"CN Eyes "Pipeline on Rail" for Oil Sands Production"

Trains Magazine General Discussion
1

The above is the headline on a short article (apparently dated “09.04.2009”) in the Railway Age “News” section, at:

http://www.railwayage.com//content/view/737/217/

Following are the notable points:

  • CN believes it can be price-competitive with a pipeline;

  • Not enough pipeline capacity presently;

  • Rail can be more flexible in delivery/ sales locations - to many different markets (inlcuding export) than a pipeline;

  • CN will be shipping 10,000 barrels daily by the end of this year.

At 42 gals. per barrel and 23,000 gallons per tank car, that would be 18 to 20 tank cars per day.

I couldn’t find anything directly related to this concept / marketing strategy or “Pipeline on Rail” at CN’s website - esp. the Media section. However, I did find a web page on “Alberta Oil Sands” in the “Shipping - Where You Can Ship - North America” section, at:

http://www.cn.ca/en/shipping-north-america-alberta-oil-sands.htm

It addresses all aspects of possible rail involvement in those projects, but seems more focused on shipping the materials and equipment in rather than the bulk liquid product out.

This reminds me of Trains columnist and author John Kneiling’s comments in the late 1960s and early 1970s - mainly in connection with the then-proposed construction of the 800-mile long Alaska oil pipeline - that when something like this was happening, the railroads usually limited themselves to asking “How do we route the pipe ?” Meaning, that the rails never tried to compete on the long-term mov

2

I believe PC had a “pipeline on rails” program to and from Oswego, NY back in the 70’s or 80’s…I think it was linked to barge from Port of Albany. Cars actually were hooked, or piped, from car to car. See Trains, etc. for the time for picturs, etc.

3

henry6 - Now that you mention it, I recall that one, too.

I’ve done a little more Internet searching on this and found a trio of much better (though similar within themselves) articles / commentary by Diane Francis of the Canadian Financial Post (“National Post”) datelined April 8, 9, and 11, 2009, at:

http://www.financialpost.com/story.html?id=1479094

http://www.financialpost.com/news-sectors/energy/story.html?id=1479470

http://www.financialpost.com/related/links/story.html?id=1486255

From one of the comments to a re-publication of one of these articles in the “Free Republic” (?) at: http://www.freerepublic.com/focus/f-news/2225734/posts

“The rail option also circumvents the problem, for Canadian producers, of reliance on monopoly markets in the United States, and on the fickleness of environmental politics south of the border.”

This is the key to the article. They do not want to rely on the US market because of our lefty loony environmentalists. With rail it is more flexible to ship their product wherever the highest bidder is.

CANADA IS MORE CAPITALISTIC THAN THE US.

  • Paul North.
4

The 1st Financial Post article linked above says that the pipeline charge from Alberta to the U.S. Gulf Coast is $17.95 (Canadian $ ?) per barrel. So let’s work with that number a bit little to see what an equivalent rail rate would be, and to assess the possible competitiveness of an all-rail move:

1 barrel = 138.8 kilograms** x 2.2* = 305.4 lbs. / 2,000 lbs. per ton = 0.153 ton.

    • Approximate or estimated value, subject to correction and refinement, as well as all subsequent figures that depend on it.

(** - For Texas Crude, per WikiAnswers - http://wiki.answers.com/Q/How_much_does_1_barrel_of_crude_oil_weigh - I couldn’t find a comparable figure for oil sands crude very quickly or easily, so I’ll use this one for the time being)

305.4 lbs. per (barrel = 42 gals.) = 7.27 lbs. per gallon (water is about 8.34 lbs. per gallon [“a pound a pint”] ), so this is a relative density of 7.27 / 8.34 = 0.872 - looks OK for now;

$17.95 / (barrel = 0.153 ton) = $117/ ton.

$117/ ton / 2,000* miles = 5.87 cents per ton-mile.

For a 23,000 gal. tank car x 7.27 lbs. per gal. = 167,210 lbs. / 2,000 lbs. per ton = 83.6 tons - OK, well within capacity for a 263,000 lb. gross wt. car = 131.5 tons.

83.6 tons x $117/ ton = $9,780 per carload revenue.

$9,780 per car / 2,000 miles = $4.89/ car-mile.

For a 100 car train, that would be $489 per train-mile; at 20 MPH, that would be $9,780 per train-hour.

Conclusion: Oh, yeah, this pipeline rate is a sitting duck for a well-organized rail move. There’s probably a good amount of "monopoly profi

5

I would hope that some of our Canadian forum members would elaborate but Canada has plenty of “green left” types in government…if power shifts back from the Conservative party to the Liberals there could be changes in Oil Sands production, ect…

6

Some more thinking about this over lunch* leads me to the conclusions that:

(* - as the westbound NS local came by with 5 cars, an SD40-2 still in CR blue on the leading end, and a GP-40 on the training end)

  1. If a significant portion of the potential production here goes by rail, it’ll become close to another Powder River Basin size operation (on the order of 18 trains per day for 1 million barrels per day to the US Gulf Coast, 73 trains per day if all of the potential 4 million barrels per day go there);

  2. The railroad can get set-up to do this for significantly less total capital investment - about only 35 to 65% of what the pipelines are reported to need - and still have all the advantages of flexibility, scalability, timing, etc.;

  3. Study and design, experience, and technical advances will refine these preliminary figures. For example, as in the PRB, the optimum train length is more likely in the 135 to 150 car range - particularly as volume increases and track capacity in terms of number of trains per day becomes more of a constraint. Also, since the tank cars have less of a cross-section and load lighter per linear foot of train length than coal cars (since the oil is less dense), a given payload weight will need a longer train length for oil than for coal;

Where do I sign up to get started on this ?

  • Paul North.

Some supporting calculations:

Consider an initial design train of 100 tank cars of 23,000 gal. capacity each, and 6 locomotives of 4,400 HP each.

Capital cost, per train: About $22 million

Cars

7

Hello Everyone.

Let us start with the above. Most of the players, but not the biggest one, are US companies working in the most conservative province in Canada where resource production is a provincial jurisdiction. So yes they’re off to a good start.

As to your estimate of Oil Sands specific gravity, my gut feeling is your a little high. Lloydminister Heavy Crude has a S.G. of about 10 I think, and Oil Sand would be much less that that. Think thick oatmeal porridge, not cold maple syrup. It has to be processed once before it can be shipped by pipeline, and to save this cost and develop some kind of tank car/hopper hybrid would make this proposal interesting. If you didn’t have to do this first processing, think coal train car weights and handling characteristics.

Now comes the difficult part. There is muskeg between Fort McMurray, AB and the rest of the world. The Alberta & Great Waterways Railway built the line up there in the 1920’s. Waterways was/is a barge loading point for cargo up the Mackenzie River to communities up to the Arctic Ocean, less than 10 miles or so from Fort McMurray. It became the Northern Alberta Railway, then CN, then the Mackenzie Northern (I think) and it is now CN again.

Brief example of muskeg. In a TRAINS or CLASSIC TRAINS article within the last year or so there was a story about a man who took a round trip up there in about 1960. He was in the Passenger Car/Caboose looking back at water covering the rails as the train went forward. Next day they come back and the track is above water as the engine passes and he sees water coming over the rails behind him. They can’t run trains there everyday, because the rails eventually wouldn’t come back above water again. Needless to say this is hard on wood ties, hence the provincial assistance to CN for a

8

Paul: Could I suggest some different values for some of the variables:

  1. Use 286K tankcars.

  2. Maximum speed 50 mph

  3. 14-day cycle time is more realistic, maybe 15 because the loading/unloading time is optimistic

  4. Have to heat the cars to get this stuff to flow – or insulated cars, which add a lot of tare weight

  5. Add $2.5 million/mile for the track and train-control infrastructure required

  6. Add $5 million/mile for the track (total $7.5 million/mile) where there is no track at present

  7. Add $1.5 million/mile for grade separations, quiet zones, line changes to avoid city cores

  8. Add $150 million each for the loading and unloading terminals

  9. Add 15% for design and permitting

RWM

9

All I can say is:

  1. It’s about time…

  2. I own shares in CN…so more power to them!

And CP …I look forward to seeing your own plan for oil in AB.

10

A question and a comment.

Paul - Did I miss how we got from 10,000 bbl/day to 1,000,000 bbl/day?

RWM - I do not think the tare weight penalty for coils and insulation will be excessive. The question as to its value is how much time on steam the insulation would save and the value of that time. My 1979 GATX Tank Car Manual has data for two 20,000 gallon net, nominal 100 ton cars for general service. Uninsulated tare is 57,800 without coils. The implied weight of the coils is 2,000 pounds. The same car with 4" of insulation and heater coils tare is 70,300 pounds. Clearly a car designed for crude would be bigger and the tank size optomized for the specific product. The car would also have to be AAR specification unless the 263,000 gross weight limit for DOT cars has been or would be changed.

Mac

11

AgentKid: How I hate muskeg!!! Since it can vary mile to mile many different solutions are needed. Of course the pipeline would also have to deal with these problems. The problems of having to heat the oil mirror thoses that the Alask pipeline has had to deal with. Either pipeline or RR will have many problems to solve.

12

See 74 FR 1802, Jan. 13, 2009. Weight in excess of 263,000 can be increased to 286,000 so long as quantity of commodity isn’t increased. I think this would apply to the insulation and coils. It does for TIH shields and shells.

RWM

13

Heating the oil so it will flow, I understand. How about churning it or even pumping it from car to car to keep it fluid? Or does it have to be heated?

14

To pump it, you heat it. Otherwise you use a shovel.

AgentKid

15

Paul:

Actually, if anyone was seriously thinking about moving 1mm bbl/day of heavy oil, one would probably say, “let’s build refineries in Edmonton and push product through pipeline instead.” It is highly economically undesirable to move large quantities of low-value unrefined commodities over land for long distances for long periods of time. Even iron ore, which ostensibly moves long distances, is usually only moving half-way to meet the coal, which is also moving half-way.

RWM

16

But on the bright side of things, when the unheated car derails, the oil doesn’t go too far.[;)]

RWM

17

[quote user=“Paul_D_North_Jr”]

The 1st Financial Post article linked above says that the pipeline charge from Alberta to the U.S. Gulf Coast is $17.95 (Canadian $ ?) per barrel. So let’s work with that number a bit little to see what an equivalent rail rate would be, and to assess the possible competitiveness of an all-rail move:

1 barrel = 138.8 kilograms** x 2.2* = 305.4 lbs. / 2,000 lbs. per ton = 0.153 ton.

    • Approximate or estimated value, subject to correction and refinement, as well as all subsequent figures that depend on it.

(** - For Texas Crude, per WikiAnswers - http://wiki.answers.com/Q/How_much_does_1_barrel_of_crude_oil_weigh - I couldn’t find a comparable figure for oil sands crude very quickly or easily, so I’ll use this one for the time being)

305.4 lbs. per (barrel = 42 gals.) = 7.27 lbs. per gallon (water is about 8.34 lbs. per gallon [“a pound a pint”] ), so this is a relative density of 7.27 / 8.34 = 0.872 - looks OK for now;

$17.95 / (barrel = 0.153 ton) = $117/ ton.

$117/ ton / 2,000* miles = 5.87 cents per ton-mile.

For a 23,000 gal. tank car x 7.27 lbs. per gal. = 167,210 lbs. / 2,000 lbs. per ton = 83.6 tons - OK, well within capacity for a 263,000 lb. gross wt. car = 131.5 tons.

83.6 tons x $117/ ton = $9,780 per carload revenue.

$9,780 per car / 2,000 miles = $4.89/ car-mile.

For a 100 car train, that would be $489 per train-mile; at 20 MPH, that would be $9,780 per train-hour.

Conclusion: Oh, yeah, this pipeline rate is a sitting duck for a well-organized rail move. There’s probably a good amount of “monopoly profit” in the pipeline rate - since pipeline capacity is apparently limited and hence constrained and priced higher, so the pipeli

18

Thanks to all for your thoughtful responses, and dialogue. My initial approach to this was, “Do we laugh at it, or take it seriously ?” I was just treating this as a “back-of-the envelope” exercise (I have big envelopes !) merely to see if this concept had any hope of working out, or if it was just wishful thinking - “Will it get off the ground ?” is a common expression that is used - and was only going as far into it as the results seemed to lead me.

It seems everyone thinks this idea has some merit, and concur that this is probably worthy of further study (by CN or someone). I’m seeing points that indicate a need for a 2nd stage of early study and refinement, not “show-stoppers”, and I pretty much agree with all of them. For example, to address just one of the comments (what I think is the most glaring omission in my little exercise) - I was seeing if just the capital cost for the equipment would put this out of the running (it didn’t). As a result, I didn’t take the time to address the infrastructure needs and costs. But even using a “worst-case” scenario of having to upgrade the whole 2,000 route-miles from Edmonton to the U.S. Gulf Coast at RWM’s figure of $2.5 Million per mile would add only $ 5.0 Billion to the $16 Billion for the equipment, for a total of $21 Billion for the rail move, as opposed to $24.7 Billion for the pipeline alternative.

If I have more time tonight or tomorrow morning, I’ll try to respond further.

By the way: The 10,000 bbl. / day is the expectation for the end of 2009; 4 million bbl. / day is the ultimate possibility mentioned in the article. But as an “intermediate” or “unit” figure just to simplify the calculations, I used 1 million bbl./ day, since of course it is simple to scale up or down from a nice round figure such as that to other values that might be more applicable as this goes o

19

20 carloads a day is a nice-to-have for a division’s revenue figures and will hit the radar screen (for a couple of minutes) on the executive floor. 200 carloads a day is an appreciable quantity, but that’s also a threshold level where serious costs start appearing for infrastructure. 20 carloads a day might not be “seen” by the infrastructure for quite a while, but 200 carloads a day will quickly have effects. Generally speaking, if the infrastructure can support a new business with 200 carloads a day during ordinary economic times without a huge cost impact on the infrastructure (on a single-track railroad), someone had really missed the mark on building and maintaining that infrastructure.

RWM

20

Meaning that the infrastructure was seriously overbuilt in some respects (kind of like the railroad verison of that proverbial glass that’s half-full - it’s twice as big as it needs to be !)

That would be yet another illustration of a principle in many parts of life - if it has a name, I have not heard it - to the effect that you can get away with almost anything on a small scale. It either falls between the cracks, no one notices, no one cares, or if they do, they can work around whatever inconvenience or disruption that it may cause, etc. But when that “exception” becomes the norm, then things change drastically - it cannot always be scaled up from a small minority to a large majority of anything without disproportionate effects showing up. This is another example of that - 20 more cars can be fit in without too much trouble, but 200 are going to have real effects.

Except that “It depends” on the “background” pre-existing configuration, condition, and usage of that infrastructure. 200 cars a day is only 2 or 3 decent sized trains. Most of the NS secondaries and mains that I’m familiar with around here could easily handle that with only a pro-rata incre