PSI Blog

Cruising through the Energy Transition with an Intelligent Autopilot for Network Management

22 Oct 2019 - Energy, Research, Artificial Intelligence, Sustainability

Voltage presented as isosurfaces or as time-lapse film. Screenshot: PSI

The energy transition has substantially changed the power generation landscape. The decentralized energy infeed, the increasing share of renewable energies, and the exit from nuclear power make high demands on the networks and network management. Network operators are hardly able anymore to meet these demands without using intelligent software.

Teamwork approach results in autopilot for network management

Several years ago, the transmission system operator Tennet TSO GmbH faced the task to manage the increasing complexity. PSI and TenneT developed in close cooperation a smart autopilot for network management. PSIsaso (SASO - Security Assessment and System Optimization) is the solution combining standard methods for network state assessment and fuzzy logic algorithms.

SASO Security Assessment & System Optimisation: Autopilot for network management

PSIsaso is designed as a stand-alone system, which exchanges all required information via interfaces to the network control system and to other information systems. Using this information, it continuously analyzes and assesses the network state. In critical situations, PSIsaso acts as an assistance system and determines operational recommendations for the network management. Furthermore, it is also used as a development platform for new tools and innovative operational and visualization concepts.

Decision support for reactive power management

The enormous changes in the power generation landscape also affect the voltage regulation in the transmission network. The reduction of reactive power reserves results in reduced voltage stability. The power plant capacity is often no longer sufficient to compensate for local reactive power deficits in the network. For specific locations in the network, the system determines the required reactive power and the supplying sources which must be made available (for example; HV DC converters, compensation plants). The determination considers all network-related and market-related measures which are available to the transmission system operator.

Control plan for reactive power management. Screenshot: PSI
Control plan for reactive power management. Screenshot: PSI

PSI fuzzy logic algorithms are used for decision support and optimization of reactive power provisioning in order to assess operational alternatives with regard to various economic and operational objectives.

This approach combines the technical reliability with human flexibility resulting in the best possible precision. In critical network situations, the system provides a set of measures validated by the network state evaluation procedures as well as optimized with regard to the process objectives.

Dynamic stability analysis

An electrical energy system is considered stable if it returns to a reliable and balanced state after a fault. In cooperation with TenneT, PSI has been developing a module for online assessment of the dynamic stability called Dynamic Security Assessment (DSA) for the SASO project:

Dynamic Security Assessment (DSA) core functions

  • Online analysis of the load angle and voltage stability
  • Quick evaluation of the network state based on a hierarchical visualization concept
  • Improved situational awareness through early identification of potential problems
  • Higher system security and more efficient network capacity utilization
  • Decision support: Determination of optimal countermeasures for critical situa
  • Standardized interfaces (including CGMES)

The focus is on the analysis of the load angle stability and the voltage stability.

The load angle stability describes the ability of the synchronous machines in the network to remain stable and synchronous after a major fault such as a three-phase short circuit. The phenomena caused by sudden imbalance of rotational mass occur within only a few seconds. Therefore, this is also called transient stability.

Control plan for reactive power management. Screenshot: PSI
Control plan for reactive power management. Screenshot: PSI

Voltage instability occurs when the physical transmission capacity of the network is exceeded due to large power transports. Then the electric system is no longer able to keep the voltage within the permitted voltage bands. This may result in a total collapse of the network voltage (voltage collapse).

Changes in the generation structure lead to increased complexity

Power flows from producers to consumers. Screenshot: PSI
Power flows from producers to consumers. Screenshot: PSI

The liberalization of the European energy markets in combination with the extensive build-up of renewable energy generation (especially wind and solar energy) results in cross-border and cross-control area power transfers causing higher utilization of the existing networks and long-distance energy transfers. Therefore, the electric transmission networks are operating ever closer to their operational and stability limits. The increasing  number of inverter-based generating power  plants for renewable energy generation, which slowly replace the large power stations, also results in reduction of the rotating mass (instant reserve). Thus, the dynamic behavior of the electric energy system is radically changed.

The changes in the power generation structure, the changed dynamic behavior and the increasing volatility result in higher complexity and a greater variety of fast-changing operational situations for the control room staff.

As an online-assistance system, PSIsaso enables the network operator to identify critical situations as early as possible and to take appropriate countermeasures.

How does dynamic stability analysis work?

The setup of the data model for the dynamic stability analysis is based on "enhancing" the already existing data model for the steady state data analysis with additional dynamic data such as detailed models of the generators, HV DC transmission lines, and FACTS elements as well as of their controllers.

The transient stability behavior is primarily analyzed using the load angle trends and voltage angle trends over time based on various predefined fault scenarios. The change over time of these variables is calculated by step-by-step numeric integration of the differential equation system which describes the system behavior.

The critical fault clearance time is a key variable for quantifying the transient load angle stability. The fault clearance time defines the maximum permitted duration of a three-phase short circuit during which the system remains stable; it is calculated for every fault scenario.
The system state with regard to the transient stability is characterized by the Transient Stability Stress Factor (TSSF) which is calculated by the ratio of a specified minimum fault clearance time (for example 150 ms) and the calculated critical fault clearance time.

The most critical state, meaning the greatest TSSF of all relevant fault scenarios is used to assess the dynamic stability state. Then the assessment of the system state for operations management can be done by, e.g., assigning warning and alarm limits to this value.

Online dynamic stability assessment will become even more important in the future

Up to now, the dynamic network stability has been assessed by offline simulation calculations based on worst case scenarios, which are performed once or twice a year as part of planning studies. The stability limits including "safety margins" for the utilization in operations management are derived thereof.

With the worst-case consideration being too pessimistic for most network situations, the previously determined stability limits, however, can lead to inefficient network operations. The efforts to increase network capacities, for example by increasing the thermal current limit values by means of weather-dependent operation of overhead power lines (overhead power line monitoring) and by using high-temperature conductor cables may, in certain situations, mean lead to the fact that the transmission power is no longer limited by the thermal limits but by the stability limits.

In the medium term, steady state network security calculations based on load flow analysis are, under certain conditions, insufficient to provide reliable analysis of the system security in critical situations. Instead, new instruments, which increase the situational awareness and support the control center staff in determining and assessing appropriate countermeasures, are will be needed for assessing the dynamic network stability, starting with the day-ahead and intra-day planning and all the way up to online monitoring.

The visualization concept of SASO

The task of the visualization component is to enable a quick recognition and evaluation of the stability state and thus to create an awareness of the current situation in a targeted and clear manner.

Visualization of the network utilization by means of a heat map. Screenshot: PSI
Visualization of the network utilization by means of a heat map. Screenshot: PSI

In order to meet these requirements, a three-level hierarchical visualization concept has been realized:

  • On the highest visualization level, general stability states are displayed on a stability monitor. A general stability state aggregates the results across all fault scenarios for a time slot. If there are critical states, the network operator is to be provided with additional information.
  • On the middle visualization level, the stability states are displayed with all relevant information, relating to individual fault scenarios.
  • The lowest visualization level enables detailed analysis of the simulation results on the level of the relevant elements and their variables such as the change over time of the load angles or generator voltages in the form of curves or tables of the respective values.

Operational recommendations for critical network situations

When PSIsaso detects critical fault scenarios, the DSA module determines appropriate countermeasures to return the system to a stable state.

Initially, these may be network-related measures to shift the power-load angle curve, such as topological measures, changing tap positions of set transformers, or adaptation of the reactive power infeeds.

If these measures are not sufficient to restore stability in all relevant scenarios, market-related measures can also be considered. By iterative redistribution of the generator active powers, the critical fault clearance times are optimized in such a way that in the subsequent validation phase, no instabilities occur for any of the fault scenarios.

Looking ahead

By the year 2030, 65% of the electricity consumption in Germany shall be covered with re-newable energies. However, due to the asymmetric state of the network expansion and the build-up of renewable energies network operations are becoming more and more challenging. This means that ensuring network and system security is becoming even more complex and expensive.

With PSIsaso, PSI makes an active contribution to the energy transition. This will help to operate the increasingly complex networks securely and economically in the future.

PSI participates in public research projects which examine innovative approaches allowing for higher utilization of the existing network with at least the same system security.

For this purpose, the network-management autopilot is used in a field test run by TenneT to demonstrate the functionality and the effectiveness of the developed concepts in a control system environment working with real time data.

Dr. Michael Heine

The author has been involved in software development for control systems for over 20 years. Since 2010, he has been manager of the NA (Network Applications) department of PSI Software AG in Aschaffenburg.