About Those Expiring Fiber IRUs and Empty Conduits

July 1st, 2011 by · 8 Comments

One of the themes I am increasingly hearing about from both inside the industry and the investment community is just what it means that many of the longhaul fiber IRUs that were sold back in the bubble are now at or beyond middle age.  You know what I mean, the fiber connecting the major cities that got swapped and traded and eventually purchased out of BK or some other restructuring 5-10 years ago, but which powers much of the internet today.  Will those IRUs be renewed and at what price?  And if not, what happens?   There are two sides to the coin. 

First off is the love/hate relationship the industry has with its own intercity dark fiber.  Those that have lots of it (Level 3 and CenturyLink/Qwest for example) really love having it, but hate selling it (or having sold it) to potential or actual competitors.  Those that don't have it love having picked it up cheap, and hate the thought of building out more of it.  Unique routes are valuable, and selling intercity dark fiber sacrifices that uniqueness.  The builders of it knew that back then, but in '98 we all thought there were enough bits for every bandwidth visionary to be a billionaire and so they shared to get there faster.  Now, the current owners of all that intercity dark fiber have spent years re-consolidating it all to help rationalize pricing.  Why would they sell again that which they wish they hadn't sold in the first place?  It depends on the actual renewal clauses of the IRU contracts of course, but there will be negotiations and overall attitude will matter.

But there's also the age issue.  When those IRUs expire, the fiber inside will be twenty years old and we will be fully in if not beyond the 100G era.  No matter how you shake it that fiber will be old if it isn't already, and to get what we need from 100G and beyond, it's going to become increasingly undesirable.  So it could be that those who own it will be simply ecstatic to get paid another 20 year IRU on the stuff - so perhaps renewal won't be a problem.  The real question might be where network operators can get some *new* fiber.

Which brings us to the conduits.  Qwest put three in the ground, Level 3 put in a full dozen on its original US build and inherited a few extra that derive from the WilTel and Broadwing buidouts too I believe, and AT&T has the former Velocita assets.  All were put in back before DWDM turned the fiber world upside down making it irrelevant how much intercity fiber you had - a couple fiber pairs was enough for almost anyone.   The extra conduits have therefore been totally irrelevant for more than a decade - but their time may finally be coming.  Assuming they have survived the last decade or more intact, it may soon be time to blow new fiber into them.

But those that do blow that new fiber will perhaps not be as sharing as they were back in the bubble, right?  That is one of the big drivers behind Hunter Newby's Allied Fiber and its buildout plans.  He's hoping to sell piles of dark fiber to big network operators with neither conduit nor interest in digging tens of thousands of miles of trench to get some.  Many were perplexed by Allied Fiber's plans, but viewed on a longer timescale they are perhaps not crazy at all.  But of course, they too will want to be paid well - so it's not as if they'll be selling it cheap, the key is that they'll be actively selling it where others might be substantially more reluctant.

While the metro business has been booming and folks are building on many fronts with solid economic underpinnings, it has been a very long time since who owned and who leased intercity longhaul fiber or conduit mattered much at all - despite the tens of billions that went into building it all.  That mismatch can't last forever, the pendulum has to swing back eventually and bring the cost of building such assets and the value derived from them back into balance.  The demands of 100G technologies and beyond along with IRU expirations will almost certainly be the trigger.

What do you think?  Whose intercity fiber will we all be using in ten years?

Categories: Fiber optic cable · Internet Backbones

Join the Discussion!

8 Comments So Far


  • mark says:

    do you have any guesses on what % of previously installed fiber routes support 100G and which don’t? how does one know if their fiber plant is 100G capable?

    • Rob Powell says:

      I think it’s less binary than that. You might be able to do 100G, but the number of channels may be limited or the cost to install or operate it may make it not helpful. Or the fiber itself may have survived the years poorly on some sections in ways that make it less usable but not others. Hard to put a simple number on that, not that I could put a complex one on it either.

  • Anon says:

    DWDM allows for more capacity over same fibers. more fibers allows for more flexibility, laterals, splices and new routes. extra ducts are “assets” but only if one is willing to sell or otherwise extract value from them.

    at the end of the day, little of this matters. what matters are customers willing to pay for the services that the assets enable. on this score, there only seem to be a couple of LH carriers that can attract, service and renew enterprise customers at a profitable clip.

  • Brad says:

    Back when we were puting in Composite fiber people thought we were nuts also. Now all of OCG’s fiber is good for 100G and is being used for it now.

    Allied’s build now will come to the rescure of mano out there that are on the old fiber and old IRU’s

  • mhammett says:

    One option (should the route owners accept it) is that old fiber that is *only* capable of 10 gig would be a boon to the independent ISPs. The cost to maintain that sort of system would likely be prohibitive to the large carriers, but would provide immense opportunity for we little guys that could spin up 1G and 10G segments going back to the new routes the big guys installed.

  • Ciena claims their equipment will do 40/100G over existing fiber:
    “With our adaptive optical engine technology, these platforms automatically adapt to a wide range of existing fiber plants, making 40G and 100G as easy to deploy as 10G”
    I’m also seeing a move toward meshed networks. For example wavelengths from SF-NY and SF-Chicago-NYC on the same fiber IRU. This reduces the number of IP hops for SF-NY traffic, improving reliability & reducing IP routing costs.

  • Brad says:

    Isnt it funny….Everyone went crazy with all these build outs new eqpt etc. But when it comes down to it the less hops and more simple reliable equipment the better.

    40G has very tight requirements to operate over long distances. Better off installing a 100G system instead.

    Ples alot of the old network has mechanical splices and/or high loss connectors. If you dont address them 40 or 100g will not work. ORL is one on the new test requirements we are seeing with all of our customers. If you have bad ORL (Connector Reflectance) then you create noice at the transmit end.

    But fiber is a mediul. If you can run something over it it can produce revenue…..We all know about revenue….

  • Chicago484 says:

    I’d be interested in seeing your current thoughts on the IRUs and empty conduits.

Leave a Comment

You may Log In to post a comment, or fill in the form to post anonymously.





  • Ramblings’ Jobs

    Post a Job - Just $99/30days
  • Event Calendar