EXCEM: Fault location techniques. (original) (raw)

1. Faults and available fault location principles

A fault appeared on the first telecommunication transatlantic subsea cable (a telegraph cable) after a few days of operation. Even though modern techniques like burying cable, information of fishermen, GPS, etc... prevent many cable damages, they are far from being ruled out. High performance fault location techniques are a key to achieve the shortest time-to-repair of submarine cables.

To further discuss this matter, it is useful to classify the faults that are most commonly encountered on a optical cable system:

• We use the term type 1 fault or cable break for the cable being cut, with a break in the electrical insulation between seawater and the power-feeding conductor.
• We use the term type 2 fault or open fault for the cable being cut, without breaking the electrical insulation between seawater and the power-feeding conductor.
• We use the term type 3 fault or shunt fault for a break in the electrical insulation between seawater and the power-feeding conductor, without this conductor itself been cut.
• We use the term type 4 fault for a damage in the optical path without significant electrical alteration of the power-feeding conductor continuity and insulation.

Some methods of locating faults that appear on subsea links have been in existence for a long time and are well known. Others, like optical methods, have been introduced recently. We may list the fault location methods which may be implemented from a cable station:

1. The direct method using single-end DC measurements is applicable to type 1 faults or type 3 faults. It consists in measuring DC resistances for several current values, from one end of the cable system. Such methods have been in use since telegraph cables, but the old procedures and equipments are no longer applicable.
2. The capacitive method using single-end measurements is applicable to type 2 faults. It consists in performing a DC capacitance measurement for determining the capacitance to ground of the insulated power-feeding conductor.
3. The conjugate method implementing current-balance measurements is applicable to type 3 faults. It operates by measuring DC resistance from the end A of the cable system with a first current source, the other end B of the faulty branch being connected to a second current source that delivers a current of the same value but opposit sign, thus canceling the current into the fault.
4. The optical path methods implementing OTDR (optical time-domain reflectometry) or COTDR (coherent optical time-domain reflectometry) are only applicable to type 4 faults.

These fault localization methods are discussed and compared in the next paragraph.

DC fault localization comprises the first three methods. They may be implemented with the L2G or with other equipments. More explanations on these methods are also included in the description of the fault localization methods available using the L2G.

Other methods for fault location were available with coaxial submarine cables, but they cannot be implemented for optical cable systems. Also we do not discuss here the "electroding" fault location technique using a low frequency tone generator (for instance at 25 Hz) and a ship travelling along the cable, because it is mostly meant to eventually spot a fault previously roughly located along the cable.

2. Discussion of the four fault location principles

Each method has its advantages and drawbacks. Fault location with the direct method is only accurate if one is able to precisely remove the contribution of the repeaters and the fault contribution to the measured current/voltage characteristics. Removing the contribution of the repeater is easy if they are well designed and characterized. Removing the contribution of the fault is also very important and much more difficult.

Fault localizations with the capacitive method require appropriate measurement techniques. Conventional capacitance measurement systems are not applicable to subsea cables, because of noise, losses along the cable (the purpose is to measure a distributed capacitance, not a lumped capacitor), imperfect isolation at the cable end, transmission line effects and frequency dependance of the capacitance itself.

A fault location with the conjugate method requires an additionnal telecommunication means between each end of the cable system, a requirement easily achieved today. It is very accurate provided good synchronization and current cancellation at the fault is achieved. Like single-ended fault location, the contribution of the repeater to the measured current/voltage characteristic must be removed.

Conventional OTDR is very accurate, but only applicable before the first repeater or regenerator. Advanced COTDR is applicable beyond optical amplifiers, provided they still work properly (they must be power-supplied) and meet the requirements of COTDR. It is potentially very accurate on type 4 faults. However, the real-world performances of this technique are often disappointing, and it requires a rather expensive hardware. Also, it may be difficult to implement by non-specialists.

Obviously, optical path measurements do not have the capability of locating all faults, depending on the submarine cable fault nature, namely in the case of:

• a shunt fault, where the optical fibers have not been damaged,
• a remote fault, when the cable technology does not comply with the requirements of the COTDR,
• any fault that does not produce a significant reflection of light at the fault location, with a good enough signal-to-noise ratio at the measurement point,
• a remote fault, when the capability of power-feeding the cable has been lost, for instance with a type 2 fault.

3. Available DC fault location equipments

There exist three ways of implementing DC fault location techniques:

• using the PFE for manual fault location,
• using a conventional DC fault location bridge for submarine cables or a capacitance meter for submarine cables,
• using an automatic apparatus for DC fault location: the L2G.

Using the PFE for manual fault location is mostly applicable to type 1 or type 3 faults. It has the advantage of not requiring additional investment. It has the following problems:

• it is prone to human errors in the measurement and/or data processing and requires skilled personnel,
• it can lead to large errors in the fault location estimates when applied to type 1 faults because it is not possible to remove the contribution of the fault from the measured voltage and currents values,
• it is therefore mostly applicable to type 3 faults, when current balancing at high current is applicable,
• even in this case, it can lead to fault degradation and lowered fault location accuracy in case of improper timing or improper accuracy or error in the current values on type 3 faults,
• a PFE is unable to give accurate reading with a noisy cable or ground or fault behaviour,
• a PFE is not a measuring instruments: they do not have a good accuracy, especially at low currents or voltages,
• a PFE are not frequently calibrated and cannot be implemented according to ISO 9000 type quality assurance.

Conventional DC fault location bridges have the following limitations:

• they are very complex to implement,
• a manual processing of the measured value is required for removing the contribution of the repeater and the fault, using fault location tables,
• they have a limited current feeding capability which renders them useless in many cases,
• they do not support current balance.

Of course, L2G fault location systems overcome all these problems. That's what they were designed for !