Abundant capacity and production to stem tide of undersea fiber-optic cable installations
Rapid growth in worldwide submarine fiber-optic cable links is expected to peak in 1996 and then slowly decrease as capacity temporarily outstrips demand for high-bandwidth applications
Thomas A. Soja
Record numbers of new submarine fiber-optic links are being installed in 1995 and 1996. However, the long-term market forecast indicates that this unprecedented growth will not be sustained. Without widespread implementation of bandwidth-demanding, internationally useful communications applications, undersea circuit demands will slacken. Accordingly, intense competition among submarine fiber-optic cable suppliers will likely result in a diminishing market share.
During the past eight years, undersea fiber-optic cable suppliers have installed an amount of lightwave cable that nearly equals that of copper cable installed during the 37 years from 1950 to 1987. Indeed, by 1998, almost 129 countries are projected to be connected to submarine fiber-optic cable networks.
In 1994, 33 fiber-optic undersea links totaling 56,896 route-kilometers were placed into service, a 22% increase over 1993 efforts. In 1995, 65 new submarine links totaling only 40,721 route-km are slated for activation. Although more links are projected for installation this year than were installed in 1994, the total amount of cable is much less because the links are shorter. In 1996, the undersea outlook is appraised at 106 new links and 103,171 route-kilometers, more than 81% over the installations made in 1994.
The forecast for undersea fiber-optic cable deployment through 1998 includes a sharp drop in cable installations to be cut over in 1997. Although more cable projects are expected, the downturn in cable deployments predicted for 1997 cannot be countered by the number of new announced systems.
The projected cumulative number of new undersea links is predicted to reach 591 by 1998. Beyond 1998, another decrease in new system cutovers in terms of route-km is forecast, with a likely recovery toward an average of approximately 75,000 route-km in the year 2000. Comparing the annual volume of new cable deployments with available production capacity suggests a large amount of idle plant capacity.
Industry undersea cable-making capacity is expected to rise from 108,000 cable-km in 1995 to more than 118,000 cable-km later this decade. These numbers dramatically exceed the projected deployment volumes that average on the order of 70,000 route-km per year through 1998.
Competitive challenges, therefore, emerge for cable and system suppliers. Leading-edge technology may no longer suffice in a near-term market where available cable circuit and production capacities abound. Strategic partnering with local companies that can provide access to growth markets has begun to shape the undersea industry. The worldwide-distributed supply of key products and services--either through ownership of those resources or through contractual arrangements with third parties--is also occurring. The key to winning market share in this business environment will be to provide market-based rather than technology-driven solutions.
Through the end of 1994, $11.3 billion was invested in submarine fiber-optic cable systems that connected 70 countries worldwide. This year, another 17 countries will be cut over onto submarine fiber on 63 new international and domestic links. Between 1995 and 1998, another $14.9 billion is expected to be invested in additional undersea fiber-optic cable systems.
Approximately 1.8 million, 64-kilobit-per-second circuits are available on international undersea fiber-optic cable links worldwide. More than half of these links connect countries in Northern Europe. For the years 1995 through 1998, developments in new circuits are projected to be more balanced among worldwide regions because 2.3 million new fiber-optic submarine circuits are planned for cutover.
A shift in the regional focus of installation activity is reflected in the submarine systems planned for the next five years. During 1995 to 1998, more new route-km of cable are anticipated to be placed into service than all of the fiber-optic undersea cable links installed through the end of 1994. However, most submarine installations will take place in Africa, Asia and Europe. The Atlantic Ocean region, the major sector of undersea installations through 1994, at 33%, is foreseen to plummet to 8% of all new cable by 1998.
Another shift has occurred in the increased number of unrepeatered links being installed as the bandwidth-distance capability of undersea systems continues to expand. Today, the unrepeatered distance limit exceeds 300 km and is estimated to reach 400 km within a year for commercial systems. Laboratory trials have already pushed the repeaterless limit beyond the 500-km mark.
In the first quarter of 1995, there were few new systems announcements. Only one new repeatered link was announced--a 339-km link that forms part of the Hawaiian Islands Fibernet System. However, in April, another major repeatered system was announced--the 2500-km Pakistan to United Arab Emirates fiber cable. And since then, the 2500-km Northstar cable, the 4000-km Guam to Hawaii cable, the 4500-km Indonesia to Guam cable and the first phase of a Southern Atlantic Ring network (the 3200-km Brazil to Senegal to Canary Islands cable) have been announced for cutovers beginning in 1996 to 1998.
Increasing fiber-cable capacity and restoration, as well as network expansion along new, unserved routes, have been major driving forces for the undersea cable industry. Activity in the Atlantic Ocean region illustrates these points as the AT&T TAT-12 and TAT-13 undersea systems come online in 1995 and 1996, respectively. These systems constitute the eighth and ninth fiber-optic transatlantic cables. They are designed to add circuit capacity and route diversity in the event of a system failure.
Both TAT links are being built in a ring configuration as a total system to provide fiber-on-fiber restoration. The TPC-5CN link in the Pacific Ocean region was designed with similar goals. These 5-Gbit/sec ring networks are capable of handling more circuit capacity than previous systems in their respective transoceanic regions.
The key issue that arises, therefore, is whether bandwidth capacity has outstripped service demand--at least for the near-term. Although studies to project circuit demand are underway, other evidence suggests a lull in large-scale system deployment. Past studies of terrestrial long-haul networks demonstrate that when major traffic routes become saturated with fiber capacity, deployment rapidly decreases.
One reason for this downturn is that cable capacities can be upgraded in terrestrial systems as time requires and technology allows--by changing the repeater equipment along the routes while leaving the fiber in place. But until recently, capacity upgrades have not been possible in repeatered fiber-optic undersea systems. Unrepeatered systems have long been promoted as upgradable because they do not employ underwater electronics. Increases in capacity are accomplished by upgrading terminal equipment to higher bit rates as distance and system considerations allow.
But the current generation of erbium-doped fiber amplifiers offers repeatered systems the potential for upgradability with proper submarine system planning at the outset, such as was done for several recent 10-Gbit/sec system experiments. Two basic methods for achieving capacity upgrades are using faster operating transmission terminal equipment and wavelength-division multiplexing technology. Although both methods have some limitations, they permit bandwidth expansion of existing erbium-doped fiber amplifier systems. For the near-term, most undersea routes cabled with these systems are expected to have adequate capacity for the next five years.
According to industry analysts, the worldwide telecommunications services market in 1994 was a $600 billion industry. The United States represented approximately one-third of the worldwide market, with $193 billion spent on services to 90 million households and 25 million businesses. Western Europe was involved with almost 21% of the worldwide market, at $128 billion. In 1993, the market in the People`s Republic of China was estimated at approximately $4.15 billion, with traffic volume up 59% over 1992; continuous growth is projected.
Carriers worldwide are increasingly investing in infrastructure to deliver services to customers and to capture a share of the global telecommunications market. But infrastructure is costly; therefore, many carriers are forming partnerships to extend their networks to more regions.
For example, the World Source partnership, originally founded as World Partners by AT&T, Kokusai Denshin Denwa and Singapore Telecom, has been expanded to include a similar European partnership--Unisource--among Telia, PTT Telecom Netherlands and Swiss Telecom. Moreover, Telefonica of Spain and General des Eaux, the French water utility, have joined the alliance. In addition, Hong Kong Telecom, Korea Telecom, Telstra (Australia) and Telecom New Zealand have joined as non-equity partners; Nippon Telegraph and Telephone (Japan) has joined as an observer. AT&T is seeking approximately 25 affiliates to join the World Source partnership.
Other global alliances include Eunetcom--consisting of Germany`s DBP Telekom and France Telecom. To extend their reach into the United States, these two carriers are trying to link up with Sprint; this alliance is awaiting U.S. regulatory approval. Another duo, British Telecom and MCI, won approval last year for their alliance, known as Concert. However, many carriers with significant assets in strategic locations have yet to sign a partnership. u
Thomas A. Soja is senior analyst at Kessler Marketing Intelligence Corp. in Newport, RI (e-mail: firstname.lastname@example.org).