Researcher in power system planning and operation issues

Current position: Assistant professor at Department of Electrical Sustainable Energy, Delft University of Technology, Delft, Netherlands.


Areas of expertise:
- Power system reliability
-  Power system planning and operation considering smart grid aspects
- Large scale integration of wind power
- HVDC links and FACTs devices
- Modal identification
- Power system modelling & data format conversion
- Probabilistic and artificial intelligence based security assessment and enhancement
                                               - Mean-variance mapping optimization (MVMO)

Ongoing projects:
- MIGRATE H2020 Project (Starting in January 2016)

Metaheuristic optimization for power system applications

The main target is to provide a thorough assessment and further development of different metaheuristic-based strategies to tackle the complexity and scalability of a variety of power system optimization problems. Among these problems are optimal power flow calculation, power system planning, large scale transmission pricing, hydrothermal scheduling, identification of dynamic equivalents from signal records, as well as location and coordinated tuning of damping devices. Initial research is being pursued to develop tests beds where existing computational techniques can be benchmarked. Development of new techniques is also targeted. So far, the outcomes of the initial research conducted in this field (as part of my Habilitation work in Germany) are expected to contribute to the technical activities of the IEEE PSACE Working Group on Modern Heuristic Optimization (WGMHO). 

Mean-variance mapping optimization (MVMO)

MVMO is a novel optimization algorithm whose conceptual framework shares some similarities to other heuristic optimization approaches. However, the most salient feature of MVMO is its use of a special mapping function applied for mutating the offspring on the basis of the statistics of a dynamically stored and updated population. 
Thanks to the well-designed balance between search diversification and intensification, MVMO possesses a fast convergence characteristic so that it can find the optimum solution quickly with minimum risk of premature convergence. To date, MVMO has been successfully applied to solve different power system optimization problems: the optimal reactive power allocation problem, the optimal dispatch of energy and reserve, the development of optimal control strategies, and the identification of dynamic equivalents.

Power system planning & operation 

In the liberalized electricity market framework, power systems are prone to be operated closer to their technical limits due to several factors such as market pressures, shortfalls in reserve margins and insufficient new investments, to name a few. Thus, from power system operation perspective, much work has been recently directed towards the development of the smart grid concept. Motivated by development experienced in the power electronic and the telecommunication technologies, the key idea is to make efficient use of sensory information to enhance system performance through timely provision of information and control options. Nevertheless, this has also led to renewed interest on several technical and economic issues involved in power system planning to make the best use of existing and planned transmission system assets in order to overcome the challenges brought about complexity of modern power systems.

Large scale integration of wind power

With increasing wind power generation, interest to fully understand and quantify the impact of this development on the performance of the interconnected systems has also grown. While individual onshore wind units in the past were only required to maintain a prescribed power factor range at the point of interconnection, now transmission system operators (TSOs) are putting more wide-ranging conditions. Some TSOs have issued grid codes spelling out a range of operational requirements, which wind farms need to fulfill upon connection to the grid. These include low voltage fault ride-through capability and voltage support following grid faults. Requirements concerning additional ancillary services are also likely to follow. It should also be borne in mind that the control systems which modern wind turbines employ are characterized by fast response time and thus open up additional, unconventional options to help meet these requirements.

Modal identification in real time

Much work has been directed towards the development of methodological approaches for identification of critical low-frequency oscillatory modes, namely model-based methods and measurement-based methods. Model-based methods (e.g. linear-analysis based eigenanalysis) are considered off-line methods, since they depend on approximate data and modeling processes. By contrast, measurement-based methods rely on actual response of a power system (i.e. signal records obtained from system measurements or from time domain simulations). Thus, these methods are of interest for monitoring and analysis power system low-frequency oscillations in real time. To date, several measurement based techniques have been developed, each one with its own advantages and limitations. Nevertheless, accurate estimation of modal parameters and their temporal changes (i.e. nonstationarities) is still a great challenge, specially for ambient data.

Wide area damping control

Many utilities have established security requirements which should be satisfied in power system planning and operation in order to ensure well damped oscillations. Amongst these is the damping ratio threshold. Moreover, it is well known that due to the inherent limitations of power systems to damp out low frequency oscillations, power system controllers and some countermeasures play a vital role in enhancing system damping performance.
From power system operation perspective, much work has been directed towards the development of methods for improving power system small-signal stability performance through operational strategies such as, for instance,  generation rescheduling methods. Other innovative methods suggest optimal power flow computation including small-signal stability constraints. 
Besides, considerable effort has gone into the research area of supplementary damping controllers. The traditional approach to damping control design in a power system is to add power system stabilizers (PSSs), which are mostly single-loop local controllers. But this kind of control design cannot always prove to be effective in damping inter-area modes. Local controllers lack global observability, mutual coordination and placement flexibility. Nowadays, there are important efforts to improve supplementary damping control, monitoring and supervision in power systems, which are mainly related to the development of the wide-area measurement system technology (WAMS).  WAMS involves distributed phasor measurements throughout the network by means of phasor measurement units (PMUs). However, further research is still needed to perform coordinated tuning as well as proper selection of the location and the input signal of the wide-area supplementary damping controllers in order to ensure robust performance.

The role of PMU signals

A strategic placement of phasor measurement units (PMUs) is strictly necessary to effectively ease real-time processing and interpretation of huge amounts of data arriving at a high sampling rate. This aspect is especially crucial for the development of online dynamic security assessment (ODSA) and real-time dynamic vulnerability assessment (RTDVA) applications, since their reliability is strongly dependent on high observability of system dynamics, which can be achieved by selecting an appropriate set of PMUs whose measurements (containing relevant features associated to fast and slow dynamic phenomena) should be trusted to ensure a satisfactory dynamic tracking performance.


Data warehousing and pattern recognition

PMU data can further entail other valuable benefits for different tasks such as real-time analysis in future time periods, post-mortem studies, as well as parameter identification and calibration of system models. Thus, research effort will be also directed for the development of a central/distributed repository of historical operational data to be interfaced with wide-area measurement and control schemes (WAMC). Remarkably, outlier detection and data compression constitute key issues to be considered in preliminary stages before recording/archiving new data. Mining the continuously/periodically updating statistics of system performance would be extremely profitable for efficient planning and operation of power systems. The application of data mining techniques would be of great value to discover the dynamic attributes (e.g. relationships/insights) of the system performance as well as to empirically estimate the system dynamic security and vulnerability regions (DSRs and DVRs) based on patterns associated to different stability phenomena.