I think every client who is considering building a fiber network asks me how long the fiber will last. Their fear is having to spend the money at some future point to rebuild the network. Recently, my response has been that fiber is a hundred-year investment—and let me explain why I say that.

We’re now seeing fiber built in the 1980s becoming opaque or developing enough microscopic cracks that impede the flow of light. A fiber built just after 1980 is now forty years old, and the fact that some fiber routes are now showing signs of aging has people worried. But fiber cable is much improved over the last forty years, and fiber purchased today will avoid many of the aging problems experienced by 1980s fiber. Newer glass is clearer and not likely to grow opaque. Newer glass is also a lot less susceptible to forming microcracks. The sheathing surrounding the fiber is vastly improved and helps to keep light transmissions on path.

We’ve also learned a lot about fiber construction since 1980. It turns out that a lot of the problems with older fiber are due to the stress imposed on the fiber during the construction process. Fiber used to be tugged and pulled too hard, and the stress from construction created the places that are now cracking. Fiber construction methods have improved, and fiber enters service today with fewer stress points.

Unfortunately, the engineers at fiber manufacturers won’t cite a life for fiber. I imagine their lawyers are worried about future lawsuits. Manufacturers also understand that factors like poor construction methods or suffering constant fiber cuts can reduce the life of a given fiber. But off the record, I’ve had lab scientists at these companies conjecture that today’s fiber cable should be good for 75 years or more if well handled.

That still doesn’t necessarily get us to one hundred years. It’s important to understand that the cost of updating fiber is far less than the cost of building the initial fiber. The biggest cost of building fiber is labor. For buried fiber, the biggest cost is getting the conduit into the ground. There is no reason to think that conduit won’t last for far more than one hundred years. If a stretch of buried fiber goes bad, a network owner can pull a second fiber through the tube as a replacement—without having to pay again for the conduit.

For aerial fiber, the biggest cost is often the make-ready effort to prepare a route for construction, along with the cost of installing a messenger wire. To replace aerial fiber usually means using the existing messenger wire and no additional make-ready, so replacing aerial fiber is also far less expensive than building new fiber.

Economists define the economic life of any asset to be the number of years before an asset must be replaced, either due to obsolescence or due to loss of functionality. It’s easy to understand the economic life of a truck or a computer—there comes a time when it’s obvious that the asset must be replaced, and replacement means buying a new truck or computer.

But fiber is a bit of an unusual asset where the asset is not ripped out and replaced when it finally starts showing end-of-life symptoms. As described above, it’s much cheaper than the original construction costs to bring a replacement fiber to an aerial or buried fiber route. Upgrading fiber is more akin to upgrading a properly constructed building—with proper care, buildings can last for a long time.

Many similar utility assets are not like this. Today, my city is upgrading a few major water mains that unbelievably used wooden water pipes a century ago. Upgrading the water system means laying down an entirely new water pipe to replace the old one.

It may sound a bit like a mathematical trick, but the fact that replacement of fiber doesn’t mean a 100% replacement cost means that the economic life is longer than with other assets. To use a simplified example, if fiber needs replacement every sixty years, and the cost of the replacement requires only half of the original cost, then the economic life of the fiber in this example is 120 years—it takes that long to have to spend as much as the original cost to replace the asset.

I know that people who build fiber want to know how long it will last, and we just don’t know. We know if fiber is constructed properly, it will last a lot longer than the 40-years we saw from 1980 fiber. We also know that in most cases, that replacement doesn’t mean starting from scratch. Hopefully, those facts will give comfort that the average economic life of fiber is something greater than 100 years—we just don’t know how much longer.