In short: this is an idea that should be discarded before you spend money on the trials because in practice it won't work, and that's before you get into the interruptions every time you want to open up a line for a cable upgrade, the problems involved in fixing anything inside a conduit carrying water under pressure, the inability to enter and exit the conduit at random spots and the various obstacles (seams, older welded joints, socks) that you'll encounter that will make inserting a continuous run of a cable through a water system already in place that wasn't designed very explicitly for that purpose.

My first thought was "eh, more plastic leeching crap into the water", but then I remembered that all modern water pipes are already plastic.. too late for that.

Most importantly, I was under the impression city-wide water pipes are rarely pressurized, the water 'runs' through it with a substantial air gap - that means entry/exit points are easy, and nothing actually needs to be in contact with the water. Is that not the case?

How would unpressurized water go from a pipe under the street to your shower head?

To be clear, it is not the case. Water Towers are elevated specifically to provide pressure.

I mean, besides making it a biohazard to repair.

But it seems there are at least a couple of working installations of that type in the world and companies offering it commercially instead of in 'perpetual trial' mode. Repairing those must be a pretty shitty job.

17 km is a pretty good stretch, that means it won't be a single run but quite a few of them with entries and exits every time they need to make a splice. It's also going to be pretty rare to find a stretch like that that just so happens to run exactly where you want fiber to go (and a nice little problem on how to route the cable around those splices so you still respect the minimum bending radius of the fiber for all phases of construction and use). Rural would be the most likely candidate, long runs, few splices. If the get the POC to work reliably across a couple of years without the water utility ripping it out at some point because it causes them trouble then more power to them, but I would not bet on this.

Also: the force of a bunch of water flowing along a fiber should not be underestimated, you won't be able to do that for a very long distance before the pull overcomes the breaking strain of the cable. It might cause the cable to get pinched on turns and rub against the inside of the turn potentially abrading the outer shell as the water flow rate varies considerably between peak and off time usage (and contaminating the water supply with the debris).

In the end the reduction in construction costs is a savings for the company that wants to lay the fiber but they will have to pay compensation/rent to the water company for being allowed to use their conduit, probably have to indemnify them for any (long term) damage and every time there is an issue with the water system they run the risk of having their stuff broken. 17 km is also long enough that you'll be passing more than one set of baffles (which are used to reduce the effect of waterhammer on the valves and the pipes), I wonder how they plan on dealing with those or if that's going to be more entries and exits.

The only remotely feasible solution that I can think of is a pipe-within-a-pipe and then to blow fibre through conventionally, that would give you the rigidity required for mechanical integrity and it will probably seal well enough that you can do it with water pipes under pressure. That's not going to be cheap but it just might come in cheaper than trenching or burrowing using remote controlled moles, it will of course still reduce the available capacity to carry water and require entry/exit every time there is a piece of infrastructure in the water system.

This is not correct. While PEX pipe has become popular, a lot of homes are still built with copper plumbing. And lots of municipal water infrastructure contains a lot of metal.

(ShitNet! Up Your's! (SCNR))

This is an incredibly pessimistic and counter-productive point of view. "You shouldn't even try because it'll never work" will never get you anywhere apart from the most obvious and incremental advances.

Even if it turns out that only 10% of the water infrastructure in the UK is usable for this sort of expansion, it has the potential to be a useful technique. Plus if it does turn out to be a relevant then pressure can be applied to the water companies by the government (using legislation if necessary). Expansion of fibre broadband is a political positive at the moment.

Being pessimistic about bad investments isn't counter productive from the point of an investor, it is a net gain because then you get to spend your money better.

"Wouldn't it be nice if we had 'x'" is rarely a good reason to invest in something, see 'Theranos', 'UBeam', the 'Moller Skycar' and a ton of other things that it would be nice to have but that are not feasible in practice.

> "Even if it turns out that only 10% of the water infrastructure in the UK is usable for this sort of expansion, it has the potential to be a useful technique."

Where does that 10% come from? All I see in the article is that they are trialing this and no mention of any percentage of the water infrastructure that might be suitable for this is made. See my other comment for one way in which you might get this to work which avoids some of the downsides but it still won't be cheap or easy to maintain long term, and it will always require coordination between the utility and the fibre company in case of a water mains issue or a fibre issue.

Edit: I've found an old article detailing a trial in Bonn, Germany, I wonder what came of that:

https://technical-news.net/bonn-fiber-optic-cable-laid-throu...

So this is definitely not a 'first' as many articles about this project say.

TL;DR: AFAIK, there is not a single trial run that has found sewer to be the best option for any of the following metrics: initial installation cost, ease of installation, lifetime cost/TCO, maintenance frequency, and network downtime/reliability rates.

I personally lead the effort to see a Fibre-to-the-home (FTTH) network deployed in my hometown of Quincy, Illinois. While this was eventually successfully implemented (with no thanks to any external firms, all done locally by our rural county telco coop) we had a false start (https://www.govtech.com/archive/quincy-running-fiber-through...) thanks to i3, the same chuckleheads who scammed Bournemouth through their Fibrecity effort. I likewise ended up with egg on my face, as the public figurehead for Quincy's fibre campaign (https://khqa.com/news/local/bringing-google-high-speed-inter...), but my only concern was the wasted time and money, and loss of public trust, for our community and its efforts.

I would simply recommend all involved be highly cautious; not because any of the underlying technology here is untrustworthy (although I'm skeptical of the technique for many reasons, most of which were well outlined by Jacquesm above) but rather because the underlying firms have their history rooted in nearly two decades of unfulfilled promises and wasted taxpayer funds.

https://archive.google.com/tisp/press.html

i3 is an infrastructure company that do this. They trialled a sewer-based fibre network in Bournemouth where I grew up back in ~2010, although it never quite got past the trial stage due to contract issues. They've done similar trials in South Africa, Australia, US, and elsewhere. Not sure if any have rolled out to full scale usage but it's been long enough that I'd expect some have.

There is a company that has been working on tidal energy for more than two decades now, they are always trialing stuff, it costs millions and it has so far produced less power than you'd need for a lightbulb. And yet they keep getting away with it. This may well be such a scheme, after all, how many complications can there be before you've seen the bulk of them and you can decide to roll out at scale and for which situations the tech is suitable (and for which it isn't).

Next few years they go to cheapest supplier in China who has trace toxic chemicals on the wrapper because they bypass the final wash process to save a few cents or their upstream supplier changes out the chemical mix to something cheaper.

But you should be using Reverse Osmosis or distilled drinking water anyway given all that's going on with lack of regulation, review or enforcement.

What are the supply considerations for pipes, anyway?

If you need to close your water main, does it not close all the way? Does it cut the fiber? Both?

Well, they are probably called sluice valves in lots of other places, but the UK specific thing is the signage.

Keep your eyes peeled for telltale SV signs (shown here alongside a fire hydrant indicator):

https://c8.alamy.com/comp/DA5TFN/fire-hydrant-water-locator-...

Maybe they plant to use the waste lines which shouldn't have valves on them past a certain point.

[0] - https://en.wikipedia.org/wiki/Cured-in-place_pipe

https://news.ycombinator.com/item?id=21070351

From that thread:

> Some of these fancy bags (e.g. from tea pigs) are made from polylactic acid. This is derived from corn starch and used in temporary medical implants (it degrades to lactic acid) so hopefully it's safe.

https://news.ycombinator.com/item?id=21072973

I'd love to see this technique become widespread. It's so sad that even in the >1 million person city where I live there still neighbourhoods without fibre-to-the-premises in <current year>.

That said all of the other complexity remains.

I think this the sewer is more suitable approach. One company in Czech republic manage to pull it off after more then decade of research.

Article in czech, but there is a lot of pictures to get the idea.

https://www.lupa.cz/galerie/pilotni-projekt-unikatni-instala...

Over half the cost in wiring a house with fiber is between the curb and the house.

Virgin Media (the other provider) just dug trenches in the pavements about 20 years ago.

Added bonus of having overhead power lines? Your lights flicker when it's windy...

And I guess depending how it’s installed it may not be that hard to pull new fiber through without actually digging anything up, just from access at two ends.