Testing ROCOF protection
Author: Andrea Bonetti
Rate of Change of Frequency (ROCOF) protection is mandatory in most countries for distributed and embedded generation schemes that are directly connected to the power distribution network. But what exactly is ROCOF protection, why is it necessary and how can it be tested? Andrea Bonetti of Megger has the answers.
While the spread of distributed generation (DG) systems supplying power to national distribution networks brings many benefits, particularly in terms of reduced environmental impact, it also creates new challenges. One of these relates to the phenomenon of ‘islanding’, which occurs when a DG system loses its network connection but continues to power local circuits – with the now isolated system operating as an ‘island’.
At first sight, this may seem like a benefit, as the local circuits will continue to receive power despite the loss of the grid connection. In practice, however, islanding is in most cases undesirable and potentially dangerous. The reasons are many: without a grid connection, the DG system may not deliver power at the correct voltage and frequency, which means that there is a risk of equipment damage. Engineers working on the system may encounter cables that are unexpectedly live, and reconnecting an islanded system to the grid is often problematic because of the need to match frequency, phase and voltage. For these reasons and others, distribution network operators (DNOs) in most countries insist that DG systems that are to be connected to their networks must incorporate equipment that will rapidly detect islanding and, when it occurs, immediately shut down the local generation. Over time, many methods have been trialed - including undervoltage and transient detection. However, achieving the necessary fast response while minimizing the risk of disruptive nuisance tripping proved challenging until an alternative approach, which looks at changes in the supply frequency, was adopted.
This is rate of change of frequency (ROCOF) protection and is based on the idea that if a local generator is suddenly disconnected from the grid, provided that the power flow to or from the grid at that time is not the zero, frequency of the supply that the generator is producing will change. Since analysis has shown that the probability of a DG system that’s connected to the grid having no net inflow or outflow of power is very small, this frequency change is a reliable and virtually instant indicator that the grid connection has been lost.
For this reason, ROCOF protection is now enshrined in national standards such as ANSI 81R and G59/3. And, since February 2019, there has been a new protection relay standard from the IEC – IEC 60255-181 – which specifies performance and test methodologies for the relays used to implement ROCOF protection.
Figure 1: Mandatory equation for generating frequency ramp
In fact, IEC 60255-181 prescribes a very specific testing regime for ROCOF relays. Fundamentally, this relies on using a test signal that takes the form of a ‘frequency ramp’, which means that the frequency of signal increases smoothly over time. Using a ramp of the correct form is essential if the performance of the relay is to be properly and reliably evaluated and, for this reason, the standard requires that the ramp must follow a defined mathematical formula, which is shown in Figure 1. The standard does not allow any other test method or form of test signal to be used.
Figure 2: Smooth frequency ramp, as mandated by IEC 60255-181
For the purposes of this article, it isn’t necessary to analyse the frequency ramp equation in detail, but it is important to note that it describes a waveform where the frequency changes are smooth and continuous. An example of such a waveform is shown in Figure 2.
The standard makes it clear that it is not acceptable to use a test signal in the form of a frequency ramp of the form shown in Figure 3, where the frequency changes in discrete steps. This applies no matter how small the frequency steps may be.
Figure 3: A ramp where the frequency changes in discrete steps cannot be used
While it may at first seem that such a precise definition of the test waveform is unnecessarily pedantic and restrictive, it is in fact of fundamental importance. This is because many unwanted operations of ROCOF relays in service have resulted from misunderstandings about what form the test waveform should take. Experience has shown that testing with a frequency ramp where, for example, the frequency changes in discrete steps of 0.04 Hz every 20 ms, which is notionally equivalent to a ramp rate of 2 Hz/s, may not give the same results as using the prescribed test waveform where the frequency changes continuously at 2 Hz/s rather than in steps.
Fortunately, relay test instruments are now available that provide facilities for testing ROCOF relays using methods that fully satisfy the requirements of the new IEC 60255-181 standard. An excellent example is the versatile and cost-effective SVERKER 900 test set from Megger, which features intuitive operation via a large color touchscreen and is suitable for testing not only ROCOF relays, but also almost all other types of relay used in modern protection schemes.
With a test set of this type, the testing and setting up of ROCOF relays become straightforward tasks, which means that the relays can be trusted to perform their essential safety and operational functions by guarding against the disruptive and potentially dangerous consequences of undetected islanding in power networks.