A 100-year-old railway Mexico hopes will rival the Panama Canal (2023) | Hacker News Reader

A 100-year-old railway Mexico hopes will rival the Panama Canal (2023)

theweek.com

75 pointsmooreds119 comments

The first time I heard about the railway was a couple months ago from a RealLifeLore YT video. It goes into a lot more detail and is worth adding to your watch queue if you're into the engineering of global trade.

https://www.youtube.com/watch?v=glR7lvtrGRI

Ha! As soon as I saw the article I thought, “I wonder if there’s a RealLifeLore video on this.”

> The restoration is part of a "bold bid to steal container traffic" away from the waterway, said the Financial Times.

I doubt that. Isn't the Canal already operating at maximum capacity? The railroad would simply enable more traffic. It's not zero sum.

Long term maybe, but it'll take years for the rest of the logistics network to fully accommodate the new capacity and do things like building more ships or reallocating routes. In the meantime the new ports would probably undercut Panama fees to attract traffic that would otherwise go via the canal.

Not really as the canal is overfull so anyone in shipping has incentive to see if they can make this work for some shipments now.

Yeah we recently had a load pay something like $1.5m to jump up the queue to transit the canal. Not even a "guaranteed transit by X" but just jumping up the line! Which at the time was something like a month and a half long, IIRC.

Being a specialized load (read, unusual cargo, not containerized, on a barge) I doubt transloading to rail and back onto a ship on the other side would have been practical in our case but for other cargoes? It might be an attractive alternative.

Why not go around if the queue is a month and a half long?

I'm not an expert, but my understanding in large freight is that Cost >>> Time. Going around probably costs a lot more in fuel etc would be my guess.

And some freight (LNG) is consumed by waiting or going around (the ship burns it to generate power to cool the LNG)!

It would have to burn that anyways.

Cape Horn is quite hazardous, so it's not a journey to prefer anyways.

Friend went through the canal and had to sit for a few days with a bunch of other ships. That's something operators hate. Because any time spent waiting is lost opportunity.

I'm assuming being able to punt and offload and swap cargo would be a welcome alternative for operators. Maybe you spend a day to offload and load new cargo but that would beat waiting two days for a slot to open up at the panama canal.

Also there is already a railway along the canal, the Panama Canal Railway which connects coast to coast.

That's a much shorter train trip than this Mexico one.

So why isn't this rail line already taking the traffic Mexico hopes for?

The Panama canal rail line does carry cargo, but its capacity is much less than the canal; every dollar that Panama spends improving the rail line is a dollar not spent improving the canal.

Yeah I remember having this realization after I ran the numbers. A canal with a massive ship beats rail easily on throughput. So much for my idea to have ships plug into a maglev railway and fire their containers to the other side.

https://news.ycombinator.com/item?id=38992104

I don't think your math works out. Let's assume you want to move 3000 containers along a 3000 mile path. Let's also assume 1: you can stack containers one high onto the railway at the rate one container per minute, 2: the containers will move at 300 miles per hour and 3: the ship moves at 20.

On thenship all 3000 containers arrive at once and the total time to move them is 3000miles / 20 = 150 hours.

On the rail the containers take 50hr to offload. Hence the first container will arrive at the other side in 3000/300 = 10 hours. The _last_ container will arrive at 60 hrs 0 minutes. At one minute per container in the 150hrs it took the ship to arrive you could have delivered 140*60 = 8400 containers to the other side.

Not that the ship is only 15 times slower than the railway.

The governing factors on your railway are the ratios for distance to travel vs the offload time. Or put another way the delay bandwidth product.

>At one minute per container in the 150hrs it took the ship to arrive you could have delivered 140*60 = 8400 containers to the other side.

If you're considering a case of 3000 miles, that's not a canal anymore and can use the full size ships that carry 18,000 containers.

https://en.wikipedia.org/wiki/Triple_E-class_container_ship?...

I’ve heard operating those cranes is a taxing job. Load and unload costs probably stack up.

It can be almost fully automated. Including 'AI-assist(FM(fake mark))' for compensating wind load/unload from/to ship.

https://en.wikipedia.org/wiki/Container_Terminal_Altenwerder

I've seen these in action many times and called it 'techno-ballet'. Rather impressive. I guess meanwhile there are even more advanced ports, but I didn't track that development.

https://en.wikipedia.org/wiki/Port_of_Rotterdam#Robotic_cont... comes to mind, though.

I think the throughput is currently quite limited because of low levels of water in the natural lakes close to Panama which are used for both Panama's drinking water and for the locks in the canal, so they're restricting the number of ships per day...

The article mentions absolutely nothing about this, but what about the security plan?

This is going to sound snarky, but I genuinely look forward to how this operates with the current state of the Mexico government, lawlessness, cartels, and in that particular area, insurgents.

Bad actors are not going to simply ignore the value, there.

I would really love to see Mexico become a rich country with a functional and less-corrupt government/military/law enforcement. That seems insurmountable within a generation, though.

Insurance companies aren't going to want to touch that with a 10 foot pole.

Simply not true. You can buy insurance for cargo to/from Mexico shipment right now on https://exfreight.com/ . Powered by Haskell. Parcels, LCL, FCL, you name it.

Insurance companies have dealt with much worse; they got their start insuring ships trying to make the hazardous journey to the Far East.

> I would really love to see Mexico become a rich country with a functional and less-corrupt government/military/law enforcement. That seems insurmountable within a generation, though.

Amen to that. That will never happen as long as we continue The War on Drugs™ -- and that appears to be an impossible habit to kick.

The War on Drugs is a failed strategy, but I don't know of another way to tackle the problem of drugs and geography, WRT the US.

Decriminalizing hard drugs has been a failed experiment in Oregon, where hard drug use shot up drastically.

>Decriminalizing hard drugs has been a failed experiment in Oregon, where hard drug use shot up drastically.

Is the problem the increased usage, or is it the secondary effects (more homelessness, etc)?

I think both, but increased usage + new addictions were specifically cited by the legislation[1] and government officials. The uptick in prevalence of homeless people in the suburban and urban areas is startling though.

1. https://olis.oregonlegislature.gov/liz/2024R1/Measures/Overv...

Oregonian here. Both, I think. However homelessness increased everywhere over the same period the legalization covered so it's kind of hard to gauge how much legalization actually increased it.

The problem was that there was no way to compel people into treatment for addiction when the drugs were essentially legal. There was an attempt to replicate what Portugal had done in the early aughts, but as I understand it they could compel people into treatment. The other problem was lack of treatment resources - even if people wanted treatment it could be difficult to come by. While I tended to favor decriminalization before, after seeing the results close up it seems that there should have been a longer period of preparation prior to the enacting legalization. And for substances like fetanyl and meth - highly addictive substances that cause people to resort to crime to keep up the habit - it's probably not be a good idea to ever legalize them because overdosing, death and permanent mental damage are common. We need a more nuanced approach where less addictive (and non-addictive) drugs like pot, LSD, shrooms, etc. are essentially legal while highly addictive drugs remain illegal with stiff penalties for distribution.

But it’s clear that there is no penalty or enforcement budget that will prevent a black market in hard drugs. You will have heroine and fent users, there is no way to prevent it.

So, how do you minimize the impact to everyone?

Decriminalization is entirely different from full legalization. That is only sensible and moral solution.

Do you have anything more to say on this point?

Given my experiences in Oregon, I'm having trouble thinking of it as sensible, but I'm open-minded.

Having seen it myself, the only people that are still talking about legal drugs this year, are people who haven’t been to Oregon.

Do you Oregon folks not think Oregon fucked up the implementation? You can’t simply decriminalize and not have adequate resources to handle the results.

Have you argued that no one has tired real socialism yet too?

This is the thing. Oregon basically decriminalized but didn't build or maintain the harm reduction infrastructure necessary to make decriminalizing drugs not a terrible idea. The two things support one another -- make it easier to stop using, and stop penalizing people so harshly they are scared to use services.

I don't know how many addicts y'all have been around, but in my experience no one wants to be an addict.

There are resources in place, but at least in my area, they aren't being utilized to capacity.

So while it seems logical to the rest of us that addicts don't want to be addicts, they're also apparently not willing to tap into that help.

You can argue this or that about something not being funded right, or set up right, but at the end of the day, these people have to make efforts. And a lot of them try! I talk to them! A lot just don't give a crap and have given up on life.

Then begs the question: who's going to pay for it, and from where? Our education is already underfunded, and personally I'd rather see funds go to education, particularly non-special education (already well-funded) and Career Technical Education.

Why don't they utilize them? What's keeping them from doing so? Addiction is brutal, and getting clean is not a simple process of rehab.

Getting clean requires a ton of energy and patience and it often hurts like hell. For someone who is spending 100% of their energy on surviving and managing their addiction, they might not have the reserves to try to get clean.

We need to figure out how to reach them, not write them off.

Also, "I'd rather fund x instead of y" is a trap. We can do both, and we can say the bar is fully funded education and fully funded support for addiction and recovery.

So, I'm somewhat tangental to these resources you describe. And you're right, they're not adequate or at the proper scale, but at the same time, the meager resources that are in place aren't even being fully tapped.

While it would seem likely that addicts do not want to be addicts, they're not exactly responding to the outreach and least in my metro area (3rd largest city in OR).

I don't think criminalizing or decriminalizing drugs really matters so much as stuff like fetty becoming commonplace in the past few decades, fetty precursors being produced by China and ending up in the US, or specific pharma companies heavily promoting anti-pain meds to encourage people to become dependent.

I think discussions on criminalizing or decriminalizing or The War on Drugs is a waste of time if the goal is to reduce drug usage because they only affect the symptoms of the drug problem. We should still discuss them, but for other reasons.

Did it shot up (relative to neighboring states, I assume, otherwise it's not very relevant) because more Oregonians started doing drugs or because drug addicts from nearby states moved in?

While the war on drugs has largely been a failure, it's too late for US drug policy changes to bring down the Mexican cartels. Even if the US government completely legalized all drugs (not something I necessarily support) the cartels would remain. Mexico is a failed state where the central government no longer even tries to exercise control over large parts of their nominal territory.

As long as the right cartels are able to skim value, I could see this being successful regardless. Cartels are businesses at the end of the day- and they won't totally destroy an industry as long as it's making them money long term. The problems occur when smaller groups start fighting over access to the resources.

That's one of the many problems with cartels: they fight over control of valuable resources. Occasionally a single cartel gets a firm hold onto something and right about then, the US government swoops in to decapitate the leadership, re-initiating a war. I.e. What happened with CDS and CJNG after El Chapo's extradition to the US.

Also, if a certain skim percentage becomes acceptable loss to all parties, the cartels will want more.

Oh those poor drug lords.

Morena is slated to win the election in June because of the perception that infrastructure projects are getting done. Sheinbaum will want to get re-elected, and if the railway gets robbed that will be difficult. It's a lot easier to protect a railway than to create a good civil society with jobs for millions of people.

https://en.wikipedia.org/wiki/Fourth_Transformation#Infrastr...

Whether it be by demographics, credit collapse, Taiwan invastion, or just authoritarian incompetence, the world and especially the US are moving away from China-centric manufacturing.

The US needs to get its drug war policy in line, but so much is on the line that I think the cartels are going to become a point of "special emphasis" in the coming decade. Mexico is going to become the center of manufacturing for the US in all likelihood.

The cartel/drug civil war will come to an end as part of a broad sweep of investment and interest by the United States. I just don't see the cartels maintaining their political/corruption power when the stakes become so much higher.

The cartels either need to legitimize, or they'll probably get a healthy dose of US military forces attention and be ground to dust.

For those not intimately familiar with Mexican politics: Morena is a political party (according to Wikipedia, "left wing and anti-neoliberal"). The current president of Mexico is from them

Sheinbaum is the current governor of Mexico city, and candidate and frontrunner for the upcoming presidential elections.

Note: Mexico doesn't do presidential reelections - even temporary holders of the post can never run for or hold it again. They had a number of civil wars and related fights around this so it's taken very seriously.

From the Financial Times article [0] cited in the OP:

Mexico’s government is bullish about prospects for the rail crossing, which will offer proximity to the US and a transit time of 6.5 hours excluding loading time — less than the eight to 10 hours it takes on the 80km canal.

Much more than transit time, I'm interested in throughput. How long does it take to move a full container ship's contents from one end to the other? How long does it take just to unload and load?

Also, regarding loading/unloading: When I think of a train, I imagine it being loaded/unloaded in semi-serial - first the front cars, then the next set of cars, as the train moves segment-by-segment past the ship. Essentially you are transforming a massive pile of containers onto a long string of containers. Is that correct? Is there any solution to speed it up? I suppose they could lay several train tracks in parallel alongside the ship and use extra train engines to pull up groups of rail cars, and then reconnect them into one long train as the train pulls out. How long is a train carrying a full container ship's capacity?

Cost seems the other essential question: Sea transport costs a fraction of rail transport, generally, but I expect that canal transport is far more expensive than open sea transport.

[0] https://www.ft.com/content/bbe8e851-e9f2-4ed1-a4a7-051a670b4...

Apparently it takes about 3 days to unload a "full size" container ship.

So this rail isn't going to be tremendously useful unless they connect it directly to the USA.

As for loading the trains, they already do it in parallel; look at an arial photo of a stockyard. For this specific use, you could parallelize it even more I suspect.

Yea, I don't really get why the 6.5 hours is particularly pertinent. The unloading time (and expense) will dominate, and any other railroads connecting a Pacific port to an Atlantic port in the Americas that are less than ~2 days transit time will be competitors.

Why does connecting directly to the US help? If you're in one ocean and just want to access the US, can't you just steam direct to a US port?

By the jones act ships cannot go US port to US port. so all ships plan to unload once. this can work aroud that by letting a ship unload some in Seattle and then send San Diego from Mexico via train (or whatever variation makes sense)

I mean ok but why not just unload at any mexican or canadian port with a rail connection to US? Why go all the way down to bottom of mexico?

By ship is still by far the cheapest way of transporting cargo, and the proposed route would involve the shortest land distances.

More importantly the Sierra Madre ranges are large geographic barriers throughout most of Mexico but are relatively flat in the area of the Chivela pass, where this rail line is.

My comment was discussing using the rail for access to the US, not for getting from one ocean to another.

Most of the US lives on the East Coast so at some point those mountains have to be crossed.

The elevation of the plateau in between the Sierra Madres is no joke; Mexico City is at 7349 ft elevation.

They can. However if the goal is the center of the US ports in Mexico may be closer (remember mountains cause are expensive to cross and so a straight line on a map may not be the best route). I'm not in logistics so I can't tell you all the considerations used to decide if they should stop in any particular port or not.

Right now the main thing controlling where the ships land is where they have ports that can receive their massive tonnage; places like Long Beach (16,618 TEU) or Houston (3,974 TEU). It's quite arguable that additional ports need to be built so that the ships can disperse cargo closer (or even farther, but more efficiently) to where it is going.

That's not quite right. Boats can't pick up cargos in one US port and move them to another. They can drop in multiple places but it is usually uneconomic.

If a boat that would go through the locks could unload in Mexico and send it to its final destination it might be worth it, but unlikely (time wise).

The whole thing comes into its own if Panama is overloaded and the other option is taking around the South America).

> How long is a train carrying a full container ship's capacity?

According to Wikipedia[1], "Container ship capacity is measured in twenty-foot equivalent units (TEU)."

And container ship size categories are (conveniently!) based on what can fit through the Panama Canal. Neopanamax ships are the category that can fit through in the canal's new (upgraded) configuration, and those can hold 10,000 to 14,500 TEU.

For round numbers I'll use 10,000 TEU. That means if you laid all the containers end to end, they would stretch out 200,000 feet, which is about 38 miles or 61 km.

Trains can stack containers two high[2], so divide that by two, and a train has to be at least 19 miles or 30 km. But really somewhat longer since one train car's containers don't butt up directly against the next car's containers.

So presumably they would need multiple trains for one ship's cargo.

---

[1] https://en.wikipedia.org/wiki/Container_ship

[2] https://en.wikipedia.org/wiki/Double-stack_rail_transport

Afaik, some container ships are made much larger than what will fit through the Panama Canal. If you are shipping from China to anywhere in the Indo-Pacific, you don't need to worry about canals.

> How long does it take just to unload and load?

If it is more than 8-10 hours, you start getting overlapped I/O.

As to parallelization, I wonder how much of the route is a single track. Parallel tracks could allow coming and going without interference. (then there are siding and crossover tracks, etc)

I also wonder - do you have to unload EVERYTHING? maybe just the express cargo, while the rest of the ship waits.

Or you could combine ship cargo and one ship could make the long trip, while others return back across the pacific.

Its maximum annual cargo capacity would be about 10.5% of the total amount of goods carried through the Panama Canal in 2022, according to calculations from the FT.

Other articles:

The upgraded Coatzacoalcos - Salina Cruz line, as well the ports in both cities, are managed by the state-owned Isthmus of Tehuantepec Railway, part of the Mexican Navy.

https://www.railjournal.com/passenger/main-line/passenger-op...

By 2028, the Isthmus of Tehuantepec Interoceanic Corridor (CIIT) should see 300,000 cargo containers transit every year, with 1.4 million — about 33 million tons — by 2033. The 80-km Panama Canal moved about 63.2 million tons in 2022.

https://www.japantimes.co.jp/business/2023/11/16/mexico-rail...

This will never work. A typical Panamax ship carries 4800 containers. A typical freight train carries 240. You need 20 trains for each ship; at 2km per train, that’s a continuous line of 40km. In reality trains need at least 10km between them, so that’s 240km. But you need slack in the system, otherwise any hiccup results in a huge jam. So realistically the maximum capacity of such a railway is one Panamax per day and direction. What fees can you charge to cover the operating costs and the capital financing costs? And to hope to be competitive with the canal?

Can’t they just build parallel railways?

They certainly can, but to build two you first build one. And if that one is not profitable, why would adding another one make it so? You could say that building 2 parallel tracks is not significantly more expensive than building one, and when you have the 2 you need less slack in the system, so you can have closer to 100% utilization rate. Yes, you could gain some economies of scale, but the fundamental problem remains: if you cross the canal, you just float on water; if you decide to use this alternative, you need to unload the ship, load 20 trains, run the trains for 300 km, unload them and load them on another ship. To be competitive with the canal, you need to do this for about $200k. The $200k should cover the operating costs (fuel, labor, equipment depreciation), the maintenance costs, the insurance costs, the capital costs and hopefully should include some profit.

The real value is if they have space and equipment to sort. That is unload the containers from one ship and put them on 10 different ships to different ports. With the jones act ships cannot go US port to US port, so small ports are bypassed when if they were used they could improve operations.

bonus if they can connect to rail heading to the middle of the us (think El Paso to Denever)

The Jones Act is one of those truly infuriating acts of protectionism that absolutely hurts everyone except a small and dwindling number protected producers and doesn't accomplish its stated purpose. It's sad that this sort of policy is once again in vogue.

Maybe those producers could be bought off to end the Jones Act.

As in, every union member gets an immediate pension and $1M cash as part of the legalization.

It's not really about the unions in this case, it's just the few owners of some decaying old ship yards. There are only maybe 90 Jones Act ships in existence (you'll see different numbers). These are the sum total output of ~150 US shipyards. Its time to rip of the bandage, not only does this arrangement produce almost no ship it hobbles the domestic shipping industry, small ports, Hawaii, Puerto Rico, and anyone who wants to move goods or people by water between US ports.

OK, these ship yard owners can surely also be bought off with the right size one time amount.

They don’t deserve a dime, they’ve been living off the government for generations.

> bonus if they can connect to rail heading to the middle of the us

I think this might be where their big advantage would be. The ports of Los Angeles and Long Beach are at capacity (and considered inefficient), and the Mexican container port of Manzanillo further down the coast is only medium-sized (1.4 million TEUs). If Mexico were to build dedicated freight corridors like India has, they could move an amazing amount of cargo into the central US.

The basic logic is convincing: the canal killed the train line's business long before containers became a thing. Ship-to-train and train-to-ship should be far more feasible today than it used to be back then.

And the train-to-ship part would double as ship-to-ship redistribution, fanning out to different ports in both Asia and on the east coast. Basically a role similar to the one served by Atlanta in the United network.

Discussion from January:

Mexico building rail rival as water shortages drain Panama Canal

https://news.ycombinator.com/item?id=38991317

Thanks! Macroexpanded:

Mexico building rail rival as water shortages drain Panama Canal - https://news.ycombinator.com/item?id=38991317 - Jan 2024 (107 comments)

An average container ship holds 15,000 containers. An average train holds 400 containers. That's 37.5 trains to move the cargo from _one_ ship.

Plus the time to unload and load all those containers (3-4days). I don't see rail competing with the canal.

Edit: On average there are 7.5 container ships transiting the Panama canal per day. And it takes 8-10 hours to transit ocean-to-ocean.

That all depends on price. If it is dirt cheap per container compared to the canal it would get used.

Sea transport costs order(s) of magnitude than land transport. I don't know about canal transport.

But the rail solution probably scales a lot better. After all, train tracks are much easier to build than canals.

If you can get dozens of trains from coast to coast in the time it takes to get one caego ship through the canal, that could end up being significantly more economical - and maybe even faster - in the long run.

It's 8-10 hours to transit the canal. Assuming a train is faster, you'd still need almost 40 trains to carry the same number of containers...are 40 trains still faster in total?

Sure, you could scale it and build 10 train tracks side-by-side (whoa!) which would support 5 trains in each direction, and then you'd only need to get 8 trains faster in total.

And that's disregarding the 3-4 days needed on either end to unload/load the ships.

You need to count only time to load/unload a train. It's not like unload the ship and then load train. Trains would go as soon as they are loaded, so 4 days for unloading a ship means 10 trains a day from one ship.

> But the rail solution probably scales a lot better.

Not really, no.

It takes 3 to 4 days to unload 1 ship, and we can assume it takes about the same time to reload them on the other side. Let's be generous, and assume that we can put the entire load on one train, that has an average speed of 40 km/h. That will be an other 8 hours gone to get from harbor A to harbor B. So, for one ship, you went from a 10 hour crossing, to 8+ days of loading-train-travel-unloading.

Now for the logistics part. If one ship holds 15,000 containers, and a train carries 400 containers, you need 38 trains to transfer 1 ship worth of containers.

In 2022, 19 ships a day passed through the Panama canal, per direction. So you would need 722 trains to process the same capacity.

You have 1440 minutes in a day, so to absorb the the incoming freight, you will need a train every 2 minutes on a single track, 24 hours a day. Which, is completely unrealistic. You can double the track, and now you have a 4 minute slot in between trains, per track. Still completely unrealistic. If you would have 5 tracks per direction, you can get to a 10 minute slot in between trains, 24 hours a day, which is still not realistic.

So, in conclusion, we can say that the rail solution does not scale better.

> If you would have 5 tracks per direction, you can get to a 10 minute slot in between trains, 24 hours a day, which is still not realistic.

10 min slot is starting to be realistic. But the unloading/reloading is still not realistic. At least, not with current trains and loading cranes. If you could make special purpose rail and cars which hold 4 containers per car (2 side by side, stacked 2 high), you get 4x capacity, now only 2 tracks per direction is enough. Special cranes which could sit above whole ship and unload several containers at once would speed unloading.

Such engineering challenge would probably cost as much as original canal, but would not require so much water (water is main bottleneck currently).

Instead of trains, how about a motorized chassis per-container - with electric motors like subway cars - with little space between containers, and fan-in and fan-out at the ends to re-sort containers into different ships.

It sounds impractical, but so does marshaling all the trucks and trains to get the very same containers from the original shippers to the ports of origin and from the ports of destination to the final recipients.

Ha! You could get a 300 km perpetual snake-train of motorized chassis following each other. And they could hit the 40 km/h traveling speeds of the current American freight trains, no problem (such a low bar..)

Could be done, I guess. The high-speed tracks in Europe are protected with fences, and have no level crossings, so the same could be done here to have a fully-automatic motorized wagon system. Without locomotive you would need a third-rail electricity pick-up system. Less safe, but the track is fenced off, so why not?

The railroad is a good idea, but the concerns on throughput are real. If it’s a single-track railroad, your consist length (at least in one direction) will be limited by the length of sidings. Every rail car heading east will inevitably head west (to pick up containers).

It’s a big operation, putting together a new container-oriented freight railroad, and I hope it goes well!

I'm also curious how security will be handled. 300km, mostly through middle-of-nowhere, with potentially high value cargo seems like organized crime's bread and butter

It seemed to work well in Fast Five until there was a plot twist.

https://en.wikipedia.org/wiki/Fast_Five

Organized crime doesn't want attention from people with power to cause them problems.

I'll have to assume you aren't too familiar with Mexico then.

Scaling a railway (horizontally and vertically) is a lot more doable than scaling a canal, right?

depends on the scale factor. can you scale it 10x to match the canal? probably not.

Well, the point is moot, because you can't scale the canal either.

How is the security situation in Mexico these days? I only hear vague "the cartels are taking over" news from time to time, but it's not really something I've looked into in depth. Security is probably a big topic for such an endeavor though. If the companies have to fear that they loose their shipments they won't use it.

Isn't there already a railway along the canal? The Panama Canal Railway? Is it not used to ship containers from ocean to Ocean?

This is a good plan. It's not so much that it's amazing on its own, but that the Panama canal is showing signs of weakness and alternatives are needed.

Already the "drought" is causing issues with the Panama canal, and not only is that drought never-ending, it's going to get worse over time (climate change).

Is it? If ocean levels rise doesn’t that just change how they need to handle the locks to balance out water end to end?

not really, there are ways to get more water stored with dams, and the drought wouldnt have happened if they hadnt decided to just take water from the lakes for consumption, they were built for the canal to work.

I wonder at what point it makes economic sense to "solve" the freshwater problem in Panama with new infrastructure. My understanding is that they lose a lot of water because it has to get flushed out to sea.

Is there an engineering solution to recycling more of the water and flushing less of it to sea?

You use big pumps to fill the lock from the low side instead of just opening a valve or the doors slightly. This would then increase the salinity on the lake side and cause whatever problems that would cause.

What I'm saying is, is there an infrastructure solution that doesn't increase the salinity in the lake? Like can we use gravity up to the last lock, but then pump all the fresh water out back to the lake and then pump seawater into the chamber after it's empty? Or something crazy like that.

The locks are all gravity powered. They certainly could operate the locks by pumping uphill instead of releasing downhill, and thus lose almost no water. I expect the expense of this makes it a non-starter.

>I expect the expense of this makes it a non-starter.

I am seeing numbers that it costs anywhere between $100k-400k per trip to go through the canal, with some tens of ships per day.

So of course, this highly profitable venture would dedicate basically nothing to improving the infrastructure until it became a tire fire.

It feels like any form of recycling water would mean pumping water back uphill, which is energy-intensive, at least relative to the current model of "just pull water from Lake Gatú and assume it never runs out".

But using less water in the first place, I'm curious what approaches might exist.

The new set of locks in the Panama canal use "water-saving basins" to reduce the amount of water lost for each cycle of the locks. https://en.wikipedia.org/wiki/Panama_Canal_expansion_project...

I guess in principle something like this could also be retrofitted to the existing locks.

The OP probably should be changed to the Financial Times article, which the The Week article is summarizing:

https://www.ft.com/content/bbe8e851-e9f2-4ed1-a4a7-051a670b4...

Gift link: https://on.ft.com/3UcOwKR

6.5 hours for 300km sounds very slow. Is that because it's only a single track, and you lose a lot of time when passing oncoming trains? Or is the track not straight enough for fast trains?

Why optimize for latency when throughout is your single dominatant metric? And with braking distances being the limiting factor of all train signaling schemes, you might even find yourself able to increase throughput by going slower. (not necessarily of course, but train signaling schemes are ... complicated)

Freight trains are almost never fast trains. It isn't worth the extra energy for saving hours for containers that have been in transit for days.

The rehabilitation increased the speed from 12mph to 43mph. They are planning passenger train that can reach 50mph. So that is probably limit to the railroad.

I don’t think cargo trains ever go “fast”. But it is slower than normal in the US anyways. I think in US max speed for cargo is like 50 mph / ~75 kph, and this train is going along at 46 kph assuming constant motion. In reality, it’s fast in some segments, dramatically slower in older segments.

According to Wikipedia, it’s an old line that has speed limitations in some areas of 20 kph. Addressing these limitations is a substantial part of the rehabilitation project, aiming to get the speed from 20 kph up to 70 kph.

https://en.m.wikipedia.org/wiki/Interoceanic_Corridor_of_the...

“the rehabilitation is reportedly expected to result in a speed increase for trains using the railway, from 20 kilometres per hour (12 mph) to 70 km/h (43 mph) for cargo trains, and of up to 80 km/h (50 mph) for passenger trains”

Depends on the cargo. Mail and fresh food goes fast as time matters. Most other freight would prefer to save money with slower speeds.

The weight and speed limits are a regulatory matter in the US, as there's a safety factor at play. Derailments of cargo tend to be... messy affairs.

Freight travels at many different speeds. There are many tracks in the US that are speed limited to 30mph or less because they have not maintained the tracks to higher standard, even though there are other freight tracks where trains move at 60+mph.

Yep, sometimes there are hills that also cause trains to go faster or slower depending on the incline. Fascinating world out there.

Speeds are regulated.. Here's some reading: https://en.wikipedia.org/wiki/Rail_speed_limits_in_the_Unite....)

No map?

Reply on news.ycombinator.com