Property researcher BIS Shrapnel has shone a torch into one of the dark corners of the NBN debate: the cost of maintaining Australia’s copper network. The company’s nearly-$AU20,000 report into infrastructure maintenance in Australia, Maintenance in Australia, 2012-2027, tags the annual cost of telecommunications network …
And thanks to Richard ...
For bring this torch out of whatever dark hole it was hiding in so we can all see it. It's nice to have numbers to quantify things.
How much does it cost...
... to fix fibre vs copper when Fred Backhoedriver sticks his bucket through the trench after not dialling before he digs?
Not being smart - I'd really like to know what's cheaper/faster to fix.
Re: How much does it cost...
First you need to repair the mechanical damage to the trench or duct - that costs the same, regardless.
Joints are the weak point of any cable so you'd normally replace the entire length back to the existing joints rather than try and stick what's still there back together.
Fibre is cheaper than copper, but the cost of having trained people available with the right kit is, longer term, a greater expense that the cost of a reel of cable. If there are high bandwidth services with guaranteed SLAs running over the fibre the incident cost of a fibre break might actually be higher if you can't fix it in time. At 11PM on the night before a public holiday it could take you some time to get the right people there to deal with it.
Net result - I'm not sure it changes the economics much for an "unplanned mechanical intervention". Where the economics do change is that all things being equal fibre is less susceptible to damage from water ingress and electrical storms. Copper networks tend to have lots of 'partial' faults - poor joints that affect DSL throughput or cause crackling or noise on audio paths - they're devils to find because disturbing the connection will often fix it., temporarily (ever made a wonky pair of headphones work again by twisting the lead about). Those faults don't occur in optical networks and poor connections cause total failure which is much easier to locate and remedy.
The ultimate service life might be shorter though - if in a decade we're using unheard of bandwidths and the fibre in the ground can't support the wavelengths required, it all has to be ripped out or augmented. Early fibre (1980's) had a tendency to go brittle or opaque over time - I presume it's better now than it used to be.
Re: Re: How much does it cost...
A couple of things to consider.
Late in the 1990s, Corning (I think) tested a 15-year-old fibre retrieved from the Seattle mountains after repeated sub-zero winters and regular flooding during the melt. Their conclusion, after destructive tests, was that the glass showed zero observable deterioration.
Today's glass could be presumed to be better than that.
Second: We're nowhere near the capacity of the physical medium yet. Single fibre capacities are still restricted by terminal equipment; the highest transmission I'm aware of was 100 petabits.kilometer (ie, longer distance, lower speed) by Alcatel a couple of years back.
Re: How much does it cost...
Temperature is less of an issue than some physical damage that can occur. Mercury's early trunk network suffered repeated failures with the fibres - they were placed alongside railway lines in trenches, and the movement of trains passing by caused slight whipping which over time broke the joints. Widespread deployment of fibre in urban environments is relatively new and there may be mechanisms of decay we've yet to discover. There's lots of 'interesting' stuff that comes out of vehicle exhausts for example - do we know long term what effect that might have on urban fibre deployments? Low-level leakage of fumes and liquid from petrol stations has, over time, a damaging effect on the plastic sheathing of copper cables in the vicinity - we might see a similar mechanism that ages fibre cable.
You're right that we're far from reaching the physical capacity of fibre, but it does work over a relatively narrow range of light frequencies. We might find that we can use cheaper terminal equipment with a new type of fibre that can easily handle a wider bandwidth, or perhaps use existing technology in a different configuration. The economics will determine how the use of fibre evolves and I'd be reluctant to say that what is in the ground now is without doubt future-proofed and will be there for as long as copper has been. It need not even be related to the physical components themselves - we might find that a different topology is more cost effective when we have a more settled view of who uses fibre and how. If deployment is to millions of premises even a small improvement in efficiency has a big impact on the companies operating these networks.
fibre - I want it
I'm stuck on the end of several km of copper so that my adsl2 provides 5 mbit/s. I'd be really happy to buy a fiber switch, which would also save media conversion costs at the Telco end.. Dedicating a few kb/s to voice from a fiber link is no hardship.
And yes, when it rains (in Melbourne) my internet used to drop out all the time. Things have improved since they replaced the wire from the road to my house and I gained an extra mbit/s when they replaced a a few yards of cable after a branch fell on it in my street. I suspect there is some really dodgy copper cable around.