Electricity distribution is designed to flow one way from a centralized system to customers. But Distributed Generation (DG) from solar PV and wind violates this.
Impacts caused by high penetration levels of intermittent renewable DG can be complex and severe and may include voltage increase, voltage fluctuation, interaction with voltage regulation and control equipment, reverse power flows, temporary over-voltage, power quality and protection concerns, and current and voltage unbalance, to name a few.
There are solutions, but they’re expensive, complicated, and add to the already insane challenges of thousands of utilities, power generators, independent system operators, and other entities trying to coordinate the largest machine in the world when cooperation isn’t always in their best interest.
IEEE. September 5, 2014. IEEE Report to DOE Quadrennial Energy Review on Priority Issues. IEEE
On the distribution system, high penetration levels of intermittent renewable Distributed Generation (DG) creates a different set of challenges than at transmission system level, given that distribution is generally designed to be operated in a radial fashion with one way flow of power to customers, and DG (including PV and wind technologies) interconnection violates this fundamental assumption. Impacts caused by high penetration levels of intermittent renewable DG can be complex and severe and may include voltage increase, voltage fluctuation, interaction with voltage regulation and control equipment, reverse power flows, temporary overvoltage, power quality and protection concerns, and current and voltage unbalance, among others.
Common impacts of DG in distribution grids are described below; this list is not exhaustive and includes operational and planning aspects50, 51.
- Voltage increase can lead to customer complaints and potentially to customer and utility equipment damage, and service disruption.
- Voltage fluctuation may lead to flicker issues, customer complaints, and undesired interactions with voltage regulation and control equipment.
- Reverse power flow may cause undesirable interactions with voltage control and
- regulation equipment and protection system misoperations.
- Line and equipment loading increase may cause damage to equipment and service disruption may occur.
- Losses increase(under high penetration levels) can reduce system efficiency.
- Power factor decrease below minimum limits set by some utilities in their contractual agreements with transmission organizations, would create economic penalties and losses for utilities.
- Current unbalance and voltage unbalance may lead to system efficiency and protection issues, customer complaints and potentially to equipment damage.
- Interaction with Load Tap Changers (LTC), line voltage regulators (VR), and switched
- capacitor banks due to voltage fluctuations can cause undesired and frequent voltage
- changes, customer complaints, reduce equipment life and increase the need for maintenance
- Temporary Overvoltage (TOV): if accidental islanding occurs and no effective reference to ground is provided then voltages in the island may increase significantly and exceed allowable operating limits. This can damage utility and customer equipment, e.g., arresters may fail, and cause service disruptions.
- Harmonic distortion caused by proliferation of power electronic equipment such as PV inverters.
The aggregate effect from hundreds or thousands of inverters may cause service disruptions, complaints or customer economic losses, particularly for those relying on the utilization of sensitive equipment for critical production processes.
- Voltage sags and swells caused by sudden connection and disconnection of large DG units may cause the tripping of sensitive equipment of end users and service disruptions.
- Interaction with protection systems including increase in fault currents, reach
- modification, sympathetic tripping, miscoordination, etc.
- Voltage and transient stability: voltage and transient stability are well-known phenomena at transmission and sub-transmission system level but until very recently were not a subject of interest for distribution systems. As DG proliferates, such concerns are becoming more common.
The severity of these impacts is a function of multiple variables, particularly of the DG penetration level and real-time monitoring, control and automation of the distribution system. However, generally speaking, it is difficult to define guidelines to determine maximum penetration limits of DG or maximum hosting capacities of distribution grids without conducting detailed studies.
From the utility perspective, high PV penetration and non-utility microgrid implementations shift the legacy, centralized, unidirectional power system to a more complex, bidirectional power system with new supply and load variables at the grid’s edge. This shift introduces operational issues such as the nature, cost, and impact of interconnections, voltage stability, frequency regulation, and personnel safety, which in turn impact resource planning and investment decisions.
NREL. 2014. Volume 4: Bulk Electric Power Systems: Operations and Transmission Planning. National Renewable Energy Laboratory.
Initial experience with PV indicates that output can vary more rapidly than wind unless aggregated over a large footprint. Further, PV installed at the distribution level (e.g., residential and commercial rooftop systems) can create challenges in management of distribution voltage.
Meier, A. May 2014. Challenges to the integration of renewable resources at high system penetration. California Energy Commission.
3.2 Distribution Level: Local Issues
A significant class of challenges to the integration of renewable resources is associated primarily with distributed siting, and only secondarily with intermittence of output. These site‐specific issues apply equally to renewable and non‐renewable resources, collectively termed distributed generation (DG). However, DG and renewable generation categories overlap to a large extent due to
- technical and environmental feasibility of siting renewables close to loads
- high public interest in owning renewable generation, especially photovoltaics (PV)
- distributed siting as an avenue to meet renewable portfolio standards (RPS), augmenting the contribution from large‐scale installations Motivation exists; therefore, to facilitate the integration of distributed generation, possibly at substantial cost and effort, if this generation is based on renewable resources.
Distributed generation may therefore be clustered, with much higher penetration on individual distribution feeders than the system‐wide average, for any number of reasons outside the utility’s control, including local government initiatives, socio‐economic factors, or neighborhood social dynamics.
The actual effects of distributed generation at high penetration levels are still unknown but are likely to be very location specific, depending on the particular characteristics of individual distribution feeders.
Technical issues associated with high local penetration of distributed generation include
- Clustering: The local effects of distributed generation depend on local, not system‐wide penetration (percent contribution). Local penetration level of distributed generation may be clustered on individual feeders for reasons outside the utility’s control, such as local government initiatives, socio‐economic factors, including neighborhood social dynamics Clustering density is relative to the distribution system’s functional connectivity, not just geographic proximity, and may therefore not be obvious to outside observers.
- Transformer capacity: Locally, the relative impact of DG is measured relative to load – specifically, current. Equipment, especially distribution transformers, may have insufficient capacity to accommodate amounts of distributed generation desired by customers. Financial responsibility for capacity upgrades may need to be negotiated politically.
- Modeling: From the grid perspective, DG is observed in terms of net load. Neither the amount of actual generation nor the unmasked load may be known to the utility or system operator. Without this information, however, it is impossible to construct an accurate model of local load, for purposes of: forecasting future load, including ramp rates, ascertaining system reliability and security in case DG fails Models of load with high local DG penetration will have to account for both generation and load explicitly in order to predict their combined behavior. • Voltage regulation: Areas of concern, explained in more detail in the Background section below, include: maintaining voltage in permissible range, wear on existing voltage regulation equipment, reactive power (VAR) support from DG
Areas of concern and strategic interest, explained in more detail in the Background section below, include: preventing unintentional islanding, application of microgrid concept, variable power quality and reliability Overall, the effect of distributed generation on distribution systems can vary widely between positive and negative, depending on specific circumstances that include
- the layout of distribution circuits
- existing voltage regulation and protection equipment
- the precise location of DG on the circuit
3.2.2 Background: Voltage Regulation Utilities are required to provide voltage at every customer service entrance within permissible range, generally ±5 percent of nominal. For example, a nominal residential service voltage of 120V means that the actual voltage at the service entrance may vary between 114 and 126 V. Due to the relative paucity of instrumentation in the legacy grid, the precise voltage at different points in the distribution system is often unknown, but estimated by engineers as a function of system characteristics and varying load conditions.
Different settings of load tap changer (LTC) or other voltage regulation equipment may be required to maintain voltage in permissible range as DG turns on and off. Potential problems include the following:
- DG drives voltage out of the range of existing equipment’s ability to control
- Due to varying output, DG provokes frequent operation of voltage regulation equipment, causing excessive wear
DG creates conditions where voltage profile status is not transparent to operators
Fundamentally, voltage regulation is a solvable problem, regardless of the level of DG penetration. However, it may not be possible to regulate voltage properly on a given distribution feeder with existing voltage regulation equipment if substantial DG is added. Thus a high level of DG may necessitate upgrading voltage regulation capabilities, possibly at significant cost. Research is needed to determine the best and most cost‐effective ways to provide voltage regulation, where utility distribution system equipment and DG complement each other.
Legacy power distribution systems generally have a radial design, meaning power flows in only one direction: outward from substations toward customers. The “outward” or “downstream” direction of power flow is intuitive on a diagram; on location, it can be defined in terms of the voltage drop (i.e., power flows from higher to lower voltage).
If distributed generation exceeds load in its vicinity at any one moment, power may flow in the opposite direction, or “upstream” on the distribution circuit. To date, interconnection standards are written with the intention to prevent such “upstream” power flow.
The function of circuit protection is to interrupt power flow in case of a fault, i.e. a dangerous electrical contact between wires, ground, trees or animals that results in an abnormal current (fault current). Protective devices include fuses (which simply melt under excessive current), circuit breakers (which are opened by a relay) and reclosers (which are designed to re‐establish contact if the fault has gone away).
The exception is a networked system, where redundant supply is always present. Networks are more complicated to protect and require special circuit breakers called “network protectors” to prevent circulating or reverse power flow. If connected within such a networked system, DG is automatically prevented from backfeeding into the grid. Due to their considerable cost, networked distribution systems are common only in dense urban areas with a high concentration of critical loads, such as downtown Sacramento or San Francisco, and account for a small percentage of distribution feeders in California.
3.2.5 Research Needs Related to Circuit Protection
The presence of distributed generation complicates protection coordination in several ways: • The fault must now be isolated not only from the substation (“upstream”) power source, but also from DG • Until the fault is isolated, DG contributes a fault current that must be modeled and safely managed
Shifting fault current contributions can compromise the safe functioning of other protective devices: it may delay or prevent their actuation (relay desensitization), and it may increase the energy (I2t) that needs to be dissipated by each device.6 Interconnection standards limit permissible fault current contributions (specifically, no more than 10 percent of total for all DG collectively on a given feeder). The complexity of protection coordination and modeling increases dramatically with increasing number of connected DG units, and innovative protection strategies are likely required to enable higher penetration of DG.
Standard utility operating procedures in the United States do not ordinarily permit power islands. The main exception is the restoration of service after an outage, during which islanded portions of the grid are re‐connected in a systematic, sequential process; in this case, each island is controlled by one or larger, utility‐operated generators. Interconnection rules for distributed generation aim to prevent unintentional islanding. To this end, they require that DG shall disconnect in response to disturbances, such as voltage or frequency excursions, that might be precursors to an event that will isolate the distribution circuit with DG from its substation source.
Disconnecting the DG is intended to assure that if the distribution circuit becomes isolated, it will not be energized. This policy is based on several risks entailed by power islands: • Safety of utility crews: If lines are unexpectedly energized by DG, they may pose an electrocution hazard, especially to line workers sent to repair the cause of the interruption. It is important to keep in mind that even though a small DG facility such as a rooftop solar array has limited capacity to provide power, it would still energize the primary distribution line with high voltage through its transformer connection, and is therefore just as potentially lethal as any larger power source. • Power quality: DG may be unable to maintain local voltage and frequency within desired or legally mandated parameters for other customers on its power island, especially without provisions for matching generation to local load. Voltage and frequency departures may cause property damage for which the utility could be held liable, although it would have no control over DG and power quality on the island. • Re‐synchronization: When energized power islands are connected to each other, the frequency and phase of the a.c. cycle must match precisely (i.e., be synchronized), or elsegenerators could be severely damaged. DG may lack the capability to synchronize its output with the grid upon re‐connection of an island.
3.2.7 Research Needs Related to Islanding
In view of the above risks, most experts agree that specifications for the behavior of DG should be sufficiently restrictive to prevent unintentional islanding. Interconnection rules aim to do this by requiring DG to disconnect within a particular time frame in response to a voltage or frequency deviation of particular magnitude, disconnecting more quickly (down to 0.16 seconds, or 10 cycles) in response to a larger deviation. At the same time, however, specifications should not be too conservative to prevent DG from supporting power quality and reliability when it is most needed.
There is no broad consensus among experts at this time about how best to reconcile the competing goals of minimizing the probability of unintentional islanding, while also maximizing the beneficial contribution from DG to distribution circuits.
As for the possibility of permitting DG to intentionally support power islands on portions of the utility distribution system, there is a lack of knowledge and empirical data concerning how power quality might be safely and effectively controlled by different types of DG, and what requirements and procedures would have to be in place to assure the safe creation and re‐connection of islands. Because of these uncertainties, the subject of islanding seems likely to remain somewhat controversial for some time.
Needed: • Modeling of DG behavior at high local penetrations, including o prevention of unintentional islanding o DG control capabilities during intentional islanding • Collaboration across utility and DG industries to facilitate DG performance standardization, reliability and trust. This means that utilities can depend on DG equipment to perform according to expectations during critical times and abnormal conditions on the distribution system, the handling of which is ultimately the utility’s responsibility.
In the long run, intentional islanding capabilities – with appropriate safety and power quality control – may be strategically desirable for reliability goals, security and optimal resource utilization. Such hypothetical power islands are related to but distinct from the concept of microgrids, in that they would be scaled up to the primary distribution system rather than limited to a single customer’s premises. A microgrid is a power island on customer premises, intermittently connected to the distribution system behind a point of common coupling (PCC) that may comprise a diversity of DG resources, energy storage, loads, and control infrastructure. Three key features of a microgrid are • Design around total system energy requirements:
Depending on their importance, time preference or sensitivity to power quality, different loads may be assigned to different primary and/or back‐up generation sources, storage, or uninterruptible power supplies (UPS). A crucial concept is that the expense of providing highly reliable, high‐quality power (i.e., very tightly controlled voltage and frequency) can be focused on those loads where it really matters to the end user (or life of the appliance), at considerable overall economic savings. However, the provision of heterogeneous power quality and reliability (PQR) requires a strategic decision of what service level is desired for each load, as well as the technical capability to discriminate among connected loads and perform appropriate switching operations. • Presentation to the macrogrid as a single controlled entity: At the point of common coupling, the microgrid appears to the utility distribution system simply as a time‐varying load. The complexity and information management involved in coordinating generation, storage and loads is thus contained within the local boundaries of the microgrid.
Note that the concepts of microgrids and power islands differ profoundly in terms of • ownership • legal responsibility (i.e. for safety and power quality) • legality of power transfers (i.e., selling power to loads behind other meters) • regulatory jurisdiction • investment incentives Nevertheless, microgrids and hypothetical power islands on distribution systems involve many of the same fundamental technical issues. In the long run, the increased application of the microgrid concept, possibly at a higher level in distribution systems, may offer a means for integrating renewable DG at high penetration levels, while managing coordination issues and optimizing resource utilization locally. Research Needs: • Empirical performance validation of microgrids • Study of the implications of applying microgrid concepts to higher levels of distribution circuits, including o time‐varying connectivity o heterogeneous power quality and reliability o local coordination of resources and end‐uses to strategically optimize local benefits of distributed renewable generation • Study of interactions among multiple microgrids