Patrick Fay
KMI Corp.
Worldwide demand for cabled fiber grew 38% last year, and demand for nonzero dispersion-shifted fiber (NZDSF) grew 159%. Bandwidth requirements for the Internet are driving these fiber deployments, as new operators build more efficient networks to compete with the legacy networks of incumbent carriers.
This unprecedented demand created a fiber shortage in 1999 that will continue through 2000. Fiber supply will continue to be tight in 2001. The announced investment targeted toward capacity expansion in 2000, however, will bring on enough capacity in 2002 to relieve the shortage.
The worldwide market in 1997 and 1998 was characterized by over-capacity and falling fiber prices due to soft demand and regional economic hardships. Since 1998, fiber demand has increased and far surpassed even the most optimistic forecast. The major producers had slowed their expansion plans during the period of excess capacity just when 1999 would show unprecedented growth. The result has been a worldwide fiber shortage, although certain regions such as North America and Europe initially were hit harder than other regions.The shortage has caused major producers thus far in 2000 to invest more than $2.5 billion in capacity expansions; this investment will increase worldwide capacity from 77 million fiber-km in 1999 to 131 million fiber-km in 2001, an increase of 70%.
Deployment of fiber-optic cable increased 38% in 1999 to 63.3 million km of cabled optical fiber, from 45.9 million km of fiber in 1998. By 2004, annual deployment will reach 181.5 million km (see Figure 1).
Figure 2 shows the segmentation of the cable market by terrestrial singlemode, multimode, international submarine, and domestic cable types. Singlemode cable (59 million fiber-km in 1999) accounts for more than 93% of the volume of cable installed worldwide and is 75% of the market value of all cable installed. In 1999, the terrestrial singlemode cable market was $6.5 billion; KMI expects this segment to represent $13.7 billion in 2004 or 78% of the market.
Multimode cable was 4% of the volume in 1999 and 9% of the market value ($816 million). The proportion of multimode volume decreases to 3% in 2004 (5.3 million fiber-km), while the market value increases to 8% ($1.4 billion).
International submarine-cable accounts for 1%-2% of cable volume installed from 1998 through 2004. But because of the extreme environmental requirements, the cable is costly, so the market value of the cable is disproportionately high-from 10% in 1998 to 15% in 1999. KMI forecasts this segment to remain at least 12% of the market for total cabled fiber demand through the forecast period. The international submarine-cable market will grow from $1.3 billion last year to $2.3 billion in 2004.
Domestic submarine cable, which includes domestic festoon systems, is less costly than international, transoceanic cable. This cable accounts for less than 1% of cable volume.
Optical-fiber demand is driven by the cable production needed to meet actual and forecast deployment worldwide. A greater volume of fiber-about 10% more-is needed to yield the annual cable-deployment volumes shown in Figure 1. Demand last year was 70.2 million km of fiber-a $3.6-billion market (see Figure 3).
In 1999, conventional singlemode fiber (CSMF) accounts for 54% of the market on 78% volume. In 2004, the market for CSMF is forecast to reach $4.5 billion or 70% of the market. Production of NZDSF increased in 1999 to 18% of total demand, up from 9% in 1998. In 1999, the market for NZDSF was $1.3 billion or 35% of the total market on a volume of 12.6 million fiber-km. As with submarine cable, NZDSF carries a premium, which accounts for its higher value proportionate to the market. In 2004, the market for NZDSF is forecast to reach $1.2 billion or 19% of the total market.
Last year, 43% of the fiber-optic cable or 28.9 million fiber-km of worldwide production was manufactured in North America, and 40% of the world's fiber was also produced there (27.9 million fiber-km). Cable makers in North America had net exports of 4.1 million km of cabled fiber but imported almost 2.1 million km of bare fiber in 1999. North America previously was a net exporter of fiber, but the dramatic increase in production of NZDSF by North American producers and its effect on capacity meant that the manufacturers had to import fiber from other regions to meet CSMF cable requirements.
Less than one-third (28%) of the fiber-optic cable (17.7 million fiber-km) installed worldwide in 1999, and more than one-quarter (29%) of the fiber (20.7 million fiber-km) was made in the Asia-Pacific region. Western Europe made 23% of the cable (14.8 million fiber-km) and 27% of the fiber (18.7 million km) in 1999. All other regions combined produced 5% or less of either cable or fiber volume in 1999.
KMI has estimated manufacturing capacity for optical fiber through 2002 based on announced expansions by several major fiber suppliers. For 2003 and 2004, KMI assumes that capacity will increase at a rate that permits fiber suppliers to meet seasonal or unexpected demand.
Figure 4 shows the impact of NZDSF on fiber production capacity. It is understood throughout the industry that NZDSF is more time-consuming to produce to meet stricter fiber geometries, meaning that in any given period, considerably more CSMF can be produced than NZDSF using the same available capacity. To determine the effect that the increased production of NZDSF has had on capacity and its contribution to the shortage, KMI assumes that for every 1 km of NZDSF produced, 1.5 km of CSMF would have been produced.
There are two columns for each year in Figure 4. The left column for each year indicates actual singlemode production. The right column for each year indicates the capacity required to produce all singlemode fiber in CSMF terms. For example, in 1996, the actual production for singlemode fiber and the capacity required to support that production are virtually identical. That's because the 619,000 km of NZDSF that was produced only required an additional 309,000 km of capacity. The contrast in capacity required to produce the 12.6 million km of NZDSF in 1999 is evident. KMI estimates that 6.3 million km of capacity that otherwise could have been used to produce CSMF was "lost" in producing NZDSF.
Announced investment in 2000 will create sufficient capacity to meet demand in 2002. An estimated $4.9 billion will have been invested in capacity expansion from 1996 through 2000. More than half of this investment is the $2.6 billion announced in 2000. KMI estimates that another $2.3 billion must be invested in 2003 and 2004 to bring capacity for those years to 213 million and 252 million km.
The fiber and cable industry responded to accelerated demand for bandwidth with introduction of new fiber products for all the segments of communication networks. In long-distance systems, new products have been introduced for use in the high-bit-rate DWDM systems. This trend is going to continue until an optimal solution is found and standardization occurs around it. In the metropolitan networks, enhanced G.652, with stringent tolerance of hydroxyl ion content, is beginning to take hold as multiple vendors start to introduce products. In premises applications, multimode fibers with improved profiles have been introduced to meet the needs of 10-Gbit/sec Ethernet applications.
Deregulation of carriers in the United States and in Europe as well as the rest of the world has had a profound change in the structure of the buyers of optical-fiber cables. The industry has had to restructure to accommodate new customers, as evidenced by the following examples:
- Corning became a cable company through forward integration by acquiring BICC and Siemens fiberoptic cable facilities.
- Draka acquired more interests in cable companies and merged with NK Cables.
- Lucent Technologies acquired SpecTran.
The battle for market share will be in cable sales, and cost-effective sourcing of preforms and fibers will be necessary for success for many of the integrated producers.
Market segmentation at the fiber level has changed, and a new segmentation in preform-level sales has developed. Last year, fibers produced from purchased preforms became greater than 10% of all fibers produced. Since preform production is a chemical vapor deposition glass production technology, core competency considerations may drive the market to continue to become further segmented at the preform stage.
Patrick Fay is an analyst at KMI Corp. (Newport, RI). This article is an excerpt from the Executive Summary of "Worldwide Optical Fiber and Fiberoptic Cable Markets-Analyses and Forecasts of Installations, Production, and Capacity," published this month. More information is available at the company's Website, www.kmicorp.com.