Microgrid Concepts and Distributed Generation Technologies ECE 2795 Microgrid Concepts and Distributed Generation Technologies Spring 2017 Week #14 © A. Kwasinski, 2017

Microgrids and grid interaction Microgrids could have a grid interconnection to Improve system economics Improve operation Improve availability With a suitable planning, grid planning can benefit from having microgrids by Reducing conductor’s size Improving availability Improving stability Tools, strategies and techniques for an effective integration of a microgrid into the main grid: Net metering – bi-directional power flow. Peak shaving Advanced communications and controls Demand response (?) © A. Kwasinski, 2017

Microgrids and grid interaction Interconnection practice / recommendation: IEEE standard 1547 Potential issues with microgrids integration into the main grid: Infrastructure long term planning / economics: There is no coordination in planning the grid and microgrids. The grid is planned on a long term basis considering traditional loads. Microgrids may “pop-up” afterwards “without notice.” Grid’s planning links economic (cost of grid’s electricity, future demand…..) and technical aspects (line congestion….) Stability: microgrids are variable loads with positive and negative impedance (they can act to the grid as generators) © A. Kwasinski, 2017

Microgrids and grid interaction More potential issues with microgrids integration into the main grid: Safety: When there is a fault in the grid, power from the microgrid into the grid should be interrupted (islanding) Availability: Microgrids can trigger protections (directional relays) upstream in the grid and interrupt service to other loads Key issue: microgrids are supposed to be independently controlled cells within the main grid. How much independence microgrids should have? Does independence apply also to planning? How much interaction / communications should be between the grid and the microgrid? © A. Kwasinski, 2017

Microgrids and grid interaction Example of microgrid development. Initial condition. Equipment and financial planning is done with all the load in the figure in mind. © A. Kwasinski, 2017

Microgrids and grid interaction Example of microgrid development. Planning issues. A microgrid is installed few years later. Transformers and conductors can now be oversized (remember this aspect for PEV and PHEV integration) Microgrid’s area © A. Kwasinski, 2017

Microgrids and grid interaction Example of microgrid development. Initial normal power flow direction Directional Relay © A. Kwasinski, 2017

Microgrids and grid interaction Example of microgrid operational issues. New power flow with a microgrid. The microgrid’s power trips open the directional relay Is it possible to change the grid’s state fast enough to prevent voltage collapse due to loss of stability caused by the sudden load changes introduced by the microgrid? Directional Relay What microgrid’s control action follows? Can the microgrid stop injecting power back into the grid (i.e. prevents islanding)? Microgrid’s area © A. Kwasinski, 2017

Microgrids and grid interaction Example of microgrid operation. Islanding. If islanding occurs the microgrid will continue to provide power to a portion of the grid even though the grid connection upstream has been interrupted. Potential issues: Utility crews safety. Power quality at the energized portion could be poor. Loads could be damaged. “Island” Microgrid’s area © A. Kwasinski, 2017

Microgrids and grid interaction Grid interconnection might be different for dc or ac microgrids For ac microgrids, grid interconnection can be done directly, with a disconnect switch, and a transformer only. For dc microgrids an inverter is necessary Examples: CERTS microgrid (ac) NTT Facilities Sendai project (ac and dc) © A. Kwasinski, 2017

Microgrids and grid interaction dc microgrids integration with the grid The interface may or may not allow for bidirectional power flow. Bidirectional power flow can be needed for: `Energy storage dc loads © A. Kwasinski, 2017

Smart grids There are two similar but not equal approaches to the smart grid concept. EU-led vision (customer and environmentally driven): Europe’s electricity networks in 2020 and beyond will be: Flexible: Fulfilling customers’ needs whilst responding to the changes and challenges ahead; Accessible: Granting connection access to all network users, particularly for renewable energy sources and high efficiency local generation with zero or low carbon emissions; Reliable: Assuring and improving security and quality of supply, consistent with the demands of the digital age; Economic: Providing best value through innovation, efficient energy management and ‘level playing field’ competition and regulation. “European Technology Platform SmartGrids. Vision and Strategy for Europe’s Electricity Networks of the Future” European Commission KI-NA-22040-EN-C EUR 22040 © A. Kwasinski, 2017

Smart grids US led vision (security and consumer driven) - Motivated by needs in availability improvements “The NETL Modern Grid Initiative A VISION FOR THE MODERN GRID”, US DOE © A. Kwasinski, 2017

The smart grid concept There are many views of what is In reality, a smart grid is not a single concept but rather a combination of technologies and methods intended to modernize the existing grid in order to improve flexibility, availability, energy efficiency, and costs. Smart Grid 1.0: Smart meters Smart Grid 2.0 (“Energy Internet” enabler): advanced autonomous controls, distributed energy storage, distributed generation, and flexible power architectures. Distributed generation (DG), flexible power architectures, autonomous controls and loads constitute local low-power grids (micro-grids). © A. Kwasinski, 2017

Smart grid evolution: Past Centralized operation and control Passive transmission and distribution. Lack of flexibility Vulnerable © A. Kwasinski, 2017

Smart grid evolution: Present/immediate future Still primarily centralized control. Limited active distribution network (distributed local generation and storage). Use of virtual storage (demand-response) Addition of communication systems More efficient loads Flexibility issues Somewhat more robust © A. Kwasinski, 2017

Smart grid evolution: Future Distributed operation and control Active distribution network (distributed local generation and storage). Integrated communications Advanced more efficient loads Flexible More robust © A. Kwasinski, 2017

Smart grids Technologies and concepts: Distributed energy resources (generation and storage) are fundamental parts. They provide the necessary active characteristics to an otherwise passive grid. Advanced and distributed communications. All the grid components are able to communicate. The grid operates like a power-Internet (distributed, multiple-redundant, interactive and autonomous). I.e. a Power-Net. Intelligent metering. Policies and regulatory actions. Necessary to achieve integration of all the parts. Inadequate pricing models is a significant barrier to introduce service-based business models (vs. energy-based). Grid modernization. © A. Kwasinski, 2017

The Power-Net DOE view for a smart grid: - “An electrical grid is a network of technologies that delivers electricity from power plants to consumers in their homes and offices.” A Power-Net expands this view based on paradigms from the Internet. Some features compared with conventional power grids: more reliable, efficient, and flexible. © A. Kwasinski, 2017

The Power-Net Like the Internet, the Power-Net involves diverse and redundant path for the power to flow from distributed generators to users. Its control resides in autonomous distributed agents. Power is generated in distributed generators, usually from alternative or renewable energy sources. Power buffers are included to match generators and loads dynamics. Energy buffers are added to make variable sources dispatchable. Contrary to the Internet, the Power-Net involves a local approach for power interactions. © A. Kwasinski, 2017

The Internet Desired Internet features: distributed and autonomous control, diverse information routing and redundant data or application storage, performance degradation instead of full failure, link transmission rate control through temporary data storage in buffers. Buffer size Link bandwidth Maximum (delay) time © A. Kwasinski, 2017

Extending the Internet into Smart grids Key aspect: add distributed generation (fuel cells, microturbines, PV modules, small wind, reciprocating engines) close to the load to make power grids distribution portion an active electric circuit. Autonomous and distributed controls can be implemented with DG. Power vs. Energy buffers: Predicted solar radiation on PV module Batteries (Energy buffer) Ultracapacitors or flywheels (power buffer) © A. Kwasinski, 2017

Control and communication issues Coordination is needed in order to integrate variable generation sources (such as PV modules) in the grid. Centralized control requires significant communication resources (i.e., large bandwidth spectrum allocation) which in general is not available. The alternative is to provide all active nodes with an autonomous control that allows controlling power interactions with the grid without dedicated communication links. These more intelligent nodes become agents. VS. © A. Kwasinski, 2017

Power Supply Resilience In the past, several issues were identified in conventional power grids that affect their availability, particularly during natural disasters. Conventional power grids were shown to be very fragile systems. Some of the issues found in conventional power grids include: Primarily centralized control and power distribution architecture. Passive power distribution grid Lack of redundancy in most sub-transmission and distribution paths. Difficulties in integrating meaningful levels of energy storage. Power supply issues during disasters is a grid’s problem transferred to the load. © A. Kwasinski, 2017

Power Supply Resilience Smart grid planning for disaster resiliency must consider disaster impact on lifelines. During disasters special attention should be paid to dissimilar ways in which disasters affect different distributed generation (DG) technologies. Renewable sources do not have lifelines but they are not dispatchable, they are expensive, and they require large footprints. Most DG technologies have availabilities lower than that of the grid. DG needs diverse power supply in order to achieve high availabilities. DG provides a technological solution to the vulnerable availability point existing in air conditioners power supply. DG provides the active component to grid’s distribution portion, essential for advanced self-healing power architectures. © A. Kwasinski, 2017

Extending the Internet into Smart grids Lifeline dependencies can be reduced by extended local energy storage. Lifeline’s effects on availability can be mitigated with diverse local power generation. PVs and wind do not require lifelines but their variable profile leads to added DG or extensive local energy buffers. Performance degradation: voltage regulation or selective load shedding. Advanced (active) distribution through power routing interfaces © A. Kwasinski, 2017

Advanced Power Architectures Power routers Monitoring points A hybrid ac (solid lines) and dc (doted lines) architecture with both centralized and distributed generation resources. © A. Kwasinski, 2017

Grid-Microgrids Interconnection Part 2 Grid-Microgrids Interconnection © A. Kwasinski, 2017

Motivation Reasons for connecting a microgrid to a main grid: Availability: Highly available power grids may act as an additional source for micro-grids. Operations/stability: Direct connection of ac microgrids to a large power grid facilitates stable operation but only if the power grid acts as a “stiff” source to the microgrid. When using renewable energy sources, a grid connection may allow reducing the need for energy storage in the microgrid. If not all loads in a microgrid are critical, a grid connection may allow to reduce the investment in local generation. Economics: Microgrids are typically planned with extra capacity with respect to the local load. This extra power capacity can be injected back into the grid in order to obtain some economic benefit. Grid interconnection allows to reduce fuel operational costs by using the grid at night when electricity costs are low. © A. Kwasinski, 2017

Definitions Point of common coupling (PCC): it is the point in the electric circuit where a microgrid is connected to a main grid. © A. Kwasinski, 2017

Standards There are several standards specifying various aspects grid interconnection of a local power generation source. Arguably the most important one is IEEE 1547. IEEE 1547 has several parts: Main body IEEE Standard 1547.1 “IEEE Standard Conformance Test Procedures for Equipment Interconnecting Distributed Resources with Electric Power Systems.” EEE Standard 1547.2 “IEEE Application Guide for IEEE Std 1547™, IEEE Standard for Interconnecting Distributed Resources with Electric Power Systems.” IEEE Standard 1547.3 “IEEE Guide for Monitoring, Information Exchange, and Control of Distributed Resources Interconnected with Electric Power Systems.” IEEE Standard 1547.4 “IEEE Guide for Design, Operation, and Integration of Distributed Resource Island Systems with Electric Power Systems.” IEEE Standard 1547.5 has not been issued, yet. Its intended scope is to address issues when interconnecting electric power sources of more than 10 MVA to the power grid. IEEE Standard 1547.6 “IEEE Recommended Practice for Interconnecting Distributed Resources with Electric Power Systems Distribution Secondary Networks.” IEEE Standard 1547.8 has not been issued, yet. Its intended scope is to provide supplemental support for implementation methods for expanded use of the previous standards, for example when addressing issues with high penetration of residential PV systems. © A. Kwasinski, 2017

Standards Main provisions from IEEE 1547: The micro-grid must “not actively regulate the voltage at the PCC.” The grounding approach chosen for the local area power and energy system (LAPES) must not create overvoltages that exceed the ratings of the equipment connected to the main grid or must not affect ground fault protection coordination in the main grid. The distributed resources in the LAPES must be able to parallel with the main grid “without causing voltage fluctuations at the PCC greater than ±5% of the prevailing voltage level of the Area electric power system (EPS) at the PCC” and flicker must be within acceptable ranges. The LAPES must not energized the main grid when the main grid is not energized. Each distributed resource (DR) “unit of 250 kVA or more or DR aggregate of 250 kVA or more at a single PCC shall have provisions for monitoring its connection status, real power output, reactive power output, and voltage at the point of DR connection.” A visible-break isolation device must be located between the main grid and a DR unit only when required by the main grid provider practices. The interconnection system must meet applicable surge and EMI standards. © A. Kwasinski, 2017

Standards Main provisions from IEEE 1547: When a fault occurs in the main grid circuit to which a LAPES is connected, then the micro-grid local power generation units must stop to power this circuit before reclosure from the main grid happens. The interconnection system must be able to measure relevant indicated voltages and frequencies at the PCC or the point of connection of DR and disconnect within a given allowed time all local power generating units in the micro-grid when these measured voltages or frequencies fall within a range specified in a table in this standard. For example, when voltages fall below 50 % of the base voltage, the LAPES must disconnect its DR within 0.16 seconds (one 60 Hz cycle). The time extends to 2 seconds for voltages between 50 and 88 % of the base voltage. Disconnection must occur within 1 second if measured voltages are between 110 and 120 % of the base voltage and within 0.16 seconds if the voltage exceeds 120 % of the base voltage. For frequency measurements, any DR of 30 kW or less must disconnect 0.16 seconds if the measured frequency is above 60.5 Hz or below 59.3 Hz. The same disconnect time applies for DR of more than 30 kW when the frequency exceeds 60.5 Hz, but for the lower range at these power levels disconnect within 0.16 seconds must occur if the frequency falls below 57 Hz, whereas disconnection is adjustable between 0.16 and 300 Hz if the frequency falls between 59.8 and 57 Hz. Reconnection of a LAPES to a main grid may occur at least 5 minutes after voltages and frequency fall within indicated required ranges. © A. Kwasinski, 2017

Standards Main provisions from IEEE 1547: Reconnection of a LAPES to a main grid may occur at least 5 minutes after voltages and frequency fall within indicated required ranges. A microgrid must “not inject dc current greater than 0.5% of the full rated output current” at the PCC. Harmonic current injection by the LAPES into the main grid measured at the PCC must not exceed certain levels both in total and for given harmonic order ranges. The total demand distortion must not be more than 5 % of the local main grid “maximum load current integrated demand (15 or 30 minutes) without the DR unit, or the DR unit rated current capacity,” whatever is greater. Base of this same base current, harmonic content for harmonics with an odd order below 11 must not exceed 4 %. If the odd harmonic order is between 11 and 17 the limit is 2 %, whereas this limit falls to 1.5 % for odd harmonics with an order between 17 and 23 and 0.6 % for odd harmonics with an order between 23 and 35. For odd harmonics with an order above 35, the harmonic content with respect to the indicated current must not exceed 0.3 %. For even harmonics their content limits are a quarter of those indicated for the odd harmonic orders. © A. Kwasinski, 2017

Standards Other important provisions from IEEE 1547.6 about network protections (NP) on the grid’s side: The presence of DR should not: - “cause any NP to exceed its fault-interrupting capability.” - “cause any NP to operate more frequently than prior to DR operation.” - “prevent or delay the NP from opening for faults on the network feeders.” - “delay or prevent NP closure.” - “require the NP settings to be adjusted except by consent of the area EPS operator.” - “cause an islanding condition within part of a grid network.” © A. Kwasinski, 2017

Interconnection methods and technologies Directly through switchgear Power electronic interfaces Static switches Directly through circuit breakers: Relatively simple and inexpensive Slow (3 to 6 cycles to achieve a complete disconnection). Since electrical characteristics on both sides of the circuit breakers must be the same, then, electrical characteristics on the micro-grid side are dependent on the grid characteristics. For example, use of a circuit breaker implicitly limits the micro-grid to have, at least partially, an ac power distribution system in order to match the grid’s electrical characteristics. Power flow through the PCC cannot be controlled © A. Kwasinski, 2017

Interconnection methods and technologies Directly through circuit breakers: Example of one of such systems: Use of static switches: Usually based on SCRs in antiparallel configuration to allow bidirectional power flow © A. Kwasinski, 2017

Interconnection methods and technologies Use of static switches: They are costlier and more complex than using circuit breakers. Usually, conventional circuit breakers are still used to provide a way to achieve full galvanic isolation. A Bypass switch is also added for maintenance reasons. © A. Kwasinski, 2017

Interconnection methods and technologies Use of static switches: They allow for many open/close operations They act much faster than conventional circuit breakers (in the order of half a cycle to a cycle). Sometimes IGBTs are used instead of SCR because IGBTs tend to be faster than SCRs and their current is inherently limited. Still power flow cannot be controlled. There are some conduction losses in the devices. © A. Kwasinski, 2017

Interconnection methods and technologies Power electronic interfaces: It is the costlier option but it is also the most flexible one. Allow for power distribution architecture characteristics on both sides of the PCC to be completely different. Both real and reactive power flow can be controlled. Reaction times to connection or disconnection commands are similar to those provided by static switches, although in the case of a power electronic circuit, it response also depends on its dynamic performance, given by its controller, topology, and internal energy storage components characteristics. Still, in many cases, a circuit breaker will still be required at the grid-side terminal of the power electronic interface with a microgrid in order to provide a way to physically disconnect the micro-grid from the grid. Also, similarly to static switches, the presence of a power electronic circuit will lead to some power losses not found in the approach using mechanical interfaces. © A. Kwasinski, 2017

Islanding In IEEE Standard 1547.4 an intentional island is said to be the result of “intentional events for which the time and duration of the planned island are agreed upon by all parties involved.” There are several reasons why intentional island operation of a micro-grid may occur, but a common one is a preemptive disconnection from the grid in anticipation of a power outage on the main grid side caused by an event that can be anticipated, such as an incoming hurricane or storm, or wildfires. The advantage of this intentional islanding operation instead of waiting for the outage in the main grid to occur in order to switch the microgrid to operate in islanding mode is that an intentional islanding allows for a controlled transition that prevents potential failures or quality issues in the micro-grid. Two phases can be distinguished in islanded operation: transition from grid connected to island operation operation isolated from the grid. © A. Kwasinski, 2017

Islanding During the transition into island operation it is important that: voltage disturbances are quickly dampened and that protection schemes both inside the microgrid and in the grid are not affected. When the transition is completed it is important that the micro-grid has sufficient local power generation and energy storage in order to ensure that loads are powered with the agreed quality level. For example, in ac micro-grids it is important that distributed resources are able to provide real and reactive power to the specified load range. This is particularly important in order to avoid loss of stability if there are large motors in the microgrid that require significant amounts of reactive power during startup Also for ac micro-grids, their control systems must be able to regulate both voltage and frequency within acceptable ranges. In dc micro-grids, neither frequency regulation nor reactive power generation are issues to consider. © A. Kwasinski, 2017

Islanding Eventually, it can be anticipated that the micro-grid would be connected to the main grid again. Grid connection of dc micro-grids or ac micro-grids with a power electronics interface with the main grid tends to be simpler than the case of ac micro-grids connected to the main grid through circuit breakers, contactors, or static switches because in the dc micro-grid and the ac micro-grid with a power electronics interface cases reconnection control resides only in this power electronic interface. That is, the controller in this power electronic interface would controlled in order to realize on its grid side some voltage waveform so its amplitude, frequency and phase angle are within specified limits to allow reconnection. In the other ac micro-grid cases—those directly connected to the main grid though mechanical switchgear or static switches—reconnection is more complicated because there is no possibility of directly controlling the voltage waveforms at the PCC. In this case, ensuring that the voltage, frequency and phase angle are within acceptable limits depend on how the microgrid distributed resources are controlled. © A. Kwasinski, 2017

Islanding According to IEEE Standard 1547.4 these approaches can be distinguished in this case of ac micro-grids in order to achieve a successful reconnection: Active synchronization: In this approach, the microgrid controller matches the voltage signal on the PCC micro-grid side to those of the PCC on the grid side immediately before closing the islanding devices, such as a circuit breaker. Implementation of this approach requires measuring these three voltage signal parameters—amplitude, frequency and phase angle—on both sides of the PCC. A communications channel in order to exchange information between the micro-grid and the main grid is also necessary. This need for sensing and communications may lead to a higher failure rate as the sensing and communications subsystems may become a single point of failure. Passive synchronization: In this approach a device is used to monitor the voltage at both sides of the PCC and allows the microgrid to connect to the main grid only when the voltage signal on the microgrid side is within some given required range of the main grid analogous voltage parameters. Like the active synchronization approach, passive synchronization requires sensing and communications, leading to the same potential reliability concerns. In addition, this method may be slower than active synchronization. Open transition: This approach is more basic than the other two, because the method involves connecting both ends of the PCC after interrupting disconnecting the microgrid load. Once the micro-grid is connected to the grid then this load is brought back online. © A. Kwasinski, 2017

Islanding According to IEEE Standard 1547.4, un-intentional islanding operations are “inadvertent events that are typically initiated by loss of area EPS or equipment failure, and the DR island system may be automatically sectionalized from the area EPS by protective equipment.” Once the island has been established, the same considerations that were considered for the intentional island condition applies to the un-intentional island. Contrary to the case of intentional islanding, during an un-intentional island it is not possible to prepare the microgrid for such transition, such as verifying that there is sufficient local generation to sustain a stable operation powering all loads. Hence, in case it is expected that local generation capacity may be insufficient to sustain the load during un-intentional islands, black start functions or standby generators with transfer switches have to be allocated within the microgrid. Once the issue in the main grid that led to loss of service to the micro-grid feeder is solved, it may be of interest to reconnect soon the main grid to the micro-grid. However, such connection cannot occur until the voltage and frequency of the grid are stable and within acceptable ranges. In order to ensure meeting such requirement, a delay of up to five minutes may be provided between the time power is restored at the PCC from the main grid and the time reconnection to the microgrid is established. © A. Kwasinski, 2017