Protection Technology. Challenges of protection power grids with distributed generating units. Technical Article

Transcription

1 Protection Technology Challenges of protection power grids with distributed generating units Technical Article

2 The energy transition changes power grids Challenges of protecting power grids with distributed generating units The energy transition implies a paradigm shift in the energy system and has an impact on the generation, transmission and distribution of electricity. An increasing number of distributed generating units based on renewable energies have to be integrated. Wind energy and photovoltaics systems are the fastest growing suppliers. This calls for a protection concept that is tailored to the new conditions, the selection of appropriate protection devices including the functions and protective settings adjusted to the changed grid conditions. This essay presents possible consequences for the protection technology. As the energy transition is being put into practice, energy is generated at all voltage levels which entails a change of classic grid tasks. A distribution grid contributes significantly to the generation of electricity and assumes transport tasks. The protection system makes a major contribution to the supply security. Marginal conditions Distributed generating units have to observe guidelines and they are subject to certification. The following guidelines have to be complied with: Gridcode (for example transmission code TC 2007) which specifies the feed-in conditions (for example, also feed-in during a fault fault ride through, reactive power provision for voltage stabilization)) [1] Guidelines for generating units at the medium-voltage grid 2008; EEC generating units at the high- and extra high-voltage grid 2004 and their update in the technical connection conditions TAB [2, 3] System service provision ordinance on wind energy farms (SDLWindV) [4]. General requirements for the protection technology Power grid studies found the following: Existing protection principles and schemes are basically suitable. However, protective settings have to be adjusted to the changed grid conditions. The contribution of converter-controlled generating units to the short-circuit current is low and is within the range of the generating unit's rated current (1 Irated) and in some cases even below, for example under weak wind conditions. This has an impact on the selection of protection principles and protective settings. It is recommended to make sample calculations to distribute the short-circuit fault during a fault. In case of unbalanced faults, the fault current contribution depends on control procedures of the inverter or frequency converter (for example only feed-in of a positive-sequence current). To cope with unbalanced faults, a negative-sequence current has to be injected. The inverter operating principle has to be inquired with the manufacturer. Due to the distributed feed-ins, the distribution grid's topology and tasks are similar to the transmission grid. For example, there are multi-end feed-ins or intermediate feed-ins. This requires established protection concepts and graded protection principles from the transmission grid to be implemented. In addition to protection tasks, grid disconnection and load shedding have to be implemented, taking the feed-in conditions into account. Digital protection devices are the optimum solution for many applications Figure 1. single-pole earth fault and no infeed of short circuit through wind farm The multifunctional digital device technology and the functions including state-of-the-art communication features are designed to enable solutions tailored to the application.

3 x Figure 2. example of additional infeed of synchronous generator into the medium-voltage grid a) situation of infeed and failure; b) impedance trajectory depending on duration of failure They are supported by a powerful engineering tool (for example Digsi 5 from Siemens). Siprotec 5, the new family of protection devices from Siemens, combines protection, control, automation, measurement, regulation and communication in one device [5] and has the following typical properties: A modular hardware and software concept enables the devices to be tailored to the grid conditions (for example free selection of the measuring points, of the I/O quantity structure, the protection functions and the communication infrastructure). Identical functions are used in various devices, increasing the flexibility of application (communication of protection data between the devices). High-precision measurements (class 0.2 for U, I and class 0.5 for P, Q, S) make additional measuring transducers redundant and additionally enable a wide setting range of the protection functions. Powerful fault recording and evaluation tools support fault recovery. Different protocols and communication infrastructures provide for the optimal connection to the substation control and protection system. The devices support IEC Edition 1 and 2. Applications We have selected some representative examples from the application options. They give an overview of possible changes: Feed-in of a wind farm Figure 1 shows a wind farm connected to the grid with a phase-to-ground fault occurring in the grid. The individual wind energy units feed energy to the grid through the frequency converter. No zero-sequence system is transmitted in case of a transformer in the star delta vector group. A phase-to-ground fault at the star side results in a 2-phase short circuit with a distinct negative-sequence current. But since the wind farm in its present design does not feed in any negative-sequence current (no negative-sequence impedance available), the wind farm neither supplies a fault current. This is equivalent to an open circuit breaker in the wind farm. The fault current is solely fed through the remaining grid. The behavior (open circuit breaker at the transformer delta side) is known as the Bauch phenomenon in protection technology. Figure 1 depicts the situation and shows the currents and voltages at the measuring point at the transformer star side. The distinct zero-sequence current in all three phases is typical of the Bauch phenomenon. A distance protection must recognize this situation, select the proper measuring loop and trip selectively. As feed-ins through inverters or frequency converters will increase, these also have to be capable of supplying a negative-sequence current during a fault from a protection perspective [6, 7]. Simulations have shown that when inverter controllers are not set correctly, inverters are capable of delivering harmonics during a fault as a result. When the setting is correct, the feeding quantities are sinusoidal (fundamental component). As faults cannot be ruled out entirely, the protection has to be resistant towards harmonics. Additional feed-in of synchronous generators in a medium-voltage grid In this example, a synchronous generator (biomass, hydro-electric power plant) additionally feeds energy into the grid. The grid connection design is depicted in Figure 2a. The busbar is fed through a strong grid. The protection in the outgoing feeders is implemented as a directional time overcurrent protection. The selected delay time t results from the grading time. Now if a short-circuit occurs on the line of protection equipment A, this fault is fed both by the grid and by the synchronous generator. Upon fault inception, a forward fault is recognized in A and timing element t is started. Protection device B, in contrast, recognizes the reverse direction. In the event of a fault, the synchronous generator is relieved in terms of active power, causing the generator to accelerate and the angular displacement to change. The current/voltage vector will thus offset over time. In line B that is feeding the fault current, the protection device will recognize the reverse fault as a forward fault after a certain period of time and also start the timing element. Depending on the fault duration and the set grading time, the outgoing feeder B may also trip. To prevent this non-selective tripping, the tripping time in outgoing feeder A has to be reduced. This can be achieved by replacing the so far directional overcurrent protection with a distance protection.

4 Figure 4 shows a connection example, the most important protection and disconnection functions and their settings. Other well-known disconnection criteria such as the phase angle jump of the voltage vector (vector jump) and the frequency rate of change function df/dt) are not recommended, because they require complex settings (grid calculations with variant consideration) and involve the risk of over- or underfunctioning (practical experience). The threshold values of the frequency functions correspond to the requirements from [1] and are identical to those of conventional power plants (type 1 acc. to [1]). Figure 3. load shedding by taking into account infeed conditions But there are other approaches to reducing the tripping time: Using the line differential protection that needs an additional backup protection to respond when the upstream protection device fails. Possible backup protection functions include a time overcurrent protection (also directional depending on the application) and a distance protection at higher voltage levels [8]. Expanding the directional time overcurrent protection to ensure protection data communication and implementation of directional comparison protection. The latter has to be designed so as to require no infeed from the remote end. This means that if the infeed side protection device picks up and the opposite side is not blocked, the device will trip short time [9]. Load shedding If consumed and generated active power are out of balance, the previous solution has been to shed loads with underfrequency based on a fixed frequency stage plan (5-stage plan [1]). If, however, distributed generating units supply at the grid connection point, these are lost. This exacerbates the situation between generation and consumption. Therefore, the international and national guidelines were adapted accordingly without any further development towards more flexibility [8]. If there are distributed feed-ins besides loads at the grid connection point, the circuit breaker must not open when energy is fed back (Figure 3). The frequency protection has to be connected to the power directional protection. This automation environment in particular contains modern, digital devices that offer various possibilities, such as the function itself, high-precision measurements and flexible communication. Adaptive load shedding becomes reality. Grid disconnection Given the rapid increase of distributed feed-ins (in particular type 2 according to [1]), the disconnection strategy has changed over the years. The goal is to keep generating units connected to the grid for as long as possible. Besides grid disconnection specifications, the guidelines also define marginal conditions for the feed-in behavior. For example, the fault ride-through capability is required. This means that during a 3-pole short-line fault generating units have to stay connected to the grid for at least 150 ms and feed in again once the fault has been cleared by the protection device (e.g. the differential protection). Moreover, the systems have to contribute to voltage stabilization resulting from the requirement to inject reactive power in the event of a fault. A pragmatic approach was chosen for the definition of the disconnection criteria. In addition to overfrequency and underfrequency (f>, f<), overvoltage and undervoltage (U>, U<) are required. The directional reactive power function with additional undervoltage release in all three phases is a new feature (evaluation of the phase-to-phase voltage). This function aims to prevent the voltage from being reduced further when reactive power is imported. The selectivity of the disconnection functions is implemented by different threshold values and timer settings. Different variants of the plant design and special cases to be considered were discussed in detail in [10]. For this reason, we will go into Figure 4 by way of example. It depicts a typical constellation: Supply to a medium-voltage grid and connection to the high-voltage grid through a radial-line connection. The solutions to protect the tapped line on the high-voltage side are sufficiently known. There is one special aspect in this application. If faults occur in the high-voltage grid, the circuit-breaker in the radial-line connection remains closed. But the circuit-breaker of the feed-in into the medium-voltage grid will be intertripped by distance protection A. The goal is to assure the supply reliability of the consumers in the medium voltage, since most faults in high voltage can be cleared by automatic reclosing. In medium-voltage faults, distance protection A assumes the function of a backup protection. Distance protection B fulfils the same function with the reverse zone. The forward zone of distance protection B clears faults on the line towards the generating unit. It is crucial for the distance protection setting that fault currents inside or below the rated current range have to be mastered. A U/I and Z pickup is a mandatory requirement. The overvoltage disconnection function protects a grid against overvoltage and in doing so contributes to the power quality. The undervoltage function protects the grid against inadmissible undervoltage and is the backup protection of the distance protection if the latter fails to pick up. The grid connection U< stage assures the grid disconnection and it is the backup protection for the undervoltage protection of the generating unit. The U< stage assures the grid disconnection in the event of undervoltage conditions and simultaneously assumes a backup function for the short-circuit protection of the generating unit (for type 2) and the U<< stage. The U<< stage is the main protection for low-current faults. The QU protection monitors the appropriate post-fault behavior and performs the separation at the grid connection point.

5 Summary We have used examples to show that the energy transition with its distributed generating units has an impact on the protection technology. Existing protection principles still have a right to exist. The significant changes in the feed-in situation call for a reassessment of existing protection concepts and settings. Hence, it is insufficient to simply connect the generating unit (as certified). A holistic consideration of the system has to be performed at any rate. In this context, generating units are not all the same. Type 1 units, feed-in via synchronous generators, behave in the well-known way. In contrast, type 2 units, such as feed-in via frequency converter/inverter, exhibit a different behavior. In particular, the low short-circuit current has to be coped with in the event of a fault. As a result, distance or differential protection schemes and modified settings (pickup and rated current) are required. Modern digital protection devices have the potential to tackle the new challenges. Future solutions will be based not only on the implementation of new functions, but also on flexible communication features including high-precision measurement technology. It is crucial that both grid operators and device manufacturers exchange their experiences and work together on developing a new and reliable electrical energy system. References [1] Transmission Code 2007: Netz-und Systemregeln der deutschen Übertragungsnetzbetreiber. VDN Berlin, Version 1.1, August [2] Technische Richtlinie Erzeugungsanlagen am Mittelspannungsnetz. Regelungen und Übergangsfristen für bestimmte Anforderungen in Ergänzung zur technischen Richtlinie. BDEW, Berlin, January 2013 edition. [3] EEG-Erzeugungsanlagen am Hoch- und Höchstspannungsnetz. VDN Berlin, August 2004 und TAB Hochspannung, VDE-AR-N 4120, VDN Berlin, January [4] Verordnung zu Systemdienstleistungen durch Windenergieanlagen (SDLWindV). German Federal Environment Ministry, Berlin, 3 July 2009, and amendments in [5] Siprotec 5 Schutz, Automatisierung und Überwachung. Übersichtskatalog C1000G220-A221, Siemens AG, Nürnberg, [6] Erlich, I.; Schegner, P.: Wind Turbine Negative Sequence Current Control and its Effect on Power System Protection. IEEE PES GM, Vancouver [7] Kühn, H.: Der Einfluss regenerativer Erzeugungsanlagen auf den Kurzschlussstrom. 7. ETG/FNN-Tutorial Schutz- und Leittechnik, Mainz, February [8] Hülshorst, H.-G.: Auswirkung der dezentralen Einspeisung auf das Lastabwurfkonzept. ETG-Tagung Schutz- und Leittechnik, Düsseldorf, February [9] Siprotec-5-Distanzschutz, Leitungsdifferentialschutz und Schaltermanagement für einpolige und dreipolige Auslösung 7SA87, 7SD87, 7SL87, 7VK87. Handbuch C53000G5000-C011-6, Siemens AG, [10] Siprotec 5 Überstromzeitschutz 7SJ82,7SJ85, Handbuch C53000G5000-C017-5, Siemens AG, [11] Hinz., K.; Schossig, W.: Schutzkonzeptionen für Verteilnetze dezentrale Energieerzeugungsanlagen. Anwendertagung Omicroncamp, Ulm, June Author Dipl. Ing. habil. Hans-Joachim Herrmann, Product Lifecycle Manager, Protection and Substation Automation, Principa Key Expert Energy Management Division, Energy Automation, Siemens AG, Nuremberg