Hacker News Reader

Comment by chris12321 | Hacker News Reader

I recently read The Ministry for the Future by Kim Stanley Robinson, and one of the ideas in it that I thought was very good was replacing our cargo ships with wind-powered ships, basically giant sailing ships. In the book, they were incredibly slow, with shipments taking months to complete, but if supply lines were set up correctly, that wouldn't matter for a lot of cargo. Cargo ships are a massive CO2 contributor, and it was interesting that a solution could be to return to sailing ships.

I know there are probably huge engineering problems preventing this from happening, so feel free to tell me why it's impossible.

View on HN

Currently, ships need human sailors. They perform maintenance aboard ship as well as have legal oversight of the craft. We are not yet able to replace the crew with automation.

It's difficult to find skilled crewmembers willing to sign up to extremely long rotations away from home.

It just takes money. $100k-$300k/yr is plenty to have your pick of pretty good people, especially if there are any perks like the food being halfway decent (should be basically a given if you have to pay the chef a lot anyway).

With a fraction of the money, you pay for energy to move faster ...

That's not even close to true for the kind of large shipping we're talking about. Crews are small (the Ever Given had 25 people on board) but the ships they crew take up to ~100k gallons of fuel per day (Ever Given has a fuel capacity of 3622168.679 gallons, 13711400 liters, and is set up for voyages of ~30 days underway).

Fuel costs are ~$2.5 USD/gallon for bunker fuel. That means a cool $200k per day (conservatively).

It is absolutely not the personnel costs that'd be the big differences in expenses.

Good point, I didn't have a handle on the fuel costs.

Backup argument: if you go at half-speed, you'll need twice as many ships for the same throughput.

Not really. The unit economics work out heavily in favor of wind even with slower trips and absurd wages and the fact that oil has its externalities pushed to other people. Ignoring local manufacturing minima, the reason we don't do more of it is that the capital outlay is important and heavily favors faster trips, much like how excess solar for refinery power isn't often enticing because the factory spends too much time idle. Combine that with manuverability in canals (so you probably need a powerful engine anyway), and the project needs a lot of TLC to make economic sense while oil is subsidized to this degree, but unit costs aren't the culprit, and even wages at that extreme are totally fine.

But, ships need far smaller crews than they did in the past. A tall ship takes a larger crew than a steamship back in the 1980s. (I've crossed the Atlantic both ways.) Today, with automation, we have unattended engine rooms (unattended machinery spaces or UMS). You'll never totally eliminate a crew, for hte reasons you mention; but, we've reduced the size significantly.

Or Nuclear Propulsion:

https://en.wikipedia.org/wiki/Nuclear_marine_propulsion#Merc...

> Nuclear ships are currently the responsibility of their own countries, but none are involved in international trade. As a result of this work in 2014 two papers on commercial nuclear marine propulsion were published by Lloyd's Register and the other members of this consortium.... > This is a small fast-neutron reactor using lead–bismuth eutectic cooling and able to operate for ten full-power years before refueling, and in service last for a 25-year operational life of the vessel. They conclude that the concept is feasible, but further maturity of nuclear technology and the development and harmonisation of the regulatory framework would be necessary before the concept would be viable.

> In December 2023, the Jiangnan Shipyard under the China State Shipbuilding Corporation officially released a design of a 24000 TEU-class container ship — known as the KUN-24AP — at Marintec China 2023, a premier maritime industry exhibition held in Shanghai. The container ship is reported to be powered by a thorium-based molten salt reactor, making it a first thorium-powered container ship and, if completed, the largest nuclear-powered container ship in the world.

Nuclear ships are technically possible, but have a massive number of downsides.

- The construction cost would be significantly higher than a conventional ship.

- Reactors are far from trivial, so you'd double or triple the crew required.

- Shipbreaking would become even more of an issue than it already is. You can't just beach a ship like this in Bangladesh and have a bunch of untrained people attack it with plasma cutters.

- The ship would be a huge target for pirates and terrorists. It's essentially a floating dirty bomb, after all, just waiting for the USS Cole treatment.

- A lot of countries would not accept nuclear ships, both due to perceived security risks and for more ideological reasons.

... and that's probably only the tip of the iceberg.

Nuclear is barely economically viable with land-based large-scale nuclear power plants running for 50+ years. They are an attractive option for some military ships, but I doubt anyone would be willing to risk it for regular commercial shipping.

> They are an attractive option for some military ships, but I doubt anyone would be willing to risk it for regular commercial shipping.

There's been a few built over the years, mostly for research.

Russia apparently still operates one.

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

Despite hurtles you've pointed to it is still being considered:

https://www.spglobal.com/commodity-insights/en/news-research...

> This source of power confers some advantages. "You will have ships going maybe 50% faster because the fuel is essentially free once you have made the upfront capex investment," Sohmen-Pao said.

You achieve ~0 emissions AND avoid increasing transit time going with pure sailing ships.

We should not under-estimate the need for speed in supply chains. Predicting future demand is hard. To be more specific, we're talking about predicting ~100M unique products (the order of magnitude that moves on the pacific) and some of them have very lumpy demand (e.g. invent a new product, but it depends on 100 other obscure products).

We should also not over-estimate the need for speed. Just because some items need speed, it does not follow that all items need speed.

> if supply lines were set up correctly, that wouldn't matter for a lot of cargo.

One of the big problems facing logistics across the board is just optimization. But at some point, you run out of intuition to uncover more efficiencies. This space is actually a really good use for AI. In fact, it's even useful for predicting what to put on that boat ahead of when it's ordered/purchased (up to a point). So yes, longer shipping times might not be that big a deal for non-perishables and frozen products.

> Cargo ships are a massive CO2 contributor

Not really. They’re 1.6% of global emissions if I multiplied the numbers on this page out right. The table says they’re 20% of shipping emissions, and the intro says shipping is 8% of global emissions (excluding ports and warehouses).

This result seems surprising until you realize that semi trucks produce 100x more CO2 per kg-mile of cargo.

https://climate.mit.edu/explainers/freight-transportation

We probably don’t have the physical space to onshore 6x the capacity in warehouse or whatever it would work out to around ports. Short of multilevel warehouses but I mean containers have been stacked 5 high since covid around the port of LA and I think that is about the limit without some significant rethinking of process. Maybe the ships could just anchor off shore for longer and function as the warehouse.

There's nothing stopping it, here's a link to an article from 2023: https://www.bbc.com/news/technology-66543643

Why can't cargo ships deploy floating solar panels to power the ship motors?

Because floating solar panels add drag proportional to their area, and it takes a lot of area of panels to power a motor that is sufficient for a cargo ship even without the added drag of the panels. Also, because oceans and the things one runs into in them aren't easy on solar panels being dragged along by cargo ships.

Would make more sense to produce chemical from solar energy harvested on the water fuels, collect the fuel and then use this with ships

i wish there was more talk about this. it seems i heard a lot about making hydrocarbons from co2 in the air + solar or algae a couple years ago. if your hydrocarbons are made this way it seems they would be carbon neutral.

i'm guessing there's more research to make it feasable since i haven't seen "carbon neutral gas alternative" at the local Chevron.

There has been quite some buzz about ammonia, as it is fairly easy to turn electricity into hydrogen, and hydrogen into ammonia. It has a reasonably high energy density, is not too nasty to handle, and already has a huge industry built around it.

My understanding is that drag is more about the "front-on" view of a craft than how long the craft is.

Since solar panels are very thin and aimed up, it feels like they add minimal cross-sectional area to the craft. Your assertion seems trivially incorrect to me?

Ships drag across sticky goop, not fly through soup.

There's a pressure drag and skin friction drag. Friction drag is supposedly a majority component unless you sail a brick. But I don't have sources to prove that.

Oceans can be extremely rough, but even mild waves make it inappropriate to approximate PV as thin.

The requisite area to power a ship is huge, something like 1.4km^2 (ballpark estimate for 20% cells, reasonable capacity factor guess, 60 MW consumption requirement). If a ship is about 30m wide, it's trailing about 45 km of PV. You're not even into 4 digits of cargo ships before the combined length is longer than the circumference of the planet.

> My understanding is that drag is more about the "front-on" view of a craft than how long the craft is.

Drag (fluid mechanics generally) is... ludicrously complicated. For the typical shapes of ships, I believe you are correct that the main factor is cross sectional area perpendicular to the direction of travel, but that’s not universally true. i think that for a floating raft of panels, it would be proportional to the panel area, similar to how for winged aicraft its the wing area and not the cross section perpendicular to direction of travel.

Economics.

The solar panels would be more expensive than bunker fuel.

Sails would be cheaper.

it might be fun to try to make a modern wooden sailing ship cargo fleet.

maybe with an emergency diesel engine in the back.

It's been done, however the scale of modern Panamax containerships is baffling and most people underestimate their size.

https://www.newscientist.com/article/2445620-worlds-largest-...

Interesting idea, but that would require more than a square kilometer (or a 100m strip 10km long) of solar panels (not accounting for the additional power required to tow the panel array).

Solar power being useful doesn’t require 100% of propulsion to come from solar panels.

You see solar panels added to a wide range of boats because even bunker fuel isn’t free and panels are light for the power they provide over even a few days. A current 399.9 * 61.3m container ship doesn’t need panels everywhere to benefit, but the potential savings is significant if they do.

This is unfortunately not true because of the dynamics of diesel engines: there is by design surplus energy relative to requirements from running them at efficient operating points. Otherwise the ship is not a good ship.

You can always scale design to fit reduced demand. Also, loss of efficiency is more than made up for with vastly lower energy demand.

“Lowering speed reduces fuel consumption because the force of drag imparted by a fluid increases quadratically with increase in speed. Thus traveling twice as fast requires four times as much energy and therefore fuel for a given distance.”

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

“Container ship Emma Mærsk in Aarhus, 5 September 2006 Mærsk Line's E-class container ships such as the Emma Mærsk can save 4 metric kilotons of fuel oil on a Europe-Singapore voyage by slow steaming.[5] At typical fuel prices of US$600-700 per tonne,[4] this works out to a saving of US$2.4-2.8 million on a typical one-way voyage. Maersk's Triple E-class container ships were designed for slow steaming and have less powerful engines than their predecessors.[5]”

Sure, but what does this have to do with what I said? You need design and operating margin, and the engine is always running.

Reducing the load is always going to save fuel. There’s no difference between energy used to move a boat and energy used to run the lights.

Put another way if there was excess torque being generated it would go somewhere such as increasing the engine RPM.

What you seem to be missing is that your understanding is not true because of the practical realities of operating large internal combustion engines.

For example, one tonne of fuel is about 11 MWh. So if you run the calculations, you will see that adding solar panels to a diesel boat, even if the energy they provide is free, essentially never ROIs, and makes the boat less reliable and useful as a boat.

These kinds of engines generate tens of megawatts when they are on, and they are always on when the ship is moving.

One tonne of fuel is more like 5.5 MWh. You did a mathematical calculation while ignoring engine efficiency.

The reality of large internal combustion engines is you still pay for every single kWh. These ships already have extremely complex electrical systems with multiple redundancies and load balancing etc.

The dealbreaker is R&D as unlike a house or sailboat you can’t just yolo where panels are placed and wires run etc, this is all bespoke engineering with few of any give design being manufactured and little available space.

No, one tonne of diesel fuel contains about 11 MWh of potential energy as determined by calorimetric methods. One tonne of fuel when consumed produces a variable amount of useful energy output depending on the efficiency of the engine.

If you said fuel was 5.5 MWh per tonne people would wonder what you cut it with.

The reality of outputting 80MW is that the power to your lights is a rounding error and you’d be better off buying a robot to regularly clean the hull.

> No, one tonne of diesel fuel contains about 11 MWh of potential energy as determined by calorimetric methods. One tonne of fuel when consumed produces a variable amount of useful energy output depending on the efficiency of the engine.

That’s almost correct, good try.

> lights is a rounding error

Ships use electrical power for far more than lighting, and no electricity is not a rounding error compared to profit it’s a significant expense for cargo ships.

I'm not an expert, but I've worked close to some of the engines that power those ships. My gut feeling is that you're vastly underestimating how much power those ships consume (and therefore produce).

Reply on news.ycombinator.com