Facts about engines

The RMS Titanic weighed almost 50,000 tons and could carry 3,500 people. Before it sunk, it was world-famous as the massive titan of the sea. Its multiple engines, powered by 159 coal furnaces, were designed to deliver 46,000 horsepower.

Compare that to today’s beastly mode of transport: the Boeing 777. Bangalore Aviation points out that a single GE90-115B engine puts out over 110,000 horsepower, or more than twice the design output of all the Titanic’s steam engines.

And that power is obviously hooked up to a much smaller vehicle. The Titanic had to carry 14,000,000 pounds of coal alone; the 777 has a total weight of only 775,000 pounds.

Here is more and for the pointer I thank Roland Stephen.


I am not an engineer, but maybe you just need more power if you want to fly something than you need for making a ship go on top of water.

You're right that you need more power to fly than to float, but power is Energy over Time...

Consider this:

An average Boeing 777 can carry ~300 passengers on a transatlantic flight in ~7.5 hours. The Titanic carried 3500 passengers/crew, and left Southhampton on April 10th, and was supposed to reach NYC by April 17th, or 7 Days.

If we assume 12 hours for the 777, to take into consideration turn-around time, the 777 could carry the same number of people from london to NYC /and back/ in the same amount of time.

Considering Fuel, here are boeing's stats for two variants:
777-200ER - Fuel 45,520 US gallons (171,160 litres). Passengers (3 classes) 301. Range 7,730 nms. Therefore 5.85 gallons per mile, 150 gallons per passenger.

747-400ER - Fuel 63,705 gallons (241,140 litres). Passengers (3 classes) 416. Range 7,670 nms. Therefore 8.3 gallons per mile, 153 gallons per passenger.

It's 5560 KM from Southhampton to NYC, or ~3000 nautical miles.

So.... 4 tons of coal per passanger, vs ~75 gallons of fuel per passanger.

I'd say that's much more energy consumed on the Titanic than the 777, per passanger. The 777 just uses it in a much shorter time. (Course, it's a teensy bit easier to move through air than water, and the ship itself was a bit heavier than, say 10 777s).

The container-ship Emma Mærsk might provide a more relevant comparison: Gross tonnage of 180,000 tonnes and total power of 150,000 hp.

Tonnage increase factor of 3.6 whereas a power increase of 3.2. The Titanic had a top speed of 23 knots; Emma Maersk gets up to 26 knots.

"it’s a teensy bit easier to move through air than water": but except at the end Titanic wasn't a submarine. It takes quite a bit of power to move along the water-air interface, because lots is lost in making waves. If you really want to drop the energy consumption of shipping, submarines is the way to go. Which effectively dictates nuclear motors too.

That ain't nessecarily true, not if the cargo has a density that is substantially less than 1.

Consider if you've got 1000m^3 worth of cargo, weighing in at 10000kg. You could transport it in a sub, but the sub would have a substantially larger wet surface than a ship that carried most of the cargo above the waterline. water-friction is substantially higher than air-friction, thus you could well end up spending more energy on the sub.

And that's not even counting the fact that subs cost atleast 1, frequently 2 orders of magnitude more to produce which will raise costs substantially.

If your cargo has a density of 1 (or more) though, you can't do better than a completely-submerged teardrop-shape vehicle.

Subs cost so much partly because they're built on defence contracts. Anyway, you're still ignoring the wave-making point. Come to think of it, a good solution might be a rather small tug towing a submerged cargo-dirigible. That would drop costs a mile.

What makes you think subs don't make waves?

There was a barge design that sailed the great lakes a century ago that sounds similar to this. Was called a Whaleback:


It's true: Submarines don't make waves and as a result it takes much less power than surface ships to drive them through the water at the same speed. Barnes Wallis was always promoting this idea.

Actually an A380-800 equipped with a Rolls Royce Trent 980 engine will generate roughly 566,660 horsepower in total, or about 141,600 per engine - an even more beastly output than the 777.

No, the 80,000 lbf thrust Trent 980 is decidedly not more "beastly" than the 115,000 lbf thrust GE90-115B.

Considering the A380 and a double decker and has 4 engines that amount of thrust is needed for the operations of the A380-800 where as theoretically the 777-300 can take off with just one engine working (although that will never happen) the A380 would need 3 of the four so thrust out put per Ron is better on the 777.

On another note, petroleum (which was just started getting used then for cars) out performs steam/coal on energy released

In 2000, I bought a VW with a nice dose of horsepower. This week, I just bought the same VW with the same number of ponies. The difference? My old VR6 engine got about 22 mph. The new 4 cylinder TSI with identical horsepower gets 34 mph. Small steps to a better world, indeed.

So are you the old fart that has been driving around my neighborhood at 22 miles per hour?

But you spent the same amount of money!! Stagnation!!

I'm assuming you meant mpg (miles per gallon) and not mph (miles per hour).

@Moser: Power requirements rise geometrically with velocity, which explains a bit about the massive growth in per capita energy use in the last century, and also the massive and quite effective drive for efficiency by jet-engine makers. I heard, a few years back, that some airfreight companies were looking into using massive prop-driven craft, instead of jets, for some shorter routes, the idea being that halving airspeed would only marginally affect delivery times, but would dramatically cut fuel costs. @JoeDog; torque is as important as hp--does your new rig have the same acceleration?

It has a smidget less torque but from a butt in the driver's sear perspective, it was indistinguishable. I took the salesman on a white-knuckle test drive. What little acceleration I may be losing is worth an extra 12 mpg.

There are at least three stories here: specific energy of fuel, overall efficiency of power conversion, specific and volumetric power of engines. All have improved radically from steam times. But economists beware: the next 100 years will not provide the same gains, physics being what it is.

I have a hunch that diminishing returns isn't an unfamiliar concept to economists.

Perhaps they're not, but dismal as they are as a group, they seem to place a lot of optimism for technological progress, perhaps because it is something of a get-out-jail-free card.

But the technology of moving physical stuff from place to place is pretty mature. Efficiency gains in the future may come more from NOT moving people are things that we used to move.

So the way to get around that is to stop having to move stuff around in the first place. That is part of technological progress. Who cares if there is no way to move faster if you no longer have to drive there.

Wrong economic concept. Diminishing returns refers to the effect on output from increasing an input, holding at least one other input fixed. This is a short-run phenomenon.

When all inputs are variable, the corresponding concept is diseconomies of scale. After some point, increasing scale results in higher long-run average cost. But that concept is inapplicable to this situation too because technology is held constant.

As Dave suggests, "Physics being what it is," there are dynamic diseconomies to increasing technology. As we advance in our technology, we will eventually bump up against limitations imposed by the physical properties of matter. Battery capacity and microprocessor speed might very well be approaching physical limits.

But we will likely find biological or subatomic solutions to those problems, pushing the upper bounds of technology thousands of years into the future. There may come a day when we see diminishing returns to technology, but it's so far into the future that we need not concern ourselves with it. It's more likely that an asteroid will render the question moot before that becomes a problem.

Power output is based on thermal efficiency and a gas turbine isn't the most efficient; diesel engines are. But to get the same increase in efficiency gain seen over the half century post Titanic would violate the laws of nature. The current maximum thermal efficiency of multiple expansion steam engines is about 17%, and only two of the Titanic's engines were triple expansion while the primary was a low pressure single stage (driven by "waste" steam from other steam consumers). Gas turbines are 35% efficient, while current equivalent marine diesel engines are 50% efficient. In other words, current ship engines are at least twice or three times the efficiency of the Titanic, so to achieve the same progress would require 70-100% efficient heat engines which exceeds theoretical maximums.

Combustion engines reached the limits of efficiency half a century ago. The only significant efficiency gains possible for combustion engines is in total system design to maximize engine efficiency and eliminating efficiency losses elsewhere in the system.

It depends on the technology employed. What is the energy-requirement for moving 1000 tons at airplane-velocity on a maglev-train travelling trough an evacuated tube ? Essentially zero except for what's needed to accelerate. (which can be partially regenerated on braking)

In principle, such a system could spend less than 1% of the energy current planes spend, while outdoing them in speed.

Engines as such though, won't see the same progress, since they'll bump up against the laws of physics sooner than that.

If you assume away friction, you don't need much energy to move things? Who knew!? But of course the energy required to lift millions of pounds off the ground isn't "essentially zero"

"But economists beware: the next 100 years will not provide the same gains, physics being what it is."

Which is also not much different than what people were saying 100 years ago. This is my favorite quote from the era:
"Everything that can be invented has been invented.. -- Charles H. Duell, Commissioner, US Patent Office, 1899"

Funny that. Back in the early days of computers there was an axiom that went: "never underestimate the bandwidth of a station wagon filled with magnetic tapes".

I guess in today's environment it would now say: "Never underestimate the bandwidth of an Anontov AN-225 filled with 64 GB microSDXC cards"....

micro SD card: 15mm x 11mm x 1mm = 1.65e-7 m^3 volume

cargo volume, AN-224: 1300 m^3.

So, 7.88e9 cards.

Assuming 64 GiB each, that's ~5.41e20 B, or ~470 EiB.

AN-225 cruises 800 km/h, so say, on a "link" from NY to LA, that's 3933 km, so ~17,700 seconds.

That works out to a bandwidth of about 27 PiB/s. Pretty wretched latency, though.

Not bad. Note that I left out takeoff and landing and the time to load and unload the ship as well as copy the data to and from the cards. I think that would cut down your bandwidth pretty radically.

A better plan might be to install hard drives in SAN/NAS boxes on the aircraft, and be able to hook up directly to networks at endpoints.

Better still-install the SAN/NAS arrays in a cargo-carrier constructed enclosure and if the primary site goes casters-up, quiesce the SAN/NAS's at the remote site, make the call and wheel out the cargo boxes for direct loading to save time of packing. Once the cargo lands, truck it to the site, drop it in, cable it up, run the usual diagnostics and bring the critter up.

You'll probably have to run a good vibration modes analysis.

The axiom is still true. If you need to restore a few Terabytes of data on a crashed server from an offsite backup location there's no way for an internet link to compete with FedEx next day delivery.

Yep. Fedex and harddrives my preferred method.

If "a few" ain't very many, then yes there is. A 400 Mbps link does 50MB/second, or 3 gigabyte/minute, 180 GB/hour. A 2 terabyte restore is thus 11 hours, which beats the FedEx-delivery, especially since clever filesystems can restore the data you need first first, odds are that you'll be able to come back up with less than 10% of the data restored - if the remaining 90% of the data is available at higher latency from the backup-site. And 400 Mbps ain't a lot - I used that example speed because it's the fastest speed offered by my ISP for residential private customers.

So yes, for "a few terabytes" some network-links can compete. If we're talking *many* terabytes however, then FedEx still wins.

Which filesystem does the restore-data-you-need-first trick? Sounds cool.

The Titanic carried a ton of non-human cargo, did it not?

Many tons, actually.

The 777 carries non-human cargo too, but you raise a good point.

Rahul offered a comparison with a modern E-Class freighter. These ships could, theoretically, carry passengers as well as cargo. But it's not economically efficient for mixed use. Airplanes are the most efficient passenger carriers, and ships are more efficient for cargo that doesn't absolutely, positively have to be there overnight.

Check out Nanofuture Chapter 6 (title: Engines). We are at just about the 300th anniversary of the Industrial Revolution's beginnings, and there has been a smooth exponential improvement of power-to-weight ratio of engines over the entire period. The improvement of efficiency has been more complex, since perforce it has to be asymptotic to 100%, but it has shown a comparable improvement given the constraint.

By the curve, 2050 should see a 50,000 HP engine that fits in the palm of your hand. We can design one today, but it requires the technology to place every atom and covalent bond in the device precisely.

Thats hard to imagine.

The big leap came in the 1920s with the "Aerodynamic Theory of Turbine Design" by Alan Arnold Griffith.

Actually, most smaller (i.e. non-aircraft carrier) U.S. naval vessels use marine versions of aircraft engines as prime movers. My Spruance-class Destroyer had four GE LM2500 engines, each with about 25,000 horsepower. The more modern Arleigh Burke class destroyers have similar set-ups.

The power of jet engines is measured in pounds of thrust, the GE-90 engine produces over 100,000 lbs of thrust. The engine would only produce 110,000 HP with the parking brakes set. As soon as the plane starts moving the horsepower declines dramatically. Incidentally, I fly for one the major express freight carriers, and I have never heard any mention of the possibility of flying slower planes. It is impossible to get from the West Coast to the middle of the country, sort freight, and continue on to the East Coast in one night without flying as fast as possible.

Not directly related, but I read yesterday that the original Newcomen engine was about 0.5% efficient. It's amazing to believe that it moved itself, let alone did any useful work. By comparison, the engine in your car is probably around 40% efficient and stationary power plants can get well over 70. 140x increase in 300 years. Not bad.

" the engine in your car is probably around 40% efficient and stationary power plants can get well over 70."

Not possible. The mobile diesel engines get 35% peak thermal efficiency, and combined cycle stationary peak at under 60%.

Or are you stating the efficiency relative to theoretical energy efficiency which for a car is 40% of about 70% theoretical efficiency or about 25% thermal efficiency....

Seems to me one would use the cruise liner Oasis of the Seas for comparison to the Titanic, which is more efficient than a 777 in its engine efficiency by 40-50%.

"The main engines of the ship are 6 diesel engines - 3 x Wartsila 16-cylinder engines, which are giving 24,290 hp each and 3 x Wartsila 12-cylinder engines, which are giving 18,590 hp each. All engines and diesel common rail and having total heavy fuel oil consumption of 2,400 gallons per hour. Large part of about 97,020 kW from these main engines goes into power for electricity for the hotel part, just like lights, electronics, elevators, galleys, water treatment plant and other systems, which are used for the operation of the vessel and propulsion. The propulsion is provided by three "Azipods" with total power of 20,000 kW."

"The world's largest diesel engine is currently a Wärtsilä Sulzer RT96-C Common Rail marine diesel of about 108,920 hp (81,220 kW) @ 102 rpm output."

By the way, credit NASA for funding the development of energy efficient turbofan designs at GE in the 70s which became the basis of the GE90 engine series. But we should credit the huge public investment in jet engine design and development of the manufacturing and service industry that produced the jet aircraft passenger and cargo services of today.

Comments for this post are closed