Liberty ships

This is how the US government came to own ships named SS Stage Door Canteen……

This very example is something that\’s running around my head at present.

Say there was a plan (entirely ephemeral, unlikely to be ever realised, but something that is at least possible in the physics of this universe) which would require putting 160 million tonnes a year of stuff through furnaces.

Currently big furnaces cost $300 tonne per year capacity (so, a 500,000 tonne a year capacity furnace is $150 million), are cheap to run (relatively) and take 2-4 years to build.

OK, so over the years you need to expand out to 320 of these furnaces (I said it was unlikely!).

Now, would it change the game entirely if you went for mass production techniques of furnaces? Let\’s say we went for 50,000 tpa furnaces. We now need 3,200 to build out. If they\’re all at $300 tpa then, well, we\’ve got our capital requirement rather more granular, possibly easier to finance, true, but almost certainly at the expense of higher running costs.

But what if that very mass production technique took our production costs down to $200 tpa? $50 tpa?

Now, yes, obviously, if you knew what that mass production cost was then working out whether to do it or not is simple, a few numbers into a spreadsheet and bingo!

But what does anyone reckon the discount for mass production would be?

Currently our $300 tpa (umm, the world\’s not mine) furnaces are built at the rate of one or two globally each year. What if we decided to build the smaller ones at a rate of one a day? Or the large ones at 3 a month? (to give an equal decade long roll out for the globe).

What does anyone think the savings on the $ per tpa number would be from proper, one design and one design only, entirely interchangeable parts etc, mass production would be?

Liberty, or Model T furnaces?


5 thoughts on “Liberty ships”

  1. ….What does anyone think the savings on the $ per tpa number would be from proper, one design and one design only, entirely interchangeable parts etc, mass production would be?……

    I think massive would be the answer.

    You would have to overcome massive inertia though, and all the competition attacking your “ill thought out plans” because such a scheme would destroy a huge amount of their profit making potential.

    Is there any technical reason why it might not work (raw material specifications for example)? In the oil industry, feedstock characteristics are extremely important.

    A little market research among small mining companies might be called for.

  2. Having done this exercise on oil rigs I’d say you really want to get the size down and volume up. The real benefits don’t kick in until you can automate manufacturing to a large degree and for that you need the larger volume, smaller unit design.

    How much you can get the price down is questionable. A furnace is a big chunk of raw material with a small amount of control system. Your base line cost isn’t going to move no matter how many you build and it’s going to be a much higher percentage of a furnace than in, say, a car.

  3. A fairly common rule of thumb is that the production cost drops 10-15% for each production doubling. So, if your first unit costs 100, the second will be about 90, the fourth 81, and so on. I understand it comes from the aviation industry, when the likes of Douglas and Lockheed had to bid for 50,000 unit production contracts when they only had a prototype.

    As David Moore says, though, this is based around fairly complex units, and is the result of cumulative learning effects. If I were you, i’d remove the raw construction material costs from the equation and apply the rule of thumb to everything else to get a better idea.

    Tim adds: So, over hundreds of units starting at $300 tpa we’d be asymptotically approaching $30 tpa?

    I think you’re right/ Raw material costs would have to come out before that could be possible.

  4. Look, we all know that come the revolution there’ll be a hell of a lot of socialists/social workers/Guardian columnists/……..(enter group of choice here)……. dangling from lamp-posts & hence a big clean up problem but isn’t this getting a bit extreme………..

  5. Oh, come on!
    Why the hell would anyone build a 500k furnace if it only cost 10x a 50k furnace and production costs per ton of product were the same?
    Elementary O level physics – heat loss is proportional to surface area not volume. Although structural members need to be sturdier for larger buildings, cost per ton declines from 1 ton pa for a long time before it starts to rise again (I don’t have the data to hand for steel furnaces but when I was a kid (admittedly in the dark ages) that was 300,00o0 ton/year for ammonia plants – ammonia is lighter than air so for heavier-than-air metals it might be a bit bigger than 300k tons rather than 50k.

Leave a Reply

Your email address will not be published. Required fields are marked *