Quote:
Originally Posted by Dado
That's $172 per hour per 3-car TRAIN* - NOT per single LRT car. That's the big mistake that the consultants made in estimating future operating costs for LRT. It's a great big gaping mistake so large the O-Train could be driven through it because it overstates LRT operating costs by more than threefold [the $172 has 4 basic cost components: 3 identical ones for each car (mainly electricity and maintenance) and a per train component (basically the operator) so using the $172 as it was used isn't just 3 times wrong it's more wrong still because it's also adding in the full per train component to each car -- so taking account of this last point and calculating for only 3-car trains it means that the $434M annual operating cost of Option #4 is in fact below $380M compared to the $485M of Option #1 - nothing like a $50M+ error on something as insignificant as annual operating costs...]. The page you cite above is pretty clear on the matter of what the $172/hour represents: back when it was written they said it cost $163 per hour to operate LRT and then they go on to say in the next sentence that "With an average of 600 boarding passengers per operating hour the average cost per LRT passenger is only $0.27 ($0.23 USD)." The stat of 600 boarding passengers can only come from an entire 3-car train (there's little en-route passenger turnover in Calgary) and one can only get a per passenger cost of $0.27 from the given data by dividing 600 into $163 [($163/hour) / (600 passengers/hour) = $0.27/passenger - the 'hour' terms drop out of the division by cancellation]. Of course that low per passenger cost only occurs at peak periods but the point of the foregoing was to show that the $163/$172 represents the per train and not the per car cost.
They then go on to say: "In comparison, the average for bus passenger boardings is approximately $1.50 ($1.28 USD) or almost 6 times the cost of
carrying an LRT passenger."
Ottawa's average annual per passenger cost was $2.69 in 2007. I would guess that the $1.50 figure for Calgary assumes full buses (makes sense to compare to a full CTrain) because Calgary's systemwide average annual per passenger cost was $2.26 in 2005 (Ottawa's was $2.54 that year). In the time since 2005 Calgary has had significant ridership gains so it's per passenger cost almost certainly went down whereas ours continued to go up even as ridership increased.
Finally, they conclude that paragraph with these words of wisdom: "Of course buses have considerably lower capital cost and have different capabilities."
And that is our problem. Ottawa seems to think that buses can do everything everywhere and always rather than recognizing the different capabilities and uses of different modes. The very existence of Options #1 and #2 in the TMP update presented earnestly in the documentation as if they were viable (14 second headways vs 16 second headways! yippee!) was a clear demonstration of that thought process.
*There is a discrepancy between the $163/hour figure in the Calgary Transit document (along with the $172/hour figure used by Delcan) and the calculated $257/hour figure using data from the Calgary Transit website. If one takes only marginal costs (vehicle maintenance, power, wages) from the website into the calculation then the cost would be $194/hour - closer, but not close enough. I can't resolve this discrepancy to any satisfaction - but whichever is the case it is still a per *TRAIN* cost, not a per LRT car cost and therefore the conclusion that the City's consultants have vastly overestimated LRT operating costs still remains.
To follow on from the chart in the Appleby's post using the derived figures:
1-car train: $159/hour
2-car train: $208/hour
3-car train: $257/hour
So, only against a 1-car train do buses stand a chance of competing. Against even a 2-car train the train comes out ahead since $208/hour is close to the cost of operating two standard buses but the capacity of 2-car train is much greater than that of two buses. And, indeed, the background documents of the N-S LRT business case came to the conclusion that a 2-car train had an operating cost of 83% of that of an equivalent bus-based service. As labour costs rise (for operators and mechanics) and since the cost of diesel will rise more than that of electricity, these numbers will increasingly favour rail as time goes on.
|
I find it hard to believe that the cost is $172 per train-hour. That works out to $57 per car-hour. The cost per car-hour for the Skytrain in Vancouver, an automated system with smaller vehicles, is $85 (the Skytrain has total operating costs of $83M per year / 940,000 car-hours per year / boardings of 73,000,000 per year / 78 boardings per car-hour). Even the $257 per train-hour estimate seems low, since it would imply the same car-hour costs as the Skytrain. I know Translink is not the most efficient organization in the world but still.
I’ve nonetheless used Appleby’s figures to see what the results would be. Under all scenarios, peak bus costs were reduced by 10% to take into account the use of “trunk expresses”. This gave an annual cost for the bus option of $99M (this total reflects the costs within the Blair-Baseline and N/S corridors, and thus excludes the cost of the rest of the network – these costs should be fairly similar under both the bus and LRT options).
Scenario 1: Under the first scenario, I assumed that all trains (i.e., both peak and off-peak) on the East-West Line would be 4-cars long while all those on the N/S line would be 2-cars long. The results were as follows:
• Scenario 1a: With the amount of annual LRT service assumed in the MRC/Delcan report (474,000 car-hours), annual operating costs would be $38M. Thus, this yielded annual savings of $61M. This would result in off-peak headways that are 6.7 times higher than for bus.
• Scenario 1b: If off-peak service levels were doubled, the amount of annual LRT service would grow to 692,000 car-hours and annual operating costs would amount to $56M. Thus, annual savings would be $44M. This would result in off-peak headways that are 3.4 times higher than bus.
• Scenario 1c: If off-peak service levels were tripled, the amount of annual LRT service would grow to 910,000 car-hours and annual operating costs would amount to $73M. Thus, the annual savings would be $26M. This would result in off-peak headways that are 2.2 times higher than bus.
Scenario 2: Under the second scenario, I assumed that peak train lengths would be 4-cars on the East-West Line and 2-cars on the N/S line, while off-peak service would be provided by 2-car and 1-car trains, respectively. The total amount of annual LRT service (measured in car-hours, not train-hours) would remain the same as in the MRC/Delcan report (i.e., 474,000 car-hours). This would have annual operating costs of $45M or $54M less than the bus option. Off-peak frequencies would be 3.4 times higher than the bus option.
Scenario 3: Under the third scenario, I assumed that peak train lengths would be 4-cars on the East-West Line and 2-cars on the N/S line, while off-peak service would be provided by 1-car trains on both lines. The total amount of annual LRT service (measured in car-hours, not train-hours) would remain the same as in the MRC/Delcan report (i.e., 474,000 car-hours). This would have annual operating costs of $55M or $44M less than the bus option. Off-peak frequencies would be 1.9 times higher than the bus option.
Scenario 4: The fourth scenario is the same as the third one, except that off-peak frequencies for the LRT would be the same as for the bus option. The total amount of annual LRT service (measured in car-hours, not train-hours) would grow to 676,000 car-hours. This would have annual operating costs of $87M or $12M less than the bus option.
If Appleby’s cost figures are reflective of what can be achieved in Ottawa, then I will admit that I was wrong in my previous post. An LRT option will have lower operating costs than a “best bus” option under all scenarios. If ridership continued to increase, and peak East-West trains became 6-car trains and the peak N/S trains were truncated at Bayview, the savings would grow over time.
In terms of which scenario would be the likeliest, I’m not sure. Scenario 1a is really terrible, but it would be attractive to a transit authority that is fixated on minimizing operating costs. I also think that pursuing scenario 1b or scenario 2 would still not be acceptable and lead to pressure to go with a scenario 1c or a scenario 3. Scenario 3 would much more attractive from a financial perspective, but for some reason Calgary does not pursue such a strategy (i.e., running shorter, more frequent trains during the off-peak period). There must be a reason why. Sadly, I don’t think that scenario 4 is likely at all.
Irrespective of the scenario, one would need to compare the savings in operating costs of the LRT option with the increase in user costs due to longer waits, slower travel times and the increase in transfers. A back-of-the-envelope calculation suggests the increase in user costs due longer waits and slower travel times would cancel at least half of the savings under scenario 3. Differences in capital costs and external benefits (e.g., the bus option should have higher ridership because of lower in-vehicle and out-of-vehicle travel times) would also need to be considered.
Prev
Linear Mode
