By Dr. Graham Dudgeon and Shripad Chandrachood, MathWorks
Microgrid systems attract interest due to enhanced reliability and self-reliance. From technical and economic perspectives, microgrid development poses challenges and opportunities. Integrating system management makes it possible to realize smart operation—to maximize system efficiency and security while lowering operational costs.
Developing microgrid systems with smart operation requires tight integration of multiple engineering disciplines: control systems, electrical power systems and communication systems. Software tools that enable engineers to share a common environment offer improved project efficiency. In that environment, the simulation model serves as an integrated reference. The reference allows engineering groups to conduct design tasks, communicate and evaluate overall system performance early in the development process. Model-based design with MATLAB and Simulink allows engineers to focus on design and test. The engineers are not expending resources on integrating multiple software tools or exporting design information from one software environment and importing it into another.
Integrated Systems, Development
Realizing the maximum operational potential of an integrated system requires strong collaboration among multiple engineering disciplines. For microgrids (and other smart grid technologies), electrical power engineers, control system engineers and communication system engineers must work together in the design process to mitigate integration and operation risks. The need for collaboration is clear, but execution and management may increase project delivery effectiveness.
Organizations need a specification common to all teams and open to little subjective interpretation to facilitate collaboration. Further, engineers require the ability to add details and enhancements to the specification as the project progresses, so that it evolves in parallel with the project. With model-based design, these needs are addressed by the simulation model, which serves as an executable specification of the system under development. Simulation models help engineers increase systems’ understanding in a cost-effective and repeatable environment. This benefit holds true at any stage of a system life cycle, from original concept through specification, design, integration, test, commission and in-service support.
Modeling and simulation can progress as an integral activity with system development. It may align with existing processes, thus providing a powerful and complementary support tool. Transferring knowledge across engineering departments via an executable specification facilitates a common reference that spans the entire engineering team. Furthermore, execution of the model via simulation offers an objective assessment of system functionality against requirements.
Smart, Robust Grids
The heart of a smart grid is a robust grid that provides its customers with a high-quality power supply, both reliably and cost effectively. To achieve this robustness, overall system performance cannot be viewed as a summation of individual component performance. There is, therefore, a paradigm shift from traditional local measurement and control response optimization to global system performance management and optimization. Global system performance management provides an extra layer of situational awareness that allows generators, loads and protection relays to be commanded under certain system contingencies to promote minimal loss-of-load probability.
Consider a simple example of such a contingency: a generator trips due to a fault, causing a decrease in system capacity. The management system responds to this by commanding a large load on the system (in this case a motor drive) to limit power consumption—so that overall system consumption remains below available capacity. This allows the system frequency to remain within required limits while the management system commands a standby generator to synchronize and connect. After connection, the management system informs the motor that it may return to nominal power requirement. Figure 1 shows motor drive power consumption being limited at two seconds after the reduction of available capacity. At 3.5 seconds, the spinning reserve synchronizes and connects, allowing the power ramp to continue. Figure 2 shows the spinning reserve power output.
The management system also may perform economic dispatch calculations to promote cost-effective generator operation for a given demand. If a contingency arises that causes priority to shift to security rather than cost-effectiveness, then the economic dispatch will be overridden in favor of supporting secure operation.
Figure 3 shows an example of a security constrained dispatch in which a parallel feeder connecting two inexpensive generators to a load has tripped at 22 seconds, limiting the ability of these generators to service the load. This, in turn, limits the ability of the system to generate power at an economic optimum. Generator 3, a more expensive generator, must increase power output to support system security. The plot of system frequency during this contingency in Figure 4 confirms that frequency remains within acceptable limits. In this example, dispatch calculations are performed every five seconds. Unlike large-scale grids, which have hundreds of generators receiving dispatch instructions every five minutes, short-duration dispatch instructions are necessary in microgrids with a small number of generators. The lower system inertia is unlikely to effectively absorb any energy imbalance incurred between longer-duration dispatch instructions.
The ability to connect to and disconnect from the main utility is a unique microgrid feature. The connection to the utility may be either AC or DC. With a DC link, the microgrid and utility frequency may remain asynchronous, but steps may be required to mitigate harmonics. Appropriate active- or static-harmonic filters or architectural design of the DC link may achieve this mitigation—the greater number of switching devices, the lower the harmonic signature, assuming appropriate switching strategies are adopted. DC links also provide more control via the power electronic interfaces, which enables flow flexibility between the microgrid and the utility.
For an AC connection with the utility grid, online generators must perform a synchronization sequence. Such a sequence requires that microgrid system frequency be elevated above utility frequency without causing internal power quality issues. Once microgrid and utility voltage waveforms are overlaid within some tolerance, then the connection is made. Figure 5 shows a comparison of active power of a microgrid for AC synchronization and DC connection between the microgrid and a utility. The DC connection enables improved transient response following the connection. The power settling at 0.5 per unit following the connection demonstrates that no power flows through the connection in the steady state in either scenario.
Microgrids lend themselves to integration of smaller-scale solar- and wind-power systems, which may be connected at the distribution level. These systems, characterized by variable power generation and unbalanced conditions, lead to operational challenges including system instability and quality of supply issues. Supplementing the system with storage technology can mitigate both variability and system imbalance. Additionally, it provides a means to enhance system stability through active power control, reactive power control or both. Simulation enables engineers to identify appropriate storage capacity, location and control strategies. Figure 6 shows an example storage device with a supplemental frequency damping control function used to enhance frequency stability in a system containing many renewable energy sources. The damping function is enabled at 0.4 seconds.
In these cases, model-based design enables efficient project execution by sharing designs through an executable specification. It allows engineers to focus on design and testing, rather than expending valuable resources on integrating multiple software systems or translating design information between disparate software environments.
The text of MathWorks’ original article in the September issue was inadvertently replaced with another author’s. POWERGRID International apologizes for this error and is delighted to run the article here in full.
Dr. Graham Dudgeon is MathWorks’ energy production industry marketing manager. He joined MathWorks in 2004 as a senior technical consultant; prior to that, he held research fellow positions at the University of Strathclyde in the Centre for Economic Renewable Power Delivery (CERPD) and the Rolls-Royce University Technology Centre in Electrical Power Systems.
Shripad Chandrachood is an application engineer for control design automation at MathWorks. He joined MathWorks in 2007 as an application support engineer, prior to that, he was an engineer at TATA Power Company.