Moving the focus from relay element testing to protection system testing

The last three decades have seen significant improvements in power system protection. With the rise of digital relays, recalibration during maintenance tests became obsolete. The model of one IEEE element per electromechanical relay is not the case anymore. Digital relays build a system of elements freely combinable with programmable logic conditions. Additionally, to be able to protect the increasingly complex power system, these relays communicate with other relays, building an even bigger system. This being said, with few exceptions, like the use of RTDS (real-time-digital simulator) to simulate real world test cases, protection testing has neither changed significantly since the days of one element per electromechanical relay, nor adapted to the new challenges.Still the ultimate goal of testing stays the same - ensuring that life and equipment is protected and power system stability is maintained. To achieve this goal we have to test the protection system, and all processes leading to it, as a whole. The protection system of today is usually the product of a process involving several engineers, technicians, and in some cases, companies. Applying a test only to a relay, checking each individual element one by one, at the very end of the process is insufficient as it cannot cover all possible sources of error. This was underlined by the 2013 NERC Misoperations Report which showed that in today's protections systems, relay failures are not the single source of error anymore, but rather incorrect settings, logic and design errors already account for more misoperations. This paper proposes quality assurance for the whole process from design to commissioning by using a system testing approach. By elaborating in detail the challenges and different sources of error in modern protection systems, we will show how system testing tools can complement current testing methods or even replace some tests that no longer add value. Different testing approaches will be measured against improved quality vs. effort to run. Finally, the paper will share experiences from several cases around the world where system testing approach was applied with great success, and cases where gaps in system traditional testing methods were identified as areas of improvement.