Power system operators and network owners rely on power system modelling and simulation to maintain secure operation of power systems in real time. Conventional power system simulation models, typically referred to as root-mean square (RMS) models, have been used worldwide by all major network owners and system operators for predicting the response of power systems subjected to credible and non-credible contingency events. These types of models represent a trade-off between accuracy and simulation speed, and have proven to be acceptable for power systems having a large share of conventional synchronous generation and a limited penetration of non-synchronous energy sources.
Current RMS-type models lose accuracy as the ratio of synchronous to inverter connected generation declines. This primarily stems from the fast control systems used in inverter-based resources, the dynamics of which, being in the order of several kHz, cannot be adequately represented in RMS simulation tools.
Figure 1 compares the response of a large-scale power system with RMS and full electromagnetic transient (EMT) models. An important difference demonstrated in this figure is the presence of sustained post-fault voltage oscillations with a peak magnitude of approximately 3% and a frequency of around 8 Hz in the EMT simulation. These oscillations cannot be seen in the RMS simulation, because the dominant frequency of oscillations is far from the system fundamental frequency. The magnitude and frequency of oscillations obtained from the EMT-type simulation are aligned with those observed in a practical incident. These oscillations are unacceptable with respect to requirements for short-term flicker and adequate damping of oscillations. Figure 1 therefore highlights a deficiency of RMS-type models in assessing power system security and quality of supply.
Figure 1 Response of RMS- and EMT-type models to a fault for low system strength conditions
These oscillations can be experienced for moderate levels of instantaneous penetration of inverter-based resources, in the order of 20%. Thus it is clear that the need for large-scale power system modelling with EMT-type tools is not limited to systems with very high penetration ratios, for example in excess of 60%; the implications of inaccurate modelling simply become more pronounced under such high penetration scenarios.
To facilitate a sustainable transition to a power system with a significantly higher penetration of inverter-based resources, the importance of appropriate methodologies and analysis tools to identify any adverse responses before connection of these resources to the grid is becoming more apparent.
There are existing examples of inverter-based resources connected on the basis of traditional models and analysis, which were unable to identify issues that subsequently emerged during commissioning or commercial operation. This could have drastic consequences, potentially resulting in physical damage to assets due to adverse interactions, and compromising system security.
New approaches in EMT-type model development have allowed the use of exact control codes from the actual product in the corresponding simulation model. This provides a higher degree of confidence in the use of EMT-type models to identify and address performance issues before connection to the grid, in circumstances where the use of a corresponding RMS-type model may not reveal any stability or quality of supply concerns.
Another advantage of EMT-type models is that they can be used to predict the response of power systems to major disturbances, for example causation chains that may result in a major supply disruption, or during extreme operating conditions where correct understanding of the response of automatic load and generation shedding and special protection schemes is essential. These concerns are exacerbated as the ratio of synchronous to inverter-based generation declines.
EMT-type models and tools have been in use for several decades. Until fairly recently, their use has been primarily limited to specialized power system studies where the dominant frequency of interest is sufficiently far from the system fundamental frequency, or when control systems respond differently to balanced and unbalanced disturbances. Examples of such studies include switching and lightning analysis, sub-synchronous control and torsional interactions, and commutation failure of line-commutated HVDC links.
EMT-type modelling tools are in use by most major original equipment manufacturers (OEMs) for designing their equipment, and for designing the balance of plant components and tuning inverter control systems, in particular for low system strength conditions. These studies are often conducted with a single-machine infinite bus system. Reduced-order EMT-type models, comprising a limited number of nearby inverter-based resources, have also been used for low system strength conditions.
However, developing EMT-type models for large-scale power systems comprising several hundreds to thousands of busbars has not been widely attempted by system operators. This is partly due to the computational burden of running large numbers of EMT models in parallel, as well as to difficulties in sourcing suitable models from the OEMs.
To address the speed of simulation issues associated with EMT models, state-of-the-art solution techniques are being progressively developed by software and hardware developers, in collaboration with system operators and network owners. Concurrently, improvements are being applied to the speed and robustness of the simulation models developed by OEMs. The ability to conduct EMT-type simulation studies for large-scale power systems could become a necessity in the next few years as several jurisdictions implement ambitious renewable energy targets.
The limitations of RMS-type models mentioned in this blog have already been observed in some systems with a very high penetration of inverter-based resources – including Australia, Texas and the UK – where system operators have developed large-scale EMT-type models for large parts of their networks that are used extensively for making operational decisions. In addition to their application in specialized power system studies discussed above, these models are extensively used for power system stability analysis where the dominant frequency of interest could be close to the system fundamental frequency.
To ensure that EMT models provided by plant owners and manufacturers present the required level of accuracy, AEMO has worked closely with plant manufacturers to progressively develop model requirement standards and has established model development methodologies in Power System Model Guidelines. Furthermore, comparison of measured and simulated responses of the overall power system and its constituent parts against actual system disturbances has proven critical in identifying modelling errors and inconsistencies. An issue sometimes identified is the need for control system tuning, in either the EMT-type model or the actual plant. This usually arises where the available system strength in the network is lower than that for which the standard model or product is designed.
A new CIGRE WG C4.56 (entitled electromagnetic transient simulation model for large-scale system impact studies in power systems having high penetration of inverter-connected generation) has been recently formed, which will look in more detail at the aspects discussed above.
Manager Operational Analysis and Engineering
Australian Energy Market Operator