Estimating frequency stability margin for flexible under-frequency relay operation
Introduction
One of the main reasons for the extreme level of efficiency of modern electrical power systems (EPSs) lies in how emerging technologies are treated and implemented by power system utilities. A very conventional approach in the process of harvesting capabilities of new technologies often seems rather unnecessary and difficult to comprehend. However, throughout the years, EPSs have become the leading infrastructure in supporting modern society's existence and so indispensable in a variety of ways. They became the most notable representatives of the so-called critical infrastructure. Many areas of human activities are so dependent on an EPS's operation that conditions emerging from blackouts are often unimaginable (social unrest, food shortage, increased crime rate, etc.) [1]. For this reason, system integrity protection schemes (SIPS) were introduced to prevent system operating conditions leading to blackouts. Major system imbalances between active power generation and consumption are being dealt with by under-frequency load-shedding (UFLS) protection. The conventional UFLS setting ([2,3]) appeared satisfactory up until the last decade or so when the penetration level of intermittent converter-based generation units began to seriously interfere with EPS inertia. Consequently, EPS frequency responses to active power imbalances became more turbulent in terms of the increased Rate of Change of Frequency (RoCoF). This led to the need for the under-frequency protection devices to enable RoCoF calculation in real time. In order for the measured RoCoF to be highly accurate, a time window of at least 100 ms is required. Nevertheless, RoCoF obtained from a shorter time window can also give some rough indications of a potentially dangerous situation, however, one should handle its value with care. Despite the above, many algorithms that extract appropriate information from RoCoF are still under research.
Several approaches to using RoCoF for UFLS purposes can be found in the existing literature. The large majority of them rely on knowing the relation between the active power imbalance and RoCoF that corresponds to a single synchronous machine (e.g. [4]) – the swing equation [5]. Substituting the entire set of individual machines with an equivalent generator enables one to study the average EPS frequency response, usually referred to as the Center of Inertia (COI) [6]. However, such solutions have a tendency to increase the UFLS complexity with the involvement of wide-area communication and the estimation of (mostly unknown) EPS inertia (such as [7], [8], [9], [10]). In contrast to such suggestions, this paper formulates the technology of a RoCoF-based modification of conventional UFLS based entirely on locally-obtained measurements and the estimated value of the frequency stability margin, which helps to recognize the need for immediate UFLS activation. A brief preliminary and elementary explanation of the concept in terms of basic principles and initial observations can be found in [11], whereas this paper presents its extensive in-depth formulation and analysis. The modification introduces an additional shedding criterion which significantly increases the scheme's flexibility to changes in operating conditions. This criterion is developed from an innovative representation of operating conditions in a newly-defined plane, which differs from commonly used frequency versus RoCoF planes ([12,13]). Apart from this, several further possibilities for user-defined UFLS settings emerge from the proposal that might open up a wide array of additional variations in the future. The concept was proven in [14] with real-time digital simulator in a hardware-in-the-loop setup that also served for improving RoCoF filtering technique and supporting the applicability of the approach.
Section snippets
Basic facts
Currently, most of UFLS schemes used in practice are of the conventional type, encompassing several stages ([5, 15,16]). The setting of each stage includes its size (amount of disconnected consumption/load, in practice achieved by tripping appropriate feeders) and the corresponding frequency threshold at which it is being activated. However, load feeders (assigned to an individual stage) are subjected to different power flows depending on different factors, such as the season within a year, day
Background
The idea for the presented RoCoF-based criterion arose from the existing set of predictive UFLS suggestions. The feasibility of the initial attempts of applying the second time derivative of the COI frequency [21] for short-term prediction of the frequency response was indeed questionable in terms of practical applications. However, they inspired several publications that followed, each of them eliminating a few questionable elements. As a result, a WAMS-based predictive UFLS was suggested in
Power system model
For a case study of the operation and efficiency of the presented approach, we used an existing EPS model. It encompasses a part of the 110 kV network of the Slovenian EPS. Since this model was successfully validated against PMU measurements before [20], the authors consider it appropriate. It consists of twelve hydro units with synchronous generators having an equivalent inertia constant of 6 s (exciter, governor) and nine substations. We modeled each out of the nine substations as having ten
Conclusions
In this paper, a transparent and effective modification of a conventional UFLS relay setting is presented. It employs RoCoF for real-time frequency stability margin calculation, which is further monitored against the potential violation of newly-proposed frequency stability margin thresholds. The authors suggest implementing this new criterion in existing under-frequency relays alongside the already existing one. As a result, conventional UFLS gains the required flexibility with minimum
CRediT Author Statement
Urban Rudez: Conceptualization, Methodology, Software, Validation, Resources, Writing - Original Draft, Writing - Review & Editing, Visualization, Supervision, Funding acquisition.
Denis Sodin: Formal analysis, Investigation, Data Curation.
Rafael Mihalic: Project administration, Funding acquisition.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This work was funded by the Slovenian Research Agency through the Electric Power Systems No. P2-0356 research program and the Resource management for low latency reliable communications in smart grids - LoLaG, J2-9232 project. The authors would like to thank the Slovenian Research Agency for the financial support. This work is subject to a pending International Patent Application No. PCT/EP2018/059048 filed on April 9, 2018.
References (24)
- et al.
A review of the state of the art of UFLS schemes for isolated power systems
Int. J. Electr. Power Energy Syst.
(2018) - et al.
Application of load shedding schemes for distribution network connected with distributed generation: a review
Renew. Sustain. Energy Rev.
(2018) - et al.
An adaptive load shedding method with time-based design for isolated power systems
Int. J. Electr. Power Energy Syst.
(2000) - et al.
Applying a wide-area measurement system to validate the dynamic model of a part of European power system
Electr. Power Syst. Res.
(2015) - et al.
A novel approach to underfrequency load shedding
Electr. Power Syst. Res.
(2011) - et al.
Power systems with high renewable energy sources: a review of inertia and frequency control strategies over time
Renew. Sustain. Energy Rev.
(2019) - et al.
Blackouts: a sociology of electrical power failure
Soc. Space
(2014) - et al.
New proposed adaptive frequency load shedding scheme for cogeneration plants
- et al.
Minimal load shedding using the swing equation
IEEE Trans. Power Syst.
(2017) Adaptive underfrequency load shedding based on the magnitude of the disturbance estimation
IEEE Trans. Power Syst.
(2006)