The programme has to cover three things that a single test cannot. The condition of the cable as delivered and installed, the integrity of the joints and terminations made on site, and the continuity and condition of the sheath and bonding system. Each is established by a different measurement at a different stage.

What an acceptance programme must cover

The first stage is a check on the cable itself. A drum length that was damaged in transit, or a section that was over-stressed during pulling, may pass every subsequent test until the point at which it fails in service. Testing each section before it is jointed isolates that risk to a specific length and a specific activity, and it means a failure found later can be attributed.

The second stage is the completed circuit. After jointing and termination, the circuit is tested as a whole, and this is the test that proves the joints and the terminations along with the cable body. Because the joints are made in field conditions, this is where most installation defects appear, and it is the test that the acceptance decision rests on.

The third stage covers the sheath and bonding system. The outer jacket, the metallic sheath, the bonding links and the earthing arrangement form a system that protects the cable from moisture and provides the fault current path. Testing it separately from the conductor insulation is necessary because a sheath defect does not affect the insulation measurements taken on the conductor.

Voltage class and route type decide the scope

The scope differences are driven by where the defects occur and how expensive they are to find. A route with many buried joints justifies the additional stage of testing before jointing, because the alternative is excavating under finished reinstatement for a defect that could have been identified earlier.

Testing before and after jointing

Testing before jointing has a practical constraint: each section has to be accessible and its far end isolated, which is straightforward while the trench is open and difficult afterwards. The measurements that suit this stage are those that can be applied to a length of cable with both ends accessible and no terminations yet made.

Testing after jointing is the acceptance test proper. It is applied from the ends of the completed circuit and it covers everything installed. The sequencing constraint is that all joints and terminations must be complete before it runs, because an unfinished end cannot be energised for the test.

The two stages complement each other in the record. A defect found at the second stage can be cross-referenced against the first-stage results for the section it falls in, which is often enough to establish whether the cable or the joint is responsible. Without the first-stage results, the attribution has to be made from the excavation.

Sequence across a multi-section route

A multi-section route has a natural sequence that follows the construction order. Each pulled section is tested when it is placed and before the next joint is made. Joints are then completed and, where the design includes them, the bonding connections are made between sections. The completed circuit is tested as a whole.

The sequence matters because the test covers what has been installed up to that point. Testing the whole route before the bonding and terminations are complete produces a result that does not describe the finished circuit, and repeating the test afterwards doubles the work. Sequencing the stages to match the construction programme avoids both.

Where a route is built in phases and energised section by section, each phase has to be tested as it is completed, because a section that is energised cannot later be subjected to an acceptance test without being removed from service. The programme should be written to match the energisation plan rather than the construction plan, since those two can differ.

Diagnostic tests worth including at acceptance

The diagnostic measurements that suit acceptance are those that produce a baseline. A partial discharge measurement taken during the withstand test establishes the discharge behaviour of the circuit at a known stress level when the circuit is new. A dielectric loss measurement, where the arrangement and cable type allow it, establishes the loss behaviour of the insulation.

Both are more valuable at acceptance than at any later point, because at acceptance the circuit is in its best state. A measurement taken years later can only be interpreted against something, and the something is either the acceptance measurement or nothing.

The measurement configuration matters as much as the value. The coupling arrangement, the detection threshold achieved in the noise environment, and the test voltage at which the measurement was taken all have to be recorded, because the next measurement will only be comparable if the configuration is reproduced.

Documentation the owner should receive

The handover package should contain the test results for each stage with the conditions recorded, the withstand test record with the method, the waveform and frequency where applicable, the voltage, the duration and the observed current behaviour, and the sheath test results with the bonding configuration used.

Alongside the electrical results, the package should contain the joint and termination schedule with the installation conditions recorded, the as-built route in plan and section with joint positions, and the cable data that determines the propagation velocity. Those items are what make the first fault on the route cheaper and faster to locate.

The installation conditions for joints deserve particular attention, because moisture and contamination during assembly are among the commonest causes of early joint failure. Recording ambient conditions during assembly is a small effort that produces a large benefit when a joint behaves unexpectedly in its first years.

Failure handling during acceptance testing

A failure during acceptance testing is a valuable event if it is handled properly. The test has found a defect while the contractor is on site, the trench may still be open, and the responsibility is still attributable. The handling that preserves those advantages is to locate the failure rather than to repeat the test on the assumption that it was spurious.

Location follows the same two-stage logic as a service fault: pre-location to establish the distance, then pinpointing to establish the position. Because the route is likely to be accessible and its features recorded, the search is usually faster than a fault on an established route, and the position found can be compared against the joint schedule immediately.

The repeat test after the repair should be the full acceptance test rather than a shortened version, and the record should show the original failure, the repair, and the result of the repeat. A circuit that has failed once during commissioning is a circuit whose history should be visible to whoever maintains it.

Building the programme for a tender

Where the programme is a tender requirement, the specification should state the stages, the method with its standard reference, the acceptance criteria for each stage, the diagnostic measurements required and the documentation to be delivered. Leaving any of those to be settled during construction transfers the decision to the contractor, who will choose the interpretation most convenient to the schedule.

The specification should also state who witnesses and who accepts each stage, and what happens when a stage fails. A failure handling procedure agreed in advance prevents the situation where a failed test is repeated until it passes without the failure being recorded.

The field test requirements for the relevant cable classes are set out in IEC 60502-2 for medium voltage extruded cables, in IEC 60840 for the higher voltage classes and in IEC 62067 for the highest extruded classes, with the general field test framework in IEC 60060-3. Cable system installation documentation from manufacturers such as Prysmian and Nexans states the mechanical and environmental limits the installation has to respect. The cable testing range is grouped on the cable fault testing hub, and comparable supply and commissioning work is listed under typical achievements.