INPOWEL: EMT models for grid forming assets

Context

Grid stability challenges in the transition to 100% renewables

Towards carbon neutrality by 2050 with 100% renewables, not only do we need to expand on renewables, but we also need to maintain the stability of the grid.  It is now well understood that integration of renewables reduces the systems inertia (MW.s) for frequency contingencies by limiting instantaneous MW response to counteract frequency/RoCoF events, lowers the short circuit level (MVA) for voltage contingencies by limiting the fast fault current injection during system fault, and also leads to more nervous response of renewables due to the higher sensitivity of their terminal voltage to its current injection. Various real-world examples of unit disconnections and oscillatory behavior and interactions are also reported. Different TSOs recognize these new stability challenges and have taken different initiatives to tackle them. Examples are German System Stability Roadmap proposed by the Federal Ministry for Economic Affairs and Climate Protection (BMWK), and also the National Grid Stability Pathfinder program of the UK.
In terms of inertia, in Europe, power system inertia reduced by 20% on average between 1996 and 2016. By 2030, it is projected that system inertia could decrease by approximately 20-30% on average. In terms of short circuit level (SCL), synchronous generators (SGs) provide the highest fault current contribution which can 5-7 times their rated armature current, while inverter-based resources (IBRs) have the much lower fault current contribution due to their limited overload handling capability which is about 1.2-1.5 times the inverter steady-state current rating. 
Currently, synchronous condensers (SCs), often combined with flywheels, are the go-to solution used by various TSOs to provide inertia as well as SCL. However, they are expensive, have high operational costs, and require substantial physical space, making them infeasible for coping with high penetration of renewables. There is a clear need for a more efficient technological solution. 

Grid-Forming Converters: A solution for future grids

According to various projections [IEA, IRENA, BloombergNEF], global installed renewable energy capacity is expected to triple (3x) between now and 2030, and then triple (3x) again from 2030 to 2050 – resulting in a nearly tenfold increase from current levels. Solar PV and wind together are projected to account for 70% - 90% of total renewable capacity additions. All these renewables are integrated into the grid through power electronic-based converters. If these converters can be controlled to mimic the stabilizing properties of the conventional SGs, it would provide significant benefits. This is precisely what grid forming converters (GFM) are designed to achieve. Unlike legacy grid following converters (GFL), which require a strong grid for synchronization via a phase-locked loop (PLL), grid forming converters can independently “form” the grid and thus operate robustly in weak grid conditions. Grid following converters work well when GFL IBRs are relatively scarce. However, as IBR levels exceed 60% - 70% of instantaneous generation, maintaining stability becomes increasingly challenging. Furthermore, the growing share of IBRs can cause oscillations and interactions within the network. 

Advanced simulations and model validation for stability studies

Conventional Root Mean Square (RMS) simulations, with a timestep of 10 ms, can effectively capture frequencies up to around 50 Hz. However, they are unable to capture transients and dynamics faster than this. In contrast, ElectroMagnetic Transient (EMT) simulations can capture high-frequency phenomena and fast transients of power electronic-based IBRs due to their much shorter timesteps, typically in the range of 1–50 μs. For example, transient switching frequencies of 2–3 kHz can be accurately modeled with timesteps of 30–50 μs. For EMT simulations involving frequency content up to 10 kHz, smaller timesteps of approximately 10 μs are typically required.
Therefore, to study IBR interactions and to develop suitable performance requirements for the future grid, generic EMT models for technology such as grid forming inverters are important. However, validation of performance of generic models are critical to establish their suitability and sufficiency for the purpose of the studies. Equally important is to have visibility and insight regarding the limitations of the generic models and their applicability.
To this end, in the framework of INPOWEL project, generic EMT and Real Time Digital Simulator (RTDS) models for Battery Energy Storage Systems (BESS) and Static Synchronous Compensators with Energy Storage (E-STATCOM) have been developed that to represent their behavior in both grid following and grid forming modes. The models are developed in a modular fashion allowing for the same model to be used to represent either technology, depending on the user input parameters. Further, extensive validation of the model's performance is carried out by comparing its response against the performance obtained from blackbox models from original equipment manufacturers (OEMs). The validation exercise has been carried out at a variety of system strength scenarios, voltage, frequency, and phase angle changes and balanced/unbalanced faults. These responses also highlight potential limitations of use of the generic models. The work carried out also looks into an application study in a larger network with grid following converters exhibiting oscillations and grid forming technology helping with the oscillation. Static evaluation metrics have been used to identify locations and ratings of the grid forming devices that can be of assistance. The purpose of simulation studies is to identify any interactions between various technologies and the grid as well as identify any limitations of EMT models and propose good practices for use of EMT models.

EMT modeling details

A unified EMT model for both BESS and E‐STATCOM units within PSCAD is developed which supports both GFM and GFL modes of operation through a user‐defined control setting. Figure 1 illustrates a simplified schematic diagram of the model which uses an aggregated plant model with a two‐stage topology. 

Figure 1. Simplified schematic diagram of the BESS/E-STATCOM plant model

Comparison of generic EMT model performance against blackbox model

Using a single source infinite bus test system, as shown in Fig. 2, the performance of the generic GFM and GFL assets across different system strength and operation points, are compared against blackbox OEM models.
The test sets provide visibility into the following operation scenarios:
o Varying SCR levels and X/R ratios
o Frequency, voltage, and phase angle changes
o Balanced and unbalanced faults

The list of tests are conducted when the IBR device is operating away from current limits and operating close to its current limits. It is expected that current limits may be hit due to the occurrence of a disturbance, or the device may assume to operate at current limits before an event occurs.

Figure 2. Single source infinite bus (SSIB) setup for a collection of identification tests

A total of 110+ tests have been carried out in order to demonstrate the performance and limitation and sensitivity of the assets under different operating conditions and against different faults. The following extensive comparisons were made: (i) generic PSCAD model vs vendor blackbox PSCAD model (ii) generic PSCAD model vs generic real time simulator model.

As an example, using the test system shown in Fig 2, when operating in GFM mode, at close to maximum power output (0.9pu), and at a high SCR value of 10.0 and X/R ratio of 10.0, the comparison of performance for step changes in infinite bus voltage magnitude are shown in Fig 3. The response is observed at the point of interconnection (high voltage side of the interconnection transformer).

The reactive power response comparison between the generic model and the OEM model provides a suitable match. However, it can be seen that there is a difference in the steady-state value of active power response and the magnitude of the current response. A reason for this difference can be the treatment of power limits within the OEM’s model.
At this operating point, the models are operating close to both their power and current limits. When the voltage drops, the reactive power increases to support the voltage. However, at the pre-disturbance value of active power, this increase in reactive power can result in a violation of the MVA limits of the inverter. As a result, the active power output from the OEM’s model reduces.
The generic model also reduces its active power output during this scenario, but not to the same extent as the OEM model. The resultant difference in active power also manifests as the difference in current magnitude. This difference in behavior near the limits of operation of these devices is expected when comparing the performance of a generic model with an OEM model. The methods used to handle limits and ensure stable and smooth recovery from limiting conditions are often considered core intellectual property (IP) by OEMs. Hence, there are few aspects regarding operation at limits that may not be captured by generic models.
However, overall, the comparison of the generic model’s performance is deemed to be suitable and acceptable for the purpose of its use.

 
Figure 3. Comparison of performance of BESS in GFM mode for step change in
voltage at an SCR of 10 and at active power of 0.9pu


As the famous quote by George Box (1976) goes, “All models are wrong, some are useful.”, it is important to note that generic models will always have room for improvement, as they are not designed to exactly match the characteristics of an OEM model. Instead, generic models are only meant to provide the general trend in a technology’s response. Therefore, further validation and testing are (often) necessary to ensure their accuracy and reliability.

 

Conclusions

• Generic models may have challenges in their operation near limits, as these are situations where original equipment manufacturers may have their core intellectual property.
• Static evaluation metrics can be used for screening studies to obtain an insight into where potential instabilities may occur. These metrics also help identify locations and rating of grid forming devices that could help the network.
• Real time simulation allows live interaction with a simulation model and allowing for the development of a digital twin which provides a live digital replica of the power system which is continuously updated from real‐time measurements.

 

Next steps

Future work includes (i) extension of the models to include additional sources of energy such as wind, PV, and/or HVDC, (ii) improvement in current limiting logic and coming out of current limits after an event, (iii) improvement in ride through behavior during an unbalanced fault.

Partners