The consequence is that testing activities are unusually sensitive to late predecessors, and the recovery options are limited because the method and the acceptance criteria cannot be compressed.

Where testing sits in a turnkey programme

Each asset follows a sequence: civil works, erection, cabling and wiring, testing, and energisation. The testing activity depends on everything before it and constrains everything after it, which is the definition of a critical activity.

The testing itself is usually fast relative to the construction activities. The reason it becomes a schedule problem is not its duration but its position: it is the last activity before the date the client is waiting for.

That position also makes it the activity that absorbs every previous delay. A cable installation that finishes a week late moves the test a week later, and the energisation date moves with it unless there is float elsewhere.

Dependencies on civil and erection works

The as-built drawings deserve particular attention. Testing against a design drawing that does not match the installation produces findings that cannot be attributed, and re-drawing after the test is more expensive than drawing before it.

Float and critical path around test activities

A test activity with float can absorb a late predecessor without affecting the energisation date. One on the critical path cannot. The programme should state which is which, because the recovery options differ.

Where testing has no float, the practical responses are to start parts of it earlier by staging, to add resources so that activities that could be sequential run in parallel, and to remove dependencies that are not essential.

Adding resources works where the constraint is the number of teams and does not work where the constraint is access. On a compressed programme the two are often confused, and the response of adding people to a single-access activity lengthens it rather than shortening it.

Staged energisation versus single commissioning

Staged energisation usually shortens the overall programme. The first stage is tested, energised and put into service while the later stages are still being completed, and the client gains capability earlier.

It requires the protection, control and interlocking arrangements to support partial operation, which is a design decision that has to be made early. Retrofitting staged operation into a design that assumes a single commissioning is expensive.

Staging also changes the testing scope. Each stage has to be proved as a system in its own right, including the interfaces to the stages not yet built, and those interfaces have to be tested again when the next stage is added. The saving in programme time is real, and the additional testing it creates should be included in the plan.

Coordinating multiple contractors

A turnkey programme involves several organisations whose work interleaves: civil, erection, cable installation, protection and control, and the testing organisation. The testing schedule has to be agreed across them rather than within each.

The practical arrangement is a single commissioning schedule with each activity’s prerequisites and completion evidence stated, owned by one coordinator. Where each contractor sequences its own work, the interfaces between them are where the programme breaks.

The coordination requirement extends to the equipment. Test instruments are shared between activities, and the schedule should show where the same equipment is required by two activities at once. That conflict is easier to resolve in planning than in the week it occurs.

Managing re-testing after defects

Re-testing after a defect consumes the window allocated to the original test, and on a compressed programme there is often no window left. The schedule should therefore include an allowance for re-testing based on experience rather than assume a first-time pass.

The allowance should be expressed as time rather than as a contingency sum, because the constraint is the window and not the money. A programme with a contingency budget and no schedule allowance will still miss the energisation date.

Where a defect is found, the sequence should be defined: stop, contain, establish the cause, repair, re-test and record. Establishing the cause before the repeat is what prevents the same defect being found again in the next stage, and it is the step most often skipped under schedule pressure.

Common schedule failures and their cost

Late civil works delay access to the area. Incomplete cable installation delays the circuit tests. Missing as-built drawings delay the verification. Access restrictions delay the mobilisation of the test team.

Each of those delays the testing rather than causing it to fail, and each of them is visible in the programme before the outage date if the dependencies are stated and monitored. The failures that surprise a programme are the ones whose dependencies were never written down.

The cost of a delay is not linear. A delay that pushes the energisation date past a contractual milestone carries the consequence of that milestone, which is usually larger than the cost of the testing activity itself. That asymmetry is the argument for protecting the test activities with float wherever the programme allows.

Reporting progress on test activities

Progress reporting on testing should be by activity and by acceptance, not by percentage. A test that is ninety percent complete produces no evidence; a test that is complete produces a result that can be accepted.

The reporting should identify the activities that are complete with their evidence, the activities in progress with their status, and the activities not started with their outstanding prerequisites. That structure shows where the programme is blocked and what has to be done about it.

Where a test has produced a finding, the reporting should show the finding and the corrective action rather than only the delay. A schedule that shows a slipped date without showing the reason gives the programme manager nothing to act on.

Handling re-testing inside the same outage window

Where a repeat test has to be performed within the same window, the plan should allow for it by sequencing the tests so that the repeat can be inserted without disturbing the rest. That usually means performing the tests that are most likely to need a repeat first.

The window should also allow for the diagnosis. A failed test that is repeated immediately, without establishing the cause, is likely to fail again, and the second failure consumes the remaining window with no more information than the first.

Where the window cannot accommodate the repeat, the decision to defer is legitimate provided it is recorded with its consequence. The requirement on the programme is then to find the time later rather than to assume the test passed. Installation and acceptance requirements that the programme ultimately demonstrates are set out in IEC 61936-1; the transformer evidence required is defined in IEC 60076-1; the switchgear requirements are in IEC 62271-1; independent acceptance practice is published by NETA; programme and asset management practice is coordinated through CIGRE study committees and published by EPRI; and the equipment and support arrangements for the work are described under testing services and OEM/ODM solutions and all electrical testing equipment.