With the growing interest in renewable energy generation, the use of small-scale, localized power generation based on solar (photovoltaic and solar thermal) resources is playing an increasing role in electricity production. The same is true of other renewable energy sources (RES), such as wind, biomass and tidal.
As electric systems around the world shift toward decentralization, opening markets to competition and allowing cross-border electricity trading, consumers are seeking more engagement with electricity providers to ensure greater competition, reliability and service security.
As these trends increase, microgrids (localized, small grids close to the consumers with the ability to disconnect from the traditional power grid and operate autonomously) can offer the flexibility that utilities and consumers require. Microgrids strengthen grid resilience, mitigating the impact of power outages and enhancing reliability by reducing electric power downtime. The integration of RES-based distributed generation (DG) and energy storage within a microgrid can potentially reduce customer costs and enhance system efficiency through on-site generation management. The use of localized DG to serve local loads helps reduce energy transmission losses, further increasing grid efficiency.
One barrier to microgrid adoption has been the clumsy transfer of power from one source to another, which may cause momentary power loss – and, thus, cause computer systems to crash or disrupt other critical loads – or place electrical stress on system components. Fortunately, new technology is emerging that smooths out the bumps in power transfers and also makes a power backup system a more attractive proposition to commercial and industrial facilities.
Now the same inverter that provides grid services, such as demand charge reduction, at a commercial campus can also act as an uninterruptible power supply, smoothing power transfers and offering reliable power. If an outage from the utility doesn’t last long, then the facility won’t need to start its diesel generators. The same is true for RES-enabled microgrids when the wind stops blowing or the day turns cloudy, making alternative energy more practical. In all cases, the goal is to minimize disturbances to the load.
This article describes how Duke Energy, a major investor-owned electric utility, put battery energy storage system (BESS) technology to the test and offers insights on how other utilities may approach the challenge of achieving seamless mode transfer.
Microgrid structure
Figure 1 illustrates a Duke Energy microgrid test bed in Mount Holly, N.C., that was implemented in partnership with 25 vendors. It incorporates several different RES, including a 250 kW/250 kWh BESS, a 100 kW three-phase solar system, a 10 kW single-phase solar carport with electric vehicle charging stations, a 500 kW resistive load bank, 20 kW of single-phase household loads, a control center, and automated distribution grid equipment, including phasor measurement units, reclosers, smart meters and sensors.
The energy storage system (including two PV inverters, the load bank and the recloser) is connected to the Open Field Message Bus (OpenFMB) microgrid controller platform (MCP) using wired and wireless communication networks. The OpenFMB, which is backed by the Smart Grid Interoperability Panel, is a distributed logical interface that sits between all of the various pieces of equipment and systems on the microgrid and orchestrates all of the device-to-device or device-to-system interactions at the grid edge; it works by translating traditional client/server SCADA protocols (e.g., DNP and Modbus) to and from open-standards-based Internet of Things publish-subscribe middleware (e.g., DDS, MQTT and AMQP). The OpenFMB MCP collects information from the RES, energy storage and utility grid, plus power consumption data from the loads, in order to reduce power losses and optimize energy efficiency. While doing this, it must satisfy the load demand and operating constraints, such as DG power scheduling, supply/demand power balance in either grid-tie or island operation, and the battery’s state of charge (SOC). The OpenFMB MCP is also in charge of controlling the operation of the recloser to connect or disconnect the microgrid to or from the utility grid at the 12.4 kV line.
System operating modes
The microgrid has two operating modes: grid-tie and island. In the grid-tie mode, the load demand is met by the DG and the utility power. Maximum renewable power generation is desirable for energy-saving and environmental conditions. When the DG power exceeds the load demand, the microgrid can deliver power to the utility grid.
In the island mode (isolated from the electric power utility), the objective of the microgrid is to maintain a stable and reliable power supply to the loads at a minimum cost, which implies maximum DG power delivery. One of the main technical challenges for microgrids is the RES power scheduling associated with their fluctuating generation. In this context, energy storage is indispensable, both for peak demand shaving and to guarantee a stable and reliable power supply.
BESS control
Figure 2 shows a simplified control diagram of the BESS inverter connected to the microgrid. In grid-tie mode, the inverter works under P – Q (active/reactive) power control, while in island or stand-alone mode, it operates under V – F (voltage/frequency) control. In both modes, an inner current control loop is employed to improve the safety and reliability of the inverter.
When the inverter is operating in grid-tie mode, the active and reactive power demands are defined by the OpenFMB MCP. These demand values depend on the batteries’ SOC, load balance and AC voltage. The output of the closed-loop P – Q controllers is the reference current for the inner current control loop of the BESS inverter.
Alternatively, power sharing among different inverters that feed the microgrid loads can be managed using a droop control running on the application control layer of the inverter.
In island mode, the BESS inverter acts as the voltage and frequency source for the microgrid. The inverter receives the voltage and frequency demands from the OpenFMB MCP. The closed-loop outer voltage control provides the current reference for the inner current loop and sets the frequency demand. The voltage control loop is aimed to control output voltage robustly under load transient events.
The BESS inverter also provides the ability to energize or black start the isolated microgrid until the external grid becomes available and then transfer seamlessly to grid-tie mode.
Setup of the experiment
It was essential that the test program be set up to evaluate how seamless the transfer strategy would be when changing between grid-tie and island modes under a variety of operating conditions. Half of the tests simulated unexpected loss of the grid (due to a downed power line, for example), while the other half simulated the effect of reconnecting the microgrid to the utility grid. Both sets of tests were done under a variety of operating conditions.
The recloser at the point of common coupling is commanded by the OpenFMB MCP to simulate a fault from the grid or to reconnect to it. The BESS inverter has no control over the recloser, nor any knowledge of its state. The load bank is considered a critical load that cannot be tripped during any of the mode transitions. In order to prevent nuisance trips, the abnormal grid thresholds of the BESS for voltage and frequency must be coordinated with the over/under voltage and over/under frequency thresholds of the relays and reclosers in the microgrid. The test results for five of the most relevant transitions include the following:
Grid loss with BESS charging
In this mode, the load is initially supplied by both the DG units and the utility grid while the BESS is charging, with power flow distribution before and after the grid loss shown in Table 1. Figure 3 shows the voltage and currents at the BESS inverter during the transition. As can be seen from the lower trace (expanded 100 times), when the grid is disconnected, the voltage drops to 0.15 p.u. in fewer than two cycles. At this time, the inverter transitions to island mode. The voltage demand is ramped in a controlled fashion from the transfer point up to nominal voltage to avoid high inrush currents in the transformer. This scenario, when the BESS is initially charging at high powers, was found to be the most stringent transfer condition.
Grid loss with BESS discharging
Table 2 shows the power flow conditions when the grid fails while the BESS is discharging and feeding power into the grid. In this case, the inverter detects the grid loss due to the frequency measurement, and the switching loss transient is less severe than the previous condition.
Grid loss to island with BESS discharging with zero power contribution from the grid
Table 3 shows the load flow condition before and after the seamless transfer from grid-tie mode to island mode when the microgrid loads are supplied by the DG and the BESS and there is no contribution from the utility grid. The switching transient was quite small.
Island to grid-tie with BESS charging
In order to test a high-power BESS charging condition with the microgrid in island mode and the maximum power delivered by the PV inverter via the solar array’s maximum power point tracking (MPPT) controller, the load bank was set to a low power demand. Table 4 shows the results and load case for this condition. Once again, the transition caused very little disturbance at the load.
Island to grid-tie with BESS discharging
Table 5 shows load flow conditions for an island to grid-tie transition with the BESS discharging. As with previous tests, the load bank did not trip, although power to the load bank current increased.
Summary
This article has presented the results of tests on the seamless mode transfer strategy between grid-tie mode and island mode on a microgrid test bed under a variety of operating conditions. Under all test conditions, the microgrid was able to transfer seamlessly from one mode of operation to the other without any load tripping or loss of any of the DG inverters. These results show that the seamless transfer strategy employed can safely and robustly transfer between grid-tie and island modes, regardless of the operating conditions before and after the transition.
As consumers have come to demand greater reliability from their electric power suppliers in terms of reduced downtime, the emerging ability to disconnect microgrids from the traditional power grid and operate on their own guarantees supply security in case of disruptions, mitigating the impact of power outages. The successful test results on the BESS technology, microgrid control and communication technologies serve to demonstrate the growing practicality of alternative energy technologies to both electricity producers and consumers. It also demonstrates that collaboration between utilities and the companies that provide equipment and support can speed advances in clean energy technology.
Maziel E. Velasquez is principal engineer of the electromechanical and drives division at Parker Hannifin; Dr. Stuart Laval is manager of technology development at Duke Energy; and Prateek Pandey and Dr. David Blood are a power electronics design engineer and market manager of the energy grid-tie division at Parker Hannifin, respectively. This article’s figures and tables are courtesy of Parker Hannifin.
