When a specialist contractor has completed a pre-commissioning or commissioning activity the associated test documentation will be submitted for review.
The commissioning manager shall check the documents for completeness and correctness (it is surprising how often incorrect data is used) and, if satisfied, arrange a witness demonstration.
Typically the commissioning manager shall witness a sample of the offered system to determine if the commissioning has been performed correctly.
If the results from the sample match the recorded values then the witness is generally considered to be successful.
If the outcome of the initial witness is not satisfactory then the sample size should be increased.
Should the demonstration continue to prove unsuccessful then it shall be marked as a failure.
The contractor shall need to re-commission and re-apply for a witness.
Sample sizes tend to be in the region 10-20% of the total system (e.g. if 200 valves are measured then a witness would typically be around 20-30).
This is not a fixed value and the most important outcome is that the commissioning manager has confidence the system has been correctly commissioned and the recorded results reflect the current status.
For some systems, such as life safety systems, we prefer to perform a 100% witness.