For containers headed to the US east coast from the Pacific, is there any advantage of using this route and incurring two extra mode switches (sea->rail->sea), over going to a US west coast port and just incurring one (and possibly zero) extra mode switches?
Given that the containers may be distributed in a variety of modes when they are unloaded at a land port (e.g. different trains, trucks, etc), it's probably reasonable to think about fractional mode switches. I.e. what's the average number of switches across all containers on the ship. Using a rail portage to get to from the Pacific to the Atlantic will add 2 to that number for every container, whereas going to a west coast port might add one or maybe zero (I.e. switch to a truck that drives cross country, vs rail->truck)
Is the increased cost of transcontinental land transport in the US so high as to make this scheme worthwhile for US East coast?
> But projects like CIIT may still struggle to entice cargo away from the Panama Canal. The largest vessels that go through it can carry 14,000 containers. Mexico’s government accurately reckons that the coast-to-coast rail journey will be quicker than passing through the canal. But it neglects to mention that the trains’ capacity and the speed at which they can be loaded and unloaded mean that the overall rate of goods’ transit between the two oceans will be much slower than the canal.
> Moreover, Niels Rasmussen, chief shipping analyst at Bimco, an industry association, says that carrying cargo by train or road has big snags. Most shippers would prefer to rack up extra miles on other maritime routes than to deal with the hassle of unloading and reloading. And if push comes to shove, many would probably prefer existing routes across the United States to untested road alternatives in Latin America.
So clearly they don’t know what they are talking about. Any sailor would tell you, you don’t go above and below 55 degrees. Granted these are steel container ships but the seas don’t care. What’s one or two lost containers of goods? They sit just below the water surface for the next vessel to crash into.
The Panama Canal has two paths. Only one can be used for container ships. Even that has been at capacity for quite some time. A more integrated shipping supply chain is needed to get goods from A to Z. Rail, truck, boat, and plane.
> The Panama Canal has two paths. Only one can be used for container ships.
I believe both Panamax and Neopanamax lanes can be used by most container ships. The Neopanamax lane can accommodate the biggest new container ships in addition to that.
Only because of the North Atlantic Gulf Stream bringing warmer waters. Everywhere else around there is closed in the winter due to ice. Murmansk is a freak of nature, an exception to the rule, like how you can’t freeze vodka in a normal freezer.
Even with global warming you cannot expect the north around Canada path to be open every summer. It will be closed most of the year even in the best case. Maybe you can attempt it some years but it isn't a reliable route worth considering.
Sea transport is cheap but slow. Mexico is much narrower than the US so they can potentially be faster than the Canal without adding much more costs. Going LA to the east cost is done, but has more expensive land transport. So this could be a useful compromise. If (big if that I have no confidence in, but still possible) Mexico puts in great sorting they can also take containers from on ship and sort them into several others to various ports on the east coast.
Is it worth it? Hard to say. Sea transport is cheap, but not free. There is a lot of opportunity to reduce the costs of move switches by automation. However thee biggest advantage to this really only an come if they can sort containers to different destinations.
I'm no engineer, but I would not imagine that the Panamax ships are built to handle that kinda of stress while still loaded with cargo. The ships are designed to have the weight of itself and its cargo spread out pushing out against the displaced water pushing back on the hull rather evenly and not a handful of highly concentrated points.
Imagine the hilarity if the giant bathtub train car derails and tips over. Has there ever been a train wreck that caused a flood? If not, there's a first time for everything.
There's also the possibility that you get 100 km inland and then the bathtub springs a leak. Not a good situation if the ship's structural integrity depends on the water in the tub.
Hard enough to snake one rail line through mountainous terrain let alone 8. I can see them building a large railyard in Mexico that sorts incoming containers into separate trains depending on which ship they are destined for.
Building the drydocks capable of submerging the rail cars and positioning the ship above them and then hauling the ship out would be the bigger challenge. Dealing with any slope with that much weight is tricky.
I’m pretty sure once you cede that you’re willing to use the energy required to move a container ships worth of weight from one end of the canal to the other, the cheaper option would be to pump water to refill the locks. Water is easier to transport than ships, and while I don’t think the efficiency of the locks helps much in this context, there might be some clever optimizations… especially if you’re willing to run a ton of high pressure pipe.
Pump it from where though? It would be easy to pump and reuse the water that already goes through the locks, except that it gets mixed with salt water, so putting it back in the lake would quickly contaminate it.
Based on the diagram of the locks and water saving basins at https://en.wikipedia.org/wiki/Panama_Canal_expansion_project... they could have a second basin D at the same level as basin C. They would pump the fresh water in C that has come down the lock chain back up into the lake. They could pump sea water into D that would be used to fill the bottom lock when a ship is acsending the chain. Then the sea water in the bottom lock would be drained to the sea when a ship descends.
Pumping sea water to the top of the lock chain or maybe the lock next to the top would save a lot of fresh water but would slowly contaminate the lake unless there were some kind of double gate to prevent sea water used to fill the top lock leaking out when a ship enters the lock from the lake.
> what if we built say ten parallel railways 10 ft apart and moved cargo ships over the rails?
Interesting concept. Never thought of it before.
The India-Middle East-Europe Economic Corridor (IMEC) has a similar issue. Containers will be offloaded in UAE, then travel via rail through Saudi Arabia and Israel, and then reloaded on to ships in the Mediterranean for a short trip to Greece.
It's a dumb idea, we already have standardized cargo containers and the technology to load/unload them efficiently. Putting a ship on 10 Train Tracks will require the development of new technology and will make every bridge (weight allowance) and tunnel (the profile would need to be big enough for the whole ship) prohibitively expensive. With normal cargo trains, which are a very efficient and cheap way to transport a lot of goods, you can just use cheap off the shelf stuff.
Also transporting ships with their huge tanks of heavy oil is a huge health and safety hazard.
Good reply. You're right, I hadn't thought about bridges / tunnels. Nor about the fact that the current system is basically "cheap off the shelf stuff".
This exchange has been a revelation in systems-design for me. The value one gets from adding another brain with a differing viewpoint.
Something similar to this idea exists in Russia. But for larger river ships only, so smaller scale.
There is a large dam, with generating station, downstream the valley is narrow, walls are steep. Differential to top of dam is large/high. Water locks impractical.
So they've built something like a dock for the ships to enter, dock closes with ship inside, and is moved on cog rails up/downhill for a few miles, including a turntable!
By electromotors, powered by > 100.000V three-phase delivered via catenary from the sides!
At the time this made the rounds trough the net (possibly +20 years) it all looked rather insane/gigantic/impressive.
Wasted 5 minutes trying to find it, seems to be gone. shrug
This is Krasnoyarsk dam. It is located in small town of Divnogorsk some 40km up the Yenisei river. I was raised in Krasnoyarsk and have seen this with my own eyes.
Weight wise it looks feasible. The hard part is getting something capable of transferring hundreds of tons that can also conform to hulls of different shapes.
If weight is ok, just put a lock on a rail: so boat enters the lock, two gates close behind it, the lock (and one gate) moves via a rail to the other side, reverse at the other ocean.
Triremes had to be regularly lifted out of the water or they would become waterlogged. They were surprisingly light weight, and could actually be carried onto a beach by their crew.
The Ever Given has a maximum speed of 23kt / 42km/h and carries 20,124 TEU or approx. 7660 54-ft containers.
Rail figures are harder to find, but a reasonable guess based on what I can find seems to be 400 containers per train if the loading gauge permits double stacking containers. Speeds over 100km/h are easily achievable on railways without at-grade crossings.
So one ship could easily require ten trains to move all its cargo in the same time, plus the time taken to offload at each end, which I guess is why this hasn't taken off. Though if it's all just transhipped to trains this could be faster than at most ports where containers have to be sorted to different destinations.
Though interestingly with a dedicated right-of-way the trains don't actually need to be manned, remote control freight trains exist.
I could foresee a parallel loading crane to train track system where the cranes take up to 10 or 20 containers at once and then put them on 10 to 20 parallel train track at once on the other side.
Hand counting how wide a panamax gets, they seem to max out at about 19 or 20 containers wide and 15 containers deep. It also looks like shipping containers already can be top latched with this video: https://www.youtube.com/watch?v=GZQPxl9zssk
So if you could automate this system at 30m per transfer of 10, a port with 15 cranes could do 300 containers an hour and clear a boat in about 15 hours. You could expand the system further into a full panamax width and do it in about 8 hours per panamax boat with 4500 TEUs.
Freight trains run usually at 80-120km/h on railways with at-grade-crossings (for higher speeds you would need better brakes, and more energy which doesn't really make sense for freight). At least in Europe you only need to eliminate at-grade-crossings if you want to run trains over 160km/h
Because it's stupidly expensive boring a Tunnel when you can just do it on the surface.
Also Railway lines take up considerably less space than every Street. Weather also rarely impacts rail
Maybe on the surface (pun intended) it appears stupidly expensive in comparison, but I bet if you actually put a value to all of the positives then the cost would at minimum be at par.
With tunnels you can also go 3D, you could go 100 tunnels deep - stacked - increasing capacity on the same line as needed; something Elon points out as part of the traffic congestion problem is that cities went 3D (skyscrapers) but traffic did not.
You build that route through nothing. There is no need for tunnels, and I don't see any positives, that you couldn't remediate much cheaper. For animal movement you can just build huge animal bridges every few hundred meters. Weather is rarely a problem for train operations, tunnels create also a whole lot of new "Negatives". You need evacuation systems, ventilation systems, maintenance etc...
I really would love to see an effort, perhaps best spearheaded by Boring Company themselves, to do a very thorough "feature matrix" of pros and cons and real-life current costs of things depending on the area, e.g. snow removal costs, how much building "huge animal bridges" that allows ecosystem to not be too disrupted, etc.
It's such a complex issue that a fully "complete" overview is necessary, that can be then dove into over weeks or months of time to first study its thoroughness to then be able to decide how accurate or credible it is, and add to or correct it if an incongruence or error or incomplete aspect found.
So one ship will have to unload on one side and then another ship will have to pick up the cargo on the other end? So the shipping company has to have two ships, one on each side, and wait for the loading and unloading?
I'm sure smarter people than me have worked this out but it seems like this would take a lot longer than the canal (which only takes 8-10 hours vs 7 hours for this train plus load and unload). Granted ships without reservations have to wait 2-20 days to transit the canal right now, but I suspect wait times for cranes at each end of this rail would be about the same as canal wait times if not longer.
See e.g. [1]. As long as the Panama Canal is at capacity, they don't need to compete with the Panama Canal. They need to compete with long detours:
> The Panama Canal has become so backlogged that the world’s largest operator of chemical tankers has decided to reroute its fleet to the Suez Canal.
> London-based Stolt-Nielsen, which has a tanker division with 166 ships, is charging customers additional costs for the longer route, it said in an email. A bottleneck at the Panama Canal due to low water levels has prompted shippers to divert to Suez, the Cape of Good Hope, or even through the Strait of Magellan off the tip of South America.
And so, presumably, they're assuming that irrespective of resolution of these immediate issues, the Panama Canal won't be able to keep up with demand.
Generally you wouldn't have two ships, you would have dozens. This only makes sense if you sort. There are many ports across the Pacific, and many other ports on the east coast (include South American and the Caribbean!). Dozens of ships from different ports cross the Pacific and unload. Then the containers get mixed and matched with dozens of other ships on the other side going to dozens of different ports. And the reverse for the return trips.
Note that the ships do not all need to be owned by the same company, the important part is the containers get where they need to be. Sometimes they are, but container logistics is often separate and even when not the shipping companies will use each other when needed to get their containers where the customer needs them.
Given multiple source ports and destination ports having the ability to sort containers is useful. Similar to Fedex/Airline hub and spoke models. Also why there is a huge railyard in North Platte. https://youtu.be/zgpMRY1gAzw?si=-zNXLdLor5bt_0vQ
Another possible way this project could be beneficial is if the Panama Canal can't widen its locks, and hence this project could enable ships larger than that (which are more fuel efficient / cheaper) to have a far faster and possibly cheaper (depends on their costs, etc.) route.
Ooooh that's an aspect I hadn't considered. Post-Panamax ships tend to be 2x-3x more TEUs and they continue to grow.
Eeeenteresting.
Plus if they're already offloading to land in the Americas, don't just put 'em onto another ship and sail up to LAX, a rail route up through Mexico itself could mean the containers arrive directly into the US without going through LAX, which has interesting implications for labor.
>I suspect wait times for cranes at each end of this rail would be about the same as canal wait times if not longer.
This is the wrong way to think about it. Wait times would naturally start at zero days, and then increase from there until there is no more demand or the comparative advantage runs out.
Sure, as long as you have an even number of ships in your fleet and they can be paired off to carry the same number of containers. It's certainly solvable but adds a whole new level of logistical planning to make sure there is a suitable ship available and empty to receive the containers at the other end.
If you're shipping from northern hemisphere to northern hemisphere, then you also save to time/fuel to go down to Panama. No idea if that's significant or not.
I'm wondering why the whole thing is an issue anyway. Why is there a water shortage? There's an entire ocean on both sides of the canal. Gatun Lake provides the water buffer for the canal. Currently it's freshwater based but it's also artificial to begin with. Why not transition to using salt water pumped from the oceans to top it off as needed? Seems a hell of a lot easier than trying to replace the canals with rails.
"Every ship traversing the canal between the Atlantic and Pacific oceans requires 50 million gallons of fresh water"[1]. And it can transit 40 ships per day[2]. So, lifting 2 billion gallons of water, just over 9 billion litres, up to the 26 metres altitude of Gatun Lake.
Wondering how plausible that is, the numbers match very closely to the Dinorwig pumped-storage hydroelectric power plant in the UK[3], which stores ~9.2GWh of energy by pumping 9.2 million cubic metres of water, 9.2 billion litres, up 536 metres to Marchlyn Mawr reservoir. And it takes 7 hours to do that[4].
If the same amount of water has to be lifted only ~5% of the height, assume 5% of the energy, that would be ~450MWh/day for the Panama Canal.
I see it wouldn't be good to dump ocean water into Gatun Lake, but could the lowest locks of freshwater be pumped back up, instead of released into the ocean?
Gatún Lake is 425 km^2. It provides the drinking water for Panama City. It has a watershed covering a large portion of the country. You might be underestimating the scale of the problem.
A refinement of the thought, then, would be to install a small modular nuclear reactor to power desalination plants to replenish the freshwater when needed. Panama City must also be threatened if the lake level is that low.
The canal forms a ladder that raises ships from one ocean, up through a series of locks, into waterways that cross over the middle of the isthmus, then back down the other side. The canal itself it not all at sea level.
Locks work by controlling the release of water from a high point to a low point. The water needs to already be at the high point, by tapping a river or something that's fed by rain; locks do not have pumps to supply their own water.
To raise a ship, you pull it into the lock, close the doors on the low side, and open a valve that lets water from the high side flow into the lock and float the ship up. Then you open the doors on the high side.
To lower a ship, you pull it into the lock from the high side, close the doors on the high side, and open a valve that lets water from the lock flow to the low side. Then open the doors on the low side and pull the ship out.
The low side of one lock is the high side of the next one, so that water basically follows the ship and gets lowered with it as it descends the ladder. But the water has to get to the high middle somehow.
Just adding a few small details here:
1. When the canal was built, it was deemed to hard and expensive to dig at the level of the two oceans so they imagined the locks system as a solution for that. Back then, the canal was built with US money in a time of conflict between Panama(which was part of Colombia) and Colombia. Today Panama is an independent country, but they don't have the finances to upgrade the canal.
2. They don't want to bring water from the ocean up the canal is because the canal gets its water from a series of fresh water lakes. Bringing salty water would kill the life in the lakes are around the lakes and cause a lot more problems long term.
Agreed, pumping sea water seems terrible, but it also seems like there'd be no reason to do that.
First, at either end of the system there's a chunk of river at sea-level-ish but flowing towards the sea, so the water up at the upstream end where the lower locks are is probably not salty at all. Second, if each lock had its own pumps, doing it in 3 short lifts rather than 1 big lift to the top, then even if the one or two closest to sea-level ended up somewhat brackish, there'd be a dilution effect going up hill, and there would still be fresh water entering the system from the lakes which would tend to flush it downward. I don't think Gatun is at risk.
But what would it cost?
Okay, bullshit math time. In mixed units just to annoy metric purists:
A quick search says each lock cycle is around 50 million gallons, or 153-acre-feet. It takes 3 steps to raise them 85 feet above sea level, in energy terms let's just model it as one big step. It takes almost exactly 1kWh of energy to raise one acre-foot of water by 1 foot, so 153 acre-feet raised by 85 feet is about 13MWh per cycle. Apparently the canal manages about 40 transits per day, which I'm interpreting as each lock cycles once, so that's 521MWh of energy being provided by the rain delivering the water to the upper lakes.
Let's say we want to double that to 80 cycles a day. Oh, also, pumps are only about 70% efficient, so we actually need to provide 745MWh of power. Divide that by 24 hours and we end up with a 31MW power plant.
That's..... tiny.
Nah. Let's do it with solar. Roughly 6 hours peak sun equivalent per day (data for bocas del toro), we need 124MW of panels. The lakes have plenty of storage so there's no need for batteries, just run the pumps only when the sun is shining. Utility-scale solar runs around $1/watt installed. Labor is cheaper in Panama but conditions may be more challenging, I don't have good estimates for either of those so I'll assume they cancel out. Hey, it's napkin math!
$124M for the panels, figure the pumps and plumbing aren't cheap either but even if that's 10x the cost, we're probably still in the single-digit billions.
If they've been charging ships to use the canal like Egypt charges for the Suez, they should have quite the rainy-day, er, not-rainy-enough-day, fund for such projects. Especially if they raise prices auction-style during times of high demand, and let the high bidder through.
I did my own estimate ( https://news.ycombinator.com/item?id=38997124 ) via a different method, before reading down to yours, and got ~450MWh of energy needed - which is close enough to your 521MWh that I'm pretty happy.
> "Nah. Let's do it with solar."
Why not with hydroelectric? Store power by pumping water up to Gatun Lake, generate power while filling the locks as the water falls down...
> which is close enough to your 521MWh that I'm pretty happy.
Nice! And Dinorwig runs around 72% round-trip efficiency, which is higher than I expect to see here simply on account of scale, but also in the ballpark.
> generate power while filling the locks as the water falls down...
Now that's an interesting one.
There are already lock systems where there's two parallel chains, and the ascending and descending locks operate in lock-step, with the two locks first equalizing water with each other, before the ascending takes water from above and the descending releases water to below. So that doubles the amount of traffic that can be handled by the same amount of water.
Presumably building twice the infrastructure would be cost-prohibitive or they would've done it over the preceding decade, but storing the water's energy could be quite a bit cheaper and smaller.
Whether you're filling or draining, you clutch that turbine to the lift pump, and lift some water upward as the working water flows downward. No electricity needed.
They use freshwater from a man made lake, and over the decades, silt has reduced capacity and then they have been in a massive drought for years. they have no real reclamation from the locks back into the lake, so each trip wastes a ton of water. Really, they should just charge enough to reflect reality and then most ship with just go to the west coast
I was also confused by this, I had always (perhaps naively) assumed that the canal used salt water throughout. I did some searching and found this article.
> Gatun Lake, which forms a key stretch of the canal system and provides fresh water for its locks, saw little rain this year, as El Niño triggered a withering drought.
There's already a trans-isthmus railroad running from Atlantic to the Pacific in Panama. There's an oil pipeline as well. The water shortage is not going to get resolved anytime soon and is a result of the third locks project, a project that did not provide for increased water for transits. Another issue is that containers stacked on top of vessel decks are not counted for the purposes of calculating vessel tolls. In essence, they ride for free. Unloading and loading individual containers means someone will have to start paying for their interoceanic trip.
After a quick glance at maps, it looks like they will also need to build docks and container facilities in both ports. This isn't a terrible idea, but it's costly and requires a lot more than "building a railway."
Given that the containers may be distributed in a variety of modes when they are unloaded at a land port (e.g. different trains, trucks, etc), it's probably reasonable to think about fractional mode switches. I.e. what's the average number of switches across all containers on the ship. Using a rail portage to get to from the Pacific to the Atlantic will add 2 to that number for every container, whereas going to a west coast port might add one or maybe zero (I.e. switch to a truck that drives cross country, vs rail->truck)
Is the increased cost of transcontinental land transport in the US so high as to make this scheme worthwhile for US East coast?