchange.com. With 1sqm=10.8 sqft, at the middle of the range (Rs. 20000 per sqft) the market cost becomes Rs. 20000*10.8*604.64= Rs. 130602240. Taking 10% of this as the yearly opportunity loss cost due to lease rentals, we come to a cost of Rs. 13060224.

19. Cost of national Productivity loss: This is one of the most important hidden costs which is not always accounted for due to various reasons. When the social commitments are harped upon, social costs should also be highlighted and it stands to reason that our investments should pass the litmus test of Social Cost benefit Analysis (SCBA). An average waiting time in queue as well as while being serviced is 21.43 as has been calculated below. Monetising this wasted time of 21.43 per person, it is seen that the Nation has suffered a calamitous productivity loss of 1835993*21.43 minutes= 39345330 minutes= 655755.5 hours= 27323.15 man days. So in one year, CCG PRS leads to a wastage of about 74.86 man years just in queue! Iterate it over the entire country and all the activities where there are lengthy queues and one can have a fare idea of why this nation ranks low on Per Capita Income and HDI! Anyway, with the latest data of GNP of 44,65 lakh crores (GDP is about 47 lakh crores) and population of 112 crores, per capita income becomes GNP/population= Rs. 39866.07. Therefore, the yearly national productivity loss due to wastage of time in queue in CCG PRS is Rs. 39866.07 * 17.47= Rs. 2984291.63.

20. Costs not considered: Cost of land, fractional cost of officer apportioned to this PRS centre, furniture, benefits of leave like paternity, maternity, various subsidized loans, cost of maintenance of diesel generator, cost of diesel, miscellaneous pensioners' allowance, zonal work order cost on PRS etc. have not been considered here.

21. Earnings due to advertisements on tickets: As the loss of advertisement earnings on these tickets are likely to be offset by having advertisements on paper tickets to be booked through IRCTC, we are not considering the beneficial effect of advertisement due to compensatory nature.

22. Number of tickets printed: Total tickets booked in last one year for the period of study is 1576970 booked and 259023 cancelled. So total tickets printed is 1835993 in one year.

The above items calculated in item 1 to 19 gives a comprehensive picture of costs incurred on CCG PRS in one year. This cost comes to Rs. 41703883.7 considering full pensionary liability on 28 years and Rs. 41476521.69 considering full pensionary benefits on completion of 33 years.

This brings us to the per unit cost of printing a ticket through CCG PRS. At the lower end, it is Rs. 22.59 and at the higher end it is Rs. 22.71 per ticket booked or cancelled.

The above calculation is under an overall caveat that the major cost of land has not been considered here. If the market cost of land on which PRS is set up is considered, the cost of ticketing would increase exponentially.

This comparison will help us evaluate the relative speed of data entry as well as help rationalize the set yardsticks so that it can contribute to overall enhancement of productivity. Comparison with Data Entry Operators of EDP Centre is done as the both of them do data entry of random digits and characters. However there is a caveat: data entry by ECRCs have financial value and as such their speed would be slightly less than DEOs of EDP Centre whose output is subject to ARDs. This comparison is being to reducing their work to the lowest common denominator called 'key depressions' which means the number of time the keys are pressed. This 'key depression' is further irreducible quantity.

Yardstick for an ECRC is 120 forms per 6 hours of shift meaning thereby that they have to service a customer in 3 minutes. Now let us examine the components and consequent key depressions (KD) of booking a ticket:

1. Train No. : 4 KD

2. Date : 4 KD

3. Station : 2-4 KD

4. Class : 2 KD

5. Destination : 2-4 KD

6. Reservation upto : 2-4 KD (hardly used)

7. Boarding : 2-4 KD (not much used)

8. Name : 3-15 KD

9. Sex : 1 KD

10. Age : 2 KD

11. Berth choice : 1 KD

12. Sr. Citizen : 1 KD

13. Concession Code : 6 KD (only for pass holder)

14. Concession Station : 4 KD (only for pass holder)

15. Concession Number : 4-10 KD (only for pass holder)

16. Class Appeal : 2 KD (only for pass holder)

17. System Command : 1 KD

18. Tatkal : 1 KD

19. Upgradation : 1 KD

20. Print Command : 2 KD

As the study is being done for CCG PRS where pass-holders do not go owing to the fact that there is a dedicated PRS terminal in CCG itself only for pass-holders hence we can safely disregard the key depressions pertaining to the pass details (items 13 to 16). Hence we see that at the highest level, the number of key depressions comes to 51 including item number 6 and 7. Apart from this, ECRCs spend time in collecting forms, collecting cash, counting cash, organizing refund and printing tickets. Even if 50% of the assigned time (i.e. 1.5 minutes out of the assigned 3 minutes) is spent in non-punching activities, then we see that in 1.5 minutes the ECRC punches money value data equivalent to 51 key depressions bringing their rate to 34 key depressions per minute. And point to note here is that out of this, item numbers 6 and 7 are most likely not used and item number 8 for Name is not used for 15 characters. So one may as well be punching only 38 characters per form having one passenger. But if number of passengers per form increases to 6, the number of key depressions will increase that much.

Let us contrast it with the yardstick set by Railway Board for Data Entry Operators which is 56000 key depressions in 7 hours i.e. 133 key depressions per minute! And the data punched at both places are random data. So even if due to money value transactions, ECRCs have to careful, there certainly is a case for rationalizing their yardstick and revising it upwards to 100 key depressions per minute. This will bring their service rate to about 30 per hour from present 20 per hour.

Calculation of Arrival Rate and Disposal Rate and Waiting Time in System

It is seen that in the probabilistic queuing models, customer arrivals follow Poisson distribution whereas the service times are exponentially distributed. For this type of queues, the arrival rate should be less than the service/disposal rate for the system to be viable or else service facility will always be busy.

The number of counter in operation here are 18. Actual service rate can be calculated by dividing the number of booked tickets by number of service counters and number of service hours in the year. Here we have considered changeover gap of half an hour i.e. total working hours of 11.5 hours per day so as to arrive at a realistic calculation. Thus we get actual service rate as 1835993/(18*11.5*365)=24.3. Considering that customers start assembling just one hour prior to booking of the ticket, one comes to an arrival rate of 1835993/(365*13*18)= 21.5. In fact in peak season one can see people sleeping in queues!

Thus the average time a customer waits in the system (time in queue plus servicing time)

Waiting time in system(Ws)= 1/(service rate-arrival rate)

Thus we see that the waiting time of a customer in our PRS system in this kind of multi queue and multi servers is 1/(24.3-21.5) hours which is 21.43 minutes. This is the average time a customer spends in the system which has been quantified as loss of national productivity in item number 19 above.

Using the same data, it can be shown that the expected time a customer waits in the system with the above kind of data is about 3.0833 mnutes and thus this mode of queuing in PRS should be preferred.

Inference and suggestions:

1. Railways will be probably the only organisation in the whole world which has invested so much money in computerisation as well as web based ticket booking through IRCTC, Yet it is neither utilizing the full commercial benefit of that nor is it passing on that benefit to the customers. And a screaming example of that is that net booking of tickets through IRCTC is closed in the nights when it should be kept open 24 hours a day, save for routine maintenance of servers by CRIS. No commercial organization of the world having a web portal for its commercial activities keeps the portal closed 'during nights' and in doing so underutilizes its own assets! IRCTC needs to be given full access to PRS servers 24 hours a day, 365 days a week. It needs to be kept in mind that night in India does not mean night all over the world. And thus future tourists from foreign countries as well as local Indians might want to utilize these services at night. This is one decision which needs no investment, has no adverse monetary impact and thus it should be set in motion as soon as possible.

2. As marginal cost is always lesser than average cost, we need to reduce our average cost by increasing the marginal utilization. For this we need to increase the productivity of our PRS assets by utilizing it full time. As of now PRS centre is being utilized at 50% of its capacity as it is closed between 8 to 8 leading to wasteful idling of assets. Hence Railways should explore the possibility of (floating an open tender for) issuing tickets during this period of 8 to 8 through private operators by paying them on per ticket basis. This will not only lead to lesser queues, but also 100% asset utilization as against 50% now and reduced average cost per ticket printed. For example, if the private party charges Rs. 5 per ticket and if 20 lakhs tickets are booked in a year during 8 to 8 (private party is likely to be more efficient), then Railways would pay that party Rs. 100 lakhs. This brings down the average cost of booking a ticket to Rs. (10000000+41703883.7/(2000000+1835993) =Rs. 13.48 at higher side and Rs. (10000000+41476521.69/(2000000+1835993)= Rs. 13.42 at lower side. Thus we see that our average cost per ticket reduces drastically, it achieves better asset utilization would lead to lesser queues and lesser national wastage due to loss of public time. Slowly, 24 hours PRS centre should be enabled all over the country with the night time being manned by outsourcing this work.

3. As the cost of issuing ticket through Railways own set up is a minimum of Rs. 20, the tendency to open more and more PRS by investing Railways' money needs to be curbed and internet booking encouraged the way IATA has banned issue of paper tickets for airline industry from 01.06.2008.

4. Internet booking can be encouraged by passing on the saving on the cost of booking ticket from PRS to customers. Thus Railways should start giving a marginal discount of Rs. 20 on every ticket booked through internet/IRCTC as one less booking through PRS centre means Rs. 20 saved by Railways on the above cost. Seen from the management perspective from the highest level, this is not likely to lead to any reduction in earnings as the expenditure will also be simultaneously reducing. This may be termed as Avoidance of Anticipated Expenditure. Equivalently and alternatively, Railways can also think of charging Rs. 20 more on tickets booked or cancelled through PRS so that a differential pricing is maintained to reflect our cost of servicing customers.

5. Tickets booked through IRCTC and paper tickets printed through the website should necessarily have advertisement which would open one more channel of advertisement revenue.

6. The advertisement cost on IRCTC website should be rationalized and pegged much higher as marginally reduced fare through IRCTC will lead to massive increase in traffic to this site, as and when and if it happens.

7. As social obligations and responsibilities of Indian Railways can never be overstressed, Railways should concentrate on opening more PRS centres in villages and underdeveloped areas and certainly not in urban areas having better net penetration, internet cafes and existing PRS at adjacent locations. After all banks do not open their branches at every nook and corners!

8. To penetrate villages, Railways should tie up with Department of Post on mutually agreeable terms and provide a computer, printer and wireless data card (along with monthly consumables cost) to every Post Master along with authorization to print tickets from IRCTC site. The paper printout of ticket should be stamped with Post Master's seal so that the need for the passenger to carry identity card is obviated. This will lead to PRS Centre reaching nook and corner of the country. A personal example would suffice. In my village, the nearest PRS centre is in the district headquarters which takes Rs. 25 to reach by bus and then rickshaw. Thus a person is spending Rs. 50 to book a ticket. If Post Master can issue the same for Rs. 10 extra per ticket, it would still be beneficial for the customer and win-win situation for both and Railways would have brought these services at the doorsteps of customers.

9. The same should progressively be done with State Bank of India and later with all other banks. This way number of PRS will increase substantially without investing Railways own resources.

10. The data punching form on the computer should have a column for punching the contact details of the customer. These 10 extra key depressions are not going to take any substantial time but would come in handy during the time of any emergency to locate the passengers or their families as against the present system or locating the PRS counter and then the forms of that counter. This can also help us in updating passengers on late arrival of trains or ticket reservation status or for commercial purposes for staring frequent travelers' scheme etc.

11. Yardstick for ECRCs can be rationalized and revised upwards to 30 forms per hour i.e. 180 forms per shift.

12. Single queue, multiple server queue should be allowed to be formed for faster clearance and lesser waiting time.

I invite the fraternity to take the pain of critically evaluating it and offer their suggestions.

scmitra