Submarine salvage: a second life for old cables
More new submarine cable capacity was deployed last year than the entire undersea bandwidth in service globally in 2011, according to market research firm TeleGeography. That astounding statistic represents international bandwidth growth of 44 per cent in 2014, to reach 211Tb/s. Wholesale carriers and private networks, particularly those of large content providers, continue to invest billions of dollars to upgrade the capacity of existing subsea cables and install new systems.
In parallel to the demand for ultra-high-capacity cables on the busiest international routes, there is a need to build new subsea connections to create physical route diversity and increased resiliency. One of the biggest challenges facing submarine network operators is that many parts of the world are still vulnerable to service disruption caused by natural disasters, such as mudslides and earthquakes.
In the past decade a new trend has emerged. New submarine routes are being installed to previously underserved areas, offering optical connectivity to small islands that did not previously have access to high-speed Internet services. Isolated island communities need to provide their inhabitants with reasonable communications services, which today play a crucial role in so many aspects of life, notably emergency services and administration, but also stimulating economic growth by supporting the flow of trade and information.
Lack of access to international connections prevents small communities from keeping up with the rest of the world. More generally speaking, it is widely recognised that economic growth is strongly linked to broadband access. In a study by Chalmers University for Ericsson, it was found that upgrading broadband speed to 4Mb/s increased household income between 2.2 and 4.7 per cent. In addition, broadband penetration has a direct correlation to a reduction in unemployment, according to a study for the International Telecommunications Union.
Price and capacity-wise, satellite communications are not in a position to meet these connectivity needs. The last bastions of satellite dependency, including the land-locked countries in Africa and islands across the Pacific and Caribbean, are now discovering that their economic development needs will be better served by optical fibre. Switching from satellite to optical infrastructure is an effective way to increase the capacity, enhance the availability and reduce the latency of the network, improving both quality and stability of service.
In recent years, Xtera has noticed an increasing number of projects that plan to extend optical connectivity to small, remote islands. For example, there are ongoing discussions to find the best way to connect small island nations in the Pacific Ocean, populated by about less than 100,000 inhabitants each. New subsea systems for connecting small communities to the rest of the world do not require huge capacity but must offer a low price point to make the business case viable, bearing in mind the intrinsic small number of customers.
Two options are available. One, build a new subsea cable system, using components carefully selected to minimise the cost. Or two, recover and redeploy a subsea system that has been decommissioned, usually because the technology has become obsolete, leading to an unfavourable ratio between capacity and operational expenses on highly competitive submarine routes. In Xtera’s experience, building regional low-capacity systems on new routes represents the application where cable redeployment makes the most sense.
Low impact, high reward
One of the major reasons for considering cable redeployment rather than new build is the lower overall project cost. The typical cost structure of a new regional submarine cable system has a number of different elements (Figure 1). The largest portions are the cable and repeaters (wet plant) and their assembly and transfer to the deployment location, which can represent more than half the cost of the project. These portions also represent the biggest opportunities for cost savings.
Of course, while cost is saved on the cable (and repeaters if they are also recovered and re-used), additional marine operations are required to recover the wet plant. The cost of recovering the cable from the sea bed is determined by the level of difficulty, which will depend on whether the original cable was surface laid or buried. But in the end, even when the cost of transporting the new cable from one of the few submarine cable factories dotted around the world is factored in, it can be significantly cheaper to reuse an old cable (Figure 2).
Another important benefit of the cable re-lay approach is the shorter lead time for project execution compared to new builds requiring new cable. While manufacturing a long optical submarine cable can take up to one year (a typical lead time is 9 to 12 months), recovering an existing cable requires much less time. In general, the project will be devised so that the recovered cable is located near the redeployed cable so the general lead time is three to four months prior to redeployment. Shorter lead times turn into earlier revenues, and more secure business cases.
Redeployment projects also have good green credentials. Manufacturing a new cable requires energy and multiple materials, some of them scarce or costly to produce. On the other hand, recovering an existing cable requires minimal energy, mainly cable ship fuel. A comparison study by KTH Royal Institute of Technology in Sweden of the potential environmental impacts of submarine cable systems has shown that the recovery and re-lay of an existing submarine cable is equivalent to about five per cent of the impact caused by the manufacturing and deployment of a new cable.
Taking early retirement
An important question is how long redeployed cables can be expected to last when pressed back into service. Submarine cable systems are built with a technical lifetime of 25 years, however in reality, most systems are retired much earlier as they no longer remain commercially viable. In practice, the working lifetime could be much longer than the design life for a couple of reasons: cable systems are usually overdesigned to ensure that the lifetime requirements are met, and the materials often have a significantly longer life than originally anticipated.
In a recently worked example, Xtera looked at an opportunity where the decommissioned undersea cable had been in use for just under 10 years. Taking pessimistic assumptions, we showed that if the cable was originally operating at 7kV and since redeployment at 3kV in a shorter system, then the cable could be expected to continue operating at this voltage for at least 600 years.
For the wet plant, including repeaters, several elements must be known before the status and expected lifetime can be assessed. There are few active optical elements in a submarine repeater; in fact, the only powered components are the semiconductor lasers for delivering the optical pump waves to the amplifiers. Their characteristics are well known. Other key pieces of information are a well-documented fault history and regular performance monitoring reports from the repeaters to indicate how well these were performing prior to decommissioning.
Taking the same system as above, and knowing that the failure mechanism for lasers follows a log-normal distribution, Xtera calculated that the total laser (and associated electronics) failure rate would increase from 85 FITs at 25 years to 128 FITs over a further 25 years. (The ‘failures in time’, or FIT, rate of a device is the number of failures that can be expected in one billion device-hours of operation.) This corresponded to 0.65 ship repairs over the next 25 years. As this rounds up to one, then the number of repairs is no different to a new build equivalent for the same link.
A further level of reassurance can be gained from the understanding that, in general, most ageing occurs in the early life of a system. This means that, several years after commissioning, the subsea cable operators have a pretty good idea of the physical condition of their wet asset: either multiple, regular failures have happened and a pattern can be deduced to predict degradation in the mid-term, or no degradation happened after one or two years of commercial service.
In the latter scenario, there is a high level of confidence that the subsea cable was properly designed, manufactured and installed, and that it will retain good characteristics and performances for the future. This was the situation for the Southern Cross Cable, a 28,900-km transpacific cable system connecting Australia, New Zealand, Fiji, Hawaii and the US mainland. After an extensive consulting process with third-party experts and the original supplier, in November 2014 the cable lifetime was extended by another five years beyond the original 25 years specification.
Of course, there is an alternative scenario when there is some doubt about the active components of the cable system. The simple answer is to reuse the cable and replace the repeaters, and this too has been shown to be cost-effective.
Management of a redeployment project is quite different from a new cable project as many more boundary conditions need to be taken into account. Since the project does not start from scratch, with brand new, clearly specified wet components, cable redeployment projects are typically more challenging to design and execute.
To make a redeployment project viable from both an economic and technical perspective, there is more to consider than simply the physical status of the candidate cable. One of the early considerations is where the cable originates from.
Cable redeployment is an appropriate approach when the cable is recovered from a less or similarly benign seabed than the destination seabed. Recovery of extensively well buried armoured cable can be a slow and risky process; hence, cable redeployment works best for a destination that requires a minimal amount of armoured cable and a maximal quantity of deep water cable. When the existing cable is simply surface laid, cable recovery is fast and the cable is unlikely to get damaged during the recovery process.
There are a number of technical challenges that a redeployment project may face. The cable recovery process has to be well controlled in order to avoid applying mechanical tensions to the cable and repeaters that exceed the upper specified limits. The aim is to maximise the cable recovery yield, i.e. the amount of the existing wet plant that can be effectively re-used. Armoured cable is required for shore ends and shallow water and, typically, this section will be a new build.
Where possible, it makes economic sense to reuse the existing landing station at one end of the cable system, in order to avoid the cost of permits, the lead time to obtain them, and all the installation activities needed for the landing itself.
If repeaters are re-used in the redeployed cable, they must be monitored in the new system. Line monitoring equipment is therefore required to generate and detect the test signals appropriate to the system, to enable any faults to be located to within one repeater section.
Xtera has been involved in several projects to retrieve and redeploy submarine cables. One of the earliest examples in 2007 was the recovery and redeployment of some pieces of the Gemini cable. Originally built by a joint venture between Cable&Wireless and Worldcom, the Gemini system was a two-leg transatlantic cable system that was phased out only six years after its commissioning in 1998 due to its obsolete terminal transmission technology.
Pieces of the decommissioned Gemini cable were recovered and redeployed (along with the original repeaters) for three re-lay projects connecting Bermuda, Tortola and Jamaica to the United States (Figure 3). Although the Gemini cable system had been originally designed for operation at 4 x 2.5Gb/s, new line terminal equipment was installed that supported 16 x 10 Gb/s on each of the two fibre pairs, thus providing additional capacity for future expansion.
Additional pieces of the Gemini system were redeployed in a separate project on the east side of the Atlantic Ocean. Completed in 2006, project HUGO connected the island of Guernsey to mainland Britain and France. HUGO had initially been an unrepeatered system (with the original Gemini repeaters taken off), but had become limited by the relatively high loss of the cable. The recent upgrade of the HUGO system wet plant represents another type of cable recovery and re-lay project with the insertion by Xtera of Raman-based optical repeater to remove these loss limitations. The marine operations required for the insertion of the repeaters were similar to those carried out for a simple cable repair. This wet upgrade demonstrates that new technologies can be used to extend the capacity and working life of redeployed systems.
Another recent cable re-lay project using the redundant Rioja 1 cable was carried out by BT in 2014 to improve connectivity to the Isles of Scilly. The 939km cable between Porthcurno, Cornwall, and Santander, Spain, had remained unused on the seabed of the Atlantic Ocean since it was taken out of service in 2006. The cable was cut at two points in the Atlantic – about 100km and 15km from Land’s End. The 85km cable section recovered from the seabed was landed at Portcressa beach on the Isles of Scilly in July 2014, re-laid back to the UK mainland and then reconnected to the original cable approximately 15km offshore.
There are many compelling reasons to use recovered cables around the world. Compared with new builds, cable re-lay offers the following potential advantages:
- It can be significantly quicker to get the system up and running with funding in place.
- In some cases, savings close to 50 per cent can be seen when compared with new build.
- There are massive environmental savings of up to 95 per cent.
Recovering a decommissioned cable system with the objective of building a new system in another location is a challenging endeavour requiring additional skills beyond those needed for standard new builds with brand new wet plant direct from the factory. Also, the commercial benefits that the purchasers can expect from the cable redeployment is strongly influenced by a number of factors, including the original marine installation of the cable to be recovered, the relative locations of existing and new systems, and project design requirements. As such, re-lay project assessment can only be carried out on a case-by-case basis, and will require a high level of expertise in order to correctly identify, quantify and optimise all the factors impacting the project cost and performance.
Bertrand Clesca is head of global marketing
for Xtera and is based in Paris, France
‘Twenty thousand leagues under the sea: A life cycle assessment of fibre optic submarine cable systems’, Craig Donovan – KTH, Stockholm, Sweden, October 2009.