Blogs, DER, Solar, Utility Integration

Distribution Grid Power Electronic Devices as Mitigation Options for Circuits with High PV Penetration

As the penetration of variable Distributed Generation (DG) is increasing due to incentives and mandates, such as the Renewables Portfolio Standard (RPS) target in California of 33% by 2020, a number of challenges to utilities have arisen. The existing electric grid was designed and created to safely and reliably distribute power from a few centralized power generation sources through highly monitored and controlled transmission lines to distributed loads typically supplied via radial distribution feeders. Some DG such as small hydro and bio-gas generators are relatively constant and predictable in output. However, most of the recent growth in DG has been to exploit renewable resources, such as solar and wind energy, which are inherently intermittent. This intermittency can cause wide and unpredictable variability of the net power flow over the distribution system. This variability, along with the frequently wide dispersal of the generation by numerous small sources, can cause a number of voltage and power quality problems, and inhibits the utility’s operational control over the voltage and loading conditions in their service territory. At the same time, DG, in particular solar photovoltaic (PV) due to its recent popularity resulting in large-scale deployment of PV installations in many distribution systems, offers opportunities for utilities that, if leveraged, can help optimize grid performance.

DG Might Cause Issues on Distribution Feeders

The detrimental impact DG potentially has when installed in large numbers on an unprepared grid includes:

  • overvoltages causing voltages on distribution feeders that are outside permissible limits and/or tripping of PV inverters;
  • voltage sags and swells
  • flicker caused by the variability of DG-generated power and inrush currents during start-up of wind turbine generators;
  • reverse fault current flow causing existing protection schemes to be uncoordinated or possibly non-functional;
  • increased wear on utility equipment such as tap-changers of voltage regulators and capacitor switches;
  • unintentional islanding;
  • harmonics;
  • and real and reactive power phase imbalances.

Many of these issues have already been observed as evident from the results of utility surveys we conducted and also from our experience with utilities we support with mitigating issues that have arisen on their high DG penetration distribution feeders. Utilities are experiencing more frequent operation of voltage regulators on their high-PV penetration feeders leading to greater maintenance requirements and PV inverters disconnect frequently due to overvoltage conditions. Simply adding DG to the system without preparing for the consequences will almost certainly result in an escalation of the DG-caused issues that already exist on the feeder, and in the emergence of additional issues once even higher DG penetration levels are reached.

Mitigation Challenges for Utilities

There are a number of (often costly) measures at the utilities’ disposal to mitigate power quality issues caused by high-penetration PV. Examples for conventional measures include:

  • reconductoring;
  • upgrading transformer sizes;
  • and adding tapchanging voltage regulators or capacitor banks.

There are also a number of “new” technologies available that have communication and control capabilities thereby facilitating a tighter control of the grid. Examples for these technologies are:

  • storage;
  • PV inverters with dynamic voltage control (sometimes called “smart inverters”);
  • solid-state transformers;
  • static VAR compensators;
  • and dynamic VAR compensators.

Even though mitigating PV-caused issues is crucial, economics and a limited or non-existing communication infrastructure introduce constraints that render many of the mitigation options listed above non-viable.

VAR Compensators Based Solutions

One “new” technology for mitigating PV caused issues and providing other benefits that is commercially available today are power electronics-based smart devices, which promise distributed Volt/Var control that is faster, more responsive, and versatile than centralized command and control architectures, such as DMS, VVO, and SCADA. These devices are of interest to today’s utilities because they are:

  • relatively inexpensive – the cost per device can be as low as $2,000 – $3,000;
  • operate autonomously (voiding the need for a communication infrastructure) or connected to gateways receiving voltage set points form appropriate management systems (frequent changes for a variety of demand management methods);
  • and can be deployed quickly (less than 30 days).

These devices are commonly advertised to have a number of benefits, including:

  • improving grid integration of PV by stabilizing grid voltage;
  • enabling Conservation Voltage Reduction (CVR);
  • reducing system losses and managing system peak demand;
  • mitigating low voltage issues voiding the need for infrastructure upgrades;
  • and providing phase balance, and system Var support through Volt/Var optimization.

Even though the cost per device is relatively low, as stated above, a large number of devices, or a small number of higher-rated, more expensive devices may be necessary to achieve the desired mitigation goal. To avoid excessive costs due to ineffective use and placement of these devices, studies that investigate feeder-specific device numbers and deployment locations are needed.

Available Technologies

Over the last several years we crossed paths with numerous devices and even conducted research studies implementing some of their functionalities in our simulation models. Currently, commercially available VAR compensators based devices can be installed on the medium distribution voltage side (primary side of the service transformer) and utilization voltage side (secondary side of the service transformer). They can be pole or pad mounted, and come in single-phase or three-phase configurations. Choosing the right solution for the given problem requires an in-depth analysis of the pros and cons (i.e., severity of the condition, budget, operational and topological changes to the circuit, etc.).

Secondary Side of the Distribution Circuit

At least two types of utilization voltage rated technologies are currently available.

Secondary Side Type I

Placed between the service transformer and the load these shunt devices inject controlled amounts of reactive current (capacitive) into the grid connection point. Usually, these devices are rated 10 kVA – 15 kVA and are designed to stepwise inject reactive current (1kVAr – 1.5 kVAr steps) into the system with a sub-cycle response time frame. From the three types of such devices considered in this document, these devices come with the lowest price tag ($2,000-$3,000) and a fairly straightforward control strategy.

Typically, the Type I devices are distributed along a circuit at locations experiencing low voltage conditions. Following specifically defined guidelines the planning engineer distributes a definite amount of units (~ 30-40 units for a mid-size circuit) to achieve set goals (i.e. minimum circuit voltage, overall voltage improvement, etc.). Usually, as a rule of thumb, enough units are deployed to keep the circuit minimum voltage above the ANSI limits by several percentage points during peak load conditions. Once the PV penetration hits a critical point and overvoltage conditions are observed, the head bus voltage can be permanently lowered. This prevents the voltage rising above ANSI limits during daylight hours while during dark hours or times of high solar variability the voltage is prevented from dropping below ANSI limits by the deployed Type I devices.

Secondary Side Type II

These devices usually consist of two separately controlled parts – a shunt current source and a series voltage source. Placed between the service transformer and the load these devices are controlled to buck or boost the load side voltage and/or inject controlled amounts of reactive current (capacitive or inductive) to maintain predefined power factor settings. Typically, these devices are rated around 50 kVA, can buck/boost the voltage up to 8%-8.5%, and continuously vary the injected reactive current within its rating at a sub-cycle time frame. The cost range for these units is around $4,000 – $5,000.

Due to the more advanced capabilities, when compared to the Type I devices, the Type II devices can be used for a wider range of problems. These units are usually deployed at or around locations with over and/or under voltage issues (and other PQ problems). The series voltage source allows to maintain the downstream voltage around a set point, while the shunt current source provides the capability of power factor support for the surrounding area. Usually, these devices are deployed to solve known and localized PQ issues rather than spread over the entire circuits to provide large-area support. Due to their superior rating, relatively few of them (10-20 units) are needed when the goal is to mitigate under/over voltage conditions at particular locations. However, with excessive PV deployment, under unaltered circuit conditions (i.e., no CVR) the number of required units can grow to 100-200. To reach the compromise between keeping the number of deployed units low (and the costs) and mitigating the potential voltage issues the circuit operator would still be required to lower the head bus voltage (albeit not as much as with the Type I devices).

Primary Side of the Distribution Circuit

These units are usually rated significantly higher than their secondary side counterparts. It is not uncommon for them to be rated 300 kVAR – 400 kVAR per phase. They can usually be placed anywhere (1- or 3-phase locations) on the middle voltage side of the circuit (usually near known problem areas) and are capable of injecting continuous amounts of reactive current (capacitive or inductive). The relatively high ratings come with a steep price tag ($100,000 – $150,000). However, it only takes 1-2 units for a mid-size feeder to prevent the voltage from violating the ANSI limits (and other PQ issues). A widely spread control method of these units is similar to the approach used in “smart PV inverters”. That is, the units follow a specific Volt-Var curve to maintain the voltage set point by injecting capacitive or inductive reactive current into the grid connection point. If multiple units are deployed, it is not unusual to use individual Volt-Var curves to achieve best results. When properly deployed and operated these units don’t require any additional intervention by the circuit operator (i.e., lowering the voltage at head bus, PV production curtailment, etc.). However, during substantial overvoltage conditions vast amounts of reactive power have to come from the substation, thus potentially lowering the head bus power factor.

Conclusion

High penetration of intermittent distributed generation such as photovoltaics (PV) can cause a number of voltage and power quality issues including violation of voltage limits and excessive voltage variations. These issues inhibit (1) the integration of PVs in the distribution systems and (2) utility’s operational control over the voltage and loading conditions in their service territory. Recently introduced power electronics-based smart devices claim to mitigate these issues thereby improving PV integration and providing other benefits. These devices come in a variety of styles, sizes, price tags, technologies, and deployment strategies. To avoid excessive costs (and turning bad situations into even worse ones) due to ineffective use and placement of these devices, studies that investigate feeder-specific conditions, device numbers and deployment strategies are necessary.

Contact:

Vadim Zheglov, Power Systems Senior Consultant, joined EnerNex in July 2010. Throughout his career, Vadim has performed numerous studies covering large variety of areas such as — electric power grid modernization, power quality issues in wind plants and other power systems, renewable generation integration in distribution systems, integration cost analysis, arc flash hazards, and others. He has conducted analytical studies of various power systems related projects, such as load flow, transient and harmonic studies using PSSE, PSLF, Matlab/Simulink, OpenDSS, CYME, EMTP-RV and various other simulation tools. Vadim also performs on-site field measurements on operating power systems ranging from solar and wind power plants to integrated backup power supplies and industrial facilities.