Uncertainty as the Telecom Boom’s Fiber IRU Flurry Closes in on the 20-year Mark
February 13, 2013 1 Comment
• Claims of network providers entering flux as telecom boom’s IRUs expire are important for wholesale; enterprise services should be unaffected.
• The telecom boom’s national fiber builds are now well over a decade old: The fiber’s transport characteristics remain serviceable today.
Starting anytime now, and through the end of the decade, we just might see network operators’ national footprints widely thrown into disarray, as the host of IRU agreements signed during the telecom boom of the 1990s begin to expire (the standard length of an IRU being 20 years, though they can vary). Some observers speculate that, just like the baby boom set the stage for a next-generation ‘baby boomlet’, expiring IRUs will spur a fresh round of telecom negotiations and deal-making. These new deals would buoy transport, network and infrastructure providers. Personally, I expect the freeing up of IRUs to be a relatively minor event. But on the other hand, there still appears to be room for strategic fiber investment.There are two important background points to understand when discussing the telecom industry’s IRUs:
First, there are two different types of IRUs. Capacity IRUs typically cover static lit services, which in the mid-1990s meant a wavelengths of up to 2.5 Gbps SONET. Fiber IRUs cover dark fiber; operators with dark fiber could participate directly in the explosion in capacity that fiber provides, adding gear as needed. Capacity IRUs are a small part of the market these days: Those older wavelengths were pricey to buy at the time, and have probably run their course. Operators who are exiting capacity IRUs free up dollars to spend elsewhere. Fiber IRUs are trickier because of a second important point.
Second, the mid-1990s saw a sea change in which operators stopped using long-haul fiber designed to carry just two or three wavelengths, in favor of non-zero dispersion shift fiber (NZ-DSF) that could support 80 or more wavelengths. Fiber manufacturers have made refinements and improvements on NZ-DSF manufacturing in the past 15+ years. But “old” Corning LEAF or Lucent TrueWave fiber from 1996 and beyond, when architected properly, is still capable of handling 80 or more wavelengths of 10 Gbps, and usually are upgradeable to support 40 Gbps and 100 Gbps wavelengths.
The crux of the second point is that the industry generally estimates its infrastructure will be obsolete in 20 years. In the past 15 to 20 years, transport technologies cycled quickly while new generations of fiber made incremental improvements – but the “old” long-haul fiber remains quite capable (and valuable). What happens to still-precious Fiber IRUs when the timer expires is up to the specific wording of the contract, or for the licensor and licensee to negotiate: in the case of fiber swaps, each party is both licensor and licensee.
In the end, we believe that most expiring fiber IRUs, even where the go-forward terms are not clear, will be resolved smoothly. The parties had plenty of notice to discuss and resolve their issues. While there are only a few national dark fiber providers, there are other partners with long-haul routes to tap for leverage. Payment terms may change, but national providers will not be forced to de-light and abandon routes – that is, unless they choose to do so for dramatic effect.