In Denmark, the green transition follows up with reconstruction of the transmission grid meaning here complete replacement of the existing 150 kV overhead lines (OHL) with underground cables. At present, the 150 kV transmission grid reconstruction is ongoing in South-Western Jutland, in Western Denmark belonging to the Continental European synchronous area of ENTSO-E. Besides complete replacement of the OHL with cables, the presented grid reconstruction includes a new 150 kV substation, grid-connection of converter-interfaced units, new connection routes between the substations, and more operation regimes. The presented project establishes 200 km 150 kV cables replacing the OHL and shall be completed due 2023.
The harmonic assessment by simulations shall identify excessive harmonic distortion in the 150 kV reconstructed grid and propose mitigation. The challenges are that the simulations are conducted for the grid, which is not yet established and differs completely from the present grid, the distributed harmonic emission sources are not explicitly described for direct inclusion in the simulation model, and the outcome shall support investment decisions of specific mitigation measures.
The harmonic assessment is conducted in steps. The initial step is analysis of the power quality measurements in the present transmission grid with (n-0) and different (n-1) conditions. Combined with the present-stage grid model, this step results in construction of the distributed harmonic source models and secures validated starting-point for the projected grid reconstruction.
The harmonic distortion depends strongly on the harmonic emission sources. Inaccurate modelling of such sources would produce misleading results and mitigation decisions. Therefore, the next step is the evaluation methods for reducing modelling uncertainty. The three evaluation methods are developed. The two methods work with a normal case of the harmonic emission sources. The first, normal-case, method applies arithmetical margins between the measured distortion in the present grid and the simulated distortion in the reconstructed grid. The second, normal-case, method applies gain factors with the present grid as reference of the reconstructed grid. The third method uses a simultaneous worst case of the harmonic emission sources, so that simulating highest realistic values of the harmonic distortion. When both normal-case and worst-case methods show excessive harmonic distortion in simulations, there will be a risk of such in the physical reconstructed grid.
The assessment has shown excessive harmonic amplification in the 150 kV reconstructed grid. The final step is mitigation, which for the presented project is a new passive harmonic filter. This is new knowledge since previous replacement of the 150 kV OHL with underground cables did not show excessive harmonic amplification. However, a single harmonic filter is a small effort in comparison to the volume of the reconstructed grid.
The system separation is a critical event caused by major incidents, often leading to severe frequency transients and actual risk of system collapse. The expected increase of transmission capacities and renewables generation share can potentially lead to a worsening of the situation: the power flows between the different parts of the system can increase considerably, determining high power imbalances in case of system split; renewable energy sources as solar and wind are typically interfaced to the grid through power electronics, so they do not provide an inherent contribution to the frequency dynamics of the system and they can be more likely subjected to disconnections. The two aspects of increasing power exchanges and increasing non-synchronous generation can have therefore a critical impact on the frequency dynamics of the system, especially in case of system split scenarios. The work takes into examination the Continental Europe (CE) power system for an actual split event occurred in 2021. First, the frequency transient registered after the system split is reproduced for both parts of the CE area, with specific dynamic models and simulations of the system, based on fundamental aspects of the frequency containment reserve (FCR) process. The validated simulation models are then used to investigate the frequency dynamics of the CE system under operating conditions characterized by high shares of renewable energy sources. The dynamics of the system is first analyzed assuming no contribution to the FCR process from the non-synchronous generation sources. In this case, the dynamic models used to represent the non-synchronous generation are controlled current sources with typical grid-following converter characteristics, without any specific control extension. This scenario is compared with the case of inverter-based generation participating in the FCR of the system. In this case, specific controls for the converters interfacing the generation sources to the system are implemented. The grid-forming control based on the virtual synchronous machine concept is considered for investigations. For sake of comparison, the grid-following control extended with virtual inertia and fast frequency response is also simulated. Different values and combinations of the relevant control parameters are studied, to get insights about possible sensitivities and realize more comprehensive examinations. Analysis and results suggest that a high share of renewable energy sources might not necessarily imply only issues and challenges for the frequency stability of the system. Due to specific controls, the inverter-based generation can in fact participate in the FCR process with an essential contribution, significantly supporting the frequency dynamics of the system.
Key words: phase shifting transformer (PST), offshore wind farm, diode-rectifier unit, HVDC, blackstart, dynamic behavior.
Abstract
Since the wind energy is renewable and environmental natural resource, the utilization of wind power plant increased quickly. In the future, the development of wind power utilization will focus on large offshore wind farms (OWFs) [1].
Due to the distance between OWF and the onshore point of common coupling (PCC), voltage source converter (VSC) HVDC technology is widely applied for large OWF grid connection. Also, in order to reduce the investment of offshore HVDC converter platform, diode-rectifier unit (DRU) is proposed [2]. Since the DRU is a passive rectifier without control possibility, in order to ensure the stable operation of OWF, the wind turbine controllers should modified (e.g. grid forming control). Moreover, during the start of OWF, there could be a period where the DRU HVDC and MVAC cable system (e.g. 33kV or 66kV) in parallel operation. Large inrush current could occur during the parallel operation.
For the stable operation and improvement of the dynamic behavior of OWF with diode DRU, a new method is proposed in this paper: application of a phase shifting transformer (as shown in Fig. 1.) on the MVAC cable for stable operation of the OWF with DRU [3].
Fig. 1. Application of PST for OWF with DRU
In this paper, dynamic behavior of the OWF and DRU with the PST are analyzed in detail.
References
[1] U. Karaagac, J. Mahseredjian, H. Saad, S. Jensen and L.J. Cai, "Examination of Fault Ride-Through Methods for Off-Shore Wind Farms Connected to the Grid Through VSC-Based HVDC Transmission", Paper published in 11th International Workshop on Large-Scale Integration of Wind Power into Power Systems, November 13-15, 2012 in Lisbon, Portugal.
[2] T. Hammer, S. Seman, P. Menke, F. Hacker, B. Szangolies, J. Meth, J. Dorn, K. Loppach, R. Zurowski, "Diode-Rectifier HVDC link to onshore power systems: Dynamic performance of wind turbine generators and Reliability of liquid immersed HVDC Diode Rectifier Units", Paper published in CIGRE 2016.
[3] Patent P26336DE: Verbindung zwischen Offshore-Energiesystemen und Landstromnetzen sowie Schwarzstartverfahren für Offshore-Energiesysteme.
This paper synthesizes a number of US studies that investigate 100% clean electricity (or similar) goals to determine the implications for transmission needs and approaches. In particular, it examines the role of a macro grid, or large-scale inter-regional transmission network, in enabling decarbonization at least cost, and the role of transmission infrastructure in maintaining a secure and reliable system over all timescales.
The National Renewable Energy Laboratory’s Interconnection Seams Study, Brown and Botterud (MIT study), Vibrant Clean Energy’s ZeroByFifty Study, and the Midcontinent Independent System Operator’s Renewable Integration Impact Assessment Study, among others, use different assumptions and datasets, employ different types of models, and have diverse goals, but all share some common findings regarding the need for significant transmission expansion.
We also examine the competition and complementarity between transmission and other resources such as battery storage, long-duration storage (hydrogen) and distributed energy resources.
Finally, we discuss recommendations for national-scale transmission planning, based on lessons learned from these studies and previous efforts.
The multi-terminal HVDC (MTDC) system based on modular multilevel converters (MMCs) is increasingly developed in the high-voltage transmission grid. The MTDC could introduce various benefits to the power system by bringing in more operational flexibility. Yet, the dynamic interaction between MMCs and the power grid may lead to system oscillations in a wide frequency range, which greatly threaten the stable and reliable operation of the power system. Therefore, it becomes relevant for transmission system operators (TSOs) to perform stability studies before commissioning MTDC in practice. Among other stability assessment methodologies, impedance-based stability analysis is commonly adopted by TSOs dealing with black-box models.
The accurate impedance (matrix) measurement of MMCs is the precondition for implementing impedance-based stability analysis, which, however, is non-trivial due to the frequency coupling nature of the MMC introduced by its internal dynamics. To tackle this challenge, a PSCAD-compatible software toolbox is developed that enables the accurate AC/DC impedance (matrix) measurement of the MMC by capturing its frequency-coupling nature. The toolbox is capable of automating the impedance (matrix) measurement process, which makes it friendly to users. The impedance (matrix) measurement accuracy is also demonstrated by comparing the measurement data with analytical results.
Finally, case studies are carried out on a generic MTDC system by using the impedance (matrix) measurement data, and the stability prediction based on impedance-based method is verified by nonlinear time-domain simulation.
Some grid codes (e.g. Sweden, Finland) now require wind power plants (WPP) to provide damping of electromechanical oscillations, caused by rotor speed oscillations among synchronous generators in the power system. In general, those requirements require WPPs to offer power oscillation damping (POD) performance with a degree of similarity to the traditional power system stabilizer (PSS) function implemented in conventional power plants. The objective of POD function is to contribute on damping externally caused power oscillations by modulating the produced active or reactive power from the WPP in response to certain monitored oscillatory input signals from the grid. Important aspects to consider are the fundamental technical characteristics of WPPs, mainly related to wind speed conditions, plant component ratings and wind turbines (WTG) mechanical eigenmodes, which impose operational boundaries in WPPs. Proper selection of input/output signals to the POD function and their coordination in terms of gain and phase are essential for providing satisfactory damping performance, while considering the operational boundaries of WPPs. Moreover, the coordination of POD control with existing voltage and frequency controllers is crucial to maintain grid code compliance. This paper provides some proposals to operators and other stakeholders of the wind industry on how to measure and verify the performance of POD offered by WPPs. Some relevant theory on active and reactive power modulation by utilizing various input signals is provided. The results of extensive simulations studies on a 12-bus power system suggest the possibility for robust control tuning for a wide range of grid conditions. The challenge of simultaneously operating with POD and voltage slope control is discussed and a possible realization suggested.
To mitigate the resonance risk of power electronics-dominated power systems with multi-vendor converters, it is essential to obtain deep insight into the source of the resonances rather than just the system stability conditions. This paper proposed a dynamic power flow-based resonance source location method, which is able to identify the resonance source preventively using the black-box impedance/admittance models. The definition of the resonance source is firstly presented based on the concept of the dynamic power flow. Then the analytical calculation method of dynamic power flows for critical resonance modes is derived based on the impedance/admittance models of the system. As the required impedance models can be measured by frequency scanning, the proposed resonance source location method can be readily applied to the large-scale power electronics dominated power system, which provides a visualization tool to identify the root causes of various resonances. Case studies from an 800MW offshore wind energy system with a real-life complexity confirm the effectiveness of the proposed source location method.
In the light of the EU energy policy objectives aiming at reaching climate-neutrality in the European Union until 2050, the European and in particular the German power generation portfolio are subject to a significant transformation from synchronously connected power generation modules towards decentralised inverter-based renewable energy sources. This transformation leads to new challenges for European transmission system operators with respect to system stability and the manageability of disturbed system states. System splits like in November 2006 or recently in January 2021 must remain controllable also under the increasing penetration of power electronic interfaced power sources. To maintain transmission system security a new type of inverter control with grid forming capabilities is needed, instead of the currently dominating grid following control strategies.
A number of renowned studies have already dealt with the technical definition of future grid forming inverters like the MIGRATE project (Massive Integration of Power Electronic Devices), the German 4-TSO project SUE (Systemeigenschaften umrichterbasierter Erzeugung) and recently the HPoPEIS-report (High Penetration of Power Electronic Interfaced Power Sources) of ENTSO-E. The basic technical requirements of grid forming power generating modules have been extensively researched in these studies.
The proposed paper follows a different approach by introducing the potential demand and highlighting the urgency of implementation of grid forming inverters by the example of the changing German generation portfolio. Due to the almost completed German nuclear exit, the already initiated withdrawal of coal-fired generation and an increase in inverter-based renewable energy sources Germany will experience the challenging impacts of the loss of synchronous generators on transmission grid stability as one of the first countries in Continental Europe.
Different scenarios of the need of grid forming inverters based upon figures of the German national grid development plan have been investigated. The results of these studies including the potential timelines for implementation of this equipment will be presented. The paper analyses by when a certain percentage level of grid forming inverters can be met in Germany in the presence of grid following inverter-based generation. It will demonstrate that each delay in the deployment of grid forming installations might have a negative impact on the stable, reliable and safe operation of the German transmission grid despite the support by remaining synchronous power generation modules like gas-fired or transitionally still operating coal-fired power plants.
This contribution presents a selection of results from the project NetzWind, funded by the DBU (Deutsche Bundesstiftung Umwelt). The basis of the investigations is provided by a grid-voltage-forming control method for grid-connected converter system, which is implemented on the grid-side converter. Without modifying the design of the wind turbine, the aim is to demonstrate the possibilities of providing uncontrolled instantaneous active power by means of simulation. Two independent models are used for this purpose, one for the line-side converter system and one for the wind turbine including the converter unit. Besides the consideration of the inherent power supply, the focus is especially on the lifetime of the components.
The progressive integration of generation plants based on renewable energy sources (RES) as wind and solar is pushing system operators towards the definition of new services within the grid codes. These grid services range in a wide area of applications, as for instance black-start capability, synthetic inertia and fast frequency response provision, reactive power capabilities for voltage regulation. The units capable of providing a specific service could access the remuneration scheme designated for that service. For a RES-based generation unit, it is then fundamental to assess and possibly achieve the ability of meeting the requirements indicated by the system operators for a specific service object of remuneration. The paper illustrates the results of a study concerning the participation of an existing wind power plant in a given voltage regulation service. The plant is located in Sicily, South Italy, in the facility of Lago Arancio. The wind park is composed by 22 wind turbines of 2 MW rated power each, distributed along four main cable feeders at 30 kV. The Italian grid operator Terna is currently examining the possibility of introducing a new service for voltage regulation, with the provision of a given amount of reactive power within a determined time window. For the facility of Lago Arancio to be compliant to the requirements specified by Terna and participate in the remuneration of that service, the transient capabilities of reactive power control need to be assessed. Experimental tests have been conducted for this purpose, as first proof of the transient response of the wind plant in its current conditions. The wind park of Lago Arancio is modeled for dynamic simulations according to the IEC 61400-27-1 Edition 2 Standard. The wind turbines are simulated in an aggregated model according to the Type 3 defined by the IEC Standard. All the controllers are implemented in the simulation model, including WP plant controllers, communication modules and single WT controls. The model is validated against field measurements obtained by the SCADA system during the experimental tests of the wind plant. As some parameters of the control system were unknown, the identification of the most relevant parameters governing the dynamics of the reactive power and voltage control has been first performed. These parameters of the wind plant controllers have been then set to proper values to match the experimental data and tune the model. The dynamic model of the wind park so developed and validated is then used to investigate and provide a set of control system parameters, which allow the wind plant to match the requirements of the voltage regulation service.
Previous works have shown the potential benefits of offshore energy hubs in the North Sea . The topic is moving fast from academic works and reports to reality, with Denmark going ahead with two multi-GW offshore hubs. Electrification and sector coupling increase electricity consumption and flexibility in the energy system; both aspects are expected to push variable renewable energy (VRE) installations towards 2050. The increased need for electricity generation is expected to increase offshore wind installations, especially if onshore wind expansion is restricted, e.g., due to social acceptance challenges. However, there are also aspects which may hinder the development of offshore energy hubs; this paper analyses the impact of wake losses on large-scale offshore wind generation at the hubs. Although shown to have potentially dramatic impacts on offshore wind capacity factors (CFs) when going to tens of GWs of installations in a limited geographical area, large-scale wake losses are often not considered in large-scale energy system optimisation models.
A combination of the Correlations in Renewable Energy Sources (CorRES) and PyWake tools is used to analyse the impact of large-scale wakes on offshore wind generation time series. Hub sizes from 2 GW to 24 GW are analysed, with the impacts on the hub’s generation variability also studied.
The results show that large-scale wake losses can be significant, and they get larger as hub size increases. Hub size impacts also the variability of the hub’s generation. It is argued that these impacts are significant and should be considered in energy system optimisation. The high capacity factors achieved in far-offshore energy hubs still make them attractive in highly sector coupled future scenarios. However, when the dependency between hub size and wake losses is considered, a more distributed placement of hubs in the North Sea is found optimal. The best wind resource hub locations far in the North Sea remain attractive for very large hubs even when the increased losses are considered. Although wake losses can be mitigated using lower installation densities, it will lead to larger sea area use, which can be significant when going towards the hundreds of GW of offshore wind installations needed to reach Europe’s decarbonization targets.
When integrating grid-forming converters (GFC) into the power grid, it is not just a matter of replacing the converter's current controller with a voltage controller; all aspects of the power system dynamics must be taken into account in order to design a suitable behavior of GFC. This is necessary in order to fully replace the grid-stabilizing properties of synchronous generators (SG) with converters.
Within the joint project VerbundnetzStabil the partners KACO new energy, TransnetBW, University of Stuttgart and Fraunhofer ISE investigated the stability of interconnected grids with a high penetration of converters over the past four years.
Whereas the dynamic behavior of a SG is defined by its physical properties, the dynamic behavior of a GFC only depends on its control concept, opening a wide range of possible implementations. Although these concepts are mainly based on the same fundamental principle of power synchronization as SG, there are many differences in their dynamic behavior and the way how synchronization with the grid is achieved. Some of the GFC control concepts discussed for the use in interconnected power systems originally have been developed for microgrid applications. Furthermore, existing developments have been modified and adapted to address the needs of interconnected power systems, ultimately resulting in a wide range of concepts that are currently under discussion. In fact, some of these are partially or even largely equivalent.
Several existing publications aim at comparing different GFC control concepts. However, these publications are often limited to a small number of control concepts or lack a direct simulation-based comparison of their dynamic system behavior. Thus, the goal of this paper is to compare a large number of today's relevant control concepts by 1) a detailed system-theoretical analysis of their design and by 2) a comparison of their dynamical behavior in simulation-based studies. The control concepts are compared with respect to their intrinsic and their quasi-stationary behavior during phase jumps, frequency changes and active power imbalances. Insight gained from these specific comparisons can be helpful for the development, improvement, and functional specification of GFC.
For the purpose of comparison, the simulation models of the various control concepts are configured such that they provide comparable changes in active power for a given phase jump and for a given RoCoF, respectively. Thus, differences in their dynamic behavior become apparent as well as whether an active power component is provided that is directly proportional to the frequency deviation.
With the rapid advancement in power electronics, the shipping industry has dramatically moved towards low-carbon emission-free technology. Moreover, a practical and cost-effective solution is required from an engineering perspective to evaluate the system performance as global trade is increasing exponentially. In contrast, various challenges being faced such as higher fuel prices, more stringent regulations for the environment, and safety concerns. To mitigate these issues, an onboard low voltage dc microgrid was proposed which provides a more efficient and state-of-the-art solution by reducing energy consumption, energy-related costs, and prolonged maintenance intervals. In this paper, a detailed simulation for a low voltage dc system was performed because of various potential advantages of dc over ac system. One of the key benefits is the neutralization of the skin effect in dc system, which is quite common in the power transmission of the AC system. Whereas, grid synchronization with renewable energy generators is not required which ultimately curtails operational complications. Finally, in case of power disruptions or outages from the onshore ac grid, the dc grid indulge reliable, and controllable solution with enhanced power quality. Moreover, system architecture and control structure for the designed system shows the feasibility of overall configuration. To evaluate system performance, renewable generators (e.g., PV generators with a fully interleaved boost converter, Battery Energy Storage System with bidirectional converter electronics, and Wind Turbines) were interfaced to a common dc-link to support propeller load profile. Two diesel Generators with constant speed profiles were providing enough initial torque to run six-phase permanent magnet synchronous generators associated with a six-phase rectifier, providing power to common dc-link. Design constraints parameters for common dc-link were chosen 1000V, which accelerates power from dc-link to six-phase inverter connected with six-phase permanent magnet synchronous machine to run the propeller load. Sizing criteria of converter ratings were performed based on mathematical modeling and load requirement. The control interface for each section was illustrated comprehensively and an on-shore grid was connected as the vessel approaches to berth. The system was developed in MATLAB/Simulink Environment which verifies the proposed network effectiveness.
In this paper, a converter test bench is presented with which grid-forming converter control methods can be easily tested by using real-time PC-based control. In combination with a previously presented synchronous generator-based variable-frequency island network, the behaviour can also be investigated during frequency fluctuations.
Since a converter of a type 4 wind power plant cannot be tested in the available laboratory in a frequency-variable island grid, the nominal power of the converter created was designed for a nominal power of 60 kVA. The converter consists of a back-to-back arrangement of two three-phase bridge circuits. It was important to use IGBT power semiconductor modules that are also commonly used for wind power plants. Due to the overdimensioning of the power semiconductors, a loss-optimised design is not possible, but fault cases can be tested without the risk of damage to the power semiconductors. Automatically configurable LCL or LC filters are available as grid filters. The mains connection is made via transformers with star-delta connection. Circuit breakers provide hardware-based protection against overcurrent.
The control is carried out via a real-time PC-based platform, which is also used in wind power plants. Usually, however, the inverter control is not implemented on such a central platform. The control can be developed using Matlab/Simulink and can be executed on the real-time platform with a cycle time of 50 µs. Via field bus, voltage reference values are transferred to an FPGA-based platform, which performs the modulation and control of the power semiconductors. In addition, this platform records the voltages and currents, which can be transmitted diritally filtered via fieldbus to the real-time PC-based platform. In order to be flexible with regard to the use of different control methods, voltages and currents are recorded in three phases at several points on the AC side. The test bed will be presented in detail in the final paper.
When integrating grid-forming converters (GFC) into the power grid, it is not just a matter of replacing the converter's current controller with a voltage controller; all aspects of the power system dynamics must be taken into account in order to design a suitable behavior of GFC. This is necessary in order to fully replace the grid-stabilizing properties of synchronous generators with converters.
Within the joint project VerbundnetzStabil the partners KACO new energy, TransnetBW, University of Stuttgart and Fraunhofer ISE investigated the stability of interconnected grids with a high penetration of converters over the past four years. Among others the following research questions were addressed.
- What are the principle requirements for GFC from an interconnected grid perspective?
- Which kind of GFC control fulfills these requirements and how can it be implemented on a converter platform?
- How can generally applicable test procedures verify the right properties of GFCs?
- What kind of aggregation methods for distribution grids with GFCs are appropriate to get reduces simulation models for grid stability studies?
- Which unique properties of GFC need to be considered for simulative transmission system studies and how do realistic EMT models of GFCs look like?
- How can grid events such as system splits or voltage/frequency fault scenarios be scaled down and tested on a megawatt-scale laboratory platform?
This paper provides answers to the aforementioned questions by summarizing the final results of the VerbundnetzStabil research project.
This research presents a supervisory optimal control framework called Oxtimal for the efficient control of Type 4 wind turbines with power quality and power system balancing metrics. This framework consists of a reduced order model (ROM) of wind turbines connected to an electrical microgrid, a discretization of the resulting constitutive equations using an orthogonal spline collocation method (OSCM), and an optimization engine to solve the resulting formulation. Using this framework, the approach is validated using wind profiles that result from high fidelity wind simulations.
Design and Operation of Energy Systems with Large Amounts of Variable Generation
IEA TCP WIND Task 25 ”Design and Operation of Energy Systems with Large Amounts of Variable Generation” has compiled a Summary report on the main issues regarding wind and solar impacts on power systems. The session will drill into the different challenges of power systems with high shares of wind and solar: Planning, Operations, Stability and Markets.
Each presentation highlights the issues and examples of mitigation, from experience and system study results:
A joint paper will be published in the proceedings, which will be a summary of the report, currently being finalized to go to IEA Wind ExCo review during summer.
Long ago, when inverter-based resource (IBR) technology did not represent a significant share of the total mix of power generation, its impact on the grid was neglectable. Therefore, transmission system operators (TSO) around the globe did not pay much attention to the impact studies coming from this minority of power plant generation. Therefore, the requirements regarding the accuracy and performance of electrical simulation model for wind turbines were not strict and, in some cases, not existing. The focus was to have a simplified generic standard library model, to model the performance of WTGs, irrespective of the control functionalities provided by different original equipment manufacturers (OEMs).
For many years, grid interconnection studies for impact assessment of wind power plants (WPP) were conducted using generic models developed by the two main groups, IEC 61400-27 and the WECC Renewable Energy Modelling Task Force. These models used to adapt quite well to the most generic and standard features of Double Feed Induction Generators (DFIG), Full-Scale Converter System (FSCS) machines, and Power Plant Controllers (PPC).
Nowadays, the situation has changed significantly. IBR represents a significant share of the online power generation in several countries and several system events have demonstrated that the IBR did not perform as predicted by the models. Therefore, TSOs have introduced new grid code requirements in terms of performance and accuracy of grid interconnection studies to demonstrate grid code compliance for new WPP. Consequently, there is a demand for dynamic models with high accuracy and which are validated against site tests to resemble the features and performance of a real WPP.
Due to the introduction of more advanced features, the available generic DFIG (known as Type 3), FSCS (known as Type 4), and PPC models are not always capable of reproducing the performance of many (OEMs) products. Due to the long lead time of standard models, it is not possible to tweak the available controllers of the generic models to achieve the desired performance. Therefore, generic models are not suitable to be used for grid interconnection studies for grid code compliance.
This paper highlights the different technical limitations of IEC and WECC generic models when compared to Vestas turbines and PPC. The paper describes the core limitations of several control block functionalities that impedes the user to parametrize models to capture specific performance in several cases such as; voltage and temperature dependencies of dynamic power curve capabilities of wind turbine generators (WTGs), lack of accuracy to represent Fault Ride Through (FRT) response of WTGs, missing active power production controller in WTG, dynamic PI controllers of PPC as several others.
In a wind turbine generator (WTG), the behaviour of the Phase-Locked Loop (PLL) under weak grid conditions becomes complex and requires nonlinear analysis to assess the WTG control performance and system stability. This paper introduces a quasi-static reduced-order VSC-grid model of a WTG and evaluates its performance for large-signal disturbances such as changes in current set points and grid short circuit ratio. The time-domain expression for a synchronous reference frame PLL is considered with its nonlinearity intact. Based on the analytical model, a V-δ mechanism is established to identify the PLL synchronisation instability events relevant to a WTG operation. Further, for PLL stable operation, a globally valid stability boundary is discussed. The complex nonlinear PLL dynamics are addressed using phase portraits, where it is shown that the results have direct implications on robust PLL gain tuning guidelines.
Keywords—Fourier transform, long short-term memory, wind power, forecasting, false data
Abstract - Forecasting of wind power is necessary to remove the system operational uncertainties via ensuring more reliable inputs for power system scheduling and control. In order to achieve accurate forecast of wind power for up to 50 seconds (very short-term forecasting), this paper proposes the method Fourier Denoising combined with Long Short-Term Memory (FD-LSTM). The FD-LSTM cascades the output of the Fast Fourier Transform (FFT)-based denoising algorithm to the input of the Long Short-Term Memory (LSTM) forecaster. In this method, first, the inclusion of the FFT-based denoising algorithm ensures better and balanced performance for different forecasting horizons by removing the frequencies. Next, short-term forecasting by employing LSTM reduces the uncertainty and improves both the quality of the operation and the planning. In this paper, we first evaluate the performance of the proposed FD-LSTM method on increased forecasting horizons based on Mean Square Error (MSE), Mean Absolute Error (MAPE), and R-Squared, and compare the results against linear regression. Afterwards, the method is tested with data sets where false data is present. The results show that FD-LSTM outperforms the other forecasting methods under the presence of false data.
Historically, root-mean-square (RMS) dynamic simulation tools have been the preferred option for grid interconnection stability studies. In many cases, the employed RMS models are based on generic models developed by IEC and WECC working groups and sometimes user-defined models developed by the original equipment manufacturers (OEMs). RMS models have been the preferred option due to their low computational burden and fast simulation speed. However, there are significant challenges in using these types of models, to represent accurately the performance of the wind power plant (WPP) under several conditions like weak grid scenarios, feeder loss, etc. The nature of RMS simulation tools forces model developers to implement simplifications over the control design of the real product, especially in the instantaneous and fast control loops, to adapt to the phasor calculation method of RMS simulation tools. Consequently, the use of RMS models to demonstrate grid code compliance for specific WPP projects may differ significantly compared to the real power plant performance.
Nowadays, due to the increasing penetration of inverter-based resources (IBR), the industry is rapidly facing new challenges inherent to this kind of technology. Therefore, System Operators (SO) around the globe are scrutinizing the accuracy and simplification of RMS models used for grid interconnection studies to demonstrate grid code compliance, by introducing strict requirements to compare simulation results of RMS models against Electromagnetic transient (EMT) simulation models. In 2020, ERCOT in the USA introduced a new set of model acceptance requirements to ensure the models produce consistent results between RMS, EMT, and on-site performance.
The purpose of this paper is to outline the challenges found during the verification process for simulation model performance in several commercial RMS and EMT simulation tools required by ERCOT. These challenges are categorized into 2 main groups: Simulation Tool Limitations (STL) and Simulation Model Simplification (SMS). STL covers intrinsic conditions of RMS tools when compared to EMT tools like software solver, time step, acceleration factor. SMS’s consist of simplifications used for the model development in RMS environment, to adapt to phasor domain calculation and greater simulation time steps. The paper highlights the lessons learned, recommended practices, and underlying reasons for discrepancies when benchmarking simulation results from RMS tools vs EMT tools. Finally, the paper will elaborate how Vestas solution overcomes the aforementioned challenges by using a full source code integrated model called Unified Model Framework (UMF), based on the original control developed by the different Vestas design teams for both wind turbine and power plant controller products, which is shared between RMS and EMT tools and maintains a 1-2-1 parametrization with the real product.
The provision of fast fault current injection is of great importance for power system protection and stability [1]. The European network code Requirement for Generators [2], in force since May 2017, gave the opportunity to Transmission System Operators (TSOs) to specify this feature in their grid codes. In August 2020 the French regulation [3] was updated to add such new requirement. However, the impact on the transmission network and interaction between the commissioned generators (not submitted to the new regulation) and new ones could be of concern, particularly when this configuration could take place within the same power park generating facility.
To address this topic, RTE (French TSO) is conducting several dynamic studies using manufacturer’s black-boxed models in Electromagnetic transient (EMT) type tool to ensure high accuracy and reliable results. ENERCON and RTE have decided to collaborate on this subject, to share their knowledge and to perform onsite tests of this new feature in an existing wind park, which is planned to be upgraded. The first step of the collaboration, which is addressed within this paper, is the conducted EMTP-RV studies on the wind park. After validating the black-boxed ENERCON model, several dynamic simulations, including fault at PCC and inside the wind park, were conducted to investigate the impact of the implementation of the fast fault current injection in one part the of the wind park.
The outcome of this study provides support for the implementation in the field of the reactive current injection on some generators of the chosen park and its monitoring.
[1] : “Fault current contribution from PPMS & HVDC”, ENTSO-E guidance document for national implementation for network codes on grid connection, 16/11/2016, https://eepublicdownloads.entsoe.eu/clean-documents/Network%20codes%20documents/NC%20RfG/161116_IGD_Fault%20Current%20Contribution%20from%20PPMs%20%20HVDC_for%20consultation_for%20publication.pdf
[2] : “COMMISSION REGULATION (EU) 2016/631 of 14 April 2016 establishing a network code on requirements for grid connection of generators”
[3] : “Documentation Technique de Références”, RTE, 03/08/20, https://www.services-rte.com/files/live/sites/services-rte/files/documentsLibrary/03-08-20%20complet_fr.7z
The pursuit for carbon neutrality has led to a significant increase of renewable energy infeed to the electrical grid over the last few decades. High penetration of such intermittent sources has made it difficult for system operators to ensure grid stability in terms of frequency, voltage and power quality. The operation of stochastic renewable resources over different timescales has increased the dispatch of emergency power reserves in some situations and led to power curtailment in others. Both of these processes have resulted in a significant loss of revenue for the grid operator. Hence, to improve power system flexibility, a large-scale energy storage scheme is required, which will help maintain a continuous balance between electrical power generation and consumption, minimizing associated penalties. The Hydrogen Storage Power Plant (HSPP) is one such solution. Such an interconnected system has been designed to not only mitigate the randomness in power generation due to renewables but also provide crucial frequency (synthetic inertia) and voltage ancillary support to the grid, especially in the absence of fossil-fired power plants in the future.
In this paper, a novel structure of the HSPP consisting of storages and DC-AC converters (HSPP-AC) is proposed. Under this renewed structure, the power plant can be operated as one unit (Combined HSPP-AC) as well as split into individual storages and converters (Distributed HSPP-AC). Distributed HSPP-AC operation would make it easier to choose optimized locations of the power plant components and lead to an easier scale-up process. This research is aimed to test the behavior of this Distributed HSPP-AC structure against its combined form as well as other types of power plants. This is achieved by implementing both HSPP variations in an isolated network containing conventional thermal and hydroelectric as well as a large share of wind power plants. The dynamic interaction of the different HSPP-AC versions with the other power plants and the roles of their internal components are analyzed in response to electrical disturbances in the three-phase grid. The results are expected to signify that the Distributed HSPP-AC, just like its combined counterpart, can ensure stable operation of a grid with a high penetration of renewable sources.
Harmonic-related issues in wind collector systems have been discussed at the Wind Integration Workshop for a number of years. In addition to background harmonic voltage from power systems and harmonic current emission from wind turbines, resonance has been identified as an important factor. A negatively damped (i.e., unstable) resonance can create harmonics at the resonance frequency that otherwise do not exist. A lightly damped resonance may also amplify an existing harmonic, leading to unexpectedly high distortion. A number of papers have been presented on the modeling and analysis of wind turbine impedances and wind collector system resonances. This paper presents techniques to manage lightly or negatively damped system resonances.
Wind collector systems include several cable arrays connecting individual turbines to a local substation. Typical voltage levels range from 20kv to 66kV. The substation that includes one or more transformers to raise the voltage to transmission level. The transmission system can include additional high-voltage ac cables or an HVDC system. All the cables and transformers are designed for low losses, which creates resonances that have very little inherent damping. The resonant frequencies vary with the number of cables, operating wind turbines, and transformers.
Wind turbines are also designed for low losses, and the converter controls must include some form of fast current regulation to ensure operation within the equipment limits. Fast current control inherently creates a damping effect at low frequencies, but leads to a negative damping effect at higher frequencies due to unavoidable latencies within the power electronic controls [CIGRE754]. HVDC converters exhibit similar negative damping effects. Such negatively damped converter impedance in combination with the complex collector system impedance can lead to lightly or negatively damped system resonances over a wide frequency range.
This presents a challenge that is frequently addressed only late in a project development. Several projects have experienced instabilities or high distortion during commissioning with subsequent delays in operation while the problem is investigated, and mitigation developed. The issue becomes more severe with offshore wind projects using large wind turbines since the collector voltage is 66kV -vs- 33kv or lower for previous projects. This doubling of voltage increases the effective cable charging by 4:1, with consequent lower resonant frequencies and higher energy associated with transients.
The proposed paper will describe the physical issues involved in wind collector system resonances, the limitations of wind turbine controls for mitigation, the design considerations for passive filtering as part of the mitigation, and a screening approach to evaluate the situation early in the development process. A case study is also presented to illustrate the design process.
This paper describes the method and results of a grid study that examines the technical and economic potential of "grid enhancing technologies" to significantly reduce congestion and its associated costs brought about by the Energy Transition. The study examines how a portfolio of different grid enhancing technologies can complement one another to mitigate grid congestion. In addition to more qualitative analysis of the opportunities offered by such technologies, the assessment is based on quantitative energy system modelling for a 2030 scenario including dispatch, load flow, and remedial action simulation. The geographic scope of the study covers the Central West Europe region together with Denmark West.
Dynamic Line Rating (DLR) is understood to be compatible with wind integration, but not yet widely adopted. Regulation is a key enabler, as is knowledge sharing of how it is implemented by different users.
This paper reports findings from projects implemented in France, Germany and Belgium. Time-correlated series data of wind power infeed, line loading levels, and measured thermal ratings were analyzed to assess the benefit of DLR for wind power integration. Not only real-time data but forecast data is also analyzed.
In cases where wind integration is viewed as the main challenge, high power flows on the lines correlated with high wind infeed periods. Furthermore these periods of high power flows correlated with extra capacity made available by DLR. With increasing amounts of wind power installations, the frequency of high load periods increase on a line. These coincide with the extra capacity made available by DLR, resulting in benefits such as saved redispatch and curtailment costs, as well as increased hosting capacity to accelerate connection.
However, it is not very clear how these benefits should be distributed in time: in the long-term planning; in day-ahead market exchange; or in close to real-time grid operations? This paper summarizes observations from various projects implemented in neighboring European countries, to shed some light on the benefit of grid capacity in different timeframes.)