Fiber penetration slowly increases in LEC networks
Fiber sheath miles and fiber counts are up in RBOC networks. But the technology still has a long way to go to catch copper.
By DONALD T. GALL and MITCH SHAPIRO
Every year, the Federal Communications Commission (FCC) conducts a survey of carriers--including incumbent and competitive local-exchange companies and interexchange carriers--about the status of their fiber deployment. The survey, which dates back to the mid-1980s, is conducted by Jonathan Kraushaar in the Industry Analysis Div. of the Common Carrier Bureau.
The FCC`s annual "Fiber Deployment Update" provides current and historical data on the total amount of copper and fiber deployed in incumbent local-exhange carrier (ILEC) plants. It also provides more limited data (especially regarding fiber) on the subscriber portion of plant, which includes the feeder and distribution lines connecting subscribers to a central office.
The regional Bell operating companies (RBOCs) have expanded their fiber base at a fairly steady pace during the past five years in terms of sheath mileage (see Table 1). Adding roughly 19,000 to 23,000 mi annually, they have built this base from less than 174,000 mi at the end of 1993 to nearly 258,000 mi in 1997. For the group, annual growth in sheath miles peaked in 1995 at 22,900 before declining to 20,200 in 1996 and 18,800 in 1997. On a percentage basis, expansion of the base of fiber-sheath miles gradually slowed from 13% in 1994 to 8% in 1997.
While annual growth for BellSouth, Pacific Telesis, and US West peaked in 1995, the following year marked the growth peak for Ameritech and SBC. Bell Atlantic (including Nynex) and US West were the only companies that added more fiber-sheath mileage in 1997 than they did in 1996, with the latter`s 1997 figure still far below its 1995 peak.
Because the average fiber cable cross section (also called "fiber count") showed a steady increase from 38.3 to 47.4 during this 1993-to-1997 period, the RBOCs` growth rate in fiber mileage was higher than it was for sheath miles--hitting 20% in 1994 and declining gradually to 13% in 1997. According to the FCC, the RBOCs together accounted for more than 12.2 million fiber-miles at the end of last year.
The RBOCs plus GTE and Sprint added 1.68 million fiber-mi in 1995, 1.65 million in 1996, and 1.67 million in 1997. Our own research suggests that this level of activity is not far from what cable operators have been doing in the past few years, especially in 1996 to 1997. Not including telephone company-overbuild hybrid fiber/coaxial-cable (HFC) activity (e.g., Ameritech`s cable deployment), we estimate annual cable industry fiber-miles at 1.38 million in 1995, 1.62 million in 1996, and 1.68 million in 1997. Since average fiber counts in the cable industry appear to be somewhat lower than in the telephone company camp, the former has been relatively more active in terms of laying sheathmiles of fiber. This activity is generally consistent with the notion that cable operators are deploying fiber deeper into their plant (the typical optical serving area in today`s HFC architecture is likely to be no more than 500 to 1000 homes) and that average fiber counts tend to decline as fiber is extended closer to individual subscribers.
Fiber per access line
Another useful measure of local-exchange carrier fiber activity is the amount of sheath and fiber-miles deployed on a per-line basis (see Table 2). Not surprisingly, the 1996-to-1997 trend in terms of miles per thousand lines was up for fiber and down for copper.
Because some RBOCs have very different average fiber counts, their ranking in terms of per-thousand-line "fiber-sheath miles" and "fiber-miles" are not always the same. For example, Bell Atlantic claims the lead in fiber-miles (100 per thousand lines in 1997), while BellSouth has deployed the largest number of fiber-sheath miles of all the RBOCs (2.3 per thousand lines). Bell Atlantic`s fiber cable carried a much higher average fiber count than BellSouth`s--59.2 versus 38.1 fibers per cable.
Bell Atlantic also boasts the richest mix of fiber in terms of sheath-miles, with fiber accounting for 12% of total sheath miles in 1997, versus 9.2% for BellSouth, 9% for both Ameritech and US West, and 8.7% for SBC. The laggard on this and related measures is PacTel, whose percentage is 6.7% and which had deployed only 27.4 fiber-mi per thousand lines at the end of 1997, versus 65 to 100 for the other Bells. While part of the reason is the company`s relatively dense plant (it has also deployed the least amount of copper per thousand lines and has a relatively short average loop length), this factor does not explain the bulk of the difference when considered in light of the density measures for other Bell companies.
GTE has an even higher 1997 per-line fiber-sheath average than Bell Atlantic (2.5 per thousand), yet its fiber mileage per thousand lines, at 63.7, is lower than all the Bells save PacTel. Again, the key difference here is average fiber counts, where GTE`s 25.6 compares to the Bells` average of 47.4 and Bell Atlantic`s high of 59.2. And even though GTE had the highest per-line fiber-sheath average, its fiber-sheath accounted for only 6.3% of total sheath-miles, even lower than PacTel`s 6.7%. This low percentage is attributable to GTE`s deployment of the greatest amount of copper sheath per thousand lines--37.3 miles versus an RBOC high of 23 for BellSouth, followed by 20.6 for SBC, and the rest of the Bells in the 9.2-to-15.3 range.
Table 3 highlights several other fiber-related data elements compiled by the FCC, including the percentage of "lit" fiber, the number of DS-3 (44.736-Mbit/sec) miles on fiber and T1 (1.554-Mbit/sec) miles on copper, and each LEC`s aggregate investment in fiber plant. "Percent lit" data for 1997 ranged from 16.4% for Ameritech and 18.4% for SBC at the low end to Bell Atlantic`s 40.1% and US West`s 37% at the high end. The middle ground here was inhabited by BellSouth (28.1%) and PacTel (31.5%).
As a group, the RBOCs reported 3.76 million DS-3 miles on fiber, an increase of 30% compared to the previous year`s 2.89 million. At the same time, the number of T1 miles on copper slipped by nearly 4%, from 6.67 million to 6.44 million. Among those with relatively dramatic shifts in this area were US West, PacTel, and Ameritech.
Collectively the RBOCs accounted for $8.57 billion in aggregate fiber-plant investment at the end of 1997, an increase of 9% over the previous year`s $7.84 billion. Individually, Ameritech (12%), PacTel (11%), and SBC (14%) experienced double-digit growth in 1997, while Bell Atlantic (8%), BellSouth (8%), and US West (5%) saw their aggregate fiber-plant investments grow more slowly.
Fiber in the subscriber plant
The FCC`s survey also includes some data on the amount of fiber in the LECs` subscriber plant. This information is fairly spotty, however, with virtually no recent data for Ameritech, Bell Atlantic, and BellSouth, and significant gaps in the data available for all the other RBOCs except for PacTel. Data for GTE is available for 1996 and 1997, but for no prior years.
Using the available "subscriber" fiber-sheath data in combination with the more complete set of data on copper-sheath mileage, we calculated a "fiber percentage" of total "subscriber" sheath miles in cases where both sets of data were available. For PacTel and GTE, the only two companies with 1997 data available, fiber accounted for 3% and 3.1% of total subscriber sheath miles, respectively, up from 1996 levels of 2.7% for PacTel and 2.8% for GTE.
For Nynex, we calculated a comparable percentage of 5.8% in 1996, the last year it reported the relevant fiber data. The most recent year in which Bell Atlantic reported this data was 1990, when its 2.2% "fiber" percentage of subscriber sheath miles put it at the head of the pack, followed by Nynex with 1.7%. SBC and US West last reported this data in 1995. At that time Nynex stood at 5.2%, followed by SBC (4.4%), US West (3.8%), and PacTel (2.4%). Ameritech stopped reporting subscriber-plant fiber data in 1992, when its percentage of fiber sheath was 1.7%, which at that time trailed Nynex (3%) and US West (2.1%), but topped the numbers for PacTel and SBC (both 1.5%).
A final measure of the depth of fiber`s penetration into the RBOCs` local exchange network is provided in Table 4. This table provides data on the number of fiber-fed pedestal locations, the number of fibers and fiber-miles serving these pedestals, and the number of customers within reach of these "fiber-deep" network devices.
Fiber-to-the-pedestal accounts for only a tiny fraction of total access lines, reaching 208,296 "accessible customers" at the end of 1996 and 308,205 in 1997. Of the 1997 installed base, Bell Atlantic and BellSouth each accounted for nearly 125,000 fiber-fed customers, with most of the former`s located in Nynex territory. Of the nearly 100,000 customers added to the fiber-to-the-pedestal population in 1997, more than 80,000 were claimed by BellSouth, with another 16,315 coming from Bell Atlantic. The average number of customers per fiber-fed pedestal for the Bell Atlantic deployments, most of which have been in dense urban areas, is more than 16, compared to about six for both BellSouth and US West.
Overall, the FCC data suggests that while the regional Bell LECs have continued to expand their fiber base and have gone far in linking central offices via fiber, the penetration of optical technology within the subscriber plant is still in its early stages. q
