4760 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 28, NO. 4, NOVEMBER 2013 Optimal Integration of Disparate C37.118 PMUs in Wide-Area PSS With Electromagnetic Transients Innocent Kamwa, Fellow, IEEE, S. R. Samantaray, Senior Member, IEEE, and Geza Joos, Fellow, IEEE Abstract—The present paper investigates the effective ap- plication of phasor measurement units (PMUs) with different technologies and filtering dynamics to the design of wide-area power system stabilizer (WAPSS). Two frequency-adaptive PMU algorithms shown to exceed the C37.118 class-M requirements based on FIR- and Kalman-bandpass filtering are selected for study. Their sharp dynamic differences are characterized using accurate fourth-order transfer functions derived from step re- sponses using system identification techniques. While FIR-based PMUs have no overshoot in their phase and amplitude step re- sponses, the adaptive Kalman filter allows for better out-of-band interference rejection at the cost of a phase step response with overshoot and a higher computational burden. These PMUs are implemented in Simulink and then embedded in electro-magnetic transients (EMT) simulations to experiment the realistic WAPSS behavior with the discrete-time devices in the loop. It is shown that WAPSSs with optimized settings can improve the damping performance of EMT models of a test system and a simplified Hydro-Québec network including PMU filtering dynamics and transmission delays. However, WAPSSs based on EKF-PMUs yielded better performances, thanks to their shorter group delay compared with FIR-PMUs. PMU devices and algorithms meeting standard C37-118 can thus behave quite differently under stressed dynamics. Index Terms—Damping control, FIR, IEEE PSS 4B, IEEE Std. C37.118.1-2011, Kalman filtering, phasor measurement unit (PMU), synchrophasor, wide-area measurement systems (WAMS), wide-area power system stabilizer (WAPSS). I. INTRODUCTION P HASOR measurement units (PMUs) built to comply with IEEE Std. C37.118 [1] are now widely deployed by utilities [2], [3] on the grounds that synchronized phasor measurements are important preconditions for wide-area mon- itoring systems (WAMs) [4], [5]. As a matter of fact, recent findings have shown that the PMU technology is fast improving in terms of interoperability [6], dynamic performances [7]–[9], and algorithmic sophistication [10]–[13]. Although useful met- rics for comparing performances of various PMU brands are Manuscript received February 22, 2013; revised April 30, 2013; accepted June 04, 2013. Date of publication June 25, 2013; date of current version Oc- tober 17, 2013. Paper no. TPWRS-00217-2013. I. Kamwa is with Hydro-Québec/IREQ, Power System and Mathematics, Varennes QC, Canada J3X 1S1 (e-mail: kamwa.innocent@ireq.ca). S. R. Samantaray is with the School of Electrical Sciences, Indian Institute of Technology, Bhubaneswar, Odisha 751 013, India (e-mail: sbh_samant@yahoo.co.in). G. Joos is with the Department of Electrical and Computer Engi- neering, McGill University, Montreal, QC, Canada H3A 2A7 (e-mail: geza.joos@mcgill.ca). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TPWRS.2013.2266694 provided in [1], PMU algorithms vary in detail according to de- signer preferences and background. The dynamic performance can therefore be expected to vary as well, even for PMUs that meet the existing standard metrics. Further discussion of the potential impacts of varying PMU dynamics on wide-area control performance is provided in [13]. Ever since its inception in the mid-1990s [14], [15], sev- eral aspects of wide-area control have remained poorly under- stood because of the difficulty in performing the relevant ana- lytic or parametric study on a large power grid. The impact of the control device location as a function of power flow direc- tion (sending- or receiving-end sitting, for instance) are among the challenging issues [16]. However, real-time simulation of power systems dynamics with WAPSS and PMUs in the loop as performed in [17], [18] seems promising for investigating the potential side effects of this more risky control paradigm. In this paper, we pick a tool which fits between simplified stability (positive-sequence) software and the hardware-inten- sive real-time digital simulator, namely, the Electro-Magnetic Transients (EMT) Simulator, to advance our understanding of a WAPSS under realistic conditions. We start by summarizing the features of two C37-118-compliant PMU algorithms recently presented in [13] with a focus on control applications. They are then implemented in Matlab/Simulink while adhering to strict rules that allow their automatic implementation on a manufac- turer’s target platform using a fast prototyping framework [19]. Based on Simulink Real-Time Workshop features, the resulting discrete-time software PMUs are finally embedded into detailed EMTS of the widely known two-area test network to allow a full simulation of EMT behavior together with a bit-level repre- sentation of the PMU dynamics. Under these assumptions, the paper focuses on two objectives, outlined here. 1) Confirm the benefit of a WAPSS under more realistic assumptions than in previous literature, including PMU model-based optimized settings. 2) Study at bit-level the impact of PMU filtering dynamics on the system dynamic performance with respect to the PMU time delay and electromagnetic stress. It turns out that using an EMTS-based assessment as done in this paper (for the first time to the best of our knowledge), the WAPSS maintained the key potential benefits found in earlier studies performed in positive-sequence stability soft- ware [14], [16], [21], [26]. However, it also appears that meeting C37.118.1 Class-M metrics is not the only criterion to factor in when selecting a PMU for a WAPSS. A PMU group delay, fault ride-through characteristics and within-band frequency response characteristics are just as important aspects to consider. 0885-8950 © 2013 IEEE