One of the truly great initiatives in providing capacity to smaller countries or communities is the OADM BU, the device which, when fitted to a submarine cable, allows a handful of wavelengths to be picked out and carried on a spur cable while allowing all the mainline traffic to be carried unaffected to its destination. Not particularly expensive, it affords a high degree of flexibility to design networks to service places at a fraction of the cost of providing a dedicated cable. The spur can be a short as a few kilometres or as long as over 100km.
There is a strong incentive to put these BUs along major cables at points convenient to connect the chosen location by a spur. Sometimes however, when the main cable is being laid, the beneficiaries of the spur only have the money for the BU and not the spur, or the logistics may not be right for laying the spur contiguously with the main cable. In such cases, it is normal practice to lay a short spur from the branching unit and then stub it.
This stubbed spur typically has a length equivalent to two water depths or more so it can be picked up in future without disturbing the BU. Naturally it has no electrical power on it. I have always believed that this stub can be picked up at a later date when the full spur is to be attached, without impacting the main cable or the traffic on it. So you can imagine my great surprise when I found that to be a myth. In order to pick up the stub to extend with a spur cable, whether repeated or not, you have to interrupt traffic on the main cable. This turns a straightforward job into a very complex one, calling for delicate negotiations to interrupt high volume traffic on the main cable for up to two days. It can raise the cost of the spur cable substantially. Why is this necessary? It is primarily for safety. Some repeaters cannot be latched to prevent them from accidentally switching the power-fed high voltage and current to the spur.
Although this risk can be obviated by grounding the cable, not all ships can do that. And there is another safety concern in optical power – the powerful laser light being sent along the fibre. Sometimes there is scope to limit the interruption to traffic on the main cable by diverting it to a pair which is not subject to coupling in the OADM BU. Then the light on the stub can be turned down to safe levels. But no matter what steps you take, the extension of a stub is not straightforward. This problem is certainly something that the manufacturers need to fix.
John Hibbard is CEO at Hibbard Consulting. He can be contacted at: jhibbard@bigpond.com
