Opportunities and Challenges for PMU Deployment in Distribution Systems

Phasor Measurement Units (PMUs), which measure voltage and current synchronized phasors (synchrophasors), along with frequency and rate of change of frequency (ROCOF), are the emerging measurement devices for power network monitoring. Introduced in the late 1980s and first deployed in experimental systems in the early 1990s, they are now assuming a key role in protection and control of modern electric grids, in particular for state estimation in transmission networks.

A PMU bases its capability of measuring absolute phase-angles on the availability of a common time reference (with respect to Coordinated Universal Time), usually provided by the Global Positioning System (GPS). The other main characteristics of the PMU are increased accuracy and high reporting rates, as well as digital communication interface for measurement transmission and command reception.

Due to the wide range of possible applications, many pilot projects have been developed in recent years and hundreds of PMUs have been installed, in particular by Transmission System Operators. For the same reason, the number of PMU manufacturers is constantly increasing from the few pioneer companies of the 1990s to tens of producers now. Furthermore, several Intelligent Electronic Devices used in electric substations, mainly for protection purposes or other measurement instruments (e.g. power quality monitoring), have been updated to introduce PMU functionalities.

PMUs have undergone a troubled standardization process, starting with the first IEEE standard published in 1995, continuing through two revisions, and (temporarily) ending with the last amendment released in March 2014 to either modify or suspend some of the previous performance requirements. As a general perspective, the standardization leaves PMU manufacturers the choice of design solutions, giving only specifications for synchrophasor, frequency and ROCOF measurements under steady state and dynamic test conditions. It defines the indices, in particular the Total Vector Error (TVE), for PMU accuracy evaluation and comparison. The standard IEEE C37.118.1 introduces two performance classes: a P-class, particularly intended for applications requiring fast responses, as the protection ones, and M-class, requiring higher accuracy for measurement applications.

Today, the rapidly evolving scenario of distribution grids, where the increasing presence of distributed generation and storage asks for increased measurement accuracy, faster reporting rates and higher communication capabilities as a prerequisite for complex control and management applications, makes PMUs a major perspective also for such networks. In this context, synchronized measurements should be integrated with traditional monitoring systems, such as Supervisory Control and Data Acquisition (SCADA), which use traditional measurement devices, without absolute phase-angle information, and operate at lower reporting rates.

However, the direct use of currently available off-the-shelf PMUs in the distribution networks should be examined closely. Indeed, the performance offered by current PMUs, which are normally suitable for applications in transmission grids, can be in some cases insufficient for the distribution systems, for the following reasons:

  • lower distances in distribution grids imply lower amplitude and phase differences between the electric quantities in the different nodes: higher accuracy could therefore be required to correctly measure these differences;
  • the faster dynamics that are expected in smart grids, owing also to the possible intermittent behavior of generators, loads and storage devices, may require time responsiveness beyond the limits indicated in the standard;
  • the higher distortion that is usually present in distribution grids may lead to a need for re-defining compliance limits with harmonic and interharmonic interferences, again beyond the current standard;
  • the lack of human supervision requires higher trustworthiness of the information coming from PMUs, so it can be used in automatic control and protection routines.

Therefore, the compliance of the PMUs to the IEEE standard does not guarantee, by itself, that the PMU is suitable for a specific practical application. This problem is intrinsic in the parameter chosen in the standard to represent the accuracy (the TVE is a synthetic index that summarizes amplitude and phase-angle errors, but does not separate these two sources of error), but is emphasized in the case of distribution grids, for the reasons listed above. Thus, it is highly advisable that commercial PMUs are accompanied by a set of detailed metrological specifications, so that the designer of the distributed measurement system can effectively assess their adequacy for the problems at hand.

To this purpose, it should be taken into account that the performance of the PMU, especially under non-stationary conditions, strongly depends on the algorithms used to evaluate synchrophasors, frequency and ROCOF. A PMU could therefore include, in its software suite, a set of possible algorithms which can be appropriate for different operating conditions. The detailed knowledge of the metrological characteristics of the different procedures would allow the measurement system to be tailored to the specific application needs: a custom PMU, while keeping compliance with the standard, and thus interoperability with other commercial devices, could also exhibit its own additional useful peculiarities.

These last considerations open the door to a further critical issue, namely the experimental characterization of PMUs, which is required to obtain their actual metrological specifications. To this purpose, the IEEE standard C37.242 was released in 2013 as a guide for PMU calibration, testing and installation. This document is, obviously, strictly related to the general IEEE standard on synchrophasors and the compliance tests defined therein. Consequently, it should again be seen as the basis to ensure PMU interoperability, but a wider exploration under more realistic conditions would be necessary to verify possible further requirements requested to PMUs employed in the distribution system, as discussed above. The assessment of the metrological characteristics should in any case take into account, besides the effects of the algorithms and the synchronized data acquisition system, those of the voltage and current transducers. These introduce additional uncertainty which is unavoidable, and their behaviour crucially important in the evaluation of the overall measurement accuracy.