The Paris Compressed-Air Power Network (2018) | Hacker News Reader

The Paris Compressed-Air Power Network (2018)

douglas-self.com

157 pointscamtarn35 comments

Paris also used to have an incredibly extensive network of pneumatic mail tubes: https://www.cix.co.uk/~mhayhurst/jdhayhurst/pneumatic/book1....

as did several other cities (NYC, Philadelphia, Prague, ...) to a lesser extent: https://en.wikipedia.org/wiki/Pneumatic_tube

Another interesting method for power transmission in the early days of industrialization was the line shaft (https://en.wikipedia.org/wiki/Line_shaft). These were used only for a single factory, not for a whole city, though. You can see these if you go visit old textile plants in New England, for example.

Also hydraulic accumulators:

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

https://99percentinvisible.org/article/running-on-water-the-...

Amusingly, modern hydraulic accumulators of a different sort may still have a role in power generation. I'm working on a couple of wave generator projects where the waves pump hydraulic rams, storing pulses of pressure in hydraulic accumulators, where it's then used to drive a generator at a more constant speed.

Intresting, though wave power is a minusculy small overall altrnative energy source with massive capital feqirements. There are better options to pursue.

Pumped hydro is ... after a fashion ... a vastly more appropriate hydraulic accumulator, as is CAES.

Back when I did a mechanical engineering BTECH (higher vocational UK quali) in the 80's - we still learnt how to design a belt drive.

Would belt drive systems still be covered in mechanical engineering?

The 4kW laser cutter I operate for work has (toothed) belts in it, so presumably the drive systems are covered under mechanical engineering?

Belt drive systems are still an essential mechanical design component, both for power transmission and motion control, and certainly still covered in university classes these days.

Yes. You'll study the shit out of old school flat belts because V-belts, serpentine and cogged belts require more math that nobody wants to grade. Maybe you'll do one problem calculating how tensioner placement affects the force needed to make a cogged belt skip teeth.

There's also "Stangenkunst" for long-distance mechanical power transmission:

https://www.lowtechmagazine.com/2013/01/mechanical-transmiss...

I stumbled upon one of its remaining manhole covers a few weeks ago, which led me to discover about this network.

I have been living in Paris for 10 years but never knew this existed before.

I loved this. Really interesting. Some of the French translation is a little off IMO. I would say that "usine" probably translates better as "plant" or "factory". "Particuliers" means individuals in this context. So these are supplies to individuals rather than businesses (theatres, cafes etc). I really enjoyed it nonetheless

I found the same with the translations, I was looking for a contact/feedback option to tell them, but couldn't find any way to contact the site owner. Throughly enjoyed this anyway! PS, I believe "usine" translates to "Oven" or "furnace", at least that's what I've learned in high school (French wasn't really my strong suit though), which makes sense in that's where the coal is burned to generate heat>pressure>power.

Unrelated, but as soon as I saw the domain name I got excited. This is indeed the same Douglas Self that has published a ton of interesting information on audio amplifiers. Really fascinating info all over his site, both about electronics and other random things (like the compressed-air power network)

He's one of my favourite ex-coworkers. His knowledge of audio electronics is superb and his thoughts on subjectivism are worth a read.

In contrast, the worst that could happen with compressed air was a burst pipe; this would not be a good thing to be standing near, but since the pipes ran through the sewers the chance of this was very small.

Even more reason to avoid standing near a burst pipe.

Actually its a burst reservoir that is nasty.

I was told at college that in the event of the small full service (electricity water and air on tap) lab block's compressed air reservoir blowing it would take most of the building with it.

Steam is even worse BTW

But reservoirs pretty much never burst unless relief valves fail or oil gets into them.

(inb4 a bunch of hand wringing about rusty compressor tanks, those just leak and are generally harmless)

Exactly the lecturer was making the point about how those lessons had been learned - and it was in thermo dynamics class so we where doing a lot of work on steam tables and the like

This is what the ASME was formed for - to create standards for safe pressure vessels. At the time it was founded, there was just over one steam boiler explosion a week in the US...

Does anyone know what pressure came out at the consumers end? The article mentions a 10 bar (145 psi) steam engine. Most modern air tools require about 90 psi.

I wonder what the efficiency of the system was. I am sure a lot of air leaked all over the place.

Compressed air is pretty terrible at energy efficiency. The main source of energy loss is due to heat at compression. When you compress air, a lot of the energy ends up just increasing the temperature of the compressed air and this is then quickly lost in the transmission process.

Of course! I forgot about the loss during generation. I have had to deal with my fair share of overheating compressors.

No mention of drying the air in the network. They had to have good driers to prevent water from entering the network or elaborate number of venting points to remove water from the low points of the network. Anyone know more?

I'm no expert on this, but could this be the reason of the confusion in the picture titled "Section of the Quai de Gare usine: 1892", where the author mentions "A = drying channels (not sure what that means- they look like sewers in cross-section. May just mean 'drainage')"?

Routing it all underground is going to do a really good job drying it at the expense of tons of condensation in the lines. The latter is easily solved with drain valves and traps like you'd find in a steam system.

This website is awesome. Internet should be like this.

I've been poking around through it and was going to say likewise.

Yes.

It's been featured on HN periodically: https://news.ycombinator.com/from?site=douglas-self.com

I really enjoy the section of the site on engine designs; some of them are even animated, like this bit:

http://www.douglas-self.com/MUSEUM/POWER/unusualICeng/axial-...

Previous discussion here: https://news.ycombinator.com/item?id=19442938

Whenever I read about ancient technology like this, I wonder if any of the equipment and infrastructure survives today.

You usually end up with overlapped left over portions of old networks. Paris is a good case study for that, the entire underbelly of the city is an nightmarish amalgamation of tunnels and pipes.

You have that in basically every city after after a certain period of time. In Europe even more as some settlements, like Paris, go way back. The city I live apparently can be crossed undergroung in everything ranging from Roman relics to whatever was added. The historic center at least

This is one of the best HN articles I've red in a long time - a fascinating look at a very well-developed infrastructure system in one of the most technologically advanced cities of its day.

Today I learned something about the city where I'm living. Thanks HN!

Reply on news.ycombinator.com