The problem of short circuit currents can be solved by substation upgrades, installing series reactors [80], or by employing fault current limiters [81].
1.2.17 Unintentional Islanding
An “island” is any segment of a power system which has its own generation and loads, and hence can operate independent of the power system for at least some time period. While intentionally formed islands can be helpful in supporting system reliability, unintentionally created islands can present several technical and safety challenges, and risks.
Distribution feeders with a high penetration level of solar PV power systems can potentially cause the problem of unintentional islanding [82, 83]. The solar PV system may continue to feed a load downstream of its location in an islanded mode, if due to a fault/disturbance the utility has isolated the feeder upstream of the DER. The PV system may not be designed to maintain voltage and frequency for customers in that island when the source of power from utility side is disconnected. If the islanded system is not provided with effective grounding systems, Transient Overvoltages will result due to the power being fed by ungrounded generators, which can have a damaging impact on the customer equipment connected in that island [67]. Moreover, the continued energization of the island by the solar PV systems may pose a safety threat to utility personnel working in that area.
In some cases, an island may be planned ahead of time by the utility to provide continued service to customers. However, when such islanded condition is not preplanned by the utility, the distribution system isolated from the utility source is referred as an “unintentional island”. In case of an unintended island, the solar PV system shall detect the island, cease to energize the area electric power system, and trip within a very short time (typically two seconds) of the formation of the island [47].
PV‐based DERs have a low fault current and especially with variable power output may not be able to correctly detect fault conditions on the grid. They are, therefore, critically dependent on anti‐islanding control signals issued by the grid to determine when to trip.
Anti‐islanding methods in grid‐connected solar PV systems can generally be classified into two major groups, which include: (i) communication‐based methods and (ii) noncommunication‐based methods [84]. The noncommunication‐based methods include passive methods, active methods, and hybrid methods. Communication‐based strategies include (i) transfer trip and (ii) power line signaling. Direct Transfer Trip is one of the communication‐based techniques for tripping solar PV systems [84]. However, with the unprecedented growth of solar PV systems in distribution systems, anti‐islanding techniques based on traditional transfer trip have become quite expensive to implement on each member of the large fleet of solar PV systems.
Alternate transfer trip strategies for disconnecting solar PV systems during events of upstream disruptions of power supply are being implemented recently. A power line carrier‐based anti‐islanding scheme is successfully operating in the network of National Grid in Massachusetts, USA. A signal indicating the upstream breaker “ON” status is continuously transmitted on the power line. This signal is received by all the PV inverters connected to that line or network, which continue to perform with their smart inverter functions, as long as they read this ON signal. If the upstream breaker opens due to a system fault or disturbance this signal is no longer available on the disconnected feeder and all DER inverters connected downstream, simply trip. This cost‐effective technique has been operating very successfully.
A study on the variation of frequency during an islanding condition with different ratios of rotating and inverter‐based generators is reported in [23]. During a high power mismatch between the generation and the load, it is shown that:
1 in a rotating generator dominated case, frequency declines if the load exceeds generation, however, frequency increases if generation exceeds the load. In other words, the frequency is impacted by the active power mismatch.
2 in an inverter dominated case, the frequency can even increase if the load exceeds generation when there is a surplus of reactive power. Alternatively stated, frequency variations may also be governed by reactive power mismatches.
It is further reported [23] that anti‐islanding protection systems will trip much faster if the island is dominated by inverter‐based generation and there is a surplus of reactive power. It is, therefore, recommended that to have an improved anti‐islanded protection in systems having a dominance of inverter‐based generation, excess reactive power must be made available in the islanded portion. This can be achieved through activation of shunt capacitors.
1.2.18 Frequency Regulation Issues due to Reduced Inertia
Increasing penetration of IBRs such as solar PV, wind, BESSs, etc., and the retirement of large conventional thermal synchronous generators is leading to a substantial decline in inertia in power systems. Reduced inertia can have the following consequences [11, 13]:
1 Larger or steeper ROCOF resulting in shorter time to reach UFLS thresholds.
2 Load shedding thresholds are reached even before the PFR or FCR becomes available.
3 Lower nadirs and activation of UFLS will result in potential disconnection of domestic customers.
4 Faster frequency variations may adversely impact system protection.
5 Low‐frequency conditions may cause generators to trip. Loss of generation can potentially lead to cascading effects and a partial or even full system blackout.
6 Risk of islanding segments of distribution networks with large number of solar PV systems.
7 Lack of rotating inertia also presents problems in microgrid environments, where there are no large conventional generators to stabilize the system frequency.
Some other important considerations regarding low inertia systems are as follows [13]:
1 In large interconnected systems, the level of electromechanical damping in different regions may also have an impact. This can result in a higher ROCOF close to the lost generation than at locations farther in the system. Hence locational impacts of system inertia and disturbance need to be examined.
2 The loss of generation may result in system islanding where a small part of the system gets separated and operates autonomously. Since the inertia of the islanded system is even lower than the main low inertia system, different control challenges such as system oscillations may be experienced. Such islands may form even in distribution networks.
1.2.18.1 Under Frequency Response
A study of the US Western Interconnection depicting the impact of extra high penetration of solar PV systems on overall system frequency response is described below [2]. This study considers a total renewable penetration of 80% (65% solar and 15% wind) at the interconnection level and 100% penetration (all PV) at a regional level.
A realistic model of the Western Interconnection is utilized. The 2022 Light Spring (LSP) planning case is considered due to the relatively low level of online synchronous generation during a light load condition. This represents an ideal (most severe) case to benchmark the frequency response performance.
The 2022 LSP case is implemented in GE’s Positive Sequence Load Flow analysis simulation software. The LSP case is modeled in detail with over 4000 generators and more than 19 000 buses. Each power plant has its own specific detailed dynamic models, including a generator, exciter, and turbine governor. HVDC systems and protective systems such as UFLS and line protection are also included in the Western Interconnection dynamic model. As PV systems are entirely based on inverters, they are represented by GE Type 4 wind power plant model. This includes