North American market retains fiber-optic cable lead
Fiber-optic cable sales in North America are expected to gain strength during the next decade because of increased installations of private data and cable-TV networks, intense competition among transport suppliers and a strong export market
Stephen Montgomery
electronicast Corp.
The worldwide production of fiber-optic cable is expected to expand impressively from $3.8 billion in 1995 to about $8.5 billion in 2000, and then nearly double to $15.5 billion by 2005. The North American fiber-optic cable market--consisting of the United States, Canada and Mexico--held the worldwide market share of 70% in 1995, at $2.7 billion. Europe, with a 17% share, or $643 million, and Japan and the Far East, with a 12% share at $462 million, follow (see Table 1).
Although the North American fiber-optic cable market share is predicted to stay flat at 70% through 2005, sales are expected to nearly double every five years, reaching $5.9 billion in 2000 and $10.8 billion in 2005. This strong growth is figured to come mainly from a surge in the installation of cable-TV and premises data networks. In addition, competition among transport equipment suppliers is forecast to drive rapid expansion in telecommunications networks and applications. Other growth areas include direct and indirect fiber-optic cable exports.
The fiber-optic cable market in Europe is predicted to drop slightly in market share over the next decade, slipping to 15% in 2000 and 14% in 2005. Sales are anticipated to nearly double twice during that period, though, totaling $1.2 billion in 2000 and $2.2 billion in 2005. Similarly, the Japan and Far East market share should remain flat at 14% in 2000 and 13% in 2005. However, sales are anticipated to be brisk, almost tripling to $1.2 billion in 2000 and nearly doubling to $2 billion by 2005.
The North American market, especially the United States, leads the world in fiber-optic component development and integration of optical components into telecommunications, data communications, cable-TV and high-performance military applications. It should be the leading consumer of fiber-optic components through 2005, and is evaluated to be a net exporter of high-performance fiber-optic components, mainly to Europe.
North American device and component producers, however, do not excel in low-cost production of high-volume, moderate-performance, consumer-grade devices. Consequently, the North American market is expected to be a net importer of many optoelectronic and passive optical components and connectors, almost entirely from Japan and other Pacific Rim countries. Moreover, by the year 2000, Japanese companies are determined to achieve a major role in the production of fiber-optic components, either by totally or dominantly owned subsidiaries in target countries using bases in the United States and Europe.
Historically, fiber-optic cable network deployment has been controlled by long-line installations, such as interexchange, interoffice, subocean and cable-TV trunks. The economic justification for using passive optical components in these networks has been minor. But the use of fiber-optic cable networks is shifting toward expensive, complex, local-area data, subscriber broadband-distribution and military local-area networks, all of which demand higher performance and reliability.
During the next decade, market economics is expected to favor the expansion of capacity on long-haul fiber-optic networks by adding higher-data-rate optoelectronics combined with dense wavelength-division multiplexing technology. Therefore, the market share of passive optical components, relative to total fiber-optic component sales, is expected to soar.
Steady growth
In the 1995 North American fiber-optic cable market, the telecommunications sector achieved the highest share, at 55% with $1.32 billion. This sector should remain flat in market share over the next decade (56% in 2000 and 55% in 2005), but sales should double in value. Telecommunications sales are estimated to total $2.86 billion in 2000 and $4.9 billion in 2005.
Next in market share is the private data local-area and wide-area network (LAN and WAN) sector, with an 18% share, or $422 million, in 1995. This sector should drop slightly to 16% in 2000 but nearly double in sales to $823 million. Then, it should climb to 23% in 2005, with sales of $2 billion.
Third in market share is the cable-TV and broadcast sector at 16% and $388 million in 1995. This sector is predicted to rise to 20% market share in 2000 and nearly triple consumption at $1 billion. However, as the cable-TV network expansion slows down, market share is expected to drop to 15% while consumption nudges upward to $1.3 billion by 2005 (see Table 2).
Sales of fiber communications cables in residential broadband networks are projected to expand from $1 billion in 1995 to more than $4 billion in 2005. Coaxial copper cables are the present market leader in these networks due to the new builds, rehabilitation and upgrades to hybrid fiber/coaxial-cable installations by cable-TV operators. By 2005, however, fiber-optic cables are expected to capture more than 80% of market share in broadband networks. Twisted-pair copper wires are predicted to grow moderately during 1995 to 2000 because of use in asymmetric digital subscriber line (adsl) and high-bit-rate digital subscriber line (hdsl) networks.
Nearly all the fiber-optic cabling being deployed now for broadband local-loop applications functions at a 1310-nm wavelength. The 1550-nm wavelength is preferred in long-haul networks. Presently, optoelectronics for 1550-nm operation cost much more than equivalent optoelectronics for 1310-nm wavelengths.
Local exchange carrier network planners have determined that for fiber spans of 50 to 80 km, the 1310-nm fiber networks and associated optoelectronics are the economical choice, even though optical-fiber losses are nearly twice those at 1550 nm. Some carriers have selected 1550-nm fiber, especially for longer trunk networks, because of the increased availability of optical amplifiers; these amplifiers were not available at 1310 nm until recently.
Over the next decade, network planners for local exchange carriers and interexchange carriers are figured to incorporate more 1550-nm fiber to accommodate higher data rates for compatibility with trunk networks. However, sales of 1310-nm fiber cables are still estimated to hold more than 60% market share by 2005.
Nearly all the fiber-optic cable deployment for broadband-to-the-home (bbth) from 1995 to 1996 was accomplished by large cable-TV multiple system operators (MSOs), who have been upgrading their networks for the delivery of more video channels at higher reliability. Simultaneously, they have been building network platforms for the future delivery of voice, data and other services. Their initial thrust has concentrated on high-population-density metropolitan areas, making use of existing underground ducts.
As cable-TV fiber deployment spreads to the suburbs and rural areas, however, the cable mix is expected to shift to a larger share of aerial installed cabling. Aerial installation on existing poles has proven relatively inexpensive. In contrast, digging trenches for cable installation is still costly. The aerial cable market share is assessed to rise from 45% in 1995 to 60% in 2005 as more rural subscribers come online. The buried cable market share should remain minimal during that period.
New contenders, such as local exchange carriers and interexchange carriers, are soon expected to enter the bbth race. They are anticipated to initially target highly populated areas that are readily accessible via existing ducts.
Coaxial-cable deployment for North American bbth networks, predominantly in cable-TV networks, became a major market in 1996. Although many large cable-TV MSOs are scrapping long coaxial-cable trunks in favor of fiber-optic cable installations, the deployment of coaxial-cable networks is still expected to grow modestly over the next several years. New market competitors, especially local exchange carriers, are expected to become big installers of hybrid fiber/coaxial-cable networks. Sales of bbth coaxial-cable networks are calculated to expand at an average annual rate of 17% from 1995 to 2000.
During 1996, telephone companies expressed intense interest in providing broadband residential services over unshielded twisted-pair (UTP) copper wires. The basis for this interest was the possible front-end cost reductions achievable via currently deployed voice-grade cables. Equipment sup pliers have responded by developing devices to handle Integrated Services Digital Network technology at 128 kbits/sec, and adsl and hdsl technologies at various formats and data rates to 10 Mbits/sec for faster Internet access speed and video transport over limited distances using voice-grade copper wires.
This method enables telephone companies to deploy a quick, inexpensive and low-risk network for customer-service upgrades. However, it uses already-deployed cables and does not represent a new market for UTP wiring. In addition, several limitations should stifle the use of voice-grade copper for bbth networks:
Bandwidth capability is foreseen as proving inadequate for handling future broadband services, such as high-definition television. Consequently, these copper networks are expected to degrade in performance by the year 2000.
Most of the currently deployed voice-grade copper wires has been in place for more than 30 years. These wires have suffered enough deterioration to make them unsuitable for broadband transport.
Most potential broadband service subscribers, chiefly in suburban and rural areas, are beyond the distance capability of high-data rate services.
In 1995, singlemode fiber-optic cable grabbed an 82% market share, or $1.96 billion, in North American consumption. Multimode fiber-optic cable achieved an 18% share, or $422 million. By the year 2000, singlemode fiber is expected to double consumption to $4.3 billion, with an 85% market share. Market share will then drop to 79% but will experience increased consumption of approximately $7 billion. Similarly, multimode fiber consumption should increase markedly during that period, to $776 million and a 15% market share in 2000, and jump to $1.9 billion and a 21% market share by 2005 (see Table 3).
Over the next decade, singlemode cable should find widespread use in the rapidly growing telecommunications feeder lines needed to support distribution networks. Private data communications networks are figured to spur the installation of multimode fiber.
The average number of fibers in cables should increase steadily over the next decade, as data traffic rates are expected to skyrocket. A large share of the traffic should be economically accommodated by retrofitting higher-speed optoelectronics and Synchronous Optical Network multiplexing, as well as by wavelength-division multiplexing.
Loosely buffered standard singlemode cable containing more than 50 fibers is predicted to climb from 36% of installations in 1995 to 46% of installations in 2005. The increase is expected to result from a spurt in local-loop network and cable-TV installations. u
Stephen Montgomery is vice president and chief operations officer at ElectroniCast Corp. in San Mateo, CA.